From 6451a863be65963d650d38d4a8017d4710098f38 Mon Sep 17 00:00:00 2001 From: MLopez-Ibanez <2620021+MLopez-Ibanez@users.noreply.github.com> Date: Wed, 10 Apr 2024 09:24:38 +0000 Subject: [PATCH] =?UTF-8?q?Deploying=20to=20gh-pages=20from=20@=20iridia-u?= =?UTF-8?q?lb/references@b83165eff2733d472e38d5964f7903ec0d9d9cd6=20?= =?UTF-8?q?=F0=9F=9A=80?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- bibtex.css | 191 + footer.htm | 2 + header.htm | 45 + index.html | 45201 ++++++++++++++++++++++++++++++++++++++++++ index_bib.html | 47702 +++++++++++++++++++++++++++++++++++++++++++++ testbib.pdf | Bin 0 -> 1377784 bytes testshortbib.pdf | Bin 0 -> 1274235 bytes 7 files changed, 93141 insertions(+) create mode 100644 bibtex.css create mode 100644 footer.htm create mode 100644 header.htm create mode 100644 index.html create mode 100644 index_bib.html create mode 100644 testbib.pdf create mode 100644 testshortbib.pdf diff --git a/bibtex.css b/bibtex.css new file mode 100644 index 00000000..6700c68d --- /dev/null +++ b/bibtex.css @@ -0,0 +1,191 @@ +.button { + font-family: inherit; + font-size: 80%; + padding: 0.5em 0.5em; + color: white; /* rgba not supported (IE 8) */ + border: 1px solid #999; /*IE 6/7/8*/ + border: none rgba(0, 0, 0, 0); /*IE9 + everything else*/ + background-color: #E6E6E6; + text-decoration: none; + border-radius: 4px; + text-shadow: 0 1px 1px rgba(0, 0, 0, 0.2); +} + +.button-hover, +.button:hover, +.button:focus { + background-image: linear-gradient(transparent, rgba(0,0,0, 0.05) 40%, rgba(0,0,0, 0.10)); + text-decoration: none; color : white +} + +a:link {text-decoration: none; color : #326fa3} +a:visited {text-decoration: none; color : #326fa3} +a:active {text-decoration: none; color: #666666} +a:hover {text-decoration: underline} + +} +.button:focus { + outline: 0; +} +.button-active, +.button:active { + box-shadow: 0 0 0 1px rgba(0,0,0, 0.15) inset, 0 0 6px rgba(0,0,0, 0.20) inset; + border-color: #000; +} +a.button, +a:link.button, +a:visited.button, +a:hover.button, +a:active.button { + color: white; + text-decoration: none; +} + +.background-green { + background: rgb(28, 184, 65); + /* this is a green */ +} + +.background-maroon { + background: rgb(202, 60, 60); + /* this is a maroon */ +} +.background-orange { + background: rgb(223, 117, 20); + /* this is an orange */ +} + +.background-blue { + background: rgb(66, 184, 221); + /* this is a light blue */ +} + +#at { + float: none; + display: inline; + vertical-align: middle; + border: 0; + width: 1em; + height: auto; +} +dt +{ + float: left; + text-align: right; + margin: 0; + width: 3.5em; + font-size: 90%; +} +dd +{ + text-align: left; + margin-left: 3.6em; + margin-bottom: .75em; +} + +h1 { color: #326fa3; font: 30px 'Lucida Grande', Geneva, Verdana, Arial, sans-serif; font-weight: bold;} + +h2 { color: #326fa3; font: 20px 'Lucida Grande', Geneva, Verdana, Arial, sans-serif; font-weight: bold;} + +h3 { color: #326fa3; font: 14px 'Lucida Grande', Geneva, Verdana, Arial, sans-serif; font-weight: bold;} + +.authors { font-weight: bold; color: #326fa3; font-size: 14px; margin-top: 0em; padding: 0px} +.authors dt { float: left;} +.authors dd { margin-left: 8em; } + + + +div#navlist { background-color: #f5f5f5; border: 1px solid #C9D1D7; padding: 1ex; padding-top: 0px; margin: 1em; margin-top: 2em; width: 20em; } + +#navlist h2 { color: black; font-size: 85%; } +#navlist ul, #navlist ol { list-style-type: decimal; color:#326fa3; font-size: 85%; } + +blockquote { font-size: smaller; } + +ol.publications { + margin-left: 3em; + padding: 0; + counter-reset: publication; +} + +ol.publications li{ + margin: 0; + padding: .5em 0em .5em 0em; + display: block; +} + +ol.publications > li:before { + display: inline-block; + min-width: 4ex; + padding-right: 1.5ex; + text-align: right; + margin-left: -5.5ex; + content: "["counter(publication)"] "; + counter-increment: publication; +} + +dd:target { + border:1px solid #dac89d; + background: #ffffd3; +} + +ul.publinks { + margin: 0 auto 1em; + margin: 0; + padding: 0; + border: none; +display:inline-block; +vertical-align: middle; +} + +ul.publinks > li { + margin: 0; + padding: 0; + border: none; + list-style: none; + display: inline-block; + margin-left: -4px; + line-height: 24px; /* minimum to fix pdfdoc icon */ +} + +a.pdfdoc { + display:inline-block; + line-height:22px; +/* padding-left:23px; + background:transparent url(../img/pdf22.png) center left no-repeat;*/ +} + +ul.publinks > li > a { + white-space: nowrap; + margin-left: 3px; + margin-right: 3px; +} +ul.publinks > li > a.pdfdoc { + white-space: nowrap; + margin-left: 3px; +/* padding-left: 23px; + background:transparent url(../img/pdf22.png) center left no-repeat;*/ + margin-right: 3px; + min-height: 22px; +} + +ul.publinks:before { + content: "[ "; +} + +ul.publinks:after { + content: "]"; +} + +.inline-citation { + max-width:50em; margin-left: 4em; margin-right: 4em; +} + +.repo-gh { + width: 30px; + height: 30px; + display: inline-block; + vertical-align: middle; + color: #9d9d9d; + text-decoration: none; +} diff --git a/footer.htm b/footer.htm new file mode 100644 index 00000000..308b1d01 --- /dev/null +++ b/footer.htm @@ -0,0 +1,2 @@ + + diff --git a/header.htm b/header.htm new file mode 100644 index 00000000..933f98c7 --- /dev/null +++ b/header.htm @@ -0,0 +1,45 @@ + + + + + +IRIDIA BibTeX Repository + + + + + + + + + + + +
+ + + + README + PDF (1.5 MB) + BibTeX + +License: CC0-1.0 +
+ + +

IRIDIA BibTeX Repository

+ +

What is this?

+ +

This list of references in automatically generated from a collection of BibTeX files organized in a way that tries to avoid redundancy, minimise mistakes and facilitate customization.

+ +

You only need to fork (or link) the git repository in your papers and sync with the main copy to send/receive updates.

+ +

Most customisations, such as shorter journal or conference names, do not require changing the existing .bib files. + You should not need to edit the entries directly unless you find mistakes. See the README for more details.

+ +

References

+ diff --git a/index.html b/index.html new file mode 100644 index 00000000..3ba2763a --- /dev/null +++ b/index.html @@ -0,0 +1,45201 @@ + + + + + +IRIDIA BibTeX Repository + + + + + + + + + + + +
+ + + + README + PDF (1.5 MB) + BibTeX + +License: CC0-1.0 +
+ + +

IRIDIA BibTeX Repository

+ +

What is this?

+ +

This list of references in automatically generated from a collection of BibTeX files organized in a way that tries to avoid redundancy, minimise mistakes and facilitate customization.

+ +

You only need to fork (or link) the git repository in your papers and sync with the main copy to send/receive updates.

+ +

Most customisations, such as shorter journal or conference names, do not require changing the existing .bib files. + You should not need to edit the entries directly unless you find mistakes. See the README for more details.

+ +

References

+ + +
+ +
+[1] +
+
+David Abramson, Mohan Krishna Amoorthy, and Henry Dang. + Simulated annealing cooling schedules for the school timetabling problem. + Asia-Pacific Journal of Operational Research, 16(1):1–22, 1999.
+[ bib ] + +
+ + +
+[2] +
+
+David Abramson. + Constructing School Timetables Using Simulated Annealing: Sequential and Parallel Algorithms. + Management Science, 37(1):98–113, 1991.
+[ bib ] + +
+ + +
+[3] +
+
+Tobias Achterberg. + SCIP: Solving constraint integer programs. + Mathematical Programming Computation, 1(1):1–41, July 2009.
+[ bib | +epub ] + +
+ + +
+[4] +
+
+Tobias Achterberg and Timo Berthold. + Improving the feasibility pump. + Discrete Optimization, 4(1):77–86, 2007.
+[ bib ] + +
+ + +
+[5] +
+
+Héctor-Gabriel Acosta-Mesa, Fernando Rechy-Ramírez, Efrén Mezura-Montes, Nicandro Cruz-Ramírez, and Rodolfo Hernández Jiménez. + Application of time series discretization using evolutionary programming for classification of precancerous cervical lesions. + Journal of Biomedical Informatics, 49:73–83, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[6] +
+
+Bernardetta Addis, Marco Locatelli, and Fabio Schoen. + Disk Packing in a Square: A New Global Optimization Approach. + INFORMS Journal on Computing, 20(4):516–524, 2008.
+[ bib | +DOI ] + +
+ + +
+[7] +
+
+B. Adenso-Díaz. + Restricted Neighborhood in the Tabu Search for the Flowshop Problem. + European Journal of Operational Research, 62(1):27–37, 1992.
+[ bib ] + +
+ + +
+[8] +
+
+B. Adenso-Díaz and Manuel Laguna. + Fine-Tuning of Algorithms Using Fractional Experimental Design and Local Search. + Operations Research, 54(1):99–114, 2006.
+[ bib ] +
+Keywords: Calibra +
+ +
+ + +
+[9] +
+
+Steven Adriaensen, André Biedenkapp, Gresa Shala, Noor Awad, Theresa Eimer, Marius Thomas Lindauer, and Frank Hutter. + Automated dynamic algorithm configuration. + Journal of Artificial Intelligence Research, 75:1633–1699, 2022.
+[ bib | +DOI ] + +
+ + +
+[10] +
+
+Bekir Afsar, Kaisa Miettinen, and Francisco Ruiz. + Assessing the Performance of Interactive Multiobjective Optimization Methods: A Survey. + ACM Computing Surveys, 54(4), 2021.
+[ bib | +DOI ] +
+Interactive methods are useful decision-making tools for + multiobjective optimization problems, because they allow a + decision-maker to provide her/his preference information + iteratively in a comfortable way at the same time as (s)he + learns about all different aspects of the problem. A wide + variety of interactive methods is nowadays available, and + they differ from each other in both technical aspects and + type of preference information employed. Therefore, assessing + the performance of interactive methods can help users to + choose the most appropriate one for a given problem. This is + a challenging task, which has been tackled from different + perspectives in the published literature. We present a + bibliographic survey of papers where interactive + multiobjective optimization methods have been assessed + (either individually or compared to other methods). Besides + other features, we collect information about the type of + decision-maker involved (utility or value functions, + artificial or human decision-maker), the type of preference + information provided, and aspects of interactive methods that + were somehow measured. Based on the survey and on our own + experiences, we identify a series of desirable properties of + interactive methods that we believe should be assessed. +
+
+Keywords: decision-makers, Interactive methods, performance assessment, + preference information, multiobjective optimization problems +
+ +
+ + +
+[11] +
+
+Bekir Afsar, Johanna Silvennoinen, Giovanni Misitano, Francisco Ruiz, Ana B. Ruiz, and Kaisa Miettinen. + Designing empirical experiments to compare interactive multiobjective optimization methods. + Journal of the Operational Research Society, 74(11):2327–2338, November 2022.
+[ bib | +DOI ] + +
+ + +
+[12] +
+
+Ralph D'Agostino and E. S. Pearson. + Tests for Departure from Normality. Empirical Results for the Distributions of b2 and √b1. + Biometrika, 60(3):613–622, December 1973.
+[ bib | +DOI ] + +
+ + +
+[13] +
+
+Per J. Agrell. + On redundancy in multi criteria decision making. + European Journal of Operational Research, 98(3):571–586, 1997.
+[ bib | +DOI ] +
+The concept of redundancy is accepted in Operations Research + and Information Theory. In Linear Programming, a constraint + is said to be redundant if the feasible decision space is + identical with or without the constraint. In Information + Theory, redundancy is used as a measure of the stability + against noise in transmission. Analogies with Multi Criteria + Decision Making (MCDM) are indicated and it is argued that + the redundancy concept should be used as a regular feature in + conditioning and analysis of Multi Criteria + Programs. Properties of a proposed conflict-based + characterisation are stated and some existence results are + derived. Redundancy is here intended for interactive methods, + when the efficient set is progressively explored. A new + redundancy test for the linear case is formulated from the + framework. A probabilistic method based on correlation is + proposed and tested for the non-linear case. Finally, some + general guidelines are given concerning the redundancy + problem. +
+
+Keywords: Multi criteria decision making, Redundancy, objective + reduction, Vector optimisation +
+ +
+ + +
+[14] +
+
+Hernán E. Aguirre and Kiyoshi Tanaka. + Working principles, behavior, and performance of MOEAs on MNK-landscapes. + European Journal of Operational Research, 181(3):1670–1690, 2007.
+[ bib | +DOI ] + +
+ + +
+[15] +
+
+Samad Ahmadi and Ibrahim H. Osman. + Density Based Problem Space Search for the Capacitated Clustering p-Median Problem. + Annals of Operations Research, 131:21–43, 2004.
+[ bib ] + +
+ + +
+[16] +
+
+Ali Ahrari, Saber Elsayed, Ruhul Sarker, Daryl Essam, and Carlos A. Coello Coello. + Weighted pointwise prediction method for dynamic multiobjective optimization. + Information Sciences, 546:349–367, 2021.
+[ bib ] + +
+ + +
+[17] +
+
+R. K. Ahuja, O. Ergun, and A. P. Punnen. + A Survey of Very Large-scale Neighborhood Search Techniques. + Discrete Applied Mathematics, 123(1–3):75–102, 2002.
+[ bib ] + +
+ + +
+[18] +
+
+Sandip Aine, Rajeev Kumar, and P. P. Chakrabarti. + Adaptive parameter control of evolutionary algorithms to improve quality-time trade-off. + Applied Soft Computing, 9(2):527–540, 2009.
+[ bib | +DOI ] +
+Keywords: anytime +
+ +
+ + +
+[19] +
+
+A. A. Albrecht, P. C. R. Lane, and K. Steinhöfel. + Analysis of Local Search Landscapes for k-SAT Instances. + Mathematics in Computer Science, 3(4):465–488, 2010.
+[ bib | +DOI ] + +
+ + +
+[20] +
+
+Susanne Albers. + Online Algorithms: A Survey. + Mathematical Programming, 97(1):3–26, 2003.
+[ bib ] + +
+ + +
+[21] +
+
+Aldeida Aleti and Irene Moser. + A systematic literature review of adaptive parameter control methods for evolutionary algorithms. + ACM Computing Surveys, 49(3, Article 56):35, October 2016.
+[ bib | +DOI ] + +
+ + +
+[22] +
+
+Pedro Alfaro-Fernández, Rubén Ruiz, Federico Pagnozzi, and Thomas Stützle. + Automatic Algorithm Design for Hybrid Flowshop Scheduling Problems. + European Journal of Operational Research, 282(3):835–845, 2020.
+[ bib | +DOI ] +
+Industrial production scheduling problems are challenges that + researchers have been trying to solve for decades. Many + practical scheduling problems such as the hybrid flowshop are + NP-hard. As a result, researchers resort to metaheuristics to + obtain effective and efficient solutions. The traditional + design process of metaheuristics is mainly manual, often + metaphor-based, biased by previous experience and prone to + producing overly tailored methods that only work well on the + tested problems and objectives. In this paper, we use an + Automatic Algorithm Design (AAD) methodology to eliminate + these limitations. AAD is capable of composing algorithms + from components with minimal human intervention. We test the + proposed AAD for three different optimization objectives in + the hybrid flowshop. Comprehensive computational and + statistical testing demonstrates that automatically designed + algorithms outperform specifically tailored state-of-the-art + methods for the tested objectives in most cases. +
+
+Keywords: Scheduling, Hybrid flowshop, Automatic algorithm + configuration, Automatic Algorithm Design +
+ +
+ + +
+[23] +
+
+Alnur Ali and Marina Meilă. + Experiments with Kemeny ranking: What Works When? + Mathematical Social Science, 64(1):28–40, July 2012.
+[ bib | +DOI ] +
+Computational Foundations of Social Choice +
+
+Keywords: Borda ranking, Kemeny ranking +
+ +
+ + +
+[24] +
+
+Ali Allahverdi and Harun Aydilek. + Algorithms for no-wait flowshops with total completion time subject to makespan. + International Journal of Advanced Manufacturing Technology, pp.  1–15, 2013.
+[ bib ] + +
+ + +
+[25] +
+
+Richard Allmendinger, Andrzej Jaszkiewicz, Arnaud Liefooghe, and Christiane Tammer. + What if we increase the number of objectives? Theoretical and empirical implications for many-objective combinatorial optimization. + Computers & Operations Research, 145:105857, 2022.
+[ bib | +DOI ] + +
+ + +
+[26] +
+
+Richard Allmendinger and Joshua D. Knowles. + On Handling Ephemeral Resource Constraints in Evolutionary Search. + Evolutionary Computation, 21(3):497–531, September 2013.
+[ bib | +DOI ] +
+We consider optimization problems where the set of solutions + available for evaluation at any given time t during + optimization is some subset of the feasible space. This model + is appropriate to describe many closed-loop optimization + settings (i.e. where physical processes or experiments are + used to evaluate solutions) where, due to resource + limitations, it may be impossible to evaluate particular + solutions at particular times (despite the solutions being + part of the feasible space). We call the constraints + determining which solutions are non-evaluable ephemeral + resource constraints (ERCs). In this paper, we investigate + two specific types of ERC: one encodes periodic resource + availabilities, the other models `commitment' constraints + that make the evaluable part of the space a function of + earlier evaluations conducted. In an experimental study, both + types of constraint are seen to impact the performance of an + evolutionary algorithm significantly. To deal with the + effects of the ERCs, we propose and test five different + constrainthandling policies (adapted from those used to + handle `standard' constraints), using a number of different + test functions including a fitness landscape from a real + closed-loop problem. We show that knowing information about + the type of resource constraint in advance may be sufficient + to select an effective policy for dealing with it, even when + advance knowledge of the fitness landscape is limited. +
+ +
+ + +
+[27] +
+
+Christian Almeder. + A hybrid optimization approach for multi-level capacitated lot-sizing problems. + European Journal of Operational Research, 200(2):599–606, 2010.
+[ bib | +DOI ] +
+Solving multi-level capacitated lot-sizing problems + is still a challenging task, in spite of increasing + computational power and faster algorithms. In this + paper a new approach combining an ant-based + algorithm with an exact solver for (mixed-integer) + linear programs is presented. A MAX-MIN ant system + is developed to determine the principal production + decisions, a LP/MIP solver is used to calculate + the corresponding production quantities and + inventory levels. Two different local search methods + and an improvement strategy based on reduced + mixed-integer problems are developed and integrated + into the ant algorithm. This hybrid approach + provides superior results for small and medium-sized + problems in comparison to the existing approaches in + the literature. For large-scale problems the + performance of this method is among the best +
+
+Keywords: Ant colony optimization, Manufacturing, Material + requirements planning, Mixed-integer programming +
+ +
+ + +
+[28] +
+
+S. Alupoaei and S. Katkoori. + Ant Colony System Application to Marcocell Overlap Removal. + IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 12(10):1118–1122, 2004.
+[ bib ] + +
+ + +
+[29] +
+
+André R. S. Amaral. + The Corridor Allocation Problem. + Computers & Operations Research, 39(12):3325–3330, 2012.
+[ bib | +DOI ] +
+The corridor allocation problem (CAP) seeks an arrangement of + facilities along a central corridor defined by two horizontal + lines parallel to the x-axis of a Cartesian coordinate + system. The objective is to minimize the total communication + cost among facilities, while respecting two main conditions: + (i) no space is allowed between two adjacent facilities; (ii) + the left-most point of the arrangement on either line should + have zero abscissa. The conditions (i) and (ii) are required + in many applications such as the arrangement of rooms at + office buildings or hospitals. The CAP is a NP-Hard + problem. In this paper, a mixed-integer programming + formulation of the CAP is proposed, which allows us to + compute optimal layouts in reasonable time for problem + instances of moderate sizes. Moreover, heuristic procedures + are presented that can handle larger instances. +
+
+Keywords: Facility layout, Double row layout, Integer programming +
+ +
+ + +
+[30] +
+
+C. Amir, A. Badr, and I Farag. + A Fuzzy Logic Controller for Ant Algorithms. + Computing and Information Systems, 11(2):26–34, 2007.
+[ bib ] + +
+ + +
+[31] +
+
+Christophe Andrieu, Nando de Freitas, Arnaud Doucet, and Michael I. Jordan. + An Introduction to MCMC for Machine Learning. + Machine Learning, 50(1-2):5–43, 2003.
+[ bib ] + +
+ + +
+[32] +
+
+K. A. Andersen, K. Jörnsten, and M. Lind. + On bicriterion minimal spanning trees: An approximation. + Computers & Operations Research, 23(12):1171–1182, 1996.
+[ bib ] + +
+ + +
+[33] +
+
+Y. P. Aneja and K. P. K. Nair. + Bicriteria Transportation Problem. + Management Science, 25(1):73–78, 1979.
+[ bib ] + +
+ + +
+[34] +
+
+Eric Angel, Evripidis Bampis, and Laurent Gourvés. + Approximating the Pareto curve with local search for the bicriteria TSP(1,2) problem. + Theoretical Computer Science, 310(1-3):135–146, 2004.
+[ bib | +DOI ] +
+Keywords: Archiving, Local search, Multicriteria TSP, + Approximation algorithms +
+ +
+ + +
+[35] +
+
+Daniel Angus and Clinton Woodward. + Multiple Objective Ant Colony Optimisation. + Swarm Intelligence, 3(1):69–85, 2009.
+[ bib | +DOI ] + +
+ + +
+[36] +
+
+Miguel F. Anjos and Manuel V. C. Vieira. + Mathematical optimization approaches for facility layout problems: The state-of-the-art and future research directions. + European Journal of Operational Research, 261(1):1–16, 2017.
+[ bib ] + +
+ + +
+[37] +
+
+Kurt Anstreicher, Nathan Brixius, Jean-Pierre Goux, and Jeff Linderoth. + Solving large quadratic assignment problems on computational grids. + Mathematical Programming Series B, 91(3):563–588, February 2002.
+[ bib | +DOI ] +
+The quadratic assignment problem (QAP) is among the hardest + combinatorial optimization problems. Some instances of size + n = 30 have remained unsolved for decades. The solution of + these problems requires both improvements in mathematical + programming algorithms and the utilization of powerful + computational platforms. In this article we describe a novel + approach to solve QAPs using a state-of-the-art + branch-and-bound algorithm running on a federation of + geographically distributed resources known as a computational + grid. Solution of QAPs of unprecedented complexity, including + the nug30, kra30b, and tho30 instances, is reported. +
+ +
+ + +
+[38] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + Implementing the Dantzig-Fulkerson-Johnson Algorithm for Large Traveling Salesman Problems. + Mathematical Programming Series B, 97(1–2):91–153, 2003.
+[ bib | +DOI ] + +
+ + +
+[39] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + On the Solution of Traveling Salesman Problems. + Documenta Mathematica, Extra Volume ICM III:645–656, 1998.
+[ bib ] + +
+ + +
+[40] +
+
+J. S. Appleby, D. V. Blake, and E. A. Newman. + Techniques for producing school timetables on a computer and their application to other scheduling problems. + The Computer Journal, 3(4):237–245, 1961.
+[ bib | +DOI ] + +
+ + +
+[41] +
+
+David Applegate and William J. Cook. + A Computational Study of the Job-Shop Scheduling Problem. + ORSA Journal on Computing, 3(2):149–156, 1991.
+[ bib ] + +
+ + +
+[42] +
+
+David Applegate, William J. Cook, and André Rohe. + Chained Lin-Kernighan for Large Traveling Salesman Problems. + INFORMS Journal on Computing, 15(1):82–92, 2003.
+[ bib | +DOI ] + +
+ + +
+[43] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, William J. Cook, D. Espinoza, M. Goycoolea, and Keld Helsgaun. + Certification of an Optimal TSP Tour Through 85,900 Cities. + Operations Research Letters, 37(1):11–15, 2009.
+[ bib ] + +
+ + +
+[44] +
+
+Claus Aranha, Christian Leonardo Camacho-Villalón, Felipe Campelo, Marco Dorigo, Rubén Ruiz, Marc Sevaux, Kenneth Sörensen, and Thomas Stützle. + Metaphor-based Metaheuristics, a Call for Action: the Elephant in the Room. + Swarm Intelligence, 16(1):1–6, 2022.
+[ bib | +DOI ] + +
+ + +
+[45] +
+
+Claudia Archetti, Martin Savelsbergh, and M. Grazia Speranza. + The Vehicle Routing Problem with Occasional Drivers. + European Journal of Operational Research, 254(2):472–480, 2016.
+[ bib | +DOI ] + +
+ + +
+[46] +
+
+Florian Arnold, Ítalo Santana, Kenneth Sörensen, and Thibaut Vidal. + PILS: Exploring high-order neighborhoods by pattern mining and injection. + Arxiv preprint arXiv:1912.11462 [cs.AI], 2019.
+[ bib | +DOI ] + +
+ + +
+[47] +
+
+Florian Arnold and Kenneth Sörensen. + Knowledge-guided local search for the vehicle routing problem. + Computers & Operations Research, 105:32–46, 2019.
+[ bib | +DOI ] + +
+ + +
+[48] +
+
+Florian Arnold and Kenneth Sörensen. + What makes a VRP solution good? The generation of problem-specific knowledge for heuristics. + Computers & Operations Research, 106:280–288, 2019.
+[ bib | +DOI ] + +
+ + +
+[49] +
+
+Marvin A. Arostegui Jr, Sukran N. Kadipasaoglu, and Basheer M. Khumawala. + An empirical comparison of tabu search, simulated annealing, and genetic algorithms for facilities location problems. + International Journal of Production Economics, 103(2):742–754, 2006.
+[ bib ] + +
+ + +
+[50] +
+
+José Elias C. Arroyo and V. A. Armentano. + A partial enumeration heuristic for multi-objective flowshop scheduling problems. + Journal of the Operational Research Society, 55(9):1000–1007, 2004.
+[ bib ] + +
+ + +
+[51] +
+
+José Elias C. Arroyo and V. A. Armentano. + Genetic local search for multi-objective flowshop scheduling problems. + European Journal of Operational Research, 167(3):717–738, 2005.
+[ bib ] +
+Keywords: Multicriteria Scheduling +
+ +
+ + +
+[52] +
+
+José Elias C. Arroyo and Joseph Y.-T. Leung. + An Effective Iterated Greedy Algorithm for Scheduling Unrelated Parallel Batch Machines with Non-identical Capacities and Unequal Ready Times. + Computers and Industrial Engineering, 105:84–100, 2017.
+[ bib ] + +
+ + +
+[53] +
+
+N. Ascheuer, Matteo Fischetti, and M. Grötschel. + Solving asymmetric travelling salesman problem with time windows by branch-and-cut. + Mathematical Programming, 90:475–506, 2001.
+[ bib ] + +
+ + +
+[54] +
+
+John-Alexander M. Assael, Ziyu Wang, and Nando de Freitas. + Heteroscedastic Treed Bayesian Optimisation. + Arxiv preprint arXiv:1410.7172, 2014.
+[ bib | +DOI ] +
+Keywords: Treed-GP +
+ +
+ + +
+[55] +
+
+Alper Atamtürk. + On the facets of the mixed–integer knapsack polyhedron. + Mathematical Programming, 98(1):145–175, 2003.
+[ bib | +DOI ] + +
+ + +
+[56] +
+
+Charles Audet, Cong-Kien Dang, and Dominique Orban. + Optimization of Algorithms with OPAL. + Mathematical Programming Computation, 6(3):233–254, 2014.
+[ bib ] + +
+ + +
+[57] +
+
+P. Audze and Vilnis Eglãjs. + New approach to the design of multifactor experiments. + Problems of Dynamics and Strengths, 35:104–107, 1977. + (in Russian).
+[ bib ] + +
+ + +
+[58] +
+
+Charles Audet and Dominique Orban. + Finding Optimal Algorithmic Parameters Using Derivative-Free Optimization. + SIAM Journal on Optimization, 17(3):642–664, 2006.
+[ bib | +DOI ] +
+Keywords: mesh adaptive direct search; pattern search +
+ +
+ + +
+[59] +
+
+Peter Auer. + Using Confidence Bounds for Exploitation-Exploration Trade-offs. + Journal of Machine Learning Research, 3:397–422, November 2002.
+[ bib ] +
+We show how a standard tool from statistics — namely + confidence bounds — can be used to elegantly deal with + situations which exhibit an exploitation-exploration + trade-off. Our technique for designing and analyzing + algorithms for such situations is general and can be applied + when an algorithm has to make exploitation-versus-exploration + decisions based on uncertain information provided by a random + process. We apply our technique to two models with such an + exploitation-exploration trade-off. For the adversarial + bandit problem with shifting our new algorithm suffers only + O((ST)1/2) regret with high probability over T trials + with S shifts. Such a regret bound was previously known + only in expectation. The second model we consider is + associative reinforcement learning with linear value + functions. For this model our technique improves the regret + from O(T3/4) to O(T1/2). +
+ +
+ + +
+[60] +
+
+Peter Auer, Nicolo Cesa-Bianchi, and Paul Fischer. + Finite-time analysis of the multiarmed bandit problem. + Machine Learning, 47(2-3):235–256, 2002.
+[ bib ] + +
+ + +
+[61] +
+
+Anne Auger, Johannes Bader, Dimo Brockhoff, and Eckart Zitzler. + Hypervolume-based multiobjective optimization: Theoretical foundations and practical implications. + Theoretical Computer Science, 425:75–103, 2012.
+[ bib | +DOI ] + +
+ + +
+[62] +
+
+Mustafa Avci and Seyda Topaloglu. + A Multi-start Iterated Local Search Algorithm for the Generalized Quadratic Multiple Knapsack Problem. + Computers & Operations Research, 83:54–65, 2017.
+[ bib ] + +
+ + +
+[63] +
+
+Andreea Avramescu, Richard Allmendinger, and Manuel López-Ibáñez. + Managing Manufacturing and Delivery of Personalised Medicine: Current and Future Models. + Arxiv preprint arXiv:2105.12699 [econ.GN], 2021.
+[ bib | +http ] + +
+ + +
+[64] +
+
+Doǧan Aydın, Gürcan Yavuz, and Thomas Stützle. + ABC-X: A Generalized, Automatically Configurable Artificial Bee Colony Framework. + Swarm Intelligence, 11(1):1–38, 2017.
+[ bib ] + +
+ + +
+[65] +
+
+Mayowa Ayodele, Richard Allmendinger, Manuel López-Ibáñez, and Matthieu Parizy. + A Study of Scalarisation Techniques for Multi-Objective QUBO Solving. + Arxiv preprint arXiv:2210.11321, 2022.
+[ bib | +DOI ] + +
+ + +
+[66] +
+
+Mahdi Aziz and Mohammad-H. Tayarani-N. + An adaptive memetic Particle Swarm Optimization algorithm for finding large-scale Latin hypercube designs. + Engineering Applications of Artificial Intelligence, 36:222–237, 2014.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[67] +
+
+François Bachoc, Céline Helbert, and Victor Picheny. + Gaussian process optimization with failures: Classification and convergence proof. + Journal of Global Optimization, 2020.
+[ bib | +DOI | +epub ] +
+We consider the optimization of a computer model where each + simulation either fails or returns a valid output + performance. We first propose a new joint Gaussian process + model for classification of the inputs (computation failure + or success) and for regression of the performance + function. We provide results that allow for a computationally + efficient maximum likelihood estimation of the covariance + parameters, with a stochastic approximation of the likelihood + gradient. We then extend the classical improvement criterion + to our setting of joint classification and regression. We + provide an efficient computation procedure for the extended + criterion and its gradient. We prove the almost sure + convergence of the global optimization algorithm following + from this extended criterion. We also study the practical + performances of this algorithm, both on simulated data and on + a real computer model in the context of automotive fan + design. +
+
+Keywords: crashed simulation; latent gaussian process; automotive fan + design; industrial application; GP classification; Expected + Feasible Improvement with Gaussian Process Classification + with signs; EFI GPC sign +
+ +
+ + +
+[68] +
+
+Johannes Bader and Eckart Zitzler. + HypE: An Algorithm for Fast Hypervolume-Based Many-Objective Optimization. + Evolutionary Computation, 19(1):45–76, 2011.
+[ bib | +DOI ] + +
+ + +
+[69] +
+
+Hossein Baharmand, Tina Comes, and Matthieu Lauras. + Bi-objective multi-layer location– allocation model for the immediate aftermath of sudden-onset disasters. + Transportation Research Part E: Logistics and Transportation Review, 127:86–110, 2019.
+[ bib | +DOI ] +
+Locating distribution centers is critical for humanitarians + in the immediate aftermath of a sudden-onset disaster. A + major challenge lies in balancing the complexity and + uncertainty of the problem with time and resource + constraints. To address this problem, we propose a + location-allocation model that divides the topography of + affected areas into multiple layers; considers constrained + number and capacity of facilities and fleets; and allows + decision-makers to explore trade-offs between response time + and logistics costs. To illustrate our theoretical work, we + apply the model to a real dataset from the 2015 Nepal + earthquake response. For this case, our method results in a + considerable reduction of logistics costs. +
+ +
+ + +
+[70] +
+
+Monya Baker. + Is there a reproducibility crisis? + Nature, 533:452–454, 2016.
+[ bib ] + +
+ + +
+[71] +
+
+Edward K. Baker. + An Exact Algorithm for the Time-Constrained Traveling Salesman Problem. + Operations Research, 31(5):938–945, 1983.
+[ bib | +DOI ] + +
+ + +
+[72] +
+
+Burcu Balcik and Benita M. Beamon. + Facility location in humanitarian relief. + International Journal of Logistics, 11(2):101–121, 2008.
+[ bib ] + +
+ + +
+[73] +
+
+Prasanna Balaprakash, Mauro Birattari, Thomas Stützle, and Marco Dorigo. + Adaptive Sampling Size and Importance Sampling in Estimation-based Local Search for the Probabilistic Traveling Salesman Problem. + European Journal of Operational Research, 199(1):98–110, 2009.
+[ bib ] + +
+ + +
+[74] +
+
+Prasanna Balaprakash, Mauro Birattari, Thomas Stützle, and Marco Dorigo. + Estimation-based Metaheuristics for the Probabilistic Travelling Salesman Problem. + Computers & Operations Research, 37(11):1939–1951, 2010.
+[ bib | +DOI ] + +
+ + +
+[75] +
+
+Prasanna Balaprakash, Mauro Birattari, Thomas Stützle, and Marco Dorigo. + Estimation-based Metaheuristics for the Single Vehicle Routing Problem with Stochastic Demands and Customers. + Computational Optimization and Applications, 61(2):463–487, 2015.
+[ bib | +DOI ] + +
+ + +
+[76] +
+
+Prasanna Balaprakash, Mauro Birattari, Thomas Stützle, Zhi Yuan, and Marco Dorigo. + Estimation-based Ant Colony Optimization Algorithms for the Probabilistic Travelling Salesman Problem. + Swarm Intelligence, 3(3):223–242, 2009.
+[ bib ] + +
+ + +
+[77] +
+
+Egon Balas and M. C. Carrera. + A Dynamic Subgradient-based Branch and Bound Procedure for Set Covering. + Operations Research, 44(6):875–890, 1996.
+[ bib ] + +
+ + +
+[78] +
+
+Egon Balas and C. Martin. + Pivot and Complement–A Heuristic for 0–1 Programming. + Management Science, 26(1):86–96, 1980.
+[ bib ] + +
+ + +
+[79] +
+
+Egon Balas and M. W. Padberg. + Set Partitioning: A Survey. + SIAM Review, 18:710–760, 1976.
+[ bib ] + +
+ + +
+[80] +
+
+Egon Balas and Neil Simonetti. + Linear Time Dynamic-Programming Algorithms for New Classes of Restricted TSPs: A Computational Study. + INFORMS Journal on Computing, 13(1):56–75, 2001.
+[ bib | +DOI ] +
+Consider the following restricted (symmetric or + asymmetric) traveling-salesman problem (TSP): + given an initial ordering of the n cities and an + integer k > 0, find a minimum-cost + feasible tour, where a feasible tour is one in which + city i precedes city j whenever j >= i + k in the + initial ordering. Balas (1996) has proposed a + dynamic-programming algorithm that solves this + problem in time linear in n, though exponential in + k. Some important real-world problems are amenable + to this model or some of its close relatives. The + algorithm of Balas (1996) constructs a layered + network with a layer of nodes for each position in + the tour, such that source-sink paths in this + network are in one-to-one correspondence with tours + that satisfy the postulated precedence + constraints. In this paper we discuss an + implementation of the dynamic-programming algorithm + for the general case when the integer k is replaced + with city-specific integers k(j), j = 1, . . ., + n. We discuss applications to, and computational + experience with, TSPs with time windows, a model + frequently used in vehicle routing as well as in + scheduling with setup, release and delivery + times. We also introduce a new model, the TSP with + target times, applicable to Just-in-Time + scheduling problems. Finally for TSPs that have no + precedence restrictions, we use the algorithm as a + heuristic that finds in linear time a local optimum + over an exponential-size neighborhood. For this + case, we implement an iterated version of our + procedure, based on contracting some arcs of the + tour produced by a first application of the + algorithm, then reapplying the algorithm to the + shrunken graph with the same k. +
+
+Keywords: tsptw +
+ +
+ + +
+[81] +
+
+Egon Balas and A. Vazacopoulos. + Guided Local Search with Shifting Bottleneck for Job Shop Scheduling. + Management Science, 44(2):262–275, 1998.
+[ bib ] + +
+ + +
+[82] +
+
+Steven C. Bankes. + Tools and techniques for developing policies for complex and uncertain systems. + Proceedings of the National Academy of Sciences, 99(suppl 3):7263–7266, 2002.
+[ bib | +DOI ] +
+Agent-based models (ABM) are examples of complex adaptive + systems, which can be characterized as those systems for + which no model less complex than the system itself can + accurately predict in detail how the system will behave at + future times. Consequently, the standard tools of policy + analysis, based as they are on devising policies that perform + well on some best estimate model of the system, cannot be + reliably used for ABM. This paper argues that policy analysis + by using ABM requires an alternative approach to decision + theory. The general characteristics of such an approach are + described, and examples are provided of its application to + policy analysis.ABM, agent-based model +
+ +
+ + +
+[83] +
+
+Eduardo Batista de Moraes Barbosa, Edson Luiz Francça Senne, and Messias Borges Silva. + Improving the Performance of Metaheuristics: An Approach Combining Response Surface Methodology and Racing Algorithms. + International Journal of Engineering Mathematics, 2015:Article ID 167031, 2015.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[84] +
+
+Alejandro Barredo Arrieta, Natalia Díaz-Rodríguez, Javier Del Ser, Adrien Bennetot, Siham Tabik, Alberto Barbado, Salvador Garcia, Sergio Gil-Lopez, Daniel Molina, Richard Benjamins, Raja Chatila, and Francisco Herrera. + Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI. + Information Fusion, 58:82–115, June 2020.
+[ bib | +DOI ] + +
+ + +
+[85] +
+
+Thomas Bartz-Beielstein, Carola Doerr, Daan van den Berg, Jakob Bossek, Sowmya Chandrasekaran, Tome Eftimov, Andreas Fischbach, Pascal Kerschke, William La Cava, Manuel López-Ibáñez, Katherine M. Malan, Jason H. Moore, Boris Naujoks, Patryk Orzechowski, Vanessa Volz, Markus Wagner, and Thomas Weise. + Benchmarking in Optimization: Best Practice and Open Issues. + Arxiv preprint arXiv:2007.03488 [cs.NE], 2020.
+[ bib | +http ] + +
+ + +
+[86] +
+
+Richard S. Barr, Bruce L. Golden, James P. Kelly, Mauricio G. C. Resende, and Jr. William R. Stewart. + Designing and Reporting on Computational Experiments with Heuristic Methods. + Journal of Heuristics, 1(1):9–32, 1995.
+[ bib | +DOI ] + +
+ + +
+[87] +
+
+Cynthia Barnhart, Ellis L. Johnson, George L. Nemhauser, Martin W. P. Savelsbergh, and Pamela H. Vance. + Branch-and-price: Column generation for solving huge integer programs. + Operations Research, 46(3):316–329, 1998.
+[ bib ] + +
+ + +
+[88] +
+
+Erin Bartholomew and Jan H. Kwakkel. + On considering robustness in the search phase of Robust Decision Making: A comparison of Many-Objective Robust Decision Making, multi-scenario Many-Objective Robust Decision Making, and Many Objective Robust Optimization. + Environmental Modelling & Software, 127:104699, 2020.
+[ bib | +DOI ] + +
+ + +
+[89] +
+
+Elias Bareinboim and Judea Pearl. + Causal inference and the data-fusion problem. + Proceedings of the National Academy of Sciences, 113(27):7345–7352, 2016.
+[ bib | +DOI ] + +
+ + +
+[90] +
+
+Thomas Bartz-Beielstein and Martin Zaefferer. + Model-based methods for continuous and discrete global optimization. + Applied Soft Computing, 55:154–167, June 2017.
+[ bib | +DOI ] + +
+ + +
+[91] +
+
+Atanu Basu and L. Neil Frazer. + Rapid Determination of the Critical Temperature in Simulated Annealing Inversion. + Science, 249(4975):1409–1412, 1990.
+[ bib ] + +
+ + +
+[92] +
+
+Roberto Battiti and Andrea Passerini. + Brain-Computer Evolutionary Multiobjective Optimization: A Genetic Algorithm Adapting to the Decision Maker. + IEEE Transactions on Evolutionary Computation, 14(5):671–687, 2010.
+[ bib | +DOI ] +
+Errata: DTLZ6 and DTLZ7 in the paper are actually DTLZ7 and + DTLZ8 in [1731] +
+
+Keywords: BC-EMOA +
+ +
+ + +
+[93] +
+
+Roberto Battiti and M. Protasi. + Reactive Search, A History-Based Heuristic for MAX-SAT. + ACM Journal of Experimental Algorithmics, 2, 1997.
+[ bib ] + +
+ + +
+[94] +
+
+Michele Battistutta, Andrea Schaerf, and Tommaso Urli. + Feature-based Tuning of Single-stage Simulated Annealing for Examination Timetabling. + Annals of Operations Research, 252(2):239–254, 2017.
+[ bib ] + +
+ + +
+[95] +
+
+Roberto Battiti and Giampietro Tecchiolli. + Simulated annealing and Tabu search in the long run: A comparison on QAP tasks. + Computer and Mathematics with Applications, 28(6):1–8, 1994.
+[ bib | +DOI ] + +
+ + +
+[96] +
+
+Roberto Battiti and Giampietro Tecchiolli. + The Reactive Tabu Search. + ORSA Journal on Computing, 6(2):126–140, 1994.
+[ bib ] + +
+ + +
+[97] +
+
+Roberto Battiti and Giampietro Tecchiolli. + The continuous reactive tabu search: blending combinatorial optimization and stochastic search for global optimization. + Annals of Operations Research, 63(2):151–188, 1996.
+[ bib | +DOI ] + +
+ + +
+[98] +
+
+J. Bautista and J. Pereira. + Ant algorithms for a time and space constrained assembly line balancing problem. + European Journal of Operational Research, 177(3):2016–2032, 2007.
+[ bib | +DOI ] + +
+ + +
+[99] +
+
+William J. Baumol. + Management models and industrial applications of linear programming. + Naval Research Logistics Quarterly, 9(1):63–64, 1962.
+[ bib | +DOI ] + +
+ + +
+[100] +
+
+John Baxter. + Local Optima Avoidance in Depot Location. + Journal of the Operational Research Society, 32(9):815–819, 1981.
+[ bib ] + +
+ + +
+[101] +
+
+John E. Beasley and P. C. Chu. + A Genetic Algorithm for the Set Covering Problem. + European Journal of Operational Research, 94(2):392–404, 1996.
+[ bib ] + +
+ + +
+[102] +
+
+John E. Beasley and P. C. Chu. + A Genetic Algorithm for the Multidimensional Knapsack Problem. + Journal of Heuristics, 4(1):63–86, 1998.
+[ bib ] + +
+ + +
+[103] +
+
+Jennifer Bealt, Duncan Shaw, Chris M. Smith, and Manuel López-Ibáñez. + Peer Reviews for Making Cities Resilient: A Systematic Literature Review. + International Journal of Emergency Management, 15(4):334–359, 2019.
+[ bib | +DOI ] +
+Peer reviews are a unique governance tool that use expertise + from one city or country to assess and strengthen the + capabilities of another. Peer review tools are gaining + momentum in disaster management and remain an important but + understudied topic in risk governance. Methodologies to + conduct a peer review are still in their infancy. To enhance + these, a systematic literature review (SLR) of academic and + non-academic literature was conducted on city resilience peer + reviews. Thirty-three attributes of resilience are + identified, which provides useful insights into how research + and practice can inform risk governance, and utilise peer + reviews, to drive meaningful change. Moreover, it situates + the challenges associated with resilience building tools + within risk governance to support the development of + interdisciplinary perspectives for integrated city resilience + frameworks. Results of this research have been used to + develop a peer review methodology and an international + standard on conducting peer reviews for disaster risk + reduction. +
+
+Keywords: city resilience, city peer review, disaster risk governance +
+ +
+ + +
+[104] +
+
+John E. Beasley. + OR-Library: distributing test problems by electronic mail. + Journal of the Operational Research Society, pp.  1069–1072, 1990. + Currently available from http://people.brunel.ac.uk/~mastjjb/jeb/info.html.
+[ bib ] + +
+ + +
+[105] +
+
+J. Behnamian and S.M.T. Fatemi Ghomi. + Hybrid Flowshop Scheduling with Machine and Resource-dependent Processing Times. + Applied Mathematical Modelling, 35(3):1107–1123, 2011.
+[ bib ] + +
+ + +
+[106] +
+
+Richard Bellman. + The theory of dynamic programming. + Bulletin of the American Mathematical Society, 60:503–515, 1954.
+[ bib ] + +
+ + +
+[107] +
+
+Ruggero Bellio, Sara Ceschia, Luca Di Gaspero, Andrea Schaerf, and Tommaso Urli. + Feature-based tuning of simulated annealing applied to the curriculum-based course timetabling problem. + Computers & Operations Research, 65:83–92, 2016.
+[ bib ] + +
+ + +
+[108] +
+
+Jon Louis Bentley. + Fast Algorithms for Geometric Traveling Salesman Problems. + ORSA Journal on Computing, 4(4):387–411, 1992.
+[ bib ] + +
+ + +
+[109] +
+
+Una Benlic and Jin-Kao Hao. + Breakout Local Search for the Quadratic Assignment Problem. + Applied Mathematics and Computation, 219(9):4800–4815, 2013.
+[ bib ] + +
+ + +
+[110] +
+
+Calem J. Bendell, Shalon Liu, Tristan Aumentado-Armstrong, Bogdan Istrate, Paul T. Cernek, Samuel Khan, Sergiu Picioreanu, Michael Zhao, and Robert A. Murgita. + Transient protein-protein interface prediction: datasets, features, algorithms, and the RAD-T predictor. + BMC Bioinformatics, 15:82, 2014.
+[ bib ] + +
+ + +
+[111] +
+
+Yoshua Bengio, Andrea Lodi, and Antoine Prouvost. + Machine learning for combinatorial optimization: A methodological tour d'horizon. + European Journal of Operational Research, 290(2):405–421, 2021.
+[ bib | +DOI ] +
+This paper surveys the recent attempts, both from the machine + learning and operations research communities, at leveraging + machine learning to solve combinatorial optimization + problems. Given the hard nature of these problems, + state-of-the-art algorithms rely on handcrafted heuristics + for making decisions that are otherwise too expensive to + compute or mathematically not well defined. Thus, machine + learning looks like a natural candidate to make such + decisions in a more principled and optimized way. We advocate + for pushing further the integration of machine learning and + combinatorial optimization and detail a methodology to do + so. A main point of the paper is seeing generic optimization + problems as data points and inquiring what is the relevant + distribution of problems to use for learning on a given + task. +
+
+Keywords: Combinatorial optimization, Machine learning, Branch and + bound, Mixed-integer programming solvers +
+ +
+ + +
+[112] +
+
+Alexander Javier Benavides and Marcus Ritt. + Two Simple and Effective Heuristics for Minimizing the Makespan in Non-permutation Flow Shops. + Computers & Operations Research, 66:160–169, 2016.
+[ bib | +DOI ] + +
+ + +
+[113] +
+
+J. F. Benders. + Partitioning Procedures for Solving Mixed-variables Programming Problems. + Numerische Mathematik, 4(3):238–252, 1962.
+[ bib ] + +
+ + +
+[114] +
+
+Jon Louis Bentley. + Multidimensional Divide-and-conquer. + Communications of the ACM, 23(4):214–229, 1980.
+[ bib | +DOI ] +
+Most results in the field of algorithm design are single + algorithms that solve single problems. In this paper we + discuss multidimensional divide-and-conquer, an algorithmic + paradigm that can be instantiated in many different ways to + yield a number of algorithms and data structures for + multidimensional problems. We use this paradigm to give + best-known solutions to such problems as the ECDF, maxima, + range searching, closest pair, and all nearest neighbor + problems. The contributions of the paper are on two + levels. On the first level are the particular algorithms and + data structures given by applying the paradigm. On the + second level is the more novel contribution of this paper: a + detailed study of an algorithmic paradigm that is specific + enough to be described precisely yet general enough to solve + a wide variety of problems. +
+ +
+ + +
+[115] +
+
+James S. Bergstra and Yoshua Bengio. + Random Search for Hyper-Parameter Optimization. + Journal of Machine Learning Research, 13:281–305, 2012.
+[ bib | +epub ] +
+Grid search and manual search are the most widely + used strategies for hyper-parameter + optimization. This paper shows empirically and + theoretically that randomly chosen trials are more + efficient for hyper-parameter optimization than + trials on a grid. Empirical evidence comes from a + comparison with a large previous study that used + grid search and manual search to configure neural + networks and deep belief networks. Compared with + neural networks configured by a pure grid search, we + find that random search over the same domain is able + to find models that are as good or better within a + small fraction of the computation time. Granting + random search the same computational budget, random + search finds better models by effectively searching + a larger, less promising configuration + space. Compared with deep belief networks configured + by a thoughtful combination of manual search and + grid search, purely random search over the same + 32-dimensional configuration space found + statistically equal performance on four of seven + data sets, and superior performance on one of + seven. A Gaussian process analysis of the function + from hyper-parameters to validation set performance + reveals that for most data sets only a few of the + hyper-parameters really matter, but that different + hyper-parameters are important on different data + sets. This phenomenon makes grid search a poor + choice for configuring algorithms for new data + sets. Our analysis casts some light on why recent + "High Throughput" methods achieve surprising + success: they appear to search through a large number + of hyper-parameters because most hyper-parameters do + not matter much. We anticipate that growing interest + in large hierarchical models will place an + increasing burden on techniques for hyper-parameter + optimization; this work shows that random search is + a natural baseline against which to judge progress + in the development of adaptive (sequential) + hyper-parameter optimization algorithms. +
+ +
+ + +
+[116] +
+
+Loïc Berger, Johannes Emmerling, and Massimo Tavoni. + Managing catastrophic climate risks under model uncertainty aversion. + Management Science, 63(3):749–765, 2017.
+[ bib ] + +
+ + +
+[117] +
+
+Livio Bertacco, Matteo Fischetti, and Andrea Lodi. + A feasibility pump heuristic for general mixed-integer problems. + Discrete Optimization, 4(1):63–76, 2007.
+[ bib ] + +
+ + +
+[118] +
+
+Dimitris Bertsimas and Nathan Kallus. + From predictive to prescriptive analytics. + Management Science, 66(3):1025–1044, 2020.
+[ bib ] + +
+ + +
+[119] +
+
+Felix Berkenkamp, Andreas Krause, and Angela P. Schoellig. + Bayesian Optimization with Safety Constraints: Safe and Automatic Parameter Tuning in Robotics. + Arxiv preprint arXiv:1602.04450, 2016.
+[ bib | +http ] +
+Keywords: Safe Optimization, SafeOpt +
+ +
+ + +
+[120] +
+
+Felix Berkenkamp, Andreas Krause, and Angela P. Schoellig. + Bayesian optimization with safety constraints: safe and automatic parameter tuning in robotics. + Machine Learning, June 2021.
+[ bib | +DOI ] +
+Selecting the right tuning parameters for algorithms is a + pravelent problem in machine learning that can significantly + affect the performance of algorithms. Data-efficient + optimization algorithms, such as Bayesian optimization, have + been used to automate this process. During experiments on + real-world systems such as robotic platforms these methods + can evaluate unsafe parameters that lead to safety-critical + system failures and can destroy the system. Recently, a safe + Bayesian optimization algorithm, called SafeOpt, has been + developed, which guarantees that the performance of the + system never falls below a critical value; that is, safety is + defined based on the performance function. However, coupling + performance and safety is often not desirable in practice, + since they are often opposing objectives. In this paper, we + present a generalized algorithm that allows for multiple + safety constraints separate from the objective. Given an + initial set of safe parameters, the algorithm maximizes + performance but only evaluates parameters that satisfy safety + for all constraints with high probability. To this end, it + carefully explores the parameter space by exploiting + regularity assumptions in terms of a Gaussian process + prior. Moreover, we show how context variables can be used to + safely transfer knowledge to new situations and tasks. We + provide a theoretical analysis and demonstrate that the + proposed algorithm enables fast, automatic, and safe + optimization of tuning parameters in experiments on a + quadrotor vehicle. +
+
+Preprint: http://arxiv.org/abs/1602.04450 +
+ +
+ + +
+[121] +
+
+Dimitri P. Bertsekas, John N. Tsitsiklis, and Cynara Wu. + Rollout Algorithms for Combinatorial Optimization. + Journal of Heuristics, 3(3):245–262, 1997.
+[ bib ] + +
+ + +
+[122] +
+
+Judith O. Berkey and Pearl Y. Wang. + Two-dimensional finite bin-packing algorithms. + Journal of the Operational Research Society, 38(5):423–429, 1987.
+[ bib | +DOI ] + +
+ + +
+[123] +
+
+Nicola Beume, Carlos M. Fonseca, Manuel López-Ibáñez, Luís Paquete, and Jan Vahrenhold. + On the complexity of computing the hypervolume indicator. + IEEE Transactions on Evolutionary Computation, 13(5):1075–1082, 2009.
+[ bib | +DOI ] +
+The goal of multi-objective optimization is to find + a set of best compromise solutions for typically + conflicting objectives. Due to the complex nature of + most real-life problems, only an approximation to + such an optimal set can be obtained within + reasonable (computing) time. To compare such + approximations, and thereby the performance of + multi-objective optimizers providing them, unary + quality measures are usually applied. Among these, + the hypervolume indicator (or + S-metric) is of particular relevance due to + its favorable properties. Moreover, this indicator + has been successfully integrated into stochastic + optimizers, such as evolutionary algorithms, where + it serves as a guidance criterion for finding good + approximations to the Pareto front. Recent results + show that computing the hypervolume indicator can be + seen as solving a specialized version of Klee's + Measure Problem. In general, Klee's Measure Problem + can be solved with O(n log n + + nd/2log n) comparisons for an input instance of + size n in d dimensions; as of this writing, it + is unknown whether a lower bound higher than + Ω(n log n) can be proven. In this article, + we derive a lower bound of Ω(nlog n) for the + complexity of computing the hypervolume indicator in + any number of dimensions d>1 by reducing the + so-called UniformGap problem to it. For + the three dimensional case, we also present a + matching upper bound of O(nlog n) + comparisons that is obtained by extending an + algorithm for finding the maxima of a point set. +
+ +
+ + +
+[124] +
+
+Nicola Beume, Boris Naujoks, and Michael T. M. Emmerich. + SMS-EMOA: Multiobjective selection based on dominated hypervolume. + European Journal of Operational Research, 181(3):1653–1669, 2007.
+[ bib | +DOI ] + +
+ + +
+[125] +
+
+Hans-Georg Beyer and Hans-Paul Schwefel. + Evolution Strategies: A Comprehensive Introduction. + Natural Computing, 1:3–52, 2002.
+[ bib ] + +
+ + +
+[126] +
+
+Hans-Georg Beyer, Hans-Paul Schwefel, and Ingo Wegener. + How to analyse evolutionary algorithms. + Theoretical Computer Science, 287(1):101–130, 2002.
+[ bib ] + +
+ + +
+[127] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Component-Wise Design of Multi-Objective Evolutionary Algorithms. + IEEE Transactions on Evolutionary Computation, 20(3):403–417, 2016.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[128] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + A Large-Scale Experimental Evaluation of High-Performing Multi- and Many-Objective Evolutionary Algorithms. + Evolutionary Computation, 26(4):621–656, 2018.
+[ bib | +DOI | +supplementary material ] +
+Research on multi-objective evolutionary algorithms (MOEAs) + has produced over the past decades a large number of + algorithms and a rich literature on performance assessment + tools to evaluate and compare them. Yet, newly proposed MOEAs + are typically compared against very few, often a decade older + MOEAs. One reason for this apparent contradiction is the lack + of a common baseline for comparison, with each subsequent + study often devising its own experimental scenario, slightly + different from other studies. As a result, the state of the + art in MOEAs is a disputed topic. This article reports a + systematic, comprehensive evaluation of a large number of + MOEAs that covers a wide range of experimental scenarios. A + novelty of this study is the separation between the + higher-level algorithmic components related to + multi-objective optimization (MO), which characterize each + particular MOEA, and the underlying parameters-such as + evolutionary operators, population size, etc.-whose + configuration may be tuned for each scenario. Instead of + relying on a common or "default" parameter configuration that + may be low-performing for particular MOEAs or scenarios and + unintentionally biased, we tune the parameters of each MOEA + for each scenario using automatic algorithm configuration + methods. Our results confirm some of the assumed knowledge in + the field, while at the same time they provide new insights + on the relative performance of MOEAs for many-objective + problems. For example, under certain conditions, + indicator-based MOEAs are more competitive for such problems + than previously assumed. We also analyze problem-specific + features affecting performance, the agreement between + performance metrics, and the improvement of tuned + configurations over the default configurations used in the + literature. Finally, the data produced is made publicly + available to motivate further analysis and a baseline for + future comparisons. +
+ +
+ + +
+[129] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatically Designing State-of-the-Art Multi- and Many-Objective Evolutionary Algorithms. + Evolutionary Computation, 28(2):195–226, 2020.
+[ bib | +DOI | +supplementary material ] +
+A recent comparison of well-established multiobjective + evolutionary algorithms (MOEAs) has helped better identify + the current state-of-the-art by considering (i) parameter + tuning through automatic configuration, (ii) a wide range of + different setups, and (iii) various performance + metrics. Here, we automatically devise MOEAs with verified + state-of-the-art performance for multi- and many-objective + continuous optimization. Our work is based on two main + considerations. The first is that high-performing algorithms + can be obtained from a configurable algorithmic framework in + an automated way. The second is that multiple performance + metrics may be required to guide this automatic design + process. In the first part of this work, we extend our + previously proposed algorithmic framework, increasing the + number of MOEAs, underlying evolutionary algorithms, and + search paradigms that it comprises. These components can be + combined following a general MOEA template, and an automatic + configuration method is used to instantiate high-performing + MOEA designs that optimize a given performance metric and + present state-of-the-art performance. In the second part, we + propose a multiobjective formulation for the automatic MOEA + design, which proves critical for the context of + many-objective optimization due to the disagreement of + established performance metrics. Our proposed formulation + leads to an automatically designed MOEA that presents + state-of-the-art performance according to a set of metrics, + rather than a single one. +
+ +
+ + +
+[130] +
+
+Leonora Bianchi, Mauro Birattari, M. Manfrin, M. Mastrolilli, Luís Paquete, O. Rossi-Doria, and Tommaso Schiavinotto. + Hybrid Metaheuristics for the Vehicle Routing Problem with Stochastic Demands. + Journal of Mathematical Modelling and Algorithms, 5(1):91–110, 2006.
+[ bib ] + +
+ + +
+[131] +
+
+Leonora Bianchi, Marco Dorigo, L. M. Gambardella, and Walter J. Gutjahr. + A survey on metaheuristics for stochastic combinatorial optimization. + Natural Computing, 8(2):239–287, 2009.
+[ bib ] + +
+ + +
+[132] +
+
+M. Binois, D. Ginsbourger, and O. Roustant. + Quantifying uncertainty on Pareto fronts with Gaussian process conditional simulations. + European Journal of Operational Research, 243(2):386–394, 2015.
+[ bib | +DOI ] +
+Multi-objective optimization algorithms aim at finding + Pareto-optimal solutions. Recovering Pareto fronts or Pareto + sets from a limited number of function evaluations are + challenging problems. A popular approach in the case of + expensive-to-evaluate functions is to appeal to + metamodels. Kriging has been shown efficient as a base for + sequential multi-objective optimization, notably through + infill sampling criteria balancing exploitation and + exploration such as the Expected Hypervolume + Improvement. Here we consider Kriging metamodels not only for + selecting new points, but as a tool for estimating the whole + Pareto front and quantifying how much uncertainty remains on + it at any stage of Kriging-based multi-objective optimization + algorithms. Our approach relies on the Gaussian process + interpretation of Kriging, and bases upon conditional + simulations. Using concepts from random set theory, we + propose to adapt the Vorob'ev expectation and deviation to + capture the variability of the set of non-dominated + points. Numerical experiments illustrate the potential of the + proposed workflow, and it is shown on examples how Gaussian + process simulations and the estimated Vorob'ev deviation can + be used to monitor the ability of Kriging-based + multi-objective optimization algorithms to accurately learn + the Pareto front. +
+
+Keywords: Attainment function, Expected Hypervolume Improvement, + Kriging, Multi-objective optimization, Vorob'ev expectation +
+ +
+ + +
+[133] +
+
+Mauro Birattari, Prasanna Balaprakash, Thomas Stützle, and Marco Dorigo. + Estimation Based Local Search for Stochastic Combinatorial Optimization. + INFORMS Journal on Computing, 20(4):644–658, 2008.
+[ bib ] + +
+ + +
+[134] +
+
+Mauro Birattari, Paola Pellegrini, and Marco Dorigo. + On the invariance of ant colony optimization. + IEEE Transactions on Evolutionary Computation, 11(6):732–742, 2007.
+[ bib | +DOI ] + +
+ + +
+[135] +
+
+Mauro Birattari, M. Zlochin, and Marco Dorigo. + Towards a theory of practice in metaheuristics design: A machine learning perspective. + Theoretical Informatics and Applications, 40(2):353–369, 2006.
+[ bib ] + +
+ + +
+[136] +
+
+Francesco Biscani, Dario Izzo, and Chit Hong Yam. + A Global Optimisation Toolbox for Massively Parallel Engineering Optimisation. + Arxiv preprint arXiv:1004.3824, 2010.
+[ bib | +http ] +
+A software platform for global optimisation, called PaGMO, + has been developed within the Advanced Concepts Team (ACT) at + the European Space Agency, and was recently released as an + open-source project. PaGMO is built to tackle + high-dimensional global optimisation problems, and it has + been successfully used to find solutions to real-life + engineering problems among which the preliminary design of + interplanetary spacecraft trajectories - both chemical + (including multiple flybys and deep-space maneuvers) and + low-thrust (limited, at the moment, to single phase + trajectories), the inverse design of nano-structured + radiators and the design of non-reactive controllers for + planetary rovers. Featuring an arsenal of global and local + optimisation algorithms (including genetic algorithms, + differential evolution, simulated annealing, particle swarm + optimisation, compass search, improved harmony search, and + various interfaces to libraries for local optimisation such + as SNOPT, IPOPT, GSL and NLopt), PaGMO is at its core a C++ + library which employs an object-oriented architecture + providing a clean and easily-extensible optimisation + framework. Adoption of multi-threaded programming ensures the + efficient exploitation of modern multi-core architectures and + allows for a straightforward implementation of the island + model paradigm, in which multiple populations of candidate + solutions asynchronously exchange information in order to + speed-up and improve the optimisation process. In addition to + the C++ interface, PaGMO's capabilities are exposed to the + high-level language Python, so that it is possible to easily + use PaGMO in an interactive session and take advantage of the + numerous scientific Python libraries available. +
+
+Keywords: PaGMO +
+ +
+ + +
+[137] +
+
+Bernd Bischl, Martin Binder, Michel Lang, Tobias Pielok, Jakob Richter, Stefan Coors, Janek Thomas, Theresa Ullmann, Marc Becker, Anne-Laure Boulesteix, Difan Deng, and Marius Thomas Lindauer. + Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges. + Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 13(2):e1484, 2023.
+[ bib ] + +
+ + +
+[138] +
+
+Bernd Bischl, Pascal Kerschke, Lars Kotthoff, Marius Thomas Lindauer, Yuri Malitsky, Alexandre Fréchette, Holger H. Hoos, Frank Hutter, Kevin Leyton-Brown, Kevin Tierney, and Joaquin Vanschoren. + ASlib: A Benchmark Library for Algorithm Selection. + Artificial Intelligence, 237:41–58, 2016.
+[ bib ] + +
+ + +
+[139] +
+
+Bernd Bischl, Michel Lang, Lars Kotthoff, Julia Schiffner, Jakob Richter, Erich Studerus, Giuseppe Casalicchio, and Zachary M. Jones. + mlr: Machine Learning in R. + Journal of Machine Learning Research, 17(170):1–5, 2016.
+[ bib | +epub ] + +
+ + +
+[140] +
+
+Xavier Blasco, Juan M. Herrero, Javier Sanchis, and Manuel Martínez. + A new graphical visualization of n-dimensional Pareto front for decision-making in multiobjective optimization. + Information Sciences, 178(20):3908–3924, 2008.
+[ bib ] + +
+ + +
+[141] +
+
+Craig Blackmore, Oliver Ray, and Kerstin Eder. + Automatically Tuning the GCC Compiler to Optimize the Performance of Applications Running on Embedded Systems. + Arxiv preprint arXiv:1703.08228, 2017.
+[ bib | +http ] + +
+ + +
+[142] +
+
+María J. Blesa and Christian Blum. + Finding edge-disjoint paths in networks by means of artificial ant colonies. + Journal of Mathematical Modelling and Algorithms, 6(3):361–391, 2007.
+[ bib ] + +
+ + +
+[143] +
+
+Laurens Bliek, Paulo da Costa, Reza Refaei Afshar, Robbert Reijnen, Yingqian Zhang, Tom Catshoek, Daniël Vos, Sicco Verwer, Fynn Schmitt-Ulms, André Hottung, Tapan Shah, Meinolf Sellmann, Kevin Tierney, Carl Perreault-Lafleur, Caroline Leboeuf, Federico Bobbio, Justine Pepin, Warley Almeida Silva, Ricardo Gama, Hugo L. Fernandes, Martin Zaefferer, Manuel López-Ibáñez, and Ekhine Irurozki. + The First AI4TSP Competition: Learning to Solve Stochastic Routing Problems. + Artificial Intelligence, 319:103918, 2023.
+[ bib | +DOI ] +
+This paper reports on the first international competition on + AI for the traveling salesman problem (TSP) at the + International Joint Conference on Artificial Intelligence + 2021 (IJCAI-21). The TSP is one of the classical + combinatorial optimization problems, with many variants + inspired by real-world applications. This first competition + asked the participants to develop algorithms to solve an + orienteering problem with stochastic weights and time windows + (OPSWTW). It focused on two learning approaches: + surrogate-based optimization and deep reinforcement + learning. In this paper, we describe the problem, the + competition setup, and the winning methods, and give an + overview of the results. The winning methods described in + this work have advanced the state-of-the-art in using AI for + stochastic routing problems. Overall, by organizing this + competition we have introduced routing problems as an + interesting problem setting for AI researchers. The simulator + of the problem has been made open-source and can be used by + other researchers as a benchmark for new learning-based + methods. The instances and code for the competition are + available at + https://github.com/paulorocosta/ai-for-tsp-competition. +
+
+Keywords: AI for TSP competition, Travelling salesman problem, Routing + problem, Stochastic combinatorial optimization, + Surrogate-based optimization, Deep reinforcement learning +
+ +
+ + +
+[144] +
+
+Christian Blum. + Beam-ACO—Hybridizing Ant Colony Optimization with Beam Search: An Application to Open Shop Scheduling. + Computers & Operations Research, 32(6):1565–1591, 2005.
+[ bib ] + +
+ + +
+[145] +
+
+Christian Blum. + Beam-ACO for simple assembly line balancing. + INFORMS Journal on Computing, 20(4):618–627, 2008.
+[ bib | +DOI ] + +
+ + +
+[146] +
+
+Christian Blum, María J. Blesa, and Manuel López-Ibáñez. + Beam search for the longest common subsequence problem. + Computers & Operations Research, 36(12):3178–3186, 2009.
+[ bib | +DOI ] +
+The longest common subsequence problem is a classical string + problem that concerns finding the common part of a set of + strings. It has several important applications, for example, + pattern recognition or computational biology. Most research + efforts up to now have focused on solving this problem + optimally. In comparison, only few works exist dealing with + heuristic approaches. In this work we present a deterministic + beam search algorithm. The results show that our algorithm + outperforms the current state-of-the-art approaches not only + in solution quality but often also in computation time. +
+ +
+ + +
+[147] +
+
+Christian Blum, Borja Calvo, and María J. Blesa. + FrogCOL and FrogMIS: New Decentralized Algorithms for Finding Large Independent Sets in Graphs. + Swarm Intelligence, 9(2-3):205–227, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[148] +
+
+Christian Blum and Marco Dorigo. + The hyper-cube framework for ant colony optimization. + IEEE Transactions on Systems, Man, and Cybernetics – Part B, 34(2):1161–1172, 2004.
+[ bib ] + +
+ + +
+[149] +
+
+Christian Blum and Marco Dorigo. + Search Bias in Ant Colony Optimization: On the Role of Competition-Balanced Systems. + IEEE Transactions on Evolutionary Computation, 9(2):159–174, 2005.
+[ bib ] + +
+ + +
+[150] +
+
+Christian Blum and Gabriela Ochoa. + A comparative analysis of two matheuristics by means of merged local optima networks. + European Journal of Operational Research, 290(1):36–56, 2021.
+[ bib ] + +
+ + +
+[151] +
+
+Christian Blum, Pedro Pinacho, Manuel López-Ibáñez, and José A. Lozano. + Construct, Merge, Solve & Adapt: A New General Algorithm for Combinatorial Optimization. + Computers & Operations Research, 68:75–88, 2016.
+[ bib | +DOI ] +
+Keywords: irace, CMSA +
+ +
+ + +
+[152] +
+
+Christian Blum, Jakob Puchinger, Günther R. Raidl, and Andrea Roli. + Hybrid Metaheuristics in Combinatorial Optimization: A Survey. + Applied Soft Computing, 11(6):4135–4151, 2011.
+[ bib ] + +
+ + +
+[153] +
+
+Christian Blum and Andrea Roli. + Metaheuristics in Combinatorial Optimization: Overview and Conceptual Comparison. + ACM Computing Surveys, 35(3):268–308, 2003.
+[ bib ] + +
+ + +
+[154] +
+
+Christian Blum and M. Sampels. + An Ant Colony Optimization Algorithm for Shop Scheduling Problems. + Journal of Mathematical Modelling and Algorithms, 3(3):285–308, 2004.
+[ bib | +DOI ] + +
+ + +
+[155] +
+
+Christian Blum, M. Yábar Vallès, and María J. Blesa. + An ant colony optimization algorithm for DNA sequencing by hybridization. + Computers & Operations Research, 35(11):3620–3635, 2008.
+[ bib ] + +
+ + +
+[156] +
+
+Andrea F. Bocchese, Chris Fawcett, Mauro Vallati, Alfonso E. Gerevini, and Holger H. Hoos. + Performance robustness of AI planners in the 2014 International Planning Competition. + AI Communications, 31(6):445–463, December 2018.
+[ bib | +DOI ] +
+Solver competitions have been used in many areas of AI to + assess the current state of the art and guide future research + and development. AI planning is no exception, and the + International Planning Competition (IPC) has been frequently + run for nearly two decades. Due to the organisational and + computational burden involved in running these competitions, + solvers are generally compared using a single homogeneous + hardware and software environment for all competitors. To + what extent does the specific choice of hardware and software + environment have an effect on solver performance, and is that + effect distributed equally across the competing solvers? In + this work, we use the competing planners and benchmark + instance sets from the 2014 IPC to investigate these two + questions. We recreate the 2014 IPC Optimal and Agile tracks + on two distinct hardware environments and eight distinct + software environments. We show that solver performance varies + significantly based on the hardware and software environment, + and that this variation is not equal for all + planners. Furthermore, the observed variation is sufficient + to change the competition rankings, including the top-ranked + planners for some tracks. +
+ +
+ + +
+[157] +
+
+Kenneth D. Boese, Andrew B. Kahng, and Sudhakar Muddu. + A New Adaptive Multi-Start Technique for Combinatorial Global Optimization. + Operations Research Letters, 16(2):101–113, 1994.
+[ bib ] +
+Keywords: big-valley hypothesis, TSP, landscape analysis +
+ +
+ + +
+[158] +
+
+Marko Bohanec. + Decision making: a computer-science and information-technology viewpoint. + Interdisciplinary Description of Complex Systems, 7(2):22–37, 2009.
+[ bib ] + +
+ + +
+[159] +
+
+Ihor O. Bohachevsky, Mark E. Johnson, and Myron L. Stein. + Generalized Simulated Annealing for Function Optimization. + Technometrics, 28(3):209–217, 1986.
+[ bib ] + +
+ + +
+[160] +
+
+P. C. Borges. + CHESS - Changing Horizon Efficient Set Search: A simple principle for multiobjective optimization. + Journal of Heuristics, 6(3):405–418, 2000.
+[ bib ] + +
+ + +
+[161] +
+
+Endre Boros, Peter L. Hammer, and Gabriel Tavares. + Local search heuristics for Quadratic Unconstrained Binary Optimization (QUBO). + Journal of Heuristics, 13(2):99–132, 2007.
+[ bib ] + +
+ + +
+[162] +
+
+Jean-Charles de Borda. + Mémoire sur les Élections au Scrutin. + Histoire de l'Académie Royal des Sciences, 1781.
+[ bib ] +
+Keywords: ranking +
+ +
+ + +
+[163] +
+
+Hozefa M. Botee and Eric Bonabeau. + Evolving Ant Colony Optimization. + Advances in Complex Systems, 1:149–159, 1998.
+[ bib ] + +
+ + +
+[164] +
+
+Marco Botte and Anita Schöbel. + Dominance for multi-objective robust optimization concepts. + European Journal of Operational Research, 273(2):430–440, 2019.
+[ bib ] + +
+ + +
+[165] +
+
+Salim Bouamama, Christian Blum, and Abdellah Boukerram. + A Population-based Iterated Greedy Algorithm for the Minimum Weight Vertex Cover Problem. + Applied Soft Computing, 12(6):1632–1639, 2012.
+[ bib ] + +
+ + +
+[166] +
+
+Géraldine Bous, Philippe Fortemps, François Glineur, and Marc Pirlot. + ACUTA: A novel method for eliciting additive value functions on the basis of holistic preference statements. + European Journal of Operational Research, 206(2):435–444, 2010.
+[ bib ] + +
+ + +
+[167] +
+
+K. Bouleimen and H. Lecocq. + A new efficient simulated annealing algorithm for the resource-constrained project scheduling problem and its multiple mode version. + European Journal of Operational Research, 149(2):268–281, 2003.
+[ bib | +DOI ] +
+This paper describes new simulated annealing (SA) + algorithms for the resource-constrained project + scheduling problem (RCPSP) and its multiple mode + version (MRCPSP). The objective function + considered is minimisation of the makespan. The + conventional SA search scheme is replaced by a new + design that takes into account the specificity of + the solution space of project scheduling + problems. For RCPSP, the search was based on an + alternated activity and time incrementing process, + and all parameters were set after preliminary + statistical experiments done on test instances. For + MRCPSP, we introduced an original approach using + two embedded search loops alternating activity and + mode neighbourhood exploration. The performance + evaluation done on the benchmark instances available + in the literature proved the efficiency of both + adaptations that are currently among the most + competitive algorithms for these problems. +
+
+Keywords: multi-mode resource-constrained project scheduling, + project scheduling, simulated annealing +
+ +
+ + +
+[168] +
+
+B. Bozkurt, J. W. Fowler, E. S. Gel, B. Kim, Murat Köksalan, and Jyrki Wallenius. + Quantitative comparison of approximate solution sets for multicriteria optimization problems with weighted Tchebycheff preference function. + Operations Research, 58(3):650–659, 2010.
+[ bib ] +
+Proposed IPF indicator +
+ +
+ + +
+[169] +
+
+Jürgen Branke, Salvatore Greco, Roman Slowiński, and P Zielniewicz. + Interactive evolutionary multiobjective optimization driven by robust ordinal regression. + Bulletin of the Polish Academy of Sciences: Technical Sciences, 58(3):347–358, 2010.
+[ bib | +DOI ] + +
+ + +
+[170] +
+
+S. C. Brailsford, Walter J. Gutjahr, M. S. Rauner, and W. Zeppelzauer. + Combined Discrete-event Simulation and Ant Colony Optimisation Approach for Selecting Optimal Screening Policies for Diabetic Retinopathy. + Computational Management Science, 4(1):59–83, 2006.
+[ bib ] + +
+ + +
+[171] +
+
+Jürgen Branke, T. Kaussler, and H. Schmeck. + Guidance in evolutionary multi-objective optimization. + Advances in Engineering Software, 32:499–507, 2001.
+[ bib ] + +
+ + +
+[172] +
+
+Jürgen Branke, S. Nguyen, C. W. Pickardt, and M. Zhang. + Automated Design of Production Scheduling Heuristics: A Review. + IEEE Transactions on Evolutionary Computation, 20(1):110–124, 2016.
+[ bib ] + +
+ + +
+[173] +
+
+Jürgen Branke and C. Schmidt. + Faster Convergence by Means of Fitness Estimation. + Soft Computing, 9(1):13–20, January 2005.
+[ bib | +DOI ] + +
+ + +
+[174] +
+
+Roland Braune and G. Zäpfel. + Shifting Bottleneck Scheduling for Total Weighted Tardiness Minimization—A Computational Evaluation of Subproblem and Re-optimization Heuristics. + Computers & Operations Research, 66:130–140, 2016.
+[ bib ] + +
+ + +
+[175] +
+
+Jürgen Branke, Salvatore Corrente, Salvatore Greco, Roman Slowiński, and P. Zielniewicz. + Using Choquet integral as preference model in interactive evolutionary multiobjective optimization. + European Journal of Operational Research, 250(3):884–901, 2016.
+[ bib | +DOI ] + +
+ + +
+[176] +
+
+Jürgen Branke, S. S. Farid, and N. Shah. + Industry 4.0: a vision for personalized medicine supply chains? + Cell and Gene Therapy Insights, 2(2):263–270, 2016.
+[ bib | +DOI ] + +
+ + +
+[177] +
+
+Jürgen Branke, Salvatore Greco, Roman Slowiński, and Piotr Zielniewicz. + Learning Value Functions in Interactive Evolutionary Multiobjective Optimization. + IEEE Transactions on Evolutionary Computation, 19(1):88–102, 2015.
+[ bib ] + +
+ + +
+[178] +
+
+Yaochu Jin and Jürgen Branke. + Evolutionary Optimization in Uncertain Environments—A Survey. + IEEE Transactions on Evolutionary Computation, 9(5):303–317, 2005.
+[ bib ] + +
+ + +
+[179] +
+
+Leo Breiman. + Random Forests. + Machine Learning, 45(1):5–32, 2001.
+[ bib | +DOI ] + +
+ + +
+[180] +
+
+Karl Bringmann, Sergio Cabello, and Michael T. M. Emmerich. + Maximum volume subset selection for anchored boxes. + Arxiv preprint arXiv:1803.00849, 2018.
+[ bib | +DOI ] +
+Let B be a set of n axis-parallel boxes in Rd + such that each box has a corner at the origin and the other + corner in the positive quadrant of Rd, and let + k be a positive integer. We study the problem of selecting + k boxes in B that maximize the volume of the union of the + selected boxes. This research is motivated by applications + in skyline queries for databases and in multicriteria + optimization, where the problem is known as the + hypervolume subset selection problem. It is known + that the problem can be solved in polynomial time in the + plane, while the best known running time in any dimension d + ≥3 is Ω(nk). We show that: The + problem is NP-hard already in 3 dimensions. In 3 dimensions, + we break the bound Ω(nk), by + providing an nO(√(k)) algorithm. For any constant + dimension d, we present an efficient polynomial-time + approximation scheme. +
+
+Keywords: hypervolume subset selection +
+ +
+ + +
+[181] +
+
+Karl Bringmann and Tobias Friedrich. + Approximating the Least Hypervolume Contributor: NP-Hard in General, But Fast in Practice. + Theoretical Computer Science, 425:104–116, 2012.
+[ bib | +DOI ] + +
+ + +
+[182] +
+
+Karl Bringmann and Tobias Friedrich. + An efficient algorithm for computing hypervolume contributions. + Evolutionary Computation, 18(3):383–402, 2010.
+[ bib ] + +
+ + +
+[183] +
+
+Karl Bringmann and Tobias Friedrich. + Convergence of hypervolume-based archiving algorithms. + IEEE Transactions on Evolutionary Computation, 18(5):643–657, 2014.
+[ bib | +DOI ] +
+Proof that all nondecreasing (μ+ λ) archiving algorithms with + λ< μ are ineffective. +
+
+Keywords: competitive ratio +
+ +
+ + +
+[184] +
+
+Charles G. Broyden. + The Convergence of a Class of Double-rank Minimization Algorithms: 2. The New Algorithm. + IMA Journal of Applied Mathematics, 6(3):222–231, September 1970.
+[ bib | +DOI ] +
+One of the four papers that proposed BFGS. +
+
+Keywords: BFGS +
+ +
+ + +
+[185] +
+
+Dimo Brockhoff, Johannes Bader, Lothar Thiele, and Eckart Zitzler. + Directed Multiobjective Optimization Based on the Weighted Hypervolume Indicator. + Journal of Multi-Criteria Decision Analysis, 20(5-6):291–317, 2013.
+[ bib | +DOI ] +
+Recently, there has been a large interest in set-based + evolutionary algorithms for multi objective + optimization. They are based on the definition of indicators + that characterize the quality of the current population while + being compliant with the concept of Pareto-optimality. It has + been shown that the hypervolume indicator, which measures the + dominated volume in the objective space, enables the design + of efficient search algorithms and, at the same time, opens + up opportunities to express user preferences in the search by + means of weight functions. The present paper contains the + necessary theoretical foundations and corresponding + algorithms to (i) select appropriate weight functions, to + (ii) transform user preferences into weight functions and to + (iii) efficiently evaluate the weighted hypervolume indicator + through Monte Carlo sampling. The algorithm W-HypE, which + implements the previous concepts, is introduced, and the + effectiveness of the search, directed towards the user's + preferred solutions, is shown using an extensive set of + experiments including the necessary statistical performance + assessment. +
+
+Keywords: hypervolume, preference-based search, multi objective + optimization, evolutionary algorithm +
+ +
+ + +
+[186] +
+
+Eric Brochu, Vlad Cora, and Nando de Freitas. + A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning. + Arxiv preprint arXiv:1012.2599, December 2010.
+[ bib | +http ] + +
+ + +
+[187] +
+
+Dimo Brockhoff, Tea Tušar, Dejan Tušar, Tobias Wagner, Nikolaus Hansen, and Anne Auger. + Biobjective performance assessment with the COCO platform. + Arxiv preprint arXiv:1605.01746, 2016.
+[ bib | +DOI ] + +
+ + +
+[188] +
+
+Dimo Brockhoff, Tobias Wagner, and Heike Trautmann. + R2 indicator-based multiobjective search. + Evolutionary Computation, 23(3):369–395, 2015.
+[ bib ] + +
+ + +
+[189] +
+
+Dimo Brockhoff and Eckart Zitzler. + Objective Reduction in Evolutionary Multiobjective Optimization: Theory and Applications. + Evolutionary Computation, 17(2):135–166, 2009.
+[ bib | +DOI ] +
+Many-objective problems represent a major challenge in the + field of evolutionary multiobjective optimization, in terms of + search efficiency, computational cost, decision making, + visualization, and so on. This leads to various research + questions, in particular whether certain objectives can be + omitted in order to overcome or at least diminish the + difficulties that arise when many, that is, more than three, + objective functions are involved. This study addresses this + question from different perspectives. First, we investigate + how adding or omitting objectives affects the problem + characteristics and propose a general notion of conflict + between objective sets as a theoretical foundation for + objective reduction. Second, we present both exact and + heuristic algorithms to systematically reduce the number of + objectives, while preserving as much as possible of the + dominance structure of the underlying optimization + problem. Third, we demonstrate the usefulness of the proposed + objective reduction method in the context of both decision + making and search for a radar waveform application as well as + for well-known test functions. +
+ +
+ + +
+[190] +
+
+C. G. Broyden. + The Convergence of a Class of Double-rank Minimization Algorithms 1. General Considerations. + IMA Journal of Applied Mathematics, 6(1):76–90, March 1970.
+[ bib | +DOI ] +
+This paper presents a more detailed analysis of a class of + minimization algorithms, which includes as a special case the + DFP (Davidon-Fletcher-Powell) method, than has previously + appeared. Only quadratic functions are considered but + particular attention is paid to the magnitude of successive + errors and their dependence upon the initial matrix. On the + basis of this a possible explanation of some of the observed + characteristics of the class is tentatively suggested. +
+
+Keywords: Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm +
+ +
+ + +
+[191] +
+
+Peter Brucker, Johann Hurink, and Frank Werner. + Improving Local Search Heuristics for some Scheduling Problems — Part I. + Discrete Applied Mathematics, 65(1–3):97–122, 1996.
+[ bib ] + +
+ + +
+[192] +
+
+Peter Brucker, Johann Hurink, and Frank Werner. + Improving Local Search Heuristics for some Scheduling Problems — Part II. + Discrete Applied Mathematics, 72(1–2):47–69, 1997.
+[ bib ] + +
+ + +
+[193] +
+
+M. J. Brusco, L. W. Jacobs, and G. M. Thompson. + A Morphing Procedure to Supplement a Simulated Annealing Heuristic for Cost- and Coverage-correlated Set Covering Problems. + Annals of Operations Research, 86:611–627, 1999.
+[ bib ] + +
+ + +
+[194] +
+
+John T. Buchanan. + An experimental evaluation of interactive MCDM methods and the decision making process. + Journal of the Operational Research Society, 45(9):1050–1059, 1994.
+[ bib ] + +
+ + +
+[195] +
+
+John T. Buchanan and James Corner. + The effects of anchoring in interactive MCDM solution methods. + Computers & Operations Research, 24(10):907–918, October 1997.
+[ bib | +DOI ] + +
+ + +
+[196] +
+
+A. L. Buchsbaum and M. T. Goodrich. + Three-Dimensional Layers of Maxima. + Algorithmica, 39:275–289, 2004.
+[ bib ] + +
+ + +
+[197] +
+
+B. Bullnheimer, Richard F. Hartl, and Christine Strauss. + An Improved Ant System Algorithm for the Vehicle Routing Problem. + Annals of Operations Research, 89:319–328, 1999.
+[ bib ] + +
+ + +
+[198] +
+
+B. Bullnheimer, Richard F. Hartl, and Christine Strauss. + A new rank-based version of the Ant System: A computational study. + Central European Journal for Operations Research and Economics, 7(1):25–38, 1999.
+[ bib ] + +
+ + +
+[199] +
+
+Edmund K. Burke and Yuri Bykov. + The Late Acceptance Hill-Climbing Heuristic. + European Journal of Operational Research, 258(1):70–78, 2017.
+[ bib ] + +
+ + +
+[200] +
+
+Rainer E. Burkard and Ulrich Fincke. + The asymptotic probabilistic behaviour of quadratic sum assignment problems. + Zeitschrift für Operations Research, 27(1):73–81, 1983.
+[ bib ] + +
+ + +
+[201] +
+
+Luciana Buriol, Paulo M. França, and Pablo Moscato. + A New Memetic Algorithm for the Asymmetric Traveling Salesman Problem. + Journal of Heuristics, 10(5):483–506, 2004.
+[ bib ] + +
+ + +
+[202] +
+
+Edmund K. Burke, Michel Gendreau, Matthew R. Hyde, Graham Kendall, Gabriela Ochoa, Ender Özcan, and Rong Qu. + Hyper-heuristics: A Survey of the State of the Art. + Journal of the Operational Research Society, 64(12):1695–1724, 2013.
+[ bib | +DOI ] + +
+ + +
+[203] +
+
+Edmund K. Burke, Matthew R. Hyde, Graham Kendall, and John R. Woodward. + A Genetic Programming Hyper-Heuristic Approach for Evolving 2-D Strip Packing Heuristics. + IEEE Transactions on Evolutionary Computation, 14(6):942–958, 2010.
+[ bib | +DOI ] + +
+ + +
+[204] +
+
+Edmund K. Burke, Matthew R. Hyde, Graham Kendall, and John R. Woodward. + Automating the Packing Heuristic Design Process with Genetic Programming. + Evolutionary Computation, 20(1):63–89, 2012.
+[ bib | +DOI ] +
+Keywords: one-, two-, or three-dimensional knapsack and bin packing + problems +
+ +
+ + +
+[205] +
+
+Edmund K. Burke, Matthew R. Hyde, and Graham Kendall. + Grammatical Evolution of Local Search Heuristics. + IEEE Transactions on Evolutionary Computation, 16(7):406–417, 2012.
+[ bib | +DOI ] + +
+ + +
+[206] +
+
+Rainer E. Burkard, Stefan E. Karisch, and Franz Rendl. + QAPLIB–a Quadratic Assignment Problem Library. + Journal of Global Optimization, 10(4):391–403, 1997.
+[ bib ] + +
+ + +
+[207] +
+
+Rainer E. Burkard and Franz Rendl. + A Thermodynamically Motivated Simulation Procedure for Combinatorial Optimization Problems. + European Journal of Operational Research, 17(2):169–174, 1984.
+[ bib | +DOI ] +
+Keywords: 2-exchange delta evaluation for QAP +
+ +
+ + +
+[208] +
+
+Erika Buson, Roberto Roberti, and Paolo Toth. + A Reduced-Cost Iterated Local Search Heuristic for the Fixed-Charge Transportation Problem. + Operations Research, 62(5):1095–1106, 2014.
+[ bib ] + +
+ + +
+[209] +
+
+R. Caballero, Mariano Luque, Julián Molina, and Francisco Ruiz. + PROMOIN: An Interactive System for Multiobjective Programming. + Information Technologies and Decision Making, 1:635–656, 2002.
+[ bib ] +
+Keywords: preferences, multi interactive methods framework +
+ +
+ + +
+[210] +
+
+Leslie Pérez Cáceres and Thomas Stützle. + Exploring Variable Neighborhood Search for Automatic Algorithm Configuration. + Electronic Notes in Discrete Mathematics, 58:167–174, 2017.
+[ bib | +DOI ] + +
+ + +
+[211] +
+
+Sebastien Cahon, Nordine Melab, and El-Ghazali Talbi. + ParadisEO: A Framework for the Reusable Design of Parallel and Distributed Metaheuristics. + Journal of Heuristics, 10(3):357–380, 2004.
+[ bib | +DOI ] + +
+ + +
+[212] +
+
+Zhaoquan Cai, Han Huang, Yong Qin, and Xianheng Ma. + Ant Colony Optimization Based on Adaptive Volatility Rate of Pheromone Trail. + International Journal of Communications, Network and System Sciences, 2(8):792–796, 2009.
+[ bib ] + +
+ + +
+[213] +
+
+Xinye Cai, Yexing Li, Zhun Fan, and Qingfu Zhang. + An external archive guided multiobjective evolutionary algorithm based on decomposition for combinatorial optimization. + IEEE Transactions on Evolutionary Computation, 19(4):508–523, 2015.
+[ bib ] + +
+ + +
+[214] +
+
+Xinye Cai, Yushun Xiao, Miqing Li, Han Hu, Hisao Ishibuchi, and Xiaoping Li. + A grid-based inverted generational distance for multi/many-objective optimization. + IEEE Transactions on Evolutionary Computation, 25(1):21–34, 2021.
+[ bib ] +
+weakly Pareto-compliant indicator +
+ +
+ + +
+[215] +
+
+Xinye Cai, Yushun Xiao, Zhenhua Li, Qi Sun, Hanchuan Xu, Miqing Li, and Hisao Ishibuchi. + A kernel-based indicator for multi/many-objective optimization. + IEEE Transactions on Evolutionary Computation, 2021.
+[ bib ] + +
+ + +
+[216] +
+
+Roberto Wolfler Calvo. + A New Heuristic for the Traveling Salesman Problem with Time Windows. + Transportation Science, 34(1):113–124, 2000.
+[ bib | +DOI ] + +
+ + +
+[217] +
+
+Felipe Campelo, Lucas S. Batista, and Claus Aranha. + The MOEADr Package: A Component-Based Framework for Multiobjective Evolutionary Algorithms Based on Decomposition. + Journal of Statistical Software, 92, 2020.
+[ bib | +DOI ] + +
+ + +
+[218] +
+
+Christian Leonardo Camacho-Villalón, Marco Dorigo, and Thomas Stützle. + The intelligent water drops algorithm: why it cannot be considered a novel algorithm. + Swarm Intelligence, 13:173–192, 2019.
+[ bib ] + +
+ + +
+[219] +
+
+Christian Leonardo Camacho-Villalón, Marco Dorigo, and Thomas Stützle. + An analysis of why cuckoo search does not bring any novel ideas to optimization. + Computers & Operations Research, p.  105747, 2022.
+[ bib ] + +
+ + +
+[220] +
+
+Christian Leonardo Camacho-Villalón, Marco Dorigo, and Thomas Stützle. + Exposing the grey wolf, moth-flame, whale, firefly, bat, and antlion algorithms: six misleading optimization techniques inspired by bestial metaphors. + International Transactions in Operational Research, 2022.
+[ bib | +DOI ] + +
+ + +
+[221] +
+
+Ann Melissa Campbell and Philip C. Jones. + Prepositioning supplies in preparation for disasters. + European Journal of Operational Research, 209(2):156–165, 2011.
+[ bib ] + +
+ + +
+[222] +
+
+E Cambria, B Schuller, Y Xia, and C Havasi. + New avenues in opinion mining and sentiment analysis. + IEEE Intelligent Systems, 28(2):15–21, 2013.
+[ bib | +DOI ] + +
+ + +
+[223] +
+
+Christian Leonardo Camacho-Villalón, Thomas Stützle, and Marco Dorigo. + PSO-X: A Component-Based Framework for the Automatic Design of Particle Swarm Optimization Algorithms. + IEEE Transactions on Evolutionary Computation, 26(3):402–416, 2021.
+[ bib | +DOI ] + +
+ + +
+[224] +
+
+Felipe Campelo and Elizabeth F. Wanner. + Sample size calculations for the experimental comparison of multiple algorithms on multiple problem instances. + Journal of Heuristics, 26(6):851–883, 2020.
+[ bib | +DOI ] + +
+ + +
+[225] +
+
+Z. Cao, S. Jiang, J. Zhang, and H. Guo. + A unified framework for vehicle rerouting and traffic light control to reduce traffic congestion. + IEEE Transactions on Intelligent Transportation Systems, 18(7):1958–1973, 2017.
+[ bib ] + +
+ + +
+[226] +
+
+Gilles Caporossi. + Variable Neighborhood Search for Extremal Vertices : The AutoGraphiX-III System. + Computers & Operations Research, 78:431–438, 2017.
+[ bib ] + +
+ + +
+[227] +
+
+J. Carlier. + The One-machine Sequencing Problem. + European Journal of Operational Research, 11(1):42–47, 1982.
+[ bib ] + +
+ + +
+[228] +
+
+William B. Carlton and J. Wesley Barnes. + Solving the traveling-salesman problem with time windows using tabu search. + IIE Transactions, 28:617–629, 1996.
+[ bib ] + +
+ + +
+[229] +
+
+Fabio Caraffini, Anna V. Kononova, and David Corne. + Infeasibility and structural bias in differential evolution. + Information Sciences, 496:161–179, 2019.
+[ bib | +DOI ] + +
+ + +
+[230] +
+
+Yves Caseau and François Laburthe. + Heuristics for large constrained vehicle routing problems. + Journal of Heuristics, 5(3):281–303, 1999.
+[ bib ] + +
+ + +
+[231] +
+
+Yves Caseau, Glenn Silverstein, and François Laburthe. + Learning Hybrid Algorithms for Vehicle Routing Problems. + Theory and Practice of Logic Programming, 1(6):779–806, 2001.
+[ bib | +epub ] + +
+ + +
+[232] +
+
+Diego Cattaruzza, Nabil Absi, Dominique Feillet, and Daniele Vigo. + An Iterated Local Search for the Multi-commodity Multi-trip Vehicle Routing Problem with Time Windows. + Computers & Operations Research, 51:257–267, 2014.
+[ bib ] + +
+ + +
+[233] +
+
+Aakil M. Caunhye, Xiaofeng Nie, and Shaligram Pokharel. + Optimization models in emergency logistics: A literature review. + Socio-Economic Planning Sciences, 46(1):4–13, 2012.
+[ bib ] + +
+ + +
+[234] +
+
+Josu Ceberio, Ekhine Irurozki, Alexander Mendiburu, and José A. Lozano. + A distance-based ranking model estimation of distribution algorithm for the flowshop scheduling problem. + IEEE Transactions on Evolutionary Computation, 18(2):286–300, 2014.
+[ bib | +DOI ] +
+The aim of this paper is two-fold. First, we introduce a + novel general estimation of distribution algorithm to deal + with permutation-based optimization problems. The algorithm + is based on the use of a probabilistic model for permutations + called the generalized Mallows model. In order to prove the + potential of the proposed algorithm, our second aim is to + solve the permutation flowshop scheduling problem. A hybrid + approach consisting of the new estimation of distribution + algorithm and a variable neighborhood search is + proposed. Conducted experiments demonstrate that the proposed + algorithm is able to outperform the state-of-the-art + approaches. Moreover, from the 220 benchmark instances + tested, the proposed hybrid approach obtains new best known + results in 152 cases. An in-depth study of the results + suggests that the successful performance of the introduced + approach is due to the ability of the generalized Mallows + estimation of distribution algorithm to discover promising + regions in the search space. +
+
+Keywords: Estimation of distribution algorithms,Generalized Mallows + model,Permutation flowshop scheduling + problem,Permutations-based optimization problems +
+ +
+ + +
+[235] +
+
+Vladimír Černý. + A Thermodynamical Approach to the Traveling Salesman Problem: An Efficient Simulation Algorithm. + Journal of Optimization Theory and Applications, 45(1):41–51, 1985.
+[ bib ] + +
+ + +
+[236] +
+
+Sara Ceschia, Luca Di Gaspero, and Andrea Schaerf. + Design, Engineering, and Experimental Analysis of a Simulated Annealing Approach to the Post-Enrolment Course Timetabling Problem. + Computers & Operations Research, 39(7):1615–1624, 2012.
+[ bib ] + +
+ + +
+[237] +
+
+Sara Ceschia and Andrea Schaerf. + Modeling and solving the dynamic patient admission scheduling problem under uncertainty. + Artificial Intelligence in Medicine, 56(3):199–205, 2012.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[238] +
+
+Sara Ceschia, Andrea Schaerf, and Thomas Stützle. + Local Search Techniques for a Routing-packing Problem. + Computers and Industrial Engineering, 66(4):1138–1149, 2013.
+[ bib ] + +
+ + +
+[239] +
+
+T.-J. Chang, N. Meade, John E. Beasley, and Y. M. Sharaiha. + Heuristics for cardinality constrained portfolio optimisation. + Computers & Operations Research, 27(13):1271–1302, 2000.
+[ bib ] +
+In this paper we consider the problem of finding the + efficient frontier associated with the standard mean-variance + portfolio optimisation model. We extend the standard model to + include cardinality constraints that limit a portfolio to + have a specified number of assets, and to impose limits on + the proportion of the portfolio held in a given asset (if any + of the asset is held). We illustrate the differences that + arise in the shape of this efficient frontier when such + constraints are present. We present three heuristic + algorithms based upon genetic algorithms, tabu search and + simulated annealing for finding the cardinality constrained + efficient frontier. Computational results are presented for + five data sets involving up to 225 assets. Scope and purpose + The standard Markowitz mean-variance approach to portfolio + selection involves tracing out an efficient frontier, a + continuous curve illustrating the tradeoff between return and + risk (variance). This frontier can be easily found via + quadratic programming. This approach is well-known and widely + applied. However, for practical purposes, it may be desirable + to limit the number of assets in a portfolio, as well as + imposing limits on the proportion of the portfolio devoted to + any particular asset. If such constraints exist, the problem + of finding the efficient frontier becomes much harder. This + paper illustrates how, in the presence of such constraints, + the efficient frontier becomes discontinuous. Three heuristic + techniques are applied to the problem of finding this + efficient frontier and computational results presented for a + number of data sets which are made publicly available. +
+
+Keywords: Portfolio optimisation, CCMVPOP, Efficient frontier +
+ +
+ + +
+[240] +
+
+Shelvin Chand and Markus Wagner. + Evolutionary many-objective optimization: A quick-start guide. + Surveys in Operations Research and Management Science, 20(2):35–42, 2015.
+[ bib | +DOI ] + +
+ + +
+[241] +
+
+Donald V. Chase and Lindell E. Ormsbee. + Computer-generated pumping schedules for satisfying operation objectives. + J. Am. Water Works Assoc., 85(7):54–61, 1993.
+[ bib ] + +
+ + +
+[242] +
+
+Shamik Chaudhuri and Kalyanmoy Deb. + An interactive evolutionary multi-objective optimization and decision making procedure. + Applied Soft Computing, 10(2):496–511, 2010.
+[ bib ] + +
+ + +
+[243] +
+
+Hsinchun Chen, Roger H. L. Chiang, and Veda C. Storey. + Business Intelligence and Analytics: From Big Data to Big Impact. + MIS Quarterly, 36(4):1165–1188, 2012.
+[ bib ] + +
+ + +
+[244] +
+
+Yuning Chen, Jin-Kao Hao, and Fred Glover. + A hybrid metaheuristic approach for the capacitated arc routing problem. + European Journal of Operational Research, 553(1):25–39, 2016.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[245] +
+
+Ruey-Maw Chen and Fu-Ren Hsieh. + An exchange local search heuristic based scheme for permutation flow shop problems. + Applied Mathematics & Information Sciences, 8(1):209–215, 2014.
+[ bib ] + +
+ + +
+[246] +
+
+F. Y. Cheng and X. S. Li. + Generalized center method for multiobjective engineering optimization. + Engineering Optimization, 31(5):641–661, 1999.
+[ bib | +DOI ] + +
+ + +
+[247] +
+
+Renzhi Chen, Ke Li, and Xin Yao. + Dynamic Multiobjectives Optimization With a Changing Number of Objectives. + IEEE Transactions on Evolutionary Computation, 22(1):157–171, 2017.
+[ bib | +DOI ] +
+two co-evolving populations (two archive) +
+ +
+ + +
+[248] +
+
+Rachid Chelouah and Patrick Siarry. + Tabu search applied to global optimization. + European Journal of Operational Research, 123(2):256–270, 2000.
+[ bib ] + +
+ + +
+[249] +
+
+Ni Chen, Wei-Neng Chen, Yue-Jiao Gong, Zhi-Hui Zhan, Jun Zhang, Yun Li, and Yu-Song Tan. + An evolutionary algorithm with double-level archives for multiobjective optimization. + IEEE Transactions on Cybernetics, 45(9):1851–1863, 2014.
+[ bib ] + +
+ + +
+[250] +
+
+Chin-Bin Cheng and Chun-Pin Mao. + A modified ant colony system for solving the travelling salesman problem with time windows. + Mathematical and Computer Modelling, 46:1225–1235, 2007.
+[ bib | +DOI ] + +
+ + +
+[251] +
+
+Marco Chiarandini, Mauro Birattari, Krzysztof Socha, and O. Rossi-Doria. + An Effective Hybrid Algorithm for University Course Timetabling. + Journal of Scheduling, 9(5):403–432, October 2006.
+[ bib | +DOI ] +
+Keywords: 2003 international timetabling competition, F-race +
+ +
+ + +
+[252] +
+
+Manuel Chica, Oscar Cordón, Sergio Damas, and Joaquín Bautista. + A New Diversity Induction Mechanism for a Multi-objective Ant Colony Algorithm to Solve a Real-world time and Space Assembly Line Balancing Problem. + Memetic Computing, 3(1):15–24, 2011.
+[ bib ] + +
+ + +
+[253] +
+
+D. S. Chivilikhin, V. I. Ulyantsev, and A. A. Shalyto. + Modified ant colony algorithm for constructing finite state machines from execution scenarios and temporal formulas. + Automation and Remote Control, 77(3):473–484, 2016.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[254] +
+
+Francisco Chicano, Darrell Whitley, and Enrique Alba. + A Methodology to Find the Elementary Landscape Decomposition of Combinatorial Optimization Problems. + Evolutionary Computation, 19(4):597–637, 2011.
+[ bib ] + +
+ + +
+[255] +
+
+Francisco Chicano, Gabriel J. Luque, and Enrique Alba. + Autocorrelation Measures for the Quadratic Assignment Problem. + Applied Mathematics Letters, 25:698–705, 2012.
+[ bib | +DOI ] + +
+ + +
+[256] +
+
+Nicos Christofides, A. Mingozzi, and Paolo Toth. + State-space relaxation procedures for the computation of bounds to routing problems. + Networks, 11(2):145–164, 1981.
+[ bib | +DOI ] + +
+ + +
+[257] +
+
+Tinkle Chugh, Yaochu Jin, Kaisa Miettinen, Jussi Hakanen, and Karthik Sindhya. + A Surrogate-Assisted Reference Vector Guided Evolutionary Algorithm for Computationally Expensive Many-Objective Optimization. + IEEE Transactions on Evolutionary Computation, 22(1):129–142, February 2018.
+[ bib ] + +
+ + +
+[258] +
+
+Tinkle Chugh, Karthik Sindhya, Jussi Hakanen, and Kaisa Miettinen. + A survey on handling computationally expensive multiobjective optimization problems with evolutionary algorithms. + Soft Computing, 23(9):3137–3166, 2019.
+[ bib | +DOI ] +
+Evolutionary algorithms are widely used for solving + multiobjective optimization problems but are often criticized + because of a large number of function evaluations + needed. Approximations, especially function approximations, + also referred to as surrogates or metamodels are commonly + used in the literature to reduce the computation time. This + paper presents a survey of 45 different recent algorithms + proposed in the literature between 2008 and 2016 to handle + computationally expensive multiobjective optimization + problems. Several algorithms are discussed based on what kind + of an approximation such as problem, function or fitness + approximation they use. Most emphasis is given to function + approximation-based algorithms. We also compare these + algorithms based on different criteria such as metamodeling + technique and evolutionary algorithm used, type and + dimensions of the problem solved, handling constraints, + training time and the type of evolution control. Furthermore, + we identify and discuss some promising elements and major + issues among algorithms in the literature related to using an + approximation and numerical settings used. In addition, we + discuss selecting an algorithm to solve a given + computationally expensive multiobjective optimization problem + based on the dimensions in both objective and decision spaces + and the computation budget available. +
+ +
+ + +
+[259] +
+
+Christian Cintrano, Javier Ferrer, Manuel López-Ibáñez, and Enrique Alba. + Hybridization of Evolutionary Operators with Elitist Iterated Racing for the Simulation Optimization of Traffic Lights Programs. + Evolutionary Computation, 31(1):31–51, 2023.
+[ bib | +DOI ] +
+In the traffic light scheduling problem the evaluation of + candidate solutions requires the simulation of a process + under various (traffic) scenarios. Thus, good solutions + should not only achieve good objective function values, but + they must be robust (low variance) across all different + scenarios. Previous work has shown that combining IRACE with + evolutionary operators is effective for this task due to the + power of evolutionary operators in numerical optimization. In + this paper, we further explore the hybridization of + evolutionary operators and the elitist iterated racing of + IRACE for the simulation-optimization of traffic light + programs. We review previous works from the literature to + find the evolutionary operators performing the best when + facing this problem to propose new hybrid algorithms. We + evaluate our approach over a realistic case study derived + from the traffic network of Málaga (Spain) with 275 traffic + lights that should be scheduled optimally. The experimental + analysis reveals that the hybrid algorithm comprising IRACE + plus differential evolution offers statistically better + results than the other algorithms when the budget of + simulations is low. In contrast, IRACE performs better than + the hybrids for high simulations budget, although the + optimization time is much longer. +
+
+Keywords: irace, Simulation optimization, Uncertainty, Traffic light + planning +
+ +
+ + +
+[260] +
+
+R. M. Clark, L. A. Rossman, and L. J. Wymer. + Modeling distribution system water quality: regulatory implications. + Journal of Water Resources Planning and Management, ASCE, 121(6):423–428, 1995.
+[ bib ] + +
+ + +
+[261] +
+
+Maurice Clerc and J. Kennedy. + The particle swarm - explosion, stability, and convergence in a multidimensional complex space. + IEEE Transactions on Evolutionary Computation, 6(1):58–73, February 2002.
+[ bib | +DOI ] + +
+ + +
+[262] +
+
+Andy Cockburn, Pierre Dragicevic, Lonni Besançon, and Carl Gutwin. + Threats of a Replication Crisis in Empirical Computer Science. + Communications of the ACM, 63(8):70–79, July 2020.
+[ bib | +DOI ] +
+Research replication only works if there is confidence built + into the results. +
+ +
+ + +
+[263] +
+
+B. Codenotti, G. Manzini, L. Margara, and G. Resta. + Perturbation: An Efficient Technique for the Solution of Very Large Instances of the Euclidean TSP. + INFORMS Journal on Computing, 8(2):125–133, 1996.
+[ bib ] + +
+ + +
+[264] +
+
+Carlos A. Coello Coello. + Theoretical and numerical constraint-handling techniques used with evolutionary algorithms: a survey of the state of the art. + Computer Methods in Applied Mechanics and Engineering, 191(11-12):1245–1287, 2002.
+[ bib | +DOI ] + +
+ + +
+[265] +
+
+Carlos A. Coello Coello. + Special Issue on Evolutionary Multiobjective Optimization. + IEEE Transactions on Evolutionary Computation, 7(2), 2003.
+[ bib ] + +
+ + +
+[266] +
+
+Carlos A. Coello Coello. + Evolutionary multi-objective optimization: a historical view of the field. + IEEE Computational Intelligence Magazine, 1(1):28–36, 2006.
+[ bib ] + +
+ + +
+[267] +
+
+Harry Cohn and Mark J. Fielding. + Simulated Annealing: Searching for an Optimal Temperature. + SIAM Journal on Optimization, 9(3):779–802, 1999.
+[ bib ] + +
+ + +
+[268] +
+
+Andrew F. Colombo and Bryan W. Karney. + Impacts of Leaks on Energy Consumption in Pumped Systems with Storage. + Journal of Water Resources Planning and Management, ASCE, 131(2):146–155, March 2005.
+[ bib ] + +
+ + +
+[269] +
+
+Alberto Colorni, Marco Dorigo, Vittorio Maniezzo, and M. Trubian. + Ant System for Job-shop Scheduling. + JORBEL — Belgian Journal of Operations Research, Statistics and Computer Science, 34(1):39–53, 1994.
+[ bib ] + +
+ + +
+[270] +
+
+Barry McCollum, Andrea Schaerf, Ben Paechter, Paul McMullan, Rhyd M. R. Lewis, Andrew J. Parkes, Luca Di Gaspero, Rong Qu, and Edmund K. Burke. + Setting the Research Agenda in Automated Timetabling: The Second International Timetabling Competition. + INFORMS, 22(1):120–130, February 2010.
+[ bib | +DOI ] + +
+ + +
+[271] +
+
+Richard K. Congram, Chris N. Potts, and Steve van de Velde. + An Iterated Dynasearch Algorithm for the Single-Machine Total Weighted Tardiness Scheduling Problem. + INFORMS Journal on Computing, 14(1):52–67, 2002.
+[ bib ] + +
+ + +
+[272] +
+
+David T. Connolly. + An Improved Annealing Scheme for the QAP. + European Journal of Operational Research, 46(1):93–100, 1990.
+[ bib ] + +
+ + +
+[273] +
+
+Richard J. Cook and Vern T. Farewell. + Multiplicity Considerations in the Design and Analysis of Clinical Trials. + Journal of the Royal Statistical Society: Series A, 159:93–110, 1996.
+[ bib ] +
+multiplicity; multiple endpoints; multiple treatments; + p-value adjustment; type I error; argues that if results are + intended to be interpreted marginally, there may be no need + for controlling experimentwise error rate +
+ +
+ + +
+[274] +
+
+Don Coppersmith, Lisa K. Fleischer, and Atri Rurda. + Ordering by Weighted Number of Wins Gives a Good Ranking for Weighted Tournaments. + ACM Transactions on Algorithms, 6(3):1–13, July 2010.
+[ bib | +DOI ] +
+Keywords: Approximation algorithms,Borda's method,feedback arc set + problem,rank aggregation,tournaments +
+ +
+ + +
+[275] +
+
+Oscar Cordón and Sergio Damas. + Image Registration with Iterated Local Search. + Journal of Heuristics, 12(1–2):73–94, 2006.
+[ bib ] + +
+ + +
+[276] +
+
+Jeroen Corstjens, Nguyen Dang, Benoît Depaire, An Caris, and Patrick De Causmaecker. + A combined approach for analysing heuristic algorithms. + Journal of Heuristics, 25(4):591–628, 2019.
+[ bib | +DOI ] + +
+ + +
+[277] +
+
+Jeroen Corstjens, Benoît Depaire, An Caris, and Kenneth Sörensen. + A multilevel evaluation method for heuristics with an application to the VRPTW. + International Transactions in Operational Research, 27(1):168–196, 2020.
+[ bib | +DOI ] + +
+ + +
+[278] +
+
+P. Corry and E. Kozan. + Ant Colony Optimisation for Machine Layout Problems. + Computational Optimization and Applications, 28(3):287–310, 2004.
+[ bib ] + +
+ + +
+[279] +
+
+Jean-François Cordeau, Gilbert Laporte, and A. Mercier. + A unified tabu search heuristic for vehicle routing problems with time windows. + Journal of the Operational Research Society, 52(8):928–936, 2001.
+[ bib ] + +
+ + +
+[280] +
+
+Jean-François Cordeau and Mirko Maischberger. + A Parallel Iterated Tabu Search Heuristic for Vehicle Routing Problems. + Computers & Operations Research, 39(9):2033–2050, 2012.
+[ bib ] + +
+ + +
+[281] +
+
+Wagner Emanoel Costa, Marco Cesar Goldbarg, and Elizabeth Ferreira Gouvêa Goldbarg. + Hybridizing VNS and path-relinking on a particle swarm framework to minimize total flowtime. + Expert Systems with Applications, 39(18):13118–13126, 2012.
+[ bib ] + +
+ + +
+[282] +
+
+D. Costa and A. Hertz. + Ants can color graphs. + Journal of the Operational Research Society, 48:295–305, 1997.
+[ bib ] + +
+ + +
+[283] +
+
+S. P. Coy, B. L. Golden, G. C. Runger, and E. A. Wasil. + Using Experimental Design to Find Effective Parameter Settings for Heuristics. + Journal of Heuristics, 7(1):77–97, 2001.
+[ bib ] + +
+ + +
+[284] +
+
+I. Barry Crabtree. + Resource Scheduling: Comparing Simulated Annealing with Constraint Programming. + BT Technology Journal, 13(1):121–127, 1995.
+[ bib ] + +
+ + +
+[285] +
+
+Douglas Edward Critchlow, Michael A. Fligner, and Joseph S. Verducci. + Probability Models on Rankings. + Journal of Mathematical Psychology, 35:294–318, 1991.
+[ bib ] + +
+ + +
+[286] +
+
+G. A. Croes. + A Method for Solving Traveling Salesman Problems. + Operations Research, 6:791–812, 1958.
+[ bib ] + +
+ + +
+[287] +
+
+Harlan P. Crowder, Ron S. Dembo, and John M. Mulvey. + Reporting computational experiments in mathematical programming. + Mathematical Programming, 15(1):316–329, 1978.
+[ bib | +DOI ] +
+Keywords: reproducibility +
+ +
+ + +
+[288] +
+
+Carlos Cruz, Juan Ramón González, and David A. Pelta. + Optimization in Dynamic Environments: A Survey on Problems, Methods and Measures. + Soft Computing, 15(7):1427–1448, 2011.
+[ bib ] + +
+ + +
+[289] +
+
+Fábio Cruz, Anand Subramanian, Bruno P. Bruck, and Manuel Iori. + A Heuristic Algorithm for a Single Vehicle Static Bike Sharing Rebalancing Problem. + Computers & Operations Research, 79:19–33, 2017.
+[ bib ] + +
+ + +
+[290] +
+
+Joseph C. Culberson. + On the Futility of Blind Search: An Algorithmic View of “No Free Lunch”. + Evolutionary Computation, 6(2):109–127, 1998.
+[ bib | +DOI ] +
+Keywords: NFL +
+ +
+ + +
+[291] +
+
+P. Czyzżak and Andrzej Jaszkiewicz. + Pareto simulated annealing – a metaheuristic technique for multiple-objective combinatorial optimization. + Journal of Multi-Criteria Decision Analysis, 7(1):34–47, 1998.
+[ bib ] + +
+ + +
+[292] +
+
+Steven B. Damelin, Fred J. Hickernell, David L. Ragozin, and Xiaoyan Zeng. + On Energy, Discrepancy and Group Invariant Measures on Measurable Subsets of Euclidean Space. + Journal of Fourier Analysis and Applications, 16(6):813–839, 2010.
+[ bib ] +
+Keywords: Capacity; Cubature; Discrepancy; Distribution; Group + invariant kernel; Group invariant measure; Energy minimizer; + Equilibrium measure; Numerical integration; Positive + definite; Potential field; Riesz kernel; Reproducing Hilbert + space; Signed measure +
+ +
+ + +
+[293] +
+
+M. Damas, M. Salmerón, J. Ortega, G. Olivares, and H. Pomares. + Parallel Dynamic Water Supply Scheduling in a Cluster of Computers. + Concurrency and Computation: Practice and Experience, 13(15):1281–1302, December 2001.
+[ bib ] + +
+ + +
+[294] +
+
+Augusto Dantas and Aurora Pozo. + On the use of fitness landscape features in meta-learning based algorithm selection for the quadratic assignment problem. + Theoretical Computer Science, 805:62–75, 2020.
+[ bib | +DOI ] + +
+ + +
+[295] +
+
+Emilie Danna, Edward Rothberg, and Claude Le Pape. + Exploring relaxation induced neighborhoods to improve MIP solutions. + Mathematical Programming, 102(1):71–90, 2005.
+[ bib ] + +
+ + +
+[296] +
+
+George B. Dantzig and Philip Wolfe. + Decomposition Principle for Linear Programs. + Operations Research, 8(1):101–111, 1960.
+[ bib ] + +
+ + +
+[297] +
+
+Fabio Daolio, Arnaud Liefooghe, Sébastien Verel, Hernán E. Aguirre, and Kiyoshi Tanaka. + Problem Features versus Algorithm Performance on Rugged Multiobjective Combinatorial Fitness Landscapes. + Evolutionary Computation, 25(4):555–585, 2017.
+[ bib | +DOI ] + +
+ + +
+[298] +
+
+Indraneel Das and John E. Dennis. + A closer look at drawbacks of minimizing weighted sums of objectives for Pareto set generation in multicriteria optimization problems. + Structural Optimization, 14(1):63–69, 1997.
+[ bib | +DOI ] + +
+ + +
+[299] +
+
+Indraneel Das and John E. Dennis. + Normal-boundary intersection: A new method for generating the Pareto surface in nonlinear multicriteria optimization problems. + SIAM Journal on Optimization, 8(3):631–657, 1998.
+[ bib ] +
+Keywords: simplex lattice design +
+ +
+ + +
+[300] +
+
+Swagatam Das, Sankha Subhra Mullick, and Ponnuthurai N. Suganthan. + Recent advances in differential evolution–An updated survey. + Swarm and Evolutionary Computation, 27:1–30, 2016.
+[ bib ] + +
+ + +
+[301] +
+
+Swagatam Das and Ponnuthurai N. Suganthan. + Differential Evolution: A Survey of the State-of-the-art. + IEEE Transactions on Evolutionary Computation, 15(1), February 2011.
+[ bib ] + +
+ + +
+[302] +
+
+Sanjeeb Dash. + Exponential Lower Bounds on the Lengths of Some Classes of Branch-and-Cut Proofs. + Mathematics of Operations Research, 30(3):678–700, 2005.
+[ bib ] + +
+ + +
+[303] +
+
+Constantinos Daskalakis, Ilias Diakonikolas, and Mihalis Yannakakis. + How good is the Chord algorithm? + SIAM Journal on Computing, 45(3):811–858, 2016.
+[ bib ] + +
+ + +
+[304] +
+
+Jean Daunizeau, Hanneke E. M. den Ouden, Matthias Pessiglione, Stefan J. Kiebel, Karl J. Friston, and Klaas E. Stephan. + Observing the observer (II): deciding when to decide. + PLoS One, 5(12):e15555, 2010.
+[ bib | +DOI ] + +
+ + +
+[305] +
+
+Jean Daunizeau, Hanneke E. M. den Ouden, Matthias Pessiglione, Klaas E. Stephan, Stefan J. Kiebel, and Karl J. Friston. + Observing the observer (I): meta-Bayesian models of learning and decision-making. + PLoS One, 5(12):e15554, 2010.
+[ bib | +DOI ] + +
+ + +
+[306] +
+
+Kalyanmoy Deb. + An efficient constraint handling method for genetic algorithms. + Computer Methods in Applied Mechanics and Engineering, 186(2/4):311–338, 2000.
+[ bib | +DOI ] + +
+ + +
+[307] +
+
+Kalyanmoy Deb, A. Pratap, S. Agarwal, and T. Meyarivan. + A fast and elitist multi-objective genetic algorithm: NSGA-II. + IEEE Transactions on Evolutionary Computation, 6(2):182–197, 2002.
+[ bib | +DOI ] + +
+ + +
+[308] +
+
+Kalyanmoy Deb. + Multi-objective genetic algorithms: problem difficulties and construction of test problems. + Evolutionary Computation, 7(3):205–230, 1999.
+[ bib ] +
+Naive definition of PLO-set +
+ +
+ + +
+[309] +
+
+Kalyanmoy Deb and Ram Bhushan Agrawal. + Simulated binary crossover for continuous search spaces. + Complex Systems, 9(2):115–148, 1995.
+[ bib | +epub ] +
+Keywords: SBX +
+ +
+ + +
+[310] +
+
+Kalyanmoy Deb and Debayan Deb. + Analysing mutation schemes for real-parameter genetic algorithms. + International Journal of Artificial Intelligence and Soft Computing, 4(1):1–28, 2014.
+[ bib ] +
+Proposed Gaussian mutation +
+ +
+ + +
+[311] +
+
+Kalyanmoy Deb, S. Gupta, D. Daum, Jürgen Branke, A. Mall, and D. Padmanabhan. + Reliability-based optimization using evolutionary algorithms. + IEEE Transactions on Evolutionary Computation, 13(5):1054–1074, October 2009.
+[ bib | +DOI ] + +
+ + +
+[312] +
+
+Kalyanmoy Deb and Himanshu Jain. + An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting Approach, Part I: Solving Problems With Box Constraints. + IEEE Transactions on Evolutionary Computation, 18(4):577–601, 2014.
+[ bib ] +
+Proposed NSGA-III +
+ +
+ + +
+[313] +
+
+Kalyanmoy Deb and Murat Köksalan. + Guest Editorial: Special Issue on Preference-based Multiobjective Evolutionary Algorithms. + IEEE Transactions on Evolutionary Computation, 14(5):669–670, October 2010.
+[ bib | +DOI ] + +
+ + +
+[314] +
+
+Kalyanmoy Deb, Manikanth Mohan, and Shikhar Mishra. + Evaluating the ε-domination based multi-objective evolutionary algorithm for a quick computation of Pareto-optimal solutions. + Evolutionary Computation, 13(4):501–525, December 2005.
+[ bib | +DOI ] +
+Keywords: ε-dominance, ε-MOEA +
+ +
+ + +
+[315] +
+
+Kalyanmoy Deb and Santosh Tiwari. + Omni-optimizer: A generic evolutionary algorithm for single and multi-objective optimization. + European Journal of Operational Research, 185(3):1062–1087, 2008.
+[ bib | +DOI ] +
+Archiving method with epsilon dominance and density in the + decision and objective spaces +
+
+Keywords: epsilon-dominance, archiving +
+ +
+ + +
+[316] +
+
+Kalyanmoy Deb, Ling Zhu, and Sandeep Kulkarni. + Handling Multiple Scenarios in Evolutionary Multi-Objective Numerical Optimization. + IEEE Transactions on Evolutionary Computation, 22(6):920–933, 2018.
+[ bib | +DOI ] +
+Solutions to most practical numerical optimization problems + must be evaluated for their performance over a number of + different loading or operating conditions, which we refer + here as scenarios. Therefore, a meaningful and resilient + optimal solution must be such that it remains feasible under + all scenarios and performs close to an individual optimal + solution corresponding to each scenario. Despite its + practical importance, multi-scenario consideration has + received a lukewarm attention, particularly in the context of + multi-objective optimization. The usual practice is to + optimize for the worst-case scenario. In this paper, we + review existing methodologies in this direction and set our + goal to suggest a new and potential population-based method + for handling multiple scenarios by defining scenario-wise + domination principle and scenario-wise diversity-preserving + operators. To evaluate, the proposed method is applied to a + number of numerical test problems and engineering design + problems with a detail explanation of the obtained results + and compared with an existing method. This first systematic + evolutionary based multi-scenario, multiobjective, + optimization study on numerical problems indicates that + multiple scenarios can be handled in an integrated manner + using an EMO framework to find a well-balanced compromise set + of solutions to multiple scenarios and maintain a tradeoff + among multiple objectives. In comparison to an existing + serial multiple optimization approach, the proposed approach + finds a set of compromised trade-off solutions + simultaneously. An achievement of multi-objective trade-off + and multi-scenario trade-off is algorithmically challenging, + but due to its practical appeal, further research and + application must be spent. +
+
+Keywords: scenario-based +
+ +
+ + +
+[317] +
+
+Annelies De Corte and Kenneth Sörensen. + Optimisation of gravity-fed water distribution network design: A critical review. + European Journal of Operational Research, 228(1):1–10, 2013.
+[ bib | +DOI ] + +
+ + +
+[318] +
+
+Annelies De Corte and Kenneth Sörensen. + An Iterated Local Search Algorithm for Water Distribution Network Design Optimization. + Networks, 67(3):187–198, 2016.
+[ bib ] + +
+ + +
+[319] +
+
+Annelies De Corte and Kenneth Sörensen. + An Iterated Local Search Algorithm for multi-period water distribution network design optimization. + Water, 8(8):359, 2016.
+[ bib | +DOI ] + +
+ + +
+[320] +
+
+V. Dekhtyarenko. + Verification of weight coefficients in multicriteria optimization problems. + Computer-Aided Design, 13(6):339–344, 1981.
+[ bib ] + +
+ + +
+[321] +
+
+X. Delorme, Xavier Gandibleux, and F. Degoutin. + Evolutionary, constructive and hybrid procedures for the bi-objective set packing problem. + European Journal of Operational Research, 204(2):206–217, 2010.
+[ bib ] +
+This paper cannot be found on internet!! Does it exist? +
+ +
+ + +
+[322] +
+
+Federico Della Croce, Thierry Garaix, and Andrea Grosso. + Iterated Local Search and Very Large Neighborhoods for the Parallel-machines Total Tardiness Problem. + Computers & Operations Research, 39(6):1213–1217, 2012.
+[ bib ] + +
+ + +
+[323] +
+
+Maxence Delorme, Manuel Iori, and Silvano Martello. + Bin packing and cutting stock problems: Mathematical models and exact algorithms. + European Journal of Operational Research, 255(1):1–20, 2016.
+[ bib | +DOI ] + +
+ + +
+[324] +
+
+Mauro Dell'Amico, Manuel Iori, Silvano Martello, and Michele Monaci. + Heuristic and Exact Algorithms for the Identical Parallel Machine Scheduling Problem. + INFORMS Journal on Computing, 20(3):333–344, 2016.
+[ bib ] + +
+ + +
+[325] +
+
+Maxence Delorme, Manuel Iori, and Silvano Martello. + BPPLIB: a library for bin packing and cutting stock problems. + Optimization Letters, 12(2):235–250, 2018.
+[ bib | +DOI ] + +
+ + +
+[326] +
+
+Mauro Dell'Amico, Manuel Iori, Stefano Novellani, and Thomas Stützle. + A destroy and repair algorithm for the Bike sharing Rebalancing Problem. + Computers & Operations Research, 71:146–162, 2016.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[327] +
+
+Robert F. Dell and Mark H. Karwan. + An interactive MCDM weight space reduction method utilizing a Tchebycheff utility function. + Naval Research Logistics, 37(2):263–277, 1990.
+[ bib ] + +
+ + +
+[328] +
+
+Mauro Dell'Amico and Marco Trubian. + Applying Tabu Search to the Job Shop Scheduling Problem. + Annals of Operations Research, 41:231–252, 1993.
+[ bib ] + +
+ + +
+[329] +
+
+Stephan Dempe, Gabriele Eichfelder, and Jörg Fliege. + On the effects of combining objectives in multi-objective optimization. + Mathematical Methods of Operations Research, 82(1):1–18, 2015.
+[ bib ] + +
+ + +
+[330] +
+
+Jean-Louis Deneubourg, S. Aron, S. Goss, and J.-M. Pasteels. + The Self-Organizing Exploratory Pattern of the Argentine Ant. + Journal of Insect Behavior, 3(2):159–168, 1990.
+[ bib | +DOI ] + +
+ + +
+[331] +
+
+Joaquín Derrac, Salvador García, Daniel Molina, and Francisco Herrera. + A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms. + Swarm and Evolutionary Computation, 1(1):3–18, 2011.
+[ bib ] + +
+ + +
+[332] +
+
+Ulrich Derigs and Ulrich Vogel. + Experience with a Framework for Developing Heuristics for Solving Rich Vehicle Routing Problems. + Journal of Heuristics, 20(1):75–106, 2014.
+[ bib ] + +
+ + +
+[333] +
+
+Aryan Deshwal, Syrine Belakaria, Janardhan Rao Doppa, and Dae Hyun Kim. + Bayesian Optimization over Permutation Spaces. + Arxiv preprint arXiv:2112.01049, 2021.
+[ bib | +DOI ] +
+Keywords: BOPS, CEGO +
+ +
+ + +
+[334] +
+
+Marcelo De Souza, Marcus Ritt, Manuel López-Ibáñez, and Leslie Pérez Cáceres. + ACVIZ: A Tool for the Visual Analysis of the Configuration of Algorithms with irace. + Operations Research Perspectives, 8:100186, 2021.
+[ bib | +DOI | +supplementary material ] +
+This paper introduces acviz, a tool that helps to analyze the + automatic configuration of algorithms with irace. It provides + a visual representation of the configuration process, + allowing users to extract useful information, e.g. how the + configurations evolve over time. When test data is available, + acviz also shows the performance of each configuration on the + test instances. Using this visualization, users can analyze + and compare the quality of the resulting configurations and + observe the performance differences on training and test + instances. +
+ +
+ + +
+[335] +
+
+Paolo Detti, Francesco Papalini, and Garazi Zabalo Manrique de Lara. + A multi-depot dial-a-ride problem with heterogeneous vehicles and compatibility constraints in healthcare. + Omega, 70:1–14, 2017.
+[ bib | +DOI ] + +
+ + +
+[336] +
+
+Sven De Vries and Rakesh V. Vohra. + Combinatorial Auctions: A Survey. + INFORMS Journal on Computing, 15(3):284–309, 2003.
+[ bib ] + +
+ + +
+[337] +
+
+Juan Esteban Diaz, Julia Handl, and Dong-Ling Xu. + Evolutionary robust optimization in production planning: interactions between number of objectives, sample size and choice of robustness measure. + Computers & Operations Research, 79:266–278, 2017.
+[ bib | +DOI ] +
+Keywords: Evolutionary multi-objective optimization, Production + planning, Robust optimization, Simulation-based optimization, + Uncertainty modelling +
+ +
+ + +
+[338] +
+
+Juan Esteban Diaz, Julia Handl, and Dong-Ling Xu. + Integrating meta-heuristics, simulation and exact techniques for production planning of a failure-prone manufacturing system. + European Journal of Operational Research, 266(3):976–989, 2018.
+[ bib | +DOI ] +
+Keywords: Genetic algorithms, Combinatorial optimization, Production + planning, Simulation-based optimization, Uncertainty + modelling +
+ +
+ + +
+[339] +
+
+Juan Esteban Diaz and Manuel López-Ibáñez. + Incorporating Decision-Maker's Preferences into the Automatic Configuration of Bi-Objective Optimisation Algorithms. + European Journal of Operational Research, 289(3):1209–1222, 2021.
+[ bib | +DOI | +supplementary material ] +
+Automatic configuration (AC) methods are increasingly used to + tune and design optimisation algorithms for problems with + multiple objectives. Most AC methods use unary quality + indicators, which assign a single scalar value to an + approximation to the Pareto front, to compare the performance + of different optimisers. These quality indicators, however, + imply preferences beyond Pareto-optimality that may differ + from those of the decision maker (DM). Although it is + possible to incorporate DM's preferences into quality + indicators, e.g., by means of the weighted hypervolume + indicator (HVw), expressing preferences in terms of weight + function is not always intuitive nor an easy task for a DM, + in particular, when comparing the stochastic outcomes of + several algorithm configurations. A more visual approach to + compare such outcomes is the visualisation of their empirical + attainment functions (EAFs) differences. This paper proposes + using such visualisations as a way of eliciting information + about regions of the objective space that are preferred by + the DM. We present a method to convert the information about + EAF differences into a HVw that will assign higher quality + values to approximation fronts that result in EAF differences + preferred by the DM. We show that the resulting HVw may be + used by an AC method to guide the configuration of + multi-objective optimisers according to the preferences of + the DM. We evaluate the proposed approach on a well-known + benchmark problem. Finally, we apply our approach to + re-configuring, according to different DM's preferences, a + multi-objective optimiser tackling a real-world production + planning problem arising in the manufacturing industry. +
+ +
+ + +
+[340] +
+
+L. C. Dias, Vincent Mousseau, José Rui Figueira, and J. N. Clímaco. + An aggregation/disaggregation approach to obtain robust conclusions with ELECTRE TRI. + European Journal of Operational Research, 138(2):332–348, April 2002.
+[ bib ] + +
+ + +
+[341] +
+
+Ilias Diakonikolas and Mihalis Yannakakis. + Small approximate Pareto sets for biobjective shortest paths and other problems. + SIAM Journal on Computing, 39(4):1340–1371, 2009.
+[ bib ] + +
+ + +
+[342] +
+
+Gianni A. Di Caro and Marco Dorigo. + AntNet: Distributed Stigmergetic Control for Communications Networks. + Journal of Artificial Intelligence Research, 9:317–365, 1998.
+[ bib ] + +
+ + +
+[343] +
+
+Gianni A. Di Caro, F. Ducatelle, and L. M. Gambardella. + AntHocNet: An adaptive nature-inspired algorithm for routing in mobile ad hoc networks. + European Transactions on Telecommunications, 16(5):443–455, 2005.
+[ bib ] + +
+ + +
+[344] +
+
+Luca Di Gaspero and Andrea Schaerf. + EasyLocal++: An object-oriented framework for flexible design of local search algorithms. + Software — Practice & Experience, 33(8):733–765, July 2003.
+[ bib | +epub ] +
+Keywords: software engineering, local search, easylocal +
+ +
+ + +
+[345] +
+
+Bistra Dilkina, Elias B. Khalil, and George L. Nemhauser. + Comments on: On learning and branching: a survey. + TOP, 25:242–246, 2017.
+[ bib ] +
+Comments on [847]. +
+ +
+ + +
+[346] +
+
+Rui Ding, Hongbin Dong, Jun He, and Tao Li. + A novel two-archive strategy for evolutionary many-objective optimization algorithm based on reference points. + Applied Soft Computing, 78:447–464, 2019.
+[ bib | +DOI ] + +
+ + +
+[347] +
+
+J.-Y. Ding, S. Song, J. N. D. Gupta, R. Zhang, R. Chiong, and C. Wu. + An Improved Iterated Greedy Algorithm with a Tabu-based Reconstruction Strategy for the No-wait Flowshop Scheduling Problem. + Applied Soft Computing, 30:604–613, 2015.
+[ bib ] + +
+ + +
+[348] +
+
+Benjamin Doerr, Carola Doerr, and Franziska Ebel. + From black-box complexity to designing new genetic algorithms. + Theoretical Computer Science, 567:87–104, 2015.
+[ bib | +DOI ] + +
+ + +
+[349] +
+
+Benjamin Doerr, Carola Doerr, and Jing Yang. + Optimal parameter choices via precise black-box analysis. + Theoretical Computer Science, 801:1–34, 2020.
+[ bib | +DOI ] + +
+ + +
+[350] +
+
+Karl F. Doerner, Guenther Fuellerer, Manfred Gronalt, Richard F. Hartl, and Manuel Iori. + Metaheuristics for the Vehicle Routing Problem with Loading Constraints. + Networks, 49(4):294–307, 2006.
+[ bib ] + +
+ + +
+[351] +
+
+Benjamin Doerr, Christian Gießen, Carsten Witt, and Jing Yang. + The (1+λ) evolutionary algorithm with self-adjusting mutation rate. + Algorithmica, 81(2):593–631, 2019.
+[ bib ] + +
+ + +
+[352] +
+
+Karl F. Doerner, Walter J. Gutjahr, Richard F. Hartl, Christine Strauss, and Christian Stummer. + Nature-Inspired Metaheuristics in Multiobjective Activity Crashing. + Omega, 36(6):1019–1037, 2008.
+[ bib ] + +
+ + +
+[353] +
+
+Karl F. Doerner, Walter J. Gutjahr, Richard F. Hartl, Christine Strauss, and Christian Stummer. + Pareto Ant Colony Optimization: A Metaheuristic Approach to Multiobjective Portfolio Selection. + Annals of Operations Research, 131:79–99, 2004.
+[ bib ] + +
+ + +
+[354] +
+
+Karl F. Doerner, Walter J. Gutjahr, Richard F. Hartl, Christine Strauss, and Christian Stummer. + Pareto ant colony optimization with ILP preprocessing in multiobjective project portfolio selection. + European Journal of Operational Research, 171:830–841, 2006.
+[ bib ] + +
+ + +
+[355] +
+
+Karl F. Doerner, Richard F. Hartl, and Marc Reimann. + Are COMPETants more competent for problem solving? The case of a multiple objective transportation problem. + Central European Journal for Operations Research and Economics, 11(2):115–141, 2003.
+[ bib ] + +
+ + +
+[356] +
+
+Benjamin Doerr, Daniel Johannsen, and Carola Winzen. + Multiplicative drift analysis. + Algorithmica, 64(4):673–697, 2012.
+[ bib ] + +
+ + +
+[357] +
+
+Benjamin Doerr, Timo Kötzing, Johannes Lengler, and Carola Winzen. + Black-box complexities of combinatorial problems. + Theoretical Computer Science, 471:84–106, 2013.
+[ bib ] + +
+ + +
+[358] +
+
+Karl F. Doerner, D. Merkle, and Thomas Stützle. + Special issue on Ant Colony Optimization. + Swarm Intelligence, 3(1), 2009.
+[ bib ] + +
+ + +
+[359] +
+
+Benjamin Doerr, Frank Neumann, Dirk Sudholt, and Carsten Witt. + Runtime analysis of the 1-ANT ant colony optimizer. + Theoretical Computer Science, 412(1):1629–1644, 2011.
+[ bib ] + +
+ + +
+[360] +
+
+Doǧan Aydın. + Composite artificial bee colony algorithms: From component-based analysis to high-performing algorithms. + Applied Soft Computing, 32:266–285, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[361] +
+
+Jean-Paul Doignon, Aleksandar Pekeč, and Michel Regenwetter. + The repeated insertion model for rankings: Missing link between two subset choice models. + Psychometrika, 69(1):33–54, March 2004.
+[ bib | +DOI ] +
+Several probabilistic models for subset choice have been + proposed in the literature, for example, to explain approval + voting data. We show that Marley et al.'s latent scale model + is subsumed by Falmagne and Regenwetter's size-independent + model, in the sense that every choice probability + distribution generated by the former can also be explained by + the latter. Our proof relies on the construction of a + probabilistic ranking model which we label the “repeated + insertion model”. This model is a special case of Marden's + orthogonal contrast model class and, in turn, includes the + classical Mallows φ-model as a special case. We + explore its basic properties as well as its relationship to + Fligner and Verducci's multistage ranking model. +
+ +
+ + +
+[362] +
+
+Elizabeth D. Dolan and Jorge J. Moré. + Benchmarking optimization software with performance profiles. + Mathematical Programming, 91(2):201–213, 2002.
+[ bib ] +
+This methodology has been criticized in https://doi.org/10.1145/2950048 +
+
+Keywords: performance profiles; convergence +
+ +
+ + +
+[363] +
+
+Xingye Dong, Ping, Houkuan Huang, and Maciek Nowak. + A Multi-restart Iterated Local Search Algorithm for the Permutation Flow Shop Problem Minimizing Total Flow Time. + Computers & Operations Research, 40(2):627–632, 2013.
+[ bib ] + +
+ + +
+[364] +
+
+X. Dong, H. Huang, and P. Chen. + An Iterated Local Search Algorithm for the Permutation Flowshop Problem with Total Flowtime Criterion. + Computers & Operations Research, 36(5):1664–1669, 2009.
+[ bib ] + +
+ + +
+[365] +
+
+A. V. Donati, Roberto Montemanni, N. Casagrande, A. E. Rizzoli, and L. M. Gambardella. + Time dependent vehicle routing problem with a multi ant colony system. + European Journal of Operational Research, 185(3):1174–1191, 2008.
+[ bib ] + +
+ + +
+[366] +
+
+Marco Dorigo. + Ant Colony Optimization. + Scholarpedia, 2(3):1461, 2007.
+[ bib | +DOI ] + +
+ + +
+[367] +
+
+Marco Dorigo. + Swarm intelligence: A few things you need to know if you want to publish in this journal. + Swarm Intelligence, November 2016.
+[ bib | +http ] + +
+ + +
+[368] +
+
+Marco Dorigo, Mauro Birattari, Xiaodong Li, Manuel López-Ibáñez, Kazuhiro Ohkura, Carlo Pinciroli, and Thomas Stützle. + ANTS 2016 Special Issue: Editorial. + Swarm Intelligence, November 2017.
+[ bib | +DOI ] + +
+ + +
+[369] +
+
+Marco Dorigo, Mauro Birattari, and Thomas Stützle. + Ant Colony Optimization: Artificial Ants as a Computational Intelligence Technique. + IEEE Computational Intelligence Magazine, 1(4):28–39, 2006.
+[ bib ] + +
+ + +
+[370] +
+
+Marco Dorigo and Christian Blum. + Ant colony optimization theory: A survey. + Theoretical Computer Science, 344(2-3):243–278, 2005.
+[ bib ] + +
+ + +
+[371] +
+
+Marco Dorigo, Gianni A. Di Caro, and L. M. Gambardella. + Ant Algorithms for Discrete Optimization. + Artificial Life, 5(2):137–172, 1999.
+[ bib ] + +
+ + +
+[372] +
+
+Marco Dorigo and L. M. Gambardella. + Ant Colonies for the Traveling Salesman Problem. + BioSystems, 43(2):73–81, 1997.
+[ bib | +DOI ] + +
+ + +
+[373] +
+
+Marco Dorigo and L. M. Gambardella. + Ant Colony System: A Cooperative Learning Approach to the Traveling Salesman Problem. + IEEE Transactions on Evolutionary Computation, 1(1):53–66, 1997.
+[ bib ] +
+Keywords: Ant Colony System +
+ +
+ + +
+[374] +
+
+Marco Dorigo, L. M. Gambardella, Martin Middendorf, and Thomas Stützle. + Guest Editorial: Special Section on Ant Colony Optimization. + IEEE Transactions on Evolutionary Computation, 6(4):317–320, 2002.
+[ bib | +DOI ] +
+Keywords: ant colony optimization, swarm intelligence +
+ +
+ + +
+[375] +
+
+Marco Dorigo, Vittorio Maniezzo, and Alberto Colorni. + Ant System: Optimization by a Colony of Cooperating Agents. + IEEE Transactions on Systems, Man, and Cybernetics – Part B, 26(1):29–41, 1996.
+[ bib ] + +
+ + +
+[376] +
+
+Marco Dorigo, Thomas Stützle, and Gianni A. Di Caro. + Special Issue on “Ant Algorithms”. + Future Generation Computer Systems, 16(8), 2000.
+[ bib ] +
+Keywords: swarm intelligence, ant colony optimization +
+ +
+ + +
+[377] +
+
+Michael Doumpos and Constantin Zopounidis. + Preference disaggregation and statistical learning for multicriteria decision support: A review. + European Journal of Operational Research, 209(3):203–214, 2011.
+[ bib ] + +
+ + +
+[378] +
+
+Erik Dovgan, Tea Tušar, and Bogdan Filipič. + Parameter tuning in an evolutionary algorithm for commodity transportation optimization. + Evolutionary Computation, pp.  1–8, 2010.
+[ bib ] + +
+ + +
+[379] +
+
+Johann Dréo and P. Siarry. + Continuous interacting ant colony algorithm based on dense heterarchy. + Future Generation Computer Systems, 20(5):841–856, 2004.
+[ bib ] + +
+ + +
+[380] +
+
+Stefan Droste, Thomas Jansen, and Ingo Wegener. + Upper and lower bounds for randomized search heuristics in black-box optimization. + Theory of Computing Systems, 39(4):525–544, 2006.
+[ bib ] + +
+ + +
+[381] +
+
+Mădălina M. Drugan and Dirk Thierens. + Stochastic Pareto local search: Pareto neighbourhood exploration and perturbation strategies. + Journal of Heuristics, 18(5):727–766, 2012.
+[ bib ] + +
+ + +
+[382] +
+
+J. Du and Joseph Y.-T. Leung. + Minimizing Total Tardiness on One Machine is NP-Hard. + Mathematics of Operations Research, 15(3):483–495, 1990.
+[ bib ] + +
+ + +
+[383] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Improving the Anytime Behavior of Two-Phase Local Search. + Annals of Mathematics and Artificial Intelligence, 61(2):125–154, 2011.
+[ bib | +DOI ] + +
+ + +
+[384] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + A Hybrid TP+PLS Algorithm for Bi-objective Flow-Shop Scheduling Problems. + Computers & Operations Research, 38(8):1219–1236, 2011.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[385] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Anytime Pareto Local Search. + European Journal of Operational Research, 243(2):369–385, 2015.
+[ bib | +DOI ] +
+Keywords: Pareto local search +
+ +
+ + +
+[386] +
+
+Jérémie Dubois-Lacoste, Federico Pagnozzi, and Thomas Stützle. + An Iterated Greedy Algorithm with Optimization of Partial Solutions for the Permutation Flowshop Problem. + Computers & Operations Research, 81:160–166, 2017.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[387] +
+
+Fabian Duddeck. + Multidisciplinary optimization of car bodies. + Structural and Multidisciplinary Optimization, 35(4):375–389, 2008.
+[ bib | +DOI ] +
+Evolutionary optimization of car bodies at General Motors +
+ +
+ + +
+[388] +
+
+Gunter Dueck and T. Scheuer. + Threshold Accepting: A General Purpose Optimization Algorithm Appearing Superior to Simulated Annealing. + Journal of Computational Physics, 90(1):161–175, 1990.
+[ bib ] + +
+ + +
+[389] +
+
+Gunter Dueck. + New Optimization Heuristics: the Great Deluge Algorithm and the Record-To-Record Travel. + Journal of Computational Physics, 104(1):86–92, 1993.
+[ bib ] + +
+ + +
+[390] +
+
+Rikky R. P. R. Duivenvoorden, Felix Berkenkamp, Nicolas Carion, Andreas Krause, and Angela P. Schoellig. + Constrained Bayesian Optimization with Particle Swarms for Safe Adaptive Controller Tuning. + IFAC-PapersOnLine, 50(1):11800–11807, 2017.
+[ bib | +DOI ] +
+Tuning controller parameters is a recurring and + time-consuming problem in control. This is especially true in + the field of adaptive control, where good performance is + typically only achieved after significant tuning. Recently, + it has been shown that constrained Bayesian optimization is a + promising approach to automate the tuning process without + risking system failures during the optimization + process. However, this approach is computationally too + expensive for tuning more than a couple of parameters. In + this paper, we provide a heuristic in order to efficiently + perform constrained Bayesian optimization in high-dimensional + parameter spaces by using an adaptive discretization based on + particle swarms. We apply the method to the tuning problem of + an L1 adaptive controller on a quadrotor vehicle and show + that we can reliably and automatically tune parameters in + experiments. +
+
+20th IFAC World Congress +
+
+Keywords: Adaptive Control, Constrained Bayesian Optimization, Safety, + Gaussian Process, Particle Swarm Optimization, Policy Search, + Reinforcement learning +
+ +
+ + +
+[391] +
+
+Cees Duin and Stefan Voß. + The Pilot Method: A Strategy for Heuristic Repetition with Application to the Steiner Problem in Graphs. + Networks, 34(3):181–191, 1999.
+[ bib ] + +
+ + +
+[392] +
+
+Y. Dumas, J. Desrosiers, E. Gelinas, and M. M. Solomon. + An Optimal Algorithm for the Traveling Salesman Problem with Time Windows. + Operations Research, 43(2):367–371, 1995.
+[ bib | +DOI ] + +
+ + +
+[393] +
+
+Olive Jean Dunn. + Multiple Comparisons Using Rank Sums. + Technometrics, 6(3):241–252, 1964.
+[ bib ] + +
+ + +
+[394] +
+
+Olive Jean Dunn. + Multiple Comparisons Among Means. + Journal of the American Statistical Association, 56(293):52–64, 1961.
+[ bib ] + +
+ + +
+[395] +
+
+Juan J. Durillo and Antonio J. Nebro. + jMetal: A Java framework for multi-objective optimization. + Advances in Engineering Software, 42(10):760–771, 2011.
+[ bib | +DOI ] + +
+ + +
+[396] +
+
+Katharina Eggensperger, Marius Thomas Lindauer, and Frank Hutter. + Pitfalls and best practices in algorithm configuration. + Journal of Artificial Intelligence Research, 64:861–893, 2019.
+[ bib ] + +
+ + +
+[397] +
+
+Richard W. Eglese. + Simulated Annealing: a Tool for Operational Research. + European Journal of Operational Research, 46(3):271–281, 1990.
+[ bib ] + +
+ + +
+[398] +
+
+Matthias Ehrgott. + A discussion of scalarization techniques for multiple objective integer programming. + Annals of Operations Research, 147(1):343–360, 2006.
+[ bib ] + +
+ + +
+[399] +
+
+Matthias Ehrgott and Xavier Gandibleux. + Approximative Solution Methods for Combinatorial Multicriteria Optimization. + TOP, 12(1):1–88, 2004.
+[ bib ] + +
+ + +
+[400] +
+
+Matthias Ehrgott and Kathrin Klamroth. + Connectedness of Efficient Solutions in Multiple Criteria Combinatorial Optimization. + European Journal of Operational Research, 97(1):159–166, 1997.
+[ bib | +DOI ] + +
+ + +
+[401] +
+
+Agoston E. Eiben, Robert Hinterding, and Zbigniew Michalewicz. + Parameter Control in Evolutionary Algorithms. + IEEE Transactions on Evolutionary Computation, 3(2):124–141, 1999.
+[ bib ] + +
+ + +
+[402] +
+
+Agoston E. Eiben and Günther Rudolph. + Theory of evolutionary algorithms: A bird's eye view. + Theoretical Computer Science, 229(1-2):3–9, 1999.
+[ bib ] + +
+ + +
+[403] +
+
+Agoston E. Eiben and Selmar K. Smit. + Parameter Tuning for Configuring and Analyzing Evolutionary Algorithms. + Swarm and Evolutionary Computation, 1(1):19–31, 2011.
+[ bib | +DOI ] + +
+ + +
+[404] +
+
+Sibel Eker and Jan H. Kwakkel. + Including robustness considerations in the search phase of Many-Objective Robust Decision Making. + Environmental Modelling & Software, 105:201–216, 2018.
+[ bib ] +
+Keywords: scenario-based +
+ +
+ + +
+[405] +
+
+Jeffrey L Elman. + Distributed representations, simple recurrent networks, and grammatical structure. + Machine Learning, 7(2-3):195–225, 1991.
+[ bib ] + +
+ + +
+[406] +
+
+V. A. Emelichev and V. A. Perepelitsa. + Complexity of Vector Optimization Problems on Graphs. + Optimization, 22(6):906–918, 1991.
+[ bib | +DOI ] + +
+ + +
+[407] +
+
+V. A. Emelichev and V. A. Perepelitsa. + On the Cardinality of the Set of Alternatives in Discrete Many-criterion Problems. + Discrete Mathematics and Applications, 2(5):461–471, 1992.
+[ bib ] + +
+ + +
+[408] +
+
+Michael T. M. Emmerich and André H. Deutz. + A tutorial on multiobjective optimization: Fundamentals and evolutionary methods. + Natural Computing, 17(3):585–609, 2018.
+[ bib ] + +
+ + +
+[409] +
+
+Michael T. M. Emmerich, K. C. Giannakoglou, and Boris Naujoks. + Single- and multiobjective evolutionary optimization assisted by Gaussian random field metamodels. + IEEE Transactions on Evolutionary Computation, 10(4):421–439, 2006.
+[ bib | +DOI ] + +
+ + +
+[410] +
+
+Alexander Engau and Margaret M. Wiecek. + 2D decision-making for multicriteria design optimization. + Structural and Multidisciplinary Optimization, 34:301–315, 2007.
+[ bib | +DOI ] + +
+ + +
+[411] +
+
+Alexander Engau and Margaret M. Wiecek. + Interactive coordination of objective decompositions in multiobjective programming. + Management Science, 54(7):1350–1363, 2008.
+[ bib ] + +
+ + +
+[412] +
+
+Imen Essafi, Yazid Mati, and Stéphane Dauzère-Pèrés. + A Genetic Local Search Algorithm for Minimizing Total Weighted Tardiness in the Job-shop Scheduling Problem. + Computers & Operations Research, 35(8):2599–2616, 2008.
+[ bib ] + +
+ + +
+[413] +
+
+Wei Fan and Albert Bifet. + Mining big data: current status, and forecast to the future. + ACM SIGKDD Explorations Newsletter, 14(2):1–5, 2013.
+[ bib ] + +
+ + +
+[414] +
+
+Daniele Fanelli. + Negative results are disappearing from most disciplines and countries. + Scientometrics, 90(3):891–904, 2012.
+[ bib | +DOI ] +
+Concerns that the growing competition for funding and + citations might distort science are frequently discussed, but + have not been verified directly. Of the hypothesized + problems, perhaps the most worrying is a worsening of + positive-outcome bias. A system that disfavours negative + results not only distorts the scientific literature directly, + but might also discourage high-risk projects and pressure + scientists to fabricate and falsify their data. This study + analysed over 4,600 papers published in all disciplines + between 1990 and 2007, measuring the frequency of papers + that, having declared to have “tested” a hypothesis, + reported a positive support for it. The overall frequency of + positive supports has grown by over 22% between 1990 and + 2007, with significant differences between disciplines and + countries. The increase was stronger in the social and some + biomedical disciplines. The United States had published, over + the years, significantly fewer positive results than Asian + countries (and particularly Japan) but more than European + countries (and in particular the United + Kingdom). Methodological artefacts cannot explain away these + patterns, which support the hypotheses that research is + becoming less pioneering and/or that the objectivity with + which results are produced and published is decreasing. +
+ +
+ + +
+[415] +
+
+H. Faria, Jr, S. Binato, Mauricio G. C. Resende, and D. J. Falcão. + Power transmission network design by a greedy randomized adaptive path relinking approach. + IEEE Transactions on Power Systems, 20(1):43–49, 2005.
+[ bib ] + +
+ + +
+[416] +
+
+Vincent E. Farrugia, Héctor P. Martínez, and Georgios N. Yannakakis. + The Preference Learning Toolbox. + Arxiv preprint arXiv:1506.01709, 2015.
+[ bib | +DOI ] + +
+ + +
+[417] +
+
+R. Farmani, Godfrey A. Walters, and Dragan A. Savic. + Evolutionary multi-objective optimization of the design and operation of water distribution network: total cost vs. reliability vs. water quality. + Journal of Hydroinformatics, 8(3):165–179, 2006.
+[ bib ] + +
+ + +
+[418] +
+
+D. Favaretto, E. Moretti, and Paola Pellegrini. + Ant colony system approach for variants of the traveling salesman problem with time windows. + Journal of Information and Optimization Sciences, 27(1):35–54, 2006.
+[ bib ] + +
+ + +
+[419] +
+
+D. Favaretto, E. Moretti, and Paola Pellegrini. + Ant Colony System for a VRP with Multiple Time Windows and Multiple Visits. + Journal of Interdisciplinary Mathematics, 10(2):263–284, 2007.
+[ bib ] + +
+ + +
+[420] +
+
+Chris Fawcett and Holger H. Hoos. + Analysing Differences Between Algorithm Configurations through Ablation. + Journal of Heuristics, 22(4):431–458, 2016.
+[ bib ] + +
+ + +
+[421] +
+
+T. A. Feo and Mauricio G. C. Resende. + A Probabilistic Heuristic for a Computationally Difficult Set Covering Problem. + Operations Research Letters, 8(2):67–71, 1989.
+[ bib ] +
+Proposed GRASP +
+ +
+ + +
+[422] +
+
+T. A. Feo and Mauricio G. C. Resende. + Greedy Randomized Adaptive Search Procedures. + Journal of Global Optimization, 6(2):109–113, 1995.
+[ bib ] + +
+ + +
+[423] +
+
+T. A. Feo, Mauricio G. C. Resende, and S. H. Smith. + A Greedy Randomized Adaptive Search Procedure for Maximum Independent Set. + Operations Research, 42:860–878, October 1994.
+[ bib ] +
+Keywords: GRASP +
+ +
+ + +
+[424] +
+
+Victor Fernandez-Viagas and Jose M. Framiñán. + On Insertion Tie-breaking Rules in Heuristics for the Permutation Flowshop Scheduling Problem. + Computers & Operations Research, 45:60–67, 2014.
+[ bib ] + +
+ + +
+[425] +
+
+Victor Fernandez-Viagas and Jose M. Framiñán. + A Beam-search-based Constructive Heuristic for the PFSP to Minimise Total Flowtime. + Computers & Operations Research, 81:167–177, 2017.
+[ bib ] + +
+ + +
+[426] +
+
+Victor Fernandez-Viagas and Jose M. Framiñán. + Iterated-greedy-based algorithms with beam search initialization for the permutation flowshop to minimise total tardiness. + Expert Systems with Applications, 94:58–69, 2018.
+[ bib ] + +
+ + +
+[427] +
+
+Javier Ferrer, José García-Nieto, Enrique Alba, and Francisco Chicano. + Intelligent Testing of Traffic Light Programs: Validation in Smart Mobility Scenarios. + Mathematical Problems in Engineering, 2016:1–19, 2016.
+[ bib | +DOI ] + +
+ + +
+[428] +
+
+Alberto Ferrer, Daniel Guimarans, Helena Ramalhinho Lourenço, and Angel A. Juan. + A BRILS Metaheuristic for Non-smooth Flow-shop Problems with Failure-risk Costs. + Expert Systems with Applications, 44:177–186, 2016.
+[ bib ] + +
+ + +
+[429] +
+
+Javier Ferrer, Manuel López-Ibáñez, and Enrique Alba. + Reliable Simulation-Optimization of Traffic Lights in a Real-World City. + Applied Soft Computing, 78:697–711, 2019.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[430] +
+
+Eduardo Fernandez, Jorge Navarro, and Sergio Bernal. + Multicriteria Sorting Using a Valued Indifference Relation Under a Preference Disaggregation Paradigm. + European Journal of Operational Research, 198(2):602–609, 2009.
+[ bib ] + +
+ + +
+[431] +
+
+Victor Fernandez-Viagas, Rubén Ruiz, and Jose M. Framiñán. + A New Vision of Approximate Methods for the Permutation Flowshop to Minimise Makespan: State-of-the-art and Computational Evaluation. + European Journal of Operational Research, 257(3):707–721, 2017.
+[ bib ] + +
+ + +
+[432] +
+
+R. Ferreira da Silva and S. Urrutia. + A General VNS Heuristic for the Traveling Salesman Problem with Time Windows. + Discrete Optimization, 7(4):203–211, 2010.
+[ bib | +DOI ] +
+Keywords: TSPTW, GVNS +
+ +
+ + +
+[433] +
+
+Victor Fernandez-Viagas, Jorge M. S. Valente, and Jose M. Framiñán. + Iterated-greedy-based algorithms with Beam Search Initialization for the Permutation Flowshop to Minimise Total Tardiness. + Expert Systems with Applications, 94:58–69, 2018.
+[ bib ] + +
+ + +
+[434] +
+
+Álvaro Fialho, Luis Da Costa, Marc Schoenauer, and Michèle Sebag. + Analyzing Bandit-based Adaptive Operator Selection Mechanisms. + Annals of Mathematics and Artificial Intelligence, 60(1–2):25–64, 2010.
+[ bib ] + +
+ + +
+[435] +
+
+Mark J. Fielding. + Simulated Annealing with an Optimal Fixed Temperature. + SIAM Journal on Optimization, 11(2):289–307, 2000.
+[ bib ] + +
+ + +
+[436] +
+
+Jonathan E. Fieldsend, Richard M. Everson, and Sameer Singh. + Using unconstrained elite archives for multiobjective optimization. + IEEE Transactions on Evolutionary Computation, 7(3):305–323, 2003.
+[ bib | +DOI ] + +
+ + +
+[437] +
+
+José Rui Figueira, Carlos M. Fonseca, Pascal Halffmann, Kathrin Klamroth, Luís Paquete, Stefan Ruzika, Britta Schulze, Michael Stiglmayr, and David Willems. + Easy to say they are Hard, but Hard to see they are Easy-Towards a Categorization of Tractable Multiobjective Combinatorial Optimization Problems. + Journal of Multi-Criteria Decision Analysis, 24(1-2):82–98, 2017.
+[ bib | +DOI ] + +
+ + +
+[438] +
+
+Andreas Fischbach and Thomas Bartz-Beielstein. + Improving the reliability of test functions generators. + Applied Soft Computing, 92:106315, 2020.
+[ bib ] + +
+ + +
+[439] +
+
+Matteo Fischetti, Fred Glover, and Andrea Lodi. + The feasibility pump. + Mathematical Programming, 104(1):91–104, 2005.
+[ bib ] + +
+ + +
+[440] +
+
+Matteo Fischetti and Andrea Lodi. + Local Branching. + Mathematical Programming Series B, 98:23–47, 2003.
+[ bib ] + +
+ + +
+[441] +
+
+Matteo Fischetti and Michele Monaci. + Proximity search for 0-1 mixed-integer convex programming. + Journal of Heuristics, 20(6):709–731, 2014.
+[ bib ] + +
+ + +
+[442] +
+
+Matteo Fischetti and Michele Monaci. + Exploiting Erraticism in Search. + Operations Research, 62(1):114–122, 2014.
+[ bib | +DOI ] +
+ High sensitivity to initial conditions is generally viewed + as a drawback of tree search methods because it leads to + erratic behavior to be mitigated somehow. In this paper we + investigate the opposite viewpoint and consider this behavior + as an opportunity to exploit. Our working hypothesis is that + erraticism is in fact just a consequence of the exponential + nature of tree search that acts as a chaotic amplifier, so it + is largely unavoidable. We propose a bet-and-run approach to + actually turn erraticism to one's advantage. The idea is to + make a number of short sample runs with randomized initial + conditions, to bet on the "most promising" run selected + according to certain simple criteria, and to bring it to + completion. Computational results on a large testbed of mixed + integer linear programs from the literature are presented, + showing the potential of this approach even when embedded in + a proof-of-concept implementation. +
+
+http://mat.tepper.cmu.edu/blog/?p=1695 +
+ +
+ + +
+[443] +
+
+Matteo Fischetti, Michele Monaci, and Domenico Salvagnin. + Three Ideas for the Quadratic Assignment Problem. + Operations Research, 60(4):954–964, 2012.
+[ bib ] + +
+ + +
+[444] +
+
+Matteo Fischetti and Domenico Salvagnin. + Feasibility pump 2.0. + Mathematical Programming Computation, 1(2–3):201–222, 2009.
+[ bib ] + +
+ + +
+[445] +
+
+Roger Fletcher. + A new approach to variable metric algorithms. + The Computer Journal, 13(3):317–322, September 1970.
+[ bib | +DOI ] +
+One of the four papers that proposed BFGS. +
+
+Keywords: BFGS +
+ +
+ + +
+[446] +
+
+Charles Fleurent and Fred Glover. + Improved constructive multistart strategies for the quadratic assignment problem using adaptive memory. + INFORMS Journal on Computing, 11(2):198–204, 1999.
+[ bib ] + +
+ + +
+[447] +
+
+Jörg Fliege. + The effects of adding objectives to an optimisation problem on the solution set. + Operations Research Letters, 35(6):782–790, 2007.
+[ bib ] + +
+ + +
+[448] +
+
+Michael A. Fligner and Joseph S. Verducci. + Distance Based Ranking Models. + Journal of the Royal Statistical Society: Series B (Methodological), 48(3):359–369, 1986.
+[ bib | +DOI ] +
+Keywords: Mallows model, ranking, probabilistic models +
+ +
+ + +
+[449] +
+
+M. M. Flood. + The Travelling Salesman Problem. + Operations Research, 4:61–75, 1956.
+[ bib ] + +
+ + +
+[450] +
+
+D. Floreano and L. Keller. + Evolution of Adaptive Behaviour in Robots by Means of Darwinian Selection. + PLoS Biology, 8(1):e1000292, 2010.
+[ bib | +DOI ] + +
+ + +
+[451] +
+
+D. Floreano and J. Urzelai. + Evolutionary robots with on-line self-organization and behavioral fitness. + Neural Networks, 13(4-5):431–443, 2000.
+[ bib ] + +
+ + +
+[452] +
+
+Benito E. Flores. + A pragmatic view of accuracy measurement in forecasting. + Omega, 14(2):93–98, 1986.
+[ bib ] +
+Proposed symmetric mean absolute percentage error (SMAPE) +
+ +
+ + +
+[453] +
+
+Filippo Focacci, Andrea Lodi, and Michela Milano. + A Hybrid Exact Algorithm for the TSPTW. + INFORMS Journal on Computing, 14:403–417, 2002.
+[ bib ] + +
+ + +
+[454] +
+
+Carlos M. Fonseca and Peter J. Fleming. + An overview of evolutionary algorithms in multiobjective optimization. + Evolutionary Computation, 3(1):1–16, 1995.
+[ bib ] +
+Proposed FON benchmark problem +
+ +
+ + +
+[455] +
+
+Carlos M. Fonseca and Peter J. Fleming. + Multiobjective Optimization and Multiple Constraint Handling with Evolutionary Algorithms (II): Application Example. + IEEE Transactions on Systems, Man, and Cybernetics – Part A, 28(1):38–44, January 1998.
+[ bib | +DOI ] + +
+ + +
+[456] +
+
+Carlos M. Fonseca and Peter J. Fleming. + Multiobjective Optimization and Multiple Constraint Handling with Evolutionary Algorithms (I): A Unified Formulation. + IEEE Transactions on Systems, Man, and Cybernetics – Part A, 28(1):26–37, January 1998.
+[ bib | +DOI ] + +
+ + +
+[457] +
+
+Alexander I. J. Forrester and Andy J. Keane. + Recent advances in surrogate-based optimization. + Progress in Aerospace Sciences, 45(1-3):50–79, 2009.
+[ bib | +DOI ] +
+Keywords: Kriging; Gaussian Process; EGO; Design of Experiments +
+ +
+ + +
+[458] +
+
+John W. Fowler, Esma S. Gel, Murat Köksalan, Pekka Korhonen, Jon L. Marquis, and Jyrki Wallenius. + Interactive evolutionary multi-objective optimization for quasi-concave preference functions. + European Journal of Operational Research, 206(2):417–425, 2010.
+[ bib | +DOI ] +
+We present a new hybrid approach to interactive evolutionary + multi-objective optimization that uses a partial preference + order to act as the fitness function in a customized genetic + algorithm. We periodically send solutions to the decision + maker (DM) for her evaluation and use the resulting + preference information to form preference cones consisting of + inferior solutions. The cones allow us to implicitly rank + solutions that the DM has not considered. This technique + avoids assuming an exact form for the preference function, + but does assume that the preference function is + quasi-concave. This paper describes the genetic algorithm and + demonstrates its performance on the multi-objective knapsack + problem. +
+
+Keywords: Interactive optimization, Multi-objective optimization, + Evolutionary optimization, Knapsack problem +
+ +
+ + +
+[459] +
+
+Bennett L. Fox. + Integrating and accelerating tabu search, simulated annealing, and genetic algorithms. + Annals of Operations Research, 41(2):47–67, 1993.
+[ bib ] + +
+ + +
+[460] +
+
+Peter I. Frazier. + A Tutorial on Bayesian Optimization. + Arxiv preprint arXiv:1807.02811, 2018.
+[ bib | +DOI ] + +
+ + +
+[461] +
+
+Alberto Franzin. + Empirical Analysis of Stochastic Local Search Behaviour: Connecting Structure, Components and Landscape. + 4OR: A Quarterly Journal of Operations Research, 2022.
+[ bib | +DOI ] + +
+ + +
+[462] +
+
+G. Francesca, M. Brambilla, A. Brutschy, Vito Trianni, and Mauro Birattari. + AutoMoDe: A Novel Approach to the Automatic Design of Control Software for Robot Swarms. + Swarm Intelligence, 8(2):89–112, 2014.
+[ bib | +DOI ] + +
+ + +
+[463] +
+
+Gianpiero Francesca, Manuele Brambilla, Arne Brutschy, Lorenzo Garattoni, Roman Miletitch, Gaetan Podevijn, Andreagiovanni Reina, Touraj Soleymani, Mattia Salvaro, Carlo Pinciroli, Franco Mascia, Vito Trianni, and Mauro Birattari. + AutoMoDe-Chocolate: Automatic Design of Control Software for Robot Swarms. + Swarm Intelligence, 2015.
+[ bib | +DOI ] +
+Keywords: Swarm robotics; Automatic design; AutoMoDe +
+ +
+ + +
+[464] +
+
+Jose M. Framiñán, Jatinder N.D. Gupta, and Rainer Leisten. + A Review and Classification of Heuristics for Permutation Flow-shop Scheduling with Makespan Objective. + Journal of the Operational Research Society, 55(12):1243–1255, 2004.
+[ bib ] + +
+ + +
+[465] +
+
+Alberto Franzin, Leslie Pérez Cáceres, and Thomas Stützle. + Effect of Transformations of Numerical Parameters in Automatic Algorithm Configuration. + Optimization Letters, 12(8):1741–1753, 2018.
+[ bib | +DOI ] + +
+ + +
+[466] +
+
+Alberto Franzin, Francesco Sambo, and Barbara Di Camillo. + bnstruct: an R package for Bayesian Network structure learning in the presence of missing data. + Bioinformatics, 33(8):1250–1252, 2016.
+[ bib ] + +
+ + +
+[467] +
+
+Alberto Franzin and Thomas Stützle. + Revisiting Simulated Annealing: A Component-Based Analysis. + Computers & Operations Research, 104:191–206, 2019.
+[ bib | +DOI ] + +
+ + +
+[468] +
+
+Alberto Franzin and Thomas Stützle. + A Landscape-based Analysis of Fixed Temperature and Simulated Annealing. + European Journal of Operational Research, 304(2):395–410, 2023.
+[ bib | +DOI ] + +
+ + +
+[469] +
+
+Brendan J. Frey and Delbert Dueck. + Clustering by Passing Messages Between Data Points. + Science, 315(5814):972–976, February 2007.
+[ bib | +DOI ] +
+Keywords: clustering; affinity propagation +
+ +
+ + +
+[470] +
+
+Alan R. R. de Freitas, Peter J. Fleming, and Frederico G. Guimarães. + Aggregation trees for visualization and dimension reduction in many-objective optimization. + Information Sciences, 298:288–314, 2015.
+[ bib ] + +
+ + +
+[471] +
+
+Hela Frikha, Habib Chabchoub, and Jean-Marc Martel. + Inferring criteria's relative importance coefficients in PROMETHEE II. + International Journal of Operational Research, 7(2):257–275, 2010.
+[ bib ] + +
+ + +
+[472] +
+
+Matteo Frigo and Steven G. Johnson. + The Design and Implementation of FFTW3. + Proceedings of the IEEE, 93(2):216–231, 2005. + Special issue on “Program Generation, Optimization, and Platform Adaptation”.
+[ bib ] + +
+ + +
+[473] +
+
+Milton Friedman. + The use of ranks to avoid the assumption of normality implicit in the analysis of variance. + Journal of the American Statistical Association, 32(200):675–701, 1937.
+[ bib ] + +
+ + +
+[474] +
+
+Z Fu, R Eglese, and L Y O Li. + A unified tabu search algorithm for vehicle routing problems with soft time windows. + Journal of the Operational Research Society, 59(5):663–673, 2008.
+[ bib ] + +
+ + +
+[475] +
+
+Guenther Fuellerer, Karl F. Doerner, Richard F. Hartl, and Manuel Iori. + Metaheuristics for vehicle routing problems with three-dimensional loading constraints. + European Journal of Operational Research, 201(3):751–759, 2009.
+[ bib | +DOI ] + +
+ + +
+[476] +
+
+Guenther Fuellerer, Karl F. Doerner, Richard F. Hartl, and Manuel Iori. + Ant colony optimization for the two-dimensional loading vehicle routing problem. + Computers & Operations Research, 36(3):655–673, 2009.
+[ bib ] + +
+ + +
+[477] +
+
+Alex S. Fukunaga. + Automated Discovery of Local Search Heuristics for Satisfiability Testing. + Evolutionary Computation, 16(1):31–61, March 2008.
+[ bib | +DOI ] +
+The development of successful metaheuristic + algorithms such as local search for a difficult + problem such as satisfiability testing (SAT) is a + challenging task. We investigate an evolutionary + approach to automating the discovery of new local + search heuristics for SAT. We show that several + well-known SAT local search algorithms such as + Walksat and Novelty are composite heuristics that + are derived from novel combinations of a set of + building blocks. Based on this observation, we + developed CLASS, a genetic programming system that + uses a simple composition operator to automatically + discover SAT local search heuristics. New + heuristics discovered by CLASS are shown to be + competitive with the best Walksat variants, + including Novelty+. Evolutionary algorithms have + previously been applied to directly evolve a + solution for a particular SAT instance. We show + that the heuristics discovered by CLASS are also + competitive with these previous, direct evolutionary + approaches for SAT. We also analyze the local + search behavior of the learned heuristics using the + depth, mobility, and coverage metrics proposed by + Schuurmans and Southey. +
+ +
+ + +
+[478] +
+
+Grigori Fursin, Yuriy Kashnikov, Abdul Wahid Memon, Zbigniew Chamski, Olivier Temam, Mircea Namolaru, Elad Yom-Tov, Bilha Mendelson, Ayal Zaks, Eric Courtois, Francois Bodin, Phil Barnard, Elton Ashton, Edwin Bonilla, John Thomson, Christopher K. I. Williams, and Michael O'Boyle. + Milepost GCC: Machine Learning Enabled Self-tuning Compiler. + International Journal of Parallel Programming, 39(3):296–327, 2011.
+[ bib | +DOI ] + +
+ + +
+[479] +
+
+Caroline Gagné, W. L. Price, and M. Gravel. + Comparing an ACO algorithm with other heuristics for the single machine scheduling problem with sequence-dependent setup times. + Journal of the Operational Research Society, 53:895–906, 2002.
+[ bib ] + +
+ + +
+[480] +
+
+Matteo Gagliolo and J. Schmidhuber. + Learning dynamic algorithm portfolios. + Annals of Mathematics and Artificial Intelligence, 47(3-4):295–328, 2007.
+[ bib | +DOI ] +
+fully dynamic and online algorithm selection technique, with + no separate training phase: all candidate algorithms are run + in parallel, while a model incrementally learns their runtime + distributions. +
+ +
+ + +
+[481] +
+
+Philippe Galinier and Jin-Kao Hao. + Hybrid evolutionary algorithms for graph coloring. + Journal of Combinatorial Optimization, 3(4):379–397, 1999.
+[ bib | +DOI ] + +
+ + +
+[482] +
+
+Tomas Gal and Heiner Leberling. + Redundant objective functions in linear vector maximum problems and their determination. + European Journal of Operational Research, 1(3):176–184, 1977.
+[ bib | +DOI ] +
+Suppose that in a multicriteria linear programming problem + among the given objective functions there are some which can + be deleted without influencing the set E of all efficient + solutions. Such objectives are said to be + redundant. Introducing systems of objective functions which + realize their individual optimum in a single vertex of the + polyhedron generated by the restriction set, the notion of + relative or absolute redundant objectives is defined. A + theory which describes properties of absolute and relative + redundant objectives is developed. A method for determining + all the relative and absolute redundant objectives, based on + this theory, is given. Illustrative examples demonstrate the + procedure. +
+ +
+ + +
+[483] +
+
+L. M. Gambardella and Marco Dorigo. + Ant Colony System Hybridized with a New Local Search for the Sequential Ordering Problem. + INFORMS Journal on Computing, 12(3):237–255, 2000.
+[ bib ] + +
+ + +
+[484] +
+
+L. M. Gambardella, Roberto Montemanni, and Dennis Weyland. + Coupling Ant Colony Systems with Strong Local Searches. + European Journal of Operational Research, 220(3):831–843, 2012.
+[ bib | +DOI ] + +
+ + +
+[485] +
+
+Xavier Gandibleux, Andrzej Jaszkiewicz, A. Fréville, and Roman Slowiński. + Special Issue on Multiple Objective Metaheuristics. + Journal of Heuristics, 6(3), 2000.
+[ bib ] + +
+ + +
+[486] +
+
+Kaizhou Gao, Yicheng Zhang, Ali Sadollah, and Rong Su. + Optimizing urban traffic light scheduling problem using harmony search with ensemble of local search. + Applied Soft Computing, 48:359–372, November 2016.
+[ bib | +DOI ] +
+Keywords: harmony search algorithm,traffic light scheduling +
+ +
+ + +
+[487] +
+
+Huiru Gao, Haifeng Nie, and Ke Li. + Visualisation of Pareto Front Approximation: A Short Survey and Empirical Comparisons. + Arxiv preprint arXiv:1903.01768, 2019.
+[ bib | +http ] + +
+ + +
+[488] +
+
+José García-Nieto, Enrique Alba, and Ana Carolina Olivera. + Swarm intelligence for traffic light scheduling: Application to real urban areas. + Engineering Applications of Artificial Intelligence, 25(2):274–283, March 2012.
+[ bib ] +
+Keywords: Cycle program optimization,Particle swarm + optimization,Realistic traffic instances,SUMO microscopic + simulator of urban mobility,Traffic light scheduling +
+ +
+ + +
+[489] +
+
+Carlos García-Martínez, Oscar Cordón, and Francisco Herrera. + A taxonomy and an empirical analysis of multiple objective ant colony optimization algorithms for the bi-criteria TSP. + European Journal of Operational Research, 180(1):116–148, 2007.
+[ bib ] + +
+ + +
+[490] +
+
+Javier García and Fernando Fernández. + A comprehensive survey on safe reinforcement learning. + Journal of Machine Learning Research, 16(1):1437–1480, 2015.
+[ bib | +epub ] + +
+ + +
+[491] +
+
+Salvador García, Alberto Fernández, Julián Luengo, and Francisco Herrera. + Advanced nonparametric tests for multiple comparisons in the design of experiments in computational intelligence and data mining: Experimental analysis of power. + Information Sciences, 180(10):2044–2064, 2010.
+[ bib ] + +
+ + +
+[492] +
+
+Carlos García-Martínez, Fred Glover, Francisco J. Rodríguez, Manuel Lozano, and Rafael Martí. + Strategic Oscillation for the Quadratic Multiple Knapsack Problem. + Computational Optimization and Applications, 58(1):161–185, 2014.
+[ bib ] + +
+ + +
+[493] +
+
+M. R. Garey, David S. Johnson, and R. Sethi. + The Complexity of Flowshop and Jobshop Scheduling. + Mathematics of Operations Research, 1:117–129, 1976.
+[ bib ] + +
+ + +
+[494] +
+
+Josselin Garnier and Leila Kallel. + Efficiency of Local Search with Multiple Local Optima. + SIAM Journal Discrete Mathematics, 15(1):122–141, 2001.
+[ bib | +DOI ] + +
+ + +
+[495] +
+
+Salvador García, Daniel Molina, Manuel Lozano, and Francisco Herrera. + A study on the use of non-parametric tests for analyzing the evolutionary algorithms' behaviour: a case study on the CEC'2005 Special Session on Real Parameter Optimization. + Journal of Heuristics, 15(617):617–644, 2009.
+[ bib | +DOI ] + +
+ + +
+[496] +
+
+José García-Nieto, Ana Carolina Olivera, and Enrique Alba. + Optimal Cycle Program of Traffic Lights With Particle Swarm Optimization. + IEEE Transactions on Evolutionary Computation, 17(6):823–839, December 2013.
+[ bib | +DOI ] + +
+ + +
+[497] +
+
+Carlos García-Martínez, Francisco J. Rodríguez, and Manuel Lozano. + Arbitrary function optimisation with metaheuristics: No free lunch and real-world problems. + Soft Computing, 16(12):2115–2133, 2012.
+[ bib | +DOI ] + +
+ + +
+[498] +
+
+Carlos García-Martínez, Francisco J. Rodríguez, and Manuel Lozano. + Tabu-enhanced Iterated Greedy Algorithm: A Case Study in the Quadratic Multiple Knapsack Problem. + European Journal of Operational Research, 232(3):454–463, 2014.
+[ bib ] + +
+ + +
+[499] +
+
+Gauci Melvin, Tony J. Dodd, and Roderich Groß. + Why `GSA: a gravitational search algorithm' is not genuinely based on the law of gravity. + Natural Computing, 11(4):719–720, 2012.
+[ bib ] + +
+ + +
+[500] +
+
+Martin Josef Geiger. + Decision Support for Multi-objective Flow Shop Scheduling by the Pareto Iterated Local Search Methodology. + Computers and Industrial Engineering, 61(3):805–812, 2011.
+[ bib ] + +
+ + +
+[501] +
+
+Martin Josef Geiger. + A Multi-threaded Local Search Algorithm and Computer Implementation for the Multi-mode, Resource-constrained Multi-project Scheduling Problem. + European Journal of Operational Research, 256:729–741, 2017.
+[ bib ] + +
+ + +
+[502] +
+
+Stuart Geman and Donald Geman. + Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images. + IEEE Transactions on Pattern Analysis and Machine Intelligence, 6(6):721–741, 1984.
+[ bib ] + +
+ + +
+[503] +
+
+Michel Gendreau, Francois Guertin, Jean-Yves Potvin, and Éric D. Taillard. + Parallel tabu search for real-time vehicle routing and dispatching. + Transportation Science, 33(4):381–390, 1999.
+[ bib ] + +
+ + +
+[504] +
+
+Michel Gendreau, Francois Guertin, Jean-Yves Potvin, and René Séguin. + Neighborhood search heuristics for a dynamic vehicle dispatching problem with pick-ups and deliveries. + Transportation Research Part C: Emerging Technologies, 14(3):157–174, 2006.
+[ bib ] + +
+ + +
+[505] +
+
+Mitsuo Gen and Lin Lin. + Multiobjective evolutionary algorithm for manufacturing scheduling problems: state-of-the-art survey. + Journal of Intelligent Manufacturing, 25(5):849–866, 2014.
+[ bib ] + +
+ + +
+[506] +
+
+Robin Genuer, Jean-Michel Poggi, and Christine Tuleau-Malot. + Variable selection using random forests. + Pattern Recognition Letters, 31(14):2225–2236, 2010.
+[ bib ] + +
+ + +
+[507] +
+
+Michel Gendreau, A. Hertz, Gilbert Laporte, and M. Stan. + A Generalized Insertion Heuristic for the Traveling Salesman Problem with Time Windows. + Operations Research, 46:330–335, 1998.
+[ bib ] + +
+ + +
+[508] +
+
+Samuel J. Gershman, Eric J. Horvitz, and Joshua B. Tenenbaum. + Computational rationality: A converging paradigm for intelligence in brains, minds, and machines. + Science, 349(6245):273–278, 2015.
+[ bib | +DOI | +epub ] +
+After growing up together, and mostly growing apart in the + second half of the 20th century, the fields of artificial + intelligence (AI), cognitive science, and neuroscience are + reconverging on a shared view of the computational + foundations of intelligence that promotes valuable + cross-disciplinary exchanges on questions, methods, and + results. We chart advances over the past several decades that + address challenges of perception and action under uncertainty + through the lens of computation. Advances include the + development of representations and inferential procedures for + large-scale probabilistic inference and machinery for + enabling reflection and decisions about tradeoffs in effort, + precision, and timeliness of computations. These tools are + deployed toward the goal of computational rationality: + identifying decisions with highest expected utility, while + taking into consideration the costs of computation in complex + real-world problems in which most relevant calculations can + only be approximated. We highlight key concepts with examples + that show the potential for interchange between computer + science, cognitive science, and neuroscience. +
+ +
+ + +
+[509] +
+
+Pierre Geurts, Damien Ernst, and Louis Wehenkel. + Extremely randomized trees. + Machine Learning, 63(1):3–42, March 2006.
+[ bib | +DOI ] +
+Proposed ExtraTrees +
+ +
+ + +
+[510] +
+
+Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung. + The Google File System. + SIGOPS Oper. Syst. Rev., 37(5):29–43, December 2003.
+[ bib ] + +
+ + +
+[511] +
+
+K. Ghoseiri and B. Nadjari. + An ant colony optimization algorithm for the bi-objective shortest path problem. + Applied Soft Computing, 10(4):1237–1246, 2010.
+[ bib ] + +
+ + +
+[512] +
+
+Nicolas Girerd, Muriel Rabilloud, Philippe Pibarot, Patrick Mathieu, and Pascal Roy. + Quantification of Treatment Effect Modification on Both an Additive and Multiplicative Scale. + PLoS One, 11(4):1–14, April 2016.
+[ bib | +DOI ] + +
+ + +
+[513] +
+
+Fred Glover. + Heuristics for Integer Programming Using Surrogate Constraints. + Decision Sciences, 8:156–166, 1977.
+[ bib ] + +
+ + +
+[514] +
+
+Fred Glover. + Future Paths for Integer Programming and Links to Artificial Intelligence. + Computers & Operations Research, 13(5):533–549, 1986.
+[ bib ] + +
+ + +
+[515] +
+
+Fred Glover. + Tabu Search – Part I. + INFORMS Journal on Computing, 1(3):190–206, 1989.
+[ bib | +DOI ] + +
+ + +
+[516] +
+
+Fred Glover. + Tabu Search – Part II. + INFORMS Journal on Computing, 2(1):4–32, 1990.
+[ bib ] + +
+ + +
+[517] +
+
+Fred Glover and Jin-Kao Hao. + The case for Strategic Oscillation. + Annals of Operations Research, 183(1):163–173, 2011.
+[ bib ] + +
+ + +
+[518] +
+
+Fred Glover, Gary A. Kochenberger, and Bahram Alidaee. + Adaptive Memory Tabu Search for Binary Quadratic Programs. + Management Science, 44(3):336–345, 1998.
+[ bib ] + +
+ + +
+[519] +
+
+Fred Glover, Zhipeng Lü, and Jin-Kao Hao. + Diversification-driven tabu search for unconstrained binary quadratic problems. + 4OR: A Quarterly Journal of Operations Research, 8(3):239–253, 2010.
+[ bib | +DOI ] + +
+ + +
+[520] +
+
+Marc Goerigk and Anita Schöbel. + Recovery-to-optimality: A new two-stage approach to robustness with an application to aperiodic timetabling. + Computers & Operations Research, 52:1–15, 2014.
+[ bib ] + +
+ + +
+[521] +
+
+Donald Goldfarb. + A Family of Variable-Metric Methods Derived by Variational Means. + Mathematics of Computation, 24(109):23–26, 1970.
+[ bib ] +
+One of the four papers that proposed BFGS. +
+
+Keywords: BFGS +
+ +
+ + +
+[522] +
+
+David E. Goldberg. + Probability matching, the magnitude of reinforcement, and classifier system bidding. + Machine Learning, 5(4):407–425, 1990.
+[ bib ] + +
+ + +
+[523] +
+
+Zaiwu Gong, Ning Zhang, and Francisco Chiclana. + The optimization ordering model for intuitionistic fuzzy preference relations with utility functions. + Knowledge-Based Systems, 162:174–184, 2018.
+[ bib | +DOI ] +
+Intuitionistic fuzzy sets describe information from the three + aspects of membership degree, non-membership degree and + hesitation degree, which has more practical significance when + uncertainty pervades qualitative decision problems. In this + paper, we investigate the problem of ranking intuitionistic + fuzzy preference relations (IFPRs) based on various + non-linear utility functions. First, we transform IFPRs into + their isomorphic interval-value fuzzy preference relations + (IVFPRs), and utilise non-linear utility functions, such as + parabolic, S-shaped, and hyperbolic absolute risk aversion, + to fit the true value of a decision-maker's + judgement. Ultimately, the optimization ordering models for + the membership and non-membership of IVFPRs based on utility + function are constructed, with objective function aiming at + minimizing the distance deviation between the multiplicative + consistency ideal judgment and the actual judgment, + represented by utility function, subject to the + decision-maker's utility constraints. The proposed models + ensure that more factual and optimal ranking of alternative + is acquired, avoiding information distortion caused by the + operations of intervals. Second, by introducing a + non-Archimedean infinitesimal, we establish the optimization + ordering model for IFPRs with the priority of utility or + deviation, which realises the goal of prioritising solutions + under multi-objective programming. Subsequently, we verify + that a close connection exists between the ranking for + membership and non-membership degree IVFPRs. Comparison + analyses with existing approaches are summarized to + demonstrate that the proposed models have advantage in + dealing with group decision making problems with IFPRs. +
+
+Special Issue on intelligent decision-making and consensus + under uncertainty in inconsistent and dynamic environments +
+
+Keywords: Intuitionistic fuzzy preference relation, Utility function, + Ranking, Multiplicative consistency, Non-archimedean + infinitesimal +
+ +
+ + +
+[524] +
+
+Jochen Gorski, Kathrin Klamroth, and Stefan Ruzika. + Connectedness of Efficient Solutions in Multiple Objective Combinatorial Optimization. + Journal of Optimization Theory and Applications, 150(3):475–497, 2011.
+[ bib | +DOI ] + +
+ + +
+[525] +
+
+Abhijit Gosavi. + Reinforcement Learning: A Tutorial Survey and Recent Advances. + INFORMS Journal on Computing, 21(2):178–192, 2009.
+[ bib | +DOI ] + +
+ + +
+[526] +
+
+N. I. M. Gould, D. Orban, and P. L. Toint. + CUTEr and SifDec: A constrained and unconstrained testing environment, revisited. + ACM Transactions on Mathematical Software, 29:373–394, 2003.
+[ bib ] + +
+ + +
+[527] +
+
+Jonathan Gratch and Steve A. Chien. + Adaptive Problem-solving for Large-scale Scheduling Problems: A Case Study. + Journal of Artificial Intelligence Research, 4:365–396, 1996.
+[ bib ] +
+Earliest hyper-heuristic? +
+ +
+ + +
+[528] +
+
+Robert B. Gramacy and Herbert K. H. Lee. + Bayesian Treed Gaussian Process Models With an Application to Computer Modeling. + Journal of the American Statistical Association, 103:1119–1130, 2008.
+[ bib | +DOI ] +
+Keywords: Treed-GP +
+ +
+ + +
+[529] +
+
+Alex Grasas, Angel A. Juan, and Helena Ramalhinho Lourenço. + SimILS: A Simulation-based Extension of the Iterated Local Search Metaheuristic for Stochastic Combinatorial Optimization. + Journal of Simulation, 10(1):69–77, 2016.
+[ bib ] + +
+ + +
+[530] +
+
+M. Gravel, W. L. Price, and Caroline Gagné. + Scheduling continuous casting of aluminum using a multiple objective ant colony optimization metaheuristic. + European Journal of Operational Research, 143(1):218–229, 2002.
+[ bib | +DOI ] + +
+ + +
+[531] +
+
+John J. Grefenstette. + Optimization of Control Parameters for Genetic Algorithms. + IEEE Transactions on Systems, Man, and Cybernetics, 16(1):122–128, 1986.
+[ bib | +DOI ] +
+Keywords: parameter tuning +
+ +
+ + +
+[532] +
+
+Salvatore Greco, Milosz Kadziński, Vincent Mousseau, and Roman Slowiński. + ELECTREGKMS: Robust ordinal regression for outranking methods. + European Journal of Operational Research, 214(1):118–135, 2011.
+[ bib ] + +
+ + +
+[533] +
+
+Salvatore Greco, Vincent Mousseau, and Roman Slowiński. + Robust ordinal regression for value functions handling interacting criteria. + European Journal of Operational Research, 239(3):711–730, 2014.
+[ bib | +DOI ] + +
+ + +
+[534] +
+
+David R. Grimes, Chris T. Bauch, and John P. A. Ioannidis. + Modelling science trustworthiness under publish or perish pressure. + Royal Society Open Science, 5:171511, 2018.
+[ bib ] + +
+ + +
+[535] +
+
+Andrea Grosso, Federico Della Croce, and R. Tadei. + An Enhanced Dynasearch Neighborhood for the Single-Machine Total Weighted Tardiness Scheduling Problem. + Operations Research Letters, 32(1):68–72, 2004.
+[ bib ] + +
+ + +
+[536] +
+
+Andrea Grosso, A. R. M. J. U. Jamali, and Marco Locatelli. + Finding Maximin Latin Hypercube Designs by Iterated Local Search Heuristics. + European Journal of Operational Research, 197(2):541–547, 2009.
+[ bib ] + +
+ + +
+[537] +
+
+Peter Groves, Basel Kayyali, David Knott, and Steve Van Kuiken. + The "big data" revolution in healthcare. + McKinsey Quarterly, 2, 2013.
+[ bib ] + +
+ + +
+[538] +
+
+Benoît Groz and Silviu Maniu. + Hypervolume subset selection with small subsets. + Evolutionary Computation, 27(4):611–637, 2019.
+[ bib ] + +
+ + +
+[539] +
+
+Viviane Grunert da Fonseca and Carlos M. Fonseca. + A link between the multivariate cumulative distribution function and the hitting function for random closed sets. + Statistics & Probability Letters, 57(2):179–182, 2002.
+[ bib | +DOI ] + +
+ + +
+[540] +
+
+Andreia P. Guerreiro, Carlos M. Fonseca, and Luís Paquete. + The Hypervolume Indicator: Computational Problems and Algorithms. + ACM Computing Surveys, 54(6):1–42, 2021.
+[ bib ] + +
+ + +
+[541] +
+
+Andreia P. Guerreiro, Vasco Manquinho, and José Rui Figueira. + Exact hypervolume subset selection through incremental computations. + Computers & Operations Research, 136:105–471, December 2021.
+[ bib | +DOI ] + +
+ + +
+[542] +
+
+Gonzalo Guillén-Gosálbez. + A novel MILP-based objective reduction method for multi-objective optimization: Application to environmental problems. + Computers & Chemical Engineering, 35(8):1469–1477, 2011.
+[ bib | +DOI ] +
+Multi-objective optimization has recently emerged as a useful + technique in sustainability analysis, as it can assist in the + study of optimal trade-off solutions that balance several + criteria. The main limitation of multi-objective optimization + is that its computational burden grows in size with the + number of objectives. This computational barrier is critical + in environmental applications in which decision-makers seek + to minimize simultaneously several environmental indicators + of concern. With the aim to overcome this limitation, this + paper introduces a systematic method for reducing the number + of objectives in multi-objective optimization with emphasis + on environmental problems. The approach presented relies on a + novel mixed-integer linear programming formulation that + minimizes the error of omitting objectives. We test the + capabilities of this technique through two environmental + problems of different nature in which we attempt to minimize + a set of life cycle assessment impacts. Numerical examples + demonstrate that certain environmental metrics tend to behave + in a non-conflicting manner, which makes it possible to + reduce the dimension of the problem without losing + information. +
+
+Keywords: Environmental engineering, Life cycle assessment, + Multi-objective optimization, Objective reduction +
+ +
+ + +
+[543] +
+
+Odd Erik Gundersen, Yolanda Gil, and David W. Aha. + On Reproducible AI: Towards Reproducible Research, Open Science, and Digital Scholarship in AI Publications. + AI Magazine, 39(3):56–68, September 2018.
+[ bib | +DOI ] +
+The reproducibility guidelines can be found here: + https://folk.idi.ntnu.no/odderik/reproducibility_guidelines.pdf + and a short how-to can be found here: + https://folk.idi.ntnu.no/odderik/reproducibility_guidelines_how_to.html +
+ +
+ + +
+[544] +
+
+Aldy Gunawan, Kien Ming Ng, and Kim Leng Poh. + A Hybridized Lagrangian Relaxation and Simulated Annealing Method for the Course Timetabling Problem. + Computers & Operations Research, 39(12):3074–3088, 2012.
+[ bib ] + +
+ + +
+[545] +
+
+J. N. D. Gupta. + Flowshop schedules with sequence dependent setup times. + Journal of Operations Research Society of Japan, 29:206–219, 1986.
+[ bib ] + +
+ + +
+[546] +
+
+Walter J. Gutjahr. + A Graph-based Ant System and its Convergence. + Future Generation Computer Systems, 16(8):873–888, 2000.
+[ bib ] + +
+ + +
+[547] +
+
+Walter J. Gutjahr. + ACO Algorithms with Guaranteed Convergence to the Optimal Solution. + Information Processing Letters, 82(3):145–153, 2002.
+[ bib ] + +
+ + +
+[548] +
+
+Walter J. Gutjahr. + On the finite-time dynamics of ant colony optimization. + Methodology and Computing in Applied Probability, 8(1):105–133, 2006.
+[ bib ] + +
+ + +
+[549] +
+
+Walter J. Gutjahr. + Mathematical runtime analysis of ACO algorithms: survey on an emerging issue. + Swarm Intelligence, 1(1):59–79, 2007.
+[ bib ] + +
+ + +
+[550] +
+
+Walter J. Gutjahr and Marion S. Rauner. + An ACO algorithm for a dynamic regional nurse-scheduling problem in Austria. + Computers & Operations Research, 34(3):642–666, 2007.
+[ bib | +DOI ] +
+To the best of our knowledge, this paper describes the first + ant colony optimization (ACO) approach applied to nurse + scheduling, analyzing a dynamic regional problem which is + currently under discussion at the Vienna hospital + compound. Each day, pool nurses have to be assigned for the + following days to public hospitals while taking into account + a variety of soft and hard constraints regarding working date + and time, working patterns, nurses qualifications, nurses + and hospitals preferences, as well as costs. Extensive + computational experiments based on a four week simulation + period were used to evaluate three different scenarios + varying the number of nurses and hospitals for six different + hospitals demand intensities. The results of our simulations + and optimizations reveal that the proposed ACO algorithm + achieves highly significant improvements compared to a greedy + assignment algorithm. +
+ +
+ + +
+[551] +
+
+Walter J. Gutjahr. + First steps to the runtime complexity analysis of ant colony optimization. + Computers & Operations Research, 35(9):2711–2727, 2008.
+[ bib ] + +
+ + +
+[552] +
+
+Walter J. Gutjahr and G. Sebastiani. + Runtime analysis of ant colony optimization with best-so-far reinforcement. + Methodology and Computing in Applied Probability, 10(3):409–433, 2008.
+[ bib ] + +
+ + +
+[553] +
+
+Gregory Gutin, Anders Yeo, and Alexey Zverovich. + Traveling salesman should not be greedy: domination analysis of greedy-type heuristics for the TSP. + Discrete Applied Mathematics, 117(1–3), 2002.
+[ bib ] + +
+ + +
+[554] +
+
+Isabelle Guyon, Jason Weston, Stephen Barnhill, and Vladimir Vapnik. + Gene selection for cancer classification using support vector machines. + Machine Learning, 46(1):389–422, 2002.
+[ bib ] +
+Keywords: recursive feature elimination +
+ +
+ + +
+[555] +
+
+Heikki Haario, Eero Saksman, and Johanna Tamminen. + An adaptive Metropolis algorithm. + Bernoulli, 7(2):223–242, 2001.
+[ bib ] + +
+ + +
+[556] +
+
+David Hadka and Patrick M. Reed. + Borg: An Auto-Adaptive Many-Objective Evolutionary Computing Framework. + Evolutionary Computation, 21(2):231–259, 2013.
+[ bib ] + +
+ + +
+[557] +
+
+David Hadka and Patrick M. Reed. + Diagnostic Assessment of Search Controls and Failure Modes in Many-Objective Evolutionary Optimization. + Evolutionary Computation, 20(3):423–452, 2012.
+[ bib ] + +
+ + +
+[558] +
+
+Josef Hadar and William R. Russell. + Rules for ordering uncertain prospects. + The American Economic Review, 59(1):25–34, 1969.
+[ bib | +epub ] +
+Keywords: stochastic dominance +
+ +
+ + +
+[559] +
+
+Y. Haimes, L. Lasdon, and D. Da Wismer. + On a bicriterion formation of the problems of integrated system identification and system optimization. + IEEE Transactions on Systems, Man, and Cybernetics, 1(3):296–297, 1971.
+[ bib | +DOI ] +
+Keywords: epsilon-constraint method +
+ +
+ + +
+[560] +
+
+Prabhat Hajela and C-Y Lin. + Genetic search strategies in multicriterion optimal design. + Structural Optimization, 4(2):99–107, 1992.
+[ bib ] + +
+ + +
+[561] +
+
+Bruce Hajek and Galen Sasaki. + Simulated annealing–to cool or not. + System & Control Letters, 12(5):443–447, 1989.
+[ bib ] + +
+ + +
+[562] +
+
+Bruce Hajek. + Cooling Schedules for Optimal Annealing. + Mathematics of Operations Research, 13(2):311–329, 1988.
+[ bib ] + +
+ + +
+[563] +
+
+George T. Hall, Pietro S. Oliveto, and Dirk Sudholt. + On the impact of the performance metric on efficient algorithm configuration. + Artificial Intelligence, 303:103629, February 2022.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[564] +
+
+Raimo P. Hämäläinen and Tuomas J. Lahtinen. + Path dependence in Operational Research–How the modeling process can influence the results. + Operations Research Perspectives, 3:14–20, January 2016.
+[ bib | +DOI ] +
+In Operational Research practice there are almost always + alternative paths that can be followed in the modeling and + problem solving process. Path dependence refers to the impact + of the path on the outcome of the process. The steps of the + path include, e.g. forming the problem solving team, the + framing and structuring of the problem, the choice of model, + the order in which the different parts of the model are + specified and solved, and the way in which data or + preferences are collected. We identify and discuss seven + possibly interacting origins or drivers of path dependence: + systemic origins, learning, procedure, behavior, motivation, + uncertainty, and external environment. We provide several + ideas on how to cope with path dependence. +
+
+Keywords: Behavioral Biases, Behavioral Operational Research, Ethics in + modelling, OR practice, Path dependence, Systems perspective +
+ +
+ + +
+[565] +
+
+Raimo P. Hämäläinen, Jukka Luoma, and Esa Saarinen. + On the importance of behavioral operational research: The case of understanding and communicating about dynamic systems. + European Journal of Operational Research, 228(3):623–634, August 2013.
+[ bib | +DOI ] +
+We point out the need for Behavioral Operational Research + (BOR) in advancing the practice of OR. So far, in OR + behavioral phenomena have been acknowledged only in + behavioral decision theory but behavioral issues are always + present when supporting human problem solving by + modeling. Behavioral effects can relate to the group + interaction and communication when facilitating with OR + models as well as to the possibility of procedural mistakes + and cognitive biases. As an illustrative example we use well + known system dynamics studies related to the understanding of + accumulation. We show that one gets completely opposite + results depending on the way the phenomenon is described and + how the questions are phrased and graphs used. The results + suggest that OR processes are highly sensitive to various + behavioral effects. As a result, we need to pay attention to + the way we communicate about models as they are being + increasingly used in addressing important problems like + climate change. +
+ +
+ + +
+[566] +
+
+Horst W. Hamacher and Günter Ruhe. + On spanning tree problems with multiple objectives. + Annals of Operations Research, 52(4):209–230, 1994.
+[ bib ] + +
+ + +
+[567] +
+
+Nikolaus Hansen, Anne Auger, Dimo Brockhoff, and Tea Tušar. + Anytime Performance Assessment in Blackbox Optimization Benchmarking. + IEEE Transactions on Evolutionary Computation, 26(6):1293–1305, December 2022.
+[ bib | +DOI ] + +
+ + +
+[568] +
+
+Nikolaus Hansen, Anne Auger, Olaf Mersmann, Tea Tušar, and Dimo Brockhoff. + COCO: A platform for comparing continuous optimizers in a black-box setting. + Arxiv preprint arXiv:1603.08785, 2016. + Published as [569].
+[ bib ] + +
+ + +
+[569] +
+
+Nikolaus Hansen, Anne Auger, Raymond Ros, Olaf Mersmann, Tea Tušar, and Dimo Brockhoff. + COCO: A platform for comparing continuous optimizers in a black-box setting. + Optimization Methods and Software, 36(1):1–31, 2020.
+[ bib | +DOI ] + +
+ + +
+[570] +
+
+Pierre Hansen and B. Jaumard. + Algorithms for the Maximum Satisfiability Problem. + Computing, 44:279–303, 1990.
+[ bib ] + +
+ + +
+[571] +
+
+Pierre Hansen and Nenad Mladenović. + Variable neighborhood search: Principles and applications. + European Journal of Operational Research, 130(3):449–467, 2001.
+[ bib ] + +
+ + +
+[572] +
+
+Nikolaus Hansen and A. Ostermeier. + Completely derandomized self-adaptation in evolution strategies. + Evolutionary Computation, 9(2):159–195, 2001.
+[ bib | +DOI ] +
+Keywords: CMA-ES +
+ +
+ + +
+[573] +
+
+Nikolaus Hansen, Raymond Ros, Nikolaus Mauny, Marc Schoenauer, and Anne Auger. + Impacts of invariance in search: When CMA-ES and PSO face ill-conditioned and non-separable problems. + Applied Soft Computing, 11(8):5755–5769, 2011.
+[ bib ] + +
+ + +
+[574] +
+
+Thomas Hanne. + On the convergence of multiobjective evolutionary algorithms. + European Journal of Operational Research, 117(3):553–564, 1999.
+[ bib | +DOI ] +
+Keywords: archiving, efficiency presserving +
+ +
+ + +
+[575] +
+
+Thomas Hanne. + A multiobjective evolutionary algorithm for approximating the efficient set. + European Journal of Operational Research, 176(3):1723–1734, 2007.
+[ bib ] + +
+ + +
+[576] +
+
+Douglas P. Hardin and Edward B. Saff. + Discretizing Manifolds via Minimum Energy Points. + Notices of the American Mathematical Society, 51(10):1186–1194, 2004.
+[ bib ] + +
+ + +
+[577] +
+
+J. P. Hart and A. W. Shogan. + Semi-greedy heuristics: An empirical study. + Operations Research Letters, 6(3):107–114, 1987.
+[ bib ] + +
+ + +
+[578] +
+
+Emma Hart and Kevin Sim. + A Hyper-Heuristic Ensemble Method for Static Job-Shop Scheduling. + Evolutionary Computation, 24(4):609–635, 2016.
+[ bib | +DOI ] + +
+ + +
+[579] +
+
+Kazuya Haraguchi. + Iterated Local Search with Trellis-Neighborhood for the Partial Latin Square Extension Problem. + Journal of Heuristics, 22(5):727–757, 2016.
+[ bib ] + +
+ + +
+[580] +
+
+Sameer Hasija and Chandrasekharan Rajendran. + Scheduling in flowshops to minimize total tardiness of jobs. + International Journal of Production Research, 42(11):2289–2301, 2004.
+[ bib | +DOI ] + +
+ + +
+[581] +
+
+Hideki Hashimoto, Mutsunori Yagiura, and Toshihide Ibaraki. + An Iterated Local Search Algorithm for the Time-dependent Vehicle Routing Problem with Time Windows. + Discrete Optimization, 5(2):434–456, 2008.
+[ bib ] + +
+ + +
+[582] +
+
+Simon Haykin. + A comprehensive foundation. + Neural Networks, 2:41, 2004.
+[ bib ] + +
+ + +
+[583] +
+
+Öncü Hazir, Yavuz Günalay, and Erdal Erel. + Customer order scheduling problem: a comparative metaheuristics study. + International Journal of Advanced Manufacturing Technology, 37(5):589–598, May 2008.
+[ bib | +DOI ] +
+The customer order scheduling problem (COSP) is defined as + to determine the sequence of tasks to satisfy the demand of + customers who order several types of products produced on a + single machine. A setup is required whenever a product type + is launched. The objective of the scheduling problem is to + minimize the average customer order flow time. Since the + customer order scheduling problem is known to be strongly + NP-hard, we solve it using four major metaheuristics and + compare the performance of these heuristics, namely, + simulated annealing, genetic algorithms, tabu search, and ant + colony optimization. These are selected to represent various + characteristics of metaheuristics: nature-inspired + vs. artificially created, population-based vs. local search, + etc. A set of problems is generated to compare the solution + quality and computational efforts of these heuristics. + Results of the experimentation show that tabu search and ant + colony perform better for large problems whereas simulated + annealing performs best in small-size problems. Some + conclusions are also drawn on the interactions between + various problem parameters and the performance of the + heuristics. +
+
+Keywords: ACO,Customer order scheduling,Genetic + algorithms,Meta-heuristics,Simulated annealing,Tabu + search +
+ +
+ + +
+[584] +
+
+Zhenan He and Gary G. Yen. + Many-Objective Evolutionary Algorithm: Objective Space Reduction and Diversity Improvement. + IEEE Transactions on Evolutionary Computation, 20(1):145–160, 2016.
+[ bib ] + +
+ + +
+[585] +
+
+Xin He, Kaiyong Zhao, and Xiaowen Chu. + AutoML: A survey of the state-of-the-art. + Knowledge-Based Systems, 212:106622, 2021.
+[ bib | +DOI ] + +
+ + +
+[586] +
+
+Sabine Helwig, Jürgen Branke, and Sanaz Mostaghim. + Experimental Analysis of Bound Handling Techniques in Particle Swarm Optimization. + IEEE Transactions on Evolutionary Computation, 17(2):259–271, April 2013.
+[ bib | +DOI ] + +
+ + +
+[587] +
+
+Michael Held and Richard M. Karp. + The Traveling-Salesman Problem and Minimum Spanning Trees. + Operations Research, 18(6):1138–1162, 1970.
+[ bib ] + +
+ + +
+[588] +
+
+Christoph Helmberg and Franz Rendl. + Solving quadratic (0,1)-problems by semidefinite programs and cutting planes. + Mathematical Programming, 82(3):291–315, 1998.
+[ bib ] + +
+ + +
+[589] +
+
+Keld Helsgaun. + An Effective Implementation of the Lin-Kernighan Traveling Salesman Heuristic. + European Journal of Operational Research, 126:106–130, 2000.
+[ bib ] + +
+ + +
+[590] +
+
+Keld Helsgaun. + General k-opt Submoves for the Lin-Kernighan TSP Heuristic. + Mathematical Programming Computation, 1(2–3):119–163, 2009.
+[ bib ] + +
+ + +
+[591] +
+
+Michael A. Heroux. + Editorial: ACM TOMS Replicated Computational Results Initiative. + ACM Transactions on Mathematical Software, 41(3):1–5, June 2015.
+[ bib | +DOI ] + +
+ + +
+[592] +
+
+H. Hernández and Christian Blum. + Ant colony optimization for multicasting in static wireless ad-hoc networks. + Swarm Intelligence, 3(2):125–148, 2009.
+[ bib ] + +
+ + +
+[593] +
+
+Alberto Herrán, J. Manuel Colmenar, and Abraham Duarte. + An efficient Variable Neighborhood Search for the Space-Free Multi-Row Facility Layout problem. + European Journal of Operational Research, 2021.
+[ bib | +DOI ] + +
+ + +
+[594] +
+
+Francisco Herrera, Manuel Lozano, and A. M. Sánchez. + A taxonomy for the crossover operator for real-coded genetic algorithms: An experimental study. + International Journal of Intelligent Systems, 18(3):309–338, 2003.
+[ bib | +DOI ] + +
+ + +
+[595] +
+
+Francisco Herrera, Manuel Lozano, and J. L. Verdegay. + Tackling Real-Coded Genetic Algorithms: Operators and Tools for Behavioural Analysis. + Artificial Intelligence Review, 12:265–319, 1998.
+[ bib ] +
+Keywords: genetic algorithms, real coding, continuous search + spaces, mutation, recombination +
+ +
+ + +
+[596] +
+
+Carlos Ignacio Hernández Castellanos and Oliver Schütze. + A Bounded Archiver for Hausdorff Approximations of the Pareto Front for Multi-Objective Evolutionary Algorithms. + Mathematical and Computational Applications, 27(3):48, 2022.
+[ bib | +DOI ] + +
+ + +
+[597] +
+
+Carlos Ignacio Hernández Castellanos, Oliver Schütze, J. Q. Sun, and S. Ober-Blöbaum. + Non-epsilon dominated evolutionary algorithm for the set of approximate solutions. + Mathematical and Computational Applications, 25(1):3, 2020.
+[ bib ] +
+Keywords: archiving, multimodal +
+ +
+ + +
+[598] +
+
+Jano I. van Hemert. + Evolving Combinatorial Problem Instances That Are Difficult to Solve. + Evolutionary Computation, 14(4):433–462, 2006.
+[ bib | +DOI ] +
+This paper demonstrates how evolutionary computation can be + used to acquire difficult to solve combinatorial problem + instances. As a result of this technique, the corresponding + algorithms used to solve these instances are + stress-tested. The technique is applied in three important + domains of combinatorial optimisation, binary constraint + satisfaction, Boolean satisfiability, and the travelling + salesman problem. The problem instances acquired through this + technique are more difficult than the ones found in popular + benchmarks. In this paper, these evolved instances are + analysed with the aim to explain their difficulty in terms of + structural properties, thereby exposing the weaknesses of + corresponding algorithms. +
+ +
+ + +
+[599] +
+
+Robert Heumüller, Sebastian Nielebock, Jacob Krüger, and Frank Ortmeier. + Publish or perish, but do not forget your software artifacts. + Empirical Software Engineering, 25(6):4585–4616, 2020.
+[ bib | +DOI ] + +
+ + +
+[600] +
+
+Christian Hicks. + A Genetic Algorithm tool for optimising cellular or functional layouts in the capital goods industry. + International Journal of Production Economics, 104(2):598–614, 2006.
+[ bib | +DOI ] + +
+ + +
+[601] +
+
+Robert M. Hierons, Miqing Li, Xiaohui Liu, Jose Antonio Parejo, Sergio Segura, and Xin Yao. + Many-objective test suite generation for software product lines. + ACM Transactions on Software Engineering and Methodology, 29(1):1–46, 2020.
+[ bib ] + +
+ + +
+[602] +
+
+Geoffrey E. Hinton and Ruslan R. Salakhutdinov. + Reducing the dimensionality of data with neural networks. + Science, 313(5786):504–507, 2006.
+[ bib ] + +
+ + +
+[603] +
+
+Wassily Hoeffding. + Probability inequalities for sums of bounded random variables. + Journal of the American Statistical Association, 58(301):13–30, 1963.
+[ bib ] + +
+ + +
+[604] +
+
+I. Hong, A. B. Kahng, and B. R. Moon. + Improved large-step Markov chain variants for the symmetric TSP. + Journal of Heuristics, 3(1):63–81, 1997.
+[ bib ] + +
+ + +
+[605] +
+
+John N. Hooker. + Needed: An Empirical Science of Algorithms. + Operations Research, 42(2):201–212, 1994.
+[ bib ] + +
+ + +
+[606] +
+
+John N. Hooker. + Testing Heuristics: We Have It All Wrong. + Journal of Heuristics, 1(1):33–42, 1996.
+[ bib | +DOI ] + +
+ + +
+[607] +
+
+Giles Hooker. + Generalized functional ANOVA diagnostics for high-dimensional functions of dependent variables. + Journal of Computational and Graphical Statistics, 16(3):709–732, 2012.
+[ bib | +DOI ] + +
+ + +
+[608] +
+
+Holger H. Hoos, Marius Thomas Lindauer, and Torsten Schaub. + Claspfolio 2: Advances in Algorithm Selection for Answer Set Programming. + Theory and Practice of Logic Programming, 14(4-5):560–585, 2014.
+[ bib ] + +
+ + +
+[609] +
+
+Holger H. Hoos and Thomas Stützle. + On the Empirical Scaling of Run-time for Finding Optimal Solutions to the Traveling Salesman Problem. + European Journal of Operational Research, 238(1):87–94, 2014.
+[ bib ] + +
+ + +
+[610] +
+
+Holger H. Hoos and Thomas Stützle. + On the Empirical Time Complexity of Finding Optimal Solutions vs. Proving Optimality for Euclidean TSP Instances. + Optimization Letters, 9(6):1247–1254, 2015.
+[ bib ] + +
+ + +
+[611] +
+
+Holger H. Hoos. + Programming by optimization. + Communications of the ACM, 55(2):70–80, February 2012.
+[ bib | +DOI ] + +
+ + +
+[612] +
+
+André Hottung, Shunji Tanaka, and Kevin Tierney. + Deep learning assisted heuristic tree search for the container pre-marshalling problem. + Computers & Operations Research, 113:104781, 2020.
+[ bib | +DOI ] + +
+ + +
+[613] +
+
+André Hottung and Kevin Tierney. + Neural large neighborhood search for routing problems. + Artificial Intelligence, 313:103786, December 2022.
+[ bib | +DOI ] + +
+ + +
+[614] +
+
+Stela Pudar Hozo, Benjamin Djulbegovic, and Iztok Hozo. + Estimating the mean and variance from the median, range, and the size of a sample. + BMC Medical Research Methodology, 5(1):13, 2005.
+[ bib ] + +
+ + +
+[615] +
+
+T. C. Hu, A. B. Kahng, and C.-W. A. Tsao. + Old Bachelor Acceptance: A New Class of Non-Monotone Threshold Accepting Methods. + ORSA Journal on Computing, 7(4):417–425, 1995.
+[ bib ] + +
+ + +
+[616] +
+
+Wenbin Hu, Huan Wang, Zhenyu Qiu, Cong Nie, and Liping Yan. + A quantum particle swarm optimization driven urban traffic light scheduling model. + Neural Computing & Applications, 2018.
+[ bib | +DOI ] +
+Keywords: BML,Optimization,Simulation,Traffic congestion,Updating + rules +
+ +
+ + +
+[617] +
+
+Wenbin Hu, Liping Yan, Huan Wang, Bo Du, and Dacheng Tao. + Real-time traffic jams prediction inspired by Biham, Middleton and Levine (BML) model. + Information Sciences, 2017.
+[ bib ] +
+Keywords: BML model,Prediction,Real-time,Traffic jam,Urban traffic + network +
+ +
+ + +
+[618] +
+
+Deng Huang, Theodore T. Allen, William I. Notz, and Ning Zeng. + Global Optimization of Stochastic Black-Box Systems via Sequential Kriging Meta-Models. + Journal of Global Optimization, 34(3):441–466, 2006.
+[ bib | +DOI ] + +
+ + +
+[619] +
+
+Changwu Huang, Yuanxiang Li, and Xin Yao. + A Survey of Automatic Parameter Tuning Methods for Metaheuristics. + IEEE Transactions on Evolutionary Computation, 24(2):201–216, 2020.
+[ bib | +DOI ] + +
+ + +
+[620] +
+
+S. Huband, P. Hingston, L. Barone, and L. While. + A Review of Multiobjective Test Problems and a Scalable Test Problem Toolkit. + IEEE Transactions on Evolutionary Computation, 10(5):477–506, 2006.
+[ bib | +DOI ] +
+Proposed the WFG benchmark suite +
+ +
+ + +
+[621] +
+
+B. Huberman, R. Lukose, and T. Hogg. + An Economic Approach to Hard Computational Problems. + Science, 275:51–54, 1997.
+[ bib ] + +
+ + +
+[622] +
+
+D. L. Huerta-Muñoz, R. Z. Ríos-Mercado, and Rubén Ruiz. + An Iterated Greedy Heuristic for a Market Segmentation Problem with Multiple Attributes. + European Journal of Operational Research, 261(1):75–87, 2017.
+[ bib ] + +
+ + +
+[623] +
+
+Jérémie Humeau, Arnaud Liefooghe, El-Ghazali Talbi, and Sébastien Verel. + ParadisEO-MO: From Fitness Landscape Analysis to Efficient Local Search Algorithms. + Journal of Heuristics, 19(6):881–915, June 2013.
+[ bib | +DOI ] + +
+ + +
+[624] +
+
+Ying Hung, V. Roshan Joseph, and Shreyes N. Melkote. + Design and Analysis of Computer Experiments With Branching and Nested Factors. + Technometrics, 51(4):354–365, 2009.
+[ bib | +DOI ] + +
+ + +
+[625] +
+
+M. Hurtgen and J.-C. Maun. + Optimal PMU placement using Iterated Local Search. + International Journal of Electrical Power & Energy Systems, 32(8):857–860, 2010.
+[ bib ] + +
+ + +
+[626] +
+
+S. H. Hurlbert. + Pseudoreplication and the Design of Ecological Field Experiments. + Ecological Monographs, 54(2):187–211, 1984.
+[ bib ] + +
+ + +
+[627] +
+
+Mohamed Saifullah Hussin and Thomas Stützle. + Tabu Search vs. Simulated Annealing for Solving Large Quadratic Assignment Instances. + Computers & Operations Research, 43:286–291, 2014.
+[ bib ] + +
+ + +
+[628] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Tradeoffs in the Empirical Evaluation of Competing Algorithm Designs. + Annals of Mathematics and Artificial Intelligence, 60(1–2):65–89, 2010.
+[ bib ] + +
+ + +
+[629] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Bayesian Optimization With Censored Response Data. + Arxiv preprint arXiv:1310.1947, 2013.
+[ bib | +http ] + +
+ + +
+[630] +
+
+Frank Hutter, Holger H. Hoos, Kevin Leyton-Brown, and Thomas Stützle. + ParamILS: An Automatic Algorithm Configuration Framework. + Journal of Artificial Intelligence Research, 36:267–306, October 2009.
+[ bib | +DOI ] + +
+ + +
+[631] +
+
+Frank Hutter, Marius Thomas Lindauer, Adrian Balint, Sam Bayless, Holger H. Hoos, and Kevin Leyton-Brown. + The Configurable SAT Solver Challenge (CSSC). + Artificial Intelligence, 243:1–25, 2017.
+[ bib | +DOI ] + +
+ + +
+[632] +
+
+Frank Hutter, Lin Xu, Holger H. Hoos, and Kevin Leyton-Brown. + Algorithm Runtime Prediction: Methods & evaluation. + Artificial Intelligence, 206:79–111, 2014.
+[ bib | +DOI ] +
+Perhaps surprisingly, it is possible to predict how long an + algorithm will take to run on a previously unseen input, + using machine learning techniques to build a model of the + algorithm's runtime as a function of problem-specific + instance features. Such models have important applications to + algorithm analysis, portfolio-based algorithm selection, and + the automatic configuration of parameterized algorithms. Over + the past decade, a wide variety of techniques have been + studied for building such models. Here, we describe + extensions and improvements of existing models, new families + of models, and—perhaps most importantly—a much more thorough + treatment of algorithm parameters as model inputs. We also + comprehensively describe new and existing features for + predicting algorithm runtime for propositional satisfiability + (SAT), travelling salesperson (TSP) and mixed integer + programming (MIP) problems. We evaluate these innovations + through the largest empirical analysis of its kind, comparing + to a wide range of runtime modelling techniques from the + literature. Our experiments consider 11 algorithms and 35 + instance distributions; they also span a very wide range of + SAT, MIP, and TSP instances, with the least structured having + been generated uniformly at random and the most structured + having emerged from real industrial applications. Overall, we + demonstrate that our new models yield substantially better + runtime predictions than previous approaches in terms of + their generalization to new problem instances, to new + algorithms from a parameterized space, and to both + simultaneously. +
+
+Keywords: Empirical performance models; Mixed integer programming; SAT +
+ +
+ + +
+[633] +
+
+Hao Wang, Diederick Vermetten, Furong Ye, Carola Doerr, and Thomas Bäck. + IOHanalyzer: Detailed Performance Analyses for Iterative Optimization Heuristics. + ACM Transactions on Evolutionary Learning and Optimization, 2(1):3:1–3:29, 2022.
+[ bib | +DOI ] + +
+ + +
+[634] +
+
+Jacob de Nobel, Furong Ye, Diederick Vermetten, Hao Wang, Carola Doerr, and Thomas Bäck. + IOHexperimenter: Benchmarking Platform for Iterative Optimization Heuristics. + Arxiv preprint arXiv:2111.04077, 2021.
+[ bib | +DOI ] +
+Published in ECJ [635] +
+ +
+ + +
+[635] +
+
+Jacob de Nobel, Furong Ye, Diederick Vermetten, Hao Wang, Carola Doerr, and Thomas Bäck. + IOHexperimenter: Benchmarking Platform for Iterative Optimization Heuristics. + Evolutionary Computation, pp.  1–6, 2024.
+[ bib | +DOI ] + +
+ + +
+[636] +
+
+Carola Doerr, Hao Wang, Furong Ye, Sander van Rijn, and Thomas Bäck. + IOHprofiler: A Benchmarking and Profiling Tool for Iterative Optimization Heuristics. + Arxiv preprint arXiv:1806.07555, October 2018.
+[ bib | +DOI ] +
+Keywords: Benchmarking; Heuristics +
+ +
+ + +
+[637] +
+
+Claudio Iacopino and Phil Palmer. + The Dynamics of Ant Colony Optimization Algorithms Applied to Binary Chains. + Swarm Intelligence, 6(4):343–377, 2012.
+[ bib ] + +
+ + +
+[638] +
+
+Claudio Iacopino, Phil Palmer, N. Policella, A. Donati, and A. Brewer. + How Ants Can Manage Your Satellites. + Acta Futura, 9:59–72, 2014.
+[ bib | +DOI ] +
+Keywords: ACO, Space +
+ +
+ + +
+[639] +
+
+Toshihide Ibaraki, Shinji Imahori, Koji Nonobe, Kensuke Sobue, Takeaki Uno, and Mutsunori Yagiura. + An Iterated Local Search Algorithm for the Vehicle Routing Problem with Convex Time Penalty Functions. + Discrete Applied Mathematics, 156(11):2050–2069, 2008.
+[ bib ] + +
+ + +
+[640] +
+
+Toshihide Ibaraki. + A Personal Perspective on Problem Solving by General Purpose Solvers. + International Transactions in Operational Research, 17(3):303–315, 2010.
+[ bib ] + +
+ + +
+[641] +
+
+Jonas Ide and Anita Schöbel. + Robustness for uncertain multi-objective optimization: a survey and analysis of different concepts. + OR Spectrum, 38(1):235–271, 2016.
+[ bib | +DOI ] +
+In this paper, we discuss various concepts of robustness for + uncertain multi-objective optimization problems. We extend + the concepts of flimsily, highly, and lightly robust + efficiency and we collect different versions of minmax robust + efficiency and concepts based on set order relations from the + literature. Altogether, we compare and analyze ten different + concepts and point out their relations to each + other. Furthermore, we present reduction results for the + class of objective-wise uncertain multi-objective + optimization problems. +
+ +
+ + +
+[642] +
+
+Christian Igel, Nikolaus Hansen, and S. Roth. + Covariance Matrix Adaptation for Multi-objective Optimization. + Evolutionary Computation, 15(1):1–28, 2007.
+[ bib ] + +
+ + +
+[643] +
+
+Christian Igel, V. Heidrich-Meisner, and T. Glasmachers. + Shark. + Journal of Machine Learning Research, 9:993–996, June 2008.
+[ bib | +http ] + +
+ + +
+[644] +
+
+Nesa Ilich and Slobodan P. Simonovic. + Evolutionary Algorithm for minimization of pumping cost. + Journal of Computing in Civil Engineering, ASCE, 12(4):232–240, October 1998.
+[ bib ] +
+This paper deals with minimizing the total cost of + pumping in a liquid pipeline. Previous experience + with the most common solution procedures in pipeline + optimization is discussed along with their strengths + and weaknesses. The proposed method is an + evolutionary algorithm with two distinct features: + (1) The search is restricted to feasible region + only; and (2) it utilizes a floating point decision + variable rather than integer or binary as is the + case with most other similar approaches. A numerical + example is presented as a basis for verification of + the proposed method and its comparison with the + existing solver that utilizes the nonlinear + Newtonian search. The proposed method provides + promising improvements in terms of optimality when + compared to the widespread gradient search methods + because it does not involve evaluation of the + gradient of the objective function. It also provides + potential to improve the performance of previous + evolutionary programs because it restricts the + search to the feasible region, thus eliminating + large overhead associated with generation and + inspection of solutions that are + infeasible. Comparison of the two solutions revealed + improvement of the solution in favor of the proposed + algorithm, which ranged up to 6% depending on the + initial values of the decision variables in the + Newtonian search. The proposed method was not + sensitive to the starting value of the decision + variables. +
+ +
+ + +
+[645] +
+
+Takashi Imamichi, Mutsunori Yagiura, and Hiroshi Nagamochi. + An Iterated Local Search Algorithm Based on Nonlinear Programming for the Irregular Strip Packing Problem. + Discrete Optimization, 6(4):345–361, 2009.
+[ bib ] + +
+ + +
+[646] +
+
+Alfred Inselberg. + The Plane with Parallel Coordinates. + The Visual Computer, 1(2):69–91, 1985.
+[ bib ] + +
+ + +
+[647] +
+
+John P. A. Ioannidis. + Why Most Published Research Findings Are False. + PLoS Medicine, 2(8):e124, 2005.
+[ bib | +DOI ] + +
+ + +
+[648] +
+
+Stefan Irnich. + A Unified Modeling and Solution Framework for Vehicle Routing and Local Search-Based Metaheuristics. + INFORMS Journal on Computing, 20(2):270–287, 2008.
+[ bib ] + +
+ + +
+[649] +
+
+Ekhine Irurozki, Borja Calvo, and José A. Lozano. + Sampling and Learning Mallows and Generalized Mallows Models Under the Cayley Distance. + Methodology and Computing in Applied Probability, 20(1):1–35, June 2016.
+[ bib | +DOI ] + +
+ + +
+[650] +
+
+Ekhine Irurozki, Borja Calvo, and José A. Lozano. + PerMallows: An R Package for Mallows and Generalized Mallows Models. + Journal of Statistical Software, 71, 2019.
+[ bib | +DOI ] +
+In this paper we present the R package PerMallows, which is a + complete toolbox to work with permutations, distances and + some of the most popular probability models for permutations: + Mallows and the Generalized Mallows models. The Mallows model + is an exponential location model, considered as analogous to + the Gaussian distribution. It is based on the definition of a + distance between permutations. The Generalized Mallows model + is its best-known extension. The package includes functions + for making inference, sampling and learning such + distributions. The distances considered in PerMallows are + Kendall's τ, Cayley, Hamming and Ulam. +
+
+Keywords: Cayley,Generalized Mallows,Hamming,Kendall's + τ,Learning,Mallows,Permutation,R,Ranking,Sampling,Ulam +
+ +
+ + +
+[651] +
+
+Ekhine Irurozki, Jesus Lobo, Aritz Perez, and Javier Del Ser. + Rank aggregation for non-stationary data streams. + Arxiv preprint arXiv:1910.08795 [stat.ML], 2020.
+[ bib | +http ] +
+Keywords: uborda +
+ +
+ + +
+[652] +
+
+Hisao Ishibuchi and T. Murata. + A multi-objective genetic local search algorithm and its application to flowshop scheduling. + IEEE Transactions on Systems, Man, and Cybernetics – Part C, 28(3):392–403, 1998.
+[ bib ] + +
+ + +
+[653] +
+
+Hisao Ishibuchi, N. Akedo, and Y. Nojima. + Behavior of Multiobjective Evolutionary Algorithms on Many-Objective Knapsack Problems. + IEEE Transactions on Evolutionary Computation, 19(2):264–283, 2015.
+[ bib | +DOI ] + +
+ + +
+[654] +
+
+Hisao Ishibuchi, Ryo Imada, Yu Setoguchi, and Yusuke Nojima. + How to specify a reference point in hypervolume calculation for fair performance comparison. + Evolutionary Computation, 26(3):411–440, 2018.
+[ bib ] + +
+ + +
+[655] +
+
+Hisao Ishibuchi, Shinta Misaki, and Hideo Tanaka. + Modified simulated annealing algorithms for the flow shop sequencing problem. + European Journal of Operational Research, 81(2):388–398, 1995.
+[ bib ] + +
+ + +
+[656] +
+
+Hisao Ishibuchi, Yu Setoguchi, Hiroyuki Masuda, and Yusuke Nojima. + Performance of decomposition-based many-objective algorithms strongly depends on Pareto front shapes. + IEEE Transactions on Evolutionary Computation, 21(2):169–190, 2017.
+[ bib ] + +
+ + +
+[657] +
+
+Peter Ivie and Douglas Thain. + Reproducibility in Scientific Computing. + ACM Computing Surveys, 51(3):1–36, 2019.
+[ bib | +DOI ] + +
+ + +
+[658] +
+
+Srikanth K. Iyer and Barkha Saxena. + Improved genetic algorithm for the permutation flowshop scheduling problem. + Computers & Operations Research, 31(4):593–606, 2004.
+[ bib | +DOI ] + +
+ + +
+[659] +
+
+Christopher H. Jackson. + Multi-State Models for Panel Data: The msm Package for R. + Journal of Statistical Software, 38(8):1–29, 2011.
+[ bib | +http ] + +
+ + +
+[660] +
+
+Richard H. F. Jackson, Paul T. Boggs, Stephen G. Nash, and Susan Powell. + Guidelines for Reporting Results of Computational Experiments. Report of the Ad Hoc Committee. + Mathematical Programming, 49(3):413–425, 1991.
+[ bib ] + +
+ + +
+[661] +
+
+Larry W. Jacobs and Michael J. Brusco. + A Local Search Heuristic for Large Set-Covering Problems. + Naval Research Logistics, 42(7):1129–1140, 1995.
+[ bib ] + +
+ + +
+[662] +
+
+Karen E. Jacowitz and Daniel Kahneman. + Measures of Anchoring in Estimation Tasks. + Personality and Social Psychology Bulletin, 21(11):1161–1166, November 1995.
+[ bib | +DOI ] +
+The authors describe a method for the quantitative study of + anchoring effects in estimation tasks. A calibration group + provides estimates of a set of uncertain quantities. Subjects + in the anchored condition first judge whether a specified + number (the anchor) is higher or lower than the true value + before estimating each quantity. The anchors are set at + predetermined percentiles of the distribution of estimates in + the calibration group (15th and 85th percentiles in this + study). This procedure permits the transformation of anchored + estimates into percentiles in the calibration group, allows + pooling of results across problems, and provides a natural + measure of the size of the effect. The authors illustrate the + method by a demonstration that the initial judgment of the + anchor is susceptible to an anchoring-like bias and by an + analysis of the relation between anchoring and subjective + confidence. +
+ +
+ + +
+[663] +
+
+Warren G. Jackson, Ender Özcan, and Robert I. John. + Move acceptance in local search metaheuristics for cross-domain search. + Expert Systems with Applications, 109:131–151, 2018.
+[ bib ] + +
+ + +
+[664] +
+
+Daniel M Jaeggi, Geoffrey T Parks, Timoleon Kipouros, and P John Clarkson. + The development of a multi-objective Tabu Search algorithm for continuous optimisation problems. + European Journal of Operational Research, 185(3):1192–1212, 2008.
+[ bib ] + +
+ + +
+[665] +
+
+Satish Jajodia, Ioannis Minis, George Harhalakis, and Jean-Marie Proth. + CLASS: computerized layout solutions using simulated annealing. + International Journal of Production Research, 30(1):95–108, 1992.
+[ bib ] + +
+ + +
+[666] +
+
+Andrzej Jaszkiewicz. + Genetic local search for multi-objective combinatorial optimization. + European Journal of Operational Research, 137(1):50–71, 2002.
+[ bib ] + +
+ + +
+[667] +
+
+Andrzej Jaszkiewicz. + Many-Objective Pareto Local Search. + European Journal of Operational Research, 271(3):1001–1013, 2018.
+[ bib | +DOI ] + +
+ + +
+[668] +
+
+Andrzej Jaszkiewicz and Thibaut Lust. + ND-tree-based update: a fast algorithm for the dynamic nondominance problem. + IEEE Transactions on Evolutionary Computation, 22(5):778–791, 2018.
+[ bib ] + +
+ + +
+[669] +
+
+Andrzej Jaszkiewicz. + On the performance of multiple-objective genetic local search on the 0/1 knapsack problem – A comparative experiment. + IEEE Transactions on Evolutionary Computation, 6(4):402–412, 2002.
+[ bib ] + +
+ + +
+[670] +
+
+M. T. Jensen. + Reducing the run-time complexity of multiobjective EAs: The NSGA-II and other algorithms. + IEEE Transactions on Evolutionary Computation, 7(5):503–515, 2003.
+[ bib ] + +
+ + +
+[671] +
+
+M. T. Jensen. + Helper-Objectives: Using Multi-Objective Evolutionary Algorithms for Single-Objective Optimisation. + Journal of Mathematical Modelling and Algorithms, 3(4):323–347, 2004.
+[ bib ] +
+Keywords: multi-objectivization +
+ +
+ + +
+[672] +
+
+Mark Jerrum and Gregory Sorkin. + The Metropolis algorithm for graph bisection. + Discrete Applied Mathematics, 82(1):155–175, 1998.
+[ bib ] + +
+ + +
+[673] +
+
+Mark Jerrum. + Large cliques elude the Metropolis process. + Random Structures & Algorithms, 3(4):347–359, 1992.
+[ bib ] + +
+ + +
+[674] +
+
+S. Jiang, Y. S. Ong, J. Zhang, and L. Feng. + Consistencies and Contradictions of Performance Metrics in Multiobjective Optimization. + IEEE Transactions on Cybernetics, 44(12):2391–2404, 2014.
+[ bib ] + +
+ + +
+[675] +
+
+Shouyong Jiang, Juan Zou, Shengxiang Yang, and Xin Yao. + Evolutionary Dynamic Multi-Objective Optimisation: A Survey. + ACM Computing Surveys, 55(4), November 2022.
+[ bib | +DOI ] +
+Keywords: evolutionary algorithm, evolutionary dynamic multi-objective + optimisation, dynamic environment, Multi-objective + optimisation +
+ +
+ + +
+[676] +
+
+Yaochu Jin. + A Comprehensive Survey of Fitness Approximation in Evolutionary Computation. + Soft Computing, 9(1):3–12, 2005.
+[ bib ] + +
+ + +
+[677] +
+
+Yaochu Jin. + Surrogate-Assisted Evolutionary Computation: Recent Advances and Future Challenges. + Swarm and Evolutionary Computation, 1(2):61–70, June 2011.
+[ bib | +DOI ] +
+Surrogate-assisted, or meta-model based evolutionary + computation uses efficient computational models, often known + as surrogates or meta-models, for approximating the fitness + function in evolutionary algorithms. Research on + surrogate-assisted evolutionary computation began over a + decade ago and has received considerably increasing interest + in recent years. Very interestingly, surrogate-assisted + evolutionary computation has found successful applications + not only in solving computationally expensive single- or + multi-objective optimization problems, but also in addressing + dynamic optimization problems, constrained optimization + problems and multi-modal optimization problems. This paper + provides a concise overview of the history and recent + developments in surrogate-assisted evolutionary computation + and suggests a few future trends in this research area. +
+
+Keywords: Evolutionary computation,Expensive optimization + problems,Machine learning,Meta-models,Model + management,Surrogates +
+ +
+ + +
+[678] +
+
+Yaochu Jin, Handing Wang, Tinkle Chugh, Dan Guo, and Kaisa Miettinen. + Data-Driven Evolutionary Optimization: An Overview and Case Studies. + IEEE Transactions on Evolutionary Computation, 23(3):442–458, June 2019.
+[ bib | +DOI ] + +
+ + +
+[679] +
+
+Huidong Jin and Man-Leung Wong. + Adaptive, convergent, and diversified archiving strategy for multiobjective evolutionary algorithms. + Expert Systems with Applications, 37(12):8462–8470, 2010.
+[ bib ] + +
+ + +
+[680] +
+
+David S. Johnson. + Optimal Two- and Three-stage Production Scheduling with Setup Times Included. + Naval Research Logistics Quarterly, 1:61–68, 1954.
+[ bib ] + +
+ + +
+[681] +
+
+David S. Johnson, Cecilia R. Aragon, Lyle A. McGeoch, and Catherine Schevon. + Optimization by Simulated Annealing: An Experimental Evaluation: Part I, Graph Partitioning. + Operations Research, 37(6):865–892, 1989.
+[ bib | +DOI ] + +
+ + +
+[682] +
+
+David S. Johnson, Cecilia R. Aragon, Lyle A. McGeoch, and Catherine Schevon. + Optimization by Simulated Annealing: An Experimental Evaluation: Part II, Graph Coloring and Number Partitioning. + Operations Research, 39(3):378–406, 1991.
+[ bib ] + +
+ + +
+[683] +
+
+Alan W. Johnson and Sheldon H. Jacobson. + On the Convergence of Generalized Hill Climbing Algorithms. + Discrete Applied Mathematics, 119(1):37–57, 2002.
+[ bib ] + +
+ + +
+[684] +
+
+Mark E. Johnson, Leslie M. Moore, and Donald Ylvisaker. + Minimax and maximin distance designs. + Journal of Statistical Planning and Inference, 26(2):131–148, 1990.
+[ bib ] +
+Keywords: Bayesian design +
+ +
+ + +
+[685] +
+
+David S. Johnson, Christos H. Papadimitriou, and Mihalis Yannakakis. + How Easy is Local Search? + Journal of Computer System Science, 37(1):79–100, 1988.
+[ bib ] + +
+ + +
+[686] +
+
+C. Joncour, J. Kritter, S. Michel, and X. Schepler. + Generalized Relax-and-Fix Heuristic. + Computers & Operations Research, 149:106038, 2023.
+[ bib ] + +
+ + +
+[687] +
+
+Donald R. Jones, Matthias Schonlau, and William J. Welch. + Efficient Global Optimization of Expensive Black-Box Functions. + Journal of Global Optimization, 13(4):455–492, 1998.
+[ bib ] +
+Proposed EGO algorithm +
+
+Keywords: EGO +
+ +
+ + +
+[688] +
+
+Kenneth A. De Jong and William M. Spears. + A formal analysis of the role of multi-point crossover in genetic algorithms. + Annals of Mathematics and Artificial Intelligence, 5(1):1–26, 1992.
+[ bib ] + +
+ + +
+[689] +
+
+Jorik Jooken, Pieter Leyman, and Patrick De Causmaecker. + A new class of hard problem instances for the 0–1 knapsack problem. + European Journal of Operational Research, 301(3):841–854, 2022.
+[ bib ] + +
+ + +
+[690] +
+
+Jorik Jooken, Pieter Leyman, Tony Wauters, and Patrick De Causmaecker. + Exploring search space trees using an adapted version of Monte Carlo tree search for combinatorial optimization problems. + Computers & Operations Research, 150:106070, 2023.
+[ bib | +DOI ] + +
+ + +
+[691] +
+
+D. E. Joslin and D. P. Clements. + Squeaky Wheel Optimization. + Journal of Artificial Intelligence Research, 10:353–373, 1999.
+[ bib ] + +
+ + +
+[692] +
+
+P. W. Jowitt and G. Germanopoulos. + Optimal pump scheduling in water supply networks. + Journal of Water Resources Planning and Management, ASCE, 118(4):406–422, 1992.
+[ bib ] +
+The electricity cost of pumping accounts for a large + part of the total operating cost for water-supply + networks. This study presents a method based on + linear programming for determining an optimal + (minimum cost) schedule of pumping on a 24-hr + basis. Both unit and maximum demand electricity + charges are considered. Account is taken of the + relative efficiencies of the available pumps, the + structure of the electricity tariff, the + consumer-demand profile, and the hydraulic + characteristics and operational constraints of the + network. The use of extended-period simulation of + the network operation in determining the parameters + of the linearized network equations and constraints + and in studying the optimized network operation is + described. An application of the method to an + existing network in the United Kingdom is presented, + showing that considerable savings are possible. The + method was found to be robust and with low + computation-time requirements, and is therefore + suitable for real-time implementation. +
+ +
+ + +
+[693] +
+
+Angel A. Juan, Javier Faulin, Scott E. Grasman, Markus Rabe, and Gonçalo Figueira. + A review of simheuristics: Extending metaheuristics to deal with stochastic combinatorial optimization problems. + Operations Research Perspectives, 2:62–72, 2015.
+[ bib | +DOI ] +
+Keywords: Metaheuristics; Simulation; Combinatorial optimization; + Stochastic problems +
+ +
+ + +
+[694] +
+
+Angel A. Juan, Helena R. Lourenço, Manuel Mateo, Rachel Luo, and Quim Castellà. + Using Iterated Local Search for Solving the Flow-shop Problem: Parallelization, Parametrization, and Randomization Issues. + International Transactions in Operational Research, 21(1):103–126, 2014.
+[ bib ] + +
+ + +
+[695] +
+
+M. Jünger, Gerhard Reinelt, and S. Thienel. + Provably Good Solutions for the Traveling Salesman Problem. + Zeitschrift für Operations Research, 40(2):183–217, 1994.
+[ bib ] + +
+ + +
+[696] +
+
+Elena A. Kabova, Jason C. Cole, Oliver Korb, Manuel López-Ibáñez, Adrian C. Williams, and Kenneth Shankland. + Improved performance of crystal structure solution from powder diffraction data through parameter tuning of a simulated annealing algorithm. + Journal of Applied Crystallography, 50(5):1411–1420, October 2017.
+[ bib | +DOI ] +
+Significant gains in the performance of the simulated + annealing algorithm in the DASH software package have + been realized by using the irace automatic + configuration tool to optimize the values of three key + simulated annealing parameters. Specifically, the success + rate in finding the global minimum in intensity χ2 + space is improved by up to an order of magnitude. The general + applicability of these revised simulated annealing parameters + is demonstrated using the crystal structure determinations of + over 100 powder diffraction datasets. +
+
+Keywords: crystal structure determination, powder diffraction, + simulated annealing, parameter tuning, irace +
+ +
+ + +
+[697] +
+
+Daniel Kahneman and Amos Tversky. + Prospect theory: An analysis of decision under risk. + Econometrica, 47(2):263–291, 1979.
+[ bib | +DOI ] + +
+ + +
+[698] +
+
+Daniel Kahneman. + Maps of bounded rationality: Psychology for behavioral economics. + The American Economic Review, 93(5):1449–1475, 2003.
+[ bib ] + +
+ + +
+[699] +
+
+Jakob Kallestad, Ramin Hasibi, Ahmad Hemmati, and Kenneth Sörensen. + A general deep reinforcement learning hyperheuristic framework for solving combinatorial optimization problems. + European Journal of Operational Research, 309(1):446–468, August 2023.
+[ bib | +DOI ] +
+Keywords: Deep RL, hyper-heuristic, ALNS +
+ +
+ + +
+[700] +
+
+Qinma Kang, Hong He, and Jun Wei. + An Effective Iterated Greedy Algorithm for Reliability-oriented Task Allocation in Distributed Computing Systems. + Journal of Parallel and Distributed Computing, 73(8):1106–1115, 2013.
+[ bib ] + +
+ + +
+[701] +
+
+Korhan Karabulut. + A hybrid iterated greedy algorithm for total tardiness minimization in permutation flowshops. + Computers and Industrial Engineering, 98(Supplement C):300 – 307, 2016.
+[ bib ] + +
+ + +
+[702] +
+
+Dervis Karaboga and Bahriye Akay. + A Survey: Algorithms Simulating Bee Swarm Intelligence. + Artificial Intelligence Review, 31(1–4):61–85, 2009.
+[ bib ] + +
+ + +
+[703] +
+
+Giorgos Karafotias, Mark Hoogendoorn, and Agoston E. Eiben. + Parameter Control in Evolutionary Algorithms: Trends and Challenges. + IEEE Transactions on Evolutionary Computation, 19(2):167–187, 2015.
+[ bib ] + +
+ + +
+[704] +
+
+İbrahim Karahan and Murat Köksalan. + A territory defining multiobjective evolutionary algorithms and preference incorporation. + IEEE Transactions on Evolutionary Computation, 14(4):636–664, 2010.
+[ bib | +DOI ] +
+Keywords: TDEA +
+ +
+ + +
+[705] +
+
+Maryam Karimi-Mamaghan, Mehrdad Mohammadi, Patrick Meyer, Amir Mohammad Karimi-Mamaghan, and El-Ghazali Talbi. + Machine learning at the service of meta-heuristics for solving combinatorial optimization problems: A state-of-the-art. + European Journal of Operational Research, 296(2):393–422, 2022.
+[ bib | +DOI ] +
+In recent years, there has been a growing research interest + in integrating machine learning techniques into + meta-heuristics for solving combinatorial optimization + problems. This integration aims to lead meta-heuristics + toward an efficient, effective, and robust search and improve + their performance in terms of solution quality, convergence + rate, and robustness. Since various integration methods with + different purposes have been developed, there is a need to + review the recent advances in using machine learning + techniques to improve meta-heuristics. To the best of our + knowledge, the literature is deprived of having a + comprehensive yet technical review. To fill this gap, this + paper provides such a review on the use of machine learning + techniques in the design of different elements of + meta-heuristics for different purposes including algorithm + selection, fitness evaluation, initialization, evolution, + parameter setting, and cooperation. First, we describe the + key concepts and preliminaries of each of these ways of + integration. Then, the recent advances in each way of + integration are reviewed and classified based on a proposed + unified taxonomy. Finally, we provide a technical discussion + on the advantages, limitations, requirements, and challenges + of implementing each of these integration ways, followed by + promising future research directions. +
+
+Keywords: Meta-heuristics, Machine learning, Combinatorial optimization + problems, State-of-the-art +
+ +
+ + +
+[706] +
+
+Oleksiy Karpenko, Jianming Shi, and Yang Dai. + Prediction of MHC class II binders using the ant colony search strategy. + Artificial Intelligence in Medicine, 35(1):147–156, 2005.
+[ bib ] + +
+ + +
+[707] +
+
+Korhan Karabulut and Fatih M. Tasgetiren. + A Variable Iterated Greedy Algorithm for the Traveling Salesman Problem with Time Windows. + Information Sciences, 279:383–395, 2014.
+[ bib ] + +
+ + +
+[708] +
+
+Joseph R. Kasprzyk, Shanthi Nataraj, Patrick M. Reed, and Robert J. Lempert. + Many objective robust decision making for complex environmental systems undergoing change. + Environmental Modelling & Software, 42:55–71, 2013.
+[ bib ] +
+Keywords: scenario-based +
+ +
+ + +
+[709] +
+
+Joseph R. Kasprzyk, Patrick M. Reed, Gregory W. Characklis, and Brian R. Kirsch. + Many-objective de Novo water supply portfolio planning under deep uncertainty. + Environmental Modelling & Software, 34:87–104, 2012.
+[ bib ] +
+Keywords: scenario-based +
+ +
+ + +
+[710] +
+
+Artem Kaznatcheev, David A. Cohen, and Peter Jeavons. + Representing Fitness Landscapes by Valued Constraints to Understand the Complexity of Local Search. + Journal of Artificial Intelligence Research, 69:1077–1102, 2020.
+[ bib | +DOI ] + +
+ + +
+[711] +
+
+Liangjun Ke, Claudia Archetti, and Zuren Feng. + Ants can solve the team orienteering problem. + Computers and Industrial Engineering, 54(3):648–665, 2008.
+[ bib | +DOI ] +
+The team orienteering problem (TOP) involves + finding a set of paths from the starting point to + the ending point such that the total collected + reward received from visiting a subset of locations + is maximized and the length of each path is + restricted by a pre-specified limit. In this paper, + an ant colony optimization (ACO) approach is + proposed for the team orienteering problem. Four + methods, i.e., the sequential, + deterministic-concurrent and random-concurrent and + simultaneous methods, are proposed to construct + candidate solutions in the framework of ACO. We + compare these methods according to the results + obtained on well-known problems from the + literature. Finally, we compare the algorithm with + several existing algorithms. The results show that + our algorithm is promising. +
+
+Keywords: Ant colony optimization, Ant system, Heuristics, + Team orienteering problem +
+ +
+ + +
+[712] +
+
+R. L. Keeney. + Analysis of preference dependencies among objectives. + Operations Research, 29:1105–1120, 1981.
+[ bib ] + +
+ + +
+[713] +
+
+Graham Kendall, Ruibin Bai, Jacek Blazewicz, Patrick De Causmaecker, Michel Gendreau, Robert John, Jiawei Li, Barry McCollum, Erwin Pesch, Rong Qu, Nasser Sabar, Greet Vanden Berghe, and Angelina Yee. + Good Laboratory Practice for Optimization Research. + Journal of the Operational Research Society, 67(4):676–689, 2016.
+[ bib | +DOI ] + +
+ + +
+[714] +
+
+Pascal Kerschke, Holger H. Hoos, Frank Neumann, and Heike Trautmann. + Automated Algorithm Selection: Survey and Perspectives. + Evolutionary Computation, 27(1):3–45, March 2019.
+[ bib | +DOI ] + +
+ + +
+[715] +
+
+B. W. Kernighan and S. Lin. + An Efficient Heuristic Procedure for Partitioning Graphs. + Bell Systems Technology Journal, 49(2):213–219, 1970.
+[ bib ] + +
+ + +
+[716] +
+
+Pascal Kerschke and Heike Trautmann. + Automated Algorithm Selection on Continuous Black-Box Problems by Combining Exploratory Landscape Analysis and Machine Learning. + Evolutionary Computation, 27(1):99–127, 2019.
+[ bib | +DOI ] +
+In this article, we build upon previous work on designing + informative and efficient Exploratory Landscape Analysis + features for characterizing problems' landscapes and show + their effectiveness in automatically constructing algorithm + selection models in continuous black-box optimization + problems. Focusing on algorithm performance results of the + COCO platform of several years, we construct a representative + set of high-performing complementary solvers and present an + algorithm selection model that, compared to the portfolio's + single best solver, on average requires less than half of the + resources for solving a given problem. Therefore, there is a + huge gain in efficiency compared to classical ensemble + methods combined with an increased insight into problem + characteristics and algorithm properties by using informative + features. The model acts on the assumption that the function + set of the Black-Box Optimization Benchmark is representative + enough for practical applications. The model allows for + selecting the best suited optimization algorithm within the + considered set for unseen problems prior to the optimization + itself based on a small sample of function evaluations. Note + that such a sample can even be reused for the initial + population of an evolutionary (optimization) algorithm so + that even the feature costs become negligible. +
+ +
+ + +
+[717] +
+
+Pascal Kerschke, Hao Wang, Mike Preuss, Christian Grimme, André H. Deutz, Heike Trautmann, and Michael T. M. Emmerich. + Search Dynamics on Multimodal Multiobjective Problems. + Evolutionary Computation, 27(4):577–609, 2019.
+[ bib | +DOI ] + +
+ + +
+[718] +
+
+Norbert L. Kerr. + HARKing: Hypothesizing After the Results are Known. + Personality and Social Psychology Review, 2(3):196–217, August 1998.
+[ bib | +DOI ] + +
+ + +
+[719] +
+
+A. R. KhudaBukhsh, Lin Xu, Holger H. Hoos, and Kevin Leyton-Brown. + SATenstein: Automatically Building Local Search SAT Solvers from Components. + Artificial Intelligence, 232:20–42, 2016.
+[ bib | +DOI ] + +
+ + +
+[720] +
+
+Philip Kilby and Tommaso Urli. + Fleet design optimisation from historical data using constraint programming and large neighbourhood search. + Constraints, pp.  1–20, 2015.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[721] +
+
+Yeong-Dae Kim. + Heuristics for Flowshop Scheduling Problems Minimizing Mean Tardiness. + Journal of the Operational Research Society, 44(1):19–28, 1993.
+[ bib | +DOI ] + +
+ + +
+[722] +
+
+Youngmin Kim, Richard Allmendinger, and Manuel López-Ibáñez. + Safe Learning and Optimization Techniques: Towards a Survey of the State of the Art. + Arxiv preprint arXiv:2101.09505 [cs.LG], 2020.
+[ bib | +http ] +
+Safe learning and optimization deals with learning and + optimization problems that avoid, as much as possible, the + evaluation of non-safe input points, which are solutions, + policies, or strategies that cause an irrecoverable loss + (e.g., breakage of a machine or equipment, or life + threat). Although a comprehensive survey of safe + reinforcement learning algorithms was published in 2015, a + number of new algorithms have been proposed thereafter, and + related works in active learning and in optimization were not + considered. This paper reviews those algorithms from a number + of domains including reinforcement learning, Gaussian process + regression and classification, evolutionary algorithms, and + active learning. We provide the fundamental concepts on which + the reviewed algorithms are based and a characterization of + the individual algorithms. We conclude by explaining how the + algorithms are connected and suggestions for future + research. +
+ +
+ + +
+[723] +
+
+Jungtaek Kim, Michael McCourt, Tackgeun You, Saehoon Kim, and Seungjin Choi. + Bayesian Optimization with Approximate Set Kernels. + Machine Learning, 2021.
+[ bib | +DOI ] +
+We propose a practical Bayesian optimization method over + sets, to minimize a black-box function that takes a set as a + single input. Because set inputs are permutation-invariant, + traditional Gaussian process-based Bayesian optimization + strategies which assume vector inputs can fall short. To + address this, we develop a Bayesian optimization method with + set kernel that is used to build surrogate + functions. This kernel accumulates similarity over set + elements to enforce permutation-invariance, but this comes at + a greater computational cost. To reduce this burden, we + propose two key components: (i) a more efficient approximate + set kernel which is still positive-definite and is an + unbiased estimator of the true set kernel with upper-bounded + variance in terms of the number of subsamples, (ii) a + constrained acquisition function optimization over sets, + which uses symmetry of the feasible region that defines a set + input. Finally, we present several numerical experiments + which demonstrate that our method outperforms other methods. +
+ +
+ + +
+[724] +
+
+J.-S. Kim, J.-H. Park, and D.-H. Lee. + Iterated Greedy Algorithms to Minimize the Total Family Flow Time for Job-shop Scheduling with Job Families and Sequence-dependent Set-ups. + Engineering Optimization, 49(10):1719–1732, 2017.
+[ bib ] + +
+ + +
+[725] +
+
+Diederik P. Kingma and Jimmy Ba. + Adam: A method for stochastic optimization. + Arxiv preprint arXiv:1412.6980 [cs.LG], 2014.
+[ bib | +http ] +
+Published as a conference paper at the 3rd International + Conference for Learning Representations, San Diego, 2015 [2095] +
+ +
+ + +
+[726] +
+
+Scott Kirkpatrick and G. Toulouse. + Configuration Space Analysis of Travelling Salesman Problems. + Journal de Physique, 46(8):1277–1292, 1985.
+[ bib ] + +
+ + +
+[727] +
+
+Scott Kirkpatrick. + Optimization by Simulated Annealing: Quantitative Studies. + Journal of Statistical Physics, 34(5-6):975–986, 1984.
+[ bib ] + +
+ + +
+[728] +
+
+Scott Kirkpatrick, C. D. Gelatt, and M. P. Vecchi. + Optimization by Simulated Annealing. + Science, 220(4598):671–680, 1983.
+[ bib | +DOI ] +
+Proposed Simulated Annealing +
+ +
+ + +
+[729] +
+
+Kathrin Klamroth, Sanaz Mostaghim, Boris Naujoks, Silvia Poles, Robin C. Purshouse, Günther Rudolph, Stefan Ruzika, Serpil Sayın, Margaret M. Wiecek, and Xin Yao. + Multiobjective optimization for interwoven systems. + Journal of Multi-Criteria Decision Analysis, 24(1-2):71–81, 2017.
+[ bib | +DOI ] + +
+ + +
+[730] +
+
+Anton J. Kleywegt, Alexander Shapiro, and Tito Homem-de-Mello. + The Sample Average Approximation Method for Stochastic Discrete Optimization. + SIAM Journal on Optimization, 12(2):479–502, 2002.
+[ bib ] + +
+ + +
+[731] +
+
+Joshua D. Knowles. + ParEGO: A hybrid algorithm with on-line landscape approximation for expensive multiobjective optimization problems. + IEEE Transactions on Evolutionary Computation, 10(1):50–66, 2006.
+[ bib | +DOI ] +
+Keywords: ParEGO, online, metamodel +
+ +
+ + +
+[732] +
+
+Joshua D. Knowles. + Closed-loop evolutionary multiobjective optimization. + IEEE Computational Intelligence Magazine, 4:77–91, 2009.
+[ bib | +DOI ] +
+Artificial evolution has been used for more than 50 years as a method of optimization in engineering, operations research and computational intelligence. In closed-loop evolution (a term used by the statistician, George Box) or, equivalently, evolutionary experimentation (Ingo Rechenberg's terminology), the “phenotypes” are evaluated in the real world by conducting a physical experiment, whilst selection and breeding is simulated. Well-known early work on artificial evolution — design engineering problems in fluid dynamics, and chemical plant process optimization — was carried out in this experimental mode. More recently, the closed-loop approach has been successfully used in much evolvable hardware and evolutionary robotics research, and in some microbiology and biochemistry applications. In this article, several further new targets for closed-loop evolutionary and multiobjective optimization are considered. Four case studies from my own collaborative work are described: (i) instrument optimization in analytical biochemistry; (ii) finding effective drug combinations in vitro; (iii) onchip synthetic biomolecule design; and (iv) improving chocolate production processes. Accurate simulation in these applications is not possible due to complexity or a lack of adequate analytical models. In these and other applications discussed, optimizing experimentally brings with it several challenges: noise; nuisance factors; ephemeral resource constraints; expensive evaluations, and evaluations that must be done in (large) batches. Evolutionary algorithms (EAs) are largely equal to these vagaries, whilst modern multiobjective EAs also enable tradeoffs among conflicting optimization goals to be explored. Nevertheless, principles from other disciplines, such as statistics, Design of Experiments, machine learning and global optimization are also relevant to aspects of the closed-loop problem, and may inspire futher development of multiobjective EAs. +
+ +
+ + +
+[733] +
+
+Joshua D. Knowles and David Corne. + Approximating the Nondominated Front Using the Pareto Archived Evolution Strategy. + Evolutionary Computation, 8(2):149–172, 2000.
+[ bib | +DOI ] +
+Proposed PAES +
+ +
+ + +
+[734] +
+
+Joshua D. Knowles and David Corne. + Properties of an Adaptive Archiving Algorithm for Storing Nondominated Vectors. + IEEE Transactions on Evolutionary Computation, 7(2):100–116, April 2003.
+[ bib ] +
+Proposed to use S-metric (hypervolume metric) for + environmental selection +
+
+Keywords: S-metric, hypervolume +
+ +
+ + +
+[735] +
+
+Mirjam J. Knol, Tyler J. VanderWeele, Rolf H. H. Groenwold, Olaf H. Klungel, Maroeska M. Rovers, and Diederick E. Grobbee. + Estimating measures of interaction on an additive scale for preventive exposures. + European Journal of Epidemiology, 26(6):433–438, 2011.
+[ bib ] + +
+ + +
+[736] +
+
+Gary A. Kochenberger, Fred Glover, Bahram Alidaee, and Cesar Rego. + A unified modeling and solution framework for combinatorial optimization problems. + OR Spektrum, 26(2):237–250, 2004.
+[ bib ] + +
+ + +
+[737] +
+
+Gary A. Kochenberger, Jin-Kao Hao, Fred Glover, Mark Lewis, Zhipeng Lü, Haibo Wang, and Yang Wang. + The unconstrained binary quadratic programming problem: a survey. + Journal of Combinatorial Optimization, 28(1):58–81, 2014.
+[ bib | +DOI ] + +
+ + +
+[738] +
+
+Murat Köksalan. + Multiobjective Combinatorial Optimization: Some Approaches. + Journal of Multi-Criteria Decision Analysis, 15:69–78, 2009.
+[ bib | +DOI ] + +
+ + +
+[739] +
+
+Murat Köksalan and İbrahim Karahan. + An Interactive Territory Defining Evolutionary Algorithm: iTDEA. + IEEE Transactions on Evolutionary Computation, 14(5):702–722, October 2010.
+[ bib | +DOI ] + +
+ + +
+[740] +
+
+Rainer Kolisch and Sönke Hartmann. + Experimental investigation of heuristics for resource-constrained project scheduling: An update. + European Journal of Operational Research, 174(1):23–37, October 2006.
+[ bib | +DOI ] +
+This paper considers heuristics for the well-known + resource-constrained project scheduling problem + (RCPSP). It provides an update of our survey which + was published in 2000. We summarize and categorize a + large number of heuristics that have recently been + proposed in the literature. Most of these heuristics + are then evaluated in a computational study and + compared on the basis of our standardized + experimental design. Based on the computational + results we discuss features of good heuristics. The + paper closes with some remarks on our test design + and a summary of the recent developments in research + on heuristics for the RCPSP. +
+
+Keywords: Computational evaluation, Heuristics, Project + scheduling, Resource constraints +
+ +
+ + +
+[741] +
+
+Vladlen Koltun and Christos H. Papadimitriou. + Approximately dominating representatives. + Theoretical Computer Science, 371(3):148–154, 2007.
+[ bib ] + +
+ + +
+[742] +
+
+A. Kolen and Erwin Pesch. + Genetic Local Search in Combinatorial Optimization. + Discrete Applied Mathematics, 48(3):273–284, 1994.
+[ bib ] + +
+ + +
+[743] +
+
+Joshua B. Kollat and Patrick M. Reed. + A framework for visually interactive decision-making and design using evolutionary multi-objective optimization (VIDEO). + Environmental Modelling & Software, 22(12):1691–1704, 2007.
+[ bib ] +
+Keywords: glyph plot +
+ +
+ + +
+[744] +
+
+Tjalling C. Koopmans and Martin J. Beckmann. + Assignment Problems and the Location of Economic Activities. + Econometrica, 25:53–76, 1957.
+[ bib ] +
+Introduced the Quadratic Assignment Problem (QAP) +
+ +
+ + +
+[745] +
+
+Jsh Kornbluth. + Sequential multi-criterion decision making. + Omega, 13(6):569–574, 1985.
+[ bib | +DOI ] +
+In this paper we consider a simple sequential + multicriterion decision making problem in which a + decision maker has to accept or reject a series of + multi-attributed outcomes. We show that using very + simple programming techniques, a great deal of the + decision making can be automated. The method might + be applicable to situations in which a dealer is + having to consider sequential offers in a trading + market. +
+
+Keywords: machine decision making +
+ +
+ + +
+[746] +
+
+Pekka Korhonen, Herbert Moskowitz, and Jyrki Wallenius. + Choice Behavior in Interactive Multiple-Criteria Decision Making. + Annals of Operations Research, 23(1):161–179, December 1990.
+[ bib | +DOI ] +
+Choice behavior in an interactive multiple-criteria decision + making environment is examined experimentally. A “free + search” discrete visual interactive reference direction + approach was used on a microcomputer by management students + to solve two realistic and relevant multiple-criteria + decision problems. The results revealed persistent patterns + of intransitive choice behavior, and an unexpectedly rapid + degree of convergence of the reference direction approach on + a preferred solution. The results can be explained using + Tversky' additive utility difference model and + Kahneman-Tversky's prospect theory. The implications of the + results for the design of interactive multiple-criteria + decision procedures are discussed. +
+ +
+ + +
+[747] +
+
+Flip Korn, B.-U. Pagel, and Christos Faloutsos. + On the “dimensionality curse” and the “self-similarity blessing”. + IEEE Transactions on Knowledge and Data Engineering, 13(1):96–111, 2001.
+[ bib | +DOI ] +
+Spatial queries in high-dimensional spaces have been studied + extensively. Among them, nearest neighbor queries are + important in many settings, including spatial databases (Find + the k closest cities) and multimedia databases (Find the k + most similar images). Previous analyses have concluded that + nearest-neighbor search is hopeless in high dimensions due to + the notorious "curse of dimensionality". We show that this + may be overpessimistic. We show that what determines the + search performance (at least for R-tree-like structures) is + the intrinsic dimensionality of the data set and not the + dimensionality of the address space (referred to as the + embedding dimensionality). The typical (and often implicit) + assumption in many previous studies is that the data is + uniformly distributed, with independence between + attributes. However, real data sets overwhelmingly disobey + these assumptions; rather, they typically are skewed and + exhibit intrinsic ("fractal") dimensionalities that are much + lower than their embedding dimension, e.g. due to subtle + dependencies between attributes. We show how the Hausdorff + and Correlation fractal dimensions of a data set can yield + extremely accurate formulas that can predict the I/O + performance to within one standard deviation on multiple real + and synthetic data sets. +
+ +
+ + +
+[748] +
+
+P. Korošec, Jurij Šilc, and B. Robič. + Solving the mesh-partitioning problem with an ant-colony algorithm. + Parallel Computing, 30:785–801, 2004.
+[ bib ] + +
+ + +
+[749] +
+
+Pekka Korhonen, Kari Silvennoinen, Jyrki Wallenius, and Anssi Öörni. + Can a linear value function explain choices? An experimental study. + European Journal of Operational Research, 219(2):360–367, June 2012.
+[ bib | +DOI ] +
+We investigate in a simple bi-criteria experimental study, + whether subjects are consistent with a linear value function + while making binary choices. Many inconsistencies appeared in + our experiment. However, the impact of inconsistencies on the + linearity vs. non-linearity of the value function was + minor. Moreover, a linear value function seems to predict + choices for bi-criteria problems quite well. This ability to + predict is independent of whether the value function is + diagnosed linear or not. Inconsistencies in responses did not + necessarily change the original diagnosis of the form of the + value function. Our findings have implications for the design + and development of decision support tools for Multiple + Criteria Decision Making problems. +
+
+Keywords: Binary choices, Inconsistency, Linear value function, + Multiple criteria, Weights +
+ +
+ + +
+[750] +
+
+Oliver Korb, Thomas Stützle, and Thomas E. Exner. + An Ant Colony Optimization Approach to Flexible Protein–Ligand Docking. + Swarm Intelligence, 1(2):115–134, 2007.
+[ bib ] + +
+ + +
+[751] +
+
+Oliver Korb, Thomas Stützle, and Thomas E. Exner. + Empirical Scoring Functions for Advanced Protein-Ligand Docking with PLANTS. + Journal of Chemical Information and Modeling, 49(2):84–96, 2009.
+[ bib ] + +
+ + +
+[752] +
+
+Oliver Korb, Peter Monecke, Gerhard Hessler, Thomas Stützle, and Thomas E. Exner. + pharmACOphore: Multiple Flexible Ligand Alignment Based on Ant Colony Optimization. + Journal of Chemical Information and Modeling, 50(9):1669–1681, 2010.
+[ bib ] + +
+ + +
+[753] +
+
+Pekka Korhonen and Jyrki Wallenius. + A pareto race. + Naval Research Logistics, 35(6):615–623, 1988.
+[ bib | +DOI ] +
+A dynamic and visual “free-search” type of interactive + procedure for multiple-objective linear programming is + presented. The method enables a decision maker to freely + search any part of the efficient frontier by controlling the + speed and direction of motion. The objective function values + are represented in numeric form and as bar graphs on a + display. The method is implemented on an IBM PC/1 + microcomputer and is illustrated using a multiple-objective + linear-programming model for managing disposal of sewage + sludge in the New York Bight. Some other applications are + also briefly discussed. +
+ +
+ + +
+[754] +
+
+Lars Kotthoff. + Algorithm Selection for Combinatorial Search Problems: A Survey. + AI Magazine, 35(3):48–60, 2014.
+[ bib ] + +
+ + +
+[755] +
+
+Timo Kötzing, Frank Neumann, Heiko Röglin, and Carsten Witt. + Theoretical Analysis of Two ACO Approaches for the Traveling Salesman Problem. + Swarm Intelligence, 6(1):1–21, 2012.
+[ bib | +DOI ] +
+Bioinspired algorithms, such as evolutionary algorithms and + ant colony optimization, are widely used for different + combinatorial optimization problems. These algorithms rely + heavily on the use of randomness and are hard to understand + from a theoretical point of view. This paper contributes to + the theoretical analysis of ant colony optimization and + studies this type of algorithm on one of the most prominent + combinatorial optimization problems, namely the traveling + salesperson problem (TSP). We present a new construction + graph and show that it has a stronger local property than one + commonly used for constructing solutions of the TSP. The + rigorous runtime analysis for two ant colony optimization + algorithms, based on these two construction procedures, shows + that they lead to good approximation in expected polynomial + time on random instances. Furthermore, we point out in which + situations our algorithms get trapped in local optima and + show where the use of the right amount of heuristic + information is provably beneficial. +
+ +
+ + +
+[756] +
+
+Lars Kotthoff, Chris Thornton, Holger H. Hoos, Frank Hutter, and Kevin Leyton-Brown. + Auto-WEKA 2.0: Automatic model selection and hyperparameter optimization in WEKA. + Journal of Machine Learning Research, 17:1–5, 2016.
+[ bib ] + +
+ + +
+[757] +
+
+Katharina Kowalski, Sigrid Stagl, Reinhard Madlener, and Ines Omann. + Sustainable energy futures: Methodological challenges in combining scenarios and participatory multi-criteria analysis. + European Journal of Operational Research, 197(3):1063–1074, 2009.
+[ bib ] + +
+ + +
+[758] +
+
+Oliver Kramer. + Iterated Local Search with Powell's Method: A Memetic Algorithm for Continuous Global Optimization. + Memetic Computing, 2(1):69–83, 2010.
+[ bib | +DOI ] + +
+ + +
+[759] +
+
+Daniel Krajzewicz, Jakob Erdmann, Michael Behrisch, and Laura Bieker. + Recent development and applications of SUMO - Simulation of Urban MObility. + International Journal On Advances in Systems and Measurements, 5(3-4):128–138, 2012.
+[ bib ] + +
+ + +
+[760] +
+
+S. Kreipl. + A Large Step Random Walk for Minimizing Total Weighted Tardiness in a Job Shop. + Journal of Scheduling, 3(3):125–138, 2000.
+[ bib ] + +
+ + +
+[761] +
+
+Stefanie Kritzinger, Fabien Tricoire, Karl F. Doerner, Richard F. Hartl, and Thomas Stützle. + A Unified Framework for Routing Problems with a Fixed Fleet Size. + International Journal of Metaheuristics, 6(3):160–209, 2017.
+[ bib ] + +
+ + +
+[762] +
+
+Joseph B Kruskal. + On the shortest spanning subtree of a graph and the traveling salesman problem. + Proceedings of the American Mathematical Society, 7(1):48–50, 1956.
+[ bib ] + +
+ + +
+[763] +
+
+J Kuhpfahl and Christian Bierwirth. + A Study on Local Search Neighborhoods for the Job Shop Scheduling Problem with Total Weighted Tardiness Objective. + Computers & Operations Research, 66:44–57, 2016.
+[ bib ] + +
+ + +
+[764] +
+
+Tobias Kuhn, Carlos M. Fonseca, Luís Paquete, Stefan Ruzika, Miguel M. Duarte, and José Rui Figueira. + Hypervolume subset selection in two dimensions: Formulations and algorithms. + Evolutionary Computation, 24(3):411–425, 2016.
+[ bib ] + +
+ + +
+[765] +
+
+Harold W. Kuhn. + The hungarian method for the assignment problem. + Naval Research Logistics Quarterly, 2(1–2):83–97, 1955.
+[ bib ] + +
+ + +
+[766] +
+
+Max Kuhn. + Building Predictive Models in R Using the caret Package. + Journal of Statistical Software, 28(5):1–26, 2008.
+[ bib ] + +
+ + +
+[767] +
+
+R. Kumar and P. K. Singh. + Pareto Evolutionary Algorithm Hybridized with Local Search for Biobjective TSP. + Studies in Computational Intelligence, 75:361–398, 2007.
+[ bib ] + +
+ + +
+[768] +
+
+H. T. Kung, F. Luccio, and F. P. Preparata. + On Finding the Maxima of a Set of Vectors. + Journal of the ACM, 22(4):469–476, 1975.
+[ bib ] + +
+ + +
+[769] +
+
+I. Kurtulus and E. W. Davis. + Multi-Project Scheduling: Categorization of Heuristic Rules Performance. + Management Science, 28(2):161–172, 1982.
+[ bib | +DOI ] +
+Application of heuristic solution procedures to the + practical problem of project scheduling has + previously been studied by numerous + researchers. However, there is little consensus + about their findings, and the practicing manager is + currently at a loss as to which scheduling rule to + use. Furthermore, since no categorization process + was developed, it is assumed that once a rule is + selected it must be used throughout the whole + project. This research breaks away from this + tradition by providing a categorization process + based on two powerful project summary measures. The + first measure identifies the location of the peak of + total resource requirements and the second measure + identifies the rate of utilization of each resource + type. The performance of the rules are classified + according to values of these two measures, and it is + shown that a rule introduced by this research + performs significantly better on most categories of + projects. +
+
+Keywords: project management, research and development +
+ +
+ + +
+[770] +
+
+H. J. Kushner. + A New Method of Locating the Maximum Point of an Arbitrary Multipeak Curve in the Presence of Noise. + Journal of Basic Engineering, 86(1):97–106, March 1964.
+[ bib | +DOI | +epub ] +
+A versatile and practical method of searching a parameter + space is presented. Theoretical and experimental results + illustrate the usefulness of the method for such problems as + the experimental optimization of the performance of a system + with a very general multipeak performance function when the + only available information is noise-distributed samples of + the function. At present, its usefulness is restricted to + optimization with respect to one system parameter. The + observations are taken sequentially; but, as opposed to the + gradient method, the observation may be located anywhere on + the parameter interval. A sequence of estimates of the + location of the curve maximum is generated. The location of + the next observation may be interpreted as the location of + the most likely competitor (with the current best estimate) + for the location of the curve maximum. A Brownian motion + stochastic process is selected as a model for the unknown + function, and the observations are interpreted with respect + to the model. The model gives the results a simple intuitive + interpretation and allows the use of simple but efficient + sampling procedures. The resulting process possesses some + powerful convergence properties in the presence of noise; it + is nonparametric and, despite its generality, is efficient in + the use of observations. The approach seems quite promising + as a solution to many of the problems of experimental system + optimization. +
+ +
+ + +
+[771] +
+
+Jan H. Kwakkel. + The Exploratory Modeling Workbench: An open source toolkit for exploratory modeling, scenario discovery, and (multi-objective) robust decision making. + Environmental Modelling & Software, 96:239–250, 2017.
+[ bib ] + +
+ + +
+[772] +
+
+Marco Laumanns, Lothar Thiele, Kalyanmoy Deb, and Eckart Zitzler. + Combining Convergence and Diversity in Evolutionary Multiobjective Optimization. + Evolutionary Computation, 10(3):263–282, 2002.
+[ bib | +DOI ] +
+Proposed ε-approx and ε-Pareto archivers +
+
+Keywords: archiving, ε-dominance, ε-approximation, + ε-Pareto +
+ +
+ + +
+[773] +
+
+Antonio LaTorre, Santiago Muelas, and José-María Peña. + A MOS-based dynamic memetic differential evolution algorithm for continuous optimization: a scalability test. + Soft Computing, 15(11):2187–2199, 2011.
+[ bib ] + +
+ + +
+[774] +
+
+Peter J. M. van Laarhoven, Emile H. L. Aarts, and Jan Karel Lenstra. + Job Shop Scheduling by Simulated Annealing. + Operations Research, 40(1):113–125, 1992.
+[ bib ] + +
+ + +
+[775] +
+
+Martine Labbé, Patrice Marcotte, and Gilles Savard. + A Bilevel Model of Taxation and Its Application to Optimal Highway Pricing. + Management Science, 44(12):1608–1622, 1998.
+[ bib | +DOI ] + +
+ + +
+[776] +
+
+Martine Labbé and Alessia Violin. + Bilevel programming and price setting problems. + 4OR: A Quarterly Journal of Operations Research, 11(1):1–30, 2013.
+[ bib | +DOI ] + +
+ + +
+[777] +
+
+Benjamin Lacroix, Daniel Molina, and Francisco Herrera. + Region based memetic algorithm for real-parameter optimisation. + Information Sciences, 262:15–31, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[778] +
+
+Manuel Laguna. + Editor's Note on the MIC 2013 Special Issue of the Journal of Heuristics (Volume 22, Issue 4, August 2016). + Journal of Heuristics, 22(5):665–666, 2016.
+[ bib ] + +
+ + +
+[779] +
+
+Xiangjing Lai and Jin-Kao Hao. + Iterated Maxima Search for the Maximally Diverse Grouping Problem. + European Journal of Operational Research, 254(3):780–800, 2016.
+[ bib ] + +
+ + +
+[780] +
+
+A. H. Land and A. G. Doig. + An Automatic Method of Solving Discrete Programming Problems. + Econometrica, 28(3):497–520, 1960.
+[ bib ] + +
+ + +
+[781] +
+
+William B. Langdon and Mark Harman. + Optimising Software with Genetic Programming. + IEEE Transactions on Evolutionary Computation, 19(1):118–135, 2015.
+[ bib ] + +
+ + +
+[782] +
+
+M. Lang, H. Kotthaus, P. Marwedel, C. Weihs, J. Rahnenführer, and Bernd Bischl. + Automatic Model Selection for High-Dimensional Survival Analysis. + Journal of Statistical Computation and Simulation, 85(1):62–76, 2014.
+[ bib | +DOI ] + +
+ + +
+[783] +
+
+A. Langevin, F. Soumis, and J. Desrosiers. + Classification of travelling salesman problem formulations. + Operations Research Letters, 9(2):127–132, 1990.
+[ bib ] + +
+ + +
+[784] +
+
+A. Langevin, M. Desrochers, J. Desrosiers, Sylvie Gélinas, and F. Soumis. + A Two-Commodity Flow Formulation for the Traveling Salesman and Makespan Problems with Time Windows. + Networks, 23(7):631–640, 1993.
+[ bib ] + +
+ + +
+[785] +
+
+Kevin E. Lansey and K. Awumah. + Optimal Pump Operations Considering Pump Switches. + Journal of Water Resources Planning and Management, ASCE, 120(1):17–35, January / February 1994.
+[ bib ] + +
+ + +
+[786] +
+
+Gilbert Laporte. + Fifty Years of Vehicle Routing. + Transportation Science, 43(4):408–416, 2009.
+[ bib ] + +
+ + +
+[787] +
+
+Marco Laumanns. + Stochastic convergence of random search to fixed size Pareto set approximations. + Arxiv preprint arXiv:0711.2949, 2007.
+[ bib | +http ] + +
+ + +
+[788] +
+
+Benoît Laurent and Jin-Kao Hao. + Iterated Local Search for the Multiple Depot Vehicle Scheduling Problem. + Computers and Industrial Engineering, 57(1):277–286, 2009.
+[ bib ] + +
+ + +
+[789] +
+
+Marco Laumanns, Lothar Thiele, and Eckart Zitzler. + Running time analysis of multiobjective evolutionary algorithms on pseudo-boolean functions. + IEEE Transactions on Evolutionary Computation, 8(2):170–182, 2004.
+[ bib ] + +
+ + +
+[790] +
+
+Marco Laumanns, Lothar Thiele, and Eckart Zitzler. + Running time analysis of evolutionary algorithms on a simplified multiobjective knapsack problem. + Natural Computing, 3(1):37–51, 2004.
+[ bib ] + +
+ + +
+[791] +
+
+Marco Laumanns and Rico Zenklusen. + Stochastic convergence of random search methods to fixed size Pareto front approximations. + European Journal of Operational Research, 213(2):414–421, 2011.
+[ bib | +DOI ] + +
+ + +
+[792] +
+
+E. L. Lawler and D. E. Wood. + Branch-and-Bound Methods: A Survey. + Operations Research, 14(4):699–719, 1966.
+[ bib | +DOI ] + +
+ + +
+[793] +
+
+S. E. Lazic. + The problem of pseudoreplication in neuroscientific studies: is it affecting your analysis? + BMC Neuroscience, 11(5):397–407, 2004.
+[ bib | +DOI ] + +
+ + +
+[794] +
+
+Yann LeCun, Yoshua Bengio, et al. + Convolutional networks for images, speech, and time series. + The handbook of brain theory and neural networks, 3361(10):255–258, 1995.
+[ bib ] + +
+ + +
+[795] +
+
+Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. + Deep learning. + Nature, 521(7553):436–444, 2015.
+[ bib ] + +
+ + +
+[796] +
+
+Vinícius Leal do Forte, Flávio Marcelo Tavares Montenegro, José André de Moura Brito, and Nelson Maculan. + Iterated Local Search Algorithms for the Euclidean Steiner Tree Problem in n Dimensions. + International Transactions in Operational Research, 23(6):1185–1199, 2016.
+[ bib ] + +
+ + +
+[797] +
+
+Per Kristian Lehre and Carsten Witt. + Black-box search by unbiased variation. + Algorithmica, 64(4):623–642, 2012.
+[ bib ] + +
+ + +
+[798] +
+
+Frank Thomson Leighton. + A Graph Coloring Algorithm for Large Scheduling Problems. + Journal of Research of the National Bureau of Standards, 84(6):489–506, 1979.
+[ bib ] + +
+ + +
+[799] +
+
+Robert J. Lempert, David G. Groves, Steven W. Popper, and Steven C. Bankes. + A general analytic method for generating robust strategies and narrative scenarios. + Management Science, 52(4):514–528, 2006.
+[ bib ] + +
+ + +
+[800] +
+
+C. Leon, S. Martin, J. M. Elena, and J. Luque. + EXPLORE: Hybrid expert system for water networks management. + Journal of Water Resources Planning and Management, ASCE, 126(2):65–74, 2000.
+[ bib ] + +
+ + +
+[801] +
+
+Leonid Levin. + Universal'nyie perebornyie zadachi. + Problemy Peredachi Informatsii, 9:265–266, 1973.
+[ bib ] + +
+ + +
+[802] +
+
+Daniel Lewandowski, Dorota Kurowicka, and Harry Joe. + Generating Random Correlation Matrices Based on Vines and Extended Onion Method. + Journal of Multivariate Analysis, 100(9):1989–2001, 2009.
+[ bib | +DOI ] +
+We extend and improve two existing methods of generating + random correlation matrices, the onion method of Ghosh and + Henderson [S. Ghosh, S.G. Henderson, Behavior of the norta + method for correlated random vector generation as the + dimension increases, ACM Transactions on Modeling and + Computer Simulation (TOMACS) 13 (3) (2003) 276-294] and the + recently proposed method of Joe [H. Joe, Generating random + correlation matrices based on partial correlations, Journal + of Multivariate Analysis 97 (2006) 2177-2189] based on + partial correlations. The latter is based on the so-called + D-vine. We extend the methodology to any regular vine and + study the relationship between the multiple correlation and + partial correlations on a regular vine. We explain the onion + method in terms of elliptical distributions and extend it to + allow generating random correlation matrices from the same + joint distribution as the vine method. The methods are + compared in terms of time necessary to generate 5000 random + correlation matrices of given dimensions. +
+
+Keywords: Correlation matrix; Dependence vines; Onion method; Partial + correlation; LKJ +
+ +
+ + +
+[803] +
+
+Jianjun David Li. + A two-step rejection procedure for testing multiple hypotheses. + Journal of Statistical Planning and Inference, 138(6):1521–1527, 2008.
+[ bib ] + +
+ + +
+[804] +
+
+Miqing Li. + Is Our Archiving Reliable? Multiobjective Archiving Methods on “Simple” Artificial Input Sequences. + ACM Transactions on Evolutionary Learning and Optimization, 1(3):1–19, 2021.
+[ bib | +DOI ] + +
+ + +
+[805] +
+
+Ke Li, Renzhi Chen, Guangtao Fu, and Xin Yao. + Two-archive evolutionary algorithm for constrained multiobjective optimization. + IEEE Transactions on Evolutionary Computation, 23(2):303–315, 2018.
+[ bib ] + +
+ + +
+[806] +
+
+Miqing Li, Tao Chen, and Xin Yao. + How to evaluate solutions in Pareto-based search-based software engineering? A critical review and methodological guidance. + IEEE Transactions on Software Engineering, 48(5):1771–1799, 2020.
+[ bib | +DOI ] + +
+ + +
+[807] +
+
+Miqing Li, Crina Grosan, Shengxiang Yang, Xiaohui Liu, and Xin Yao. + Multi-line distance minimization: A visualized many-objective test problem suite. + IEEE Transactions on Evolutionary Computation, 22(1):61–78, 2018.
+[ bib ] +
+highly degenerate Pareto fronts +
+ +
+ + +
+[808] +
+
+Zhiyi Li, Mohammad Shahidehpour, Shay Bahramirad, and Amin Khodaei. + Optimizing Traffic Signal Settings in Smart Cities. + IEEE Transactions on Smart Grid, 3053(4):1–1, 2016.
+[ bib | +DOI ] +
+Traffic signals play a critical role in smart cities for + mitigating traffic congestions and reducing the emission in + metropolitan areas. This paper proposes a bi-level + optimization framework to settle the optimal traffic signal + setting problem. The upper-level problem determines the + traffic signal settings to minimize the drivers' average + travel time, while the lower-level problem aims for achieving + the network equilibrium using the settings calculated at the + upper level. Genetic algorithm is employed with the + integration of microscopic-traffic-simulation based dynamic + traffic assignment (DTA) to decouple the complex bi-level + problem into tractable single-level problems which are solved + sequentially. Case studies on a synthetic traffic network and + a real-world traffic subnetwork are conducted to examine the + effectiveness of the proposed model and relevant solution + methods. Additional strategies are provided for the extension + of the proposed model and the acceleration solution process + in large-area traffic network applications. +
+ +
+ + +
+[809] +
+
+Xiaoping Li, Long Chen, Haiyan Xu, and Jatinder N.D. Gupta. + Trajectory Scheduling Methods for Minimizing Total Tardiness in a Flowshop. + Operations Research Perspectives, 2:13–23, 2015.
+[ bib | +DOI ] + +
+ + +
+[810] +
+
+Lisha Li, Kevin Jamieson, Giulia DeSalvo, Afshin Rostamizadeh, and Ameet Talwalkar. + Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization. + Journal of Machine Learning Research, 18(185):1–52, 2018.
+[ bib | +epub ] +
+Performance of machine learning algorithms depends critically on identifying a good set of hyperparameters. While recent approaches use Bayesian optimization to adaptively select configurations, we focus on speeding up random search through adaptive resource allocation and early-stopping. We formulate hyperparameter optimization as a pure-exploration non-stochastic infinite-armed bandit problem where a predefined resource like iterations, data samples, or features is allocated to randomly sampled configurations. We introduce a novel algorithm, our algorithm, for this framework and analyze its theoretical properties, providing several desirable guarantees. Furthermore, we compare our algorithm with popular Bayesian optimization methods on a suite of hyperparameter optimization problems. We observe that our algorithm can provide over an order-of-magnitude speedup over our competitor set on a variety of deep-learning and kernel-based learning problems. +
+
+Keywords: racing +
+ +
+ + +
+[811] +
+
+Y. Li and W. Li. + Adaptive Ant Colony Optimization Algorithm Based on Information Entropy: Foundation and Application. + Fundamenta Informaticae, 77(3):229–242, 2007.
+[ bib ] + +
+ + +
+[812] +
+
+Bingdong Li, Jinlong Li, Ke Tang, and Xin Yao. + Many-Objective Evolutionary Algorithms: A Survey. + ACM Computing Surveys, 48(1):1–35, 2015.
+[ bib | +DOI ] + +
+ + +
+[813] +
+
+Miqing Li, Manuel López-Ibáñez, and Xin Yao. + Multi-Objective Archiving. + IEEE Transactions on Evolutionary Computation, 2023.
+[ bib | +DOI ] +
+Most multi-objective optimisation algorithms maintain an + archive explicitly or implicitly during their search. Such an + archive can be solely used to store high-quality solutions + presented to the decision maker, but in many cases may + participate in the search process (e.g., as the population in + evolutionary computation). Over the last two decades, + archiving, the process of comparing new solutions with + previous ones and deciding how to update the + archive/population, stands as an important issue in + evolutionary multi-objective optimisation (EMO). This is + evidenced by constant efforts from the community on + developing various effective archiving methods, ranging from + conventional Pareto-based methods to more recent + indicator-based and decomposition-based ones. However, the + focus of these efforts is on empirical performance comparison + in terms of specific quality indicators; there is lack of + systematic study of archiving methods from a general + theoretical perspective. In this paper, we attempt to conduct + a systematic overview of multi-objective archiving, in the + hope of paving the way to understand archiving algorithms + from a holistic perspective of theory and practice, and more + importantly providing a guidance on how to design + theoretically desirable and practically useful archiving + algorithms. In doing so, we also present that archiving + algorithms based on weakly Pareto compliant indicators (e.g., + ε-indicator), as long as designed properly, can + achieve the same theoretical desirables as archivers based on + Pareto compliant indicators (e.g., hypervolume + indicator). Such desirables include the property + limit-optimal, the limit form of the possible optimal + property that a bounded archiving algorithm can have with + respect to the most general form of superiority between + solution sets. +
+ +
+ + +
+[814] +
+
+Bingdong Li, Ke Tang, Jinlong Li, and Xin Yao. + Stochastic ranking algorithm for many-objective optimization based on multiple indicators. + IEEE Transactions on Evolutionary Computation, 20(6):924–938, 2016.
+[ bib ] + +
+ + +
+[815] +
+
+Miqing Li, Shengxiang Yang, and Xiaohui Liu. + Shift-based density estimation for Pareto-based algorithms in many-objective optimization. + IEEE Transactions on Evolutionary Computation, 18(3):348–365, 2014.
+[ bib ] +
+Proposed SDE indicator algorithm +
+ +
+ + +
+[816] +
+
+Miqing Li, Shengxiang Yang, and Xiaohui Liu. + Pareto or non-Pareto: Bi-criterion evolution in multiobjective optimization. + IEEE Transactions on Evolutionary Computation, 20(5):645–665, 2016.
+[ bib ] + +
+ + +
+[817] +
+
+Miqing Li and Xin Yao. + Quality Evaluation of Solution Sets in Multiobjective Optimisation: A Survey. + ACM Computing Surveys, 52(2):1–38, 2019.
+[ bib | +DOI ] + +
+ + +
+[818] +
+
+Miqing Li and Xin Yao. + Dominance Move: A Measure of Comparing Solution Sets in Multiobjective Optimization. + arXiv preprint arXiv:1702.00477, 2017.
+[ bib ] + +
+ + +
+[819] +
+
+Miqing Li and Xin Yao. + What weights work for you? Adapting weights for any Pareto front shape in decomposition-based evolutionary multiobjective optimisation. + Evolutionary Computation, 28(2):227–253, 2020.
+[ bib ] + +
+ + +
+[820] +
+
+Hui Li and Qingfu Zhang. + Multiobjective Optimization Problems with Complicated Pareto sets, MOEA/D and NSGA-II. + IEEE Transactions on Evolutionary Computation, 13(2):284–302, 2009.
+[ bib ] + +
+ + +
+[821] +
+
+Zhipan Li, Juan Zou, Shengxiang Yang, and Jinhua Zheng. + A two-archive algorithm with decomposition and fitness allocation for multi-modal multi-objective optimization. + Information Sciences, 574:413–430, 2021.
+[ bib ] + +
+ + +
+[822] +
+
+Tianjun Liao, Doǧan Aydın, and Thomas Stützle. + Artificial Bee Colonies for Continuous Optimization: Experimental Analysis and Improvements. + Swarm Intelligence, 7(4):327–356, 2013.
+[ bib ] + +
+ + +
+[823] +
+
+Tianjun Liao, Daniel Molina, Marco A. Montes de Oca, and Thomas Stützle. + A Note on the Effects of Enforcing Bound Constraints on Algorithm Comparisons using the IEEE CEC'05 Benchmark Function Suite. + Evolutionary Computation, 22(2):351–359, 2014.
+[ bib ] + +
+ + +
+[824] +
+
+Tianjun Liao, Daniel Molina, and Thomas Stützle. + Performance Evaluation of Automatically Tuned Continuous Optimizers on Different Benchmark Sets. + Applied Soft Computing, 27:490–503, 2015.
+[ bib ] + +
+ + +
+[825] +
+
+Tianjun Liao, Marco A. Montes de Oca, and Thomas Stützle. + Computational results for an automatically tuned CMA-ES with increasing population size on the CEC'05 benchmark set. + Soft Computing, 17(6):1031–1046, 2013.
+[ bib | +DOI ] + +
+ + +
+[826] +
+
+Tianjun Liao, Krzysztof Socha, Marco A. Montes de Oca, Thomas Stützle, and Marco Dorigo. + Ant Colony Optimization for Mixed-Variable Optimization Problems. + IEEE Transactions on Evolutionary Computation, 18(4):503–518, 2014.
+[ bib ] +
+Keywords: ACOR +
+ +
+ + +
+[827] +
+
+Tianjun Liao, Thomas Stützle, Marco A. Montes de Oca, and Marco Dorigo. + A Unified Ant Colony Optimization Algorithm for Continuous Optimization. + European Journal of Operational Research, 234(3):597–609, 2014.
+[ bib ] + +
+ + +
+[828] +
+
+C.-J. Liao, C.-T. Tseng, and P. Luarn. + A Discrete Version of Particle Swarm Optimization for Flowshop Scheduling Problems. + Computers & Operations Research, 34(10):3099–3111, 2007.
+[ bib ] + +
+ + +
+[829] +
+
+Arnaud Liefooghe, Fabio Daolio, Bilel Derbel, Sébastien Verel, Hernán E. Aguirre, and Kiyoshi Tanaka. + Landscape-Aware Performance Prediction for Evolutionary Multi-objective Optimization. + IEEE Transactions on Evolutionary Computation, 24(6):1063–1077, 2020.
+[ bib ] + +
+ + +
+[830] +
+
+Arnaud Liefooghe, Jérémie Humeau, Salma Mesmoudi, Laetitia Jourdan, and El-Ghazali Talbi. + On dominance-based multiobjective local search: design, implementation and experimental analysis on scheduling and traveling salesman problems. + Journal of Heuristics, 18(2):317–352, 2012.
+[ bib | +DOI ] +
+This paper discusses simple local search approaches for + approximating the efficient set of multiobjective + combinatorial optimization problems. We focus on algorithms + defined by a neighborhood structure and a dominance relation + that iteratively improve an archive of nondominated + solutions. Such methods are referred to as dominance-based + multiobjective local search. We first provide a concise + overview of existing algorithms, and we propose a model + trying to unify them through a fine-grained + decomposition. The main problem-independent search components + of dominance relation, solution selection, neighborhood + exploration and archiving are largely discussed. Then, a + number of state-of-the-art and original strategies are + experimented on solving a permutation flowshop scheduling + problem and a traveling salesman problem, both on a two- and + a three-objective formulation. Experimental results and a + statistical comparison are reported in the paper, and some + directions for future research are highlighted. +
+ +
+ + +
+[831] +
+
+Arnaud Liefooghe, Laetitia Jourdan, and El-Ghazali Talbi. + A Software Framework Based on a Conceptual Unified Model for Evolutionary Multiobjective Optimization: ParadisEO-MOEO. + European Journal of Operational Research, 209(2):104–112, 2011.
+[ bib ] + +
+ + +
+[832] +
+
+Arnaud Liefooghe, Sébastien Verel, and Jin-Kao Hao. + A hybrid metaheuristic for multiobjective unconstrained binary quadratic programming. + Applied Soft Computing, 16:10–19, 2014.
+[ bib ] + +
+ + +
+[833] +
+
+Bojan Likar and Juš Kocijan. + Predictive control of a gas–liquid separation plant based on a Gaussian process model. + Computers & Chemical Engineering, 31(3):142–152, 2007.
+[ bib | +DOI ] + +
+ + +
+[834] +
+
+Marius Thomas Lindauer, Katharina Eggensperger, Matthias Feurer, André Biedenkapp, Difan Deng, Carolin Benjamins, Tim Ruhkopf, René Sass, and Frank Hutter. + SMAC3: A Versatile Bayesian Optimization Package for Hyperparameter Optimization. + Journal of Machine Learning Research, 23:1–9, 2022.
+[ bib | +epub ] + +
+ + +
+[835] +
+
+Marius Thomas Lindauer, Holger H. Hoos, Frank Hutter, and Torsten Schaub. + AutoFolio: An Automatically Configured Algorithm Selector. + Journal of Artificial Intelligence Research, 53:745–778, 2015.
+[ bib ] + +
+ + +
+[836] +
+
+S. Lin and B. W. Kernighan. + An Effective Heuristic Algorithm for the Traveling Salesman Problem. + Operations Research, 21(2):498–516, 1973.
+[ bib ] + +
+ + +
+[837] +
+
+Marius Thomas Lindauer, Jan N. van Rijn, and Lars Kotthoff. + The algorithm selection competitions 2015 and 2017. + Artificial Intelligence, 272:86–100, 2019.
+[ bib ] + +
+ + +
+[838] +
+
+Andrei Lissovoi and Carsten Witt. + Runtime Analysis of Ant Colony Optimization on Dynamic Shortest Path Problems. + Theoretical Computer Science, 561(Part A):73–85, 2015.
+[ bib | +DOI ] +
+A simple ACO algorithm called λ-MMAS for dynamic + variants of the single-destination shortest paths problem is + studied by rigorous runtime analyses. Building upon previous + results for the special case of 1-MMAS, it is studied to what + extent an enlarged colony using λ ants per vertex + helps in tracking an oscillating optimum. It is shown that + easy cases of oscillations can be tracked by a constant + number of ants. However, the paper also identifies more + involved oscillations that with overwhelming probability + cannot be tracked with any polynomial-size colony. Finally, + parameters of dynamic shortest-path problems which make the + optimum difficult to track are discussed. Experiments + illustrate theoretical findings and conjectures. +
+ +
+ + +
+[839] +
+
+J. D. C. Little, K. G. Murty, D. W. Sweeney, and C. Karel. + An Algorithm for the Traveling Salesman Problem. + Operations Research, 11:972–989, 1963.
+[ bib ] + +
+ + +
+[840] +
+
+Shusen Liu, Dan Maljovec, Bei Wang, Peer-Timo Bremer, and Valerio Pascucci. + Visualizing High-Dimensional Data: Advances in the Past Decade. + IEEE Transactions on Visualization and Computer Graphics, 23(3), 2017.
+[ bib | +DOI ] + +
+ + +
+[841] +
+
+Jiyin Liu and Colin R. Reeves. + Constructive and Composite Heuristic Solutions to the P//ΣCi Scheduling Problem. + European Journal of Operational Research, 132(2):439–452, 2001.
+[ bib | +DOI ] + +
+ + +
+[842] +
+
+Yiping Liu, Gary G. Yen, and Dunwei Gong. + A multimodal multiobjective evolutionary algorithm using two-archive and recombination strategies. + IEEE Transactions on Evolutionary Computation, 23(4):660–674, 2018.
+[ bib ] + +
+ + +
+[843] +
+
+Marco Locatelli and Fabio Schoen. + Random Linkage: a family of acceptance/rejection algorithms for global optimisation. + Mathematical Programming, 85(2), 1999.
+[ bib ] +
+Keywords: Multi-Level Single-Linkage (MLSL) +
+ +
+ + +
+[844] +
+
+Andrea Lodi, Silvano Martello, and Michele Monaci. + Two-dimensional packing problems: A survey. + European Journal of Operational Research, 141(2):241–252, 2002.
+[ bib | +DOI ] + +
+ + +
+[845] +
+
+Andrea Lodi, Silvano Martello, and Daniele Vigo. + Heuristic and metaheuristic approaches for a class of two-dimensional bin packing problems. + INFORMS Journal on Computing, 11(4):345–357, 1999.
+[ bib | +DOI ] + +
+ + +
+[846] +
+
+Andrea Lodi, Silvano Martello, and Daniele Vigo. + TSpack: a unified tabu search code for multi-dimensional bin packing problems. + Annals of Operations Research, 131(1-4):203–213, 2004.
+[ bib | +DOI ] + +
+ + +
+[847] +
+
+Andrea Lodi and Giulia Zarpellon. + On Learning and Branching: A Survey. + TOP, 25:207–236, 2017.
+[ bib ] + +
+ + +
+[848] +
+
+Jason D. Lohn, Gregory S. Hornby, and Derek S. Linden. + Human-competitive Evolved Antennas. + Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 22(3):235–247, 2008.
+[ bib | +DOI ] +
+Evolutionary optimization of antennas for NASA +
+ +
+ + +
+[849] +
+
+Manuel López-Ibáñez and Christian Blum. + Beam-ACO for the travelling salesman problem with time windows. + Computers & Operations Research, 37(9):1570–1583, 2010.
+[ bib | +DOI ] +
+The travelling salesman problem with time windows is + a difficult optimization problem that arises, for + example, in logistics. This paper deals with the + minimization of the travel-cost. For solving this + problem, this paper proposes a Beam-ACO algorithm, + which is a hybrid method combining ant colony + optimization with beam search. In general, Beam-ACO + algorithms heavily rely on accurate and + computationally inexpensive bounding information for + differentiating between partial solutions. This work + uses stochastic sampling as a useful alternative. An + extensive experimental evaluation on seven benchmark + sets from the literature shows that the proposed + Beam-ACO algorithm is currently a state-of-the-art + technique for the travelling salesman problem with + time windows when travel-cost optimization is + concerned. +
+
+Keywords: Ant colony optimization, Travelling salesman problem with + time windows, Hybridization +
+ +
+ + +
+[850] +
+
+Manuel López-Ibáñez, Christian Blum, Jeffrey W. Ohlmann, and Barrett W. Thomas. + The Travelling Salesman Problem with Time Windows: Adapting Algorithms from Travel-time to Makespan Optimization. + Applied Soft Computing, 13(9):3806–3815, 2013.
+[ bib | +DOI | +epub ] + +
+ + +
+[851] +
+
+Manuel López-Ibáñez, Jürgen Branke, and Luís Paquete. + Reproducibility in Evolutionary Computation. + Arxiv preprint arXiv:20102.03380 [cs.AI], 2021.
+[ bib | +http ] +
+Experimental studies are prevalent in Evolutionary + Computation (EC), and concerns about the reproducibility and + replicability of such studies have increased in recent times, + reflecting similar concerns in other scientific fields. In + this article, we suggest a classification of different types + of reproducibility that refines the badge system of the + Association of Computing Machinery (ACM) adopted by TELO. We + discuss, within the context of EC, the different types of + reproducibility as well as the concepts of artifact and + measurement, which are crucial for claiming + reproducibility. We identify cultural and technical obstacles + to reproducibility in the EC field. Finally, we provide + guidelines and suggest tools that may help to overcome some + of these reproducibility obstacles. +
+
+Keywords: Evolutionary Computation, Reproducibility, Empirical study, + Benchmarking +
+ +
+ + +
+[852] +
+
+Manuel López-Ibáñez, Jürgen Branke, and Luís Paquete. + Reproducibility in Evolutionary Computation. + ACM Transactions on Evolutionary Learning and Optimization, 1(4):1–21, 2021.
+[ bib | +DOI | +epub ] +
+Experimental studies are prevalent in Evolutionary + Computation (EC), and concerns about the reproducibility and + replicability of such studies have increased in recent times, + reflecting similar concerns in other scientific fields. In + this article, we suggest a classification of different types + of reproducibility that refines the badge system of the + Association of Computing Machinery (ACM) adopted by TELO. We + discuss, within the context of EC, the different types of + reproducibility as well as the concepts of artifact and + measurement, which are crucial for claiming + reproducibility. We identify cultural and technical obstacles + to reproducibility in the EC field. Finally, we provide + guidelines and suggest tools that may help to overcome some + of these reproducibility obstacles. +
+
+Keywords: Evolutionary Computation, Reproducibility, Empirical study, + Benchmarking +
+ +
+ + +
+[853] +
+
+Manuel López-Ibáñez, Jérémie Dubois-Lacoste, Leslie Pérez Cáceres, Thomas Stützle, and Mauro Birattari. + The irace Package: Iterated Racing for Automatic Algorithm Configuration. + Operations Research Perspectives, 3:43–58, 2016.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[854] +
+
+Manuel López-Ibáñez, Marie-Eléonore Kessaci, and Thomas Stützle. + Automatic Design of Hybrid Metaheuristics from Algorithmic Components. + Submitted, 2017.
+[ bib ] + +
+ + +
+[855] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + Hybrid Population-based Algorithms for the Bi-objective Quadratic Assignment Problem. + Journal of Mathematical Modelling and Algorithms, 5(1):111–137, 2006.
+[ bib | +DOI ] +
+We present variants of an ant colony optimization + (MO-ACO) algorithm and of an evolutionary algorithm + (SPEA2) for tackling multi-objective combinatorial + optimization problems, hybridized with an iterative + improvement algorithm and the robust tabu search + algorithm. The performance of the resulting hybrid + stochastic local search (SLS) algorithms is + experimentally investigated for the bi-objective + quadratic assignment problem (bQAP) and compared + against repeated applications of the underlying + local search algorithms for several + scalarizations. The experiments consider structured + and unstructured bQAP instances with various degrees + of correlation between the flow matrices. We do a + systematic experimental analysis of the algorithms + using outperformance relations and the attainment + functions methodology to asses differences in the + performance of the algorithms. The experimental + results show the usefulness of the hybrid algorithms + if the available computation time is not too limited + and identify SPEA2 hybridized with very short tabu + search runs as the most promising variant. +
+ +
+ + +
+[856] +
+
+Manuel López-Ibáñez, Leslie Pérez Cáceres, and Thomas Stützle. + irace: A Tool for the Automatic Configuration of Algorithms. + International Federation of Operational Research Societies (IFORS) News, 14(2):30–32, June 2020.
+[ bib | +http ] + +
+ + +
+[857] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Ant Colony Optimisation for the Optimal Control of Pumps in Water Distribution Networks. + Journal of Water Resources Planning and Management, ASCE, 134(4):337–346, 2008.
+[ bib | +DOI | +epub ] +
+Reducing energy consumption of water distribution + networks has never had more significance than today. The greatest + energy savings can be obtained by careful scheduling of operation of + pumps. Schedules can be defined either implicitly, in terms of other + elements of the network such as tank levels, or explicitly by + specifying the time during which each pump is on/off. The + traditional representation of explicit schedules is a string of + binary values with each bit representing pump on/off status during a + particular time interval. In this paper a new explicit + representation is presented. It is based on time controlled + triggers, where the maximum number of pump switches is specified + beforehand. In this representation a pump schedule is divided into a + series of integers with each integer representing the number of + hours for which a pump is active/inactive. This reduces the number + of potential schedules (search space) compared to the binary + representation. Ant colony optimization (ACO) is a stochastic + meta-heuristic for combinatorial optimization problems that is + inspired by the foraging behavior of some species of ants. In this + paper, an application of the ACO framework was developed for the + optimal scheduling of pumps. The proposed representation was adapted + to an ant colony Optimization framework and solved for the optimal + pump schedules. Minimization of electrical cost was considered as + the objective, while satisfying system constraints. Instead of using + a penalty function approach for constraint violations, constraint + violations were ordered according to their importance and solutions + were ranked based on this order. The proposed approach was tested on + a small test network and on a large real-world network. Results are + compared with those obtained using a simple genetic algorithm based + on binary representation and a hybrid genetic algorithm that uses + level-based triggers. +
+ +
+ + +
+[858] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Representations and Evolutionary Operators for the Scheduling of Pump Operations in Water Distribution Networks. + Evolutionary Computation, 19(3):429–467, 2011.
+[ bib | +DOI ] +
+Reducing the energy consumption of water + distribution networks has never had more + significance. The greatest energy savings can be + obtained by carefully scheduling the operations of + pumps. Schedules can be defined either implicitly, + in terms of other elements of the network such as + tank levels, or explicitly by specifying the time + during which each pump is on/off. The traditional + representation of explicit schedules is a string of + binary values with each bit representing pump on/off + status during a particular time interval. In this + paper, we formally define and analyze two new + explicit representations based on time-controlled + triggers, where the maximum number of pump switches + is established beforehand and the schedule may + contain less switches than the maximum. In these + representations, a pump schedule is divided into a + series of integers with each integer representing + the number of hours for which a pump is + active/inactive. This reduces the number of + potential schedules compared to the binary + representation, and allows the algorithm to operate + on the feasible region of the search space. We + propose evolutionary operators for these two new + representations. The new representations and their + corresponding operations are compared with the two + most-used representations in pump scheduling, + namely, binary representation and level-controlled + triggers. A detailed statistical analysis of the + results indicates which parameters have the greatest + effect on the performance of evolutionary + algorithms. The empirical results show that an + evolutionary algorithm using the proposed + representations improves over the results obtained + by a recent state-of-the-art Hybrid Genetic + Algorithm for pump scheduling using level-controlled + triggers. +
+ +
+ + +
+[859] +
+
+Manuel López-Ibáñez and Thomas Stützle. + An experimental analysis of design choices of multi-objective ant colony optimization algorithms. + Swarm Intelligence, 6(3):207–232, 2012.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[860] +
+
+Manuel López-Ibáñez and Thomas Stützle. + The Automatic Design of Multi-Objective Ant Colony Optimization Algorithms. + IEEE Transactions on Evolutionary Computation, 16(6):861–875, 2012.
+[ bib | +DOI ] +
+ +Multi-objective optimization problems are problems with several, +typically conflicting criteria for evaluating solutions. Without +any a priori preference information, the Pareto optimality +principle establishes a partial order among solutions, and the +output of the algorithm becomes a set of nondominated solutions +rather than a single one. Various ant colony optimization (ACO) +algorithms have been proposed in recent years for solving such +problems. These multi-objective ACO (MOACO) algorithms exhibit +different design choices for dealing with the particularities of +the multi-objective context. This paper proposes a formulation of +algorithmic components that suffices to describe most MOACO +algorithms proposed so far. This formulation also shows that +existing MOACO algorithms often share equivalent design choices +but they are described in different terms. Moreover, this +formulation is synthesized into a flexible algorithmic framework, +from which not only existing MOACO algorithms may be +instantiated, but also combinations of components that were never +studied in the literature. In this sense, this paper goes beyond +proposing a new MOACO algorithm, but it rather introduces a +family of MOACO algorithms. The flexibility of the proposed MOACO +framework facilitates the application of automatic algorithm +configuration techniques. The experimental results presented in +this paper show that the automatically configured MOACO framework +outperforms the MOACO algorithms that inspired the framework +itself. This paper is also among the first to apply automatic +algorithm configuration techniques to multi-objective algorithms. +
+ +
+ + +
+[861] +
+
+Manuel López-Ibáñez and Thomas Stützle. + Automatically Improving the Anytime Behaviour of Optimisation Algorithms. + European Journal of Operational Research, 235(3):569–582, 2014.
+[ bib | +DOI | +supplementary material ] +
+Optimisation algorithms with good anytime behaviour try to + return as high-quality solutions as possible independently of + the computation time allowed. Designing algorithms with good + anytime behaviour is a difficult task, because performance is + often evaluated subjectively, by plotting the trade-off curve + between computation time and solution quality. Yet, the + trade-off curve may be modelled also as a set of mutually + nondominated, bi-objective points. Using this model, we + propose to combine an automatic configuration tool and the + hypervolume measure, which assigns a single quality measure + to a nondominated set. This allows us to improve the anytime + behaviour of optimisation algorithms by means of + automatically finding algorithmic configurations that produce + the best nondominated sets. Moreover, the recently proposed + weighted hypervolume measure is used here to incorporate the + decision-maker's preferences into the automatic tuning + procedure. We report on the improvements reached when + applying the proposed method to two relevant scenarios: (i) + the design of parameter variation strategies for MAX-MIN Ant + System and (ii) the tuning of the anytime behaviour of SCIP, + an open-source mixed integer programming solver with more + than 200 parameters. +
+ +
+ + +
+[862] +
+
+Eunice López-Camacho, Hugo Terashima-Marin, Peter Ross, and Gabriela Ochoa. + A unified hyper-heuristic framework for solving bin packing problems. + Expert Systems with Applications, 41(15):6876–6889, 2014.
+[ bib | +DOI ] + +
+ + +
+[863] +
+
+Samir Loudni and Patrice Boizumault. + Combining VNS with constraint programming for solving anytime optimization problems. + European Journal of Operational Research, 191:705–735, 2008.
+[ bib | +DOI ] + +
+ + +
+[864] +
+
+Helena R. Lourenço. + Job-Shop Scheduling: Computational Study of Local Search and Large-Step Optimization Methods. + European Journal of Operational Research, 83(2):347–364, 1995.
+[ bib ] + +
+ + +
+[865] +
+
+Antonio Lova and Pilar Tormos. + Analysis of Scheduling Schemes and Heuristic Rules Performance in Resource-Constrained Multiproject Scheduling. + Annals of Operations Research, 102(1-4):263–286, February 2001.
+[ bib | +DOI ] +
+Frequently, the availability of resources assigned + to a project is limited and not sufficient to + execute all the concurrent activities. In this + situation, decision making about their schedule is + necessary. Many times this schedule supposes an + increase in the project completion + time. Additionally, companies commonly manage + various projects simultaneously, sharing a pool of + renewable resources. Given these resource + constraints, we often can only apply heuristic + methods to solve the scheduling problem. In this + work the effect of the schedule generation schemes - + serial or parallel - and priority rules - MINLFT, + MINSLK, MAXTWK, SASP or FCFS - with two + approaches - multi-project and single-project - are + analysed. The time criteria considered are the mean + project delay and the multiproject duration + increase. Through an extensive computational study, + results show that with the parallel scheduling + generation scheme and the multi-project approach the + project manager can obtain a good multiproject + schedule with the time criterion selected: + minimising mean project delay or minimising + multiproject duration increase. New heuristics - + based on priority rules with a two-phase approach - + that outperform classical ones are proposed to + minimise mean project delay with a multi-project + approach. Finally, the best heuristics analysed are + evaluated together with a representative sample of + commercial project management software. +
+
+Keywords: Combinatorics, heuristic based on priority rules, + Multiproject scheduling, Operation + Research/Decision Theory, Project management, + project management software, Resource allocation, + Theory of Computation +
+ +
+ + +
+[866] +
+
+Antonio Lova, Pilar Tormos, Mariamar Cervantes, and Federico Barber. + An efficient hybrid genetic algorithm for scheduling projects with resource constraints and multiple execution modes. + International Journal of Production Economics, 117(2):302–316, 2009.
+[ bib | +DOI ] +
+Multi-mode Resource Constrained Project Scheduling + Problem (MRCPSP) aims at finding the start times + and execution modes for the activities of a project + that optimize a given objective function while + verifying a set of precedence and resource + constraints. In this paper, we focus on this problem + and develop a hybrid Genetic Algorithm (MM-HGA) to + solve it. Its main contributions are the mode + assignment procedure, the fitness function and the + use of a very efficient improving method. Its + performance is demonstrated by extensive + computational results obtained on a set of standard + instances and against the best currently available + algorithms. +
+
+Keywords: genetic algorithm, multi-mode resource-constrained + project scheduling +
+ +
+ + +
+[867] +
+
+Manuel Lozano, Fred Glover, Carlos García-Martínez, Francisco J. Rodríguez, and Rafael Martí. + Tabu Search with Strategic Oscillation for the Quadratic Minimum Spanning Tree. + IIE Transactions, 46(4):414–428, 2014.
+[ bib ] + +
+ + +
+[868] +
+
+Manuel Lozano, Daniel Molina, and Carlos García-Martínez. + Iterated Greedy for the Maximum Diversity Problem. + European Journal of Operational Research, 214(1):31–38, 2011.
+[ bib ] + +
+ + +
+[869] +
+
+Zhipeng Lü, Fred Glover, and Jin-Kao Hao. + A hybrid metaheuristic approach to solving the UBQP problem. + European Journal of Operational Research, 207(3):1254–1262, 2010.
+[ bib | +DOI ] + +
+ + +
+[870] +
+
+Andrew Lucas. + Ising formulations of many NP problems. + Frontiers in Physics, 2:5, 2014.
+[ bib | +DOI ] + +
+ + +
+[871] +
+
+M. Lundy and A. Mees. + Convergence of an Annealing Algorithm. + Mathematical Programming, 34(1):111–124, 1986.
+[ bib ] + +
+ + +
+[872] +
+
+Thibaut Lust and Jacques Teghem. + Two-phase Pareto local search for the biobjective traveling salesman problem. + Journal of Heuristics, 16(3):475–510, 2010.
+[ bib | +DOI ] +
+In this work, we present a method, called Two-Phase + Pareto Local Search, to find a good approximation of the + efficient set of the biobjective traveling salesman + problem. In the first phase of the method, an initial + population composed of a good approximation of the extreme + supported efficient solutions is generated. We use as second + phase a Pareto Local Search method applied to each solution + of the initial population. We show that using the combination + of these two techniques: good initial population generation + plus Pareto Local Search gives better results than + state-of-the-art algorithms. Two other points are introduced: + the notion of ideal set and a simple way to produce + near-efficient solutions of multiobjective problems, by using + an efficient single-objective solver with a data perturbation + technique. +
+ +
+ + +
+[873] +
+
+Thibaut Lust and Jacques Teghem. + The multiobjective multidimensional knapsack problem: a survey and a new approach. + Arxiv preprint arXiv:1007.4063, 2010. + Published as [874].
+[ bib ] + +
+ + +
+[874] +
+
+Thibaut Lust and Jacques Teghem. + The multiobjective multidimensional knapsack problem: a survey and a new approach. + International Transactions in Operational Research, 19(4):495–520, 2012.
+[ bib | +DOI ] + +
+ + +
+[875] +
+
+Thibaut Lust and Andrzej Jaszkiewicz. + Speed-up techniques for solving large-scale biobjective TSP. + Computers & Operations Research, 37(3):521–533, 2010.
+[ bib | +DOI ] +
+Keywords: Multiobjective combinatorial optimization, Hybrid + metaheuristics, TSP, Local search, Speed-up techniques +
+ +
+ + +
+[876] +
+
+C. von Lücken, Benjamín Barán, and Carlos Brizuela. + A survey on multi-objective evolutionary algorithms for many-objective problems. + Computational Optimization and Applications, 58(3):707–756, 2014.
+[ bib ] + +
+ + +
+[877] +
+
+Laurens van der Maaten and Geoffrey Hinton. + Visualizing Data using t-SNE. + Journal of Machine Learning Research, 9(86):2579–2605, 2008.
+[ bib | +epub ] + +
+ + +
+[878] +
+
+Marlos C. Machado, Marc G. Bellemare, Erik Talvitie, Joel Veness, Matthew Hausknecht, and Michael Bowling. + Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents. + Journal of Artificial Intelligence Research, 61(1):523–562, January 2018.
+[ bib ] +
+The Arcade Learning Environment (ALE) is an evaluation + platform that poses the challenge of building AI agents with + general competency across dozens of Atari 2600 games. It + supports a variety of different problem settings and it has + been receiving increasing attention from the scientific + community, leading to some high-pro_le success stories such + as the much publicized Deep Q-Networks (DQN). In this article + we take a big picture look at how the ALE is being used by + the research community. We show how diverse the evaluation + methodologies in the ALE have become with time, and highlight + some key concerns when evaluating agents in the ALE. We use + this discussion to present some methodological best practices + and provide new benchmark results using these best + practices. To further the progress in the field, we introduce + a new version of the ALE that supports multiple game modes + and provides a form of stochasticity we call sticky + actions. We conclude this big picture look by revisiting + challenges posed when the ALE was introduced, summarizing the + state-of-the-art in various problems and highlighting + problems that remain open. +
+ +
+ + +
+[879] +
+
+Sam Madden. + From Databases to Big Data. + IEEE Internet Computing, 16(3), 2012.
+[ bib ] + +
+ + +
+[880] +
+
+M. Mahdavi, M. Fesanghary, and E. Damangir. + An improved harmony search algorithm for solving optimization problems. + Applied Mathematics and Computation, 188(2):1567–1579, 2007.
+[ bib | +DOI ] +
+This paper develops an Improved harmony search (IHS) + algorithm for solving optimization problems. IHS employs a + novel method for generating new solution vectors that + enhances accuracy and convergence rate of harmony search (HS) + algorithm. In this paper the impacts of constant parameters + on harmony search algorithm are discussed and a strategy for + tuning these parameters is presented. The IHS algorithm has + been successfully applied to various benchmarking and + standard engineering optimization problems. Numerical results + reveal that the proposed algorithm can find better solutions + when compared to HS and other heuristic or deterministic + methods and is a powerful search algorithm for various + engineering optimization problems. +
+
+Keywords: Global optimization, Heuristics, Harmony search algorithm, + Mathematical programming +
+ +
+ + +
+[881] +
+
+Guilherme B. Mainieri and Débora P. Ronconi. + New heuristics for total tardiness minimization in a flexible flowshop. + Optimization Letters, pp.  1–20, 2012.
+[ bib ] + +
+ + +
+[882] +
+
+Holger R. Maier, Angus R. Simpson, Aaron C. Zecchin, Wai Kuan Foong, Kuang Yeow Phang, Hsin Yeow Seah, and Chan Lim Tan. + Ant Colony Optimization for Design of Water Distribution Systems. + Journal of Water Resources Planning and Management, ASCE, 129(3):200–209, May / June 2003.
+[ bib ] + +
+ + +
+[883] +
+
+Sri Srinivasa Raju M, Rammohan Mallipeddi, and Kedar Nath Das. + A twin-archive guided decomposition based multi/many-objective evolutionary algorithm. + Swarm and Evolutionary Computation, 71:101082, 2022.
+[ bib | +DOI ] + +
+ + +
+[884] +
+
+Katherine M. Malan and Andries Engelbrecht. + A survey of techniques for characterising fitness landscapes and some possible ways forward. + Information Sciences, 241:148–163, 2013.
+[ bib | +DOI ] + +
+ + +
+[885] +
+
+R. M. Males, R. M. Clark, P. J. Wehrman, and W. E. Gateset. + Algorithm for mixing problems in water systems. + Journal of Hydraulic Engineering, ASCE, 111(2):206–219, 1985.
+[ bib ] + +
+ + +
+[886] +
+
+Vittorio Maniezzo. + Exact and Approximate Nondeterministic Tree-Search Procedures for the Quadratic Assignment Problem. + INFORMS Journal on Computing, 11(4):358–369, 1999.
+[ bib ] + +
+ + +
+[887] +
+
+Vittorio Maniezzo and A. Carbonaro. + An ANTS Heuristic for the Frequency Assignment Problem. + Future Generation Computer Systems, 16(8):927–935, 2000.
+[ bib ] + +
+ + +
+[888] +
+
+Vittorio Maniezzo and Alberto Colorni. + The Ant System Applied to the Quadratic Assignment Problem. + IEEE Transactions on Knowledge and Data Engineering, 11(5):769–778, 1999.
+[ bib ] + +
+ + +
+[889] +
+
+E. Q. V. Martins. + On a multicritera shortest path problem. + European Journal of Operational Research, 16:236–245, 1984.
+[ bib ] + +
+ + +
+[890] +
+
+R. T. Marler and J. S. Arora. + Survey of multi-objective optimization methods for engineering. + Structural and Multidisciplinary Optimization, 26(6):369–395, April 2004.
+[ bib | +DOI ] +
+Discusses a priori (scalarized) methods. +
+ +
+ + +
+[891] +
+
+Raul Martín-Santamaría, Sergio Cavero, Alberto Herrán, Abraham Duarte, and J. Manuel Colmenar. + A Practical Methodology for Reproducible Experimentation: An Application to the Double-Row Facility Layout Problem. + Evolutionary Computation, pp.  1–35, 11 2023.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[892] +
+
+D. Martens, M. De Backer, R. Haesen, J. Vanthienen, M. Snoeck, and B. Baesens. + Classification With Ant Colony Optimization. + IEEE Transactions on Evolutionary Computation, 11(5):651–665, 2007.
+[ bib ] + +
+ + +
+[893] +
+
+Hugues Marchand, Alexander Martin, Robert Weismantel, and Laurence Wolsey. + Cutting planes in integer and mixed integer programming. + Discrete Applied Mathematics, 123(1–3):397–446, 2002.
+[ bib ] + +
+ + +
+[894] +
+
+O. Maron and A. W. Moore. + The Racing Algorithm: Model Selection for Lazy Learners. + Artificial Intelligence Research, 11(1–5):193–225, 1997.
+[ bib ] + +
+ + +
+[895] +
+
+Olivier Martin and S. W. Otto. + Partitioning of Unstructured Meshes for Load Balancing. + Concurrency: Practice and Experience, 7(4):303–314, 1995.
+[ bib ] + +
+ + +
+[896] +
+
+Olivier Martin and S. W. Otto. + Combining Simulated Annealing with Local Search Heuristics. + Annals of Operations Research, 63:57–75, 1996.
+[ bib ] + +
+ + +
+[897] +
+
+Olivier Martin, S. W. Otto, and E. W. Felten. + Large-Step Markov Chains for the Traveling Salesman Problem. + Complex Systems, 5(3):299–326, 1991.
+[ bib ] + +
+ + +
+[898] +
+
+Olivier Martin, S. W. Otto, and E. W. Felten. + Large-step Markov Chains for the TSP Incorporating Local Search Heuristics. + Operations Research Letters, 11(4):219–224, 1992.
+[ bib ] + +
+ + +
+[899] +
+
+Rafael Martí, Gerhard Reinelt, and Abraham Duarte. + A Benchmark Library and a Comparison of Heuristic Methods for the Linear Ordering Problem. + Computational Optimization and Applications, 51(3):1297–1317, 2012.
+[ bib ] + +
+ + +
+[900] +
+
+Raul Martín-Santamaría, Jesús Sánchez-Oro, S. Pérez-Peló, and Abraham Duarte. + Strategic oscillation for the balanced minimum sum-of-squares clustering problem. + Information Sciences, 585:529–542, 2022.
+[ bib | +DOI ] + +
+ + +
+[901] +
+
+Silvano Martello and Paolo Toth. + Lower bounds and reduction procedures for the bin packing problem. + Discrete Applied Mathematics, 28(1):59–70, 1990.
+[ bib | +DOI ] + +
+ + +
+[902] +
+
+Silvano Martello and Daniele Vigo. + Exact solution of the two-dimensional finite bin packing problem. + Management Science, 44(3):388–399, 1998.
+[ bib | +DOI ] + +
+ + +
+[903] +
+
+Abu S. Masud and C. L. Hwang. + Interactive Sequential Goal Programming. + Journal of the Operational Research Society, 32(5):391–400, May 1981.
+[ bib | +DOI ] +
+This paper introduces a new solution method based on Goal + Programming for Multiple Objective Decision Making (MODM) + problems. The method, called Interactive Sequential Goal + Programming (ISGP), combines and extends the attractive + features of both Goal Programming and interactive solution + approaches for MODM problems. ISGP is applicable to both + linear and non-linear problems. It uses existing single + objective optimization techniques and, hence, available + computer codes utilizing these techniques can be adapted for + use in ISGP. The non-dominance of the "best-compromise" + solution is assured. The information required from the + decision maker in each iteration is simple. The proposed + method is illustrated by solving a nutrition problem. +
+ +
+ + +
+[904] +
+
+Franco Mascia, Manuel López-Ibáñez, Jérémie Dubois-Lacoste, and Thomas Stützle. + Grammar-Based Generation of Stochastic Local Search Heuristics through Automatic Algorithm Configuration Tools. + Computers & Operations Research, 51:190–199, 2014.
+[ bib | +DOI ] + +
+ + +
+[905] +
+
+Franco Mascia, Paola Pellegrini, Thomas Stützle, and Mauro Birattari. + An Analysis of Parameter Adaptation in Reactive Tabu Search. + International Transactions in Operational Research, 21(1):127–152, 2014.
+[ bib ] + +
+ + +
+[906] +
+
+Renaud Masson, Thibaut Vidal, Julien Michallet, Puca Huachi Vaz Penna, Vinicius Petrucci, Anand Subramanian, and Hugues Dubedout. + An Iterated Local Search Heuristic for Multi-capacity Bin Packing and Machine Reassignment Problems. + Expert Systems with Applications, 40(13):5266–5275, 2013.
+[ bib ] + +
+ + +
+[907] +
+
+Yazid Mati, Stéphane Dauzère-Pèrés, and Chams Lahlou. + A General Approach for Optimizing Regular Criteria in the Job-shop Scheduling Problem. + European Journal of Operational Research, 212(1):33–42, 2011.
+[ bib ] + +
+ + +
+[908] +
+
+Atanu Mazumdar, Manuel López-Ibáñez, Tinkle Chugh, and Kaisa Miettinen. + Treed Gaussian Process Regression for Solving Offline Data-Driven Continuous Multiobjective Optimization Problems. + Evolutionary Computation, 31(4):375–399, 2023.
+[ bib | +DOI ] +
+For offline data-driven multiobjective optimization problems + (MOPs), no new data is available during the optimization + process. Approximation models (or surrogates) are first built + using the provided offline data and an optimizer, e.g. a + multiobjective evolutionary algorithm, can then be utilized + to find Pareto optimal solutions to the problem with + surrogates as objective functions. In contrast to online + data-driven MOPs, these surrogates cannot be updated with new + data and, hence, the approximation accuracy cannot be + improved by considering new data during the optimization + process. Gaussian process regression (GPR) models are widely + used as surrogates because of their ability to provide + uncertainty information. However, building GPRs becomes + computationally expensive when the size of the dataset is + large. Using sparse GPRs reduces the computational cost of + building the surrogates. However, sparse GPRs are not + tailored to solve offline data-driven MOPs, where good + accuracy of the surrogates is needed near Pareto optimal + solutions. Treed GPR (TGPR-MO) surrogates for offline + data-driven MOPs with continuous decision variables are + proposed in this paper. The proposed surrogates first split + the decision space into subregions using regression trees and + build GPRs sequentially in regions close to Pareto optimal + solutions in the decision space to accurately approximate + tradeoffs between the objective functions. TGPR-MO surrogates + are computationally inexpensive because GPRs are built only + in a smaller region of the decision space utilizing a subset + of the data. The TGPR-MO surrogates were tested on + distance-based visualizable problems with various data sizes, + sampling strategies, numbers of objective functions, and + decision variables. Experimental results showed that the + TGPR-MO surrogates are computationally cheaper and can handle + datasets of large size. Furthermore, TGPR-MO surrogates + produced solutions closer to Pareto optimal solutions + compared to full GPRs and sparse GPRs. +
+
+Keywords: Gaussian processes, Kriging, Regression trees, Metamodelling, + Surrogate, Pareto optimality +
+ +
+ + +
+[909] +
+
+Ross M. McConnell, Kurt Mehlhorn, Stefan Näher, and Pascal Schweitzer. + Certifying algorithms. + Computer Science Review, 5(2):119–161, 2011.
+[ bib | +DOI ] +
+A certifying algorithm is an algorithm that produces, with + each output, a certificate or witness (easy-to-verify proof) + that the particular output has not been compromised by a + bug. A user of a certifying algorithm inputs x, receives the + output y and the certificate w, and then checks, either + manually or by use of a program, that w proves that y is a + correct output for input x. In this way, he/she can be sure + of the correctness of the output without having to trust the + algorithm. We put forward the thesis that certifying + algorithms are much superior to non-certifying algorithms, + and that for complex algorithmic tasks, only certifying + algorithms are satisfactory. Acceptance of this thesis would + lead to a change of how algorithms are taught and how + algorithms are researched. The widespread use of certifying + algorithms would greatly enhance the reliability of + algorithmic software. We survey the state of the art in + certifying algorithms and add to it. In particular, we start + a theory of certifying algorithms and prove that the concept + is universal. +
+
+Keywords: Algorithms, Software reliability, Certification +
+ +
+ + +
+[910] +
+
+G. McCormick and R. S. Powell. + Optimal Pump Scheduling in Water Supply Systems with Maximum Demand Charges. + Journal of Water Resources Planning and Management, ASCE, 129(5):372–379, 2003.
+[ bib | +DOI | +epub ] +
+Keywords: water supply; pumps; Markov processes; cost optimal + control +
+ +
+ + +
+[911] +
+
+G. McCormick and R. S. Powell. + Derivation of near-optimal pump schedules for water distribution by simulated annealing. + Journal of the Operational Research Society, 55(7):728–736, July 2004.
+[ bib | +DOI ] +
+The scheduling of pumps for clean water distribution is a + partially discrete non-linear problem with many + variables. The scheduling method described in this paper + typically produces costs within 1% of a linear program-based + solution, and can incorporate realistic non-linear costs that + may be hard to incorporate in linear programming + formulations. These costs include pump switching and maximum + demand charges. A simplified model is derived from a standard + hydraulic simulator. An initial schedule is produced by a + descent method. Two-stage simulated annealing then produces + solutions in a few minutes. Iterative recalibration ensures + that the solution agrees closely with the results from a full + hydraulic simulation. +
+ +
+ + +
+[912] +
+
+James McDermott. + When and Why Metaheuristics Researchers can Ignore "No Free Lunch" Theorems. + SN Computer Science, 1(60):1–18, 2020.
+[ bib | +DOI ] + +
+ + +
+[913] +
+
+Catherine C. McGeoch. + Analyzing Algorithms by Simulation: Variance Reduction Techniques and Simulation Speedups. + ACM Computing Surveys, 24(2):195–212, 1992.
+[ bib | +DOI ] +
+Although experimental studies have been widely applied to the + investigation of algorithm performance, very little attention + has been given to experimental method in this area. This is + unfortunate, since much can be done to improve the quality of + the data obtained; often, much improvement may be needed for + the data to be useful. This paper gives a tutorial discussion + of two aspects of good experimental technique: the use of + variance reduction techniques and simulation speedups in + algorithm studies. In an illustrative study, application of + variance reduction techniques produces a decrease in variance + by a factor 1000 in one case, giving a dramatic improvement + in the precision of experimental results. Furthermore, the + complexity of the simulation program is improved from + Θ(m n/Hn) to Θ(m + n log n) (where m is + typically much larger than n), giving a much faster + simulation program and therefore more data per unit of + computation time. The general application of variance + reduction techniques is also discussed for a variety of + algorithm problem domains. +
+
+Keywords: experimental analysis of algorithms, move-to-front rule, + self-organizing sequential search, statistical analysis of + algorithms, transpose rule, variance reduction techniques +
+ +
+ + +
+[914] +
+
+Catherine C. McGeoch. + Toward an Experimental Method for Algorithm Simulation. + INFORMS Journal on Computing, 8(1):1–15, 1996.
+[ bib | +DOI ] + +
+ + +
+[915] +
+
+Michael D. McKay, Richard J. Beckman, and W. J. Conover. + A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code. + Technometrics, 21(2):239–245, 1979.
+[ bib | +DOI ] +
+Two types of sampling plans are examined as alternatives to + simple random sampling in Monte Carlo studies. These plans + are shown to be improvements over simple random sampling with + respect to variance for a class of estimators which includes + the sample mean and the empirical distribution function. +
+ +
+ + +
+[916] +
+
+Russell McKenna, Valentin Bertsch, Kai Mainzer, and Wolf Fichtner. + Combining local preferences with multi-criteria decision analysis and linear optimization to develop feasible energy concepts in small communities. + European Journal of Operational Research, 268(3):1092–1110, 2018.
+[ bib ] + +
+ + +
+[917] +
+
+Robert I. Mckay, Nguyen Xuan Hoai, Peter Alexander Whigham, Yin Shan, and Michael O'Neill. + Grammar-based Genetic Programming: A Survey. + Genetic Programming and Evolvable Machines, 11(3-4):365–396, September 2010.
+[ bib | +DOI ] + +
+ + +
+[918] +
+
+Klaus Meer. + Simulated annealing versus Metropolis for a TSP instance. + Information Processing Letters, 104(6):216–219, 2007.
+[ bib ] + +
+ + +
+[919] +
+
+Gábor Melis, Chris Dyer, and Phil Blunsom. + On the State of the Art of Evaluation in Neural Language Models. + Arxiv preprint arXiv:1807.02811, 2017.
+[ bib | +http ] + +
+ + +
+[920] +
+
+M. T. Melo, S. Nickel, and F. Saldanha-da Gama. + Facility location and supply chain management: A review. + European Journal of Operational Research, 196(2):401–412, 2009.
+[ bib | +DOI ] + +
+ + +
+[921] +
+
+Ole J. Mengshoel. + Understanding the role of noise in stochastic local search: Analysis and experiments. + Artificial Intelligence, 172(8):955–990, 2008.
+[ bib ] + +
+ + +
+[922] +
+
+Juan-Julián Merelo and Carlos Cotta. + Building bridges: the role of subfields in metaheuristics. + SIGEVOlution, 1(4):9–15, 2006.
+[ bib | +DOI ] + +
+ + +
+[923] +
+
+Peter Merz and Bernd Freisleben. + Memetic Algorithms for the Traveling Salesman Problem. + Complex Systems, 13(4):297–345, 2001.
+[ bib ] + +
+ + +
+[924] +
+
+Peter Merz and Bernd Freisleben. + Fitness Landscape Analysis and Memetic Algorithms for the Quadratic Assignment Problem. + IEEE Transactions on Evolutionary Computation, 4(4):337–352, 2000.
+[ bib ] + +
+ + +
+[925] +
+
+Peter Merz and Kengo Katayama. + Memetic algorithms for the unconstrained binary quadratic programming problem. + BioSystems, 78(1):99–118, 2004.
+[ bib | +DOI ] + +
+ + +
+[926] +
+
+D. Merkle and Martin Middendorf. + Ant Colony Optimization with Global Pheromone Evaluation for Scheduling a Single Machine. + Applied Intelligence, 18(1):105–111, 2003.
+[ bib ] + +
+ + +
+[927] +
+
+D. Merkle and Martin Middendorf. + Modeling the Dynamics of Ant Colony Optimization. + Evolutionary Computation, 10(3):235–262, 2002.
+[ bib ] + +
+ + +
+[928] +
+
+D. Merkle, Martin Middendorf, and Hartmut Schmeck. + Ant Colony Optimization for Resource-Constrained Project Scheduling. + IEEE Transactions on Evolutionary Computation, 6(4):333–346, 2002.
+[ bib ] + +
+ + +
+[929] +
+
+Peter Merz and Bernd Freisleben. + Greedy and Local Search Heuristics for Unconstrained Binary Quadratic Programming. + Journal of Heuristics, 8(2):197–213, 2002.
+[ bib | +DOI ] + +
+ + +
+[930] +
+
+Rafael G. Mesquita, Ricardo M. A. Silva, Carlos A. B. Mello, and Péricles B. C. Miranda. + Parameter tuning for document image binarization using a racing algorithm. + Expert Systems with Applications, 42(5):2593–2603, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[931] +
+
+N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. Teller, and E. Teller. + Equation of State Calculations by Fast Computing Machines. + Journal of Chemical Physics, 21:1087–1092, 1953.
+[ bib ] + +
+ + +
+[932] +
+
+Nicolas Meuleau and Marco Dorigo. + Ant Colony Optimization and Stochastic Gradient Descent. + Artificial Life, 8(2):103–121, 2002.
+[ bib ] + +
+ + +
+[933] +
+
+Laurent Meunier, Herilalaina Rakotoarison, Pak-Kan Wong, Baptiste Rozière, Jérémy Rapin, Olivier Teytaud, Antoine Moreau, and Carola Doerr. + Black-Box Optimization Revisited: Improving Algorithm Selection Wizards through Massive Benchmarking. + Arxiv preprint arXiv:2010.04542, 2020.
+[ bib | +DOI ] +
+Keywords: Nevergrad, NGOpt +
+ +
+ + +
+[934] +
+
+Laurent Meunier, Herilalaina Rakotoarison, Pak-Kan Wong, Baptiste Rozière, Jérémy Rapin, Olivier Teytaud, Antoine Moreau, and Carola Doerr. + Black-Box Optimization Revisited: Improving Algorithm Selection Wizards Through Massive Benchmarking. + IEEE Transactions on Evolutionary Computation, 26(3):490–500, 2022.
+[ bib | +DOI ] +
+Keywords: nevergrad, NGOpt +
+ +
+ + +
+[935] +
+
+R. M'Hallah. + An iterated local search variable neighborhood descent hybrid heuristic for the total earliness tardiness permutation flow shop. + International Journal of Production Research, 52(13):3802–3819, 2014.
+[ bib ] + +
+ + +
+[936] +
+
+Zbigniew Michalewicz, Dipankar Dasgupta, Rodolphe G. Le Riche, and Marc Schoenauer. + Evolutionary algorithms for constrained engineering problems. + Computers and Industrial Engineering, 30(4):851–870, 1996.
+[ bib | +DOI ] + +
+ + +
+[937] +
+
+Julien Michallet, Christian Prins, Farouk Yalaoui, and Grégoire Vitry. + Multi-start Iterated Local Search for the Periodic Vehicle Routing Problem with Time Windows and Time Spread Constraints on Services. + Computers & Operations Research, 41:196–207, 2014.
+[ bib ] + +
+ + +
+[938] +
+
+Kaisa Miettinen. + Survey of methods to visualize alternatives in multiple criteria decision making problems. + OR Spectrum, 36(1):3–37, 2014.
+[ bib | +DOI ] + +
+ + +
+[939] +
+
+Kaisa Miettinen, Jyri Mustajoki, and T. J. Stewart. + Interactive multiobjective optimization with NIMBUS for decision making under uncertainty. + OR Spectrum, 36(1):39–56, 2014.
+[ bib ] + +
+ + +
+[940] +
+
+R. B. Millar and M. J. Anderson. + Remedies for pseudoreplication. + Fisheries Research, 70(2–3):397–407, 2004.
+[ bib | +DOI ] + +
+ + +
+[941] +
+
+George A. Miller. + The magical number seven, plus or minus two: Some limits on our capacity for processing information. + Psychological Review, 63(2):81–97, 1956.
+[ bib | +DOI ] + +
+ + +
+[942] +
+
+Steven Minton. + Automatically configuring constraint satisfaction programs: A case study. + Constraints, 1(1):7–43, 1996.
+[ bib | +DOI ] + +
+ + +
+[943] +
+
+Gerardo Minella, Rubén Ruiz, and M. Ciavotta. + A Review and Evaluation of Multiobjective Algorithms for the Flowshop Scheduling Problem. + INFORMS Journal on Computing, 20(3):451–471, 2008.
+[ bib ] + +
+ + +
+[944] +
+
+Giovanni Misitano, Bekir Afsar, Giomara Larraga, and Kaisa Miettinen. + Towards explainable interactive multiobjective optimization: R-XIMO. + Autonomous Agents and Multi-Agent Systems, 36(42), 2022.
+[ bib | +DOI ] + +
+ + +
+[945] +
+
+Alfonsas Misevičius and Dovilė Kuznecovaitė. + Investigating some strategies for construction of initial populations in genetic algorithms. + Computational Science and Techniques, 5(1):560–573, 2018.
+[ bib ] + +
+ + +
+[946] +
+
+Alfonsas Misevičius. + Genetic Algorithm Hybridized with Ruin and Recreate Procedure: Application to the Quadratic Assignment Problem. + Knowledge-Based Systems, 16(5–6):261–268, 2003.
+[ bib ] + +
+ + +
+[947] +
+
+Alfonsas Misevičius. + A modified simulated annealing algorithm for the quadratic assignment problem. + Informatica, 14(4):497–514, 2003.
+[ bib ] + +
+ + +
+[948] +
+
+P. Mitra, C. A. Murthy, and S. K. Pal. + Unsupervised feature selection using feature similarity. + IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(3):301–312, 2002.
+[ bib | +DOI ] + +
+ + +
+[949] +
+
+Alfonsas Misevičius, Dovilė Kuznecovaitė, and Jūratė Platužienė. + Some Further Experiments with Crossover Operators for Genetic Algorithms. + Informatica, 29(3):499–516, 2018.
+[ bib ] + +
+ + +
+[950] +
+
+Nenad Mladenović and Pierre Hansen. + Variable Neighborhood Search. + Computers & Operations Research, 24(11):1097–1100, 1997.
+[ bib ] + +
+ + +
+[951] +
+
+Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Andrei A. Rusu, Joel Veness, Marc G. Bellemare, Alex Graves, Martin Riedmiller, Andreas K. Fidjeland, Georg Ostrovski, et al. + Human-level control through deep reinforcement learning. + Nature, 518(7540):529, 2015.
+[ bib ] + +
+ + +
+[952] +
+
+Jonas Močkus, Vytautas Tiesis, and Antanas Zilinskas. + The application of bayesian methods for seeking the extremum. + Towards global optimization, pp.  117–129, 1978.
+[ bib ] +
+Proposed Bayesian optimization (but later than + [2276]) +
+ +
+ + +
+[953] +
+
+Julián Molina, Luis V. Santana, Alfredo G. Hernández-Díaz, Carlos A. Coello Coello, and Rafael Caballero. + g-Dominance: Reference point based dominance for Multiobjective Metaheuristics. + European Journal of Operational Research, 197(2):685–692, September 2009.
+[ bib | +DOI ] +
+Proposed g-NSGA-II +
+ +
+ + +
+[954] +
+
+Marco A. Montes de Oca, Doǧan Aydın, and Thomas Stützle. + An Incremental Particle Swarm for Large-Scale Continuous Optimization Problems: An Example of Tuning-in-the-loop (Re)Design of Optimization Algorithms. + Soft Computing, 15(11):2233–2255, 2011.
+[ bib | +DOI ] + +
+ + +
+[955] +
+
+Alysson Mondoro, Dan M. Frangopol, and Liang Liu. + Multi-criteria robust optimization framework for bridge adaptation under climate change. + Structural Safety, 74:14–23, 2018.
+[ bib ] + +
+ + +
+[956] +
+
+Roberto Montemanni, L. M. Gambardella, A. E. Rizzoli, and A. V. Donati. + Ant colony system for a dynamic vehicle routing problem. + Journal of Combinatorial Optimization, 10:327–343, 2005.
+[ bib ] + +
+ + +
+[957] +
+
+James Montgomery, Marcus Randall, and Tim Hendtlass. + Solution bias in ant colony optimisation: Lessons for selecting pheromone models. + Computers & Operations Research, 35(9):2728–2749, 2008.
+[ bib | +DOI ] + +
+ + +
+[958] +
+
+Elizabeth Montero, María-Cristina Riff, and Bertrand Neveu. + A Beginner's Buide to Tuning Methods. + Applied Soft Computing, 17:39–51, 2014.
+[ bib | +DOI ] + +
+ + +
+[959] +
+
+Marco A. Montes de Oca, Thomas Stützle, Mauro Birattari, and Marco Dorigo. + Frankenstein's PSO: A Composite Particle Swarm Optimization Algorithm. + IEEE Transactions on Evolutionary Computation, 13(5):1120–1132, 2009.
+[ bib | +DOI ] + +
+ + +
+[960] +
+
+Nicolas Monmarché, G. Venturini, and M. Slimane. + On how pachycondyla apicalis ants suggest a new search algorithm. + Future Generation Computer Systems, 16(8):937–946, 2000.
+[ bib ] + +
+ + +
+[961] +
+
+Peter D. Morgan. + Simulation of an adaptive behavior mechanism in an expert decision-maker. + IEEE Transactions on Systems, Man, and Cybernetics, 23(1):65–76, 1993.
+[ bib ] + +
+ + +
+[962] +
+
+J. N. Morse. + Reducing the size of the nondominated set: Pruning by clustering. + Computers & Operations Research, 7(1-2):55–66, 1980.
+[ bib ] + +
+ + +
+[963] +
+
+Mouad Morabit, Guy Desaulniers, and Andrea Lodi. + Machine-learning–based column selection for column generation. + Transportation Science, 55(4):815–831, 2021.
+[ bib ] +
+Keywords: graph neural networks +
+ +
+ + +
+[964] +
+
+Sara Morin, Caroline Gagné, and Marc Gravel. + Ant colony optimization with a specialized pheromone trail for the car-sequencing problem. + European Journal of Operational Research, 197(3):1185–1191, 2009.
+[ bib | +DOI ] +
+ This paper studies the learning process in an ant + colony optimization algorithm designed to solve the + problem of ordering cars on an assembly line + (car-sequencing problem). This problem has been + shown to be NP-hard and evokes a great deal of + interest among practitioners. Learning in an ant + algorithm is achieved by using an artificial + pheromone trail, which is a central element of this + metaheuristic. Many versions of the algorithm are + found in literature, the main distinction among them + being the management of the pheromone + trail. Nevertheless, few of them seek to perfect + learning by modifying the internal structure of the + trail. In this paper, a new pheromone trail + structure is proposed that is specifically adapted + to the type of constraints in the car-sequencing + problem. The quality of the results obtained when + solving three sets of benchmark problems is superior + to that of the best solutions found in literature + and shows the efficiency of the specialized trail. +
+
+Keywords: Ant colony optimization,Car-sequencing + problem,Pheromone trail,Scheduling +
+ +
+ + +
+[965] +
+
+A. M. Mora, Juan-Julián Merelo, Juan Luis Jiménez Laredo, C. Millan, and J. Torrecillas. + CHAC, a MOACO algorithm for computation of bi-criteria military unit path in the battlefield: Presentation and first results. + International Journal of Intelligent Systems, 24(7):818–843, 2009.
+[ bib ] + +
+ + +
+[966] +
+
+Max D. Morris and Toby J. Mitchell. + Exploratory designs for computational experiments. + Journal of Statistical Planning and Inference, 43(3):381–402, 1995.
+[ bib | +DOI ] +
+Keywords: Bayesian prediction +
+ +
+ + +
+[967] +
+
+Pablo Moscato and José F. Fontanari. + Stochastic Versus Deterministic Update in Simulated Annealing. + Physics Letters A, 146(4):204–208, 1990.
+[ bib ] + +
+ + +
+[968] +
+
+John Mote, Ishwar Murthy, and David L. Olson. + A parametric approach to solving bicriterion shortest path problems. + European Journal of Operational Research, 53(1):81–92, 1991.
+[ bib | +DOI ] + +
+ + +
+[969] +
+
+John Mote, David L. Olson, and M. A. Venkataramanan. + A comparative multiobjective programming study. + Mathematical and Computer Modelling, 10(10):719–729, 1988.
+[ bib | +DOI ] +
+The purpose of this study was to systematically evaluate a + number of multiobjective programming concepts relative to + reflection of utility, assurance of nondominated solutions + and practicality for larger problems using conventional + software. In the problem used, the nonlinear simulated DM + utility function applied resulted in a nonextreme point + solution. Very often, the preferred solution could end up + being an extreme point solution, in which case the techniques + relying upon LP concepts would work as well if not better + than utilizing constrained objective attainments. The point + is that there is no reason to expect linear or near linear + utility. +
+
+Keywords: artificial DM, interactive +
+ +
+ + +
+[970] +
+
+Sébastien Mouthuy, Yves Deville, and Pascal van Hentenryck. + Constraint-based Very Large-Scale Neighborhood Search. + Constraints, 17(2):87–122, 2012.
+[ bib | +DOI ] + +
+ + +
+[971] +
+
+Lucien Mousin, Marie-Eléonore Kessaci, and Clarisse Dhaenens. + Exploiting Promising Sub-Sequences of Jobs to solve the No-Wait Flowshop Scheduling Problem. + Arxiv preprint arXiv:1903.09035, 2019.
+[ bib | +http ] + +
+ + +
+[972] +
+
+Noura Al Moubayed, Andrei Petrovski, and John McCall. + D2MOPSO: MOPSO based on decomposition and dominance with archiving using crowding distance in objective and solution spaces. + Evolutionary Computation, 22(1):47–77, 2014.
+[ bib ] + +
+ + +
+[973] +
+
+Vincent Mousseau and Roman Slowiński. + Inferring an ELECTRE TRI model from assignment examples. + Journal of Global Optimization, 12(2):157–174, 1998.
+[ bib ] + +
+ + +
+[974] +
+
+Christian L. Müller and Ivos F. Sbalzarini. + Energy Landscapes of Atomic Clusters as Black Box Optimization Benchmarks. + Evolutionary Computation, 20(4):543–573, 2012.
+[ bib | +DOI ] + +
+ + +
+[975] +
+
+H. Mühlenbein and D. Schlierkamp-Voosen. + Predictive models for the breeder genetic algorithm. + Evolutionary Computation, 1(1):25–49, 1993.
+[ bib ] +
+Keywords: crossover, intermediate, line +
+ +
+ + +
+[976] +
+
+Mario A. Muñoz and Kate Smith-Miles. + Generating New Space-Filling Test Instances for Continuous Black-Box Optimization. + Evolutionary Computation, 28(3):379–404, September 2020.
+[ bib | +DOI ] + +
+ + +
+[977] +
+
+Mario A. Muñoz, Yuan Sun, Michael Kirley, and Saman K. Halgamuge. + Algorithm selection for black-box continuous optimization problems: a survey on methods and challenges. + Information Sciences, 317:224–245, 2015.
+[ bib ] + +
+ + +
+[978] +
+
+Mario A. Muñoz, Laura Villanova, Davaatseren Baatar, and Kate Smith-Miles. + Instance Spaces for Machine Learning Classification. + Machine Learning, 107(1):109–147, 2018.
+[ bib | +DOI ] + +
+ + +
+[979] +
+
+Yuichi Nagata and Shigenobu Kobayashi. + A Powerful Genetic Algorithm Using Edge Assembly Crossover for the Traveling Salesman Problem. + INFORMS Journal on Computing, 25(2):346–363, 2013.
+[ bib | +DOI ] +
+This paper presents a genetic algorithm (GA) for solving the + traveling salesman problem (TSP). To construct a powerful GA, + we use edge assembly crossover (EAX) and make substantial + enhancements to it: (i) localization of EAX together with its + efficient implementation and (ii) the use of a local search + procedure in EAX to determine good combinations of building + blocks of parent solutions for generating even better + offspring solutions from very high-quality parent + solutions. In addition, we develop (iii) an innovative + selection model for maintaining population diversity at a + negligible computational cost. Experimental results on + well-studied TSP benchmarks demonstrate that the proposed GA + outperforms state-of-the-art heuristic algorithms in finding + very high-quality solutions on instances with up to 200,000 + cities. In contrast to the state-of-the-art TSP heuristics, + which are all based on the Lin-Kernighan (LK) algorithm, our + GA achieves top performance without using an LK-based + algorithm. +
+
+Keywords: TSP, EAX, evolutionary algorithms +
+ +
+ + +
+[980] +
+
+Marcelo S. Nagano, Fernando L. Rossi, and Nádia J. Martarelli. + High-performing heuristics to minimize flowtime in no-idle permutation flowshop. + Engineering Optimization, 51(2):185–198, 2019.
+[ bib ] + +
+ + +
+[981] +
+
+Yuichi Nagata and David Soler. + A New Genetic Algorithm for the Asymmetric TSP. + Expert Systems with Applications, 39(10):8947–8953, 2012.
+[ bib ] + +
+ + +
+[982] +
+
+Samadhi Nallaperuma, Pietro S. Oliveto, Jorge Pérez Heredia, and Dirk Sudholt. + On the Analysis of Trajectory-Based Search Algorithms: When is it Beneficial to Reject Improvements? + Algorithmica, 81(2):858–885, 2019.
+[ bib ] + +
+ + +
+[983] +
+
+Yang Nan, Ke Shang, Hisao Ishibuchi, and Linjun He. + Reverse strategy for non-dominated archiving. + IEEE Access, 8:119458–119469, 2020.
+[ bib ] + +
+ + +
+[984] +
+
+Kaname Narukawa, Yu Setoguchi, Yuki Tanigaki, Markus Olhofer, Bernhard Sendhoff, and Hisao Ishibuchi. + Preference representation using Gaussian functions on a hyperplane in evolutionary multi-objective optimization. + Soft Computing, 20(7):2733–2757, July 2016.
+[ bib | +DOI ] + +
+ + +
+[985] +
+
+John Nash and Ravi Varadhan. + Unifying Optimization Algorithms to Aid Software System Users: optimx for R. + Journal of Statistical Software, 43(9):1–14, 2011.
+[ bib ] + +
+ + +
+[986] +
+
+M. Nawaz, E. Enscore, Jr, and I. Ham. + A Heuristic Algorithm for the m-Machine, n-Job Flow-Shop Sequencing Problem. + Omega, 11(1):91–95, 1983.
+[ bib ] +
+Keywords: NEH heuristic +
+ +
+ + +
+[987] +
+
+Antonio J. Nebro, Manuel López-Ibáñez, José García-Nieto, and Carlos A. Coello Coello. + On the automatic design of multi-objective particle swarm optimizers: experimentation and analysis. + Swarm Intelligence, 2023.
+[ bib | +DOI ] +
+Research in multi-objective particle swarm optimizers + (MOPSOs) progresses by proposing one new MOPSO at a time. In + spite of the commonalities among different MOPSOs, it is + often unclear which algorithmic components are crucial for + explaining the performance of a particular MOPSO + design. Moreover, it is expected that different designs may + perform best on different problem families and identifying a + best overall MOPSO is a challenging task. We tackle this + challenge here by: (1) proposing AutoMOPSO, a flexible + algorithmic template for designing MOPSOs with a design space + that can instantiate thousands of potential MOPSOs; and (2) + searching for good-performing MOPSO designs given a family of + training problems by means of an automatic configuration tool + (irace). We apply this automatic design methodology to + generate a MOPSO that significantly outperforms two + state-of-the-art MOPSOs on four well-known bi-objective + problem families. We also identify the key design choices and + parameters of the winning MOPSO by means of + ablation. AutoMOPSO is publicly available as part of the + jMetal framework. +
+ +
+ + +
+[988] +
+
+Antonio J. Nebro, F. Luna, Enrique Alba, Bernabé Dorronsoro, Juan J. Durillo, and A. Beham. + AbYSS: Adapting Scatter Search to Multiobjective Optimization. + IEEE Transactions on Evolutionary Computation, 12(4):439–457, August 2008.
+[ bib | +DOI ] + +
+ + +
+[989] +
+
+F. Nerri and Carlos Cotta. + Memetic algorithms and memetic computing optimization: A literature review. + Swarm and Evolutionary Computation, 2:1–14, 2012.
+[ bib | +DOI ] + +
+ + +
+[990] +
+
+Frank Neumann, Dirk Sudholt, and Carsten Witt. + Analysis of different MMAS ACO algorithms on unimodal functions and plateaus. + Swarm Intelligence, 3(1):35–68, 2009.
+[ bib ] + +
+ + +
+[991] +
+
+Frank Neumann and Carsten Witt. + Runtime Analysis of a Simple Ant Colony Optimization Algorithm. + Electronic Colloquium on Computational Complexity (ECCC), 13(084), 2006.
+[ bib ] + +
+ + +
+[992] +
+
+Allen Newell and Herbert A. Simon. + Computer Science as Empirical Inquiry: Symbols and Search. + Communications of the ACM, 19(3):113–126, March 1976.
+[ bib | +DOI ] +
+Computer science is the study of the phenomena surrounding + computers. The founders of this society understood this very + well when they called themselves the Association for + Computing Machinery. The machine-not just the hardware, but + the programmed, living machine-is the organism we study. +
+
+Keywords: cognition, Turing, search, problem solving, symbols, + heuristics, list processing, computer science, artificial + intelligence, science, empirical +
+ +
+ + +
+[993] +
+
+Viet-Phuong Nguyen, Christian Prins, and Caroline Prodhon. + A Multi-start Iterated Local Search with Tabu List and Path Relinking for the Two-echelon Location-routing Problem. + Engineering Applications of Artificial Intelligence, 25(1):56–71, 2012.
+[ bib ] + +
+ + +
+[994] +
+
+Anh-Tuan Nguyen, Sigrid Reiter, and Philippe Rigo. + A review on simulation-based optimization methods applied to building performance analysis. + Applied Energy, 113:1043–1058, 2014.
+[ bib | +DOI ] + +
+ + +
+[995] +
+
+Trung Thanh Nguyen, Shengxiang Yang, and Jürgen Branke. + Evolutionary Dynamic Optimization: A Survey of the State of the Art. + Swarm and Evolutionary Computation, 6:1–24, 2012.
+[ bib ] + +
+ + +
+[996] +
+
+Su Nguyen, Mengjie Zhang, Mark Johnston, and Kay Chen Tan. + Genetic Programming for Evolving Due-Date Assignment Models in Job Shop Environments. + Evolutionary Computation, 22(1):105–138, 2014.
+[ bib ] + +
+ + +
+[997] +
+
+Su Nguyen, Mengjie Zhang, Mark Johnston, and Kay Chen Tan. + Automatic Design of Scheduling Policies for Dynamic Multi-objective Job Shop Scheduling via Cooperative Coevolution Genetic Programming. + IEEE Transactions on Evolutionary Computation, 18(2):193–208, 2014.
+[ bib ] + +
+ + +
+[998] +
+
+Peter Nightingale, Özguür Akgün, Ian P. Gent, Christopher Jefferson, Ian Miguel, and Patrick Spracklen. + Automatically Improving Constraint Models in Savile Row. + Artificial Intelligence, 251:35–61, 2017.
+[ bib ] + +
+ + +
+[999] +
+
+Chao Ning and Fengqi You. + Optimization under uncertainty in the era of big data and deep learning: When machine learning meets mathematical programming. + Computers & Chemical Engineering, 125:434–448, 2019.
+[ bib | +DOI ] +
+This paper reviews recent advances in the field of + optimization under uncertainty via a modern data lens, + highlights key research challenges and promise of data-driven + optimization that organically integrates machine learning and + mathematical programming for decision-making under + uncertainty, and identifies potential research + opportunities. A brief review of classical mathematical + programming techniques for hedging against uncertainty is + first presented, along with their wide spectrum of + applications in Process Systems Engineering. A comprehensive + review and classification of the relevant publications on + data-driven distributionally robust optimization, data-driven + chance constrained program, data-driven robust optimization, + and data-driven scenario-based optimization is then + presented. This paper also identifies fertile avenues for + future research that focuses on a closed-loop data-driven + optimization framework, which allows the feedback from + mathematical programming to machine learning, as well as + scenario-based optimization leveraging the power of deep + learning techniques. Perspectives on online learning-based + data-driven multistage optimization with a + learning-while-optimizing scheme are presented. +
+
+Keywords: Data-driven optimization, Decision making under uncertainty, + Big data, Machine learning, Deep learning +
+ +
+ + +
+[1000] +
+
+Naoki Nishimura, Kotaro Tanahashi, Koji Suganuma, Masamichi J. Miyama, and Masayuki Ohzeki. + Item listing optimization for e-commerce websites based on diversity. + Frontiers in Computer Science, 1:2, 2019.
+[ bib ] +
+Keywords: Quantum Annealing +
+ +
+ + +
+[1001] +
+
+Vilas Nitivattananon, Elaine C. Sadowski, and Rafael G. Quimpo. + Optimization of Water Supply System Operation. + Journal of Water Resources Planning and Management, ASCE, 122(5):374–384, September / October 1996.
+[ bib ] + +
+ + +
+[1002] +
+
+Bruno Nogueira, Rian G. S. Pinheiro, and Anand Subramanian. + A Hybrid Iterated Local Search Heuristic for the Maximum Weight Independent Set Problem. + Optimization Letters, 12(3):567–583, 2018.
+[ bib | +DOI ] + +
+ + +
+[1003] +
+
+B. A. Nosek, G. Alter, G. C. Banks, D. Borsboom, S. D. Bowman, S. J. Breckler, S. Buck, C. D. Chambers, G. Chin, G. Christensen, M. Contestabile, A. Dafoe, E. Eich, J. Freese, R. Glennerster, D. Goroff, D. P. Green, B. Hesse, M. Humphreys, J. Ishiyama, D. Karlan, A. Kraut, A. Lupia, P. Mabry, T. Madon, N. Malhotra, E. Mayo-Wilson, M. McNutt, E. Miguel, E. L. Paluck, U. Simonsohn, C. Soderberg, B. A. Spellman, J. Turitto, G. VandenBos, S. Vazire, E. J. Wagenmakers, R. Wilson, and T. Yarkoni. + Promoting an open research culture. + Science, 348(6242):1422–1425, June 2015.
+[ bib | +DOI ] + +
+ + +
+[1004] +
+
+Brian A. Nosek, Charles R. Ebersole, Alexander C. DeHaven, and David T. Mellor. + The Preregistration Revolution. + Proceedings of the National Academy of Sciences, 115(11):2600–2606, March 2018.
+[ bib | +DOI ] +
+Progress in science relies in part on generating hypotheses + with existing observations and testing hypotheses with new + observations. This distinction between postdiction and + prediction is appreciated conceptually but is not respected + in practice. Mistaking generation of postdictions with + testing of predictions reduces the credibility of research + findings. However, ordinary biases in human reasoning, such + as hindsight bias, make it hard to avoid this mistake. An + effective solution is to define the research questions and + analysis plan before observing the research outcomes–a + process called preregistration. Preregistration distinguishes + analyses and outcomes that result from predictions from those + that result from postdictions. A variety of practical + strategies are available to make the best possible use of + preregistration in circumstances that fall short of the ideal + application, such as when the data are preexisting. Services + are now available for preregistration across all disciplines, + facilitating a rapid increase in the practice. Widespread + adoption of preregistration will increase distinctiveness + between hypothesis generation and hypothesis testing and will + improve the credibility of research findings. +
+ +
+ + +
+[1005] +
+
+Yaghout Nourani and Bjarne Andresen. + A Comparison of Simulated Annealing Cooling Strategies. + Journal of Physics A, 31(41):8373–8385, 1998.
+[ bib ] + +
+ + +
+[1006] +
+
+Eugeniusz Nowicki and Czeslaw Smutnicki. + A Fast Taboo Search Algorithm for the Job Shop Problem. + Management Science, 42(6):797–813, 1996.
+[ bib ] + +
+ + +
+[1007] +
+
+Eugeniusz Nowicki and Czeslaw Smutnicki. + A fast tabu search algorithm for the permutation flow-shop problem. + European Journal of Operational Research, 91(1):160–175, 1996.
+[ bib ] + +
+ + +
+[1008] +
+
+Open Science Collaboration. + Estimating the reproducibility of psychological science. + Science, 349(6251):aac4716, 2015.
+[ bib | +DOI ] + +
+ + +
+[1009] +
+
+Gabriela Ochoa and Nadarajen Veerapen. + Mapping the global structure of TSP fitness landscapes. + Journal of Heuristics, 24(3):265–294, 2018.
+[ bib ] + +
+ + +
+[1010] +
+
+Angelo Oddi, Amadeo Cesta, Nicola Policella, and Stephen F. Smith. + Combining Variants of Iterative Flattening Search. + Engineering Applications of Artificial Intelligence, 21(5):683–690, 2008.
+[ bib ] + +
+ + +
+[1011] +
+
+Angelo Oddi, Amadeo Cesta, Nicola Policella, and Stephen F. Smith. + Iterative Flattening Search for Resource Constrained Scheduling. + Journal of Intelligent Manufacturing, 21(1):17–30, 2010.
+[ bib ] + +
+ + +
+[1012] +
+
+F. A. Ogbu and David K. Smith. + The Application of the Simulated Annealing Algorithm to the Solution of the n/m/C Max Flowshop Problem. + Computers & Operations Research, 17(3):243–253, 1990.
+[ bib ] + +
+ + +
+[1013] +
+
+Jeffrey W. Ohlmann and Barrett W. Thomas. + A Compressed-Annealing Heuristic for the Traveling Salesman Problem with Time Windows. + INFORMS Journal on Computing, 19(1):80–90, 2007.
+[ bib | +DOI ] + +
+ + +
+[1014] +
+
+Pietro S. Oliveto, Jun He, and Xin Yao. + Time complexity of evolutionary algorithms for combinatorial optimization: A decade of results. + International Journal of Automation and Computing, 4(3):281–293, 2007.
+[ bib ] + +
+ + +
+[1015] +
+
+Pietro S. Oliveto and Carsten Witt. + Improved time complexity analysis of the Simple Genetic Algorithm. + Theoretical Computer Science, 605:21–41, 2015.
+[ bib | +DOI ] + +
+ + +
+[1016] +
+
+David L. Olson. + Review of Empirical Studies in Multiobjective Mathematical Programming: Subject Reflection of Nonlinear Utility and Learning. + Decision Sciences, 23(1):1–20, 1992.
+[ bib | +DOI ] +
+Multiple objective programming provides a means of + aiding decision makers facing complex decisions where + trade-offs among conflicting objectives must be + reconciled. Interactive multiobjective programming provides a + means for decision makers to learn what these trade-offs + involve, while the mathematical program generates solutions + that seek improvement of the implied utility of the decision + maker. A variety of multiobjective programming techniques + have been presented in the multicriteria decision-making + literature. This study reviews published studies with human + subjects where some of these techniques were applied. While + all of the techniques have the ability to support decision + makers under conditions of multiple objectives, a number of + features in applying these systems have been tested by these + studies. A general evolution of techniques is traced, + starting with methods relying upon linear combinations of + value, to more recent methods capable of reflecting nonlinear + trade-offs of value. Support of nonlinear utility and + enhancing decision-maker learning are considered. +
+
+Keywords: Decision Analysis, Human Information Processing, Linear + Programming +
+ +
+ + +
+[1017] +
+
+Roland Olsson and Arne Løkketangen. + Using Automatic Programming to Generate State-of-the-art Algorithms for Random 3-SAT. + Journal of Heuristics, 19(5):819–844, 2013.
+[ bib | +DOI ] +
+Uses evolution but it is not genetic programming, nor + grammatical evolution. +
+ +
+ + +
+[1018] +
+
+Mihai Oltean. + Evolving Evolutionary Algorithms Using Linear Genetic Programming. + Evolutionary Computation, 13(3):387–410, 2005.
+[ bib | +DOI ] + +
+ + +
+[1019] +
+
+Michael O'Neill and Conor Ryan. + Grammatical Evolution. + IEEE Transactions on Evolutionary Computation, 5(4):349–358, 2001.
+[ bib ] + +
+ + +
+[1020] +
+
+Lindell E. Ormsbee, Thomas M. Walski, Donald V. Chase, and W. W. Sharp. + Methodology for improving pump operation efficiency. + Journal of Water Resources Planning and Management, ASCE, 115(2):148–164, 1989.
+[ bib ] + +
+ + +
+[1021] +
+
+Lindell E. Ormsbee and Kevin E. Lansey. + Optimal Control of Water Supply Pumping Systems. + Journal of Water Resources Planning and Management, ASCE, 120(2):237–252, 1994.
+[ bib ] + +
+ + +
+[1022] +
+
+Lindell E. Ormsbee and Srinivasa L. Reddy. + Nonlinear Heuristic for Pump Operations. + Journal of Water Resources Planning and Management, ASCE, 121(4):302–309, July / August 1995.
+[ bib ] + +
+ + +
+[1023] +
+
+Jeffrey E. Orosz and Sheldon H. Jacobson. + Analysis of Static Simulated Annealing Algorithms. + Journal of Optimization Theory and Applications, 115(1):165–182, 2002.
+[ bib ] + +
+ + +
+[1024] +
+
+Ibrahim H. Osman and Chris N. Potts. + Simulated Annealing for Permutation Flow-Shop Scheduling. + Omega, 17(6):551–557, 1989.
+[ bib ] + +
+ + +
+[1025] +
+
+P. S. Ow and T. E. Morton. + Filtered Beam Search in Scheduling. + International Journal of Production Research, 26:297–307, 1988.
+[ bib ] + +
+ + +
+[1026] +
+
+Gül Özerol and Esra Karasakal. + Interactive outranking approaches for multicriteria decision-making problems with imprecise information. + Journal of the Operational Research Society, 59:1253–1268, 2007.
+[ bib ] + +
+ + +
+[1027] +
+
+Manfred Padberg and Giovanni Rinaldi. + A branch-and-cut algorithm for the resolution of large-scale symmetric traveling salesman problems. + SIAM Review, 33(1):60–100, 1991.
+[ bib ] + +
+ + +
+[1028] +
+
+Federico Pagnozzi and Thomas Stützle. + Speeding up Local Search for the Insert Neighborhood in the Weighted Tardiness Permutation Flowshop Problem. + Optimization Letters, 11:1283–1292, 2017.
+[ bib | +DOI ] + +
+ + +
+[1029] +
+
+Federico Pagnozzi and Thomas Stützle. + Automatic Design of Hybrid Stochastic Local Search Algorithms for Permutation Flowshop Problems. + European Journal of Operational Research, 276:409–421, 2019.
+[ bib | +DOI ] +
+Stochastic local search methods are at the core of many + effective heuristics for tackling different permutation + flowshop problems (PFSPs). Usually, such algorithms require a + careful, manual algorithm engineering effort to reach high + performance. An alternative to the manual algorithm + engineering is the automated design of effective SLS + algorithms through building flexible algorithm frameworks and + using automatic algorithm configuration techniques to + instantiate high-performing algorithms. In this paper, we + automatically generate new high-performing algorithms for + some of the most widely studied variants of the PFSP. More in + detail, we (i) developed a new algorithm framework, EMILI, + that implements algorithm-specific and problem-specific + building blocks; (ii) define the rules of how to compose + algorithms from the building blocks; and (iii) employ an + automatic algorithm configuration tool to search for high + performing algorithm configurations. With these ingredients, + we automatically generate algorithms for the PFSP with the + objectives makespan, total completion time and total + tardiness, which outperform the best algorithms obtained by a + manual algorithm engineering process. +
+
+Keywords: EMILI +
+ +
+ + +
+[1030] +
+
+Federico Pagnozzi and Thomas Stützle. + Evaluating the impact of grammar complexity in automatic algorithm design. + International Transactions in Operational Research, pp.  1–26, 2020.
+[ bib | +DOI ] + +
+ + +
+[1031] +
+
+Federico Pagnozzi and Thomas Stützle. + Automatic design of hybrid stochastic local search algorithms for permutation flowshop problems with additional constraints. + Operations Research Perspectives, 8:100180, 2021.
+[ bib | +DOI ] +
+Automatic design of stochastic local search algorithms has + been shown to be very effective in generating algorithms for + the permutation flowshop problem for the most studied + objectives including makespan, flowtime and total + tardiness. The automatic design system uses a configuration + tool to combine algorithmic components following a set of + rules defined as a context-free grammar. In this paper we use + the same system to tackle two of the most studied additional + constraints for these objectives: sequence dependent setup + times and no-idle constraint. Additional components have been + added to adapt the system to the new problems while keeping + intact the grammar structure and the experimental setup. The + experiments show that the generated algorithms outperform the + state of the art in each case. +
+ +
+ + +
+[1032] +
+
+Alberto Pajares, Xavier Blasco, Juan Manuel Herrero, and Miguel A. Martínez. + A Comparison of Archiving Strategies for Characterization of Nearly Optimal Solutions under Multi-Objective Optimization. + Mathematics, 9(9):999, 2021.
+[ bib | +DOI ] +
+In a multi-objective optimization problem, in addition to + optimal solutions, multimodal and/or nearly optimal + alternatives can also provide additional useful information + for the decision maker. However, obtaining all nearly optimal + solutions entails an excessive number of + alternatives. Therefore, to consider the nearly optimal + solutions, it is convenient to obtain a reduced set, putting + the focus on the potentially useful alternatives. These + solutions are the alternatives that are close to the optimal + solutions in objective space, but which differ significantly + in the decision space. To characterize this set, it is + essential to simultaneously analyze the decision and + objective spaces. One of the crucial points in an + evolutionary multi-objective optimization algorithm is the + archiving strategy. This is in charge of keeping the solution + set, called the archive, updated during the optimization + process. The motivation of this work is to analyze the three + existing archiving strategies proposed in the literature + (ArchiveUpdatePQ,εDxy, Archive_nevMOGA, and + targetSelect) that aim to characterize the potentially useful + solutions. The archivers are evaluated on two benchmarks and + in a real engineering example. The contribution clearly shows + the main differences between the three archivers. This + analysis is useful for the design of evolutionary algorithms + that consider nearly optimal solutions. +
+
+Keywords: multi-objective optimization; nearly optimal solutions; + non-epsilon dominance; multimodality; decision space + diversity; archiving strategy; evolutionary algorithm; + non-linear parametric identification +
+ +
+ + +
+[1033] +
+
+Daniel Palhazi Cuervo, Peter Goos, Kenneth Sörensen, and Emely Arráiz. + An Iterated Local Search Algorithm for the Vehicle Routing Problem with Backhauls. + European Journal of Operational Research, 237(2):454–464, 2014.
+[ bib ] + +
+ + +
+[1034] +
+
+Gintaras Palubeckis. + Iterated tabu search for the unconstrained binary quadratic optimization problem. + Informatica, 17(2):279–296, 2006.
+[ bib | +DOI ] + +
+ + +
+[1035] +
+
+Quan-Ke Pan and Rubén Ruiz. + Local Search Methods for the Flowshop Scheduling Problem with Flowtime Minimization. + European Journal of Operational Research, 222(1):31–43, 2012.
+[ bib ] + +
+ + +
+[1036] +
+
+Quan-Ke Pan and Rubén Ruiz. + A Comprehensive Review and Evaluation of Permutation Flowshop Heuristics to Minimize Flowtime. + Computers & Operations Research, 40(1):117–128, 2013.
+[ bib ] + +
+ + +
+[1037] +
+
+Quan-Ke Pan, Rubén Ruiz, and Pedro Alfaro-Fernández. + Iterated Search Methods for Earliness and Tardiness Minimization in Hybrid Flowshops with Due Windows. + Computers & Operations Research, 80:50–60, 2017.
+[ bib ] + +
+ + +
+[1038] +
+
+Quan-Ke Pan, Mehmet Fatih Tasgetiren, and Yun-Chia Liang. + A Discrete Differential Evolution Algorithm for the Permutation Flowshop Scheduling Problem. + Computers and Industrial Engineering, 55(4):795 – 816, 2008.
+[ bib ] + +
+ + +
+[1039] +
+
+Quan-Ke Pan, Ling Wang, and Bao-Hua Zhao. + An improved iterated greedy algorithm for the no-wait flow shop scheduling problem with makespan criterion. + International Journal of Advanced Manufacturing Technology, 38(7-8):778–786, 2008.
+[ bib ] + +
+ + +
+[1040] +
+
+Sinno Jialin Pan and Qiang Yang. + A survey on transfer learning. + IEEE Transactions on Knowledge and Data Engineering, 22(10):1345–1359, 2009.
+[ bib ] + +
+ + +
+[1041] +
+
+Luís Paquete, Tommaso Schiavinotto, and Thomas Stützle. + On Local Optima in Multiobjective Combinatorial Optimization Problems. + Annals of Operations Research, 156:83–97, 2007.
+[ bib | +DOI ] +
+In this article, local optimality in multiobjective + combinatorial optimization is used as a baseline for + the design and analysis of two iterative improvement + algorithms. Both algorithms search in a neighborhood + that is defined on a collection of sets of feasible + solutions and their acceptance criterion is based on + outperformance relations. Proofs of the soundness + and completeness of these algorithms are given. +
+
+Keywords: Pareto local search, PLS +
+ +
+ + +
+[1042] +
+
+Luís Paquete and Thomas Stützle. + A study of stochastic local search algorithms for the biobjective QAP with correlated flow matrices. + European Journal of Operational Research, 169(3):943–959, 2006.
+[ bib ] + +
+ + +
+[1043] +
+
+Luís Paquete and Thomas Stützle. + Design and analysis of stochastic local search for the multiobjective traveling salesman problem. + Computers & Operations Research, 36(9):2619–2631, 2009.
+[ bib | +DOI ] + +
+ + +
+[1044] +
+
+S. N. Parragh, Karl F. Doerner, Richard F. Hartl, and Xavier Gandibleux. + A heuristic two-phase solution approach for the multi-objective dial-a-ride problem. + Networks, 54(4):227–242, 2009.
+[ bib ] + +
+ + +
+[1045] +
+
+Rebecca Parsons and Mark Johnson. + A Case Study in Experimental Design Applied to Genetic Algorithms with Applications to DNA Sequence Assembly. + American Journal of Mathematical and Management Sciences, 17(3-4):369–396, 1997.
+[ bib | +DOI ] + +
+ + +
+[1046] +
+
+Moon-Won Park and Yeong-Dae Kim. + A systematic procedure for setting parameters in simulated annealing algorithms. + Computers & Operations Research, 25(3):207–217, 1998.
+[ bib | +DOI ] + +
+ + +
+[1047] +
+
+R. S. Parpinelli, H. S. Lopes, and A. A. Freitas. + Data Mining with an Ant Colony Optimization Algorithm. + IEEE Transactions on Evolutionary Computation, 6(4):321–332, 2002.
+[ bib ] + +
+ + +
+[1048] +
+
+Terence J. Parr and Russell W. Quong. + ANTLR: A predicated-LL (k) parser generator. + Software — Practice & Experience, 25(7):789–810, 1995.
+[ bib ] + +
+ + +
+[1049] +
+
+R. O. Parreiras and J. A. Vascocelos. + A multiplicative version of PROMETHEE II applied to multiobjective optimization problems. + European Journal of Operational Research, 183:729–740, 2007.
+[ bib ] + +
+ + +
+[1050] +
+
+Gerald Paul. + Comparative performance of tabu search and simulated annealing heuristics for the quadratic assignment problem. + Operations Research Letters, 38(6):577–581, 2010.
+[ bib ] + +
+ + +
+[1051] +
+
+Judea Pearl. + The seven tools of causal inference, with reflections on machine learning. + Communications of the ACM, 62(3):54–60, 2019.
+[ bib ] + +
+ + +
+[1052] +
+
+Martín Pedemonte, Sergio Nesmachnow, and Héctor Cancela. + A survey on parallel ant colony optimization. + Applied Soft Computing, 11(8):5181–5197, 2011.
+[ bib ] + +
+ + +
+[1053] +
+
+Paola Pellegrini, Mauro Birattari, and Thomas Stützle. + A Critical Analysis of Parameter Adaptation in Ant Colony Optimization. + Swarm Intelligence, 6(1):23–48, 2012.
+[ bib | +DOI ] + +
+ + +
+[1054] +
+
+Paola Pellegrini, L. Castelli, and R. Pesenti. + Metaheuristic algorithms for the simultaneous slot allocation problem. + IET Intelligent Transport Systems, 6(4):453–462, December 2012.
+[ bib | +DOI ] + +
+ + +
+[1055] +
+
+Paola Pellegrini, Franco Mascia, Thomas Stützle, and Mauro Birattari. + On the Sensitivity of Reactive Tabu Search to its Meta-parameters. + Soft Computing, 18(11):2177–2190, 2014.
+[ bib | +DOI ] + +
+ + +
+[1056] +
+
+Puca Huachi Vaz Penna, Anand Subramanian, and Luiz Satoru Ochi. + An Iterated Local Search Heuristic for the Heterogeneous Fleet Vehicle Routing Problem. + Journal of Heuristics, 19(2):201–232, 2013.
+[ bib ] + +
+ + +
+[1057] +
+
+Jeffrey M. Perkel. + Challenge to scientists: does your ten-year-old code still run? + Nature, 584:556–658, 2020.
+[ bib | +DOI ] +
+Keywords: reproducibility; software engineering; ReScience C; Ten Years + Reproducibility Challenge; code reusability +
+ +
+ + +
+[1058] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, and Thomas Stützle. + Ant colony optimization on a limited budget of evaluations. + Swarm Intelligence, 9(2-3):103–124, 2015.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[1059] +
+
+Matias Péres, Germán Ruiz, Sergio Nesmachnow, and Ana C. Olivera. + Multiobjective evolutionary optimization of traffic flow and pollution in Montevideo, Uruguay. + Applied Soft Computing, 70:472–485, 2018.
+[ bib ] +
+Keywords: Multiobjective evolutionary + algorithms,Pollution,Simulation,Traffic flow +
+ +
+ + +
+[1060] +
+
+A. Pessoa, E. Uchoa, M. Aragão, and R. Rodrigues. + Exact Algorithm over an Arc-time-indexed formulation for Parallel Machine Scheduling Problems. + Mathematical Programming Computation, 2(3–4):259–290, 2010.
+[ bib ] + +
+ + +
+[1061] +
+
+Gilles Pesant, Michel Gendreau, Jean-Yves Potvin, and J.-M. Rousseau. + An Exact Constraint Logic Programming Algorithm for the Traveling Salesman Problem with Time Windows. + Transportation Science, 32:12–29, 1998.
+[ bib ] + +
+ + +
+[1062] +
+
+Charles W. Petit. + Touched by nature: putting evolution to work on the assembly line. + U.S. News & World Report, 125(4):43–45, July 1998.
+[ bib | +http ] +
+Evolutionary optimization of turbine design of the + Boeing 777 GE +
+ +
+ + +
+[1063] +
+
+Justyna Petke, Saemundur O. Haraldsson, Mark Harman, William B. Langdon, David R. White, and John R. Woodward. + Genetic Improvement of Software: A Comprehensive Survey. + IEEE Transactions on Evolutionary Computation, 22(3):415–432, 2018.
+[ bib | +DOI ] + +
+ + +
+[1064] +
+
+Marek Petrik and Shlomo Zilberstein. + Learning parallel portfolios of algorithms. + Annals of Mathematics and Artificial Intelligence, 48(1):85–106, 2006.
+[ bib ] +
+Keywords: algorithm selection +
+ +
+ + +
+[1065] +
+
+S. Pezeshk and O. J. Helweg. + Adaptative Search Optimisation in reducing pump operation costs. + Journal of Water Resources Planning and Management, ASCE, 122(1):57–63, January / February 1996.
+[ bib ] + +
+ + +
+[1066] +
+
+Selcen Phelps and Murat Köksalan. + An interactive evolutionary metaheuristic for multiobjective combinatorial optimization. + Management Science, 49(12):1726–1738, 2003.
+[ bib ] + +
+ + +
+[1067] +
+
+Francesco di Pierro, Soon-Thiam Khu, and Dragan A. Savic. + An investigation on preference order ranking scheme for multiobjective evolutionary optimization. + IEEE Transactions on Evolutionary Computation, 11(1):17–45, 2007.
+[ bib ] + +
+ + +
+[1068] +
+
+Joelle Pineau, Philippe Vincent-Lamarre, Koustuv Sinha, Vincent Larivière, Alina Beygelzimer, Florence d'Alché Buc, Emily Fox, and Hugo Larochelle. + Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program). + Arxiv preprint arXiv:2003.12206 [cs.LG], 2020.
+[ bib | +http ] + +
+ + +
+[1069] +
+
+David Pisinger. + Where are the hard knapsack problems? + Computers & Operations Research, 32(9):2271–2284, 2005.
+[ bib ] + +
+ + +
+[1070] +
+
+David Pisinger and Stefan Ropke. + A General Heuristic for Vehicle Routing Problems. + Computers & Operations Research, 34(8):2403–2435, 2007.
+[ bib ] + +
+ + +
+[1071] +
+
+Rapeepan Pitakaso, Christian Almeder, Karl F. Doerner, and Richard F. Hartl. + Combining exact and population-based methods for the Constrained Multilevel Lot Sizing Problem. + International Journal of Production Research, 44(22):4755–4771, 2006.
+[ bib ] + +
+ + +
+[1072] +
+
+Rapeepan Pitakaso, Christian Almeder, Karl F. Doerner, and Richard F. Hartl. + A Max-Min Ant System for unconstrained multi-level lot-sizing problems. + Computers & Operations Research, 34(9):2533–2552, 2007.
+[ bib | +DOI ] +
+ In this paper, we present an ant-based algorithm + for solving unconstrained multi-level lot-sizing + problems called ant system for multi-level + lot-sizing algorithm (ASMLLS). We apply a hybrid + approach where we use ant colony optimization in + order to find a good lot-sizing sequence, i.e. a + sequence of the different items in the product + structure in which we apply a modified + Wagner-Whitin algorithm for each item + separately. Based on the setup costs each ant + generates a sequence of items. Afterwards a simple + single-stage lot-sizing rule is applied with + modified setup costs. This modification of the setup + costs depends on the position of the item in the + lot-sizing sequence, on the items which have been + lot-sized before, and on two further parameters, + which are tried to be improved by a systematic + search. For small-sized problems ASMLLS is among + the best algorithms, but for most medium- and + large-sized problems it outperforms all other + approaches regarding solution quality as well as + computational time. +
+
+Keywords: Ant colony optimization, Material requirements + planning, Multi-level lot-sizing, Wagner-Whitin + algorithm +
+ +
+ + +
+[1073] +
+
+Hans E. Plesser. + Reproducibility vs. Replicability: A Brief History of a Confused Terminology. + Frontiers in Neuroinformatics, 11, January 2018.
+[ bib | +DOI ] + +
+ + +
+[1074] +
+
+Daniel Porumbel, Gilles Goncalves, Hamid Allaoui, and Tienté Hsu. + Iterated Local Search and Column Generation to solve Arc-Routing as a Permutation Set-Covering Problem. + European Journal of Operational Research, 256(2):349–367, 2017.
+[ bib ] + +
+ + +
+[1075] +
+
+Juan Porta, Jorge Parapar, Ramón Doallo, Vasco Barbosa, Inés Santé, Rafael Crecente, and Carlos Díaz. + A Population-based Iterated Greedy Algorithm for the Delimitation and Zoning of Rural Settlements. + Computers, Environment and Urban Systems, 39:12–26, 2013.
+[ bib ] + +
+ + +
+[1076] +
+
+Jean-Yves Potvin and S. Bengio. + The Vehicle Routing Problem with Time Windows Part II: Genetic Search. + INFORMS Journal on Computing, 8:165–172, 1996.
+[ bib ] + +
+ + +
+[1077] +
+
+T. Devi Prasad. + Design of pumped water distribution networks with storage. + Journal of Water Resources Planning and Management, ASCE, 136(4):129–136, 2009.
+[ bib ] + +
+ + +
+[1078] +
+
+Marco Pranzo and D. Pacciarelli. + An Iterated Greedy Metaheuristic for the Blocking Job Shop Scheduling Problem. + Journal of Heuristics, 22(4):587–611, 2016.
+[ bib | +DOI ] + +
+ + +
+[1079] +
+
+Marcelo Prates, Pedro H. C. Avelar, Henrique Lemos, Luis C. Lamb, and Moshe Y. Vardi. + Learning to Solve NP-Complete Problems: A Graph Neural Network for Decision TSP. + Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):4731–4738, July 2019.
+[ bib | +DOI ] + +
+ + +
+[1080] +
+
+Kenneth V. Price, Abhishek Kumar, and Ponnuthurai N. Suganthan. + Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests. + Swarm and Evolutionary Computation, 78:101287, 2023.
+[ bib | +DOI ] +
+Keywords: Benchmarking, Two-variable non-parametric tests, Evolutionary + algorithms, Dominance, Stochastic optimization, Numerical + optimization, Mann-Whitney test +
+ +
+ + +
+[1081] +
+
+Robert Clay Prim. + Shortest connection networks and some generalizations. + Bell System Technical Journal, 36(6):1389–1401, 1957.
+[ bib ] + +
+ + +
+[1082] +
+
+Philipp Probst, Bernd Bischl, and Anne-Laure Boulesteix. + Tunability: Importance of Hyperparameters of Machine Learning Algorithms. + Arxiv preprint arXiv:1802.09596, 2018.
+[ bib | +http ] +
+Keywords: parameter importance +
+ +
+ + +
+[1083] +
+
+Philipp Probst, Bernd Bischl, and Anne-Laure Boulesteix. + Tunability: Importance of Hyperparameters of Machine Learning Algorithms. + Journal of Machine Learning Research, 20(53):1–32, 2019.
+[ bib ] + +
+ + +
+[1084] +
+
+Luc Pronzato and Werner G. Müller. + Design of computer experiments: space filling and beyond. + Statistics and Computing, 22(3):681–701, 2012.
+[ bib ] +
+Keywords: Kriging; Entropy; Design of experiments; Space-filling; + Sphere packing; Maximin design; Minimax design +
+ +
+ + +
+[1085] +
+
+Harilaos N. Psaraftis. + Dynamic Vehicle Routing: Status and Prospects. + Annals of Operations Research, 61:143–164, 1995.
+[ bib ] + +
+ + +
+[1086] +
+
+Timo Pukkala and Tero Heinonen. + Optimizing heuristic search in forest planning. + Nonlinear Analysis: Real World Applications, 7(5):1284–1297, 2006.
+[ bib ] + +
+ + +
+[1087] +
+
+Luca Pulina and Armando Tacchella. + A self-adaptive multi-engine solver for quantified Boolean formulas. + Constraints, 14(1):80–116, 2009.
+[ bib ] + +
+ + +
+[1088] +
+
+Robin C. Purshouse and Peter J. Fleming. + On the Evolutionary Optimization of Many Conflicting Objectives. + IEEE Transactions on Evolutionary Computation, 11(6):770–784, 2007.
+[ bib | +DOI ] + +
+ + +
+[1089] +
+
+Yutao Qi, Xiaoliang Ma, Fang Liu, Licheng Jiao, Jianyong Sun, and Jianshe Wu. + MOEA/D with adaptive weight adjustment. + Evolutionary Computation, 22(2):231–264, 2014.
+[ bib | +DOI ] +
+Uses an external population +
+ +
+ + +
+[1090] +
+
+Julianne D. Quinn, Patrick M. Reed, and Klaus Keller. + Direct policy search for robust multi-objective management of deeply uncertain socio-ecological tipping points. + Environmental Modelling & Software, 92:125–141, 2017.
+[ bib ] + +
+ + +
+[1091] +
+
+Shahriar Farahmand Rad, Rubén Ruiz, and Naser Boroojerdian. + New High Performing Heuristics for Minimizing Makespan in Permutation Flowshops. + Omega, 37(2):331–345, 2009.
+[ bib ] + +
+ + +
+[1092] +
+
+C. Rajendran. + Heuristic algorithm for scheduling in a flowshop to minimize total flowtime. + International Journal of Production Economics, 29(1):65–73, 1993.
+[ bib ] + +
+ + +
+[1093] +
+
+C. Rajendran and H. Ziegler. + Ant-colony algorithms for permutation flowshop scheduling to minimize makespan/total flowtime of jobs. + European Journal of Operational Research, 155(2):426–438, 2004.
+[ bib ] + +
+ + +
+[1094] +
+
+C. Rajendran and H. Ziegler. + An efficient heuristic for scheduling in a flowshop to minimize total weighted flowtime of jobs. + European Journal of Operational Research, 103(1):129–138, 1997.
+[ bib | +DOI ] + +
+ + +
+[1095] +
+
+David Garzón Ramos and Mauro Birattari. + Automatic Design of Collective Behaviors for Robots that Can Display and Perceive Colors. + Applied Sciences, 10(13):4654, 2020.
+[ bib ] + +
+ + +
+[1096] +
+
+Juan-Manuel Ramos-Pérez, Gara Miranda, Eduardo Segredo, Coromoto León, and Casiano Rodríguez-León. + Application of Multi-Objective Evolutionary Algorithms for Planning Healthy and Balanced School Lunches. + Mathematics, 9(1):80, December 2021.
+[ bib | +DOI ] +
+A multi-objective formulation of the Menu Planning Problem, + which is termed the Multi-objective Menu Planning Problem, is + presented herein. Menu planning is of great interest in the + health field due to the importance of proper nutrition in + today's society, and particularly, in school + canteens. In addition to considering the cost of the meal + plan as the classic objective to be minimized, we also + introduce a second objective aimed at minimizing the degree + of repetition of courses and food groups that a particular + meal plan consists of. The motivation behind this particular + multi-objective formulation is to offer a meal plan that is + not only affordable but also varied and balanced from a + nutritional standpoint. The plan is designed for a given + number of days and ensures that the specific nutritional + requirements of school-age children are satisfied. The main + goal of the current work is to demonstrate the + multi-objective nature of the said formulation, through a + comprehensive experimental assessment carried out over a set + of multi-objective evolutionary algorithms applied to + different instances. At the same time, we are also interested + in validating the multi-objective formulation by performing + quantitative and qualitative analyses of the solutions + attained when solving it. Computational results show the + multi-objective nature of the said formulation, as well as + that it allows suitable meal plans to be obtained. +
+ +
+ + +
+[1097] +
+
+Camelia Ram, Gilberto Montibeller, and Alec Morton. + Extending the use of scenario planning and MCDA for the evaluation of strategic options. + Journal of the Operational Research Society, 62(5):817–829, 2011.
+[ bib ] + +
+ + +
+[1098] +
+
+Zhengfu Rao and Elad Salomons. + Development of a real-time, near-optimal control process for water-distribution networks. + Journal of Hydroinformatics, 9(1):25–37, 2007.
+[ bib | +DOI ] + +
+ + +
+[1099] +
+
+Ronald L. Rardin and Reha Uzsoy. + Experimental Evaluation of Heuristic Optimization Algorithms: A Tutorial. + Journal of Heuristics, 7(3):261–304, 2001.
+[ bib ] + +
+ + +
+[1100] +
+
+Jussi Rasku, Nysret Musliu, and Tommi Kärkkäinen. + On automatic algorithm configuration of vehicle routing problem solvers. + Journal on Vehicle Routing Algorithms, 2(1-4):1–22, February 2019.
+[ bib | +DOI ] +
+Keywords: irace, SMAC, GGA, REVAC, VRP +
+ +
+ + +
+[1101] +
+
+Ingo Rechenberg. + Case studies in evolutionary experimentation and computation. + Computer Methods in Applied Mechanics and Engineering, 186(2-4):125–140, 2000.
+[ bib | +DOI ] + +
+ + +
+[1102] +
+
+Colin R. Reeves and A. V. Eremeev. + Statistical analysis of local search landscapes. + Journal of the Operational Research Society, 55(7):687–693, 2004.
+[ bib | +epub ] + +
+ + +
+[1103] +
+
+Gary R. Reeves and Juan J. Gonzalez. + A comparison of two interactive MCDM procedures. + European Journal of Operational Research, 41(2):203–209, 1989.
+[ bib | +DOI ] +
+Keywords: artificial DM, interactive +
+ +
+ + +
+[1104] +
+
+Patrick M. Reed, David Hadka, Jonathan D. Herman, Joseph R. Kasprzyk, and Joshua B. Kollat. + Evolutionary multiobjective optimization in water resources: The past, present, and future. + Advances in Water Resources, 51:438–456, 2013.
+[ bib ] + +
+ + +
+[1105] +
+
+Tao Chen, Miqing Li, and Xin Yao. + Standing on the shoulders of giants: Seeding search-based multi-objective optimization with prior knowledge for software service composition. + Information and Software Technology, 114:155–175, 2019.
+[ bib ] +
+Example of deteroriation in archiving +
+ +
+ + +
+[1106] +
+
+Frederik Rehbach, Martin Zaefferer, Andreas Fischbach, Günther Rudolph, and Thomas Bartz-Beielstein. + Benchmark-Driven Configuration of a Parallel Model-Based Optimization Algorithm. + IEEE Transactions on Evolutionary Computation, 26(6):1365–1379, 2022.
+[ bib | +DOI ] + +
+ + +
+[1107] +
+
+Gerhard Reinelt. + TSPLIB — A Traveling Salesman Problem Library. + ORSA Journal on Computing, 3(4):376–384, 1991.
+[ bib ] + +
+ + +
+[1108] +
+
+Marc Reimann, Karl F. Doerner, and Richard F. Hartl. + D-ants: Savings based ants divide and conquer the vehicle routing problems. + Computers & Operations Research, 31(4):563–591, 2004.
+[ bib ] + +
+ + +
+[1109] +
+
+Marc Reimann and Marco Laumanns. + Savings based ant colony optimization for the capacitated minimum spanning tree problem. + Computers & Operations Research, 33(6):1794–1822, 2006.
+[ bib | +DOI ] +
+ The problem of connecting a set of client nodes + with known demands to a root node through a minimum + cost tree network, subject to capacity constraints + on all links is known as the capacitated minimum + spanning tree (CMST) problem. As the problem is + NP-hard, we propose a hybrid ant colony + optimization (ACO) algorithm to tackle it + heuristically. The algorithm exploits two important + problem characteristics: (i) the CMST problem is + closely related to the capacitated vehicle routing + problem (CVRP), and (ii) given a clustering of + client nodes that satisfies capacity constraints, + the solution is to find a MST for each cluster, + which can be done exactly in polynomial time. Our + ACO exploits these two characteristics of the + CMST by a solution construction originally + developed for the CVRP. Given the CVRP solution, + we then apply an implementation of Prim's algorithm + to each cluster to obtain a feasible CMST + solution. Results from a comprehensive computational + study indicate the efficiency and effectiveness of + the proposed approach. +
+
+Keywords: Ant colony Optimization, Capacitated minimum + spanning tree problem +
+ +
+ + +
+[1110] +
+
+Zhi-Gang Ren, Zu-Ren Feng, Liang-Jun Ke, and Zhao-Jun Zhang. + New Ideas for Applying Ant Colony Optimization to the Set Covering Problem. + Computers and Industrial Engineering, 58(4):774–784, 2010.
+[ bib ] + +
+ + +
+[1111] +
+
+M. Reyes-Sierra and Carlos A. Coello Coello. + Multi-objective particle swarm optimizers: A survey of the state-of-the-art. + International Journal of Computational Intelligence Research, 2(3):287–308, 2006.
+[ bib ] + +
+ + +
+[1112] +
+
+Craig W. Reynolds. + Flocks, Herds, and Schools: A Distributed Behavioral Model. + ACM Computer Graphics, 21(4):25–34, 1987.
+[ bib ] + +
+ + +
+[1113] +
+
+Jafar Rezaei, Alireza Arab, and Mohammadreza Mehregan. + Analyzing anchoring bias in attribute weight elicitation of SMART, Swing, and best-worst method. + International Transactions in Operational Research, 2022.
+[ bib | +DOI ] +
+Keywords: anchoring bias, best-worst method, cognitive bias, MADM, + multi-attribute weighting, SMART, Swing +
+ +
+ + +
+[1114] +
+
+S. Reza Hejazi and S. Saghafian. + Flowshop-scheduling Problems with Makespan Criterion: A Review. + International Journal of Production Research, 43(14):2895–2929, 2005.
+[ bib ] + +
+ + +
+[1115] +
+
+Imma Ribas, Ramon Companys, and Xavier Tort-Martorell. + An iterated greedy algorithm for the flowshop scheduling problem with blocking. + Omega, 39(3):293 – 301, 2011.
+[ bib ] + +
+ + +
+[1116] +
+
+Imma Ribas, Ramon Companys, and Xavier Tort-Martorell. + An Efficient Iterated Local Search Algorithm for the Total Tardiness Blocking Flow Shop Problem. + International Journal of Production Research, 51(17):5238–5252, 2013.
+[ bib ] + +
+ + +
+[1117] +
+
+Celso C. Ribeiro and Sebastián Urrutia. + Heuristics for the Mirrored Traveling Tournament Problem. + European Journal of Operational Research, 179(3):775–787, 2007.
+[ bib ] + +
+ + +
+[1118] +
+
+A. J. Richmond and John E. Beasley. + An Iterative Construction Heuristic for the Ore Selection Problem. + Journal of Heuristics, 10(2):153–167, 2004.
+[ bib ] + +
+ + +
+[1119] +
+
+John R. Rice. + The Algorithm Selection Problem. + Advances in Computers, 15:65–118, 1976.
+[ bib | +DOI ] +
+The problem of selecting an effective algorithm arises in a + wide variety of situations. This chapter starts with a + discussion on abstract models: the basic model and associated + problems, the model with selection based on features, and the + model with variable performance criteria. One objective of + this chapter is to explore the applicability of the + approximation theory to the algorithm selection + problem. There is an intimate relationship here and that the + approximation theory forms an appropriate base upon which to + develop a theory of algorithm selection methods. The + approximation theory currently lacks much of the necessary + machinery for the algorithm selection problem. There is a + need to develop new results and apply known techniques to + these new circumstances. The final pages of this chapter form + a sort of appendix, which lists 15 specific open problems and + questions in this area. There is a close relationship between + the algorithm selection problem and the general optimization + theory. This is not surprising since the approximation + problem is a special form of the optimization problem. Most + realistic algorithm selection problems are of moderate to + high dimensionality and thus one should expect them to be + quite complex. One consequence of this is that most + straightforward approaches (even well-conceived ones) are + likely to lead to enormous computations for the best + selection. The single most important part of the solution of + a selection problem is the appropriate choice of the form for + selection mapping. It is here that theories give the least + guidance and that the art of problem solving is most + crucial. +
+ +
+ + +
+[1120] +
+
+Juan Carlos Rivera, H. Murat Afsar, and Christian Prins. + A Multistart Iterated Local Search for the Multitrip Cumulative Capacitated Vehicle Routing Problem. + Computational Optimization and Applications, 61(1):159–187, 2015.
+[ bib ] + +
+ + +
+[1121] +
+
+Lucía Rivadeneira, Jian-Bo Yang, and Manuel López-Ibáñez. + Predicting tweet impact using a novel evidential reasoning prediction method. + Expert Systems with Applications, 169:114400, May 2021.
+[ bib | +DOI ] +
+This study presents a novel evidential reasoning (ER) + prediction model called MAKER-RIMER to examine how different + features embedded in Twitter posts (tweets) can predict the + number of retweets achieved during an electoral campaign. The + tweets posted by the two most voted candidates during the + official campaign for the 2017 Ecuadorian Presidential + election were used for this research. For each tweet, five + features including type of tweet, emotion, URL, hashtag, and + date are identified and coded to predict if tweets are of + either high or low impact. The main contributions of the new + proposed model include its suitability to analyse tweet + datasets based on likelihood analysis of data. The model is + interpretable, and the prediction process relies only on the + use of available data. The experimental results show that + MAKER-RIMER performed better, in terms of misclassification + error, when compared against other predictive machine + learning approaches. In addition, the model allows observing + which features of the candidates' tweets are linked to high + and low impact. Tweets containing allusions to the contender + candidate, either with positive or negative connotations, + without hashtags, and written towards the end of the + campaign, were persistently those with the highest + impact. URLs, on the other hand, is the only variable that + performs differently for the two candidates in terms of + achieving high impact. MAKER-RIMER can provide campaigners of + political parties or candidates with a tool to measure how + features of tweets are predictors of their impact, which can + be useful to tailor Twitter content during electoral + campaigns. +
+
+Keywords: Evidential reasoning rule,Belief rule-based inference,Maximum + likelihood data analysis,Twitter,Retweet,Prediction +
+ +
+ + +
+[1122] +
+
+C. P. Robert. + Simulation of truncated normal variables. + Statistics and Computing, 5(2):121–125, June 1995.
+[ bib ] + +
+ + +
+[1123] +
+
+P. A. Romero, A. Krause, and F. H. Arnold. + Navigating the Protein Fitness Landscape with Gaussian Processes. + Proceedings of the National Academy of Sciences, 110(3):E193–E201, December 2012.
+[ bib | +DOI ] +
+Keywords: Combinatorial Black-box Expensive +
+ +
+ + +
+[1124] +
+
+Fabio Romeo and Alberto Sangiovanni-Vincentelli. + A Theoretical Framework for Simulated Annealing. + Algorithmica, 6(1-6):302–345, 1991.
+[ bib ] + +
+ + +
+[1125] +
+
+David S. Roos. + Bioinformatics–trying to swim in a sea of data. + Science, 291(5507):1260–1261, 2001.
+[ bib ] + +
+ + +
+[1126] +
+
+Stefan Ropke and David Pisinger. + A Unified Heuristic for a Large Class of Vehicle Routing Problems with Backhauls. + European Journal of Operational Research, 171(3):750–775, 2006.
+[ bib ] + +
+ + +
+[1127] +
+
+Stefan Ropke and David Pisinger. + An Adaptive Large Neighborhood Search Heuristic for the Pickup and Delivery Problme with Time Windows. + Transportation Science, 40(4):455–472, 2006.
+[ bib ] + +
+ + +
+[1128] +
+
+Brian C. Ross. + Mutual Information between Discrete and Continuous Data Sets. + PLoS One, 9(2):1–5, February 2014.
+[ bib | +DOI ] +
+Mutual information (MI) is a powerful method for detecting + relationships between data sets. There are accurate methods + for estimating MI that avoid problems with “binning” when + both data sets are discrete or when both data sets are + continuous. We present an accurate, non-binning MI estimator + for the case of one discrete data set and one continuous data + set. This case applies when measuring, for example, the + relationship between base sequence and gene expression level, + or the effect of a cancer drug on patient survival time. We + also show how our method can be adapted to calculate the + Jensen-Shannon divergence of two or more data sets. +
+ +
+ + +
+[1129] +
+
+Jonathan Rose, Wolfgang Klebsch, and Jürgen Wolf. + Temperature measurement and equilibrium dynamics of simulated annealing placements. + IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 9(3):253–259, 1990.
+[ bib ] + +
+ + +
+[1130] +
+
+Edward Rothberg. + An evolutionary algorithm for polishing mixed integer programming solutions. + INFORMS Journal on Computing, 19(4):534–541, 2007.
+[ bib ] + +
+ + +
+[1131] +
+
+Daniel H. Rothman. + Nonlinear inversion, statistical mechanics, and residual statics estimation. + Geophysics, 50(12):2784–2796, 1985.
+[ bib ] + +
+ + +
+[1132] +
+
+Daniel H. Rothman. + Automatic estimation of large residual statics corrections. + Geophysics, 51(2):332–346, 1986.
+[ bib ] + +
+ + +
+[1133] +
+
+Bernard Roy. + Robustness in operational research and decision aiding: A multi-faceted issue. + European Journal of Operational Research, 200(3):629–638, 2010.
+[ bib | +DOI ] + +
+ + +
+[1134] +
+
+Günther Rudolph, Oliver Schütze, Christian Grimme, Christian Domínguez-Medina, and Heike Trautmann. + Optimal averaged Hausdorff archives for bi-objective problems: theoretical and numerical results. + Computational Optimization and Applications, 64(2):589–618, 2016.
+[ bib ] + +
+ + +
+[1135] +
+
+Günther Rudolph. + Convergence analysis of canonical genetic algorithms. + IEEE Transactions on Neural Networks, 5(1):96–101, 1994.
+[ bib ] + +
+ + +
+[1136] +
+
+Rubén Ruiz and C. Maroto. + A Comprehensive Review and Evaluation of Permutation Flowshop Heuristics. + European Journal of Operational Research, 165(2):479–494, 2005.
+[ bib ] + +
+ + +
+[1137] +
+
+Rubén Ruiz, C. Maroto, and Javier Alcaraz. + Two new robust genetic algorithms for the flowshop scheduling problem. + Omega, 34(5):461–476, 2006.
+[ bib | +DOI ] + +
+ + +
+[1138] +
+
+Ana Belén Ruiz, Rubén Saborido, and Mariano Luque. + A preference-based evolutionary algorithm for multiobjective optimization: the weighting achievement scalarizing function genetic algorithm. + Journal of Global Optimization, 62(1):101–129, May 2015.
+[ bib | +DOI ] +
+When solving multiobjective optimization problems, + preference-based evolutionary multiobjective optimization + (EMO) algorithms introduce preference information into an + evolutionary algorithm in order to focus the search for + objective vectors towards the region of interest of the + Pareto optimal front. In this paper, we suggest a + preference-based EMO algorithm called weighting achievement + scalarizing function genetic algorithm (WASF-GA), which + considers the preferences of the decision maker (DM) + expressed by means of a reference point. The main purpose of + WASF-GA is to approximate the region of interest of the + Pareto optimal front determined by the reference point, which + contains the Pareto optimal objective vectors that obey the + preferences expressed by the DM in the best possible way. The + proposed approach is based on the use of an achievement + scalarizing function (ASF) and on the classification of the + individuals into several fronts. At each generation of + WASF-GA, this classification is done according to the values + that each solution takes on the ASF for the reference point + and using different weight vectors. These vectors of weights + are selected so that the vectors formed by their inverse + components constitute a well-distributed representation of + the weight vectors space. The efficiency and usefulness of + WASF-GA is shown in several test problems in comparison to + other preference-based EMO algorithms. Regarding a metric + based on the hypervolume, we can say that WASF-GA has + outperformed the other algorithms considered in most of the + problems. +
+
+Proposed WASF-GA +
+ +
+ + +
+[1139] +
+
+Rubén Ruiz and Thomas Stützle. + A Simple and Effective Iterated Greedy Algorithm for the Permutation Flowshop Scheduling Problem. + European Journal of Operational Research, 177(3):2033–2049, 2007.
+[ bib ] + +
+ + +
+[1140] +
+
+Rubén Ruiz and Thomas Stützle. + An Iterated Greedy heuristic for the sequence dependent setup times flowshop problem with makespan and weighted tardiness objectives. + European Journal of Operational Research, 187(3):1143 – 1159, 2008.
+[ bib ] + +
+ + +
+[1141] +
+
+Robert A. Russell. + Hybrid Heuristics for the Vehicle Routing Problem with Time Windows. + Transportation Science, 29(2):156–166, 1995.
+[ bib ] + +
+ + +
+[1142] +
+
+N. R. Sabar, M. Ayob, Graham Kendall, and R. Qu. + Grammatical Evolution Hyper-Heuristic for Combinatorial Optimization Problems. + IEEE Transactions on Evolutionary Computation, 17(6):840–861, 2013.
+[ bib ] + +
+ + +
+[1143] +
+
+N. R. Sabar, M. Ayob, Graham Kendall, and R. Qu. + A Dynamic Multiarmed Bandit-Gene Expression Programming Hyper-Heuristic for Combinatorial Optimization Problems. + IEEE Transactions on Cybernetics, 45(2):217–228, 2015.
+[ bib ] + +
+ + +
+[1144] +
+
+N. R. Sabar, M. Ayob, Graham Kendall, and R. Qu. + Automatic Design of a Hyper-Heuristic Framework With Gene Expression Programming for Combinatorial Optimization Problems. + IEEE Transactions on Evolutionary Computation, 19(3):309–325, 2015.
+[ bib ] + +
+ + +
+[1145] +
+
+Matthieu Sacher, Régis Duvigneau, Olivier Le Maitre, Mathieu Durand, Elisa Berrini, Frédéric Hauville, and Jacques-André Astolfi. + A classification approach to efficient global optimization in presence of non-computable domains. + Structural and Multidisciplinary Optimization, 58(4):1537–1557, 2018.
+[ bib | +DOI ] +
+Proposed EGO-LS-SVM +
+
+Keywords: Safe optimization; CMA-ES, Gaussian processes; Least-Squares + Support Vector Machine +
+ +
+ + +
+[1146] +
+
+Pramod J. Sadalage and Martin Fowler. + NoSQL distilled. + AddisonWesley Professional, 2012.
+[ bib ] + +
+ + +
+[1147] +
+
+A. Burcu Altan Sakarya and Larry W. Mays. + Optimal Operation of Water Distribution Pumps Considering Water Quality. + Journal of Water Resources Planning and Management, ASCE, 126(4):210–220, July / August 2000.
+[ bib ] + +
+ + +
+[1148] +
+
+Marcela Samà, Paola Pellegrini, Andrea D'Ariano, Joaquin Rodriguez, and Dario Pacciarelli. + Ant colony optimization for the real-time train routing selection problem. + Transportation Research Part B: Methodological, 85:89–108, 2016.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[1149] +
+
+Malcolm Sambridge. + Geophysical inversion with a neighbourhood algorithm–I. Searching a parameter space. + Geophysical Journal International, 138(2):479–494, 1999.
+[ bib ] + +
+ + +
+[1150] +
+
+Alejandro Santiago, Bernabé Dorronsoro, Antonio J. Nebro, Juan J. Durillo, Oscar Castillo, and Héctor J. Fraire. + A novel multi-objective evolutionary algorithm with fuzzy logic based adaptive selection of operators: FAME. + Information Sciences, 471:233–251, 2019.
+[ bib | +DOI ] +
+Keywords: Multi-objective optimization, density estimation, + evolutionary algorithm, adaptive algorithm, fuzzy logic, spatial spread deviation +
+ +
+ + +
+[1151] +
+
+Javier Sánchez, Manuel Galán, and Enrique Rubio. + Applying a traffic lights evolutionary optimization technique to a real case: “Las Ramblas” area in Santa Cruz de Tenerife. + IEEE Transactions on Evolutionary Computation, 12(1):25–40, 2008.
+[ bib ] +
+Keywords: Cellular automata, Combinatorial optimization, Genetic + algorithms, Microscopic traffic simulator, Traffic lights + optimization +
+ +
+ + +
+[1152] +
+
+J. J. Sánchez-Medina, M. J. Galán-Moreno, and E. Rubio-Royo. + Traffic Signal Optimization in “La Almozara” District in Saragossa Under Congestion Conditions, Using Genetic Algorithms, Traffic Microsimulation, and Cluster Computing. + IEEE Transactions on Intelligent Transportation Systems, 11(1):132–141, March 2010.
+[ bib | +DOI ] +
+Keywords: cellular automata; genetic algorithms; road traffic;traffic + light programming;urban traffic congestion +
+ +
+ + +
+[1153] +
+
+Nathan Sankary and Avi Ostfeld. + Stochastic Scenario Evaluation in Evolutionary Algorithms Used for Robust Scenario-Based Optimization. + Water Resources Research, 54(4):2813–2833, 2018.
+[ bib ] + +
+ + +
+[1154] +
+
+Alberto Santini, Stefan Ropke, and Lars Magnus Hvattum. + A comparison of acceptance criteria for the adaptive large neighbourhood search metaheuristic. + Journal of Heuristics, 24:783–815, 2018.
+[ bib | +DOI ] + +
+ + +
+[1155] +
+
+E. Sandgren. + Nonlinear integer and discrete programming in mechanical design optimization. + Journal of Mechanical Design, 112(2):223–229, 1990.
+[ bib | +DOI ] + +
+ + +
+[1156] +
+
+René Sass, Eddie Bergman, André Biedenkapp, Frank Hutter, and Marius Thomas Lindauer. + DeepCAVE: An Interactive Analysis Tool for Automated Machine Learning. + Arxiv preprint arXiv:2206.03493 [cs.LG], 2022.
+[ bib | +DOI ] + +
+ + +
+[1157] +
+
+Martin W. P. Savelsbergh. + Local search in routing problems with time windows. + Annals of Operations Research, 4(1):285–305, December 1985.
+[ bib | +DOI ] +
+We develop local search algorithms for routing + problems with time windows. The presented algorithms + are based on thek-interchange concept. The presence + of time windows introduces feasibility constraints, + the checking of which normally requires O(N) + time. Our method reduces this checking effort to + O(1) time. We also consider the problem of finding + initial solutions. A complexity result is given and + an insertion heuristic is described. +
+ +
+ + +
+[1158] +
+
+Dhish Kumar Saxena, João A. Duro, Anish Tiwari, Kalyanmoy Deb, and Qingfu Zhang. + Objective Reduction in Many-Objective Optimization: Linear and Nonlinear Algorithms. + IEEE Transactions on Evolutionary Computation, 17(1):77–99, 2013.
+[ bib | +DOI ] + +
+ + +
+[1159] +
+
+Michael Schilde, Karl F. Doerner, Richard F. Hartl, and Guenter Kiechle. + Metaheuristics for the bi-objective orienteering problem. + Swarm Intelligence, 3(3):179–201, 2009.
+[ bib | +DOI ] +
+In this paper, heuristic solution + techniques for the multi-objective orienteering + problem are developed. The motivation stems from the + problem of planning individual tourist routes in a + city. Each point of interest in a city provides + different benefits for different categories (e.g., + culture, shopping). Each tourist has different + preferences for the different categories when + selecting and visiting the points of interests + (e.g., museums, churches). Hence, a multi-objective + decision situation arises. To determine all the + Pareto optimal solutions, two metaheuristic search + techniques are developed and applied. We use the + Pareto ant colony optimization algorithm and extend + the design of the variable neighborhood search + method to the multi-objective case. Both methods are + hybridized with path relinking procedures. The + performances of the two algorithms are tested on + several benchmark instances as well as on real world + instances from different Austrian regions and the + cities of Vienna and Padua. The computational + results show that both implemented methods are well + performing algorithms to solve the multi-objective + orienteering problem. +
+ +
+ + +
+[1160] +
+
+Martin Schlüter, Jose A. Egea, and Julio R. Banga. + Extended ant colony optimization for non-convex mixed integer nonlinear programming. + Computers & Operations Research, 36(7):2217–2229, 2009.
+[ bib | +DOI ] + +
+ + +
+[1161] +
+
+Oliver Schütze, X. Esquivel, A. Lara, and Carlos A. Coello Coello. + Using the Averaged Hausdorff Distance as a Performance Measure in Evolutionary Multiobjective Optimization. + IEEE Transactions on Evolutionary Computation, 16(4):504–522, 2012.
+[ bib ] + +
+ + +
+[1162] +
+
+Josef Schmee and Gerald J. Hahn. + A Simple Method for Regression Analysis with Censored Data. + Technometrics, 21(4):417–432, 1979.
+[ bib | +DOI ] + +
+ + +
+[1163] +
+
+Mark Schillinger, Benjamin Hartmann, Patric Skalecki, Mona Meister, Duy Nguyen-Tuong, and Oliver Nelles. + Safe active learning and safe Bayesian optimization for tuning a PI-controller. + IFAC-PapersOnLine, 50(1):5967–5972, 2017.
+[ bib | +DOI ] + +
+ + +
+[1164] +
+
+Julie R. Schames, Richard H. Henchman, Jay S. Siegel, Christoph A. Sotriffer, Haihong Ni, and J. Andrew McCammon. + Discovery of a Novel Binding Trench in HIV Integrase. + Journal of Medicinal Chemistry, 47(8):1879–1881, 2004.
+[ bib | +DOI ] +
+Evolutionary optimization of the first clinically approved + anti-viral drug for HIV +
+ +
+ + +
+[1165] +
+
+Oliver Schütze, Carlos Hernández, El-Ghazali Talbi, Jian-Qiao Sun, Yousef Naranjani, and F-R Xiong. + Archivers for the representation of the set of approximate solutions for MOPs. + Journal of Heuristics, 25:71–105, 2019.
+[ bib | +DOI ] +
+Keywords: archiving, nearly optimality, epsilon-dominance, epsilon-approximation, hausdorff convergence +
+ +
+ + +
+[1166] +
+
+Jeffrey C. Schank and Thomas J. Koehnle. + Pseudoreplication is a pseudoproblem. + Journal of Comparative Psychology, 123(4):421–433, 2009.
+[ bib ] + +
+ + +
+[1167] +
+
+Oliver Schütze, A. Lara, and Carlos A. Coello Coello. + On the Influence of the Number of Objectives on the Hardness of a Multiobjective Optimization Problem. + IEEE Transactions on Evolutionary Computation, 15(4):444–455, 2011.
+[ bib ] + +
+ + +
+[1168] +
+
+Oliver Schütze, Marco Laumanns, Carlos A. Coello Coello, Michael Dellnitz, and El-Ghazali Talbi. + Convergence of stochastic search algorithms to finite size Pareto set approximations. + Journal of Global Optimization, 41(4):559–577, 2008.
+[ bib ] + +
+ + +
+[1169] +
+
+Oliver Schütze, Marco Laumanns, Emilia Tantar, Carlos A. Coello Coello, and El-Ghazali Talbi. + Computing gap free Pareto front approximations with stochastic search algorithms. + Evolutionary Computation, 18(1):65–96, 2010.
+[ bib ] + +
+ + +
+[1170] +
+
+G. R. Schreiber and Olivier Martin. + Cut Size Statistics of Graph Bisection Heuristics. + SIAM Journal on Optimization, 10(1):231–251, 1999.
+[ bib ] + +
+ + +
+[1171] +
+
+Gerhard Schrimpf, Johannes Schneider, Hermann Stamm-Wilbrandt, and Gunter Dueck. + Record Breaking Optimization Results Using the Ruin and Recreate Principle. + Journal of Computational Physics, 159(2):139–171, 2000.
+[ bib ] + +
+ + +
+[1172] +
+
+Marie Schmidt, Anita Schöbel, and Lisa Thom. + Min-ordering and max-ordering scalarization methods for multi-objective robust optimization. + European Journal of Operational Research, 275(2):446–459, 2019.
+[ bib ] + +
+ + +
+[1173] +
+
+Eric Schulz, Maarten Speekenbrink, and Andreas Krause. + A tutorial on Gaussian process regression: Modelling, exploring, and exploiting functions. + Journal of Mathematical Psychology, 85:1–16, August 2018.
+[ bib | +DOI ] + +
+ + +
+[1174] +
+
+Tommaso Schiavinotto and Thomas Stützle. + The Linear Ordering Problem: Instances, Search Space Analysis and Algorithms. + Journal of Mathematical Modelling and Algorithms, 3(4):367–402, 2004.
+[ bib ] + +
+ + +
+[1175] +
+
+Tommaso Schiavinotto and Thomas Stützle. + A Review of Metrics on Permutations for Search Space Analysis. + Computers & Operations Research, 34(10):3143–3153, 2007.
+[ bib ] + +
+ + +
+[1176] +
+
+Tom Schrijvers, Guido Tack, Pieter Wuille, Horst Samulowitz, and Peter J. Stuckey. + Search Combinators. + Constraints, 18(2):269–305, 2013.
+[ bib ] + +
+ + +
+[1177] +
+
+Oliver Schütze, Massimiliano Vasile, and Carlos A. Coello Coello. + Computing the Set of Epsilon-Efficient Solutions in Multiobjective Space Mission Design. + Journal of Aerospace Computing, Information, and Communication, 8(3):53–70, 2011.
+[ bib | +DOI ] + +
+ + +
+[1178] +
+
+Matthias Schonlau, William J. Welch, and Donald R. Jones. + Global versus Local Search in Constrained Optimization of Computer Models. + Lecture Notes-Monograph Series, 34:11–25, 1998.
+[ bib | +DOI ] + +
+ + +
+[1179] +
+
+Elias Schede, Jasmin Brandt, Alexander Tornede, Marcel Wever, Viktor Bengs, Eyke Hüllermeier, and Kevin Tierney. + A survey of methods for automated algorithm configuration. + Journal of Artificial Intelligence Research, 75:425–487, 2022.
+[ bib | +DOI ] + +
+ + +
+[1180] +
+
+Pauli Virtanen et al. + SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python. + Nature Methods, 17:261–272, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[1181] +
+
+David F. Shanno. + Conditioning of Quasi-Newton Methods for Function Minimization. + Mathematics of Computation, 24(111):647–656, 1970.
+[ bib ] +
+One of the four papers that proposed BFGS. +
+
+Keywords: BFGS +
+ +
+ + +
+[1182] +
+
+Seyed Mahdi Shavarani, Manuel López-Ibáñez, and Richard Allmendinger. + Detecting Hidden and Irrelevant Objectives in Interactive Multi-Objective Optimization. + IEEE Transactions on Evolutionary Computation, 2023.
+[ bib | +DOI ] +
+Evolutionary multi-objective optimization algorithms (EMOAs) + typically assume that all objectives that are relevant to the + decision-maker (DM) are optimized by the EMOA. In some + scenarios, however, there are irrelevant objectives that are + optimized by the EMOA but ignored by the DM, as well as, + hidden objectives that the DM considers when judging the + utility of solutions but are not optimized. This discrepancy + between the EMOA and the DM's preferences may impede the + search for the most-preferred solution and waste resources + evaluating irrelevant objectives. Research on objective + reduction has focused so far on the structure of the problem + and correlations between objectives and neglected the role of + the DM. We formally define here the concepts of irrelevant + and hidden objectives and propose methods for detecting them, + based on uni-variate feature selection and recursive feature + elimination, that use the preferences already elicited when a + DM interacts with a ranking-based interactive EMOA + (iEMOA). We incorporate the detection methods into an iEMOA + capable of dynamically switching the objectives being + optimized. Our experiments show that this approach can + efficiently identify which objectives are relevant to the DM + and reduce the number of objectives being optimized, while + keeping and often improving the utility, according to the DM, + of the best solution found. +
+ +
+ + +
+[1183] +
+
+Seyed Mahdi Shavarani, Manuel López-Ibáñez, and Joshua D. Knowles. + On Benchmarking Interactive Evolutionary Multi-Objective Algorithms. + IEEE Transactions on Evolutionary Computation, 2023.
+[ bib | +DOI ] +
+We carry out a detailed performance assessment of two + interactive evolutionary multi-objective algorithms (EMOAs) + using a machine decision maker that enables us to repeat + experiments and study specific behaviours modeled after human + decision makers (DMs). Using the same set of benchmark test + problems as in the original papers on these interactive EMOAs + (in up to 10 objectives), we bring to light interesting + effects when we use a machine DM based on sigmoidal utility + functions that have support from the psychology literature + (replacing the simpler utility functions used in the original + papers). Our machine DM enables us to go further and simulate + human biases and inconsistencies as well. Our results from + this study, which is the most comprehensive assessment of + multiple interactive EMOAs so far conducted, suggest that + current well-known algorithms have shortcomings that need + addressing. These results further demonstrate the value of + improving the benchmarking of interactive EMOAs +
+ +
+ + +
+[1184] +
+
+Babooshka Shavazipour, Manuel López-Ibáñez, and Kaisa Miettinen. + Visualizations for Decision Support in Scenario-based Multiobjective Optimization. + Information Sciences, 578:1–21, 2021.
+[ bib | +DOI | +supplementary material ] +
+We address challenges of decision problems when managers need + to optimize several conflicting objectives simultaneously + under uncertainty. We propose visualization tools to support + the solution of such scenario-based multiobjective + optimization problems. Suitable graphical visualizations are + necessary to support managers in understanding, evaluating, + and comparing the performances of management decisions + according to all objectives in all plausible scenarios. To + date, no appropriate visualization has been suggested. This + paper fills this gap by proposing two visualization methods: + a novel extension of empirical attainment functions for + scenarios and an adapted version of heatmaps. They help a + decision-maker in gaining insight into realizations of + trade-offs and comparisons between objective functions in + different scenarios. Some fundamental questions that a + decision-maker may wish to answer with the help of + visualizations are also identified. Several examples are + utilized to illustrate how the proposed visualizations + support a decision-maker in evaluating and comparing + solutions to be able to make a robust decision by answering + the questions. Finally, we validate the usefulness of the + proposed visualizations in a real-world problem with a real + decision-maker. We conclude with guidelines regarding which + of the proposed visualizations are best suited for different + problem classes. +
+ +
+ + +
+[1185] +
+
+Weishi Shao, Dechang Pi, and Zhongshi Shao. + Memetic algorithm with node and edge histogram for no-idle flow shop scheduling problem to minimize the makespan criterion. + Applied Soft Computing, 54:164–182, 2017.
+[ bib ] + +
+ + +
+[1186] +
+
+Weishi Shao, Dechang Pi, and Zhongshi Shao. + A hybrid discrete teaching-learning based meta-heuristic for solving no-idle flow shop scheduling problem with total tardiness criterion. + Computers & Operations Research, 94:89–105, 2018.
+[ bib ] + +
+ + +
+[1187] +
+
+Ke Shang, Tianye Shu, Hisao Ishibuchi, Yang Nan, and Lie Meng Pang. + Benchmarking large-scale subset selection in evolutionary multi-objective optimization. + Information Sciences, 622:755–770, 2023.
+[ bib | +DOI ] + +
+ + +
+[1188] +
+
+Babooshka Shavazipour and T. J. Stewart. + Multi-objective optimisation under deep uncertainty. + Operational Research, September 2019.
+[ bib | +DOI ] +
+This paper presents a scenario-based Multi-Objective + structure to handle decision problems under deep + uncertainty. Most of the decisions in real-life problems need + to be made in the absence of complete knowledge about the + consequences of the decision and/or are characterised by + uncertainties about the future which is unpredictable. These + uncertainties are almost impossible to reduce by gathering + more information and are not statistical in + nature. Therefore, classical probability-based approaches, + such as stochastic programming, do not address these + problems; as they require a correctly-defined complete sample + space, strong assumptions (e.g. normality), or both. The + proposed method extends the concept of two-stage stochastic + programming with recourse to address the capability of + dealing with deep uncertainty through the use of scenario + planning rather than statistical expectation. In this + research, scenarios are used as a dimension of preference to + avoid problems relating to the assessment and use of + probabilities under deep uncertainty. Such scenario-based + thinking involved a multi-objective representation of + performance under different future conditions as an + alternative to expectation. To the best of our knowledge, + this is the first attempt of performing a multi-criteria + evaluation under deep uncertainty through a structured + optimisation model. The proposed structure replacing + probabilities (in dynamic systems with deep uncertainties) by + aspirations within a goal programming structure. In fact, + this paper also proposes an extension of the goal programming + paradigm to deal with deep uncertainty. Furthermore, we will + explain how this structure can be modelled, implemented, and + solved by Goal Programming using some simple, but not + trivial, examples. Further discussion and comparisons with + some popular existing methods will also provided to highlight + the superiorities of the proposed structure. +
+ +
+ + +
+[1189] +
+
+Babooshka Shavazipour, Jonas Stray, and T. J. Stewart. + Sustainable planning in sugar-bioethanol supply chain under deep uncertainty: A case study of South African sugarcane industry. + Computers & Chemical Engineering, 143:107091, 2020.
+[ bib | +DOI ] +
+In this paper, the strategic planning of sugar-bioethanol + supply chains (SCs) under deep uncertainty has been addressed + by applying a two-stage scenario-based multiobjective + optimisation methodology. In practice, the depth of + uncertainty is very high, potential outcomes are not + precisely enumerable, and probabilities of outcomes are not + properly definable. To date, no appropriate framework has + been suggested for dealing with deep uncertainty in supply + chain management and energy-related problems. This study is + the first try to fills this gap. Particularly, the + sustainability of the whole infrastructure of the + sugar-bioethanol SCs is analysed in such a way that the final + solutions are sustainable, robust and adaptable for a broad + range of plausible futures. Three objectives are considered + in this problem under six uncertain parameters. A case study + of South African sugarcane industry is utilised to study and + examine the proposed model. The results prove the economic + profitability and sustainability of the project. +
+
+Keywords: Supply chain management, Multi-objective optimisation, Deep + uncertainty, Scenario planning, Renewable energy, +
+ +
+ + +
+[1190] +
+
+B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and Nando de Freitas. + Taking the human out of the loop: A review of Bayesian optimization. + Proceedings of the IEEE, 104(1):148–175, 2016.
+[ bib ] + +
+ + +
+[1191] +
+
+Bobak Shahriari, Kevin Swersky, Ziyu Wang, Ryan P. Adams, and Nando de Freitas. + Taking the Human Out of the Loop: A Review of Bayesian Optimization. + Proceedings of the IEEE, 104(1):148–175, 2016.
+[ bib ] + +
+ + +
+[1192] +
+
+Ofer M. Shir and Thomas Bäck. + Niching with derandomized evolution strategies in artificial and real-world landscapes. + Natural Computing, 8(1):171–196, 2009.
+[ bib | +DOI ] + +
+ + +
+[1193] +
+
+David Shilane, Jarno Martikainen, Sandrine Dudoit, and Seppo J. Ovaska. + A general framework for statistical performance comparison of evolutionary computation algorithms. + Information Sciences, 178(14):2870–2879, 2008.
+[ bib | +DOI ] + +
+ + +
+[1194] +
+
+Michael D. Shields and Jiaxin Zhang. + The generalization of Latin hypercube sampling. + Reliability Engineering & System Safety, 148:96–108, 2016.
+[ bib ] + +
+ + +
+[1195] +
+
+A. Shmygelska and Holger H. Hoos. + An Ant Colony Optimisation Algorithm for the 2D and 3D Hydrophobic Polar Protein Folding Problem. + BMC Bioinformatics, 6:30, 2005.
+[ bib | +DOI ] + +
+ + +
+[1196] +
+
+Moisés Silva-Muñoz, Alberto Franzin, and Hughes Bersini. + Automatic configuration of the Cassandra database using irace. + PeerJ Computer Science, 7:e634, 2021.
+[ bib | +DOI ] + +
+ + +
+[1197] +
+
+Paulo Vitor Silvestrin and Marcus Ritt. + An Iterated Tabu Search for the Multi-compartment Vehicle Routing Problem. + Computers & Operations Research, 81:192–202, 2017.
+[ bib ] + +
+ + +
+[1198] +
+
+Marcos Melo Silva, Anand Subramanian, and Luiz Satoru Ochi. + An Iterated Local Search Heuristic for the Split Delivery Vehicle Routing Problem. + Computers & Operations Research, 53:234–249, 2015.
+[ bib ] + +
+ + +
+[1199] +
+
+Olivier Simonin, François Charpillet, and Eric Thierry. + Revisiting wavefront construction with collective agents: an approach to foraging. + Swarm Intelligence, 9(2):113–138, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[1200] +
+
+Kevin Sim, Emma Hart, and Ben Paechter. + A Lifelong Learning Hyper-heuristic Method for Bin Packing. + Evolutionary Computation, 23(1):37–67, 2015.
+[ bib | +DOI ] + +
+ + +
+[1201] +
+
+Joseph P. Simmons, Leif D. Nelson, and Uri Simonsohn. + False-Positive Psychology: Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant. + Psychological Science, 2011.
+[ bib | +http ] +
+Proposed the term p-hacking +
+ +
+ + +
+[1202] +
+
+Herbert A. Simon and Allen Newell. + Heuristic Problem Solving: The Next Advance in Operations Research. + Operations Research, 6(1):1–10, 1958.
+[ bib | +DOI ] + +
+ + +
+[1203] +
+
+Joseph P. Simmons, Robyn A. LeBoeuf, and Leif D. Nelson. + The effect of accuracy motivation on anchoring and adjustment: Do people adjust from provided anchors? + Journal of Personality and Social Psychology, 99(6):917–932, 2010.
+[ bib | +DOI ] +
+Increasing accuracy motivation (e.g., by providing monetary + incentives for accuracy) often fails to increase adjustment + away from provided anchors, a result that has led researchers + to conclude that people do not effortfully adjust away from + such anchors. We challenge this conclusion. First, we show + that people are typically uncertain about which way to adjust + from provided anchors and that this uncertainty often causes + people to believe that they have initially adjusted too far + away from such anchors (Studies 1a and 1b). Then, we show + that although accuracy motivation fails to increase the gap + between anchors and final estimates when people are uncertain + about the direction of adjustment, accuracy motivation does + increase anchor-estimate gaps when people are certain about + the direction of adjustment, and that this is true regardless + of whether the anchors are provided or self-generated + (Studies 2, 3a, 3b, and 5). These results suggest that people + do effortfully adjust away from provided anchors but that + uncertainty about the direction of adjustment makes that + adjustment harder to detect than previously assumed. This + conclusion has important theoretical implications, suggesting + that currently emphasized distinctions between anchor types + (self-generated vs. provided) are not fundamental and that + ostensibly competing theories of anchoring (selective + accessibility and anchoring-and-adjustment) are + complementary. +
+ +
+ + +
+[1204] +
+
+Herbert A. Simon. + A Behavioral Model of Rational Choice. + The Quarterly Journal of Economics, 69(1):99–118, 1955.
+[ bib | +epub ] + +
+ + +
+[1205] +
+
+Hemant Kumar Singh, Amitay Isaacs, and Tapabrata Ray. + A Pareto Corner Search Evolutionary Algorithm and Dimensionality Reduction in Many-Objective Optimization Problems. + IEEE Transactions on Evolutionary Computation, 15(4):539–556, 2011.
+[ bib | +DOI ] +
+Many-objective optimization refers to the optimization + problems containing large number of objectives, typically + more than four. Non-dominance is an inadequate strategy for + convergence to the Pareto front for such problems, as almost + all solutions in the population become non-dominated, + resulting in loss of convergence pressure. However, for some + problems, it may be possible to generate the Pareto front + using only a few of the objectives, rendering the rest of the + objectives redundant. Such problems may be reducible to a + manageable number of relevant objectives, which can be + optimized using conventional multiobjective evolutionary + algorithms (MOEAs). For dimensionality reduction, most + proposals in the paper rely on analysis of a representative + set of solutions obtained by running a conventional MOEA for + a large number of generations, which is computationally + overbearing. A novel algorithm, Pareto corner search + evolutionary algorithm (PCSEA), is introduced in this paper, + which searches for the corners of the Pareto front instead of + searching for the complete Pareto front. The solutions + obtained using PCSEA are then used for dimensionality + reduction to identify the relevant objectives. The potential + of the proposed approach is demonstrated by studying its + performance on a set of benchmark test problems and two + engineering examples. While the preliminary results obtained + using PCSEA are promising, there are a number of areas that + need further investigation. This paper provides a number of + useful insights into dimensionality reduction and, in + particular, highlights some of the roadblocks that need to be + cleared for future development of algorithms attempting to + use few selected solutions for identifying relevant + objectives +
+ +
+ + +
+[1206] +
+
+Marcos Singer and Michael L. Pinedo. + A Computational Study of Branch and Bound Techniques for Minimizing the Total Weighted Tardiness in Job Shops. + IIE Transactions, 30(2):109–118, 1998.
+[ bib ] + +
+ + +
+[1207] +
+
+Ankur Sinha, Dhish Kumar Saxena, Kalyanmoy Deb, and Ashutosh Tiwari. + Using objective reduction and interactive procedure to handle many-objective optimization problems. + Applied Soft Computing, 13(1):415–427, 2013.
+[ bib | +DOI ] +
+A number of practical optimization problems are posed as + many-objective (more than three objectives) problems. Most of + the existing evolutionary multi-objective optimization + algorithms, which target the entire Pareto-front are not + equipped to handle many-objective problems. Though there have + been copious efforts to overcome the challenges posed by such + problems, there does not exist a generic procedure to + effectively handle them. This paper presents a simplify and + solve framework for handling many-objective optimization + problems. In that, a given problem is simplified by + identification and elimination of the redundant objectives, + before interactively engaging the decision maker to converge + to the most preferred solution on the Pareto-optimal + front. The merit of performing objective reduction before + interacting with the decision maker is two fold. Firstly, the + revelation that certain objectives are redundant, + significantly reduces the complexity of the optimization + problem, implying lower computational cost and higher search + efficiency. Secondly, it is well known that human beings are + not efficient in handling several factors (objectives in the + current context) at a time. Hence, simplifying the problem a + priori addresses the fundamental issue of cognitive overload + for the decision maker, which may help avoid inconsistent + preferences during the different stages of interactive + engagement. The implementation of the proposed framework is + first demonstrated on a three-objective problem, followed by + its application on two real-world engineering problems. +
+
+Keywords: Evolutionary algorithms, Evolutionary multi- and + many-objective optimization, Multi-criteria decision making, + Machine learning, Interactive optimization +
+ +
+ + +
+[1208] +
+
+Hemant Kumar Singh, Kalyan Shankar Bhattacharjee, and Tapabrata Ray. + Distance-based subset selection for benchmarking in evolutionary multi/many-objective optimization. + IEEE Transactions on Evolutionary Computation, 23(5):904–912, 2019.
+[ bib ] + +
+ + +
+[1209] +
+
+Aymen Sioud and Caroline Gagné. + Enhanced migrating birds optimization algorithm for the permutation flow shop problem with sequence dependent setup times. + European Journal of Operational Research, 264(1):66–73, 2018.
+[ bib ] + +
+ + +
+[1210] +
+
+Ben G. Small, Barry W. McColl, Richard Allmendinger, Jürgen Pahle, Gloria López-Castejón, Nancy J. Rothwell, Joshua D. Knowles, Pedro Mendes, David Brough, and Douglas B. Kell. + Efficient discovery of anti-inflammatory small-molecule combinations using evolutionary computing. + Nature Chemical Biology, 7(12):902–908, 2011.
+[ bib ] + +
+ + +
+[1211] +
+
+Kate Smith-Miles, Davaatseren Baatar, Brendan Wreford, and Rhyd M. R. Lewis. + Towards Objective Measures of Algorithm Performance across Instance Space. + Computers & Operations Research, 45:12–24, 2014.
+[ bib | +DOI ] +
+This paper tackles the difficult but important task of + objective algorithm performance assessment for + optimization. Rather than reporting average performance of + algorithms across a set of chosen instances, which may bias + conclusions, we propose a methodology to enable the strengths + and weaknesses of different optimization algorithms to be + compared across a broader instance space. The results + reported in a recent Computers and Operations Research paper + comparing the performance of graph coloring heuristics are + revisited with this new methodology to demonstrate (i) how + pockets of the instance space can be found where algorithm + performance varies significantly from the average performance + of an algorithm; (ii) how the properties of the instances can + be used to predict algorithm performance on previously unseen + instances with high accuracy; and (iii) how the relative + strengths and weaknesses of each algorithm can be visualized + and measured objectively. +
+
+Keywords: Algorithm selection; Instance Space Analysis; Graph coloring; + Heuristics; Performance prediction +
+ +
+ + +
+[1212] +
+
+Kate Smith-Miles and Simon Bowly. + Generating New Test Instances by Evolving in Instance Space. + Computers & Operations Research, 63:102–113, 2015.
+[ bib | +DOI ] +
+Our confidence in the future performance of any algorithm, + including optimization algorithms, depends on how carefully + we select test instances so that the generalization of + algorithm performance on future instances can be inferred. In + recent work, we have established a methodology to generate a + 2-d representation of the instance space, comprising a set of + known test instances. This instance space shows the + similarities and differences between the instances using + measurable features or properties, and enables the + performance of algorithms to be viewed across the instance + space, where generalizations can be inferred. The power of + this methodology is the insights that can be generated into + algorithm strengths and weaknesses by examining the regions + in instance space where strong performance can be + expected. The representation of the instance space is + dependent on the choice of test instances however. In this + paper we present a methodology for generating new test + instances with controllable properties, by filling observed + gaps in the instance space. This enables the generation of + rich new sets of test instances to support better the + understanding of algorithm strengths and weaknesses. The + methodology is demonstrated on graph colouring as a case + study. +
+
+Keywords: Benchmarking; Evolving instances; Graph colouring; Instance + space; Test instances +
+ +
+ + +
+[1213] +
+
+Kate Smith-Miles, Jeffrey Christiansen, and Mario A. Muñoz. + Revisiting Where Are the Hard Knapsack Problems? Via Instance Space Analysis. + Computers & Operations Research, 128:105184, 2021.
+[ bib | +DOI ] +
+Keywords: 0-1 Knapsack problem; Algorithm portfolios; Algorithm + selection; Instance difficulty; Instance generation; Instance + Space Analysis; Performance evaluation +
+ +
+ + +
+[1214] +
+
+Kate Smith-Miles and Leo Lopes. + Measuring instance difficulty for combinatorial optimization problems. + Computers & Operations Research, 39:875–889, 2012.
+[ bib ] + +
+ + +
+[1215] +
+
+Kate Smith-Miles and Mario A. Muñoz. + Instance Space Analysis for Algorithm Testing: Methodology and Software Tools. + ACM Computing Surveys, 55(12), March 2023.
+[ bib | +DOI ] +
+Instance Space Analysis (ISA) is a recently developed + methodology to (a) support objective testing of algorithms + and (b) assess the diversity of test instances. Representing + test instances as feature vectors, the ISA methodology + extends Rice's 1976 Algorithm Selection Problem framework to + enable visualization of the entire space of possible test + instances, and gain insights into how algorithm performance + is affected by instance properties. Rather than reporting + algorithm performance on average across a chosen set of test + problems, as is standard practice, the ISA methodology offers + a more nuanced understanding of the unique strengths and + weaknesses of algorithms across different regions of the + instance space that may otherwise be hidden on average. It + also facilitates objective assessment of any bias in the + chosen test instances and provides guidance about the + adequacy of benchmark test suites. This article is a + comprehensive tutorial on the ISA methodology that has been + evolving over several years, and includes details of all + algorithms and software tools that are enabling its worldwide + adoption in many disciplines. A case study comparing + algorithms for university timetabling is presented to + illustrate the methodology and tools. +
+
+Keywords: test instance diversity, benchmarking, timetabling, Algorithm + footprints, MATLAB, software as a service, meta-heuristics, + algorithm selection, meta-learning +
+ +
+ + +
+[1216] +
+
+Kate Smith-Miles. + Cross-disciplinary Perspectives on Meta-learning for Algorithm Selection. + ACM Computing Surveys, 41(1):1–25, 2008.
+[ bib ] + +
+ + +
+[1217] +
+
+Krzysztof Socha and Christian Blum. + An ant colony optimization algorithm for continuous optimization: An application to feed-forward neural network training. + Neural Computing & Applications, 16(3):235–247, 2007.
+[ bib ] + +
+ + +
+[1218] +
+
+Krzysztof Socha and Marco Dorigo. + Ant Colony Optimization for Continuous Domains. + European Journal of Operational Research, 185(3):1155–1173, 2008.
+[ bib | +DOI ] +
+Proposed ACOR (ACOR) +
+
+Keywords: ACOR +
+ +
+ + +
+[1219] +
+
+Christine Solnon. + Ants Can Solve Constraint Satisfaction Problems. + IEEE Transactions on Evolutionary Computation, 6(4):347–357, 2002.
+[ bib ] + +
+ + +
+[1220] +
+
+D. Soler, E. Martínez, and J. C. Micó. + A Transformation for the Mixed General Routing Problem with Turn Penalties. + Journal of the Operational Research Society, 59:540–547, 2008.
+[ bib ] + +
+ + +
+[1221] +
+
+M. M. Solomon. + Algorithms for the Vehicle Routing and Scheduling Problems with Time Windows. + Operations Research, 35:254–265, 1987.
+[ bib ] + +
+ + +
+[1222] +
+
+Zhenshou Song, Handing Wang, Cheng He, and Yaochu Jin. + A Kriging-assisted two-archive evolutionary algorithm for expensive many-objective optimization. + IEEE Transactions on Evolutionary Computation, 25(6):1013–1027, 2021.
+[ bib ] + +
+ + +
+[1223] +
+
+Kenneth Sörensen. + Metaheuristics—the metaphor exposed. + International Transactions in Operational Research, 22(1):3–18, 2015.
+[ bib | +DOI ] + +
+ + +
+[1224] +
+
+Kenneth Sörensen, Florian Arnold, and Daniel Palhazi Cuervo. + A critical analysis of the “improved Clarke and Wright savings algorithm”. + International Transactions in Operational Research, 26(1):54–63, 2017.
+[ bib | +DOI ] +
+Keywords: reproducibility, vehicle routing +
+ +
+ + +
+[1225] +
+
+Jorge A. Soria-Alcaraz, Gabriela Ochoa, Marco A. Sotelo-Figeroa, and Edmund K. Burke. + A Methodology for Determining an Effective Subset of Heuristics in Selection Hyper-heuristics. + European Journal of Operational Research, 260:972–983, 2017.
+[ bib ] + +
+ + +
+[1226] +
+
+Marcelo De Souza, Marcus Ritt, and Manuel López-Ibáñez. + Capping Methods for the Automatic Configuration of Optimization Algorithms. + Computers & Operations Research, 139:105615, 2022.
+[ bib | +DOI | +supplementary material ] +
+Automatic configuration techniques are widely and + successfully used to find good parameter settings for + optimization algorithms. Configuration is costly, because it + is necessary to evaluate many configurations on different + instances. For decision problems, when the objective is to + minimize the running time of the algorithm, many + configurators implement capping methods to discard poor + configurations early. Such methods are not directly + applicable to optimization problems, when the objective is to + optimize the cost of the best solution found, given a + predefined running time limit. We propose new capping methods + for the automatic configuration of optimization + algorithms. They use the previous executions to determine a + performance envelope, which is used to evaluate new + executions and cap those that do not satisfy the envelope + conditions. We integrate the capping methods into the irace + configurator and evaluate them on different optimization + scenarios. Our results show that the proposed methods can + save from about 5% to 78% of the configuration effort, + while finding configurations of the same quality. Based on + the computational analysis, we identify two conservative and + two aggressive methods, that save an average of about 20% + and 45% of the configuration effort, respectively. We also + provide evidence that capping can help to better use the + available budget in scenarios with a configuration time + limit. +
+ +
+ + +
+[1227] +
+
+Abdelghani Souilah. + Simulated annealing for manufacturing systems layout design. + European Journal of Operational Research, 82(3):592–614, 1995.
+[ bib ] + +
+ + +
+[1228] +
+
+Charles Spearman. + The proof and measurement of association between two things. + The American journal of psychology, 15(1):72–101, 1904.
+[ bib ] + +
+ + +
+[1229] +
+
+J. L. Henning. + SPEC CPU2000: measuring CPU performance in the New Millennium. + Computer, 33(7):28–35, 2000.
+[ bib | +DOI ] + +
+ + +
+[1230] +
+
+Daniel A. Spielman and Shang-Hua Teng. + Smoothed analysis of algorithms: Why the simplex algorithm usually takes polynomial time. + Journal of the ACM, 51(3):385–463, 2004.
+[ bib ] + +
+ + +
+[1231] +
+
+Arno Sprecher, Sönke Hartmann, and Andreas Drexl. + An exact algorithm for project scheduling with multiple modes. + OR Spektrum, 19(3):195–203, 1997.
+[ bib | +DOI ] +
+Keywords: branch-and-bound, multi-mode resource-constrained + project scheduling, project scheduling +
+ +
+ + +
+[1232] +
+
+Arno Sprecher, Rainer Kolisch, and Andreas Drexl. + Semi-active, active, and non-delay schedules for the resource-constrained project scheduling problem. + European Journal of Operational Research, 80(1):94–102, 1995.
+[ bib | +DOI ] +
+We consider the resource-constrained project + scheduling problem (RCPSP). The focus of the paper + is on a formal definition of semi-active, active, + and non-delay schedules. Traditionally these + schedules establish basic concepts within the job + shop scheduling literature. There they are usually + defined in a rather informal way which does not + create any substantial problems. Using these + concepts in the more general RCPSP without giving + a formal definition may cause serious + problems. After providing a formal definition of + semi-active, active, and non-delay schedules for the + RCPSP we outline some of these problems occurring + within the disjunctive arc concept. +
+
+Keywords: active schedules, Branch-and-bound methods, + non-delay schedules, Resource-constrained project + scheduling, Semi-active schedules +
+ +
+ + +
+[1233] +
+
+N. Srinivas and Kalyanmoy Deb. + Multiobjective Optimization Using Nondominated Sorting in Genetic Algorithms. + Evolutionary Computation, 2(3):221–248, 1994.
+[ bib ] + +
+ + +
+[1234] +
+
+T. J. Stewart. + Robustness of Additive Value Function Methods in MCDM. + Journal of Multi-Criteria Decision Analysis, 5(4):301–309, 1996.
+[ bib ] +
+Keywords: machine decision-making +
+ +
+ + +
+[1235] +
+
+T. J. Stewart. + Evaluation and refinement of aspiration-based methods in MCDM. + European Journal of Operational Research, 113(3):643–652, 1999.
+[ bib ] +
+Keywords: machine decision-making +
+ +
+ + +
+[1236] +
+
+T. J. Stewart. + Goal programming and cognitive biases in decision-making. + Journal of the Operational Research Society, 56(10):1166–1175, 2005.
+[ bib | +DOI ] +
+Keywords: machine decision making +
+ +
+ + +
+[1237] +
+
+T. J. Stewart, Simon French, and Jesus Rios. + Integrating multicriteria decision analysis and scenario planning: Review and extension. + Omega, 41(4):679–688, 2013.
+[ bib | +DOI ] +
+Keywords: Multicriteria decision analysis +
+ +
+ + +
+[1238] +
+
+Helena Stegherr, Michael Heider, and Jörg Hähner. + Classifying Metaheuristics: Towards a unified multi-level classification system. + Natural Computing, 2020.
+[ bib | +DOI ] + +
+ + +
+[1239] +
+
+Sarah Steiner and Tomasz Radzik. + Computing all efficient solutions of the biobjective minimum spanning tree problem. + Computers & Operations Research, 35(1):198–211, 2008.
+[ bib ] + +
+ + +
+[1240] +
+
+Victoria Stodden. + What scientific idea is ready for retirement? Reproducibility. + Edge, 2014.
+[ bib | +http ] +
+Introduces computational reproducibility, empirical + reproducibility and statistical reproducibility +
+ +
+ + +
+[1241] +
+
+Daniel H. Stolfi and Enrique Alba. + Red Swarm: Reducing travel times in smart cities by using bio-inspired algorithms. + Applied Soft Computing, 24:181–195, 2014.
+[ bib | +DOI ] +
+This article presents an innovative approach to solve one of + the most relevant problems related to smart mobility: the + reduction of vehicles' travel time. Our original approach, + called Red Swarm, suggests a potentially customized route to + each vehicle by using several spots located at traffic lights + in order to avoid traffic jams by using {V2I} + communications. That is quite different from other existing + proposals, as it deals with real maps and actual streets, as + well as several road traffic distributions. We propose an + evolutionary algorithm (later efficiently parallelized) to + optimize our case studies which have been imported from + OpenStreetMap into {SUMO} as they belong to a real city. We + have also developed a Rerouting Algorithm which accesses the + configuration of the Red Swarm and communicates the route + chosen to vehicles, using the spots (via WiFi + link). Moreover, we have developed three competing algorithms + in order to compare their results to those of Red Swarm and + have observed that Red Swarm not only achieved the best + results, but also outperformed the experts' solutions in a + total of 60 scenarios tested, with up to 19% shorter travel + times. +
+
+Keywords: Evolutionary algorithm,Road traffic,Smart city,Smart + mobility,Traffic light,WiFi connections +
+ +
+ + +
+[1242] +
+
+Victoria Stodden, Marcia McNutt, David H. Bailey, Ewa Deelman, Yolanda Gil, Brooks Hanson, Michael A. Heroux, John P. A. Ioannidis, and Michela Taufer. + Enhancing reproducibility for computational methods. + Science, 354(6317):1240–1241, December 2016.
+[ bib | +DOI ] + +
+ + +
+[1243] +
+
+Rainer Storn and Kenneth Price. + Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces. + Journal of Global Optimization, 11(4):341–359, 1997.
+[ bib | +DOI ] +
+Proposed differential evolution +
+ +
+ + +
+[1244] +
+
+Victoria Stodden, Jennifer Seiler, and Zhaokun Ma. + An empirical analysis of journal policy effectiveness for computational reproducibility. + Proceedings of the National Academy of Sciences, 115(11):2584–2589, March 2018.
+[ bib | +DOI ] + +
+ + +
+[1245] +
+
+Philip N. Strenski and Scott Kirkpatrick. + Analysis of Finite Length Annealing Schedules. + Algorithmica, 6(1-6):346–366, 1991.
+[ bib ] + +
+ + +
+[1246] +
+
+Patrycja Strycharczuk, Manuel López-Ibáñez, Georgina Brown, and Adrian Leemann. + General Northern English: Exploring regional variation in the North of England with machine learning. + Frontiers in Artificial Intelligence, 3(48), 2020.
+[ bib | +DOI ] +
+In this paper, we present a novel computational approach to + the analysis of accent variation. The case study is dialect + leveling in the North of England, manifested as reduction of + accent variation across the North and emergence of General + Northern English (GNE), a pan-regional standard accent + associated with middle-class speakers. We investigated this + instance of dialect leveling using random forest + classification, with audio data from a crowd-sourced corpus + of 105 urban, mostly highly-educated speakers from five + northern UK cities: Leeds, Liverpool, Manchester, Newcastle + upon Tyne, and Sheffield. We trained random forest models to + identify individual northern cities from a sample of other + northern accents, based on first two formant measurements of + full vowel systems. We tested the models using unseen + data. We relied on undersampling, bagging (bootstrap + aggregation) and leave-one-out cross-validation to address + some challenges associated with the data set, such as + unbalanced data and relatively small sample size. The + accuracy of classification provides us with a measure of + relative similarity between different pairs of cities, while + calculating conditional feature importance allows us to + identify which input features (which vowels and which + formants) have the largest influence in the prediction. We do + find a considerable degree of leveling, especially between + Manchester, Leeds and Sheffield, although some differences + persist. The features that contribute to these differences + most systematically are typically not the ones discussed in + previous dialect descriptions. We propose that the most + systematic regional features are also not salient, and as + such, they serve as sociolinguistic regional indicators. We + supplement the random forest results with a more traditional + variationist description of by-city vowel systems, and we use + both sources of evidence to inform a description of the + vowels of General Northern English. +
+
+Keywords: vowels, accent features, dialect leveling, Random forest + (bagging), Feature selecion +
+ +
+ + +
+[1247] +
+
+Thomas Stützle. + Iterated Local Search for the Quadratic Assignment Problem. + European Journal of Operational Research, 174(3):1519–1539, 2006.
+[ bib ] + +
+ + +
+[1248] +
+
+Thomas Stützle and Marco Dorigo. + A Short Convergence Proof for a Class of ACO Algorithms. + IEEE Transactions on Evolutionary Computation, 6(4):358–365, 2002.
+[ bib ] + +
+ + +
+[1249] +
+
+Thomas Stützle and Holger H. Hoos. + Max-Min Ant System. + Future Generation Computer Systems, 16(8):889–914, 2000.
+[ bib ] + +
+ + +
+[1250] +
+
+Zhaopin Su, Guofu Zhang, Feng Yue, Dezhi Zhan, Miqing Li, Bin Li, and Xin Yao. + Enhanced Constraint Handling for Reliability-Constrained Multiobjective Testing Resource Allocation. + IEEE Transactions on Evolutionary Computation, 25(3):537–551, 2021.
+[ bib ] + +
+ + +
+[1251] +
+
+Anand Subramanian and Maria Battarra. + An Iterated Local Search Algorithm for the Travelling Salesman Problem with Pickups and Deliveries. + Journal of the Operational Research Society, 64(3):402–409, 2013.
+[ bib ] + +
+ + +
+[1252] +
+
+Anand Subramanian, Maria Battarra, and Chris N. Potts. + An Iterated Local Search Heuristic for the Single Machine Total Weighted Tardiness Scheduling Problem with Sequence-dependent Setup Times. + International Journal of Production Research, 52(9):2729–2742, 2014.
+[ bib ] + +
+ + +
+[1253] +
+
+Yanan Sui, Vincent Zhuang, Joel W. Burdick, and Yisong Yue. + Stagewise Safe Bayesian Optimization with Gaussian Processes. + Arxiv preprint arXiv:1806.07555, 2018. + Published as [2563].
+[ bib | +http ] +
+Enforcing safety is a key aspect of many problems pertaining + to sequential decision making under uncertainty, which + require the decisions made at every step to be both + informative of the optimal decision and also safe. For + example, we value both efficacy and comfort in medical + therapy, and efficiency and safety in robotic control. We + consider this problem of optimizing an unknown utility + function with absolute feedback or preference feedback + subject to unknown safety constraints. We develop an + efficient safe Bayesian optimization algorithm, StageOpt, + that separates safe region expansion and utility function + maximization into two distinct stages. Compared to existing + approaches which interleave between expansion and + optimization, we show that StageOpt is more efficient and + naturally applicable to a broader class of problems. We + provide theoretical guarantees for both the satisfaction of + safety constraints as well as convergence to the optimal + utility value. We evaluate StageOpt on both a variety of + synthetic experiments, as well as in clinical practice. We + demonstrate that StageOpt is more effective than existing + safe optimization approaches, and is able to safely and + effectively optimize spinal cord stimulation therapy in our + clinical experiments. +
+
+Keywords: Safe Optimization, StageOpt +
+ +
+ + +
+[1254] +
+
+Yanan Sun, Gary G. Yen, and Zhang Yi. + IGD Indicator-based Evolutionary Algorithm for Many-objective Optimization Problems. + IEEE Transactions on Evolutionary Computation, 23(2):173–187, 2019.
+[ bib | +DOI ] + +
+ + +
+[1255] +
+
+A. Suppapitnarm, K. A. Seffen, G. T. Parks, and P. J. Clarkson. + A simulated annealing algorithm for multiobjective optimization. + Engineering Optimization, 33(1):59–85, 2000.
+[ bib ] + +
+ + +
+[1256] +
+
+Johan A. K. Suykens and Joos Vandewalle. + Least Squares Support Vector Machine Classifiers. + Neural Processing Letters, 9(3):293–300, 1999.
+[ bib | +DOI ] +
+Keywords: LS-SVM +
+ +
+ + +
+[1257] +
+
+Jerry Swan, Steven Adriaensen, Adam D. Barwell, Kevin Hammond, and David R. White. + Extending the “Open-Closed Principle” to Automated Algorithm Configuration. + Evolutionary Computation, 27(1):173–193, 2019.
+[ bib | +DOI ] + +
+ + +
+[1258] +
+
+Jerry Swan, Steven Adriaensen, Alexander E. I. Brownlee, Kevin Hammond, Colin G. Johnson, Ahmed Kheiri, Faustyna Krawiec, Juan-Julián Merelo, Leandro L. Minku, Ender Özcan, Gisele Pappa, Pablo García-Sánchez, Kenneth Sörensen, Stefan Voß, Markus Wagner, and David R. White. + Metaheuristics “In the Large”. + European Journal of Operational Research, 297(2):393–406, March 2022.
+[ bib | +DOI ] + +
+ + +
+[1259] +
+
+Jerry Swan, John R. Woodward, Ender Özcan, Graham Kendall, and Edmund K. Burke. + Searching the Hyper-heuristic Design Space. + Cognitive Computation, 6(1):66–73, March 2014.
+[ bib | +DOI ] + +
+ + +
+[1260] +
+
+Harold Szu and Ralph Hartley. + Fast Simulated Annealing. + Physics Letters A, 122(3):157–162, 1987.
+[ bib ] + +
+ + +
+[1261] +
+
+Éric D. Taillard. + Some Efficient Heuristic Methods for the Flow Shop Sequencing Problem. + European Journal of Operational Research, 47(1):65–74, 1990.
+[ bib ] + +
+ + +
+[1262] +
+
+Éric D. Taillard. + Robust Taboo Search for the Quadratic Assignment Problem. + Parallel Computing, 17(4-5):443–455, 1991.
+[ bib ] +
+faster 2-exchange delta evaluation in QAP +
+ +
+ + +
+[1263] +
+
+Éric D. Taillard. + Benchmarks for Basic Scheduling Problems. + European Journal of Operational Research, 64(2):278–285, 1993.
+[ bib ] + +
+ + +
+[1264] +
+
+Éric D. Taillard. + Comparison of Iterative Searches for the Quadratic Assignment Problem. + Location Science, 3(2):87–105, 1995.
+[ bib ] + +
+ + +
+[1265] +
+
+El-Ghazali Talbi. + A Taxonomy of Hybrid Metaheuristics. + Journal of Heuristics, 8(5):541–564, 2002.
+[ bib ] + +
+ + +
+[1266] +
+
+Kar Yan Tam. + A Simulated Annealing Algorithm for Allocating Space to Manufacturing Cells. + International Journal of Production Research, 30(1):63–87, 1992.
+[ bib ] + +
+ + +
+[1267] +
+
+M. Tamiz, D. F. Jones, and E. El-Darzi. + A review of Goal Programming and its applications. + Annals of Operations Research, 58(1):39–53, January 1995.
+[ bib | +DOI ] +
+This paper presents a review of the current literature on the + branch of multi-criteria decision modelling known as Goal + Programming (GP). The result of our indepth investigations of + the two main GP methods, lexicographic and weighted GP + together with their distinct application areas is + reported. Some guidelines to the scope of GP as an + application tool are given and methods of determining which + problem areas are best suited to the different GP approaches + are proposed. The correlation between the method of assigning + weights and priorities and the standard of the results is + also ascertained. +
+
+Keywords: Goal Programming, lexicographic, weighted +
+ +
+ + +
+[1268] +
+
+Shunji Tanaka and Mituhiko Araki. + An Exact Algorithm for the Single-machine Total Weighted Tardiness Problem with Sequence-dependent Setup Times. + Computers & Operations Research, 40(1):344–352, 2013.
+[ bib ] + +
+ + +
+[1269] +
+
+Ryoji Tanabe and Hisao Ishibuchi. + An easy-to-use real-world multi-objective optimization problem suite. + Applied Soft Computing, 89:106078, 2020.
+[ bib ] +
+Proposed the RE benchmark suite +
+ +
+ + +
+[1270] +
+
+Ryoji Tanabe, Hisao Ishibuchi, and Akira Oyama. + Benchmarking Multi- and Many-Objective Evolutionary Algorithms Under Two Optimization Scenarios. + IEEE Access, 5:19597–19619, 2017.
+[ bib ] +
+compared a number of MOEAs using a wide range of numbers of + objectives and stopping criteria, with and without archivers; unbounded archive +
+ +
+ + +
+[1271] +
+
+Lixin Tang and Xianpeng Wang. + Iterated local search algorithm based on very large-scale neighborhood for prize-collecting vehicle routing problem. + International Journal of Advanced Manufacturing Technology, 29(11):1246–1258, 2006.
+[ bib ] + +
+ + +
+[1272] +
+
+A. J. Tarquin and J. Dowdy. + Optimal pump operation in water distribution. + Journal of Hydraulic Engineering, ASCE, 115(2):158–169 or 496–501, February 1989.
+[ bib ] + +
+ + +
+[1273] +
+
+M. F. Tasgetiren, D. Kizilay, Quan-Ke Pan, and Ponnuthurai N. Suganthan. + Iterated Greedy Algorithms for the Blocking Flowshop Scheduling Problem with Makespan Criterion. + Computers & Operations Research, 77:111–126, 2017.
+[ bib ] + +
+ + +
+[1274] +
+
+M. Fatih Tasgetiren, Yun-Chia Liang, Mehmet Sevkli, and Gunes Gencyilmaz. + A particle swarm optimization algorithm for makespan and total flowtime minimization in the permutation flowshop sequencing problem. + European Journal of Operational Research, 177(3):1930–1947, 2007.
+[ bib | +DOI ] + +
+ + +
+[1275] +
+
+M. Fatih Tasgetiren, Quan-Ke Pan, Ponnuthurai N. Suganthan, and Ozge Buyukdagli. + A variable iterated greedy algorithm with differential evolution for the no-idle permutation flowshop scheduling problem. + Computers & Operations Research, 40(7):1729–1743, 2013.
+[ bib ] + +
+ + +
+[1276] +
+
+Joc Cing Tay and Nhu Binh Ho. + Evolving dispatching rules using genetic programming for solving multi-objective flexible job-shop problems. + Computers and Industrial Engineering, 54(3):453 – 473, 2008.
+[ bib | +DOI ] + +
+ + +
+[1277] +
+
+Cristina Teixeira, José Covas, Thomas Stützle, and António Gaspar-Cunha. + Engineering an Efficient Two-Phase Local Search for the Co-Rotating Twin-Screw Configuration Problem. + International Transactions in Operational Research, 18(2):271–291, 2011.
+[ bib ] + +
+ + +
+[1278] +
+
+Cristina Teixeira, José Covas, Thomas Stützle, and António Gaspar-Cunha. + Multi-Objective Ant Colony Optimization for Solving the Twin-Screw Extrusion Configuration Problem. + Engineering Optimization, 44(3):351–371, 2012.
+[ bib ] + +
+ + +
+[1279] +
+
+Cristina Teixeira, José Covas, Thomas Stützle, and António Gaspar-Cunha. + Hybrid Algorithms for the Twin-Screw Extrusion Configuration Problem. + Applied Soft Computing, 23:298–307, 2014.
+[ bib ] + +
+ + +
+[1280] +
+
+Fitsum Teklu, Agachai Sumalee, and David Watling. + A Genetic Algorithm Approach for Optimizing Traffic Control Signals Considering Routing. + Computer-Aided Civil and Infrastructure Engineering, 22(1):31–43, January 2007.
+[ bib | +DOI ] + +
+ + +
+[1281] +
+
+J. B. Tenenbaum, V. D. Silva, and J. C. Langford. + A global geometric framework for nonlinear dimensionality reduction. + Science, 290(5500):2319–2323, 2000.
+[ bib ] + +
+ + +
+[1282] +
+
+J. Teo and Hussein A. Abbass. + Automatic generation of controllers for embodied legged organisms: A Pareto evolutionary multi-objective approach. + Evolutionary Computation, 12(3):355–394, 2004.
+[ bib | +DOI ] + +
+ + +
+[1283] +
+
+Kei Terayama, Masato Sumita, Ryo Tamura, and Koji Tsuda. + Black-Box Optimization for Automated Discovery. + Accounts of Chemical Research, 54(6):1334–1346, March 2021.
+[ bib | +DOI ] +
+In chemistry and materials science, researchers and engineers + discover, design, and optimize chemical compounds or + materials with their professional knowledge and + techniques. At the highest level of abstraction, this process + is formulated as black-box optimization. For instance, the + trial-and-error process of synthesizing various molecules for + better material properties can be regarded as optimizing a + black-box function describing the relation between a chemical + formula and its properties. Various black-box optimization + algorithms have been developed in the machine learning and + statistics communities. Recently, a number of researchers + have reported successful applications of such algorithms to + chemistry. They include the design of photofunctional + molecules and medical drugs, optimization of thermal emission + materials and high Li-ion conductive solid electrolytes, and + discovery of a new phase in inorganic thin films for solar + cells.There are a wide variety of algorithms available for + black-box optimization, such as Bayesian optimization, + reinforcement learning, and active learning. Practitioners + need to select an appropriate algorithm or, in some cases, + develop novel algorithms to meet their demands. It is also + necessary to determine how to best combine machine learning + techniques with quantum mechanics- and molecular + mechanics-based simulations, and experiments. In this + Account, we give an overview of recent studies regarding + automated discovery, design, and optimization based on + black-box optimization. The Account covers the following + algorithms: Bayesian optimization to optimize the chemical or + physical properties, an optimization method using a quantum + annealer, best-arm identification, gray-box optimization, and + reinforcement learning. In addition, we introduce active + learning and boundless objective-free exploration, which may + not fall into the category of black-box optimization.Data + quality and quantity are key for the success of these + automated discovery techniques. As laboratory automation and + robotics are put forward, automated discovery algorithms + would be able to match human performance at least in some + domains in the near future. +
+ +
+ + +
+[1284] +
+
+Patrick Thibodeau. + Machine-based decision-making is coming. + Computer World, November 2011. + Last accessed: 15 January 2014.
+[ bib | +http ] + +
+ + +
+[1285] +
+
+Lothar Thiele, Kaisa Miettinen, Pekka Korhonen, and Julián Molina. + A Preference-Based Evolutionary Algorithm for Multi-Objective Optimization. + Evolutionary Computation, 17(3):411–436, 2009.
+[ bib | +DOI ] +
+ Abstract In this paper, we discuss the idea of incorporating + preference information into evolutionary multi-objective + optimization and propose a preference-based evolutionary + approach that can be used as an integral part of an + interactive algorithm. One algorithm is proposed in the + paper. At each iteration, the decision maker is asked to give + preference information in terms of his or her reference point + consisting of desirable aspiration levels for objective + functions. The information is used in an evolutionary + algorithm to generate a new population by combining the + fitness function and an achievement scalarizing function. In + multi-objective optimization, achievement scalarizing + functions are widely used to project a given reference point + into the Pareto optimal set. In our approach, the next + population is thus more concentrated in the area where more + preferred alternatives are assumed to lie and the whole + Pareto optimal set does not have to be generated with equal + accuracy. The approach is demonstrated by numerical + examples. +
+ +
+ + +
+[1286] +
+
+Ye Tian, Ran Cheng, Xingyi Zhang, Fan Cheng, and Yaochu Jin. + An Indicator-Based Multiobjective Evolutionary Algorithm With Reference Point Adaptation for Better Versatility. + IEEE Transactions on Evolutionary Computation, 22(4):609–622, 2018.
+[ bib | +DOI ] +
+IGD-based archiver +
+ +
+ + +
+[1287] +
+
+Tiew-On Ting, M. V. C. Rao, C. K. Loo, and S. S. Ngu. + Solving Unit Commitment Problem Using Hybrid Particle Swarm Optimization. + Journal of Heuristics, 9(6):507–520, 2003.
+[ bib | +DOI ] + +
+ + +
+[1288] +
+
+Santosh Tiwari, Georges Fadel, and Kalyanmoy Deb. + AMGA2: Improving the performance of the archive-based micro-genetic algorithm for multi-objective optimization. + Engineering Optimization, 43(4):377–401, 2011.
+[ bib ] + +
+ + +
+[1289] +
+
+V. T'Kindt, Nicolas Monmarché, F. Tercinet, and D. Laügt. + An ant colony optimization algorithm to solve a 2-machine bicriteria flowshop scheduling problem. + European Journal of Operational Research, 142(2):250–257, 2002.
+[ bib ] + +
+ + +
+[1290] +
+
+Michal K Tomczyk and Milosz Kadziński. + Decomposition-based interactive evolutionary algorithm for multiple objective optimization. + IEEE Transactions on Evolutionary Computation, 24(2):320–334, 2019.
+[ bib | +DOI ] +
+We propose a decomposition-based interactive evolutionary + algorithm (EA) for multiple objective optimization. During an + evolutionary search, a decision maker (DM) is asked to + compare pairwise solutions from the current population. Using + the Monte Carlo simulation, the proposed algorithm generates + from a uniform distribution a set of instances of the + preference model compatible with such an indirect preference + information. These instances are incorporated as the search + directions with the aim of systematically converging a + population toward the DMs most preferred region of the Pareto + front. The experimental comparison proves that the proposed + decomposition-based method outperforms the state-of-the-art + interactive counterparts of the dominance-based EAs. We also + show that the quality of constructed solutions is highly + affected by the form of the incorporated preference model. +
+
+Keywords: interactive multi-objective; decision-making +
+ +
+ + +
+[1291] +
+
+Michal K Tomczyk and Milosz Kadziński. + EMOSOR: Evolutionary multiple objective optimization guided by interactive stochastic ordinal regression. + Computers & Operations Research, 108:134–154, 2019.
+[ bib | +DOI ] +
+We propose a family of algorithms, called EMOSOR, combining + Evolutionary Multiple Objective Optimization with Stochastic + Ordinal Regression. The proposed methods ask the Decision + Maker (DM) to holistically compare, at regular intervals, a + pair of solutions, and use the Monte Carlo simulation to + construct a set of preference model instances compatible with + such indirect and incomplete information. The specific + variants of EMOSOR are distinguished by the following three + aspects. Firstly, they make use of two different preference + models, i.e., either an additive value function or a + Chebyshev function. Secondly, they aggregate the + acceptability indices derived from the stochastic analysis in + various ways, and use thus constructed indicators or + relations to sort the solutions obtained in each + generation. Thirdly, they incorporate different active + learning strategies for selecting pairs of solutions to be + critically judged by the DM. The extensive computational + experiments performed on a set of benchmark optimization + problems reveal that EMOSOR is able to bias an evolutionary + search towards a part of the Pareto front being the most + relevant to the DM, outperforming in this regard the + state-of-the-art interactive evolutionary hybrids. Moreover, + we demonstrate that the performance of EMOSOR improves in + case the forms of a preference model used by the method and + the DM's value system align. Furthermore, we discuss how + vastly incorporation of different indicators based on the + stochastic acceptability indices influences the quality of + both the best constructed solution and an entire + population. Finally, we demonstrate that our novel + questioning strategies allow to reduce a number of + interactions with the DM until a high-quality solution is + constructed or, alternatively, to discover a better solution + after the same number of interactions. +
+
+Keywords: Multiple objective optimization, Interactive evolutionary + hybrids, Stochastic ordinal regression, Preference + disaggregation, Pairwise comparisons, Active learning +
+ +
+ + +
+[1292] +
+
+Michal K Tomczyk and Milosz Kadziński. + Decomposition-based co-evolutionary algorithm for interactive multiple objective optimization. + Information Sciences, 549:178–199, 2021.
+[ bib | +DOI ] +
+We propose a novel co-evolutionary algorithm for interactive + multiple objective optimization, named CIEMO/D. It aims at + finding a region in the Pareto front that is highly relevant + to the Decision Maker (DM). For this reason, CIEMO/D asks the + DM, at regular intervals, to compare pairs of solutions from + the current population and uses such preference information + to bias the evolutionary search. Unlike the existing + interactive evolutionary algorithms dealing with just a + single population, CIEMO/D co-evolves a pool of + subpopulations in a steady-state decomposition-based + evolutionary framework. The evolution of each subpopulation + is driven by the use of a different preference model. In this + way, the algorithm explores various regions in the objective + space, thus increasing the chances of finding DM's most + preferred solution. To improve the pace of the evolutionary + search, CIEMO/D allows for the migration of solutions between + different subpopulations. It also dynamically alters the + subpopulations' size based on compatibility between the + incorporated preference models and the decision examples + supplied by the DM. The extensive experimental evaluation + reveals that CIEMO/D can successfully adjust to different + DM's decision policies. We also compare CIEMO/D with selected + state-of-the-art interactive evolutionary hybrids that make + use of the DM's pairwise comparisons, demonstrating its high + competitiveness. +
+
+Keywords: Evolutionary multiple objective optimization, Co-evolution, + Decomposition, Indirect preference information, Preference + learning +
+ +
+ + +
+[1293] +
+
+C. E. Torres, L. F. Rossi, J. Keffer, K. Li, and C.-C. Shen. + Modeling, analysis and simulation of ant-based network routing protocols. + Swarm Intelligence, 4(3):221–244, 2010.
+[ bib ] + +
+ + +
+[1294] +
+
+Heike Trautmann and Jörn Mehnen. + Preference-based Pareto optimization in certain and noisy environments. + Engineering Optimization, 41(1):23–38, January 2009.
+[ bib ] + +
+ + +
+[1295] +
+
+Vito Trianni and Manuel López-Ibáñez. + Advantages of Task-Specific Multi-Objective Optimisation in Evolutionary Robotics. + PLoS One, 10(8):e0136406, 2015.
+[ bib | +DOI ] +
+The application of multi-objective optimisation to + evolutionary robotics is receiving increasing attention. A + survey of the literature reveals the different possibilities + it offers to improve the automatic design of efficient and + adaptive robotic systems, and points to the successful + demonstrations available for both task-specific and + task-agnostic approaches (i.e., with or without reference to + the specific design problem to be tackled). However, the + advantages of multi-objective approaches over + single-objective ones have not been clearly spelled out and + experimentally demonstrated. This paper fills this gap for + task-specific approaches: starting from well-known results in + multi-objective optimisation, we discuss how to tackle + commonly recognised problems in evolutionary robotics. In + particular, we show that multi-objective optimisation (i) + allows evolving a more varied set of behaviours by exploring + multiple trade-offs of the objectives to optimise, (ii) + supports the evolution of the desired behaviour through the + introduction of objectives as proxies, (iii) avoids the + premature convergence to local optima possibly introduced by + multi-component fitness functions, and (iv) solves the + bootstrap problem exploiting ancillary objectives to guide + evolution in the early phases. We present an experimental + demonstration of these benefits in three different case + studies: maze navigation in a single robot domain, flocking + in a swarm robotics context, and a strictly collaborative + task in collective robotics. +
+ +
+ + +
+[1296] +
+
+Vito Trianni and S. Nolfi. + Engineering the evolution of self-organizing behaviors in swarm robotics: A case study. + Artificial Life, 17(3):183–202, 2011.
+[ bib ] + +
+ + +
+[1297] +
+
+Anupam Trivedi, Dipti Srinivasan, Krishnendu Sanyal, and Abhiroop Ghosh. + A survey of multiobjective evolutionary algorithms based on decomposition. + IEEE Transactions on Evolutionary Computation, 21(3):440–462, 2016.
+[ bib ] + +
+ + +
+[1298] +
+
+L.-Y. Tseng and Y.-T. Lin. + A hybrid genetic local search algorithm for the permutation flowshop scheduling problem. + European Journal of Operational Research, 198(1):84–92, 2009.
+[ bib ] + +
+ + +
+[1299] +
+
+S. Tsutsui. + Ant Colony Optimization with Cunning Ants. + Transactions of the Japanese Society for Artificial Intelligence, 22:29–36, 2007.
+[ bib | +DOI ] +
+Keywords: ant colony optimization, traveling salesman problem, cunning + ant, donor ant, local search +
+ +
+ + +
+[1300] +
+
+Alexis Tugilimana, Ashley P. Thrall, and Rajan Filomeno Coelho. + Conceptual Design of Modular Bridges Including Layout Optimization and Component Reusability. + Journal of Bridge Engineering, 22(11):04017094, 2017.
+[ bib | +DOI ] +
+Keywords: scenario-based +
+ +
+ + +
+[1301] +
+
+Renata Turkeš, Kenneth Sörensen, and Lars Magnus Hvattum. + Meta-analysis of metaheuristics: Quantifying the effect of adaptiveness in adaptive large neighborhood search. + European Journal of Operational Research, 292(2):423–42, 2021.
+[ bib | +DOI ] +
+Keywords: Metaheuristics, Meta-analysis, Adaptive large neighborhood + search +
+ +
+ + +
+[1302] +
+
+Tea Tušar and Bogdan Filipič. + Visualizing Exact and Approximated 3D Empirical Attainment Functions. + Mathematical Problems in Engineering, 2014, 2014. + Article ID 569346, 18 pages.
+[ bib | +DOI ] + +
+ + +
+[1303] +
+
+Tea Tušar and Bogdan Filipič. + Visualization of Pareto front approximations in evolutionary multiobjective optimization: A critical review and the prosection method. + IEEE Transactions on Evolutionary Computation, 19(2):225–245, 2015.
+[ bib | +DOI ] + +
+ + +
+[1304] +
+
+D. Tuyttens, Jacques Teghem, Philippe Fortemps, and K. Van Nieuwenhuyze. + Performance of the MOSA Method for the Bicriteria Assignment Problem. + Journal of Heuristics, 6:295–310, 2000.
+[ bib ] + +
+ + +
+[1305] +
+
+Amos Tversky and Daniel Kahneman. + Judgment under uncertainty: Heuristics and biases. + Science, 185(4157):1124–1131, 1974.
+[ bib ] + +
+ + +
+[1306] +
+
+Amos Tversky and Daniel Kahneman. + Loss aversion in riskless choice: a reference-dependent model. + The Quarterly Journal of Economics, 106(4):1039–1061, 1991.
+[ bib ] + +
+ + +
+[1307] +
+
+Amos Tversky. + Choice by elimination. + Journal of Mathematical Psychology, 9(4):341–367, 1972.
+[ bib ] + +
+ + +
+[1308] +
+
+Colin Twomey, Thomas Stützle, Marco Dorigo, Max Manfrin, and Mauro Birattari. + An Analysis of Communication Policies for Homogeneous Multi-colony ACO Algorithms. + Information Sciences, 180(12):2390–2404, 2010.
+[ bib | +DOI ] + +
+ + +
+[1309] +
+
+E. Ulungu and Jacques Teghem. + The two phases method: An efficient procedure to solve bi-objective combinatorial optimization problems. + Foundations of Computing and Decision Sciences, 20(2):149–165, 1995.
+[ bib ] + +
+ + +
+[1310] +
+
+E. Ulungu, Jacques Teghem, P. H. Fortemps, and D. Tuyttens. + MOSA method: a tool for solving multiobjective combinatorial optimization problems. + Journal of Multi-Criteria Decision Analysis, 8(4):221–236, 1999.
+[ bib ] + +
+ + +
+[1311] +
+
+Thijs Urlings, Rubén Ruiz, and F. Sivrikaya-Şerifoğlu. + Genetic Algorithms for Complex Hybrid Flexible Flow Line Problems. + International Journal of Metaheuristics, 1(1):30–54, 2010.
+[ bib ] + +
+ + +
+[1312] +
+
+Thijs Urlings, Rubén Ruiz, and Thomas Stützle. + Shifting Representation Search for Hybrid Flexible Flowline Problems. + European Journal of Operational Research, 207(2):1086–1095, 2010.
+[ bib | +DOI ] + +
+ + +
+[1313] +
+
+Rob J. M. Vaessens, Emile H. L. Aarts, and Jan Karel Lenstra. + A Local Search Template. + Computers & Operations Research, 25(11):969–979, 1998.
+[ bib | +DOI ] + +
+ + +
+[1314] +
+
+Claudio Lucio do Val Lopes, Flávio Vinícius Cruzeiro Martins, Elizabeth F. Wanner, and Kalyanmoy Deb. + Analyzing dominance move (MIP-DoM) indicator for multi-and many-objective optimization. + IEEE Transactions on Evolutionary Computation, 2021.
+[ bib ] + +
+ + +
+[1315] +
+
+Eva Vallada and Rubén Ruiz. + Genetic algorithms with path relinking for the minimum tardiness permutation flowshop problem. + Omega, 38(1–2):57–67, 2010.
+[ bib | +DOI ] + +
+ + +
+[1316] +
+
+Eva Vallada, Rubén Ruiz, and Jose M. Framiñán. + New hard benchmark for flowshop scheduling problems minimising makespan. + European Journal of Operational Research, 240(3):666–677, 2015.
+[ bib | +DOI ] + +
+ + +
+[1317] +
+
+Eva Vallada, Rubén Ruiz, and Gerardo Minella. + Minimising total tardiness in the m-machine flowshop problem: A review and evaluation of heuristics and metaheuristics. + Computers & Operations Research, 35(4):1350–1373, 2008.
+[ bib ] + +
+ + +
+[1318] +
+
+Pieter Vansteenwegen and Manuel Mateo. + An Iterated Local Search Algorithm for the Single-vehicle Cyclic Inventory Routing Problem. + European Journal of Operational Research, 237(3):802–813, 2014.
+[ bib ] + +
+ + +
+[1319] +
+
+Pieter Vansteenwegen, Wouter Souffriau, Greet Vanden Berghe, and Dirk Van Oudheusden. + Iterated Local Search for the Team Orienteering Problem with Time Tindows. + Computers & Operations Research, 36(12):3281–3290, 2009.
+[ bib ] + +
+ + +
+[1320] +
+
+Joaquin Vanschoren, Jan N. van Rijn, Bernd Bischl, and Luis Torgo. + OpenML: Networked Science in Machine Learning. + ACM SIGKDD Explorations Newsletter, 15(2):49–60, June 2014.
+[ bib | +DOI ] + +
+ + +
+[1321] +
+
+A. Vargha and H. D. Delaney. + A critique and improvement of the CL common language effect size statistics of McGraw and Wong. + Journal of Educational and Behavioral Statistics, 25(2):101–132, 2000.
+[ bib ] +
+Keywords: effect size test, A12 test +
+ +
+ + +
+[1322] +
+
+T. K. Varadharajan and C. Rajendran. + A multi-objective simulated-annealing algorithm for scheduling in flowshops to minimize the makespan and total flowtime of jobs. + European Journal of Operational Research, 167(3):772–795, 2005.
+[ bib ] + +
+ + +
+[1323] +
+
+Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. + Attention Is All You Need. + Arxiv preprint arXiv:1706.03762, 2017.
+[ bib | +http ] +
+The dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 English-to-German translation task, improving over the existing best results, including ensembles by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.8 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data. +
+ +
+ + +
+[1324] +
+
+A. Vasan and Slobodan P. Simonovic. + Optimization of Water Distribution Network Design Using Differential Evolution. + Journal of Water Resources Planning and Management, ASCE, 136(2):279–287, 2010.
+[ bib ] + +
+ + +
+[1325] +
+
+Sergei Vassilvitskii and Mihalis Yannakakis. + Efficiently computing succinct trade-off curves. + Theoretical Computer Science, 348(2-3):334–356, 2005.
+[ bib ] + +
+ + +
+[1326] +
+
+J. A. Vázquez-Rodríguez and Gabriela Ochoa. + On the Automatic Discovery of Variants of the NEH Procedure for Flow Shop Scheduling Using Genetic Programming. + Journal of the Operational Research Society, 62(2):381–396, 2010.
+[ bib ] + +
+ + +
+[1327] +
+
+Daniel Vaz, Luís Paquete, Carlos M. Fonseca, Kathrin Klamroth, and Michael Stiglmayr. + Representation of the non-dominated set in biobjective discrete optimization. + Computers & Operations Research, 63:172–186, 2015.
+[ bib | +DOI ] + +
+ + +
+[1328] +
+
+David A. Van Veldhuizen and Gary B. Lamont. + Multiobjective Evolutionary Algorithms: Analyzing the State-of-the-art. + Evolutionary Computation, 8(2):125–147, 2000.
+[ bib | +DOI ] + +
+ + +
+[1329] +
+
+Amit Verma and Mark Lewis. + Penalty and partitioning techniques to improve performance of QUBO solvers. + Discrete Optimization, p.  100594, 2020.
+[ bib | +DOI ] +
+Keywords: Quadratic Unconstrained Binary Optimization, Nonlinear + optimization, Pseudo-Boolean optimization, Equality + constraint, Inequality constraint +
+ +
+ + +
+[1330] +
+
+Sébastien Verel, Arnaud Liefooghe, Laetitia Jourdan, and Clarisse Dhaenens. + On the Structure of Multiobjective Combinatorial Search Space: MNK-landscapes with Correlated Objectives. + European Journal of Operational Research, 227(2):331–342, 2013.
+[ bib | +DOI ] + +
+ + +
+[1331] +
+
+Paolo Viappiani, Boi Faltings, and Pearl Pu. + Preference-based Search using Example-Critiquing with Suggestions. + Journal of Artificial Intelligence Research, 27:465–503, 2006.
+[ bib ] + +
+ + +
+[1332] +
+
+Paolo Viappiani, Pearl Pu, and Boi Faltings. + Preference-based Search with Adaptive Recommendations. + AI Communications, 21(2):155–175, 2008.
+[ bib ] + +
+ + +
+[1333] +
+
+Thibaut Vidal, Teodor Gabriel Crainic, Michel Gendreau, and Christian Prins. + Heuristics for Multi-attribute Vehicle Routing Problems: A Survey and Synthesis. + European Journal of Operational Research, 231(1):1–21, 2013.
+[ bib ] + +
+ + +
+[1334] +
+
+Thibaut Vidal, Teodor Gabriel Crainic, Michel Gendreau, and Christian Prins. + A Unified Solution Framework for Multi-attribute Vehicle Routing Problems. + European Journal of Operational Research, 234(3):658–673, 2014.
+[ bib ] + +
+ + +
+[1335] +
+
+Matheus Guedes Vilas Boas, Haroldo Gambini Santos, Luiz Henrique de Campos Merschmann, and Greet Vanden Berghe. + Optimal decision trees for the algorithm selection problem: integer programming based approaches. + International Transactions in Operational Research, 28(5):2759–2781, 2021.
+[ bib | +DOI ] + +
+ + +
+[1336] +
+
+Christos Voudouris and Edward P. K. Tsang. + Guided Local Search and its Application to the Travelling Salesman Problem. + European Journal of Operational Research, 113(2):469–499, 1999.
+[ bib ] + +
+ + +
+[1337] +
+
+Jyrki Wallenius. + Comparative Evaluation of Some Interactive Approaches to Multicriterion Optimization. + Management Science, 21(12):1387–1396, 1975.
+[ bib ] + +
+ + +
+[1338] +
+
+C. Walshaw and M. Cross. + Mesh Partitioning: A Multilevel Balancing and Refinement Algorithm. + SIAM Journal on Scientific Computing, 22(1):63–80, 2000.
+[ bib | +DOI ] + +
+ + +
+[1339] +
+
+David J. Walker, Richard M. Everson, and Jonathan E. Fieldsend. + Visualizing mutually nondominating solution sets in many-objective optimization. + IEEE Transactions on Evolutionary Computation, 17(2):165–184, 2012.
+[ bib ] + +
+ + +
+[1340] +
+
+Chengen Wang, Chengbin Chu, and Jean-Marie Proth. + Heuristic Approaches for n/m/F/ΣCi Scheduling Problems. + European Journal of Operational Research, 96(3):636–644, 1997.
+[ bib | +DOI ] + +
+ + +
+[1341] +
+
+Handing Wang, Licheng Jiao, and Xin Yao. + TwoArch2: An improved two-archive algorithm for many-objective optimization. + IEEE Transactions on Evolutionary Computation, 19(4):524–541, 2015.
+[ bib ] + +
+ + +
+[1342] +
+
+Xilu Wang, Yaochu Jin, Sebastian Schmitt, Markus Olhofer, and Richard Allmendinger. + Transfer learning based surrogate assisted evolutionary bi-objective optimization for objectives with different evaluation times. + Knowledge-Based Systems, 227:107190, 2021.
+[ bib | +DOI ] + +
+ + +
+[1343] +
+
+Yang Wang, Zhipeng Lü, Fred Glover, and Jin-Kao Hao. + Probabilistic GRASP-Tabu Search algorithms for the UBQP problem. + Computers & Operations Research, 40(12):3100–3107, 2013.
+[ bib | +DOI ] + +
+ + +
+[1344] +
+
+Yang Wang, Zhipeng Lü, Fred Glover, and Jin-Kao Hao. + Backbone Guided Tabu Search for Solving the UBQP Problem. + Journal of Heuristics, 19(4):679–695, 2013.
+[ bib | +DOI ] + +
+ + +
+[1345] +
+
+Rui Wang, Robin C. Purshouse, and Peter J. Fleming. + Preference-Inspired Coevolutionary Algorithms for Many-Objective Optimization. + IEEE Transactions on Evolutionary Computation, 17(4):474–494, 2013.
+[ bib ] + +
+ + +
+[1346] +
+
+Rui Wang, Jian Xiong, Min-fan He, Liang Gao, and Ling Wang. + Multi-objective optimal design of hybrid renewable energy system under multiple scenarios. + Renewable Energy, 151:226–237, 2020.
+[ bib | +DOI ] + +
+ + +
+[1347] +
+
+Yang Wang, Zhipeng Lü, Fred Glover, and Jin-Kao Hao. + Path relinking for unconstrained binary quadratic programming. + European Journal of Operational Research, 223(3):595–604, 2012.
+[ bib | +DOI ] + +
+ + +
+[1348] +
+
+Jean-Paul Watson, L. Barbulescu, Darrell Whitley, and Adele E. Howe. + Contrasting Structured and Random Permutation Flow-Shop Scheduling Problems: Search Space Topology and Algorithm Performance. + INFORMS Journal on Computing, 14(2):98–123, 2002.
+[ bib ] + +
+ + +
+[1349] +
+
+Jean-Paul Watson, J. C. Beck, A. E. Howe, and Darrell Whitley. + Problem Difficulty for Tabu Search in Job-Shop Scheduling. + Artificial Intelligence, 143(2):189–217, 2003.
+[ bib ] + +
+ + +
+[1350] +
+
+Jean-Paul Watson, Adele E Howe, and Darrell Whitley. + Deconstructing Nowicki and Smutnicki's i-TSAB tabu search algorithm for the job-shop scheduling problem. + Computers & Operations Research, 33(9):2623–2644, 2006.
+[ bib ] + +
+ + +
+[1351] +
+
+Abigail A. Watson and Joseph R. Kasprzyk. + Incorporating deeply uncertain factors into the many objective search process. + Environmental Modelling & Software, 89:159–171, 2017.
+[ bib ] +
+Keywords: scenario-based +
+ +
+ + +
+[1352] +
+
+Edward J. Wegman. + Hyperdimensional data analysis using parallel coordinates. + Journal of the American Statistical Association, 85(411):664–675, 1990.
+[ bib ] + +
+ + +
+[1353] +
+
+Edward D. Weinberger. + Local properties of Kauffman's N-k model: A tunably rugged energy landscape. + Physical Review A, 44(10):6399, 1991.
+[ bib ] + +
+ + +
+[1354] +
+
+Karl Weiss, Taghi M. Khoshgoftaar, and DingDing Wang. + A survey of transfer learning. + Journal of Big Data, 3(1):1–40, 2016.
+[ bib ] + +
+ + +
+[1355] +
+
+Bernard L. Welch. + The significance of the difference between two means when the population variances are unequal. + Biometrika, 29(3/4):350–362, 1938.
+[ bib ] + +
+ + +
+[1356] +
+
+Simon Wessing and Manuel López-Ibáñez. + Latin Hypercube Designs with Branching and Nested Factors for Initialization of Automatic Algorithm Configuration. + Evolutionary Computation, 27(1):129–145, 2018.
+[ bib | +DOI ] + +
+ + +
+[1357] +
+
+Dennis Weyland. + A Rigorous Analysis of the Harmony Search Algorithm: How the Research Community can be misled by a “novel” Methodology. + International Journal of Applied Metaheuristic Computing, 12(2):50–60, 2010.
+[ bib ] + +
+ + +
+[1358] +
+
+Dennis Weyland. + A critical analysis of the harmony search algorithm: How not to solve Sudoku. + Operations Research Perspectives, 2:97–105, 2015.
+[ bib ] + +
+ + +
+[1359] +
+
+D. R. White, A. Arcuri, and J. A. Clark. + Evolutionary Improvement of Programs. + IEEE Transactions on Evolutionary Computation, 15(4):515–538, 2011.
+[ bib ] + +
+ + +
+[1360] +
+
+L. While, L. Bradstreet, and L. Barone. + A Fast Way of Calculating Exact Hypervolumes. + IEEE Transactions on Evolutionary Computation, 16(1):86–95, 2012.
+[ bib ] + +
+ + +
+[1361] +
+
+Darrell Whitley, Soraya Rana, John Dzubera, and Keith E. Mathias. + Evaluating Evolutionary Algorithms. + Artificial Intelligence, 85:245–296, 1996.
+[ bib ] + +
+ + +
+[1362] +
+
+R. J. Williams. + Simple Statistical Gradient-Following Algorithms for Connectionist Reinforcement Learning. + Machine Learning, 8(3):229–256, 1992.
+[ bib ] + +
+ + +
+[1363] +
+
+P. Winkler. + Random Orders. + Order, 1:317–331, 1985.
+[ bib ] +
+Showed that fraction of Pareto-optimal increases with number + of objectives +
+ +
+ + +
+[1364] +
+
+Carsten Witt. + Analysis of an Iterated Local Search Algorithm for Vertex Cover in Sparse Random Graphs. + Theoretical Computer Science, 425:117–125, 2012.
+[ bib ] + +
+ + +
+[1365] +
+
+D. H. Wolpert and W. G. Macready. + No Free Lunch Theorems for Optimization. + IEEE Transactions on Evolutionary Computation, 1(1):67–82, 1997.
+[ bib | +DOI ] + +
+ + +
+[1366] +
+
+Matthew J. Woodruff, Patrick M. Reed, and Timothy W. Simpson. + Many objective visual analytics: rethinking the design of complex engineered systems. + Structural and Multidisciplinary Optimization, 48(1):201–219, 2013.
+[ bib | +DOI ] + +
+ + +
+[1367] +
+
+David L. Woodruff, Ulrike Ritzinger, and Johan Oppen. + Research Note: The Point of Diminishing Returns in Heuristic Search. + International Journal of Metaheuristics, 1(3):222–231, 2011.
+[ bib | +DOI ] +
+Keywords: anytime +
+ +
+ + +
+[1368] +
+
+H. S. Woo and D. S. Yim. + A Heuristic Algorithm for Mean Flowtime Objective in Flowshop Scheduling. + Computers & Operations Research, 25(3):175–182, 1998.
+[ bib ] + +
+ + +
+[1369] +
+
+Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, et al. + Google's neural machine translation system: Bridging the gap between human and machine translation. + Arxiv preprint arXiv:1609.08144 [cs.CL], 2016.
+[ bib | +http ] + +
+ + +
+[1370] +
+
+Xindong Wu, Xingquan Zhu, Gong-Qing Wu, and Wei Ding. + Data mining with big data. + IEEE Transactions on Knowledge and Data Engineering, 26(1):97–107, 2014.
+[ bib ] + +
+ + +
+[1371] +
+
+Adilson Elias Xavier and Vinicius Layter Xavier. + Solving the minimum sum-of-squares clustering problem by hyperbolic smoothing and partition into boundary and gravitational regions. + Pattern Recognition, 44(1):70–77, 2011.
+[ bib | +DOI ] +
+Keywords: Cluster analysis, Min-sum-min problems, Nondifferentiable + programming, Smoothing +
+ +
+ + +
+[1372] +
+
+B. Xin, L. Chen, J. Chen, Hisao Ishibuchi, K. Hirota, and B. Liu. + Interactive Multiobjective Optimization: A Review of the State-of-the-Art. + IEEE Access, 6:41256–41279, 2018.
+[ bib | +DOI ] +
+Interactive multiobjective optimization (IMO) aims at finding + the most preferred solution of a decision maker with the + guidance of his/her preferences which are provided + progressively. During the process, the decision maker can + adjust his/her preferences and explore only interested + regions of the search space. In recent decades, IMO has + gradually become a common interest of two distinct + communities, namely, the multiple criteria decision making + (MCDM) and the evolutionary multiobjective optimization + (EMO). The IMO methods developed by the MCDM community + usually use the mathematical programming methodology to + search for a single preferred Pareto optimal solution, while + those which are rooted in EMO often employ evolutionary + algorithms to generate a representative set of solutions in + the decision maker's preferred region. This paper aims to + give a review of IMO research from both MCDM and EMO + perspectives. Taking into account four classification + criteria including the interaction pattern, preference + information, preference model, and search engine (i.e., + optimization algorithm), a taxonomy is established to + identify important IMO factors and differentiate various IMO + methods. According to the taxonomy, state-of-the-art IMO + methods are categorized and reviewed and the design ideas + behind them are summarized. A collection of important issues, + e.g., the burdens, cognitive biases and preference + inconsistency of decision makers, and the performance + measures and metrics for evaluating IMO methods, are + highlighted and discussed. Several promising directions + worthy of future research are also presented. +
+
+Keywords: Decision making, Evolutionary computation, Pareto + optimization, Evolutionary multiobjective optimization, + interactive multiobjective optimization, multiple criteria + decision making, preference information, preference models +
+ +
+ + +
+[1373] +
+
+Jiefeng Xu, Steve Y. Chiu, and Fred Glover. + Fine-tuning a tabu search algorithm with statistical tests. + International Transactions in Operational Research, 5(3):233–244, 1998.
+[ bib | +DOI ] + +
+ + +
+[1374] +
+
+Lin Xu, Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + SATzilla: Portfolio-based Algorithm Selection for SAT. + Journal of Artificial Intelligence Research, 32:565–606, June 2008.
+[ bib | +DOI | +epub ] + +
+ + +
+[1375] +
+
+Hongyun Xu, Zhipeng Lü, and T. C. E. Cheng. + Iterated Local Search for Single-machine Scheduling with Sequence-dependent Setup Times to Minimize Total Weighted Tardiness. + Journal of Scheduling, 17(3):271–287, 2014.
+[ bib ] + +
+ + +
+[1376] +
+
+Dong-Ling Xu and Jian-Bo Yang. + Intelligent Decision System for Self-Assessment. + Journal of Multi-Criteria Decision Analysis, 12(1):43–60, 2003.
+[ bib | +DOI ] +
+Many small and medium enterprises (SMEs) in the UK use the + beta (Business Excellence Through Action) approach to the + EFQM Excellence Model to conduct business excellence + self-assessment, which is in essence a multiple criteria + decision analysis (MCDA) problem. This paper introduces a + decision support software package called Intelligent Decision + System (IDS) to implement the beta approach. It is + demonstrated in the paper that the IDS-beta package can + provide not only average scores but also the following + numerical results and graphical displays on: Distributed + assessment results to demonstrate the diversity of company + performances The performance range to cater for incomplete + assessment information Comparisons between current + performances and past performances, among different companies + among different action plans. Strengths and weaknesses The + IDS-beta package also provides a structured knowledge base to + help assessors to make judgements more objectively. The + knowledge base contains guidelines provided by the developers + of the beta approach, best practices gathered from research + on award winning organizations, evidence collected from + companies being assessed and comments provided by assessors + to record the reasons why a specific criterion is assessed to + a certain grade for a company. Four small UK companies, the + industry partners of the research project, have carried out + the preliminary self-assessment using the package. The + results and experience of the application are discussed at + the end of the paper. +
+
+Keywords: decision support system, business excellence, MCDA, quality + award, self-assessment, the evidential reasoning approach +
+ +
+ + +
+[1377] +
+
+Mutsunori Yagiura, M. Kishida, and Toshihide Ibaraki. + A 3-Flip Neighborhood Local Search for the Set Covering Problem. + European Journal of Operational Research, 172(2):472–499, 2006.
+[ bib ] + +
+ + +
+[1378] +
+
+Yuki Yamada. + How to Crack Pre-registration: Toward Transparent and Open Science. + Frontiers in Psychology, 9, September 2018.
+[ bib | +DOI ] +
+Keywords: HARKing; PARKing +
+ +
+ + +
+[1379] +
+
+Kaifeng Yang, Michael T. M. Emmerich, André H. Deutz, and Thomas Bäck. + Multi-Objective Bayesian Global Optimization using Expected Hypervolume Improvement Gradient. + Swarm and Evolutionary Computation, 44:945–956, February 2019.
+[ bib | +DOI ] +
+Keywords: Bayesian Optimisation with preferences +
+ +
+ + +
+[1380] +
+
+Y. Yang, S. Kreipl, and M. L. Pinedo. + Heuristics for Minimizing Total Weighted Tardiness in Flexible Flow Shops. + Journal of Scheduling, 3(2):89–108, 2000.
+[ bib ] + +
+ + +
+[1381] +
+
+S. Yang, Miqing Li, X. Liu, and J. Zheng. + A Grid-Based Evolutionary Algorithm for Many-Objective Optimization. + IEEE Transactions on Evolutionary Computation, 17(5):721–736, 2013.
+[ bib | +DOI ] +
+epsilon-grid +
+ +
+ + +
+[1382] +
+
+Furong Ye, Carola Doerr, Hao Wang, and Thomas Bäck. + Automated Configuration of Genetic Algorithms by Tuning for Anytime Performance. + IEEE Transactions on Evolutionary Computation, 26(6):1526–1538, 2022.
+[ bib | +DOI ] + +
+ + +
+[1383] +
+
+Vincent F. Yu and Shih-Wei Lin. + Iterated Greedy Heuristic for the Time-dependent Prize-collecting Arc Routing Problem. + Computers and Industrial Engineering, 90:54–66, 2015.
+[ bib ] + +
+ + +
+[1384] +
+
+G. Yu, R. S. Powell, and M. J. H. Sterling. + Optimized Pump Scheduling in Water Distribution Systems. + Journal of Optimization Theory and Applications, 83(3):463–488, 1994.
+[ bib ] + +
+ + +
+[1385] +
+
+Zhi Yuan, Marco A. Montes de Oca, Thomas Stützle, and Mauro Birattari. + Continuous Optimization Algorithms for Tuning Real and Integer Algorithm Parameters of Swarm Intelligence Algorithms. + Swarm Intelligence, 6(1):49–75, 2012.
+[ bib ] + +
+ + +
+[1386] +
+
+Q. Zeng and Z. Yang. + Integrating Simulation and Optimization to Schedule Loading Operations in Container Terminals. + Computers & Operations Research, 36(6):1935–1944, 2009.
+[ bib | +DOI ] + +
+ + +
+[1387] +
+
+Tiantian Zhang, Michael Georgiopoulos, and Georgios C. Anagnostopoulos. + Multi-Objective Model Selection via Racing. + IEEE Transactions on Cybernetics, 46(8):1863–1876, 2016.
+[ bib ] + +
+ + +
+[1388] +
+
+Qingfu Zhang and Hui Li. + MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition. + IEEE Transactions on Evolutionary Computation, 11(6):712–731, 2007.
+[ bib | +DOI ] +
+Introduces penalty-based boundary intersection (PBI) + function +
+ +
+ + +
+[1389] +
+
+Jingqiao Zhang and Arthur C. Sanderson. + JADE: Adaptive differential evolution with optional external archive. + IEEE Transactions on Evolutionary Computation, 13(5):945–958, 2009.
+[ bib | +DOI ] + +
+ + +
+[1390] +
+
+H. Zhao and Sudha Ram. + Constrained cascade generalization of decision trees. + IEEE Transactions on Knowledge and Data Engineering, 16(6):727–739, 2004.
+[ bib | +DOI ] + +
+ + +
+[1391] +
+
+Lu Zhen and Dao-Fang Chang. + A bi-objective model for robust berth allocation scheduling. + Computers and Industrial Engineering, 63(1):262–273, 2012.
+[ bib ] + +
+ + +
+[1392] +
+
+A. Zhou, Qingfu Zhang, and Yaochu Jin. + Approximating the set of Pareto-optimal solutions in both the decision and objective spaces by an estimation of distribution algorithm. + IEEE Transactions on Evolutionary Computation, 13(5):1167–1189, 2009.
+[ bib | +DOI ] +
+Keywords: multi-modal, IGDX +
+ +
+ + +
+[1393] +
+
+Shlomo Zilberstein. + Using Anytime Algorithms in Intelligent Systems. + AI Magazine, 17(3):73–83, 1996.
+[ bib | +DOI | +epub ] +
+Anytime algorithms give intelligent systems the capability to trade deliberation time for quality of results. This capability is essential for successful operation in domains such as signal interpretation, real-time diagnosis and repair, and mobile robot control. What characterizes these domains is that it is not feasible (computationally) or desirable (economically) to compute the optimal answer. This article surveys the main control problems that arise when a system is composed of several anytime algorithms. These problems relate to optimal management of uncertainty and precision. After a brief introduction to anytime computation, I outline a wide range of existing solutions to the metalevel control problem and describe current work that is aimed at increasing the applicability of anytime computation. +
+
+Keywords: performance profiles +
+ +
+ + +
+[1394] +
+
+Stanley Zionts and Jyrki Wallenius. + An interactive multiple objective linear programming method for a class of underlying nonlinear utility functions. + Management Science, 29(5):519–529, 1983.
+[ bib ] + +
+ + +
+[1395] +
+
+Eckart Zitzler and Lothar Thiele. + Multiobjective Evolutionary Algorithms: A Comparative Case Study and the Strength Pareto Approach. + IEEE Transactions on Evolutionary Computation, 3(4):257–271, 1999.
+[ bib | +DOI ] +
+Proposed SPEA, + http://www.tik.ee.ethz.ch/sop/publicationListFiles/zt1999a.pdf +
+ +
+ + +
+[1396] +
+
+Eckart Zitzler, Lothar Thiele, and Johannes Bader. + On Set-Based Multiobjective Optimization. + IEEE Transactions on Evolutionary Computation, 14(1):58–79, 2010.
+[ bib | +DOI ] +
+Proposed SPAM and explores combination of quality indicators +
+
+Keywords: Performance assessment; Preference articulation; refinement; + Set Partitioning; Set-preference +
+ +
+ + +
+[1397] +
+
+Eckart Zitzler, Lothar Thiele, and Kalyanmoy Deb. + Comparison of Multiobjective Evolutionary Algorithms: Empirical Results. + Evolutionary Computation, 8(2):173–195, 2000.
+[ bib | +DOI ] +
+Keywords: ZDT benchmark +
+ +
+ + +
+[1398] +
+
+Eckart Zitzler, Lothar Thiele, Marco Laumanns, Carlos M. Fonseca, and Viviane Grunert da Fonseca. + Performance Assessment of Multiobjective Optimizers: an Analysis and Review. + IEEE Transactions on Evolutionary Computation, 7(2):117–132, 2003.
+[ bib | +DOI ] +
+Proposed the combination of quality indicators; proposed epsilon-indicator +
+ +
+ + +
+[1399] +
+
+M. Zlochin, Mauro Birattari, N. Meuleau, and Marco Dorigo. + Model-Based Search for Combinatorial Optimization: A Critical Survey. + Annals of Operations Research, 131(1–4):373–395, 2004.
+[ bib ] + +
+ + +
+[1400] +
+
+Bernd Bischl, Jakob Richter, Jakob Bossek, Daniel Horn, Janek Thomas, and Michel Lang. + mlrMBO: A Modular Framework for Model-Based Optimization of Expensive Black-Box Functions. + Arxiv preprint arXiv:1703.03373 [stat.ML], 2017.
+[ bib | +http ] + +
+ + +
+[1401] +
+
+Oscar Cordón, Francisco Herrera, and Thomas Stützle. + Special Issue on Ant Colony Optimization: Models and Applications. + Mathware & Soft Computing, 9(3):137–268, 2002.
+[ bib ] + +
+ + +
+[1402] +
+
+G. McCormick and R. S. Powell. + Optimal Pump Scheduling in Water Supply Systems with Maximum Demand Charges. + Journal of Water Resources Planning and Management, ASCE, 129(5):372–379, September / October 2003.
+[ bib ] + +
+ + +
+[1403] +
+
+Gang Quan, Garrison W. Greenwood, Donglin Liu, and Sharon Hu. + Searching for multiobjective preventive maintenance schedules: Combining preferences with evolutionary algorithms. + European Journal of Operational Research, 177(3):1969–1984, 2007.
+[ bib | +DOI ] +
+Heavy industry maintenance facilities at aircraft service + centers or railroad yards must contend with scheduling + preventive maintenance tasks to ensure critical equipment + remains available. The workforce that performs these tasks + are often high-paid, which means the task scheduling should + minimize worker idle time. Idle time can always be minimized + by reducing the workforce. However, all preventive + maintenance tasks should be completed as quickly as possible + to make equipment available. This means the completion time + should be also minimized. Unfortunately, a small workforce + cannot complete many maintenance tasks per hour. Hence, there + is a tradeoff: should the workforce be small to reduce idle + time or should it be large so more maintenance can be + performed each hour? A cost effective schedule should strike + some balance between a minimum schedule and a minimum size + workforce. This paper uses evolutionary algorithms to solve + this multiobjective problem. However, rather than conducting + a conventional dominance-based Pareto search, we introduce a + form of utility theory to find Pareto optimal solutions. The + advantage of this method is the user can target specific + subsets of the Pareto front by merely ranking a small set of + initial solutions. A large example problem is used to + demonstrate our method. +
+
+Keywords: Evolutionary computations, Scheduling, Utility theory, + Preventive maintenance, Multi-objective optimization, + ranking-based, interactive +
+ +
+ + +
+[1404] +
+
+Marvin N. Wright and Andreas Ziegler. + ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. + Arxiv preprint arXiv:1508.04409 [stat.ML], 2015.
+[ bib | +http ] + +
+ + +
+[1405] +
+
+Marvin N. Wright and Andreas Ziegler. + ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. + Journal of Statistical Software, 77(1):1–17, 2017.
+[ bib | +DOI ] + +
+ + +
+[1406] +
+
+F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. + Scikit-learn: Machine learning in Python. + Journal of Machine Learning Research, 12:2825–2830, 2011.
+[ bib ] + +
+ + +
+[1407] +
+
+Jakobus E. van Zyl, Dragan A. Savic, and Godfrey A. Walters. + Operational Optimization of Water Distribution Systems using a Hybrid Genetic Algorithm. + Journal of Water Resources Planning and Management, ASCE, 130(2):160–170, March 2004.
+[ bib ] + +
+ + +
+[1408] +
+
+AAAI. + 35th AAAI Conference on Artificial Intelligence: Reproducibility Checklist. + https://aaai.org/Conferences/AAAI-21/reproducibility-checklist/, 2021. + Last accessed: June 6th, 2021.
+[ bib ] + +
+ + +
+[1409] +
+
+ACM. + Artifact Review and Badging Version 1.1. + https://www.acm.org/publications/policies/artifact-review-and-badging-current, August 2020.
+[ bib ] + +
+ + +
+[1410] +
+
+Emile H. L. Aarts, Jan H. M. Korst, and Wil Michiels. + Simulated Annealing. + In E. K. Burke and G. Kendall, editors, Search Methodologies, pp.  187–210. Springer, Boston, MA, 2005.
+[ bib | +DOI ] + +
+ + +
+[1411] +
+
+Hussein A. Abbass. + The self-adaptive Pareto differential evolution algorithm. + In Proceedings of the 2002 Congress on Evolutionary Computation (CEC'02), pp.  831–836, Piscataway, NJ, 2002. IEEE Press.
+[ bib ] + +
+ + +
+[1412] +
+
+Ricardo Henrique Remes de Lima and Aurora Trinidad Ramirez Pozo. + A study on auto-configuration of Multi-Objective Particle Swarm Optimization Algorithm. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  718–725, Piscataway, NJ, 2017. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1413] +
+
+Hussein A. Abbass, Ruhul Sarker, and Charles Newton. + PDE: a Pareto-frontier differential evolution approach for multi-objective optimization problems. + In Proceedings of the 2001 Congress on Evolutionary Computation (CEC'01), pp.  971–978, Piscataway, NJ, 2001. IEEE Press.
+[ bib ] + +
+ + +
+[1414] +
+
+F. Ben Abdelaziz, S. Krichen, and J. Chaouachi. + A hybrid heuristic for multiobjective knapsack problems. + In M. G. C. Resende and J. Pinho de Souza, editors, Proceedings of MIC 1997, the 2nd Metaheuristics International Conference, pp.  205–212, 1997.
+[ bib | +DOI ] + +
+ + +
+[1415] +
+
+A. Acan. + An external memory implementation in ant colony optimization. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  73–84. Springer, Heidelberg, Germany, 2004.
+[ bib ] +
+Keywords: memory-based ACO +
+ +
+ + +
+[1416] +
+
+A. Acan. + An external partial permutations memory for ant colony optimization. + In G. R. Raidl and J. Gottlieb, editors, Proceedings of EvoCOP 2005 – 5th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3448 of Lecture Notes in Computer Science, pp.  1–11. Springer, Heidelberg, Germany, 2005.
+[ bib ] +
+Keywords: memory-based ACO +
+ +
+ + +
+[1417] +
+
+Hernán E. Aguirre, Saúl Zapotecas, Arnaud Liefooghe, Sébastien Verel, and Kiyoshi Tanaka. + Approaches for Many-Objective Optimization: Analysis and Comparison on MNK-Landscapes. + In S. Bonnevay et al., editors, Artificial Evolution: 12th International Conference, Evolution Artificielle, EA, 2015, volume 9554 of Lecture Notes in Computer Science, pp.  14–28. Springer, Cham, Switzerland, 2016.
+[ bib | +DOI ] + +
+ + +
+[1418] +
+
+Weiwei Cheng, Jens Hühn, and Eyke Hüllermeier. + Decision Tree and Instance-Based Learning for Label Ranking. + In A. P. Danyluk, L. Bottou, and M. L. Littman, editors, Proceedings of the 26th International Conference on Machine Learning, ICML 2009, pp.  161–168, New York, NY, 2009. ACM Press.
+[ bib | +DOI ] + +
+ + +
+[1419] +
+
+Hernán E. Aguirre and Kiyoshi Tanaka. + Many-Objective Optimization by Space Partitioning and Adaptive ε-Ranking on MNK-Landscapes. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  407–422. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[1420] +
+
+Hernán E. Aguirre. + Advances on Many-objective Evolutionary Optimization. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2013, pp.  641–666. ACM Press, New York, NY, 2013.
+[ bib ] +
+Keywords: many-objective evolutionary optimization +
+ +
+ + +
+[1421] +
+
+A. Aho, J. Hopcroft, and J. Ullman. + Data structures and algorithms. + Addison-Wesley, Reading, MA, 1983.
+[ bib ] + +
+ + +
+[1422] +
+
+R. K. Ahuja, T. Magnanti, and J. B. Orlin. + Network Flows: Theory, Algorithms and Applications. + Prentice-Hall, 1993.
+[ bib ] + +
+ + +
+[1423] +
+
+Uwe Aickelin, Edmund K. Burke, and Jingpeng Li. + Improved Squeaky Wheel Optimisation for Driver Scheduling. + In T. P. Runarsson, H.-G. Beyer, E. K. Burke, J.-J. Merelo, D. Whitley, and X. Yao, editors, Parallel Problem Solving from Nature – PPSN IX, volume 4193 of Lecture Notes in Computer Science, pp.  182–191. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[1424] +
+
+Hassene Aissi and Bernard Roy. + Robustness in Multi-criteria Decision Aiding. + In M. Ehrgott, J. R. Figueira, and S. Greco, editors, Trends in Multiple Criteria Decision Analysis, volume 142 of International Series in Operations Research & Management Science, chapter 4, pp.  87–121. Springer, US, 2010.
+[ bib ] + +
+ + +
+[1425] +
+
+Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, and Masanori Koyama. + Optuna: A Next-generation Hyperparameter Optimization Framework. + In Teredesai et al., editors, 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.  2623–2631. ACM Press, New York, NY, July 2019.
+[ bib | +DOI ] + +
+ + +
+[1426] +
+
+S. M. Aktürk, Alper Atamtürk, and S. Gürel. + A Strong Conic Quadratic Reformulation for Machine-Job Assignment with Controllable Processing Times. + Research Report BCOL.07.01, University of California-Berkeley, 2007.
+[ bib ] + +
+ + +
+[1427] +
+
+I. Alaya, Christine Solnon, and Khaled Ghédira. + Ant Colony Optimization for Multi-Objective Optimization Problems. + In 19th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2007), volume 1, pp.  450–457. IEEE Computer Society Press, Los Alamitos, CA, 2007.
+[ bib ] + +
+ + +
+[1428] +
+
+I. Alaya, Christine Solnon, and Khaled Ghédira. + Ant algorithm for the multi-dimensional knapsack problem. + In B. Filipič and J. Šilc, editors, International Conference on Bioinspired Optimization Methods and their Applications (BIOMA 2004), pp.  63–72, 2004.
+[ bib | +http ] + +
+ + +
+[1429] +
+
+Enrique Alba and Francisco Chicano. + ACOhg: dealing with huge graphs. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  10–17. ACM Press, New York, NY, 2007.
+[ bib | +DOI ] + +
+ + +
+[1430] +
+
+Mohamad Alissa, Kevin Sim, and Emma Hart. + Algorithm Selection Using Deep Learning without Feature Extraction. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  198–206. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[1431] +
+
+Sam Allen, Edmund K. Burke, Matthew R. Hyde, and Graham Kendall. + Evolving reusable 3d packing heuristics with genetic programming. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2009, pp.  931–938. ACM Press, New York, NY, 2009.
+[ bib | +DOI ] +
+Keywords: hyper-heuristic +
+ +
+ + +
+[1432] +
+
+Richard Allmendinger and Joshua D. Knowles. + Evolutionary Optimization on Problems Subject to Changes of Variables. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  151–160. Springer, Heidelberg, Germany, 2010.
+[ bib ] +
+Motivated by an experimental problem involving the + identification of effective drug combinations drawn from a + non-static drug library, this paper examines evolutionary + algorithm strategies for dealing with changes of + variables. We consider four standard techniques from dynamic + optimization, and propose one new technique. The results show + that only little additional diversity needs to be introduced + into the population when changing a small number of + variables, while changing many variables or optimizing a + rugged landscape requires often a restart of the optimization + process +
+ +
+ + +
+[1433] +
+
+Richard Allmendinger and Joshua D. Knowles. + Evolutionary Search in Lethal Environments. + In International Conference on Evolutionary Computation Theory and Applications, pp.  63–72. SciTePress, 2011.
+[ bib | +DOI | +epub ] + +
+ + +
+[1434] +
+
+Richard Allmendinger and Joshua D. Knowles. + Policy Learning in Resource-Constrained Optimization. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  1971–1979. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] +
+We consider an optimization scenario in which resources are + required in the evaluation process of candidate + solutions. The challenge we are focussing on is that certain + resources have to be committed to for some period of time + whenever they are used by an optimizer. This has the effect + that certain solutions may be temporarily non-evaluable + during the optimization. Previous analysis revealed that + evolutionary algorithms (EAs) can be effective against this + resourcing issue when augmented with static strategies for + dealing with non-evaluable solutions, such as repairing, + waiting, or penalty methods. Moreover, it is possible to + select a suitable strategy for resource-constrained problems + offline if the resourcing issue is known in advance. In this + paper we demonstrate that an EA that uses a reinforcement + learning (RL) agent, here Sarsa(λ), to learn + offline when to switch between static strategies, can be more + effective than any of the static strategies themselves. We + also show that learning the same task as the RL agent but + online using an adaptive strategy selection method, here + D-MAB, is not as effective; nevertheless, online learning is + an alternative to static strategies. +
+ +
+ + +
+[1435] +
+
+Joseph Allen, Ahmed Moussa, and Xudong Liu. + Human-in-the-Loop Learning of Qualitative Preference Models. + In R. Barták and K. W. Brawner, editors, Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, pp.  108–111. AAAI Press, 2019.
+[ bib | +DOI ] + +
+ + +
+[1436] +
+
+Richard Allmendinger. + Tuning Evolutionary Search for Closed-Loop Optimization. + PhD thesis, The University of Manchester, UK, January 2012.
+[ bib ] + +
+ + +
+[1437] +
+
+A. Alsheddy and E. Tsang. + Guided Pareto local search and its application to the 0/1 multi-objective knapsack problems. + In M. Caserta and S. Voß, editors, Proceedings of MIC 2009, the 8th Metaheuristics International Conference, Hamburg, Germany, 2010. University of Hamburg.
+[ bib ] + +
+ + +
+[1438] +
+
+Sanae Amani, Mahnoosh Alizadeh, and Christos Thrampoulidis. + Linear Stochastic Bandits Under Safety Constraints. + In H. M. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. B. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems (NeurIPS 32), pp.  9256–9266, 2019.
+[ bib | +epub ] + +
+ + +
+[1439] +
+
+Klaus Andersen, René Victor Valqui Vidal, and Villy Bæk Iversen. + Design of a Teleprocessing Communication Network Using Simulated Annealing. + In R. V. V. Vidal, editor, Applied Simulated Annealing, pp.  201–215. Springer, 1993.
+[ bib ] + +
+ + +
+[1440] +
+
+J. H. Andersen and R. S. Powell. + The Use of Continuous Decision Variables in an Optimising Fixed Speed Pump Scheduling Algorithm. + In R. S. Powell and K. S. Hindi, editors, Computing and Control for the Water Industry, pp.  119–128. Research Studies Press Ltd., 1999.
+[ bib ] + +
+ + +
+[1441] +
+
+D. Anghinolfi, A. Boccalatte, M. Paolucci, and C. Vecchiola. + Performance Evaluation of an Adaptive Ant Colony Optimization Applied to Single Machine Scheduling. + In X. Li et al., editors, Simulated Evolution and Learning, 7th International Conference, SEAL 2008, volume 5361 of Lecture Notes in Computer Science, pp.  411–420. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[1442] +
+
+Daniel Angus. + Population-Based Ant Colony Optimisation for Multi-objective Function Optimisation. + In M. Randall, H. A. Abbass, and J. Wiles, editors, Progress in Artificial Life (ACAL), volume 4828 of Lecture Notes in Computer Science, pp.  232–244. Springer, Heidelberg, Germany, 2007.
+[ bib | +DOI ] + +
+ + +
+[1443] +
+
+J. Ansel, S. Kamil, K. Veeramachaneni, J. Ragan-Kelley, J. Bosboom, Una-May O'Reilly, and S. Amarasinghe. + OpenTuner: An extensible framework for program autotuning. + In Proceedings of the 23rd International Conference on Parallel Architectures and Compilation, pp.  303–315, New York, NY, 2014. ACM Press.
+[ bib | +DOI ] + +
+ + +
+[1444] +
+
+Carlos Ansótegui, Yuri Malitsky, Horst Samulowitz, Meinolf Sellmann, and Kevin Tierney. + Model-Based Genetic Algorithms for Algorithm Configuration. + In Q. Yang and M. Wooldridge, editors, Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI-15), pp.  733–739. IJCAI/AAAI Press, Menlo Park, CA, 2015.
+[ bib | +epub ] +
+Keywords: GGA++ +
+ +
+ + +
+[1445] +
+
+Carlos Ansótegui, Yuri Malitsky, and Meinolf Sellmann. + MaxSAT by Improved Instance-Specific Algorithm Configuration. + In D. Stracuzzi et al., editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  2594–2600. AAAI Press, 2014.
+[ bib ] + +
+ + +
+[1446] +
+
+Carlos Ansótegui, Meinolf Sellmann, and Kevin Tierney. + A Gender-Based Genetic Algorithm for the Automatic Configuration of Algorithms. + In I. P. Gent, editor, Principles and Practice of Constraint Programming, CP 2009, volume 5732 of Lecture Notes in Computer Science, pp.  142–157. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] +
+Keywords: GGA +
+ +
+ + +
+[1447] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + Finding Cuts in the TSP. + Technical Report 95–05, DIMACS Center, Rutgers University, Piscataway, NJ, USA, March 1995.
+[ bib ] + +
+ + +
+[1448] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + Finding Tours in the TSP. + Technical Report 99885, Forschungsinstitut für Diskrete Mathematik, University of Bonn, Germany, 1999.
+[ bib ] + +
+ + +
+[1449] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + The Traveling Salesman Problem: A Computational Study. + Princeton University Press, Princeton, NJ, 2006.
+[ bib ] + +
+ + +
+[1450] +
+
+Jay April, Fred Glover, James P. Kelly, and Manuel Laguna. + Simulation-based optimization: Practical introduction to simulation optimization. + In S. E. Chick, P. J. Sanchez, D. M. Ferrin, and D. J. Morrice, editors, Proceedings of the 35th Winter Simulation Conference: Driving Innovation, volume 1, pp.  71–78, New York, NY, December 2003. ACM Press.
+[ bib | +DOI ] + +
+ + +
+[1451] +
+
+Sanjeev Arora and Boaz Barak. + Computational complexity: a modern approach. + Cambridge University Press, 2009.
+[ bib ] + +
+ + +
+[1452] +
+
+Etor Arza, Josu Ceberio, Aritz Pérez, and Ekhine Irurozki. + Approaching the quadratic assignment problem with kernels of mallows models under the hamming distance. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+Keywords: QAP, EDA, Mallows +
+ +
+ + +
+[1453] +
+
+Y. Asahiro, K. Iwama, and E. Miyano. + Random Generation of Test Instances with Controlled Attributes. + In D. S. Johnson and M. A. Trick, editors, Cliques, Coloring, and Satisfiability: Second DIMACS Implementation Challenge, volume 26 of DIMACS Series on Discrete Mathematics and Theoretical Computer Science, pp.  377–393. American Mathematical Society, Providence, RI, 1996.
+[ bib ] + +
+ + +
+[1454] +
+
+N. Ascheuer. + Hamiltonian Path Problems in the On-line Optimization of Flexible Manufacturing Systems. + PhD thesis, Technische Universität Berlin, Berlin, Germany, 1995.
+[ bib ] + +
+ + +
+[1455] +
+
+R. Atkinson, Jakobus E. van Zyl, Godfrey A. Walters, and Dragan A. Savic. + Genetic algorithm optimisation of level-controlled pumping station operation. + In Water network modelling for optimal design and management, pp.  79–90. Centre for Water Systems, Exeter, UK, 2000.
+[ bib ] + +
+ + +
+[1456] +
+
+Charles Audet, Cong-Kien Dang, and Dominique Orban. + Algorithmic Parameter Optimization of the DFO Method with the OPAL Framework. + In K. Naono, K. Teranishi, J. Cavazos, and R. Suda, editors, Software Automatic Tuning: From Concepts to State-of-the-Art Results, pp.  255–274. Springer, 2010.
+[ bib ] + +
+ + +
+[1457] +
+
+Anne Auger, Johannes Bader, Dimo Brockhoff, and Eckart Zitzler. + Articulating User Preferences in Many-Objective Problems by Sampling the Weighted Hypervolume. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2009, pp.  555–562. ACM Press, New York, NY, 2009.
+[ bib ] + +
+ + +
+[1458] +
+
+Anne Auger, Johannes Bader, Dimo Brockhoff, and Eckart Zitzler. + Investigating and Exploiting the Bias of the Weighted Hypervolume to Articulate User Preferences. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2009, pp.  563–570. ACM Press, New York, NY, 2009.
+[ bib ] + +
+ + +
+[1459] +
+
+Anne Auger, Johannes Bader, Dimo Brockhoff, and Eckart Zitzler. + Theory of the hypervolume indicator: optimal μ-distributions and the choice of the reference point. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2009, pp.  87–102. ACM Press, New York, NY, 2009.
+[ bib ] + +
+ + +
+[1460] +
+
+Anne Auger, Dimo Brockhoff, Manuel López-Ibáñez, Kaisa Miettinen, Boris Naujoks, and Günther Rudolph. + Which questions should be asked to find the most appropriate method for decision making and problem solving? (Working Group “Algorithm Design Methods”). + In S. Greco, J. D. Knowles, K. Miettinen, and E. Zitzler, editors, Learning in Multiobjective Optimization (Dagstuhl Seminar 12041), volume 2(1) of Dagstuhl Reports, pp.  92–93. Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[1461] +
+
+A. Auger and B. Doerr, editors. + Theory of Randomized Search Heuristics: Foundations and Recent Developments, volume 1 of Series on Theoretical Computer Science. + World Scientific Publishing Co., Singapore, 2011.
+[ bib ] + +
+ + +
+[1462] +
+
+Anne Auger and Nikolaus Hansen. + A restart CMA evolution strategy with increasing population size. + In Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), pp.  1769–1776, Piscataway, NJ, September 2005. IEEE Press.
+[ bib | +DOI ] +
+Keywords: IPOP-CMA-ES +
+ +
+ + +
+[1463] +
+
+Anne Auger and Nikolaus Hansen. + Performance evaluation of an advanced local search evolutionary algorithm. + In Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), pp.  1777–1784, Piscataway, NJ, September 2005. IEEE Press.
+[ bib ] +
+Keywords: LR-CMAES +
+ +
+ + +
+[1464] +
+
+Andreea Avramescu, Richard Allmendinger, and Manuel López-Ibáñez. + A Multi-Objective Multi-Type Facility Location Problem for the Delivery of Personalised Medicine. + In P. Castillo and J. L. Jiménez Laredo, editors, Applications of Evolutionary Computation, volume 12694 of Lecture Notes in Computer Science, pp.  388–403. Springer, Cham, Switzerland, 2021.
+[ bib | +DOI | +supplementary material ] +
+Advances in personalised medicine targeting specific + sub-populations and individuals pose a challenge to the + traditional pharmaceutical industry. With a higher level of + personalisation, an already critical supply chain is facing + additional demands added by the very sensitive nature of its + products. Nevertheless, studies concerned with the efficient + development and delivery of these products are scarce. Thus, + this paper presents the case of personalised medicine and the + challenges imposed by its mass delivery. We propose a + multi-objective mathematical model for the + location-allocation problem with two interdependent facility + types in the case of personalised medicine products. We show + its practical application through a cell and gene therapy + case study. A multi-objective genetic algorithm with a novel + population initialisation procedure is used as solution + method. +
+
+Keywords: Personalised medicine, Biopharmaceuticals Supply chain, + Facility location-allocation, Evolutionary multi-objective + optimisation +
+ +
+ + +
+[1465] +
+
+Doǧan Aydın, Gürcan Yavuz, Serdar Özyön, Celal Yasar, and Thomas Stützle. + Artificial Bee Colony Framework to Non-convex Economic Dispatch Problem with Valve Point Effects: A Case Study. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2017, pp.  1311–1318. ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[1466] +
+
+Mayowa Ayodele, Richard Allmendinger, Manuel López-Ibáñez, Matthieu Parizy, and Arnaud Liefooghe. + Applying Ising Machines to Multi-Objective QUBOs. + In S. Silva and L. Paquete, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2023, pp.  2166–2174. ACM Press, New York, NY, 2023.
+[ bib | +DOI ] +
+Multi-objective optimisation problems involve finding + solutions with varying trade-offs between multiple and often + conflicting objectives. Ising machines are physical devices + that aim to find the absolute or approximate ground states of + an Ising model. To apply Ising machines to multi-objective + problems, a weighted sum objective function is used to + convert multi-objective into single-objective + problems. However, deriving scalarisation weights that + archives evenly distributed solutions across the Pareto front + is not trivial. Previous work has shown that adaptive weights + based on dichotomic search, and one based on averages of + previously explored weights can explore the Pareto front + quicker than uniformly generated weights. However, these + adaptive methods have only been applied to bi-objective + problems in the past. In this work, we extend the adaptive + method based on averages in two ways: (i) we extend the + adaptive method of deriving scalarisation weights for + problems with two or more objectives, and (ii) we use an + alternative measure of distance to improve performance. We + compare the proposed method with existing ones and show that + it leads to the best performance on multi-objective + Unconstrained Binary Quadratic Programming (mUBQP) instances + with 3 and 4 objectives and that it is competitive with the + best one for instances with 2 objectives. +
+
+ISBN: 979-8-4007-0120-7 +
+
+Keywords: digital annealer, multi-objective, bi-objective QAP, QUBO +
+ +
+ + +
+[1467] +
+
+Mayowa Ayodele, Richard Allmendinger, Manuel López-Ibáñez, and Matthieu Parizy. + Multi-Objective QUBO Solver: Bi-Objective Quadratic Assignment Problem. + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, pp.  467–475. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] +
+Quantum and quantum-inspired optimisation algorithms are + designed to solve problems represented in binary, quadratic + and unconstrained form. Combinatorial optimisation problems + are therefore often formulated as Quadratic Unconstrained + Binary Optimisation Problems (QUBO) to solve them with these + algorithms. Moreover, these QUBO solvers are often + implemented using specialised hardware to achieve enormous + speedups, e.g. Fujitsu's Digital Annealer (DA) and D-Wave's + Quantum Annealer. However, these are single-objective + solvers, while many real-world problems feature multiple + conflicting objectives. Thus, a common practice when using + these QUBO solvers is to scalarise such multi-objective + problems into a sequence of single-objective problems. Due to + design trade-offs of these solvers, formulating each + scalarisation may require more time than finding a local + optimum. We present the first attempt to extend the algorithm + supporting a commercial QUBO solver as a multi-objective + solver that is not based on scalarisation. The proposed + multi-objective DA algorithm is validated on the bi-objective + Quadratic Assignment Problem. We observe that algorithm + performance significantly depends on the archiving strategy + adopted, and that combining DA with non-scalarisation methods + to optimise multiple objectives outperforms the current + scalarised version of the DA in terms of final solution + quality. +
+
+Keywords: digital annealer, multi-objective, bi-objective QAP, QUBO +
+ +
+ + +
+[1468] +
+
+Mayowa Ayodele, Richard Allmendinger, Manuel López-Ibáñez, and Matthieu Parizy. + A Study of Scalarisation Techniques for Multi-objective QUBO Solving. + In O. Grothe, S. Nickel, S. Rebennack, and O. Stein, editors, Operations Research Proceedings 2022, OR 2022, Lecture Notes in Operations Research, pp.  393–399. Springer, Cham, Switzerland, 2022.
+[ bib | +DOI ] + +
+ + +
+[1469] +
+
+Mayowa Ayodele. + Penalty Weights in QUBO Formulations: Permutation Problems. + In L. Pérez Cáceres and S. Verel, editors, Proceedings of EvoCOP 2022 – 22nd European Conference on Evolutionary Computation in Combinatorial Optimization, Lecture Notes in Computer Science, pp.  159–174. Springer, Cham, Switzerland, 2022.
+[ bib ] + +
+ + +
+[1470] +
+
+Amine Aziz-Alaoui, Carola Doerr, and Johann Dréo. + Towards Large Scale Automated Algorithm Design by Integrating Modular Benchmarking Frameworks. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1365–1374. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[1471] +
+
+Ilya Loshchilov and T. Glasmachers. + Black Box Optimization Competition, 2017.
+[ bib | +http ] + +
+ + +
+[1472] +
+
+Anne Auger, Dimo Brockhoff, Nikolaus Hansen, Dejan Tusar, Tea Tušar, and Tobias Wagner. + GECCO Workshop on Real-Parameter Black-Box Optimization Benchmarking (BBOB 2016): Focus on multi-objective problems. + https://numbbo.github.io/workshops/BBOB-2016/, 2016.
+[ bib ] + +
+ + +
+[1473] +
+
+Eckart Zitzler, Marco Laumanns, and S. Bleuler. + A tutorial on evolutionary multiobjective optimization. + In X. Gandibleux, M. Sevaux, K. Sörensen, and V. T'Kindt, editors, Metaheuristics for Multiobjective Optimisation, volume 535 of Lecture Notes in Economics and Mathematical Systems, pp.  3–37. Springer, Berlin/Heidelberg, 2004.
+[ bib ] + +
+ + +
+[1474] +
+
+S. Bleuler, Marco Laumanns, Lothar Thiele, and Eckart Zitzler. + PISA – A Platform and Programming Language Independent Interface for Search Algorithms. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  494–508. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1475] +
+
+Domagoj Babić. + Spear theorem prover. + https://www.domagoj-babic.com/index.php/ResearchProjects/Spear, 2008.
+[ bib ] + +
+ + +
+[1476] +
+
+Domagoj Babić and Alan J. Hu. + Structural Abstraction of Software Verification Conditions. + In Computer Aided Verification: 19th International Conference, CAV 2007, pp.  366–378, 2007.
+[ bib ] +
+Spear-swv instances, + http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/benchmark_instances/SpearSWV/SWV-scrambled-first302.tar.gz, + http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/benchmark_instances/SpearSWV/SWV-scrambled-last302.tar.gz +
+ +
+ + +
+[1477] +
+
+Domagoj Babić and Frank Hutter. + Spear Theorem Prover. + In SAT'08: Proceedings of the SAT 2008 Race, 2008.
+[ bib | +epub | +supplementary material ] +
+Unreviewed paper +
+ +
+ + +
+[1478] +
+
+Thomas Bäck, David B. Fogel, and Zbigniew Michalewicz. + Handbook of evolutionary computation. + IOP Publishing, 1997.
+[ bib ] + +
+ + +
+[1479] +
+
+Achim Bachem, Barthel Steckemetz, and Michael Wottawa. + An efficient parallel cluster-heuristic for large Traveling Salesman Problems. + Technical Report 94-150, University of Koln, Germany, 1994.
+[ bib ] +
+Keywords: Genetic Edge Recombination (ERX) +
+ +
+ + +
+[1480] +
+
+Thomas Bäck. + Evolutionary algorithms in theory and practice: evolution strategies, evolutionary programming, genetic algorithms. + Oxford University Press, 1996.
+[ bib ] + +
+ + +
+[1481] +
+
+Prasanna Balaprakash, Mauro Birattari, Thomas Stützle, and Marco Dorigo. + Incremental local search in ant colony optimization: Why it fails for the quadratic assignment problem. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  156–166. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[1482] +
+
+Prasanna Balaprakash, Mauro Birattari, and Thomas Stützle. + Improvement Strategies for the F-Race Algorithm: Sampling Design and Iterative Refinement. + In T. Bartz-Beielstein, M. J. Blesa, C. Blum, B. Naujoks, A. Roli, G. Rudolph, and M. Sampels, editors, Hybrid Metaheuristics, volume 4771 of Lecture Notes in Computer Science, pp.  108–122. Springer, Heidelberg, Germany, 2007.
+[ bib | +DOI ] +
+Keywords: Iterated Race +
+ +
+ + +
+[1483] +
+
+Egon Balas and Andrew Ho. + Set Covering Algorithms Using Cutting Planes, Heuristics, and Subgradient Optimization: A Computational Study. + In M. W. Padberg, editor, Combinatorial optimization, volume 12 of Mathematical Programming Studies, pp.  37–60. Springer, Berlin/Heidelberg, 1980.
+[ bib | +DOI ] + +
+ + +
+[1484] +
+
+P. Baptiste and L. K. Hguny. + A branch and bound algorithm for the F/no_idle/Cmax. + In Proceedings of the international conference on industrial engineering and production management, IEPM'97, pp.  429–438, Lyon, 1997.
+[ bib ] + +
+ + +
+[1485] +
+
+Thomas Bartz-Beielstein. + Experimental Research in Evolutionary Computation: The New Experimentalism. + Springer, Berlin, Germany, 2006.
+[ bib ] +
+Keywords: SPO +
+ +
+ + +
+[1486] +
+
+Thomas Bartz-Beielstein. + How to Create Generalizable Results. + In J. Kacprzyk and W. Pedrycz, editors, Springer Handbook of Computational Intelligence, pp.  1127–1142. Springer, Berlin/Heidelberg, 2015.
+[ bib ] +
+Keywords: Mixed-effects models, random-effects model, problem instance + generation +
+ +
+ + +
+[1487] +
+
+Thomas Bartz-Beielstein, Oliver Flasch, Patrick Koch, and Wolfgang Konen. + SPOT: A Toolbox for Interactive and Automatic Tuning in the R Environment. + In Proceedings 20. Workshop Computational Intelligence, pp.  264–273, Karlsruhe, 2010. KIT Scientific Publishing.
+[ bib ] + +
+ + +
+[1488] +
+
+Thomas Bartz-Beielstein, C. Lasarczyk, and Mike Preuss. + Sequential Parameter Optimization. + In Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), pp.  773–780, Piscataway, NJ, September 2005. IEEE Press.
+[ bib ] + +
+ + +
+[1489] +
+
+Thomas Bartz-Beielstein, C. Lasarczyk, and Mike Preuss. + The Sequential Parameter Optimization Toolbox. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  337–360. Springer, Berlin, Germany, 2010.
+[ bib | +DOI ] +
+Keywords: SPOT +
+ +
+ + +
+[1490] +
+
+Thomas Bartz-Beielstein and Sandor Markon. + Tuning search algorithms for real-world applications: A regression tree based approach. + In Proceedings of the 2004 Congress on Evolutionary Computation (CEC 2004), pp.  1111–1118, Piscataway, NJ, September 2004. IEEE Press.
+[ bib ] + +
+ + +
+[1491] +
+
+Elias Bareinboim and Judea Pearl. + Transportability of causal effects: Completeness results. + In J. Hoffmann and B. Selman, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  698,704. AAAI Press, 2012.
+[ bib ] + +
+ + +
+[1492] +
+
+Thomas Bartz-Beielstein and Mike Preuss. + Considerations of budget allocation for sequential parameter optimization (SPO). + In L. Paquete, M. Chiarandini, and D. Basso, editors, Empirical Methods for the Analysis of Algorithms, Workshop EMAA 2006, Proceedings, pp.  35–40, Reykjavik, Iceland, 2006.
+[ bib ] + +
+ + +
+[1493] +
+
+Thomas Bartz-Beielstein and Mike Preuss. + Experimental Analysis of Optimization Algorithms: Tuning and Beyond. + In Y. Borenstein and A. Moraglio, editors, Theory and Principled Methods for the Design of Metaheuristics, Natural Computing Series, pp.  205–245. Springer, Berlin/Heidelberg, 2014.
+[ bib | +DOI ] + +
+ + +
+[1494] +
+
+Benjamín Barán and Matilde Schaerer. + A multiobjective ant colony system for vehicle routing problem with time windows. + In Proceedings of the Twenty-first IASTED International Conference on Applied Informatics, pp.  97–102, Insbruck, Austria, 2003.
+[ bib ] + +
+ + +
+[1495] +
+
+Matthieu Basseur, Adrien Goëffon, Arnaud Liefooghe, and Sébastien Verel. + On Set-based Local Search for Multiobjective Combinatorial Optimization. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2013, pp.  471–478. ACM Press, New York, NY, 2013.
+[ bib | +DOI ] + +
+ + +
+[1496] +
+
+Vitor Basto-Fernandes, Iryna Yevseyeva, André Deutz, and Michael T. M. Emmerich. + A survey of diversity oriented optimization: Problems, indicators, and algorithms. + In EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation VII, volume 662 of Studies in Computational Intelligence, pp.  3–23. Springer, Cham, Switzerland, 2017.
+[ bib | +DOI ] + +
+ + +
+[1497] +
+
+Roberto Battiti, M. Brunato, and Franco Mascia. + Reactive Search and Intelligent Optimization, volume 45 of Operations Research/Computer Science Interfaces. + Springer, New York, NY, 2008.
+[ bib | +DOI ] + +
+ + +
+[1498] +
+
+Roberto Battiti and Paolo Campigotto. + Reactive search optimization: Learning while optimizing. An experiment in interactive multi-objective optimization. + In M. Caserta and S. Voß, editors, Proceedings of MIC 2009, the 8th Metaheuristics International Conference, Hamburg, Germany, 2010. University of Hamburg.
+[ bib ] + +
+ + +
+[1499] +
+
+Michele Battistutta, Andrea Schaerf, and Tommaso Urli. + Feature-based tuning of single-stage simulated annealing for examination timetabling. + In E. Özcan, E. K. Burke, and B. McCollum, editors, PATAT 2014: Proceedings of the 10th International Conference of the Practice and Theory of Automated Timetabling, pp.  53–61. PATAT, 2014.
+[ bib ] +
+Keywords: F-race +
+ +
+ + +
+[1500] +
+
+E. B. Baum. + Iterated Descent: A Better Algorithm for Local Search in Combinatorial Optimization Problems. + Manuscript, 1986.
+[ bib ] + +
+ + +
+[1501] +
+
+E. B. Baum. + Towards Practical “Neural” Computation for Combinatorial Optimization Problems. + In Neural Networks for Computing, AIP Conference Proceedings, pp.  53–64, 1986.
+[ bib ] + +
+ + +
+[1502] +
+
+A. Baykasoglu, T. Dereli, and I. Sabuncu. + A multiple objective ant colony optimization approach to assembly line balancing problems. + In 35th International Conference on Computers and Industrial Engineering (CIE35), pp.  263–268, Istanbul, Turkey, 2005.
+[ bib ] + +
+ + +
+[1503] +
+
+Brian Beachkofski and Ramana Grandhi. + Improved Distributed Hypercube Sampling. + In Proceedings of the 43rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference. AIAA paper 2002-1274, American Institute of Aeronautics and Astronautics, 2002.
+[ bib ] + +
+ + +
+[1504] +
+
+John E. Beasley. + Heuristic algorithms for the unconstrained binary quadratic programming problem. + Technical report, The Management School, Imperial College, London, England, 1998.
+[ bib | +epub ] + +
+ + +
+[1505] +
+
+S. Becker, J. Gottlieb, and Thomas Stützle. + Applications of Racing Algorithms: An Industrial Perspective. + In E.-G. Talbi, P. Liardet, P. Collet, E. Lutton, and M. Schoenauer, editors, Artificial Evolution, volume 3871 of Lecture Notes in Computer Science, pp.  271–283. Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[1506] +
+
+David D. Bedworth and James E. Bailey. + Integrated Production Control Systems: Management, Analysis, Design, volume 2. + John Wiley & Sons, New York, NY, 1982.
+[ bib ] + +
+ + +
+[1507] +
+
+Andreas Beham, Michael Affenzeller, and Stefan Wagner. + Instance-based algorithm selection on quadratic assignment problem landscapes. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2017, pp.  1471–1478. ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[1508] +
+
+Valerie Belton, Jürgen Branke, Petri Eskelinen, Salvatore Greco, Julián Molina, Francisco Ruiz, and Roman Slowiński. + Interactive Multiobjective Optimization from a Learning Perspective. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science, pp.  405–433. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[1509] +
+
+Nacim Belkhir, Johann Dréo, Pierre Savéant, and Marc Schoenauer. + Feature Based Algorithm Configuration: A Case Study with Differential Evolution. + In J. Handl, E. Hart, P. R. Lewis, M. López-Ibáñez, G. Ochoa, and B. Paechter, editors, Parallel Problem Solving from Nature – PPSN XIV, volume 9921 of Lecture Notes in Computer Science, pp.  156–166. Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[1510] +
+
+Nacim Belkhir, Johann Dréo, Pierre Savéant, and Marc Schoenauer. + Per Instance Algorithm Configuration of CMA-ES with Limited Budget. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  681–688. ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[1511] +
+
+Jon Louis Bentley. + Experiments on Traveling Salesman Heuristics. + In D. S. Johnson, editor, Proceedings of the First Annual ACM-SIAM Symposium on Discrete Algorithms, pp.  91–99. Society for Industrial and Applied Mathematics, Philadelphia, PA, USA, 1990.
+[ bib ] + +
+ + +
+[1512] +
+
+Nawal Benabbou, Cassandre Leroy, and Thibaut Lust. + An Interactive Regret-Based Genetic Algorithm for Solving Multi-Objective Combinatorial Optimization Problems. + In Proceedings of the AAAI Conference on Artificial Intelligence, pp.  2335–2342. AAAI Press, 2020.
+[ bib | +DOI ] +
+Keywords: interactive, multi-objective, decision-makers +
+ +
+ + +
+[1513] +
+
+Alessio Benavoli, Giorgio Corani, Francesca Mangili, and Marco Zaffalon. + A Bayesian nonparametric procedure for comparing algorithms. + In F. Bach and D. Blei, editors, Proceedings of the 32nd International Conference on Machine Learning, ICML 2015, volume 37, pp.  1264–1272. PMLR, 2015.
+[ bib | +epub ] +
+Keywords: racing +
+ +
+ + +
+[1514] +
+
+Alexander Javier Benavides and Marcus Ritt. + Iterated Local Search Heuristics for Minimizing Total Completion Time in Permutation and Non-permutation Flow Shops. + In R. I. Brafman, C. Domshlak, P. Haslum, and S. Zilberstein, editors, Proceedings of the Twenty-Fifth International Conference on Automated Planning and Scheduling, ICAPS 2015, pp.  34–41. AAAI Press, Menlo Park, CA, 2015.
+[ bib ] +
+Keywords: irace +
+ +
+ + +
+[1515] +
+
+Stefano Benedettini, Andrea Roli, and Christian Blum. + A Randomized Iterated Greedy Algorithm for the Founder Sequence Reconstruction Problem. + In C. Blum and R. Battiti, editors, Learning and Intelligent Optimization, 4th International Conference, LION 4, volume 6073 of Lecture Notes in Computer Science, pp.  37–51. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[1516] +
+
+Stefano Benedettini, Andrea Roli, and Luca Di Gaspero. + Two-level ACO for Haplotype Inference under Pure Parsimony. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 6th International Conference, ANTS 2008, volume 5217 of Lecture Notes in Computer Science, pp.  179–190. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[1517] +
+
+D. Bertsekas. + Dynamic Programming and Optimal Control. + Athena Scientific, Belmont, MA, 1995.
+[ bib ] + +
+ + +
+[1518] +
+
+D. Bertsekas. + Network Optimization: Continuous and Discrete Models. + Athena Scientific, Belmont, MA, 1998.
+[ bib ] + +
+ + +
+[1519] +
+
+James S. Bergstra, Rémi Bardenet, Yoshua Bengio, and Balázs Kégl. + Algorithms for Hyper-Parameter Optimization. + In J. Shawe-Taylor, R. S. Zemel, P. L. Bartlett, F. Pereira, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems (NIPS 24), pp.  2546–2554. Curran Associates, Red Hook, NY, 2011.
+[ bib | +http ] + +
+ + +
+[1520] +
+
+Hughes Bersini, Marco Dorigo, S. Langerman, G. Seront, and L. M. Gambardella. + Results of the First International Contest on Evolutionary Optimisation. + In T. Bäck, T. Fukuda, and Z. Michalewicz, editors, Proceedings of the 1996 IEEE International Conference on Evolutionary Computation (ICEC'96), pp.  611–615. IEEE Press, Piscataway, NJ, 1996.
+[ bib ] + +
+ + +
+[1521] +
+
+Felix Berkenkamp, Angela P. Schoellig, and Andreas Krause. + Safe controller optimization for quadrotors with Gaussian processes. + In 2016 IEEE International Conference on Robotics and Automation (ICRA), pp.  491–496. IEEE, 2016.
+[ bib | +DOI ] +
+Keywords: Safe Optimization, SafeOpt +
+ +
+ + +
+[1522] +
+
+James S. Bergstra, Daniel Yasmin, and David Cox. + Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures. + In S. Dasgupta and D. McAllester, editors, Proceedings of the 30th International Conference on Machine Learning, ICML 2013, volume 28, pp.  115–123, 2013.
+[ bib | +http ] + +
+ + +
+[1523] +
+
+Matthijs L. den Besten, Thomas Stützle, and Marco Dorigo. + Ant Colony Optimization for the Total Weighted Tardiness Problem. + In M. Schoenauer et al., editors, Parallel Problem Solving from Nature – PPSN VI, volume 1917 of Lecture Notes in Computer Science, pp.  611–620. Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[1524] +
+
+Matthijs L. den Besten, Thomas Stützle, and Marco Dorigo. + Design of Iterated Local Search Algorithms: An Example Application to the Single Machine Total Weighted Tardiness Problem. + In E. J. W. Boers et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2001, volume 2037 of Lecture Notes in Computer Science, pp.  441–452. Springer, Heidelberg, Germany, 2001.
+[ bib ] + +
+ + +
+[1525] +
+
+Nicola Beume and Günther Rudolph. + Faster S-Metric Calculation by Considering Dominated Hypervolume as Klee's Measure Problem. + In B. Kovalerchuk, editor, Proceedings of the Second IASTED Conference on Computational Intelligence, pp.  231–236. ACTA Press, Anaheim, 2006.
+[ bib ] + +
+ + +
+[1526] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Generation of Multi-Objective ACO Algorithms for the Biobjective Knapsack. + In M. Dorigo et al., editors, Swarm Intelligence, 8th International Conference, ANTS 2012, volume 7461 of Lecture Notes in Computer Science, pp.  37–48. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[1527] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Generation of MOACO Algorithms for the Biobjective Bidimensional Knapsack Problem: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2012-008/, 2012.
+[ bib ] + +
+ + +
+[1528] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + An Analysis of Local Search for the Bi-objective Bidimensional Knapsack: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2012-016/, 2013.
+[ bib ] + +
+ + +
+[1529] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Deconstructing Multi-Objective Evolutionary Algorithms: An Iterative Analysis on the Permutation Flowshop: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2013-010/, 2013.
+[ bib ] + +
+ + +
+[1530] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + An Analysis of Local Search for the Bi-objective Bidimensional Knapsack Problem. + In M. Middendorf and C. Blum, editors, Proceedings of EvoCOP 2013 – 13th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 7832 of Lecture Notes in Computer Science, pp.  85–96. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[1531] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Component-Wise Design of Multi-Objective Evolutionary Algorithms. + Technical Report TR/IRIDIA/2014-012, IRIDIA, Université Libre de Bruxelles, Belgium, August 2014.
+[ bib ] + +
+ + +
+[1532] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Deconstructing Multi-Objective Evolutionary Algorithms: An Iterative Analysis on the Permutation Flowshop. + In P. M. Pardalos, M. G. C. Resende, C. Vogiatzis, and J. L. Walteros, editors, Learning and Intelligent Optimization, 8th International Conference, LION 8, volume 8426 of Lecture Notes in Computer Science, pp.  57–172. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[1533] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Design of Evolutionary Algorithms for Multi-Objective Combinatorial Optimization. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  508–517. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[1534] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Design of Evolutionary Algorithms for Multi-Objective Combinatorial Optimization. + http://iridia.ulb.ac.be/supp/IridiaSupp2014-007/, 2014.
+[ bib ] + +
+ + +
+[1535] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Component-Wise Design of Multi-Objective Evolutionary Algorithms. + https://github.com/iridia-ulb/automoea-tevc-2016, 2015.
+[ bib ] + +
+ + +
+[1536] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + To DE or Not to DE? Multi-objective Differential Evolution Revisited from a Component-Wise Perspective: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2015-001/, 2015.
+[ bib ] + +
+ + +
+[1537] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + To DE or Not to DE? Multi-objective Differential Evolution Revisited from a Component-Wise Perspective. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  48–63. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] + +
+ + +
+[1538] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Comparing Decomposition-Based and Automatically Component-Wise Designed Multi-Objective Evolutionary Algorithms. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  396–410. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] + +
+ + +
+[1539] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + A Large-Scale Experimental Evaluation of High-Performing Multi- and Many-Objective Evolutionary Algorithms. + http://iridia.ulb.ac.be/supp/IridiaSupp2015-007/, 2017.
+[ bib ] + +
+ + +
+[1540] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + A Large-Scale Experimental Evaluation of High-Performing Multi- and Many-Objective Evolutionary Algorithms. + Technical Report TR/IRIDIA/2017-005, IRIDIA, Université Libre de Bruxelles, Belgium, February 2017.
+[ bib ] + +
+ + +
+[1541] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + An Empirical Assessment of the Properties of Inverted Generational Distance Indicators on Multi- and Many-objective Optimization. + In H. Trautmann, G. Rudolph, K. Klamroth, O. Schütze, M. M. Wiecek, Y. Jin, and C. Grimme, editors, Evolutionary Multi-criterion Optimization, EMO 2017, volume 10173 of Lecture Notes in Computer Science, pp.  31–45. Springer International Publishing, Cham, Switzerland, 2017.
+[ bib | +DOI ] + +
+ + +
+[1542] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatically Designing State-of-the-Art Multi- and Many-Objective Evolutionary Algorithms: Supplementary material. + https://github.com/iridia-ulb/automoea-ecj-2020, 2019.
+[ bib ] + +
+ + +
+[1543] +
+
+Hao Wang, Chaoli Sun, Yaochu Jin, Shufen Qin, and Haibo Yu. + A Multi-indicator based Selection Strategy for Evolutionary Many-objective Optimization. + In Proceedings of the 2019 Congress on Evolutionary Computation (CEC 2019), pp.  2042–2049, Piscataway, NJ, 2019. IEEE Press.
+[ bib ] +
+unbounded archive +
+ +
+ + +
+[1544] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Archiver Effects on the Performance of State-of-the-art Multi- and Many-objective Evolutionary Algorithms. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub | +supplementary material ] + +
+ + +
+[1545] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Archiver Effects on the Performance of State-of-the-art Multi- and Many-objective Evolutionary Algorithms: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2019-004/, 2019.
+[ bib ] + +
+ + +
+[1546] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Configuration of Multi-objective Optimizers and Multi-objective Configuration. + In T. Bartz-Beielstein, B. Filipič, P. Korošec, and E.-G. Talbi, editors, High-Performance Simulation-Based Optimization, pp.  69–92. Springer International Publishing, Cham, Switzerland, 2020.
+[ bib | +DOI ] +
+Heuristic optimizers are an important tool in academia and industry, and their performance-optimizing configuration requires a significant amount of expertise. As the proper configuration of algorithms is a crucial aspect in the engineering of heuristic algorithms, a significant research effort has been dedicated over the last years towards moving this step to the computer and, thus, make it automatic. These research efforts go way beyond tuning only numerical parameters of already fully defined algorithms, but exploit automatic configuration as a means for automatic algorithm design. In this chapter, we review two main aspects where the research on automatic configuration and multi-objective optimization intersect. The first is the automatic configuration of multi-objective optimizers, where we discuss means and specific approaches. In addition, we detail a case study that shows how these approaches can be used to design new, high-performing multi-objective evolutionary algorithms. The second aspect is the research on multi-objective configuration, that is, the possibility of using multiple performance metrics for the evaluation of algorithm configurations. We highlight some few examples in this direction. +
+ +
+ + +
+[1547] +
+
+Leonardo C. T. Bezerra. + A component-wise approach to multi-objective evolutionary algorithms: from flexible frameworks to automatic design. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2016.
+[ bib ] +
+Supervised by Thomas Stützle and Manuel López-Ibáñez +
+ +
+ + +
+[1548] +
+
+Leonora Bianchi, L. M. Gambardella, and Marco Dorigo. + An Ant Colony Optimization Approach to the Probabilistic Traveling Salesman Problem. + In J.-J. Merelo et al., editors, Parallel Problem Solving from Nature – PPSN VII, volume 2439 of Lecture Notes in Computer Science, pp.  883–892. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1549] +
+
+Armin Biere. + Yet another Local Search Solver and Lingeling and Friends Entering the SAT Competition 2014. + In A. Belov, D. Diepold, M. Heule, and M. Järvisalo, editors, Proceedings of SAT Competition 2014: Solver and Benchmark Descriptions, volume B-2014-2 of Science Series of Publications B, pp.  39–40. University of Helsinki, 2014.
+[ bib ] + +
+ + +
+[1550] +
+
+André Biedenkapp, H. Furkan Bozkurt, Theresa Eimer, Frank Hutter, and Marius Thomas Lindauer. + Dynamic Algorithm Configuration: Foundation of a New Meta-Algorithmic Framework. + In G. D. Giacomo, A. Catala, B. Dilkina, M. Milano, S. Barro, A. Bugarín, and J. Lang, editors, Proceedings of the 24th European Conference on Artificial Intelligence (ECAI), volume 325 of Frontiers in Artificial Intelligence and Applications, pp.  427–434. IOS Press, 2020.
+[ bib | +epub ] + +
+ + +
+[1551] +
+
+André Biedenkapp, Marius Thomas Lindauer, Katharina Eggensperger, Frank Hutter, Chris Fawcett, and Holger H. Hoos. + Efficient Parameter Importance Analysis via Ablation with Surrogates. + In S. P. Singh and S. Markovitch, editors, Proceedings of the AAAI Conference on Artificial Intelligence. AAAI Press, February 2017.
+[ bib | +DOI ] + +
+ + +
+[1552] +
+
+André Biedenkapp, Joshua Marben, Marius Thomas Lindauer, and Frank Hutter. + CAVE: Configuration assessment, visualization and evaluation. + In R. Battiti, M. Brunato, I. Kotsireas, and P. M. Pardalos, editors, Learning and Intelligent Optimization, 12th International Conference, LION 12, volume 11353 of Lecture Notes in Computer Science, pp.  115–130. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] + +
+ + +
+[1553] +
+
+George Bilchev and Ian C. Parmee. + The Ant Colony Metaphor for Searching Continuous Design Spaces. + In T. C. Fogarty, editor, Evolutionary Computing, AISB Workshop, volume 993 of Lecture Notes in Computer Science, pp.  25–39. Springer, Berlin, Germany, 1995.
+[ bib | +DOI ] + +
+ + +
+[1554] +
+
+Mauro Birattari, Prasanna Balaprakash, and Marco Dorigo. + The ACO/F-RACE algorithm for combinatorial optimization under uncertainty. + In K. F. Doerner, M. Gendreau, P. Greistorfer, W. J. Gutjahr, R. F. Hartl, and M. Reimann, editors, Metaheuristics – Progress in Complex Systems Optimization, volume 39 of Operations Research/Computer Science Interfaces Series, pp.  189–203. Springer, New York, NY, 2006.
+[ bib ] + +
+ + +
+[1555] +
+
+Mauro Birattari, Marco Chiarandini, Marco Saerens, and Thomas Stützle. + Learning Graphical Models for Algorithm Configuration. + In T. Berthold, A. M. Gleixner, S. Heinz, and T. Koch, editors, Integration of AI and OR Techniques in Contraint Programming for Combinatorial Optimization Problems, Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[1556] +
+
+Mauro Birattari, Gianni A. Di Caro, and Marco Dorigo. + Toward the formal foundation of Ant Programming. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  188–201. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1557] +
+
+Steven Bird, Ewan Klein, and Edward Loper. + Natural language processing with Python: analyzing text with the natural language toolkit. + O'Reilly Media, Inc., 2009.
+[ bib ] + +
+ + +
+[1558] +
+
+Mauro Birattari, Thomas Stützle, Luís Paquete, and Klaus Varrentrapp. + A Racing Algorithm for Configuring Metaheuristics. + In W. B. Langdon et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2002, pp.  11–18. Morgan Kaufmann Publishers, San Francisco, CA, 2002.
+[ bib | +epub ] +
+Keywords: F-race +
+ +
+ + +
+[1559] +
+
+Mauro Birattari, Zhi Yuan, Prasanna Balaprakash, and Thomas Stützle. + F-Race and Iterated F-Race: An Overview. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  311–336. Springer, Berlin, Germany, 2010.
+[ bib | +DOI ] +
+Keywords: F-race, iterated F-race, irace, tuning +
+ +
+ + +
+[1560] +
+
+Mauro Birattari, Zhi Yuan, Prasanna Balaprakash, and Thomas Stützle. + Parameter Adaptation in Ant Colony Optimization. + In M. Caserta and S. Voß, editors, Proceedings of MIC 2009, the 8th Metaheuristics International Conference, Hamburg, Germany, 2010. University of Hamburg.
+[ bib ] + +
+ + +
+[1561] +
+
+Mauro Birattari. + Tuning Metaheuristics: A Machine Learning Perspective, volume 197 of Studies in Computational Intelligence. + Springer, Berlin/Heidelberg, 2009.
+[ bib | +DOI ] +
+Based on the PhD thesis [1562] +
+ +
+ + +
+[1562] +
+
+Mauro Birattari. + The Problem of Tuning Metaheuristics as Seen from a Machine Learning Perspective. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2004.
+[ bib ] +
+Supervised by Marco Dorigo +
+ +
+ + +
+[1563] +
+
+Francesco Biscani, Dario Izzo, and Chit Hong Yam. + A Global Optimisation Toolbox for Massively Parallel Engineering Optimisation. + In Astrodynamics Tools and Techniques (ICATT 2010), 4th International Conference on, 2010.
+[ bib | +http ] +
+Keywords: PaGMO +
+ +
+ + +
+[1564] +
+
+Bernd Bischl, Olaf Mersmann, Heike Trautmann, and Mike Preuss. + Algorithm Selection Based on Exploratory Landscape Analysis and Cost-sensitive Learning. + In T. Soule and J. H. Moore, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2012, pp.  313–320. ACM Press, New York, NY, 2012.
+[ bib ] +
+Keywords: continuous optimization, landscape analysis, algorithm selection +
+ +
+ + +
+[1565] +
+
+Christopher M. Bishop. + Pattern recognition and machine learning. + Springer, 2006.
+[ bib ] + +
+ + +
+[1566] +
+
+Erdem Bıyık, Jonathan Margoliash, Shahrouz Ryan Alimo, and Dorsa Sadigh. + Efficient and Safe Exploration in Deterministic Markov Decision Processes with Unknown Transition Models. + In 2019 American Control Conference (ACC), pp.  1792–1799. IEEE, 2019.
+[ bib | +DOI ] + +
+ + +
+[1567] +
+
+María J. Blesa and Christian Blum. + Ant Colony Optimization for the Maximum Edge-Disjoint Paths Problem. + In G. R. Raidl et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004, volume 3005 of Lecture Notes in Computer Science, pp.  160–169. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[1568] +
+
+John Blitzer, Ryan McDonald, and Fernando Pereira. + Domain adaptation with structural correspondence learning. + In D. Jurafsky and E. Gaussier, editors, Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP2006, Empirical Methods in Natural Language Processing, pp.  120–128, 2006.
+[ bib ] + +
+ + +
+[1569] +
+
+Aymeric Blot, Holger H. Hoos, Laetitia Jourdan, Marie-Eléonore Kessaci-Marmion, and Heike Trautmann. + MO-ParamILS: A Multi-objective Automatic Algorithm Configuration Framework. + In P. Festa, M. Sellmann, and J. Vanschoren, editors, Learning and Intelligent Optimization, 10th International Conference, LION 10, volume 10079 of Lecture Notes in Computer Science, pp.  32–47. Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[1570] +
+
+Aymeric Blot, Laetitia Jourdan, and Marie-Eléonore Kessaci-Marmion. + Automatic design of multi-objective local search algorithms: case study on a bi-objective permutation flowshop scheduling problem. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  227–234. ACM Press, New York, NY, 2017.
+[ bib | +DOI ] + +
+ + +
+[1571] +
+
+Aymeric Blot, Manuel López-Ibáñez, Marie-Eléonore Kessaci-Marmion, and Laetitia Jourdan. + New Initialisation Techniques for Multi-Objective Local Search: Application to the Bi-objective Permutation Flowshop. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11101 of Lecture Notes in Computer Science, pp.  323–334. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] + +
+ + +
+[1572] +
+
+Aymeric Blot, Alexis Pernet, Laetitia Jourdan, Marie-Eléonore Kessaci-Marmion, and Holger H. Hoos. + Automatically Configuring Multi-objective Local Search Using Multi-objective Optimisation. + In H. Trautmann, G. Rudolph, K. Klamroth, O. Schütze, M. M. Wiecek, Y. Jin, and C. Grimme, editors, Evolutionary Multi-criterion Optimization, EMO 2017, volume 10173 of Lecture Notes in Computer Science, pp.  61–76. Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[1573] +
+
+Christian Blum, J. Bautista, and J. Pereira. + Beam-ACO applied to assembly line balancing. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  96–107. Springer, Heidelberg, Germany, 2006.
+[ bib | +DOI ] + +
+ + +
+[1574] +
+
+Christian Blum, María J. Blesa, and Manuel López-Ibáñez. + Beam Search for the Longest Common Subsequence Problem. + Technical Report LSI-08-29, Department LSI, Universitat Politècnica de Catalunya, 2008. + Published in Computers & Operations Research [146].
+[ bib ] + +
+ + +
+[1575] +
+
+Christian Blum, Carlos Cotta, Antonio J. Fernández, and J. E. Gallardo. + A probabilistic beam search algorithm for the shortest common supersequence problem. + In C. Cotta et al., editors, Proceedings of EvoCOP 2007 – Seventh European Conference on Evolutionary Computation in Combinatorial Optimisation, volume 4446 of Lecture Notes in Computer Science, pp.  36–47. Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[1576] +
+
+Christian Blum and Manuel López-Ibáñez. + Ant Colony Optimization. + In The Industrial Electronics Handbook: Intelligent Systems. CRC Press, 2nd edition, 2011.
+[ bib | +http ] + +
+ + +
+[1577] +
+
+Christian Blum and M. Mastrolilli. + Using Branch & Bound Concepts in Construction-Based Metaheuristics: Exploiting the Dual Problem Knowledge. + In T. Bartz-Beielstein, M. J. Blesa, C. Blum, B. Naujoks, A. Roli, G. Rudolph, and M. Sampels, editors, Hybrid Metaheuristics, volume 4771 of Lecture Notes in Computer Science, pp.  123–139. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[1578] +
+
+C. Blum and D. Merkle, editors. + Swarm Intelligence–Introduction and Applications. + Natural Computing Series. Springer Verlag, Berlin, Germany, 2008.
+[ bib ] + +
+ + +
+[1579] +
+
+Christian Blum and Günther R. Raidl. + Hybrid Metaheuristics—Powerful Tools for Optimization. + Artificial Intelligence: Foundations, Theory, and Algorithms. Springer, Berlin, Germany, 2016.
+[ bib ] + +
+ + +
+[1580] +
+
+Christian Blum and Andrea Roli. + Hybrid metaheuristics: an introduction. + In C. Blum, M. J. Blesa, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics: An emergent approach for optimization, volume 114 of Studies in Computational Intelligence, pp.  1–30. Springer, Berlin, Germany, 2008.
+[ bib ] + +
+ + +
+[1581] +
+
+Christian Blum and M. Yábar Vallès. + Multi-level ant colony optimization for DNA sequencing by hybridization. + In F. Almeida et al., editors, Hybrid Metaheuristics, volume 4030 of Lecture Notes in Computer Science, pp.  94–109. Springer, Heidelberg, Germany, 2006.
+[ bib | +DOI ] + +
+ + +
+[1582] +
+
+K. D. Boese. + Models for Iterative Global Optimization. + PhD thesis, University of California, Computer Science Department, Los Angeles, CA, 1996.
+[ bib ] + +
+ + +
+[1583] +
+
+Béla Bollobás. + Random Graphs. + Cambridge University Press, New York, NY, 2nd edition, 2001.
+[ bib ] + +
+ + +
+[1584] +
+
+Grady Booch, James E. Rumbaugh, and Ivar Jacobson. + The Unified Modeling Language User Guide. + Addison-Wesley, 2nd edition, 2005.
+[ bib ] + +
+ + +
+[1585] +
+
+P. C. Borges and Michael Pilegaard Hansen. + A basis for future successes in multiobjective combinatorial optimization. + Technical Report IMM-REP-1998-8, Institute of Mathematical Modelling, Technical University of Denmark, Lyngby, Denmark, 1998.
+[ bib ] + +
+ + +
+[1586] +
+
+Allan Borodin and Ran El-Yaniv. + Online computation and competitive analysis. + Cambridge University Press, New York, NY, 1998.
+[ bib ] + +
+ + +
+[1587] +
+
+Michael Borenstein, Larry V. Hedges, Julian P. T. Higgins, and Hannah R. Rothstein. + Introduction to Meta-Analysis. + Wiley, 2009.
+[ bib ] + +
+ + +
+[1588] +
+
+Bernhard E. Boser, Isabelle Guyon, and Vladimir Vapnik. + A Training Algorithm for Optimal Margin Classifiers. + In D. Haussler, editor, COLT'92, pp.  144–152. ACM Press, 1992.
+[ bib | +DOI ] +
+Proposed SVM +
+ +
+ + +
+[1589] +
+
+Jakob Bossek, Pascal Kerschke, Aneta Neumann, Markus Wagner, Frank Neumann, and Heike Trautmann. + Evolving Diverse TSP Instances by Means of Novel and Creative Mutation Operators. + In T. Friedrich, C. Doerr, and D. V. Arnold, editors, Proceedings of the 15th ACM/SIGEVO Conference on Foundations of Genetic Algorithms, pp.  58–71. ACM, 2019.
+[ bib ] + +
+ + +
+[1590] +
+
+Paul F. Boulos, Chun Hou Orr, Werner de Schaetzen, J. G. Chatila, Michael Moore, Paul Hsiung, and Devan Thomas. + Optimal pump operation of water distribution systems using genetic algorithms. + In AWWA Distribution System Symp., Denver, USA, 2001. American Water Works Association.
+[ bib ] + +
+ + +
+[1591] +
+
+V. Bowman and Jr. Joseph. + On the Relationship of the Tchebycheff Norm and the Efficient Frontier of Multiple-Criteria Objectives. + In H. Thiriez and S. Zionts, editors, Multiple Criteria Decision Making, volume 130 of Lecture Notes in Economics and Mathematical Systems, pp.  76–86. Springer, Berlin/Heidelberg, 1976.
+[ bib | +DOI ] + +
+ + +
+[1592] +
+
+George E. P. Box and Norman R. Draper. + Response surfaces, mixtures, and ridge analyses. + John Wiley & Sons, 2007.
+[ bib ] + +
+ + +
+[1593] +
+
+G. E. P. Box, W. G. Hunter, and J. S. Hunter. + Statistics for experimenters: an introduction to design, data analysis, and model building. + John Wiley & Sons, New York, NY, 1978.
+[ bib ] + +
+ + +
+[1594] +
+
+A. Brandt. + Multilevel Computations: Review and Recent Developments. + In S. F. McCormick, editor, Multigrid Methods: Theory, Applications, and Supercomputing, Proceedings of the 3rd Copper Mountain Conference on Multigrid Methods, volume 110 of Lecture Notes in Pure and Applied Mathematics, pp.  35–62. Marcel Dekker, New York, NY, 1988.
+[ bib ] + +
+ + +
+[1595] +
+
+L. Bradstreet, L. Barone, L. While, S. Huband, and P. Hingston. + Use of the WFG Toolkit and PISA for Comparison of MOEAs. + In IEEE Symposium on Computational Intelligence in Multicriteria Decision-Making, IEEE MCDM, pp.  382–389, 2007.
+[ bib ] + +
+ + +
+[1596] +
+
+Cristóbal Barba-González, Vesa Ojalehto, José García-Nieto, Antonio J. Nebro, Kaisa Miettinen, and José F. Aldana-Montes. + Artificial Decision Maker Driven by PSO: An Approach for Testing Reference Point Based Interactive Methods. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11101 of Lecture Notes in Computer Science, pp.  274–285. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] +
+Keywords: machine decision-maker +
+ +
+ + +
+[1597] +
+
+Jürgen Branke, Salvatore Corrente, Salvatore Greco, Milosz Kadziński, Manuel López-Ibáñez, Vincent Mousseau, Mauro Munerato, and Roman Slowiński. + Behavior-Realistic Artificial Decision-Makers to Test Preference-Based Multi-objective Optimization Method (Working Group “Machine Decision-Making”). + In S. Greco, K. Klamroth, J. D. Knowles, and G. Rudolph, editors, Understanding Complexity in Multiobjective Optimization (Dagstuhl Seminar 15031), volume 5(1) of Dagstuhl Reports, pp.  110–116. Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2015.
+[ bib | +DOI ] +
+Keywords: multiple criteria decision making, evolutionary + multiobjective optimization +
+ +
+ + +
+[1598] +
+
+Jürgen Branke and Kalyanmoy Deb. + Integrating User Preferences into Evolutionary Multi-Objective Optimization. + In Y. Jin, editor, Knowledge Incorporation in Evolutionary Computation, pp.  461–477. Springer, Berlin/Heidelberg, 2005.
+[ bib | +DOI ] +
+Many real-world optimization problems involve multiple, + typically conflicting objectives. Often, it is very difficult + to weigh the different criteria exactly before alternatives + are known. Evolutionary multi-objective optimization usually + solves this predicament by searching for the whole + Pareto-optimal front of solutions. However, often the user + has at least a vague idea about what kind of solutions might + be preferred. In this chapter, we argue that such knowledge + should be used to focus the search on the most interesting + (from a user's perspective) areas of the Paretooptimal + front. To this end, we present and compare two methods which + allow to integrate vague user preferences into evolutionary + multi-objective algorithms. As we show, such methods may + speed up the search and yield a more fine-grained selection + of alternatives in the most relevant parts of the + Pareto-optimal front. +
+ +
+ + +
+[1599] +
+
+Yesnier Bravo, Javier Ferrer, Gabriel J. Luque, and Enrique Alba. + Smart Mobility by Optimizing the Traffic Lights: A New Tool for Traffic Control Centers. + In E. Alba, F. Chicano, and G. J. Luque, editors, Smart Cities (Smart-CT 2016), Lecture Notes in Computer Science, pp.  147–156. Springer, Cham, Switzerland, 2016.
+[ bib | +DOI ] +
+Urban traffic planning is a fertile area of Smart Cities to + improve efficiency, environmental care, and safety, since the + traffic jams and congestion are one of the biggest sources of + pollution and noise. Traffic lights play an important role in + solving these problems since they control the flow of the + vehicular network at the city. However, the increasing number + of vehicles makes necessary to go from a local control at one + single intersection to a holistic approach considering a + large urban area, only possible using advanced computational + resources and techniques. Here we propose HITUL, a system + that supports the decisions of the traffic control managers + in a large urban area. HITUL takes the real traffic + conditions and compute optimal traffic lights plans using + bio-inspired techniques and micro-simulations. We compare our + system against plans provided by experts. Our solutions not + only enable continuous traffic flows but reduce the + pollution. A case study of Málaga city allows us to + validate the approach and show its benefits for other cities + as well. +
+
+Keywords: Multi-objective optimization, Smart mobility, Traffic lights + planning +
+ +
+ + +
+[1600] +
+
+Jean-Pierre Brans and Bertrand Mareschal. + PROMETHEE-GAIA. Une méthode d'aide à la décision en présence de critères multiples. + Editions Ellipses, Paris, France, 2002.
+[ bib ] + +
+ + +
+[1601] +
+
+Jean-Pierre Brans and Bertrand Mareschal. + PROMETHEE Methods. + In J. R. Figueira, S. Greco, and M. Ehrgott, editors, Multiple Criteria Decision Analysis, State of the Art Surveys, chapter 5, pp.  163–195. Springer, 2005.
+[ bib ] + +
+ + +
+[1602] +
+
+Jürgen Branke, C. Schmidt, and H. Schmeck. + Efficient fitness estimation in noisy environments. + In E. D. Goodman, editor, Proceedings of the 3rd Annual Conference on Genetic and Evolutionary Computation, GECCO 2001, pp.  243–250. Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib ] + +
+ + +
+[1603] +
+
+Jürgen Branke, Salvatore Corrente, Salvatore Greco, Roman Slowiński, and P. Zielniewicz. + Using Choquet integral as preference model in interactive evolutionary multiobjective optimization. + Technical report, WBS, University of Warwick, 2014.
+[ bib ] + +
+ + +
+[1604] +
+
+Jürgen Branke and Jawad Elomari. + Simultaneous tuning of metaheuristic parameters for various computing budgets. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  263–264. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] +
+Keywords: meta-optimization, offline parameter optimization +
+ +
+ + +
+[1605] +
+
+Jürgen Branke and Jawad Elomari. + Racing with a Fixed Budget and a Self-Adaptive Significance Level. + In P. M. Pardalos and G. Nicosia, editors, Learning and Intelligent Optimization, 7th International Conference, LION 7, volume 7997 of Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[1606] +
+
+Leo Breiman, Jerome Friedman, Charles J. Stone, and Richard A. Olshen. + Classification and regression trees. + CRC Press, 1984.
+[ bib ] + +
+ + +
+[1607] +
+
+Mátyás Brendel and Marc Schoenauer. + Learn-and-Optimize: A Parameter Tuning Framework for Evolutionary AI Planning. + In J.-K. Hao, P. Legrand, P. Collet, N. Monmarché, E. Lutton, and M. Schoenauer, editors, Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011, volume 7401 of Lecture Notes in Computer Science, pp.  145–155. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[1608] +
+
+Mátyás Brendel and Marc Schoenauer. + Instance-based Parameter Tuning for Evolutionary AI Planning. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2011, pp.  591–598. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] + +
+ + +
+[1609] +
+
+Karl Bringmann and Tobias Friedrich. + Approximating the Least Hypervolume Contributor: NP-Hard in General, But Fast in Practice. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  6–20. Springer, Heidelberg, Germany, 2009.
+[ bib ] +
+Extended version published in [181] +
+ +
+ + +
+[1610] +
+
+Karl Bringmann and Tobias Friedrich. + The Maximum Hypervolume Set Yields Near-optimal Approximation. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  511–518. ACM Press, New York, NY, 2010.
+[ bib ] +
+Proved that hypervolume approximates the additive + ε-indicator, converging quickly as N increases, + that is, sets that maximize hypervolume are near optimal on + additive ε too, with the gap diminishing as quickly + as O(1/N). +
+ +
+ + +
+[1611] +
+
+Karl Bringmann and Tobias Friedrich. + Tight bounds for the approximation ratio of the hypervolume indicator. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  607–616. Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[1612] +
+
+Karl Bringmann and Tobias Friedrich. + Convergence of Hypervolume-Based Archiving Algorithms I: Effectiveness. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  745–752. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] +
+Extended version published as [183] +
+ +
+ + +
+[1613] +
+
+Karl Bringmann and Tobias Friedrich. + Convergence of Hypervolume-Based Archiving Algorithms II: Competitiveness. + In T. Soule and J. H. Moore, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2012, pp.  457–464. ACM Press, New York, NY, 2012.
+[ bib | +DOI ] +
+Extended version published as [183] +
+ +
+ + +
+[1614] +
+
+Karl Bringmann, Tobias Friedrich, and Patrick Klitzke. + Generic postprocessing via subset selection for hypervolume and epsilon-indicator. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  518–527. Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[1615] +
+
+Karl Bringmann, Tobias Friedrich, and Patrick Klitzke. + Two-dimensional subset selection for hypervolume and epsilon-indicator. + In C. Igel and D. V. Arnold, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2014. ACM Press, New York, NY, 2014.
+[ bib | +DOI ] + +
+ + +
+[1616] +
+
+Andre Britto and Aurora Pozo. + Using archiving methods to control convergence and diversity for many-objective problems in particle swarm optimization. + In Proceedings of the 2012 Congress on Evolutionary Computation (CEC 2012), pp.  1–8, Piscataway, NJ, 2012. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1617] +
+
+Karl Bringmann and Tobias Friedrich. + Don't be greedy when calculating hypervolume contributions. + In I. I. Garibay, T. Jansen, R. P. Wiegand, and A. S. Wu, editors, Proceedings of the Tenth ACM SIGEVO Workshop on Foundations of Genetic Algorithms (FOGA), pp.  103–112. ACM, 2009.
+[ bib ] +
+Extended version published in [182] +
+ +
+ + +
+[1618] +
+
+Karl Bringmann, Tobias Friedrich, Frank Neumann, and Markus Wagner. + Approximation-guided Evolutionary Multi-objective Optimization. + In T. Walsh, editor, Proceedings of the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11), pp.  1198–1203. IJCAI/AAAI Press, Menlo Park, CA, 2011.
+[ bib ] + +
+ + +
+[1619] +
+
+Dimo Brockhoff. + A Bug in the Multiobjective Optimizer IBEA: Salutary Lessons for Code Release and a Performance Re-Assessment. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  187–201. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] + +
+ + +
+[1620] +
+
+Dimo Brockhoff, Roberto Calandra, Manuel López-Ibáñez, Frank Neumann, and Selvakumar Ulaganathan. + Meta-modeling for (interactive) multi-objective optimization (WG5). + In K. Klamroth, J. D. Knowles, G. Rudolph, and M. M. Wiecek, editors, Personalized Multiobjective Optimization: An Analytics Perspective (Dagstuhl Seminar 18031), volume 8(1) of Dagstuhl Reports, pp.  85–94. Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2018.
+[ bib | +DOI ] +
+Keywords: multiple criteria decision making, evolutionary + multiobjective optimization +
+ +
+ + +
+[1621] +
+
+Dimo Brockhoff, Tobias Friedrich, N. Hebbinghaus, C. Klein, Frank Neumann, and Eckart Zitzler. + Do Additional Objectives Make a Problem Harder? + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  765–772. ACM Press, New York, NY, 2007.
+[ bib | +DOI ] + +
+ + +
+[1622] +
+
+Dimo Brockhoff, Manuel López-Ibáñez, Boris Naujoks, and Günther Rudolph. + Runtime Analysis of Simple Interactive Evolutionary Biobjective Optimization Algorithms. + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part I, volume 7491 of Lecture Notes in Computer Science, pp.  123–132. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] +
+Development and deployment of interactive evolutionary + multiobjective optimization algorithms (EMOAs) have recently + gained broad interest. In this study, first steps towards a + theory of interactive EMOAs are made by deriving bounds on + the expected number of function evaluations and queries to a + decision maker. We analyze randomized local search and the + (1+1)-EA on the biobjective problems LOTZ and COCZ under the + scenario that the decision maker interacts with these + algorithms by providing a subjective preference whenever + solutions are incomparable. It is assumed that this decision + is based on the decision maker's internal utility + function. We show that the performance of the interactive + EMOAs may dramatically worsen if the utility function is + non-linear instead of linear. +
+ +
+ + +
+[1623] +
+
+Dimo Brockhoff, Dhish Kumar Saxena, Kalyanmoy Deb, and Eckart Zitzler. + On Handling a Large Number of Objectives A Posteriori and During Optimization. + In J. D. Knowles, D. Corne, K. Deb, and D. R. Chair, editors, Multiobjective Problem Solving from Nature, Natural Computing Series, pp.  377–403. Springer, Berlin/Heidelberg, 2008.
+[ bib | +DOI ] +
+Dimensionality reduction methods are used routinely in + statistics, pattern recognition, data mining, and machine + learning to cope with high-dimensional spaces. Also in the + case of high-dimensional multiobjective optimization + problems, a reduction of the objective space can be + beneficial both for search and decision making. New questions + arise in this context, e.g., how to select a subset of + objectives while preserving most of the problem structure. In + this chapter, two different approaches to the task of + objective reduction are developed, one based on assessing + explicit conflicts, the other based on principal component + analysis (PCA). Although both methods use different + principles and preserve different properties of the + underlying optimization problems, they can be effectively + utilized either in an a posteriori scenario or during + search. Here, we demonstrate the usability of the + conflict-based approach in a decision-making scenario after + the search and show how the principal-component-based + approach can be integrated into an evolutionary + multicriterion optimization (EMO) procedure. +
+ +
+ + +
+[1624] +
+
+Dimo Brockhoff and Tea Tušar. + Benchmarking algorithms from the platypus framework on the biobjective bbob-biobj testbed. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1905–1911. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+Keywords: unbounded archive +
+ +
+ + +
+[1625] +
+
+Dimo Brockhoff, Tobias Wagner, and Heike Trautmann. + On the properties of the R2 indicator. + In T. Soule and J. H. Moore, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2012, pp.  465–472. ACM Press, New York, NY, 2012.
+[ bib ] +
+Proof that R2 is weakly Pareto compliant. +
+ +
+ + +
+[1626] +
+
+Dimo Brockhoff and Eckart Zitzler. + Are All Objectives Necessary? On Dimensionality Reduction in Evolutionary Multiobjective Optimization. + In T. P. Runarsson, H.-G. Beyer, E. K. Burke, J.-J. Merelo, D. Whitley, and X. Yao, editors, Parallel Problem Solving from Nature – PPSN IX, volume 4193 of Lecture Notes in Computer Science, pp.  533–542. Springer, Heidelberg, Germany, 2006.
+[ bib ] +
+Most of the available multiobjective evolutionary algorithms + (MOEA) for approximating the Pareto set have been designed + for and tested on low dimensional problems (≤3 + objectives). However, it is known that problems with a high + number of objectives cause additional difficulties in terms + of the quality of the Pareto set approximation and running + time. Furthermore, the decision making process becomes the + harder the more objectives are involved. In this context, the + question arises whether all objectives are necessary to + preserve the problem characteristics. One may also ask under + which conditions such an objective reduction is feasible, and + how a minimum set of objectives can be computed. In this + paper, we propose a general mathematical framework, suited to + answer these three questions, and corresponding algorithms, + exact and heuristic ones. The heuristic variants are geared + towards direct integration into the evolutionary search + process. Moreover, extensive experiments for four well-known + test problems show that substantial dimensionality reductions + are possible on the basis of the proposed methodology. +
+ +
+ + +
+[1627] +
+
+Dimo Brockhoff and Eckart Zitzler. + Dimensionality Reduction in Multiobjective Optimization: The Minimum Objective Subset Problem. + In K.-H. Waldmann and U. M. Stocker, editors, Operations Research Proceedings 2006, pp.  423–429. Springer, Berlin/Heidelberg, 2007.
+[ bib | +DOI ] +
+The number of objectives in a multiobjective optimization + problem strongly influences both the performance of + generating methods and the decision making process in + general. On the one hand, with more objectives, more + incomparable solutions can arise, the number of which affects + the generating method's performance. On the other hand, the + more objectives are involved the more complex is the choice + of an appropriate solution for a (human) decision maker. In + this context, the question arises whether all objectives are + actually necessary and whether some of the objectives may be + omitted; this question in turn is closely linked to the + fundamental issue of conflicting and non-conflicting + optimization criteria. Besides a general definition of + conflicts between objective sets, we here introduce the + NP-hard problem of computing a minimum subset of objectives + without losing information (MOSS). Furthermore, we present + for MOSS both an approximation algorithm with optimum + approximation ratio and an exact algorithm which works well + for small input instances. We conclude with experimental + results for a random problem and the multiobjective + 0/1-knapsack problem +
+
+Keywords: objective reduction +
+ +
+ + +
+[1628] +
+
+Dimo Brockhoff and Eckart Zitzler. + Improving hypervolume-based multiobjective evolutionary algorithms by using objective reduction methods. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  2086–2093, Piscataway, NJ, 2007. IEEE Press.
+[ bib | +DOI ] +
+Keywords: objective reduction +
+ +
+ + +
+[1629] +
+
+Artur Brum and Marcus Ritt. + Automatic Design of Heuristics for Minimizing the Makespan in Permutation Flow Shops. + In Proceedings of the 2018 Congress on Evolutionary Computation (CEC 2018), pp.  1–8, Piscataway, NJ, 2018. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1630] +
+
+Artur Brum and Marcus Ritt. + Automatic Algorithm Configuration for the Permutation Flow Shop Scheduling Problem Minimizing Total Completion Time. + In A. Liefooghe and M. López-Ibáñez, editors, Proceedings of EvoCOP 2018 – 18th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 10782 of Lecture Notes in Computer Science, pp.  85–100. Springer, Heidelberg, Germany, 2018.
+[ bib | +DOI ] + +
+ + +
+[1631] +
+
+T. N. Bui and J. R. Rizzo, Jr. + Finding Maximum Cliques with Distributed Ants. + In K. Deb et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2004, Part I, volume 3102 of Lecture Notes in Computer Science, pp.  24–35. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[1632] +
+
+Edmund K. Burke and Yuri Bykov. + The Late Acceptance Hill-Climbing Heuristic. + Technical Report CSM-192, University of Stirling, 2012.
+[ bib ] + +
+ + +
+[1633] +
+
+Edmund K. Burke, Matthew R. Hyde, Graham Kendall, and John R. Woodward. + Automatic Heuristic Generation with Genetic Programming: Evolving a Jack-of-all-trades or a Master of One. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  1559–1565. ACM Press, New York, NY, 2007.
+[ bib | +DOI ] + +
+ + +
+[1634] +
+
+Edmund K. Burke, Matthew R. Hyde, Graham Kendall, Gabriela Ochoa, Ender Özcan, and John R. Woodward. + A Classification of Hyper-Heuristic Approaches: Revisited. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 272 of International Series in Operations Research & Management Science, chapter 14, pp.  453–477. Springer, 2019.
+[ bib | +DOI ] + +
+ + +
+[1635] +
+
+Rainer E. Burkard, Eranda Çela, Panos M. Pardalos, and L. S. Pitsoulis. + The quadratic assignment problem. + In P. M. Pardalos and D.-Z. Du, editors, Handbook of Combinatorial Optimization, volume 2, pp.  241–338. Kluwer Academic Publishers, 1998.
+[ bib ] + +
+ + +
+[1636] +
+
+Maxim Buzdalov. + Towards better estimation of statistical significance when comparing evolutionary algorithms. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1782–1788. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[1637] +
+
+Nicola Beume, Carlos M. Fonseca, Manuel López-Ibáñez, Luís Paquete, and Jan Vahrenhold. + On the Complexity of Computing the Hypervolume Indicator. + Technical Report CI-235/07, University of Dortmund, December 2007. + Published in IEEE Transactions on Evolutionary Computation [123].
+[ bib ] + +
+ + +
+[1638] +
+
+COnfiguration and SElection of ALgorithms. + http://www.coseal.net, 2017.
+[ bib ] + +
+ + +
+[1639] +
+
+IBM. + ILOG CPLEX Optimizer. + http://www.ibm.com/software/integration/optimization/cplex-optimizer/, 2017.
+[ bib ] + +
+ + +
+[1640] +
+
+Borja Calvo, Ofer M. Shir, Josu Ceberio, Carola Doerr, Hao Wang, Thomas Bäck, and José A. Lozano. + Bayesian Performance Analysis for Black-box Optimization Benchmarking. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1789–1797. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+Keywords: bayesian inference, benchmarking, black-box optimization, + evolutionary algorithms, performance measures, plackett-luce + model +
+ +
+ + +
+[1641] +
+
+Christian Leonardo Camacho-Villalón, Marco Dorigo, and Thomas Stützle. + Why the Intelligent Water Drops Cannot Be Considered as a Novel Algorithm. + In M. Dorigo, M. Birattari, A. L. Christensen, A. Reina, and V. Trianni, editors, Swarm Intelligence, 11th International Conference, ANTS 2018, volume 11172 of Lecture Notes in Computer Science, pp.  302–314. Springer, Heidelberg, Germany, 2018.
+[ bib ] + +
+ + +
+[1642] +
+
+Paolo Campigotto and Andrea Passerini. + Adapting to a realistic decision maker: experiments towards a reactive multi-objective optimizer. + In C. Blum and R. Battiti, editors, Learning and Intelligent Optimization, 4th International Conference, LION 4, volume 6073 of Lecture Notes in Computer Science, pp.  338–341. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[1643] +
+
+Christian Leonardo Camacho-Villalón, Thomas Stützle, and Marco Dorigo. + Grey Wolf, Firefly and Bat Algorithms: Three Widespread Algorithms that Do Not Contain Any Novelty. + In M. Dorigo, T. Stützle, M. J. Blesa, C. Blum, H. Hamann, and M. K. Heinrich, editors, Swarm Intelligence, 12th International Conference, ANTS 2020, volume 12421 of Lecture Notes in Computer Science, pp.  121–133. Springer, Heidelberg, Germany, 2020.
+[ bib ] + +
+ + +
+[1644] +
+
+Felipe Campelo, Áthila R. Trindade, and Manuel López-Ibáñez. + Pseudoreplication in Racing Methods for Tuning Metaheuristics. + In preparation, 2017.
+[ bib ] + +
+ + +
+[1645] +
+
+E. Cantú-Paz. + Efficient and Accurate Parallel Genetic Algorithms. + Kluwer Academic Publishers, Boston, MA, 2000.
+[ bib ] + +
+ + +
+[1646] +
+
+P. Cardoso, M. Jesus, and A. Marquez. + MONACO: multi-objective network optimisation based on an ACO. + In Proc. X Encuentros de Geometría Computacional, Seville, Spain, 2003.
+[ bib ] + +
+ + +
+[1647] +
+
+Alex Guimarães Cardoso de Sá, Walter José G. S. Pinto, Luiz Otávio Vilas Boas Oliveira, and Gisele Pappa. + RECIPE: A Grammar-Based Framework for Automatically Evolving Classification Pipelines. + In J. McDermott, M. Castelli, L. Sekanina, E. Haasdijk, and P. García-Sánchez, editors, Proceedings of the 20th European Conference on Genetic Programming, EuroGP 2017, volume 10196 of Lecture Notes in Computer Science, pp.  246–261. Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[1648] +
+
+Ioannis Caragiannis, Ariel D. Procaccia, and Nisarg Shah. + When Do Noisy Votes Reveal the Truth? + In M. J. Kearns, R. P. McAfee, and É. Tardos, editors, Proceedings of the Fourteenth ACM Conference on Electronic Commerce, pp.  143–160. ACM Press, New York, NY, 2013.
+[ bib | +DOI ] +
+A well-studied approach to the design of voting rules views + them as maximum likelihood estimators; given votes that are + seen as noisy estimates of a true ranking of the + alternatives, the rule must reconstruct the most likely true + ranking. We argue that this is too stringent a requirement, + and instead ask: How many votes does a voting rule need to + reconstruct the true ranking? We define the family of + pairwise-majority consistent rules, and show that for all + rules in this family the number of samples required from the + Mallows noise model is logarithmic in the number of + alternatives, and that no rule can do asymptotically better + (while some rules like plurality do much worse). Taking a + more normative point of view, we consider voting rules that + surely return the true ranking as the number of samples tends + to infinity (we call this property accuracy in the limit); + this allows us to move to a higher level of abstraction. We + study families of noise models that are parametrized by + distance functions, and find voting rules that are accurate + in the limit for all noise models in such general + families. We characterize the distance functions that induce + noise models for which pairwise-majority consistent rules are + accurate in the limit, and provide a similar result for + another novel family of position-dominance consistent + rules. These characterizations capture three well-known + distance functions. +
+
+Keywords: computer social choice, mallows model, sample complexity +
+ +
+ + +
+[1649] +
+
+Josu Ceberio, Alexander Mendiburu, and José A. Lozano. + Kernels of Mallows Models for Solving Permutation-based Problems. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  505–512. ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[1650] +
+
+Eranda Çela. + The Quadratic Assignment Problem: Theory and Algorithms. + Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998.
+[ bib ] + +
+ + +
+[1651] +
+
+Amadeo Cesta, Angelo Oddi, and Stephen F. Smith. + Iterative Flattening: A Scalable Method for Solving Multi-Capacity Scheduling Problems. + In H. A. Kautz and B. W. Porter, editors, Proceedings of AAAI 2000 – Seventeenth National Conference on Artificial Intelligence, pp.  742–747. AAAI Press/MIT Press, Menlo Park, CA, 2000.
+[ bib ] + +
+ + +
+[1652] +
+
+S. T. H. Chang. + Optimizing the Real Time Operation of a Pumping Station at a Water Filtration Plant using Genetic Algorithms. + Honors thesis, Department of Civil and Environmental Engineering, The University of Adelaide, 1999.
+[ bib ] + +
+ + +
+[1653] +
+
+Donald V. Chase and Lindell E. Ormsbee. + Optimal pump operation of water distribution systems with multiple storage tanks. + In Proceedings of American Water Works Association Computer Specialty Conference, pp.  205–214, Denver, USA, 1989. AWWA.
+[ bib ] + +
+ + +
+[1654] +
+
+Donald V. Chase and Lindell E. Ormsbee. + An alternate formulation of time as a decision variable to facilitate real-time operation of water supply systems. + In Proceedings of the 18th Annual Conference of Water Resources Planning and Management, pp.  923–927, New York, NY, 1991. ASCE.
+[ bib ] + +
+ + +
+[1655] +
+
+Deyao Chen, Maxim Buzdalov, Carola Doerr, and Nguyen Dang. + Using Automated Algorithm Configuration for Parameter Control. + In F. Chicano, T. Friedrich, T. Kötzing, and F. Rothlauf, editors, Proceedings of the 17th ACM/SIGEVO Conference on Foundations of Genetic Algorithms, pp.  38–49. ACM, 2023.
+[ bib | +DOI ] + +
+ + +
+[1656] +
+
+Fei Chen, Yang Gao, Zhao-qian Chen, and Shi-fu Chen. + SCGA: Controlling genetic algorithms with Sarsa(0). + In Computational Intelligence for Modelling, Control and Automation, 2005 and International Conference on Intelligent Agents, Web Technologies and Internet Commerce, International Conference on, volume 1, pp.  1177–1183. IEEE, 2005.
+[ bib | +DOI ] + +
+ + +
+[1657] +
+
+Clément Chevalier, David Ginsbourger, Julien Bect, and Ilya Molchanov. + Estimating and Quantifying Uncertainties on Level Sets Using the Vorob'ev Expectation and Deviation with Gaussian Process Models. + In D. Ucinski, A. C. Atkinson, and M. Patan, editors, mODa 10–Advances in Model-Oriented Design and Analysis, pp.  35–43. Springer International Publishing, Heidelberg, Germany, 2013.
+[ bib | +DOI ] +
+Several methods based on Kriging have recently been proposed + for calculating a probability of failure involving + costly-to-evaluate functions. A closely related problem is to + estimate the set of inputs leading to a response exceeding a + given threshold. Now, estimating such a level set—and not + solely its volume—and quantifying uncertainties on it are + not straightforward. Here we use notions from random set + theory to obtain an estimate of the level set, together with + a quantification of estimation uncertainty. We give explicit + formulae in the Gaussian process set-up and provide a + consistency result. We then illustrate how space-filling + versus adaptive design strategies may sequentially reduce + level set estimation uncertainty. +
+ +
+ + +
+[1658] +
+
+Weiyu Chen, Hisao Ishibuchi, and Ke Shang. + Clustering-Based Subset Selection in Evolutionary Multiobjective Optimization. + In 2021 IEEE International Conference on Systems, Man, and Cybernetics, pp.  468–475. IEEE, 2021.
+[ bib ] + +
+ + +
+[1659] +
+
+Peter C. Cheeseman, Bob Kanefsky, and William M. Taylor. + Where the Really Hard Problems Are. + In J. Mylopoulos and R. Reiter, editors, Proceedings of the 12th International Joint Conference on Artificial Intelligence (IJCAI-91), pp.  331–340. Morgan Kaufmann Publishers, 1995.
+[ bib ] + +
+ + +
+[1660] +
+
+L. Chen, X. H. Xu, and Y. X. Chen. + An adaptive ant colony clustering algorithm. + In I. Cloete, K.-P. Wong, and M. Berthold, editors, Proceedings of the International Conference on Machine Learning and Cybernetics, pp.  1387–1392. IEEE Press, 2004.
+[ bib ] + +
+ + +
+[1661] +
+
+Weiyu Chen, Hisao Ishibuchi, and Ke Shang. + Modified Distance-based Subset Selection for Evolutionary Multi-objective Optimization Algorithms. + In Proceedings of the 2020 Congress on Evolutionary Computation (CEC 2020), pp.  1–8, Piscataway, NJ, 2020. IEEE Press.
+[ bib ] +
+Keywords: IGD+ +
+ +
+ + +
+[1662] +
+
+Lu Chen, Bin Xin, Jie Chen, and Juan Li. + A virtual-decision-maker library considering personalities and dynamically changing preference structures for interactive multiobjective optimization. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  636–641, Piscataway, NJ, 2017. IEEE Press.
+[ bib | +DOI ] +
+Keywords: machine DM, interactive EMOA +
+ +
+ + +
+[1663] +
+
+Marco Chiarandini and Yuri Goegebeur. + Mixed Models for the Analysis of Optimization Algorithms. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  225–264. Springer, Berlin, Germany, 2010.
+[ bib | +DOI ] +
+Preliminary version available as Tech. Rep. MF-2009-07-001 at the The Danish Mathematical Society +
+ +
+ + +
+[1664] +
+
+Marco Chiarandini. + Stochastic Local Search Methods for Highly Constrained Combinatorial Optimisation Problems. + PhD thesis, FB Informatik, TU Darmstadt, Germany, 2005.
+[ bib ] + +
+ + +
+[1665] +
+
+Tsung-Che Chiang. + nsga3cpp: A C++ implementation of NSGA-III. + http://web.ntnu.edu.tw/~tcchiang/publications/nsga3cpp/nsga3cpp.htm, 2014.
+[ bib ] + +
+ + +
+[1666] +
+
+Matthias Christen, Olaf Schenk, and Helmar Burkhart. + PATUS: A Code Generation and Autotuning Framework for Parallel Iterative Stencil Computations on Modern Microarchitectures. + In F. Mueller, editor, Proceedings of the 2011 IEEE International Parallel & Distributed Processing Symposium, IPDPS '11, pp.  676–687. IEEE Computer Society, 2011.
+[ bib | +DOI ] + +
+ + +
+[1667] +
+
+Jan Christiaens and Greet Vanden Berghe. + Slack Induction by String Removals for Vehicle Routing Problems. + Technical Report 7-05-2018, Department of Computing Science, KU Leuven, Gent, Belgium, 2018.
+[ bib ] + +
+ + +
+[1668] +
+
+Nicos Christofides. + Worst-case analysis of a new heuristic for the travelling salesman problem. + Technical Report 388, Graduate School of Industrial Administration, Carnegie-Mellon University, Pittsburgh, PA, 1976.
+[ bib ] + +
+ + +
+[1669] +
+
+Tinkle Chugh and Manuel López-Ibáñez. + Maximising Hypervolume and Minimising ε-Indicators using Bayesian Optimisation over Sets. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1326–1334. ACM Press, New York, NY, 2021.
+[ bib | +DOI | +supplementary material ] +
+Keywords: multi-objective, surrogate models, epsilon, hypervolume +
+ +
+ + +
+[1670] +
+
+S. Chusanapiputt, D. Nualhong, S. Jantarang, and S. Phoomvuthisarn. + Selective self-adaptive approach to ant system for solving unit commitment problem. + In M. Cattolico et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2006, pp.  1729–1736. ACM Press, New York, NY, 2006.
+[ bib ] + +
+ + +
+[1671] +
+
+Tinkle Chugh. + Handling expensive multiobjective optimization problems with evolutionary algorithms. + PhD thesis, University of Jyväskylä, 2017.
+[ bib ] + +
+ + +
+[1672] +
+
+Tinkle Chugh. + Scalarizing Functions in Bayesian Multiobjective Optimization. + In Proceedings of the 2020 Congress on Evolutionary Computation (CEC 2020), pp.  1–8, Piscataway, NJ, 2020. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1673] +
+
+Christian Cintrano, Javier Ferrer, Manuel López-Ibáñez, and Enrique Alba. + Hybridization of Racing Methods with Evolutionary Operators for Simulation Optimization of Traffic Lights Programs. + In C. Zarges and S. Verel, editors, Proceedings of EvoCOP 2021 – 21th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 12692 of Lecture Notes in Computer Science, pp.  17–33. Springer, Cham, Switzerland, 2021.
+[ bib | +DOI ] +
+In many real-world optimization problems, like the traffic + light scheduling problem tackled here, the evaluation of + candidate solutions requires the simulation of a process + under various scenarios. Thus, good solutions should not only + achieve good objective function values, but they must be + robust (low variance) across all different scenarios. + Previous work has revealed the effectiveness of IRACE for + this task. However, the operators used by IRACE to generate + new solutions were designed for configuring algorithmic + parameters, that have various data types (categorical, + numerical, etc.). Meanwhile, evolutionary algorithms have + powerful operators for numerical optimization, which could + help to sample new solutions from the best ones found in the + search. Therefore, in this work, we propose a hybridization + of the elitist iterated racing mechanism of IRACE with + evolutionary operators from differential evo- lution and + genetic algorithms. We consider a realistic case study + derived from the traffic network of Malaga (Spain) with 275 + traffic lights that should be scheduled optimally. After a + meticulous study, we discovered that the hybrid algorithm + comprising IRACE plus differential evolution offers + statistically better results than conventional algorithms and + also improves travel times and reduces pollution. +
+
+Extended version published in Evolutionary Computation journal [259]. +
+
+Keywords: Hybrid algorithms, Evolutionary algorithms, Simulation + optimization, Uncertainty, Traffic light planning +
+ +
+ + +
+[1674] +
+
+Jill Cirasella, David S. Johnson, Lyle A. McGeoch, and Weixiong Zhang. + The Asymmetric Traveling Salesman Problem: Algorithms, Instance Generators, and Tests. + In A. L. Buchsbaum and J. Snoeyink, editors, Algorithm Engineering and Experimentation, Third International Workshop, ALENEX 2001, Washington, DC, USA, January 5-6, 2001, Revised Papers, volume 2153 of Lecture Notes in Computer Science, pp.  32–59, Berlin, Germany, 2001. Springer.
+[ bib | +DOI ] + +
+ + +
+[1675] +
+
+Jon Claerbout and Martin Karrenbach. + Electronic documents give reproducible research a new meaning. + In SEG Technical Program Expanded Abstracts 1992, pp.  601–604. Society of Exploration Geophysicists, 1992.
+[ bib | +DOI ] +
+Proposed a reproducibility taxonomy, defined reproducibility + and taxonomy +
+ +
+ + +
+[1676] +
+
+Maurice Clerc and J. Kennedy. + Standard PSO 2011. + Particle Swarm Central, 2011.
+[ bib | +http ] + +
+ + +
+[1677] +
+
+Maurice Clerc. + Standard Particle Swarm Optimisation. + hal-00764996, September 2012.
+[ bib | +http ] +
+Since 2006, three successive standard PSO versions have been + put on line on the Particle Swarm Central + (http://particleswarm.info), namely SPSO 2006, 2007, + and 2011. The basic principles of all three versions can be + informally described the same way, and in general, this + statement holds for almost all PSO variants. However, the + exact formulae are slightly different, because they took + advantage of latest theoretical analysis available at the + time they were designed. +
+
+Keywords: particle swarm optimisation +
+ +
+ + +
+[1678] +
+
+Carlos A. Coello Coello. + Multi-objective Evolutionary Algorithms in Real-World Applications: Some Recent Results and Current Challenges. + In Advances in Evolutionary and Deterministic Methods for Design, Optimization and Control in Engineering and Sciences, pp.  3–18. Springer, 2015.
+[ bib | +DOI ] + +
+ + +
+[1679] +
+
+Carlos A. Coello Coello, Gary B. Lamont, and David A. Van Veldhuizen. + Evolutionary Algorithms for Solving Multi-Objective Problems. + Springer, New York, NY, 2nd edition, 2007.
+[ bib | +DOI ] + +
+ + +
+[1680] +
+
+Carlos A. Coello Coello and Margarita Reyes-Sierra. + A Study of the Parallelization of a Coevolutionary Multi-objective Evolutionary Algorithm. + In R. Monroy, G. Arroyo-Figueroa, L. E. Sucar, and H. Sossa, editors, Proceedings of MICAI, volume 2972 of Lecture Notes in Artificial Intelligence, pp.  688–697. Springer, Heidelberg, Germany, 2004.
+[ bib ] +
+Introduces Inverted Generational Distance (IGD) +
+
+Keywords: IGD +
+ +
+ + +
+[1681] +
+
+Carlos A. Coello Coello. + Handling Preferences in Evolutionary Multiobjective Optimization: A Survey. + In Proceedings of the 2000 Congress on Evolutionary Computation (CEC'00), pp.  30–37, Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[1682] +
+
+Carlos A. Coello Coello. + Recent Results and Open Problems in Evolutionary Multiobjective Optimization. + In C. Martín-Vide, R. Neruda, and M. A. Vega-Rodríguez, editors, Theory and Practice of Natural Computing - 6th International Conference, TPNC 2017, volume 10687 of Lecture Notes in Computer Science, pp.  3–21. Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[1683] +
+
+Paul R. Cohen. + Empirical Methods for Artificial Intelligence. + MIT Press, Cambridge, MA, 1995.
+[ bib ] + +
+ + +
+[1684] +
+
+G. Cohen. + Optimal Control of Water Supply Networks. + In S. G. Tzafestas, editor, Optimization and Control of Dynamic Operational Research Models, volume 4, chapter 8, pp.  251–276. North-Holland Publishing Company, Amsterdam, 1982.
+[ bib ] + +
+ + +
+[1685] +
+
+Alberto Colorni, Marco Dorigo, and Vittorio Maniezzo. + Distributed Optimization by Ant Colonies. + In F. J. Varela and P. Bourgine, editors, Proceedings of the First European Conference on Artificial Life, pp.  134–142. MIT Press, Cambridge, MA, 1992.
+[ bib ] + +
+ + +
+[1686] +
+
+Sonia Colas, Nicolas Monmarché, Pierre Gaucher, and Mohamed Slimane. + Artificial Ants for the Optimization of Virtual Keyboard Arrangement for Disabled People. + In N. Monmarché, E.-G. Talbi, P. Collet, M. Schoenauer, and E. Lutton, editors, Artificial Evolution, volume 4926 of Lecture Notes in Computer Science, pp.  87–99. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[1687] +
+
+Andrew R. Conn, Katya Scheinberg, and Luis N. Vicente. + Introduction to Derivative-Free Optimization. + MPS–SIAM Series on Optimization. Society for Industrial and Applied Mathematics, Philadelphia, PA, USA, 2009.
+[ bib ] + +
+ + +
+[1688] +
+
+David Applegate, Robert E. Bixby, Vašek Chvátal, and William J. Cook. + Concorde TSP Solver. + http://www.math.uwaterloo.ca/tsp/concorde.html, 2014. + Version visited last on 15 April 2014.
+[ bib ] + +
+ + +
+[1689] +
+
+W. J. Conover. + Practical Nonparametric Statistics. + John Wiley & Sons, New York, NY, 3rd edition, 1999.
+[ bib ] + +
+ + +
+[1690] +
+
+Stephen A. Cook. + The Complexity of Theorem-proving Procedures. + In Proceedings of the Third Annual ACM Symposium on Theory of Computing, STOC '71, pp.  151–158. ACM, 1971.
+[ bib | +DOI ] + +
+ + +
+[1691] +
+
+William J. Cook. + In Pursuit of the Traveling Salesman. + Princeton University Press, Princeton, NJ, 2012.
+[ bib ] + +
+ + +
+[1692] +
+
+William J. Cook. + Computing in Combinatorial Optimization. + In B. Steffen and G. Woeginger, editors, Computing and Software Science: State of the Art and Perspectives, volume 10000 of Lecture Notes in Computer Science, pp.  27–47. Springer, Cham, Switzerland, 2019.
+[ bib | +DOI ] + +
+ + +
+[1693] +
+
+David Corne, Nick R. Jerram, Joshua D. Knowles, and Martin J. Oates. + PESA-II: Region-Based Selection in Evolutionary Multiobjective Optimization. + In E. D. Goodman, editor, Proceedings of the 3rd Annual Conference on Genetic and Evolutionary Computation, GECCO 2001, pp.  283–290. Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib | +DOI ] + +
+ + +
+[1694] +
+
+David Corne and Joshua D. Knowles. + Some Multiobjective Optimizers are Better than Others. + In Proceedings of the 2003 Congress on Evolutionary Computation (CEC'03), pp.  2506–2512, Piscataway, NJ, December 2003. IEEE Press.
+[ bib ] + +
+ + +
+[1695] +
+
+David Corne and Joshua D. Knowles. + No free lunch and free leftovers theorems for multiobjective optimisation problems. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  327–341. Springer, Heidelberg, Germany, 2003.
+[ bib | +DOI ] + +
+ + +
+[1696] +
+
+David Corne, Joshua D. Knowles, and M. J. Oates. + The Pareto Envelope-Based Selection Algorithm for Multiobjective Optimization. + In M. Schoenauer et al., editors, Parallel Problem Solving from Nature – PPSN VI, volume 1917 of Lecture Notes in Computer Science, pp.  839–848. Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[1697] +
+
+Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, and Clifford Stein. + Introduction to algorithms. + MIT Press, Cambridge, MA, 2009.
+[ bib ] + +
+ + +
+[1698] +
+
+David Corne and Alan Reynolds. + Evaluating optimization algorithms: bounds on the performance of optimizers on unseen problems. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  707–710. ACM Press, New York, NY, 2011.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[1699] +
+
+Oscar Cordón, I. Fernández de Viana, Francisco Herrera, and L. Moreno. + A New ACO Model Integrating Evolutionary Computation Concepts: The Best-Worst Ant System. + In M. Dorigo et al., editors, Abstract proceedings of ANTS 2000 – From Ant Colonies to Artificial Ants: Second International Workshop on Ant Algorithms, pp.  22–29. IRIDIA, Université Libre de Bruxelles, Belgium, September 7–9 2000.
+[ bib ] + +
+ + +
+[1700] +
+
+Peter I. Cowling, Graham Kendall, and Eric Soubeiga. + A Hyperheuristic Approach to Scheduling a Sales Summit. + In E. K. Burke and W. Erben, editors, PATAT 2000: Proceedings of the 3rd International Conference of the Practice and Theory of Automated Timetabling, volume 2079 of Lecture Notes in Computer Science, pp.  176–190. Springer, 2000.
+[ bib | +DOI ] +
+First mention of the term hyper-heuristic. +
+ +
+ + +
+[1701] +
+
+M. J. Crawley. + The R Book. + Wiley, 2nd edition, 2012.
+[ bib ] + +
+ + +
+[1702] +
+
+W. B. Crowston, F. Glover, G. L. Thompson, and J. D. Trawick. + Probabilistic and Parametric Learning Combinations of Local Job Shop Scheduling Rules. + ONR Research Memorandum No. 117, GSIA, Carnegie-Mellon University, Pittsburgh, PA, USA, 1963.
+[ bib ] + +
+ + +
+[1703] +
+
+Joseph C. Culberson. + Iterated Greedy Graph Coloring and the Difficulty Landscape. + Technical Report 92-07, Department of Computing Science, The University of Alberta, Edmonton, Alberta, Canada, 1992.
+[ bib ] + +
+ + +
+[1704] +
+
+Joseph C. Culberson, A. Beacham, and D. Papp. + Hiding our Colors. + In Proceedings of the CP'95 Workshop on Studying and Solving Really Hard Problems, pp.  31–42, Cassis, France, September 1995.
+[ bib ] + +
+ + +
+[1705] +
+
+Joseph C. Culberson and F. Luo. + Exploring the k-colorable Landscape with Iterated Greedy. + In D. S. Johnson and M. A. Trick, editors, Cliques, Coloring, and Satisfiability: Second DIMACS Implementation Challenge, volume 26 of DIMACS Series on Discrete Mathematics and Theoretical Computer Science, pp.  245–284. American Mathematical Society, Providence, RI, 1996.
+[ bib ] + +
+ + +
+[1706] +
+
+Jeff Cumming. + Understanding the New Statistics – Effect Sizes, Confidence Intervals, and Meta-analysis. + Taylor & Francis, 2012.
+[ bib ] + +
+ + +
+[1707] +
+
+Nguyen Dang Thi Thanh and Patrick De Causmaecker. + Motivations for the Development of a Multi-objective Algorithm Configurator. + In B. Vitoriano, E. Pinson, and F. Valente, editors, ICORES 2014 - Proceedings of the 3rd International Conference on Operations Research and Enterprise Systems, pp.  328–333. SciTePress, 2014.
+[ bib ] + +
+ + +
+[1708] +
+
+Nguyen Dang Thi Thanh and Patrick De Causmaecker. + Characterization of Neighborhood Behaviours in a Multi-neighborhood Local Search Algorithm. + In P. Festa, M. Sellmann, and J. Vanschoren, editors, Learning and Intelligent Optimization, 10th International Conference, LION 10, volume 10079 of Lecture Notes in Computer Science, pp.  234–239. Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[1709] +
+
+Nguyen Dang and Patrick De Causmaecker. + Analysis of Algorithm Components and Parameters: Some Case Studies. + In N. F. Matsatsinis, Y. Marinakis, and P. M. Pardalos, editors, Learning and Intelligent Optimization, 13th International Conference, LION 13, volume 11968 of Lecture Notes in Computer Science, pp.  288–303. Springer, Cham, Switzerland, 2019.
+[ bib | +DOI ] +
+Modern high-performing algorithms are usually highly + parameterised, and can be configured either manually or by an + automatic algorithm configurator. The algorithm performance + dataset obtained after the configuration step can be used to + gain insights into how different algorithm parameters + influence algorithm performance. This can be done by a number + of analysis methods that exploit the idea of learning + prediction models from an algorithm performance dataset and + then using them for the data analysis on the importance of + variables. In this paper, we demonstrate the complementary + usage of three methods along this line, namely forward + selection, fANOVA and ablation analysis with surrogates on + three case studies, each of which represents some special + situations that the analyses can fall into. By these + examples, we illustrate how to interpret analysis results and + discuss the advantage of combining different analysis + methods. +
+ +
+ + +
+[1710] +
+
+Nguyen Dang and Carola Doerr. + Hyper-parameter tuning for the (1 + (λ, λ)) GA. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  889–897. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+Keywords: irace; theory +
+ +
+ + +
+[1711] +
+
+Nguyen Dang Thi Thanh, Leslie Pérez Cáceres, Patrick De Causmaecker, and Thomas Stützle. + Configuring irace Using Surrogate Configuration Benchmarks. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  243–250. ACM Press, New York, NY, 2017.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[1712] +
+
+Augusto Lopez Dantas and Aurora Trinidad Ramirez Pozo. + A Meta-Learning Algorithm Selection Approach for the Quadratic Assignment Problem. + In Proceedings of the 2018 Congress on Evolutionary Computation (CEC 2018), pp.  1–8, Piscataway, NJ, 2018. IEEE Press.
+[ bib ] + +
+ + +
+[1713] +
+
+Graeme C. Dandy and Matthew S. Gibbs. + Optimizing System Operations and Water Quality. + In P. Bizier and P. DeBarry, editors, Proceedings of World Water and Environmental Resources Congress. ASCE, Philadelphia, USA, 2003. + on CD-ROM.
+[ bib | +DOI ] + +
+ + +
+[1714] +
+
+Nguyen Dang Thi Thanh. + Data analytics for algorithm design. + PhD thesis, KU Leuven, Belgium, 2018.
+[ bib ] +
+Supervised by Patrick De Causmaecker +
+ +
+ + +
+[1715] +
+
+Fabio Daolio, Sébastien Verel, Gabriela Ochoa, and Marco Tomassini. + Local Optima Networks and the Performance of Iterated Local Search. + In T. Soule and J. H. Moore, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2012, pp.  369–376. ACM Press, New York, NY, 2012.
+[ bib ] + +
+ + +
+[1716] +
+
+Samuel Daulton, Maximilian Balandat, and Eytan Bakshy. + Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization. + In H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems (NeurIPS 33), pp.  9851–9864, 2020.
+[ bib | +epub ] + +
+ + +
+[1717] +
+
+Werner de Schaetzen, Dragan A. Savic, and Godfrey A. Walters. + A genetic algorithm approach to pump scheduling in water supply. + In V. Babovic and L. C. Larsen, editors, Hydroinformatics '98, pp.  897–899, Rotterdam, Balkema, 1998.
+[ bib ] + +
+ + +
+[1718] +
+
+Thomas Dean and Mark S. Boddy. + An Analysis of Time-Dependent Planning. + In H. E. Shrobe, T. M. Mitchell, and R. G. Smith, editors, Proceedings of the 7th National Conference on Artificial Intelligence, AAAI-88, pp.  49–54. AAAI Press/MIT Press, Menlo Park, CA, 1988.
+[ bib | +http ] +
+Keywords: anytime, performance profiles +
+ +
+ + +
+[1719] +
+
+Angela Dean and Daniel Voss. + Design and Analysis of Experiments. + Springer, London, UK, 1999.
+[ bib | +DOI ] + +
+ + +
+[1720] +
+
+Kalyanmoy Deb. + Introduction to evolutionary multiobjective optimization. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science, pp.  59–96. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] +
+In its current state, evolutionary multiobjective + optimization (EMO) is an established field of research and + application with more than 150 PhD theses, more than ten + dedicated texts and edited books, commercial softwares and + numerous freely downloadable codes, a biannual conference + series running successfully since 2001, special sessions and + workshops held at all major evolutionary computing + conferences, and full-time researchers from universities and + industries from all around the globe. In this chapter, we + provide a brief introduction to EMO principles, illustrate + some EMO algorithms with simulated results, and outline the + current research and application potential of EMO. For + solving multiobjective optimization problems, EMO procedures + attempt to find a set of well-distributed Pareto-optimal + points, so that an idea of the extent and shape of the + Pareto-optimal front can be obtained. Although this task was + the early motivation of EMO research, EMO principles are now + being found to be useful in various other problem solving + tasks, enabling one to treat problems naturally as they + are. One of the major current research thrusts is to combine + EMO procedures with other multiple criterion decision making + (MCDM) tools so as to develop hybrid and interactive + multiobjective optimization algorithms for finding a set of + trade-off optimal solutions and then choose a preferred + solution for implementation. This chapter provides the + background of EMO principles and their potential to launch + such collaborative studies with MCDM researchers in the + coming years. +
+ +
+ + +
+[1721] +
+
+Kalyanmoy Deb. + Multi-objective optimization. + In E. K. Burke and G. Kendall, editors, Search Methodologies, pp.  273–316. Springer, Boston, MA, 2005.
+[ bib | +DOI ] + +
+ + +
+[1722] +
+
+Kalyanmoy Deb. + Multi-Objective Optimization Using Evolutionary Algorithms. + Wiley, Chichester, UK, 2001.
+[ bib ] + +
+ + +
+[1723] +
+
+Kalyanmoy Deb and S. Agrawal. + A Niched-Penalty Approach for Constraint Handling in Genetic Algorithms. + In A. Dobnikar, N. C. Steele, D. W. Pearson, and R. F. Albrecht, editors, Artificial Neural Nets and Genetic Algorithms (ICANNGA-99), pp.  235–243. Springer Verlag, 1999.
+[ bib | +DOI ] +
+Keywords: polynomial mutation +
+ +
+ + +
+[1724] +
+
+Kalyanmoy Deb, S. Agarwal, A. Pratap, and T. Meyarivan. + A fast elitist non-dominated sorting genetic algorithm for multi-objective optimization: NSGA-II. + In M. Schoenauer et al., editors, Parallel Problem Solving from Nature – PPSN VI, volume 1917 of Lecture Notes in Computer Science, pp.  849–858. Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[1725] +
+
+Kalyanmoy Deb and Sachin Jain. + Multi-Speed Gearbox Design Using Multi-Objective Evolutionary Algorithms. + Technical Report 2002001, KanGAL, February 2002.
+[ bib ] + +
+ + +
+[1726] +
+
+Kalyanmoy Deb and Christie Myburgh. + Breaking the billion-variable barrier in real-world optimization using a customized evolutionary algorithm. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2016, pp.  653–660. ACM Press, New York, NY, 2016.
+[ bib ] + +
+ + +
+[1727] +
+
+Kalyanmoy Deb and Ankur Sinha. + Solving Bilevel Multi-Objective Optimization Problems Using Evolutionary Algorithms. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  110–124. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[1728] +
+
+Kalyanmoy Deb and J. Sundar. + Reference point based multi-objective optimization using evolutionary algorithms. + In M. Cattolico et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2006, pp.  635–642. ACM Press, New York, NY, 2006.
+[ bib | +DOI ] +
+Proposed R-NSGA-II +
+ +
+ + +
+[1729] +
+
+Kalyanmoy Deb, Rahul Tewari, Mayur Dixit, and Joydeep Dutta. + Finding trade-off solutions close to KKT points using evolutionary multi-objective optimization. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  2109–2116, Piscataway, NJ, 2007. IEEE Press.
+[ bib ] + +
+ + +
+[1730] +
+
+Kalyanmoy Deb, Lothar Thiele, Marco Laumanns, and Eckart Zitzler. + Scalable Test Problems for Evolutionary Multi-Objective Optimization. + Technical Report 112, Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Zürich, Switzerland, 2001. + Do not cite this TR! It is incorrect and it is superseeded by [1731].
+[ bib ] +
+Keywords: DTLZ benchmark +
+ +
+ + +
+[1731] +
+
+Kalyanmoy Deb, Lothar Thiele, Marco Laumanns, and Eckart Zitzler. + Scalable Test Problems for Evolutionary Multiobjective Optimization. + In A. Abraham, L. Jain, and R. Goldberg, editors, Evolutionary Multiobjective Optimization, Advanced Information and Knowledge Processing, pp.  105–145. Springer, London, UK, January 2005.
+[ bib | +DOI ] +
+Keywords: DTLZ benchmark +
+ +
+ + +
+[1732] +
+
+William A. Dees, Jr. and Patrick G. Karger. + Automated Rip-up and Reroute Techniques. + In DAC'82, Proceedings of the 19th Design Automation Workshop, pp.  432–439. IEEE Press, 1982.
+[ bib ] + +
+ + +
+[1733] +
+
+Matthijs L. den Besten. + Simple Metaheuristics for Scheduling. + PhD thesis, FB Informatik, TU Darmstadt, Germany, 2004.
+[ bib | +http ] + +
+ + +
+[1734] +
+
+Roman Denysiuk, Lino Costa, and Isabel Espírito Santo. + Many-objective optimization using differential evolution with variable-wise mutation restriction. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2013, pp.  591–598. ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[1735] +
+
+Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. + Imagenet: A large-scale hierarchical image database. + In Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Conference on, pp.  248–255. IEEE, 2009.
+[ bib ] + +
+ + +
+[1736] +
+
+Marcelo De Souza and Marcus Ritt. + An Automatically Designed Recombination Heuristic for the Test-Assignment Problem. + In Proceedings of the 2018 Congress on Evolutionary Computation (CEC 2018), pp.  1–8, Piscataway, NJ, 2018. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1737] +
+
+Marcelo De Souza and Marcus Ritt. + Automatic Grammar-Based Design of Heuristic Algorithms for Unconstrained Binary Quadratic Programming. + In A. Liefooghe and M. López-Ibáñez, editors, Proceedings of EvoCOP 2018 – 18th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 10782 of Lecture Notes in Computer Science, pp.  67–84. Springer, Heidelberg, Germany, 2018.
+[ bib | +DOI ] + +
+ + +
+[1738] +
+
+Marcelo De Souza and Marcus Ritt. + Hybrid Heuristic for Unconstrained Binary Quadratic Programming – Source Code of HHBQP. + https://github.com/souzamarcelo/hhbqp, 2018.
+[ bib ] + +
+ + +
+[1739] +
+
+Marcelo De Souza, Marcus Ritt, Manuel López-Ibáñez, and Leslie Pérez Cáceres. + ACVIZ: A Tool for the Visual Analysis of the Configuration of Algorithms with irace – Source Code. + https://github.com/souzamarcelo/acviz, 2020.
+[ bib ] + +
+ + +
+[1740] +
+
+Marcelo De Souza, Marcus Ritt, Manuel López-Ibáñez, and Leslie Pérez Cáceres. + ACVIZ: Algorithm Configuration Visualizations for irace: Supplementary material. + http://doi.org/10.5281/zenodo.4714582, September 2020.
+[ bib ] + +
+ + +
+[1741] +
+
+Sophie Dewez. + On the toll setting problem. + PhD thesis, Faculté de Sciences, Université Libre de Bruxelles, 2014.
+[ bib ] +
+Supervised by Dr. Martine Labbé +
+ +
+ + +
+[1742] +
+
+Ilias Diakonikolas and Mihalis Yannakakis. + Succinct approximate convex Pareto curves. + In Proceedings of the nineteenth annual ACM-SIAM symposium on Discrete algorithms, pp.  74–83. Society for Industrial and Applied Mathematics, 2008.
+[ bib ] + +
+ + +
+[1743] +
+
+Diego Díaz, Pablo Valledor, Paula Areces, Jorge Rodil, and Montserrat Suárez. + An ACO Algorithm to Solve an Extended Cutting Stock Problem for Scrap Minimization in a Bar Mill. + In M. Dorigo et al., editors, Swarm Intelligence, 9th International Conference, ANTS 2014, volume 8667 of Lecture Notes in Computer Science, pp.  13–24. Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[1744] +
+
+Luca Di Gaspero, Marco Chiarandini, and Andrea Schaerf. + A Study on the Short-Term Prohibition Mechanisms in Tabu Search. + In G. Brewka, S. Coradeschi, A. Perini, and P. Traverso, editors, Proceedings of the 17th European Conference on Artificial Intelligence, ECAI 2006, Riva del Garda, Italy, August29 - September 1, 2006, pp.  83–87. IOS Press, 2006.
+[ bib ] + +
+ + +
+[1745] +
+
+Luca Di Gaspero, Andrea Rendl, and Tommaso Urli. + Constraint-Based Approaches for Balancing Bike Sharing Systems. + In C. Schulte, editor, Principles and Practice of Constraint Programming, volume 8124 of Lecture Notes in Computer Science, pp.  758–773. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[1746] +
+
+Luca Di Gaspero, Andrea Rendl, and Tommaso Urli. + A Hybrid ACO+CP for Balancing Bicycle Sharing Systems. + In M. J. Blesa, C. Blum, P. Festa, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics, volume 7919 of Lecture Notes in Computer Science, pp.  198–212. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] +
+Keywords: F-race +
+ +
+ + +
+[1747] +
+
+Daniel Doblas, Antonio J. Nebro, Manuel López-Ibáñez, José García-Nieto, and Carlos A. Coello Coello. + Automatic Design of Multi-objective Particle Swarm Optimizers. + In M. Dorigo, H. Hamann, M. López-Ibáñez, J. García-Nieto, A. Engelbrecht, C. Pinciroli, V. Strobel, and C. L. Camacho-Villalón, editors, Swarm Intelligence, 13th International Conference, ANTS 2022, volume 13491 of Lecture Notes in Computer Science, pp.  28–40. Springer, Cham, Switzerland, 2022.
+[ bib | +DOI ] + +
+ + +
+[1748] +
+
+Pedro Domingos and Geoff Hulten. + Mining high-speed data streams. + In R. Ramakrishnan, S. J. Stolfo, R. J. Bayardo, and I. Parsa, editors, The 6th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2000, pp.  71–80. ACM Press, New York, NY, 2000.
+[ bib | +epub ] + +
+ + +
+[1749] +
+
+Marco Dorigo and Gianni A. Di Caro. + The Ant Colony Optimization Meta-Heuristic. + In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in Optimization, pp.  11–32. McGraw Hill, London, UK, 1999.
+[ bib ] + +
+ + +
+[1750] +
+
+Marco Dorigo and L. M. Gambardella. + Ant Colony System. + Technical Report IRIDIA/96-05, IRIDIA, Université Libre de Bruxelles, Belgium, 1996.
+[ bib ] + +
+ + +
+[1751] +
+
+Marco Dorigo, Vittorio Maniezzo, and Alberto Colorni. + The Ant System: An autocatalytic optimizing process. + Technical Report 91-016 Revised, Dipartimento di Elettronica, Politecnico di Milano, Italy, 1991.
+[ bib ] + +
+ + +
+[1752] +
+
+Marco Dorigo, Vittorio Maniezzo, and Alberto Colorni. + Positive Feedback as a Search Strategy. + Technical Report 91-016, Dipartimento di Elettronica, Politecnico di Milano, Italy, 1991.
+[ bib ] + +
+ + +
+[1753] +
+
+Marco Dorigo, Marco A. Montes de Oca, Sabrina Oliveira, and Thomas Stützle. + Ant Colony Optimization. + In J. J. Cochran, editor, Wiley Encyclopedia of Operations Research and Management Science, volume 1, pp.  114–125. John Wiley & Sons, 2011.
+[ bib | +DOI ] + +
+ + +
+[1754] +
+
+Marco Dorigo and Thomas Stützle. + The Ant Colony Optimization Metaheuristic: Algorithms, Applications and Advances. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  251–285. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[1755] +
+
+Marco Dorigo and Thomas Stützle. + Ant Colony Optimization. + MIT Press, Cambridge, MA, 2004.
+[ bib ] + +
+ + +
+[1756] +
+
+Marco Dorigo. + Optimization, Learning and Natural Algorithms. + PhD thesis, Dipartimento di Elettronica, Politecnico di Milano, Italy, 1992. + In Italian.
+[ bib ] + +
+ + +
+[1757] +
+
+Johann Dréo. + Using performance fronts for parameter setting of stochastic metaheuristics. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2009, pp.  2197–2200. ACM Press, New York, NY, 2009.
+[ bib | +DOI ] + +
+ + +
+[1758] +
+
+Johann Dréo, Carola Doerr, and Yann Semet. + Coupling the design of benchmark with algorithm in landscape-aware solver design. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1419–1420. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[1759] +
+
+Johann Dréo, Arnaud Liefooghe, Sébastien Verel, Marc Schoenauer, Juan-Julián Merelo, Alexandre Quemy, Benjamin Bouvier, and Jan Gmys. + Paradiseo: from a modular framework for evolutionary computation to the automated design of metaheuristics: 22 years of Paradiseo. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1522–1530. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] +
+Keywords: metaheuristics, evolutionary computation, software framework, + automated algorithm design +
+ +
+ + +
+[1760] +
+
+Johann Dréo and P. Siarry. + A New Ant Colony Algorithm Using the Heterarchical Concept Aimed at Optimization of Multiminima Continuous Functions. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  216–221. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1761] +
+
+Johann Dréo. + Adaptation de la métaheuristique des colonies de fourmis pour l'optimisation difficile en variables continues: Application en génie biologique et médical. + PhD thesis, Université Paris XII Val de Marne, December 2003.
+[ bib | +http ] +
+Keywords: metaheuristic ; continuous optimization ; global optimization + ; imagery ; registration ; ant colony algorithm ; estimation + of distribution algorithm ; evolutionary computation ; + métaheuristique ; optimisation continue ; optimisation + globale ; imagerie ; biomédical ; recalage ; algorithme + de colonie de fourmis ; algorithme à estimation de + distribution ; algorithme évolutionnaire +
+ +
+ + +
+[1762] +
+
+Stefan Droste, Thomas Jansen, and Ingo Wegener. + A new framework for the valuation of algorithms for black-box-optimization. + In K. A. De Jong, R. Poli, and J. E. Rowe, editors, Proceedings of the Seventh Workshop on Foundations of Genetic Algorithms (FOGA), pp.  253–270. Morgan Kaufmann Publishers, 2002.
+[ bib ] + +
+ + +
+[1763] +
+
+Hisao Ishibuchi, Lie Meng Pang, and Ke Shang. + A new framework of evolutionary multi-objective algorithms with an unbounded external archive. + In G. D. Giacomo, A. Catala, B. Dilkina, M. Milano, S. Barro, A. Bugarín, and J. Lang, editors, Proceedings of the 24th European Conference on Artificial Intelligence (ECAI), volume 325 of Frontiers in Artificial Intelligence and Applications. IOS Press, 2020.
+[ bib ] + +
+ + +
+[1764] +
+
+Chris Drummond. + Replicability is not Reproducibility: Nor is it Good Science. + In Proceedings of the Evaluation Methods for Machine Learning Workshop at the 26th ICML, Montreal, Canada, 2009.
+[ bib | +http ] + +
+ + +
+[1765] +
+
+Mădălina M. Drugan and Dirk Thierens. + Path-Guided Mutation for Stochastic Pareto Local Search Algorithms. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  485–495. Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[1766] +
+
+Abraham Duarte, Jesús Sánchez-Oro, Nenad Mladenović, and Raca Todosijević. + Variable Neighborhood Descent. + In R. Martí, P. M. Pardalos, and M. G. C. Resende, editors, Handbook of Heuristics, pp.  341–367. Springer International Publishing, 2018.
+[ bib | +DOI ] + +
+ + +
+[1767] +
+
+Jérémie Dubois-Lacoste. + Weight Setting Strategies for Two-Phase Local Search: A Study on Biobjective Permutation Flowshop Scheduling. + Technical Report TR/IRIDIA/2009-024, IRIDIA, Université Libre de Bruxelles, Belgium, 2009.
+[ bib ] + +
+ + +
+[1768] +
+
+Jérémie Dubois-Lacoste, Holger H. Hoos, and Thomas Stützle. + On the Empirical Scaling Behaviour of State-of-the-art Local Search Algorithms for the Euclidean TSP. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  377–384. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] + +
+ + +
+[1769] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Effective Hybrid Stochastic Local Search Algorithms for Biobjective Permutation Flowshop Scheduling. + In M. J. Blesa, C. Blum, L. Di Gaspero, A. Roli, M. Sampels, and A. Schaerf, editors, Hybrid Metaheuristics, volume 5818 of Lecture Notes in Computer Science, pp.  100–114. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[1770] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Supplementary material: Improving the Anytime Behavior of Two-Phase Local Search. + http://iridia.ulb.ac.be/supp/IridiaSupp2010-012, 2010.
+[ bib ] + +
+ + +
+[1771] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Supplementary material: A Hybrid TP+PLS Algorithm for Bi-objective Flow-shop Scheduling Problems. + http://iridia.ulb.ac.be/supp/IridiaSupp2010-001, 2010.
+[ bib ] + +
+ + +
+[1772] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Adaptive “Anytime” Two-Phase Local Search. + In C. Blum and R. Battiti, editors, Learning and Intelligent Optimization, 4th International Conference, LION 4, volume 6073 of Lecture Notes in Computer Science, pp.  52–67. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[1773] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Configuration of State-of-the-art Multi-Objective Optimizers Using the TP+PLS Framework. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  2019–2026. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] + +
+ + +
+[1774] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Pareto Local Search Algorithms for Anytime Bi-objective Optimization. + In J.-K. Hao and M. Middendorf, editors, Proceedings of EvoCOP 2012 – 12th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 7245 of Lecture Notes in Computer Science, pp.  206–217. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[1775] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Combining Two Search Paradigms for Multi-objective Optimization: Two-Phase and Pareto Local Search. + In E.-G. Talbi, editor, Hybrid Metaheuristics, volume 434 of Studies in Computational Intelligence, pp.  97–117. Springer Verlag, 2013.
+[ bib | +DOI | +http ] + +
+ + +
+[1776] +
+
+Jérémie Dubois-Lacoste, Federico Pagnozzi, and Thomas Stützle. + Supplementary material: An iterated greedy algorithm with optimization of partial solutions for the permutation flowshop problem. + http://iridia.ulb.ac.be/supp/IridiaSupp2013-006, 2017.
+[ bib ] + +
+ + +
+[1777] +
+
+Jérémie Dubois-Lacoste and Thomas Stützle. + Tuning of a Stigmergy-based Traffic Light Controller as a Dynamic Optimization Problem. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  1–8, Piscataway, NJ, 2017. IEEE Press.
+[ bib ] + +
+ + +
+[1778] +
+
+Jérémie Dubois-Lacoste. + A study of Pareto and Two-Phase Local Search Algorithms for Biobjective Permutation Flowshop Scheduling. + Master's thesis, IRIDIA, Université Libre de Bruxelles, Belgium, 2009.
+[ bib ] + +
+ + +
+[1779] +
+
+Jérémie Dubois-Lacoste. + Effective Stochastic Local Search Algorithms For Bi-Objective Permutation Flowshop Scheduling. + Rapport d'avancement de recherches présenté pour la formation doctorale en sciences de l'ingénieur, IRIDIA, Université Libre de Bruxelles, Belgium, 2010.
+[ bib ] + +
+ + +
+[1780] +
+
+Jérémie Dubois-Lacoste. + Anytime Local Search for Multi-Objective Combinatorial Optimization: Design, Analysis and Automatic Configuration. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2014.
+[ bib ] +
+Supervised by Thomas Stützle and Manuel López-Ibáñez +
+ +
+ + +
+[1781] +
+
+Gunter Dueck, Martin Maehler, Johannes Schneider, Gerhard Schrimpf, and Hermann Stamm-Wilbrandt. + Optimization with Ruin Recreate. + US Patent 6,418,398 B1, July 2002. + Filed on October 1, 1999 and granted on July 9, 2002; Assignee is IBM Corporation, Armonk, NY (US).
+[ bib ] + +
+ + +
+[1782] +
+
+Irina Dumitrescu and Thomas Stützle. + Combinations of Local Search and Exact Algorithms. + In G. R. Raidl and J. Gottlieb, editors, Proceedings of EvoCOP 2003 – 3rd European Conference on Evolutionary Computation in Combinatorial Optimization, volume 2611 of Lecture Notes in Computer Science, pp.  211–223. Springer, Heidelberg, Germany, 2003.
+[ bib | +DOI ] + +
+ + +
+[1783] +
+
+Irina Dumitrescu and Thomas Stützle. + Usage of Exact Algorithms to Enhance Stochastic Local Search Algorithms. + In V. Maniezzo, T. Stützle, and S. Voß, editors, Matheuristics—Hybridizing Metaheuristics and Mathematical Programming, volume 10 of Annals of Information Systems, pp.  103–134. Springer, New York, NY, 2009.
+[ bib ] + +
+ + +
+[1784] +
+
+Juan J. Durillo, José García-Nieto, Antonio J. Nebro, Carlos A. Coello Coello, Francisco Luna, and Enrique Alba. + Multi-Objective Particle Swarm Optimizers: An Experimental Comparison. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  495–509. Springer, Heidelberg, Germany, 2009.
+[ bib ] +
+Particle Swarm Optimization (PSO) has received increasing + attention in the optimization research community since its + first appearance in the mid-1990s. Regarding multi-objective + optimization, a considerable number of algorithms based on + Multi-Objective Particle Swarm Optimizers (MOPSOs) can be + found in the specialized literature. Unfortunately, no + experimental comparisons have been made in order to clarify + which MOPSO version shows the best performance. In this + paper, we use a benchmark composed of three well-known + problem families (ZDT, DTLZ, and WFG) with the aim of + analyzing the search capabilities of six representative + state-of-the-art MOPSOs, namely, NSPSO, SigmaMOPSO, OMOPSO, + AMOPSO, MOPSOpd, and CLMOPSO. We additionally propose a new + MOPSO algorithm, called SMPSO, characterized by including a + velocity constraint mechanism, obtaining promising results + where the rest perform inadequately. +
+ +
+ + +
+[1785] +
+
+Juan J. Durillo, Antonio J. Nebro, Francisco Luna, and Enrique Alba. + On the Effect of the Steady-State Selection Scheme in Multi-Objective Genetic Algorithms. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  183–197. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[1786] +
+
+Cynthia Dwork, Ravi Kumar, Moni Naor, and D. Sivakumar. + Rank aggregation methods for the Web. + In V. Y. Shen, N. Saito, M. R. Lyu, and M. E. Zurko, editors, Proceedings of the Tenth International World Wide Web Conference, WWW 10, pp.  613–622. ACM Press, New York, NY, 2001.
+[ bib | +DOI ] +
+Keywords: Kemeny ranking,multi-word queries,rank aggregation,ranking + functions,spam +
+ +
+ + +
+[1787] +
+
+L. A. Rossman. + EPANET 2 Users Manual. + U.S. Environmental Protection Agency, Cincinnati, USA, 2000.
+[ bib ] + +
+ + +
+[1788] +
+
+L. A. Rossman. + EPANET User's Guide. + Risk Reduction Engineering Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, USA, 1994.
+[ bib ] + +
+ + +
+[1789] +
+
+L. A. Rossman. + The EPANET Programmer's Toolkit for Analysis of Water Distribution Systems. + In Proceedings of the Annual Water Resources Planning and Management Conference, Reston, USA, 1999. ASCE.
+[ bib ] + +
+ + +
+[1790] +
+
+Russell C. Eberhart and J. Kennedy. + A New Optimizer Using Particle Swarm Theory. + In Proceedings of the Sixth International Symposium on Micro Machine and Human Science, pp.  39–43, 1995.
+[ bib ] + +
+ + +
+[1791] +
+
+Katharina Eggensperger, Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Efficient Benchmarking of Hyperparameter Optimizers via Surrogates. + In B. Bonet and S. Koenig, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  1114–1120. AAAI Press, 2015.
+[ bib | +DOI ] + +
+ + +
+[1792] +
+
+Werner Ehm. + Reproducibility from the perspective of meta-analysis. + In H. Atmanspacher and S. Maasen, editors, Reproducibility – Principles, problems, practices and prospects, pp.  141–168. Wiley, 2016.
+[ bib ] + +
+ + +
+[1793] +
+
+Matthias Ehrgott and Xavier Gandibleux. + Hybrid Metaheuristics for Multi-objective Combinatorial Optimization. + In C. Blum, M. J. Blesa, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics: An emergent approach for optimization, volume 114 of Studies in Computational Intelligence, pp.  221–259. Springer, Berlin, Germany, 2008.
+[ bib | +DOI ] +
+Many real-world optimization problems can be + modelled as combinatorial optimization + problems. Often, these problems are characterized by + their large size and the presence of multiple, + conflicting objectives. Despite progress in solving + multi-objective combinatorial optimization problems + exactly, the large size often means that heuristics + are required for their solution in acceptable time. + Since the middle of the nineties the trend is + towards heuristics that “pick and choose” elements + from several of the established metaheuristic + schemes. Such hybrid approximation techniques may + even combine exact and heuristic approaches. In this + chapter we give an overview over approximation + methods in multi-objective combinatorial + optimization. We briefly summarize “classical” + metaheuristics and focus on recent approaches, where + metaheuristics are hybridized and/or combined with + exact methods. +
+ +
+ + +
+[1794] +
+
+Matthias Ehrgott. + Multicriteria Optimization, volume 491 of Lecture Notes in Economics and Mathematical Systems. + Springer, Berlin, Germany, 2000.
+[ bib ] + +
+ + +
+[1795] +
+
+Matthias Ehrgott. + Multicriteria Optimization. + Springer, Berlin, Germany, 2nd edition, 2005.
+[ bib | +DOI ] + +
+ + +
+[1796] +
+
+Agoston E. Eiben, Mark Horvath, Wojtek Kowalczyk, and Martijn C. Schut. + Reinforcement learning for online control of evolutionary algorithms. + In International Workshop on Engineering Self-Organising Applications, pp.  151–160. Springer, 2006.
+[ bib ] + +
+ + +
+[1797] +
+
+Agoston E. Eiben and M. Jelasity. + A critical note on experimental research methodology in EC. + In Proceedings of the 2002 Congress on Evolutionary Computation (CEC'02), pp.  582–587, Piscataway, NJ, 2002. IEEE Press.
+[ bib | +DOI ] +
+Discusses reproducibility, generalizability and separation + between training (for tuning and experimentation) and testing + instances (for comparisons). +
+ +
+ + +
+[1798] +
+
+Agoston E. Eiben, Zbigniew Michalewicz, Marc Schoenauer, and James E. Smith. + Parameter Control in Evolutionary Algorithms. + In F. Lobo, C. F. Lima, and Z. Michalewicz, editors, Parameter Setting in Evolutionary Algorithms, pp.  19–46. Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[1799] +
+
+Agoston E. Eiben and James E. Smith. + Introduction to Evolutionary Computing. + Springer, 2003.
+[ bib ] + +
+ + +
+[1800] +
+
+Agoston E. Eiben and James E. Smith. + Introduction to Evolutionary Computing. + Natural Computing Series. Springer, 2nd edition, 2007.
+[ bib ] + +
+ + +
+[1801] +
+
+Mohammed El-Abd. + Opposition-based Artificial Bee Colony Algorithm. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  109–116. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[1802] +
+
+Nada Elsokkary, Faisal Shah Khan, Davide La Torre, Travis S. Humble, and Joel Gottlieb. + Financial Portfolio Management using D-Wave's Quantum Optimizer: The Case of Abu Dhabi Securities Exchange. + Technical report, Oak Ridge National Lab, Oak Ridge, TN, USA, 2017.
+[ bib | +http ] + +
+ + +
+[1803] +
+
+Michael T. M. Emmerich, André H. Deutz, and J. W. Klinkenberg. + Hypervolume-based expected improvement: Monotonicity properties and exact computation. + In Proceedings of the 2011 Congress on Evolutionary Computation (CEC 2011), pp.  2147–2154, Piscataway, NJ, 2011. IEEE Press.
+[ bib | +DOI ] +
+Proposed Expected Hypervolume Improvement (EHVI) +
+ +
+ + +
+[1804] +
+
+Michael T. M. Emmerich and Carlos M. Fonseca. + Computing Hypervolume Contributions in Low Dimensions: Asymptotically Optimal Algorithm and Complexity Results. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  121–135. Springer, Berlin/Heidelberg, 2011.
+[ bib | +DOI ] +
+Given a finite set Y ⊂Rd of n mutually + non-dominated vectors in d ≥2 dimensions, the + hypervolume contribution of a point y ∈Y is the + difference between the hypervolume indicator of Y + and the hypervolume indicator of Y ∖ {y}. In + multi-objective metaheuristics, hypervolume + contributions are computed in several selection and + bounded-size archiving procedures. This paper + presents new results on the (time) complexity of + computing all hypervolume contributions. It is + proved that for d = 2 and 3 the problem has time + complexity Θ(n logn), and, for d > 3, + the time complexity is bounded below by Ω(n + logn). Moreover, complexity bounds are derived for + computing a single hypervolume contribution. A + dimension sweep algorithm with time complexity + O (n logn) and space + complexity O(n) is + proposed for computing all hypervolume contributions + in three dimensions. It improves the complexity of + the best known algorithm for d = 3 by a factor of + √(n) . Theoretical results + are complemented by performance tests on randomly + generated test-problems. +
+ +
+ + +
+[1805] +
+
+Stefan Eppe, Yves De Smet, and Thomas Stützle. + A bi-objective optimization model to eliciting decision maker's preferences for the PROMETHEE II method. + In R. I. Brafman, F. Roberts, and A. Tsoukiàs, editors, Algorithmic Decision Theory, Third International Conference, ADT 2011, volume 6992 of Lecture Notes in Artificial Intelligence, pp.  56–66. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[1806] +
+
+Stefan Eppe, Manuel López-Ibáñez, Thomas Stützle, and Yves De Smet. + An Experimental Study of Preference Model Integration into Multi-Objective Optimization Heuristics. + In Proceedings of the 2011 Congress on Evolutionary Computation (CEC 2011), pp.  2751–2758, Piscataway, NJ, 2011. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1807] +
+
+David Eriksson, Michael Pearce, Jacob Gardner, Ryan D. Turner, and Matthias Poloczek. + Scalable Global Optimization via Local Bayesian Optimization. + In H. M. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. B. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems (NeurIPS 32), pp.  5496–5507, 2019.
+[ bib | +epub ] +
+Arxiv preprint arXiv: https://arxiv.org/abs/1910.01739 +
+ +
+ + +
+[1808] +
+
+Emre Ertin, Anthony N. Dean, Mathew L. Moore, and Kevin L. Priddy. + Dynamic Optimization for Optimal Control of Water Distribution Systems. + In K. L. Priddy, P. E. Keller, and P. J. Angeline, editors, Applications and Science of Computational Intelligence IV, Proceedings of SPIE, volume 4390, pp.  142–149, March 2001.
+[ bib ] + +
+ + +
+[1809] +
+
+V. Esat and M. Hall. + Water resources system optimization using genetic algorithms. + In A. Verwey, A. Minns, V. Babovic, and C. Maksimović, editors, Hydroinformatics'94, pp.  225–231, Balkema, Rotterdam, The Netherlands, 1994.
+[ bib ] + +
+ + +
+[1810] +
+
+Larry J. Eshelman and J. David Schaffer. + Real-Coded Genetic Algorithms and Interval-Schemata. + In D. Whitley, editor, Foundations of Genetic Algorithms (FOGA), pp.  187–202. Morgan Kaufmann Publishers, 1993.
+[ bib ] + +
+ + +
+[1811] +
+
+Larry J. Eshelman, A. Caruana, and J. David Schaffer. + Biases in the Crossover Landscape. + In J. D. Schaffer, editor, Proceedings of the Third International Conference on Genetic Algorithms (ICGA'89), pp.  86–91. Morgan Kaufmann Publishers, San Mateo, CA, 1989.
+[ bib ] + +
+ + +
+[1812] +
+
+Richard M. Everson, Jonathan E. Fieldsend, and Sameer Singh. + Full Elite Sets for Multi-objective Optimisation. + In Adaptive Computing in Design and Manufacture V, pp.  343–354. Springer, London, UK, 2002.
+[ bib | +DOI ] +
+Extended version published as [436] +
+ +
+ + +
+[1813] +
+
+C. J. Eyckelhof and M. Snoek. + Ant Systems for a Dynamic TSP: Ants Caught in a Traffic Jam. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  88–99. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1814] +
+
+Stefan Falkner, Marius Thomas Lindauer, and Frank Hutter. + SpySMAC: Automated configuration and performance analysis of SAT solvers. + In M. Heule and S. Weaver, editors, Theory and Applications of Satisfiability Testing – SAT 2015, volume 9340 of Lecture Notes in Computer Science, pp.  215–222. Springer, Cham, Switzerland, 2015.
+[ bib | +DOI ] + +
+ + +
+[1815] +
+
+Jesús Guillermo Falcón-Cardona, Saúl Zapotecas-Martínez, and Abel García-Nájera. + Pareto compliance from a practical point of view. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  395–402. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[1816] +
+
+M. Farina and P. Amato. + On the Optimal Solution Definition for Many-criteria Optimization Problems. + In Proceedings of the NAFIPS-FLINT International Conference'2002, pp.  233–238, Piscataway, New Jersey, June 2002. IEEE Service Center.
+[ bib | +DOI ] +
+First to mention exponential number of nondominated solutions + with respect to many objectives? +
+ +
+ + +
+[1817] +
+
+D. Favaretto, E. Moretti, and Paola Pellegrini. + On the explorative behavior of Max-Min Ant System. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2009, volume 5752 of Lecture Notes in Computer Science, pp.  115–119. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[1818] +
+
+Chris Fawcett, Malte Helmert, Holger H. Hoos, Erez Karpas, Gabriele Röger, and Jendrik Seipp. + FD-Autotune: Domain-Specific Configuration using Fast-Downward. + In E. Karpas, S. Jiménez Celorrio, and S. Kambhampati, editors, Proceedings of ICAPS-PAL11, 2011.
+[ bib ] + +
+ + +
+[1819] +
+
+Chris Fawcett and Holger H. Hoos. + Analysing Differences between Algorithm Configurations through Ablation. + In Proceedings of MIC 2013, the 10th Metaheuristics International Conference, pp.  123–132, 2013.
+[ bib | +epub ] + +
+ + +
+[1820] +
+
+Silvino Fernández, Segundo Álvarez, Diego Díaz, Miguel Iglesias, and Borja Ena. + Scheduling a Galvanizing Line by Ant Colony Optimization. + In M. Dorigo et al., editors, Swarm Intelligence, 9th International Conference, ANTS 2014, volume 8667 of Lecture Notes in Computer Science, pp.  146–157. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[1821] +
+
+Silvino Fernández, Segundo Álvarez, Eneko Malatsetxebarria, Pablo Valledor, and Diego Díaz. + Performance Comparison of Ant Colony Algorithms for the Scheduling of Steel Production Lines. + In J. L. Jiménez Laredo, S. Silva, and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2015. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[1822] +
+
+José C. Ferreira, Carlos M. Fonseca, and António Gaspar-Cunha. + Methodology to select solutions from the Pareto-optimal set: a comparative study. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  789–796. ACM Press, New York, NY, 2007.
+[ bib ] + +
+ + +
+[1823] +
+
+F. J. Ferri, P. Pudil, M. Hatef, and J. Kittler. + Comparative study of techniques for large-scale feature selection. + In E. S. Gelsema and L. S. Kanal, editors, Pattern Recognition in Practice IV, volume 16 of Machine Intelligence and Pattern Recognition, pp.  403–413. North-Holland, 1994.
+[ bib | +DOI ] +
+The combinatorial search problem arising in feature selection + in high dimensional spaces is considered. Recently developed + techniques based on the classical sequential methods and the + (l, r) search called Floating search algorithms are compared + against the Genetic approach to feature subset search. Both + approaches have been designed with the view to give a good + compromise between efficiency and effectiveness for large + problems. The purpose of this paper is to investigate the + applicability of these techniques to high dimensional + problems of feature selection. The aim is to establish + whether the properties inferred for these techniques from + medium scale experiments involving up to a few tens of + dimensions extend to dimensionalities of one order of + magnitude higher. Further, relative merits of these + techniques vis-a-vis such high dimensional problems are + explored and the possibility of exploiting the best aspects + of these methods to create a composite feature selection + procedure with superior properties is considered. +
+
+Describes sequential forward / backward selection +
+ +
+ + +
+[1824] +
+
+Silvino Fernández, Pablo Valledor, Diego Díaz, Eneko Malatsetxebarria, and Miguel Iglesias. + Criticality of Response Time in the usage of Metaheuristics in Industry. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2016, pp.  937–940. ACM Press, New York, NY, 2016.
+[ bib ] + +
+ + +
+[1825] +
+
+Matthias Feurer, Aaron Klein, Katharina Eggensperger, Jost Springenberg, Manuel Blum, and Frank Hutter. + Efficient and robust automated machine learning. + In C. Cortes, N. D. Lawrence, D. D. Lee, M. Sugiyama, and R. Garnett, editors, Advances in Neural Information Processing Systems (NIPS 28), pp.  2962–2970, 2015.
+[ bib | +epub | +http ] + +
+ + +
+[1826] +
+
+Matthias Feurer, Aaron Klein, Katharina Eggensperger, Jost Springenberg, Manuel Blum, and Frank Hutter. + Auto-sklearn: Efficient and Robust Automated Machine Learning. + In F. Hutter, L. Kotthoff, and J. Vanschoren, editors, Automated Machine Learning, pp.  113–134. Springer, 2019.
+[ bib | +DOI | +epub ] +
+The success of machine learning in a broad range of + applications has led to an ever-growing demand for machine + learning systems that can be used off the shelf by + non-experts. To be effective in practice, such systems need + to automatically choose a good algorithm and feature + preprocessing steps for a new dataset at hand, and also set + their respective hyperparameters. Recent work has started to + tackle this automated machine learning (AutoML) problem with + the help of efficient Bayesian optimization methods. Building + on this, we introduce a robust new AutoML system based on the + Python machine learning package scikit-learn (using 15 + classifiers, 14 feature preprocessing methods, and 4 data + preprocessing methods, giving rise to a structured hypothesis + space with 110 hyperparameters). This system, which we dub + Auto-sklearn, improves on existing AutoML methods by + automatically taking into account past performance on similar + datasets, and by constructing ensembles from the models + evaluated during the optimization. Our system won six out of + ten phases of the first ChaLearn AutoML challenge, and our + comprehensive analysis on over 100 diverse datasets shows + that it substantially outperforms the previous state of the + art in AutoML. We also demonstrate the performance gains due + to each of our contributions and derive insights into the + effectiveness of the individual components of Auto-sklearn. +
+ +
+ + +
+[1827] +
+
+Álvaro Fialho, Raymond Ros, Marc Schoenauer, and Michèle Sebag. + Comparison-based adaptive strategy selection with bandits in differential evolution. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  194–203. Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[1828] +
+
+Álvaro Fialho, Marc Schoenauer, and Michèle Sebag. + Fitness-AUC bandit adaptive strategy selection vs. the probability matching one within differential evolution: an empirical comparison on the BBOB-2010 noiseless testbed. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2010, pp.  1535–1542. ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[1829] +
+
+Álvaro Fialho, Marc Schoenauer, and Michèle Sebag. + Toward comparison-based adaptive operator selection. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  767–774. ACM Press, New York, NY, 2010.
+[ bib ] +
+Proposed F-AUC and F-SR +
+ +
+ + +
+[1830] +
+
+Álvaro Fialho. + Adaptive operator selection for optimization. + PhD thesis, Université Paris Sud-Paris XI, 2010.
+[ bib ] + +
+ + +
+[1831] +
+
+Jonathan E. Fieldsend. + University staff teaching allocation: formulating and optimising a many-objective problem. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  1097–1104. ACM Press, New York, NY, 2017.
+[ bib | +DOI ] +
+Example of NSGA-III deteriorating. +
+ +
+ + +
+[1832] +
+
+Jonathan E. Fieldsend and Richard M. Everson. + Visualising high-dimensional Pareto relationships in two-dimensional scatterplots. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  558–572. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[1833] +
+
+Jonathan E. Fieldsend. + Data structures for non-dominated sets: implementations and empirical assessment of two decades of advances. + In C. A. Coello Coello, editor, Proceedings of the 2020 Genetic and Evolutionary Computation Conference, pp.  489–497. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] +
+unbounded archives +
+ +
+ + +
+[1834] +
+
+Andreas Fink and Stefan Voß. + HotFrame: A Heuristic Optimization Framework. + In S. Voß and D. L. Woodruff, editors, Optimization Software Class Libraries, pp.  81–154. Kluwer Academic Publishers, Boston, MA, 2002.
+[ bib ] + +
+ + +
+[1835] +
+
+Benjamin Fisset, Clarisse Dhaenens, and Laetitia Jourdan. + MO-Mineclust: A Framework for Multi-Objective Clustering. + In C. Dhaenens, L. Jourdan, and M.-E. Marmion, editors, Learning and Intelligent Optimization, 9th International Conference, LION 9, volume 8994 of Lecture Notes in Computer Science, pp.  293–305. Springer, Heidelberg, Germany, 2015.
+[ bib ] +
+Keywords: irace +
+ +
+ + +
+[1836] +
+
+Peter J. Fleming, Robin C. Purshouse, and Robert J. Lygoe. + Many-objective optimization: An engineering design perspective. + In C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors, Evolutionary Multi-criterion Optimization, EMO 2005, volume 3410 of Lecture Notes in Computer Science, pp.  14–32. Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[1837] +
+
+Robin C. Purshouse, Cezar Jalbă, and Peter J. Fleming. + Preference-Driven Co-Evolutionary Algorithms Show Promise for Many-Objective optimisation. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  136–150. Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[1838] +
+
+M. Fleischer. + The Measure of Pareto Optima. Applications to Multi-objective Metaheuristics. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  519–533. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1839] +
+
+R. Fletcher. + Practical methods of optimization. + John Wiley & Sons, New York, NY, 1987.
+[ bib ] +
+BFGS +
+ +
+ + +
+[1840] +
+
+Dario Floreano and Francesco Mondada. + Automatic creation of an autonomous agent: Genetic evolution of a neural network driven robot. + In D. Cliff, P. Husbands, J.-A. Meyer, and S. Wilson, editors, Proceedings of the third international conference on Simulation of adaptive behavior: From Animals to Animats 3, pp.  421–430. MIT Press, Cambridge, MA, 1994.
+[ bib ] +
+LIS-CONF-1994-003 +
+ +
+ + +
+[1841] +
+
+Filippo Focacci, François Laburthe, and Andrea Lodi. + Local Search and Constraint Programming. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  369–403. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[1842] +
+
+David B. Fogel, Alvin J. Owens, and Michael J. Walsh. + Artificial Intelligence Through Simulated Evolution. + John Wiley & Sons, 1966.
+[ bib ] + +
+ + +
+[1843] +
+
+David B. Fogel. + Evolutionary Computation. Toward a New Philosophy of Machine Intelligence. + IEEE Press, 1995.
+[ bib ] + +
+ + +
+[1844] +
+
+Carlos M. Fonseca and Peter J. Fleming. + Genetic Algorithms for Multiobjective Optimization: Formulation, Discussion and Generalization. + In S. Forrest, editor, Proceedings of the Fifth International Conference on Genetic Algorithms (ICGA'93), pp.  416–423. Morgan Kaufmann Publishers, 1993.
+[ bib | +epub ] +
+Proposes MOGA and P-MOGA +
+ +
+ + +
+[1845] +
+
+Carlos M. Fonseca and Peter J. Fleming. + On the Performance Assessment and Comparison of Stochastic Multiobjective Optimizers. + In H.-M. Voigt et al., editors, Parallel Problem Solving from Nature – PPSN IV, volume 1141 of Lecture Notes in Computer Science, pp.  584–593. Springer, Heidelberg, Germany, 1996.
+[ bib ] + +
+ + +
+[1846] +
+
+Viviane Grunert da Fonseca and Carlos M. Fonseca. + The Relationship between the Covered Fraction, Completeness and Hypervolume Indicators. + In J.-K. Hao, P. Legrand, P. Collet, N. Monmarché, E. Lutton, and M. Schoenauer, editors, Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011, volume 7401 of Lecture Notes in Computer Science, pp.  25–36. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[1847] +
+
+Carlos M. Fonseca, Viviane Grunert da Fonseca, and Luís Paquete. + Exploring the Performance of Stochastic Multiobjective Optimisers with the Second-Order Attainment Function. + In C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors, Evolutionary Multi-criterion Optimization, EMO 2005, volume 3410 of Lecture Notes in Computer Science, pp.  250–264. Springer, Heidelberg, Germany, 2005.
+[ bib | +DOI ] +
+The attainment function has been proposed as a + measure of the statistical performance of stochastic + multiobjective optimisers which encompasses both the + quality of individual non-dominated solutions in + objective space and their spread along the trade-off + surface. It has also been related to results from + random closed-set theory, and cast as a mean-like, + first-order moment measure of the outcomes of + multiobjective optimisers. In this work, the use of + more informative, second-order moment measures for + the evaluation and comparison of multiobjective + optimiser performance is explored experimentally, + with emphasis on the interpretability of the + results. +
+ +
+ + +
+[1848] +
+
+Carlos M. Fonseca, Andreia P. Guerreiro, Manuel López-Ibáñez, and Luís Paquete. + On the Computation of the Empirical Attainment Function. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  106–120. Springer, Berlin/Heidelberg, 2011.
+[ bib | +DOI ] +
+The attainment function provides a description of the + location of the distribution of a random non-dominated point + set. This function can be estimated from experimental data + via its empirical counterpart, the empirical attainment + function (EAF). However, computation of the EAF in more than + two dimensions is a non-trivial task. In this article, the + problem of computing the empirical attainment function is + formalised, and upper and lower bounds on the corresponding + number of output points are presented. In addition, efficient + algorithms for the two and three-dimensional cases are + proposed, and their time complexities are related to lower + bounds derived for each case. +
+ +
+ + +
+[1849] +
+
+Carlos M. Fonseca, Luís Paquete, and Manuel López-Ibáñez. + An improved dimension- sweep algorithm for the hypervolume indicator. + In Proceedings of the 2006 Congress on Evolutionary Computation (CEC 2006), pp.  1157–1163, Piscataway, NJ, July 2006. IEEE Press.
+[ bib | +DOI ] +
+This paper presents a recursive, dimension-sweep + algorithm for computing the hypervolume indicator of + the quality of a set of n non-dominated points in + d>2 dimensions. It improves upon the existing HSO + (Hypervolume by Slicing Objectives) algorithm by + pruning the recursion tree to avoid repeated + dominance checks and the recalculation of partial + hypervolumes. Additionally, it incorporates a recent + result for the three-dimensional special case. The + proposed algorithm achieves O(nd-2 log n) time + and linear space complexity in the worst-case, but + experimental results show that the pruning + techniques used may reduce the time complexity + exponent even further. +
+ +
+ + +
+[1850] +
+
+Jorge Ramón Fonseca Cacho and Kazem Taghva. + The State of Reproducible Research in Computer Science. + In S. Latifi, editor, 17th International Conference on Information Technology-New Generations (ITNG 2020), Advances in Intelligent Systems and Computing, pp.  519–524. Springer International Publishing, 2020.
+[ bib | +DOI ] +
+Reproducible research is the cornerstone of cumulative + science and yet is one of the most serious crisis that we + face today in all fields. This paper aims to describe the + ongoing reproducible research crisis along with + counter-arguments of whether it really is a crisis, suggest + solutions to problems limiting reproducible research along + with the tools to implement such solutions by covering the + latest publications involving reproducible research. +
+
+Keywords: Docker, Improving transparency, OCR, Open science, + Replicability, Reproducibility +
+ +
+ + +
+[1851] +
+
+Manuel Förster, Bettina Bickel, Bernd Hardung, and Gabriella Kókai. + Self-adaptive ant colony optimisation applied to function allocation in vehicle networks. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  1991–1998. ACM Press, New York, NY, 2007.
+[ bib ] + +
+ + +
+[1852] +
+
+Michael Foster, Matthew Hughes, George O'Brien, Pietro S. Oliveto, James Pyle, Dirk Sudholt, and James Williams. + Do sophisticated evolutionary algorithms perform better than simple ones? + In C. A. Coello Coello, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020, pp.  184–192. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[1853] +
+
+Robert Fourer, David M. Gay, and Brian W. Kernighan. + AMPL: A Modeling Language for Mathematical Programming. + Duxbury, 2nd edition, 2002.
+[ bib ] + +
+ + +
+[1854] +
+
+Bennett L. Fox. + Uniting probabilistic methods for optimization. + In Proceedings of the 24th conference on Winter simulation, pp.  500–505. ACM, 1992.
+[ bib ] + +
+ + +
+[1855] +
+
+Bennett L. Fox. + Simulated annealing: folklore, facts, and directions. + In Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing, pp.  17–48. Springer, 1995.
+[ bib ] + +
+ + +
+[1856] +
+
+Alberto Franzin. + Empirical Analysis of Stochastic Local Search Behaviour: Connecting Structure, Components and Landscape. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2021.
+[ bib ] + +
+ + +
+[1857] +
+
+C. B. Fraser. + Subsequences and Supersequences of Strings. + PhD thesis, University of Glasgow, 1995.
+[ bib ] + +
+ + +
+[1858] +
+
+Alberto Franzin, Raphaël Gyory, Jean-Charles Nadé, Guillaume Aubert, Georges Klenkle, and Hughes Bersini. + Philéas: Anomaly Detection for IoT Monitoring. + In L. Cao, W. Kosters, and J. Lijffijt, editors, Proceedings of the 32nd Benelux Conference on Artificial Intelligence, BNAIC 2020, Leiden, The Netherlands, 19-20 November 2020, pp.  56–70, 2020.
+[ bib | +http ] + +
+ + +
+[1859] +
+
+Jose M. Framiñán, Rainer Leisten, and Rubén Ruiz. + Manufacturing Scheduling Systems: An Integrated View on Models, Methods, and Tools. + Springer, New York, NY, 2014.
+[ bib ] + +
+ + +
+[1860] +
+
+Alberto Franzin and Thomas Stützle. + Exploration of Metaheuristics through Automatic Algorithm Configuration Techniques and Algorithmic Frameworks. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2016, pp.  1341–1347. ACM Press, New York, NY, 2016.
+[ bib ] + +
+ + +
+[1861] +
+
+Alberto Franzin and Thomas Stützle. + Comparison of Acceptance Criteria in Randomized Local Searches. + In E. Lutton, P. Legrand, P. Parrend, N. Monmarché, and M. Schoenauer, editors, EA 2017: Artificial Evolution, volume 10764 of Lecture Notes in Computer Science, pp.  16–29. Springer, Heidelberg, Germany, 2017.
+[ bib ] + +
+ + +
+[1862] +
+
+Alberto Franzin and Thomas Stützle. + Revisiting Simulated Annealing: a Component-Based Analysis: Supplementaty Material. + http://iridia.ulb.ac.be/supp/IridiaSupp2018-001, 2018.
+[ bib ] + +
+ + +
+[1863] +
+
+Alberto Franzin and Thomas Stützle. + Towards transferring algorithm configurations across problems. + In M. Vlastelica, J. Song, A. Ferber, B. Amos, G. Martius, B. Dilkina, and Y. Yue, editors, Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020, LMCA 2020, Vancouver, Canada, December 12, 2020, pp.  1–6, 2020.
+[ bib ] + +
+ + +
+[1864] +
+
+Alberto Franzin and Thomas Stützle. + A causal framework for understanding optimisation algorithms. + In F. Heintz, M. Milano, and B. O'Sullivan, editors, Trustworthy AI – Integrating Learning, Optimization and Reasoning. TAILOR 2020, volume 12641 of Lecture Notes in Computer Science, pp.  140–145. Springer, Cham, Switzerland, 2021.
+[ bib ] + +
+ + +
+[1865] +
+
+Alberto Franzin and Thomas Stützle. + A Landscape-based Analysis of Fixed Temperature and Simulated Annealing: Supplementaty Material. + http://iridia.ulb.ac.be/supp/IridiaSupp2021-002, 2021.
+[ bib ] + +
+ + +
+[1866] +
+
+A. R. R. Freitas, Peter J. Fleming, and Frederico G. Guimarães. + A Non-parametric Harmony-Based Objective Reduction Method for Many-Objective Optimization. + In 2013 IEEE International Conference on Systems, Man, and Cybernetics, pp.  651–656. IEEE Press, 2013.
+[ bib | +DOI ] + +
+ + +
+[1867] +
+
+B. Freisleben and P. Merz. + A Genetic Local Search Algorithm for Solving Symmetric and Asymmetric Traveling Salesman Problems. + In T. Bäck, T. Fukuda, and Z. Michalewicz, editors, Proceedings of the 1996 IEEE International Conference on Evolutionary Computation (ICEC'96), pp.  616–621. IEEE Press, Piscataway, NJ, 1996.
+[ bib ] + +
+ + +
+[1868] +
+
+Tobias Friedrich, Andreas Göbel, Francesco Quinzan, and Markus Wagner. + Heavy-Tailed Mutation Operators in Single-Objective Combinatorial Optimization. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11101 of Lecture Notes in Computer Science, pp.  134–145. Springer, Cham, Switzerland, 2018.
+[ bib ] +
+A core feature of evolutionary algorithms is their mutation + operator. Recently, much attention has been devoted to the + study of mutation operators with dynamic and non-uniform + mutation rates. Following up on this line of work, we propose + a new mutation operator and analyze its performance on the + (1+1) Evolutionary Algorithm (EA). Our analyses show that + this mutation operator competes with pre-existing ones, when + used by the (1+1)-EA on classes of problems for which + results on the other mutation operators are available. We + present a “jump” function for which the performance of the + (1+1)-EA using any static uniform mutation and any restart + strategy can be worse than the performance of the (1+1)-EA + using our mutation operator with no restarts. We show that + the (1+1)-EA using our mutation operator finds a + (1/3)-approximation ratio on any non-negative submodular + function in polynomial time. This performance matches that of + combinatorial local search algorithms specifically designed + to solve this problem. +
+ +
+ + +
+[1869] +
+
+Tobias Friedrich, Timo Kötzing, Martin S. Krejca, and Andrew M. Sutton. + Robustness of Ant Colony Optimization to Noise. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  17–24. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] +
+Keywords: ant colony optimization, noisy fitness, run time analysis, + theory +
+ +
+ + +
+[1870] +
+
+Tobias Friedrich, Timo Kötzing, and Markus Wagner. + A Generic Bet-and-Run Strategy for Speeding Up Stochastic Local Search. + In S. P. Singh and S. Markovitch, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  801–807. AAAI Press, February 2017.
+[ bib ] + +
+ + +
+[1871] +
+
+Tobias Friedrich, Francesco Quinzan, and Markus Wagner. + Escaping Large Deceptive Basins of Attraction with Heavy-tailed Mutation Operators. + In H. E. Aguirre and K. Takadama, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2018, pp.  293–300. ACM Press, New York, NY, 2018.
+[ bib | +DOI ] +
+Keywords: combinatorial optimization, heavy-tailed mutation, + single-objective optimization, experiments-motivated theory, + irace +
+ +
+ + +
+[1872] +
+
+Michael Friendly. + Statistical graphics for multivariate data. + In SAS Conference Proceedings: SAS Users Group International 16 (SUGI 16), 1991.
+[ bib ] +
+February 17-20, 1991, New Orleans, Louisiana, 297 papers +
+ +
+ + +
+[1873] +
+
+D. Fudenberg and J. Tirole. + Game Theory. + MIT Press, Cambridge, MA, 1983.
+[ bib ] + +
+ + +
+[1874] +
+
+Noriyuki Fujimoto and Kouki Nanai. + Solving QUBO with GPU parallel MOPSO. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1788–1794. ACM Press, New York, NY, 2021.
+[ bib ] + +
+ + +
+[1875] +
+
+Alex S. Fukunaga. + Evolving Local Search Heuristics for SAT Using Genetic Programming. + In K. Deb et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2004, Part II, volume 3103 of Lecture Notes in Computer Science, pp.  483–494. Springer, Heidelberg, Germany, 2004.
+[ bib ] +
+Satisfiability testing (SAT) is a very active area + of research today, with numerous real-world + applications. We describe CLASS2.0, a genetic + programming system for semi-automatically designing + SAT local search heuristics. An empirical + comparison shows that that the heuristics generated + by our GP system outperform the state of the art + human-designed local search algorithms, as well as + previously proposed evolutionary approaches, with + respect to both runtime as well as search efficiency + (number of variable flips to solve a problem). +
+ +
+ + +
+[1876] +
+
+Nancy E. Furlong, Eugene A. Lovelace, and Kristin L. Lovelace. + Research Methods and Statistics: An Integrated Approach. + Harcourt College Publishers, 2000.
+[ bib ] + +
+ + +
+[1877] +
+
+D. Gaertner and K. Clark. + On Optimal Parameters for Ant Colony Optimization Algorithms. + In H. R. Arabnia and R. Joshua, editors, Proceedings of the 2005 International Conference on Artificial Intelligence, ICAI 2005, pp.  83–89. CSREA Press, 2005.
+[ bib ] + +
+ + +
+[1878] +
+
+Matteo Gagliolo and Catherine Legrand. + Algorithm Survival Analysis. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  161–184. Springer, Berlin, Germany, 2010.
+[ bib | +DOI ] +
+Algorithm selection is typically based on models of + algorithm performance,learned during a separate + offline training sequence, which can be + prohibitively expensive. In recent work, we adopted + an online approach, in which models of the runtime + distributions of the available algorithms are + iteratively updated and used to guide the allocation + of computational resources, while solving a sequence + of problem instances. The models are estimated using + survival analysis techniques, which allow us to + reduce computation time, censoring the runtimes of + the slower algorithms. Here, we review the + statistical aspects of our online selection method, + discussing the bias induced in the runtime + distributions (RTD) models by the competition of + different algorithms on the same problem instances. +
+ +
+ + +
+[1879] +
+
+L. M. Gambardella and Marco Dorigo. + Ant-Q: A Reinforcement Learning Approach to the Traveling Salesman Problem. + In A. Prieditis and S. Russell, editors, Proceedings of the Twelfth International Conference on Machine Learning (ML-95), pp.  252–260. Morgan Kaufmann Publishers, Palo Alto, CA, 1995.
+[ bib ] + +
+ + +
+[1880] +
+
+L. M. Gambardella and Marco Dorigo. + Solving Symmetric and Asymmetric TSPs by Ant Colonies. + In T. Bäck, T. Fukuda, and Z. Michalewicz, editors, Proceedings of the 1996 IEEE International Conference on Evolutionary Computation (ICEC'96), pp.  622–627. IEEE Press, Piscataway, NJ, 1996.
+[ bib ] + +
+ + +
+[1881] +
+
+L. M. Gambardella, Éric D. Taillard, and G. Agazzi. + MACS-VRPTW: A Multiple Ant Colony System for Vehicle Routing Problems with Time Windows. + In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in Optimization, pp.  63–76. McGraw Hill, London, UK, 1999.
+[ bib ] + +
+ + +
+[1882] +
+
+Xavier Gandibleux, X. Delorme, and V. T'Kindt. + An Ant Colony Optimisation Algorithm for the Set Packing Problem. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  49–60. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[1883] +
+
+Xavier Gandibleux, N. Mezdaoui, and A. Fréville. + A tabu search procedure to solve multiobjective combinatorial optimization problem. + In R. Caballero, F. Ruiz, and R. Steuer, editors, Advances in Multiple Objective and Goal Programming, volume 455 of Lecture Notes in Economics and Mathematical Systems, pp.  291–300. Springer, Heidelberg, Germany, 1997.
+[ bib ] + +
+ + +
+[1884] +
+
+Xavier Gandibleux, H. Morita, and N. Katoh. + Use of a genetic heritage for solving the assignment problem with two objectives. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  43–57. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1885] +
+
+Huiru Gao, Haifeng Nie, and Ke Li. + Visualisation of Pareto Front Approximation: A Short Survey and Empirical Comparisons. + In Proceedings of the 2019 Congress on Evolutionary Computation (CEC 2019), pp.  1750–1757, Piscataway, NJ, 2019. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1886] +
+
+Deon Garrett and Dipankar Dasgupta. + Multiobjective landscape analysis and the generalized assignment problem. + In V. Maniezzo, R. Battiti, and J.-P. Watson, editors, Learning and Intelligent Optimization, Second International Conference, LION 2, volume 5313 of Lecture Notes in Computer Science, pp.  110–124. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[1887] +
+
+M. R. Garey and David S. Johnson. + Computers and Intractability: A Guide to the Theory of NP-Completeness. + Freeman & Co, San Francisco, CA, 1979.
+[ bib ] + +
+ + +
+[1888] +
+
+José García-Nieto, Esteban López-Camacho, María Jesús Godoy García, Antonio J. Nebro, Juan J. Durillo, and José F. Aldana-Montes. + A study of archiving strategies in multi-objective PSO for molecular docking. + In M. Dorigo, M. Birattari, X. Li, M. López-Ibáñez, K. Ohkura, C. Pinciroli, and T. Stützle, editors, Swarm Intelligence, 10th International Conference, ANTS 2016, volume 9882 of Lecture Notes in Computer Science, pp.  40–52. Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[1889] +
+
+Beatriz A. Garro, Humberto Sossa, and Roberto A. Vazquez. + Evolving ant colony system for optimizing path planning in mobile robots. + In Electronics, Robotics and Automotive Mechanics Conference, pp.  444–449, Los Alamitos, CA, 2007. IEEE Computer Society.
+[ bib | +DOI ] + +
+ + +
+[1890] +
+
+Luca Di Gaspero and Andrea Schaerf. + Easysyn++: A tool for automatic synthesis of stochastic local search algorithms. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2007, volume 4638 of Lecture Notes in Computer Science, pp.  177–181. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[1891] +
+
+Martin Gebser, Roland Kaminski, Benjamin Kaufmann, Torsten Schaub, Marius Thomas Schneider, and Stefan Ziller. + A portfolio solver for answer set programming: Preliminary report. + In P. Calabar and T. C. Son, editors, Logic Programming and Nonmonotonic Reasoning, volume 8148 of Lecture Notes in Artificial Intelligence, pp.  352–357. Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[1892] +
+
+Peter Geibel. + Reinforcement Learning for MDPs with Constraints. + In J. Fürnkranz, T. Scheffer, and M. Spiliopoulou, editors, Machine Learning: ECML 2006, volume 4212 of Lecture Notes in Computer Science, pp.  646–653, 2006.
+[ bib | +DOI ] +
+In this article, I will consider Markov Decision Processes + with two criteria, each defined as the expected value of an + infinite horizon cumulative return. The second criterion is + either itself subject to an inequality constraint, or there + is maximum allowable probability that the single returns + violate the constraint. I describe and discuss three new + reinforcement learning approaches for solving such control + problems. +
+
+Keywords: Safe RL +
+ +
+ + +
+[1893] +
+
+Ian P. Gent, Stuart A. Grant, Ewen MacIntyre, Patrick Prosser, Paul Shaw, Barbara M. Smith, and Toby Walsh. + How Not To Do It. + Technical Report 97.27, School of Computer Studies, University of Leeds, May 1997.
+[ bib ] +
+We give some dos and don'ts for those analysing algorithms + experimentally. We illustrate these with many examples from + our own research on the study of algorithms for NP-complete + problems such as satisfiability and constraint + satisfaction. Where we have not followed these maxims, we + have suffered as a result. +
+ +
+ + +
+[1894] +
+
+Ian P. Gent, Holger H. Hoos, P. Prosser, and T. Walsh. + Morphing: Combining Structure and Randomness. + In Proceedings of the Sixteenth National Conference on Artificial Intelligence, pp.  654–660, 1999.
+[ bib ] + +
+ + +
+[1895] +
+
+Michel Gendreau and Jean-Yves Potvin. + Tabu Search. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, pp.  41–59. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[1896] +
+
+Daniel Geschwender, Frank Hutter, Lars Kotthoff, Yuri Malitsky, Holger H. Hoos, and Kevin Leyton-Brown. + Algorithm Configuration in the Cloud: A Feasibility Study. + In P. M. Pardalos, M. G. C. Resende, C. Vogiatzis, and J. L. Walteros, editors, Learning and Intelligent Optimization, 8th International Conference, LION 8, volume 8426 of Lecture Notes in Computer Science, pp.  41–46. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[1897] +
+
+Matthew S. Gibbs, Graeme C. Dandy, Holger R. Maier, and John B. Nixon. + Calibrating genetic algorithms for water distribution system optimisation. + In 7th Annual Symposium on Water Distribution Systems Analysis. ASCE, May 2005.
+[ bib ] + +
+ + +
+[1898] +
+
+T. Glasmachers. + A fast incremental BSP tree archive for non-dominated points. + In H. Trautmann, G. Rudolph, K. Klamroth, O. Schütze, M. M. Wiecek, Y. Jin, and C. Grimme, editors, Evolutionary Multi-criterion Optimization, EMO 2017, volume 10173 of Lecture Notes in Computer Science, pp.  252–266. Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] +
+Keywords: archiving +
+ +
+ + +
+[1899] +
+
+Fred Glover. + A Template for Scatter Search and Path Relinking. + In J.-K. Hao, E. Lutton, E. M. A. Ronald, M. Schoenauer, and D. Snyers, editors, Artificial Evolution, volume 1363 of Lecture Notes in Computer Science, pp.  1–51. Springer, Heidelberg, Germany, 1998.
+[ bib | +DOI ] + +
+ + +
+[1900] +
+
+Xavier Glorot and Yoshua Bengio. + Understanding the difficulty of training deep feedforward neural networks. + In Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp.  249–256, 2010.
+[ bib ] + +
+ + +
+[1901] +
+
+Fred Glover and Gary A. Kochenberger. + Critical Even Tabu Search for Multidimensional Knapsack Problems. + In I. H. Osman and J. P. Kelly, editors, Metaheuristics: Theory & Applications, pp.  407–427. Kluwer Academic Publishers, Norwell, MA, 1996.
+[ bib ] + +
+ + +
+[1902] +
+
+Fred Glover and Manuel Laguna. + Tabu Search. + Kluwer Academic Publishers, Boston, MA, USA, 1997.
+[ bib ] + +
+ + +
+[1903] +
+
+Fred Glover, Manuel Laguna, and Rafael Martí. + Scatter Search and Path Relinking: Advances and Applications. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  1–35. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[1904] +
+
+Daniel Golovin, Benjamin Solnik, Subhodeep Moitra, Greg Kochanski, John Karro, and D. Sculley. + Google Vizier: A Service for Black-Box Optimization. + In S. Matwin, S. Yu, and F. Farooq, editors, 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.  1487–1495. ACM Press, 2017.
+[ bib | +DOI ] + +
+ + +
+[1905] +
+
+Elizabeth Ferreira Gouvêa Goldbarg, Givanaldo R. Souza, and Marco Cesar Goldbarg. + Particle Swarm for the Traveling Salesman Problem. + In J. Gottlieb and G. R. Raidl, editors, Proceedings of EvoCOP 2006 – 6th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3906 of Lecture Notes in Computer Science, pp.  99–110. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[1906] +
+
+David E. Goldberg. + Genetic Algorithms in Search, Optimization and Machine Learning. + Addison-Wesley, Boston, MA, 1989.
+[ bib ] + +
+ + +
+[1907] +
+
+Fred E. Goldman and Larry W. Mays. + The Application of Simulated Annealing to the Optimal Operation of Water Systems. + In Proceedings of 26th Annual Water Resources Planning and Management Conference, Tempe, USA, June 2000. ASCE.
+[ bib ] + +
+ + +
+[1908] +
+
+Ralph E. Gomory. + An algorithm for integer solutions to linear programs. + In R. Graves and P. Wolfe, editors, Recent Advances in Mathematical Programming, pp.  260–302. McGraw Hill, New York, NY, 1963.
+[ bib ] + +
+ + +
+[1909] +
+
+Wenyin Gong, Álvaro Fialho, and Zhihua Cai. + Adaptive strategy selection in differential evolution. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  409–416. ACM Press, New York, NY, 2010.
+[ bib | +DOI ] + +
+ + +
+[1910] +
+
+M. Gorges-Schleuter. + Asparagos96 and the Travelling Salesman Problem. + In T. Bäck, Z. Michalewicz, and X. Yao, editors, Proceedings of the 1997 IEEE International Conference on Evolutionary Computation (ICEC'97), pp.  171–174. IEEE Press, Piscataway, NJ, 1997.
+[ bib ] + +
+ + +
+[1911] +
+
+J. Gottlieb, M. Puchta, and Christine Solnon. + A Study of Greedy, Local Search, and Ant Colony Optimization Approaches for Car Sequencing Problems. + In S. Cagnoni et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2003, volume 2611 of Lecture Notes in Computer Science, pp.  246–257. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1912] +
+
+Jonathan Gratch and Gerald DeJong. + COMPOSER: A probabilistic solution to the utility problem in speed-up learning. + In W. R. Swartout, editor, Proceedings of the 10th National Conference on Artificial Intelligence, pp.  235–240. AAAI Press/MIT Press, Menlo Park, CA, 1992.
+[ bib ] +
+Eearliest hyper-heuristic? +
+ +
+ + +
+[1913] +
+
+Alex Graves, Abdel-rahman Mohamed, and Geoffrey Hinton. + Speech recognition with deep recurrent neural networks. + In Acoustics, speech and signal processing (icassp), 2013 ieee international conference on, pp.  6645–6649. IEEE, 2013.
+[ bib ] + +
+ + +
+[1914] +
+
+Garrison W. Greenwood, Xiaobo Hu, and Joseph G. D'Ambrosio. + Fitness functions for multiple objective optimization problems: Combining preferences with Pareto rankings. + In R. K. Belew and M. D. Vose, editors, Foundations of Genetic Algorithms (FOGA), pp.  437–455. Morgan Kaufmann Publishers, 1996.
+[ bib ] + +
+ + +
+[1915] +
+
+Salvatore Greco, Benedetto Matarazzo, and Roman Slowiński. + Interactive evolutionary multiobjective optimization using dominance-based rough set approach. + In H. Ishibuchi et al., editors, Proceedings of the 2010 Congress on Evolutionary Computation (CEC 2010), pp.  1–8, Piscataway, NJ, 2010. IEEE Press.
+[ bib ] + +
+ + +
+[1916] +
+
+Viviane Grunert da Fonseca and Carlos M. Fonseca. + A characterization of the outcomes of stochastic multiobjective optimizers through a reduction of the hitting function test sets. + Technical report, CSI, Universidade do Algarve, 2004.
+[ bib ] +
+Keywords: high-order EAF +
+ +
+ + +
+[1917] +
+
+Viviane Grunert da Fonseca and Carlos M. Fonseca. + The Attainment-Function Approach to Stochastic Multiobjective Optimizer Assessment and Comparison. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  103–130. Springer, Berlin, Germany, 2010.
+[ bib ] + +
+ + +
+[1918] +
+
+Viviane Grunert da Fonseca, Carlos M. Fonseca, and Andreia O. Hall. + Inferential Performance Assessment of Stochastic Optimisers and the Attainment Function. + In E. Zitzler, K. Deb, L. Thiele, C. A. Coello Coello, and D. Corne, editors, Evolutionary Multi-criterion Optimization, EMO 2001, volume 1993 of Lecture Notes in Computer Science, pp.  213–225. Springer, Berlin/Heidelberg, 2001.
+[ bib | +DOI ] +
+The performance of stochastic optimisers can be assessed + experimentally on given problems by performing multiple + optimisation runs, and analysing the results. Since an + optimiser may be viewed as an estimator for the (Pareto) + minimum of a (vector) function, stochastic optimiser + performance is discussed in the light of the criteria + applicable to more usual statistical + estimators. Multiobjective optimisers are shown to deviate + considerably from standard point estimators, and to require + special statistical methodology. The attainment function is + formulated, and related results from random closed-set theory + are presented, which cast the attainment function as a + mean-like measure for the outcomes of multiobjective + optimisers. Finally, a covariance-measure is defined, which + should bring additional insight into the stochastic behaviour + of multiobjective optimisers. Computational issues and + directions for further work are discussed at the end of the + paper. +
+
+Proposed looking at anytime behavior as a multi-objective + problem +
+
+Keywords: EAF +
+ +
+ + +
+[1919] +
+
+C. Guéret, Nicolas Monmarché, and M. Slimane. + Ants Can Play Music. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  310–317. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[1920] +
+
+M. Guntsch and Jürgen Branke. + New Ideas for Applying Ant Colony Optimization to the Probabilistic TSP. + In S. Cagnoni et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2003, volume 2611 of Lecture Notes in Computer Science, pp.  165–175. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1921] +
+
+M. Guntsch and Martin Middendorf. + Pheromone Modification Strategies for Ant Algorithms Applied to Dynamic TSP. + In E. J. W. Boers et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2001, volume 2037 of Lecture Notes in Computer Science, pp.  213–222. Springer, Heidelberg, Germany, 2001.
+[ bib ] + +
+ + +
+[1922] +
+
+M. Guntsch and Martin Middendorf. + A Population Based Approach for ACO. + In S. Cagnoni et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2002, volume 2279 of Lecture Notes in Computer Science, pp.  71–80. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1923] +
+
+M. Guntsch and Martin Middendorf. + Solving Multi-Objective Permutation Problems with Population Based ACO. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  464–478. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[1924] +
+
+M. Guntsch and Martin Middendorf. + Applying Population Based ACO to Dynamic Optimization Problems. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  111–122. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[1925] +
+
+Gurobi. + Gurobi Optimizer. + http://www.gurobi.com/products/gurobi-optimizer, 2017.
+[ bib ] + +
+ + +
+[1926] +
+
+D. Gusfield. + Algorithms on Strings, Trees, and Sequences. + In Computer Science and Computational Biology. Cambridge University Press, 1997.
+[ bib ] + +
+ + +
+[1927] +
+
+Walter J. Gutjahr. + S-ACO: An Ant-Based Approach to Combinatorial Optimization Under Uncertainty. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  238–249. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[1928] +
+
+Walter J. Gutjahr. + A converging ACO algorithm for stochastic combinatorial optimization. + In A. Albrecht and K. Steinhöfel, editors, Stochastic Algorithms: Foundations and Applications, volume 2827 of Lecture Notes in Computer Science, pp.  10–25. Springer Verlag, 2003.
+[ bib | +DOI ] + +
+ + +
+[1929] +
+
+Evert Haasdijk, Arif Atta-ul Qayyum, and Agoston E. Eiben. + Racing to improve on-line, on-board evolutionary robotics. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  187–194. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[1930] +
+
+S. Häckel, M. Fischer, D. Zechel, and T. Teich. + A multi-objective ant colony approach for Pareto-optimization using dynamic programming. + In C. Ryan, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2008, pp.  33–40. ACM Press, New York, NY, 2008.
+[ bib ] + +
+ + +
+[1931] +
+
+David Hadka, Patrick M. Reed, and T. W. Simpson. + Diagnostic assessment of the Borg MOEA for many-objective product family design problems. + In Proceedings of the 2012 Congress on Evolutionary Computation (CEC 2012), pp.  1–10, Piscataway, NJ, 2012. IEEE Press.
+[ bib ] + +
+ + +
+[1932] +
+
+Apache Software Foundation. + Hadoop, 2008.
+[ bib | +http ] + +
+ + +
+[1933] +
+
+Thomas M. Walski, Donald V. Chase, Dragan A. Savic, Walter Grayman, Stephen Beckwith, and Edmundo Koelle. + Advanced Water Distribution Modeling and Management. + Haestad Methods, Inc., Haestad Press, 1st edition, 2003.
+[ bib ] + +
+ + +
+[1934] +
+
+George T. Hall, Pietro S. Oliveto, and Dirk Sudholt. + On the impact of the cutoff time on the performance of algorithm configurators. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  907–915. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+Keywords: theory, automatic configuration, capping +
+ +
+ + +
+[1935] +
+
+George T. Hall, Pietro S. Oliveto, and Dirk Sudholt. + Fast Perturbative Algorithm Configurators. + In T. Bäck, M. Preuss, A. Deutz, H. Wang, C. Doerr, M. T. M. Emmerich, and H. Trautmann, editors, Parallel Problem Solving from Nature – PPSN XVI, volume 12269 of Lecture Notes in Computer Science, pp.  19–32. Springer, Cham, Switzerland, 2020.
+[ bib | +DOI ] + +
+ + +
+[1936] +
+
+George T. Hall, Pietro S. Oliveto, and Dirk Sudholt. + Analysis of the performance of algorithm configurators for search heuristics with global mutation operators. + In C. A. Coello Coello, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020, pp.  823–831. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[1937] +
+
+Greg Hamerly and Charles Elkan. + Learning the k in k-means. + In S. Thrun, L. Saul, and B. Schölkopf, editors, Advances in Neural Information Processing Systems (NIPS 16). MIT Press, 2003.
+[ bib | +epub ] + +
+ + +
+[1938] +
+
+Hayfa Hammami and Thomas Stützle. + A Computational Study of Neighborhood Operators for Job-Shop Scheduling Problems with Regular Objectives. + In B. Hu and M. López-Ibáñez, editors, Proceedings of EvoCOP 2017 – 17th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 10197 of Lecture Notes in Computer Science, pp.  1–17. Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[1939] +
+
+Michael Pilegaard Hansen. + Tabu search for multiobjective optimization: MOTS. + In J. Climaco, editor, Proceedings of the 13th International Conference on Multiple Criteria Decision Making (MCDM'97), pp.  574–586. Springer Verlag, 1997.
+[ bib ] + +
+ + +
+[1940] +
+
+Nikolaus Hansen, Youhei Akimoto, and Petr Baudis. + CMA-ES/pycma on Github. + Zenodo, February 2019.
+[ bib | +DOI ] + +
+ + +
+[1941] +
+
+Nikolaus Hansen, Anne Auger, S. Finck, and R. Ros. + Real-Parameter Black-Box Optimization Benchmarking 2009: Experimental setup. + Technical Report RR-6828, INRIA, France, 2009.
+[ bib | +supplementary material ] + +
+ + +
+[1942] +
+
+Nikolaus Hansen, Anne Auger, Raymond Ros, Steffen Finck, and Petr Pošík. + Comparing Results of 31 Algorithms from the Black-Box Optimization Benchmarking BBOB-2009. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2010, pp.  1689–1696. ACM Press, New York, NY, 2010.
+[ bib | +DOI ] +
+This paper presents results of the BBOB-2009 benchmarking of + 31 search algorithms on 24 noiseless functions in a black-box + optimization scenario in continuous domain. The runtime of + the algorithms, measured in number of function evaluations, + is investigated and a connection between a single convergence + graph and the runtime distribution is uncovered. Performance + is investigated for different dimensions up to 40-D, for + different target precision values, and in different subgroups + of functions. Searching in larger dimension and multi-modal + functions appears to be more difficult. The choice of the + best algorithm also depends remarkably on the available + budget of function evaluations. +
+
+Keywords: benchmarking, black-box optimization +
+ +
+ + +
+[1943] +
+
+Nikolaus Hansen, Steffen Finck, Raymond Ros, and Anne Auger. + Real-Parameter Black-Box Optimization Benchmarking 2009: Noiseless Functions Definitions. + Technical Report RR-6829, INRIA, France, 2009. + Updated February 2010.
+[ bib | +epub ] +
+http://coco.gforge.inria.fr/bbob2012-downloads +
+ +
+ + +
+[1944] +
+
+Michael Pilegaard Hansen and Andrzej Jaszkiewicz. + Evaluating the quality of approximations to the non-dominated set. + Technical Report IMM-REP-1998-7, Institute of Mathematical Modelling, Technical University of Denmark, Lyngby, Denmark, 1998.
+[ bib ] +
+Proposed R2 indicator +
+ +
+ + +
+[1945] +
+
+Julia Handl and Joshua D. Knowles. + Modes of Problem Solving with Multiple Objectives: Implications for Interpreting the Pareto Set and for Decision Making. + In J. D. Knowles, D. Corne, K. Deb, and D. R. Chair, editors, Multiobjective Problem Solving from Nature, Natural Computing Series, pp.  131–151. Springer, Berlin/Heidelberg, 2008.
+[ bib | +DOI ] + +
+ + +
+[1946] +
+
+Pierre Hansen and Nenad Mladenović. + Variable Neighborhood Search. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  145–184. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[1947] +
+
+Pierre Hansen, Nenad Mladenović, Jack Brimberg, and José A. Moreno Pérez. + Variable Neighborhood Search. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, pp.  61–86. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[1948] +
+
+Nikolaus Hansen and Andreas Ostermeier. + Adapting Arbitrary Normal Mutation Distributions in Evolution Strategies: The Covariance Matrix Adaptation. + In T. Bäck, T. Fukuda, and Z. Michalewicz, editors, Proceedings of the 1996 IEEE International Conference on Evolutionary Computation (ICEC'96), pp.  312–317. IEEE Press, Piscataway, NJ, 1996.
+[ bib | +DOI ] +
+A new formulation for coordinate system independent + adaptation of arbitrary normal mutation distributions with + zero mean is presented. This enables the evolution strategy + (ES) to adapt the correct scaling of a given problem and also + ensures invariance with respect to any rotation of the + fitness function (or the coordinate system). Especially + rotation invariance, here resulting directly from the + coordinate system independent adaptation of the mutation + distribution, is an essential feature of the ES with regard + to its general applicability to complex fitness + functions. Compared to previous work on this subject, the + introduced formulation facilitates an interpretation of the + resulting mutation distribution, making sensible manipulation + by the user possible (if desired). Furthermore it enables a + more effective control of the overall mutation variance + (expected step length) +
+
+Proposed CMA-ES +
+
+Keywords: Evolution strategies, Evolutionary algorithms, + self-adaptation, stochastic processes, Covariance matrix, + matrix algebra, derandomised adaptation, mutation + distribution, rotation invariance, electronic switching + systems +
+ +
+ + +
+[1949] +
+
+Thomas Hanne. + Global Multiobjective Optimization with Evolutionary Algorithms: Selection Mechanisms and Mutation Control. + In E. Zitzler, K. Deb, L. Thiele, C. A. Coello Coello, and D. Corne, editors, Evolutionary Multi-criterion Optimization, EMO 2001, volume 1993 of Lecture Notes in Computer Science, pp.  197–212. Springer, Berlin/Heidelberg, 2001.
+[ bib ] + +
+ + +
+[1950] +
+
+Michael Pilegaard Hansen. + Metaheuristics for multiple objective combinatorial optimization. + PhD thesis, Institute of Mathematical Modelling, Technical University of Denmark, March 1998.
+[ bib ] + +
+ + +
+[1951] +
+
+Nikolaus Hansen. + The CMA evolution strategy: a comparing review. + In Towards a new evolutionary computation, pp.  75–102. Springer, 2006.
+[ bib ] + +
+ + +
+[1952] +
+
+Nikolaus Hansen. + Benchmarking a BI-population CMA-ES on the BBOB-2009 function testbed. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2009, pp.  2389–2396. ACM Press, New York, NY, 2009.
+[ bib ] +
+Keywords: bipop-cma-es +
+ +
+ + +
+[1953] +
+
+Zhifeng Hao, Ruichu Cai, and Han Huang. + An Adaptive Parameter Control Strategy for ACO. + In Proceedings of the International Conference on Machine Learning and Cybernetics, pp.  203–206. IEEE Press, 2006.
+[ bib ] + +
+ + +
+[1954] +
+
+Zhifeng Hao, Han Huang, Yong Qin, and Ruichu Cai. + An ACO Algorithm with Adaptive Volatility Rate of Pheromone Trail. + In Y. Shi, G. D. van Albada, J. Dongarra, and P. M. A. Sloot, editors, Computational Science – ICCS 2007, 7th International Conference, Proceedings, Part IV, volume 4490 of Lecture Notes in Computer Science, pp.  1167–1170. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[1955] +
+
+Jin-Kao Hao and Jêrome Pannier. + Simulated Annealing and Tabu Search for Constraint Solving. + In M. C. Golumbic et al., editors, Fifth International Symposium on Artificial Intelligence and Mathematics, AIM 1998, Fort Lauderdale, Florida, USA, January 4-6, 1998, pp.  1–15, 1998.
+[ bib ] + +
+ + +
+[1956] +
+
+William D. Harvey and Matthew L. Ginsberg. + Limited Discrepancy Search. + In C. S. Mellish, editor, Proceedings of the 14th International Joint Conference on Artificial Intelligence (IJCAI-95), pp.  607–615. Morgan Kaufmann Publishers, 1995.
+[ bib ] + +
+ + +
+[1957] +
+
+Hado van Hasselt, Arthur Guez, and David Silver. + Deep Reinforcement Learning with Double Q-Learning. + In D. Schuurmans and M. P. Wellman, editors, Proceedings of the AAAI Conference on Artificial Intelligence. AAAI Press, 2016.
+[ bib | +epub ] + +
+ + +
+[1958] +
+
+Verena Heidrich-Meisner and Christian Igel. + Hoeffding and Bernstein races for selecting policies in evolutionary direct policy search. + In A. P. Danyluk, L. Bottou, and M. L. Littman, editors, Proceedings of the 26th International Conference on Machine Learning, ICML 2009, pp.  401–408, New York, NY, 2009. ACM Press.
+[ bib | +DOI ] +
+Keywords: automated algorithm configuration, CMA-ES, racing +
+ +
+ + +
+[1959] +
+
+Keld Helsgaun. + Source Code of the Lin-Kernighan-Helsgaun Traveling Salesman Heuristic. + http://webhotel4.ruc.dk/~keld/research/LKH/, 2018.
+[ bib ] + +
+ + +
+[1960] +
+
+Keld Helsgaun. + Efficient Recombination in the Lin-Kernighan-Helsgaun Traveling Salesman Heuristic. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11101 of Lecture Notes in Computer Science, pp.  95–107. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] + +
+ + +
+[1961] +
+
+Pascal van Hentenryck. + The OPL optimization programming language. + MIT Press, Cambridge, MA, 1999.
+[ bib ] + +
+ + +
+[1962] +
+
+Darrall Henderson, Sheldon H. Jacobson, and Alan W. Johnson. + The Theory and Practice of Simulated Annealing. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  287–319. Springer, Boston, MA, 2003.
+[ bib | +DOI ] + +
+ + +
+[1963] +
+
+Pascal van Hentenryck and Laurent D. Michel. + Constraint-based Local Search. + MIT Press, Cambridge, MA, 2005.
+[ bib ] + +
+ + +
+[1964] +
+
+Pascal van Hentenryck and Laurent D. Michel. + Synthesis of constraint-based local search algorithms from high-level models. + In R. C. Holte and A. Howe, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  273–278. AAAI Press/MIT Press, Menlo Park, CA, 2007.
+[ bib ] + +
+ + +
+[1965] +
+
+R. Herbrich, T. Graepel, and K. Obermayer. + Support vector learning for ordinal regression. + In ICANN'99: Proceedings of the 9th International Conference on Artificial Neural Networks, pp.  97–102, 1999.
+[ bib | +DOI ] +
+We investigate the problem of predicting variables of ordinal + scale. This task is referred to as ordinal regression and is + complementary to the standard machine learning tasks of + classification and metric regression. In contrast to + statistical models we present a distribution independent + formulation of the problem together with uniform bounds of + the risk functional. The approach presented is based on a + mapping from objects to scalar utility values. Similar to + support vector methods we derive a new learning algorithm for + the task of ordinal regression based on large margin rank + boundaries. We give experimental results for an information + retrieval task: learning the order of documents with respect + to an initial query. Experimental results indicate that the + presented algorithm outperforms more naive approaches to + ordinal regression such as support vector classification and + support vector regression in the case of more than two + ranks. +
+
+Proposed the pairwise transform for learning-to-rank +
+
+Keywords: support vector machine;metric regression;support vector + learning;ordinal regression;information retrieval;risk + functional;machine learning;pattern classification; +
+ +
+ + +
+[1966] +
+
+Francisco Herrera, Manuel Lozano, and Daniel Molina. + Test suite for the special issue of Soft Computing on scalability of evolutionary algorithms and other metaheuristics for large scale continuous optimization problems. + http://sci2s.ugr.es/eamhco/, 2010.
+[ bib ] +
+Keywords: SOCO benchmark +
+ +
+ + +
+[1967] +
+
+Carlos Hernández and Oliver Schütze. + A bounded archive based for bi-objective problems based on distance and e-dominance to avoid cyclic behavior. + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, pp.  583–591. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] + +
+ + +
+[1968] +
+
+Daniel P Heyman and Matthew J Sobel. + Stochastic models in operations research: stochastic optimization, volume 2. + Courier Corporation, 2003.
+[ bib ] + +
+ + +
+[1969] +
+
+J. Holland. + Adaptation in Natural and Artificial Systems. + University of Michigan Press, 1975.
+[ bib ] + +
+ + +
+[1970] +
+
+Myle Hollander and Douglas A. Wolfe. + Nonparametric statistical inference. + John Wiley & Sons, New York, NY, 1973. + Second edition (1999).
+[ bib ] + +
+ + +
+[1971] +
+
+Giles Hooker. + Discovering Additive Structure in Black Box Functions. + In W. Kim, R. Kohavi, J. Gehrke, and W. DuMouchel, editors, Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, KDD'04, pp.  575–580. ACM Press, New York, NY, 2004.
+[ bib | +DOI ] +
+Many automated learning procedures lack interpretability, + operating effectively as a black box: providing a prediction + tool but no explanation of the underlying dynamics that drive + it. A common approach to interpretation is to plot the + dependence of a learned function on one or two predictors. We + present a method that seeks not to display the behavior of a + function, but to evaluate the importance of non-additive + interactions within any set of variables. Should the function + be close to a sum of low dimensional components, these + components can be viewed and even modeled + parametrically. Alternatively, the work here provides an + indication of where intrinsically high-dimensional behavior + takes place.The calculations used in this paper correspond + closely with the functional ANOVA decomposition; a + well-developed construction in Statistics. In particular, the + proposed score of interaction importance measures the loss + associated with the projection of the prediction function + onto a space of additive models. The algorithm runs in linear + time and we present displays of the output as a graphical + model of the function for interpretation purposes. +
+
+Keywords: diagnostics, functional ANOVA, feature selection, + interpretation, visualization, additive models, draphical + models +
+ +
+ + +
+[1972] +
+
+Holger H. Hoos, Frank Hutter, and Kevin Leyton-Brown. + Automated Configuration and Selection of SAT Solvers. + In Handbook of Satisfiability, pp.  481–507. IOS Press, February 2021.
+[ bib | +DOI ] + +
+ + +
+[1973] +
+
+Holger H. Hoos and Thomas Stützle. + Stochastic Local Search: Foundations and Applications. + Elsevier, Amsterdam, The Netherlands, 2004.
+[ bib ] + +
+ + +
+[1974] +
+
+Holger H. Hoos and Thomas Stützle. + Stochastic Local Search—Foundations and Applications. + Morgan Kaufmann Publishers, San Francisco, CA, 2005.
+[ bib ] + +
+ + +
+[1975] +
+
+Holger H. Hoos and Thomas Stützle. + Evaluating Las Vegas Algorithms — Pitfalls and Remedies. + In G. F. Cooper and S. Moral, editors, Proceedings of the Fourteenth Conference on Uncertainty in Artificial Intelligence, pp.  238–245. Morgan Kaufmann Publishers, San Francisco, CA, 1998.
+[ bib ] + +
+ + +
+[1976] +
+
+Holger H. Hoos. + Programming by Optimisation: Towards a new Paradigm for Developing High-Performance Software. + In MIC 2011, the 9th Metaheuristics International Conference, 2011. + Plenary talk.
+[ bib | +http ] + +
+ + +
+[1977] +
+
+Holger H. Hoos. + Automated Algorithm Configuration and Parameter Tuning. + In Y. Hamadi, E. Monfroy, and F. Saubion, editors, Autonomous Search, pp.  37–71. Springer, Berlin, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[1978] +
+
+Christian Horoba and Frank Neumann. + Benefits and drawbacks for the use of epsilon-dominance in evolutionary multi-objective optimization. + In C. Ryan, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2008, pp.  641–648. ACM Press, New York, NY, 2008.
+[ bib ] +
+Proposed ε-box +
+ +
+ + +
+[1979] +
+
+J. Horn, N. Nafpliotis, and David E. Goldberg. + A niched Pareto genetic algorithm for multiobjective optimization. + In Proceedings of the 1994 World Congress on Computational Intelligence (WCCI 1994), pp.  82–87, Piscataway, NJ, June 1994. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[1980] +
+
+Kenneth Hoste and Lieven Eeckhout. + Cole: Compiler Optimization Level Exploration. + In M. L. Soffa and E. Duesterwald, editors, Proceedings of the 6th Annual IEEE/ACM International Symposium on Code Generation and Optimization, CGO '08, pp.  165–174, New York, NY, 2008. ACM Press.
+[ bib | +DOI ] + +
+ + +
+[1981] +
+
+Han Huang, Xiaowei Yang, Zhifeng Hao, and Ruichu Cai. + A Novel ACO Algorithm with Adaptive Parameter. + In D.-S. Huang, K. Li, and G. W. Irwin, editors, International Conference on Computational Science (3), volume 4115 of Lecture Notes in Computer Science, pp.  12–21. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[1982] +
+
+Kuo-Si Huang, Chang-Biau Yang, and Kuo tsung Tseng. + Fast algorithms for finding the common subsequences of multiple sequences. + In Proceedings of the International Computer Symposium, pp.  1006–1011. IEEE Press, 2004.
+[ bib ] + +
+ + +
+[1983] +
+
+Evan J. Hughes. + Multiple single objective Pareto sampling. + In Proceedings of the 2003 Congress on Evolutionary Computation (CEC'03), pp.  2678–2684, Piscataway, NJ, December 2003. IEEE Press.
+[ bib ] + +
+ + +
+[1984] +
+
+Evan J. Hughes. + MSOPS-II: A general-purpose many-objective optimiser. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  3944–3951, Piscataway, NJ, 2007. IEEE Press.
+[ bib ] + +
+ + +
+[1985] +
+
+Evan J. Hughes. + Many-objective directed evolutionary line search. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  761–768. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[1986] +
+
+Maura Hunt and Manuel López-Ibáñez. + Modeling a Decision-Maker in Goal Programming by means of Computational Rationality. + In I. Palomares, editor, International Alan Turing Conference on Decision Support and Recommender systems, pp.  17–20, London, UK, November 21–22 2019. Alan Turing Institute.
+[ bib | +epub ] +
+This paper extends a simulation of cognitive mechanisms in + the context of multi-criteria decision-making by using ideas + from computational rationality. Specifically, this paper + improves the simulation of a human decision-maker (DM) by + considering how resource constraints impact their evaluation + process in an interactive Goal Programming problem. Our + analysis confirms and emphasizes a previous simulation study + by showing key areas that could be effected by cognitive + mechanisms. While the results are promising, the effects + should be validated by future experiments with human DMs. +
+ +
+ + +
+[1987] +
+
+Mohamed Saifullah Hussin and Thomas Stützle. + Hierarchical Iterated Local Search for the Quadratic Assignment Problem. + In M. J. Blesa, C. Blum, L. Di Gaspero, A. Roli, M. Sampels, and A. Schaerf, editors, Hybrid Metaheuristics, volume 5818 of Lecture Notes in Computer Science, pp.  115–129. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[1988] +
+
+Frank Hutter, Domagoj Babić, Holger H. Hoos, and Alan J. Hu. + Boosting Verification by Automatic Tuning of Decision Procedures. + In J. Baumgartner and M. Sheeran, editors, FMCAD'07: Proceedings of the 7th International Conference Formal Methods in Computer Aided Design, pp.  27–34, Austin, Texas, USA, 2007. IEEE Computer Society, Washington, DC, USA.
+[ bib ] + +
+ + +
+[1989] +
+
+Frank Hutter, Holger H. Hoos, Kevin Leyton-Brown, and Kevin P. Murphy. + An experimental investigation of model-based parameter optimisation: SPO and beyond. + In F. Rothlauf, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2009, pp.  271–278. ACM Press, New York, NY, 2009.
+[ bib | +DOI ] + +
+ + +
+[1990] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Automated Configuration of Mixed Integer Programming Solvers. + In A. Lodi, M. Milano, and P. Toth, editors, Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 7th International Conference, CPAIOR 2010, volume 6140 of Lecture Notes in Computer Science, pp.  186–202. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] +
+Keywords: MIP, ParamILS +
+ +
+ + +
+[1991] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Sequential Model-Based Optimization for General Algorithm Configuration. + In C. A. Coello Coello, editor, Learning and Intelligent Optimization, 5th International Conference, LION 5, volume 6683 of Lecture Notes in Computer Science, pp.  507–523. Springer, Heidelberg, Germany, 2011.
+[ bib | +DOI ] +
+Keywords: SMAC,ROAR +
+ +
+ + +
+[1992] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Parallel Algorithm Configuration. + In Y. Hamadi and M. Schoenauer, editors, Learning and Intelligent Optimization, 6th International Conference, LION 6, volume 7219 of Lecture Notes in Computer Science, pp.  55–70. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[1993] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Identifying Key Algorithm Parameters and Instance Features using Forward Selection. + In P. M. Pardalos and G. Nicosia, editors, Learning and Intelligent Optimization, 7th International Conference, LION 7, volume 7997 of Lecture Notes in Computer Science, pp.  364–381. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] +
+Most state-of-the-art algorithms for large-scale optimization + problems expose free parameters, giving rise to combinatorial + spaces of possible configurations. Typically, these spaces + are hard for humans to understand. In this work, we study a + model-based approach for identifying a small set of both + algorithm parameters and instance features that suffices for + predicting empirical algorithm performance well. Our + empirical analyses on a wide variety of hard combinatorial + problem benchmarks spanning SAT, MIP, and TSP show that–for + parameter configurations sampled uniformly at random–very + good performance predictions can typically be obtained based + on just two key parameters, and that similarly, few instance + features and algorithm parameters suffice to predict the most + salient algorithm performance characteristics in the combined + configuration/feature space. We also use these models to + identify settings of these key parameters that are predicted + to achieve the best overall performance, both on average + across instances and in an instance-specific way. This serves + as a further way of evaluating model quality and also + provides a tool for further understanding the parameter + space. We provide software for carrying out this analysis on + arbitrary problem domains and hope that it will help + algorithm developers gain insights into the key parameters of + their algorithms, the key features of their instances, and + their interactions. +
+
+Keywords: parameter importance +
+ +
+ + +
+[1994] +
+
+Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + An Efficient Approach for Assessing Hyperparameter Importance. + In E. P. Xing and T. Jebara, editors, Proceedings of the 31st International Conference on Machine Learning, ICML 2014, volume 32, pp.  754–762. PMLR, 2014.
+[ bib | +http ] +
+Keywords: fANOVA, parameter importance +
+ +
+ + +
+[1995] +
+
+Frank Hutter, Holger H. Hoos, Kevin Leyton-Brown, and Kevin Murphy. + Time-Bounded Sequential Parameter Optimization. + In C. Blum and R. Battiti, editors, Learning and Intelligent Optimization, 4th International Conference, LION 4, volume 6073 of Lecture Notes in Computer Science, pp.  281–298. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[1996] +
+
+Frank Hutter, Holger H. Hoos, and Thomas Stützle. + Automatic Algorithm Configuration Based on Local Search. + In R. C. Holte and A. Howe, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  1152–1157. AAAI Press/MIT Press, Menlo Park, CA, 2007.
+[ bib ] + +
+ + +
+[1997] +
+
+Frank Hutter, Manuel López-Ibáñez, Chris Fawcett, Marius Thomas Lindauer, Holger H. Hoos, Kevin Leyton-Brown, and Thomas Stützle. + AClib: A Benchmark Library for Algorithm Configuration. + In P. M. Pardalos, M. G. C. Resende, C. Vogiatzis, and J. L. Walteros, editors, Learning and Intelligent Optimization, 8th International Conference, LION 8, volume 8426 of Lecture Notes in Computer Science, pp.  36–40. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[1998] +
+
+Frank Hutter. + SAT benchmarks used in automated algorithm configuration. + http://www.cs.ubc.ca/labs/beta/Projects/AAC/SAT-benchmarks.html, 2007.
+[ bib ] + +
+ + +
+[1999] +
+
+Frank Hutter. + Automated Configuration of Algorithms for Solving Hard Computational Problems. + PhD thesis, University of British Columbia, Department of Computer Science, Vancouver, Canada, October 2009.
+[ bib ] + +
+ + +
+[2000] +
+
+Zhiyuan Liu and Jian Tang. + IJCAI 2021 Reproducibility Guidelines, 35th International Joint Conference on Artificial Intelligence. + https://ijcai-21.org/wp-content/uploads/2020/12/20201226-IJCAI-Reproducibility.pdf, 2021.
+[ bib ] + +
+ + +
+[2001] +
+
+Jérémie Humeau, Arnaud Liefooghe, El-Ghazali Talbi, and Sébastien Verel. + ParadisEO-MO: From Fitness Landscape Analysis to Efficient Local Search Algorithms. + Rapport de recherche RR-7871, INRIA, France, 2012.
+[ bib | +epub ] + +
+ + +
+[2002] +
+
+Mauro Birattari. + The race Package for R: Racing Methods for the Selection of the Best. + Technical Report TR/IRIDIA/2003-037, IRIDIA, Université Libre de Bruxelles, Belgium, 2003.
+[ bib ] + +
+ + +
+[2003] +
+
+Mauro Birattari. + On the Estimation of the Expected Performance of a Metaheuristic on a Class of Instances. How Many Instances, How Many Runs? + Technical Report TR/IRIDIA/2004-001, IRIDIA, Université Libre de Bruxelles, Belgium, 2004.
+[ bib ] + +
+ + +
+[2004] +
+
+Krzysztof Socha and Marco Dorigo. + Ant Colony Optimization for Mixed-Variable Optimization Problems. + Technical Report TR/IRIDIA/2007-019, IRIDIA, Université Libre de Bruxelles, Belgium, October 2007.
+[ bib ] + +
+ + +
+[2005] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + Exploratory Analysis of Stochastic Local Search Algorithms in Biobjective Optimization. + Technical Report TR/IRIDIA/2009-015, IRIDIA, Université Libre de Bruxelles, Belgium, May 2009. + Published as a book chapter [2191].
+[ bib ] + +
+ + +
+[2006] +
+
+Manuel López-Ibáñez and Thomas Stützle. + An Analysis of Algorithmic Components for Multiobjective Ant Colony Optimization: A Case Study on the Biobjective TSP. + Technical Report TR/IRIDIA/2009-019, IRIDIA, Université Libre de Bruxelles, Belgium, June 2009. + Published in the proceedings of Evolution Artificielle, 2009 [2198].
+[ bib ] + +
+ + +
+[2007] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Effective Hybrid Stochastic Local Search Algorithms for Biobjective Permutation Flowshop Scheduling. + Technical Report TR/IRIDIA/2009-020, IRIDIA, Université Libre de Bruxelles, Belgium, June 2009. + Published in the proceedings of Hybrid Metaheuristics 2009 [1769].
+[ bib | +http ] + +
+ + +
+[2008] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Adaptive “Anytime” Two-Phase Local Search. + Technical Report TR/IRIDIA/2009-026, IRIDIA, Université Libre de Bruxelles, Belgium, 2010. + Published in the proceedings of LION 4 [1772].
+[ bib | +http ] + +
+ + +
+[2009] +
+
+Thomas Stützle, Manuel López-Ibáñez, Paola Pellegrini, Michael Maur, Marco A. Montes de Oca, Mauro Birattari, and Marco Dorigo. + Parameter Adaptation in Ant Colony Optimization. + Technical Report TR/IRIDIA/2010-002, IRIDIA, Université Libre de Bruxelles, Belgium, January 2010. + Published as a book chapter [2555].
+[ bib ] + +
+ + +
+[2010] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + A Hybrid TP+PLS Algorithm for Bi-objective Flow-Shop Scheduling Problems. + Technical Report TR/IRIDIA/2010-019, IRIDIA, Université Libre de Bruxelles, Belgium, 2010. + Published in Computers & Operations Research [384].
+[ bib | +http ] + +
+ + +
+[2011] +
+
+M. S. Hussin and Thomas Stützle. + Tabu Search vs. Simulated Annealing for Solving Large Quadratic Assignment Instances. + Technical Report TR/IRIDIA/2010-020, IRIDIA, Université Libre de Bruxelles, Belgium, 2010.
+[ bib ] + +
+ + +
+[2012] +
+
+Jérémie Dubois-Lacoste, Manuel López-Ibáñez, and Thomas Stützle. + Improving the Anytime Behavior of Two-Phase Local Search. + Technical Report TR/IRIDIA/2010-022, IRIDIA, Université Libre de Bruxelles, Belgium, 2010. + Published in Annals of Mathematics and Artificial Intelligence [383].
+[ bib | +http ] + +
+ + +
+[2013] +
+
+Manuel López-Ibáñez, Joshua D. Knowles, and Marco Laumanns. + On Sequential Online Archiving of Objective Vectors. + Technical Report TR/IRIDIA/2011-001, IRIDIA, Université Libre de Bruxelles, Belgium, 2011. + This is a revised version of the paper published in EMO 2011 [2184].
+[ bib | +http ] + +
+ + +
+[2014] +
+
+Mauro Birattari, Marco Chiarandini, Marco Saerens, and Thomas Stützle. + Learning graphical models for parameter tuning. + Technical Report TR/IRIDIA/2011-002, IRIDIA, Université Libre de Bruxelles, Belgium, 2011.
+[ bib | +http ] + +
+ + +
+[2015] +
+
+Manuel López-Ibáñez and Thomas Stützle. + The Automatic Design of Multi-Objective Ant Colony Optimization Algorithms. + Technical Report TR/IRIDIA/2011-003, IRIDIA, Université Libre de Bruxelles, Belgium, 2011. + Published in IEEE Transactions on Evolutionary Computation [860].
+[ bib | +http ] + +
+ + +
+[2016] +
+
+Tianjun Liao, Daniel Molina, Marco A. Montes de Oca, and Thomas Stützle. + A Note on the Effects of Enforcing Bound Constraints on Algorithm Comparisons using the IEEE CEC'05 Benchmark Function Suite. + Technical Report TR/IRIDIA/2011-010, IRIDIA, Université Libre de Bruxelles, Belgium, 2011. + Published in Evolutionary Computation [823].
+[ bib | +http ] + +
+ + +
+[2017] +
+
+Tianjun Liao, Daniel Molina, Marco A. Montes de Oca, and Thomas Stützle. + Computational Results for an Automatically Tuned IPOP-CMA-ES on the CEC'05 Benchmark Set. + Technical Report TR/IRIDIA/2011-022, IRIDIA, Université Libre de Bruxelles, Belgium, 2011.
+[ bib ] + +
+ + +
+[2018] +
+
+Manuel López-Ibáñez and Thomas Stützle. + Automatically Improving the Anytime Behaviour of Optimisation Algorithms. + Technical Report TR/IRIDIA/2012-012, IRIDIA, Université Libre de Bruxelles, Belgium, May 2012. + Published in European Journal of Operational Research [861].
+[ bib ] + +
+ + +
+[2019] +
+
+Andreea Radulescu, Manuel López-Ibáñez, and Thomas Stützle. + Automatically Improving the Anytime Behaviour of Multiobjective Evolutionary Algorithms. + Technical Report TR/IRIDIA/2012-019, IRIDIA, Université Libre de Bruxelles, Belgium, 2012. + Published in the proceedings of EMO 2013 [2418].
+[ bib ] + +
+ + +
+[2020] +
+
+Tianjun Liao, Thomas Stützle, Marco A. Montes de Oca, and Marco Dorigo. + A Unified Ant Colony Optimization Algorithm for Continuous Optimization. + Technical Report TR/IRIDIA/2013-002, IRIDIA, Université Libre de Bruxelles, Belgium, 2013.
+[ bib ] + +
+ + +
+[2021] +
+
+Franco Mascia, Manuel López-Ibáñez, Jérémie Dubois-Lacoste, and Thomas Stützle. + Grammar-based generation of stochastic local search heuristics through automatic algorithm configuration tools. + Technical Report TR/IRIDIA/2013-015, IRIDIA, Université Libre de Bruxelles, Belgium, 2013.
+[ bib ] + +
+ + +
+[2022] +
+
+Manuel López-Ibáñez, Arnaud Liefooghe, and Sébastien Verel. + Local Optimal Sets and Bounded Archiving on Multi-objective NK-Landscapes with Correlated Objectives. + Technical Report TR/IRIDIA/2014-009, IRIDIA, Université Libre de Bruxelles, Belgium, 2014.
+[ bib ] + +
+ + +
+[2023] +
+
+Vito Trianni and Manuel López-Ibáñez. + Advantages of Multi-Objective Optimisation in Evolutionary Robotics: Survey and Case Studies. + Technical Report TR/IRIDIA/2014-014, IRIDIA, Université Libre de Bruxelles, Belgium, 2014.
+[ bib | +http ] + +
+ + +
+[2024] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + A Large-Scale Experimental Evaluation of High-Performing Multi- and Many-Objective Evolutionary Algorithms. + Technical Report TR/IRIDIA/2017-005, IRIDIA, Université Libre de Bruxelles, Belgium, November 2017.
+[ bib ] + +
+ + +
+[2025] +
+
+Alberto Franzin, Leslie Pérez Cáceres, and Thomas Stützle. + Effect of Transformations of Numerical Parameters in Automatic Algorithm Configuration. + Technical Report TR/IRIDIA/2017-006, IRIDIA, Université Libre de Bruxelles, Belgium, March 2017.
+[ bib | +http ] + +
+ + +
+[2026] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Configuration of Multi-objective Optimizers and Multi-objective Configuration. + Technical Report TR/IRIDIA/2017-011, IRIDIA, Université Libre de Bruxelles, Belgium, November 2017. + Published as a book chapter [1546].
+[ bib | +http ] + +
+ + +
+[2027] +
+
+Manuel López-Ibáñez, Marie-Eléonore Kessaci, and Thomas Stützle. + Automatic Design of Hybrid Metaheuristics from Algorithmic Components. + Technical Report TR/IRIDIA/2017-012, IRIDIA, Université Libre de Bruxelles, Belgium, December 2017.
+[ bib | +http ] + +
+ + +
+[2028] +
+
+Leonardo C. T. Bezerra, Manuel López-Ibáñez, and Thomas Stützle. + Automatically Designing State-of-the-Art Multi- and Many-Objective Evolutionary Algorithms. + Technical Report TR/IRIDIA/2018-001, IRIDIA, Université Libre de Bruxelles, Belgium, January 2018. + Published in Evolutionary Computation journal [129].
+[ bib | +http ] + +
+ + +
+[2029] +
+
+Alberto Franzin and Thomas Stützle. + Revisiting Simulated Annealing: a Component-Based Analysis. + Technical Report TR/IRIDIA/2018-010, IRIDIA, Université Libre de Bruxelles, Belgium, 2018.
+[ bib | +http ] + +
+ + +
+[2030] +
+
+Christian Leonardo Camacho-Villalón, Thomas Stützle, and Marco Dorigo. + PSO-X: A Component-Based Framework for the Automatic Design of Particle Swarm Optimization Algorithms. + Technical Report TR/IRIDIA/2021-002, IRIDIA, Université Libre de Bruxelles, Belgium, 2021.
+[ bib | +http ] +
+Published as [223] +
+ +
+ + +
+[2031] +
+
+Alberto Franzin and Thomas Stützle. + A Landscape-based Analysis of Fixed Temperature and Simulated Annealing. + Technical Report TR/IRIDIA/2021-005, IRIDIA, Université Libre de Bruxelles, Belgium, 2021.
+[ bib | +http ] + +
+ + +
+[2032] +
+
+Christian Leonardo Camacho-Villalón, Thomas Stützle, and Marco Dorigo. + Cuckoo Search ≡(μ+ λ)-Evolution Strategy – A Rigorous Analysis of an Algorithm That Has Been Misleading the Research Community for More Than 10 Years and Nobody Seems to Have Noticed. + Technical Report TR/IRIDIA/2021-006, IRIDIA, Université Libre de Bruxelles, Belgium, 2021.
+[ bib | +http ] + +
+ + +
+[2033] +
+
+Christian Igel. + Multi-objective Model Selection for Support Vector Machines. + In C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors, Evolutionary Multi-criterion Optimization, EMO 2005, volume 3410 of Lecture Notes in Computer Science, pp.  534–546. Springer, Heidelberg, Germany, 2005.
+[ bib | +DOI ] +
+Early work on multi-objective hyper-parameter optimization + (AutoML) +
+ +
+ + +
+[2034] +
+
+Kokolo Ikeda, Hajime Kita, and Shigenobu Kobayashi. + Failure of Pareto-based MOEAs: Does non-dominated really mean near to optimal? + In Proceedings of the 2001 Congress on Evolutionary Computation (CEC'01), pp.  957–962, Piscataway, NJ, 2001. IEEE Press.
+[ bib ] +
+Keywords: dominance resistance +
+ +
+ + +
+[2035] +
+
+Janine Illian, Antti Penttinen, Helga Stoyan, and Dietrich Stoyan. + Statistical Analysis and Modelling of Spatial Point Patterns. + Wiley, 2008.
+[ bib ] + +
+ + +
+[2036] +
+
+S. Iredi, D. Merkle, and Martin Middendorf. + Bi-Criterion Optimization with Multi Colony Ant Algorithms. + In E. Zitzler, K. Deb, L. Thiele, C. A. Coello Coello, and D. Corne, editors, Evolutionary Multi-criterion Optimization, EMO 2001, volume 1993 of Lecture Notes in Computer Science, pp.  359–372. Springer, Berlin/Heidelberg, 2001.
+[ bib ] +
+Keywords: BicriterionAnt +
+ +
+ + +
+[2037] +
+
+Manuel López-Ibáñez and Thomas Stützle. + Automatically Improving the Anytime Behaviour of Optimisation Algorithms: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2012-011/, 2012.
+[ bib ] + +
+ + +
+[2038] +
+
+Ekhine Irurozki and Manuel López-Ibáñez. + Unbalanced Mallows Models for Optimizing Expensive Black-Box Permutation Problems. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  225–233. ACM Press, New York, NY, 2021.
+[ bib | +DOI | +supplementary material ] +
+Expensive black-box combinatorial optimization problems arise + in practice when the objective function is evaluated by means + of a simulator or a real-world experiment. Since each fitness + evaluation is expensive in terms of time or resources, only a + limited number of evaluations is possible, typically several + orders of magnitude smaller than in non-expensive + problems. In this scenario, classical optimization methods + such as mixed-integer programming and local search are not + useful. In the continuous case, Bayesian optimization, in + particular using Gaussian processes, has proven very + effective under these conditions. Much less research is + available in the combinatorial case. In this paper, we + propose and analyze UMM, an estimation-of-distribution (EDA) + algorithm based on a Mallows probabilistic model and + unbalanced rank aggregation (uBorda). Experimental results on + black-box versions of LOP and PFSP show that UMM is able to + match, and sometimes surpass, the solutions obtained by CEGO, + a Bayesian optimization algorithm for combinatorial + optimization. Moreover, the computational complexity of UMM + increases linearly with both the number of function + evaluations and the permutation size. +
+
+Keywords: UMM, Permutation, Expensive, Black-box +
+ +
+ + +
+[2039] +
+
+Hisao Ishibuchi, Hiroyuki Masuda, and Yusuke Nojima. + A Study on Performance Evaluation Ability of a Modified Inverted Generational Distance Indicator. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  695–702. ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[2040] +
+
+Hisao Ishibuchi, Hiroyuki Masuda, Yuki Tanigaki, and Yusuke Nojima. + Modified Distance Calculation in Generational Distance and Inverted Generational Distance. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  110–125. Springer, Heidelberg, Germany, 2015.
+[ bib ] +
+Proposed IGD+ +
+
+Keywords: Performance metrics, multi-objective, IGD, IGD+ +
+ +
+ + +
+[2041] +
+
+Hisao Ishibuchi, N. Tsukamoto, and Y. Nojima. + Evolutionary many-objective optimization: A short review. + In Proceedings of the 2008 Congress on Evolutionary Computation (CEC 2008), pp.  2419–2426, Piscataway, NJ, 2008. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2042] +
+
+Sophie Jacquin, Laetitia Jourdan, and El-Ghazali Talbi. + Dynamic Programming Based Metaheuristic for Energy Planning Problems. + In A. I. Esparcia-Alcázar and A. M. Mora, editors, Applications of Evolutionary Computation, volume 8602 of Lecture Notes in Computer Science, pp.  165–176. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[2043] +
+
+Antonio López Jaimes, Carlos A. Coello Coello, and Debrup Chakraborty. + Objective reduction using a feature selection technique. + In C. Ryan, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2008, pp.  673–680. ACM Press, New York, NY, 2008.
+[ bib ] + +
+ + +
+[2044] +
+
+Antonio López Jaimes, Carlos A. Coello Coello, and Jesús E. Urías Barrientos. + Online Objective Reduction to Deal with Many-Objective Problems. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  423–437. Springer, Heidelberg, Germany, 2009.
+[ bib ] +
+In this paper, we propose and analyze two schemes to + integrate an objective reduction technique into a + multi-objective evolutionary algorithm (moea) in order to + cope with many-objective problems. One scheme reduces + periodically the number objectives during the search until + the required objective subset size has been reached and, + towards the end of the search, the original objective set is + used again. The second approach is a more conservative scheme + that alternately uses the reduced and the entire set of + objectives to carry out the search. Besides improving + computational efficiency by removing some objectives, the + experimental results showed that both objective reduction + schemes also considerably improve the convergence of a moea + in many-objective problems. +
+ +
+ + +
+[2045] +
+
+Kevin G. Jamieson and Ameet Talwalkar. + Non-stochastic Best Arm Identification and Hyperparameter Optimization. + In A. Gretton and C. C. Robert, editors, Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, AISTATS 2016, Cadiz, Spain, May 9-11, 2016, volume 51 of JMLR Workshop and Conference Proceedings, pp.  240–248. JMLR.org, 2016.
+[ bib | +http ] + +
+ + +
+[2046] +
+
+Andrzej Jaszkiewicz and Jürgen Branke. + Interactive Multiobjective Evolutionary Algorithms. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science, pp.  179–193. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[2047] +
+
+Andrzej Jaszkiewicz, Hisao Ishibuchi, and Qingfu Zhang. + Multiobjective memetic algorithms. + In F. Neri, C. Cotta, and P. Moscato, editors, Handbook of Memetic Algorithms, volume 379 of Studies in Computational Intelligence, pp.  201–217. Springer, 2011.
+[ bib ] + +
+ + +
+[2048] +
+
+Frank Hutter and Steve Ramage. + Manual for SMAC. + University of British Columbia, 2015. + SMAC version 2.10.03.
+[ bib | +http ] + +
+ + +
+[2049] +
+
+Mark Jerrum and Alistair Sinclair. + The Markov chain Monte Carlo method: an approach to approximate counting and integration. + In D. S. Hochbaum, editor, Approximation Algorithms For NP-hard Problems, pp.  482–520. PWS Publishing Co., 1996.
+[ bib ] + +
+ + +
+[2050] +
+
+Alexandre D. Jesus, Arnaud Liefooghe, Bilel Derbel, and Luís Paquete. + Algorithm Selection of Anytime Algorithms. + In C. A. Coello Coello, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020, pp.  850—858. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[2051] +
+
+Journal of Heuristics. Policies on Heuristic Search Research. + http://www.springer.com/journal/10732, 2015. + Version visited last on June 10, 2015.
+[ bib ] + +
+ + +
+[2052] +
+
+David S. Johnson, G. Gutin, Lyle A. McGeoch, A. Yeo, W. Zhang, and A. Zverovitch. + Experimental Analysis of Heuristics for the ATSP. + In G. Gutin and A. Punnen, editors, The Traveling Salesman Problem and its Variations, pp.  445–487. Kluwer Academic Publishers, Dordrecht, The Netherlands, 2002.
+[ bib ] + +
+ + +
+[2053] +
+
+David S. Johnson and Lyle A. McGeoch. + Experimental Analysis of Heuristics for the STSP. + In G. Gutin and A. Punnen, editors, The Traveling Salesman Problem and its Variations, pp.  369–443. Kluwer Academic Publishers, Dordrecht, The Netherlands, 2002.
+[ bib ] + +
+ + +
+[2054] +
+
+David S. Johnson and Lyle A. McGeoch. + The Traveling Salesman Problem: A Case Study in Local Optimization. + In E. H. L. Aarts and J. K. Lenstra, editors, Local Search in Combinatorial Optimization, pp.  215–310. John Wiley & Sons, Chichester, UK, 1997.
+[ bib ] + +
+ + +
+[2055] +
+
+David S. Johnson. + Local Optimization and the Traveling Salesman Problem. + In M. Paterson, editor, Automata, Languages and Programming, 17th International Colloquium, volume 443 of Lecture Notes in Computer Science, pp.  446–461, Heidelberg, Germany, 1990. Springer.
+[ bib ] + +
+ + +
+[2056] +
+
+David S. Johnson, Lyle A. McGeoch, C. Rego, and Fred Glover. + 8th DIMACS Implementation Challenge: The Traveling Salesman Problem. + http://dimacs.rutgers.edu/archive/Challenges/TSP, 2001.
+[ bib ] +
+Keywords: TSP Challenge, RUE, RCE, generators +
+ +
+ + +
+[2057] +
+
+David S. Johnson. + A Theoretician's Guide to the Experimental Analysis of Algorithms. + In M. H. Goldwasser, D. S. Johnson, and C. C. McGeoch, editors, Data Structures, Near Neighbor Searches, and Methodology: Fifth and Sixth DIMACS Implementation Challenges, volume 59 of DIMACS Series in Discrete Mathematics and Theoretical Computer Science, pp.  215–250. American Mathematical Society, Providence, RI, 2002.
+[ bib | +DOI ] + +
+ + +
+[2058] +
+
+Kenneth A. De Jong. + Evolutionary computation: a unified approach. + MIT Press, Cambridge, MA, 2006.
+[ bib ] + +
+ + +
+[2059] +
+
+Terry Jones and Stephanie Forrest. + Fitness Distance Correlation as a Measure of Problem Difficulty for Genetic Algorithms. + In L. J. Eshelman, editor, Proceedings of the Sixth International Conference on Genetic Algorithms (ICGA'95), pp.  184–192. Morgan Kaufmann Publishers, San Francisco, CA, Pittsburgh, PA, 1995.
+[ bib ] + +
+ + +
+[2060] +
+
+Neil C. Jones and Pavel A. Pevzner. + An introduction to bioinformatics algorithms. + MIT Press, Cambridge, MA, 2004.
+[ bib ] + +
+ + +
+[2061] +
+
+H. Juillé and J. B. Pollack. + A Sampling-Based Heuristic for Tree Search Applied to Grammar Induction. + In J. Mostow and C. Rich, editors, Proceedings of AAAI 1998 – Fifteenth National Conference on Artificial Intelligence, pp.  776–783. AAAI Press/MIT Press, Menlo Park, CA, 1998.
+[ bib ] + +
+ + +
+[2062] +
+
+Bryant A. Julstrom. + What Have You Done for Me Lately? Adapting Operator Probabilities in a Steady-State Genetic Algorithm. + In L. J. Eshelman, editor, Proceedings of the Sixth International Conference on Genetic Algorithms (ICGA'95), pp.  81–87. Morgan Kaufmann Publishers, San Francisco, CA, Pittsburgh, PA, 1995.
+[ bib ] + +
+ + +
+[2063] +
+
+Serdar Kadioglu, Yuri Malitsky, Meinolf Sellmann, and Kevin Tierney. + ISAC: Instance-Specific Algorithm Configuration. + In H. Coelho, R. Studer, and M. Wooldridge, editors, Proceedings of the 19th European Conference on Artificial Intelligence, pp.  751–756. IOS Press, 2010.
+[ bib ] + +
+ + +
+[2064] +
+
+H. Kaji, Kokolo Ikeda, and Hajime Kita. + Avoidance of constraint violation for experiment-based evolutionary multi-objective optimization. + In Proceedings of the 2009 Congress on Evolutionary Computation (CEC 2009), pp.  2756–2763, Piscataway, NJ, 2009. IEEE Press.
+[ bib | +DOI ] +
+Keywords: Safe Optimization, evolutionary computation, constraint + violation, experiment-based evolutionary multiobjective + optimization, evolutionary algorithm, risky-constraint + violation, Constraint optimization, Diesel engines, + Calibration, Evolutionary computation, Electric breakdown, + Optimization methods, Uncertainty, Computational fluid + dynamics +
+ +
+ + +
+[2065] +
+
+Giorgos Karafotias, Agoston E. Eiben, and Mark Hoogendoorn. + Generic parameter control with reinforcement learning. + In C. Igel and D. V. Arnold, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2014, pp.  1319–1326. ACM Press, New York, NY, 2014.
+[ bib ] + +
+ + +
+[2066] +
+
+Giorgos Karafotias, Mark Hoogendoorn, and Agoston E. Eiben. + Evaluating reward definitions for parameter control. + In A. M. Mora and G. Squillero, editors, Applications of Evolutionary Computation, volume 9028 of Lecture Notes in Computer Science, pp.  667–680. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] + +
+ + +
+[2067] +
+
+Zohar Karnin, Tomer Koren, and Oren Somekh. + Almost optimal exploration in multi-armed bandits. + In S. Dasgupta and D. McAllester, editors, Proceedings of the 30th International Conference on Machine Learning, ICML 2013, volume 28, pp.  1238–1246, 2013.
+[ bib | +http ] +
+Sequential Halving, Successive Halving +
+ +
+ + +
+[2068] +
+
+Daniel Karapetyan, Andrew J. Parkes, and Thomas Stützle. + Algorithm Configuration: Learning policies for the quick termination of poor performers. + In R. Battiti, M. Brunato, I. Kotsireas, and P. M. Pardalos, editors, Learning and Intelligent Optimization, 12th International Conference, LION 12, volume 11353 of Lecture Notes in Computer Science, pp.  220–224. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] + +
+ + +
+[2069] +
+
+Giorgos Karafotias, Selmar K. Smit, and Agoston E. Eiben. + A generic approach to parameter control. + In C. Di Chio et al., editors, Applications of Evolutionary Computation, volume 7248 of Lecture Notes in Computer Science, pp.  366–375. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2070] +
+
+Narendra Karmarkar. + A new polynomial-time algorithm for linear programming. + In R. A. DeMillo, editor, Proceedings of the sixteenth annual ACM Symposium on Theory of Computing, pp.  302–311. ACM Press, 1984.
+[ bib ] + +
+ + +
+[2071] +
+
+Richard M. Karp. + Reducibility among combinatorial problems. + In R. E. Miller and W. Thatcher, James, editors, Proceedings of a symposium on the Complexity of Computer Computations, held March 20-22, 1972, at the IBM Thomas J. Watson Research Center, Yorktown Heights, New York, USA, The IBM Research Symposia Series, pp.  85–103. Springer, 1972.
+[ bib ] + +
+ + +
+[2072] +
+
+K. Katayama and H. Narihisa. + Iterated Local Search Approach using Genetic Transformation to the Traveling Salesman Problem. + In W. Banzhaf, J. M. Daida, A. E. Eiben, M. H. Garzon, V. Honavar, M. J. Jakiela, and R. E. Smith, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 1999, volume 1, pp.  321–328. Morgan Kaufmann Publishers, San Francisco, CA, 1999.
+[ bib ] + +
+ + +
+[2073] +
+
+S. A. Kauffman. + The Origins of Order. + Oxford University Press, 1993.
+[ bib ] + +
+ + +
+[2074] +
+
+Michael D. Kazantzis, Angus R. Simpson, David Kwong, and Shyh Min Tan. + A new methodology for optimizing the daily operations of a pumping plant. + In Proceedings of 2002 Conference on Water Resources Planning, Roanoke, USA, May 2002. ASCE.
+[ bib ] + +
+ + +
+[2075] +
+
+Liangjun Ke, Zuren Feng, Zongben Xu, Ke Shang, and Yonggang Wang. + A multiobjective ACO algorithm for rough feature selection. + In Circuits, Communications and System (PACCS), 2010 Second Pacific-Asia Conference on, volume 1, pp.  207–210, 2010.
+[ bib ] + +
+ + +
+[2076] +
+
+Eric Kee, Sarah Airey, and Walling Cyre. + An adaptive genetic algorithm. + In E. D. Goodman, editor, Proceedings of the 3rd Annual Conference on Genetic and Evolutionary Computation, GECCO 2001, pp.  391–397. Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib ] + +
+ + +
+[2077] +
+
+Hans Kellerer, Ulrich Pferschy, and David Pisinger. + Knapsack problems. + Springer, 2004.
+[ bib ] + +
+ + +
+[2078] +
+
+Robert E. Keller and Riccardo Poli. + Linear genetic programming of parsimonious metaheuristics. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  4508–4515, Piscataway, NJ, 2007. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2079] +
+
+Robert E. Keller and Riccardo Poli. + Cost-Benefit Investigation of a Genetic-Programming Hyperheuristic. + In E. Lutton, P. Legrand, P. Parrend, N. Monmarché, and M. Schoenauer, editors, EA 2017: Artificial Evolution, volume 10764 of Lecture Notes in Computer Science, pp.  13–24. Springer, Heidelberg, Germany, 2017.
+[ bib ] + +
+ + +
+[2080] +
+
+J. Kennedy and Russell C. Eberhart. + Particle Swarm Optimization. + In Proceedings of International Conference on Neural Networks (ICNN'95), pp.  1942–1948, Piscataway, NJ, 1995. IEEE Press.
+[ bib | +DOI ] +
+Proposed PSO +
+ +
+ + +
+[2081] +
+
+J. Kennedy and Russell C. Eberhart. + A discrete binary version of the particle swarm algorithm. + In Proceedings of the 1997 IEEE International Conference on Systems, Man, and Cybernetics, pp.  4104–4108, Piscataway, NJ, 1997. IEEE Press.
+[ bib ] + +
+ + +
+[2082] +
+
+J. Kennedy, Russell C. Eberhart, and Yuhui Shi. + Swarm Intelligence. + Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib ] + +
+ + +
+[2083] +
+
+Maurice G. Kendall. + Rank correlation methods. + Griffin, London, 1948.
+[ bib ] + +
+ + +
+[2084] +
+
+Pascal Kerschke and Heike Trautmann. + The R-package FLACCO for exploratory landscape analysis with applications to multi-objective optimization problems. + In Proceedings of the 2016 Congress on Evolutionary Computation (CEC 2016), pp.  5262–5269, Piscataway, NJ, 2016. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2085] +
+
+Pascal Kerschke, Hao Wang, Mike Preuss, Christian Grimme, André H. Deutz, Heike Trautmann, and Michael T. M. Emmerich. + Towards Analyzing Multimodality of Continuous Multiobjective Landscapes. + In J. Handl, E. Hart, P. R. Lewis, M. López-Ibáñez, G. Ochoa, and B. Paechter, editors, Parallel Problem Solving from Nature – PPSN XIV, volume 9921 of Lecture Notes in Computer Science, pp.  962–972. Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[2086] +
+
+François Chollet et al. + Keras. + https://keras.io, 2015.
+[ bib ] + +
+ + +
+[2087] +
+
+M. Kerrisk. + pthreads - POSIX Threads. + In Linux Programmer's Manual, Section 7. https://man7.org/linux/man-pages/man7/pthreads.7.html, 2021. + (Last accessed Feb 22 2023).
+[ bib ] + +
+ + +
+[2088] +
+
+V. Khare, Xin Yao, and Kalyanmoy Deb. + Performance Scaling of Multi-objective Evolutionary Algorithms. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  376–390. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2089] +
+
+M. Khichane, P. Albert, and Christine Solnon. + Integration of ACO in a Constraint Programming Language. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 6th International Conference, ANTS 2008, volume 5217 of Lecture Notes in Computer Science, pp.  84–95. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2090] +
+
+M. Khichane, P. Albert, and Christine Solnon. + An ACO-Based Reactive Framework for Ant Colony Optimization: First Experiments on Constraint Satisfaction Problems. + In T. Stützle, editor, Learning and Intelligent Optimization, Third International Conference, LION 3, volume 5851 of Lecture Notes in Computer Science, pp.  119–133. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2091] +
+
+A. R. KhudaBukhsh, Lin Xu, Holger H. Hoos, and Kevin Leyton-Brown. + SATenstein: Automatically Building Local Search SAT Solvers from Components. + In C. Boutilier, editor, Proceedings of the 21st International Joint Conference on Artificial Intelligence (IJCAI-09), pp.  517–524. AAAI Press, Menlo Park, CA, 2009.
+[ bib | +epub ] + +
+ + +
+[2092] +
+
+Youngmin Kim, Richard Allmendinger, and Manuel López-Ibáñez. + Safe Learning and Optimization Techniques: Towards a Survey of the State of the Art. + In F. Heintz, M. Milano, and B. O'Sullivan, editors, Trustworthy AI – Integrating Learning, Optimization and Reasoning. TAILOR 2020, volume 12641 of Lecture Notes in Computer Science, pp.  123–139. Springer, Cham, Switzerland, 2021.
+[ bib | +DOI ] +
+Safe learning and optimization deals with learning and + optimization problems that avoid, as much as possible, the + evaluation of non-safe input points, which are solutions, + policies, or strategies that cause an irrecoverable loss + (e.g., breakage of a machine or equipment, or life + threat). Although a comprehensive survey of safe + reinforcement learning algorithms was published in 2015, a + number of new algorithms have been proposed thereafter, and + related works in active learning and in optimization were not + considered. This paper reviews those algorithms from a number + of domains including reinforcement learning, Gaussian process + regression and classification, evolutionary computing, and + active learning. We provide the fundamental concepts on which + the reviewed algorithms are based and a characterization of + the individual algorithms. We conclude by explaining how the + algorithms are connected and suggestions for future + research. +
+ +
+ + +
+[2093] +
+
+Youngmin Kim, Richard Allmendinger, and Manuel López-Ibáñez. + Are Evolutionary Algorithms Safe Optimizers? + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, pp.  814–822. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] +
+We consider a type of constrained optimization problem, where + the violation of a constraint leads to an irrevocable loss, + such as breakage of a valuable experimental resource/platform + or loss of human life. Such problems are referred to as safe + optimization problems (SafeOPs). While SafeOPs have received + attention in the machine learning community in recent years, + there was little interest in the evolutionary computation + (EC) community despite some early attempts between 2009 and + 2011. Moreover, there is a lack of acceptable guidelines on + how to benchmark different algorithms for SafeOPs, an area + where the EC community has significant experience in. Driven + by the need for more eficient algorithms and benchmark + guidelines for SafeOPs, the objective of this paper is to + reignite the interest of the EC community in this problem + class. To achieve this we (i) provide a formal definition of + SafeOPs and contrast it to other types of optimization + problems that the EC community is familiar with, (ii) + investigate the impact of key SafeOP parameters on the + performance of selected safe optimization algorithms, (iii) + benchmark EC against state-of-the-art safe optimization + algorithms from the machine learning community, and (iv) + provide an open-source Python framework to replicate and + extend our work. +
+
+Keywords: Bayesian optimization, constrained optimization, + benchmarking, safety constraints, safe optimization +
+ +
+ + +
+[2094] +
+
+Minsu Kim, Jinkyoo Park, and Joungho Kim. + Learning Collaborative Policies to Solve NP-hard Routing Problems. + In M. Ranzato, A. Beygelzimer, Y. Dauphin, P. Liang, and J. W. Vaughan, editors, Advances in Neural Information Processing Systems 34 (NeurIPS 2021), 2021.
+[ bib | +epub ] +
+Keywords: Deep RL, TSP, prize collecting, PCTSP, CVRP, routing, + attention model +
+ +
+ + +
+[2095] +
+
+Diederik P. Kingma and Jimmy Ba. + Adam: A Method for Stochastic Optimization. + In Y. Bengio and Y. LeCun, editors, 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015.
+[ bib ] + +
+ + +
+[2096] +
+
+Joshua D. Knowles. + A summary-attainment-surface plotting method for visualizing the performance of stochastic multiobjective optimizers. + In A. Abraham and M. Paprzycki, editors, Proceedings of the 5th International Conference on Intelligent Systems Design and Applications, pp.  552–557, 2005.
+[ bib | +DOI | +supplementary material ] +
+When evaluating the performance of a stochastic optimizer it + is sometimes desirable to express performance in terms of the + quality attained in a certain fraction of sample runs. For + example, the sample median quality is the best estimator of + what one would expect to achieve in 50% of runs, and + similarly for other quantiles. In multiobjective + optimization, the notion still applies but the outcome of a + run is measured not as a scalar (i.e. the cost of the best + solution), but as an attainment surface in k-dimensional + space (where k is the number of objectives). In this paper + we report an algorithm that can be conveniently used to plot + summary attainment surfaces in any number of dimensions + (though it is particularly suited for three). A summary + attainment surface is defined as the union of all tightest + goals that have been attained (independently) in precisely + s of the runs of a sample of n runs, for any s ∈ + 1...n, and for any k. We also discuss the + computational complexity of the algorithm and give some + examples of its use. C code for the algorithm is available + from the author. +
+ +
+ + +
+[2097] +
+
+Joshua D. Knowles and David Corne. + The Pareto Archived Evolution Strategy: A New Baseline Algorithm for Multiobjective Optimisation. + In Proceedings of the 1999 Congress on Evolutionary Computation (CEC 1999), pp.  98–105, Piscataway, NJ, 1999. IEEE Press.
+[ bib ] +
+first mention of Adaptive Grid Archiving +
+ +
+ + +
+[2098] +
+
+Joshua D. Knowles and David Corne. + M-PAES: A memetic algorithm for multiobjective optimization. + In Proceedings of the 2000 Congress on Evolutionary Computation (CEC'00), pp.  325–332, Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[2099] +
+
+Joshua D. Knowles and David Corne. + On Metrics for Comparing Non-Dominated Sets. + In Proceedings of the 2002 Congress on Evolutionary Computation (CEC'02), pp.  711–716, Piscataway, NJ, 2002. IEEE Press.
+[ bib ] + +
+ + +
+[2100] +
+
+Joshua D. Knowles and David Corne. + Instance Generators and Test Suites for the Multiobjective Quadratic Assignment Problem. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  295–310. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2101] +
+
+Joshua D. Knowles and David Corne. + Bounded Pareto Archiving: Theory and Practice. + In X. Gandibleux, M. Sevaux, K. Sörensen, and V. T'Kindt, editors, Metaheuristics for Multiobjective Optimisation, volume 535 of Lecture Notes in Economics and Mathematical Systems, pp.  39–64. Springer, Berlin/Heidelberg, 2004.
+[ bib | +DOI ] + +
+ + +
+[2102] +
+
+Joshua D. Knowles and David Corne. + Memetic algorithms for multiobjective optimization: issues, methods and prospects. + In H. W. E., S. J. E., and K. N., editors, Recent Advances in Memetic Algorithms, volume 166 of Studies in Fuzziness and Soft Computing, pp.  313–352. Springer, Berlin/Heidelberg, 2005.
+[ bib | +DOI ] + +
+ + +
+[2103] +
+
+Joshua D. Knowles and David Corne. + Quantifying the Effects of Objective Space Dimension in Evolutionary Multiobjective Optimization. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  757–771. Springer, Heidelberg, Germany, 2007.
+[ bib ] +
+The scalability of EMO algorithms is an issue of significant + concern for both algorithm developers and users. A key aspect + of the issue is scalability to objective space dimension, + other things being equal. Here, we make some observations + about the efficiency of search in discrete spaces as a + function of the number of objectives, considering both + uncorrelated and correlated objective values. Efficiency is + expressed in terms of a cardinality-based + (scaling-independent) performance indicator. Considering + random sampling of the search space, we measure, empirically, + the fraction of the true PF covered after p iterations, as + the number of objectives grows, and for different + correlations. A general analytical expression for the + expected performance of random search is derived, and is + shown to agree with the empirical results. We postulate that + for even moderately large numbers of objectives, random + search will be competitive with an EMO algorithm and show + that this is the case empirically: on a function where each + objective is relatively easy for an EA to optimize (an + NK-landscape with K=2), random search compares favourably to + a well-known EMO algorithm when objective space dimension is + ten, for a range of inter-objective correlation values. The + analytical methods presented here may be useful for + benchmarking of other EMO algorithms. +
+ +
+ + +
+[2104] +
+
+Joshua D. Knowles, David Corne, and Kalyanmoy Deb. + Introduction: Problem solving, EC and EMO. + In J. D. Knowles, D. Corne, K. Deb, and D. R. Chair, editors, Multiobjective Problem Solving from Nature, Natural Computing Series, pp.  1–28. Springer, Berlin/Heidelberg, 2008.
+[ bib | +DOI ] + +
+ + +
+[2105] +
+
+Joshua D. Knowles, David Corne, and Mark Fleischer. + Bounded archiving using the Lebesgue measure. + In Proceedings of the 2003 Congress on Evolutionary Computation (CEC'03), pp.  2490–2497, Piscataway, NJ, December 2003. IEEE Press.
+[ bib ] + +
+ + +
+[2106] +
+
+Joshua D. Knowles, David Corne, and Alan P. Reynolds. + Noisy Multiobjective Optimization on a Budget of 250 Evaluations. + In M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors, Evolutionary Multi-criterion Optimization, EMO 2009, volume 5467 of Lecture Notes in Computer Science, pp.  36–50. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2107] +
+
+Joshua D. Knowles, Lothar Thiele, and Eckart Zitzler. + A tutorial on the performance assessment of stochastic multiobjective optimizers. + TIK-Report 214, Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Zürich, Switzerland, February 2006. + Revised version.
+[ bib | +epub ] + +
+ + +
+[2108] +
+
+Joshua D. Knowles, Richard A. Watson, and David Corne. + Reducing Local Optima in Single-Objective Problems by Multi-objectivization. + In E. Zitzler, K. Deb, L. Thiele, C. A. Coello Coello, and D. Corne, editors, Evolutionary Multi-criterion Optimization, EMO 2001, volume 1993 of Lecture Notes in Computer Science, pp.  269–283. Springer, Berlin/Heidelberg, 2001.
+[ bib | +DOI ] +
+Proposed multi-objectivization +
+ +
+ + +
+[2109] +
+
+Joshua D. Knowles. + Local-Search and Hybrid Evolutionary Algorithms for Pareto Optimization. + PhD thesis, University of Reading, UK, 2002.
+[ bib ] +
+(Examiners: Prof. K. Deb and Prof. K. Warwick) +
+ +
+ + +
+[2110] +
+
+Daphne Koller and Nir Friedman. + Probabilistic graphical models: principles and techniques. + MIT Press, 2009.
+[ bib ] + +
+ + +
+[2111] +
+
+Mario Koppen and Kaori Yoshida. + Visualization of Pareto-sets in evolutionary multi-objective optimization. + In 7th International Conference on Hybrid Intelligent Systems (HIS 2007), pp.  156–161. IEEE, 2007.
+[ bib ] + +
+ + +
+[2112] +
+
+P. Korošec, Jurij Šilc, K. Oblak, and F. Kosel. + The differential ant-stigmergy algorithm: an experimental evaluation and a real-world application. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  157–164, Piscataway, NJ, 2007. IEEE Press.
+[ bib ] + +
+ + +
+[2113] +
+
+P. Korošec, Jurij Šilc, and B. Robič. + Mesh-Partitioning with the Multiple Ant-Colony Algorithm. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  430–431. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2114] +
+
+Oliver Korb, Thomas Stützle, and Thomas E. Exner. + PLANTS: Application of ant colony optimization to structure-based drug design. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  247–258. Springer, Heidelberg, Germany, 2006.
+[ bib | +DOI ] + +
+ + +
+[2115] +
+
+Pekka Korhonen and Jyrki Wallenius. + Behavioral Issues in MCDM: Neglected Research Questions. + In J. Clímaco, editor, Multicriteria Analysis, pp.  412–422. Springer, Berlin/Heidelberg, 1997.
+[ bib | +DOI ] +
+Behavior decision theorists have studied human decision + making in great detail. Since the late 1960's, Einhorn, + Edwards, Kahneman, Roy, Trevsky, and others have developed + new thoeries to explain choice and decision behavior. Thus + far this behavior research has had little impact on multiple + criteria decision making (MCDM). Only a handful of + MCDM-research have critically examined the behavioral + underpinnings of our field. To improve the success of + decision tools in practice, MCDM-research should pay more + attention to the behavioral realities of decision making. In + this paper, we discuss various behavioral issues relevent for + MCDM based on our personal observations and experiments with + human subjects. The spirit of our paper is to pose questions + rather than provide definite answers. +
+
+Keywords: Aspiration Level, Decision Tool, Nondominated Solution, + Prefer Solution, Prospect Theory +
+ +
+ + +
+[2116] +
+
+Ana Kostovska, Diederick Vermetten, Carola Doerr, Sašo Džeroski, Panče Panov, and Tome Eftimov. + OPTION: optimization algorithm benchmarking ontology. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  239–240. ACM Press, New York, NY, 2021.
+[ bib ] + +
+ + +
+[2117] +
+
+Ana Kostovska, Diederick Vermetten, Sašo Džeroski, Panče Panov, Tome Eftimov, and Carola Doerr. + Using Knowledge Graphs for Performance Prediction of Modular Optimization Algorithms. + In J. a. Correia, S. Smith, and R. Qaddoura, editors, EvoApplications 2023: Applications of Evolutionary Computation, volume 13989 of Lecture Notes in Computer Science, pp.  253–268. Springer Nature, Switzerland, 2023.
+[ bib ] + +
+ + +
+[2118] +
+
+P. Kouvelis and G. Yu. + Robust discrete optimization and its applications. + Nonconvex optimization and its applications. Kluwer Academic Publishers, Dordrecht, The Netherlands, 1997.
+[ bib ] + +
+ + +
+[2119] +
+
+O. Kovářík and M. Skrbek. + Ant Colony Optimization with Castes. + In V. Kurkova-Pohlova and J. Koutnik, editors, ICANN'08: Proceedings of the 18th International Conference on Artificial Neural Networks, Part I, volume 5163 of Lecture Notes in Computer Science, pp.  435–442. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2120] +
+
+Slawomir Koziel, David Echeverría Ciaurri, and Leifur Leifsson. + Surrogate-Based Methods. + In S. Koziel and X.-S. Yang, editors, Computational Optimization, Methods and Algorithms, volume 356 of Studies in Computational Intelligence, pp.  33–59. Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[2121] +
+
+J. Koza. + Genetic Programming: On the Programming of Computers By the Means of Natural Selection. + MIT Press, Cambridge, MA, 1992.
+[ bib ] + +
+ + +
+[2122] +
+
+Oswin Krause, T. Glasmachers, Nikolaus Hansen, and Christian Igel. + Unbounded population MO-CMA-ES for the bi-objective BBOB test suite. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2016, pp.  1177–1184. ACM Press, New York, NY, 2016.
+[ bib ] +
+Keywords: archiving +
+ +
+ + +
+[2123] +
+
+Oliver Kramer, Bartek Gloger, and Andreas Goebels. + An Experimental Analysis of Evolution Strategies and Particle Swarm Optimisers Using Design of Experiments. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, pp.  674–681. ACM Press, New York, NY, 2007.
+[ bib ] + +
+ + +
+[2124] +
+
+Daniel Krajzewicz, Marek Heinrich, Michela Milano, Paolo Bellavista, Thomas Stützle, Jérôme Härri, Thrasyvoulos Spyropoulos, Robbin Blokpoel, Stefan Hausberger, and Martin Fellendorf. + COLOMBO: Investigating the Potential of V2X for Traffic Management Purposes assuming low penetration Rates. + In Proceedings of ITS Europe 2013, Dublin, Ireland, 2013.
+[ bib ] + +
+ + +
+[2125] +
+
+Daniel Krajzewicz, Andreas Leich, Robbin Blokpoel, Michela Milano, and Thomas Stützle. + COLOMBO: Exploiting Vehicular Communications at Low Equipment Rates for Traffic Management Purposes. + In T. Schulze, B. Müller, and G. Meyer, editors, Advanced Microsystems for Automotive Applications 2015: Smart Systems for Green and Automated Driving, pp.  117–130. Springer International Publishing, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[2126] +
+
+Jakob Krarup and Peter Mark Pruzan. + Computer-aided Layout Design. + In M. L. Balinski and C. Lemarechal, editors, Mathematical Programming in Use, volume 9 of Mathematical Programming Studies, pp.  75–94. Springer, Berlin/Heidelberg, 1978.
+[ bib ] + +
+ + +
+[2127] +
+
+Johannes Krettek, Jan Braun, Frank Hoffmann, and Torsten Bertram. + Interactive Incorporation of User Preferences in Multiobjective Evolutionary Algorithms. + In J. Mehnen, M. Köppen, A. Saad, and A. Tiwari, editors, Applications of Soft Computing, volume 58 of Advances in Intelligent and Soft Computing, pp.  379–388. Springer, Berlin/Heidelberg, 2009.
+[ bib ] + +
+ + +
+[2128] +
+
+Johannes Krettek, Jan Braun, Frank Hoffmann, and Torsten Bertram. + Preference Modeling and Model Management for Interactive Multi-objective Evolutionary Optimization. + In E. Hüllermeier, R. Kruse, and F. Hoffmann, editors, Information Processing and Management of Uncertainty, 13th International Conference, IPMU2010, volume 6178 of Lecture Notes in Artificial Intelligence, pp.  574–583. Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2129] +
+
+William H. Kruskal and Judith M. Tanur. + Linear Hypotheses, volume 1. + Free Press, 1978.
+[ bib ] + +
+ + +
+[2130] +
+
+S. Kukkonen and J. Lampinen. + GDE3: the third evolution step of generalized differential evolution. + In Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), pp.  443–450, Piscataway, NJ, September 2005. IEEE Press.
+[ bib ] + +
+ + +
+[2131] +
+
+Ravi Kumar and Sergei Vassilvitskii. + Generalized Distances between Rankings. + In M. Rappa, P. Jones, J. Freire, and S. Chakrabarti, editors, Proceedings of the 19th International Conference on World Wide Web, WWW 2010. ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2132] +
+
+Frank Kursawe. + A variant of evolution strategies for vector optimization. + In H.-P. Schwefel and R. Männer, editors, Parallel Problem Solving from Nature – PPSN I, pp.  193–197. Springer, Berlin/Heidelberg, 1991.
+[ bib | +DOI ] +
+Proposed KUR benchmark +
+ +
+ + +
+[2133] +
+
+Benjamin Lacroix, Daniel Molina, and Francisco Herrera. + Dynamically updated region based memetic algorithm for the 2013 CEC Special Session and Competition on Real Parameter Single Objective Optimization. + In Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), pp.  1945–1951, Piscataway, NJ, 2013. IEEE Press.
+[ bib ] + +
+ + +
+[2134] +
+
+R. M. Lark and D. J. Lapworth. + A new statistic to express the uncertainty of kriging predictions for purposes of survey planning. + In EGU General Assembly Conference Abstracts, May 2014.
+[ bib | +http ] + +
+ + +
+[2135] +
+
+Pedro Larrañaga and José A. Lozano. + Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation. + Kluwer Academic Publishers, Boston, MA, 2002.
+[ bib ] + +
+ + +
+[2136] +
+
+Craig Larman. + Applying UML and Patterns: An Introduction to Object-Oriented Analysis and Design and Iterative Development. + Prentice Hall, Englewood Cliffs, NJ, 3rd edition, 2004.
+[ bib ] + +
+ + +
+[2137] +
+
+Marco Laumanns, Lothar Thiele, Eckart Zitzler, and Kalyanmoy Deb. + Archiving with guaranteed convergence and diversity in multi-objective optimization. + In W. B. Langdon et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2002, pp.  439–447. Morgan Kaufmann Publishers, San Francisco, CA, 2002.
+[ bib ] + +
+ + +
+[2138] +
+
+Marco Laumanns and Rico Zenklusen. + Stochastic convergence of random search methods to fixed size Pareto front approximations. + (submitted), November 2010.
+[ bib ] +
+Published as [791]. Keep this reference for historical reasons. +
+ +
+ + +
+[2139] +
+
+Marco Laumanns, Eckart Zitzler, and Lothar Thiele. + A unified model for multi-objective evolutionary algorithms with elitism. + In Proceedings of the 2000 Congress on Evolutionary Computation (CEC'00), pp.  46–53, Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[2140] +
+
+E. L. Lawler, J. K. Lenstra, A. H. G. Rinnooy Kan, and D. B. Shmoys. + The Traveling Salesman Problem. + John Wiley & Sons, Chichester, UK, 1985.
+[ bib ] + +
+ + +
+[2141] +
+
+Guillermo Leguizamón and Enrique Alba. + Ant Colony Based Algorithms for Dynamic Optimization Problems. + In E. Alba, A. Nakib, and P. Siarry, editors, Metaheuristics for Dynamic Optimization, volume 433 of Studies in Computational Intelligence, pp.  189–210. Springer, Berlin/Heidelberg, 2013.
+[ bib | +DOI ] + +
+ + +
+[2142] +
+
+Guillermo Leguizamón and Zbigniew Michalewicz. + A New Version of Ant System for Subset Problems. + In Proceedings of the 1999 Congress on Evolutionary Computation (CEC 1999), pp.  1459–1464, Piscataway, NJ, 1999. IEEE Press.
+[ bib ] + +
+ + +
+[2143] +
+
+R. J. Lempert, S. Popper, and Steven C. Bankes. + Shaping the Next One Hundred Years: New Methods for Quantitative, Long Term Policy Analysis. + RAND, 2003.
+[ bib ] + +
+ + +
+[2144] +
+
+L. Lessing, Irina Dumitrescu, and Thomas Stützle. + A Comparison Between ACO Algorithms for the Set Covering Problem. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  1–12. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2145] +
+
+Rhyd M. R. Lewis. + A Guide to Graph Colouring: Algorithms and Applications. + Springer, Cham, Switzerland, 2016.
+[ bib | +DOI ] +
+Supplementary material available at [2146] +
+ +
+ + +
+[2146] +
+
+Rhyd M. R. Lewis. + Suite of Graph Colouring Algorithms – Supplementary Material to the Book “A Guide to Graph Colouring: Algorithms and Applications”. + http://rhydlewis.eu/resources/gCol.zip, 2016.
+[ bib ] + +
+ + +
+[2147] +
+
+Kevin Leyton-Brown, Eugene Nudelman, Galen Andrew, Jim McFadden, and Yoav Shoham. + A Portfolio Approach to Algorithm Selection. + In G. Gottlob and T. Walsh, editors, Proceedings of the 18th International Joint Conference on Artificial Intelligence (IJCAI-03), pp.  1542–1543. Morgan Kaufmann Publishers, 2003.
+[ bib | +epub ] +
+First example of modern algorithm selection for + optimisation? +
+ +
+ + +
+[2148] +
+
+Kevin Leyton-Brown, Eugene Nudelman, and Yoav Shoham. + Learning the Empirical Hardness of Optimization Problems: The Case of Combinatorial Auctions. + In P. van Hentenryck, editor, Principles and Practice of Constraint Programming, CP 2002, Lecture Notes in Computer Science, pp.  556–572. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2149] +
+
+Kevin Leyton-Brown, M. Pearson, and Y. Shoham. + Towards a Universal Test Suite for Combinatorial Auction Algorithms. + In A. Jhingran et al., editors, ACM Conference on Electronic Commerce (EC-00), pp.  66–76. ACM Press, New York, NY, 2000.
+[ bib | +DOI ] +
+CPLEX-regions200 benchmark set, + http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/results.html +
+ +
+ + +
+[2150] +
+
+Bingdong Li, Jinlong Li, Ke Tang, and Xin Yao. + An Improved Two Archive Algorithm for Many-Objective Optimization. + In Proceedings of the 2014 Congress on Evolutionary Computation (CEC 2014), pp.  2869–2876, Piscataway, NJ, 2014. IEEE Press.
+[ bib ] + +
+ + +
+[2151] +
+
+Z. Li, Y. Wang, J. Yu, Y. Zhang, and X. Li. + A Novel Cloud-Based Fuzzy Self-Adaptive Ant Colony System. + In ICNC'08: Proceedings of the 2008 Fourth International Conference on Natural Computation, volume 7, pp.  460–465, Washington, DC, 2008. IEEE Computer Society.
+[ bib ] + +
+ + +
+[2152] +
+
+Miqing Li, Shengxiang Yang, and Xiaohui Liu. + A performance comparison indicator for Pareto front approximations in many-objective optimization. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  703–710. ACM Press, New York, NY, 2015.
+[ bib ] +
+Proposed PCI indicator +
+ +
+ + +
+[2153] +
+
+Miqing Li, Shengxiang Yang, Xiaohui Liu, and Ruimin Shen. + A Comparative Study on Evolutionary Algorithms for Many-Objective Optimization. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  261–275. Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2154] +
+
+Miqing Li and Xin Yao. + An Empirical Investigation of the Optimality and Monotonicity Properties of Multiobjective Archiving Methods. + In K. Deb, E. D. Goodman, C. A. Coello Coello, K. Klamroth, K. Miettinen, S. Mostaghim, and P. Reed, editors, Evolutionary Multi-criterion Optimization, EMO 2019, volume 11411 of Lecture Notes in Computer Science, pp.  15–26. Springer International Publishing, Cham, Switzerland, 2019.
+[ bib | +DOI ] + +
+ + +
+[2155] +
+
+Longmei Li, Iryna Yevseyeva, Vitor Basto-Fernandes, Heike Trautmann, Ning Jing, and Michael T. M. Emmerich. + Building and using an ontology of preference-based multiobjective evolutionary algorithms. + In H. Trautmann, G. Rudolph, K. Klamroth, O. Schütze, M. M. Wiecek, Y. Jin, and C. Grimme, editors, Evolutionary Multi-criterion Optimization, EMO 2017, volume 10173 of Lecture Notes in Computer Science, pp.  406–421. Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[2156] +
+
+Tianjun Liao, Marco A. Montes de Oca, Doǧan Aydın, Thomas Stützle, and Marco Dorigo. + An Incremental Ant Colony Algorithm with Local Search for Continuous Optimization. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  125–132. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2157] +
+
+Tianjun Liao, Marco A. Montes de Oca, and Thomas Stützle. + Tuning Parameters across Mixed Dimensional Instances: A Performance Scalability Study of Sep-G-CMA-ES. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2011, pp.  703–706. ACM Press, New York, NY, 2011.
+[ bib ] +
+Workshop on Scaling Behaviours of Landscapes, + Parameters and Algorithms +
+ +
+ + +
+[2158] +
+
+Tianjun Liao and Thomas Stützle. + Benchmark results for a simple hybrid algorithm on the CEC 2013 benchmark set for real-parameter optimization. + In Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), pp.  1938–1944, Piscataway, NJ, 2013. IEEE Press.
+[ bib ] + +
+ + +
+[2159] +
+
+Tianjun Liao. + Population-based Heuristic Algorithms for Continuous and Mixed Discrete-Continuous Optimization Problem. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2013.
+[ bib ] + +
+ + +
+[2160] +
+
+Arnaud Liefooghe, Bilel Derbel, Sébastien Verel, Hernán E. Aguirre, and Kiyoshi Tanaka. + Towards Landscape-Aware Automatic Algorithm Configuration: Preliminary Experiments on Neutral and Rugged Landscapes. + In B. Hu and M. López-Ibáñez, editors, Proceedings of EvoCOP 2017 – 17th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 10197 of Lecture Notes in Computer Science, pp.  215–232. Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[2161] +
+
+Arnaud Liefooghe, Bilel Derbel, Sébastien Verel, Manuel López-Ibáñez, Hernán E. Aguirre, and Kiyoshi Tanaka. + On Pareto Local Optimal Solutions Networks. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11102 of Lecture Notes in Computer Science, pp.  232–244. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI ] + +
+ + +
+[2162] +
+
+Arnaud Liefooghe and Manuel López-Ibáñez. + Many-objective (Combinatorial) Optimization is Easy. + In S. Silva and L. Paquete, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2023, pp.  704–712. ACM Press, New York, NY, 2023.
+[ bib | +DOI ] +
+It is a common held assumption that problems with many + objectives are harder to optimize than problems with two or + three objectives. In this paper, we challenge this assumption + and provide empirical evidence that increasing the number of + objectives tends to reduce the difficulty of the landscape + being optimized. Of course, increasing the number of + objectives brings about other challenges, such as an increase + in the computational effort of many operations, or the memory + requirements for storing non-dominated solutions. More + precisely, we consider a broad range of multi- and + many-objective combinatorial benchmark problems, and we + measure how the number of objectives impacts the dominance + relation among solutions, the connectedness of the Pareto + set, and the landscape multimodality in terms of local + optimal solutions and sets. Our analysis shows the limit + behavior of various landscape features when adding more + objectives to a problem. Our conclusions do not contradict + previous observations about the inability of + Pareto-optimality to drive search, but we explain these + observations from a different perspective. Our findings have + important implications for the design and analysis of + many-objective optimization algorithms. +
+
+ISBN: 9798400701191 +
+ +
+ + +
+[2163] +
+
+Arnaud Liefooghe, Manuel López-Ibáñez, Luís Paquete, and Sébastien Verel. + Dominance, Epsilon, and Hypervolume Local Optimal Sets in Multi-objective Optimization, and How to Tell the Difference. + In H. E. Aguirre and K. Takadama, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2018, pp.  324–331. ACM Press, New York, NY, 2018.
+[ bib | +DOI ] + +
+ + +
+[2164] +
+
+Arnaud Liefooghe, Salma Mesmoudi, Jérémie Humeau, Laetitia Jourdan, and El-Ghazali Talbi. + A Study on Dominance-based Local Search Approaches for Multiobjective Combinatorial Optimization. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2009, volume 5752 of Lecture Notes in Computer Science, pp.  120–124. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2165] +
+
+Arnaud Liefooghe, Luís Paquete, Marco Simōes, and José Rui Figueira. + Connectedness and Local Search for Bicriteria Knapsack Problems. + In P. Merz and J.-K. Hao, editors, Proceedings of EvoCOP 2011 – 11th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 6622 of Lecture Notes in Computer Science, pp.  48–59. Springer, Heidelberg, Germany, 2011.
+[ bib | +DOI ] + +
+ + +
+[2166] +
+
+Arnaud Liefooghe, Bilel Derbel, Sébastien Verel, Hernán E. Aguirre, and Kiyoshi Tanaka. + What Makes an Instance Difficult for Black-box 0–1 Evolutionary Multiobjective Optimizers? + In P. Legrand et al., editors, Artificial Evolution: 11th International Conference, Evolution Artificielle, EA, 2013, volume 8752 of Lecture Notes in Computer Science, pp.  3–15. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2167] +
+
+Arnaud Liefooghe, Sébastien Verel, Benjamin Lacroix, Alexandru-Ciprian Zavoianu, and John McCall. + Landscape features and automated algorithm selection for multi-objective interpolated continuous optimisation problems. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  421–429. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2168] +
+
+Arnaud Liefooghe, Sébastien Verel, Luís Paquete, and Jin-Kao Hao. + Experiments on Local Search for Bi-objective Unconstrained Binary Quadratic Programming. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  171–186. Springer, Heidelberg, Germany, 2015.
+[ bib ] +
+This article reports an experimental analysis on stochastic + local search for approximating the Pareto set of bi-objective + unconstrained binary quadratic programming problems. First, + we investigate two scalarizing strategies that iteratively + identify a high-quality solution for a sequence of + sub-problems. Each sub-problem is based on a static or + adaptive definition of weighted-sum aggregation coefficients, + and is addressed by means of a state-of-the-art + single-objective tabu search procedure. Next, we design a + Pareto local search that iteratively improves a set of + solutions based on a neighborhood structure and on the Pareto + dominance relation. At last, we hybridize both classes of + algorithms by combining a scalarizing and a Pareto local + search in a sequential way. A comprehensive experimental + analysis reveals the high performance of the proposed + approaches, which substantially improve upon previous + best-known solutions. Moreover, the obtained results show the + superiority of the hybrid algorithm over non-hybrid ones in + terms of solution quality, while requiring a competitive + computational cost. In addition, a number of structural + properties of the problem instances allow us to explain the + main difficulties that the different classes of local search + algorithms have to face. +
+ +
+ + +
+[2169] +
+
+David J. Lilja. + Measuring Computer Performance: A Practitioner's Guide. + Cambridge University Press, 2000.
+[ bib | +DOI ] +
+Measuring Computer Performance sets out the fundamental + techniques used in analyzing and understanding the + performance of computer systems. Throughout the book, the + emphasis is on practical methods of measurement, simulation, + and analytical modeling. The author discusses performance + metrics and provides detailed coverage of the strategies used + in benchmark programmes. He gives intuitive explanations of + the key statistical tools needed to interpret measured + performance data. He also describes the general 'design of + experiments' technique, and shows how the maximum amount of + information can be obtained for the minimum effort. The book + closes with a chapter on the technique of queueing + analysis. Appendices listing common probability distributions + and statistical tables are included, along with a glossary of + important technical terms. This practically-oriented book + will be of great interest to anyone who wants a detailed, yet + intuitive, understanding of computer systems performance + analysis. +
+ +
+ + +
+[2170] +
+
+Marius Thomas Lindauer, Holger H. Hoos, Frank Hutter, and Torsten Schaub. + AutoFolio: Algorithm Configuration for Algorithm Selection. + In B. Bonet and S. Koenig, editors, Proceedings of the AAAI Conference on Artificial Intelligence. AAAI Press, 2015.
+[ bib ] + +
+ + +
+[2171] +
+
+W. Ling and H. Luo. + An Adaptive Parameter Control Strategy for Ant Colony Optimization. + In CIS'07: Proceedings of the 2007 International Conference on Computational Intelligence and Security, pp.  142–146, Washington, DC, 2007. IEEE Computer Society.
+[ bib ] + +
+ + +
+[2172] +
+
+Innovation 24. + LocalSolver. + http://www.localsolver.com, 2016. + Last visited, August 15, 2016.
+[ bib ] + +
+ + +
+[2173] +
+
+Andrea Lodi and Andrea Tramontani. + Performance Variability in Mixed-Integer Programming. + In H. Topaluglu, editor, Theory Driven by Influential Applications, pp.  1–12. INFORMS, 2013.
+[ bib ] + +
+ + +
+[2174] +
+
+Andrea Lodi, Silvano Martello, and Daniele Vigo. + Two- and Three-Dimensional Bin Packing – Source Code of TSpack. + https://site.unibo.it/operations-research/en/research/library-of-codes-and-instances-1/tspack-tar.gz/@@download/file/TSpack.tar.gz, 2004.
+[ bib ] + +
+ + +
+[2175] +
+
+Po-Ling Loh and Sebastian Nowozin. + Faster Hoeffding Racing: Bernstein Races via Jackknife Estimates. + In S. Jain, R. Munos, F. Stephan, and T. Zeugmann, editors, Proceedings of Algorithmic Learning Theory, volume 8139 of Lecture Notes in Computer Science, pp.  203–217. Springer, Berlin, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2176] +
+
+Manuel López-Ibáñez. + High Performance Ant Colony Optimisation of the Pump Scheduling Problem. + In P. Alberigo, G. Erbacci, F. Garofalo, and S. Monfardini, editors, Science and Sumpercomputing in Europe, pp.  371–375. CINECA, 2007.
+[ bib ] + +
+ + +
+[2177] +
+
+Manuel López-Ibáñez and Christian Blum. + Beam-ACO Based on Stochastic Sampling: A Case Study on the TSP with Time Windows. + Technical Report LSI-08-28, Department LSI, Universitat Politècnica de Catalunya, 2008. + Extended version published in Computers & Operations Research [849].
+[ bib ] + +
+ + +
+[2178] +
+
+Manuel López-Ibáñez, Christian Blum, Dhananjay Thiruvady, Andreas T. Ernst, and Bernd Meyer. + Beam-ACO based on stochastic sampling for makespan optimization concerning the TSP with time windows. + In C. Cotta and P. Cowling, editors, Proceedings of EvoCOP 2009 – 9th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 5482 of Lecture Notes in Computer Science, pp.  97–108. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2179] +
+
+Manuel López-Ibáñez and Christian Blum. + Beam-ACO Based on Stochastic Sampling: A Case Study on the TSP with Time Windows. + In T. Stützle, editor, Learning and Intelligent Optimization, Third International Conference, LION 3, volume 5851 of Lecture Notes in Computer Science, pp.  59–73. Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2180] +
+
+Manuel López-Ibáñez, Francisco Chicano, and Rodrigo Gil-Merino. + The Asteroid Routing Problem: A Benchmark for Expensive Black-Box Permutation Optimization. + In J. L. Jiménez Laredo et al., editors, EvoApplications 2022: Applications of Evolutionary Computation, volume 13224 of Lecture Notes in Computer Science, pp.  124–140. Springer Nature, Switzerland, 2022.
+[ bib | +DOI | +epub | +supplementary material ] +
+Inspired by the recent 11th Global Trajectory Optimisation + Competition, this paper presents the asteroid routing problem + (ARP) as a realistic benchmark of algorithms for expensive + bound-constrained black-box optimization in permutation + space. Given a set of asteroids' orbits and a departure + epoch, the goal of the ARP is to find the optimal sequence + for visiting the asteroids, starting from Earth's orbit, in + order to minimize both the cost, measured as the sum of the + magnitude of velocity changes required to complete the trip, + and the time, measured as the time elapsed from the departure + epoch until visiting the last asteroid. We provide + open-source code for generating instances of arbitrary sizes + and evaluating solutions to the problem. As a preliminary + analysis, we compare the results of two methods for expensive + black-box optimization in permutation spaces, namely, + Combinatorial Efficient Global Optimization (CEGO), a + Bayesian optimizer based on Gaussian processes, and + Unbalanced Mallows Model (UMM), an estimation-of-distribution + algorithm based on probabilistic Mallows models. We + investigate the best permutation representation for each + algorithm, either rank-based or order-based. Moreover, we + analyze the effect of providing a good initial solution, + generated by a greedy nearest neighbor heuristic, on the + performance of the algorithms. The results suggest directions + for improvements in the algorithms being compared. +
+
+Keywords: Spacecraft Trajectory Optimization, Unbalanced Mallows Model, + Combinatorial Efficient Global Optimization, Estimation of + Distribution Algorithms, Bayesian Optimization +
+ +
+ + +
+[2181] +
+
+Manuel López-Ibáñez, Jérémie Dubois-Lacoste, Leslie Pérez Cáceres, Thomas Stützle, and Mauro Birattari. + The irace Package: Iterated Racing for Automatic Algorithm Configuration (Supplementary Material). + http://iridia.ulb.ac.be/supp/IridiaSupp2016-003, 2016.
+[ bib ] + +
+ + +
+[2182] +
+
+Manuel López-Ibáñez, Jérémie Dubois-Lacoste, Thomas Stützle, and Mauro Birattari. + The irace package, Iterated Race for Automatic Algorithm Configuration. + Technical Report TR/IRIDIA/2011-004, IRIDIA, Université Libre de Bruxelles, Belgium, 2011. + Published in Operations Research Perspectives [853].
+[ bib | +http ] + +
+ + +
+[2183] +
+
+Manuel López-Ibáñez and Joshua D. Knowles. + Machine Decision Makers as a Laboratory for Interactive EMO. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part II, volume 9019 of Lecture Notes in Computer Science, pp.  295–309. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] +
+A key challenge, perhaps the central challenge, of + multi-objective optimization is how to deal with candidate + solutions that are ultimately evaluated by the hidden or + unknown preferences of a human decision maker (DM) who + understands and cares about the optimization problem. + Alternative ways of addressing this challenge exist but + perhaps the favoured one currently is the interactive + approach (proposed in various forms). Here, an evolutionary + multi-objective optimization algorithm (EMOA) is controlled + by a series of interactions with the DM so that preferences + can be elicited and the direction of search controlled. MCDM + has a key role to play in designing and evaluating these + approaches, particularly in testing them with real DMs, but + so far quantitative assessment of interactive EMOAs has been + limited. In this paper, we propose a conceptual framework + for this problem of quantitative assessment, based on the + definition of machine decision makers (machine DMs), made + somewhat realistic by the incorporation of various + non-idealities. The machine DM proposed here draws from + earlier models of DM biases and inconsistencies in the MCDM + literature. As a practical illustration of our approach, we + use the proposed machine DM to study the performance of an + interactive EMOA, and discuss how this framework could help + in the evaluation and development of better interactive + EMOAs. +
+ +
+ + +
+[2184] +
+
+Manuel López-Ibáñez, Joshua D. Knowles, and Marco Laumanns. + On Sequential Online Archiving of Objective Vectors. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  46–60. Springer, Berlin/Heidelberg, 2011.
+[ bib | +DOI ] +
+In this paper, we examine the problem of maintaining + an approximation of the set of nondominated points + visited during a multiobjective optimization, a + problem commonly known as archiving. Most of the + currently available archiving algorithms are + reviewed, and what is known about their convergence + and approximation properties is summarized. The main + scenario considered is the restricted case where the + archive must be updated online as points are + generated one by one, and at most a fixed number of + points are to be stored in the archive at any one + time. In this scenario, the better-monotonicity of + an archiving algorithm is proposed as a weaker, but + more practical, property than negative efficiency + preservation. This paper shows that + hypervolume-based archivers and a recently proposed + multi-level grid archiver have this property. On the + other hand, the archiving methods used by SPEA2 and + NSGA-II do not, and they may better-deteriorate with + time. The better-monotonicity property has meaning + on any input sequence of points. We also classify + archivers according to limit properties, + i.e. convergence and approximation properties of the + archiver in the limit of infinite (input) samples + from a finite space with strictly positive + generation probabilities for all points. This paper + establishes a number of research questions, and + provides the initial framework and analysis for + answering them. +
+
+Revised version available at http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-001.pdf +
+ +
+ + +
+[2185] +
+
+Manuel López-Ibáñez, Tianjun Liao, and Thomas Stützle. + On the anytime behavior of IPOP-CMA-ES. + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part I, volume 7491 of Lecture Notes in Computer Science, pp.  357–366. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2186] +
+
+Manuel López-Ibáñez, Tianjun Liao, and Thomas Stützle. + On the anytime behavior of IPOP-CMA-ES: Supplementary material. + https://iridia.ulb.ac.be/supp/IridiaSupp2012-010/IridiaSupp2012-010.pdf, 2012.
+[ bib ] + +
+ + +
+[2187] +
+
+Manuel López-Ibáñez, Arnaud Liefooghe, and Sébastien Verel. + Local Optimal Sets and Bounded Archiving on Multi-objective NK-Landscapes with Correlated Objectives. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  621–630. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] +
+The properties of local optimal solutions in multi-objective + combinatorial optimization problems are crucial for the + effectiveness of local search algorithms, particularly when + these algorithms are based on Pareto dominance. Such local + search algorithms typically return a set of mutually + nondominated Pareto local optimal (PLO) solutions, that is, a + PLO-set. This paper investigates two aspects of PLO-sets by + means of experiments with Pareto local search (PLS). First, + we examine the impact of several problem characteristics on + the properties of PLO-sets for multi-objective NK-landscapes + with correlated objectives. In particular, we report that + either increasing the number of objectives or decreasing the + correlation between objectives leads to an exponential + increment on the size of PLO-sets, whereas the variable + correlation has only a minor effect. Second, we study the + running time and the quality reached when using bounding + archiving methods to limit the size of the archive handled by + PLS, and thus, the maximum size of the PLO-set found. We + argue that there is a clear relationship between the running + time of PLS and the difficulty of a problem instance. +
+ +
+ + +
+[2188] +
+
+Manuel López-Ibáñez, Franco Mascia, Marie-Eléonore Marmion, and Thomas Stützle. + Automatic Design of a Hybrid Iterated Local Search for the Multi-Mode Resource-Constrained Multi-Project Scheduling Problem. + In G. Kendall, G. Vanden Berghe, and B. McCollum, editors, Multidisciplinary International Conference on Scheduling: Theory and Applications (MISTA 2013), pp.  1–6, Gent, Belgium, 2013.
+[ bib | +epub ] + +
+ + +
+[2189] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + On the Design of ACO for the Biobjective Quadratic Assignment Problem. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  214–225. Springer, Heidelberg, Germany, 2004.
+[ bib | +DOI ] + +
+ + +
+[2190] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + Hybrid Population-based Algorithms for the Bi-objective Quadratic Assignment Problem. + Technical Report AIDA–04–11, FG Intellektik, FB Informatik, TU Darmstadt, December 2004. + Published in Journal of Mathematical Modelling and Algorithms [855].
+[ bib ] + +
+ + +
+[2191] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + Exploratory Analysis of Stochastic Local Search Algorithms in Biobjective Optimization. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  209–222. Springer, Berlin, Germany, 2010.
+[ bib | +DOI ] +
+This chapter introduces two Perl programs that + implement graphical tools for exploring the + performance of stochastic local search algorithms + for biobjective optimization problems. These tools + are based on the concept of the empirical attainment + function (EAF), which describes the probabilistic + distribution of the outcomes obtained by a + stochastic algorithm in the objective space. In + particular, we consider the visualization of + attainment surfaces and differences between the + first-order EAFs of the outcomes of two + algorithms. This visualization allows us to identify + certain algorithmic behaviors in a graphical way. + We explain the use of these visualization tools and + illustrate them with examples arising from + practice. +
+ +
+ + +
+[2192] +
+
+Manuel López-Ibáñez, Luís Paquete, and Thomas Stützle. + EAF Graphical Tools. + http://lopez-ibanez.eu/eaftools, 2010. + These tools are described in the book chapter “Exploratory analysis of stochastic local search algorithms in biobjective optimization” [2191].
+[ bib ] +
+Please cite the book chapter, not this. +
+ +
+ + +
+[2193] +
+
+Manuel López-Ibáñez, Leslie Pérez Cáceres, Jérémie Dubois-Lacoste, Thomas Stützle, and Mauro Birattari. + The irace package: User Guide. + Technical Report TR/IRIDIA/2016-004, IRIDIA, Université Libre de Bruxelles, Belgium, 2016.
+[ bib | +http ] + +
+ + +
+[2194] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Parallel Optimisation Of Pump Schedules With A Thread-Safe Variant Of EPANET Toolkit. + In J. E. van Zyl, A. A. Ilemobade, and H. E. Jacobs, editors, Proceedings of the 10th Annual Water Distribution Systems Analysis Conference (WDSA 2008). ASCE, August 2008.
+[ bib | +DOI ] + +
+ + +
+[2195] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Solving Optimal Pump Control Problem using Max-Min Ant System. + In D. Thierens et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2007, volume 1, p.  176. ACM Press, New York, NY, 2007.
+[ bib | +DOI ] + +
+ + +
+[2196] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Multi-objective Optimisation of the Pump Scheduling Problem using SPEA2. + In Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), volume 1, pp.  435–442, Piscataway, NJ, September 2005. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2197] +
+
+Manuel López-Ibáñez, T. Devi Prasad, and Ben Paechter. + Optimal Pump Scheduling: Representation and Multiple Objectives. + In D. A. Savic, G. A. Walters, R. King, and S. Thiam-Khu, editors, Proceedings of the Eighth International Conference on Computing and Control for the Water Industry (CCWI 2005), volume 1, pp.  117–122, University of Exeter, UK, September 2005.
+[ bib ] + +
+ + +
+[2198] +
+
+Manuel López-Ibáñez and Thomas Stützle. + An Analysis of Algorithmic Components for Multiobjective Ant Colony Optimization: A Case Study on the Biobjective TSP. + In P. Collet, N. Monmarché, P. Legrand, M. Schoenauer, and E. Lutton, editors, Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009, volume 5975 of Lecture Notes in Computer Science, pp.  134–145. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2199] +
+
+Manuel López-Ibáñez and Thomas Stützle. + Automatic Configuration of Multi-Objective ACO Algorithms. + In M. Dorigo et al., editors, Swarm Intelligence, 7th International Conference, ANTS 2010, volume 6234 of Lecture Notes in Computer Science, pp.  95–106. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] +
+In the last few years a significant number of ant + colony optimization (ACO) algorithms have been + proposed for tackling multi-objective optimization + problems. In this paper, we propose a software + framework that allows to instantiate the most + prominent multi-objective ACO (MOACO) + algorithms. More importantly, the flexibility of + this MOACO framework allows the application of + automatic algorithm configuration techniques. The + experimental results presented in this paper show + that such an automatic configuration of MOACO + algorithms is highly desirable, given that our + automatically configured algorithms clearly + outperform the best performing MOACO algorithms that + have been proposed in the literature. As far as we + are aware, this paper is also the first to apply + automatic algorithm configuration techniques to + multi-objective stochastic local search algorithms. +
+ +
+ + +
+[2200] +
+
+Manuel López-Ibáñez and Thomas Stützle. + The impact of design choices of multi-objective ant colony optimization algorithms on performance: An experimental study on the biobjective TSP. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  71–78. ACM Press, New York, NY, 2010.
+[ bib | +DOI ] +
+Over the last few years, there have been a number of + proposals of ant colony optimization (ACO) + algorithms for tackling multiobjective combinatorial + optimization problems. These proposals adapt ACO + concepts in various ways, for example, some use + multiple pheromone matrices and multiple heuristic + matrices and others use multiple ant colonies.
In + this article, we carefully examine several of the + most prominent of these proposals. In particular, we + identify commonalities among the approaches by + recasting the original formulation of the algorithms + in different terms. For example, several proposals + described in terms of multiple colonies can be cast + equivalently using a single ant colony, where ants + use different weights for aggregating the pheromone + and/or the heuristic information. We study + algorithmic choices for the various proposals and we + identify previously undetected trade-offs in their + performance. +
+ +
+ + +
+[2201] +
+
+Manuel López-Ibáñez and Thomas Stützle. + The impact of design choices of multi-objective ant colony optimization algorithms on performance: An experimental study on the biobjective TSP. + http://iridia.ulb.ac.be/supp/IridiaSupp2010-003/, 2010. + Supplementary material of [2200].
+[ bib ] + +
+ + +
+[2202] +
+
+Manuel López-Ibáñez and Thomas Stützle. + The Automatic Design of Multi-Objective Ant Colony Optimization Algorithms: Supplementary material, 2011.
+[ bib | +http ] + +
+ + +
+[2203] +
+
+Manuel López-Ibáñez and Thomas Stützle. + An experimental analysis of design choices of multi-objective ant colony optimization algorithms: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2012-006/, 2012.
+[ bib ] + +
+ + +
+[2204] +
+
+Manuel López-Ibáñez, Thomas Stützle, and Marco Dorigo. + Ant Colony Optimization: A Component-Wise Overview. + In R. Martí, P. M. Pardalos, and M. G. C. Resende, editors, Handbook of Heuristics, pp.  371–407. Springer International Publishing, 2018.
+[ bib | +DOI | +supplementary material ] +
+Proposed ACOTSPQAP software +
+ +
+ + +
+[2205] +
+
+Manuel López-Ibáñez. + Multi-objective Ant Colony Optimization. + Diploma thesis, Intellectics Group, Computer Science Department, Technische Universität Darmstadt, Germany, 2004.
+[ bib ] + +
+ + +
+[2206] +
+
+Manuel López-Ibáñez. + Operational Optimisation of Water Distribution Networks. + PhD thesis, School of Engineering and the Built Environment, Edinburgh Napier University, UK, 2009.
+[ bib | +http ] + +
+ + +
+[2207] +
+
+Ilya Loshchilov, Marc Schoenauer, and Michèle Sebag. + Alternative Restart Strategies for CMA-ES. + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part I, volume 7491 of Lecture Notes in Computer Science, pp.  296–305. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2208] +
+
+A. V. Lotov and Kaisa Miettinen. + Visualizing the Pareto Frontier. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science, pp.  213–243. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2209] +
+
+Helena R. Lourenço, Olivier Martin, and Thomas Stützle. + Iterated Local Search. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  321–353. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib | +DOI ] + +
+ + +
+[2210] +
+
+Helena R. Lourenço, Olivier Martin, and Thomas Stützle. + Iterated Local Search: Framework and Applications. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, chapter 9, pp.  363–397. Springer, New York, NY, 2nd edition, 2010.
+[ bib | +DOI ] + +
+ + +
+[2211] +
+
+Helena R. Lourenço, Olivier Martin, and Thomas Stützle. + Iterated Local Search: Framework and Applications. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 272 of International Series in Operations Research & Management Science, chapter 5, pp.  129–168. Springer, 2019.
+[ bib | +DOI ] + +
+ + +
+[2212] +
+
+Scott M. Lundberg and Su-In Lee. + A unified approach to interpreting model predictions. + In I. Guyon, U. von Luxburg, S. Bengio, H. M. Wallach, R. Fergus, S. V. N. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems (NIPS 30), pp.  4765–4774, 2016.
+[ bib | +epub ] +
+Keywords: SHAP, interpretable AI +
+ +
+ + +
+[2213] +
+
+Hoang N. Luong and Peter A. N. Bosman. + Elitist Archiving for Multi-Objective Evolutionary Algorithms: To Adapt or Not to Adapt. + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part II, volume 7492 of Lecture Notes in Computer Science, pp.  72–81. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2214] +
+
+Thibaut Lust and Jacques Teghem. + The multiobjective traveling salesman problem: A survey and a new approach. + In C. A. Coello Coello, C. Dhaenens, and L. Jourdan, editors, Advances in Multi-Objective Nature Inspired Computing, volume 272 of Studies in Computational Intelligence, pp.  119–141. Springer, 2010.
+[ bib ] + +
+ + +
+[2215] +
+
+Khin Lwin, Rong Qu, and Jianhua Zheng. + Multi-objective Scatter Search with External Archive for Portfolio Optimization. + In Proceedings of the 5th International Joint Conference on Computational Intelligence - ECTA (IJCCI 2013), pp.  111–119. SciTePress, 2013.
+[ bib | +DOI ] +
+Crowding archive +
+ +
+ + +
+[2216] +
+
+Robert John Lygoe. + Complexity reduction in high-dimensional multi-objective optimisation. + PhD thesis, University of Sheffield Sheffield, UK, 2010.
+[ bib ] + +
+ + +
+[2217] +
+
+Kate Smith-Miles, Mario A. Muñoz, and Neelofar. + Melbourne Algorithm Test Instance Library with Data Analytics (MATILDA), 2020.
+[ bib | +http ] + +
+ + +
+[2218] +
+
+Gunther Mäckle, Dragan A. Savic, and Godfrey A. Walters. + Application of Genetic Algorithms to Pump Scheduling for Water Supply. + In Genetic Algorithms in Engineering Systems: Innovations and Applications, GALESIA'95, volume 414, pp.  400–405, Sheffield, UK, September 1995. IEE Conference Publication.
+[ bib | +DOI ] +
+ A simple Genetic Algorithm has been applied to the + scheduling of multiple pumping units in a water supply system + with the objective of minimising the overall cost of the + pumping operation, taking advantage of storage capacity in + the system and the availability of off peak electricity + tariffs. A simple example shows that the method is easy to + apply and has produced encouraging preliminary results +
+ +
+ + +
+[2219] +
+
+Nateri K. Madavan. + Multiobjective optimization using a Pareto differential evolution approach. + In D. B. Fogel et al., editors, Proceedings of the 2002 World Congress on Computational Intelligence (WCCI 2002), pp.  1145–1150, Piscataway, NJ, 2002. IEEE Press.
+[ bib ] + +
+ + +
+[2220] +
+
+D. R. Broad, Graeme C. Dandy, and Holger R. Maier. + A Metamodeling Approach to Water Distribution System Optimization. + In 6th Annual Symposium on Water Distribution Systems Analysis. ASCE, June 2004.
+[ bib ] + +
+ + +
+[2221] +
+
+Yuri Malitsky and Meinolf Sellmann. + Instance-specific algorithm configuration as a method for non-model-based portfolio generation. + In N. Beldiceanu, N. Jussien, and E. Pinson, editors, Integration of AI and OR Techniques in Contraint Programming for Combinatorial Optimization Problems, volume 7298 of Lecture Notes in Computer Science, pp.  244–259. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2222] +
+
+Yuri Malitsky, Deepak Mehta, Barry O'Sullivan, and Helmut Simonis. + Tuning parameters of large neighborhood search for the machine reassignment problem. + In C. Gomes and M. Sellmann, editors, Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, CPAIOR 2010, volume 7874 of Lecture Notes in Computer Science, pp.  176–192. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2223] +
+
+Vittorio Maniezzo, M. Boschetti, and M. Jelasity. + An Ant Approach to Membership Overlay Design. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  37–48. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2224] +
+
+Vittorio Maniezzo and M. Milandri. + An Ant-Based Framework for Very Strongly Constrained Problems. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  222–227. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2225] +
+
+Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Rose Finkel, Steven J. Bethard, and David McClosky. + The Stanford CoreNLP Natural Language Processing Toolkit. + In Association for Computational Linguistics (ACL) System Demonstrations, pp.  55–60, 2014.
+[ bib ] +
+http://www.aclweb.org/anthology/P/P14/P14-5010 +
+ +
+ + +
+[2226] +
+
+F. Martínez, V. Bou, V. Hernández, F. Alvarruiz, and J. M. Alonso. + ANN Architectures for Simulating Water Distribution Networks. + In D. A. Savic, G. A. Walters, R. King, and S. Thiam-Khu, editors, Proceedings of the Eighth International Conference on Computing and Control for the Water Industry (CCWI 2005), volume 1, pp.  251–256, University of Exeter, UK, September 2005.
+[ bib ] + +
+ + +
+[2227] +
+
+Marie-Eléonore Marmion, Clarisse Dhaenens, Laetitia Jourdan, Arnaud Liefooghe, and Sébastien Verel. + NILS: A Neutrality-Based Iterated Local Search and Its Application to Flowshop Scheduling. + In P. Merz and J.-K. Hao, editors, Proceedings of EvoCOP 2011 – 11th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 6622 of Lecture Notes in Computer Science, pp.  191–202. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2228] +
+
+Marie-Eléonore Marmion, Franco Mascia, Manuel López-Ibáñez, and Thomas Stützle. + Automatic Design of Hybrid Stochastic Local Search Algorithms. + In M. J. Blesa, C. Blum, P. Festa, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics, volume 7919 of Lecture Notes in Computer Science, pp.  144–158. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2229] +
+
+Oded Maron and Andrew W. Moore. + Hoeffding races: Accelerating model selection search for classification and function approximation. + In J. D. Cowan, G. Tesauro, and J. Alspector, editors, Advances in Neural Information Processing Systems, volume 6, pp.  59–66. Morgan Kaufmann Publishers, San Francisco, CA, 1994.
+[ bib ] + +
+ + +
+[2230] +
+
+C. E. Mariano and E. Morales. + MOAQ: An Ant-Q Algorithm for Multiple Objective Optimization Problems. + In W. Banzhaf, J. M. Daida, A. E. Eiben, M. H. Garzon, V. Honavar, M. J. Jakiela, and R. E. Smith, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 1999, pp.  894–901. Morgan Kaufmann Publishers, San Francisco, CA, 1999.
+[ bib ] + +
+ + +
+[2231] +
+
+Elena Marchiori and Adri G. Steenbeek. + An Iterated Heuristic Algorithm for the Set Covering Problem. + In K. Mehlhorn, editor, Algorithm Engineering, 2nd International Workshop, WAE'92, pp.  155–166. Max-Planck-Institut für Informatik, Saarbrücken, Germany, 1998.
+[ bib ] + +
+ + +
+[2232] +
+
+Elena Marchiori and Adri G. Steenbeek. + An Evolutionary Algorithm for Large Scale Set Covering Problems with Application to Airline Crew Scheduling. + In S. Cagnoni et al., editors, Real-World Applications of Evolutionary Computing, EvoWorkshops 2000, volume 1803 of Lecture Notes in Computer Science, pp.  367–381. Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[2233] +
+
+K. Marriott and P. Stuckey. + Programming With Constraints. + MIT Press, Cambridge, MA, 1998.
+[ bib ] + +
+ + +
+[2234] +
+
+Silvano Martello and Paolo Toth. + Knapsack Problems: Algorithms and Computer Implementations. + John Wiley & Sons, Chichester, UK, 1990.
+[ bib ] +
+Keywords: bin packing +
+ +
+ + +
+[2235] +
+
+Oded Maron. + Hoeffding Races: Model selection for MRI classification. + Master's thesis, Massachusetts Institute of Technology, 1994.
+[ bib ] + +
+ + +
+[2236] +
+
+Franco Mascia, Mauro Birattari, and Thomas Stützle. + Tuning Algorithms for Tackling Large Instances: An Experimental Protocol. + In P. M. Pardalos and G. Nicosia, editors, Learning and Intelligent Optimization, 7th International Conference, LION 7, volume 7997 of Lecture Notes in Computer Science, pp.  410–422. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2237] +
+
+Florence Massen, Yves Deville, and Pascal van Hentenryck. + Pheromone-Based Heuristic Column Generation for Vehicle Routing Problems with Black Box Feasibility. + In N. Beldiceanu, N. Jussien, and E. Pinson, editors, Integration of AI and OR Techniques in Contraint Programming for Combinatorial Optimization Problems, volume 7298 of Lecture Notes in Computer Science, pp.  260–274. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2238] +
+
+Franco Mascia, Manuel López-Ibáñez, Jérémie Dubois-Lacoste, and Thomas Stützle. + Grammar-based generation of stochastic local search heuristics through automatic algorithm configuration tools: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2013-009/, 2013.
+[ bib ] + +
+ + +
+[2239] +
+
+Franco Mascia, Manuel López-Ibáñez, Jérémie Dubois-Lacoste, Marie-Eléonore Marmion, and Thomas Stützle. + Algorithm Comparison by Automatically Configurable Stochastic Local Search Frameworks: A Case Study Using Flow-Shop Scheduling Problems. + In M. J. Blesa, C. Blum, and S. Voß, editors, Hybrid Metaheuristics, volume 8457 of Lecture Notes in Computer Science, pp.  30–44. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[2240] +
+
+Franco Mascia, Manuel López-Ibáñez, Jérémie Dubois-Lacoste, and Thomas Stützle. + From Grammars to Parameters: Automatic Iterated Greedy Design for the Permutation Flow-shop Problem with Weighted Tardiness. + In P. M. Pardalos and G. Nicosia, editors, Learning and Intelligent Optimization, 7th International Conference, LION 7, volume 7997 of Lecture Notes in Computer Science, pp.  321–334. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2241] +
+
+Florence Massen, Manuel López-Ibáñez, Thomas Stützle, and Yves Deville. + Experimental Analysis of Pheromone-Based Heuristic Column Generation Using irace. + In M. J. Blesa, C. Blum, P. Festa, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics, volume 7919 of Lecture Notes in Computer Science, pp.  92–106. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2242] +
+
+Renzo Massobrio, Sergio Nesmachnow, and Bernabé Dorronsoro. + Virtual Savant: learning for optimization. + In M. Vlastelica, J. Song, A. Ferber, B. Amos, G. Martius, B. Dilkina, and Y. Yue, editors, Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020, LMCA 2020, Vancouver, Canada, December 12, 2020, pp.  1–5, 2020.
+[ bib ] + +
+ + +
+[2243] +
+
+Satoshi Matsubara, Motomu Takatsu, Toshiyuki Miyazawa, Takayuki Shibasaki, Yasuhiro Watanabe, Kazuya Takemoto, and Hirotaka Tamura. + Digital Annealer for High-Speed Solving of Combinatorial optimization Problems and Its Applications. + In 2020 25th Asia and South Pacific Design Automation Conference (ASP-DAC), pp.  667–672. IEEE, 2020.
+[ bib | +DOI ] +
+A Digital Annealer (DA) is a dedicated architecture for + high-speed solving of combinatorial optimization problems + mapped to an Ising model. With fully coupled bit connectivity + and high coupling resolution as a major feature, it can be + used to express a wide variety of combinatorial optimization + problems. The DA uses Markov Chain Monte Carlo as a basic + search mechanism, accelerated by the hardware implementation + of multiple speed-enhancement techniques such as parallel + search, escape from a local solution, and replica + exchange. It is currently being offered as a cloud service + using a second-generation chip operating on a scale of 8,192 + bits. This paper presents an overview of the DA, its + performance against benchmarks, and application examples. +
+ +
+ + +
+[2244] +
+
+Michael Maur, Manuel López-Ibáñez, and Thomas Stützle. + Pre-scheduled and adaptive parameter variation in Max-Min Ant System. + In H. Ishibuchi et al., editors, Proceedings of the 2010 Congress on Evolutionary Computation (CEC 2010), pp.  3823–3830, Piscataway, NJ, 2010. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2245] +
+
+Atanu Mazumdar, Tinkle Chugh, Kaisa Miettinen, and Manuel López-Ibáñez. + On Dealing with Uncertainties from Kriging Models in Offline Data-Driven Evolutionary Multiobjective Optimization. + In K. Deb, E. D. Goodman, C. A. Coello Coello, K. Klamroth, K. Miettinen, S. Mostaghim, and P. Reed, editors, Evolutionary Multi-criterion Optimization, EMO 2019, volume 11411 of Lecture Notes in Computer Science, pp.  463–474. Springer International Publishing, Cham, Switzerland, 2019.
+[ bib | +DOI ] + +
+ + +
+[2246] +
+
+G. McCormick and R. S. Powell. + A progressive mixed integer-programming method for pump scheduling. + In C. Maksimović, D. Butler, and F. A. Memon, editors, Advances in Water Supply Management, pp.  307–313. CRC Press, 2003.
+[ bib ] + +
+ + +
+[2247] +
+
+Catherine C. McGeoch. + A Guide to Experimental Algorithmics. + Cambridge University Press, 2012.
+[ bib ] + +
+ + +
+[2248] +
+
+Catherine C. McGeoch and Pau Farré. + The D-Wave Advantage System: An Overview. + Technical report, D-Wave Systems Inc., Burnaby, BC, Canada, 2020.
+[ bib | +http ] + +
+ + +
+[2249] +
+
+Hudson Geovane de Medeiros, Elizabeth Ferreira Gouvêa Goldbarg, and Marco Cesar Goldbarg. + Analyzing Limited Size Archivers of Multi-objective Optimizers. + In 2014 Brazilian Conference on Intelligent Systems, pp.  85–90, 2014.
+[ bib | +DOI ] +
+Keywords: archiving +
+ +
+ + +
+[2250] +
+
+J. Fabian Meier and Uwe Clausen. + A versatile heuristic approach for generalized hub location problems. + Preprint, Provided upon personal request, 2014.
+[ bib | +http ] +
+Keywords: irace +
+ +
+ + +
+[2251] +
+
+L. Melo, F. Pereira, and E. Costa. + MC-ANT: a Multi-colony Ant Algorithm. + In P. Collet, N. Monmarché, P. Legrand, M. Schoenauer, and E. Lutton, editors, Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009, volume 5975 of Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2252] +
+
+Adriana Menchaca-Mendez and Carlos A. Coello Coello. + GD-MOEA: A New Multi-Objective Evolutionary Algorithm Based on the Generational Distance Indicator. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  156–170. Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2253] +
+
+Adriana Menchaca-Mendez and Carlos A. Coello Coello. + GDE-MOEA: A New MOEA based on the generational distance indicator and ε-dominance. + In Proceedings of the 2015 Congress on Evolutionary Computation (CEC 2015), pp.  947–955, Piscataway, NJ, 2015. IEEE Press.
+[ bib ] + +
+ + +
+[2254] +
+
+Hector Mendoza, Aaron Klein, Matthias Feurer, Jost Tobias Springenberg, and Frank Hutter. + Towards automatically-tuned neural networks. + In Workshop on Automatic Machine Learning, pp.  58–65, 2016.
+[ bib ] + +
+ + +
+[2255] +
+
+Olaf Mersmann, Bernd Bischl, Heike Trautmann, Mike Preuss, Claus Weihs, and Günther Rudolph. + Exploratory Landscape Analysis. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  829–836. ACM Press, New York, NY, 2011.
+[ bib ] +
+Keywords: continuous optimization, landscape analysis, instance features +
+ +
+ + +
+[2256] +
+
+Peter Merz and Jutta Huhse. + An Iterated Local Search Approach for Finding Provably Good Solutions for Very Large TSP Instances. + In G. Rudolph et al., editors, Parallel Problem Solving from Nature – PPSN X, volume 5199 of Lecture Notes in Computer Science, pp.  929–939. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2257] +
+
+D. Merkle and Martin Middendorf. + Prospects for Dynamic Algorithm Control: Lessons from the Phase Structure of Ant Scheduling Algorithms. + In R. B. Heckendorn, editor, Proceedings of the 2001 Genetic and Evolutionary Computation Conference – Workshop Program. Workshop “The Next Ten Years of Scheduling Research”, pp.  121–126. Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib ] + +
+ + +
+[2258] +
+
+D. Merkle, Martin Middendorf, and Hartmut Schmeck. + Ant Colony Optimization for Resource-Constrained Project Scheduling. + In D. Whitley et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2000, pp.  893–900. Morgan Kaufmann Publishers, San Francisco, CA, 2000.
+[ bib ] + +
+ + +
+[2259] +
+
+Olaf Mersmann, Heike Trautmann, Boris Naujoks, and Claus Weihs. + Benchmarking Evolutionary Multiobjective Optimization Algorithms. + In H. Ishibuchi et al., editors, Proceedings of the 2010 Congress on Evolutionary Computation (CEC 2010), pp.  1–8, Piscataway, NJ, 2010. IEEE Press.
+[ bib ] +
+TR: http://hdl.handle.net/2003/26671 +
+ +
+ + +
+[2260] +
+
+Bernd Meyer. + Convergence control in ACO. + In Genetic and Evolutionary Computation Conference (GECCO), Seattle, WA, 2004. + Late-breaking paper available on CD.
+[ bib ] + +
+ + +
+[2261] +
+
+Bernd Meyer and Andreas T. Ernst. + Integrating ACO and Constraint Propagation. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  166–177. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2262] +
+
+Efrén Mezura-Montes, M. Reyes-Sierra, and Carlos A. Coello Coello. + Multi-objective optimization using differential evolution: a survey of the state-of-the-art. + In U. K. Chakraborty, editor, Advances in differential evolution, pp.  173–196. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[2263] +
+
+Efrén Mezura-Montes, Jesús Velázquez-Reyes, and Carlos A. Coello Coello. + A comparative study of differential evolution variants for global optimization. + In M. Cattolico et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2006, pp.  485–492. ACM Press, New York, NY, 2006.
+[ bib | +DOI ] + +
+ + +
+[2264] +
+
+Zbigniew Michalewicz and David B. Fogel. + How to Solve It: Modern Heuristics. + Springer, 2nd edition, 2004.
+[ bib ] + +
+ + +
+[2265] +
+
+Laurent D. Michel and Pascal van Hentenryck. + Iterative Relaxations for Iterative Flattening in Cumulative Scheduling. + In S. Zilberstein, J. Koehler, and S. Koenig, editors, Proceedings of the Fourteenth International Conference on Automated Planning and Scheduling (ICAPS 2004), pp.  200–208. AAAI Press/MIT Press, Menlo Park, CA, 2004.
+[ bib ] + +
+ + +
+[2266] +
+
+R. Michel and M. Middendorf. + An Island Model based Ant System with Lookahead for the Shortest Supersequence Problem. + In A. E. Eiben, T. Bäck, M. Schoenauer, and H.-P. Schwefel, editors, Parallel Problem Solving from Nature – PPSN V, volume 1498 of Lecture Notes in Computer Science, pp.  692–701. Springer, Heidelberg, Germany, 1998.
+[ bib ] + +
+ + +
+[2267] +
+
+Zbigniew Michalewicz. + Genetic Algorithms + Data Structures = Evolution Programs. + Springer, Berlin, Germany, 3rd edition, 1996.
+[ bib ] + +
+ + +
+[2268] +
+
+Kaisa Miettinen. + Nonlinear Multiobjective Optimization. + Kluwer Academic Publishers, Boston, MA, 1999.
+[ bib ] +
+Nonlinear Multiobjective Optimization provides an extensive, + up-to-date, self-contained and consistent survey and review + of the literature and of the state of the art on nonlinear + (deterministic) multiobjective optimization, its methods, its + theory and its background. This book is intended for both + researchers and students in the areas of (applied) + mathematics, engineering, economics, operations research and + management science; it is meant for both professionals and + practitioners in many different fields of application. The + intention is to provide a consistent summary that may help in + selecting an appropriate method for the problem to be + solved. The extensive bibliography will be of value to + researchers. +
+ +
+ + +
+[2269] +
+
+Kaisa Miettinen, Francisco Ruiz, and Andrzej P. Wierzbicki. + Introduction to Multiobjective Optimization: Interactive Approaches. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] +
+We give an overview of interactive methods developed + for solving nonlinear multiobjective optimization + problems. In interactive methods, a decision maker + plays an important part and the idea is to support + her/him in the search for the most preferred + solution. In interactive methods, steps of an + iterative solution algorithm are repeated and the + decision maker progressively provides preference + information so that the most preferred solution can + be found. We identify three types of specifying + preference information in interactive methods and + give some examples of methods representing each + type. The types are methods based on trade-off + information, reference points and classification of + objective functions. +
+ +
+ + +
+[2270] +
+
+Péricles Miranda, Ricardo M. Silva, and Ricardo B. Prudêncio. + Fine-Tuning of Support Vector Machine Parameters Using Racing Algorithms. + In European Symposium on Artificial Neural Networks, ESSAN, pp.  325–330, 2014.
+[ bib | +epub ] +
+Keywords: irace +
+ +
+ + +
+[2271] +
+
+Péricles Miranda, Ricardo M. Silva, and Ricardo B. Prudêncio. + I/S-Race: An Iterative Multi-objective Racing Algorithm for the SVM Parameter Selection Problem. + In European Symposium on Artificial Neural Networks, ESSAN, pp.  573–578, 2015.
+[ bib | +epub ] + +
+ + +
+[2272] +
+
+Alfonsas Misevičius. + Ruin and Recreate Principle Based Approach for the Quadratic Assignment Problem. + In E. Cantú-Paz et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2003, Part I, volume 2723 of Lecture Notes in Computer Science, pp.  598–609. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2273] +
+
+Debasis Mitra, Fabio Romeo, and Alberto Sangiovanni-Vincentelli. + Convergence and Finite-Time Behavior of Simulated Annealing. + In Decision and Control, 1985 24th IEEE Conference on, pp.  761–767. IEEE, 1985.
+[ bib ] + +
+ + +
+[2274] +
+
+David G. Mitchell, Bart Selman, and Hector J. Levesque. + Hard and Easy Distributions of SAT Problems. + In W. R. Swartout, editor, Proceedings of the 10th National Conference on Artificial Intelligence, pp.  459–465. AAAI Press/MIT Press, Menlo Park, CA, 1992.
+[ bib ] + +
+ + +
+[2275] +
+
+Volodymyr Mnih, Csaba Szepesvári, and Jean-Yves Audibert. + Empirical Bernstein stopping. + In W. W. Cohen, A. McCallum, and S. T. Roweis, editors, Proceedings of the 25th International Conference on Machine Learning, ICML 2008, pp.  672–679, New York, NY, 2008. ACM Press.
+[ bib ] + +
+ + +
+[2276] +
+
+Jonas Močkus. + On Bayesian Methods for Seeking the Extremum. + In G. I. Marchuk, editor, Optimization Techniques IFIP Technical Conference Novosibirsk, July 1–7, 1974, volume 27 of Lecture Notes in Computer Science, pp.  400–404. Springer, Berlin/Heidelberg, 1975.
+[ bib | +DOI ] +
+Proposed Bayesian optimization +
+ +
+ + +
+[2277] +
+
+Jonas Močkus. + Bayesian Approach to Global Optimization: Theory and Applications. + Kluwer Academic Publishers, 1989.
+[ bib ] + +
+ + +
+[2278] +
+
+Sander van Rijn. + Modular CMA-ES framework from [2439], v0.3.0. + https://github.com/sjvrijn/ModEA, 2018. + Available also as pypi package at https://pypi.org/project/ModEA/0.3.0/.
+[ bib ] + +
+ + +
+[2279] +
+
+Atefeh Moghaddam, Farouk Yalaoui, and Lionel Amodeo. + Lorenz versus Pareto Dominance in a Single Machine Scheduling Problem with Rejection. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  520–534. Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[2280] +
+
+Teodor Mihai Moldovan and Pieter Abbeel. + Safe Exploration in Markov Decision Processes. + In J. Langford and J. Pineau, editors, Proceedings of the 29th International Conference on Machine Learning, ICML 2012, pp.  1451–1458. Omnipress, 2012.
+[ bib | +epub ] + +
+ + +
+[2281] +
+
+Ilya Molchanov. + Theory of Random Sets. + Springer, 2005.
+[ bib ] +
+Keywords: Vorob'ev expectation +
+ +
+ + +
+[2282] +
+
+Jean-Noël Monette, Yves Deville, and Pascal van Hentenryck. + Aeon: Synthesizing Scheduling Algorithms from High-Level Models. + In J. W. Chinneck, B. Kristjansson, and M. J. Saltzman, editors, Operations Research and Cyber-Infrastructure, volume 47 of Operations Research/Computer Science Interfaces, pp.  43–59. Springer, New York, NY, 2009.
+[ bib ] + +
+ + +
+[2283] +
+
+Elizabeth Montero, Leslie Pérez Cáceres, María-Cristina Riff, and Carlos A. Coello Coello. + Are State-of-the-Art Fine-Tuning Algorithms Able to Detect a Dummy Parameter? + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part I, volume 7491 of Lecture Notes in Computer Science, pp.  306–315. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2284] +
+
+María-Cristina Riff and Elizabeth Montero. + A new algorithm for reducing metaheuristic design effort. + In Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), pp.  3283–3290, Piscataway, NJ, 2013. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2285] +
+
+Elizabeth Montero and María-Cristina Riff. + Towards a Method for Automatic Algorithm Configuration: A Design Evaluation Using Tuners. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  90–99. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[2286] +
+
+Elizabeth Montero, María-Cristina Riff, and Bertrand Neveu. + An Evaluation of Off-line Calibration Techniques for Evolutionary Algorithms. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  299–300. ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2287] +
+
+Gilberto Montibeller and Hugo Yoshizaki. + A Framework for Locating Logistic Facilities with Multi-Criteria Decision Analysis. + In R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors, Evolutionary Multi-criterion Optimization, EMO 2011, volume 6576 of Lecture Notes in Computer Science, pp.  505–519. Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[2288] +
+
+Marco A. Montes de Oca. + Incremental Social Learning in Swarm Intelligence Systems. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2011.
+[ bib ] +
+Supervised by Marco Dorigo +
+ +
+ + +
+[2289] +
+
+James Montgomery. + Solution Biases and Pheromone Representation Selection in Ant Colony Optimisation. + PhD thesis, School of Information Technology, Bond University, Australia, 2005.
+[ bib ] + +
+ + +
+[2290] +
+
+Douglas C. Montgomery. + Design and Analysis of Experiments. + John Wiley & Sons, New York, NY, 8th edition, 2012.
+[ bib ] + +
+ + +
+[2291] +
+
+Andrew W. Moore and Mary S. Lee. + Efficient Algorithms for Minimizing Cross Validation Error. + In W. W. Cohen and H. Hirsh, editors, Proceedings of the 11th International Conference on Machine Learning, ICML 1994, pp.  190–198, San Francisco, CA, 1994. Morgan Kaufmann Publishers.
+[ bib ] + +
+ + +
+[2292] +
+
+A. Moraglio and A. Kattan. + Geometric Generalisation of Surrogate Model Based Optimization to Combinatorial Spaces. + In P. Merz and J.-K. Hao, editors, Proceedings of EvoCOP 2011 – 11th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 6622 of Lecture Notes in Computer Science, pp.  142–154. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2293] +
+
+A. Moraglio, Yong-Hyuk Kim, and Yourim Yoon. + Geometric Surrogate-based Optimisation for Permutation-based Problems. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2011, pp.  133–134. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2294] +
+
+Pail Morris. + The Breakout Method for Escaping from Local Minima. + In R. Fikes and W. G. Lehnert, editors, Proceedings of the 11th National Conference on Artificial Intelligence, pp.  40–45. AAAI Press/MIT Press, Menlo Park, CA, 1993.
+[ bib ] + +
+ + +
+[2295] +
+
+J. D. Moss and C. G. Johnson. + An ant colony algorithm for multiple sequence alignment in bioinformatics. + In D. W. Pearson, N. C. Steele, and R. F. Albrecht, editors, Artificial Neural Networks and Genetic Algorithms, pp.  182–186. Springer Verlag, 2003.
+[ bib ] + +
+ + +
+[2296] +
+
+Pablo Moscato. + On Evolution, Search, Optimization, Genetic Algorithms and Martial Arts: Towards Memetic Algorithms. + Caltech Concurrent Computation Program, C3P Report 826, Caltech, 1989.
+[ bib ] + +
+ + +
+[2297] +
+
+Pablo Moscato. + Memetic algorithms: a short introduction. + In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in Optimization, pp.  219–234. McGraw Hill, London, UK, 1999.
+[ bib ] + +
+ + +
+[2298] +
+
+Vincent Mousseau. + Elicitation des préférences pour l'aide multicritère à la décision. + PhD thesis, Université Paris-Dauphine, Paris, France, 2003.
+[ bib ] + +
+ + +
+[2299] +
+
+J. Moy. + RFC 1583: Open shortest path first protocol, 1994.
+[ bib ] + +
+ + +
+[2300] +
+
+Zongxu Mu, Jérémie Dubois-Lacoste, Holger H. Hoos, and Thomas Stützle. + On the Empirical Scaling of Running Time for Finding Optimal Solutions to the TSP: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2017-010/, 2017.
+[ bib ] + +
+ + +
+[2301] +
+
+Zongxu Mu, Holger H. Hoos, and Thomas Stützle. + The Impact of Automated Algorithm Configuration on the Scaling Behaviour of State-of-the-Art Inexact TSP Solvers. + In P. Festa, M. Sellmann, and J. Vanschoren, editors, Learning and Intelligent Optimization, 10th International Conference, LION 10, volume 10079 of Lecture Notes in Computer Science, pp.  157–172. Springer, Cham, Switzerland, 2016.
+[ bib | +DOI ] + +
+ + +
+[2302] +
+
+Mudita Sharma, Manuel López-Ibáñez, and Dimitar Kazakov. + Deep Reinforcement Learning Based Parameter Control in Differential Evolution: Supplementary material. + https://github.com/mudita11/DE-DDQN, 2019.
+[ bib | +DOI ] + +
+ + +
+[2303] +
+
+H. Mühlenbein and Jörg Zimmermann. + Size of neighborhood more important than temperature for stochastic local search. + In Proceedings of the 2000 Congress on Evolutionary Computation (CEC'00), pp.  1017–1024, Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[2304] +
+
+H. Mühlenbein. + Evolution in Time and Space—The Parallel Genetic Algorithm. + In G. Rawlins, editor, Foundations of Genetic Algorithms (FOGA), pp.  316–337. Morgan Kaufmann Publishers, San Mateo, CA, 1991.
+[ bib ] + +
+ + +
+[2305] +
+
+Moritz Mühlenthaler. + Fairness in Academic Course Timetabling. + Springer, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[2306] +
+
+L. J. Murphy, Graeme C. Dandy, and Angus R. Simpson. + Optimum Design and Operation of Pumped Water Distribution Systems. + In 1994 International Conference on Hydraulics and Civil Engineering, Hidraulic working with the Environment, pp.  149–155, Brisbane, Australia, February 1994. The Institution of Engineers.
+[ bib ] + +
+ + +
+[2307] +
+
+Yuichi Nagata and Shigenobu Kobayashi. + Edge Assembly Crossover: A High-power Genetic Algorithm for the Traveling Salesman Problem. + In T. Bäck, editor, ICGA, pp.  450–457. Morgan Kaufmann Publishers, San Francisco, CA, 1997.
+[ bib ] + +
+ + +
+[2308] +
+
+Yuichi Nagata and Shigenobu Kobayashi. + An analysis of edge assembly crossover for the traveling salesman problem. + In K. Ito, F. Harashima, and K. Tanie, editors, IEEE SMC'99 Conference Proceedings, 1999 IEEE International Conference on Systems, Man, and Cybernetics, pp.  628–633. IEEE Press, 1999.
+[ bib | +DOI ] + +
+ + +
+[2309] +
+
+R. Nagy, M. Suciu, and D. Dumitrescu. + Exploring Lorenz Dominance. + In Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2012 14th International Symposium on, pp.  254–259, 2012.
+[ bib ] + +
+ + +
+[2310] +
+
+Vinod Nair and Geoffrey E. Hinton. + Rectified linear units improve restricted boltzmann machines. + In J. Fürnkranz and T. Joachims, editors, Proceedings of the 27th International Conference on Machine Learning, ICML 2010, pp.  807–814, New York, NY, 2010. ACM Press.
+[ bib ] + +
+ + +
+[2311] +
+
+Samadhi Nallaperuma, Markus Wagner, and Frank Neumann. + Parameter Prediction Based on Features of Evolved Instances for Ant Colony Optimization and the Traveling Salesperson Problem. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  100–109. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[2312] +
+
+V. Nannen and Agoston E. Eiben. + A Method for Parameter Calibration and Relevance Estimation in Evolutionary Algorithms. + In M. Cattolico et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2006, pp.  183–190. ACM Press, New York, NY, 2006.
+[ bib | +DOI ] +
+Keywords: REVAC +
+ +
+ + +
+[2313] +
+
+V. Nannen and Agoston E. Eiben. + Relevance Estimation and Value Calibration of Evolutionary Algorithm Parameters. + In M. M. Veloso, editor, Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI-07), pp.  975–980. AAAI Press, Menlo Park, CA, 2007.
+[ bib ] +
+Keywords: REVAC +
+ +
+ + +
+[2314] +
+
+Youssef S. G. Nashed, Pablo Mesejo, Roberto Ugolotti, Jérémie Dubois-Lacoste, and Stefano Cagnoni. + A Comparative Study of Three GPU-Based Metaheuristics. + In C. A. Coello Coello et al., editors, Parallel Problem Solving from Nature – PPSN XII, Part II, volume 7492 of Lecture Notes in Computer Science, pp.  398–407. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2315] +
+
+Antonio J. Nebro, Juan J. Durillo, and Carlos A. Coello Coello. + Analysis of leader selection strategies in a multi-objective Particle Swarm Optimizer. + In Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), pp.  3153–3160, Piscataway, NJ, 2013. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2316] +
+
+Antonio J. Nebro, Juan J. Durillo, José García-Nieto, Carlos A. Coello Coello, F. Luna, and Enrique Alba. + SMPSO: A new PSO-based metaheuristic for multi-objective optimization. + In 2009 IEEE Symposium on Computational Intelligence in Multi-Criteria Decision-Making (MCDM), pp.  66–73, 2009.
+[ bib | +DOI ] + +
+ + +
+[2317] +
+
+Antonio J. Nebro, Juan J. Durillo, and Matthieu Vergne. + Redesigning the jMetal Multi-Objective Optimization Framework. + In J. L. Jiménez Laredo, S. Silva, and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2015, pp.  1093–1100. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] +
+Keywords: JMetal, Multi-objective metaheuristics, Open source, + Optimization framework +
+ +
+ + +
+[2318] +
+
+Antonio J. Nebro, Manuel López-Ibáñez, Cristóbal Barba-González, and José García-Nieto. + Automatic Configuration of NSGA-II with jMetal and irace. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1374–1381. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2319] +
+
+G. L. Nemhauser and L. A. Wolsey. + Integer and Combinatorial Optimization. + John Wiley & Sons, New York, NY, 1988.
+[ bib ] + +
+ + +
+[2320] +
+
+Alexander G. Nikolaev and Sheldon H. Jacobson. + Simulated Annealing. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, pp.  1–39. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[2321] +
+
+Mladen Nikolić, Filip Marić, and Predrag Janičić. + Instance-based selection of policies for SAT solvers. + In International Conference on Theory and Applications of Satisfiability Testing, pp.  326–340. Springer, 2009.
+[ bib ] + +
+ + +
+[2322] +
+
+Y. Nishio, A. Oyama, Y. Akimoto, Hernán E. Aguirre, and Kiyoshi Tanaka. + Many-objective Optimization of Trajectory Design for DESTINY Mission. + In P. M. Pardalos, M. G. C. Resende, C. Vogiatzis, and J. L. Walteros, editors, Learning and Intelligent Optimization, 8th International Conference, LION 8, volume 8426 of Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2323] +
+
+Mark S. Nixon and Alberto S. Aguado. + Feature extraction & image processing for computer vision. + Academic Press, New York, NY, 2012.
+[ bib ] + +
+ + +
+[2324] +
+
+Jacob de Nobel, Diederick Vermetten, Hao Wang, Carola Doerr, and Thomas Bäck. + Tuning as a means of assessing the benefits of new ideas in interplay with existing algorithmic modules. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1375–1384. ACM Press, New York, NY, 2021.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[2325] +
+
+Jacob de Nobel, Diederick Vermetten, Hao Wang, Carola Doerr, and Thomas Bäck. + Data and Code from Tuning as a means of assessing the benefits of new ideas in interplay with existing algorithmic modules, February 2021.
+[ bib | +DOI ] + +
+ + +
+[2326] +
+
+Jorge Nocedal and Stephen J. Wright. + Numerical Optimization. + Springer Series in Operations Research and Financial Engineering. Springer, 2nd edition, 2006.
+[ bib ] + +
+ + +
+[2327] +
+
+Houssem Eddine Nouri, Olfa Belkahla Driss, and Khaled Ghédira. + A Classification Schema for the Job Shop Scheduling Problem with Transportation Resources: State-of-the-Art Review. + In R. Silhavy, R. Senkerik, Z. K. Oplatkova, P. Silhavy, and Z. Prokopova, editors, Artificial Intelligence Perspectives in Intelligent Systems, volume 464 of Advances in Intelligent Systems and Computing, pp.  1–11. Springer International Publishing, 2016.
+[ bib ] + +
+ + +
+[2328] +
+
+Krzysztof Nowak, Marcus Märtens, and Dario Izzo. + Empirical Performance of the Approximation of the Least Hypervolume Contributor. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  662–671. Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2329] +
+
+Eoin O'Mahony, Emmanuel Hebrard, Alan Holland, Conor Nugent, and Barry O'Sullivan. + Using case-based reasoning in an algorithm portfolio for constraint solving. + In Bridge et al., editors, Irish Conference on Artificial Intelligence and Cognitive Science, pp.  210–216, 2008.
+[ bib ] + +
+ + +
+[2330] +
+
+Shigeru Obayashi and Daisuke Sasaki. + Visualization and data mining of Pareto solutions using self-organizing map. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  796–809. Springer, Heidelberg, Germany, 2003.
+[ bib ] +
+Keywords: objective reduction +
+ +
+ + +
+[2331] +
+
+Gabriela Ochoa, Matthew Hyde, Tim Curtois, Jose A. Vazquez-Rodriguez, James Walker, Michel Gendreau, Graham Kendall, Barry McCollum, Andrew J. Parkes, Sanja Petrovic, and Edmund K. Burke. + Hyflex: A benchmark framework for cross-domain heuristic search. + In J.-K. Hao and M. Middendorf, editors, Proceedings of EvoCOP 2012 – 12th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 7245 of Lecture Notes in Computer Science, pp.  136–147. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2332] +
+
+Gabriela Ochoa, Marco Tomassini, Sébastien Verel, and Christian Darabos. + A Study of NK Landscapes' Basins and Local Optima Networks. + In C. Ryan, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2008, pp.  555–562. ACM Press, New York, NY, 2008.
+[ bib ] + +
+ + +
+[2333] +
+
+Angelo Oddi, Riccardo Rasconi, Amadeo Cesta, and Stephen F. Smith. + Iterative Flattening Search for the Flexible Job Shop Scheduling Problem. + In T. Walsh, editor, Proceedings of the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11), pp.  1991–1996. IJCAI/AAAI Press, Menlo Park, CA, 2011.
+[ bib ] + +
+ + +
+[2334] +
+
+Vesa Ojalehto, Dmitry Podkopaev, and Kaisa Miettinen. + Towards Automatic Testing of Reference Point Based Interactive Methods. + In J. Handl, E. Hart, P. R. Lewis, M. López-Ibáñez, G. Ochoa, and B. Paechter, editors, Parallel Problem Solving from Nature – PPSN XIV, volume 9921 of Lecture Notes in Computer Science, pp.  483–492. Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] +
+In this research, we proposed to build an automated framework + for testing interactive multiobjective optimization methods, + without utilizing a value function to represent the DM's + preferences. This was achieved by replacing the human DM with + an artificial DM constructed from two distinct parts: the + steady part and the current context. With the steady part the + artificial DM tries to maintain the search towards its + preferences, while at the same time the current context + allows changing the direction as well as ending the solution + process prematurely, mimicking actions of a human DM. +
+
+Keywords: artificial DMs +
+ +
+ + +
+[2335] +
+
+Sabrina M. Oliveira, Mohamed Saifullah Hussin, Andrea Roli, Marco Dorigo, and Thomas Stützle. + Analysis of the Population-based Ant Colony Optimization Algorithm for the TSP and the QAP. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  1734–1741, Piscataway, NJ, 2017. IEEE Press.
+[ bib ] + +
+ + +
+[2336] +
+
+Randal S. Olson, Nathan Bartley, Ryan J. Urbanowicz, and Jason H. Moore. + Evaluation of a Tree-based Pipeline Optimization Tool for Automating Data Science. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2016, pp.  485–492. ACM Press, New York, NY, 2016.
+[ bib | +DOI ] +
+Keywords: TPOT +
+ +
+ + +
+[2337] +
+
+Randal S. Olson, Ryan J. Urbanowicz, Peter C. Andrews, Nicole A. Lavender, La Creis Kidd, and Jason H. Moore. + Automating Biomedical Data Science Through Tree-Based Pipeline Optimization. + In G. Squillero and P. Burelli, editors, Applications of Evolutionary Computation, volume 9597 of Lecture Notes in Computer Science, pp.  123–137. Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] +
+Keywords: TPOT +
+ +
+ + +
+[2338] +
+
+Avi Ostfeld and Elad Salomons. + Optimal Scheduling of Pumping and Chlorine Injections under Unsteady Hydraulics. + In G. Sehlke, D. F. Hayes, and D. K. Stevens, editors, Critical Transitions In Water And Environmental Resources Management, pp.  1–9, July 2004.
+[ bib ] +
+This paper describes the methodology and application + of a genetic algorithm (GA) scheme, tailor-made to + EPANET for simultaneously optimizing the scheduling + of existing pumping and booster disinfection units, + as well as the design of new disinfection booster + chlorination stations, under unsteady + hydraulics. The objective is to minimize the total + cost of operating the pumping units and the chlorine + booster operation and design for a selected + operational time horizon, while delivering the + consumers required water quantities, at acceptable + pressures and chlorine residual concentrations. The + decision variables, for each of the time steps that + encompass the total operational time horizon, + include: the scheduling of the pumping units, + settings of the water distribution system control + valves, and the mass injection rates at each of the + booster chlorination stations. The constraints are + domain heads and chlorine concentrations at the + consumer nodes, maximum injection rates at the + chlorine injection stations, maximum allowable + amounts of water withdraws at the sources, and + returning at the end of the operational time horizon + to a prescribed total volume in the tanks. The model + is demonstrated through an example application. +
+ +
+ + +
+[2339] +
+
+Meltem Öztürk, Alexis Tsoukiàs, and Philippe Vincke. + Preference Modelling. + In J. R. Figueira, S. Greco, and M. Ehrgott, editors, Multiple Criteria Decision Analysis, State of the Art Surveys, chapter 2, pp.  27–72. Springer, 2005.
+[ bib ] + +
+ + +
+[2340] +
+
+Federico Pagnozzi and Thomas Stützle. + Automatic Design of Hybrid Stochastic Local Search Algorithms for Permutation Flowshop Problems. + Technical Report TR/IRIDIA/2018-005, IRIDIA, Université Libre de Bruxelles, Belgium, April 2018.
+[ bib | +http ] + +
+ + +
+[2341] +
+
+Federico Pagnozzi and Thomas Stützle. + Automatic Design of Hybrid Stochastic Local Search Algorithms for Permutation Flowshop Problems: Supplementary Material. + http://iridia.ulb.ac.be/supp/IridiaSupp2018-002/, 2018.
+[ bib ] + +
+ + +
+[2342] +
+
+Federico Pagnozzi and Thomas Stützle. + Automatic design of hybrid stochastic local search algorithms for permutation flowshop problems with additional constraints. + http://iridia.ulb.ac.be/supp/IridiaSupp2018-002/, 2019.
+[ bib ] + +
+ + +
+[2343] +
+
+Federico Pagnozzi. + Automatic Design of Hybrid Stochastic Local Search Algorithms. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2019.
+[ bib ] +
+Supervised by Thomas Stützle +
+ +
+ + +
+[2344] +
+
+Lie Meng Pang, Hisao Ishibuchi, and Ke Shang. + Algorithm configurations of MOEA/D with an unbounded external archive. + In 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp.  1087–1094. IEEE, 2020.
+[ bib ] + +
+ + +
+[2345] +
+
+Christos H. Papadimitriou and K. Steiglitz. + Combinatorial Optimization – Algorithms and Complexity. + Prentice Hall, Englewood Cliffs, NJ, 1982.
+[ bib ] + +
+ + +
+[2346] +
+
+Christos H. Papadimitriou and Mihalis Yannakakis. + On the Approximability of Trade-offs and Optimal Access of Web Sources. + In A. Blum, editor, 41st Annual Symposium on Foundations of Computer Science, pp.  86–92. IEEE Computer Society Press, 2000.
+[ bib | +DOI ] + +
+ + +
+[2347] +
+
+Luís Paquete. + Algoritmos Evolutivos Multiobjectivo para Afectação de Recursos e sua Aplicação à Geração de Horários em Universidades (Multiobjective Evolutionary Algorithms for Resource Allocation and their Application to University Timetabling). + Master's thesis, University of Algarve, 2001. + In Portuguese.
+[ bib ] +
+The aim of this study is the application of + multiobjective evolutionary algorithms to resource + allocation problems, such as university examination + timetabling and course timetabling + problems. Usually, these problems are characterized + by multiple conflicting objectives. A multiobjective + formalization of these problems is presented, based + on goals and priorities. Various aspects of + evolutionary algorithms are proposed and studied for + these problems, particulary, selection methods and + types and parameters of mutation operator. The + choice of both representation and operators is made + so as not to favour excessively certain objectives + with respect to others at the level of the + exploration mechanism. A comparative study of + performance is presented for the proposed algorithms + by means of statistical inference, based on real + problems of the University of Algarve. The notion of + attainment functions is used as a base for the + assessment of performance of multiobjective + evolutionary algorithms. Finally, the evolution of + the solution cost during the runs is analysed by + means of attainment functions, as well. +
+ +
+ + +
+[2348] +
+
+Luís Paquete. + Stochastic Local Search Algorithms for Multiobjective Combinatorial Optimization: Methods and Analysis. + PhD thesis, FB Informatik, TU Darmstadt, Germany, 2005.
+[ bib ] + +
+ + +
+[2349] +
+
+Luís Paquete, Marco Chiarandini, and Thomas Stützle. + Pareto Local Optimum Sets in the Biobjective Traveling Salesman Problem: An Experimental Study. + In X. Gandibleux, M. Sevaux, K. Sörensen, and V. T'Kindt, editors, Metaheuristics for Multiobjective Optimisation, volume 535 of Lecture Notes in Economics and Mathematical Systems, pp.  177–199. Springer, Berlin/Heidelberg, 2004.
+[ bib | +DOI ] +
+In this article, we study Pareto local optimum sets for the + biobjective Traveling Salesman Problem applying + straightforward extensions of local search algorithms for the + single objective case. The performance of the local search + algorithms is illustrated by experimental results obtained + for well known benchmark instances and comparisons to methods + from literature. In fact, a 3-opt local search is able to + compete with the best performing metaheuristics in terms of + solution quality. Finally, we also present an empirical study + of the features of the solutions found by 3-opt on a set of + randomly generated instances. The results indicate the + existence of several clusters of near-optimal solutions that + are separated by only a few edges. +
+
+Keywords: Pareto local search, PLS +
+ +
+ + +
+[2350] +
+
+Luís Paquete, Carlos M. Fonseca, and Manuel López-Ibáñez. + An optimal algorithm for a special case of Klee's measure problem in three dimensions. + Technical Report CSI-RT-I-01/2006, CSI, Universidade do Algarve, 2006. + Superseded by paper in IEEE Transactions on Evolutionary Computation [123].
+[ bib ] +
+The measure of the region dominated by (the maxima + of) a set of n points in the positive d-orthant + has been proposed as an indicator of performance in + multiobjective optimization, known as the + hypervolume indicator, and the problem of computing + it efficiently is attracting increasing + attention. In this report, this problem is + formulated as a special case of Klee's measure + problem in d dimensions, which immediately + establishes O(nd/2log n) as a, possibly + conservative, upper bound on the required + computation time. Then, an O(n log n) algorithm + for the 3-dimensional version of this special case + is constructed, based on an existing dimension-sweep + algorithm for the related maxima problem. Finally, + O(n log n) is shown to remain a lower bound on the + time required by the hypervolume indicator for + d>1, which attests the optimality of the algorithm + proposed. +
+
+Proof of Theorem 3.1 is incorrect +
+ +
+ + +
+[2351] +
+
+Luís Paquete and Thomas Stützle. + Clusters of non-dominated solutions in multiobjective combinatorial optimization: An experimental analysis. + In V. Barichard, M. Ehrgott, X. Gandibleux, and V. T'Kindt, editors, Multiobjective Programming and Goal Programming: Theoretical Results and Practical Applications, volume 618 of Lecture Notes in Economics and Mathematical Systems, pp.  69–77. Springer, Berlin, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2352] +
+
+Luís Paquete and Thomas Stützle. + An Experimental Investigation of Iterated Local Search for Coloring Graphs. + In S. Cagnoni et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2002, volume 2279 of Lecture Notes in Computer Science, pp.  122–131. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2353] +
+
+Luís Paquete and Thomas Stützle. + A Two-Phase Local Search for the Biobjective Traveling Salesman Problem. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  479–493. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2354] +
+
+Luís Paquete and Thomas Stützle. + Stochastic Local Search Algorithms for Multiobjective Combinatorial Optimization: A Review. + In T. F. Gonzalez, editor, Handbook of Approximation Algorithms and Metaheuristics, pp.  411–425. Chapman & Hall/CRC, Boca Raton, FL, 2018.
+[ bib | +DOI ] + +
+ + +
+[2355] +
+
+Luís Paquete, Thomas Stützle, and Manuel López-Ibáñez. + On the design and analysis of SLS algorithms for multiobjective combinatorial optimization problems. + Technical Report TR/IRIDIA/2005-029, IRIDIA, Université Libre de Bruxelles, Belgium, 2005.
+[ bib | +http ] +
+Effective Stochastic Local Search (SLS) algorithms + can be seen as being composed of several algorithmic + components, each of which plays some specific role + with respect to overall performance. In this + article, we explore the application of experimental + design techniques to analyze the effect of different + choices for these algorithmic components on SLS + algorithms applied to Multiobjective Combinatorial + Optimization Problems that are solved in terms of + Pareto optimality. This analysis is done using the + example application of SLS algorithms to the + biobjective Quadratic Assignment Problem and we show + also that the same choices for algorithmic + components can lead to different behavior in + dependence of various instance features, such as the + structure of input data and the correlation between + objectives. +
+ +
+ + +
+[2356] +
+
+Luís Paquete, Thomas Stützle, and Manuel López-Ibáñez. + Towards the Empirical Analysis of SLS Algorithms for Multiobjective Combinatorial Optimization Problems through Experimental Design. + In K. F. Doerner, M. Gendreau, P. Greistorfer, W. J. Gutjahr, R. F. Hartl, and M. Reimann, editors, 6th Metaheuristics International Conference (MIC 2005), pp.  739–746, Vienna, Austria, 2005.
+[ bib ] +
+ Stochastic Local Search (SLS) algorithms for + Multiobjective Combinatorial Optimization Problems + (MCOPs) typically involve the selection and + parameterization of many algorithm components whose + role with respect to their overall performance and + relation to certain instance features is often not + clear. In this abstract, we use a modular approach + for the design of SLS algorithms for MCOPs defined + in terms of Pareto optimality and we present an + extensive analysis of SLS algorithms through + experimental design techniques, where each algorithm + component is considered a factor. The experimental + analysis is based on a sound experimental + methodology for analyzing the output of algorithms + for MCOPs. We show that different choices for + algorithm components can lead to different behavior + in dependence of various instance features. +
+ +
+ + +
+[2357] +
+
+Luís Paquete, Thomas Stützle, and Manuel López-Ibáñez. + Using experimental design to analyze stochastic local search algorithms for multiobjective problems. + In K. F. Doerner, M. Gendreau, P. Greistorfer, W. J. Gutjahr, R. F. Hartl, and M. Reimann, editors, Metaheuristics: Progress in Complex Systems Optimization, volume 39 of Operations Research / Computer Science Interfaces, pp.  325–344. Springer, New York, NY, 2007.
+[ bib | +DOI ] +
+Stochastic Local Search (SLS) algorithms can be seen as being + composed of several algorithmic components, each playing some + specific role with respect to overall performance. This + article explores the application of experimental design + techniques to analyze the effect of components of SLS + algorithms for Multiobjective Combinatorial Optimization + problems, in particular for the Biobjective Quadratic + Assignment Problem. The analysis shows that there exists a + strong dependence between the choices for these components + and various instance features, such as the structure of the + input data and the correlation between the objectives. +
+
+Post-Conference Proceedings of the 6th Metaheuristics + International Conference (MIC 2005) +
+ +
+ + +
+[2358] +
+
+J Paulli. + A computational comparison of simulated annealing and tabu search applied to the quadratic assignment problem. + In R. V. V. Vidal, editor, Applied Simulated Annealing, pp.  85–102. Springer, 1993.
+[ bib ] + +
+ + +
+[2359] +
+
+Lucas Marcondes Pavelski, Myriam Regattieri Delgado, and Marie-Eléonore Kessaci. + Meta-Learning on Flowshop Using Fitness Landscape Analysis. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  925–933. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2360] +
+
+Judea Pearl. + Heuristics: Intelligent Search Strategies for Computer Problem Solving. + Addison-Wesley, Reading, MA, 1984.
+[ bib ] + +
+ + +
+[2361] +
+
+Glen S. Peace. + Taguchi Methods: A Hands-On Approach. + Addison-Wesley, 1993.
+[ bib ] + +
+ + +
+[2362] +
+
+Judea Pearl. + The do-calculus revisited. + In N. de Freitas and K. Murphy, editors, Proceedings of the Twenty-Eighth Conference on Uncertainty in Artificial Intelligence (UAI'12), Catalina Island, CA August 14-18 2012, pp.  4–11. AUAI Press, 2013.
+[ bib ] + +
+ + +
+[2363] +
+
+Judea Pearl and Elias Bareinboim. + Transportability of causal and statistical relations: A formal approach. + In W. Burgard and D. Roth, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  247–254. AAAI Press, 2011.
+[ bib ] + +
+ + +
+[2364] +
+
+Judea Pearl and Dana Mackenzie. + The book of why: the new science of cause and effect. + Basic books, 2018.
+[ bib ] + +
+ + +
+[2365] +
+
+Judea Pearl. + Causality: Models, Reasoning and Inference. + Cambridge University Press, 2nd edition, 2009.
+[ bib ] + +
+ + +
+[2366] +
+
+Juan A. Pedraza, Carlos García-Martínez, Alberto Cano, and Sebastián Ventura. + Classification Rule Mining with Iterated Greedy. + In M. M. Polycarpou, A. C. P. L. F. de Carvalho, J. Pan, M. Wozniak, H. Quintián, and E. Corchado, editors, Hybrid Artificial Intelligence Systems - 9th International Conference, HAIS 2014, Salamanca, Spain, June 11-13, 2014. Proceedings, volume 8480 of Lecture Notes in Computer Science, pp.  585–596. Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2367] +
+
+Luciana R. Pedro and R. H. C. Takahashi. + Decision-Maker Preference Modeling in Interactive Multiobjective Optimization. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  811–824. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] +
+Keywords: decision-maker, interactive, neural networks +
+ +
+ + +
+[2368] +
+
+Paola Pellegrini and Mauro Birattari. + Implementation Effort and Performance. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2007, volume 4638 of Lecture Notes in Computer Science, pp.  31–45. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2369] +
+
+Paola Pellegrini, D. Favaretto, and E. Moretti. + On Max-Min Ant System's Parameters. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  203–214. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2370] +
+
+Paola Pellegrini, D. Favaretto, and E. Moretti. + Exploration in stochastic algorithms: An application on Max-Min Ant System. + In N. Krasnogor, B. Melián-Batista, J. A. Moreno-Pérez, J. M. Moreno-Vega, and D. A. Pelta, editors, Nature Inspired Cooperative Strategies for Optimization (NICSO 2008), volume 236 of Studies in Computational Intelligence, pp.  1–13. Springer, Berlin, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2371] +
+
+Paola Pellegrini, Thomas Stützle, and Mauro Birattari. + Off-line vs. On-line Tuning: A Study on Max-Min Ant System for the TSP. + In M. Dorigo et al., editors, Swarm Intelligence, 7th International Conference, ANTS 2010, volume 6234 of Lecture Notes in Computer Science, pp.  239–250. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2372] +
+
+Leslie Pérez Cáceres, Bernd Bischl, and Thomas Stützle. + Evaluating random forest models for irace. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2017, pp.  1146–1153. ACM Press, New York, NY, 2017.
+[ bib | +DOI ] + +
+ + +
+[2373] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, Holger H. Hoos, and Thomas Stützle. + An Experimental Study of Adaptive Capping in irace. + In R. Battiti, D. E. Kvasov, and Y. D. Sergeyev, editors, Learning and Intelligent Optimization, 11th International Conference, LION 11, volume 10556 of Lecture Notes in Computer Science, pp.  235–250. Springer, Cham, Switzerland, 2017.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[2374] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, Holger H. Hoos, and Thomas Stützle. + An experimental study of adaptive capping in irace: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2016-007/, 2017.
+[ bib ] + +
+ + +
+[2375] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, and Thomas Stützle. + Ant Colony Optimization on a Budget of 1000. + In M. Dorigo et al., editors, Swarm Intelligence, 9th International Conference, ANTS 2014, volume 8667 of Lecture Notes in Computer Science, pp.  50–61. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[2376] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, and Thomas Stützle. + An Analysis of Parameters of irace. + In C. Blum and G. Ochoa, editors, Proceedings of EvoCOP 2014 – 14th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 8600 of Lecture Notes in Computer Science, pp.  37–48. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] + +
+ + +
+[2377] +
+
+Leslie Pérez Cáceres, Manuel López-Ibáñez, and Thomas Stützle. + Ant Colony Optimization on a Budget of 1000: Supplementary material, 2015.
+[ bib | +http ] + +
+ + +
+[2378] +
+
+Leslie Pérez Cáceres, Federico Pagnozzi, Alberto Franzin, and Thomas Stützle. + Automatic Configuration of GCC Using irace. + In E. Lutton, P. Legrand, P. Parrend, N. Monmarché, and M. Schoenauer, editors, EA 2017: Artificial Evolution, volume 10764 of Lecture Notes in Computer Science, pp.  202–216. Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] +
+Automatic algorithm configuration techniques have proved to + be successful in finding performance-optimizing parameter + settings of many search-based decision and optimization + algorithms. A recurrent, important step in software + development is the compilation of source code written in some + programming language into machine-executable code. The + generation of performance-optimized machine code itself is a + difficult task that can be parametrized in many different + possible ways. While modern compilers usually offer different + levels of optimization as possible defaults, they have a + larger number of other flags and numerical parameters that + impact properties of the generated machine-code. While the + generation of performance-optimized machine code has received + large attention and is dealt with in the research area of + auto-tuning, the usage of standard automatic algorithm + configuration software has not been explored, even though, as + we show in this article, the performance of the compiled code + has significant stochasticity, just as standard optimization + algorithms. As a practical case study, we consider the + configuration of the well-known GNU compiler collection (GCC) + for minimizing the run-time of machine code for various + heuristic search methods. Our experimental results show that, + depending on the specific code to be optimized, improvements + of up to 40% of execution time when compared to the -O2 + and -O3 optimization flags is possible. +
+ +
+ + +
+[2379] +
+
+Leslie Pérez Cáceres, Federico Pagnozzi, Alberto Franzin, and Thomas Stützle. + Automatic configuration of GCC using irace: Supplementary material. + http://iridia.ulb.ac.be/supp/IridiaSupp2017-009/, 2017.
+[ bib ] + +
+ + +
+[2380] +
+
+Leslie Pérez Cáceres and Thomas Stützle. + Automatic Algorithm Configuration: Analysis, Improvements and Applications. + PhD thesis, IRIDIA, École polytechnique, Université Libre de Bruxelles, Belgium, 2017.
+[ bib | +epub ] +
+Supervised by Thomas Stützle and Manuel López-Ibáñez +
+ +
+ + +
+[2381] +
+
+James E. Pettinger and Richard M. Everson. + Controlling genetic algorithms with reinforcement learning. + In W. B. Langdon et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2002, pp.  692–692. Morgan Kaufmann Publishers, San Francisco, CA, 2002.
+[ bib ] + +
+ + +
+[2382] +
+
+Jonas Peters, Dominik Janzing, and Bernhard Schölkopf. + Elements of causal inference: foundations and learning algorithms. + MIT Press, 2017.
+[ bib ] + +
+ + +
+[2383] +
+
+Frank Phillipson and Harshil Singh Bhatia. + Portfolio Optimisation Using the D-Wave Quantum Annealer. + In M. Paszynski, D. Kranzlmüller, V. V. Krzhizhanovskaya, J. J. Dongarra, and P. M. A. Sloot, editors, Computational Science – ICCS 2021, pp.  45–59. Springer International Publishing, Cham, Switzerland, 2021.
+[ bib ] + +
+ + +
+[2384] +
+
+Josef Pihera and Nysret Musliu. + Application of Machine Learning to Algorithm Selection for TSP. + In G. A. Papadopoulos, editor, 26th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2014, Limassol, Cyprus, November 10-12, 2014, pp.  47–54. IEEE Press, 2014.
+[ bib ] + +
+ + +
+[2385] +
+
+M. L. Pilat and T. White. + Using Genetic Algorithms to optimize ACS-TSP. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  282–287. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2386] +
+
+Michael L. Pinedo. + Scheduling: Theory, Algorithms, and Systems. + Springer, New York, NY, 4th edition, 2012.
+[ bib ] + +
+ + +
+[2387] +
+
+Pedro Pinto, Thomas Runkler, and João Sousa. + Ant Colony Optimization and its Application to Regular and Dynamic MAX-SAT Problems. + In Advances in Biologically Inspired Information Systems, volume 69 of Studies in Computational Intelligence, pp.  285–304. Springer, Berlin, Germany, 2007.
+[ bib | +DOI ] +
+In this chapter we discuss the ant colony + optimization meta-heuristic (ACO) and its + application to static and dynamic constraint + satisfaction optimization problems, in particular + the static and dynamic maximum satisfiability + problems (MAX-SAT). In the first part of the + chapter we give an introduction to meta-heuristics + in general and ant colony optimization in + particular, followed by an introduction to + constraint satisfaction and static and dynamic + constraint satisfaction optimization problems. + Then, we describe how to apply the ACO algorithm + to the problems, and do an analysis of the results + obtained for several benchmarks. The adapted ant + colony optimization accomplishes very well the task + of dealing with systematic changes of dynamic + MAX-SAT instances derived from static problems. +
+ +
+ + +
+[2388] +
+
+Joelle Pineau and Koustuv Sinha. + The Machine Learning Reproducibility Checklist (v2.0). + https://www.cs.mcgill.ca/~jpineau/ReproducibilityChecklist-v2.0.pdf, 2020.
+[ bib ] +
+Used in NeurIPS 2020 +
+ +
+ + +
+[2389] +
+
+David Pisinger and Stefan Ropke. + Large Neighborhood Search. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, pp.  399–419. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[2390] +
+
+Erik Pitzer, Andreas Beham, and Michael Affenzeller. + Automatic Algorithm Selection for the Quadratic Assignment Problem Using Fitness Landscape Analysis. + In M. Middendorf and C. Blum, editors, Proceedings of EvoCOP 2013 – 13th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 7832 of Lecture Notes in Computer Science, pp.  109–120. Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2391] +
+
+Dmitry Plotnikov, Dmitry Melnik, Mamikon Vardanyan, Ruben Buchatskiy, Roman Zhuykov, and Je-Hyung Lee. + Automatic Tuning of Compiler Optimizations and Analysis of their Impact. + In V. Alexandrov, M. Lees, V. Krzhizhanovskaya, J. Dongarra, and P. M. Sloot, editors, 2013 International Conference on Computational Science, volume 18 of Procedia Computer Science, pp.  1312–1321. Elsevier, 2013.
+[ bib | +DOI ] + +
+ + +
+[2392] +
+
+Robert Plotnick. + The Genie in the Machine: How Computer-Automated Inventing Is Revolutionizing Law and Business. + Stanford Law Books, 2009.
+[ bib ] +
+Mentions evolutionary optimization of Oral-B toothbrushes +
+ +
+ + +
+[2393] +
+
+Michael J. D. Powell. + The BOBYQA algorithm for bound constrained optimization without derivatives. + Technical Report Cambridge NA Report NA2009/06, University of Cambridge, UK, 2009.
+[ bib | +epub ] + +
+ + +
+[2394] +
+
+Michael J. D. Powell. + A Direct Search Optimization Method That Models the Objective and Constraint Functions by Linear Interpolation. + In Advances in Optimization and Numerical Analysis, pp.  51–67. Springer, 1994.
+[ bib | +DOI ] +
+Proposed COBYLA +
+ +
+ + +
+[2395] +
+
+Raphael Patrick Prager, Heike Trautmann, Hao Wang, Thomas Bäck, and Pascal Kerschke. + Per-Instance Configuration of the Modularized CMA-ES by Means of Classifier Chains and Exploratory Landscape Analysis. + In C. A. Coello Coello, editor, 2020 IEEE Symposium Series on Computational Intelligence, SSCI 2020, Canberra, Australia, December 1-4, 2020, pp.  996–1003. IEEE Press, 2020.
+[ bib ] + +
+ + +
+[2396] +
+
+Kata Praditwong and Xin Yao. + A new multi-objective evolutionary optimisation algorithm: the two-archive algorithm. + In International Conference on Computational Intelligence and Security, volume 1, pp.  286–291. IEEE, 2006.
+[ bib ] + +
+ + +
+[2397] +
+
+T. Devi Prasad and Godfrey A. Walters. + Optimal rerouting to minimise residence times in water distribution networks. + In C. Maksimović, D. Butler, and F. A. Memon, editors, Advances in Water Supply Management, pp.  299–306. CRC Press, 2003.
+[ bib ] + +
+ + +
+[2398] +
+
+F. P. Preparata and M. I. Shamos. + Computational Geometry. An Introduction. + Springer, Berlin, Germany, 2nd edition, 1988.
+[ bib ] + +
+ + +
+[2399] +
+
+Stefan Pricopie, Richard Allmendinger, Manuel López-Ibáñez, Clyde Fare, Matt Benatan, and Joshua D. Knowles. + Expensive Optimization with Production-Graph Resource Constraints: A First Look at a New Problem Class. + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, pp.  840–848. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] +
+We consider a new class of expensive, resource-constrained + optimization problems (here arising from molecular discovery) + where costs are associated with the experiments (or + evaluations) to be carried out during the optimization + process. In the molecular discovery problem, candidate + compounds to be optimized must be synthesized in an iterative + process that starts from a set of purchasable items and + builds up to larger molecules. To produce target molecules, + their required resources are either used from + already-synthesized items in storage or produced themselves + on-demand at an additional cost. Any remaining resources from + the production process are stored for reuse for the next + evaluations. We model these resource dependencies with a + directed acyclic production graph describing the development + process from granular purchasable items to evaluable target + compounds. Moreover, we develop several resource-eficient + algorithms to address this problem. In particular, we develop + resource-aware variants of Random Search heuristics and of + Bayesian Optimization and analyze their performance in terms + of anytime behavior. The experimental results were obtained + from a real-world molecular optimization problem. Our results + suggest that algorithms that encourage exploitation by + reusing existing resources achieve satisfactory results while + using fewer resources overall. +
+
+Keywords: molecular discovery, resource constraints, expensive + optimization, production costs +
+ +
+ + +
+[2400] +
+
+Kenneth Price, Rainer M. Storn, and Jouni A. Lampinen. + Differential Evolution: A Practical Approach to Global Optimization. + Springer, New York, NY, 2005.
+[ bib | +DOI ] + +
+ + +
+[2401] +
+
+Andy Pryke, Sanaz Mostaghim, and Alireza Nazemi. + Heatmap visualization of population based multi objective algorithms. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  361–375. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2402] +
+
+Gregorio Toscano Pulido and Carlos A. Coello Coello. + The Micro Genetic Algorithm 2: Towards Online Adaptation in Evolutionary Multiobjective Optimization. + In C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors, Evolutionary Multi-criterion Optimization, EMO 2003, volume 2632 of Lecture Notes in Computer Science, pp.  252–266. Springer, Heidelberg, Germany, 2003.
+[ bib | +DOI ] + +
+ + +
+[2403] +
+
+Robin C. Purshouse, Kalyanmoy Deb, Maszatul M. Mansor, Sanaz Mostaghim, and Rui Wang. + A review of hybrid evolutionary multiple criteria decision making methods. + COIN Report 2014005, Computational Optimization and Innovation (COIN) Laboratory, University of Michigan, USA, January 2014.
+[ bib ] + +
+ + +
+[2404] +
+
+Robin C. Purshouse and Peter J. Fleming. + Evolutionary many-objective optimisation: an exploratory analysis. + In Proceedings of the 2003 Congress on Evolutionary Computation (CEC'03), pp.  2066–2073, Piscataway, NJ, December 2003. IEEE Press.
+[ bib | +DOI ] +
+First to mention NSGA-II failure to deal with + many-objectives. Mentions exponential number of nondominated + solutions with respect to many objectives (but + [1816] already did). +
+ +
+ + +
+[2405] +
+
+Markus Püschel, Franz Franchetti, and Yevgen Voronenko. + Spiral. + In D. Padua, editor, Encyclopedia of Parallel Computing, pp.  1920–1933. Springer, US, 2011.
+[ bib | +DOI ] + +
+ + +
+[2406] +
+
+Yasha Pushak and Holger H. Hoos. + Algorithm Configuration Landscapes: More Benign Than Expected? + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11101 of Lecture Notes in Computer Science, pp.  271–283. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI | +supplementary material ] +
+Best paper award at PPSN2018 +
+ +
+ + +
+[2407] +
+
+Yasha Pushak and Holger H. Hoos. + Golden parameter search: exploiting structure to quickly configure parameters in parallel. + In C. A. Coello Coello, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020, pp.  245–253. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] +
+Keywords: algorithm configuration +
+ +
+ + +
+[2408] +
+
+Bernd Bischl, Michel Lang, Jakob Bossek, Daniel Horn, Karin Schork, Jakob Richter, and Pascal Kerschke. + ParamHelpers : Helpers for Parameters in Black-Box Optimization, Tuning and Machine Learning, 2017. + R package version 1.10.
+[ bib | +http ] + +
+ + +
+[2409] +
+
+Hao Yu. + Rmpi: Interface (Wrapper) to MPI (Message-Passing Interface), 2010. + R package version 0.5-8.
+[ bib | +http ] + +
+ + +
+[2410] +
+
+Thomas Bartz-Beielstein, J. Ziegenhirt, W. Konen, O. Flasch, P. Koch, and Martin Zaefferer. + SPOT: Sequential Parameter Optimization, 2011. + R package.
+[ bib | +http ] + +
+ + +
+[2411] +
+
+Heike Trautmann, Olaf Mersmann, and David Arnu. + cmaes: Covariance Matrix Adapting Evolutionary Strategy, 2011. + R package.
+[ bib | +http ] + +
+ + +
+[2412] +
+
+Rob Carnell. + lhs: Latin Hypercube Samples, 2016. + R package version 0.14.
+[ bib | +http ] + +
+ + +
+[2413] +
+
+Olaf Mersmann. + mco: Multiple Criteria Optimization Algorithms and Related Functions, 2014. + R package version 1.0-15.1.
+[ bib | +http ] + +
+ + +
+[2414] +
+
+Bernd Bischl, Michel Lang, Jakob Bossek, Leonard Judt, Jakob Richter, Tobias Kuehn, and Erich Studerus. + mlr: Machine Learning in R, 2013. + R package.
+[ bib | +http ] + +
+ + +
+[2415] +
+
+Simon Urbanek. + multicore: Parallel Processing of R Code on Machines with Multiple Cores or CPUs, 2010. + R package version 0.1-3.
+[ bib | +http ] + +
+ + +
+[2416] +
+
+Jakob Bossek. + smoof: Single and Multi-Objective Optimization Test Functions, 2016. + R package version 1.2.
+[ bib | +http ] + +
+ + +
+[2417] +
+
+L. Rachmawati and D. Srinivasan. + Preference incorporation in multiobjective evolutionary algorithms: A survey. + In Proceedings of the 2006 Congress on Evolutionary Computation (CEC 2006), pp.  3385–3391, Piscataway, NJ, July 2006. IEEE Press.
+[ bib ] + +
+ + +
+[2418] +
+
+Andreea Radulescu, Manuel López-Ibáñez, and Thomas Stützle. + Automatically Improving the Anytime Behaviour of Multiobjective Evolutionary Algorithms. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  825–840. Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2419] +
+
+Alma A. M. Rahat, Richard M. Everson, and Jonathan E. Fieldsend. + Alternative infill strategies for expensive multi-objective optimisation. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  873–880. ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[2420] +
+
+Marcus Randall. + Near Parameter Free Ant Colony Optimisation. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  374–381. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2421] +
+
+Marcus Randall and James Montgomery. + Candidate Set Strategies for Ant Colony Optimisation. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  243–249. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2422] +
+
+Zhengfu Rao, Jon Wicks, and Sue West. + ENCOMS - An Energy Cost Minimisation System for Real-Time, Operational Control of Water Distribution Networks. + In D. A. Savic, G. A. Walters, R. King, and S. Thiam-Khu, editors, Proceedings of the Eighth International Conference on Computing and Control for the Water Industry (CCWI 2005), volume 1, pp.  85–90, University of Exeter, UK, September 2005.
+[ bib ] + +
+ + +
+[2423] +
+
+Elena Raponi, Hao Wang, Mariusz Bujny, Simonetta Boria, and Carola Doerr. + High Dimensional Bayesian Optimization Assisted by Principal Component Analysis. + In T. Bäck, M. Preuss, A. Deutz, H. Wang, C. Doerr, M. T. M. Emmerich, and H. Trautmann, editors, Parallel Problem Solving from Nature – PPSN XVI, volume 12269 of Lecture Notes in Computer Science, pp.  169–183. Springer, Cham, Switzerland, 2020.
+[ bib | +DOI ] + +
+ + +
+[2424] +
+
+Jussi Rasku, Nysret Musliu, and Tommi Kärkkäinen. + Automating the Parameter Selection in VRP: An Off-line Parameter Tuning Tool Comparison. + In W. Fitzgibbon, Y. A. Kuznetsov, P. Neittaanmäki, and O. Pironneau, editors, Modeling, Simulation and Optimization for Science and Technology, volume 34 of Computational Methods in Applied Sciences, pp.  191–209. Springer, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[2425] +
+
+Carl Edward Rasmussen and Christopher K. I. Williams. + Gaussian Processes for Machine Learning. + MIT Press, Cambridge, MA, 2006.
+[ bib ] +
+Keywords: Gaussian processes, data processing +
+ +
+ + +
+[2426] +
+
+N. Rayner. + Maverick Research: Judgment Day, or Why We Should Let Machines Automate Decision Making. + Gartner research note, Gartner, Inc, October 2011.
+[ bib ] + +
+ + +
+[2427] +
+
+Ingo Rechenberg. + Evolutionsstrategie: Optimierung technischer Systeme nach Prinzipien der biologischen Evolution. + PhD thesis, Department of Process Engineering, Technical University of Berlin, 1971.
+[ bib ] + +
+ + +
+[2428] +
+
+Ingo Rechenberg. + Evolutionsstrategie: Optimierung technischer Systeme nach Prinzipien der biologischen Evolution. + Frommann-Holzboog, Stuttgart, Germany, 1973.
+[ bib ] + +
+ + +
+[2429] +
+
+Colin R. Reeves. + Genetic algorithms. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, chapter 5, pp.  109–140. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[2430] +
+
+Patrick M. Reed. + Many-Objective Visual Analytics: Rethinking the Design of Complex Engineered Systems. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  1–1. Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2431] +
+
+Marc Reimann. + Guiding ACO by Problem Relaxation: A Case Study on the Symmetric TSP. + In T. Bartz-Beielstein, M. J. Blesa, C. Blum, B. Naujoks, A. Roli, G. Rudolph, and M. Sampels, editors, Hybrid Metaheuristics, volume 4771 of Lecture Notes in Computer Science, pp.  45–56. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2432] +
+
+Gerhard Reinelt. + The Traveling Salesman: Computational Solutions for TSP Applications, volume 840 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1994.
+[ bib ] + +
+ + +
+[2433] +
+
+Mauricio G. C. Resende and Celso C. Ribeiro. + Greedy Randomized Adaptive Search Procedures. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  219–249. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[2434] +
+
+Mauricio G. C. Resende and Celso C. Ribeiro. + Greedy Randomized Adaptive Search Procedures: Advances, Hybridizations, and Applications. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science, pp.  283–319. Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[2435] +
+
+Margarita Reyes Sierra and Carlos A. Coello Coello. + Improving PSO-Based Multi-objective Optimization Using Crowding, Mutation and ε-Dominance. + In C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors, Evolutionary Multi-criterion Optimization, EMO 2005, volume 3410 of Lecture Notes in Computer Science, pp.  505–519. Springer, Heidelberg, Germany, 2005.
+[ bib ] +
+Keywords: OMOPSO +
+ +
+ + +
+[2436] +
+
+Mona Riabacke, Mats Danielson, Love Ekenberg, and Aron Larsson. + A Prescriptive Approach for Eliciting Imprecise Weight Statements in an MCDA Process. + In F. Rossi and A. Tsoukiàs, editors, Algorithmic Decision Theory, First International Conference, ADT 2009, volume 5783 of Lecture Notes in Computer Science, pp.  168–179. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2437] +
+
+Enda Ridge and Daniel Kudenko. + Tuning the Performance of the MMAS Heuristic. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2007, volume 4638 of Lecture Notes in Computer Science, pp.  46–60. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2438] +
+
+Enda Ridge and Daniel Kudenko. + Tuning an Algorithm Using Design of Experiments. + In T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors, Experimental Methods for the Analysis of Optimization Algorithms, pp.  265–286. Springer, Berlin, Germany, 2010.
+[ bib ] + +
+ + +
+[2439] +
+
+Sander van Rijn, Hao Wang, Matthijs van Leeuwen, and Thomas Bäck. + Evolving the structure of Evolution Strategies. + In X. Chen and A. Stafylopatis, editors, Computational Intelligence (SSCI), 2016 IEEE Symposium Series on, pp.  1–8, 2016.
+[ bib | +DOI ] +
+Keywords: automated design, automatic configuration, CMA-ES, Gaussian + distribution +
+ +
+ + +
+[2440] +
+
+R Development Core Team. + R: A Language and Environment for Statistical Computing. + R Foundation for Statistical Computing, Vienna, Austria, 2008.
+[ bib | +http ] + +
+ + +
+[2441] +
+
+Tea Robič and Bogdan Filipič. + DEMO: Differential Evolution for Multiobjective Optimization. + In C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors, Evolutionary Multi-criterion Optimization, EMO 2005, volume 3410 of Lecture Notes in Computer Science, pp.  520–533. Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2442] +
+
+Francisco J. Rodríguez, Christian Blum, Manuel Lozano, and Carlos García-Martínez. + Iterated Greedy Algorithms for the Maximal Covering Location Problem. + In J.-K. Hao and M. Middendorf, editors, Proceedings of EvoCOP 2012 – 12th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 7245 of Lecture Notes in Computer Science, pp.  172–181. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2443] +
+
+Cynthia A. Rodríguez Villalobos and Carlos A. Coello Coello. + A new multi-objective evolutionary algorithm based on a performance assessment indicator. + In T. Soule and J. H. Moore, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2012, pp.  505–512. ACM Press, New York, NY, 2012.
+[ bib ] + +
+ + +
+[2444] +
+
+Peter Ross. + Hyper-Heuristics. + In E. K. Burke and G. Kendall, editors, Search Methodologies, pp.  529–556. Springer, Boston, MA, 2005.
+[ bib | +DOI ] + +
+ + +
+[2445] +
+
+Frank Rubin. + An Iterative Technique for Printed Wire Routing. + In DAC'74, Proceedings of the 11th Design Automation Workshop, pp.  308–313. IEEE Press, 1974.
+[ bib ] + +
+ + +
+[2446] +
+
+Günther Rudolph and Alexandru Agapie. + Convergence Properties of Some Multi-Objective Evolutionary Algorithms. + In Proceedings of the 2000 Congress on Evolutionary Computation (CEC'00), volume 2, pp.  1010–1016, Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[2447] +
+
+Günther Rudolph, Heike Trautmann, Soumyadip Sengupta, and Oliver Schütze. + Evenly spaced Pareto front approximations for tricriteria problems based on triangulation. + In R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors, Evolutionary Multi-criterion Optimization, EMO 2013, volume 7811 of Lecture Notes in Computer Science, pp.  443–458. Springer, Heidelberg, Germany, 2013.
+[ bib ] +
+unbounded archiver, AAΔ_1 +
+ +
+ + +
+[2448] +
+
+Günther Rudolph. + Globale Optimierung mit parallelen Evolutionsstrategien. + Diplomarbeit, Department of Computer Science, University of Dortmund, July 1990.
+[ bib ] +
+Proposed the generalized Rastrigin function +
+ +
+ + +
+[2449] +
+
+Günther Rudolph. + On Correlated Mutations in Evolution Strategies. + In R. Männer and B. Manderick, editors, Parallel Problem Solving from Nature – PPSN II, pp.  107–116. Elsevier, 1992.
+[ bib ] + +
+ + +
+[2450] +
+
+Günther Rudolph. + Convergence of non-elitist strategies. + In Z. Michalewicz, editor, Proceedings of the First IEEE International Conference on Evolutionary Computation (ICEC'94), pp.  63–66. IEEE Press, Piscataway, NJ, 1994.
+[ bib ] + +
+ + +
+[2451] +
+
+Günther Rudolph. + Evolutionary Search for Minimal Elements in Partially Ordered Finite Sets. + In V. W. Porto, N. Saravanan, D. E. Waagen, and A. E. Eiben, editors, International Conference on Evolutionary Programming, volume 1447 of Lecture Notes in Computer Science, pp.  345–353. Springer, 1998.
+[ bib | +DOI ] + +
+ + +
+[2452] +
+
+Ana Belén Ruiz, Mariano Luque, Kaisa Miettinen, and Rubén Saborido. + An Interactive Evolutionary Multiobjective Optimization Method: Interactive WASF-GA. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part II, volume 9019 of Lecture Notes in Computer Science, pp.  249–263. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] + +
+ + +
+[2453] +
+
+Rubén Ruiz, Eva Vallada, and Carlos Fernández-Martínez. + Scheduling in flowshops with no-idle machines. + In Computational intelligence in flow shop and job shop scheduling, pp.  21–51. Springer, 2009.
+[ bib ] + +
+ + +
+[2454] +
+
+W. Ruml. + Incomplete Tree Search using Adaptive Probing. + In B. Nebel, editor, Proceedings of the 17th International Joint Conference on Artificial Intelligence (IJCAI-01), pp.  235–241. IEEE Press, 2001.
+[ bib ] + +
+ + +
+[2455] +
+
+Stuart J. Russell and Peter Norvig. + Artificial Intelligence: A Modern Approach, volume 2. + Prentice Hall, Englewood Cliffs, NJ, 2003.
+[ bib ] + +
+ + +
+[2456] +
+
+John Rust. + Structural estimation of Markov decision processes. + In Handbook of Econometrics, volume 4, pp.  3081–3143. Elsevier, 1994.
+[ bib | +DOI ] + +
+ + +
+[2457] +
+
+Michael Behrisch, Laura Bieker, Jakob Erdmann, and Daniel Krajzewicz. + SUMO - Simulation of Urban MObility: An Overview. + In SIMUL 2011, The Third International Conference on Advances in System Simulation, pp.  63–68, Barcelona, Spain, 2011. ThinkMind.
+[ bib ] + +
+ + +
+[2458] +
+
+Bhupinder Singh Saini, Manuel López-Ibáñez, and Kaisa Miettinen. + Automatic Surrogate Modelling Technique Selection based on Features of Optimization Problems. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2019, pp.  1765–1772. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+A typical scenario when solving industrial single or + multiobjective optimization problems is that no explicit + formulation of the problem is available. Instead, a dataset + containing vectors of decision variables together with their + objective function value(s) is given and a surrogate model + (or metamodel) is build from the data and used for + optimization and decision-making. This data-driven + optimization process strongly depends on the ability of the + surrogate model to predict the objective value of decision + variables not present in the original dataset. Therefore, the + choice of surrogate modelling technique is crucial. While + many surrogate modelling techniques have been discussed in + the literature, there is no standard procedure that will + select the best technique for a given problem. In this work, + we propose the automatic selection of a surrogate modelling + technique based on exploratory landscape features of the + optimization problem that underlies the given dataset. The + overall idea is to learn offline from a large pool of + benchmark problems, on which we can evaluate a large number + of surrogate modelling techniques. When given a new dataset, + features are used to select the most appropriate surrogate + modelling technique. The preliminary experiments reported + here suggest that the proposed automatic selector is able to + identify high-accuracy surrogate models as long as an + appropriate classifier is used for selection. +
+ +
+ + +
+[2459] +
+
+Yoshitaka Sakurai, Kouhei Takada, Takashi Kawabe, and Setsuo Tsuruta. + A method to control parameters of evolutionary algorithms by using reinforcement learning. + In 2010 Sixth International Conference on Signal-Image Technology and Internet Based Systems, pp.  74–79. IEEE, 2010.
+[ bib ] + +
+ + +
+[2460] +
+
+A. Burcu Altan Sakarya, Fred E. Goldman, and Larry W. Mays. + Models for the optimal scheduling of pumps to meet water quality. + In D. A. Savic and G. A. Walters, editors, Water Industry Systems: Modelling and Optimization Applications, volume 2, pp.  379–391. Research Studies Press Ltd., Baldock, United Kingdom, 1999.
+[ bib ] + +
+ + +
+[2461] +
+
+Francesco Sambo, Barbara Di Camillo, Alberto Franzin, Andrea Facchinetti, Liisa Hakaste, Jasmina Kravic, Giuseppe Fico, Jaakko Tuomilehto, Leif Groop, Rafael Gabriel, Tiinamaija Tuomi, and Claudio Cobelli. + A Bayesian Network analysis of the probabilistic relations between risk factors in the predisposition to type 2 diabetes. + In N. Lovell and L. Mainardi, editors, 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2015, Proceedings, Annual International Conference of the IEEE Engineering in Medicine and Biology, pp.  2119–2122. IEEE Press, 2015.
+[ bib ] + +
+ + +
+[2462] +
+
+Valentino Santucci and Marco Baioletti. + A Fast Randomized Local Search for Low Budget Optimization in Black-Box Permutation Problems. + In Proceedings of the 2022 World Congress on Computational Intelligence (WCCI 2022), Piscataway, NJ, 2022. IEEE Press.
+[ bib ] +
+Keywords: UMM, CEGO +
+ +
+ + +
+[2463] +
+
+Thomas J. Santner, Brian J. Williams, and William I. Notz. + The Design and Analysis of Computer Experiments. + Springer Verlag, New York, NY, 2003.
+[ bib | +DOI ] + +
+ + +
+[2464] +
+
+Kaz Sato and Cliff Young. + An in-depth look at Google's first Tensor Processing Unit (TPU). + https://cloud.google.com/blog/products/ai-machine-learning/an-in-depth-look-at-googles-first-tensor-processing-unit-tpu, 2017.
+[ bib ] + +
+ + +
+[2465] +
+
+Dragan A. Savic, Godfrey A. Walters, and Martin Schwab. + Multiobjective Genetic Algorithms for Pump Scheduling in Water Supply. + In D. Corne and J. L. Shapiro, editors, Evolutionary Computing Workshop, AISB'97, volume 1305 of Lecture Notes in Computer Science, pp.  227–236. Berlin, Germany, 1997.
+[ bib | +.ps ] + +
+ + +
+[2466] +
+
+Y. Sawaragi, H. Nakayama, and T. Tanino. + Theory of multiobjective optimization. + Elsevier, 1985.
+[ bib ] + +
+ + +
+[2467] +
+
+Dhish Kumar Saxena and Kalyanmoy Deb. + Non-linear Dimensionality Reduction Procedures for Certain Large-Dimensional Multi-objective Optimization Problems: Employing Correntropy and a Novel Maximum Variance Unfolding. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  772–787. Springer, Heidelberg, Germany, 2007.
+[ bib | +DOI ] +
+In our recent publication, we began with an understanding + that many real-world applications of multi-objective + optimization involve a large number (10 or more) of + objectives but then, existing evolutionary multi-objective + optimization (EMO) methods have primarily been applied to + problems having smaller number of objectives (5 or + less). After highlighting the major impediments in handling + large number of objectives, we proposed a principal component + analysis (PCA) based EMO procedure, for dimensionality + reduction, whose efficacy was demonstrated by solving upto + 50-objective optimization problems. Here, we are addressing + the fact that, when the data points live on a non-linear + manifold or that the data structure is non-gaussian, PCA + which yields a smaller dimensional 'linear' subspace may be + ineffective in revealing the underlying dimensionality. To + overcome this, we propose two new non-linear dimensionality + reduction algorithms for evolutionary multi-objective + optimization, namely C-PCA-NSGA-II and MVU-PCA-NSGA-II. While + the former is based on the newly introduced correntropy PCA + [2], the later implements maximum variance unfolding + principle [3,4,5] in a novel way. We also establish the + superiority of these new EMO procedures over the earlier + PCA-based procedure, both in terms of accuracy and + computational time, by solving upto 50-objective optimization + problems. +
+ +
+ + +
+[2468] +
+
+Dhish Kumar Saxena and Kalyanmoy Deb. + Trading on infeasibility by exploiting constraint's criticality through multi-objectivization: A system design perspective. + In Proceedings of the 2007 Congress on Evolutionary Computation (CEC 2007), pp.  919–926, Piscataway, NJ, 2007. IEEE Press.
+[ bib | +DOI ] +
+Keywords: multi-objectivization +
+ +
+ + +
+[2469] +
+
+Dhish Kumar Saxena and Kalyanmoy Deb. + Dimensionality reduction of objectives and constraints in multi-objective optimization problems: A system design perspective. + In Proceedings of the 2008 Congress on Evolutionary Computation (CEC 2008), pp.  3204–3211, Piscataway, NJ, 2008. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2470] +
+
+J. David Schaffer. + Multiple Objective Optimization with Vector Evaluated Genetic Algorithms. + In J. J. Grefenstette, editor, Proceedings of the First International Conference on Genetic Algorithms (ICGA'85), pp.  93–100. Lawrence Erlbaum Associates, 1985.
+[ bib ] +
+Keywords: VEGA +
+ +
+ + +
+[2471] +
+
+Jeff G. Schneider. + Exploiting model uncertainty estimates for safe dynamic control learning. + In M. Mozer, M. I. Jordan, and T. Petsche, editors, Advances in Neural Information Processing Systems (NIPS 9), pp.  1047–1053. MIT Press, 1996.
+[ bib | +epub ] + +
+ + +
+[2472] +
+
+Oliver Schütze, X. Esquivel, A. Lara, and Carlos A. Coello Coello. + Some Comments on GD and IGD and Relations to the Hausdorff Distance. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  1971–1974. ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2473] +
+
+Oliver Schütze and Carlos Hernández. + Archiving Strategies for Evolutionary Multi-objective Optimization Algorithms. + Springer, 2021.
+[ bib ] + +
+ + +
+[2474] +
+
+Marius Schneider and Holger H. Hoos. + Quantifying Homogeneity of Instance Sets for Algorithm Configuration. + In Y. Hamadi and M. Schoenauer, editors, Learning and Intelligent Optimization, 6th International Conference, LION 6, volume 7219 of Lecture Notes in Computer Science, pp.  190–204. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] +
+Keywords: Quantifying Homogeneity; Empirical Analysis; Parameter + Optimization; Algorithm Configuration +
+ +
+ + +
+[2475] +
+
+Florian Schroff, Dmitry Kalenichenko, and James Philbin. + Facenet: A unified embedding for face recognition and clustering. + In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.  815–823, 2015.
+[ bib ] + +
+ + +
+[2476] +
+
+Michael Schmidt and Hod Lipson. + Age-Fitness Pareto Optimization. + In Genetic Programming Theory and Practice VIII. Genetic and Evolutionary Computation, pp.  129–146. Springer, 2011.
+[ bib | +DOI ] + +
+ + +
+[2477] +
+
+Jens Schreiter, Duy Nguyen-Tuong, Mona Eberts, Bastian Bischoff, Heiner Markert, and Marc Toussaint. + Safe Exploration for Active Learning with Gaussian Processes. + In Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2015, volume 9286 of Lecture Notes in Computer Science, pp.  133–149. Springer, 2015.
+[ bib | +DOI ] +
+Proposed Safe Active Learning (SAL) algorithm +
+ +
+ + +
+[2478] +
+
+Mark Schillinger, Benedikt Ortelt, Benjamin Hartmann, Jens Schreiter, Mona Meister, Duy Nguyen-Tuong, and Oliver Nelles. + Safe active learning of a high pressure fuel supply system. + In Proceedings of the 9th EUROSIM Congress on Modelling and Simulation, EUROSIM 2016 and the 57th SIMS Conference on Simulation and Modelling SIMS 2016, pp.  286–292. Linköping University Electronic Press, 2018.
+[ bib | +DOI ] + +
+ + +
+[2479] +
+
+Andrea Schaerf. + Combining Local Search and Look-Ahead for Scheduling and Constraint Satisfaction Problems. + In M. E. Pollack, editor, Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), volume 2, pp.  1254–1259. Morgan Kaufmann Publishers, 1997.
+[ bib ] + +
+ + +
+[2480] +
+
+Henry Scheffe. + The Analysis of Variance. + John Wiley & Sons, New York, NY, 1st edition, 1959.
+[ bib ] + +
+ + +
+[2481] +
+
+Hans-Paul Schwefel. + Numerische Optimierung von Computer–Modellen mittels der Evolutionsstrategie. + Birkhäuser, Basel, Switzerland, 1977.
+[ bib ] + +
+ + +
+[2482] +
+
+Sam Scott and Stan Matwin. + Feature engineering for text classification. + In ICML, volume 99, pp.  379–388, 1999.
+[ bib ] + +
+ + +
+[2483] +
+
+D. Sculley, Jasper Snoek, Ali Rahimi, and Alexander B. Wiltschko. + Winner's Curse? On Pace, Progress and Empirical Rigor. + In I. Murray, M. Ranzato, and O. Vinyals, editors, 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings, pp.  1–4. OpenReview.net, 2018.
+[ bib | +http ] + +
+ + +
+[2484] +
+
+Haitham Seada and Kalyanmoy Deb. + U-NSGA-III: A Unified Evolutionary Optimization Procedure for Single, Multiple, and Many Objectives: Proof-of-Principle Results. + In A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors, Evolutionary Multi-criterion Optimization, EMO 2015 Part I, volume 9018 of Lecture Notes in Computer Science, pp.  34–49. Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2485] +
+
+Moritz Seiler, Janina Pohl, Jakob Bossek, Pascal Kerschke, and Heike Trautmann. + Deep Learning as a Competitive Feature-Free Approach for Automated Algorithm Selection on the Traveling Salesperson Problem. + In T. Bäck, M. Preuss, A. Deutz, H. Wang, C. Doerr, M. T. M. Emmerich, and H. Trautmann, editors, Parallel Problem Solving from Nature – PPSN XVI, volume 12269 of Lecture Notes in Computer Science, pp.  48–64. Springer, Cham, Switzerland, 2020.
+[ bib ] + +
+ + +
+[2486] +
+
+Jendrik Seipp, Silvan Sievers, Malte Helmert, and Frank Hutter. + Automatic Configuration of Sequential Planning Portfolios. + In B. Bonet and S. Koenig, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  3364–3370. AAAI Press, 2015.
+[ bib ] + +
+ + +
+[2487] +
+
+P. Serafini. + Simulated annealing for multiple objective optimization problems. + In G. H. Tzeng and P. L. Yu, editors, Proceedings of the 10th International Conference on Multiple Criteria Decision Making (MCDM'91), volume 1, pp.  87–96. Springer Verlag, 1992.
+[ bib ] + +
+ + +
+[2488] +
+
+P. Serafini. + Some Considerations About Computational Complexity for Multiobjective Combinatorial Problems. + In J. Jahn and W. Krabs, editors, Recent Advances and Historical Development of Vector Optimization, volume 294 of Lecture Notes in Economics and Mathematical Systems, pp.  222–231. Springer, Berlin, Germany, 1986.
+[ bib ] + +
+ + +
+[2489] +
+
+K. J. Shaw, Carlos M. Fonseca, A. L. Nortcliffe, M. Thompson, J. Love, and Peter J. Fleming. + Assessing the performance of multiobjective genetic algorithms for optimization of a batch process scheduling problem. + In Proceedings of the 1999 Congress on Evolutionary Computation (CEC 1999), volume 1, pp.  34–75, Piscataway, NJ, 1999. IEEE Press.
+[ bib ] + +
+ + +
+[2490] +
+
+Ke Shang, Hisao Ishibuchi, and Weiyu Chen. + Greedy approximated hypervolume subset selection for many-objective optimization. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  448–456. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2491] +
+
+Ke Shang, Hisao Ishibuchi, and Yang Nan. + Distance-based subset selection revisited. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  439–447. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2492] +
+
+Mudita Sharma, Alexandros Komninos, Manuel López-Ibáñez, and Dimitar Kazakov. + Deep Reinforcement Learning-Based Parameter Control in Differential Evolution. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  709–717. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub | +supplementary material ] +
+Keywords: DE-DDQN +
+ +
+ + +
+[2493] +
+
+Mudita Sharma, Manuel López-Ibáñez, and Dimitar Kazakov. + Performance Assessment of Recursive Probability Matching for Adaptive Operator Selection in Differential Evolution. + In A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors, Parallel Problem Solving from Nature – PPSN XV, volume 11102 of Lecture Notes in Computer Science, pp.  321–333. Springer, Cham, Switzerland, 2018.
+[ bib | +DOI | +supplementary material ] +
+Keywords: Rec-PM +
+ +
+ + +
+[2494] +
+
+Mudita Sharma, Manuel López-Ibáñez, and Dimitar Kazakov. + Performance Assessment of Recursive Probability Matching for Adaptive Operator Selection in Differential Evolution: Supplementary material. + https://github.com/mudita11/AOS-comparisons, 2018.
+[ bib | +DOI ] + +
+ + +
+[2495] +
+
+Seyed Mahdi Shavarani, Manuel López-Ibáñez, and Joshua D. Knowles. + Realistic Utility Functions Prove Difficult for State-of-the-Art Interactive Multiobjective Optimization Algorithms. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  457–465. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2496] +
+
+Seyed Mahdi Shavarani, Manuel López-Ibáñez, and Joshua D. Knowles. + On Benchmarking Interactive Evolutionary Multi-Objective Algorithms: Supplementary material. + https://doi.org/10.5281/zenodo.7863301, 2023.
+[ bib ] + +
+ + +
+[2497] +
+
+Babooshka Shavazipour. + Multi-Objective Optimisation under Deep Uncertainty. + PhD thesis, UCT Statistical sciences, South Africa, 2018.
+[ bib | +epub ] + +
+ + +
+[2498] +
+
+Paul Shaw. + Using Constraint Programming and Local Search Methods to Solve Vehicle Routing Problems. + In M. Maher and J.-F. Puget, editors, Principles and Practice of Constraint Programming, CP98, volume 1520 of Lecture Notes in Computer Science, pp.  417–431. Springer, Heidelberg, Germany, 1998.
+[ bib ] + +
+ + +
+[2499] +
+
+David J. Sheskin. + Handbook of Parametric and Nonparametric Statistical Procedures. + Chapman & Hall/CRC, 2nd edition, 2000.
+[ bib ] + +
+ + +
+[2500] +
+
+David J. Sheskin. + Handbook of Parametric and Nonparametric Statistical Procedures. + Chapman & Hall/CRC, 5th edition, 2011.
+[ bib ] + +
+ + +
+[2501] +
+
+Yuhui Shi and Russell C. Eberhart. + Parameter selection in particle swarm optimization. + In V. W. Porto, N. Saravanan, D. Waagen, and A. E. Eiben, editors, Evolutionary Programming VII, volume 1447 of Lecture Notes in Computer Science, pp.  591–600, Heidelberg, Germany, 1998. Springer.
+[ bib | +DOI ] + +
+ + +
+[2502] +
+
+B. Shipley. + Cause and Correlation in Biology: a User's Guide to Path Analysis, Structural Equations and Causal Inference. + Cambridge University Press, 1st edition, 2000.
+[ bib ] + +
+ + +
+[2503] +
+
+A. Shmygelska, R. Aguirre-Hernández, and Holger H. Hoos. + An Ant Colony Optimization Algorithm for the 2D HP Protein Folding Problem. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  40–52. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2504] +
+
+James N. Siddall. + Optimal Engineering Design: Principles and Applications. + Marcel Dekker Inc., New York, NY, 1982.
+[ bib ] + +
+ + +
+[2505] +
+
+Sydney Siegel and N. John Castellan, Jr. + Non Parametric Statistics for the Behavioral Sciences. + McGraw Hill, New York, NY, 2nd edition, 1988.
+[ bib ] + +
+ + +
+[2506] +
+
+Moisés Silva-Muñoz, Gonzalo Calderon, Alberto Franzin, and Hughes Bersini. + Determining a consistent experimental setup for benchmarking and optimizing databases. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2021, pp.  1614–1621. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2507] +
+
+C. A. Silva, T. A. Runkler, J. M. Sousa, and R. Palm. + Ant Colonies as Logistic Processes Optimizers. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  76–87. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2508] +
+
+Angus R. Simpson, D. C. Sutton, D. S. Keane, and S. J. Sherriff. + Optimal control of pumping at a water filtration plant using genetic algorithms. + In D. A. Savic and G. A. Walters, editors, Water Industry Systems: Modelling and Optimization Applications, volume 2. Research Studies Press Ltd., Baldock, United Kingdom, 1999.
+[ bib ] + +
+ + +
+[2509] +
+
+Roman Slowiński. + Inducing preference models from pairwise comparisons: implications for preference-guided EMO. + Evolutionary Multi-Criterion Optimization, EMO 2011, 2011. + Keynote talk.
+[ bib ] + +
+ + +
+[2510] +
+
+Selmar K. Smit and Agoston E. Eiben. + Comparing Parameter Tuning Methods for Evolutionary Algorithms. + In Proceedings of the 2009 Congress on Evolutionary Computation (CEC 2009), pp.  399–406, Piscataway, NJ, 2009. IEEE Press.
+[ bib ] + +
+ + +
+[2511] +
+
+Selmar K. Smit and Agoston E. Eiben. + Beating the 'world champion' evolutionary algorithm via REVAC tuning. + In H. Ishibuchi et al., editors, Proceedings of the 2010 Congress on Evolutionary Computation (CEC 2010), pp.  1–8, Piscataway, NJ, 2010. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2512] +
+
+Selmar K. Smit and Agoston E. Eiben. + Parameter Tuning of Evolutionary Algorithms: Generalist vs. Specialist. + In C. D. Chio, S. Cagnoni, C. Cotta, M. Ebner, A. Ekárt, A. I. Esparcia-Alcázar, C. K. Goh, J.-J. Merelo, F. Neri, M. Preuss, J. Togelius, and G. N. Yannakakis, editors, Applications of Evolutionary Computation, volume 6024 of Lecture Notes in Computer Science, pp.  542–551. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2513] +
+
+Selmar K. Smit and Agoston E. Eiben. + Multi-Problem Parameter Tuning using BONESA. + In J.-K. Hao, P. Legrand, P. Collet, N. Monmarché, E. Lutton, and M. Schoenauer, editors, Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011, volume 7401 of Lecture Notes in Computer Science, pp.  222–233. Springer, Heidelberg, Germany, 2012.
+[ bib ] +
+For some reason, this was not actually published in the LNCS + Proceedings of EA +
+ +
+ + +
+[2514] +
+
+Selmar K. Smit, Agoston E. Eiben, and Z. Szlávik. + An MOEA-based Method to Tune EA Parameters on Multiple Objective Functions. + In J. Filipe and J. Kacprzyk, editors, Proceedings of the International Joint Conference on Computational Intelligence (IJCCI-2010), pp.  261–268. SciTePress, 2010.
+[ bib ] + +
+ + +
+[2515] +
+
+Tobiah E. Smith and Dorothy E. Setliff. + Knowledge-based constraint-driven software synthesis. + In Proceedings of the Seventh Knowledge-Based Software Engineering Conference, pp.  18–27. IEEE, 1992.
+[ bib | +DOI ] + +
+ + +
+[2516] +
+
+Jim Smith, Christopher Stone, and Martin Serpell. + Exploiting Diverse Distance Metrics for Surrogate-Based Optimisation of Ordering Problems. + In T. Friedrich, F. Neumann, and A. M. Sutton, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2016, pp.  701–708. ACM Press, New York, NY, 2016.
+[ bib | +DOI ] + +
+ + +
+[2517] +
+
+Kate Smith-Miles, Jano I. van Hemert, and Xin Yu Lim. + Understanding TSP difficulty by Learning from evolved instances. + In C. Blum and R. Battiti, editors, Learning and Intelligent Optimization, 4th International Conference, LION 4, volume 6073 of Lecture Notes in Computer Science, pp.  266–280. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2518] +
+
+Kate Smith-Miles. + Towards insightful algorithm selection for optimisation using meta-learning concepts. + In D. Liu et al., editors, Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008), Hong Kong, China, June 1-6, 2008, pp.  4118–4124. IEEE Press, 2008.
+[ bib ] + +
+ + +
+[2519] +
+
+George W. Snedecor and William G. Cochran. + Statistical Methods. + Iowa State University Press, Ames, IA, USA, 6th edition, 1967.
+[ bib ] + +
+ + +
+[2520] +
+
+Jasper Snoek, Hugo Larochelle, and Ryan P. Adams. + Practical Bayesian Optimization of Machine Learning Algorithms. + In P. L. Bartlett, F. C. N. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems (NIPS 25), pp.  2960–2968. Curran Associates, Red Hook, NY, 2012.
+[ bib ] + +
+ + +
+[2521] +
+
+Jasper Snoek, Kevin Swersky, Richard Zemel, and Ryan P. Adams. + Input Warping for Bayesian Optimization of Non-Stationary Functions. + In E. P. Xing and T. Jebara, editors, Proceedings of the 31st International Conference on Machine Learning, ICML 2014, volume 32, pp.  1674–1682. PMLR, 2014.
+[ bib | +http ] + +
+ + +
+[2522] +
+
+Krzysztof Socha, Joshua D. Knowles, and M. Sampels. + A Max-Min Ant System for the University Course Timetabling Problem. + In M. Dorigo et al., editors, Ant Algorithms, Third International Workshop, ANTS 2002, volume 2463 of Lecture Notes in Computer Science, pp.  1–13. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2523] +
+
+Krzysztof Socha, M. Sampels, and M. Manfrin. + Ant algorithms for the university course timetabling problem with regard to the state-of-the-art. + In S. Cagnoni et al., editors, Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2003, volume 2611 of Lecture Notes in Computer Science, pp.  334–345. Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2524] +
+
+Krzysztof Socha. + ACO for Continuous and Mixed-Variable Optimization. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science, pp.  25–36. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2525] +
+
+Christine Solnon. + Ant Colony Optimization and Constraint Programming. + Wiley, 2010.
+[ bib | +DOI ] + +
+ + +
+[2526] +
+
+Kenneth Sörensen, Marc Sevaux, and Fred Glover. + A history of metaheuristics. + In R. Martí, P. M. Pardalos, and M. G. C. Resende, editors, Handbook of Heuristics, pp.  1–27. Springer International Publishing, 2018.
+[ bib ] + +
+ + +
+[2527] +
+
+Aldo Sotelo, Julio Basulado, Pedro Doldán, and Benjamín Barán. + Algoritmos Evolutivos Multiobjetivo Combinados para la Optimización de la Programación de Bombeo en Sistemas de Suministro de Agua. + In Congreso Internacional de Tecnologías y Aplicaciones Informáticas, JIT-CITA 2001, Asunción, Paraguay, 2001. + (In Spanish).
+[ bib ] + +
+ + +
+[2528] +
+
+Aldo Sotelo, C. von Lücken, and Benjamín Barán. + Multiobjective Evolutionary Algorithms in Pump Scheduling Optimisation. + In B. H. V. Topping and Z. Bittnar, editors, Proceedings of the Third International Conference on Engineering Computational Technology. Civil-Comp Press, Stirling, Scotland, 2002.
+[ bib ] +
+Operation of pumping stations represents high costs + to water supply companies. Therefore, reducing such + costs through an optimal pump scheduling becomes an + important issue. This work presents the use of + Multiobjective Evolutionary Algorithms (MOEAs) to + solve an optimal pump-scheduling problem. For the + first time, six different approaches were + implemented and compared. These algorithms aim to + minimise four objectives: electric energy cost, + pumps' maintenance cost, maximum power peak, and + level variation in the reservoir. In order to + consider hydraulic and technical constrains, a + heuristic constrain algorithm was developed and + combined with each MOEA utilised. Evaluation of + experimental results of a set of metrics shows that + the Strength Pareto Evolutionary Algorithm (SPEA) + achieves the best performance for this + problem. Moreover, SPEA's set of solutions provide + pumping station operation engineers with a wide + range of optimal pump schedules to chose from. +
+ +
+ + +
+[2529] +
+
+Marcelo De Souza, Marcus Ritt, and Manuel López-Ibáñez. + CAPOPT: Capping Methods for the Automatic Configuration of Optimization Algorithms. + https://github.com/souzamarcelo/capopt, 2020.
+[ bib ] + +
+ + +
+[2530] +
+
+Marcelo De Souza, Marcus Ritt, and Manuel López-Ibáñez. + Capping Methods for the Automatic Configuration of Optimization Algorithms – Supplementary Material. + https://github.com/souzamarcelo/supp-cor-capopt, 2021.
+[ bib ] + +
+ + +
+[2531] +
+
+Apache Software Foundation. + Spark, 2012.
+[ bib | +http ] + +
+ + +
+[2532] +
+
+Suvrit Sra, Sebastian Nowozin, and Stephen J. Wright. + Optimization for machine learning. + MIT Press, Cambridge, MA, 2012.
+[ bib ] + +
+ + +
+[2533] +
+
+P. F. Stadler. + Toward a theory of landscapes. + In R. López-Peña, R. Capovilla, R. García-Pelayo, H. Waelbroeck, and F. Zertruche, editors, Complex Systems and Binary Networks, pp.  77–163. Springer, 1995.
+[ bib ] + +
+ + +
+[2534] +
+
+Martin Kenneth Starr. + Product design and decision theory. + Prentice-Hall Series in Engineering Design, Fundamentals of Engineering Design. Prentice-Hall, Englewood, Cliffs, NJ, 1963.
+[ bib ] + +
+ + +
+[2535] +
+
+Fernando Stefanello, Vaneet Aggarwal, Luciana Salete Buriol, José Fernando Gonçalves, and Mauricio G. C. Resende. + A Biased Random-key Genetic Algorithm for Placement of Virtual Machines Across Geo-Separated Data Centers. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  919–926. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[2536] +
+
+R. E. Steuer and Lorraine Gardiner. + On the Computational Testing of Procedures for Interactive Multiple Objective Linear Programming. + In G. Fandel and H. Gehring, editors, Operations Research, pp.  121–131. Springer, Berlin/Heidelberg, 1991.
+[ bib | +DOI ] +
+Proposed difference between ad hoc and non-ad hoc interactive + multi-objective optimization methods +
+ +
+ + +
+[2537] +
+
+Christian Steinruecken, Emma Smith, David Janz, James Lloyd, and Zoubin Ghahramani. + The Automatic Statistician. + In F. Hutter, L. Kotthoff, and J. Vanschoren, editors, Automated Machine Learning, pp.  161–173. Springer, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2538] +
+
+R. E. Steuer. + Multiple Criteria Optimization: Theory, Computation and Application. + Wiley Series in Probability and Mathematical Statistics. John Wiley & Sons, New York, NY, 1986.
+[ bib ] +
+Keywords: Maximally dispersed weights +
+ +
+ + +
+[2539] +
+
+Daniel H. Stolfi and Enrique Alba. + An Evolutionary Algorithm to Generate Real Urban Traffic Flows. + In J. M. Puerta, J. A. Gámez, B. Dorronsoro, E. Barrenechea, A. Troncoso, B. Baruque, and M. Galar, editors, Advances in Artificial Intelligence, CAEPIA 2015, volume 9422 of Lecture Notes in Computer Science, pp.  332–343. Springer, Heidelberg, Germany, 2015.
+[ bib | +DOI ] +
+In this article we present a strategy based on an evolution- + ary algorithm to calculate the real vehicle flows in cities + according to data from sensors placed in the streets. We have + worked with a map imported from OpenStreetMap into the SUMO + traffic simulator so that the resulting scenarios can be used + to perform different optimizations with the confidence of + being able to work with a traffic distribution close to + reality. We have compared the results of our algorithm to + other competitors and achieved results that replicate the + real traffic distribution with a precision higher than + 90%. +
+
+Keywords: Evolutionary algorithm,SUMO,Smart city,Smart mobility,Traffic + simulation +
+ +
+ + +
+[2540] +
+
+Thomas Stützle. + Applying Iterated Local Search to the Permutation Flow Shop Problem. + Technical Report AIDA–98–04, FG Intellektik, FB Informatik, TU Darmstadt, Germany, August 1998.
+[ bib ] + +
+ + +
+[2541] +
+
+Thomas Stützle. + ACOTSP: A Software Package of Various Ant Colony Optimization Algorithms Applied to the Symmetric Traveling Salesman Problem, 2002.
+[ bib | +http ] +
+http://www.aco-metaheuristic.org/aco-code +
+ +
+ + +
+[2542] +
+
+Thomas Stützle. + Some Thoughts on Engineering Stochastic Local Search Algorithms. + In A. Viana et al., editors, Proceedings of the EU/MEeting 2009: Debating the future: new areas of application and innovative approaches, pp.  47–52, 2009.
+[ bib ] + +
+ + +
+[2543] +
+
+Thomas Stützle. + Max-Min Ant System for the Quadratic Assignment Problem. + Technical Report AIDA–97–4, FG Intellektik, FB Informatik, TU Darmstadt, Germany, July 1997.
+[ bib ] + +
+ + +
+[2544] +
+
+Thomas Stützle. + An Ant Approach to the Flow Shop Problem. + In Proceedings of the 6th European Congress on Intelligent Techniques & Soft Computing (EUFIT'98), volume 3, pp.  1560–1564. Verlag Mainz, Aachen, Germany, 1998.
+[ bib ] + +
+ + +
+[2545] +
+
+Thomas Stützle and Marco Dorigo. + ACO Algorithms for the Quadratic Assignment Problem. + In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in Optimization, pp.  33–50. McGraw Hill, London, UK, 1999.
+[ bib ] + +
+ + +
+[2546] +
+
+Thomas Stützle and Susana Fernandes. + New Benchmark Instances for the QAP and the Experimental Analysis of Algorithms. + In J. Gottlieb and G. R. Raidl, editors, Proceedings of EvoCOP 2004 – 4th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3004 of Lecture Notes in Computer Science, pp.  199–209. Springer, Heidelberg, Germany, 2004.
+[ bib | +DOI ] +
+The quadratic assignment problem arises in a variety of + practical settings. It is known to be among the hardest + combinatorial problems for exact algorithms. Therefore, a + large number of heuristic approaches have been proposed for + its solution. In this article we introduce a new, large set + of QAP instances that is intended to allow the systematic + study of the performance of metaheuristics in dependence of + QAP instance characteristics. Additionally, we give + computational results with several high performing algorithms + known from literature and give exemplary results on the + influence of instance characteristics on the performance of + these algorithms. +
+ +
+ + +
+[2547] +
+
+Thomas Stützle and Holger H. Hoos. + Analysing the Run-time Behaviour of Iterated Local Search for the Travelling Salesman Problem. + In P. Hansen and C. Ribeiro, editors, Essays and Surveys on Metaheuristics, Operations Research/Computer Science Interfaces Series, pp.  589–611. Kluwer Academic Publishers, Boston, MA, 2001.
+[ bib ] + +
+ + +
+[2548] +
+
+Thomas Stützle and Holger H. Hoos. + Improving the Ant System: A Detailed Report on the Max-Min Ant System. + Technical Report AIDA–96–12, FG Intellektik, FB Informatik, TU Darmstadt, Germany, August 1996.
+[ bib ] + +
+ + +
+[2549] +
+
+Thomas Stützle and Holger H. Hoos. + The Max-Min Ant System and Local Search for the Traveling Salesman Problem. + In T. Bäck, Z. Michalewicz, and X. Yao, editors, Proceedings of the 1997 IEEE International Conference on Evolutionary Computation (ICEC'97), pp.  309–314. IEEE Press, Piscataway, NJ, 1997.
+[ bib ] + +
+ + +
+[2550] +
+
+Thomas Stützle and Holger H. Hoos. + Max-Min Ant System and Local Search for Combinatorial Optimization Problems. + In S. Voß, S. Martello, I. H. Osman, and C. Roucairol, editors, Meta-Heuristics: Advances and Trends in Local Search Paradigms for Optimization, pp.  137–154. Kluwer Academic Publishers, Dordrecht, The Netherlands, 1999.
+[ bib ] + +
+ + +
+[2551] +
+
+Thomas Stützle and Manuel López-Ibáñez. + Automatic (Offline) Configuration of Algorithms. + In J. L. Jiménez Laredo, S. Silva, and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2015, pp.  681–702. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] + +
+ + +
+[2552] +
+
+Thomas Stützle and Manuel López-Ibáñez. + Automated Offline Design of Algorithms. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2017, pp.  1038–1065. ACM Press, New York, NY, 2017.
+[ bib | +DOI ] + +
+ + +
+[2553] +
+
+Thomas Stützle and Manuel López-Ibáñez. + Automated Design of Metaheuristic Algorithms. + In M. Gendreau and J.-Y. Potvin, editors, Handbook of Metaheuristics, volume 272 of International Series in Operations Research & Management Science, pp.  541–579. Springer, 2019.
+[ bib | +DOI ] +
+Keywords: automatic design, automatic configuration +
+ +
+ + +
+[2554] +
+
+Thomas Stützle, Manuel López-Ibáñez, and Marco Dorigo. + A Concise Overview of Applications of Ant Colony Optimization. + In J. J. Cochran, editor, Wiley Encyclopedia of Operations Research and Management Science, volume 2, pp.  896–911. John Wiley & Sons, 2011.
+[ bib | +DOI ] + +
+ + +
+[2555] +
+
+Thomas Stützle, Manuel López-Ibáñez, Paola Pellegrini, Michael Maur, Marco A. Montes de Oca, Mauro Birattari, and Marco Dorigo. + Parameter Adaptation in Ant Colony Optimization. + In Y. Hamadi, E. Monfroy, and F. Saubion, editors, Autonomous Search, pp.  191–215. Springer, Berlin, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2556] +
+
+Thomas Stützle and Rubén Ruiz. + Iterated Greedy. + In R. Martí, P. M. Pardalos, and M. G. C. Resende, editors, Handbook of Heuristics, pp.  1–31. Springer International Publishing, 2018.
+[ bib | +DOI ] + +
+ + +
+[2557] +
+
+Thomas Stützle and Rubén Ruiz. + Iterated Local Search. + In R. Martí, P. M. Pardalos, and M. G. C. Resende, editors, Handbook of Heuristics, pp.  1–27. Springer International Publishing, 2018.
+[ bib | +DOI ] + +
+ + +
+[2558] +
+
+Thomas Stützle. + Local Search Algorithms for Combinatorial Problems — Analysis, Improvements, and New Applications. + PhD thesis, FB Informatik, TU Darmstadt, Germany, 1998.
+[ bib ] + +
+ + +
+[2559] +
+
+James Styles and Holger H. Hoos. + Ordered racing protocols for automatically configuring algorithms for scaling performance. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2013, pp.  551–558. ACM Press, New York, NY, 2013.
+[ bib | +DOI ] + +
+ + +
+[2560] +
+
+James Styles, Holger H. Hoos, and Martin Müller. + Automatically Configuring Algorithms for Scaling Performance. + In Y. Hamadi and M. Schoenauer, editors, Learning and Intelligent Optimization, 6th International Conference, LION 6, volume 7219 of Lecture Notes in Computer Science, pp.  205–219. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2561] +
+
+Ponnuthurai N. Suganthan, Nikolaus Hansen, J. J. Liang, Kalyanmoy Deb, Y. P. Chen, Anne Auger, and S. Tiwari. + Problem definitions and evaluation criteria for the CEC 2005 special session on real-parameter optimization. + Technical report, Nanyang Technological University, Singapore, 2005.
+[ bib ] +
+Also known as KanGAL Report Number 2005005 (Kanpur Genetic Algorithms + Laboratory, IIT Kanpur) +
+
+Keywords: CEC'05 benchmark +
+ +
+ + +
+[2562] +
+
+Yanan Sui, Alkis Gotovos, Joel W. Burdick, and Andreas Krause. + Safe Exploration for Optimization with Gaussian Processes. + In F. Bach and D. Blei, editors, Proceedings of the 32nd International Conference on Machine Learning, ICML 2015, volume 37, pp.  997–1005. PMLR, 2015.
+[ bib | +epub ] +
+We consider sequential decision problems under uncertainty, + where we seek to optimize an unknown function from noisy + samples. This requires balancing exploration (learning about + the objective) and exploitation (localizing the maximum), a + problem well-studied in the multi-armed bandit literature. In + many applications, however, we require that the sampled + function values exceed some prespecified "safety" threshold, + a requirement that existing algorithms fail to meet. Examples + include medical applications where patient comfort must be + guaranteed, recommender systems aiming to avoid user + dissatisfaction, and robotic control, where one seeks to + avoid controls causing physical harm to the platform. We + tackle this novel, yet rich, set of problems under the + assumption that the unknown function satisfies regularity + conditions expressed via a Gaussian process prior. We develop + an efficient algorithm called SafeOpt, and theoretically + guarantee its convergence to a natural notion of optimum + reachable under safety constraints. We evaluate SafeOpt on + synthetic data, as well as two real applications: movie + recommendation, and therapeutic spinal cord stimulation. +
+
+Keywords: Safe Optimization, SafeOpt +
+ +
+ + +
+[2563] +
+
+Yanan Sui, Vincent Zhuang, Joel W. Burdick, and Yisong Yue. + Stagewise Safe Bayesian Optimization with Gaussian Processes. + In J. G. Dy and A. Krause, editors, Proceedings of the 35th International Conference on Machine Learning, ICML 2018, volume 80 of Proceedings of Machine Learning Research, pp.  4788–4796. PMLR, 2018.
+[ bib | +epub ] +
+Keywords: StageOpt +
+ +
+ + +
+[2564] +
+
+Zhaoxu Sun and Min Han. + Multi-criteria Decision Making Based on PROMETHEE Method. + In Proceedings of the 2010 International Conference on Computing, Control and Industrial Engineering, pp.  416–418, Los Alamitos, CA, 2010. IEEE Computer Society Press.
+[ bib ] + +
+ + +
+[2565] +
+
+Richard S. Sutton and Andrew G. Barto. + Reinforcement Learning: An Introduction. + MIT Press, Cambridge, MA, 1998.
+[ bib ] + +
+ + +
+[2566] +
+
+Richard S. Sutton and Andrew G. Barto. + Reinforcement Learning: An Introduction. + MIT Press, Cambridge, MA, 2nd edition, 2018.
+[ bib ] + +
+ + +
+[2567] +
+
+D. C. Sutton, D. S. Keane, and S. J. Sherriff. + Optimizing the Real Time Operation of a Pumping Station at a Water Filtration Plant using Genetic Algorithms. + Honors thesis, Department of Civil and Environmental Engineering, The University of Adelaide, 1998.
+[ bib ] + +
+ + +
+[2568] +
+
+Jerry Swan, Ender Özcan, and Graham Kendall. + Hyperion: A Recursive Hyper-heuristic Framework. + In C. A. Coello Coello, editor, Learning and Intelligent Optimization, 5th International Conference, LION 5, volume 6683 of Lecture Notes in Computer Science, pp.  616–630. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2569] +
+
+Jerry Swan et al. + A Research Agenda for Metaheuristic Standardization. + In E.-G. Talbi, editor, Proceedings of MIC 2015, the 11th Metaheuristics International Conference, 2015.
+[ bib ] + +
+ + +
+[2570] +
+
+Gilbert Syswerda. + Uniform Crossover in Genetic Algorithms. + In J. D. Schaffer, editor, Proceedings of the Third International Conference on Genetic Algorithms (ICGA'89), pp.  2–9. Morgan Kaufmann Publishers, San Mateo, CA, 1989.
+[ bib ] +
+Keywords: uniform crossover +
+ +
+ + +
+[2571] +
+
+Kiyoharu Tagawa, Hidehito Shimizu, and Hiroyuki Nakamura. + Indicator-based Differential Evolution Using Exclusive Hypervolume Approximation and Parallelization for Multi-core Processors. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  657–664. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2572] +
+
+Yaniv Taigman, Ming Yang, Marc'Aurelio Ranzato, and Lior Wolf. + Deepface: Closing the gap to human-level performance in face verification. + In Proceedings of the IEEE conference on computer vision and pattern recognition, pp.  1701–1708, 2014.
+[ bib ] + +
+ + +
+[2573] +
+
+Ryoji Tanabe and Akira Oyama. + Benchmarking MOEAs for multi-and many-objective optimization using an unbounded external archive. + In P. A. N. Bosman, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2017, pp.  633–640. ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[2574] +
+
+M. Fatih Tasgetiren, Ozge Buyukdagli, Quan-Ke Pan, and Ponnuthurai N. Suganthan. + A general variable neighborhood search algorithm for the no-idle permutation flowshop scheduling problem. + In B. K. Panigrahi, P. N. Suganthan, S. Das, and S. S. Dash, editors, Swarm, Evolutionary, and Memetic Computing, volume 8298 of Theoretical Computer Science and General Issues, pp.  24–34. Springer International Publishing, 2013.
+[ bib ] + +
+ + +
+[2575] +
+
+Jorge Tavares and Francisco B. Pereira. + Automatic Design of Ant Algorithms with Grammatical Evolution. + In A. Moraglio, S. Silva, K. Krawiec, P. Machado, and C. Cotta, editors, Proceedings of the 15th European Conference on Genetic Programming, EuroGP 2012, volume 7244 of Lecture Notes in Computer Science, pp.  206–217. Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2576] +
+
+Cristina Teixeira, José Covas, Thomas Stützle, and António Gaspar-Cunha. + Application of Pareto Local Search and Multi-Objective Ant Colony Algorithms to the Optimization of Co-Rotating Twin Screw Extruders. + In A. Viana et al., editors, Proceedings of the EU/MEeting 2009: Debating the future: new areas of application and innovative approaches, pp.  115–120, 2009.
+[ bib ] + +
+ + +
+[2577] +
+
+Google. + TensorFlow. + https://www.tensorflow.org, 2017.
+[ bib ] + +
+ + +
+[2578] +
+
+K. T. K. Teo, W. Y. Kow, and Y. K. Chin. + Optimization of traffic flow within an urban traffic light intersection with genetic algorithm. + In Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010, pp.  172–177. IEEE, IEEE Press, 2010.
+[ bib ] +
+Keywords: Genetic algorithm,T-junction,Traffic control system,Traffic + flows +
+ +
+ + +
+[2579] +
+
+Hugo Terashima-Marín, Peter Ross, and Manuel Valenzuela-Rendón. + Evolution of Constraint Satisfaction Strategies in Examination Timetabling. + In W. Banzhaf, J. M. Daida, A. E. Eiben, M. H. Garzon, V. Honavar, M. J. Jakiela, and R. E. Smith, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 1999, pp.  635–642. Morgan Kaufmann Publishers, San Francisco, CA, 1999.
+[ bib ] + +
+ + +
+[2580] +
+
+Dirk Thierens. + Adaptive strategies for operator allocation. + In F. Lobo, C. F. Lima, and Z. Michalewicz, editors, Parameter Setting in Evolutionary Algorithms, pp.  77–90. Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[2581] +
+
+Dirk Thierens. + Adaptive operator selection for iterated local search. + In T. Stützle, M. Birattari, and H. H. Hoos, editors, Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2009, volume 5752 of Lecture Notes in Computer Science, pp.  140–144. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2582] +
+
+Dirk Thierens. + Population-based Iterated Local Search: Restricting the Neighborhood Search by Crossover. + In K. Deb et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2004, Part II, volume 3103 of Lecture Notes in Computer Science, pp.  234–245. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2583] +
+
+Dirk Thierens. + An Adaptive Pursuit Strategy for Allocating Operator Probabilities. + In H.-G. Beyer and U.-M. O'Reilly, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2005, pp.  1539–1546. ACM Press, New York, NY, 2005.
+[ bib ] + +
+ + +
+[2584] +
+
+Chris Thornton, Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Auto-WEKA: Combined Selection and Hyperparameter Optimization of Classification Algorithms. + In I. S. Dhillon, Y. Koren, R. Ghani, T. E. Senator, P. Bradley, R. Parekh, J. He, R. L. Grossman, and R. Uthurusamy, editors, The 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2013, pp.  847–855. ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[2585] +
+
+Sebastian Thrun and Lorien Pratt. + Learning to learn. + Springer, 1998.
+[ bib ] + +
+ + +
+[2586] +
+
+Renato Tinós, Darrell Whitley, and Gabriela Ochoa. + Generalized Asymmetric Partition Crossover (GAPX) for the Asymmetric TSP. + In C. Igel and D. V. Arnold, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2014, pp.  501–508. ACM Press, New York, NY, 2014.
+[ bib ] + +
+ + +
+[2587] +
+
+Michal K Tomczyk and Milosz Kadziński. + Robust Indicator-Based Algorithm for Interactive Evolutionary Multiple Objective Optimization. + In M. López-Ibáñez, A. Auger, and T. Stützle, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, pp.  629–637. ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] +
+We propose a novel robust indicator-based algorithm, called + IEMO/I, for interactive evolutionary multiple objective + optimization. During the optimization run, IEMO/I selects at + regular intervals a pair of solutions from the current + population to be compared by the Decision Maker. The + successively provided holistic judgements are employed to + divide the population into fronts of potential + optimality. These fronts are, in turn, used to bias the + evolutionary search toward a subset of Pareto-optimal + solutions being most relevant to the Decision Maker. To + ensure a fine approximation of such a subset, IEMO/I employs + a hypervolume metric within a steady-state indicator-based + evolutionary framework. The extensive experimental evaluation + involving a number of benchmark problems confirms that IEMO/I + is able to construct solutions being highly preferred by the + Decision Maker after a reasonable number of interactions. We + also compare IEMO/I with some selected state-of-the-art + interactive evolutionary hybrids incorporating preference + information in form of pairwise comparisons, proving its + competitiveness. +
+
+Keywords: preference learning, indicator-based algorithms, interactive + algorithms, multiple objective optimization, pairwise + comparisons, evolutionary algorithms +
+ +
+ + +
+[2588] +
+
+Paolo Toth and Daniele Vigo. + The vehicle routing problem. + Society for Industrial and Applied Mathematics, Philadelphia, PA, USA, 2002.
+[ bib ] + +
+ + +
+[2589] +
+
+F. Toyama, K. Shoji, H. Mori, and J. Miyamichi. + An Iterated Greedy Algorithm for the Binary Quadratic Programming Problem. + In Joint 6th International Conference on Soft Computing and Intelligent Systems (SCIS) and 13th International Symposium on Advanced Intelligent Systems (ISIS), 2012, pp.  2183–2188. IEEE Press, 2012.
+[ bib ] + +
+ + +
+[2590] +
+
+Risto Trajanov, Ana Nikolikj, Gjorgjina Cenikj, Fabien Teytaud, Mathurin Videau, Olivier Teytaud, Tome Eftimov, Manuel López-Ibáñez, and Carola Doerr. + Improving Nevergrad's Algorithm Selection Wizard NGOpt Through Automated Algorithm Configuration. + In G. Rudolph, A. V. Kononova, H. E. Aguirre, P. Kerschke, G. Ochoa, and T. Tušar, editors, Parallel Problem Solving from Nature – PPSN XVII, volume 13398 of Lecture Notes in Computer Science, pp.  18–31. Springer, Cham, Switzerland, 2022.
+[ bib | +DOI ] +
+Algorithm selection wizards are effective and versatile tools + that automatically select an optimization algorithm given + high-level information about the problem and available + computational resources, such as number and type of decision + variables, maximal number of evaluations, possibility to + parallelize evaluations, etc. State-of-the-art algorithm + selection wizards are complex and difficult to improve. We + propose in this work the use of automated configuration + methods for improving their performance by finding better + configurations of the algorithms that compose them. In + particular, we use elitist iterated racing (irace) to find + CMA configurations for specific artificial benchmarks that + replace the hand-crafted CMA configurations currently used in + the NGOpt wizard provided by the Nevergrad platform. We + discuss in detail the setup of irace for the purpose of + generating configurations that work well over the diverse set + of problem instances within each benchmark. Our approach + improves the performance of the NGOpt wizard, even on + benchmark suites that were not part of the tuning by irace. +
+ +
+ + +
+[2591] +
+
+Christoph Treude and Markus Wagner. + Predicting Good Configurations for GitHub and Stack Overflow Topic Models. + In Proceedings of the 16th International Conference on Mining Software Repositories, MSR '19, pp.  84–95, Piscataway, NJ, 2019. IEEE Press.
+[ bib | +DOI ] +
+Keywords: algorithm portfolio, corpus features, topic modelling +
+ +
+ + +
+[2592] +
+
+Michael A. Trick. + Graph Coloring Instances. + https://mat.gsia.cmu.edu/COLOR/instances.html, 2018.
+[ bib ] + +
+ + +
+[2593] +
+
+S. Tsutsui. + An Enhanced Aggregation Pheromone System for Real-Parameter Optimization in the ACO Metaphor. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  60–71. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2594] +
+
+S. Tsutsui. + cAS: Ant Colony Optimization with Cunning Ants. + In T. P. Runarsson, H.-G. Beyer, E. K. Burke, J.-J. Merelo, D. Whitley, and X. Yao, editors, Parallel Problem Solving from Nature – PPSN IX, volume 4193 of Lecture Notes in Computer Science, pp.  162–171. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2595] +
+
+Edward R. Tufte. + The Visual Display of Quantitative Information. + Graphics Press, Cheshire, CT, 2nd edition, 2001.
+[ bib ] +
+The classic book on statistical graphics, charts, + tables. Theory and practice in the design of data graphics, + 250 illustrations of the best (and a few of the worst) + statistical graphics, with detailed analysis of how to + display data for precise, effective, quick analysis. Design + of the high-resolution displays, small multiples. Editing and + improving graphics. The data-ink ratio. Time-series, + relational graphics, data maps, multivariate + designs. Detection of graphical deception: design variation + vs. data variation. Sources of deception. Aesthetics and data + graphical displays. This new edition provides excellent color + reproductions of the many graphics of William Playfair, adds + color to other images, and includes all the changes and + corrections accumulated during 17 printings of the first + edition. +
+
+Keywords: data visualization, information graphics, cognitive science +
+ +
+ + +
+[2596] +
+
+Matteo Turchetta, Felix Berkenkamp, and Andreas Krause. + Safe Exploration in Finite Markov Decision Processes with Gaussian Processes. + In D. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, and R. Garnett, editors, Advances in Neural Information Processing Systems (NIPS 29), pp.  4312–4320, 2016.
+[ bib | +DOI | +epub ] +
+Keywords: SafeMDP +
+ +
+ + +
+[2597] +
+
+Matteo Turchetta, Felix Berkenkamp, and Andreas Krause. + Safe Exploration for Interactive Machine Learning. + In H. M. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. B. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems (NeurIPS 32), pp.  2887–2897, 2019.
+[ bib | +epub ] +
+Keywords: Reinforcement Learning; Markov Decision Process; SafeML +
+ +
+ + +
+[2598] +
+
+The Turing Way Community, Becky Arnold, Louise Bowler, Sarah Gibson, Patricia Herterich, Rosie Higman, Anna Krystalli, Alexander Morley, Martin O'Reilly, and Kirstie Whitaker. + The Turing Way: A Handbook for Reproducible Data Science. + Zenodo, March 2019.
+[ bib | +DOI ] +
+Available from https://the-turing-way.netlify.app. This work was supported by The UKRI Strategic Priorities Fund + under the EPSRC Grant EP/T001569/1, particularly the "Tools, + Practices and Systems" theme within that grant, and by The + Alan Turing Institute under the EPSRC grant EP/N510129/1. +
+ +
+ + +
+[2599] +
+
+Tea Tušar and Bogdan Filipič. + Differential Evolution versus Genetic Algorithms in Multiobjective Optimization. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  257–271. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2600] +
+
+Tea Tušar and Bogdan Filipič. + Visualizing 4D approximation sets of multiobjective optimizers with prosections. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  737–744. ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2601] +
+
+Tea Tušar. + Design of an Algorithm for Multiobjective Optimization with Differential Evolution. + M.sc. thesis, Faculty of Computer and Information Science, University of Ljubljana, 2007.
+[ bib ] + +
+ + +
+[2602] +
+
+N. L. J. Ulder, Emile H. L. Aarts, H.-J. Bandelt, Peter J. M. van Laarhoven, and Erwin Pesch. + Genetic Local Search Algorithms for the Travelling Salesman Problem. + In H.-P. Schwefel and R. Männer, editors, Parallel Problem Solving from Nature – PPSN I, pp.  109–116. Springer, Berlin/Heidelberg, 1991.
+[ bib | +DOI ] + +
+ + +
+[2603] +
+
+Tamara Ulrich, Johannes Bader, and Lothar Thiele. + Defining and Optimizing Indicator-Based Diversity Measures in Multiobjective Search. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  707–717. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] +
+Two archive; two populations; decision space diversity +
+ +
+ + +
+[2604] +
+
+Andrea Valsecchi, Jérémie Dubois-Lacoste, Thomas Stützle, Sergio Damas, José Santamaría, and Linda Marrakchi-Kacem. + Evolutionary Medical Image Registration using Automatic Parameter Tuning. + In Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), pp.  1326–1333, Piscataway, NJ, 2013. IEEE Press.
+[ bib ] + +
+ + +
+[2605] +
+
+Mauro Vallati, Chris Fawcett, Alfonso E. Gerevini, Holger H. Hoos, and Alessandro Saetti. + Generating Fast Domain-Optimized Planners by Automatically Configuring a Generic Parameterised Planner. + In E. Karpas, S. Jiménez Celorrio, and S. Kambhampati, editors, Proceedings of ICAPS-PAL11, 2011.
+[ bib ] + +
+ + +
+[2606] +
+
+Peter J. M. van Laarhoven and Emile H. L. Aarts. + Simulated Annealing: Theory and Applications, volume 37. + Springer, 1987.
+[ bib ] + +
+ + +
+[2607] +
+
+Jan N. van Rijn and Frank Hutter. + Hyperparameter Importance Across Datasets. + In Y. Guo and F. Farooq, editors, 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.  2367–2376. ACM Press, New York, NY, July 2018.
+[ bib | +DOI ] +
+With the advent of automated machine learning, automated + hyperparameter optimization methods are by now routinely used + in data mining. However, this progress is not yet matched by + equal progress on automatic analyses that yield information + beyond performance-optimizing hyperparameter settings. In + this work, we aim to answer the following two questions: + Given an algorithm, what are generally its most important + hyperparameters, and what are typically good values for + these? We present methodology and a framework to answer these + questions based on meta-learning across many datasets. We + apply this methodology using the experimental meta-data + available on OpenML to determine the most important + hyperparameters of support vector machines, random forests + and Adaboost, and to infer priors for all their + hyperparameters. The results, obtained fully automatically, + provide a quantitative basis to focus efforts in both manual + algorithm design and in automated hyperparameter + optimization. The conducted experiments confirm that the + hyperparameters selected by the proposed method are indeed + the most important ones and that the obtained priors also + lead to statistically significant improvements in + hyperparameter optimization. +
+
+Keywords: hyperparameter optimization, meta-learning, hyperparameter + importance +
+ +
+ + +
+[2608] +
+
+Elia Van Wolputte, Evgeniya Korneva, and Hendrik Blockeel. + MERCS: multi-directional ensembles of regression and classification trees. + In S. A. McIlraith and K. Q. Weinberger, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  4276–4283. AAAI Press, February 2018.
+[ bib ] + +
+ + +
+[2609] +
+
+Andrea Vedaldi and Brian Fulkerson. + VLFeat: An open and portable library of computer vision algorithms. + In Proceedings of the 18th ACM international conference on Multimedia, pp.  1469–1472. ACM, 2010.
+[ bib ] + +
+ + +
+[2610] +
+
+David A. Van Veldhuizen and Gary B. Lamont. + Evolutionary Computation and Convergence to a Pareto Front. + In J. R. Koza, editor, Late Breaking Papers at the Genetic Programming 1998 Conference, pp.  221–228, Stanford University, California, July 1998. Stanford University Bookstore.
+[ bib ] +
+Keywords: generational distance +
+ +
+ + +
+[2611] +
+
+David A. Van Veldhuizen. + Multiobjective evolutionary algorithms: Classifications, analyses, and new innovations. + PhD thesis, Department of Electrical and Computer Engineering, Graduate School of Engineering, Air Force Institute of Technology, Wright-Patterson AFB, Ohio, 1999.
+[ bib ] + +
+ + +
+[2612] +
+
+Diederick Vermetten, Fabio Caraffini, Bas van Stein, and Anna V. Kononova. + Using Structural Bias to Analyse the Behaviour of Modular CMA-ES. + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO Companion 2022, pp.  1674–1682. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] + +
+ + +
+[2613] +
+
+Sébastien Verel, Arnaud Liefooghe, and Clarisse Dhaenens. + Set-based Multiobjective Fitness Landscapes: A Preliminary Study. + In N. Krasnogor and P. L. Lanzi, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2011, pp.  769–776. ACM Press, New York, NY, 2011.
+[ bib | +DOI ] + +
+ + +
+[2614] +
+
+Diederick Vermetten, Hao Wang, Carola Doerr, and Thomas Bäck. + Integrated vs. Sequential Approaches for Selecting and Tuning CMA-ES Variants. + In C. A. Coello Coello, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020. ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[2615] +
+
+Diederick Vermetten, Hao Wang, Manuel López-Ibáñez, Carola Doerr, and Thomas Bäck. + Analyzing the Impact of Undersampling on the Benchmarking and Configuration of Evolutionary Algorithms. + In J. E. Fieldsend and M. Wagner, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, pp.  867–875. ACM Press, New York, NY, 2022.
+[ bib | +DOI ] +
+The stochastic nature of iterative optimization heuristics + leads to inherently noisy performance measurements. Since + these measurements are often gathered once and then used + repeatedly, the number of collected samples will have a + significant impact on the reliability of algorithm + comparisons. We show that care should be taken when making + decisions based on limited data. Particularly, we show that + the number of runs used in many benchmarking studies, e.g., + the default value of 15 suggested by the COCO environment, + can be insufficient to reliably rank algorithms on well-known + numerical optimization benchmarks.Additionally, methods for + automated algorithm configuration are sensitive to + insufficient sample sizes. This may result in the + configurator choosing a "lucky" but poor-performing + configuration despite exploring better ones. We show that + relying on mean performance values, as many configurators do, + can require a large number of runs to provide accurate + comparisons between the considered configurations. Common + statistical tests can greatly improve the situation in most + cases but not always. We show examples of performance losses + of more than 20%, even when using statistical races to + dynamically adjust the number of runs, as done by irace. Our + results underline the importance of appropriately considering + the statistical distribution of performance values. +
+
+Keywords: parameter tuning, evolution strategies, algorithm + configuration, performance measures +
+ +
+ + +
+[2616] +
+
+Mathurin Videau, Alessandro Leite, Olivier Teytaud, and Marc Schoenauer. + Multi-Objective Genetic Programming for Explainable Reinforcement Learning. + In E. Medvet, G. Pappa, and B. Xue, editors, Proceedings of the 25th European Conference on Genetic Programming, EuroGP 2022, Lecture Notes in Computer Science, pp.  256–281. Springer Nature, Cham, Switzerland, 2022.
+[ bib ] +
+Keywords: genetic algorithms, genetic programming: Poster +
+ +
+ + +
+[2617] +
+
+Carlos Vieira, Leslie Pérez Cáceres, and Leonardo C. T. Bezerra. + Evaluating Anytime Performance on NAS-Bench-101. + In Proceedings of the 2021 Congress on Evolutionary Computation (CEC 2021), pp.  1249–1256, Piscataway, NJ, 2021. IEEE Press.
+[ bib | +DOI ] + +
+ + +
+[2618] +
+
+Alessia Violin. + Mathematical Programming Approaches to Pricing Problems. + PhD thesis, Faculté de Sciences, Université Libre de Bruxelles and Dipartimento di Ingegneria e Architettura, Università degli studi di Trieste, 2014.
+[ bib ] +
+Supervised by Dr. Martine Labbé and Dr. Lorenzo Castelli +
+ +
+ + +
+[2619] +
+
+Thomas Voß, Nikolaus Hansen, and Christian Igel. + Improved Step Size Adaptation for the MO-CMA-ES. + In M. Pelikan and J. Branke, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2010, pp.  487–494. ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2620] +
+
+Christos Voudouris and Edward P. K. Tsang. + Guided Local Search. + In F. Glover and G. A. Kochenberger, editors, Handbook of Metaheuristics, pp.  185–218. Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[2621] +
+
+D. A. Savic and G. A. Walters, editors. + Water Industry Systems: Modelling and Optimization Applications, volume 2. + Research Studies Press Ltd., Baldock, United Kingdom, 1999.
+[ bib ] + +
+ + +
+[2622] +
+
+Akifumi Wachi, Yanan Sui, Yisong Yue, and Masahiro Ono. + Safe Exploration and Optimization of Constrained MDPs Using Gaussian Processes. + In S. A. McIlraith and K. Q. Weinberger, editors, Proceedings of the AAAI Conference on Artificial Intelligence, pp.  6548–6556. AAAI Press, February 2018.
+[ bib | +DOI ] +
+We present a reinforcement learning approach to explore and + optimize a safety-constrained Markov Decision + Process(MDP). In this setting, the agent must maximize + discounted cumulative reward while constraining the + probability of entering unsafe states, defined using a safety + function being within some tolerance. The safety values of + all states are not known a priori, and we probabilistically + model them via a Gaussian Process (GP) prior. As such, + properly behaving in such an environment requires balancing a + three-way trade-off of exploring the safety function, + exploring the reward function, and exploiting acquired + knowledge to maximize reward. We propose a novel approach to + balance this trade-off. Specifically, our approach explores + unvisited states selectively; that is, it prioritizes the + exploration of a state if visiting that state significantly + improves the knowledge on the achievable cumulative + reward. Our approach relies on a novel information gain + criterion based on Gaussian Process representations of the + reward and safety functions. We demonstrate the effectiveness + of our approach on a range of experiments, including a + simulation using the real Martian terrain data. +
+
+Keywords: Markov Decision Process, Gaussian Processes +
+ +
+ + +
+[2623] +
+
+Tobias Wagner, Nicola Beume, and Boris Naujoks. + Pareto-, Aggregation-, and Indicator-Based Methods in Many-Objective Optimization. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  742–756. Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2624] +
+
+Markus Wagner, Tobias Friedrich, and Marius Thomas Lindauer. + Improving local search in a minimum vertex cover solver for classes of networks. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  1704–1711, Piscataway, NJ, 2017. IEEE Press.
+[ bib | +DOI ] +
+Keywords: graph theory;search problems;local search;minimum vertex + cover solver;network classes;straightforward alternative + approach;benchmark sets;graphs;algorithm portfolio;single + integrated approach;Training;Portfolios;Algorithm design and + analysis;Prediction algorithms;Machine learning + algorithms;Optimization;Benchmark testing,smac,paramils +
+ +
+ + +
+[2625] +
+
+Markus Wagner and Frank Neumann. + A Fast Approximation-guided Evolutionary Multi-objective Algorithm. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  687–694. ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[2626] +
+
+Benjamin W. Wah and Yi Xin Chen. + Optimal Anytime Constrained Simulated Annealing for Constrained Global Optimization. + In R. Dechter, editor, Principles and Practice of Constraint Programming, CP 2000, volume 1894 of Lecture Notes in Computer Science, pp.  425–440. Springer, Heidelberg, Germany, 2000.
+[ bib | +DOI ] + +
+ + +
+[2627] +
+
+J. P. Walser. + Solving Linear Pseudo-Boolean Constraint Problems with Local Search. + In B. Kuipers and B. L. Webber, editors, Proceedings of AAAI 1997 – Fourteenth National Conference on Artificial Intelligence, pp.  269–274. AAAI Press/MIT Press, Menlo Park, CA, 1997.
+[ bib ] + +
+ + +
+[2628] +
+
+J. P. Walser. + Integer Optimization by Local Search: A Domain-Independent Approach, volume 1637 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1999.
+[ bib ] + +
+ + +
+[2629] +
+
+J. P. Walser, R. Iyer, and N. Venkatasubramanyan. + An Integer Local Search Method with Application to Capacitated Production Planning. + In J. Mostow and C. Rich, editors, Proceedings of AAAI 1998 – Fifteenth National Conference on Artificial Intelligence, pp.  373–379. AAAI Press/MIT Press, Menlo Park, CA, 1998.
+[ bib ] + +
+ + +
+[2630] +
+
+Toby Walsh. + Depth-bounded Discrepancy Search. + In M. E. Pollack, editor, Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), pp.  1388–1395. Morgan Kaufmann Publishers, 1997.
+[ bib ] + +
+ + +
+[2631] +
+
+Handing Wang, John Doherty, and Yaochu Jin. + Hierarchical surrogate-assisted evolutionary multi-scenario airfoil shape optimization. + In Proceedings of the 2018 Congress on Evolutionary Computation (CEC 2018), pp.  1–8, Piscataway, NJ, 2018. IEEE Press.
+[ bib ] +
+Keywords: scenario-based +
+ +
+ + +
+[2632] +
+
+Yanqi Wang, Xingye Dong, Ping Chen, and Youfang Lin. + Iterated local search algorithms for the sequence-dependent setup times flow shop scheduling problem minimizing makespan. + In Foundations of Intelligent Systems, pp.  329–338. Springer, 2014.
+[ bib ] + +
+ + +
+[2633] +
+
+Shaolin Wang, Yi Mei, and Mengjie Zhang. + Two-stage multi-objective genetic programming with archive for uncertain capacitated arc routing problem. + In F. Chicano and K. Krawiec, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, pp.  287–295. ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2634] +
+
+Matthew O. Ward. + Multivariate data glyphs: Principles and practice. + In C.-h. Chen, W. K. Härdle, and A. Unwin, editors, Handbook of Data Visualization, pp.  179–198. Springer, 2008.
+[ bib ] + +
+ + +
+[2635] +
+
+Tony Wauters. + 10 years of Eternity II–from $2 million puzzle to challenging optimization problem. + In International Workshop on Cutting, Packing and Related Topics, Gent, Belgium, 2017.
+[ bib | +http ] +
+The Eternity II (EII) puzzle is a commercial edge matching + puzzle in which 256 square tiles with four coloured edges + must be arranged on a 16 by 16 grid such that all tile edges + are matched. In addition, a complete solution requires that + the `grey' patterns, which appear only on a subset of the + tiles, should be matched to the outer edges of the grid. The + puzzle belongs to the more general class of Edge Matching + Puzzles, which have been shown to be NP-complete. In July + 2007, toy distributor Tomy UK Ltd. released this challenging + edge matching puzzle with a $2 million prize. However, to + the best of our knowledge, no complete solution has ever been + found. Meanwhile, the final scrutiny date for the cash price, + 31 December 2010, has passed, leaving the large money prize + unclaimed. In its 10 years of existence many people tried to + solve EII and some are still trying. Many approaches to Edge + Matching Puzzles are reported in the literature. Among these + approaches are constraint programming and backtracking, + metaheuristics, and evolutionary methods. Other approaches + translate the problem into SAT, MILP or max-clique and then + solve it with appropriate state of the art solvers. Some + approaches have also been implemented on parallel computing + or dedicated hardware. +
+ +
+ + +
+[2636] +
+
+Ingo Wegener. + Simulated annealing beats metropolis in combinatorial optimization. + In L. Caires, G. F. Italiano, L. Monteiro, C. Palamidessi, and M. Yung, editors, Proceedings of the 32nd International Colloquium on Automata, Languages and Programming, ICALP 2005, volume 3580 of Lecture Notes in Computer Science, pp.  589–601. Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2637] +
+
+Chad Wegley, Muzaffar Eusuff, and Kevin E. Lansey. + Determining Pump Operations Using Particle Swarm Optimization. + In R. H. Hotchkiss and M. Glade, editors, Building Partnerships, Proceedings of the Joint Conference on Water Resources Engineering and Water Resources Planning and Management, Minneapolis, USA, 2000.
+[ bib ] + +
+ + +
+[2638] +
+
+Peter Wegner. + Research paradigms in computer science. + In ICSE'76: Proceedings of the 2nd international conference on Software engineering, pp.  322–330, October 1976.
+[ bib ] + +
+ + +
+[2639] +
+
+Kilian Q. Weinberger and Lawrence K. Saul. + An Introduction to Nonlinear Dimensionality Reduction by Maximum Variance Unfolding. + In A. Cohn, editor, Proceedings of the 21st National Conference on Artificial Intelligence, volume 6, pp.  1683–1686. AAAI Press/MIT Press, Menlo Park, CA, 2006.
+[ bib ] + +
+ + +
+[2640] +
+
+Kilian Q. Weinberger, Fei Sha, and Lawrence K. Saul. + Learning a kernel matrix for nonlinear dimensionality reduction. + In C. E. Brodley, editor, Proceedings of the 21st International Conference on Machine Learning, ICML 2004, New York, NY, 2004. ACM Press.
+[ bib | +DOI ] +
+We investigate how to learn a kernel matrix for high + dimensional data that lies on or near a low dimensional + manifold. Noting that the kernel matrix implicitly maps the + data into a nonlinear feature space, we show how to discover + a mapping that "unfolds" the underlying manifold from which + the data was sampled. The kernel matrix is constructed by + maximizing the variance in feature space subject to local + constraints that preserve the angles and distances between + nearest neighbors. The main optimization involves an instance + of semidefinite programming—a fundamentally different + computation than previous algorithms for manifold learning, + such as Isomap and locally linear embedding. The optimized + kernels perform better than polynomial and Gaussian kernels + for problems in manifold learning, but worse for problems in + large margin classification. We explain these results in + terms of the geometric properties of different kernels and + comment on various interpretations of other manifold learning + algorithms as kernel methods. +
+ +
+ + +
+[2641] +
+
+Simon Wessing, Nicola Beume, Günther Rudolph, and Boris Naujoks. + Parameter Tuning Boosts Performance of Variation Operators in Multiobjective Optimization. + In R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors, Parallel Problem Solving from Nature, PPSN XI, volume 6238 of Lecture Notes in Computer Science, pp.  728–737. Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2642] +
+
+Clint R. Whaley. + ATLAS: Automatically Tuned Linear Algebra Software. + In D. Padua, editor, Encyclopedia of Parallel Computing, pp.  95–101. Springer, US, 2011.
+[ bib | +DOI ] + +
+ + +
+[2643] +
+
+L. While and L. Bradstreet. + Applying the WFG Algorithm to Calculate Incremental Hypervolumes. + In Proceedings of the 2012 Congress on Evolutionary Computation (CEC 2012), pp.  1–8, Piscataway, NJ, 2012. IEEE Press.
+[ bib ] + +
+ + +
+[2644] +
+
+T. White, B. Pagurek, and F. Oppacher. + Connection Management Using Adaptive Mobile Agents. + In H. R. Arabnia, editor, Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA'98), pp.  802–809. CSREA Press, 1998.
+[ bib ] + +
+ + +
+[2645] +
+
+W. Wiesemann and Thomas Stützle. + Iterated Ants: An Experimental Study for the Quadratic Assignment Problem. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science, pp.  179–190. Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2646] +
+
+Angelika Wiegele. + Biq Mac Library – A collection of Max-Cut and quadratic 0-1 programming instances of medium size. + Technical report, Institut für Mathematik, Alpen-Adria-Universität Klagenfurt, 2007.
+[ bib | +http ] + +
+ + +
+[2647] +
+
+Angelika Wiegele. + Biq Mac Library – Binary Quadratic and Max Cut Library. + http://biqmac.aau.at/biqmaclib.html, 2007.
+[ bib ] + +
+ + +
+[2648] +
+
+Andrzej P. Wierzbicki. + The Use of Reference Objectives in Multiobjective Optimisation. + In G. Fandel and T. Gal, editors, Multiple Criteria Decision Making Theory and Application, number 177 in Lecture Notes in Economics and Mathematical Systems, pp.  468–486. Springer, Heidelberg, Germany, 1980.
+[ bib | +DOI ] + +
+ + +
+[2649] +
+
+David P. Williamson and David B. Shmoys. + The design of approximation algorithms. + Cambridge University Press, 2011.
+[ bib ] + +
+ + +
+[2650] +
+
+Steffen Wolf and Peter Merz. + Iterated Local Search for Minimum Power Symmetric Connectivity in Wireless Networks. + In C. Cotta and P. Cowling, editors, Proceedings of EvoCOP 2009 – 9th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 5482 of Lecture Notes in Computer Science, pp.  192–203. Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2651] +
+
+Lin Xu, Holger H. Hoos, and Kevin Leyton-Brown. + Hydra: Automatically Configuring Algorithms for Portfolio-Based Selection. + In M. Fox and D. Poole, editors, Proceedings of the AAAI Conference on Artificial Intelligence. AAAI Press, 2010.
+[ bib | +DOI ] +
+Keywords: automated algorithm design; portfolio-based algorithm + selection; automated algorithm configuration; SAT; stochastic + local search +
+ +
+ + +
+[2652] +
+
+Lin Xu, Frank Hutter, Holger H. Hoos, and Kevin Leyton-Brown. + Hydra-MIP: Automated Algorithm Configuration and Selection for Mixed Integer Programming. + Technical Report TR-2011-01, Department of Computer Science, University of British Columbia, Canada, 2011.
+[ bib | +http ] + +
+ + +
+[2653] +
+
+Lin Xu, A. R. KhudaBukhsh, Holger H. Hoos, and Kevin Leyton-Brown. + Quantifying the similarity of algorithm configurations. + In P. Festa, M. Sellmann, and J. Vanschoren, editors, Learning and Intelligent Optimization, 10th International Conference, LION 10, volume 10079 of Lecture Notes in Computer Science, pp.  203–217. Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[2654] +
+
+Jian-Wu Xu, Puskal P. Pokharel, António R. C. Paiva, and José C. Príncipe. + Nonlinear Component Analysis Based on Correntropy. + In Proceedings of the International Joint Conference on Neural Networks, IJCNN 2006, pp.  1851–1855. IEEE, 2006.
+[ bib | +DOI ] + +
+ + +
+[2655] +
+
+Anil Yaman, Ahmed Hallawa, Matt Coler, and Giovanni Iacca. + Presenting the ECO: evolutionary computation ontology. + In G. Squillero and K. Sim, editors, Applications of Evolutionary Computation, volume 10199 of Lecture Notes in Computer Science, pp.  603–619. Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[2656] +
+
+Xin Yao. + Evolutionary Computation: Theory and Applications. + World Scientific Singapore, River Edge, NJ, 1999.
+[ bib ] +
+Keywords: Evolutionary programming (Computer science); Neural networks + (Computer science); Evolutionary computation +
+ +
+ + +
+[2657] +
+
+A. Yarimcam, S. Asta, Ender Özcan, and Andrew J. Parkes. + Heuristic Generation via Parameter Tuning for Online Bin Packing. + In P. Angelov et al., editors, Evolving and Autonomous Learning Systems (EALS), 2014 IEEE Symposium on, pp.  102–108. IEEE, 2014.
+[ bib | +DOI ] +
+Keywords: irace +
+ +
+ + +
+[2658] +
+
+Carlos Yasojima, Tiago Araújo, Bianchi Meiguins, Nelson Neto, and Jefferson Morais. + A Comparison of Genetic Algorithms and Particle Swarm Optimization to Estimate Cluster-Based Kriging Parameters. + In P. Moura Oliveira, P. Novais, and L. P. Reis, editors, Progress in Artificial Intelligence, pp.  750–761. Springer International Publishing, Cham, Switzerland, 2019.
+[ bib ] +
+Kriging is one of the most used spatial estimation methods in + real-world applications. Some kriging parameters must be + estimated in order to reach a good accuracy in the + interpolation process, however, this task remains a + challenge. Various optimization methods have been tested to + find good parameters of the kriging process. In recent years, + many authors are using bio-inspired techniques and achieving + good results in estimating these parameters in comparison + with traditional techniques. This paper presents a comparison + between well known bio-inspired techniques such as Genetic + Algorithms and Particle Swarm Optimization in the estimation + of the essential kriging parameters: nugget, sill, range, + angle, and factor. In order to perform the tests, we proposed + a methodology based on the cluster-based kriging + method. Considering the Friedman test, the results showed no + statistical difference between the evaluated algorithms in + optimizing kriging parameters. On the other hand, the + Particle Swarm Optimization approach presented a faster + convergence, which is important in this high-cost + computational problem. +
+ +
+ + +
+[2659] +
+
+Gürcan Yavuz, Doǧan Aydın, and Thomas Stützle. + Self-adaptive Search Equation-based Artificial Bee Colony Algorithm on the CEC 2014 Benchmark Functions. + In Proceedings of the 2016 Congress on Evolutionary Computation (CEC 2016), pp.  1173–1180, Piscataway, NJ, 2016. IEEE Press.
+[ bib ] + +
+ + +
+[2660] +
+
+Cliff Young, David S. Johnson, David R. Karger, and Michael D. Smith. + Near-optimal Intraprocedural Branch Alignment. + In M. C. Chen, R. K. Cytron, and A. M. Berman, editors, Proceedings of the ACM SIGPLAN'97 Conference on Programming Language Design and Implementation (PLDI), Las Vegas, Nevada, pp.  183–193. ACM Press, 1997.
+[ bib ] + +
+ + +
+[2661] +
+
+Philip L. H. Yu, Wai Ming Wan, and Paul H. Lee. + Decision Tree Modeling for Ranking Data. + In J. Fürnkranz and E. Hüllermeier, editors, Preference Learning, pp.  83–106. Springer, Heidelberg, Germany, 2011.
+[ bib | +DOI ] + +
+ + +
+[2662] +
+
+Zhi Yuan, Armin Fügenschuh, Henning Homfeld, Prasanna Balaprakash, Thomas Stützle, and Michael Schoch. + Iterated Greedy Algorithms for a Real-World Cyclic Train Scheduling Problem. + In M. J. Blesa, C. Blum, C. Cotta, A. J. Fernández, J. E. Gallardo, A. Roli, and M. Sampels, editors, Hybrid Metaheuristics, volume 5296 of Lecture Notes in Computer Science, pp.  102–116. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2663] +
+
+Bo Yuan and Marcus Gallagher. + Statistical Racing Techniques for Improved Empirical Evaluation of Evolutionary Algorithms. + In X. Yao et al., editors, Parallel Problem Solving from Nature – PPSN VIII, volume 3242 of Lecture Notes in Computer Science, pp.  172–181. Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2664] +
+
+Bo Yuan and Marcus Gallagher. + Combining Meta-EAs and racing for difficult EA parameter tuning tasks. + In F. Lobo, C. F. Lima, and Z. Michalewicz, editors, Parameter Setting in Evolutionary Algorithms, pp.  121–142. Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[2665] +
+
+Zhi Yuan, Marco A. Montes de Oca, Thomas Stützle, Hoong Chuin Lau, and Mauro Birattari. + An Analysis of Post-selection in Automatic Configuration. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2013, pp.  1557–1564. ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[2666] +
+
+Lin Yuefeng, Wenli Du, and Thomas Stützle. + Three L-SHADE Based Algorithms on Mixed-variables Optimization Problems. + In Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), pp.  2274–2281, Piscataway, NJ, 2017. IEEE Press.
+[ bib ] + +
+ + +
+[2667] +
+
+Joseph Yuen, Sophia Gao, Markus Wagner, and Frank Neumann. + An adaptive data structure for evolutionary multi-objective algorithms with unbounded archives. + In Proceedings of the 2012 Congress on Evolutionary Computation (CEC 2012), pp.  1–8, Piscataway, NJ, 2012. IEEE Press.
+[ bib ] + +
+ + +
+[2668] +
+
+Xi Yun and Susan L. Epstein. + Learning Algorithm Portfolios for Parallel Execution. + In Y. Hamadi and M. Schoenauer, editors, Learning and Intelligent Optimization, 6th International Conference, LION 6, volume 7219 of Lecture Notes in Computer Science, pp.  323–338. Springer, Heidelberg, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2669] +
+
+Martin Zaefferer, J. Stork, and Thomas Bartz-Beielstein. + Distance Measures for Permutations in Combinatorial Efficient Global Optimization. + In T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors, Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science, pp.  373–383. Springer, Heidelberg, Germany, 2014.
+[ bib | +DOI ] +
+Keywords: CEGO, Bayesian optimization +
+ +
+ + +
+[2670] +
+
+Martin Zaefferer, J. Stork, M. Friese, Andreas Fischbach, Boris Naujoks, and Thomas Bartz-Beielstein. + Efficient Global Optimization for Combinatorial Problems. + In C. Igel and D. V. Arnold, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2014, pp.  871–878. ACM Press, New York, NY, 2014.
+[ bib | +DOI ] +
+Proposed CEGO algorithm +
+
+Keywords: CEGO, Bayesian optimization +
+ +
+ + +
+[2671] +
+
+Emmanuel Zarpas. + Benchmarking SAT solvers for bounded model checking. + In F. Bacchus and T. Walsh, editors, International Conference on Theory and Applications of Satisfiability Testing, volume 3569, pp.  340–354, 2005.
+[ bib ] + +
+ + +
+[2672] +
+
+Tiantian Zhang, Michael Georgiopoulos, and Georgios C. Anagnostopoulos. + S-Race: A Multi-Objective Racing Algorithm. + In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2013, pp.  1565–1572. ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[2673] +
+
+Tiantian Zhang, Michael Georgiopoulos, and Georgios C. Anagnostopoulos. + SPRINT: Multi-Objective Model Racing. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  1383–1390. ACM Press, New York, NY, 2015.
+[ bib | +DOI ] +
+Extended version published as [1387] +
+
+Keywords: model selection, multi-objective optimization, racing + algorithm, sequential probability ratio test +
+ +
+ + +
+[2674] +
+
+Qingfu Zhang, Wudong Liu, and Hui Li. + The Performance of a New Version of MOEA/D on CEC09 Unconstrained MOP Test Instances. + In Proceedings of the 2009 Congress on Evolutionary Computation (CEC 2009), pp.  203–208, Piscataway, NJ, 2009. IEEE Press.
+[ bib ] + +
+ + +
+[2675] +
+
+Qingfu Zhang, A. Zhou, S. Zhao, Ponnuthurai N. Suganthan, W. Liu, and S. Tiwari. + Multiobjective Optimization Test Instances for the CEC 2009 Special Session and Competition. + Working Report CES-487, School of Computer Science and Electronic Engieering, University of Essex, April 2009.
+[ bib ] +
+Proposed UF benchmark +
+ +
+ + +
+[2676] +
+
+Qingfu Zhang and Ponnuthurai N. Suganthan. + Special Session on Performance Assessment of Multiobjective Optimization Algorithms/CEC'09 MOEA Competition. + https://www3.ntu.edu.sg/home/epnsugan/index_files/CEC09-MOEA/CEC09-MOEA.htm, 2009.
+[ bib ] +
+Previously available at http://dces.essex.ac.uk/staff/qzhang/moeacompetition09.htm +
+ +
+ + +
+[2677] +
+
+Heiner Zille, Hisao Ishibuchi, Sanaz Mostaghim, and Yusuke Nojima. + Mutation operators based on variable grouping for multi-objective large-scale optimization. + In X. Chen and A. Stafylopatis, editors, Computational Intelligence (SSCI), 2016 IEEE Symposium Series on, pp.  1–8, 2016.
+[ bib | +DOI ] +
+linked polynomial mutation +
+ +
+ + +
+[2678] +
+
+Eckart Zitzler, Dimo Brockhoff, and Lothar Thiele. + The Hypervolume Indicator Revisited: On the Design of Pareto-compliant Indicators Via Weighted Integration. + In S. Obayashi et al., editors, Evolutionary Multi-criterion Optimization, EMO 2007, volume 4403 of Lecture Notes in Computer Science, pp.  862–876. Springer, Heidelberg, Germany, 2007.
+[ bib | +DOI | +supplementary material ] + +
+ + +
+[2679] +
+
+Eckart Zitzler, Joshua D. Knowles, and Lothar Thiele. + Quality Assessment of Pareto Set Approximations. + In J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors, Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science, pp.  373–404. Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[2680] +
+
+Eckart Zitzler and Simon Künzli. + Indicator-based Selection in Multiobjective Search. + In X. Yao et al., editors, Parallel Problem Solving from Nature – PPSN VIII, volume 3242 of Lecture Notes in Computer Science, pp.  832–842. Springer, Heidelberg, Germany, 2004.
+[ bib ] +
+Keywords: IBEA +
+ +
+ + +
+[2681] +
+
+Eckart Zitzler, Marco Laumanns, and Lothar Thiele. + SPEA2: Improving the Strength Pareto Evolutionary Algorithm. + Technical Report 103, Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Zürich, Switzerland, 2001.
+[ bib ] +
+Published as [2682] +
+ +
+ + +
+[2682] +
+
+Eckart Zitzler, Marco Laumanns, and Lothar Thiele. + SPEA2: Improving the Strength Pareto Evolutionary Algorithm for Multiobjective Optimization. + In K. C. Giannakoglou, D. T. Tsahalis, J. Periaux, K. D. Papaliliou, and T. Fogarty, editors, Evolutionary Methods for Design, Optimisation and Control, pp.  95–100. CIMNE, Barcelona, Spain, 2002.
+[ bib ] +
+Proposed SPEA2. +
+ +
+ + +
+[2683] +
+
+Eckart Zitzler and Lothar Thiele. + Multiobjective Optimization Using Evolutionary Algorithms - A Comparative Case Study. + In A. E. Eiben, T. Bäck, M. Schoenauer, and H.-P. Schwefel, editors, Parallel Problem Solving from Nature – PPSN V, volume 1498 of Lecture Notes in Computer Science, pp.  292–301. Springer, Heidelberg, Germany, 1998.
+[ bib | +DOI ] +
+Proposed hypervolume measure +
+ +
+ + +
+[2684] +
+
+Eckart Zitzler, Lothar Thiele, and Johannes Bader. + SPAM: Set Preference Algorithm for Multiobjective Optimization. + In G. Rudolph et al., editors, Parallel Problem Solving from Nature – PPSN X, volume 5199 of Lecture Notes in Computer Science, pp.  847–858. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2685] +
+
+Santosh Tiwari, Patrick Koch, Georges Fadel, and Kalyanmoy Deb. + AMGA: An archive-based micro genetic algorithm for multi-objective optimization. + In C. Ryan, editor, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2008, pp.  729–736. ACM Press, New York, NY, 2008.
+[ bib | +DOI ] + +
+ + +
+[2686] +
+
+Eckart Zitzler. + Evolutionary Algorithms for Multiobjective Optimization: Methods and Applications. + PhD thesis, ETH Zürich, Switzerland, 1999.
+[ bib ] + +
+ + +
+[2687] +
+
+Andrejs Zujevs and Janis Eiduks. + New decision maker model for multiobjective optimization interactive methods. + In 17th International Conference on Information and Software Technologies, Kaunas, Lithuania, pp.  51–58, 2011.
+[ bib | +epub ] +
+Keywords: Machine Decision Maker +
+ +
+ + +
+[2688] +
+
+F. E. B. Otero, A. A. Freitas, and C. G. Johnson. + cAnt-Miner: An Ant Colony Classification Algorithm to Cope with Continuous Attributes. + In M. Dorigo et al., editors, Ant Colony Optimization and Swarm Intelligence, 6th International Conference, ANTS 2008, volume 5217 of Lecture Notes in Computer Science, pp.  48–59. Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2689] +
+
+Axel de Perthuis de Laillevault, Benjamin Doerr, and Carola Doerr. + Money for Nothing: Speeding Up Evolutionary Algorithms Through Better Initialization. + In S. Silva and A. I. Esparcia-Alcázar, editors, Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2015, pp.  815–822. ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[2690] +
+
+Jeremy Rapin and Olivier Teytaud. + Nevergrad: A gradient-free optimization platform. + https://github.com/FacebookResearch/Nevergrad, 2018.
+[ bib ] + +
+ + +
+[2691] +
+
+OscaR Team. + OscaR: Scala in OR, 2012. + Available from https://bitbucket.org/oscarlib/oscar.
+[ bib ] + +
+ + +
+[2692] +
+
+Juan Luis Cano Rodríguez et al. + poliastro: Astrodynamics in Python. + Zenodo, 2015.
+[ bib | +DOI ] + +
+ + +
+[2693] +
+
+Scott Robert Ladd. + ACOVEA (Analysis of Compiler Options via Evolutionary Algorithm). + https://github.com/Acovea/libacovea, 2000.
+[ bib ] + +
+ + +
+[2694] +
+
+GNU Project, Free Software Foundation. + GCC, the GNU Compiler Collection. + https://gcc.gnu.org, 1987.
+[ bib ] + +
+ + +
+[2695] +
+
+Carlos Ansótegui, Meinolf Sellmann, and Kevin Tierney. + GGA: Gender-based Genetic Algorithm Configurator. + https://bitbucket.org/gga_ac/, 2017. + Version visited last on July 2017.
+[ bib ] + +
+ + +
+[2696] +
+
+Intel. + Intel Software Autotuning Tool. + https://software.intel.com/en-us/articles/intel-software-autotuning-tool/, 2010.
+[ bib ] + +
+ + +
+[2697] +
+
+Qingfu Zhang. + MOEA/D homepage. + https://sites.google.com/view/moead/, 2007.
+[ bib ] +
+Previous URL was at + https://dces.essex.ac.uk/staff/zhang/webofmoead.htm. +
+ +
+ + +
+[2698] +
+
+ML4AAD Group. + SMAC v3 Project. + https://github.com/automl/SMAC3, 2017. + Version visited last on August 2017.
+[ bib ] + +
+ + +
+[2699] +
+
+Gerhard Reinelt. + TSPLIB. + http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/, 1995. + Version visited last on 24 February 2023.
+[ bib ] + +
+ + +
+[2700] +
+
+William J. Cook. + The Traveling Salesman Problem. + http://www.math.uwaterloo.ca/tsp, 2010. + Version visited last on 15 April 2014.
+[ bib ] + +
+ + +
+[2701] +
+
+Frank Hutter, Holger H. Hoos, Kevin Leyton-Brown, and Thomas Stützle. + ParamILS. + http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/, 2017. + Version visited last on July 2017.
+[ bib ] + +
+ + +
+[2702] +
+
+Jakobus E. van Zyl. + A Methodology for Improved Operational Optimization of Water Distribution Systems. + PhD thesis, School of Engineering and Computer Science, University of Exeter, UK, 2001.
+[ bib ] + +
+ + +
+[2703] +
+
+H. E. Shrobe, T. M. Mitchell, and R. G. Smith, editors. + Proceedings of the 7th National Conference on Artificial Intelligence, St. Paul, MN, August 21-26, AAAI-88. AAAI Press/MIT Press, Menlo Park, CA, 1988.
+[ bib | +http ] + +
+ + +
+[2704] +
+
+W. R. Swartout, editor. + Proceedings of the 10th National Conference on Artificial Intelligence. AAAI Press/MIT Press, Menlo Park, CA, 1992.
+[ bib ] + +
+ + +
+[2705] +
+
+R. Fikes and W. G. Lehnert, editors. + Proceedings of the 11th National Conference on Artificial Intelligence. AAAI Press/MIT Press, Menlo Park, CA, 1993.
+[ bib ] + +
+ + +
+[2706] +
+
+B. Kuipers and B. L. Webber, editors. + Proceedings of the Fourteenth National Conference on Artificial Intelligence and Ninth Innovative Applications of Artificial Intelligence Conference, AAAI 97, IAAI 97, July 27-31, 1997, Providence, Rhode Island. AAAI Press/MIT Press, Menlo Park, CA, 1997.
+[ bib ] + +
+ + +
+[2707] +
+
+J. Mostow and C. Rich, editors. + Proceedings of the Fifteenth National Conference on Artificial Intelligence and Tenth Innovative Applications of Artificial Intelligence Conference, AAAI 98, IAAI 98, July 26-30, 1998, Madison, Wisconsin, USA. AAAI Press/MIT Press, Menlo Park, CA, 1998.
+[ bib ] + +
+ + +
+[2708] +
+
+H. A. Kautz and B. W. Porter, editors. + Proceedings of the Seventeenth National Conference on Artificial Intelligence and Twelfth Conference on on Innovative Applications of Artificial Intelligence, July 30 – August 3, 2000, Austin, Texas, USA. AAAI Press/MIT Press, Menlo Park, CA, 2000.
+[ bib ] + +
+ + +
+[2709] +
+
+A. Cohn, editor. + Proceedings, The Twenty-First National Conference on Artificial Intelligence and the Eighteenth Innovative Applications of Artificial Intelligence Conference, July 16-20, 2006, Boston, Massachusetts, USA, volume 6. AAAI Press/MIT Press, Menlo Park, CA, 2006.
+[ bib ] + +
+ + +
+[2710] +
+
+R. C. Holte and A. Howe, editors. + Proceedings of the Twenty-Second AAAI Conference on Artificial Intelligence, July 22-26, 2007, Vancouver, British Columbia, Canada. AAAI Press/MIT Press, Menlo Park, CA, 2007.
+[ bib ] + +
+ + +
+[2711] +
+
+M. Fox and D. Poole, editors. + Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2010, Atlanta, Georgia, USA, July 11-15, 2010. AAAI Press, 2010.
+[ bib ] + +
+ + +
+[2712] +
+
+W. Burgard and D. Roth, editors. + Proceedings of the Twenty-Fifth AAAI Conference on Artificial Intelligence, AAAI 2011, San Francisco, California, USA, August 07-11, 2011. AAAI Press, 2011.
+[ bib ] + +
+ + +
+[2713] +
+
+J. Hoffmann and B. Selman, editors. + Proceedings of the Twenty-Sixth AAAI Conference on Artificial Intelligence, AAAI 2012, Toronto, Ontario, Canada, July 22-26, 2012. AAAI Press, 2012.
+[ bib ] + +
+ + +
+[2714] +
+
+D. Stracuzzi et al., editors. + Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence, AAAI 2014, Québec City, Québec, Canada, July 27-31, 2014. AAAI Press, 2014.
+[ bib ] + +
+ + +
+[2715] +
+
+B. Bonet and S. Koenig, editors. + Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence, AAAI 2015, Austin, Texas, USA, January 25-30, 2015. AAAI Press, 2015.
+[ bib ] + +
+ + +
+[2716] +
+
+D. Schuurmans and M. P. Wellman, editors. + Proceedings of the Thirtieth AAAI Conference on Artificial Intelligence, AAAI 2016, February 12-17, 2016, Phoenix, Arizona, USA. + AAAI Press, 2016.
+[ bib | +epub ] + +
+ + +
+[2717] +
+
+S. P. Singh and S. Markovitch, editors. + Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, February 4-9, 2017, San Francisco, California, USA. + AAAI Press, February 2017.
+[ bib ] + +
+ + +
+[2718] +
+
+S. A. McIlraith and K. Q. Weinberger, editors. + Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence, February 2-7, 2018, New Orleans, Louisiana, USA. + AAAI Press, February 2018.
+[ bib ] + +
+ + +
+[2719] +
+
+The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, The Thirty-Second Innovative Applications of Artificial Intelligence Conference, IAAI 2020, The Tenth AAAI Symposium on Educational Advances in Artificial Intelligence, EAAI 2020, New York, NY, USA, February 7-12, 2020. AAAI Press, 2020.
+[ bib ] + +
+ + +
+[2720] +
+
+M. Randall, H. A. Abbass, and J. Wiles, editors. + Progress in Artificial Life, Third Australian Conference, ACAL 2007, volume 4828 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2721] +
+
+2019 American Control Conference, ACC 2019, Philadelphia, PA, USA, July 10-12, 2019. IEEE, 2019.
+[ bib ] + +
+ + +
+[2722] +
+
+F. Rossi and A. Tsoukiàs, editors. + Algorithmic Decision Theory, First International Conference, ADT 2009, Venice, Italy, October 20-23, 2009, volume 5783 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2723] +
+
+R. I. Brafman, F. Roberts, and A. Tsoukiàs, editors. + Algorithmic Decision Theory, Third International Conference, ADT 2011, Piscataway, New Jersey, USA, October 26-28, 2011, volume 6992 of Lecture Notes in Artificial Intelligence. + Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2724] +
+
+M. C. Golumbic et al., editors. + Fifth International Symposium on Artificial Intelligence and Mathematics, AIM 1998, Fort Lauderdale, Florida, USA, January 4-6, 1998, 1998.
+[ bib ] + +
+ + +
+[2725] +
+
+R. Silhavy, R. Senkerik, Z. K. Oplatkova, P. Silhavy, and Z. Prokopova, editors. + Artificial Intelligence Perspectives in Intelligent Systems, volume 464 of Advances in Intelligent Systems and Computing. + Springer International Publishing, 2016.
+[ bib ] + +
+ + +
+[2726] +
+
+T. C. Fogarty, editor. + Evolutionary Computing, AISB Workshop, Sheffield, UK, April 3-4, 1995, Selected Papers, volume 993 of Lecture Notes in Computer Science. + Springer, Berlin, Germany, 1995.
+[ bib ] + +
+ + +
+[2727] +
+
+A. Gretton and C. C. Robert, editors. + Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, AISTATS 2016, Cadiz, Spain, May 9-11, 2016, volume 51 of JMLR Workshop and Conference Proceedings. JMLR.org, 2016.
+[ bib ] + +
+ + +
+[2728] +
+
+S. Jain, R. Munos, F. Stephan, and T. Zeugmann, editors. + Algorithmic Learning Theory - 24th International Conference, ALT 2013, Singapore, October 6-9, 2013. Proceedings, volume 8139 of Lecture Notes in Computer Science. + Springer, Berlin, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2729] +
+
+C. A. Coello Coello, C. Dhaenens, and L. Jourdan, editors. + Advances in Multi-Objective Nature Inspired Computing, volume 272 of Studies in Computational Intelligence. + Springer, 2010.
+[ bib ] + +
+ + +
+[2730] +
+
+D. Cliff, P. Husbands, J.-A. Meyer, and S. Wilson, editors. + Proceedings of the third international conference on Simulation of adaptive behavior: From Animals to Animats 3. + MIT Press, Cambridge, MA, 1994.
+[ bib ] + +
+ + +
+[2731] +
+
+M. Dorigo et al., editors. + Abstract proceedings of ANTS 2000 – From Ant Colonies to Artificial Ants: Second International Workshop on Ant Algorithms. IRIDIA, Université Libre de Bruxelles, Belgium, September 7–9 2000.
+[ bib ] + +
+ + +
+[2732] +
+
+M. Dorigo et al., editors. + Ant Algorithms, Third International Workshop, ANTS 2002, Brussels, Belgium, September 12-14, 2002, Proceedings, volume 2463 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2733] +
+
+M. Dorigo et al., editors. + Ant Colony Optimization and Swarm Intelligence, 4th International Workshop, ANTS 2004, volume 3172 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2734] +
+
+M. Dorigo et al., editors. + Ant Colony Optimization and Swarm Intelligence, 5th International Workshop, ANTS 2006, volume 4150 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2735] +
+
+M. Dorigo et al., editors. + Ant Colony Optimization and Swarm Intelligence, 6th International Conference, ANTS 2008, volume 5217 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2736] +
+
+M. Dorigo et al., editors. + Ant Colony Optimization and Swarm Intelligence, 7th International Conference, ANTS 2010, volume 6234 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2737] +
+
+M. Dorigo et al., editors. + Swarm Intelligence, 8th International Conference, ANTS 2012, volume 7461 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2738] +
+
+M. Dorigo et al., editors. + Swarm Intelligence, 9th International Conference, ANTS 2014, volume 8667 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2739] +
+
+M. Dorigo, M. Birattari, X. Li, M. López-Ibáñez, K. Ohkura, C. Pinciroli, and T. Stützle, editors. + Swarm Intelligence, 10th International Conference, ANTS 2016, Brussels, Belgium, September 7-9, 2016, Proceedings, volume 9882 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[2740] +
+
+M. Dorigo, M. Birattari, A. L. Christensen, A. Reina, and V. Trianni, editors. + Swarm Intelligence, 11th International Conference, ANTS 2018, Rome, Italy, October 29-31, 2018, Proceedings, volume 11172 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2018.
+[ bib ] + +
+ + +
+[2741] +
+
+M. Dorigo, T. Stützle, M. J. Blesa, C. Blum, H. Hamann, and M. K. Heinrich, editors. + Swarm Intelligence, 12th International Conference, ANTS 2020, Barcelona, Spain, October 26-28, 2020, Proceedings, volume 12421 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2020.
+[ bib ] + +
+ + +
+[2742] +
+
+M. Dorigo, H. Hamann, M. López-Ibáñez, J. García-Nieto, A. Engelbrecht, C. Pinciroli, V. Strobel, and C. L. Camacho-Villalón, editors. + Swarm Intelligence, 13th International Conference, ANTS 2022, Málaga, Spain, November 2-4, 2022, Proceedings, volume 13491 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2022.
+[ bib | +DOI ] + +
+ + +
+[2743] +
+
+Y. Hamadi, E. Monfroy, and F. Saubion, editors. + Autonomous Search. + Springer, Berlin, Germany, 2012.
+[ bib ] + +
+ + +
+[2744] +
+
+C. Maksimović, D. Butler, and F. A. Memon, editors. + Advances in Water Supply Management: Proceedings of the CCWI '03 Conference, London, 15-17 September 2003. + CRC Press, 2003.
+[ bib ] + +
+ + +
+[2745] +
+
+E. H. L. Aarts and J. K. Lenstra, editors. + Local Search in Combinatorial Optimization. + John Wiley & Sons, Chichester, UK, 1997.
+[ bib ] + +
+ + +
+[2746] +
+
+A. Abraham, L. Jain, and R. Goldberg, editors. + Evolutionary Multiobjective Optimization. + Advanced Information and Knowledge Processing. Springer, London, UK, January 2005.
+[ bib ] + +
+ + +
+[2747] +
+
+U. K. Chakraborty, editor. + Advances in differential evolution. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2748] +
+
+B. Filipič and J. Šilc, editors. + Bioinspired optimization methods and their applications: Proceedings of the International Conference on Bioinspired Optimization Methods and their Applications - BIOMA 2004, 11-12 October 2004, Ljubljana, Slovenia, 2004.
+[ bib | +http ] + +
+ + +
+[2749] +
+
+L. Cao, W. Kosters, and J. Lijffijt, editors. + Proceedings of the 32nd Benelux Conference on Artificial Intelligence, BNAIC 2020, Leiden, The Netherlands, 19-20 November 2020, 2020.
+[ bib | +http ] + +
+ + +
+[2750] +
+
+T. Bartz-Beielstein, M. Chiarandini, L. Paquete, and M. Preuss, editors. + Experimental Methods for the Analysis of Optimization Algorithms. + Springer, Berlin, Germany, 2010.
+[ bib ] + +
+ + +
+[2751] +
+
+T. Bartz-Beielstein, B. Filipič, P. Korošec, and E.-G. Talbi, editors. + High-Performance Simulation-Based Optimization. + Springer International Publishing, Cham, Switzerland, 2020.
+[ bib ] + +
+ + +
+[2752] +
+
+C. Blum, M. J. Blesa, A. Roli, and M. Sampels, editors. + Hybrid Metaheuristics: An emergent approach for optimization, volume 114 of Studies in Computational Intelligence. + Springer, Berlin, Germany, 2008.
+[ bib ] + +
+ + +
+[2753] +
+
+Y. Borenstein and A. Moraglio, editors. + Theory and Principled Methods for the Design of Metaheuristics. + Natural Computing Series. Springer, Berlin/Heidelberg, 2014.
+[ bib ] + +
+ + +
+[2754] +
+
+J. M. Puerta, J. A. Gámez, B. Dorronsoro, E. Barrenechea, A. Troncoso, B. Baruque, and M. Galar, editors. + Advances in Artificial Intelligence: 16th Conference of the Spanish Association for Artificial Intelligence, CAEPIA 2015 Albacete, Spain, November 9-12, 2015 Proceedings, volume 9422 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2755] +
+
+R. E. Miller and W. Thatcher, James, editors. + Complexity of Computer Computations, The IBM Research Symposia Series. Springer, 1972.
+[ bib ] + +
+ + +
+[2756] +
+
+Proceedings of the 2010 International Conference on Computing, Control and Industrial Engineering, Los Alamitos, CA, 2010. IEEE Computer Society Press.
+[ bib ] + +
+ + +
+[2757] +
+
+D. A. Savic, G. A. Walters, R. King, and S. Thiam-Khu, editors. + Proceedings of the Eighth International Conference on Computing and Control for the Water Industry (CCWI 2005), volume 1, University of Exeter, UK, September 2005.
+[ bib ] + +
+ + +
+[2758] +
+
+Proceedings of the 1999 Congress on Evolutionary Computation (CEC 1999), Piscataway, NJ, 1999. IEEE Press.
+[ bib ] + +
+ + +
+[2759] +
+
+Proceedings of the 2000 Congress on Evolutionary Computation (CEC 2000), Piscataway, NJ, July 2000. IEEE Press.
+[ bib ] + +
+ + +
+[2760] +
+
+Proceedings of the 2001 Congress on Evolutionary Computation (CEC 2001), Piscataway, NJ, 2001. IEEE Press.
+[ bib ] + +
+ + +
+[2761] +
+
+Proceedings of the 2002 Congress on Evolutionary Computation (CEC'02), Piscataway, NJ, 2002. IEEE Press.
+[ bib ] + +
+ + +
+[2762] +
+
+Proceedings of the 2003 Congress on Evolutionary Computation (CEC 2003), Canberra, ACT, Australia, Piscataway, NJ, December 2003. IEEE Press.
+[ bib ] + +
+ + +
+[2763] +
+
+Proceedings of the 2004 Congress on Evolutionary Computation (CEC 2004), Piscataway, NJ, September 2004. IEEE Press.
+[ bib ] + +
+ + +
+[2764] +
+
+Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), Piscataway, NJ, September 2005. IEEE Press.
+[ bib ] + +
+ + +
+[2765] +
+
+Proceedings of the 2006 Congress on Evolutionary Computation (CEC 2006), Piscataway, NJ, July 2006. IEEE Press.
+[ bib ] + +
+ + +
+[2766] +
+
+Proceedings of the IEEE Congress on Evolutionary Computation, CEC 2007, 25-28 September 2007, Singapore, Piscataway, NJ, 2007. IEEE Press.
+[ bib ] + +
+ + +
+[2767] +
+
+Proceedings of the IEEE Congress on Evolutionary Computation, CEC 2008, June 1-6, 2008, Hong Kong, China, Piscataway, NJ, 2008. IEEE Press.
+[ bib ] + +
+ + +
+[2768] +
+
+Proceedings of the 2009 Congress on Evolutionary Computation (CEC 2009), Piscataway, NJ, 2009. IEEE Press.
+[ bib ] + +
+ + +
+[2769] +
+
+H. Ishibuchi et al., editors. + Proceedings of the 2010 Congress on Evolutionary Computation (CEC 2010), Piscataway, NJ, 2010. IEEE Press.
+[ bib ] + +
+ + +
+[2770] +
+
+Proceedings of the 2011 Congress on Evolutionary Computation (CEC 2011), New Orleans, LA, USA, Piscataway, NJ, 2011. IEEE Press.
+[ bib ] + +
+ + +
+[2771] +
+
+Proceedings of the 2012 Congress on Evolutionary Computation (CEC 2012), Piscataway, NJ, 2012. IEEE Press.
+[ bib ] + +
+ + +
+[2772] +
+
+Proceedings of the 2013 Congress on Evolutionary Computation (CEC 2013), Piscataway, NJ, 2013. IEEE Press.
+[ bib ] + +
+ + +
+[2773] +
+
+Proceedings of the 2014 Congress on Evolutionary Computation (CEC 2014), Piscataway, NJ, 2014. IEEE Press.
+[ bib ] + +
+ + +
+[2774] +
+
+Proceedings of the 2015 Congress on Evolutionary Computation (CEC 2015), Piscataway, NJ, 2015. IEEE Press.
+[ bib ] + +
+ + +
+[2775] +
+
+IEEE Congress on Evolutionary Computation, CEC 2016, Vancouver, BC, Canada, July 24-29, 2016, Piscataway, NJ, 2016. IEEE Press.
+[ bib ] + +
+ + +
+[2776] +
+
+Proceedings of the 2017 Congress on Evolutionary Computation (CEC 2017), Piscataway, NJ, 2017. IEEE Press.
+[ bib ] + +
+ + +
+[2777] +
+
+Proceedings of the 2018 Congress on Evolutionary Computation (CEC 2018), Piscataway, NJ, 2018. IEEE Press.
+[ bib ] + +
+ + +
+[2778] +
+
+Proceedings of the 2019 Congress on Evolutionary Computation (CEC 2019), Piscataway, NJ, 2019. IEEE Press.
+[ bib ] + +
+ + +
+[2779] +
+
+Proceedings of the 2020 Congress on Evolutionary Computation (CEC 2020), Piscataway, NJ, 2020. IEEE Press.
+[ bib ] + +
+ + +
+[2780] +
+
+Proceedings of the 2021 Congress on Evolutionary Computation (CEC 2021), Piscataway, NJ, 2021. IEEE Press.
+[ bib ] + +
+ + +
+[2781] +
+
+M. L. Soffa and E. Duesterwald, editors. + Proceedings of the 6th Annual IEEE/ACM International Symposium on Code Generation and Optimization, CGO '08, New York, NY, 2008. ACM Press.
+[ bib ] + +
+ + +
+[2782] +
+
+S. Koziel and X.-S. Yang, editors. + Computational Optimization, Methods and Algorithms, volume 356 of Studies in Computational Intelligence. + Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[2783] +
+
+D. Haussler, editor. + Proceedings of the Fifth Annual ACM Conference on Computational Learning Theory, COLT 1992, Pittsburgh, PA, USA, July 27-29, 1992. ACM Press, 1992.
+[ bib ] + +
+ + +
+[2784] +
+
+M. Maher and J.-F. Puget, editors. + Principles and Practice of Constraint Programming, CP98, volume 1520 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1998.
+[ bib ] + +
+ + +
+[2785] +
+
+R. Dechter, editor. + Principles and Practice of Constraint Programming, CP 2000, 6th International Conference, Singapore, September 18-21, 2000, Proceedings, volume 1894 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[2786] +
+
+P. van Hentenryck, editor. + Principles and Practice of Constraint Programming, CP 2002. + Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[2787] +
+
+I. P. Gent, editor. + Principles and Practice of Constraint Programming – CP 2009, 15th International Conference, CP 2009, Lisbon, Portugal, September 20-24, 2009, Proceedings, volume 5732 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[2788] +
+
+C. Schulte, editor. + Principles and Practice of Constraint Programming – CP 2013, 19th International Conference, CP 2013, Uppsala, Sweden, September 16-20, 2013, Proceedings, volume 8124 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib | +DOI ] + +
+ + +
+[2789] +
+
+A. Lodi, M. Milano, and P. Toth, editors. + Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 7th International Conference, CPAIOR 2010, volume 6140 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2790] +
+
+T. Berthold, A. M. Gleixner, S. Heinz, and T. Koch, editors. + Integration of AI and OR Techniques in Contraint Programming for Combinatorial Optimization Problems – 8th International Conference, CPAIOR 2011, Berlin, Germany, May 23 – 27, 2011. Proceedings. + Lecture Notes in Computer Science. Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2791] +
+
+N. Beldiceanu, N. Jussien, and E. Pinson, editors. + Integration of AI and OR Techniques in Contraint Programming for Combinatorial Optimization Problems – 9th International Conference, CPAIOR 2012, Nantes, France, May 28 – June 1, 2012. Proceedings, volume 7298 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2792] +
+
+C. Gomes and M. Sellmann, editors. + Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 10th International Conference, CPAIOR 2013, Yorktown Heights, NY, USA, May 18-22, 2013. Proceedings, volume 7874 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2793] +
+
+M. H. Goldwasser, D. S. Johnson, and C. C. McGeoch, editors. + Data Structures, Near Neighbor Searches, and Methodology: Fifth and Sixth DIMACS Implementation Challenges, Proceedings of a DIMACS Workshop, USA, 1999, volume 59 of DIMACS Series in Discrete Mathematics and Theoretical Computer Science. + American Mathematical Society, Providence, RI, 2002.
+[ bib ] + +
+ + +
+[2794] +
+
+I. Palomares, editor. + International Alan Turing Conference on Decision Support and Recommender systems (DSRC-Turing'19), London, UK, November 21–22 2019. Alan Turing Institute.
+[ bib ] + +
+ + +
+[2795] +
+
+S. Greco, J. D. Knowles, K. Miettinen, and E. Zitzler, editors. + Learning in Multiobjective Optimization (Dagstuhl Seminar 12041), volume 2(1) of Dagstuhl Reports. + Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2012.
+[ bib | +DOI ] + +
+ + +
+[2796] +
+
+S. Greco, K. Klamroth, J. D. Knowles, and G. Rudolph, editors. + Understanding Complexity in Multiobjective Optimization (Dagstuhl Seminar 15031), volume 5(1) of Dagstuhl Reports. + Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2015.
+[ bib | +DOI ] +
+Keywords: multiple criteria decision making, evolutionary + multiobjective optimization +
+ +
+ + +
+[2797] +
+
+K. Klamroth, J. D. Knowles, G. Rudolph, and M. M. Wiecek, editors. + Personalized Multiobjective Optimization: An Analytics Perspective (Dagstuhl Seminar 18031), volume 8(1) of Dagstuhl Reports. + Schloss Dagstuhl–Leibniz-Zentrum für Informatik, Germany, 2018.
+[ bib | +DOI ] +
+Keywords: multiple criteria decision making, evolutionary + multiobjective optimization +
+ +
+ + +
+[2798] +
+
+K. F. Doerner, M. Gendreau, P. Greistorfer, W. J. Gutjahr, R. F. Hartl, and M. Reimann, editors. + Metaheuristics – Progress in Complex Systems Optimization, volume 39 of Operations Research/Computer Science Interfaces Series. + Springer, New York, NY, 2006.
+[ bib ] + +
+ + +
+[2799] +
+
+B. Doerr and F. Neumann, editors. + Theory of Evolutionary Computation. + Springer International Publishing, 2020.
+[ bib | +DOI ] + +
+ + +
+[2800] +
+
+J.-K. Hao, E. Lutton, E. M. A. Ronald, M. Schoenauer, and D. Snyers, editors. + Artificial Evolution, Third European Conference, AE'97, Nîmes, France, 22-24 October 1997, Selected Papers, volume 1363 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1998.
+[ bib | +DOI ] + +
+ + +
+[2801] +
+
+E.-G. Talbi, P. Liardet, P. Collet, E. Lutton, and M. Schoenauer, editors. + Artificial Evolution: 7th International Conference, Evolution Artificielle, EA 2005, Lille, France, volume 3871 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2802] +
+
+N. Monmarché, E.-G. Talbi, P. Collet, M. Schoenauer, and E. Lutton, editors. + Artificial Evolution, 8th International Conference, Evolution Artificielle, EA 2007, Tours, France, October 29-31, 2007 Revised Selected Papers, volume 4926 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib | +DOI ] + +
+ + +
+[2803] +
+
+P. Collet, N. Monmarché, P. Legrand, M. Schoenauer, and E. Lutton, editors. + Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009, Strasbourg, France, October 26-28, 2009. Revised Selected Papers, volume 5975 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2804] +
+
+J.-K. Hao, P. Legrand, P. Collet, N. Monmarché, E. Lutton, and M. Schoenauer, editors. + Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011, Angers, France, October 24-26, 2011. Revised Selected Papers, volume 7401 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2805] +
+
+P. Legrand et al., editors. + Artificial Evolution: 11th International Conference, Evolution Artificielle, EA 2013, Bordeaux, France, October 21-23, 2013. Revised Selected Papers, volume 8752 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2806] +
+
+S. Bonnevay et al., editors. + Artificial Evolution: 12th International Conference, Evolution Artificielle, EA 2015, Lyon, France, October 26-28, 2015. Revised Selected Papers, volume 9554 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[2807] +
+
+E. Lutton, P. Legrand, P. Parrend, N. Monmarché, and M. Schoenauer, editors. + Artificial Evolution: 13th International Conference, Évolution Artificielle, EA 2017, Paris, France, October 25-27, 2017, Revised Selected, volume 10764 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2017.
+[ bib ] + +
+ + +
+[2808] +
+
+P. Angelov et al., editors. + Evolving and Autonomous Learning Systems (EALS), 2014 IEEE Symposium on. IEEE, 2014.
+[ bib ] + +
+ + +
+[2809] +
+
+A. Jhingran et al., editors. + ACM Conference on Electronic Commerce (EC-00). + ACM Press, New York, NY, 2000.
+[ bib ] + +
+ + +
+[2810] +
+
+M. J. Kearns, R. P. McAfee, and É. Tardos, editors. + Proceedings of the fourteenth ACM Conference on Electronic Commerce, EC 2013, Philadelphia, PA, USA, June 16-20, 2013. + ACM Press, New York, NY, 2013.
+[ bib | +DOI ] + +
+ + +
+[2811] +
+
+G. Brewka, S. Coradeschi, A. Perini, and P. Traverso, editors. + Proceedings of the 17th European Conference on Artificial Intelligence, ECAI 2006, Riva del Garda, Italy, August29 - September 1, 2006. + IOS Press, 2006.
+[ bib ] + +
+ + +
+[2812] +
+
+H. Coelho, R. Studer, and M. Wooldridge, editors. + Proceedings of the 19th European Conference on Artificial Intelligence. + IOS Press, 2010.
+[ bib ] + +
+ + +
+[2813] +
+
+G. D. Giacomo, A. Catala, B. Dilkina, M. Milano, S. Barro, A. Bugarín, and J. Lang, editors. + Proceedings of the 24th European Conference on Artificial Intelligence (ECAI), volume 325 of Frontiers in Artificial Intelligence and Applications. + IOS Press, 2020.
+[ bib ] + +
+ + +
+[2814] +
+
+F. J. Varela and P. Bourgine, editors. + Proceedings of the First European Conference on Artificial Life. MIT Press, Cambridge, MA, 1992.
+[ bib ] + +
+ + +
+[2815] +
+
+J. Fürnkranz, T. Scheffer, and M. Spiliopoulou, editors. + 17th European Conference on Machine Learning, Berlin, Germany, September 18-22, 2006 Proceedings, volume 4212 of Lecture Notes in Computer Science, 2006.
+[ bib ] + +
+ + +
+[2816] +
+
+Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2015, Porto, Portugal, September 7-11, 2015, Proceedings, Part III, volume 9286 of Lecture Notes in Computer Science. Springer, 2015.
+[ bib ] + +
+ + +
+[2817] +
+
+L. Paquete, M. Chiarandini, and D. Basso, editors. + Empirical Methods for the Analysis of Algorithms, Workshop EMAA 2006, Proceedings, Reykjavik, Iceland, 2006.
+[ bib ] + +
+ + +
+[2818] +
+
+N. Lovell and L. Mainardi, editors. + 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2015, Proceedings, Annual International Conference of the IEEE Engineering in Medicine and Biology. IEEE Press, 2015.
+[ bib ] + +
+ + +
+[2819] +
+
+D. Jurafsky and E. Gaussier, editors. + Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP2006, Empirical Methods in Natural Language Processing, 2006.
+[ bib ] + +
+ + +
+[2820] +
+
+E. Zitzler, K. Deb, L. Thiele, C. A. Coello Coello, and D. Corne, editors. + Evolutionary Multi-Criterion Optimization, First International Conference, EMO 2001, Zurich, Switzerland, March 7-9, 2001, Proceedings, volume 1993 of Lecture Notes in Computer Science. + Springer, Berlin/Heidelberg, 2001.
+[ bib ] + +
+ + +
+[2821] +
+
+C. M. Fonseca, P. J. Fleming, E. Zitzler, K. Deb, and L. Thiele, editors. + Evolutionary Multi-Criterion Optimization, Second International Conference, EMO 2003, Faro, Portugal, April 2003: proceedings, volume 2632 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2822] +
+
+C. A. Coello Coello, A. Hernández Aguirre, and E. Zitzler, editors. + Evolutionary Multi-Criterion Optimization, Third International Conference, EMO 2005, Guanajuato, Mexico, March 9-11, 2005. Proceedings, volume 3410 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2823] +
+
+S. Obayashi et al., editors. + Evolutionary Multi-Criterion Optimization, 4th International Conference, EMO 2007, Matsushima, Japan, March 5-8, 2007, Proceedings, volume 4403 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2824] +
+
+M. Ehrgott, C. M. Fonseca, X. Gandibleux, J.-K. Hao, and M. Sevaux, editors. + Evolutionary Multi-Criterion Optimization. 5th International Conference, EMO 2009, volume 5467 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2825] +
+
+R. H. C. Takahashi, K. Deb, E. F. Wanner, and S. Greco, editors. + Evolutionary Multi-Criterion Optimization. 6th International Conference, EMO 2011, Ouro Preto, Brazil, April 5-8, 2011, Proceedings, volume 6576 of Lecture Notes in Computer Science. + Springer, Berlin/Heidelberg, 2011.
+[ bib ] + +
+ + +
+[2826] +
+
+R. C. Purshouse, P. J. Fleming, C. M. Fonseca, S. Greco, and J. Shaw, editors. + Evolutionary Multi-Criterion Optimization – 7th International Conference, EMO 2013, Sheffield, UK, March 19-22, 2013. Proceedings, volume 7811 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2827] +
+
+A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors. + Evolutionary Multi-Criterion Optimization – 8th International Conference, EMO 2015, Guimarães, Portugal, March 29 – April 1, 2015. Proceedings, Part I, volume 9018 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2828] +
+
+A. Gaspar-Cunha, C. H. Antunes, and C. A. Coello Coello, editors. + Evolutionary Multi-Criterion Optimization – 8th International Conference, EMO 2015, Guimarães, Portugal, March 29 – April 1, 2015. Proceedings, Part II, volume 9019 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2829] +
+
+H. Trautmann, G. Rudolph, K. Klamroth, O. Schütze, M. M. Wiecek, Y. Jin, and C. Grimme, editors. + Evolutionary Multi-Criterion Optimization – 9th International Conference, EMO 2017, Münster, Germany, March 19 - 22, 2017. Proceedings, volume 10173 of Lecture Notes in Computer Science. + Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[2830] +
+
+K. Deb, E. D. Goodman, C. A. Coello Coello, K. Klamroth, K. Miettinen, S. Mostaghim, and P. Reed, editors. + Evolutionary Multi-Criterion Optimization – 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings, volume 11411 of Lecture Notes in Computer Science. + Springer International Publishing, Cham, Switzerland, 2019.
+[ bib | +DOI ] + +
+ + +
+[2831] +
+
+D. Padua, editor. + Encyclopedia of Parallel Computing. + Springer, US, 2011.
+[ bib | +DOI ] + +
+ + +
+[2832] +
+
+J. J. Cochran, editor. + Wiley Encyclopedia of Operations Research and Management Science. + John Wiley & Sons, 2011.
+[ bib | +DOI ] + +
+ + +
+[2833] +
+
+V. W. Porto, N. Saravanan, D. E. Waagen, and A. E. Eiben, editors. + Evolutionary Programming VII, 7th International Conference, EP98, San Diego, CA, USA, March 25-27, 1998, Proceedings, volume 1447 of Lecture Notes in Computer Science. + Springer, 1998.
+[ bib ] + +
+ + +
+[2834] +
+
+Proceedings of 22th European Symposium on Artificial Neural Networks, ESANN 2014, Bruges, Belgium, April 23-25, 2014, 2014.
+[ bib | +epub ] + +
+ + +
+[2835] +
+
+Proceedings of 23rd European Symposium on Artificial Neural Networks, ESANN 2015, Bruges, Belgium, April 22-24, 2015, 2015.
+[ bib | +epub ] + +
+ + +
+[2836] +
+
+A. Viana et al., editors. + Proceedings of the EU/MEeting 2009: Debating the future: new areas of application and innovative approaches, 2009.
+[ bib ] + +
+ + +
+[2837] +
+
+K. C. Giannakoglou, D. T. Tsahalis, J. Periaux, K. D. Papaliliou, and T. Fogarty, editors. + Evolutionary Methods for Design, Optimisation and Control with Application to Industrial Problems. Proceedings of the EUROGEN 2001 Conference. + CIMNE, Barcelona, Spain, 2002.
+[ bib ] + +
+ + +
+[2838] +
+
+A. Moraglio, S. Silva, K. Krawiec, P. Machado, and C. Cotta, editors. + Genetic Programming, 15th European Conference on Genetic Programming, EuroGP 2012, Proceedings, volume 7244 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2839] +
+
+J. McDermott, M. Castelli, L. Sekanina, E. Haasdijk, and P. García-Sánchez, editors. + Genetic Programming, 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, volume 10196 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[2840] +
+
+E. Medvet, G. Pappa, and B. Xue, editors. + Genetic Programming, 25th European Conference, EuroGP 2022, Held as Part of EvoStar 2022, Madrid, Spain, April 20-22, 2022, Proceedings. + Lecture Notes in Computer Science. Springer Nature, Cham, Switzerland, 2022.
+[ bib ] + +
+ + +
+[2841] +
+
+C. D. Chio, S. Cagnoni, C. Cotta, M. Ebner, A. Ekárt, A. I. Esparcia-Alcázar, C. K. Goh, J.-J. Merelo, F. Neri, M. Preuss, J. Togelius, and G. N. Yannakakis, editors. + Applications of Evolutionary Computation, EvoApplicatons 2010: EvoCOMPLEX, EvoGAMES, EvoIASP, EvoINTELLIGENCE, EvoNUM, and EvoSTOC, Istanbul, Turkey, April 7-9, 2010, Proceedings, Part I, volume 6024 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2842] +
+
+C. Di Chio et al., editors. + EvoApplications 2012: EvoCOMNET, EvoCOMPLEX, EvoFIN, EvoGAMES, EvoHOT, EvoIASP, EvoNUM, EvoPAR, EvoRISK, EvoSTIM, and EvoSTOC, Málaga, Spain, April 11-13, 2012, Proceedings, volume 7248 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2843] +
+
+A. I. Esparcia-Alcázar and A. M. Mora, editors. + 17th European Conference, EvoApplications 2014, Granada, Spain, April 23-25, 2014, Revised Selected Papers, volume 8602 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2844] +
+
+A. M. Mora and G. Squillero, editors. + Applications of Evolutionary Computation - 18th European Conference, EvoApplications 2015, Copenhagen, Denmark, April 8 – 10, 2015, Proceedings, volume 9028 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2845] +
+
+G. Squillero and P. Burelli, editors. + Applications of Evolutionary Computation: 19th European Conference, EvoApplications 2016, Porto, Portugal, March 30 – April 1, 2016, Proceedings, Part I, volume 9597 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[2846] +
+
+G. Squillero and K. Sim, editors. + Applications of Evolutionary Computation: 20th European Conference, EvoApplications 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, Part I, volume 10199 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[2847] +
+
+P. Castillo and J. L. Jiménez Laredo, editors. + Applications of Evolutionary Computation – 24th International Conference, EvoApplications 2021, Held as Part of EvoStar 2021, Virtual Event, April 7-9, 2021, Proceedings, volume 12694 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2021.
+[ bib ] + +
+ + +
+[2848] +
+
+J. L. Jiménez Laredo et al., editors. + Applications of Evolutionary Computation – 25th European Conference, EvoApplications 2022, Held as Part of EvoStar 2022, Madrid, Spain, April 20-22, 2022, Proceedings, volume 13224 of Lecture Notes in Computer Science. + Springer Nature, Switzerland, 2022.
+[ bib ] + +
+ + +
+[2849] +
+
+J. a. Correia, S. Smith, and R. Qaddoura, editors. + Applications of Evolutionary Computation – 26th European Conference, EvoApplications 2023, Held as Part of EvoStar 2023, Brno, Czech Republic, April 12-14, 2023, Proceedings, volume 13989 of Lecture Notes in Computer Science. + Springer Nature, Switzerland, 2023.
+[ bib ] + +
+ + +
+[2850] +
+
+G. R. Raidl and J. Gottlieb, editors. + Proceedings of EvoCOP 2003 – 3rd European Conference on Evolutionary Computation in Combinatorial Optimization, volume 2611 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2851] +
+
+J. Gottlieb and G. R. Raidl, editors. + Proceedings of EvoCOP 2004 – 4th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3004 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2852] +
+
+G. R. Raidl and J. Gottlieb, editors. + Proceedings of EvoCOP 2005 – 5th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3448 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2853] +
+
+J. Gottlieb and G. R. Raidl, editors. + Proceedings of EvoCOP 2006 – 6th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 3906 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2854] +
+
+C. Cotta et al., editors. + Proceedings of EvoCOP 2007 – Seventh European Conference on Evolutionary Computation in Combinatorial Optimisation, volume 4446 of Lecture Notes in Computer Science. + Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[2855] +
+
+C. Cotta and P. Cowling, editors. + Proceedings of EvoCOP 2009 – 9th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 5482 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2856] +
+
+P. Merz and J.-K. Hao, editors. + Proceedings of EvoCOP 2011 – 11th European Conference on Evolutionary Computation in Combinatorial Optimization, volume 6622 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2857] +
+
+J.-K. Hao and M. Middendorf, editors. + Evolutionary Computation in Combinatorial Optimization – 12th European Conference, EvoCOP 2012, Málaga, Spain, April 11-13, 2012, Proceedings, volume 7245 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2858] +
+
+M. Middendorf and C. Blum, editors. + Evolutionary Computation in Combinatorial Optimization – 13th European Conference, EvoCOP 2013, Vienna, Austria, April 3-5, 2013, Proceedings, volume 7832 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2859] +
+
+C. Blum and G. Ochoa, editors. + Evolutionary Computation in Combinatorial Optimization – 14th European Conference, EvoCOP 2014, Granada, Spain, April 24-25, 2014, Proceedings, volume 8600 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2860] +
+
+B. Hu and M. López-Ibáñez, editors. + Evolutionary Computation in Combinatorial Optimization – 17th European Conference, EvoCOP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, volume 10197 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2017.
+[ bib | +DOI ] + +
+ + +
+[2861] +
+
+A. Liefooghe and M. López-Ibáñez, editors. + Evolutionary Computation in Combinatorial Optimization – 18th European Conference, EvoCOP 2018, Parma, Italy, April 4-6, 2018, Proceedings, volume 10782 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2018.
+[ bib | +DOI ] + +
+ + +
+[2862] +
+
+C. Zarges and S. Verel, editors. + Evolutionary Computation in Combinatorial Optimization – 21st European Conference, EvoCOP 2021, Held as Part of EvoStar 2021, Virtual Event, April 7-9, 2021, Proceedings, volume 12692 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2021.
+[ bib ] + +
+ + +
+[2863] +
+
+L. Pérez Cáceres and S. Verel, editors. + Evolutionary Computation in Combinatorial Optimization – 22nd European Conference, EvoCOP 2022, Held as Part of EvoStar 2022, April 20-22, 2022, Proceedings. + Lecture Notes in Computer Science. Springer, Cham, Switzerland, 2022.
+[ bib ] + +
+ + +
+[2864] +
+
+Michael T. M. Emmerich, André Deutz, Oliver Schütze, Pierrick Legrand, Emilia Tantar, and Alexandru-Adrian Tantar. + EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation VII, volume 662 of Studies in Computational Intelligence. + Springer, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[2865] +
+
+V. W. Porto, N. Saravanan, D. Waagen, and A. E. Eiben, editors. + 7th International Conference, EP98 San Diego, California, USA, March 25–27, 1998 Proceedings, volume 1447 of Lecture Notes in Computer Science, Heidelberg, Germany, 1998. Springer.
+[ bib | +DOI ] + +
+ + +
+[2866] +
+
+M. Ehrgott, J. R. Figueira, and S. Greco, editors. + Trends in Multiple Criteria Decision Analysis, volume 142 of International Series in Operations Research & Management Science. + Springer, US, 2010.
+[ bib ] + +
+ + +
+[2867] +
+
+R. Barták and K. W. Brawner, editors. + Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019. AAAI Press, 2019.
+[ bib ] + +
+ + +
+[2868] +
+
+J. Baumgartner and M. Sheeran, editors. + FMCAD'07: Proceedings of the 7th International Conference Formal Methods in Computer Aided Design, Austin, Texas, USA, 2007. IEEE Computer Society, Washington, DC, USA.
+[ bib ] + +
+ + +
+[2869] +
+
+A. Blum, editor. + 41st Annual Symposium on Foundations of Computer Science, FOCS 2000, 12-14 November 2000, Redondo Beach, California, USA. IEEE Computer Society Press, 2000.
+[ bib ] + +
+ + +
+[2870] +
+
+G. Rawlins, editor. + Foundations of Genetic Algorithms. + Morgan Kaufmann Publishers, San Mateo, CA, 1991.
+[ bib ] + +
+ + +
+[2871] +
+
+D. Whitley, editor. + Proceedings of the Second Workshop on Foundations of Genetic Algorithms. + Morgan Kaufmann Publishers, 1993.
+[ bib ] + +
+ + +
+[2872] +
+
+R. K. Belew and M. D. Vose, editors. + Proceedings of the 4th Workshop on Foundations of Genetic Algorithms, San Diego, CA, USA, August 5 1996. + Morgan Kaufmann Publishers, 1996.
+[ bib ] + +
+ + +
+[2873] +
+
+K. A. De Jong, R. Poli, and J. E. Rowe, editors. + Foundations of Genetic Algorithms, 7th International Workshop, FOGA 2002, Torremolinos, Spain, September 2-4, 2002, Proceedings. + Morgan Kaufmann Publishers, 2002.
+[ bib ] + +
+ + +
+[2874] +
+
+I. I. Garibay, T. Jansen, R. P. Wiegand, and A. S. Wu, editors. + Foundations of Genetic Algorithms, 10th ACM SIGEVO International Workshop, FOGA 2009, Orlando, Florida, USA, January 9-11, 2009, Proceedings. + ACM, 2009.
+[ bib ] + +
+ + +
+[2875] +
+
+T. Friedrich, C. Doerr, and D. V. Arnold, editors. + Foundations of Genetic Algorithms, 15th ACM/SIGEVO International Workshop, FOGA 2019, Potsdam, Germany. + ACM, 2019.
+[ bib ] + +
+ + +
+[2876] +
+
+F. Chicano, T. Friedrich, T. Kötzing, and F. Rothlauf, editors. + Foundations of Genetic Algorithms, 17th ACM/SIGEVO International Workshop, FOGA 2023, Potsdam, Germany. + ACM, 2023.
+[ bib ] + +
+ + +
+[2877] +
+
+J. R. Figueira, S. Greco, and M. Ehrgott, editors. + Multiple Criteria Decision Analysis, State of the Art Surveys. + Springer, 2005.
+[ bib ] + +
+ + +
+[2878] +
+
+J. Fürnkranz and E. Hüllermeier, editors. + Preference Learning. + Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2879] +
+
+W. Banzhaf, J. M. Daida, A. E. Eiben, M. H. Garzon, V. Honavar, M. J. Jakiela, and R. E. Smith, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 1999, 13-17 July 1999, Orlando, Florida, USA. + Morgan Kaufmann Publishers, San Francisco, CA, 1999.
+[ bib ] + +
+ + +
+[2880] +
+
+D. Whitley et al., editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2000. + Morgan Kaufmann Publishers, San Francisco, CA, 2000.
+[ bib ] + +
+ + +
+[2881] +
+
+E. D. Goodman, editor. + Proceedings of the 3rd Annual Conference on Genetic and Evolutionary Computation, GECCO 2001. + Morgan Kaufmann Publishers, San Francisco, CA, 2001.
+[ bib ] + +
+ + +
+[2882] +
+
+W. B. Langdon et al., editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2002. + Morgan Kaufmann Publishers, San Francisco, CA, 2002.
+[ bib ] + +
+ + +
+[2883] +
+
+E. Cantú-Paz et al., editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2003, Part I, volume 2723 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[2884] +
+
+K. Deb et al., editors. + Genetic and Evolutionary Computation Conference, GECCO 2004, Seattle, WA, USA, June 26-30, 2004, Proceedings, Part I, volume 3102 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2885] +
+
+K. Deb et al., editors. + Genetic and Evolutionary Computation Conference, GECCO 2004, Seattle, WA, USA, June 26-30, 2004, Proceedings, Part II, volume 3103 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[2886] +
+
+H.-G. Beyer and U.-M. O'Reilly, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2005. + ACM Press, New York, NY, 2005.
+[ bib ] + +
+ + +
+[2887] +
+
+M. Cattolico et al., editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2006. + ACM Press, New York, NY, 2006.
+[ bib ] + +
+ + +
+[2888] +
+
+D. Thierens et al., editors. + Genetic and Evolutionary Computation Conference, GECCO 2007, Proceedings, London, England, UK, July 7-11, 2007. + ACM Press, New York, NY, 2007.
+[ bib ] + +
+ + +
+[2889] +
+
+C. Ryan, editor. + Genetic and Evolutionary Computation Conference, GECCO 2008, Proceedings, Atlanta, Georgia, USA July 12-16, 2008. + ACM Press, New York, NY, 2008.
+[ bib ] + +
+ + +
+[2890] +
+
+F. Rothlauf, editor. + Genetic and Evolutionary Computation Conference, GECCO 2009, Proceedings, Montreal, Québec, Canada, July 8-12, 2009. + ACM Press, New York, NY, 2009.
+[ bib ] + +
+ + +
+[2891] +
+
+F. Rothlauf, editor. + Genetic and Evolutionary Computation Conference, GECCO 2009, Proceedings, Montreal, Québec, Canada, July 8-12, 2009, Companion Material. + ACM Press, New York, NY, 2009.
+[ bib ] + +
+ + +
+[2892] +
+
+M. Pelikan and J. Branke, editors. + Genetic and Evolutionary Computation Conference, GECCO 2010, Proceedings, Portland, Oregon, USA, July 7-11, 2010. + ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2893] +
+
+M. Pelikan and J. Branke, editors. + Genetic and Evolutionary Computation Conference, GECCO 2010, Companion Material Proceedings, Portland, Oregon, USA, July 7-11, 2010. + ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[2894] +
+
+N. Krasnogor and P. L. Lanzi, editors. + Genetic and Evolutionary Computation Conference, GECCO 2011, Proceedings, Dublin, Ireland, July 12-16, 2011. + ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2895] +
+
+N. Krasnogor and P. L. Lanzi, editors. + 13th Annual Genetic and Evolutionary Computation Conference, GECCO 2011, Companion Material Proceedings, Dublin, Ireland, July 12-16, 2011. + ACM Press, New York, NY, 2011.
+[ bib ] + +
+ + +
+[2896] +
+
+T. Soule and J. H. Moore, editors. + Genetic and Evolutionary Computation Conference, GECCO 2012, Proceedings, Philadelphia, PA, USA, July 7-11, 2012. + ACM Press, New York, NY, 2012.
+[ bib ] + +
+ + +
+[2897] +
+
+T. Soule and J. H. Moore, editors. + Genetic and Evolutionary Computation Conference, GECCO 2012, Companion Material Proceedings, Philadelphia, PA, USA, July 7-11, 2012. + ACM Press, New York, NY, 2012.
+[ bib ] + +
+ + +
+[2898] +
+
+C. Blum and E. Alba, editors. + Genetic and Evolutionary Computation Conference, GECCO 2013, Proceedings, Amsterdam, The Netherlands, July 6-10, 2013. + ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[2899] +
+
+C. Blum and E. Alba, editors. + Genetic and Evolutionary Computation Conference, GECCO 2013, Companion Material Proceedings, Amsterdam, The Netherlands, July 6-10, 2013. + ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[2900] +
+
+C. Igel and D. V. Arnold, editors. + Genetic and Evolutionary Computation Conference, GECCO 2014, Proceedings, Vancouver, BC, Canada, July 12-16, 2014. + ACM Press, New York, NY, 2014.
+[ bib ] + +
+ + +
+[2901] +
+
+S. Silva and A. I. Esparcia-Alcázar, editors. + Genetic and Evolutionary Computation Conference, GECCO 2015, Proceedings, Madrid, Spain, July 11-15, 2015. + ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[2902] +
+
+J. L. Jiménez Laredo, S. Silva, and A. I. Esparcia-Alcázar, editors. + Genetic and Evolutionary Computation Conference, GECCO 2015, Madrid, Spain, July 11-15, 2015, Companion Material Proceedings. + ACM Press, New York, NY, 2015.
+[ bib ] + +
+ + +
+[2903] +
+
+T. Friedrich, F. Neumann, and A. M. Sutton, editors. + Genetic and Evolutionary Computation Conference, GECCO 2016, Proceedings, Denver, CO, USA, July 20-24, 2016. + ACM Press, New York, NY, 2016.
+[ bib ] + +
+ + +
+[2904] +
+
+T. Friedrich, F. Neumann, and A. M. Sutton, editors. + Genetic and Evolutionary Computation Conference, GECCO 2016, Denver, CO, USA, July 20-24, 2016, Companion Material Proceedings. + ACM Press, New York, NY, 2016.
+[ bib ] + +
+ + +
+[2905] +
+
+P. A. N. Bosman, editor. + Genetic and Evolutionary Computation Conference, GECCO 2017, Berlin, Germany, July 15-19, 2017. + ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[2906] +
+
+P. A. N. Bosman, editor. + Genetic and Evolutionary Computation Conference, GECCO 2017, Berlin, Germany, July 15-19, 2017. + ACM Press, New York, NY, 2017.
+[ bib ] + +
+ + +
+[2907] +
+
+H. E. Aguirre and K. Takadama, editors. + Genetic and Evolutionary Computation Conference, GECCO 2018, Kyoto, Japan, July 15-19, 2018. + ACM Press, New York, NY, 2018.
+[ bib | +DOI ] + +
+ + +
+[2908] +
+
+M. López-Ibáñez, A. Auger, and T. Stützle, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2019, Prague, Czech Republic, July 13-17, 2019. + ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2909] +
+
+M. López-Ibáñez, A. Auger, and T. Stützle, editors. + Genetic and Evolutionary Computation Conference Companion, GECCO 2019, Prague, Czech Republic, July 13-17, 2019. + ACM Press, New York, NY, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2910] +
+
+C. A. Coello Coello, editor. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2020, Cancún, Mexico, July 8-12, 2020. + ACM Press, New York, NY, 2020.
+[ bib | +DOI | +epub ] + +
+ + +
+[2911] +
+
+F. Chicano and K. Krawiec, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2021, Lille, France, July 10-14, 2021. + ACM Press, New York, NY, 2021.
+[ bib | +DOI ] + +
+ + +
+[2912] +
+
+F. Chicano and K. Krawiec, editors. + Genetic and Evolutionary Computation Conference Companion, GECCO 2021, Lille, France, July 10-14, 2021. + ACM Press, New York, NY, 2021.
+[ bib ] + +
+ + +
+[2913] +
+
+J. E. Fieldsend and M. Wagner, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2022, Boston, Massachusetts, July 9-13, 2022. + ACM Press, New York, NY, 2022.
+[ bib | +DOI ] + +
+ + +
+[2914] +
+
+J. E. Fieldsend and M. Wagner, editors. + Genetic and Evolutionary Computation Conference Companion, GECCO 2022, Boston, Massachusetts, July 9-13, 2022. + ACM Press, New York, NY, 2022.
+[ bib | +DOI ] + +
+ + +
+[2915] +
+
+S. Silva and L. Paquete, editors. + Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2023, Lisbon, Portugal, July 15-19, 2023. + ACM Press, New York, NY, 2023.
+[ bib | +DOI ] +
+ISBN: 9798400701191 +
+ +
+ + +
+[2916] +
+
+S. Silva and L. Paquete, editors. + Genetic and Evolutionary Computation Conference Companion, GECCO 2023, Lisbon, Portugal, July 15-19, 2023. + ACM Press, New York, NY, 2023.
+[ bib | +DOI ] +
+ISBN: 979-8-4007-0120-7 +
+ +
+ + +
+[2917] +
+
+J. R. Koza, editor. + Genetic Programming 1998: Proceedings of the Third Annual Conference, Late Breaking Papers, Stanford University, California, July 1998. Stanford University Bookstore.
+[ bib ] + +
+ + +
+[2918] +
+
+R. Graves and P. Wolfe, editors. + Recent Advances in Mathematical Programming. + McGraw Hill, New York, NY, 1963.
+[ bib ] + +
+ + +
+[2919] +
+
+G. Gutin and A. Punnen, editors. + The Traveling Salesman Problem and its Variations. + Kluwer Academic Publishers, Dordrecht, The Netherlands, 2002.
+[ bib ] + +
+ + +
+[2920] +
+
+F. Almeida et al., editors. + Proceedings of HM 2006 – 3rd International Workshop on Hybrid Metaheuristics, volume 4030 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2921] +
+
+T. Bartz-Beielstein, M. J. Blesa, C. Blum, B. Naujoks, A. Roli, G. Rudolph, and M. Sampels, editors. + Hybrid Metaheuristics HM 2007, 4th International Workshop, volume 4771 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2922] +
+
+M. J. Blesa, C. Blum, C. Cotta, A. J. Fernández, J. E. Gallardo, A. Roli, and M. Sampels, editors. + Hybrid Metaheuristics HM 2008, 5th International Workshop, volume 5296 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2923] +
+
+M. J. Blesa, C. Blum, L. Di Gaspero, A. Roli, M. Sampels, and A. Schaerf, editors. + Hybrid Metaheuristics, 6th International Workshop, HM 2009, Udine, Italy, October 16-17, 2009. Proceedings, volume 5818 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2924] +
+
+M. J. Blesa, C. Blum, P. Festa, A. Roli, and M. Sampels, editors. + Hybrid Metaheuristics, 8th International Workshop, HM 2013, Ischia, Italy, May 23-25, 2013. Proceedings, volume 7919 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2925] +
+
+M. J. Blesa, C. Blum, and S. Voß, editors. + Hybrid Metaheuristics, 9th International Workshop, HM 2014, Hamburg, Germany, June 11-13, 2014. Proceedings, volume 8457 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2926] +
+
+F. Glover and G. A. Kochenberger, editors. + Handbook of Metaheuristics. + Kluwer Academic Publishers, Norwell, MA, 2002.
+[ bib ] + +
+ + +
+[2927] +
+
+F. Glover and G. A. Kochenberger, editors. + Handbook of Metaheuristics. + Springer, Boston, MA, 2003.
+[ bib | +DOI ] + +
+ + +
+[2928] +
+
+M. Gendreau and J.-Y. Potvin, editors. + Handbook of Metaheuristics, volume 146 of International Series in Operations Research & Management Science. + Springer, New York, NY, 2nd edition, 2010.
+[ bib ] + +
+ + +
+[2929] +
+
+M. Gendreau and J.-Y. Potvin, editors. + Handbook of Metaheuristics, volume 272 of International Series in Operations Research & Management Science. + Springer, 2019.
+[ bib ] + +
+ + +
+[2930] +
+
+J. Kacprzyk and W. Pedrycz, editors. + Springer Handbook of Computational Intelligence. + Springer, Berlin/Heidelberg, 2015.
+[ bib ] + +
+ + +
+[2931] +
+
+P. M. Pardalos and D.-Z. Du, editors. + Handbook of Combinatorial Optimization, volume 2. + Kluwer Academic Publishers, 1998.
+[ bib ] + +
+ + +
+[2932] +
+
+H. W. E., S. J. E., and K. N., editors. + Recent Advances in Memetic Algorithms, volume 166 of Studies in Fuzziness and Soft Computing. + Springer, Berlin/Heidelberg, 2005.
+[ bib ] + +
+ + +
+[2933] +
+
+R. Martí, P. M. Pardalos, and M. G. C. Resende, editors. + Handbook of Heuristics. + Springer International Publishing, 2018.
+[ bib ] + +
+ + +
+[2934] +
+
+D. S. Hochbaum, editor. + Approximation Algorithms For NP-hard Problems. + PWS Publishing Co., 1996.
+[ bib ] + +
+ + +
+[2935] +
+
+F. Hutter, L. Kotthoff, and J. Vanschoren, editors. + Automated Machine Learning: Methods, Systems, Challenges. + Springer, 2019.
+[ bib | +DOI | +epub ] + +
+ + +
+[2936] +
+
+H. R. Arabnia and R. Joshua, editors. + Proceedings of the 2005 International Conference on Artificial Intelligence, ICAI 2005. + CSREA Press, 2005.
+[ bib ] + +
+ + +
+[2937] +
+
+L. Caires, G. F. Italiano, L. Monteiro, C. Palamidessi, and M. Yung, editors. + Proceedings of the 32nd International Colloquium on Automata, Languages and Programming, ICALP 2005, volume 3580 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2005.
+[ bib ] + +
+ + +
+[2938] +
+
+Proceedings of the 9th International Conference on Artificial Neural Networks: ICANN '99, Location: Edinburgh, UK, 7-10 Sept. 1999, 1999.
+[ bib ] + +
+ + +
+[2939] +
+
+V. Kurkova-Pohlova and J. Koutnik, editors. + ICANN'08: Proceedings of the 18th International Conference on Artificial Neural Networks, Part I, volume 5163 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2940] +
+
+V. Kurkova-Pohlova and J. Koutnik, editors. + ICANN'08: Proceedings of the 18th International Conference on Artificial Neural Networks, Part II, volume 5164 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2941] +
+
+A. Dobnikar, N. C. Steele, D. W. Pearson, and R. F. Albrecht, editors. + Artificial Neural Nets and Genetic Algorithms (ICANNGA-99), Proceedings of the International Conference in Portorož, Slovenia, 1999. + Springer Verlag, 1999.
+[ bib | +DOI ] + +
+ + +
+[2942] +
+
+D. W. Pearson, N. C. Steele, and R. F. Albrecht, editors. + Artificial Neural Networks and Genetic Algorithms. + Springer Verlag, 2003.
+[ bib ] + +
+ + +
+[2943] +
+
+E. Karpas, S. Jiménez Celorrio, and S. Kambhampati, editors. + Proceedings of the 3rd Workshop on Learning and Planning, collocated with the 21st International Conference on Automated Planning and Scheduling (ICAPS-PAL'11), 2011.
+[ bib ] + +
+ + +
+[2944] +
+
+S. Zilberstein, J. Koehler, and S. Koenig, editors. + Proceedings of the Fourteenth International Conference on Automated Planning and Scheduling (ICAPS 2004). + AAAI Press/MIT Press, Menlo Park, CA, 2004.
+[ bib ] + +
+ + +
+[2945] +
+
+R. I. Brafman, C. Domshlak, P. Haslum, and S. Zilberstein, editors. + Proceedings of the Twenty-Fifth International Conference on Automated Planning and Scheduling, ICAPS 2015, Jerusalem, Israel, June 7-11, 2015. + AAAI Press, Menlo Park, CA, 2015.
+[ bib ] + +
+ + +
+[2946] +
+
+Z. Michalewicz, editor. + Proceedings of the First IEEE International Conference on Evolutionary Computation (ICEC'94). + IEEE Press, Piscataway, NJ, 1994.
+[ bib ] + +
+ + +
+[2947] +
+
+T. Bäck, T. Fukuda, and Z. Michalewicz, editors. + Proceedings of the 1996 IEEE International Conference on Evolutionary Computation (ICEC'96). + IEEE Press, Piscataway, NJ, 1996.
+[ bib ] + +
+ + +
+[2948] +
+
+T. Bäck, Z. Michalewicz, and X. Yao, editors. + Proceedings of the 1997 IEEE International Conference on Evolutionary Computation (ICEC'97). + IEEE Press, Piscataway, NJ, 1997.
+[ bib ] + +
+ + +
+[2949] +
+
+J. J. Grefenstette, editor. + Proceedings of the 1st International Conference on Genetic Algorithms (ICGA), Pittsburgh, PA, USA, July 1985. Lawrence Erlbaum Associates, 1985.
+[ bib ] + +
+ + +
+[2950] +
+
+J. D. Schaffer, editor. + Proceedings of the 3rd International Conference on Genetic Algorithms (ICGA), George Mason University, Fairfax, Virginia, USA, June 1989. Morgan Kaufmann Publishers, San Mateo, CA, 1989.
+[ bib ] + +
+ + +
+[2951] +
+
+S. Forrest, editor. + Proceedings of the 5th International Conference on Genetic Algorithms (ICGA), Urbana-Champaign, IL, USA, June 1993. Morgan Kaufmann Publishers, 1993.
+[ bib ] + +
+ + +
+[2952] +
+
+L. J. Eshelman, editor. + Proceedings of the 6th International Conference on Genetic Algorithms (ICGA), Pittsburgh, PA, USA, July 15-19, 1995. + Morgan Kaufmann Publishers, San Francisco, CA, Pittsburgh, PA, 1995.
+[ bib ] + +
+ + +
+[2953] +
+
+T. Bäck, editor. + Proceedings of the 7th International Conference on Genetic Algorithms (ICGA), East Lansing, MI, USA, July 19-23, 1997. + Morgan Kaufmann Publishers, San Francisco, CA, 1997.
+[ bib ] + +
+ + +
+[2954] +
+
+D.-S. Huang, K. Li, and G. W. Irwin, editors. + International Conference on Computational Science (3), volume 4115 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[2955] +
+
+Y. Shi, G. D. van Albada, J. Dongarra, and P. M. A. Sloot, editors. + Computational Science – ICCS 2007, 7th International Conference, Proceedings, Part IV, volume 4490 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[2956] +
+
+Y. Bengio and Y. LeCun, editors. + 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015.
+[ bib ] + +
+ + +
+[2957] +
+
+I. Murray, M. Ranzato, and O. Vinyals, editors. + 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings. OpenReview.net, 2018.
+[ bib ] + +
+ + +
+[2958] +
+
+W. W. Cohen and H. Hirsh, editors. + Proceedings of the 11th International Conference on Machine Learning, ICML 1994, New Brunswick, NJ, USA, San Francisco, CA, 1994. Morgan Kaufmann Publishers.
+[ bib ] + +
+ + +
+[2959] +
+
+C. E. Brodley, editor. + Machine Learning, Proceedings of the Twenty-first International Conference, ICML 2004, Banff, Alberta, Canada, July 4-8, 2004, New York, NY, 2004. ACM Press.
+[ bib ] + +
+ + +
+[2960] +
+
+W. W. Cohen, A. McCallum, and S. T. Roweis, editors. + Proceedings of the 25th International Conference on Machine Learning, ICML 2008, Helsinki, Finland, July 05-09, 2008, New York, NY, 2008. ACM Press.
+[ bib ] + +
+ + +
+[2961] +
+
+A. P. Danyluk, L. Bottou, and M. L. Littman, editors. + Proceedings of the 26th Annual International Conference on Machine Learning, ICML 2009, Montreal, Quebec, Canada, June 14-18, 2009, New York, NY, 2009. ACM Press.
+[ bib ] + +
+ + +
+[2962] +
+
+J. Fürnkranz and T. Joachims, editors. + Proceedings of the 27th international conference on machine learning, ICML 2010, New York, NY, 2010. ACM Press.
+[ bib ] + +
+ + +
+[2963] +
+
+J. Langford and J. Pineau, editors. + Proceedings of the 29th International Conference on Machine Learning, ICML 2012, Edinburgh, Scotland, UK, June 26 - July 1, 2012. Omnipress, 2012.
+[ bib ] + +
+ + +
+[2964] +
+
+S. Dasgupta and D. McAllester, editors. + Proceedings of the 30th International Conference on Machine Learning, ICML 2013, Atlanta, GA, USA, 16-21 June 2013, volume 28, 2013.
+[ bib | +http ] + +
+ + +
+[2965] +
+
+E. P. Xing and T. Jebara, editors. + Proceedings of the 31st International Conference on Machine Learning, ICML 2014, Beijing, China, 21-26 June 2014, volume 32. PMLR, 2014.
+[ bib | +http ] + +
+ + +
+[2966] +
+
+F. Bach and D. Blei, editors. + Proceedings of the 32nd International Conference on Machine Learning, ICML 2015, Lille, France, 7-9 July 2015, volume 37. PMLR, 2015.
+[ bib | +epub ] + +
+ + +
+[2967] +
+
+J. G. Dy and A. Krause, editors. + Proceedings of the 35th International Conference on Machine Learning, ICML 2018, Stockholmsmässan, Stockholm, Sweden, July 10-15, 2018, volume 80 of Proceedings of Machine Learning Research. PMLR, 2018.
+[ bib ] + +
+ + +
+[2968] +
+
+I. Cloete, K.-P. Wong, and M. Berthold, editors. + Proceedings of the 3rd International Conference on Machine Learning and Cybernetics. IEEE Press, 2004.
+[ bib ] + +
+ + +
+[2969] +
+
+Proceedings of the International Conference on Machine Learning and Cybernetics. IEEE Press, 2006.
+[ bib ] + +
+ + +
+[2970] +
+
+B. Vitoriano, E. Pinson, and F. Valente, editors. + ICORES 2014 - Proceedings of the 3rd International Conference on Operations Research and Enterprise Systems, Angers, Loire Valley, France. + SciTePress, 2014.
+[ bib ] + +
+ + +
+[2971] +
+
+K. Ito, F. Harashima, and K. Tanie, editors. + 1999 IEEE International Conference on Systems, Man, and Cybernetics, October 12–15, 1999, Tokyo, Japan. IEEE Press, 1999.
+[ bib ] + +
+ + +
+[2972] +
+
+IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013, Manchester, United Kingdom, October 13-16, 2013. IEEE Press, 2013.
+[ bib ] + +
+ + +
+[2973] +
+
+G. A. Papadopoulos, editor. + 26th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2014, Limassol, Cyprus, November 10-12, 2014. IEEE Press, 2014.
+[ bib ] + +
+ + +
+[2974] +
+
+J. Mylopoulos and R. Reiter, editors. + Proceedings of the 12th International Joint Conference on Artificial Intelligence, IJCAI 91, Sydney, Australia, August 24-30, 1991. + Morgan Kaufmann Publishers, 1995.
+[ bib ] + +
+ + +
+[2975] +
+
+C. S. Mellish, editor. + Proceedings of the 14th International Joint Conference on Artificial Intelligence, IJCAI 95, Montréal Québec, Canada, August 20-25, 1995, 2 Volumes. + Morgan Kaufmann Publishers, 1995.
+[ bib ] + +
+ + +
+[2976] +
+
+M. E. Pollack, editor. + IJCAI 1997, Proceedings of the 15th International Joint Conference on Artificial Intelligence, IJCAI 97, Nagoya, Japan, August 23-29, 1997, 2 Volumes. + Morgan Kaufmann Publishers, 1997.
+[ bib ] + +
+ + +
+[2977] +
+
+B. Nebel, editor. + IJCAI 2001, Proceedings of the 17th International Joint Conference on Artificial Intelligence. IEEE Press, 2001.
+[ bib ] + +
+ + +
+[2978] +
+
+G. Gottlob and T. Walsh, editors. + IJCAI-03, Proceedings of the Eighteenth International Joint Conference on Artificial Intelligence, Acapulco, Mexico, August 9-15, 2003. Morgan Kaufmann Publishers, 2003.
+[ bib | +epub ] + +
+ + +
+[2979] +
+
+M. M. Veloso, editor. + IJCAI 2007, Proceedings of the 20th International Joint Conference on Artificial Intelligence, Hyderabad, India, January 6-12, 2007. AAAI Press, Menlo Park, CA, 2007.
+[ bib ] + +
+ + +
+[2980] +
+
+C. Boutilier, editor. + IJCAI 2009, Proceedings of the 21st International Joint Conference on Artificial Intelligence, Pasadena, California, USA, July 11-17, 2009. AAAI Press, Menlo Park, CA, 2009.
+[ bib ] + +
+ + +
+[2981] +
+
+T. Walsh, editor. + IJCAI 2011, Proceedings of the 22nd International Joint Conference on Artificial Intelligence, Barcelona, Spain, July 16-22, 2011. IJCAI/AAAI Press, Menlo Park, CA, 2011.
+[ bib ] + +
+ + +
+[2982] +
+
+Q. Yang and M. Wooldridge, editors. + IJCAI 2015, Proceedings of the 24th International Joint Conference on Artificial Intelligence, Buenos Aires, Argentina, July 25-31, 2015. IJCAI/AAAI Press, Menlo Park, CA, 2015.
+[ bib ] + +
+ + +
+[2983] +
+
+J. Filipe and J. Kacprzyk, editors. + Proceedings of the International Joint Conference on Computational Intelligence (IJCCI-2010). SciTePress, 2010.
+[ bib ] + +
+ + +
+[2984] +
+
+Proceedings of the International Joint Conference on Neural Networks, IJCNN 2006, part of the IEEE World Congress on Computational Intelligence, WCCI 2006, Vancouver, BC, Canada, 16-21 July 2006. IEEE, 2006.
+[ bib | +DOI ] + +
+ + +
+[2985] +
+
+D. Liu et al., editors. + Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008), Hong Kong, China, June 1-6, 2008. IEEE Press, 2008.
+[ bib ] + +
+ + +
+[2986] +
+
+E. Hüllermeier, R. Kruse, and F. Hoffmann, editors. + 13th International Conference on Information Processing and Management of Uncertainty, IPMU 2010, Germany, June 28-July 2, 2010. Proceedings, volume 6178 of Lecture Notes in Artificial Intelligence. + Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[2987] +
+
+A. Abraham and M. Paprzycki, editors. + Proceedings of the 5th International Conference on Intelligent Systems Design and Applications, 2005.
+[ bib ] + +
+ + +
+[2988] +
+
+D. S. Johnson and M. A. Trick, editors. + Cliques, Coloring, and Satisfiability: Second DIMACS Implementation Challenge, volume 26 of DIMACS Series on Discrete Mathematics and Theoretical Computer Science. + American Mathematical Society, Providence, RI, 1996.
+[ bib ] + +
+ + +
+[2989] +
+
+J. Kallrath, editor. + Modeling Languages in Mathematical Optimization, volume 88 of Applied Optimization. + Kluwer Academic Publishers, 2004.
+[ bib ] + +
+ + +
+[2990] +
+
+V. Maniezzo, R. Battiti, and J.-P. Watson, editors. + Learning and Intelligent Optimization, Second International Conference, LION 2007, Trento, Italy, December 8-12, 2007. Selected Papers, volume 5313 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[2991] +
+
+T. Stützle, editor. + Third International Conference, LION 3, Trento, Italy, January 14-18, 2009. Selected Papers, volume 5851 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[2992] +
+
+C. Blum and R. Battiti, editors. + 4th International Conference, LION 4, Venice, Italy, January 18-22, 2010. Selected Papers, volume 6073 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib | +DOI ] + +
+ + +
+[2993] +
+
+C. A. Coello Coello, editor. + 5th International Conference, LION 5, Rome, Italy, January 17-21, 2011. Selected Papers, volume 6683 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2011.
+[ bib ] + +
+ + +
+[2994] +
+
+Y. Hamadi and M. Schoenauer, editors. + 6th International Conference, LION 6, Paris, France, January 16-20, 2012. Selected Papers, volume 7219 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[2995] +
+
+P. M. Pardalos and G. Nicosia, editors. + 7th International Conference, LION 7, Catania, Italy, January 7-11, 2013. Selected Papers, volume 7997 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[2996] +
+
+P. M. Pardalos, M. G. C. Resende, C. Vogiatzis, and J. L. Walteros, editors. + 8th International Conference, LION 8, Gainesville, Florida, USA, February 16-21, 2014. Revised Selected Papers, volume 8426 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[2997] +
+
+C. Dhaenens, L. Jourdan, and M.-E. Marmion, editors. + 9th International Conference, LION 9, Lille, France, January 12-15, 2015. Revised Selected Papers, volume 8994 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2015.
+[ bib ] + +
+ + +
+[2998] +
+
+P. Festa, M. Sellmann, and J. Vanschoren, editors. + 10th International Conference, LION 10, Ischia, Italy, May 29 - June 1, 2016. Revised Selected Papers, volume 10079 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[2999] +
+
+R. Battiti, D. E. Kvasov, and Y. D. Sergeyev, editors. + 11th International Conference, LION 11, Nizhny Novgorod, Russia, June 19-21, 2017, Revised Selected Papers, volume 10556 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[3000] +
+
+R. Battiti, M. Brunato, I. Kotsireas, and P. M. Pardalos, editors. + 12th International Conference, LION 12, Kalamata, Greece, June 10-15, 2018, volume 11353 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2018.
+[ bib ] + +
+ + +
+[3001] +
+
+N. F. Matsatsinis, Y. Marinakis, and P. M. Pardalos, editors. + 13th International Conference, LION 13, Chania, Crete, Greece, May 27-31, 2019, Revised Selected Papers, volume 11968 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2019.
+[ bib ] + +
+ + +
+[3002] +
+
+M. Vlastelica, J. Song, A. Ferber, B. Amos, G. Martius, B. Dilkina, and Y. Yue, editors. + Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020, LMCA 2020, Vancouver, Canada, December 12, 2020, 2020.
+[ bib ] + +
+ + +
+[3003] +
+
+P. Calabar and T. C. Son, editors. + 12th International Conference, LPNMR 2013, Corunna, Spain, September 15-19, 2013. Proceedings, volume 8148 of Lecture Notes in Artificial Intelligence. + Springer, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[3004] +
+
+F. Lobo, C. F. Lima, and Z. Michalewicz, editors. + Parameter Setting in Evolutionary Algorithms. + Springer, Berlin, Germany, 2007.
+[ bib ] + +
+ + +
+[3005] +
+
+G. H. Tzeng and P. L. Yu, editors. + Proceedings of the 10th International Conference on Multiple Criteria Decision Making (MCDM'91). + Springer Verlag, 1992.
+[ bib ] + +
+ + +
+[3006] +
+
+J. Climaco, editor. + Proceedings of the 13th International Conference on Multiple Criteria Decision Making (MCDM'97). + Springer Verlag, 1997.
+[ bib ] + +
+ + +
+[3007] +
+
+G. Fandel and T. Gal, editors. + Multiple Criteria Decision Making Theory and Application, Proceedings of the Third Conference Hagen/Königswinter, West Germany, August 20-24, 1979. + Number 177 in Lecture Notes in Economics and Mathematical Systems. Springer, Heidelberg, Germany, 1980.
+[ bib ] + +
+ + +
+[3008] +
+
+M. G. C. Resende and J. Pinho de Souza, editors. + Proceedings of MIC 1997, the 2nd Metaheuristics International Conference, Sophia-Antipolis, France, July 21-24, 1997, 1997.
+[ bib ] + +
+ + +
+[3009] +
+
+K. F. Doerner, M. Gendreau, P. Greistorfer, W. J. Gutjahr, R. F. Hartl, and M. Reimann, editors. + 6th Metaheuristics International Conference (MIC 2005), Vienna, Austria, 2005.
+[ bib ] + +
+ + +
+[3010] +
+
+M. Caserta and S. Voß, editors. + Proceedings of MIC 2009, the 8th Metaheuristics International Conference, Hamburg, Germany, 2010. University of Hamburg.
+[ bib ] + +
+ + +
+[3011] +
+
+L. Di Gaspero, A. Schaerf, and T. Stützle, editors. + Proceedings of MIC 2011, the 9th Metaheuristics International Conference, 2011.
+[ bib ] + +
+ + +
+[3012] +
+
+Proceedings of MIC 2013, the 10th Metaheuristics International Conference, 2013.
+[ bib ] + +
+ + +
+[3013] +
+
+E.-G. Talbi, editor. + Proceedings of MIC 2015, the 11th Metaheuristics International Conference, 2015.
+[ bib ] + +
+ + +
+[3014] +
+
+R. Monroy, G. Arroyo-Figueroa, L. E. Sucar, and H. Sossa, editors. + MICAI 2004: Advances in Artificial Intelligence: Third Mexican International Conference on Artificial Intelligence, Mexico City, Mexico, April 26-30, 2004. Proceedings, volume 2972 of Lecture Notes in Artificial Intelligence. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[3015] +
+
+G. Kendall, G. Vanden Berghe, and B. McCollum, editors. + Multidisciplinary International Conference on Scheduling: Theory and Applications (MISTA 2013), Gent, Belgium, 2013.
+[ bib ] + +
+ + +
+[3016] +
+
+A. Prieditis and S. Russell, editors. + Proceedings of the Twelfth International Conference on Machine Learning (ML-95). + Morgan Kaufmann Publishers, Palo Alto, CA, 1995.
+[ bib ] + +
+ + +
+[3017] +
+
+X. Gandibleux, M. Sevaux, K. Sörensen, and V. T'Kindt, editors. + Metaheuristics for Multiobjective Optimisation, volume 535 of Lecture Notes in Economics and Mathematical Systems. + Springer, Berlin/Heidelberg, 2004.
+[ bib ] + +
+ + +
+[3018] +
+
+D. Ucinski, A. C. Atkinson, and M. Patan, editors. + mODa 10 – Advances in Model-Oriented Design and Analysis, Proceedings of the 10th International Workshop in Model-Oriented Design and Analysis Held in Łagów Lubuski, Poland, June 10-14, 2013. + Springer International Publishing, Heidelberg, Germany, 2013.
+[ bib ] + +
+ + +
+[3019] +
+
+J. Branke, K. Deb, K. Miettinen, and R. Slowiński, editors. + Multiobjective Optimization: Interactive and Evolutionary Approaches, volume 5252 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[3020] +
+
+R. Caballero, F. Ruiz, and R. Steuer, editors. + Advances in Multiple Objective and Goal Programming, volume 455 of Lecture Notes in Economics and Mathematical Systems. + Springer, Heidelberg, Germany, 1997.
+[ bib ] + +
+ + +
+[3021] +
+
+J. D. Knowles, D. Corne, K. Deb, and D. R. Chair, editors. + Multiobjective Problem Solving from Nature. + Natural Computing Series. Springer, Berlin/Heidelberg, 2008.
+[ bib ] + +
+ + +
+[3022] +
+
+W. Fitzgibbon, Y. A. Kuznetsov, P. Neittaanmäki, and O. Pironneau, editors. + Modeling, Simulation and Optimization for Science and Technology, volume 34 of Computational Methods in Applied Sciences. + Springer, 2014.
+[ bib ] + +
+ + +
+[3023] +
+
+V. Maniezzo, T. Stützle, and S. Voß, editors. + Matheuristics—Hybridizing Metaheuristics and Mathematical Programming, volume 10 of Annals of Information Systems. + Springer, New York, NY, 2009.
+[ bib ] + +
+ + +
+[3024] +
+
+J. Mehnen, M. Köppen, A. Saad, and A. Tiwari, editors. + Applications of Soft Computing, volume 58 of Advances in Intelligent and Soft Computing. + Springer, Berlin/Heidelberg, 2009.
+[ bib ] + +
+ + +
+[3025] +
+
+Proceedings of the NAFIPS-FLINT International Conference'2002, Piscataway, New Jersey, June 2002. IEEE Service Center.
+[ bib ] + +
+ + +
+[3026] +
+
+N. Krasnogor, B. Melián-Batista, J. A. Moreno-Pérez, J. M. Moreno-Vega, and D. A. Pelta, editors. + Nature Inspired Cooperative Strategies for Optimization (NICSO 2008), volume 236 of Studies in Computational Intelligence. + Springer, Berlin, Germany, 2009.
+[ bib | +DOI ] + +
+ + +
+[3027] +
+
+D. Corne, M. Dorigo, and F. Glover, editors. + New Ideas in Optimization. + McGraw Hill, London, UK, 1999.
+[ bib ] + +
+ + +
+[3028] +
+
+J. D. Cowan, G. Tesauro, and J. Alspector, editors. + Advances in Neural Information Processing Systems, volume 6. + Morgan Kaufmann Publishers, San Francisco, CA, 1994.
+[ bib ] + +
+ + +
+[3029] +
+
+M. Mozer, M. I. Jordan, and T. Petsche, editors. + Advances in Neural Information Processing Systems 9, NIPS, Denver, CO, USA, December 2-5, 1996. + MIT Press, 1996.
+[ bib ] + +
+ + +
+[3030] +
+
+S. Thrun, L. Saul, and B. Schölkopf, editors. + Proceedings of the 16th International Conference on Neural Information Processing Systems, NIPS. + MIT Press, 2003.
+[ bib ] + +
+ + +
+[3031] +
+
+J. Shawe-Taylor, R. S. Zemel, P. L. Bartlett, F. Pereira, and K. Q. Weinberger, editors. + Advances in Neural Information Processing Systems 24: Annual Conference on Neural Information Processing Systems 2011. + Curran Associates, Red Hook, NY, 2011.
+[ bib ] + +
+ + +
+[3032] +
+
+P. L. Bartlett, F. C. N. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger, editors. + Advances in Neural Information Processing Systems 25: 26th Annual Conference on Neural Information Processing Systems 2012. + Curran Associates, Red Hook, NY, 2012.
+[ bib ] + +
+ + +
+[3033] +
+
+C. Cortes, N. D. Lawrence, D. D. Lee, M. Sugiyama, and R. Garnett, editors. + Advances in Neural Information Processing Systems 28: Annual Conference on Neural Information Processing Systems 2015, December 7-12, 2015, Montreal, Quebec, Canada, 2015.
+[ bib | +http ] + +
+ + +
+[3034] +
+
+D. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, and R. Garnett, editors. + Advances in Neural Information Processing Systems 29: Annual Conference on Neural Information Processing Systems 2016, December 5-10, 2016, Barcelona, Spain, 2016.
+[ bib ] + +
+ + +
+[3035] +
+
+I. Guyon, U. von Luxburg, S. Bengio, H. M. Wallach, R. Fergus, S. V. N. Vishwanathan, and R. Garnett, editors. + Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, December 4-9, 2017, Long Beach, CA, USA, 2016.
+[ bib ] + +
+ + +
+[3036] +
+
+H. M. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. B. Fox, and R. Garnett, editors. + Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019, NeurIPS 2019, 8-14 December 2019, Vancouver, BC, Canada, 2019.
+[ bib | +epub ] + +
+ + +
+[3037] +
+
+H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, and H. Lin, editors. + Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, Virtual, 2020.
+[ bib | +epub ] + +
+ + +
+[3038] +
+
+M. Ranzato, A. Beygelzimer, Y. Dauphin, P. Liang, and J. W. Vaughan, editors. + Advances in Neural Information Processing Systems 34 (NeurIPS 2021), 2021.
+[ bib | +epub ] + +
+ + +
+[3039] +
+
+K. Naono, K. Teranishi, J. Cavazos, and R. Suda, editors. + Software Automatic Tuning: From Concepts to State-of-the-Art Results. + Springer, 2010.
+[ bib ] + +
+ + +
+[3040] +
+
+F. Neri, C. Cotta, and P. Moscato, editors. + Handbook of Memetic Algorithms, volume 379 of Studies in Computational Intelligence. + Springer, 2011.
+[ bib ] + +
+ + +
+[3041] +
+
+O. Grothe, S. Nickel, S. Rebennack, and O. Stein, editors. + Operations Research 2022, Selected Papers of the Annual International Conference of the German Operations Research Society (GOR), Karlsruhe, Germany, September 6-9, 2022. + Lecture Notes in Operations Research. Springer, Cham, Switzerland, 2022.
+[ bib ] + +
+ + +
+[3042] +
+
+Proceedings of the 23rd International Conference on Parallel Architectures and Compilation, New York, NY, 2014. ACM Press.
+[ bib ] + +
+ + +
+[3043] +
+
+E. K. Burke and W. Erben, editors. + Practice and Theory of Automated Timetabling III, Third International Conference, PATAT 2000, Konstanz, Germany, August 16-18, 2000, Selected Papers, volume 2079 of Lecture Notes in Computer Science. + Springer, 2000.
+[ bib ] + +
+ + +
+[3044] +
+
+E. Özcan, E. K. Burke, and B. McCollum, editors. + PATAT 2014: Proceedings of the 10th International Conference of the Practice and Theory of Automated Timetabling. PATAT, 2014.
+[ bib ] + +
+ + +
+[3045] +
+
+F. Mueller, editor. + Proceedings of the 2011 IEEE International Parallel & Distributed Processing Symposium, IPDPS '11. IEEE Computer Society, 2011.
+[ bib ] + +
+ + +
+[3046] +
+
+H. R. Arabnia, editor. + Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA'98). + CSREA Press, 1998.
+[ bib ] + +
+ + +
+[3047] +
+
+H.-P. Schwefel and R. Männer, editors. + Parallel Problem Solving from Nature, 1st Workshop, PPSN I Dortmund, FRG, October 1-3, 1990. Proceedings. + Springer, Berlin/Heidelberg, 1991.
+[ bib | +DOI ] + +
+ + +
+[3048] +
+
+R. Männer and B. Manderick, editors. + Parallel Problem Solving from Nature 2, PPSN-II, Brussels, Belgium, September 28-30, 1992. + Elsevier, 1992.
+[ bib ] + +
+ + +
+[3049] +
+
+H.-M. Voigt et al., editors. + The 4th International Conference on Parallel Problem Solving from Nature Berlin, Germany, September 22 - 26, 1996. Proceedings, volume 1141 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1996.
+[ bib ] + +
+ + +
+[3050] +
+
+A. E. Eiben, T. Bäck, M. Schoenauer, and H.-P. Schwefel, editors. + Parallel Problem Solving from Nature – PPSN V, 5th International Conference Amsterdam, The Netherlands September 27-30, 1998 Proceedings, volume 1498 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 1998.
+[ bib ] + +
+ + +
+[3051] +
+
+M. Schoenauer et al., editors. + Parallel Problem Solving from Nature – PPSN VI, volume 1917 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[3052] +
+
+J.-J. Merelo et al., editors. + Parallel Problem Solving from Nature – PPSN VII, volume 2439 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[3053] +
+
+X. Yao et al., editors. + Proceedings of PPSN-VIII, Eighth International Conference on Parallel Problem Solving from Nature, Birmingham, UK, volume 3242 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[3054] +
+
+T. P. Runarsson, H.-G. Beyer, E. K. Burke, J.-J. Merelo, D. Whitley, and X. Yao, editors. + Proceedings of PPSN-IX, Ninth International Conference on Parallel Problem Solving from Nature, volume 4193 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2006.
+[ bib ] + +
+ + +
+[3055] +
+
+G. Rudolph et al., editors. + Proceedings of PPSN-X, Tenth International Conference on Parallel Problem Solving from Nature, volume 5199 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[3056] +
+
+R. Schaefer, C. Cotta, J. Kolodziej, and G. Rudolph, editors. + Parallel Problem Solving from Nature – PPSN XI, volume 6238 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2010.
+[ bib ] + +
+ + +
+[3057] +
+
+C. A. Coello Coello et al., editors. + Parallel Problem Solving from Nature, PPSN XII, 12th International Conference, Taormina, Italy, September 1-5, 2012, Proceedings, Part I, volume 7491 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[3058] +
+
+C. A. Coello Coello et al., editors. + Parallel Problem Solving from Nature - PPSN XII - 12th International Conference, Taormina, Italy, September 1-5, 2012, Proceedings, Part II, volume 7492 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2012.
+[ bib ] + +
+ + +
+[3059] +
+
+T. Bartz-Beielstein, J. Branke, B. Filipič, and J. Smith, editors. + Parallel Problem Solving from Nature – PPSN XIII, volume 8672 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2014.
+[ bib ] + +
+ + +
+[3060] +
+
+J. Handl, E. Hart, P. R. Lewis, M. López-Ibáñez, G. Ochoa, and B. Paechter, editors. + Parallel Problem Solving from Nature - PPSN XIV 14th International Conference, Edinburgh, UK, September 17-21, 2016, Proceedings, volume 9921 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2016.
+[ bib | +DOI ] + +
+ + +
+[3061] +
+
+A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors. + Parallel Problem Solving from Nature - PPSN XV 15th International Conference, Coimbra, Portugal, September 8-12, 2018, Proceedings, Part I, volume 11101 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2018.
+[ bib ] + +
+ + +
+[3062] +
+
+A. Auger, C. M. Fonseca, N. Lourenço, P. Machado, L. Paquete, and D. Whitley, editors. + Parallel Problem Solving from Nature - PPSN XV 15th International Conference, Coimbra, Portugal, September 8-12, 2018, Proceedings, Part II, volume 11102 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2018.
+[ bib ] + +
+ + +
+[3063] +
+
+T. Bäck, M. Preuss, A. Deutz, H. Wang, C. Doerr, M. T. M. Emmerich, and H. Trautmann, editors. + Parallel Problem Solving from Nature - PPSN XVI 16th International Conference, Leiden, The Netherlands, September 5-9, 2020, Proceedings, volume 12269 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2020.
+[ bib ] + +
+ + +
+[3064] +
+
+G. Rudolph, A. V. Kononova, H. E. Aguirre, P. Kerschke, G. Ochoa, and T. Tušar, editors. + Parallel Problem Solving from Nature - PPSN XVII, 17th International Conference, PPSN 2022, Dortmund, Germany, September 10-14, 2022, Proceedings, Part I, volume 13398 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2022.
+[ bib ] + +
+ + +
+[3065] +
+
+V. Alexandrov, M. Lees, V. Krzhizhanovskaya, J. Dongarra, and P. M. Sloot, editors. + 2013 International Conference on Computational Science, volume 18 of Procedia Computer Science. Elsevier, 2013.
+[ bib ] + +
+ + +
+[3066] +
+
+A. Albrecht and K. Steinhöfel, editors. + Second International Symposium, SAGA 2003, Hatfield, UK, September 22-23, 2003, Proceedings, volume 2827 of Lecture Notes in Computer Science. Springer Verlag, 2003.
+[ bib | +DOI ] + +
+ + +
+[3067] +
+
+F. Bacchus and T. Walsh, editors. + International Conference on Theory and Applications of Satisfiability Testing, volume 3569, 2005.
+[ bib ] + +
+ + +
+[3068] +
+
+M. Heule and S. Weaver, editors. + Theory and Applications of Satisfiability Testing – SAT 2015, volume 9340 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2015.
+[ bib ] + +
+ + +
+[3069] +
+
+A. Belov, D. Diepold, M. Heule, and M. Järvisalo, editors. + Proceedings of SAT Competition 2014: Solver and Benchmark Descriptions, volume B-2014-2 of Science Series of Publications B. University of Helsinki, 2014.
+[ bib ] + +
+ + +
+[3070] +
+
+X. Li et al., editors. + Simulated Evolution and Learning, 7th International Conference, SEAL 2008, volume 5361 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2008.
+[ bib ] + +
+ + +
+[3071] +
+
+B. K. Panigrahi, P. N. Suganthan, S. Das, and S. S. Dash, editors. + International Conference on Swarm, Evolutionary, and Memetic Computing, volume 8298 of Theoretical Computer Science and General Issues. Springer International Publishing, 2013.
+[ bib ] + +
+ + +
+[3072] +
+
+R. Ramakrishnan, S. J. Stolfo, R. J. Bayardo, and I. Parsa, editors. + Proceedings of the sixth ACM SIGKDD international conference on Knowledge discovery and data mining, Boston, MA, USA, August 20-23, 2000. + ACM Press, New York, NY, 2000.
+[ bib | +epub ] + +
+ + +
+[3073] +
+
+W. Kim, R. Kohavi, J. Gehrke, and W. DuMouchel, editors. + KDD04: ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Seattle WA USA, August 22-25, 2004. + ACM Press, New York, NY, 2004.
+[ bib ] + +
+ + +
+[3074] +
+
+I. S. Dhillon, Y. Koren, R. Ghani, T. E. Senator, P. Bradley, R. Parekh, J. He, R. L. Grossman, and R. Uthurusamy, editors. + The 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2013. + ACM Press, New York, NY, 2013.
+[ bib ] + +
+ + +
+[3075] +
+
+S. Matwin, S. Yu, and F. Farooq, editors. + KDD'17: The 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, NS, Canada, August 13-17, 2017. + ACM Press, 2017.
+[ bib ] + +
+ + +
+[3076] +
+
+Y. Guo and F. Farooq, editors. + KDD'18: The 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, London United Kingdom, August 19-23, 2018. + ACM Press, New York, NY, July 2018.
+[ bib ] + +
+ + +
+[3077] +
+
+Teredesai et al., editors. + KDD'19: The 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Anchorage, AK, USA, August 4-8, 2019. + ACM Press, New York, NY, July 2019.
+[ bib ] + +
+ + +
+[3078] +
+
+S. E. Chick, P. J. Sanchez, D. M. Ferrin, and D. J. Morrice, editors. + Proceedings of the 35th Winter Simulation Conference: Driving Innovation, volume 1, New York, NY, December 2003. ACM Press.
+[ bib ] + +
+ + +
+[3079] +
+
+T. Stützle, M. Birattari, and H. H. Hoos, editors. + Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2007, volume 4638 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2007.
+[ bib ] + +
+ + +
+[3080] +
+
+T. Stützle, M. Birattari, and H. H. Hoos, editors. + Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS 2009, volume 5752 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2009.
+[ bib ] + +
+ + +
+[3081] +
+
+X. Chen and A. Stafylopatis, editors. + Computational Intelligence (SSCI), 2016 IEEE Symposium Series on, 2016.
+[ bib ] + +
+ + +
+[3082] +
+
+C. A. Coello Coello, editor. + 2020 IEEE Symposium Series on Computational Intelligence, SSCI 2020, Canberra, Australia, December 1-4, 2020. IEEE Press, 2020.
+[ bib ] + +
+ + +
+[3083] +
+
+R. A. DeMillo, editor. + Proceedings of the sixteenth annual ACM Symposium on Theory of Computing. ACM Press, 1984.
+[ bib ] + +
+ + +
+[3084] +
+
+E. K. Burke and G. Kendall, editors. + Search Methodologies: Introductory Tutorials in Optimization and Decision Support Techniques. + Springer, Boston, MA, 2005.
+[ bib | +DOI ] + +
+ + +
+[3085] +
+
+E. Alba, F. Chicano, and G. J. Luque, editors. + Smart Cities: First International Conference, Smart-CT 2016, Málaga, Spain, June 15-17, 2016, Proceedings. + Lecture Notes in Computer Science. Springer, Cham, Switzerland, 2016.
+[ bib ] + +
+ + +
+[3086] +
+
+S. Staab and R. Studer, editors. + Handbook on Ontologies. + International Handbooks on Information Systems. Springer, 2009.
+[ bib ] + +
+ + +
+[3087] +
+
+B. Steffen and G. Woeginger, editors. + Computing and Software Science: State of the Art and Perspectives, volume 10000 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2019.
+[ bib ] + +
+ + +
+[3088] +
+
+F. Heintz, M. Milano, and B. O'Sullivan, editors. + Trustworthy AI - Integrating Learning, Optimization and Reasoning First International Workshop, TAILOR 2020, Virtual Event, September 4-5, 2020, Revised Selected Papers, volume 12641 of Lecture Notes in Computer Science. + Springer, Cham, Switzerland, 2021.
+[ bib ] + +
+ + +
+[3089] +
+
+C. Martín-Vide, R. Neruda, and M. A. Vega-Rodríguez, editors. + Theory and Practice of Natural Computing - 6th International Conference, TPNC 2017, volume 10687 of Lecture Notes in Computer Science. + Springer International Publishing, Cham, Switzerland, 2017.
+[ bib ] + +
+ + +
+[3090] +
+
+E.-G. Talbi, editor. + Hybrid Metaheuristics, volume 434 of Studies in Computational Intelligence. + Springer Verlag, 2013.
+[ bib | +http ] + +
+ + +
+[3091] +
+
+H. Topaluglu, editor. + Theory Driven by Influential Applications. + INFORMS, 2013.
+[ bib ] + +
+ + +
+[3092] +
+
+N. de Freitas and K. Murphy, editors. + Proceedings of the Twenty-Eighth Conference on Uncertainty in Artificial Intelligence (UAI'12), Catalina Island, CA August 14-18 2012. AUAI Press, 2013.
+[ bib ] + +
+ + +
+[3093] +
+
+R. V. V. Vidal, editor. + Applied Simulated Annealing. + Springer, 1993.
+[ bib ] + +
+ + +
+[3094] +
+
+S. Voß and D. L. Woodruff, editors. + Optimization Software Class Libraries. + Kluwer Academic Publishers, Boston, MA, 2002.
+[ bib ] + +
+ + +
+[3095] +
+
+Proceedings of the First IEEE Conference on Evolutionary Computation, IEEE World Congress on Computational Intelligence, Orlando, Florida, USA, June 27-29, 1994, Piscataway, NJ, June 1994. IEEE Press.
+[ bib ] + +
+ + +
+[3096] +
+
+D. B. Fogel et al., editors. + Proceedings of the 2002 World Congress on Computational Intelligence (WCCI 2002), Piscataway, NJ, 2002. IEEE Press.
+[ bib ] + +
+ + +
+[3097] +
+
+Proceedings of the 2022 World Congress on Computational Intelligence (WCCI 2022), Piscataway, NJ, 2022. IEEE Press.
+[ bib ] + +
+ + +
+[3098] +
+
+V. Y. Shen, N. Saito, M. R. Lyu, and M. E. Zurko, editors. + Proceedings of the Tenth International World Wide Web Conference, WWW 10, Hong Kong, China, May 1-5, 2001. + ACM Press, New York, NY, 2001.
+[ bib ] + +
+ + +
+[3099] +
+
+M. Rappa, P. Jones, J. Freire, and S. Chakrabarti, editors. + World Wide Web Conference, WWW 2010, Proceedings, Raleigh, North Carolina, USA, April 26-30, 2010. + ACM Press, New York, NY, 2010.
+[ bib ] + +
+ + +
+[3100] +
+
+S. Cagnoni et al., editors. + Real-World Applications of Evolutionary Computing, EvoWorkshops 2000: EvoIASP, EvoSCONDI, EvoTel, EvoSTIM, EvoROB, and EvoFlight, Edinburgh, Scotland, UK, April 17, 2000, Proceedings, volume 1803 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2000.
+[ bib ] + +
+ + +
+[3101] +
+
+E. J. W. Boers et al., editors. + Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2001, volume 2037 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2001.
+[ bib ] + +
+ + +
+[3102] +
+
+S. Cagnoni et al., editors. + Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2002, volume 2279 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2002.
+[ bib ] + +
+ + +
+[3103] +
+
+S. Cagnoni et al., editors. + Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2003, volume 2611 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2003.
+[ bib ] + +
+ + +
+[3104] +
+
+G. R. Raidl et al., editors. + Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004, volume 3005 of Lecture Notes in Computer Science. + Springer, Heidelberg, Germany, 2004.
+[ bib ] + +
+ + +
+[3105] +
+
+K. Mehlhorn, editor. + Algorithm Engineering, 2nd International Workshop, WAE'92. Max-Planck-Institut für Informatik, Saarbrücken, Germany, 1998.
+[ bib ] + +
+

This file was generated by +bibtex2html 1.99.

+ + diff --git a/index_bib.html b/index_bib.html new file mode 100644 index 00000000..f9daf700 --- /dev/null +++ b/index_bib.html @@ -0,0 +1,47702 @@ + + + + + +IRIDIA BibTeX Repository + + + + + + + + + + + +
+ + + + README + PDF (1.5 MB) + BibTeX + +License: CC0-1.0 +
+ + +

IRIDIA BibTeX Repository

+ +

What is this?

+ +

This list of references in automatically generated from a collection of BibTeX files organized in a way that tries to avoid redundancy, minimise mistakes and facilitate customization.

+ +

You only need to fork (or link) the git repository in your papers and sync with the main copy to send/receive updates.

+ +

Most customisations, such as shorter journal or conference names, do not require changing the existing .bib files. + You should not need to edit the entries directly unless you find mistakes. See the README for more details.

+ +

References

+ +

tmpnDWtSPEOFf.bib

+@comment{{This file has been generated by bib2bib 1.99}}
+
+ +
+@comment{{Command line: bib2bib --warn-error --expand --expand-xrefs authors.bib abbrev.bib journals.bib articles.bib biblio.bib crossref.bib --remove pdf -ob /tmp/tmpnDWtSPEOFf.bib -oc /tmp/citefileXQgB4Np6Iv}}
+
+ +
+@preamble{{\providecommand{\MaxMinAntSystem}{{$\cal MAX$--$\cal MIN$} {Ant} {System}} } # {\providecommand{\rpackage}[1]{{#1}} } # {\providecommand{\softwarepackage}[1]{{#1}} } # {\providecommand{\proglang}[1]{{#1}} }}
+
+ +
+@article{AbrAmoDan1999,
+  title = {Simulated annealing cooling schedules for the school timetabling problem},
+  author = { David Abramson  and Amoorthy, Mohan Krishna and Dang, Henry},
+  journal = {Asia-Pacific Journal of Operational Research},
+  volume = 16,
+  number = 1,
+  pages = {1--22},
+  year = 1999
+}
+
+ +
+@article{Abramson1991,
+  title = {Constructing School Timetables Using Simulated Annealing: Sequential and Parallel Algorithms},
+  author = { David Abramson },
+  journal = {Management Science},
+  volume = 37,
+  number = 1,
+  pages = {98--113},
+  year = 1991,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{Ach2009mpc,
+  author = {Tobias Achterberg},
+  title = {{SCIP}: {Solving} constraint integer programs},
+  journal = {Mathematical Programming Computation},
+  year = 2009,
+  volume = 1,
+  number = 1,
+  month = jul,
+  pages = {1--41},
+  epub = {http://mpc.zib.de/archive/2009/1/Achterberg2009_Article_SCIPSolvingConstraintIntegerPr.pdf}
+}
+
+ +
+@article{AchBer2007,
+  title = {Improving the feasibility pump},
+  author = {Achterberg, Tobias and Berthold, Timo},
+  journal = {Discrete Optimization},
+  volume = 4,
+  number = 1,
+  pages = {77--86},
+  year = 2007,
+  publisher = {Elsevier}
+}
+
+ +
+@article{AcoMes2014jbi,
+  author = {H{\'e}ctor-Gabriel Acosta-Mesa and Fernando Rechy-Ram{\'i}rez and  Efr{\'e}n Mezura-Montes  and Nicandro Cruz-Ram{\'i}rez and
+                  Hern{\'a}ndez Jim{\'e}nez, Rodolfo},
+  title = {Application of time series discretization using evolutionary programming for classification of precancerous cervical lesions},
+  journal = {Journal of Biomedical Informatics},
+  volume = 49,
+  pages = {73--83},
+  year = 2014,
+  doi = {10.1016/j.jbi.2014.03.004},
+  keywords = {irace}
+}
+
+ +
+@article{AddLocSch2008,
+  title = {Disk Packing in a Square: A New Global Optimization Approach},
+  author = {Addis, Bernardetta and Locatelli, Marco and Schoen, Fabio},
+  journal = {INFORMS Journal on Computing},
+  year = 2008,
+  number = 4,
+  pages = {516--524},
+  volume = 20,
+  doi = {10.1287/ijoc.1080.0263},
+  alias = {Addis2008}
+}
+
+ +
+@article{Ade92,
+  author = {B. Adenso-D{\'i}az},
+  title = {Restricted Neighborhood in the Tabu Search for the
+                  Flowshop Problem},
+  journal = {European Journal of Operational Research},
+  year = 1992,
+  volume = 62,
+  number = 1,
+  pages = {27--37}
+}
+
+ +
+@article{AdeLag06tuning,
+  author = {B. Adenso-D{\'i}az and  Manuel Laguna },
+  title = {Fine-Tuning of Algorithms Using Fractional
+                  Experimental Design and Local Search},
+  journal = {Operations Research},
+  year = 2006,
+  volume = 54,
+  number = 1,
+  pages = {99--114},
+  keywords = {Calibra}
+}
+
+ +
+@article{AdrBieSha2022jair,
+  title = {Automated dynamic algorithm configuration},
+  author = { Steven Adriaensen  and  Biedenkapp, Andr{\'e}  and Shala, Gresa and Awad, Noor and Eimer, Theresa and  Marius Thomas Lindauer  and  Frank Hutter },
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 75,
+  pages = {1633--1699},
+  year = 2022,
+  doi = {10.1613/jair.1.13922}
+}
+
+ +
+@article{AfsMieRui2021survey,
+  author = {Afsar, Bekir and  Kaisa Miettinen  and  Francisco Ruiz },
+  title = {Assessing the Performance of Interactive Multiobjective
+                  Optimization Methods: A Survey},
+  year = 2021,
+  volume = 54,
+  number = 4,
+  doi = {10.1145/3448301},
+  abstract = {Interactive methods are useful decision-making tools for
+                  multiobjective optimization problems, because they allow a
+                  decision-maker to provide her/his preference information
+                  iteratively in a comfortable way at the same time as (s)he
+                  learns about all different aspects of the problem. A wide
+                  variety of interactive methods is nowadays available, and
+                  they differ from each other in both technical aspects and
+                  type of preference information employed. Therefore, assessing
+                  the performance of interactive methods can help users to
+                  choose the most appropriate one for a given problem. This is
+                  a challenging task, which has been tackled from different
+                  perspectives in the published literature. We present a
+                  bibliographic survey of papers where interactive
+                  multiobjective optimization methods have been assessed
+                  (either individually or compared to other methods). Besides
+                  other features, we collect information about the type of
+                  decision-maker involved (utility or value functions,
+                  artificial or human decision-maker), the type of preference
+                  information provided, and aspects of interactive methods that
+                  were somehow measured. Based on the survey and on our own
+                  experiences, we identify a series of desirable properties of
+                  interactive methods that we believe should be assessed.},
+  journal = {{ACM} Computing Surveys},
+  numpages = 27,
+  keywords = {decision-makers, Interactive methods, performance assessment,
+                  preference information, multiobjective optimization problems}
+}
+
+ +
+@article{AfsSilMis2022design,
+  author = {Afsar, Bekir and Silvennoinen, Johanna and Misitano,
+                  Giovanni and  Francisco Ruiz  and Ruiz, Ana B. and  Kaisa Miettinen },
+  title = {Designing empirical experiments to compare interactive
+                  multiobjective optimization methods},
+  journal = {Journal of the Operational Research Society},
+  year = 2022,
+  volume = 74,
+  number = 11,
+  pages = {2327--2338},
+  month = nov,
+  doi = {10.1080/01605682.2022.2141145}
+}
+
+ +
+@article{AgoPea1973normality,
+  doi = {10.2307/2335012},
+  year = 1973,
+  month = dec,
+  publisher = {{JSTOR}},
+  volume = 60,
+  number = 3,
+  pages = {613--622},
+  author = {Ralph {D'Agostino} and E. S. Pearson},
+  title = {Tests for Departure from Normality. Empirical Results for the
+                  Distributions of $b_2$ and $\surd b_1$},
+  journal = {Biometrika}
+}
+
+ +
+@article{Agrell1997ejor,
+  title = {On redundancy in multi criteria decision making},
+  journal = {European Journal of Operational Research},
+  volume = 98,
+  number = 3,
+  pages = {571--586},
+  year = 1997,
+  doi = {10.1016/0377-2217(95)00340-1},
+  author = {Per J. Agrell},
+  keywords = {Multi criteria decision making, Redundancy, objective
+                  reduction, Vector optimisation},
+  abstract = {The concept of redundancy is accepted in Operations Research
+                  and Information Theory. In Linear Programming, a constraint
+                  is said to be redundant if the feasible decision space is
+                  identical with or without the constraint. In Information
+                  Theory, redundancy is used as a measure of the stability
+                  against noise in transmission. Analogies with Multi Criteria
+                  Decision Making (MCDM) are indicated and it is argued that
+                  the redundancy concept should be used as a regular feature in
+                  conditioning and analysis of Multi Criteria
+                  Programs. Properties of a proposed conflict-based
+                  characterisation are stated and some existence results are
+                  derived. Redundancy is here intended for interactive methods,
+                  when the efficient set is progressively explored. A new
+                  redundancy test for the linear case is formulated from the
+                  framework. A probabilistic method based on correlation is
+                  proposed and tested for the non-linear case. Finally, some
+                  general guidelines are given concerning the redundancy
+                  problem.}
+}
+
+ +
+@article{AguTan2007ejor,
+  title = {Working principles, behavior, and performance of {MOEAs} on
+                  {MNK}-landscapes},
+  author = { Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  journal = {European Journal of Operational Research},
+  volume = 181,
+  number = 3,
+  year = 2007,
+  pages = {1670--1690},
+  doi = {10.1016/j.ejor.2006.08.004}
+}
+
+ +
+@article{AhmOsm2004:aor,
+  author = {Samad Ahmadi and  Ibrahim H. Osman },
+  title = {Density Based Problem Space Search for the Capacitated Clustering $p$-Median Problem},
+  journal = {Annals of Operations Research},
+  year = 2004,
+  volume = 131,
+  pages = {21--43}
+}
+
+ +
+@article{AhrElsSarEss2021weighted,
+  title = {Weighted pointwise prediction method for dynamic
+                  multiobjective optimization},
+  journal = {Information Sciences},
+  volume = 546,
+  pages = {349--367},
+  year = 2021,
+  author = {Ali Ahrari and Saber Elsayed and Ruhul Sarker and Daryl
+                  Essam and  Carlos A. {Coello Coello} }
+}
+
+ +
+@article{AhuErgOrlPun2002:dam,
+  author = { R. K. Ahuja   and O. Ergun and A. P. Punnen},
+  title = {A Survey of Very Large-scale Neighborhood Search
+                  Techniques},
+  journal = {Discrete Applied Mathematics},
+  year = 2002,
+  volume = 123,
+  number = {1--3},
+  pages = {75--102}
+}
+
+ +
+@article{AinKumCha2009asc,
+  author = { Sandip Aine  and  Rajeev Kumar  and  P. P. Chakrabarti },
+  title = {Adaptive parameter control of evolutionary
+                  algorithms to improve quality-time trade-off},
+  journal = {Applied Soft Computing},
+  volume = 9,
+  number = 2,
+  year = 2009,
+  pages = {527--540},
+  doi = {10.1016/j.asoc.2008.07.001},
+  keywords = {anytime}
+}
+
+ +
+@article{AlbLanSte2010,
+  author = {Albrecht, A. A. and Lane, P. C. R. and Steinh{\"o}fel, K.},
+  title = {Analysis of Local Search Landscapes for k-{SAT} Instances},
+  journal = {Mathematics in Computer Science},
+  number = 4,
+  pages = {465--488},
+  volume = 3,
+  year = 2010,
+  doi = {10.1007/s11786-010-0040-7}
+}
+
+ +
+@article{Albers2003online,
+  title = {Online Algorithms: A Survey},
+  author = {Albers, Susanne},
+  journal = {Mathematical Programming},
+  year = 2003,
+  number = 1,
+  pages = {3--26},
+  volume = 97
+}
+
+ +
+@article{AleMos2016slr,
+  author = {Aldeida Aleti and Irene Moser},
+  year = 2016,
+  title = {A systematic literature review of adaptive parameter control
+                  methods for evolutionary algorithms},
+  journal = {{ACM} Computing Surveys},
+  volume = 49,
+  number = {3, Article 56},
+  month = oct,
+  pages = 35,
+  doi = {10.1145/2996355}
+}
+
+ +
+@article{AlfRuiPagStu2020hybrid,
+  title = {Automatic Algorithm Design for Hybrid Flowshop Scheduling
+                  Problems},
+  author = { Pedro Alfaro-Fern{\'a}ndez  and  Rub{\'e}n Ruiz  and  Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  journal = {European Journal of Operational Research},
+  volume = 282,
+  number = 3,
+  pages = {835--845},
+  year = 2020,
+  doi = {10.1016/j.ejor.2019.10.004},
+  keywords = {Scheduling, Hybrid flowshop, Automatic algorithm
+                  configuration, Automatic Algorithm Design},
+  abstract = {Industrial production scheduling problems are challenges that
+                  researchers have been trying to solve for decades. Many
+                  practical scheduling problems such as the hybrid flowshop are
+                  NP-hard. As a result, researchers resort to metaheuristics to
+                  obtain effective and efficient solutions. The traditional
+                  design process of metaheuristics is mainly manual, often
+                  metaphor-based, biased by previous experience and prone to
+                  producing overly tailored methods that only work well on the
+                  tested problems and objectives. In this paper, we use an
+                  Automatic Algorithm Design (AAD) methodology to eliminate
+                  these limitations. AAD is capable of composing algorithms
+                  from components with minimal human intervention. We test the
+                  proposed AAD for three different optimization objectives in
+                  the hybrid flowshop. Comprehensive computational and
+                  statistical testing demonstrates that automatically designed
+                  algorithms outperform specifically tailored state-of-the-art
+                  methods for the tested objectives in most cases.}
+}
+
+ +
+@article{AliMei2011kemeny,
+  annote = {Computational Foundations of Social Choice},
+  year = 2012,
+  month = jul,
+  publisher = {Elsevier {BV}},
+  volume = 64,
+  number = 1,
+  pages = {28--40},
+  author = {Alnur Ali and Marina Meil{\u{a}}},
+  doi = {10.1016/j.mathsocsci.2011.08.008},
+  journal = { Mathematical Social Science },
+  title = {Experiments with {Kemeny} ranking: What Works When?},
+  keywords = {Borda ranking, Kemeny ranking}
+}
+
+ +
+@article{AllAyd2013,
+  title = {Algorithms for no-wait flowshops with total
+                  completion time subject to makespan},
+  author = {Allahverdi, Ali and Aydilek, Harun},
+  journal = {International Journal of Advanced Manufacturing Technology},
+  pages = {1--15},
+  year = 2013
+}
+
+ +
+@article{AllJasLieTam2022cor,
+  title = {What if we increase the number of objectives? {Theoretical}
+                  and empirical implications for many-objective combinatorial
+                  optimization},
+  author = { Allmendinger, Richard  and  Andrzej Jaszkiewicz  and  Arnaud Liefooghe  and Tammer,
+                  Christiane},
+  doi = {10.1016/j.cor.2022.105857},
+  journal = {Computers \& Operations Research},
+  volume = 145,
+  pages = 105857,
+  year = 2022,
+  publisher = {Elsevier}
+}
+
+ +
+@article{AllKno2013ephemeral,
+  title = {On Handling Ephemeral Resource Constraints in Evolutionary
+                  Search},
+  author = { Allmendinger, Richard  and  Joshua D. Knowles },
+  year = 2013,
+  month = sep,
+  journal = {Evolutionary Computation},
+  volume = 21,
+  number = 3,
+  pages = {497--531},
+  issn = {1063-6560, 1530-9304},
+  doi = {10.1162/EVCO_a_00097},
+  abstract = {We consider optimization problems where the set of solutions
+                  available for evaluation at any given time t during
+                  optimization is some subset of the feasible space. This model
+                  is appropriate to describe many closed-loop optimization
+                  settings (i.e. where physical processes or experiments are
+                  used to evaluate solutions) where, due to resource
+                  limitations, it may be impossible to evaluate particular
+                  solutions at particular times (despite the solutions being
+                  part of the feasible space). We call the constraints
+                  determining which solutions are non-evaluable ephemeral
+                  resource constraints (ERCs). In this paper, we investigate
+                  two specific types of ERC: one encodes periodic resource
+                  availabilities, the other models `commitment' constraints
+                  that make the evaluable part of the space a function of
+                  earlier evaluations conducted. In an experimental study, both
+                  types of constraint are seen to impact the performance of an
+                  evolutionary algorithm significantly. To deal with the
+                  effects of the ERCs, we propose and test five different
+                  constrainthandling policies (adapted from those used to
+                  handle `standard' constraints), using a number of different
+                  test functions including a fitness landscape from a real
+                  closed-loop problem. We show that knowing information about
+                  the type of resource constraint in advance may be sufficient
+                  to select an effective policy for dealing with it, even when
+                  advance knowledge of the fitness landscape is limited.},
+  langid = {english}
+}
+
+ +
+@article{Alm10,
+  author = { Christian Almeder },
+  title = {A hybrid optimization approach for multi-level
+                  capacitated lot-sizing problems},
+  number = 2,
+  journal = {European Journal of Operational Research},
+  year = 2010,
+  keywords = {Ant colony optimization, Manufacturing, Material
+                  requirements planning, Mixed-integer programming},
+  pages = {599--606},
+  volume = 200,
+  doi = {10.1016/j.ejor.2009.01.019},
+  abstract = {Solving multi-level capacitated lot-sizing problems
+                  is still a challenging task, in spite of increasing
+                  computational power and faster algorithms. In this
+                  paper a new approach combining an ant-based
+                  algorithm with an exact solver for (mixed-integer)
+                  linear programs is presented. A {MAX-MIN} ant system
+                  is developed to determine the principal production
+                  decisions, a {LP/MIP} solver is used to calculate
+                  the corresponding production quantities and
+                  inventory levels. Two different local search methods
+                  and an improvement strategy based on reduced
+                  mixed-integer problems are developed and integrated
+                  into the ant algorithm. This hybrid approach
+                  provides superior results for small and medium-sized
+                  problems in comparison to the existing approaches in
+                  the literature. For large-scale problems the
+                  performance of this method is among the best}
+}
+
+ +
+@article{AluKat2004:ieee,
+  author = {S. Alupoaei and S. Katkoori},
+  title = {Ant Colony System Application to Marcocell Overlap Removal},
+  journal = {IEEE Transactions on Very Large Scale Integration (VLSI) Systems},
+  year = 2004,
+  volume = 12,
+  number = 10,
+  pages = {1118--1122}
+}
+
+ +
+@article{Amaral2012corridor,
+  title = {The Corridor Allocation Problem},
+  journal = {Computers \& Operations Research},
+  volume = 39,
+  number = 12,
+  pages = {3325--3330},
+  year = 2012,
+  author = {Amaral, Andr{\'e} R. S.},
+  doi = {10.1016/j.cor.2012.04.016},
+  keywords = {Facility layout, Double row layout, Integer programming},
+  abstract = {The corridor allocation problem (CAP) seeks an arrangement of
+                  facilities along a central corridor defined by two horizontal
+                  lines parallel to the x-axis of a Cartesian coordinate
+                  system. The objective is to minimize the total communication
+                  cost among facilities, while respecting two main conditions:
+                  (i) no space is allowed between two adjacent facilities; (ii)
+                  the left-most point of the arrangement on either line should
+                  have zero abscissa. The conditions (i) and (ii) are required
+                  in many applications such as the arrangement of rooms at
+                  office buildings or hospitals. The CAP is a NP-Hard
+                  problem. In this paper, a mixed-integer programming
+                  formulation of the CAP is proposed, which allows us to
+                  compute optimal layouts in reasonable time for problem
+                  instances of moderate sizes. Moreover, heuristic procedures
+                  are presented that can handle larger instances.}
+}
+
+ +
+@article{AmiBadFar2007:cis,
+  author = {Amir, C. and Badr, A. and Farag, I},
+  title = {A Fuzzy Logic Controller for Ant Algorithms},
+  journal = {Computing and Information Systems},
+  year = 2007,
+  volume = 11,
+  number = 2,
+  pages = {26--34}
+}
+
+ +
+@article{AndDefDouJor2003,
+  title = {An Introduction to {MCMC} for Machine Learning},
+  author = { Christophe Andrieu  and  Nando de Freitas  and  Arnaud Doucet  and  Michael I. Jordan },
+  journal = {Machine Learning},
+  volume = 50,
+  number = {1-2},
+  pages = {5--43},
+  year = 2003,
+  publisher = {Springer}
+}
+
+ +
+@article{AndJorLin1996,
+  author = {Andersen, K. A. and J{\"o}rnsten, K. and Lind, M.},
+  title = {On bicriterion minimal spanning trees: An approximation},
+  journal = {Computers \& Operations Research},
+  volume = 23,
+  number = 12,
+  pages = {1171--1182},
+  year = 1996
+}
+
+ +
+@article{AnejaNair79,
+  author = {Aneja, Y. P. and Nair, K. P. K.},
+  title = {Bicriteria Transportation Problem},
+  journal = {Management Science},
+  volume = 25,
+  number = 1,
+  pages = {73--78},
+  year = 1979
+}
+
+ +
+@article{AngBamGou2004tcs,
+  author = {Eric Angel and Evripidis Bampis and Laurent
+                  Gourv{\'e}s},
+  title = {Approximating the {Pareto} curve with local search
+                  for the bicriteria {TSP}(1,2) problem},
+  journal = {Theoretical Computer Science},
+  number = {1-3},
+  pages = {135--146},
+  volume = 310,
+  year = 2004,
+  doi = {10.1016/S0304-3975(03)00376-1},
+  keywords = {Archiving, Local search, Multicriteria TSP,
+                  Approximation algorithms}
+}
+
+ +
+@article{AngWoo09,
+  author = { Daniel Angus  and Clinton Woodward},
+  title = {Multiple Objective Ant Colony Optimisation},
+  journal = {Swarm Intelligence},
+  year = 2009,
+  volume = 3,
+  number = 1,
+  pages = {69--85},
+  doi = {10.1007/s11721-008-0022-4}
+}
+
+ +
+@article{AnjVie2017flp,
+  title = {Mathematical optimization approaches for facility layout
+                  problems: The state-of-the-art and future research
+                  directions},
+  author = {Anjos, Miguel F. and Vieira, Manuel V. C.},
+  journal = {European Journal of Operational Research},
+  volume = 261,
+  number = 1,
+  pages = {1--16},
+  year = 2017,
+  publisher = {Elsevier}
+}
+
+ +
+@article{AnsBriGou2002qap,
+  author = {Kurt Anstreicher and Nathan Brixius and Jean-Pierre Goux and
+                  Jeff Linderoth},
+  title = {Solving large quadratic assignment problems on computational
+                  grids},
+  doi = {10.1007/s101070100255},
+  year = 2002,
+  month = feb,
+  volume = 91,
+  number = 3,
+  pages = {563--588},
+  journal = {Mathematical Programming Series B},
+  abstract = {The quadratic assignment problem (QAP) is among the hardest
+                  combinatorial optimization problems. Some instances of size
+                  $n = 30$ have remained unsolved for decades. The solution of
+                  these problems requires both improvements in mathematical
+                  programming algorithms and the utilization of powerful
+                  computational platforms. In this article we describe a novel
+                  approach to solve QAPs using a state-of-the-art
+                  branch-and-bound algorithm running on a federation of
+                  geographically distributed resources known as a computational
+                  grid. Solution of QAPs of unprecedented complexity, including
+                  the nug30, kra30b, and tho30 instances, is reported.}
+}
+
+ +
+@article{AppBixChvCoo03:mp,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {Implementing the {Dantzig}-{Fulkerson}-{Johnson} Algorithm for Large Traveling Salesman Problems},
+  journal = {Mathematical Programming Series B},
+  year = 2003,
+  volume = 97,
+  number = {1--2},
+  pages = {91--153},
+  doi = {10.1007/s10107-003-0440-4}
+}
+
+ +
+@article{AppBixChvCoo98,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {On the Solution of Traveling Salesman Problems},
+  journal = {Documenta Mathematica},
+  year = 1998,
+  volume = {Extra Volume ICM III},
+  pages = {645--656}
+}
+
+ +
+@article{AppBlaNew1961,
+  author = {Appleby, J. S. and Blake, D. V. and Newman, E. A.},
+  title = {Techniques for producing school timetables on a computer and
+                  their application to other scheduling problems},
+  journal = {The Computer Journal},
+  year = 1961,
+  volume = 3,
+  number = 4,
+  pages = {237--245},
+  doi = {10.1093/comjnl/3.4.237}
+}
+
+ +
+@article{AppCoo91,
+  author = { David Applegate  and  William J. Cook },
+  title = {A Computational Study of the Job-Shop Scheduling
+                  Problem},
+  journal = {ORSA Journal on Computing},
+  year = 1991,
+  volume = 3,
+  number = 2,
+  pages = {149--156}
+}
+
+ +
+@article{AppCooRoh2003,
+  title = {Chained {Lin}-{Kernighan} for Large Traveling Salesman
+                  Problems},
+  author = { David Applegate  and  William J. Cook  and  Andr{\'e} Rohe},
+  journal = {INFORMS Journal on Computing},
+  volume = 15,
+  number = 1,
+  pages = {82--92},
+  year = 2003,
+  alias = {AppCooRoh99},
+  doi = {10.1287/ijoc.15.1.82.15157}
+}
+
+ +
+@article{AppEtAl09,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook  and 
+                  D. Espinoza and M. Goycoolea  and  Keld Helsgaun },
+  title = {Certification of an Optimal {TSP} Tour Through 85,900 Cities},
+  journal = {Operations Research Letters},
+  volume = 37,
+  number = 1,
+  year = 2009,
+  pages = {11--15}
+}
+
+ +
+@article{AraCamCam2022openletter,
+  author = {Claus Aranha and Camacho-Villal\'{o}n, Christian Leonardo and Felipe Campelo and  Marco Dorigo  and  Rub{\'e}n Ruiz  and  Marc Sevaux  and  Kenneth S{\"o}rensen  and  Thomas St{\"u}tzle },
+  title = {Metaphor-based Metaheuristics, a Call for Action: the Elephant in the Room},
+  journal = {Swarm Intelligence},
+  pages = {1--6},
+  volume = 16,
+  number = 1,
+  doi = {10.1007/s11721-021-00202-9},
+  year = 2022
+}
+
+ +
+@article{ArcSavSpe2016vehicle,
+  title = {The Vehicle Routing Problem with Occasional Drivers},
+  author = {Archetti, Claudia and  Martin Savelsbergh  and  Speranza, M. Grazia },
+  journal = {European Journal of Operational Research},
+  volume = 254,
+  number = 2,
+  pages = {472--480},
+  year = 2016,
+  doi = {10.1016/j.ejor.2016.03.049},
+  publisher = {Elsevier}
+}
+
+ +
+@article{ArnSanSorVid2019,
+  author = {Florian Arnold and Santana, \'{I}talo and  Kenneth S{\"o}rensen  and  Thibaut Vidal },
+  title = {{PILS}: Exploring high-order neighborhoods by pattern mining
+                  and injection},
+  journal = {Arxiv preprint arXiv:1912.11462 [cs.AI]},
+  year = 2019,
+  doi = {10.48550/arXiv.1912.11462}
+}
+
+ +
+@article{ArnSor2019knowledge,
+  author = {Florian Arnold and  Kenneth S{\"o}rensen },
+  title = {Knowledge-guided local search for the vehicle routing
+                  problem},
+  journal = {Computers \& Operations Research},
+  year = 2019,
+  volume = 105,
+  pages = {32--46},
+  doi = {10.1016/j.cor.2019.01.002}
+}
+
+ +
+@article{ArnSor2019vrp,
+  author = {Florian Arnold and  Kenneth S{\"o}rensen },
+  title = {What makes a {VRP} solution good? The generation of
+                  problem-specific knowledge for heuristics},
+  journal = {Computers \& Operations Research},
+  year = 2019,
+  volume = 106,
+  pages = {280--288},
+  doi = {10.1016/j.cor.2018.02.007}
+}
+
+ +
+@article{AroKadKhu2006,
+  title = {An empirical comparison of tabu search, simulated annealing,
+                  and genetic algorithms for facilities location problems},
+  author = {Arostegui Jr, Marvin A. and Kadipasaoglu, Sukran N. and Khumawala, Basheer M.},
+  journal = {International Journal of Production Economics},
+  volume = 103,
+  number = 2,
+  pages = {742--754},
+  year = 2006,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Arr04,
+  title = {A partial enumeration heuristic for multi-objective
+                  flowshop scheduling problems},
+  author = { Jos{\'e} Elias C. Arroyo  and V. A. Armentano},
+  journal = {Journal of the Operational Research Society},
+  volume = 55,
+  number = 9,
+  pages = {1000--1007},
+  year = 2004
+}
+
+ +
+@article{ArrArm05,
+  author = { Jos{\'e} Elias C. Arroyo  and V. A. Armentano},
+  title = {Genetic local search for multi-objective flowshop
+                  scheduling problems},
+  journal = {European Journal of Operational Research},
+  volume = 167,
+  number = 3,
+  pages = {717--738},
+  year = 2005,
+  keywords = {Multicriteria Scheduling}
+}
+
+ +
+@article{ArrLeu2017,
+  author = { Jos{\'e} Elias C. Arroyo  and  Joseph Y.-T. Leung },
+  title = {An Effective Iterated Greedy Algorithm for Scheduling Unrelated Parallel Batch Machines with Non-identical Capacities and Unequal Ready Times},
+  journal = {Computers and Industrial Engineering},
+  year = 2017,
+  volume = 105,
+  pages = {84--100}
+}
+
+ +
+@article{Asch01tsptw,
+  author = { N. Ascheuer  and  Matteo Fischetti  and  M. Gr{\"o}tschel },
+  title = {Solving asymmetric travelling salesman problem with
+                  time windows by branch-and-cut},
+  journal = {Mathematical Programming},
+  year = 2001,
+  volume = 90,
+  pages = {475--506}
+}
+
+ +
+@article{AssWanFre2014hetero,
+  author = {John{-}Alexander M. Assael and Ziyu Wang and  Nando de Freitas },
+  title = {Heteroscedastic Treed Bayesian Optimisation},
+  journal = {Arxiv preprint arXiv:1410.7172},
+  doi = {10.48550/arXiv.1410.7172},
+  year = 2014,
+  eprinttype = {arXiv},
+  eprint = {1410.7172},
+  keywords = {Treed-GP}
+}
+
+ +
+@article{Ata2003mik,
+  author = { Alper Atamt{\"u}rk },
+  title = {On the facets of the mixed--integer knapsack polyhedron},
+  journal = {Mathematical Programming},
+  year = 2003,
+  volume = 98,
+  number = 1,
+  pages = {145--175},
+  doi = {10.1007/s10107-003-0400-z}
+}
+
+ +
+@article{AudDanOrb2014,
+  author = { Charles Audet  and  Cong-Kien Dang  and  Dominique Orban },
+  title = {Optimization of Algorithms with {OPAL}},
+  journal = {Mathematical Programming Computation},
+  year = 2014,
+  volume = 6,
+  number = 3,
+  pages = {233--254}
+}
+
+ +
+@article{AudEgla1977,
+  title = {New approach to the design of multifactor experiments},
+  author = {Audze, P. and Egl{\~a}js, Vilnis},
+  journal = {Problems of Dynamics and Strengths},
+  year = 1977,
+  note = {(in Russian)},
+  pages = {104--107},
+  volume = 35,
+  publisher = {Zinatne Publishing House, Riga},
+  alias = {Audze1977}
+}
+
+ +
+@article{AudOrb06:mads,
+  author = { Charles Audet  and  Dominique Orban },
+  title = {Finding Optimal Algorithmic Parameters Using Derivative-Free
+                  Optimization},
+  journal = {SIAM Journal on Optimization},
+  year = 2006,
+  volume = 17,
+  number = 3,
+  pages = {642--664},
+  keywords = {mesh adaptive direct search; pattern search},
+  doi = {10.1137/0406208}
+}
+
+ +
+@article{Aue2002using,
+  title = {Using Confidence Bounds for Exploitation-Exploration
+                  Trade-offs},
+  author = {Auer, Peter},
+  journal = {Journal of Machine Learning Research},
+  volume = 3,
+  month = nov,
+  pages = {397--422},
+  year = 2002,
+  abstract = {We show how a standard tool from statistics --- namely
+                  confidence bounds --- can be used to elegantly deal with
+                  situations which exhibit an exploitation-exploration
+                  trade-off. Our technique for designing and analyzing
+                  algorithms for such situations is general and can be applied
+                  when an algorithm has to make exploitation-versus-exploration
+                  decisions based on uncertain information provided by a random
+                  process.  We apply our technique to two models with such an
+                  exploitation-exploration trade-off. For the adversarial
+                  bandit problem with shifting our new algorithm suffers only
+                  $O((ST)^{1/2})$ regret with high probability over $T$ trials
+                  with $S$ shifts. Such a regret bound was previously known
+                  only in expectation. The second model we consider is
+                  associative reinforcement learning with linear value
+                  functions. For this model our technique improves the regret
+                  from $O(T^{3/4})$ to $O(T^{1/2})$.}
+}
+
+ +
+@article{AueCesFis2002finite,
+  title = {Finite-time analysis of the multiarmed bandit problem},
+  author = {Auer, Peter and Cesa-Bianchi, Nicolo and Fischer, Paul},
+  journal = {Machine Learning},
+  volume = 47,
+  number = {2-3},
+  pages = {235--256},
+  year = 2002
+}
+
+ +
+@article{AugBadBroZit2012tcs,
+  author = { Anne Auger  and  Johannes Bader  and  Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Hypervolume-based multiobjective optimization:
+                  Theoretical foundations and practical implications},
+  journal = {Theoretical Computer Science},
+  volume = 425,
+  year = 2012,
+  pages = {75--103},
+  doi = {10.1016/j.tcs.2011.03.012}
+}
+
+ +
+@article{AvcTop2017:cor,
+  author = {Mustafa Avci and Seyda Topaloglu},
+  title = {A Multi-start Iterated Local Search Algorithm for the Generalized Quadratic Multiple Knapsack Problem},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  volume = 83,
+  pages = {54--65}
+}
+
+ +
+@article{AvrAllLop2021arxiv,
+  author = { Andreea Avramescu  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Managing Manufacturing and Delivery of Personalised Medicine:
+                  Current and Future Models},
+  year = 2021,
+  journal = {Arxiv preprint arXiv:2105.12699 [econ.GN]},
+  url = {https://arxiv.org/abs/2105.12699}
+}
+
+ +
+@article{AydYavStu2017:si,
+  author = { Do\v{g}an Ayd{\i}n  and  G{\"{u}}rcan Yavuz  and  Thomas St{\"u}tzle },
+  title = {{ABC-X:} A Generalized, Automatically Configurable Artificial Bee
+               Colony Framework},
+  journal = {Swarm Intelligence},
+  year = 2017,
+  volume = 11,
+  number = 1,
+  pages = {1--38}
+}
+
+ +
+@article{AyoAllLopPar2022scalarisation,
+  author = { Ayodele, Mayowa  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Parizy, Matthieu },
+  title = {A Study of Scalarisation Techniques for Multi-Objective
+                  {QUBO} Solving},
+  journal = {Arxiv preprint arXiv:2210.11321},
+  year = 2022,
+  doi = {10.48550/arXiv.2210.11321}
+}
+
+ +
+@article{AziTay2014eaai,
+  author = {Mahdi Aziz and {Tayarani-N}, Mohammad-H.},
+  title = {An adaptive memetic Particle Swarm Optimization algorithm for finding large-scale Latin hypercube designs},
+  journal = {Engineering Applications of Artificial Intelligence},
+  volume = 36,
+  pages = {222--237},
+  year = 2014,
+  doi = {10.1016/j.engappai.2014.07.021},
+  keywords = {F-race}
+}
+
+ +
+@article{BacHelPic2020gaussian,
+  title = {Gaussian process optimization with failures: Classification
+                  and convergence proof},
+  author = {Bachoc, Fran{\c c}ois and Helbert, C{\'e}line and Picheny,
+                  Victor},
+  journal = {Journal of Global Optimization},
+  year = 2020,
+  epub = {https://hal.archives-ouvertes.fr/hal-02100819/file/optimwithfailurerevised_hal.pdf},
+  keywords = {crashed simulation; latent gaussian process; automotive fan
+                  design; industrial application; GP classification; Expected
+                  Feasible Improvement with Gaussian Process Classification
+                  with signs; EFI GPC sign},
+  doi = {10.1007/s10898-020-00920-0},
+  abstract = {We consider the optimization of a computer model where each
+                  simulation either fails or returns a valid output
+                  performance. We first propose a new joint Gaussian process
+                  model for classification of the inputs (computation failure
+                  or success) and for regression of the performance
+                  function. We provide results that allow for a computationally
+                  efficient maximum likelihood estimation of the covariance
+                  parameters, with a stochastic approximation of the likelihood
+                  gradient. We then extend the classical improvement criterion
+                  to our setting of joint classification and regression. We
+                  provide an efficient computation procedure for the extended
+                  criterion and its gradient. We prove the almost sure
+                  convergence of the global optimization algorithm following
+                  from this extended criterion. We also study the practical
+                  performances of this algorithm, both on simulated data and on
+                  a real computer model in the context of automotive fan
+                  design.}
+}
+
+ +
+@article{BadZit2011ec,
+  author = { Johannes Bader  and  Eckart Zitzler },
+  title = {{HypE}: An Algorithm for Fast Hypervolume-Based
+                  Many-Objective Optimization},
+  journal = {Evolutionary Computation},
+  volume = 19,
+  number = 1,
+  year = 2011,
+  pages = {45--76},
+  doi = {10.1162/EVCO_a_00009}
+}
+
+ +
+@article{BahComLau2019tre,
+  title = {Bi-objective multi-layer location--\hspace{0pt}allocation model for the
+                  immediate aftermath of sudden-onset disasters},
+  author = {Baharmand, Hossein and Comes, Tina and Lauras, Matthieu},
+  journal = {Transportation Research Part E: Logistics and Transportation Review},
+  volume = 127,
+  pages = {86--110},
+  year = 2019,
+  doi = {10.1016/j.tre.2019.05.002},
+  abstract = {Locating distribution centers is critical for humanitarians
+                  in the immediate aftermath of a sudden-onset disaster. A
+                  major challenge lies in balancing the complexity and
+                  uncertainty of the problem with time and resource
+                  constraints. To address this problem, we propose a
+                  location-allocation model that divides the topography of
+                  affected areas into multiple layers; considers constrained
+                  number and capacity of facilities and fleets; and allows
+                  decision-makers to explore trade-offs between response time
+                  and logistics costs. To illustrate our theoretical work, we
+                  apply the model to a real dataset from the 2015 Nepal
+                  earthquake response. For this case, our method results in a
+                  considerable reduction of logistics costs.}
+}
+
+ +
+@article{Baker2016reprod,
+  title = {Is there a reproducibility crisis?},
+  author = {Monya Baker},
+  journal = {Nature},
+  volume = 533,
+  pages = {452--454},
+  year = 2016
+}
+
+ +
+@article{Baker83:tsptw,
+  author = {Edward K. Baker},
+  title = {An Exact Algorithm for the Time-Constrained
+                  Traveling Salesman Problem},
+  volume = 31,
+  doi = {10.1287/opre.31.5.938},
+  number = 5,
+  journal = {Operations Research},
+  year = 1983,
+  pages = {938--945},
+  anote = {makespan optimization}
+}
+
+ +
+@article{BalBea2008,
+  title = {Facility location in humanitarian relief},
+  author = {Balcik, Burcu and Beamon, Benita M.},
+  journal = {International Journal of Logistics},
+  volume = 11,
+  number = 2,
+  pages = {101--121},
+  year = 2008,
+  publisher = {Taylor \& Francis}
+}
+
+ +
+@article{BalBirStuDor2009ejor,
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Adaptive Sampling Size and Importance Sampling in Estimation-based
+Local Search for the Probabilistic Traveling Salesman Problem},
+  journal = {European Journal of Operational Research},
+  year = 2009,
+  volume = 199,
+  number = 1,
+  pages = {98--110}
+}
+
+ +
+@article{BalBirStuDor2010cor,
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Estimation-based Metaheuristics for the Probabilistic Travelling Salesman Problem},
+  journal = {Computers \& Operations Research},
+  year = 2010,
+  volume = 37,
+  number = 11,
+  pages = {1939--1951},
+  doi = {10.1016/j.cor.2009.12.005}
+}
+
+ +
+@article{BalBirStuDor2015coa,
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Estimation-based Metaheuristics for the Single Vehicle Routing Problem with Stochastic Demands and Customers},
+  journal = {Computational Optimization and Applications},
+  year = 2015,
+  volume = 61,
+  number = 2,
+  pages = {463--487},
+  doi = {10.1007/s10589-014-9719-z}
+}
+
+ +
+@article{BalBirStuYuaDor09,
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle  and  Zhi Yuan  and  Marco Dorigo },
+  title = {Estimation-based Ant Colony Optimization Algorithms
+                  for the Probabilistic Travelling Salesman Problem},
+  journal = {Swarm Intelligence},
+  volume = 3,
+  number = 3,
+  year = 2009,
+  pages = {223--242}
+}
+
+ +
+@article{BalCar1996,
+  author = { Egon Balas  and M. C. Carrera},
+  title = {A Dynamic Subgradient-based Branch and Bound
+                  Procedure for Set Covering},
+  journal = {Operations Research},
+  year = 1996,
+  volume = 44,
+  number = 6,
+  pages = {875--890}
+}
+
+ +
+@article{BalMar1980,
+  author = { Egon Balas  and C. Martin},
+  title = {Pivot and Complement--A Heuristic for 0--1 Programming},
+  journal = {Management Science},
+  year = 1980,
+  volume = 26,
+  number = 1,
+  pages = {86--96}
+}
+
+ +
+@article{BalPad1976,
+  author = { Egon Balas  and M. W. Padberg},
+  title = {Set Partitioning: A Survey},
+  journal = {SIAM Review},
+  year = 1976,
+  volume = 18,
+  pages = {710--760}
+}
+
+ +
+@article{BalSim01tsptw,
+  author = { Egon Balas  and  Neil Simonetti },
+  title = {Linear Time Dynamic-Programming Algorithms for New
+                  Classes of Restricted {TSP}s: {A} Computational Study},
+  volume = 13,
+  doi = {10.1287/ijoc.13.1.56.9748},
+  abstract = {Consider the following restricted (symmetric or
+                  asymmetric) traveling-salesman problem {(TSP):}
+                  given an initial ordering of the n cities and an
+                  integer $k > 0$, find a minimum-cost
+                  feasible tour, where a feasible tour is one in which
+                  city $i$ precedes city $j$ whenever $j >= i + k$ in the
+                  initial ordering. Balas (1996) has proposed a
+                  dynamic-programming algorithm that solves this
+                  problem in time linear in n, though exponential in
+                  k. Some important real-world problems are amenable
+                  to this model or some of its close relatives. The
+                  algorithm of Balas (1996) constructs a layered
+                  network with a layer of nodes for each position in
+                  the tour, such that source-sink paths in this
+                  network are in one-to-one correspondence with tours
+                  that satisfy the postulated precedence
+                  constraints. In this paper we discuss an
+                  implementation of the dynamic-programming algorithm
+                  for the general case when the integer k is replaced
+                  with city-specific integers k(j), j = 1, . . .,
+                  n. We discuss applications to, and computational
+                  experience with, {TSPs} with time windows, a model
+                  frequently used in vehicle routing as well as in
+                  scheduling with setup, release and delivery
+                  times. We also introduce a new model, the {TSP} with
+                  target times, applicable to {Just-in-Time}
+                  scheduling problems. Finally for {TSPs} that have no
+                  precedence restrictions, we use the algorithm as a
+                  heuristic that finds in linear time a local optimum
+                  over an exponential-size neighborhood. For this
+                  case, we implement an iterated version of our
+                  procedure, based on contracting some arcs of the
+                  tour produced by a first application of the
+                  algorithm, then reapplying the algorithm to the
+                  shrunken graph with the same k.},
+  number = 1,
+  journal = {INFORMS Journal on Computing},
+  year = 2001,
+  keywords = {tsptw},
+  pages = {56--75}
+}
+
+ +
+@article{BalVaz1998,
+  author = { Egon Balas  and A. Vazacopoulos},
+  title = {Guided Local Search with Shifting Bottleneck for Job
+                  Shop Scheduling},
+  journal = {Management Science},
+  year = 1998,
+  volume = 44,
+  number = 2,
+  pages = {262--275}
+}
+
+ +
+@article{Bankes2002,
+  title = {Tools and techniques for developing policies for complex and
+                  uncertain systems},
+  author = { Bankes, Steven C. },
+  volume = 99,
+  number = {suppl 3},
+  pages = {7263--7266},
+  year = 2002,
+  abstract = {Agent-based models (ABM) are examples of complex adaptive
+                  systems, which can be characterized as those systems for
+                  which no model less complex than the system itself can
+                  accurately predict in detail how the system will behave at
+                  future times. Consequently, the standard tools of policy
+                  analysis, based as they are on devising policies that perform
+                  well on some best estimate model of the system, cannot be
+                  reliably used for ABM. This paper argues that policy analysis
+                  by using ABM requires an alternative approach to decision
+                  theory. The general characteristics of such an approach are
+                  described, and examples are provided of its application to
+                  policy analysis.ABM, agent-based model},
+  journal = {Proceedings of the National Academy of Sciences},
+  doi = {10.1073/pnas.092081399}
+}
+
+ +
+@article{BarBatSenSil2015ijem,
+  title = {Improving the Performance of Metaheuristics: An Approach
+                  Combining Response Surface Methodology and Racing Algorithms},
+  author = {Eduardo Batista de Moraes Barbosa and Edson Luiz
+                  Franc{\c{c}}a Senne and Messias Borges Silva},
+  journal = {International Journal of Engineering Mathematics},
+  year = 2015,
+  volume = 2015,
+  pages = {Article ID 167031},
+  doi = {10.1155/2015/167031},
+  keywords = {F-race}
+}
+
+ +
+@article{BarDiaSerBen2020xai,
+  doi = {10.1016/j.inffus.2019.12.012},
+  year = 2020,
+  month = jun,
+  publisher = {Elsevier {BV}},
+  volume = 58,
+  pages = {82--115},
+  author = {Alejandro Barredo Arrieta and Natalia
+                  D{\'{i}}az-Rodr{\'{i}}guez and Javier Del Ser and Adrien
+                  Bennetot and Siham Tabik and Alberto Barbado and Salvador
+                  Garcia and Sergio Gil-Lopez and Daniel Molina and Richard
+                  Benjamins and Raja Chatila and Francisco Herrera},
+  title = {Explainable Artificial Intelligence ({XAI}): Concepts,
+                  taxonomies, opportunities and challenges toward responsible
+                  {AI}},
+  journal = {Information Fusion}
+}
+
+ +
+@article{BarDoeBer2020benchmarking,
+  title = {Benchmarking in Optimization: Best Practice and Open Issues},
+  author = { Thomas Bartz-Beielstein  and  Carola Doerr  and Daan van den Berg and  Jakob Bossek  and Sowmya Chandrasekaran and  Tome Eftimov  and Andreas Fischbach and  Pascal Kerschke  and William {La
+                  Cava} and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Katherine M. Malan and Jason H. Moore and  Boris Naujoks  and Patryk Orzechowski and Vanessa Volz and  Markus Wagner  and Thomas Weise},
+  year = 2020,
+  journal = {Arxiv preprint arXiv:2007.03488 [cs.NE]},
+  url = {https://arxiv.org/abs/2007.03488}
+}
+
+ +
+@article{BarGolKel1995,
+  author = {Richard S. Barr and Bruce L. Golden and James P. Kelly and  Mauricio G. C. Resende  and William R. Stewart, Jr.},
+  title = {Designing and Reporting on Computational Experiments with
+                  Heuristic Methods},
+  journal = {Journal of Heuristics},
+  year = 1995,
+  volume = 1,
+  number = 1,
+  pages = {9--32},
+  doi = {10.1007/BF02430363}
+}
+
+ +
+@article{BarJohNem1998or,
+  title = {Branch-and-price: Column generation for solving huge integer
+                  programs},
+  author = {Barnhart, Cynthia and Johnson, Ellis L. and Nemhauser, George
+                  L. and  Martin W. P. Savelsbergh  and Vance, Pamela H.},
+  journal = {Operations Research},
+  volume = 46,
+  number = 3,
+  pages = {316--329},
+  year = 1998
+}
+
+ +
+@article{BarKwa2020,
+  title = {On considering robustness in the search phase of Robust
+                  Decision Making: A comparison of Many-Objective Robust
+                  Decision Making, multi-scenario Many-Objective Robust
+                  Decision Making, and Many Objective Robust Optimization},
+  author = {Bartholomew, Erin and Kwakkel, Jan H.},
+  journal = {Environmental Modelling \& Software},
+  pages = 104699,
+  year = 2020,
+  publisher = {Elsevier},
+  volume = 127,
+  doi = {10.1016/j.envsoft.2020.104699}
+}
+
+ +
+@article{BarPea2016,
+  author = { Elias Bareinboim  and  Judea Pearl },
+  title = {Causal inference and the data-fusion problem},
+  volume = 113,
+  number = 27,
+  pages = {7345--7352},
+  year = 2016,
+  doi = {10.1073/pnas.1510507113},
+  journal = {Proceedings of the National Academy of Sciences}
+}
+
+ +
+@article{BarZae2017model,
+  author = { Thomas Bartz-Beielstein  and  Martin Zaefferer },
+  title = {Model-based methods for continuous and discrete global
+                  optimization},
+  doi = {10.1016/j.asoc.2017.01.039},
+  year = 2017,
+  month = jun,
+  volume = 55,
+  pages = {154--167},
+  journal = {Applied Soft Computing}
+}
+
+ +
+@article{BasFra1990,
+  author = {Basu, Atanu and Frazer, L. Neil},
+  title = {Rapid Determination of the Critical Temperature in Simulated Annealing Inversion},
+  journal = {Science},
+  year = 1990,
+  volume = 249,
+  number = 4975,
+  pages = {1409--1412}
+}
+
+ +
+@article{BatPas2010tec,
+  author = { Roberto Battiti  and  Andrea Passerini },
+  title = {Brain-Computer Evolutionary Multiobjective Optimization: A
+                  Genetic Algorithm Adapting to the Decision Maker},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 14,
+  number = 5,
+  year = 2010,
+  pages = {671--687},
+  doi = {10.1109/TEVC.2010.2058118},
+  keywords = {BC-EMOA},
+  annote = {Errata: DTLZ6 and DTLZ7 in the paper are actually DTLZ7 and
+                  DTLZ8 in \cite{DebThiLau2005dtlz}}
+}
+
+ +
+@article{BatPro1996:jea,
+  author = { Roberto Battiti  and M. Protasi},
+  title = {Reactive Search, A History-Based Heuristic for
+                  {MAX}-{SAT}},
+  journal = {ACM Journal of Experimental Algorithmics},
+  year = 1997,
+  volume = 2
+}
+
+ +
+@article{BatSchUrl2017,
+  author = {Michele Battistutta and Andrea Schaerf and  Tommaso Urli },
+  title = {Feature-based Tuning of Single-stage Simulated Annealing for
+                  Examination Timetabling},
+  journal = {Annals of Operations Research},
+  year = 2017,
+  volume = 252,
+  number = 2,
+  pages = {239--254}
+}
+
+ +
+@article{BatTec1994:cma,
+  title = {Simulated annealing and Tabu search in the long run: A
+                  comparison on {QAP} tasks},
+  author = { Roberto Battiti  and  Tecchiolli, Giampietro },
+  journal = {Computer and Mathematics with Applications},
+  volume = 28,
+  number = 6,
+  pages = {1--8},
+  year = 1994,
+  publisher = {Elsevier},
+  doi = {10.1016/0898-1221(94)00147-2}
+}
+
+ +
+@article{BatTec1994:orsa,
+  author = { Roberto Battiti  and  Tecchiolli, Giampietro },
+  title = {The Reactive Tabu Search},
+  journal = {ORSA Journal on Computing},
+  year = 1994,
+  volume = 6,
+  number = 2,
+  pages = {126--140}
+}
+
+ +
+@article{BatTec1996aor,
+  author = { Roberto Battiti  and  Tecchiolli, Giampietro },
+  title = {The continuous reactive tabu search: blending combinatorial
+                  optimization and stochastic search for global optimization},
+  journal = {Annals of Operations Research},
+  year = 1996,
+  volume = 63,
+  number = 2,
+  pages = {151--188},
+  doi = {10.1007/BF02125453}
+}
+
+ +
+@article{BauPer2007:ejor,
+  author = { J. Bautista  and  J. Pereira },
+  title = {Ant algorithms for a time and space constrained
+                  assembly line balancing problem},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 177,
+  number = 3,
+  pages = {2016--2032},
+  doi = {10.1016/j.ejor.2005.12.017}
+}
+
+ +
+@article{Baumol1962,
+  author = {Baumol, William J.},
+  title = {Management models and industrial applications of linear
+                  programming},
+  journal = {Naval Research Logistics Quarterly},
+  volume = 9,
+  number = 1,
+  doi = {10.1002/nav.3800090109},
+  pages = {63--64},
+  year = 1962
+}
+
+ +
+@article{Baxter1981,
+  author = {John Baxter},
+  title = {Local Optima Avoidance in Depot Location},
+  journal = {Journal of the Operational Research Society},
+  year = 1981,
+  volume = 32,
+  number = 9,
+  pages = {815--819}
+}
+
+ +
+@article{BeaChu1996,
+  author = { John E. Beasley  and P. C. Chu},
+  title = {A Genetic Algorithm for the Set Covering Problem},
+  journal = {European Journal of Operational Research},
+  year = 1996,
+  volume = 94,
+  number = 2,
+  pages = {392--404}
+}
+
+ +
+@article{BeaChu1998,
+  author = { John E. Beasley  and P. C. Chu},
+  title = {A Genetic Algorithm for the Multidimensional Knapsack Problem},
+  journal = {Journal of Heuristics},
+  year = 1998,
+  volume = 4,
+  number = 1,
+  pages = {63--86}
+}
+
+ +
+@article{BeaShaSmiLop2018review,
+  author = {Bealt, Jennifer and Shaw, Duncan and Smith, Chris M. and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  year = 2019,
+  title = {Peer Reviews for Making Cities Resilient: A Systematic
+                  Literature Review},
+  journal = {International Journal of Emergency Management},
+  volume = 15,
+  number = 4,
+  pages = {334--359},
+  doi = {10.1504/IJEM.2019.104201},
+  abstract = {Peer reviews are a unique governance tool that use expertise
+                  from one city or country to assess and strengthen the
+                  capabilities of another. Peer review tools are gaining
+                  momentum in disaster management and remain an important but
+                  understudied topic in risk governance. Methodologies to
+                  conduct a peer review are still in their infancy. To enhance
+                  these, a systematic literature review (SLR) of academic and
+                  non-academic literature was conducted on city resilience peer
+                  reviews. Thirty-three attributes of resilience are
+                  identified, which provides useful insights into how research
+                  and practice can inform risk governance, and utilise peer
+                  reviews, to drive meaningful change. Moreover, it situates
+                  the challenges associated with resilience building tools
+                  within risk governance to support the development of
+                  interdisciplinary perspectives for integrated city resilience
+                  frameworks. Results of this research have been used to
+                  develop a peer review methodology and an international
+                  standard on conducting peer reviews for disaster risk
+                  reduction.},
+  keywords = {city resilience, city peer review, disaster risk governance}
+}
+
+ +
+@article{Beasley1990orlib,
+  author = { John E. Beasley },
+  title = {{OR}-{Library:} distributing test problems by electronic
+                  mail},
+  journal = {Journal of the Operational Research Society},
+  year = 1990,
+  pages = {1069--1072},
+  note = {Currently available from
+                  \url{http://people.brunel.ac.uk/~mastjjb/jeb/info.html}}
+}
+
+ +
+@article{BehFat2011,
+  author = {J. Behnamian and S.M.T. {Fatemi Ghomi}},
+  title = {Hybrid Flowshop Scheduling with Machine and Resource-dependent Processing Times},
+  journal = {Applied Mathematical Modelling},
+  year = 2011,
+  volume = 35,
+  number = 3,
+  pages = {1107--1123}
+}
+
+ +
+@article{Bel1954,
+  author = {Richard Bellman},
+  title = {The theory of dynamic programming},
+  journal = {Bulletin of the American Mathematical Society},
+  volume = 60,
+  year = 1954,
+  pages = {503--515}
+}
+
+ +
+@article{BelCesDigSchUrl2016,
+  author = {Ruggero Bellio and  Sara Ceschia  and Luca {Di Gaspero} and Andrea Schaerf and  Tommaso Urli },
+  title = {Feature-based tuning of simulated annealing applied to
+                  the curriculum-based course timetabling problem},
+  journal = {Computers \& Operations Research},
+  volume = 65,
+  pages = {83--92},
+  year = 2016,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Ben92,
+  author = { Jon Louis Bentley },
+  title = {Fast Algorithms for Geometric Traveling Salesman
+                  Problems},
+  journal = {ORSA Journal on Computing},
+  year = 1992,
+  volume = 4,
+  number = 4,
+  pages = {387--411}
+}
+
+ +
+@article{BenKao2013,
+  author = {Una Benlic and  Jin-Kao Hao },
+  title = {Breakout Local Search for the Quadratic Assignment Problem},
+  journal = {Applied Mathematics and Computation},
+  year = 2013,
+  volume = 219,
+  number = 9,
+  pages = {4800--4815}
+}
+
+ +
+@article{BenLiuAuIst2014transient,
+  title = {Transient protein-protein interface prediction: datasets,
+                  features, algorithms, and the {RAD-T} predictor},
+  author = {Bendell, Calem J. and Liu, Shalon and Aumentado-Armstrong,
+                  Tristan and Istrate, Bogdan and Cernek, Paul T. and Khan,
+                  Samuel and Picioreanu, Sergiu and Zhao, Michael and Murgita,
+                  Robert A.},
+  journal = {BMC Bioinformatics},
+  volume = 15,
+  pages = 82,
+  year = 2014
+}
+
+ +
+@article{BenLodPro2021ml,
+  author = { Bengio, Yoshua  and  Andrea Lodi  and Antoine Prouvost},
+  title = {Machine learning for combinatorial optimization: A
+                  methodological tour d'horizon},
+  journal = {European Journal of Operational Research},
+  year = 2021,
+  volume = 290,
+  number = 2,
+  pages = {405--421},
+  doi = {10.1016/j.ejor.2020.07.063},
+  keywords = {Combinatorial optimization, Machine learning, Branch and
+                  bound, Mixed-integer programming solvers},
+  abstract = {This paper surveys the recent attempts, both from the machine
+                  learning and operations research communities, at leveraging
+                  machine learning to solve combinatorial optimization
+                  problems. Given the hard nature of these problems,
+                  state-of-the-art algorithms rely on handcrafted heuristics
+                  for making decisions that are otherwise too expensive to
+                  compute or mathematically not well defined. Thus, machine
+                  learning looks like a natural candidate to make such
+                  decisions in a more principled and optimized way. We advocate
+                  for pushing further the integration of machine learning and
+                  combinatorial optimization and detail a methodology to do
+                  so. A main point of the paper is seeing generic optimization
+                  problems as data points and inquiring what is the relevant
+                  distribution of problems to use for learning on a given
+                  task.}
+}
+
+ +
+@article{BenRit2016:cor,
+  author = {Alexander Javier Benavides and  Marcus Ritt},
+  title = {Two Simple and Effective Heuristics for Minimizing the
+                  Makespan in Non-permutation Flow Shops},
+  journal = {Computers \& Operations Research},
+  year = 2016,
+  volume = 66,
+  pages = {160--169},
+  doi = {10.1016/j.cor.2015.08.001}
+}
+
+ +
+@article{Benders1962,
+  author = {Benders, J. F.},
+  title = {Partitioning Procedures for Solving Mixed-variables Programming Problems},
+  journal = {Numerische Mathematik},
+  year = 1962,
+  volume = 4,
+  number = 3,
+  pages = {238--252}
+}
+
+ +
+@article{Bentley1980,
+  author = { Jon Louis Bentley },
+  title = {Multidimensional Divide-and-conquer},
+  journal = {Communications of the ACM},
+  year = 1980,
+  volume = 23,
+  number = 4,
+  doi = {10.1145/358841.358850},
+  pages = {214--229},
+  abstract = {Most results in the field of algorithm design are single
+                  algorithms that solve single problems. In this paper we
+                  discuss multidimensional divide-and-conquer, an algorithmic
+                  paradigm that can be instantiated in many different ways to
+                  yield a number of algorithms and data structures for
+                  multidimensional problems. We use this paradigm to give
+                  best-known solutions to such problems as the ECDF, maxima,
+                  range searching, closest pair, and all nearest neighbor
+                  problems. The contributions of the paper are on two
+                  levels. On the first level are the particular algorithms and
+                  data structures given by applying the paradigm.  On the
+                  second level is the more novel contribution of this paper: a
+                  detailed study of an algorithmic paradigm that is specific
+                  enough to be described precisely yet general enough to solve
+                  a wide variety of problems.}
+}
+
+ +
+@article{BerBen2012jmlr,
+  author = { James S. Bergstra  and  Bengio, Yoshua },
+  title = {Random Search for Hyper-Parameter Optimization},
+  journal = {Journal of Machine Learning Research},
+  year = 2012,
+  volume = 13,
+  pages = {281--305},
+  abstract = {Grid search and manual search are the most widely
+                  used strategies for hyper-parameter
+                  optimization. This paper shows empirically and
+                  theoretically that randomly chosen trials are more
+                  efficient for hyper-parameter optimization than
+                  trials on a grid. Empirical evidence comes from a
+                  comparison with a large previous study that used
+                  grid search and manual search to configure neural
+                  networks and deep belief networks. Compared with
+                  neural networks configured by a pure grid search, we
+                  find that random search over the same domain is able
+                  to find models that are as good or better within a
+                  small fraction of the computation time. Granting
+                  random search the same computational budget, random
+                  search finds better models by effectively searching
+                  a larger, less promising configuration
+                  space. Compared with deep belief networks configured
+                  by a thoughtful combination of manual search and
+                  grid search, purely random search over the same
+                  32-dimensional configuration space found
+                  statistically equal performance on four of seven
+                  data sets, and superior performance on one of
+                  seven. A Gaussian process analysis of the function
+                  from hyper-parameters to validation set performance
+                  reveals that for most data sets only a few of the
+                  hyper-parameters really matter, but that different
+                  hyper-parameters are important on different data
+                  sets. This phenomenon makes grid search a poor
+                  choice for configuring algorithms for new data
+                  sets. Our analysis casts some light on why recent
+                  "High Throughput" methods achieve surprising
+                  success: they appear to search through a large number
+                  of hyper-parameters because most hyper-parameters do
+                  not matter much. We anticipate that growing interest
+                  in large hierarchical models will place an
+                  increasing burden on techniques for hyper-parameter
+                  optimization; this work shows that random search is
+                  a natural baseline against which to judge progress
+                  in the development of adaptive (sequential)
+                  hyper-parameter optimization algorithms.},
+  epub = {http://www.jmlr.org/papers/volume13/bergstra12a/bergstra12a.pdf}
+}
+
+ +
+@article{BerEmmTav2017managing,
+  title = {Managing catastrophic climate risks under model uncertainty
+                  aversion},
+  author = {Berger, Lo{\"i}c and Emmerling, Johannes and Tavoni,
+                  Massimo},
+  journal = {Management Science},
+  volume = 63,
+  number = 3,
+  pages = {749--765},
+  year = 2017,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{BerFisLod2007,
+  title = {A feasibility pump heuristic for general mixed-integer problems},
+  author = {Bertacco, Livio and  Matteo Fischetti  and  Andrea Lodi },
+  journal = {Discrete Optimization},
+  volume = 4,
+  number = 1,
+  pages = {63--76},
+  year = 2007,
+  publisher = {Elsevier}
+}
+
+ +
+@article{BerKal2020,
+  title = {From predictive to prescriptive analytics},
+  author = {Bertsimas, Dimitris and Kallus, Nathan},
+  journal = {Management Science},
+  volume = 66,
+  number = 3,
+  pages = {1025--1044},
+  year = 2020,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{BerKraSch2016bayesian,
+  author = {Felix Berkenkamp and Andreas Krause and Angela P. Schoellig},
+  title = {Bayesian Optimization with Safety Constraints: Safe and
+                  Automatic Parameter Tuning in Robotics},
+  journal = {Arxiv preprint arXiv:1602.04450},
+  year = 2016,
+  url = {http://arxiv.org/abs/1602.04450},
+  keywords = {Safe Optimization, SafeOpt}
+}
+
+ +
+@article{BerKraSch2021bayesian,
+  author = {Berkenkamp, Felix and Krause, Andreas and Schoellig, Angela
+                  P.},
+  title = {Bayesian optimization with safety constraints: safe and
+                  automatic parameter tuning in robotics},
+  journal = {Machine Learning},
+  year = 2021,
+  month = jun,
+  annote = {Preprint: \url{http://arxiv.org/abs/1602.04450}},
+  doi = {10.1007/s10994-021-06019-1},
+  abstract = {Selecting the right tuning parameters for algorithms is a
+                  pravelent problem in machine learning that can significantly
+                  affect the performance of algorithms. Data-efficient
+                  optimization algorithms, such as Bayesian optimization, have
+                  been used to automate this process. During experiments on
+                  real-world systems such as robotic platforms these methods
+                  can evaluate unsafe parameters that lead to safety-critical
+                  system failures and can destroy the system. Recently, a safe
+                  Bayesian optimization algorithm, called SafeOpt, has been
+                  developed, which guarantees that the performance of the
+                  system never falls below a critical value; that is, safety is
+                  defined based on the performance function. However, coupling
+                  performance and safety is often not desirable in practice,
+                  since they are often opposing objectives. In this paper, we
+                  present a generalized algorithm that allows for multiple
+                  safety constraints separate from the objective. Given an
+                  initial set of safe parameters, the algorithm maximizes
+                  performance but only evaluates parameters that satisfy safety
+                  for all constraints with high probability. To this end, it
+                  carefully explores the parameter space by exploiting
+                  regularity assumptions in terms of a Gaussian process
+                  prior. Moreover, we show how context variables can be used to
+                  safely transfer knowledge to new situations and tasks. We
+                  provide a theoretical analysis and demonstrate that the
+                  proposed algorithm enables fast, automatic, and safe
+                  optimization of tuning parameters in experiments on a
+                  quadrotor vehicle.}
+}
+
+ +
+@article{BerTsiWu1997:joh,
+  author = {Dimitri P. Bertsekas and John N. Tsitsiklis and Cynara Wu},
+  title = {Rollout Algorithms for Combinatorial Optimization},
+  journal = {Journal of Heuristics},
+  year = 1997,
+  volume = 3,
+  number = 3,
+  pages = {245--262}
+}
+
+ +
+@article{BerWan1987binpack,
+  title = {Two-dimensional finite bin-packing algorithms},
+  author = {Berkey, Judith O. and Wang, Pearl Y.},
+  journal = {Journal of the Operational Research Society},
+  volume = 38,
+  number = 5,
+  pages = {423--429},
+  year = 1987,
+  doi = {10.2307/2582731}
+}
+
+ +
+@article{BeuFonLopPaqVah09:tec,
+  author = { Nicola Beume  and  Carlos M. Fonseca  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Jan Vahrenhold },
+  title = {On the complexity of computing the hypervolume
+                  indicator},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2009,
+  volume = 13,
+  number = 5,
+  pages = {1075--1082},
+  doi = {10.1109/TEVC.2009.2015575},
+  abstract = {The goal of multi-objective optimization is to find
+                  a set of best compromise solutions for typically
+                  conflicting objectives. Due to the complex nature of
+                  most real-life problems, only an approximation to
+                  such an optimal set can be obtained within
+                  reasonable (computing) time. To compare such
+                  approximations, and thereby the performance of
+                  multi-objective optimizers providing them, unary
+                  quality measures are usually applied. Among these,
+                  the \emph{hypervolume indicator} (or
+                  \emph{S-metric}) is of particular relevance due to
+                  its favorable properties. Moreover, this indicator
+                  has been successfully integrated into stochastic
+                  optimizers, such as evolutionary algorithms, where
+                  it serves as a guidance criterion for finding good
+                  approximations to the Pareto front. Recent results
+                  show that computing the hypervolume indicator can be
+                  seen as solving a specialized version of Klee's
+                  Measure Problem.  In general, Klee's Measure Problem
+                  can be solved with $\mathcal{O}(n \log n +
+                  n^{d/2}\log n)$ comparisons for an input instance of
+                  size $n$ in $d$ dimensions; as of this writing, it
+                  is unknown whether a lower bound higher than
+                  $\Omega(n \log n)$ can be proven. In this article,
+                  we derive a lower bound of $\Omega(n\log n)$ for the
+                  complexity of computing the hypervolume indicator in
+                  any number of dimensions $d>1$ by reducing the
+                  so-called \textsc{UniformGap} problem to it.  For
+                  the three dimensional case, we also present a
+                  matching upper bound of $\mathcal{O}(n\log n)$
+                  comparisons that is obtained by extending an
+                  algorithm for finding the maxima of a point set.}
+}
+
+ +
+@article{BeuNauEmm2007ejor,
+  author = { Nicola Beume  and  Boris Naujoks  and  Emmerich, Michael T. M. },
+  title = {{SMS-EMOA}: Multiobjective selection based on
+                  dominated hypervolume},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 181,
+  number = 3,
+  pages = {1653--1669},
+  doi = {10.1016/j.ejor.2006.08.008}
+}
+
+ +
+@article{BeySch2002:es,
+  author = {  Hans-Georg Beyer  and  Hans-Paul Schwefel },
+  title = {Evolution Strategies: A Comprehensive Introduction},
+  journal = {Natural Computing},
+  volume = 1,
+  pages = {3--52},
+  year = 2002
+}
+
+ +
+@article{BeySchWeg2002,
+  title = {How to analyse evolutionary algorithms},
+  author = {  Hans-Georg Beyer  and  Hans-Paul Schwefel  and  Ingo Wegener },
+  journal = {Theoretical Computer Science},
+  volume = 287,
+  number = 1,
+  pages = {101--130},
+  year = 2002,
+  publisher = {Elsevier}
+}
+
+ +
+@article{BezLopStu2015tec,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Component-Wise Design of Multi-Objective
+                  Evolutionary Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2016,
+  volume = 20,
+  number = 3,
+  pages = {403--417},
+  doi = {10.1109/TEVC.2015.2474158},
+  supplement = {https://github.com/iridia-ulb/automoea-tevc-2016}
+}
+
+ +
+@article{BezLopStu2017assessment,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Large-Scale Experimental Evaluation of High-Performing
+                  Multi- and Many-Objective Evolutionary Algorithms},
+  year = 2018,
+  journal = {Evolutionary Computation},
+  doi = {10.1162/evco_a_00217},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2015-007/},
+  volume = 26,
+  number = 4,
+  pages = {621--656},
+  alias = {BezLopStu2016assessment},
+  abstract = {Research on multi-objective evolutionary algorithms (MOEAs)
+                  has produced over the past decades a large number of
+                  algorithms and a rich literature on performance assessment
+                  tools to evaluate and compare them. Yet, newly proposed MOEAs
+                  are typically compared against very few, often a decade older
+                  MOEAs. One reason for this apparent contradiction is the lack
+                  of a common baseline for comparison, with each subsequent
+                  study often devising its own experimental scenario, slightly
+                  different from other studies. As a result, the state of the
+                  art in MOEAs is a disputed topic. This article reports a
+                  systematic, comprehensive evaluation of a large number of
+                  MOEAs that covers a wide range of experimental scenarios. A
+                  novelty of this study is the separation between the
+                  higher-level algorithmic components related to
+                  multi-objective optimization (MO), which characterize each
+                  particular MOEA, and the underlying parameters-such as
+                  evolutionary operators, population size, etc.-whose
+                  configuration may be tuned for each scenario. Instead of
+                  relying on a common or "default" parameter configuration that
+                  may be low-performing for particular MOEAs or scenarios and
+                  unintentionally biased, we tune the parameters of each MOEA
+                  for each scenario using automatic algorithm configuration
+                  methods. Our results confirm some of the assumed knowledge in
+                  the field, while at the same time they provide new insights
+                  on the relative performance of MOEAs for many-objective
+                  problems. For example, under certain conditions,
+                  indicator-based MOEAs are more competitive for such problems
+                  than previously assumed. We also analyze problem-specific
+                  features affecting performance, the agreement between
+                  performance metrics, and the improvement of tuned
+                  configurations over the default configurations used in the
+                  literature. Finally, the data produced is made publicly
+                  available to motivate further analysis and a baseline for
+                  future comparisons.}
+}
+
+ +
+@article{BezLopStu2019ec,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Designing State-of-the-Art Multi- and
+                  Many-Objective Evolutionary Algorithms},
+  journal = {Evolutionary Computation},
+  year = 2020,
+  volume = 28,
+  number = 2,
+  pages = {195--226},
+  doi = {10.1162/evco_a_00263},
+  supplement = {https://github.com/iridia-ulb/automoea-ecj-2020},
+  abstract = {A recent comparison of well-established multiobjective
+                  evolutionary algorithms (MOEAs) has helped better identify
+                  the current state-of-the-art by considering (i) parameter
+                  tuning through automatic configuration, (ii) a wide range of
+                  different setups, and (iii) various performance
+                  metrics. Here, we automatically devise MOEAs with verified
+                  state-of-the-art performance for multi- and many-objective
+                  continuous optimization. Our work is based on two main
+                  considerations. The first is that high-performing algorithms
+                  can be obtained from a configurable algorithmic framework in
+                  an automated way. The second is that multiple performance
+                  metrics may be required to guide this automatic design
+                  process. In the first part of this work, we extend our
+                  previously proposed algorithmic framework, increasing the
+                  number of MOEAs, underlying evolutionary algorithms, and
+                  search paradigms that it comprises. These components can be
+                  combined following a general MOEA template, and an automatic
+                  configuration method is used to instantiate high-performing
+                  MOEA designs that optimize a given performance metric and
+                  present state-of-the-art performance. In the second part, we
+                  propose a multiobjective formulation for the automatic MOEA
+                  design, which proves critical for the context of
+                  many-objective optimization due to the disagreement of
+                  established performance metrics. Our proposed formulation
+                  leads to an automatically designed MOEA that presents
+                  state-of-the-art performance according to a set of metrics,
+                  rather than a single one.}
+}
+
+ +
+@article{BiaBirMan2006jmma,
+  author = { Leonora Bianchi  and  Mauro Birattari  and  M. Manfrin and
+                  M. Mastrolilli  and  Lu{\'i}s Paquete  and  O. Rossi-Doria  and  Tommaso Schiavinotto },
+  title = {Hybrid Metaheuristics for the Vehicle Routing Problem with
+                  Stochastic Demands},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2006,
+  volume = 5,
+  number = 1,
+  pages = {91--110},
+  alias = {Bia++06}
+}
+
+ +
+@article{BiaDorGam2009survey,
+  title = {A survey on metaheuristics for stochastic combinatorial
+                  optimization},
+  author = { Leonora Bianchi  and  Marco Dorigo  and  L. M. Gambardella  and  Gutjahr, Walter J. },
+  journal = {Natural Computing},
+  volume = 8,
+  number = 2,
+  pages = {239--287},
+  year = 2009
+}
+
+ +
+@article{BinGinRou2015gaupar,
+  title = {Quantifying uncertainty on {Pareto} fronts with {Gaussian}
+                  process conditional simulations},
+  volume = 243,
+  doi = {10.1016/j.ejor.2014.07.032},
+  abstract = {Multi-objective optimization algorithms aim at finding
+                  Pareto-optimal solutions. Recovering Pareto fronts or Pareto
+                  sets from a limited number of function evaluations are
+                  challenging problems. A popular approach in the case of
+                  expensive-to-evaluate functions is to appeal to
+                  metamodels. Kriging has been shown efficient as a base for
+                  sequential multi-objective optimization, notably through
+                  infill sampling criteria balancing exploitation and
+                  exploration such as the Expected Hypervolume
+                  Improvement. Here we consider Kriging metamodels not only for
+                  selecting new points, but as a tool for estimating the whole
+                  Pareto front and quantifying how much uncertainty remains on
+                  it at any stage of Kriging-based multi-objective optimization
+                  algorithms. Our approach relies on the Gaussian process
+                  interpretation of Kriging, and bases upon conditional
+                  simulations. Using concepts from random set theory, we
+                  propose to adapt the Vorob'ev expectation and deviation to
+                  capture the variability of the set of non-dominated
+                  points. Numerical experiments illustrate the potential of the
+                  proposed workflow, and it is shown on examples how Gaussian
+                  process simulations and the estimated Vorob'ev deviation can
+                  be used to monitor the ability of Kriging-based
+                  multi-objective optimization algorithms to accurately learn
+                  the Pareto front.},
+  number = 2,
+  journal = {European Journal of Operational Research},
+  author = {Binois, M. and Ginsbourger, D. and Roustant, O.},
+  year = 2015,
+  keywords = {Attainment function, Expected Hypervolume Improvement,
+                  Kriging, Multi-objective optimization, Vorob'ev expectation},
+  pages = {386--394}
+}
+
+ +
+@article{BirBalStuDor07:informs,
+  author = { Mauro Birattari  and   Prasanna Balaprakash  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Estimation Based Local Search for Stochastic Combinatorial Optimization},
+  journal = {INFORMS Journal on Computing},
+  year = 2008,
+  volume = 20,
+  number = 4,
+  pages = {644--658}
+}
+
+ +
+@article{BirPelDor2007:tec,
+  author = { Mauro Birattari  and  Paola Pellegrini  and  Marco Dorigo },
+  title = {On the invariance of ant colony optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2007,
+  volume = 11,
+  number = 6,
+  pages = {732--742},
+  doi = {10.1109/TEVC.2007.892762},
+  alias = {BirPelDor2007ieee-tevc}
+}
+
+ +
+@article{BirZloDor06meta_design,
+  author = { Mauro Birattari  and Zlochin, M. and  Marco Dorigo },
+  title = {Towards a theory of practice in metaheuristics design: A machine learning perspective},
+  journal = {Theoretical Informatics and Applications},
+  year = 2006,
+  volume = 40,
+  number = 2,
+  pages = {353--369}
+}
+
+ +
+@article{BisIzzYam2010:pagmo-arxiv,
+  title = {A Global Optimisation Toolbox for Massively Parallel
+                  Engineering Optimisation},
+  author = {Biscani, Francesco and Izzo, Dario and Yam, Chit Hong},
+  journal = {Arxiv preprint arXiv:1004.3824},
+  year = 2010,
+  url = {http://arxiv.org/abs/1004.3824},
+  keywords = {PaGMO},
+  abstract = {A software platform for global optimisation, called PaGMO,
+                  has been developed within the Advanced Concepts Team (ACT) at
+                  the European Space Agency, and was recently released as an
+                  open-source project. PaGMO is built to tackle
+                  high-dimensional global optimisation problems, and it has
+                  been successfully used to find solutions to real-life
+                  engineering problems among which the preliminary design of
+                  interplanetary spacecraft trajectories - both chemical
+                  (including multiple flybys and deep-space maneuvers) and
+                  low-thrust (limited, at the moment, to single phase
+                  trajectories), the inverse design of nano-structured
+                  radiators and the design of non-reactive controllers for
+                  planetary rovers. Featuring an arsenal of global and local
+                  optimisation algorithms (including genetic algorithms,
+                  differential evolution, simulated annealing, particle swarm
+                  optimisation, compass search, improved harmony search, and
+                  various interfaces to libraries for local optimisation such
+                  as SNOPT, IPOPT, GSL and NLopt), PaGMO is at its core a C++
+                  library which employs an object-oriented architecture
+                  providing a clean and easily-extensible optimisation
+                  framework. Adoption of multi-threaded programming ensures the
+                  efficient exploitation of modern multi-core architectures and
+                  allows for a straightforward implementation of the island
+                  model paradigm, in which multiple populations of candidate
+                  solutions asynchronously exchange information in order to
+                  speed-up and improve the optimisation process. In addition to
+                  the C++ interface, PaGMO's capabilities are exposed to the
+                  high-level language Python, so that it is possible to easily
+                  use PaGMO in an interactive session and take advantage of the
+                  numerous scientific Python libraries available.}
+}
+
+ +
+@article{BisBinLan2023wirdmkd,
+  title = {Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges},
+  author = { Bernd Bischl  and Binder, Martin and Lang, Michel and Pielok, Tobias and Richter, Jakob and Coors, Stefan and Thomas, Janek and Ullmann, Theresa and Becker, Marc and Boulesteix, Anne-Laure and Deng, Difan and  Marius Thomas Lindauer },
+  journal = {Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery},
+  volume = 13,
+  number = 2,
+  pages = {e1484},
+  year = 2023,
+  publisher = {Wiley Online Library}
+}
+
+ +
+@article{BisKerKot++16:ASlib,
+  author = { Bernd Bischl  and  Pascal Kerschke  and Kotthoff, Lars and  Marius Thomas Lindauer  and  Yuri Malitsky  and Alexandre Fr{\'{e}}chette and  Holger H. Hoos  and  Frank Hutter  and  Kevin Leyton-Brown  and  Kevin Tierney  and  Joaquin Vanschoren },
+  title = {{ASlib}: A Benchmark Library for Algorithm Selection},
+  journal = {Artificial Intelligence},
+  year = 2016,
+  volume = 237,
+  pages = {41--58}
+}
+
+ +
+@article{BisLanKot2016mlr,
+  title = {{\rpackage{mlr}}: Machine Learning in \proglang{R}},
+  author = { Bernd Bischl  and Michel Lang and Kotthoff, Lars and Julia
+                  Schiffner and Jakob Richter and Erich Studerus and Giuseppe
+                  Casalicchio and Zachary M. Jones},
+  journal = {Journal of Machine Learning Research},
+  year = 2016,
+  volume = 17,
+  number = 170,
+  pages = {1--5},
+  epub = {http://jmlr.org/papers/v17/15-066.html}
+}
+
+ +
+@article{BlaHerSanMar2008vis,
+  title = {A new graphical visualization of n-dimensional {Pareto} front
+                  for decision-making in multiobjective optimization},
+  author = {Blasco, Xavier and Herrero, Juan M. and Sanchis, Javier and
+                  Mart{\'i}nez, Manuel},
+  journal = {Information Sciences},
+  volume = 178,
+  number = 20,
+  pages = {3908--3924},
+  year = 2008,
+  publisher = {Elsevier}
+}
+
+ +
+@article{BlaRayEde2017:corr,
+  author = {Craig Blackmore and Oliver Ray and Kerstin Eder},
+  title = {Automatically Tuning the {GCC} Compiler to Optimize the
+                  Performance of Applications Running on Embedded Systems},
+  journal = {Arxiv preprint arXiv:1703.08228},
+  url = {https://arxiv.org/abs/1703.08228},
+  year = 2017
+}
+
+ +
+@article{BleBlu2007:jmma,
+  author = { Mar{\'i}a J. Blesa  and  Christian Blum },
+  title = {Finding edge-disjoint paths in networks by means of
+                  artificial ant colonies},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2007,
+  volume = 6,
+  number = 3,
+  pages = {361--391}
+}
+
+ +
+@article{BliCosRefZha2023aitsp,
+  title = {The First {AI4TSP} Competition: Learning to Solve Stochastic
+                  Routing Problems},
+  journal = {Artificial Intelligence},
+  pages = 103918,
+  volume = 319,
+  year = 2023,
+  issn = {0004-3702},
+  doi = {10.1016/j.artint.2023.103918},
+  author = {Laurens Bliek and Paulo {da Costa} and Reza {Refaei Afshar}
+                  and Robbert Reijnen and Yingqian Zhang and Tom Catshoek and
+                  Dani{\"e}l Vos and Sicco Verwer and Fynn Schmitt-Ulms and
+                  Andr{\'e} Hottung and Tapan Shah and  Meinolf Sellmann  and  Kevin Tierney  and Carl Perreault-Lafleur and Caroline Leboeuf
+                  and Federico Bobbio and Justine Pepin and Warley Almeida
+                  Silva and Ricardo Gama and Hugo L. Fernandes and  Martin Zaefferer  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Irurozki, Ekhine },
+  keywords = {AI for TSP competition, Travelling salesman problem, Routing
+                  problem, Stochastic combinatorial optimization,
+                  Surrogate-based optimization, Deep reinforcement learning},
+  abstract = {This paper reports on the first international competition on
+                  AI for the traveling salesman problem (TSP) at the
+                  International Joint Conference on Artificial Intelligence
+                  2021 (IJCAI-21). The TSP is one of the classical
+                  combinatorial optimization problems, with many variants
+                  inspired by real-world applications. This first competition
+                  asked the participants to develop algorithms to solve an
+                  orienteering problem with stochastic weights and time windows
+                  (OPSWTW). It focused on two learning approaches:
+                  surrogate-based optimization and deep reinforcement
+                  learning. In this paper, we describe the problem, the
+                  competition setup, and the winning methods, and give an
+                  overview of the results. The winning methods described in
+                  this work have advanced the state-of-the-art in using AI for
+                  stochastic routing problems. Overall, by organizing this
+                  competition we have introduced routing problems as an
+                  interesting problem setting for AI researchers. The simulator
+                  of the problem has been made open-source and can be used by
+                  other researchers as a benchmark for new learning-based
+                  methods. The instances and code for the competition are
+                  available at
+                  \url{https://github.com/paulorocosta/ai-for-tsp-competition}.}
+}
+
+ +
+@article{Blu05:cor,
+  author = { Christian Blum },
+  title = {{Beam-ACO}---{Hybridizing} Ant Colony Optimization
+                  with Beam Search: {An} Application to Open Shop
+                  Scheduling},
+  journal = {Computers \& Operations Research},
+  year = 2005,
+  volume = 32,
+  number = 6,
+  pages = {1565--1591},
+  alias = {BlumCOR05}
+}
+
+ +
+@article{Blu08:informs,
+  author = { Christian Blum },
+  title = {Beam-{ACO} for simple assembly line balancing},
+  journal = {INFORMS Journal on Computing},
+  year = 2008,
+  volume = 20,
+  number = 4,
+  pages = {618--627},
+  doi = {10.1287/ijoc.1080.0271},
+  alias = {Blu08:ijoc}
+}
+
+ +
+@article{BluBleLop09-BeamSearch-LCS,
+  author = { Christian Blum  and  Mar{\'i}a J. Blesa  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Beam search for the longest common subsequence
+                  problem},
+  number = 12,
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  pages = {3178--3186},
+  volume = 36,
+  doi = {10.1016/j.cor.2009.02.005},
+  abstract = {The longest common subsequence problem is a classical string
+                  problem that concerns finding the common part of a set of
+                  strings. It has several important applications, for example,
+                  pattern recognition or computational biology. Most research
+                  efforts up to now have focused on solving this problem
+                  optimally. In comparison, only few works exist dealing with
+                  heuristic approaches. In this work we present a deterministic
+                  beam search algorithm. The results show that our algorithm
+                  outperforms the current state-of-the-art approaches not only
+                  in solution quality but often also in computation time.}
+}
+
+ +
+@article{BluCaBle2015swarm,
+  author = { Christian Blum  and Borja Calvo and  Mar{\'i}a J. Blesa },
+  title = {{FrogCOL} and {FrogMIS}: New Decentralized Algorithms for Finding Large Independent Sets in Graphs},
+  journal = {Swarm Intelligence},
+  year = 2015,
+  volume = 9,
+  number = {2-3},
+  pages = {205--227},
+  doi = {10.1007/s11721-015-0110-1},
+  keywords = {irace}
+}
+
+ +
+@article{BluDor03:ieee_tsmcb,
+  author = { Christian Blum  and  Marco Dorigo },
+  title = {The hyper-cube framework for ant colony optimization},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics -- Part B},
+  year = 2004,
+  volume = 34,
+  number = 2,
+  pages = {1161--1172}
+}
+
+ +
+@article{BluDor2005:tec,
+  author = { Christian Blum  and  Marco Dorigo },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  number = 2,
+  pages = {159--174},
+  title = {Search Bias in Ant Colony Optimization: On the Role
+                  of Competition-Balanced Systems},
+  volume = 9,
+  year = 2005
+}
+
+ +
+@article{BluOch2021,
+  author = { Christian Blum  and  Gabriela Ochoa },
+  title = {A comparative analysis of two matheuristics by means of merged local optima networks},
+  journal = {European Journal of Operational Research},
+  volume = 290,
+  number = 1,
+  pages = {36--56},
+  year = 2021
+}
+
+ +
+@article{BluPinLopLoz2015cor,
+  author = { Christian Blum  and  Pedro Pinacho  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Jos{\'e} A. Lozano },
+  title = {Construct, Merge, Solve \& Adapt: A New General Algorithm for
+                  Combinatorial Optimization},
+  journal = {Computers \& Operations Research},
+  year = 2016,
+  volume = 68,
+  pages = {75--88},
+  doi = {10.1016/j.cor.2015.10.014},
+  keywords = {irace, CMSA}
+}
+
+ +
+@article{BluPucRaiRol11:asc,
+  author = { Christian Blum  and  Jakob Puchinger  and  G{\"u}nther R. Raidl  and  Andrea Roli },
+  title = {Hybrid Metaheuristics in Combinatorial Optimization: A Survey},
+  journal = {Applied Soft Computing},
+  year = 2011,
+  volume = 11,
+  number = 6,
+  pages = {4135--4151}
+}
+
+ +
+@article{BluRol03:acm-cs,
+  author = { Christian Blum  and  Andrea Roli },
+  title = {Metaheuristics in Combinatorial Optimization:
+                  Overview and Conceptual Comparison},
+  journal = {{ACM} Computing Surveys},
+  year = 2003,
+  volume = 35,
+  number = 3,
+  pages = {268--308}
+}
+
+ +
+@article{BluSam2004:jmma,
+  author = { Christian Blum  and  M. Sampels },
+  title = {An Ant Colony Optimization Algorithm for Shop
+                  Scheduling Problems},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2004,
+  volume = 3,
+  number = 3,
+  pages = {285--308},
+  doi = {10.1023/B:JMMA.0000038614.39977.6f}
+}
+
+ +
+@article{BluYabBle08:cor,
+  author = { Christian Blum  and  M. {Y{\'a}bar Vall{\`e}s}  and  Mar{\'i}a J. Blesa },
+  title = {An ant colony optimization algorithm for {DNA} sequencing by hybridization},
+  journal = {Computers \& Operations Research},
+  year = 2008,
+  volume = 35,
+  number = 11,
+  pages = {3620--3635},
+  alias = {BluYabBle08}
+}
+
+ +
+@article{BocFawVal2018performance,
+  author = {Bocchese, Andrea F. and  Chris Fawcett  and Vallati, Mauro and Gerevini, Alfonso E. and  Holger H. Hoos },
+  title = {Performance robustness of {AI} planners in the 2014
+                  International Planning Competition},
+  volume = 31,
+  doi = {10.3233/AIC-170537},
+  abstract = {Solver competitions have been used in many areas of AI to
+                  assess the current state of the art and guide future research
+                  and development. AI planning is no exception, and the
+                  International Planning Competition (IPC) has been frequently
+                  run for nearly two decades. Due to the organisational and
+                  computational burden involved in running these competitions,
+                  solvers are generally compared using a single homogeneous
+                  hardware and software environment for all competitors. To
+                  what extent does the specific choice of hardware and software
+                  environment have an effect on solver performance, and is that
+                  effect distributed equally across the competing solvers? In
+                  this work, we use the competing planners and benchmark
+                  instance sets from the 2014 IPC to investigate these two
+                  questions. We recreate the 2014 IPC Optimal and Agile tracks
+                  on two distinct hardware environments and eight distinct
+                  software environments. We show that solver performance varies
+                  significantly based on the hardware and software environment,
+                  and that this variation is not equal for all
+                  planners. Furthermore, the observed variation is sufficient
+                  to change the competition rankings, including the top-ranked
+                  planners for some tracks.},
+  number = 6,
+  journal = {AI Communications},
+  publisher = {IOS Press},
+  year = 2018,
+  month = dec,
+  pages = {445--463}
+}
+
+ +
+@article{BoeKahMud1994,
+  author = {Kenneth D. Boese and Andrew B. Kahng and Sudhakar Muddu},
+  title = {A New Adaptive Multi-Start Technique for Combinatorial Global
+                  Optimization},
+  journal = {Operations Research Letters},
+  year = 1994,
+  volume = 16,
+  number = 2,
+  pages = {101--113},
+  keywords = {big-valley hypothesis, TSP, landscape analysis}
+}
+
+ +
+@article{Boh2009idcs,
+  author = {Marko Bohanec},
+  title = {Decision making: a computer-science and
+                  information-technology viewpoint},
+  journal = {Interdisciplinary Description of Complex Systems},
+  year = 2009,
+  volume = 7,
+  number = 2,
+  pages = {22--37}
+}
+
+ +
+@article{BohJohSte1986,
+  title = {Generalized Simulated Annealing for Function Optimization},
+  author = { Ihor O. Bohachevsky  and Mark E. Johnson and  Myron L. Stein },
+  journal = {Technometrics},
+  volume = 28,
+  number = 3,
+  pages = {209--217},
+  year = 1986,
+  publisher = {Taylor \& Francis}
+}
+
+ +
+@article{Bor2000,
+  title = {{CHESS} - Changing Horizon Efficient Set Search: A
+                  simple principle for multiobjective optimization},
+  author = {Borges, P. C.},
+  journal = {Journal of Heuristics},
+  volume = 6,
+  number = 3,
+  pages = {405--418},
+  year = 2000
+}
+
+ +
+@article{BorHamTav2007joh,
+  author = {Boros, Endre and Hammer, Peter L.  and Tavares, Gabriel},
+  title = {Local search heuristics for Quadratic Unconstrained Binary
+                  Optimization ({QUBO})},
+  journal = {Journal of Heuristics},
+  year = 2007,
+  volume = 13,
+  number = 2,
+  pages = {99--132}
+}
+
+ +
+@article{Borda1781,
+  author = {Jean-Charles de Borda},
+  journal = {Histoire de l'Acad{\'e}mie Royal des Sciences},
+  title = {M{\'e}moire sur les {\'E}lections au Scrutin},
+  year = 1781,
+  keywords = {ranking}
+}
+
+ +
+@article{BotBon98,
+  author = {Hozefa M. Botee and Eric Bonabeau},
+  title = {Evolving Ant Colony Optimization},
+  year = 1998,
+  journal = {Advances in Complex Systems},
+  volume = 1,
+  pages = {149--159}
+}
+
+ +
+@article{BotSch2019dominance,
+  title = {Dominance for multi-objective robust optimization concepts},
+  author = {Botte, Marco and  Sch{\"o}bel, Anita },
+  journal = {European Journal of Operational Research},
+  volume = 273,
+  number = 2,
+  pages = {430--440},
+  year = 2019,
+  publisher = {Elsevier}
+}
+
+ +
+@article{BouBluBou2012,
+  author = {Salim Bouamama and  Christian Blum  and Abdellah Boukerram},
+  title = {A Population-based Iterated Greedy Algorithm for the Minimum Weight Vertex Cover Problem},
+  journal = {Applied Soft Computing},
+  year = 2012,
+  volume = 12,
+  number = 6,
+  pages = {1632--1639}
+}
+
+ +
+@article{BouForGliPir2010:ejor,
+  author = { G{\'e}raldine Bous  and  Philippe Fortemps  and  Fran\c{c}ois Glineur  and  Marc Pirlot },
+  title = {{ACUTA}: {A} novel method for eliciting additive value functions on the basis of holistic preference statements},
+  journal = {European Journal of Operational Research},
+  year = 2010,
+  volume = 206,
+  number = 2,
+  pages = {435--444}
+}
+
+ +
+@article{BouLec2003ejor,
+  author = {Bouleimen, K. and Lecocq, H.},
+  title = {A new efficient simulated annealing algorithm for
+                  the resource-constrained project scheduling problem
+                  and its multiple mode version},
+  volume = 149,
+  doi = {10.1016/S0377-2217(02)00761-0},
+  abstract = {This paper describes new simulated annealing ({SA)}
+                  algorithms for the resource-constrained project
+                  scheduling problem ({RCPSP)} and its multiple mode
+                  version ({MRCPSP).} The objective function
+                  considered is minimisation of the makespan. The
+                  conventional {SA} search scheme is replaced by a new
+                  design that takes into account the specificity of
+                  the solution space of project scheduling
+                  problems. For {RCPSP}, the search was based on an
+                  alternated activity and time incrementing process,
+                  and all parameters were set after preliminary
+                  statistical experiments done on test instances. For
+                  {MRCPSP}, we introduced an original approach using
+                  two embedded search loops alternating activity and
+                  mode neighbourhood exploration. The performance
+                  evaluation done on the benchmark instances available
+                  in the literature proved the efficiency of both
+                  adaptations that are currently among the most
+                  competitive algorithms for these problems.},
+  number = 2,
+  journal = {European Journal of Operational Research},
+  year = 2003,
+  keywords = {multi-mode resource-constrained project scheduling,
+                  project scheduling, simulated annealing},
+  pages = {268--281}
+}
+
+ +
+@article{BozFowGelKim2010or,
+  title = {Quantitative comparison of approximate solution sets for
+                  multicriteria optimization problems with weighted
+                  {Tchebycheff} preference function},
+  author = {Bozkurt, B. and Fowler, J. W. and Gel, E. S. and Kim, B. and  Murat K{\"o}ksalan  and  Wallenius, Jyrki },
+  journal = {Operations Research},
+  year = 2010,
+  number = 3,
+  pages = {650--659},
+  volume = 58,
+  publisher = {INFORMS},
+  annote = {Proposed IPF indicator}
+}
+
+ +
+@article{BraGreSlo2010bpas,
+  title = {Interactive evolutionary multiobjective optimization driven
+                  by robust ordinal regression},
+  author = { J{\"u}rgen Branke  and  Salvatore Greco  and  Roman S{\l}owi{\'n}ski  and Zielniewicz, P},
+  journal = {Bulletin of the Polish Academy of Sciences: Technical Sciences},
+  volume = 58,
+  number = 3,
+  pages = {347--358},
+  year = 2010,
+  doi = {10.2478/v10175-010-0033-3}
+}
+
+ +
+@article{BraGutRAu2006cms,
+  author = {S. C. Brailsford and  Gutjahr, Walter J.  and M. S. Rauner and
+                  W. Zeppelzauer},
+  title = {Combined Discrete-event Simulation and Ant Colony
+                  Optimisation Approach for Selecting Optimal Screening
+                  Policies for Diabetic Retinopathy},
+  journal = {Computational Management Science},
+  year = 2006,
+  volume = 4,
+  number = 1,
+  pages = {59--83},
+  alias = {Bra++06}
+}
+
+ +
+@article{BraKauSch2001aes,
+  author = { J{\"u}rgen Branke  and Kaussler, T. and Schmeck, H.},
+  title = {Guidance in evolutionary multi-objective optimization},
+  journal = {Advances in Engineering Software},
+  year = 2001,
+  volume = 32,
+  pages = {499--507}
+}
+
+ +
+@article{BraNguPic2016tec,
+  author = { J{\"u}rgen Branke  and S. Nguyen and C. W. Pickardt and M. Zhang},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Automated Design of Production Scheduling Heuristics: A
+                  Review},
+  year = 2016,
+  volume = 20,
+  number = 1,
+  pages = {110--124}
+}
+
+ +
+@article{BraSch2005faster,
+  title = {Faster Convergence by Means of Fitness Estimation},
+  author = { J{\"u}rgen Branke  and Schmidt, C.},
+  year = 2005,
+  month = jan,
+  journal = {Soft Computing},
+  volume = 9,
+  number = 1,
+  pages = {13--20},
+  issn = {1432-7643, 1433-7479},
+  doi = {10.1007/s00500-003-0329-4},
+  langid = {english}
+}
+
+ +
+@article{BraZap2016:cor,
+  author = {Roland Braune and G. Z{\"a}pfel},
+  title = {Shifting Bottleneck Scheduling for Total Weighted Tardiness Minimization---A Computational Evaluation of Subproblem and Re-optimization Heuristics},
+  journal = {Computers \& Operations Research},
+  year = 2016,
+  volume = 66,
+  pages = {130--140}
+}
+
+ +
+@article{BranCorrGreSlow2016ejor,
+  author = { J{\"u}rgen Branke  and  Salvatore Corrente  and  Salvatore Greco  and  Roman S{\l}owi{\'n}ski  and Zielniewicz, P.},
+  title = {Using {Choquet} integral as preference model in interactive
+                  evolutionary multiobjective optimization},
+  journal = {European Journal of Operational Research},
+  volume = 250,
+  number = 3,
+  pages = {884--901},
+  year = 2016,
+  doi = {10.1016/j.ejor.2015.10.027}
+}
+
+ +
+@article{BranFarSha2016cgti,
+  author = { J{\"u}rgen Branke  and  Farid, S. S. and Shah, N.},
+  title = {Industry 4.0: a vision for personalized medicine supply
+                  chains?},
+  journal = {Cell and Gene Therapy Insights},
+  year = 2016,
+  volume = 2,
+  number = 2,
+  pages = {263--270},
+  doi = {10.18609/cgti.2016.027}
+}
+
+ +
+@article{BranGreSlow2015,
+  author = { J{\"u}rgen Branke  and  Salvatore Greco  and  Roman S{\l}owi{\'n}ski  and Piotr Zielniewicz},
+  title = {Learning Value Functions in Interactive Evolutionary
+                  Multiobjective Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  volume = 19,
+  pages = {88--102},
+  number = 1
+}
+
+ +
+@article{BranJin2005tec,
+  author = { Yaochu Jin  and  J{\"u}rgen Branke },
+  title = {Evolutionary Optimization in Uncertain Environments---A
+                  Survey},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2005,
+  volume = 9,
+  number = 5,
+  pages = {303--317}
+}
+
+ +
+@article{Breiman2001,
+  author = {Leo Breiman},
+  title = {Random Forests},
+  journal = {Machine Learning},
+  year = 2001,
+  volume = 45,
+  number = 1,
+  pages = {5--32},
+  doi = {10.1023/A:1010933404324}
+}
+
+ +
+@article{BriCabEmm2018maximum,
+  title = {Maximum volume subset selection for anchored boxes},
+  author = { Karl Bringmann  and Cabello, Sergio and  Emmerich, Michael T. M. },
+  journal = {Arxiv preprint arXiv:1803.00849},
+  year = 2018,
+  doi = {10.48550/arXiv.1803.00849},
+  abstract = {Let $B$ be a set of $n$ axis-parallel boxes in $\mathbb{R}^d$
+                  such that each box has a corner at the origin and the other
+                  corner in the positive quadrant of $\mathbb{R}^d$, and let
+                  $k$ be a positive integer.  We study the problem of selecting
+                  $k$ boxes in $B$ that maximize the volume of the union of the
+                  selected boxes.  This research is motivated by applications
+                  in skyline queries for databases and in multicriteria
+                  optimization, where the problem is known as the
+                  \emph{hypervolume subset selection problem}.  It is known
+                  that the problem can be solved in polynomial time in the
+                  plane, while the best known running time in any dimension $d
+                  \ge 3$ is $\Omega\big(\binom{n}{k}\big)$.  We show that: The
+                  problem is NP-hard already in 3 dimensions. In 3 dimensions,
+                  we break the bound $\Omega\big(\binom{n}{k}\big)$, by
+                  providing an $n^{O(\sqrt{k})}$ algorithm. For any constant
+                  dimension $d$, we present an efficient polynomial-time
+                  approximation scheme.},
+  keywords = {hypervolume subset selection}
+}
+
+ +
+@article{BriFri2012tcs,
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  title = {Approximating the Least Hypervolume Contributor: {NP}-Hard in
+                  General, But Fast in Practice},
+  pages = {104--116},
+  year = 2012,
+  volume = 425,
+  journal = {Theoretical Computer Science},
+  doi = {10.1016/j.tcs.2010.09.026}
+}
+
+ +
+@article{BriFri2010eff,
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  title = {An efficient algorithm for computing hypervolume
+                  contributions},
+  journal = {Evolutionary Computation},
+  volume = 18,
+  number = 3,
+  pages = {383--402},
+  year = 2010
+}
+
+ +
+@article{BriFri2014convergence,
+  title = {Convergence of hypervolume-based archiving algorithms},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2014,
+  number = 5,
+  pages = {643--657},
+  volume = 18,
+  publisher = {IEEE},
+  keywords = {competitive ratio},
+  doi = {10.1109/TEVC.2014.2341711},
+  annote = {Proof that all nondecreasing $(\mu + \lambda)$ archiving algorithms with
+                  $\lambda < \mu$ are ineffective.}
+}
+
+ +
+@article{Bro1970bfgs,
+  author = {Broyden, Charles G.},
+  title = {The Convergence of a Class of Double-rank Minimization
+                  Algorithms: 2. The New Algorithm},
+  journal = {IMA Journal of Applied Mathematics},
+  year = 1970,
+  volume = 6,
+  number = 3,
+  pages = {222--231},
+  month = sep,
+  annote = {One of the four papers that proposed BFGS.},
+  doi = {10.1093/imamat/6.3.222},
+  eprint = {https://academic.oup.com/imamat/article-pdf/6/3/222/1848059/6-3-222.pdf},
+  keywords = {BFGS}
+}
+
+ +
+@article{BroBadThiZit2013directed,
+  title = {Directed Multiobjective Optimization Based on the Weighted
+                  Hypervolume Indicator},
+  volume = 20,
+  doi = {10.1002/mcda.1502},
+  abstract = {Recently, there has been a large interest in set-based
+                  evolutionary algorithms for multi objective
+                  optimization. They are based on the definition of indicators
+                  that characterize the quality of the current population while
+                  being compliant with the concept of Pareto-optimality. It has
+                  been shown that the hypervolume indicator, which measures the
+                  dominated volume in the objective space, enables the design
+                  of efficient search algorithms and, at the same time, opens
+                  up opportunities to express user preferences in the search by
+                  means of weight functions. The present paper contains the
+                  necessary theoretical foundations and corresponding
+                  algorithms to (i) select appropriate weight functions, to
+                  (ii) transform user preferences into weight functions and to
+                  (iii) efficiently evaluate the weighted hypervolume indicator
+                  through Monte Carlo sampling. The algorithm W-HypE, which
+                  implements the previous concepts, is introduced, and the
+                  effectiveness of the search, directed towards the user's
+                  preferred solutions, is shown using an extensive set of
+                  experiments including the necessary statistical performance
+                  assessment.},
+  number = {5-6},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  author = { Dimo Brockhoff  and  Johannes Bader  and  Lothar Thiele  and  Eckart Zitzler },
+  year = 2013,
+  keywords = {hypervolume, preference-based search, multi objective
+                  optimization, evolutionary algorithm},
+  pages = {291--317}
+}
+
+ +
+@article{BroCorFre2010tutorial,
+  author = {Brochu, Eric and Cora, Vlad and  Nando de Freitas },
+  year = 2010,
+  month = dec,
+  title = {A Tutorial on {Bayesian} Optimization of Expensive Cost
+                  Functions, with Application to Active User Modeling and
+                  Hierarchical Reinforcement Learning},
+  journal = {Arxiv preprint arXiv:1012.2599},
+  url = {https://arxiv.org/abs/1012.2599}
+}
+
+ +
+@article{BroTusTusWag2016biobj,
+  author = { Dimo Brockhoff  and  Tea Tu{\v s}ar  and Dejan Tu{\v s}ar and  Tobias Wagner  and  Nikolaus Hansen  and  Anne Auger },
+  title = {Biobjective performance assessment with the {COCO} platform},
+  journal = {Arxiv preprint arXiv:1605.01746},
+  year = 2016,
+  doi = {10.48550/arXiv.1605.01746}
+}
+
+ +
+@article{BroWagTrau2015r2,
+  title = {{R2} indicator-based multiobjective search},
+  author = { Dimo Brockhoff  and  Tobias Wagner  and  Heike Trautmann },
+  journal = {Evolutionary Computation},
+  year = 2015,
+  number = 3,
+  pages = {369--395},
+  volume = 23
+}
+
+ +
+@article{BroZit2009ec,
+  author = { Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Objective Reduction in Evolutionary Multiobjective
+                  Optimization: Theory and Applications},
+  journal = {Evolutionary Computation},
+  volume = 17,
+  number = 2,
+  pages = {135--166},
+  year = 2009,
+  abstract = {Many-objective problems represent a major challenge in the
+                  field of evolutionary multiobjective optimization, in terms of
+                  search efficiency, computational cost, decision making,
+                  visualization, and so on. This leads to various research
+                  questions, in particular whether certain objectives can be
+                  omitted in order to overcome or at least diminish the
+                  difficulties that arise when many, that is, more than three,
+                  objective functions are involved. This study addresses this
+                  question from different perspectives. First, we investigate
+                  how adding or omitting objectives affects the problem
+                  characteristics and propose a general notion of conflict
+                  between objective sets as a theoretical foundation for
+                  objective reduction. Second, we present both exact and
+                  heuristic algorithms to systematically reduce the number of
+                  objectives, while preserving as much as possible of the
+                  dominance structure of the underlying optimization
+                  problem. Third, we demonstrate the usefulness of the proposed
+                  objective reduction method in the context of both decision
+                  making and search for a radar waveform application as well as
+                  for well-known test functions.},
+  doi = {10.1162/evco.2009.17.2.135}
+}
+
+ +
+@article{Broyden1970bfgs,
+  author = {Broyden, C. G.},
+  title = {The Convergence of a Class of Double-rank Minimization
+                  Algorithms 1. General Considerations},
+  journal = {IMA Journal of Applied Mathematics},
+  year = 1970,
+  volume = 6,
+  number = 1,
+  pages = {76--90},
+  month = mar,
+  keywords = {Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm},
+  abstract = {This paper presents a more detailed analysis of a class of
+                  minimization algorithms, which includes as a special case the
+                  DFP (Davidon-Fletcher-Powell) method, than has previously
+                  appeared. Only quadratic functions are considered but
+                  particular attention is paid to the magnitude of successive
+                  errors and their dependence upon the initial matrix. On the
+                  basis of this a possible explanation of some of the observed
+                  characteristics of the class is tentatively suggested.},
+  doi = {10.1093/imamat/6.1.76}
+}
+
+ +
+@article{BruHurWer1996,
+  author = {Peter Brucker and Johann Hurink and Frank Werner},
+  title = {Improving Local Search Heuristics for some Scheduling Problems --- {Part} {I}},
+  journal = {Discrete Applied Mathematics},
+  year = 1996,
+  volume = 65,
+  number = {1--3},
+  pages = {97--122}
+}
+
+ +
+@article{BruHurWer1997,
+  author = {Peter Brucker and Johann Hurink and Frank Werner},
+  title = {Improving Local Search Heuristics for some Scheduling Problems --- {Part} {II}},
+  journal = {Discrete Applied Mathematics},
+  year = 1997,
+  volume = 72,
+  number = {1--2},
+  pages = {47--69}
+}
+
+ +
+@article{BruJacTho1999:aor,
+  author = {M. J. Brusco and L. W. Jacobs and G. M. Thompson},
+  title = {A Morphing Procedure to Supplement a Simulated
+                  Annealing Heuristic for Cost- and
+                  Coverage-correlated Set Covering Problems},
+  journal = {Annals of Operations Research},
+  year = 1999,
+  volume = 86,
+  pages = {611--627}
+}
+
+ +
+@article{Buc1994jors,
+  title = {An experimental evaluation of interactive {MCDM}
+                  methods and the decision making process},
+  author = {Buchanan, John T.},
+  journal = {Journal of the Operational Research Society},
+  pages = {1050--1059},
+  volume = 45,
+  number = 9,
+  year = 1994
+}
+
+ +
+@article{BucCor1997anchoring,
+  title = {The effects of anchoring in interactive {MCDM} solution
+                  methods},
+  volume = 24,
+  doi = {10.1016/S0305-0548(97)00014-2},
+  number = 10,
+  journal = {Computers \& Operations Research},
+  author = {Buchanan, John T. and Corner, James},
+  month = oct,
+  year = 1997,
+  pages = {907--918}
+}
+
+ +
+@article{BucGoo2004maxima,
+  author = {A. L. Buchsbaum and M. T. Goodrich},
+  title = {Three-Dimensional Layers of Maxima},
+  journal = {Algorithmica},
+  year = 2004,
+  volume = 39,
+  pages = {275--289},
+  alias = {Buchsbaum04}
+}
+
+ +
+@article{BulHarStr99:aor,
+  author = { B. Bullnheimer  and  Richard F. Hartl  and  Christine Strauss },
+  title = {An Improved Ant System Algorithm for the Vehicle Routing
+                  Problem},
+  journal = {Annals of Operations Research},
+  year = 1999,
+  volume = 89,
+  pages = {319--328},
+  alias = {ASrank99}
+}
+
+ +
+@article{BulHarStr99:cejore,
+  author = { B. Bullnheimer  and  Richard F. Hartl  and  Christine Strauss },
+  title = {A new rank-based version of the {Ant} {System}: {A}
+                  computational study},
+  journal = {Central European Journal for Operations Research and Economics},
+  year = 1999,
+  volume = 7,
+  number = 1,
+  pages = {25--38},
+  alias = {BulHarStr99:ras}
+}
+
+ +
+@article{BurByk2017,
+  title = {The Late Acceptance Hill-Climbing Heuristic},
+  author = { Edmund K. Burke  and  Yuri Bykov },
+  journal = {European Journal of Operational Research},
+  volume = 258,
+  number = 1,
+  pages = {70--78},
+  year = 2017
+}
+
+ +
+@article{BurFin1983,
+  title = {The asymptotic probabilistic behaviour of quadratic sum assignment problems},
+  author = { Burkard, Rainer E.  and Fincke, Ulrich},
+  journal = {Zeitschrift f{\"u}r Operations Research},
+  volume = 27,
+  number = 1,
+  pages = {73--81},
+  year = 1983,
+  publisher = {Springer}
+}
+
+ +
+@article{BurFraMos2004:joh,
+  author = {Luciana Buriol and Paulo M. Fran{\c c}a and  Pablo Moscato },
+  title = {A New Memetic Algorithm for the Asymmetric Traveling Salesman Problem},
+  journal = {Journal of Heuristics},
+  year = 2004,
+  volume = 10,
+  number = 5,
+  pages = {483--506}
+}
+
+ +
+@article{BurGenHyd2013,
+  author = { Edmund K. Burke  and  Michel Gendreau  and  Matthew R. Hyde  and  Graham Kendall  and  Gabriela Ochoa  and  Ender {\"O}zcan  and  Rong Qu },
+  title = {Hyper-heuristics: A Survey of the State of the Art},
+  journal = {Journal of the Operational Research Society},
+  year = 2013,
+  volume = 64,
+  number = 12,
+  pages = {1695--1724},
+  doi = {10.1057/jors.2013.71}
+}
+
+ +
+@article{BurHydKen2010tec,
+  author = { Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall  and John R. Woodward},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {A Genetic Programming Hyper-Heuristic Approach for Evolving
+                  {2-D} Strip Packing Heuristics},
+  year = 2010,
+  volume = 14,
+  number = 6,
+  pages = {942--958},
+  doi = {10.1109/TEVC.2010.2041061}
+}
+
+ +
+@article{BurHydKen2012ec,
+  title = {Automating the Packing Heuristic Design Process with Genetic
+                  Programming},
+  author = { Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall  and John R. Woodward},
+  doi = {10.1162/evco_a_00044},
+  year = 2012,
+  volume = 20,
+  number = 1,
+  pages = {63--89},
+  journal = {Evolutionary Computation},
+  keywords = {one-, two-, or three-dimensional knapsack and bin packing
+                  problems}
+}
+
+ +
+@article{BurHydKen2012tec,
+  author = { Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall },
+  title = {Grammatical Evolution of Local Search Heuristics},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 16,
+  number = 7,
+  year = 2012,
+  pages = {406--417},
+  doi = {10.1109/TEVC.2011.2160401}
+}
+
+ +
+@article{BurKarRen1997,
+  title = {{QAPLIB}--a Quadratic Assignment Problem Library},
+  author = { Burkard, Rainer E.  and  Stefan E. Karisch  and  Franz Rendl },
+  journal = {Journal of Global Optimization},
+  volume = 10,
+  number = 4,
+  pages = {391--403},
+  year = 1997,
+  publisher = {Springer}
+}
+
+ +
+@article{BurRen1984,
+  author = { Burkard, Rainer E.  and  Franz Rendl },
+  title = {A Thermodynamically Motivated Simulation Procedure for
+                  Combinatorial Optimization Problems},
+  journal = {European Journal of Operational Research},
+  year = 1984,
+  volume = 17,
+  number = 2,
+  pages = {169--174},
+  doi = {10.1016/0377-2217(84)90231-5},
+  keywords = {2-exchange delta evaluation for QAP}
+}
+
+ +
+@article{BusRobTot2014,
+  author = {Erika Buson and Roberto Roberti and  Paolo Toth },
+  title = {A Reduced-Cost Iterated Local Search Heuristic for the Fixed-Charge Transportation Problem},
+  journal = {Operations Research},
+  year = 2014,
+  volume = 62,
+  number = 5,
+  pages = {1095--1106}
+}
+
+ +
+@article{CabLuqMol2002promoin,
+  author = {Caballero, R. and  Mariano Luque  and  Molina, Juli{\'a}n  and  Francisco Ruiz },
+  title = {{PROMOIN}: An Interactive System for Multiobjective
+                  Programming},
+  journal = {Information Technologies and Decision Making},
+  year = 2002,
+  volume = 1,
+  pages = {635--656},
+  keywords = {preferences, multi interactive methods framework}
+}
+
+ +
+@article{CacStu2017:endm,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle },
+  title = {Exploring Variable Neighborhood Search for Automatic
+                  Algorithm Configuration},
+  journal = {Electronic Notes in Discrete Mathematics},
+  year = 2017,
+  volume = 58,
+  pages = {167--174},
+  doi = {10.1016/j.endm.2017.03.022}
+}
+
+ +
+@article{CahMelTal2004,
+  author = {Cahon, Sebastien and Melab, Nordine and  Talbi, El-Ghazali },
+  title = {{ParadisEO}: A Framework for the Reusable Design of
+                  Parallel and Distributed Metaheuristics},
+  doi = {10.1023/B:HEUR.0000026900.92269.ec},
+  journal = {Journal of Heuristics},
+  number = 3,
+  pages = {357--380},
+  volume = 10,
+  year = 2004,
+  alias = {CahMelTal04paradiseo}
+}
+
+ +
+@article{CaiHuaQuiMa2009,
+  author = {Zhaoquan Cai and Han Huang and Yong Qin and Xianheng
+                  Ma},
+  title = {Ant Colony Optimization Based on Adaptive Volatility
+                  Rate of Pheromone Trail},
+  year = 2009,
+  journal = {International Journal of Communications, Network and System Sciences},
+  volume = 2,
+  number = 8,
+  pages = {792--796}
+}
+
+ +
+@article{CaiLiFan2015archive,
+  title = {An external archive guided multiobjective evolutionary
+                  algorithm based on decomposition for combinatorial
+                  optimization},
+  author = {Cai, Xinye and Li, Yexing and Fan, Zhun and  Zhang, Qingfu },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  number = 4,
+  pages = {508--523},
+  volume = 19
+}
+
+ +
+@article{CaiXiaLiHu2021grid,
+  title = {A grid-based inverted generational distance for
+                  multi/many-objective optimization},
+  author = {Cai, Xinye and Xiao, Yushun and  Li, Miqing  and Hu,
+                  Han and  Ishibuchi, Hisao  and Li, Xiaoping},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2021,
+  number = 1,
+  pages = {21--34},
+  volume = 25,
+  publisher = {IEEE},
+  annote = {weakly Pareto-compliant indicator}
+}
+
+ +
+@article{CaiXiaLiSun2021kernel,
+  title = {A kernel-based indicator for multi/many-objective
+                  optimization},
+  author = {Cai, Xinye and Xiao, Yushun and Li, Zhenhua and Sun, Qi and
+                  Xu, Hanchuan and  Li, Miqing  and  Ishibuchi, Hisao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2021
+}
+
+ +
+@article{Calvo00tsptw,
+  author = { Roberto {Wolfler Calvo} },
+  title = {A New Heuristic for the Traveling Salesman Problem
+                  with Time Windows},
+  journal = {Transportation Science},
+  volume = 34,
+  number = 1,
+  year = 2000,
+  pages = {113--124},
+  doi = {10.1287/trsc.34.1.113.12284},
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{CamBatAra2020jss,
+  title = {The \rpackage{MOEADr} Package: A Component-Based Framework for Multiobjective Evolutionary Algorithms Based on Decomposition},
+  author = {Felipe Campelo and Lucas S. Batista and Claus Aranha},
+  year = 2020,
+  volume = 92,
+  issue = 6,
+  journal = {Journal of Statistical Software},
+  doi = {10.18637/jss.v092.i06}
+}
+
+ +
+@article{CamDorStu2019si,
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {The intelligent water drops algorithm: why it cannot be considered a novel algorithm},
+  journal = {Swarm Intelligence},
+  year = 2019,
+  volume = 13,
+  pages = {173--192}
+}
+
+ +
+@article{CamDorStu2022cuckooSearch,
+  title = {An analysis of why cuckoo search does not bring any novel ideas to optimization},
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Marco Dorigo  and  Thomas St{\"u}tzle },
+  journal = {Computers \& Operations Research},
+  pages = {105747},
+  year = 2022
+}
+
+ +
+@article{CamDorStu2022exposing,
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {Exposing the grey wolf, moth-flame, whale, firefly, bat, and antlion algorithms: six misleading optimization techniques inspired by bestial metaphors},
+  journal = {International Transactions in Operational Research},
+  doi = {10.1111/itor.13176},
+  year = 2022
+}
+
+ +
+@article{CamJon2011,
+  title = {Prepositioning supplies in preparation for disasters},
+  author = {Campbell, Ann Melissa and Jones, Philip C.},
+  journal = {European Journal of Operational Research},
+  volume = 209,
+  number = 2,
+  pages = {156--165},
+  year = 2011,
+  publisher = {Elsevier}
+}
+
+ +
+@article{CamSchXia2013sentim,
+  author = {Cambria, E and Schuller, B and Xia, Y and Havasi, C},
+  year = 2013,
+  title = {New avenues in opinion mining and sentiment analysis},
+  journal = {IEEE Intelligent Systems},
+  volume = 28,
+  number = 2,
+  pages = {15--21},
+  doi = {10.1109/MIS.2013.30}
+}
+
+ +
+@article{CamStuDor2021psox,
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {{PSO-X}: A Component-Based Framework for the Automatic Design
+                  of Particle Swarm Optimization Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2021,
+  volume = 26,
+  number = 3,
+  pages = {402--416},
+  doi = {10.1109/TEVC.2021.3102863}
+}
+
+ +
+@article{CamWan2020sample,
+  author = {Felipe Campelo and  Wanner, Elizabeth F. },
+  title = {Sample size calculations for the experimental comparison of
+                  multiple algorithms on multiple problem instances},
+  journal = {Journal of Heuristics},
+  year = 2020,
+  volume = 26,
+  number = 6,
+  pages = {851--883},
+  doi = {10.1007/s10732-020-09454-w}
+}
+
+ +
+@article{CaoJiaZhan2017its,
+  author = {Cao, Z. and Jiang, S. and Zhang, J. and Guo, H.},
+  title = {A unified framework for vehicle rerouting and traffic light
+                  control to reduce traffic congestion},
+  journal = {IEEE Transactions on Intelligent Transportation Systems},
+  pages = {1958--1973},
+  volume = 18,
+  number = 7,
+  year = 2017
+}
+
+ +
+@article{Cap2017,
+  title = {Variable Neighborhood Search for Extremal Vertices : The {AutoGraphiX-III} System},
+  author = {Gilles Caporossi},
+  journal = {Computers \& Operations Research},
+  volume = 78,
+  pages = {431--438},
+  year = 2017
+}
+
+ +
+@article{Car1982:ejor,
+  author = {J. Carlier},
+  title = {The One-machine Sequencing Problem},
+  journal = {European Journal of Operational Research},
+  year = 1982,
+  volume = 11,
+  number = 1,
+  pages = {42--47}
+}
+
+ +
+@article{CarBar96tabusearch_tsptw,
+  author = { William B. Carlton  and  J. Wesley Barnes },
+  title = {Solving the traveling-salesman problem with time
+                  windows using tabu search},
+  journal = {IIE Transactions},
+  year = 1996,
+  volume = 28,
+  pages = {617--629}
+}
+
+ +
+@article{CarKonCor2019inf,
+  author = {Fabio Caraffini and  Anna V. Kononova  and  David Corne },
+  title = {Infeasibility and structural bias in differential evolution},
+  journal = {Information Sciences},
+  volume = 496,
+  pages = {161--179},
+  year = 2019,
+  doi = {10.1016/j.ins.2019.05.019}
+}
+
+ +
+@article{CasLab99,
+  title = {Heuristics for large constrained vehicle routing problems},
+  author = {Caseau, Yves and Laburthe, Fran{\c{c}}ois},
+  journal = {Journal of Heuristics},
+  volume = 5,
+  number = 3,
+  pages = {281--303},
+  year = 1999
+}
+
+ +
+@article{CasSilLab2001tplp,
+  author = {Yves Caseau and Glenn Silverstein and Fran{\c{c}}ois
+                  Laburthe},
+  title = {Learning Hybrid Algorithms for Vehicle Routing
+                  Problems},
+  journal = {Theory and Practice of Logic Programming},
+  volume = 1,
+  number = 6,
+  year = 2001,
+  pages = {779--806},
+  epub = {http://arxiv.org/abs/cs/0405092}
+}
+
+ +
+@article{CatAbsFeiVig2014:cor,
+  author = {Diego Cattaruzza and Nabil Absi and Dominique Feillet and  Vigo, Daniele },
+  title = {An Iterated Local Search for the Multi-commodity Multi-trip Vehicle Routing Problem with Time Windows},
+  journal = {Computers \& Operations Research},
+  year = 2014,
+  volume = 51,
+  pages = {257--267}
+}
+
+ +
+@article{CauNiePok2012,
+  title = {Optimization models in emergency logistics: A literature review},
+  author = {Caunhye, Aakil M. and Nie, Xiaofeng and Pokharel, Shaligram},
+  journal = {Socio-Economic Planning Sciences},
+  volume = 46,
+  number = 1,
+  pages = {4--13},
+  year = 2012,
+  publisher = {Elsevier}
+}
+
+ +
+@article{CebIruMen2014eda,
+  author = { Josu Ceberio  and  Irurozki, Ekhine  and  Alexander Mendiburu  and  Jos{\'e} A. Lozano },
+  title = {A distance-based ranking model estimation of distribution
+                  algorithm for the flowshop scheduling problem},
+  abstract = {The aim of this paper is two-fold. First, we introduce a
+                  novel general estimation of distribution algorithm to deal
+                  with permutation-based optimization problems. The algorithm
+                  is based on the use of a probabilistic model for permutations
+                  called the generalized Mallows model. In order to prove the
+                  potential of the proposed algorithm, our second aim is to
+                  solve the permutation flowshop scheduling problem. A hybrid
+                  approach consisting of the new estimation of distribution
+                  algorithm and a variable neighborhood search is
+                  proposed. Conducted experiments demonstrate that the proposed
+                  algorithm is able to outperform the state-of-the-art
+                  approaches. Moreover, from the 220 benchmark instances
+                  tested, the proposed hybrid approach obtains new best known
+                  results in 152 cases. An in-depth study of the results
+                  suggests that the successful performance of the introduced
+                  approach is due to the ability of the generalized Mallows
+                  estimation of distribution algorithm to discover promising
+                  regions in the search space.},
+  doi = {10.1109/TEVC.2013.2260548},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  keywords = {Estimation of distribution algorithms,Generalized Mallows
+                  model,Permutation flowshop scheduling
+                  problem,Permutations-based optimization problems},
+  number = 2,
+  pages = {286--300},
+  volume = 18,
+  year = 2014
+}
+
+ +
+@article{Cer85,
+  author = {Vladim{\'i}r \v{C}ern\'y},
+  title = {A Thermodynamical Approach to the Traveling Salesman Problem: An Efficient Simulation Algorithm},
+  journal = {Journal of Optimization Theory and Applications},
+  year = 1985,
+  volume = 45,
+  number = 1,
+  pages = {41--51}
+}
+
+ +
+@article{CesDiGSch2012,
+  title = {Design, Engineering, and Experimental Analysis of a Simulated Annealing Approach to the Post-Enrolment Course Timetabling Problem},
+  author = { Sara Ceschia  and Luca {Di Gaspero} and Andrea Schaerf},
+  journal = {Computers \& Operations Research},
+  volume = 39,
+  number = 7,
+  pages = {1615--1624},
+  year = 2012,
+  publisher = {Elsevier}
+}
+
+ +
+@article{CesSch2012aim,
+  author = { Sara Ceschia  and Andrea Schaerf},
+  title = {Modeling and solving the dynamic patient admission scheduling problem under uncertainty},
+  journal = {Artificial Intelligence in Medicine},
+  volume = 56,
+  number = 3,
+  pages = {199--205},
+  year = 2012,
+  doi = {10.1016/j.artmed.2012.09.001},
+  keywords = {F-race}
+}
+
+ +
+@article{CesSchStu2013:cie,
+  author = { Sara Ceschia  and Andrea Schaerf and  Thomas St{\"u}tzle },
+  title = {Local Search Techniques for a Routing-packing Problem},
+  journal = {Computers and Industrial Engineering},
+  year = 2013,
+  volume = 66,
+  number = 4,
+  pages = {1138--1149}
+}
+
+ +
+@article{ChaMeaBea2000cor,
+  author = {T.-J. Chang and N. Meade and  John E. Beasley  and Y. M. Sharaiha},
+  title = {Heuristics for cardinality constrained portfolio
+                  optimisation},
+  journal = {Computers \& Operations Research},
+  year = 2000,
+  volume = 27,
+  number = 13,
+  pages = {1271--1302},
+  keywords = {Portfolio optimisation, CCMVPOP, Efficient frontier},
+  abstract = {In this paper we consider the problem of finding the
+                  efficient frontier associated with the standard mean-variance
+                  portfolio optimisation model. We extend the standard model to
+                  include cardinality constraints that limit a portfolio to
+                  have a specified number of assets, and to impose limits on
+                  the proportion of the portfolio held in a given asset (if any
+                  of the asset is held). We illustrate the differences that
+                  arise in the shape of this efficient frontier when such
+                  constraints are present. We present three heuristic
+                  algorithms based upon genetic algorithms, tabu search and
+                  simulated annealing for finding the cardinality constrained
+                  efficient frontier. Computational results are presented for
+                  five data sets involving up to 225 assets.  Scope and purpose
+                  The standard Markowitz mean-variance approach to portfolio
+                  selection involves tracing out an efficient frontier, a
+                  continuous curve illustrating the tradeoff between return and
+                  risk (variance). This frontier can be easily found via
+                  quadratic programming. This approach is well-known and widely
+                  applied. However, for practical purposes, it may be desirable
+                  to limit the number of assets in a portfolio, as well as
+                  imposing limits on the proportion of the portfolio devoted to
+                  any particular asset. If such constraints exist, the problem
+                  of finding the efficient frontier becomes much harder. This
+                  paper illustrates how, in the presence of such constraints,
+                  the efficient frontier becomes discontinuous. Three heuristic
+                  techniques are applied to the problem of finding this
+                  efficient frontier and computational results presented for a
+                  number of data sets which are made publicly available.}
+}
+
+ +
+@article{ChaWag2015many,
+  author = {Shelvin Chand and  Markus Wagner },
+  title = {Evolutionary many-objective optimization: A quick-start
+                  guide},
+  journal = {Surveys in Operations Research and Management Science},
+  volume = 20,
+  number = 2,
+  pages = {35--42},
+  year = 2015,
+  doi = {10.1016/j.sorms.2015.08.001}
+}
+
+ +
+@article{Chase93,
+  author = { Donald V. Chase  and  Lindell E. Ormsbee },
+  title = {Computer-generated pumping schedules for satisfying
+                  operation objectives},
+  journal = {J. Am. Water Works Assoc.},
+  year = 1993,
+  volume = 85,
+  number = 7,
+  pages = {54--61}
+}
+
+ +
+@article{ChauDeb2010asoft,
+  title = {An interactive evolutionary multi-objective optimization and
+                  decision making procedure},
+  author = {Chaudhuri, Shamik and  Kalyanmoy Deb },
+  journal = {Applied Soft Computing},
+  volume = 10,
+  number = 2,
+  pages = {496--511},
+  year = 2010
+}
+
+ +
+@article{CheChiSto2012,
+  title = {Business Intelligence and Analytics: From Big Data to Big
+                  Impact},
+  author = {Chen, Hsinchun and Chiang, Roger H. L. and Storey, Veda C.},
+  journal = {MIS Quarterly},
+  volume = 36,
+  number = 4,
+  pages = {1165--1188},
+  year = 2012
+}
+
+ +
+@article{CheHaoGlo2016,
+  title = {A hybrid metaheuristic approach for the capacitated arc
+                  routing problem},
+  author = {Yuning Chen and  Jin-Kao Hao  and  Fred Glover },
+  journal = {European Journal of Operational Research},
+  volume = 553,
+  number = 1,
+  pages = {25--39},
+  year = 2016,
+  doi = {10.1016/j.ejor.2016.02.015},
+  keywords = {irace}
+}
+
+ +
+@article{CheHsi2014,
+  author = {Chen, Ruey-Maw and Hsieh, Fu-Ren},
+  title = {An exchange local search heuristic based scheme for
+                  permutation flow shop problems},
+  journal = {Applied Mathematics \& Information Sciences},
+  volume = 8,
+  number = 1,
+  pages = {209--215},
+  year = 2014
+}
+
+ +
+@article{CheLi1999gen,
+  author = { F. Y.  Cheng and X. S.  Li },
+  title = {Generalized center method for multiobjective engineering
+                  optimization},
+  journal = {Engineering Optimization},
+  volume = 31,
+  number = 5,
+  pages = {641--661},
+  year = 1999,
+  doi = {10.1080/03052159908941390}
+}
+
+ +
+@article{CheLiYao2017dynamic,
+  author = {Chen, Renzhi and Li, Ke and  Xin Yao },
+  title = {Dynamic Multiobjectives Optimization With a Changing Number
+                  of Objectives},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2017,
+  volume = 22,
+  number = 1,
+  pages = {157--171},
+  annote = {two co-evolving populations (two archive)},
+  doi = {10.1109/TEVC.2017.2669638}
+}
+
+ +
+@article{CheSia2000ejor,
+  title = {Tabu search applied to global optimization},
+  author = {Rachid Chelouah and Patrick Siarry},
+  journal = {European Journal of Operational Research},
+  volume = 123,
+  number = 2,
+  pages = {256--270},
+  year = 2000
+}
+
+ +
+@article{ChenChenGon2014tcyb,
+  title = {An evolutionary algorithm with double-level archives for
+                  multiobjective optimization},
+  author = {Chen, Ni and Chen, Wei-Neng and Gong, Yue-Jiao and Zhan,
+                  Zhi-Hui and Zhang, Jun and Li, Yun and Tan, Yu-Song},
+  journal = {IEEE Transactions on Cybernetics},
+  year = 2014,
+  number = 9,
+  pages = {1851--1863},
+  volume = 45,
+  publisher = {IEEE}
+}
+
+ +
+@article{ChengMao07:acs_tsptw,
+  author = {Chin-Bin Cheng and Chun-Pin Mao},
+  title = {A modified ant colony system for solving the
+                  travelling salesman problem with time windows},
+  journal = {Mathematical and Computer Modelling},
+  year = 2007,
+  volume = 46,
+  pages = {1225--1235},
+  doi = {10.1016/j.mcm.2006.11.035}
+}
+
+ +
+@article{ChiBirSocRos2006jos,
+  author = { Marco Chiarandini  and  Mauro Birattari  and  Krzysztof Socha  and  O. Rossi-Doria },
+  title = {An Effective Hybrid Algorithm for University Course
+                  Timetabling},
+  journal = {Journal of Scheduling},
+  year = 2006,
+  volume = 9,
+  pages = {403--432},
+  number = 5,
+  month = oct,
+  doi = {10.1007/s10951-006-8495-8},
+  keywords = {2003 international timetabling competition, F-race}
+}
+
+ +
+@article{ChiCorDamBau11,
+  author = {Chica, Manuel and  Oscar Cord{\'o}n  and Damas,
+                  Sergio and Bautista, Joaqu{\'i}n},
+  year = 2011,
+  issn = {1865-9284},
+  journal = {Memetic Computing},
+  volume = 3,
+  number = 1,
+  title = {A New Diversity Induction Mechanism for a
+                  Multi-objective Ant Colony Algorithm to Solve a
+                  Real-world time and Space Assembly Line Balancing
+                  Problem},
+  pages = {15--24}
+}
+
+ +
+@article{ChiUlySha2016,
+  author = {D. S. Chivilikhin and V. I. Ulyantsev and A. A. Shalyto},
+  title = {Modified ant colony algorithm for constructing finite state machines from execution scenarios and temporal formulas},
+  journal = {Automation and Remote Control},
+  year = 2016,
+  volume = 77,
+  number = 3,
+  pages = {473--484},
+  doi = {10.1134/S0005117916030097},
+  keywords = {irace}
+}
+
+ +
+@article{Chicano2011ecj,
+  author = { Chicano, Francisco  and  Darrell Whitley  and  Alba, Enrique },
+  title = {A Methodology to Find the Elementary Landscape Decomposition
+                  of Combinatorial Optimization Problems},
+  journal = {Evolutionary Computation},
+  year = 2011,
+  volume = 19,
+  pages = {597--637},
+  number = 4
+}
+
+ +
+@article{Chicano2012aml,
+  author = { Chicano, Francisco  and  Gabriel J. Luque  and  Alba, Enrique },
+  title = {Autocorrelation Measures for the Quadratic Assignment
+                  Problem},
+  journal = {Applied Mathematics Letters},
+  year = 2012,
+  volume = 25,
+  pages = {698--705},
+  doi = {10.1016/j.aml.2011.09.053}
+}
+
+ +
+@article{ChrMinTot81tsptw,
+  author = { Christofides, Nicos  and A. Mingozzi and  Paolo Toth },
+  title = {State-space relaxation procedures for the
+                  computation of bounds to routing problems},
+  volume = 11,
+  number = 2,
+  journal = {Networks},
+  year = 1981,
+  pages = {145--164},
+  doi = {10.1002/net.3230110207},
+  anote = {makespan optimization}
+}
+
+ +
+@article{ChuJinMie2018surrogate,
+  author = { Tinkle Chugh  and  Yaochu Jin  and  Kaisa Miettinen  and Hakanen,
+                  Jussi and Sindhya, Karthik},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {A Surrogate-Assisted Reference Vector Guided Evolutionary
+                  Algorithm for Computationally Expensive Many-Objective
+                  Optimization},
+  year = 2018,
+  volume = 22,
+  number = 1,
+  pages = {129--142},
+  month = feb
+}
+
+ +
+@article{ChuSinHak2019surv,
+  author = { Tinkle Chugh  and Sindhya, Karthik and Hakanen, Jussi and  Kaisa Miettinen },
+  title = {A survey on handling computationally expensive multiobjective
+                  optimization problems with evolutionary algorithms},
+  journal = {Soft Computing},
+  pages = {3137--3166},
+  volume = 23,
+  number = 9,
+  year = 2019,
+  abstract = {Evolutionary algorithms are widely used for solving
+                  multiobjective optimization problems but are often criticized
+                  because of a large number of function evaluations
+                  needed. Approximations, especially function approximations,
+                  also referred to as surrogates or metamodels are commonly
+                  used in the literature to reduce the computation time. This
+                  paper presents a survey of 45 different recent algorithms
+                  proposed in the literature between 2008 and 2016 to handle
+                  computationally expensive multiobjective optimization
+                  problems. Several algorithms are discussed based on what kind
+                  of an approximation such as problem, function or fitness
+                  approximation they use. Most emphasis is given to function
+                  approximation-based algorithms. We also compare these
+                  algorithms based on different criteria such as metamodeling
+                  technique and evolutionary algorithm used, type and
+                  dimensions of the problem solved, handling constraints,
+                  training time and the type of evolution control. Furthermore,
+                  we identify and discuss some promising elements and major
+                  issues among algorithms in the literature related to using an
+                  approximation and numerical settings used. In addition, we
+                  discuss selecting an algorithm to solve a given
+                  computationally expensive multiobjective optimization problem
+                  based on the dimensions in both objective and decision spaces
+                  and the computation budget available.},
+  doi = {10.1007/s00500-017-2965-0}
+}
+
+ +
+@article{CinFerLopAlb2022irace,
+  author = { Christian Cintrano  and  Javier Ferrer  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Alba, Enrique },
+  title = {Hybridization of Evolutionary Operators with Elitist Iterated
+                  Racing for the Simulation Optimization of Traffic Lights
+                  Programs},
+  journal = {Evolutionary Computation},
+  year = 2023,
+  volume = 31,
+  number = 1,
+  pages = {31--51},
+  doi = {10.1162/evco_a_00314},
+  abstract = {In the traffic light scheduling problem the evaluation of
+                  candidate solutions requires the simulation of a process
+                  under various (traffic) scenarios. Thus, good solutions
+                  should not only achieve good objective function values, but
+                  they must be robust (low variance) across all different
+                  scenarios. Previous work has shown that combining IRACE with
+                  evolutionary operators is effective for this task due to the
+                  power of evolutionary operators in numerical optimization. In
+                  this paper, we further explore the hybridization of
+                  evolutionary operators and the elitist iterated racing of
+                  IRACE for the simulation-optimization of traffic light
+                  programs. We review previous works from the literature to
+                  find the evolutionary operators performing the best when
+                  facing this problem to propose new hybrid algorithms. We
+                  evaluate our approach over a realistic case study derived
+                  from the traffic network of Málaga (Spain) with 275 traffic
+                  lights that should be scheduled optimally. The experimental
+                  analysis reveals that the hybrid algorithm comprising IRACE
+                  plus differential evolution offers statistically better
+                  results than the other algorithms when the budget of
+                  simulations is low. In contrast, IRACE performs better than
+                  the hybrids for high simulations budget, although the
+                  optimization time is much longer.},
+  keywords = {irace, Simulation optimization, Uncertainty, Traffic light
+                  planning}
+}
+
+ +
+@article{Clark95,
+  author = { R. M. Clark  and  L. A. Rossman  and  L. J. Wymer },
+  title = {Modeling distribution system water quality:
+                  regulatory implications},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 1995,
+  volume = 121,
+  number = 6,
+  pages = {423--428},
+  note = {}
+}
+
+ +
+@article{CleKen2022tec,
+  author = { Clerc, Maurice  and  J. Kennedy },
+  title = {The particle swarm - explosion, stability, and convergence in
+                  a multidimensional complex space},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 6,
+  number = 1,
+  pages = {58--73},
+  month = feb,
+  year = 2002,
+  doi = {10.1109/4235.985692}
+}
+
+ +
+@article{CocDraBes2020threats,
+  author = {Cockburn, Andy and Dragicevic, Pierre and Besan\c{c}on, Lonni
+                  and Gutwin, Carl},
+  title = {Threats of a Replication Crisis in Empirical Computer
+                  Science},
+  year = 2020,
+  volume = 63,
+  number = 8,
+  doi = {10.1145/3360311},
+  abstract = {Research replication only works if there is confidence built
+                  into the results.},
+  journal = {Communications of the ACM},
+  month = jul,
+  pages = {70--79},
+  numpages = 10
+}
+
+ +
+@article{CodManMarRes96:informs,
+  author = {B. Codenotti and G. Manzini and L. Margara and G. Resta},
+  title = {Perturbation: An Efficient Technique for the
+                  Solution of Very Large Instances of the
+                  {Euclidean} {TSP}},
+  journal = {INFORMS Journal on Computing},
+  year = 1996,
+  volume = 8,
+  number = 2,
+  pages = {125--133}
+}
+
+ +
+@article{Coe2002constraint,
+  author = { Carlos A. {Coello Coello} },
+  title = {Theoretical and numerical constraint-handling techniques used
+                  with evolutionary algorithms: a survey of the state of the
+                  art},
+  journal = {Computer Methods in Applied Mechanics and Engineering},
+  year = 2002,
+  volume = 191,
+  number = {11-12},
+  pages = {1245--1287},
+  doi = {10.1016/S0045-7825(01)00323-1}
+}
+
+ +
+@article{Coello2003tec-emo,
+  title = {Special Issue on {Evolutionary} {Multiobjective}
+                  {Optimization}},
+  author = { Carlos A. {Coello Coello} },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2003,
+  volume = 7,
+  number = 2,
+  alias = {Coello03}
+}
+
+ +
+@article{Coello2006evolutionary,
+  title = {Evolutionary multi-objective optimization: a historical view
+                  of the field},
+  author = { Carlos A. {Coello Coello} },
+  journal = {IEEE Computational Intelligence Magazine},
+  volume = 1,
+  number = 1,
+  pages = {28--36},
+  year = 2006
+}
+
+ +
+@article{CohFie1999:siamo,
+  author = { Harry Cohn  and  Mark J. Fielding },
+  title = {Simulated Annealing: Searching for an Optimal Temperature},
+  journal = {SIAM Journal on Optimization},
+  year = 1999,
+  volume = 9,
+  number = 3,
+  pages = {779--802}
+}
+
+ +
+@article{ColBry05:jwrpm,
+  author = { Andrew F. Colombo  and  Bryan W. Karney },
+  title = {Impacts of Leaks on Energy Consumption in Pumped
+                  Systems with Storage},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2005,
+  volume = 131,
+  number = 2,
+  pages = {146--155},
+  month = mar
+}
+
+ +
+@article{ColDorManTru1994:jorbel,
+  author = { Alberto Colorni  and  Marco Dorigo  and  Vittorio Maniezzo  and M. Trubian},
+  title = {{Ant} {System} for Job-shop Scheduling},
+  journal = {JORBEL --- Belgian Journal of Operations Research, Statistics and Computer Science},
+  year = 1994,
+  volume = 34,
+  number = 1,
+  pages = {39--53},
+  alias = {ColDorManTru94:jorbel}
+}
+
+ +
+@article{ColSchPae2010setting,
+  title = {Setting the Research Agenda in Automated Timetabling: The
+                  Second International Timetabling Competition},
+  author = { Barry McCollum  and Andrea Schaerf and  Ben Paechter  and  Paul McMullan  and  Lewis, Rhyd M. R.  and  Andrew J. Parkes  and Luca {Di Gaspero} and  Rong Qu  and  Edmund K. Burke },
+  doi = {10.1287/ijoc.1090.0320},
+  year = 2010,
+  month = feb,
+  volume = 22,
+  number = 1,
+  pages = {120--130},
+  journal = {{INFORMS}}
+}
+
+ +
+@article{ConPotVel02,
+  author = {Richard K. Congram and  Chris N. Potts  and Steve van de Velde},
+  title = {An Iterated Dynasearch Algorithm for the
+                  Single-Machine Total Weighted Tardiness Scheduling
+                  Problem},
+  journal = {INFORMS Journal on Computing},
+  year = 2002,
+  volume = 14,
+  number = 1,
+  pages = {52--67}
+}
+
+ +
+@article{Connolly1990,
+  author = { David T. Connolly },
+  title = {An Improved Annealing Scheme for the {QAP}},
+  journal = {European Journal of Operational Research},
+  volume = 46,
+  number = 1,
+  pages = {93--100},
+  year = 1990,
+  publisher = {Elsevier}
+}
+
+ +
+@article{CooFar1996mul,
+  author = {Cook, Richard J. and Farewell, Vern T.},
+  year = 1996,
+  title = {Multiplicity Considerations in the Design and Analysis of
+                  Clinical Trials},
+  journal = {Journal of the Royal Statistical Society: Series A},
+  volume = 159,
+  pages = {93--110},
+  annote = {multiplicity; multiple endpoints; multiple treatments;
+                  p-value adjustment; type I error; argues that if results are
+                  intended to be interpreted marginally, there may be no need
+                  for controlling experimentwise error rate}
+}
+
+ +
+@article{CopFleRur2010ordering,
+  title = {Ordering by Weighted Number of Wins Gives a Good Ranking for
+                  Weighted Tournaments},
+  author = {Coppersmith, Don and Fleischer, Lisa K. and Rurda, Atri},
+  doi = {10.1145/1798596.1798608},
+  journal = {{ACM} Transactions on Algorithms },
+  keywords = {Approximation algorithms,Borda's method,feedback arc set
+                  problem,rank aggregation,tournaments},
+  month = jul,
+  number = 3,
+  volume = 6,
+  pages = {1--13},
+  publisher = {ACM},
+  year = 2010
+}
+
+ +
+@article{CorDam2006,
+  author = { Oscar Cord{\'o}n  and Sergio Damas},
+  title = {Image Registration with Iterated Local Search},
+  journal = {Journal of Heuristics},
+  year = 2006,
+  volume = 12,
+  number = {1--2},
+  pages = {73--94}
+}
+
+ +
+@article{CorDanDep2019,
+  author = {Jeroen Corstjens and Nguyen Dang and  Depaire, Beno{\^i}t  and Caris, An and Patrick {De Causmaecker}},
+  title = {A combined approach for analysing heuristic algorithms},
+  journal = {Journal of Heuristics},
+  year = 2019,
+  volume = 25,
+  number = 4,
+  pages = {591--628},
+  doi = {10.1007/s10732-018-9388-7}
+}
+
+ +
+@article{CorDepCarSor2020,
+  title = {A multilevel evaluation method for heuristics with an application to the {VRPTW}},
+  author = {Jeroen Corstjens and  Depaire, Beno{\^i}t  and Caris, An and  Kenneth S{\"o}rensen },
+  journal = {International Transactions in Operational Research},
+  year = 2020,
+  volume = 27,
+  number = 1,
+  pages = {168--196},
+  doi = {10.1111/itor.12631}
+}
+
+ +
+@article{CorKoz04:compu,
+  author = {P. Corry and E. Kozan},
+  title = {Ant Colony Optimisation for Machine Layout Problems},
+  journal = {Computational Optimization and Applications},
+  year = 2004,
+  volume = 28,
+  number = 3,
+  pages = {287--310}
+}
+
+ +
+@article{CorLapMer2001,
+  author = {Jean{-}Fran{\c{c}}ois Cordeau and  Gilbert Laporte  and A. Mercier},
+  title = {A unified tabu search heuristic for vehicle routing problems with time windows},
+  journal = {Journal of the Operational Research Society},
+  year = 2001,
+  volume = 52,
+  number = 8,
+  pages = {928--936}
+}
+
+ +
+@article{CorMai2012:cor,
+  author = {Jean{-}Fran{\c{c}}ois Cordeau and Mirko Maischberger},
+  title = {A Parallel Iterated Tabu Search Heuristic for Vehicle Routing Problems},
+  journal = {Computers \& Operations Research},
+  year = 2012,
+  volume = 39,
+  number = 9,
+  pages = {2033--2050}
+}
+
+ +
+@article{CosGolGol2012:esa,
+  title = {Hybridizing {VNS} and path-relinking on a particle swarm
+                  framework to minimize total flowtime },
+  journal = {Expert Systems with Applications},
+  volume = 39,
+  number = 18,
+  pages = {13118--13126},
+  year = 2012,
+  author = {Wagner Emanoel Costa and  Goldbarg, Marco Cesar  and  Goldbarg, Elizabeth Ferreira Gouv{\^e}a }
+}
+
+ +
+@article{CosHer97:jors,
+  author = {D. Costa and A. Hertz},
+  title = {Ants can color graphs},
+  journal = {Journal of the Operational Research Society},
+  year = 1997,
+  volume = 48,
+  pages = {295--305},
+  alias = {CosHer97}
+}
+
+ +
+@article{CoyGolRunWas2001,
+  author = {S. P. Coy and B. L. Golden and G. C. Runger and E. A. Wasil},
+  title = {Using Experimental Design to Find Effective
+             Parameter Settings for Heuristics},
+  journal = {Journal of Heuristics},
+  year = 2001,
+  volume = 7,
+  number = 1,
+  pages = {77--97}
+}
+
+ +
+@article{Crabtree1995,
+  title = {Resource Scheduling: Comparing Simulated Annealing with Constraint Programming},
+  author = { I. Barry Crabtree },
+  journal = {BT Technology Journal},
+  volume = 13,
+  number = 1,
+  pages = {121--127},
+  year = 1995,
+  publisher = {Springer}
+}
+
+ +
+@article{CriFliVer1996rankings,
+  author = {Critchlow, Douglas Edward and Fligner, Michael A. and
+                  Verducci, Joseph S.},
+  journal = {Journal of Mathematical Psychology},
+  pages = {294--318},
+  title = {Probability Models on Rankings},
+  volume = 35,
+  year = 1991
+}
+
+ +
+@article{Cro1958,
+  author = {G. A. Croes},
+  title = {A Method for Solving Traveling Salesman Problems},
+  journal = {Operations Research},
+  year = 1958,
+  volume = 6,
+  pages = {791--812}
+}
+
+ +
+@article{CroDemMul1978reporting,
+  author = {Harlan P. Crowder and Ron S. Dembo and John M. Mulvey},
+  title = {Reporting computational experiments in mathematical
+                  programming},
+  journal = {Mathematical Programming},
+  volume = 15,
+  number = 1,
+  pages = {316--329},
+  year = 1978,
+  doi = {10.1007/BF01609036},
+  keywords = {reproducibility}
+}
+
+ +
+@article{CruGonPel2011,
+  author = {Carlos Cruz and Juan Ram{\'{o}}n Gonz{\'{a}}lez and David A. Pelta},
+  title = {Optimization in Dynamic Environments: A Survey on Problems, Methods
+               and Measures},
+  journal = {Soft Computing},
+  year = 2011,
+  volume = 15,
+  number = 7,
+  pages = {1427--1448}
+}
+
+ +
+@article{CruSubBruOir2017:cor,
+  author = {F{\'a}bio Cruz and  Anand Subramanian  and Bruno P. Bruck and  Manuel Iori },
+  title = {A Heuristic Algorithm for a Single Vehicle Static Bike Sharing Rebalancing Problem},
+  journal = {Computers \& Operations Research},
+  volume = 79,
+  year = 2017,
+  pages = {19--33}
+}
+
+ +
+@article{Cul1998ec,
+  author = { Joseph C. Culberson },
+  title = {On the Futility of Blind Search: An Algorithmic View of ``No
+                  Free Lunch''},
+  journal = {Evolutionary Computation},
+  year = 1998,
+  volume = 6,
+  number = 2,
+  pages = {109--127},
+  doi = {10.1162/evco.1998.6.2.109},
+  keywords = {NFL}
+}
+
+ +
+@article{CzyJas1998,
+  title = {{Pareto} simulated annealing -- a metaheuristic
+                  technique for multiple-objective combinatorial
+                  optimization},
+  author = {Czyz{\.z}ak, P.  and  Andrzej Jaszkiewicz },
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  volume = 7,
+  number = 1,
+  pages = {34--47},
+  year = 1998
+}
+
+ +
+@article{DamHickRago2010,
+  title = {On Energy, Discrepancy and Group Invariant Measures on
+                  Measurable Subsets of {Euclidean} Space},
+  author = {Damelin, Steven B. and Hickernell, Fred J. and Ragozin, David
+                  L. and Zeng, Xiaoyan},
+  journal = {Journal of Fourier Analysis and Applications},
+  year = 2010,
+  number = 6,
+  pages = {813--839},
+  volume = 16,
+  keywords = {Capacity; Cubature; Discrepancy; Distribution; Group
+                  invariant kernel; Group invariant measure; Energy minimizer;
+                  Equilibrium measure; Numerical integration; Positive
+                  definite; Potential field; Riesz kernel; Reproducing Hilbert
+                  space; Signed measure},
+  publisher = {SP Birkhäuser Verlag Boston},
+  alias = {Damelin2010}
+}
+
+ +
+@article{Damas:2001:PDW,
+  author = {M. Damas and M. Salmer{\'o}n and J. Ortega and
+                  G. Olivares and H. Pomares},
+  title = {Parallel Dynamic Water Supply Scheduling in a
+                  Cluster of Computers},
+  journal = {Concurrency and Computation: Prac\-tice and Experience},
+  volume = 13,
+  number = 15,
+  pages = {1281--1302},
+  day = 25,
+  month = dec,
+  year = 2001,
+  coden = {CCPEBO},
+  issn = {1532-0626 (print), 1532-0634 (electronic)}
+}
+
+ +
+@article{DanPoz2020,
+  author = {Augusto Dantas and Aurora Pozo},
+  title = {On the use of fitness landscape features in meta-learning
+                  based algorithm selection for the quadratic assignment
+                  problem},
+  journal = {Theoretical Computer Science},
+  year = 2020,
+  volume = 805,
+  pages = {62--75},
+  doi = {10.1016/j.tcs.2019.10.033}
+}
+
+ +
+@article{DanRotLep2005,
+  title = {Exploring relaxation induced neighborhoods to improve {MIP} solutions},
+  author = {Danna, Emilie and Rothberg, Edward and Le Pape, Claude},
+  journal = {Mathematical Programming},
+  volume = 102,
+  number = 1,
+  pages = {71--90},
+  year = 2005,
+  publisher = {Springer}
+}
+
+ +
+@article{DanWol1960,
+  title = {Decomposition Principle for Linear Programs},
+  author = {George B. Dantzig and Philip Wolfe},
+  journal = {Operations Research},
+  volume = 8,
+  number = 1,
+  pages = {101--111},
+  year = 1960
+}
+
+ +
+@article{DaoLieVerAgu2017ec,
+  author = {Fabio Daolio and  Arnaud Liefooghe  and  Verel, S{\'e}bastien  and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  journal = {Evolutionary Computation},
+  number = 4,
+  pages = {555--585},
+  title = {Problem Features versus Algorithm Performance on Rugged
+                  Multiobjective Combinatorial Fitness Landscapes},
+  volume = 25,
+  year = 2017,
+  doi = {10.1162/evco_a_00193}
+}
+
+ +
+@article{DasDen1997,
+  author = {Das, Indraneel and Dennis, John E.},
+  title = {A closer look at drawbacks of minimizing weighted
+                  sums of objectives for {Pareto} set generation in
+                  multicriteria optimization problems},
+  journal = {Structural Optimization},
+  year = 1997,
+  volume = 14,
+  number = 1,
+  pages = {63--69},
+  doi = {10.1007/BF01197559}
+}
+
+ +
+@article{DasDen1998normal,
+  author = {Das, Indraneel and Dennis, John E.},
+  title = {Normal-boundary intersection: A new method for generating the
+                  {Pareto} surface in nonlinear multicriteria optimization
+                  problems},
+  journal = {SIAM Journal on Optimization},
+  year = 1998,
+  volume = 8,
+  number = 3,
+  pages = {631--657},
+  keywords = {simplex lattice design}
+}
+
+ +
+@article{DasMulSug2016de,
+  title = {Recent advances in differential evolution--{An} updated
+                  survey},
+  author = { Swagatam Das  and Mullick, Sankha Subhra and  Ponnuthurai N. Suganthan },
+  journal = {Swarm and Evolutionary Computation},
+  volume = 27,
+  pages = {1--30},
+  year = 2016
+}
+
+ +
+@article{DasSug2011:tec,
+  author = { Swagatam Das  and  Ponnuthurai N. Suganthan },
+  title = {Differential Evolution: A Survey of the State-of-the-art},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2011,
+  volume = 15,
+  number = 1,
+  month = feb
+}
+
+ +
+@article{Dash05,
+  author = {Sanjeeb Dash},
+  title = {Exponential Lower Bounds on the Lengths of Some Classes
+               of Branch-and-Cut Proofs},
+  journal = {Mathematics of Operations Research},
+  year = 2005,
+  volume = 30,
+  number = 3,
+  pages = {678--700}
+}
+
+ +
+@article{DasDiaYan2016chord,
+  title = {How good is the {Chord} algorithm?},
+  author = {Daskalakis, Constantinos and Diakonikolas, Ilias and  Mihalis Yannakakis },
+  journal = {SIAM Journal on Computing},
+  year = 2016,
+  number = 3,
+  pages = {811--858},
+  volume = 45
+}
+
+ +
+@article{DauDouPesKie2010:po2,
+  author = { Jean Daunizeau  and  Hanneke E. M. den Ouden  and  Matthias Pessiglione  and  Stefan J. Kiebel  and  Karl J. Friston  and  Klaas E. Stephan },
+  title = {Observing the observer ({II}): deciding when to decide},
+  journal = {PLoS One},
+  year = 2010,
+  volume = 5,
+  number = 12,
+  pages = {e15555},
+  doi = {10.1371/journal.pone.0015555}
+}
+
+ +
+@article{DauDouPesSte2010:po1,
+  author = { Jean Daunizeau  and  Hanneke E. M. den Ouden  and  Matthias Pessiglione  and  Klaas E. Stephan  and  Stefan J. Kiebel  and  Karl J. Friston },
+  title = {Observing the observer ({I}): meta-{Bayesian} models of learning and decision-making},
+  journal = {PLoS One},
+  year = 2010,
+  volume = 5,
+  number = 12,
+  pages = {e15554},
+  doi = {10.1371/journal.pone.0015554}
+}
+
+ +
+@article{Deb00constraint,
+  author = { Kalyanmoy Deb },
+  title = {An efficient constraint handling method for genetic
+                  algorithms},
+  journal = {Computer Methods in Applied Mechanics and Engineering},
+  year = 2000,
+  volume = 186,
+  number = {2/4},
+  pages = {311--338},
+  doi = {10.1016/S0045-7825(99)00389-8}
+}
+
+ +
+@article{Deb02nsga2,
+  author = { Kalyanmoy Deb  and A. Pratap and S. Agarwal and T. Meyarivan},
+  title = {A fast and elitist multi-objective genetic
+                  algorithm: {NSGA-II}},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002,
+  volume = 6,
+  number = 2,
+  pages = {182--197},
+  doi = {10.1109/4235.996017}
+}
+
+ +
+@article{Deb1999ec,
+  author = { Kalyanmoy Deb },
+  title = {Multi-objective genetic algorithms: problem
+                  difficulties and construction of test problems},
+  journal = {Evolutionary Computation},
+  year = 1999,
+  volume = 7,
+  number = 3,
+  pages = {205--230},
+  annote = {Naive definition of PLO-set}
+}
+
+ +
+@article{DebAgr1995sbx,
+  author = { Kalyanmoy Deb  and  Ram Bhushan Agrawal },
+  title = {Simulated binary crossover for continuous search
+                  spaces},
+  journal = {Complex Systems},
+  volume = 9,
+  number = 2,
+  pages = {115--148},
+  year = 1995,
+  epub = {http://www.complex-systems.com/abstracts/v09_i02_a02.html},
+  keywords = {SBX}
+}
+
+ +
+@article{DebDeb2014,
+  author = { Kalyanmoy Deb  and Debayan Deb},
+  title = {Analysing mutation schemes for real-parameter genetic
+                  algorithms},
+  journal = {International Journal of Artificial Intelligence and Soft Computing},
+  year = 2014,
+  volume = 4,
+  number = 1,
+  pages = {1--28},
+  annote = {Proposed Gaussian mutation}
+}
+
+ +
+@article{DebGupDauBran2009reliab,
+  author = { Kalyanmoy Deb  and S. Gupta and D. Daum and  J{\"u}rgen Branke  and A. Mall and D. Padmanabhan},
+  title = {Reliability-based optimization using evolutionary algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 13,
+  number = 5,
+  pages = {1054--1074},
+  month = oct,
+  year = 2009,
+  doi = {10.1109/TEVC.2009.2014361}
+}
+
+ +
+@article{DebJain2014:nsga3-part1,
+  author = { Kalyanmoy Deb  and  Himanshu Jain },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {An Evolutionary Many-Objective Optimization Algorithm Using
+                  Reference-Point-Based Nondominated Sorting Approach, Part
+                  {I}: Solving Problems With Box Constraints},
+  year = 2014,
+  volume = 18,
+  number = 4,
+  pages = {577--601},
+  annote = {Proposed NSGA-III}
+}
+
+ +
+@article{DebKok2010tec-ged,
+  author = { Kalyanmoy Deb  and  Murat K{\"o}ksalan },
+  title = {Guest Editorial: Special Issue on Preference-based
+                  Multiobjective Evolutionary Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 14,
+  number = 5,
+  month = oct,
+  year = 2010,
+  pages = {669--670},
+  doi = {10.1109/TEVC.2010.2070371}
+}
+
+ +
+@article{DebMohMis2005epsilon,
+  title = {Evaluating the {$\epsilon$}-domination based multi-objective
+                  evolutionary algorithm for a quick computation of
+                  {Pareto}-optimal solutions},
+  author = { Kalyanmoy Deb  and Mohan, Manikanth and Mishra, Shikhar},
+  journal = {Evolutionary Computation},
+  year = 2005,
+  month = dec,
+  number = 4,
+  pages = {501--525},
+  volume = 13,
+  doi = {10.1162/106365605774666895},
+  keywords = {$\epsilon$-dominance, $\epsilon$-MOEA}
+}
+
+ +
+@article{DebTiw2008omni,
+  author = { Kalyanmoy Deb  and  Santosh Tiwari },
+  title = {Omni-optimizer: {A} generic evolutionary algorithm for single
+                  and multi-objective optimization},
+  journal = {European Journal of Operational Research},
+  year = 2008,
+  volume = 185,
+  number = 3,
+  pages = {1062--1087},
+  annote = {Archiving method with epsilon dominance and density in the
+                  decision and objective spaces},
+  keywords = {epsilon-dominance, archiving},
+  doi = {10.1016/j.ejor.2006.06.042}
+}
+
+ +
+@article{DebZhuKul2018tec,
+  author = { Kalyanmoy Deb  and Zhu, Ling and Kulkarni, Sandeep},
+  title = {Handling Multiple Scenarios in Evolutionary Multi-Objective
+                  Numerical Optimization},
+  doi = {10.1109/TEVC.2017.2776921},
+  abstract = {Solutions to most practical numerical optimization problems
+                  must be evaluated for their performance over a number of
+                  different loading or operating conditions, which we refer
+                  here as scenarios. Therefore, a meaningful and resilient
+                  optimal solution must be such that it remains feasible under
+                  all scenarios and performs close to an individual optimal
+                  solution corresponding to each scenario. Despite its
+                  practical importance, multi-scenario consideration has
+                  received a lukewarm attention, particularly in the context of
+                  multi-objective optimization. The usual practice is to
+                  optimize for the worst-case scenario. In this paper, we
+                  review existing methodologies in this direction and set our
+                  goal to suggest a new and potential population-based method
+                  for handling multiple scenarios by defining scenario-wise
+                  domination principle and scenario-wise diversity-preserving
+                  operators. To evaluate, the proposed method is applied to a
+                  number of numerical test problems and engineering design
+                  problems with a detail explanation of the obtained results
+                  and compared with an existing method. This first systematic
+                  evolutionary based multi-scenario, multiobjective,
+                  optimization study on numerical problems indicates that
+                  multiple scenarios can be handled in an integrated manner
+                  using an EMO framework to find a well-balanced compromise set
+                  of solutions to multiple scenarios and maintain a tradeoff
+                  among multiple objectives. In comparison to an existing
+                  serial multiple optimization approach, the proposed approach
+                  finds a set of compromised trade-off solutions
+                  simultaneously. An achievement of multi-objective trade-off
+                  and multi-scenario trade-off is algorithmically challenging,
+                  but due to its practical appeal, further research and
+                  application must be spent.},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  volume = 22,
+  number = 6,
+  pages = {920--933},
+  keywords = {scenario-based}
+}
+
+ +
+@article{DecSor2012ejor,
+  author = {Annelies De Corte and  Kenneth S{\"o}rensen },
+  title = {Optimisation of gravity-fed water distribution network
+                  design: A critical review},
+  journal = {European Journal of Operational Research},
+  volume = 228,
+  number = 1,
+  pages = {1--10},
+  doi = {10.1016/j.ejor.2012.11.046},
+  year = 2013
+}
+
+ +
+@article{DecSor2016,
+  author = {Annelies De Corte and  Kenneth S{\"o}rensen },
+  title = {An Iterated Local Search Algorithm for Water Distribution
+                  Network Design Optimization},
+  journal = {Networks},
+  year = 2016,
+  volume = 67,
+  number = 3,
+  pages = {187--198}
+}
+
+ +
+@article{DecSor2016water,
+  author = {Annelies De Corte and  Kenneth S{\"o}rensen },
+  title = {An Iterated Local Search Algorithm for multi-period water
+                  distribution network design optimization},
+  journal = {Water},
+  volume = 8,
+  number = 8,
+  pages = 359,
+  doi = {10.3390/w8080359},
+  year = 2016
+}
+
+ +
+@article{Dek1981:cad,
+  author = { V. Dekhtyarenko },
+  title = {Verification of weight coefficients in multicriteria optimization problems},
+  journal = {Computer-Aided Design},
+  volume = 13,
+  number = 6,
+  pages = {339--344},
+  year = 1981
+}
+
+ +
+@article{DelGanDeg2010,
+  title = {Evolutionary, constructive and hybrid procedures for
+                  the bi-objective set packing problem},
+  author = {Delorme, X.  and  Xavier Gandibleux  and  Degoutin, F.},
+  journal = {European Journal of Operational Research},
+  volume = 204,
+  number = 2,
+  pages = {206--217},
+  year = 2010,
+  annote = {This paper cannot be found on internet!! Does it exist?}
+}
+
+ +
+@article{DelGarGro2012:cor,
+  author = { Federico {Della Croce}  and Thierry Garaix and  Andrea Grosso },
+  title = {Iterated Local Search and Very Large Neighborhoods for the Parallel-machines
+               Total Tardiness Problem},
+  journal = {Computers \& Operations Research},
+  year = 2012,
+  volume = 39,
+  number = 6,
+  pages = {1213--1217}
+}
+
+ +
+@article{DelIorMar2016binpack,
+  title = {Bin packing and cutting stock problems: Mathematical models
+                  and exact algorithms},
+  author = {Delorme, Maxence and  Manuel Iori  and  Silvano Martello },
+  journal = {European Journal of Operational Research},
+  volume = 255,
+  number = 1,
+  pages = {1--20},
+  year = 2016,
+  publisher = {Elsevier},
+  doi = {10.1016/j.ejor.2016.04.030}
+}
+
+ +
+@article{DelIorMarMon2016,
+  author = {Mauro Dell'Amico and  Manuel Iori  and  Silvano Martello  and  Monaci, Michele },
+  title = {Heuristic and Exact Algorithms for the Identical Parallel Machine Scheduling Problem},
+  journal = {INFORMS Journal on Computing},
+  year = 2016,
+  volume = 20,
+  number = 3,
+  pages = {333--344}
+}
+
+ +
+@article{DelIorMarc2018bpplib,
+  title = {{BPPLIB}: a library for bin packing and cutting stock problems},
+  author = {Delorme, Maxence and  Manuel Iori  and  Silvano Martello },
+  journal = {Optimization Letters},
+  volume = 12,
+  number = 2,
+  pages = {235--250},
+  year = 2018,
+  doi = {10.1007/s11590-017-1192-z}
+}
+
+ +
+@article{DelIorNovStu2016,
+  author = {Mauro Dell'Amico and  Manuel Iori  and Stefano Novellani and  Thomas St{\"u}tzle },
+  title = {A destroy and repair algorithm for the Bike sharing Rebalancing Problem},
+  journal = {Computers \& Operations Research},
+  volume = 71,
+  pages = {146--162},
+  year = 2016,
+  doi = {10.1016/j.cor.2016.01.011},
+  keywords = {irace}
+}
+
+ +
+@article{DelKar1990interactive,
+  title = {An interactive {MCDM} weight space reduction method utilizing
+                  a {Tchebycheff} utility function},
+  author = {Dell, Robert F. and Karwan, Mark H.},
+  journal = {Naval Research Logistics},
+  volume = 37,
+  number = 2,
+  pages = {263--277},
+  year = 1990
+}
+
+ +
+@article{DelTru1993,
+  author = {Mauro Dell'Amico and Marco Trubian},
+  title = {Applying Tabu Search to the Job Shop Scheduling Problem},
+  journal = {Annals of Operations Research},
+  year = 1993,
+  volume = 41,
+  pages = {231--252}
+}
+
+ +
+@article{DemEichFli2015effects,
+  title = {On the effects of combining objectives in multi-objective
+                  optimization},
+  author = {Dempe, Stephan and Eichfelder, Gabriele and Fliege, J{\"o}rg},
+  journal = {Mathematical Methods of Operations Research},
+  volume = 82,
+  number = 1,
+  pages = {1--18},
+  year = 2015,
+  publisher = {Springer}
+}
+
+ +
+@article{DenAroGosPas1990,
+  author = { Jean-Louis Deneubourg  and S. Aron and S. Goss and
+                  J.-M. Pasteels},
+  title = {The Self-Organizing Exploratory Pattern of the
+                  {Argentine} Ant},
+  journal = {Journal of Insect Behavior},
+  year = 1990,
+  volume = 3,
+  number = 2,
+  pages = {159--168},
+  doi = {10.1007/BF01417909},
+  alias = {DAGP90}
+}
+
+ +
+@article{DerGarMolHer2011stats,
+  title = {A practical tutorial on the use of nonparametric statistical
+                  tests as a methodology for comparing evolutionary and swarm
+                  intelligence algorithms},
+  author = {Derrac, Joaqu{\'i}n and Garc{\'i}a, Salvador and  Daniel Molina  and  Francisco Herrera },
+  journal = {Swarm and Evolutionary Computation},
+  volume = 1,
+  number = 1,
+  pages = {3--18},
+  year = 2011
+}
+
+ +
+@article{DerVog2014:joh,
+  author = {Ulrich Derigs and Ulrich Vogel},
+  title = {Experience with a Framework for Developing Heuristics for
+                  Solving Rich Vehicle Routing Problems},
+  journal = {Journal of Heuristics},
+  year = 2014,
+  volume = 20,
+  number = 1,
+  pages = {75--106}
+}
+
+ +
+@article{DesBelDop2021bops,
+  author = {Aryan Deshwal and Syrine Belakaria and Janardhan Rao Doppa
+                  and Dae Hyun Kim},
+  title = {Bayesian Optimization over Permutation Spaces},
+  journal = {Arxiv preprint arXiv:2112.01049},
+  year = 2021,
+  doi = {10.48550/arXiv.2112.01049},
+  keywords = {BOPS, CEGO}
+}
+
+ +
+@article{DesRitLopPer2021acviz,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and   P{\'e}rez C{\'a}ceres, Leslie},
+  title = {{\softwarepackage{ACVIZ}}: A Tool for the Visual Analysis of
+                  the Configuration of Algorithms with {\rpackage{irace}}},
+  journal = {Operations Research Perspectives},
+  year = 2021,
+  doi = {10.1016/j.orp.2021.100186},
+  supplement = {https://zenodo.org/record/4714582},
+  abstract = {This paper introduces acviz, a tool that helps to analyze the
+                  automatic configuration of algorithms with irace. It provides
+                  a visual representation of the configuration process,
+                  allowing users to extract useful information, e.g. how the
+                  configurations evolve over time. When test data is available,
+                  acviz also shows the performance of each configuration on the
+                  test instances. Using this visualization, users can analyze
+                  and compare the quality of the resulting configurations and
+                  observe the performance differences on training and test
+                  instances.},
+  volume = 8,
+  pages = 100186
+}
+
+ +
+@article{DetPapZab2017omega,
+  title = {A multi-depot dial-a-ride problem with heterogeneous vehicles
+                  and compatibility constraints in healthcare},
+  author = {Paolo Detti and Francesco Papalini and Garazi Zabalo {Manrique
+                  de Lara}},
+  journal = {Omega},
+  volume = 70,
+  pages = {1--14},
+  year = 2017,
+  doi = {10.1016/j.omega.2016.08.008}
+}
+
+ +
+@article{DevVoh2003informs,
+  title = {Combinatorial Auctions: A Survey},
+  author = { Sven {De Vries}  and  Rakesh V. Vohra },
+  journal = {INFORMS Journal on Computing},
+  volume = 15,
+  number = 3,
+  pages = {284--309},
+  year = 2003,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{DiaHanXu2017,
+  title = {Evolutionary robust optimization in production planning:
+                  interactions between number of objectives, sample size and
+                  choice of robustness measure},
+  journal = {Computers \& Operations Research},
+  volume = 79,
+  pages = {266--278},
+  year = 2017,
+  doi = {10.1016/j.cor.2016.06.020},
+  author = { Juan Esteban Diaz  and  Julia Handl  and  Dong-Ling Xu },
+  keywords = {Evolutionary multi-objective optimization, Production
+                  planning, Robust optimization, Simulation-based optimization,
+                  Uncertainty modelling},
+  alias = {DIAZ2017266}
+}
+
+ +
+@article{DiaHanXu2018,
+  title = {Integrating meta-heuristics, simulation and exact techniques
+                  for production planning of a failure-prone manufacturing
+                  system},
+  journal = {European Journal of Operational Research},
+  volume = 266,
+  number = 3,
+  pages = {976--989},
+  year = 2018,
+  issn = {0377-2217},
+  doi = {10.1016/j.ejor.2017.10.062},
+  author = { Juan Esteban Diaz  and  Julia Handl  and  Dong-Ling Xu },
+  keywords = {Genetic algorithms, Combinatorial optimization, Production
+                  planning, Simulation-based optimization, Uncertainty
+                  modelling},
+  alias = {DIAZ2018976}
+}
+
+ +
+@article{DiaLop2020ejor,
+  author = { Juan Esteban Diaz  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Incorporating Decision-Maker's Preferences into the Automatic
+                  Configuration of Bi-Objective Optimisation Algorithms},
+  journal = {European Journal of Operational Research},
+  year = 2021,
+  volume = 289,
+  number = 3,
+  pages = {1209--1222},
+  doi = {10.1016/j.ejor.2020.07.059},
+  abstract = {Automatic configuration (AC) methods are increasingly used to
+                  tune and design optimisation algorithms for problems with
+                  multiple objectives. Most AC methods use unary quality
+                  indicators, which assign a single scalar value to an
+                  approximation to the Pareto front, to compare the performance
+                  of different optimisers. These quality indicators, however,
+                  imply preferences beyond Pareto-optimality that may differ
+                  from those of the decision maker (DM). Although it is
+                  possible to incorporate DM's preferences into quality
+                  indicators, e.g., by means of the weighted hypervolume
+                  indicator (HV$^w$), expressing preferences in terms of weight
+                  function is not always intuitive nor an easy task for a DM,
+                  in particular, when comparing the stochastic outcomes of
+                  several algorithm configurations. A more visual approach to
+                  compare such outcomes is the visualisation of their empirical
+                  attainment functions (EAFs) differences. This paper proposes
+                  using such visualisations as a way of eliciting information
+                  about regions of the objective space that are preferred by
+                  the DM. We present a method to convert the information about
+                  EAF differences into a HV$^w$ that will assign higher quality
+                  values to approximation fronts that result in EAF differences
+                  preferred by the DM. We show that the resulting HV$^w$ may be
+                  used by an AC method to guide the configuration of
+                  multi-objective optimisers according to the preferences of
+                  the DM. We evaluate the proposed approach on a well-known
+                  benchmark problem. Finally, we apply our approach to
+                  re-configuring, according to different DM's preferences, a
+                  multi-objective optimiser tackling a real-world production
+                  planning problem arising in the manufacturing industry.},
+  supplement = {https://doi.org/10.5281/zenodo.3749288}
+}
+
+ +
+@article{DiaMouFigCli2002:ejor,
+  author = { L. C. Dias  and  Vincent Mousseau  and  Jos{\'e} Rui Figueira  and  J. N. Cl{\'i}maco },
+  title = {An aggregation/disaggregation approach to obtain robust conclusions with {ELECTRE TRI}},
+  year = 2002,
+  journal = {European Journal of Operational Research},
+  volume = 138,
+  number = 2,
+  month = apr,
+  pages = {332--348 }
+}
+
+ +
+@article{DiaYan2009small,
+  title = {Small approximate {Pareto} sets for biobjective shortest
+                  paths and other problems},
+  author = {Diakonikolas, Ilias and  Mihalis Yannakakis },
+  journal = {SIAM Journal on Computing},
+  year = 2009,
+  number = 4,
+  pages = {1340--1371},
+  volume = 39
+}
+
+ +
+@article{DicDor1998:jair,
+  author = { Gianni A. {Di Caro}  and  Marco Dorigo },
+  title = {Ant{Net}: Distributed Stigmergetic Control for Communications
+                  Networks},
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 9,
+  pages = {317--365},
+  year = 1998,
+  alias = {DicDor98:jair}
+}
+
+ +
+@article{DicDucGam2005,
+  author = { Gianni A. {Di Caro}  and  F. Ducatelle  and  L. M. Gambardella },
+  title = {{AntHocNet}: An adaptive nature-inspired algorithm for routing in mobile ad hoc networks},
+  journal = {European Transactions on Telecommunications},
+  year = 2005,
+  volume = 16,
+  number = 5,
+  pages = {443--455}
+}
+
+ +
+@article{DigSch2003,
+  author = {Luca {Di Gaspero} and Andrea Schaerf},
+  citations = 36,
+  journal = {Software --- Practice \& Experience},
+  keywords = {software engineering, local search, easylocal},
+  month = jul,
+  number = 8,
+  pages = {733--765},
+  publisher = {John Wiley \& Sons},
+  title = {\textsc{EasyLocal++}: An object-oriented framework
+                  for flexible design of local search algorithms},
+  epub = {http://www.diegm.uniud.it/satt/papers/DiSc03.pdf},
+  volume = 33,
+  year = 2003
+}
+
+ +
+@article{DilKhaNem2017comments,
+  title = {Comments on: On learning and branching: a survey},
+  author = {Dilkina, Bistra and Khalil, Elias B. and Nemhauser, George
+                  L.},
+  journal = {TOP},
+  volume = 25,
+  pages = {242--246},
+  year = 2017,
+  publisher = {Springer},
+  annote = {Comments on \cite{LodZar2017learning}.}
+}
+
+ +
+@article{DinDonHeLi2019twoarch,
+  title = {A novel two-archive strategy for evolutionary many-objective
+                  optimization algorithm based on reference points},
+  author = {Ding, Rui and Dong, Hongbin and He, Jun and Li, Tao},
+  journal = {Applied Soft Computing},
+  year = 2019,
+  pages = {447--464},
+  volume = 78,
+  doi = {10.1016/j.asoc.2019.02.040},
+  publisher = {Elsevier}
+}
+
+ +
+@article{DinSonGup2015,
+  author = {Ding, J.-Y. and Song, S. and Gupta, J. N. D. and Zhang, R. and Chiong, R. and Wu, C.},
+  title = {An Improved Iterated Greedy Algorithm with a Tabu-based Reconstruction Strategy for the No-wait Flowshop Scheduling Problem},
+  journal = {Applied Soft Computing},
+  year = 2015,
+  volume = 30,
+  pages = {604--613}
+}
+
+ +
+@article{DoeDoeEbe2015,
+  author = { Benjamin Doerr  and  Carola Doerr  and Franziska Ebel},
+  title = {From black-box complexity to designing new genetic algorithms},
+  journal = {Theoretical Computer Science},
+  volume = 567,
+  pages = {87--104},
+  year = 2015,
+  doi = {10.1016/j.tcs.2014.11.028}
+}
+
+ +
+@article{DoeDoeYan2020,
+  author = { Benjamin Doerr  and  Carola Doerr  and Yang, Jing},
+  title = {Optimal parameter choices via precise black-box analysis},
+  journal = {Theoretical Computer Science},
+  volume = 801,
+  pages = {1--34},
+  year = 2020,
+  doi = {10.1016/j.tcs.2019.06.014}
+}
+
+ +
+@article{DoeFueGro06,
+  author = { Karl F. Doerner  and  Guenther Fuellerer  and  Manfred Gronalt  and  Richard F. Hartl  and  Manuel Iori },
+  title = {Metaheuristics for the Vehicle Routing Problem with Loading Constraints},
+  journal = {Networks},
+  year = 2006,
+  volume = 49,
+  number = 4,
+  pages = {294--307}
+}
+
+ +
+@article{DoeGieWitYan2019,
+  title = {The ({1+\(\lambda\)}) evolutionary
+                  algorithm with self-adjusting mutation rate},
+  author = { Benjamin Doerr  and Gie{\ss}en, Christian and  Carsten Witt  and Yang, Jing},
+  journal = {Algorithmica},
+  volume = 81,
+  number = 2,
+  pages = {593--631},
+  year = 2019,
+  publisher = {Springer}
+}
+
+ +
+@article{DoeGutHar08,
+  author = { Karl F. Doerner  and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Christine Strauss  and  Christian Stummer },
+  title = {Nature-Inspired Metaheuristics in Multiobjective
+                  Activity Crashing},
+  journal = {Omega},
+  year = 2008,
+  volume = 36,
+  number = 6,
+  pages = {1019--1037}
+}
+
+ +
+@article{DoeGutHarStrStu04:aor,
+  author = { Karl F. Doerner  and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Christine Strauss  and  Christian Stummer },
+  title = {{Pareto} Ant Colony Optimization: A Metaheuristic
+                  Approach to Multiobjective Portfolio Selection},
+  journal = {Annals of Operations Research},
+  year = 2004,
+  volume = 131,
+  pages = {79--99},
+  alias = {DoeGutHarStrStu03}
+}
+
+ +
+@article{DoeGutHarStrStu06:ejor,
+  author = { Karl F. Doerner  and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Christine Strauss  and  Christian Stummer },
+  title = {{Pareto} ant colony optimization with ILP preprocessing in
+                   multiobjective project portfolio selection},
+  journal = {European Journal of Operational Research},
+  year = 2006,
+  volume = 171,
+  pages = {830--841}
+}
+
+ +
+@article{DoeHarRei03,
+  author = { Karl F. Doerner  and  Richard F. Hartl  and  Marc Reimann },
+  title = {Are {COMPETants} more competent for problem solving?
+                  {The} case of a multiple objective transportation
+                  problem},
+  journal = {Central European Journal for Operations Research and Economics},
+  pages = {115--141},
+  volume = 11,
+  number = 2,
+  year = 2003
+}
+
+ +
+@article{DoeJohWin2012,
+  title = {Multiplicative drift analysis},
+  author = { Benjamin Doerr  and Johannsen, Daniel and Winzen, Carola},
+  journal = {Algorithmica},
+  volume = 64,
+  number = 4,
+  pages = {673--697},
+  year = 2012
+}
+
+ +
+@article{DoeKotLenWin2013,
+  title = {Black-box complexities of combinatorial problems},
+  author = { Benjamin Doerr  and K{\"o}tzing, Timo  and Lengler, Johannes and Winzen, Carola},
+  journal = {Theoretical Computer Science},
+  volume = 471,
+  pages = {84--106},
+  year = 2013
+}
+
+ +
+@article{DoeMerStu2009:swarm,
+  author = { Karl F. Doerner  and  D. Merkle  and  Thomas St{\"u}tzle },
+  title = {Special issue on Ant Colony Optimization},
+  journal = {Swarm Intelligence},
+  year = 2009,
+  volume = 3,
+  number = 1
+}
+
+ +
+@article{DoeNeuSudWit2011:tcs,
+  author = { Benjamin Doerr  and  Frank Neumann  and  Dirk Sudholt  and  Carsten Witt },
+  title = {Runtime analysis of the 1-{ANT} ant colony optimizer},
+  journal = {Theoretical Computer Science},
+  year = 2011,
+  volume = 412,
+  number = 1,
+  pages = {1629--1644}
+}
+
+ +
+@article{Dog2015asoco,
+  author = { Do\v{g}an Ayd{\i}n },
+  title = {Composite artificial bee colony algorithms: From
+                  component-based analysis to high-performing algorithms},
+  journal = {Applied Soft Computing},
+  volume = 32,
+  pages = {266--285},
+  year = 2015,
+  doi = {10.1016/j.asoc.2015.03.051},
+  keywords = {irace}
+}
+
+ +
+@article{DoiPekReg2004rank,
+  author = {Jean-Paul Doignon and Aleksandar Peke{\v{c}} and Michel
+                  Regenwetter},
+  title = {The repeated insertion model for rankings: Missing link
+                  between two subset choice models},
+  doi = {10.1007/bf02295838},
+  year = 2004,
+  month = mar,
+  volume = 69,
+  number = 1,
+  pages = {33--54},
+  journal = {Psychometrika},
+  abstract = {Several probabilistic models for subset choice have been
+                  proposed in the literature, for example, to explain approval
+                  voting data. We show that Marley et al.'s latent scale model
+                  is subsumed by Falmagne and Regenwetter's size-independent
+                  model, in the sense that every choice probability
+                  distribution generated by the former can also be explained by
+                  the latter. Our proof relies on the construction of a
+                  probabilistic ranking model which we label the ``repeated
+                  insertion model''. This model is a special case of Marden's
+                  orthogonal contrast model class and, in turn, includes the
+                  classical Mallows $\varphi$-model as a special case. We
+                  explore its basic properties as well as its relationship to
+                  Fligner and Verducci's multistage ranking model.}
+}
+
+ +
+@article{DolMor2002benchmarking,
+  author = {Dolan, Elizabeth D. and Mor{\'e}, Jorge J.},
+  journal = {Mathematical Programming},
+  number = 2,
+  pages = {201--213},
+  title = {Benchmarking optimization software with performance profiles},
+  volume = 91,
+  year = 2002,
+  keywords = {performance profiles; convergence},
+  annote = {This methodology has been criticized in \url{https://doi.org/10.1145/2950048}}
+}
+
+ +
+@article{DonCheHua2013,
+  author = {Xingye Dong and Ping and Houkuan Huang and Maciek Nowak},
+  title = {A Multi-restart Iterated Local Search Algorithm for the Permutation Flow Shop Problem Minimizing Total Flow Time},
+  journal = {Computers \& Operations Research},
+  year = 2013,
+  volume = 40,
+  number = 2,
+  pages = {627--632}
+}
+
+ +
+@article{DonHuaChe2009:cor,
+  author = {X. Dong and H. Huang and P. Chen},
+  title = {An Iterated Local Search Algorithm for the Permutation Flowshop Problem with Total Flowtime Criterion},
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  volume = 36,
+  number = 5,
+  pages = {1664--1669}
+}
+
+ +
+@article{DonMonCasRizGam08,
+  author = {A. V. Donati and  Roberto Montemanni  and N. Casagrande  and A. E. Rizzoli  and  L. M. Gambardella },
+  title = {Time dependent vehicle routing problem with a multi ant
+                  colony system},
+  journal = {European Journal of Operational Research},
+  volume = 185,
+  number = 3,
+  year = 2008,
+  pages = {1174--1191}
+}
+
+ +
+@article{Dor2007:scholarpedia,
+  author = { Marco Dorigo },
+  title = {Ant {Colony} {Optimization}},
+  year = 2007,
+  journal = {Scholarpedia},
+  volume = 2,
+  number = 3,
+  pages = 1461,
+  doi = {10.4249/scholarpedia.1461},
+  alias = {Dor2007:sch-aco}
+}
+
+ +
+@article{Dor2016sipolicy,
+  author = { Marco Dorigo },
+  title = {Swarm intelligence: A few things you need to know if you want
+                  to publish in this journal},
+  journal = {Swarm Intelligence},
+  year = 2016,
+  month = nov,
+  url = {https://static.springer.com/sgw/documents/1593723/application/pdf/Additional_submission_instructions.pdf}
+}
+
+ +
+@article{DorBirLiLop2017si,
+  author = { Marco Dorigo  and  Mauro Birattari  and  Li, Xiaodong  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kazuhiro Ohkura  and  Carlo Pinciroli  and  Thomas St{\"u}tzle },
+  title = {{ANTS} 2016 Special Issue: Editorial},
+  journal = {Swarm Intelligence},
+  year = 2017,
+  month = nov,
+  doi = {10.1007/s11721-017-0146-5}
+}
+
+ +
+@article{DorBirStu06:ci,
+  author = { Marco Dorigo  and  Mauro Birattari  and  Thomas St{\"u}tzle },
+  title = {Ant Colony Optimization: Artificial Ants as a Computational Intelligence Technique},
+  journal = {IEEE Computational Intelligence Magazine},
+  year = 2006,
+  volume = 1,
+  number = 4,
+  pages = {28--39}
+}
+
+ +
+@article{DorBlu2005:tcs,
+  author = { Marco Dorigo  and  Christian Blum },
+  title = {Ant colony optimization theory: A survey},
+  journal = {Theoretical Computer Science},
+  volume = 344,
+  number = {2-3},
+  year = 2005,
+  pages = {243--278}
+}
+
+ +
+@article{DorDicGam99:al,
+  author = { Marco Dorigo  and  Gianni A. {Di Caro}  and  L. M. Gambardella },
+  title = {Ant Algorithms for Discrete Optimization},
+  journal = {Artificial Life},
+  volume = 5,
+  number = 2,
+  pages = {137--172},
+  year = 1999
+}
+
+ +
+@article{DorGam1997:biosys,
+  author = { Marco Dorigo  and  L. M. Gambardella },
+  title = {Ant Colonies for the Traveling Salesman Problem},
+  journal = {BioSystems},
+  year = 1997,
+  volume = 43,
+  number = 2,
+  pages = {73--81},
+  alias = {DorGam1997:bio},
+  doi = {10.1016/S0303-2647(97)01708-5}
+}
+
+ +
+@article{DorGam1997:tec,
+  key = {DorGam97:tec},
+  alias = {DorGam97:tec},
+  author = { Marco Dorigo  and  L. M. Gambardella },
+  title = {{Ant} {Colony} {System}: A Cooperative Learning
+                  Approach to the Traveling Salesman Problem},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 1997,
+  volume = 1,
+  number = 1,
+  pages = {53--66},
+  keywords = {Ant Colony System}
+}
+
+ +
+@article{DorGamMidStu2002:tec,
+  doi = {10.1109/TEVC.2002.802446},
+  author = { Marco Dorigo  and  L. M. Gambardella  and  Martin Middendorf  and  Thomas St{\"u}tzle },
+  title = {Guest Editorial: Special Section on Ant Colony
+                  Optimization},
+  year = 2002,
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 6,
+  number = 4,
+  pages = {317--320},
+  keywords = {ant colony optimization, swarm intelligence}
+}
+
+ +
+@article{DorManCol1996:tsmcb,
+  author = { Marco Dorigo  and  Vittorio Maniezzo  and  Alberto Colorni },
+  title = {{Ant} {System}: Optimization by a Colony of
+                  Cooperating Agents},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics -- Part B},
+  year = 1996,
+  volume = 26,
+  number = 1,
+  pages = {29--41},
+  alias = {DorManCol96:tsmcb}
+}
+
+ +
+@article{DorStuDic2000:fgcs,
+  author = { Marco Dorigo  and  Thomas St{\"u}tzle  and  Gianni A. {Di Caro} },
+  title = {Special Issue on ``{Ant} {Algorithms}''},
+  year = 2000,
+  journal = {Future Generation Computer Systems},
+  volume = 16,
+  number = 8,
+  keywords = {swarm intelligence, ant colony optimization}
+}
+
+ +
+@article{DouZop2010:ejor,
+  author = { Michael Doumpos  and  Constantin Zopounidis },
+  title = {Preference disaggregation and statistical learning for multicriteria decision support: A review},
+  journal = {European Journal of Operational Research},
+  volume = 209,
+  number = 3,
+  pages = {203--214},
+  year = 2011
+}
+
+ +
+@article{DovTusFil2010ec,
+  author = {Erik Dovgan and  Tea Tu{\v s}ar  and Bogdan Filipi{\v c}},
+  title = {Parameter tuning in an evolutionary algorithm for commodity
+                  transportation optimization},
+  journal = {Evolutionary Computation},
+  alias = {ParamTunComm},
+  year = 2010,
+  pages = {1--8}
+}
+
+ +
+@article{DreSia2004,
+  author = { Johann Dr{\'e}o  and P. Siarry},
+  title = {Continuous interacting ant colony algorithm based on dense
+                  heterarchy},
+  journal = {Future Generation Computer Systems},
+  volume = 20,
+  number = 5,
+  year = 2004,
+  pages = {841--856},
+  alias = {DreSia04}
+}
+
+ +
+@article{DroJanWeg2006,
+  title = {Upper and lower bounds for randomized search heuristics in black-box optimization},
+  author = {Droste, Stefan and Jansen, Thomas and  Ingo Wegener },
+  journal = {Theory of Computing Systems},
+  volume = 39,
+  number = 4,
+  pages = {525--544},
+  year = 2006,
+  publisher = {Springer}
+}
+
+ +
+@article{DruThi2012,
+  author = { M{\u{a}}d{\u{a}}lina M. Drugan  and  Dirk Thierens },
+  title = {Stochastic {Pareto} local search: {Pareto} neighbourhood
+              exploration and perturbation strategies},
+  journal = {Journal of Heuristics},
+  volume = 18,
+  number = 5,
+  year = 2012,
+  pages = {727--766}
+}
+
+ +
+@article{DuLeu90,
+  author = { J. Du  and  Joseph Y.-T. Leung },
+  title = {Minimizing Total Tardiness on One Machine is
+                  {NP}-Hard},
+  journal = {Mathematics of Operations Research},
+  year = 1990,
+  volume = 15,
+  number = 3,
+  pages = {483--495}
+}
+
+ +
+@article{DubLopStu2011amai,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Improving the Anytime Behavior of Two-Phase Local
+                  Search},
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  year = 2011,
+  volume = 61,
+  number = 2,
+  pages = {125--154},
+  doi = {10.1007/s10472-011-9235-0},
+  alias = {DubLopStu2010amai}
+}
+
+ +
+@article{DubLopStu2011cor,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Hybrid {TP$+$PLS} Algorithm for Bi-objective
+                  Flow-Shop Scheduling Problems},
+  journal = {Computers \& Operations Research},
+  year = 2011,
+  volume = 38,
+  number = 8,
+  pages = {1219--1236},
+  doi = {10.1016/j.cor.2010.10.008},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2010-001/}
+}
+
+ +
+@article{DubLopStu2015ejor,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Anytime {Pareto} Local Search},
+  journal = {European Journal of Operational Research},
+  year = 2015,
+  volume = 243,
+  number = 2,
+  pages = {369--385},
+  doi = {10.1016/j.ejor.2014.10.062},
+  alias = {DubLopStu2013cor},
+  keywords = {Pareto local search}
+}
+
+ +
+@article{DubPagStu2017cor,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {An Iterated Greedy Algorithm with Optimization of Partial
+  Solutions for the Permutation Flowshop Problem},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2013-006},
+  volume = 81,
+  pages = {160--166},
+  doi = {10.1016/j.cor.2016.12.021}
+}
+
+ +
+@article{Dud2008carbodies,
+  author = {Fabian Duddeck},
+  title = {Multidisciplinary optimization of car bodies},
+  journal = {Structural and Multidisciplinary Optimization},
+  year = 2008,
+  volume = 35,
+  number = 4,
+  pages = {375--389},
+  doi = {10.1007/s00158-007-0130-6},
+  annote = {Evolutionary optimization of car bodies at General Motors}
+}
+
+ +
+@article{DueSch90,
+  author = { Gunter Dueck  and T. Scheuer},
+  title = {Threshold Accepting: A General Purpose Optimization
+                  Algorithm Appearing Superior to Simulated Annealing},
+  journal = {Journal of Computational Physics},
+  year = 1990,
+  volume = 90,
+  number = 1,
+  pages = {161--175}
+}
+
+ +
+@article{Dueck1993,
+  title = {New Optimization Heuristics: the Great Deluge Algorithm and the Record-To-Record Travel},
+  author = { Gunter Dueck },
+  journal = {Journal of Computational Physics},
+  volume = 104,
+  number = 1,
+  pages = {86--92},
+  year = 1993,
+  publisher = {Elsevier}
+}
+
+ +
+@article{DuiBerCar2017constrained,
+  title = {Constrained {Bayesian} Optimization with Particle Swarms for
+                  Safe Adaptive Controller Tuning},
+  journal = {{IFAC}-{PapersOnLine}},
+  volume = 50,
+  number = 1,
+  pages = {11800--11807},
+  year = 2017,
+  annote = {20th IFAC World Congress},
+  doi = {10.1016/j.ifacol.2017.08.1991},
+  author = {Rikky R. P. R. Duivenvoorden and Felix Berkenkamp and Nicolas
+                  Carion and Andreas Krause and Angela P. Schoellig},
+  keywords = {Adaptive Control, Constrained Bayesian Optimization, Safety,
+                  Gaussian Process, Particle Swarm Optimization, Policy Search,
+                  Reinforcement learning},
+  abstract = {Tuning controller parameters is a recurring and
+                  time-consuming problem in control. This is especially true in
+                  the field of adaptive control, where good performance is
+                  typically only achieved after significant tuning. Recently,
+                  it has been shown that constrained Bayesian optimization is a
+                  promising approach to automate the tuning process without
+                  risking system failures during the optimization
+                  process. However, this approach is computationally too
+                  expensive for tuning more than a couple of parameters. In
+                  this paper, we provide a heuristic in order to efficiently
+                  perform constrained Bayesian optimization in high-dimensional
+                  parameter spaces by using an adaptive discretization based on
+                  particle swarms. We apply the method to the tuning problem of
+                  an L1 adaptive controller on a quadrotor vehicle and show
+                  that we can reliably and automatically tune parameters in
+                  experiments.}
+}
+
+ +
+@article{DuiVos1999,
+  author = {Cees Duin and  Stefan Vo{\ss} },
+  title = {The Pilot Method: A Strategy for Heuristic Repetition with Application to the {Steiner} Problem in Graphs},
+  journal = {Networks},
+  year = 1999,
+  volume = 34,
+  number = 3,
+  pages = {181--191}
+}
+
+ +
+@article{Dumas95tsptw,
+  author = { Y. Dumas  and  J. Desrosiers  and  E. Gelinas  and  M. M. Solomon },
+  title = {An Optimal Algorithm for the Traveling Salesman Problem with
+                  Time Windows},
+  journal = {Operations Research},
+  year = 1995,
+  volume = 43,
+  number = 2,
+  pages = {367--371},
+  doi = {10.1287/opre.43.2.367}
+}
+
+ +
+@article{Dun1964tech,
+  title = {Multiple Comparisons Using Rank Sums},
+  author = { Olive Jean Dunn },
+  journal = {Technometrics},
+  volume = 6,
+  number = 3,
+  pages = {241--252},
+  year = 1964,
+  publisher = {Taylor \& Francis Group}
+}
+
+ +
+@article{Dunn1961jasa,
+  title = {Multiple Comparisons Among Means},
+  author = { Olive Jean Dunn },
+  journal = {Journal of the American Statistical Association},
+  volume = 56,
+  number = 293,
+  pages = {52--64},
+  year = 1961,
+  publisher = {Taylor \& Francis Group}
+}
+
+ +
+@article{DurNeb2011jmetal,
+  title = {{jMetal}: A {Java} framework for multi-objective
+                  optimization},
+  journal = {Advances in Engineering Software},
+  volume = 42,
+  number = 10,
+  pages = {760--771},
+  year = 2011,
+  issn = {0965-9978},
+  doi = {10.1016/j.advengsoft.2011.05.014},
+  author = { Durillo, Juan J.  and  Nebro, Antonio J. }
+}
+
+ +
+@article{EggLinHut2019jair,
+  title = {Pitfalls and best practices in algorithm configuration},
+  author = { Katharina Eggensperger  and  Marius Thomas Lindauer  and  Frank Hutter },
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 64,
+  pages = {861--893},
+  year = 2019
+}
+
+ +
+@article{Eglese1990,
+  title = {Simulated Annealing: a Tool for Operational Research},
+  author = { Richard W. Eglese },
+  journal = {European Journal of Operational Research},
+  volume = 46,
+  number = 3,
+  pages = {271--281},
+  year = 1990,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Ehr2006discussion,
+  title = {A discussion of scalarization techniques for multiple
+                  objective integer programming},
+  author = { Matthias Ehrgott },
+  journal = {Annals of Operations Research},
+  volume = 147,
+  number = 1,
+  pages = {343--360},
+  year = 2006,
+  publisher = {Springer}
+}
+
+ +
+@article{EhrGan04,
+  author = { Matthias Ehrgott  and  Xavier Gandibleux },
+  title = {Approximative Solution Methods for Combinatorial
+                  Multicriteria Optimization},
+  journal = {TOP},
+  year = 2004,
+  volume = 12,
+  number = 1,
+  pages = {1--88}
+}
+
+ +
+@article{EhrKla1997connectedness,
+  author = { Matthias Ehrgott  and  Kathrin Klamroth },
+  journal = {European Journal of Operational Research},
+  number = 1,
+  pages = {159--166},
+  title = {Connectedness of Efficient Solutions in Multiple Criteria
+                  Combinatorial Optimization},
+  volume = 97,
+  year = 1997,
+  doi = {10.1016/S0377-2217(96)00116-6}
+}
+
+ +
+@article{EibHinMich1999tec,
+  author = { Agoston E. Eiben  and Robert Hinterding and  Zbigniew Michalewicz },
+  title = {Parameter Control in Evolutionary Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 3,
+  number = 2,
+  pages = {124--141},
+  year = 1999
+}
+
+ +
+@article{EibRud1999,
+  title = {Theory of evolutionary algorithms: A bird's eye view},
+  author = { Agoston E. Eiben  and  G{\"u}nther Rudolph },
+  journal = {Theoretical Computer Science},
+  volume = 229,
+  number = {1-2},
+  pages = {3--9},
+  year = 1999
+}
+
+ +
+@article{EibSmi2011swec,
+  author = { Agoston E. Eiben  and  Smit, Selmar K. },
+  title = {Parameter Tuning for Configuring and Analyzing
+                  Evolutionary Algorithms},
+  journal = {Swarm and Evolutionary Computation},
+  year = 2011,
+  volume = 1,
+  number = 1,
+  pages = {19--31},
+  doi = {10.1016/j.swevo.2011.02.001}
+}
+
+ +
+@article{EkeKwa2018ems,
+  title = {Including robustness considerations in the search phase of
+                  Many-Objective Robust Decision Making},
+  author = {Eker, Sibel and  Kwakkel, Jan H. },
+  journal = {Environmental Modelling \& Software},
+  volume = 105,
+  pages = {201--216},
+  year = 2018,
+  keywords = {scenario-based}
+}
+
+ +
+@article{Elm1991rnn,
+  title = {Distributed representations, simple recurrent networks, and
+                  grammatical structure},
+  author = {Elman, Jeffrey L},
+  journal = {Machine Learning},
+  volume = 7,
+  number = {2-3},
+  pages = {195--225},
+  year = 1991,
+  publisher = {Springer}
+}
+
+ +
+@article{EmePer1991,
+  author = {V. A. Emelichev and V. A. Perepelitsa},
+  title = {Complexity of Vector Optimization Problems on Graphs},
+  journal = {Optimization},
+  year = 1991,
+  volume = 22,
+  number = 6,
+  doi = {10.1080/02331939108843732},
+  pages = {906--918}
+}
+
+ +
+@article{EmePer1992,
+  author = {V. A. Emelichev and V. A. Perepelitsa},
+  title = {On the Cardinality of the Set of Alternatives in Discrete Many-criterion Problems},
+  journal = {Discrete Mathematics and Applications},
+  year = 1992,
+  volume = 2,
+  number = 5,
+  pages = {461--471}
+}
+
+ +
+@article{EmmDeuw2018tutorial,
+  title = {A tutorial on multiobjective optimization: Fundamentals and
+                  evolutionary methods},
+  author = { Emmerich, Michael T. M.  and   Andr{\'{e}} H. Deutz },
+  journal = {Natural Computing},
+  year = 2018,
+  number = 3,
+  pages = {585--609},
+  volume = 17,
+  publisher = {Springer}
+}
+
+ +
+@article{EmmGiaNau2006,
+  author = { Emmerich, Michael T. M.  and K. C. Giannakoglou and  Boris Naujoks },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Single- and multiobjective evolutionary optimization assisted
+                  by {Gaussian} random field metamodels},
+  year = 2006,
+  volume = 10,
+  number = 4,
+  pages = {421--439},
+  doi = {10.1109/TEVC.2005.859463}
+}
+
+ +
+@article{EngWie2007twod,
+  title = {{2D} decision-making for multicriteria design optimization},
+  author = {Engau, Alexander and  Margaret M. Wiecek },
+  journal = {Structural and Multidisciplinary Optimization},
+  volume = 34,
+  pages = {301--315},
+  year = 2007,
+  doi = {10.1007/s00158-006-0078-y},
+  publisher = {Springer}
+}
+
+ +
+@article{EngWie2008interactive,
+  title = {Interactive coordination of objective decompositions in
+                  multiobjective programming},
+  author = {Engau, Alexander and  Margaret M. Wiecek },
+  journal = {Management Science},
+  volume = 54,
+  number = 7,
+  pages = {1350--1363},
+  year = 2008,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{EssMatDau2008,
+  author = {Imen Essafi and Yazid Mati and  St{\'e}phane Dauz{\`e}re-P{\`e}r{\'e}s },
+  title = {A Genetic Local Search Algorithm for Minimizing Total Weighted
+  Tardiness in the Job-shop Scheduling Problem},
+  journal = {Computers \& Operations Research},
+  year = 2008,
+  volume = 35,
+  number = 8,
+  pages = {2599--2616}
+}
+
+ +
+@article{FanBif2013,
+  author = {Fan, Wei and Bifet, Albert},
+  title = {Mining big data: current status, and forecast to the future},
+  journal = {{ACM} {SIGKDD} Explorations Newsletter},
+  year = 2013,
+  volume = 14,
+  number = 2,
+  pages = {1--5},
+  publisher = {ACM}
+}
+
+ +
+@article{Fanelli2012,
+  author = {Fanelli, Daniele},
+  title = {Negative results are disappearing from most disciplines and
+                  countries},
+  journal = {Scientometrics},
+  year = 2012,
+  volume = 90,
+  number = 3,
+  pages = {891--904},
+  doi = {10.1007/s11192-011-0494-7},
+  abstract = {Concerns that the growing competition for funding and
+                  citations might distort science are frequently discussed, but
+                  have not been verified directly. Of the hypothesized
+                  problems, perhaps the most worrying is a worsening of
+                  positive-outcome bias. A system that disfavours negative
+                  results not only distorts the scientific literature directly,
+                  but might also discourage high-risk projects and pressure
+                  scientists to fabricate and falsify their data. This study
+                  analysed over 4,600 papers published in all disciplines
+                  between 1990 and 2007, measuring the frequency of papers
+                  that, having declared to have ``tested'' a hypothesis,
+                  reported a positive support for it. The overall frequency of
+                  positive supports has grown by over 22{\%} between 1990 and
+                  2007, with significant differences between disciplines and
+                  countries. The increase was stronger in the social and some
+                  biomedical disciplines. The United States had published, over
+                  the years, significantly fewer positive results than Asian
+                  countries (and particularly Japan) but more than European
+                  countries (and in particular the United
+                  Kingdom). Methodological artefacts cannot explain away these
+                  patterns, which support the hypotheses that research is
+                  becoming less pioneering and/or that the objectivity with
+                  which results are produced and published is decreasing.}
+}
+
+ +
+@article{FarBinResFal2005tps,
+  author = { Faria, Jr, H.  and S. Binato and  Mauricio G. C. Resende  and  D. J. Falc{\~a}o },
+  title = {Power transmission network design by a greedy randomized
+                  adaptive path relinking approach},
+  journal = {IEEE Transactions on Power Systems},
+  year = 2005,
+  volume = 20,
+  number = 1,
+  pages = {43--49}
+}
+
+ +
+@article{FarMarYan2015pltoolbox,
+  author = {Farrugia, Vincent E. and Mart{\'i}nez, H{\'e}ctor P. and
+                  Yannakakis, Georgios N.},
+  title = {The Preference Learning Toolbox},
+  journal = {Arxiv preprint arXiv:1506.01709},
+  year = 2015,
+  doi = {10.48550/arXiv.1506.01709}
+}
+
+ +
+@article{FarWalSav2006hydroinf,
+  author = {R. Farmani and  Godfrey A. Walters  and  Dragan A. Savic },
+  title = {Evolutionary multi-objective optimization of the
+                  design and operation of water distribution network:
+                  total cost vs. reliability vs. water quality},
+  journal = { Journal of Hydroinformatics },
+  year = 2006,
+  volume = 8,
+  number = 3,
+  pages = {165--179}
+}
+
+ +
+@article{FavMorPel06tsptw,
+  author = { D. Favaretto  and  E. Moretti  and  Paola Pellegrini },
+  title = {Ant colony system approach for variants of the
+                  traveling salesman problem with time windows},
+  journal = {Journal of Information and Optimization Sciences},
+  year = 2006,
+  volume = 27,
+  number = 1,
+  pages = {35--54}
+}
+
+ +
+@article{FavMorPel07,
+  author = { D. Favaretto  and  E. Moretti  and  Paola Pellegrini },
+  title = {Ant Colony System for a {VRP} with Multiple Time Windows and Multiple Visits},
+  journal = {Journal of Interdisciplinary Mathematics},
+  year = 2007,
+  volume = 10,
+  number = 2,
+  pages = {263--284}
+}
+
+ +
+@article{FawHoos2015ablation,
+  title = {Analysing Differences Between Algorithm Configurations
+                  through Ablation},
+  author = { Chris Fawcett  and  Holger H. Hoos },
+  journal = {Journal of Heuristics},
+  pages = {431--458},
+  volume = 22,
+  number = 4,
+  year = 2016
+}
+
+ +
+@article{FeoRes1989,
+  author = { T. A. Feo  and  Mauricio G. C. Resende },
+  title = {A Probabilistic Heuristic for a Computationally Difficult Set
+                  Covering Problem},
+  journal = {Operations Research Letters},
+  year = 1989,
+  volume = 8,
+  number = 2,
+  pages = {67--71},
+  annote = {Proposed GRASP}
+}
+
+ +
+@article{FeoRes1995,
+  author = { T. A. Feo  and  Mauricio G. C. Resende },
+  title = {Greedy Randomized Adaptive Search Procedures},
+  journal = {Journal of Global Optimization},
+  year = 1995,
+  volume = 6,
+  number = 2,
+  pages = {109--113},
+  alias = {FR95}
+}
+
+ +
+@article{FeoResSmi1994,
+  author = { T. A. Feo  and  Mauricio G. C. Resende  and S. H. Smith},
+  title = {A Greedy Randomized Adaptive Search Procedure for Maximum
+                  Independent Set},
+  journal = {Operations Research},
+  year = 1994,
+  volume = 42,
+  pages = {860--878},
+  month = oct,
+  keywords = {GRASP}
+}
+
+ +
+@article{FerFra2014,
+  title = {On Insertion Tie-breaking Rules in Heuristics for the Permutation Flowshop Scheduling Problem},
+  author = { Victor Fernandez-Viagas  and  Jose M. Frami{\~n}{\'a}n },
+  journal = {Computers \& Operations Research},
+  year = 2014,
+  pages = {60--67},
+  volume = 45
+}
+
+ +
+@article{FerFra2017,
+  author = { Victor Fernandez-Viagas  and  Jose M. Frami{\~n}{\'a}n },
+  title = {A Beam-search-based Constructive Heuristic for the {PFSP} to Minimise
+               Total Flowtime},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  volume = 81,
+  pages = {167--177}
+}
+
+ +
+@article{FerFra2018,
+  author = { Victor Fernandez-Viagas  and  Jose M. Frami{\~n}{\'a}n },
+  title = {Iterated-greedy-based algorithms with beam search initialization for the permutation flowshop to minimise total tardiness},
+  journal = {Expert Systems with Applications},
+  year = 2018,
+  volume = 94,
+  pages = {58--69}
+}
+
+ +
+@article{FerGarAl2016mpe,
+  author = { Javier Ferrer  and  Jos{\'e} Garc{\'i}a-Nieto  and  Alba, Enrique  and  Chicano, Francisco },
+  doi = {10.1155/2016/3871046},
+  journal = {Mathematical Problems in Engineering},
+  pages = {1--19},
+  title = {Intelligent Testing of Traffic Light Programs: Validation in
+                  Smart Mobility Scenarios},
+  volume = 2016,
+  year = 2016
+}
+
+ +
+@article{FerGuiRamJua2016,
+  author = {Alberto Ferrer and Daniel Guimarans and  Helena {Ramalhinho Louren{\c c}o}  and Angel A. Juan},
+  title = {A {BRILS} Metaheuristic for Non-smooth Flow-shop Problems with Failure-risk
+               Costs},
+  journal = {Expert Systems with Applications},
+  year = 2016,
+  volume = 44,
+  pages = {177--186}
+}
+
+ +
+@article{FerLopAlb2019asoc,
+  author = { Javier Ferrer  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Alba, Enrique },
+  title = {Reliable Simulation-Optimization of Traffic Lights in a
+                  Real-World City},
+  journal = {Applied Soft Computing},
+  year = 2019,
+  volume = 78,
+  pages = {697--711},
+  doi = {10.1016/j.asoc.2019.03.016},
+  supplement = {https://github.com/MLopez-Ibanez/irace-sumo}
+}
+
+ +
+@article{FerNavBer2009:ejor,
+  author = { Eduardo Fernandez  and  Jorge Navarro  and  Sergio Bernal },
+  title = {Multicriteria Sorting Using a Valued Indifference Relation Under a Preference Disaggregation Paradigm},
+  journal = {European Journal of Operational Research},
+  volume = 198,
+  number = 2,
+  pages = {602--609},
+  year = 2009
+}
+
+ +
+@article{FerRuiFra2016,
+  author = { Victor Fernandez-Viagas  and  Rub{\'e}n Ruiz  and  Jose M. Frami{\~n}{\'a}n },
+  title = {A New Vision of Approximate Methods for the Permutation Flowshop to Minimise Makespan: State-of-the-art and Computational Evaluation},
+  journal = {European Journal of Operational Research},
+  volume = 257,
+  number = 3,
+  pages = {707--721},
+  year = 2017
+}
+
+ +
+@article{FerUrr2010do,
+  author = {R. {Ferreira da Silva} and S. Urrutia},
+  title = {A {General} {VNS} Heuristic for the Traveling Salesman
+                  Problem with Time Windows},
+  journal = {Discrete Optimization},
+  year = 2010,
+  volume = 7,
+  number = 4,
+  pages = {203--211},
+  keywords = {TSPTW, GVNS},
+  doi = {10.1016/j.disopt.2010.04.002}
+}
+
+ +
+@article{FerValFra2018,
+  author = { Victor Fernandez-Viagas  and  Jorge M. S. Valente  and  Jose M. Frami{\~n}{\'a}n },
+  title = {Iterated-greedy-based algorithms with Beam Search Initialization for the Permutation Flowshop to Minimise Total Tardiness},
+  journal = {Expert Systems with Applications},
+  volume = 94,
+  pages = {58--69},
+  year = 2018
+}
+
+ +
+@article{FiaDaCSchSeb2010:amai,
+  author = { {\'A}lvaro Fialho  and Luis {Da Costa} and  Marc Schoenauer  and  Mich{\`e}le Sebag },
+  title = {Analyzing Bandit-based Adaptive Operator Selection
+                  Mechanisms},
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  year = 2010,
+  volume = 60,
+  number = {1--2},
+  pages = {25--64}
+}
+
+ +
+@article{Fie2000:siamo,
+  author = { Mark J. Fielding },
+  title = {Simulated Annealing with an Optimal Fixed Temperature},
+  journal = {SIAM Journal on Optimization},
+  year = 2000,
+  volume = 11,
+  number = 2,
+  pages = {289--307}
+}
+
+ +
+@article{FieEveSing2003tec,
+  title = {Using unconstrained elite archives for multiobjective
+                  optimization},
+  author = { Jonathan E. Fieldsend  and  Everson, Richard M.  and Singh, Sameer},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 7,
+  number = 3,
+  pages = {305--323},
+  year = 2003,
+  doi = {10.1109/TEVC.2003.810733}
+}
+
+ +
+@article{FigFonHalKla2017easy,
+  title = {Easy to say they are Hard, but Hard to see they are
+                  Easy-Towards a Categorization of Tractable Multiobjective
+                  Combinatorial Optimization Problems},
+  author = { Jos{\'e} Rui Figueira  and  Carlos M. Fonseca  and Halffmann, Pascal and  Kathrin Klamroth  and  Lu{\'i}s Paquete  and Ruzika, Stefan and Schulze,
+                  Britta and Stiglmayr, Michael and Willems, David},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  volume = 24,
+  number = {1-2},
+  pages = {82--98},
+  year = 2017,
+  publisher = {Wiley Online Library},
+  doi = {10.1002/mcda.1574}
+}
+
+ +
+@article{FisBar2020improv,
+  author = { Andreas Fischbach  and  Thomas Bartz-Beielstein },
+  title = {Improving the reliability of test functions generators},
+  journal = {Applied Soft Computing},
+  year = 2020,
+  volume = 92,
+  pages = 106315
+}
+
+ +
+@article{FisGloLod2005,
+  title = {The feasibility pump},
+  author = { Matteo Fischetti  and  Fred Glover  and  Andrea Lodi },
+  journal = {Mathematical Programming},
+  volume = 104,
+  number = 1,
+  pages = {91--104},
+  year = 2005,
+  publisher = {Springer}
+}
+
+ +
+@article{FisLod2003:mp,
+  author = { Matteo Fischetti  and  Andrea Lodi },
+  title = {Local Branching},
+  journal = {Mathematical Programming Series B},
+  year = 2003,
+  volume = 98,
+  pages = {23--47}
+}
+
+ +
+@article{FisMon2014joh,
+  title = {Proximity search for 0-1 mixed-integer convex programming},
+  author = { Matteo Fischetti  and  Monaci, Michele },
+  journal = {Journal of Heuristics},
+  volume = 20,
+  number = 6,
+  pages = {709--731},
+  year = 2014,
+  publisher = {Springer}
+}
+
+ +
+@article{FisMon2014or,
+  author = { Matteo Fischetti  and  Monaci, Michele },
+  title = {Exploiting Erraticism in Search},
+  journal = {Operations Research},
+  volume = 62,
+  number = 1,
+  pages = {114--122},
+  year = 2014,
+  doi = {10.1287/opre.2013.1231},
+  annote = {\url{http://mat.tepper.cmu.edu/blog/?p=1695}},
+  abstract = { High sensitivity to initial conditions is generally viewed
+                  as a drawback of tree search methods because it leads to
+                  erratic behavior to be mitigated somehow. In this paper we
+                  investigate the opposite viewpoint and consider this behavior
+                  as an opportunity to exploit. Our working hypothesis is that
+                  erraticism is in fact just a consequence of the exponential
+                  nature of tree search that acts as a chaotic amplifier, so it
+                  is largely unavoidable. We propose a bet-and-run approach to
+                  actually turn erraticism to one's advantage. The idea is to
+                  make a number of short sample runs with randomized initial
+                  conditions, to bet on the "most promising" run selected
+                  according to certain simple criteria, and to bring it to
+                  completion. Computational results on a large testbed of mixed
+                  integer linear programs from the literature are presented,
+                  showing the potential of this approach even when embedded in
+                  a proof-of-concept implementation. }
+}
+
+ +
+@article{FisMonSal2012:or,
+  author = { Matteo Fischetti  and  Monaci, Michele  and Domenico Salvagnin},
+  title = {Three Ideas for the Quadratic Assignment Problem},
+  journal = {Operations Research},
+  year = 2012,
+  volume = 60,
+  number = 4,
+  pages = {954--964}
+}
+
+ +
+@article{FisSal2009,
+  title = {Feasibility pump 2.0},
+  author = { Matteo Fischetti  and Salvagnin, Domenico},
+  journal = {Mathematical Programming Computation},
+  volume = 1,
+  number = {2--3},
+  pages = {201--222},
+  year = 2009,
+  publisher = {Springer}
+}
+
+ +
+@article{Fle1970bfgs,
+  author = {Fletcher, Roger},
+  title = {A new approach to variable metric algorithms},
+  journal = {The Computer Journal},
+  year = 1970,
+  volume = 13,
+  number = 3,
+  pages = {317--322},
+  month = sep,
+  annote = {One of the four papers that proposed BFGS.},
+  doi = {10.1093/comjnl/13.3.317},
+  eprint = {https://academic.oup.com/comjnl/article-pdf/13/3/317/988678/130317.pdf},
+  keywords = {BFGS}
+}
+
+ +
+@article{FleGlo99,
+  title = {Improved constructive multistart strategies for the
+                  quadratic assignment problem using adaptive memory},
+  author = {Fleurent, Charles and  Fred Glover },
+  journal = {INFORMS Journal on Computing},
+  volume = 11,
+  number = 2,
+  pages = {198--204},
+  year = 1999
+}
+
+ +
+@article{Fli2007effects,
+  title = {The effects of adding objectives to an optimisation problem
+                  on the solution set},
+  author = {Fliege, J{\"o}rg},
+  journal = {Operations Research Letters},
+  volume = 35,
+  number = 6,
+  pages = {782--790},
+  year = 2007,
+  publisher = {Elsevier}
+}
+
+ +
+@article{FliVer1986,
+  title = {Distance Based Ranking Models},
+  author = {Fligner, Michael A. and Verducci, Joseph S.},
+  journal = {Journal of the Royal Statistical Society: Series B (Methodological)},
+  number = 3,
+  volume = 48,
+  pages = {359--369},
+  year = 1986,
+  keywords = {Mallows model, ranking, probabilistic models},
+  doi = {10.1111/j.2517-6161.1986.tb01420.x}
+}
+
+ +
+@article{Flo1956,
+  author = {M. M. Flood},
+  title = {The Travelling Salesman Problem},
+  journal = {Operations Research},
+  year = 1956,
+  volume = 4,
+  pages = {61--75}
+}
+
+ +
+@article{FloKel2010plos,
+  author = {Floreano, D. and Keller, L.},
+  journal = {PLoS Biology},
+  pages = {e1000292},
+  title = {Evolution of Adaptive Behaviour in Robots by Means
+                  of {Darwinian} Selection},
+  volume = 8,
+  number = 1,
+  year = 2010,
+  doi = {10.1371/journal.pbio.1000292},
+  alias = {Floreano:2010eg}
+}
+
+ +
+@article{FloUrz2000nn,
+  author = {Floreano, D. and Urzelai, J.},
+  title = {Evolutionary robots with on-line self-organization
+                  and behavioral fitness},
+  journal = {Neural Networks},
+  year = 2000,
+  volume = 13,
+  number = {4-5},
+  pages = {431--443},
+  alias = {Floreano:2000uy}
+}
+
+ +
+@article{Flores1986pragmatic,
+  author = {Flores, Benito E.},
+  title = {A pragmatic view of accuracy measurement in forecasting},
+  journal = {Omega},
+  year = 1986,
+  volume = 14,
+  number = 2,
+  pages = {93--98},
+  annote = {Proposed symmetric mean absolute percentage error (SMAPE)}
+}
+
+ +
+@article{FocLodMil02tsptw,
+  author = { Filippo Focacci  and  Andrea Lodi  and  Michela Milano },
+  title = {A Hybrid Exact Algorithm for the {TSPTW}},
+  journal = {INFORMS Journal on Computing},
+  year = 2002,
+  volume = 14,
+  pages = {403--417}
+}
+
+ +
+@article{FonFle1995ec,
+  title = {An overview of evolutionary algorithms in multiobjective
+                  optimization},
+  author = { Carlos M. Fonseca  and  Peter J. Fleming },
+  journal = {Evolutionary Computation},
+  year = 1995,
+  number = 1,
+  pages = {1--16},
+  volume = 3,
+  annote = {Proposed FON benchmark problem}
+}
+
+ +
+@article{FonFle1998:tsmca,
+  author = { Carlos M. Fonseca  and  Peter J. Fleming },
+  title = {Multiobjective Optimization and Multiple Constraint
+                  Handling with Evolutionary Algorithms ({II}):
+                  {Application} Example},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics -- Part A},
+  year = 1998,
+  volume = 28,
+  number = 1,
+  pages = {38--44},
+  month = jan,
+  doi = {10.1109/3468.650320}
+}
+
+ +
+@article{FonFle1998:tsmca1,
+  author = { Carlos M. Fonseca  and  Peter J. Fleming },
+  title = {Multiobjective Optimization and Multiple Constraint
+                  Handling with Evolutionary Algorithms ({I}): {A}
+                  Unified Formulation},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics -- Part A},
+  year = 1998,
+  volume = 28,
+  number = 1,
+  pages = {26--37},
+  month = jan,
+  doi = {10.1109/3468.650319}
+}
+
+ +
+@article{ForKea2009surrogate,
+  author = {Forrester, Alexander I. J. and Keane, Andy J.},
+  title = {Recent advances in surrogate-based optimization},
+  journal = {Progress in Aerospace Sciences},
+  volume = 45,
+  number = {1-3},
+  pages = {50--79},
+  doi = {10.1016/j.paerosci.2008.11.001},
+  year = 2009,
+  keywords = {Kriging; Gaussian Process; EGO; Design of Experiments}
+}
+
+ +
+@article{FowGelKok2010ejor,
+  title = {Interactive evolutionary multi-objective optimization for
+                  quasi-concave preference functions},
+  journal = {European Journal of Operational Research},
+  volume = 206,
+  number = 2,
+  pages = {417--425},
+  year = 2010,
+  doi = {10.1016/j.ejor.2010.02.027},
+  author = {John W. Fowler and Esma S. Gel and  Murat K{\"o}ksalan  and  Pekka Korhonen  and Jon L. Marquis and  Wallenius, Jyrki },
+  keywords = {Interactive optimization, Multi-objective optimization,
+                  Evolutionary optimization, Knapsack problem},
+  abstract = {We present a new hybrid approach to interactive evolutionary
+                  multi-objective optimization that uses a partial preference
+                  order to act as the fitness function in a customized genetic
+                  algorithm. We periodically send solutions to the decision
+                  maker (DM) for her evaluation and use the resulting
+                  preference information to form preference cones consisting of
+                  inferior solutions. The cones allow us to implicitly rank
+                  solutions that the DM has not considered. This technique
+                  avoids assuming an exact form for the preference function,
+                  but does assume that the preference function is
+                  quasi-concave. This paper describes the genetic algorithm and
+                  demonstrates its performance on the multi-objective knapsack
+                  problem.}
+}
+
+ +
+@article{Fox1993integrating,
+  author = { Bennett L. Fox },
+  title = {Integrating and accelerating tabu search, simulated
+                  annealing, and genetic algorithms},
+  journal = {Annals of Operations Research},
+  volume = 41,
+  number = 2,
+  pages = {47--67},
+  year = 1993,
+  publisher = {Springer}
+}
+
+ +
+@article{Fra2018tutorial,
+  author = {Peter I. Frazier},
+  title = {A Tutorial on {Bayesian} Optimization},
+  journal = {Arxiv preprint arXiv:1807.02811},
+  year = 2018,
+  doi = {10.48550/arXiv.1807.02811}
+}
+
+ +
+@article{Fra2022:4or,
+  title = {Empirical Analysis of Stochastic Local Search Behaviour: Connecting Structure, Components and Landscape},
+  author = { Alberto Franzin },
+  journal = {{4OR}: A Quarterly Journal of Operations Research},
+  year = 2022,
+  doi = {10.1007/s10288-022-00511-7}
+}
+
+ +
+@article{FraBraBru2014automode,
+  author = {G. Francesca and  M. Brambilla and 
+                  A. Brutschy and  Vito Trianni  and  Mauro Birattari },
+  title = {{AutoMoDe}: A Novel Approach to the Automatic Design
+                  of Control Software for Robot Swarms},
+  journal = {Swarm Intelligence},
+  year = 2014,
+  volume = 8,
+  number = 2,
+  pages = {89--112},
+  doi = {10.1007/s11721-014-0092-4}
+}
+
+ +
+@article{FraBraBru2015automode,
+  author = {Francesca, Gianpiero and Brambilla, Manuele and Brutschy,
+                  Arne and Garattoni, Lorenzo and Miletitch, Roman and
+                  Podevijn, Gaetan and Reina, Andreagiovanni and Soleymani,
+                  Touraj and Salvaro, Mattia and Pinciroli, Carlo and Mascia,
+                  Franco and  Vito Trianni  and  Mauro Birattari },
+  title = {{AutoMoDe-Chocolate}: Automatic Design of Control Software
+                  for Robot Swarms},
+  year = 2015,
+  journal = {Swarm Intelligence},
+  doi = {10.1007/s11721-015-0107-9},
+  keywords = {Swarm robotics; Automatic design; AutoMoDe},
+  language = {English}
+}
+
+ +
+@article{FraGupLei2004,
+  title = {A Review and Classification of Heuristics for Permutation Flow-shop Scheduling with Makespan Objective},
+  author = { Jose M. Frami{\~n}{\'a}n  and Jatinder N.D. Gupta and  Rainer Leisten },
+  journal = {Journal of the Operational Research Society},
+  year = 2004,
+  number = 12,
+  pages = {1243--1255},
+  volume = 55
+}
+
+ +
+@article{FraPerStu2018ol,
+  author = { Alberto Franzin  and   P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle },
+  title = {Effect of Transformations of Numerical Parameters
+                 in Automatic Algorithm Configuration},
+  journal = {Optimization Letters},
+  year = 2018,
+  volume = 12,
+  number = 8,
+  pages = {1741--1753},
+  doi = {10.1007/s11590-018-1240-3}
+}
+
+ +
+@article{FraSamDiC2016,
+  author = { Alberto Franzin  and Sambo, Francesco and Di Camillo, Barbara},
+  title = {\rpackage{bnstruct}: an {R} package for {Bayesian} Network structure learning
+                  in the presence of missing data},
+  journal = {Bioinformatics},
+  year = 2016,
+  volume = 33,
+  number = 8,
+  pages = {1250--1252}
+}
+
+ +
+@article{FraStu2019:cor,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Revisiting Simulated Annealing: A Component-Based Analysis},
+  journal = {Computers \& Operations Research},
+  volume = 104,
+  pages = {191--206},
+  year = 2019,
+  doi = {10.1016/j.cor.2018.12.015}
+}
+
+ +
+@article{FraStu2022:ejor,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {A Landscape-based Analysis of Fixed Temperature and Simulated Annealing},
+  year = 2023,
+  journal = {European Journal of Operational Research},
+  volume = 304,
+  number = 2,
+  pages = {395--410},
+  doi = {10.1016/j.ejor.2022.04.014}
+}
+
+ +
+@article{FreDue2007clust,
+  author = {Brendan J. Frey and Delbert Dueck},
+  title = {Clustering by Passing Messages Between Data Points},
+  year = 2007,
+  month = feb,
+  publisher = {American Association for the Advancement of Science ({AAAS})},
+  volume = 315,
+  number = 5814,
+  pages = {972--976},
+  doi = {10.1126/science.1136800},
+  journal = {Science},
+  keywords = {clustering; affinity propagation}
+}
+
+ +
+@article{FreFleGui2015aggregation,
+  title = {Aggregation trees for visualization and dimension reduction
+                  in many-objective optimization},
+  author = {de Freitas, Alan R. R. and  Peter J. Fleming  and Guimar{\~a}es, Frederico G.},
+  journal = {Information Sciences},
+  volume = 298,
+  pages = {288--314},
+  year = 2015,
+  publisher = {Elsevier}
+}
+
+ +
+@article{FriChaMar2010:ijor,
+  author = { Hela Frikha  and  Habib Chabchoub  and  Jean-Marc Martel },
+  title = {Inferring criteria's relative importance coefficients
+                  in {PROMETHEE II}},
+  journal = {International Journal of Operational Research},
+  year = 2010,
+  volume = 7,
+  number = 2,
+  pages = {257--275}
+}
+
+ +
+@article{FriJoh05FFTW,
+  author = {Matteo Frigo and Steven G. Johnson},
+  title = {The Design and Implementation of {FFTW3}},
+  journal = {Proceedings of the IEEE},
+  year = 2005,
+  volume = 93,
+  number = 2,
+  pages = {216--231},
+  note = {Special issue on ``Program Generation, Optimization, and Platform Adaptation''}
+}
+
+ +
+@article{Friedman1937use,
+  title = {The use of ranks to avoid the assumption of normality
+                  implicit in the analysis of variance},
+  author = {Friedman, Milton},
+  journal = {Journal of the American Statistical Association},
+  volume = 32,
+  number = 200,
+  pages = {675--701},
+  year = 1937
+}
+
+ +
+@article{FuEglLi2008,
+  author = {Z Fu and R Eglese and L Y O Li},
+  title = {A unified tabu search algorithm for vehicle routing problems with soft time windows},
+  journal = {Journal of the Operational Research Society},
+  volume = 59,
+  number = 5,
+  pages = {663--673},
+  year = 2008
+}
+
+ +
+@article{FueDoeHar09ejor,
+  author = { Guenther Fuellerer  and  Karl F. Doerner  and  Richard F. Hartl  and  Manuel Iori },
+  title = {Metaheuristics for vehicle routing problems with
+                  three-dimensional loading constraints},
+  journal = {European Journal of Operational Research},
+  year = 2009,
+  volume = 201,
+  number = 3,
+  pages = {751--759},
+  doi = {10.1016/j.ejor.2009.03.046}
+}
+
+ +
+@article{FueDoeHarIor09,
+  author = { Guenther Fuellerer  and  Karl F. Doerner  and  Richard F. Hartl  and  Manuel Iori },
+  title = {Ant colony optimization for the two-dimensional
+                  loading vehicle routing problem},
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  volume = 36,
+  number = 3,
+  pages = {655--673}
+}
+
+ +
+@article{Fuk2008ec,
+  title = {Automated Discovery of Local Search Heuristics for
+                  Satisfiability Testing},
+  author = { Fukunaga, Alex S. },
+  number = 1,
+  journal = {Evolutionary Computation},
+  month = mar,
+  year = 2008,
+  pages = {31--61},
+  volume = 16,
+  doi = {10.1162/evco.2008.16.1.31},
+  abstract = {The development of successful metaheuristic
+                  algorithms such as local search for a difficult
+                  problem such as satisfiability testing ({SAT)} is a
+                  challenging task. We investigate an evolutionary
+                  approach to automating the discovery of new local
+                  search heuristics for {SAT}. We show that several
+                  well-known {SAT} local search algorithms such as
+                  Walksat and Novelty are composite heuristics that
+                  are derived from novel combinations of a set of
+                  building blocks. Based on this observation, we
+                  developed {CLASS}, a genetic programming system that
+                  uses a simple composition operator to automatically
+                  discover {SAT} local search heuristics. New
+                  heuristics discovered by {CLASS} are shown to be
+                  competitive with the best Walksat variants,
+                  including Novelty+. Evolutionary algorithms have
+                  previously been applied to directly evolve a
+                  solution for a particular {SAT} instance. We show
+                  that the heuristics discovered by {CLASS} are also
+                  competitive with these previous, direct evolutionary
+                  approaches for {SAT}. We also analyze the local
+                  search behavior of the learned heuristics using the
+                  depth, mobility, and coverage metrics proposed by
+                  Schuurmans and Southey.}
+}
+
+ +
+@article{Fursin2011milepost,
+  author = {Grigori Fursin and Yuriy Kashnikov and Abdul Wahid Memon and Zbigniew Chamski and Olivier Temam and Mircea Namolaru and Elad Yom-Tov and Bilha Mendelson and Ayal Zaks and Eric Courtois and Francois Bodin and Phil Barnard and Elton Ashton and Edwin Bonilla and John Thomson and  Christopher K. I. Williams and Michael O'Boyle},
+  title = {Milepost {GCC}: Machine Learning Enabled Self-tuning Compiler},
+  journal = {International Journal of Parallel Programming},
+  year = 2011,
+  volume = 39,
+  number = 3,
+  pages = {296--327},
+  publisher = {Springer, US},
+  doi = {10.1007/s10766-010-0161-2}
+}
+
+ +
+@article{GagPriGra02:jors,
+  author = { Caroline Gagn{\'e}  and W. L. Price and M. Gravel},
+  title = {Comparing an {ACO} algorithm with other heuristics
+                  for the single machine scheduling problem with
+                  sequence-dependent setup times},
+  journal = {Journal of the Operational Research Society},
+  year = 2002,
+  volume = 53,
+  pages = {895--906}
+}
+
+ +
+@article{GagSch2007,
+  author = { Matteo Gagliolo  and J. Schmidhuber},
+  title = {Learning dynamic algorithm portfolios},
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  year = 2007,
+  volume = 47,
+  number = {3-4},
+  pages = {295--328},
+  doi = {10.1007/s10472-006-9036-z},
+  annote = {fully dynamic and online algorithm selection technique, with
+                  no separate training phase: all candidate algorithms are run
+                  in parallel, while a model incrementally learns their runtime
+                  distributions.}
+}
+
+ +
+@article{GalHao1999heacol,
+  title = {Hybrid evolutionary algorithms for graph coloring},
+  author = {Galinier, Philippe and  Jin-Kao Hao },
+  journal = {Journal of Combinatorial Optimization},
+  volume = 3,
+  number = 4,
+  pages = {379--397},
+  year = 1999,
+  publisher = {Springer},
+  doi = {10.1023/A:1009823419804}
+}
+
+ +
+@article{GalLeb1977ejor,
+  title = {Redundant objective functions in linear vector maximum
+                  problems and their determination},
+  journal = {European Journal of Operational Research},
+  volume = 1,
+  number = 3,
+  pages = {176--184},
+  year = 1977,
+  doi = {10.1016/0377-2217(77)90025-X},
+  author = {Tomas Gal and Heiner Leberling},
+  abstract = {Suppose that in a multicriteria linear programming problem
+                  among the given objective functions there are some which can
+                  be deleted without influencing the set E of all efficient
+                  solutions. Such objectives are said to be
+                  redundant. Introducing systems of objective functions which
+                  realize their individual optimum in a single vertex of the
+                  polyhedron generated by the restriction set, the notion of
+                  relative or absolute redundant objectives is defined. A
+                  theory which describes properties of absolute and relative
+                  redundant objectives is developed. A method for determining
+                  all the relative and absolute redundant objectives, based on
+                  this theory, is given. Illustrative examples demonstrate the
+                  procedure.}
+}
+
+ +
+@article{GamDor00:informs,
+  author = { L. M. Gambardella  and  Marco Dorigo },
+  title = {Ant {Colony} {System} Hybridized with a New Local
+                  Search for the Sequential Ordering Problem},
+  volume = 12,
+  number = 3,
+  pages = {237--255},
+  journal = {INFORMS Journal on Computing},
+  year = 2000,
+  anote = {IJ.26}
+}
+
+ +
+@article{GamMonWey12:ejor,
+  author = { L. M. Gambardella  and  Roberto Montemanni  and  Dennis Weyland },
+  title = {Coupling Ant Colony Systems with Strong Local Searches},
+  journal = {European Journal of Operational Research},
+  volume = 220,
+  number = 3,
+  year = 2012,
+  pages = {831--843},
+  doi = {10.1016/j.ejor.2012.02.038}
+}
+
+ +
+@article{GanJasFre2000joh,
+  author = { Xavier Gandibleux  and  Andrzej Jaszkiewicz  and  A. Fr{\'e}ville  and  Roman S{\l}owi{\'n}ski },
+  title = {Special Issue on {Multiple} {Objective} {Metaheuristics}},
+  journal = {Journal of Heuristics},
+  year = 2000,
+  volume = 6,
+  number = 3,
+  alias = {Gandibleux00}
+}
+
+ +
+@article{Gao2016,
+  author = {Gao, Kaizhou and Zhang, Yicheng and Sadollah, Ali and Su,
+                  Rong},
+  doi = {10.1016/j.asoc.2016.07.029},
+  journal = {Applied Soft Computing},
+  keywords = {harmony search algorithm,traffic light scheduling},
+  month = nov,
+  pages = {359--372},
+  title = {Optimizing urban traffic light scheduling problem using
+                  harmony search with ensemble of local search},
+  volume = 48,
+  year = 2016
+}
+
+ +
+@article{GaoNieLi2019visarxiv,
+  title = {Visualisation of {Pareto} Front Approximation: A Short Survey
+                  and Empirical Comparisons},
+  author = {Gao, Huiru and Nie, Haifeng and Li, Ke},
+  journal = {Arxiv preprint arXiv:1903.01768},
+  year = 2019,
+  url = {http://arxiv.org/abs/1903.01768}
+}
+
+ +
+@article{GarAlbOli2012,
+  author = { Jos{\'e} Garc{\'i}a-Nieto  and  Alba, Enrique  and  Olivera, Ana Carolina },
+  journal = {Engineering Applications of Artificial Intelligence},
+  keywords = {Cycle program optimization,Particle swarm
+                  optimization,Realistic traffic instances,SUMO microscopic
+                  simulator of urban mobility,Traffic light scheduling},
+  month = mar,
+  number = 2,
+  pages = {274--283},
+  title = {Swarm intelligence for traffic light scheduling: Application
+                  to real urban areas},
+  volume = 25,
+  year = 2012
+}
+
+ +
+@article{GarCorHer07,
+  author = { Carlos Garc{\'i}a-Mart{\'i}nez  and  Oscar Cord{\'o}n  and  Francisco Herrera },
+  title = {A taxonomy and an empirical analysis of multiple
+                  objective ant colony optimization algorithms for the
+                  bi-criteria {TSP}},
+  journal = {European Journal of Operational Research},
+  volume = 180,
+  number = 1,
+  year = 2007,
+  pages = {116--148}
+}
+
+ +
+@article{GarFer2015saferl,
+  title = {A comprehensive survey on safe reinforcement learning},
+  author = {Garc{\'i}a, Javier and Fern{\'a}ndez, Fernando},
+  journal = {Journal of Machine Learning Research},
+  volume = 16,
+  number = 1,
+  pages = {1437--1480},
+  year = 2015,
+  epub = {http://jmlr.org/papers/volume16/garcia15a/garcia15a.pdf}
+}
+
+ +
+@article{GarFerLueHer2010tests,
+  title = {Advanced nonparametric tests for multiple comparisons in the
+                  design of experiments in computational intelligence and data
+                  mining: Experimental analysis of power},
+  author = {Garc{\'i}a, Salvador and Fern{\'a}ndez, Alberto and Luengo,
+                  Juli{\'a}n and  Francisco Herrera },
+  journal = {Information Sciences},
+  volume = 180,
+  number = 10,
+  pages = {2044--2064},
+  year = 2010
+}
+
+ +
+@article{GarGloRodLozMar2014,
+  author = { Carlos Garc{\'i}a-Mart{\'i}nez  and  Fred Glover  and Francisco J. Rodr{\'i}guez and  Manuel Lozano  and  Rafael Mart{\'i} },
+  title = {Strategic Oscillation for the Quadratic Multiple Knapsack Problem},
+  journal = {Computational Optimization and Applications},
+  year = 2014,
+  volume = 58,
+  number = 1,
+  pages = {161--185}
+}
+
+ +
+@article{GarJohSet76,
+  author = {M. R. Garey and David S. Johnson and R. Sethi},
+  title = {The Complexity of Flowshop and Jobshop Scheduling},
+  journal = {Mathematics of Operations Research},
+  year = 1976,
+  volume = 1,
+  pages = {117--129}
+}
+
+ +
+@article{GarKal2001eff,
+  author = {Garnier, Josselin and Kallel, Leila},
+  title = {Efficiency of Local Search with Multiple Local Optima},
+  journal = {SIAM Journal Discrete Mathematics},
+  volume = 15,
+  number = 1,
+  pages = {122--141},
+  year = 2001,
+  doi = {10.1137/S0895480199355225}
+}
+
+ +
+@article{GarMolLoz2009joh,
+  author = {Garc{\'i}a, Salvador and  Daniel Molina  and  Manuel Lozano  and  Francisco Herrera },
+  journal = {Journal of Heuristics},
+  number = 617,
+  title = {A study on the use of non-parametric tests for analyzing the
+                  evolutionary algorithms' behaviour: a case study on the
+                  {CEC'2005} {Special} {Session} on {Real} {Parameter}
+                  {Optimization}},
+  volume = 15,
+  year = 2009,
+  pages = {617--644},
+  doi = {10.1007/s10732-008-9080-4}
+}
+
+ +
+@article{GarOliAlb2013tec,
+  author = { Jos{\'e} Garc{\'i}a-Nieto  and  Olivera, Ana Carolina  and  Alba, Enrique },
+  doi = {10.1109/TEVC.2013.2260755},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  month = dec,
+  number = 6,
+  pages = {823--839},
+  title = {Optimal Cycle Program of Traffic Lights With Particle Swarm
+                  Optimization},
+  volume = 17,
+  year = 2013
+}
+
+ +
+@article{GarRodLoz2012soco,
+  author = { Carlos Garc{\'i}a-Mart{\'i}nez  and Francisco J. Rodr{\'i}guez and  Manuel Lozano },
+  title = {Arbitrary function optimisation with metaheuristics: No free
+                  lunch and real-world problems},
+  journal = {Soft Computing},
+  year = 2012,
+  volume = 16,
+  number = 12,
+  pages = {2115--2133},
+  doi = {10.1007/s00500-012-0881-x}
+}
+
+ +
+@article{GarRodLoz2014,
+  author = { Carlos Garc{\'i}a-Mart{\'i}nez  and Francisco J. Rodr{\'i}guez and  Manuel Lozano },
+  title = {Tabu-enhanced Iterated Greedy Algorithm: A Case Study in the Quadratic Multiple Knapsack Problem},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 232,
+  number = 3,
+  pages = {454--463}
+}
+
+ +
+@article{GauDodGro2012gsa,
+  author = {Melvin, Gauci and Tony J. Dodd and  Roderich Gro{\ss} },
+  title = {Why `{GSA}: a gravitational search algorithm' is not
+                  genuinely based on the law of gravity},
+  journal = {Natural Computing},
+  year = 2012,
+  volume = 11,
+  number = 4,
+  pages = {719--720}
+}
+
+ +
+@article{Gei2011,
+  author = {Martin Josef Geiger},
+  title = {Decision Support for Multi-objective Flow Shop Scheduling
+               by the {Pareto} Iterated Local Search Methodology},
+  journal = {Computers and Industrial Engineering},
+  volume = 61,
+  number = 3,
+  pages = {805--812},
+  year = 2011
+}
+
+ +
+@article{Geiger2017:ejor,
+  author = {Martin Josef Geiger},
+  title = {A Multi-threaded Local Search Algorithm and Computer Implementation for the Multi-mode, Resource-constrained Multi-project Scheduling Problem},
+  journal = {European Journal of Operational Research},
+  year = 2017,
+  volume = 256,
+  pages = {729--741}
+}
+
+ +
+@article{GemGem1984,
+  author = { Stuart Geman  and  Donald Geman },
+  title = {Stochastic Relaxation, {Gibbs} Distributions, and the {Bayesian} Restoration of Images},
+  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
+  volume = 6,
+  number = 6,
+  pages = {721--741},
+  year = 1984,
+  publisher = {IEEE Press}
+}
+
+ +
+@article{GenGuePot1999ts,
+  title = {Parallel tabu search for real-time vehicle routing and dispatching},
+  author = { Michel Gendreau  and Francois Guertin and  Jean-Yves Potvin  and  {\'E}ric D. Taillard },
+  journal = {Transportation Science},
+  volume = 33,
+  number = 4,
+  pages = {381--390},
+  year = 1999
+}
+
+ +
+@article{GenGuePot2006trc,
+  title = {Neighborhood search heuristics for a dynamic vehicle dispatching problem with pick-ups and deliveries},
+  author = { Michel Gendreau  and Francois Guertin and  Jean-Yves Potvin  and S{\'e}guin, Ren{\'e}},
+  journal = {Transportation Research Part C: Emerging Technologies},
+  volume = 14,
+  number = 3,
+  pages = {157--174},
+  year = 2006,
+  publisher = {Elsevier}
+}
+
+ +
+@article{GenMit2014,
+  title = {Multiobjective evolutionary algorithm for manufacturing scheduling problems: state-of-the-art survey},
+  author = {Gen, Mitsuo and Lin, Lin},
+  journal = {Journal of Intelligent Manufacturing},
+  volume = 25,
+  number = 5,
+  pages = {849--866},
+  year = 2014,
+  publisher = {Springer}
+}
+
+ +
+@article{GenPogTul2010,
+  title = {Variable selection using random forests},
+  author = {Genuer, Robin and Poggi, Jean-Michel and Tuleau-Malot, Christine},
+  journal = {Pattern Recognition Letters},
+  volume = 31,
+  number = 14,
+  pages = {2225--2236},
+  year = 2010,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Gendreau98tsptw,
+  author = { Michel Gendreau  and A. Hertz  and  Gilbert Laporte  and  M. Stan },
+  title = {A Generalized Insertion Heuristic for the Traveling
+                  Salesman Problem with Time Windows},
+  journal = {Operations Research},
+  year = 1998,
+  volume = 46,
+  pages = {330--335}
+}
+
+ +
+@article{GerHorTen2015comp,
+  author = {Gershman, Samuel J. and Horvitz, Eric J. and Tenenbaum,
+                  Joshua B.},
+  title = {Computational rationality: A converging paradigm for
+                  intelligence in brains, minds, and machines},
+  volume = 349,
+  number = 6245,
+  pages = {273--278},
+  year = 2015,
+  doi = {10.1126/science.aac6076},
+  publisher = {American Association for the Advancement of Science},
+  abstract = {After growing up together, and mostly growing apart in the
+                  second half of the 20th century, the fields of artificial
+                  intelligence (AI), cognitive science, and neuroscience are
+                  reconverging on a shared view of the computational
+                  foundations of intelligence that promotes valuable
+                  cross-disciplinary exchanges on questions, methods, and
+                  results. We chart advances over the past several decades that
+                  address challenges of perception and action under uncertainty
+                  through the lens of computation. Advances include the
+                  development of representations and inferential procedures for
+                  large-scale probabilistic inference and machinery for
+                  enabling reflection and decisions about tradeoffs in effort,
+                  precision, and timeliness of computations. These tools are
+                  deployed toward the goal of computational rationality:
+                  identifying decisions with highest expected utility, while
+                  taking into consideration the costs of computation in complex
+                  real-world problems in which most relevant calculations can
+                  only be approximated. We highlight key concepts with examples
+                  that show the potential for interchange between computer
+                  science, cognitive science, and neuroscience.},
+  epub = {https://science.sciencemag.org/content/349/6245/273.full.pdf},
+  journal = {Science}
+}
+
+ +
+@article{GeuDamWeh2006extratrees,
+  author = {Pierre Geurts and Damien Ernst and Louis Wehenkel},
+  title = {Extremely randomized trees},
+  doi = {10.1007/s10994-006-6226-1},
+  year = 2006,
+  month = mar,
+  publisher = {Springer Science and Business Media {LLC}},
+  volume = 63,
+  number = 1,
+  pages = {3--42},
+  journal = {Machine Learning},
+  annote = {Proposed ExtraTrees}
+}
+
+ +
+@article{GheGobLeu2003gfs,
+  author = {Ghemawat, Sanjay and Gobioff, Howard and Leung, Shun-Tak},
+  title = {The {Google} {File} {System}},
+  journal = {SIGOPS Oper. Syst. Rev.},
+  volume = 37,
+  number = 5,
+  year = 2003,
+  month = dec,
+  pages = {29--43},
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@article{GhoNad09:asc,
+  author = { K. Ghoseiri  and  B. Nadjari },
+  title = {An ant colony optimization algorithm for the bi-objective
+                  shortest path problem},
+  journal = {Applied Soft Computing},
+  volume = 10,
+  number = 4,
+  year = 2010,
+  pages = {1237--1246}
+}
+
+ +
+@article{GirRabPib2016,
+  title = {Quantification of Treatment Effect Modification on Both an Additive and Multiplicative Scale},
+  author = {Girerd, Nicolas and Rabilloud, Muriel and Pibarot, Philippe
+                  and Mathieu, Patrick and Roy, Pascal},
+  journal = {PLoS One},
+  year = 2016,
+  month = apr,
+  volume = 11,
+  pages = {1--14},
+  number = 4,
+  doi = {10.1371/journal.pone.0153010}
+}
+
+ +
+@article{Glo1977,
+  author = { Fred Glover },
+  title = {Heuristics for Integer Programming Using Surrogate Constraints},
+  journal = {Decision Sciences},
+  year = 1977,
+  volume = 8,
+  pages = {156--166}
+}
+
+ +
+@article{Glo1986,
+  author = { Fred Glover },
+  title = {Future Paths for Integer Programming and Links to
+                  Artificial Intelligence},
+  journal = {Computers \& Operations Research},
+  year = 1986,
+  volume = 13,
+  number = 5,
+  pages = {533--549}
+}
+
+ +
+@article{Glo1989,
+  title = {Tabu Search -- {Part} {I}},
+  author = { Fred Glover },
+  journal = {INFORMS Journal on Computing},
+  volume = 1,
+  number = 3,
+  pages = {190--206},
+  year = 1989,
+  doi = {10.1287/ijoc.1.3.190}
+}
+
+ +
+@article{Glo1990,
+  author = { Fred Glover },
+  title = {Tabu Search -- {Part} {II}},
+  journal = {INFORMS Journal on Computing},
+  year = 1990,
+  number = 1,
+  volume = 2,
+  pages = {4--32}
+}
+
+ +
+@article{GloHao2011so,
+  author = { Fred Glover  and  Jin-Kao Hao },
+  title = {The case for Strategic Oscillation},
+  journal = {Annals of Operations Research},
+  year = 2011,
+  volume = 183,
+  number = 1,
+  pages = {163--173}
+}
+
+ +
+@article{GloKocAli98,
+  author = { Fred Glover  and  Gary A. Kochenberger  and Bahram Alidaee},
+  title = {Adaptive Memory Tabu Search for Binary Quadratic Programs},
+  journal = {Management Science},
+  year = 1998,
+  volume = 44,
+  number = 3,
+  pages = {336--345}
+}
+
+ +
+@article{GloLuHao2010diversif,
+  title = {Diversification-driven tabu search for unconstrained binary
+                  quadratic problems},
+  author = { Fred Glover  and L{\"u}, Zhipeng and  Jin-Kao Hao },
+  journal = {{4OR}: A Quarterly Journal of Operations Research},
+  volume = 8,
+  number = 3,
+  pages = {239--253},
+  year = 2010,
+  doi = {10.1007/s10288-009-0115-y}
+}
+
+ +
+@article{GoeScho2014recovery,
+  title = {Recovery-to-optimality: A new two-stage approach to
+                  robustness with an application to aperiodic timetabling},
+  author = {Goerigk, Marc and  Sch{\"o}bel, Anita },
+  journal = {Computers \& Operations Research},
+  volume = 52,
+  pages = {1--15},
+  year = 2014,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Gol1970bfgs,
+  author = {Donald Goldfarb},
+  title = {A Family of Variable-Metric Methods Derived by Variational
+                  Means},
+  journal = {Mathematics of Computation},
+  year = 1970,
+  volume = 24,
+  number = 109,
+  pages = {23--26},
+  annote = {One of the four papers that proposed BFGS.},
+  publisher = {American Mathematical Society},
+  eprint = {http://www.jstor.org/stable/2004873},
+  keywords = {BFGS}
+}
+
+ +
+@article{Gold1990pm,
+  title = {Probability matching, the magnitude of reinforcement, and
+                  classifier system bidding},
+  author = { David E. Goldberg },
+  journal = {Machine Learning},
+  volume = 5,
+  number = 4,
+  pages = {407--425},
+  year = 1990
+}
+
+ +
+@article{GonZhaChi2018kbs,
+  title = {The optimization ordering model for intuitionistic fuzzy
+                  preference relations with utility functions},
+  journal = {Knowledge-Based Systems},
+  volume = 162,
+  pages = {174--184},
+  year = 2018,
+  annote = {Special Issue on intelligent decision-making and consensus
+                  under uncertainty in inconsistent and dynamic environments},
+  issn = {0950-7051},
+  doi = {10.1016/j.knosys.2018.07.012},
+  author = {Zaiwu Gong and Ning Zhang and Francisco Chiclana},
+  keywords = {Intuitionistic fuzzy preference relation, Utility function,
+                  Ranking, Multiplicative consistency, Non-archimedean
+                  infinitesimal},
+  abstract = {Intuitionistic fuzzy sets describe information from the three
+                  aspects of membership degree, non-membership degree and
+                  hesitation degree, which has more practical significance when
+                  uncertainty pervades qualitative decision problems. In this
+                  paper, we investigate the problem of ranking intuitionistic
+                  fuzzy preference relations (IFPRs) based on various
+                  non-linear utility functions. First, we transform IFPRs into
+                  their isomorphic interval-value fuzzy preference relations
+                  (IVFPRs), and utilise non-linear utility functions, such as
+                  parabolic, S-shaped, and hyperbolic absolute risk aversion,
+                  to fit the true value of a decision-maker's
+                  judgement. Ultimately, the optimization ordering models for
+                  the membership and non-membership of IVFPRs based on utility
+                  function are constructed, with objective function aiming at
+                  minimizing the distance deviation between the multiplicative
+                  consistency ideal judgment and the actual judgment,
+                  represented by utility function, subject to the
+                  decision-maker's utility constraints. The proposed models
+                  ensure that more factual and optimal ranking of alternative
+                  is acquired, avoiding information distortion caused by the
+                  operations of intervals. Second, by introducing a
+                  non-Archimedean infinitesimal, we establish the optimization
+                  ordering model for IFPRs with the priority of utility or
+                  deviation, which realises the goal of prioritising solutions
+                  under multi-objective programming. Subsequently, we verify
+                  that a close connection exists between the ranking for
+                  membership and non-membership degree IVFPRs. Comparison
+                  analyses with existing approaches are summarized to
+                  demonstrate that the proposed models have advantage in
+                  dealing with group decision making problems with IFPRs.}
+}
+
+ +
+@article{GorKlaRuz2011connectedness,
+  author = {Gorski, Jochen and  Kathrin Klamroth  and Ruzika, Stefan},
+  journal = {Journal of Optimization Theory and Applications},
+  number = 3,
+  pages = {475--497},
+  publisher = {Springer},
+  title = {Connectedness of Efficient Solutions in Multiple Objective
+                  Combinatorial Optimization},
+  volume = 150,
+  year = 2011,
+  doi = {10.1007/s10957-011-9849-8}
+}
+
+ +
+@article{Gos2009rl,
+  author = {Gosavi, Abhijit},
+  title = {Reinforcement Learning: A Tutorial Survey and Recent
+                  Advances},
+  journal = {INFORMS Journal on Computing},
+  volume = 21,
+  number = 2,
+  pages = {178--192},
+  year = 2009,
+  doi = {10.1287/ijoc.1080.0305}
+}
+
+ +
+@article{GouOrbToi03,
+  author = {N. I. M. Gould and D. Orban and P. L. Toint},
+  title = {{CUTEr} and {SifDec}: A constrained and unconstrained testing
+                  environment, revisited},
+  journal = {ACM Transactions on Mathematical Software},
+  year = 2003,
+  volume = 29,
+  pages = {373--394}
+}
+
+ +
+@article{GraChi1996jair,
+  author = {Jonathan Gratch and Steve A. Chien},
+  title = {Adaptive Problem-solving for Large-scale Scheduling Problems:
+                  A Case Study},
+  journal = {Journal of Artificial Intelligence Research},
+  year = 1996,
+  volume = 4,
+  pages = {365--396},
+  annote = {Earliest hyper-heuristic?}
+}
+
+ +
+@article{GraHer2009treedgp,
+  author = {Robert B. Gramacy and Lee, Herbert K. H.},
+  title = {Bayesian Treed {Gaussian} Process Models With an Application
+                  to Computer Modeling},
+  journal = {Journal of the American Statistical Association},
+  volume = 103,
+  pages = {1119--1130},
+  year = 2008,
+  doi = {10.1198/016214508000000689},
+  keywords = {Treed-GP}
+}
+
+ +
+@article{GraJuaLou2016,
+  author = {Alex Grasas and Angel A. Juan and  Helena {Ramalhinho Louren{\c c}o} },
+  title = {{SimILS}: A Simulation-based Extension of the Iterated Local Search Metaheuristic for Stochastic Combinatorial Optimization},
+  journal = {Journal of Simulation},
+  year = 2016,
+  volume = 10,
+  number = 1,
+  pages = {69--77}
+}
+
+ +
+@article{GraPriGag02,
+  author = {M. Gravel and W. L. Price and  Caroline Gagn{\'e} },
+  title = {Scheduling continuous casting of aluminum using a
+                  multiple objective ant colony optimization
+                  metaheuristic},
+  journal = {European Journal of Operational Research},
+  year = 2002,
+  volume = 143,
+  number = 1,
+  pages = {218--229},
+  doi = {10.1016/S0377-2217(01)00329-0},
+  alias = {GraPriGag:ejor}
+}
+
+ +
+@article{Gre86,
+  author = {John J. Grefenstette},
+  title = {Optimization of Control Parameters for Genetic Algorithms},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics},
+  year = 1986,
+  volume = 16,
+  number = 1,
+  pages = {122--128},
+  keywords = {parameter tuning},
+  doi = {10.1109/TSMC.1986.289288}
+}
+
+ +
+@article{GreKadMouSlo2011:ejor,
+  author = { Salvatore Greco  and  Kadzi{\'n}ski, Mi{\l}osz   and  Vincent Mousseau  and  Roman S{\l}owi{\'n}ski },
+  title = {{ELECTRE}$^\mathrm{{GKMS}}$: Robust ordinal regression for outranking methods},
+  journal = {European Journal of Operational Research},
+  volume = 214,
+  number = 1,
+  pages = {118--135},
+  year = 2011
+}
+
+ +
+@article{GreMouSlo2014ejor,
+  author = { Salvatore Greco  and  Vincent Mousseau  and  Roman S{\l}owi{\'n}ski },
+  title = {Robust ordinal regression for value functions handling interacting
+               criteria},
+  journal = {European Journal of Operational Research},
+  volume = 239,
+  number = 3,
+  pages = {711--730},
+  year = 2014,
+  doi = {10.1016/j.ejor.2014.05.022}
+}
+
+ +
+@article{GriBauIoa2018modeling,
+  title = {Modelling science trustworthiness under publish or perish pressure},
+  author = {David R. Grimes and Chris T. Bauch and  John P. A. Ioannidis },
+  journal = {Royal Society Open Science},
+  volume = 5,
+  pages = {171511},
+  year = 2018
+}
+
+ +
+@article{GroDelTad2004,
+  author = { Andrea Grosso  and  Federico {Della Croce}  and R. Tadei},
+  title = {An Enhanced Dynasearch Neighborhood for the
+                  Single-Machine Total Weighted Tardiness Scheduling
+                  Problem},
+  journal = {Operations Research Letters},
+  year = 2004,
+  volume = 32,
+  number = 1,
+  pages = {68--72}
+}
+
+ +
+@article{GroJamLoc2009,
+  author = { Andrea Grosso  and A. R. M. J. U. Jamali and Marco Locatelli},
+  title = {Finding Maximin Latin Hypercube Designs by Iterated Local Search Heuristics},
+  journal = {European Journal of Operational Research},
+  year = 2009,
+  volume = 197,
+  number = 2,
+  pages = {541--547}
+}
+
+ +
+@article{GroKayKnoVan2013,
+  title = {The "big data" revolution in healthcare},
+  author = {Groves, Peter and Kayyali, Basel and Knott, David and Van
+                  Kuiken, Steve},
+  journal = {McKinsey Quarterly},
+  volume = 2,
+  year = 2013
+}
+
+ +
+@article{GroMan2019hvsubset,
+  title = {Hypervolume subset selection with small subsets},
+  author = {Groz, Beno{\^i}t and Maniu, Silviu},
+  journal = {Evolutionary Computation},
+  year = 2019,
+  number = 4,
+  pages = {611--637},
+  volume = 27
+}
+
+ +
+@article{GruFon2002spl,
+  author = { Viviane {Grunert da Fonseca}  and  Carlos M. Fonseca },
+  title = {A link between the multivariate cumulative distribution
+                  function and the hitting function for random closed sets},
+  journal = {Statistics \& Probability Letters},
+  year = 2002,
+  volume = 57,
+  number = 2,
+  pages = {179--182},
+  alias = {Fonseca02a},
+  doi = {10.1016/S0167-7152(02)00046-9}
+}
+
+ +
+@article{GueFonPaq2021hv,
+  title = {The Hypervolume Indicator: Computational Problems and Algorithms},
+  author = { Andreia P. Guerreiro  and  Carlos M. Fonseca  and  Lu{\'i}s Paquete },
+  journal = {{ACM} Computing Surveys},
+  year = 2021,
+  number = 6,
+  pages = {1--42},
+  volume = 54
+}
+
+ +
+@article{GueManFig2021exacthv,
+  author = { Andreia P. Guerreiro  and Vasco Manquinho and  Jos{\'e} Rui Figueira },
+  title = {Exact hypervolume subset selection through incremental
+                  computations},
+  doi = {10.1016/j.cor.2021.105471},
+  year = 2021,
+  month = dec,
+  volume = 136,
+  pages = {105--471},
+  journal = {Computers \& Operations Research}
+}
+
+ +
+@article{Gui2011objred,
+  author = {Gonzalo Guill{\'e}n-Gos{\'a}lbez},
+  title = {A novel {MILP}-based objective reduction method for
+                  multi-objective optimization: Application to environmental
+                  problems},
+  journal = {Computers \& Chemical Engineering},
+  volume = 35,
+  number = 8,
+  pages = {1469--1477},
+  year = 2011,
+  issn = {0098-1354},
+  doi = {10.1016/j.compchemeng.2011.02.001},
+  keywords = {Environmental engineering, Life cycle assessment,
+                  Multi-objective optimization, Objective reduction},
+  abstract = {Multi-objective optimization has recently emerged as a useful
+                  technique in sustainability analysis, as it can assist in the
+                  study of optimal trade-off solutions that balance several
+                  criteria. The main limitation of multi-objective optimization
+                  is that its computational burden grows in size with the
+                  number of objectives. This computational barrier is critical
+                  in environmental applications in which decision-makers seek
+                  to minimize simultaneously several environmental indicators
+                  of concern. With the aim to overcome this limitation, this
+                  paper introduces a systematic method for reducing the number
+                  of objectives in multi-objective optimization with emphasis
+                  on environmental problems. The approach presented relies on a
+                  novel mixed-integer linear programming formulation that
+                  minimizes the error of omitting objectives. We test the
+                  capabilities of this technique through two environmental
+                  problems of different nature in which we attempt to minimize
+                  a set of life cycle assessment impacts. Numerical examples
+                  demonstrate that certain environmental metrics tend to behave
+                  in a non-conflicting manner, which makes it possible to
+                  reduce the dimension of the problem without losing
+                  information.}
+}
+
+ +
+@article{GunGilAha2018repro,
+  author = {Odd Erik Gundersen and Yolanda Gil and David W. Aha},
+  title = {On Reproducible {AI}: Towards Reproducible Research, Open
+                  Science, and Digital Scholarship in {AI} Publications},
+  doi = {10.1609/aimag.v39i3.2816},
+  year = 2018,
+  month = sep,
+  publisher = {Association for the Advancement of Artificial Intelligence
+                  ({AAAI})},
+  volume = 39,
+  number = 3,
+  pages = {56--68},
+  journal = {{AI} Magazine},
+  annote = {The reproducibility guidelines can be found here:
+                  \url{https://folk.idi.ntnu.no/odderik/reproducibility_guidelines.pdf}
+                  and a short how-to can be found here:
+                  \url{https://folk.idi.ntnu.no/odderik/reproducibility_guidelines_how_to.html}}
+}
+
+ +
+@article{GunNgPoh2012,
+  title = {A Hybridized {Lagrangian} Relaxation and Simulated Annealing
+                  Method for the Course Timetabling Problem},
+  author = { Aldy Gunawan  and  Ng, Kien Ming  and  Poh, Kim Leng },
+  journal = {Computers \& Operations Research},
+  volume = 39,
+  number = 12,
+  pages = {3074--3088},
+  year = 2012,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Gup1986,
+  title = {Flowshop schedules with sequence dependent setup times},
+  author = {J. N. D. Gupta},
+  journal = {Journal of Operations Research Society of Japan},
+  volume = 29,
+  year = 1986,
+  pages = {206--219}
+}
+
+ +
+@article{Gut2000:fgcs,
+  author = { Gutjahr, Walter J. },
+  title = {A {Graph}-based {Ant} {System} and its Convergence},
+  journal = {Future Generation Computer Systems},
+  year = 2000,
+  volume = 16,
+  number = 8,
+  pages = {873--888}
+}
+
+ +
+@article{Gut2002:ipl,
+  author = { Gutjahr, Walter J. },
+  title = {{ACO} Algorithms with Guaranteed Convergence to the
+                  Optimal Solution},
+  journal = {Information Processing Letters},
+  year = 2002,
+  volume = 82,
+  number = 3,
+  pages = {145--153}
+}
+
+ +
+@article{Gut2006:mcap,
+  author = { Gutjahr, Walter J. },
+  title = {On the finite-time dynamics of ant colony
+                  optimization},
+  journal = {Methodology and Computing in Applied Probability},
+  year = 2006,
+  volume = 8,
+  number = 1,
+  pages = {105--133}
+}
+
+ +
+@article{Gut2007:swarm,
+  author = { Gutjahr, Walter J. },
+  title = {Mathematical runtime analysis of {ACO} algorithms:
+                  survey on an emerging issue},
+  journal = {Swarm Intelligence},
+  volume = 1,
+  number = 1,
+  year = 2007,
+  pages = {59--79}
+}
+
+ +
+@article{Gut2007cor,
+  title = {An {ACO} algorithm for a dynamic regional nurse-scheduling
+                  problem in {Austria} },
+  journal = {Computers \& Operations Research},
+  volume = 34,
+  number = 3,
+  pages = {642--666},
+  year = 2007,
+  anote = {Logistics of Health Care Management Part Special Issue:
+                  Logistics of Health Care Management },
+  doi = {10.1016/j.cor.2005.03.018},
+  author = { Gutjahr, Walter J.  and  Marion S. Rauner},
+  abstract = {To the best of our knowledge, this paper describes the first
+                  ant colony optimization (ACO) approach applied to nurse
+                  scheduling, analyzing a dynamic regional problem which is
+                  currently under discussion at the Vienna hospital
+                  compound. Each day, pool nurses have to be assigned for the
+                  following days to public hospitals while taking into account
+                  a variety of soft and hard constraints regarding working date
+                  and time, working patterns, nurses qualifications, nurses
+                  and hospitals preferences, as well as costs. Extensive
+                  computational experiments based on a four week simulation
+                  period were used to evaluate three different scenarios
+                  varying the number of nurses and hospitals for six different
+                  hospitals demand intensities. The results of our simulations
+                  and optimizations reveal that the proposed {ACO} algorithm
+                  achieves highly significant improvements compared to a greedy
+                  assignment algorithm.}
+}
+
+ +
+@article{Gut2008:cor,
+  author = { Gutjahr, Walter J. },
+  title = {First steps to the runtime complexity analysis of ant colony
+               optimization},
+  journal = {Computers \& Operations Research},
+  volume = 35,
+  number = 9,
+  year = 2008,
+  pages = {2711--2727}
+}
+
+ +
+@article{GutSeb2008,
+  author = { Gutjahr, Walter J.  and G. Sebastiani},
+  title = {Runtime analysis of ant colony optimization with best-so-far
+                  reinforcement},
+  journal = {Methodology and Computing in Applied Probability},
+  year = 2008,
+  volume = 10,
+  number = 3,
+  pages = {409--433}
+}
+
+ +
+@article{GutYeoZve2002,
+  author = {Gutin, Gregory and Yeo, Anders and Zverovich, Alexey},
+  title = {Traveling salesman should not be greedy: domination analysis
+                  of greedy-type heuristics for the {TSP}},
+  journal = {Discrete Applied Mathematics},
+  volume = 117,
+  number = {1--3},
+  year = 2002
+}
+
+ +
+@article{GuyWesBar2002rfe,
+  title = {Gene selection for cancer classification using support vector
+                  machines},
+  author = {Guyon, Isabelle and Weston, Jason and Barnhill, Stephen and
+                  Vapnik, Vladimir},
+  journal = {Machine Learning},
+  volume = 46,
+  number = 1,
+  pages = {389--422},
+  year = 2002,
+  publisher = {Springer},
+  keywords = {recursive feature elimination}
+}
+
+ +
+@article{HaaSakTam2001,
+  title = {An adaptive {Metropolis} algorithm},
+  author = {Haario, Heikki and Saksman, Eero and Tamminen, Johanna},
+  journal = {Bernoulli},
+  volume = 7,
+  number = 2,
+  pages = {223--242},
+  year = 2001
+}
+
+ +
+@article{HadRee2013borg,
+  author = { David Hadka  and  Patrick M. Reed },
+  title = {Borg: An Auto-Adaptive Many-Objective Evolutionary Computing
+                  Framework},
+  journal = {Evolutionary Computation},
+  number = 2,
+  pages = {231--259},
+  volume = 21,
+  year = 2013,
+  alias = {Hadka13borg}
+}
+
+ +
+@article{HadReed2012ec,
+  author = { David Hadka  and  Patrick M. Reed },
+  title = {Diagnostic Assessment of Search Controls and Failure Modes in
+                  Many-Objective Evolutionary Optimization},
+  journal = {Evolutionary Computation},
+  volume = 20,
+  number = 3,
+  year = 2012,
+  pages = {423--452}
+}
+
+ +
+@article{HadRus1969rules,
+  title = {Rules for ordering uncertain prospects},
+  author = {Hadar, Josef and Russell, William R.},
+  journal = {The American Economic Review},
+  volume = 59,
+  number = 1,
+  pages = {25--34},
+  year = 1969,
+  epub = {https://www.jstor.org/stable/1811090},
+  keywords = {stochastic dominance}
+}
+
+ +
+@article{HaiLasWis1971bicriterion,
+  title = {On a bicriterion formation of the problems of integrated
+                  system identification and system optimization},
+  author = {Haimes, Y. and Lasdon, L. and Da Wismer, D.},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics},
+  volume = 1,
+  number = 3,
+  pages = {296--297},
+  year = 1971,
+  doi = {10.1109/TSMC.1971.4308298},
+  keywords = {epsilon-constraint method}
+}
+
+ +
+@article{HajLin1992decomposition,
+  title = {Genetic search strategies in multicriterion optimal design},
+  author = {Hajela, Prabhat and Lin, C-Y},
+  journal = {Structural Optimization},
+  volume = 4,
+  number = 2,
+  pages = {99--107},
+  year = 1992,
+  publisher = {Springer}
+}
+
+ +
+@article{HajSas1989,
+  title = {Simulated annealing--to cool or not},
+  author = { Bruce Hajek  and Galen Sasaki},
+  journal = {System \& Control Letters},
+  volume = 12,
+  number = 5,
+  pages = {443--447},
+  year = 1989,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Hajek1988,
+  title = {Cooling Schedules for Optimal Annealing},
+  author = { Bruce Hajek },
+  journal = {Mathematics of Operations Research},
+  volume = 13,
+  number = 2,
+  pages = {311--329},
+  year = 1988,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{HalOliSud2022ai,
+  author = { George T. Hall  and  Oliveto, Pietro S.  and  Dirk Sudholt },
+  title = {On the impact of the performance metric on efficient
+                  algorithm configuration},
+  doi = {10.1016/j.artint.2021.103629},
+  year = 2022,
+  month = feb,
+  publisher = {Elsevier {BV}},
+  volume = 303,
+  pages = 103629,
+  journal = {Artificial Intelligence},
+  keywords = {irace}
+}
+
+ +
+@article{HamLah2016path,
+  title = {Path dependence in {Operational} {Research}--{How} the
+                  modeling process can influence the results},
+  author = { H{\"a}m{\"a}l{\"a}inen, Raimo P.  and Lahtinen, Tuomas J.},
+  doi = {10.1016/j.orp.2016.03.001},
+  abstract = {In Operational Research practice there are almost always
+                  alternative paths that can be followed in the modeling and
+                  problem solving process. Path dependence refers to the impact
+                  of the path on the outcome of the process. The steps of the
+                  path include, e.g. forming the problem solving team, the
+                  framing and structuring of the problem, the choice of model,
+                  the order in which the different parts of the model are
+                  specified and solved, and the way in which data or
+                  preferences are collected. We identify and discuss seven
+                  possibly interacting origins or drivers of path dependence:
+                  systemic origins, learning, procedure, behavior, motivation,
+                  uncertainty, and external environment. We provide several
+                  ideas on how to cope with path dependence.},
+  journal = {Operations Research Perspectives},
+  month = jan,
+  volume = 3,
+  year = 2016,
+  keywords = {Behavioral Biases, Behavioral Operational Research, Ethics in
+                  modelling, OR practice, Path dependence, Systems perspective},
+  pages = {14--20}
+}
+
+ +
+@article{HamLuoSaa2013bor,
+  author = { H{\"a}m{\"a}l{\"a}inen, Raimo P.  and Luoma, Jukka and Saarinen, Esa},
+  title = {On the importance of behavioral operational research: {The}
+                  case of understanding and communicating about dynamic
+                  systems},
+  volume = 228,
+  shorttitle = {On the importance of behavioral operational research},
+  doi = {10.1016/j.ejor.2013.02.001},
+  abstract = {We point out the need for Behavioral Operational Research
+                  (BOR) in advancing the practice of OR. So far, in OR
+                  behavioral phenomena have been acknowledged only in
+                  behavioral decision theory but behavioral issues are always
+                  present when supporting human problem solving by
+                  modeling. Behavioral effects can relate to the group
+                  interaction and communication when facilitating with OR
+                  models as well as to the possibility of procedural mistakes
+                  and cognitive biases. As an illustrative example we use well
+                  known system dynamics studies related to the understanding of
+                  accumulation. We show that one gets completely opposite
+                  results depending on the way the phenomenon is described and
+                  how the questions are phrased and graphs used. The results
+                  suggest that OR processes are highly sensitive to various
+                  behavioral effects. As a result, we need to pay attention to
+                  the way we communicate about models as they are being
+                  increasingly used in addressing important problems like
+                  climate change.},
+  number = 3,
+  journal = {European Journal of Operational Research},
+  month = aug,
+  year = 2013,
+  pages = {623--634}
+}
+
+ +
+@article{HamRuh1994,
+  author = {Hamacher, Horst W. and Ruhe, G\"{u}nter},
+  title = {On spanning tree problems with multiple objectives},
+  journal = {Annals of Operations Research},
+  year = 1994,
+  volume = 52,
+  number = 4,
+  pages = {209--230}
+}
+
+ +
+@article{HanAugBroTus2022anytime,
+  author = { Nikolaus Hansen  and  Anne Auger  and  Dimo Brockhoff  and  Tea Tu{\v s}ar },
+  title = {Anytime Performance Assessment in Blackbox Optimization
+                  Benchmarking},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2022,
+  volume = 26,
+  number = 6,
+  pages = {1293--1305},
+  month = dec,
+  doi = {10.1109/TEVC.2022.3210897}
+}
+
+ +
+@article{HanAugMer2016coco,
+  title = {{COCO}: A platform for comparing continuous optimizers in a
+                  black-box setting},
+  author = { Nikolaus Hansen  and  Anne Auger  and Mersmann, Olaf and  Tea Tu{\v s}ar  and  Dimo Brockhoff },
+  journal = {Arxiv preprint arXiv:1603.08785},
+  year = 2016,
+  note = {Published as \cite{HanAugMer2020coco}}
+}
+
+ +
+@article{HanAugMer2020coco,
+  title = {{COCO}: A platform for comparing continuous optimizers in a
+                  black-box setting},
+  author = { Nikolaus Hansen  and  Anne Auger  and Ros, Raymond and Mersmann,
+                  Olaf and  Tea Tu{\v s}ar  and  Dimo Brockhoff },
+  journal = {Optimization Methods and Software},
+  pages = {1--31},
+  year = 2020,
+  volume = 36,
+  number = 1,
+  doi = {10.1080/10556788.2020.1808977},
+  publisher = {Taylor \& Francis}
+}
+
+ +
+@article{HanJau90,
+  author = { Pierre Hansen  and B. Jaumard},
+  title = {Algorithms for the Maximum Satisfiability Problem},
+  journal = {Computing},
+  year = 1990,
+  volume = 44,
+  pages = {279--303}
+}
+
+ +
+@article{HanMla01:ejor,
+  title = {Variable neighborhood search: Principles and applications},
+  author = { Pierre Hansen  and  Nenad Mladenovi{\'c} },
+  journal = {European Journal of Operational Research},
+  volume = 130,
+  number = 3,
+  pages = {449--467},
+  year = 2001
+}
+
+ +
+@article{HanOst2001ec,
+  author = { Nikolaus Hansen  and Ostermeier, A.},
+  title = {Completely derandomized self-adaptation in evolution
+                  strategies},
+  journal = {Evolutionary Computation},
+  year = 2001,
+  volume = 9,
+  pages = {159--195},
+  number = 2,
+  doi = {10.1162/106365601750190398},
+  keywords = {CMA-ES}
+}
+
+ +
+@article{HanRosMauSchAug2011,
+  author = { Nikolaus Hansen  and Raymond Ros and Nikolaus Mauny and  Marc Schoenauer  and  Anne Auger },
+  title = {Impacts of invariance in search: When {CMA-ES} and {PSO} face ill-conditioned and non-separable problems},
+  journal = {Applied Soft Computing},
+  year = 2011,
+  volume = 11,
+  number = 8,
+  pages = {5755--5769}
+}
+
+ +
+@article{Hanne1999ejor,
+  author = {Thomas Hanne},
+  journal = {European Journal of Operational Research},
+  title = {On the convergence of multiobjective evolutionary algorithms},
+  volume = 117,
+  number = 3,
+  pages = {553--564},
+  year = 1999,
+  doi = {10.1016/S0377-2217(98)00262-8},
+  keywords = {archiving, efficiency presserving}
+}
+
+ +
+@article{Hanne2007ejor,
+  title = {A multiobjective evolutionary algorithm for approximating the
+                  efficient set},
+  author = {Hanne, Thomas},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  number = 3,
+  pages = {1723--1734},
+  volume = 176
+}
+
+ +
+@article{HarSaf2004,
+  author = {Hardin, Douglas P. and Saff, Edward B.},
+  title = {Discretizing Manifolds via Minimum Energy Points},
+  journal = {Notices of the American Mathematical Society},
+  year = 2004,
+  number = 10,
+  pages = {1186--1194},
+  volume = 51,
+  alias = {Hardin2004}
+}
+
+ +
+@article{HarSho87a,
+  author = {J. P. Hart and A. W. Shogan},
+  title = {Semi-greedy heuristics: An empirical study},
+  journal = {Operations Research Letters},
+  volume = 6,
+  number = 3,
+  pages = {107--114},
+  year = 1987
+}
+
+ +
+@article{HarSim2016ec,
+  author = { Emma Hart  and Kevin Sim},
+  title = {A Hyper-Heuristic Ensemble Method for Static Job-Shop Scheduling},
+  journal = {Evolutionary Computation},
+  volume = 24,
+  number = 4,
+  pages = {609--635},
+  year = 2016,
+  doi = {10.1162/EVCO_a_00183 }
+}
+
+ +
+@article{Haraguchi2016:joh,
+  author = {Kazuya Haraguchi},
+  title = {Iterated Local Search with {Trellis}-Neighborhood for the Partial {Latin} Square Extension Problem},
+  journal = {Journal of Heuristics},
+  year = 2016,
+  volume = 22,
+  number = 5,
+  pages = {727--757}
+}
+
+ +
+@article{HasRaj2004,
+  author = { Sameer   Hasija  and  Chandrasekharan   Rajendran },
+  title = {Scheduling in flowshops to minimize total tardiness of jobs},
+  journal = {International Journal of Production Research},
+  volume = 42,
+  number = 11,
+  pages = {2289--2301},
+  year = 2004,
+  doi = {10.1080/00207540310001657595}
+}
+
+ +
+@article{HasYagIba2008:do,
+  author = {Hideki Hashimoto and  Mutsunori Yagiura  and  Toshihide Ibaraki },
+  title = {An Iterated Local Search Algorithm for the Time-dependent Vehicle
+               Routing Problem with Time Windows},
+  journal = {Discrete Optimization},
+  year = 2008,
+  volume = 5,
+  number = 2,
+  pages = {434--456}
+}
+
+ +
+@article{Haykin2004nn,
+  title = {A comprehensive foundation},
+  author = {Haykin, Simon},
+  journal = {Neural Networks},
+  volume = 2,
+  pages = 41,
+  year = 2004
+}
+
+ +
+@article{HazGunEre08customer,
+  title = {Customer order scheduling problem: a comparative
+                  metaheuristics study},
+  number = 5,
+  journal = {International Journal of Advanced Manufacturing Technology},
+  author = {{\"O}nc{\"u} Hazir and Yavuz G{\"u}nalay and Erdal Erel},
+  month = may,
+  year = 2008,
+  keywords = {{ACO,Customer} order {scheduling,Genetic}
+                  {algorithms,Meta-heuristics,Simulated} {annealing,Tabu}
+                  search},
+  pages = {589--598},
+  volume = 37,
+  doi = {10.1007/s00170-007-0998-8},
+  abstract = {The customer order scheduling problem {(COSP)} is defined as
+                  to determine the sequence of tasks to satisfy the demand of
+                  customers who order several types of products produced on a
+                  single machine. A setup is required whenever a product type
+                  is launched.  The objective of the scheduling problem is to
+                  minimize the average customer order flow time. Since the
+                  customer order scheduling problem is known to be strongly
+                  {NP-hard,} we solve it using four major metaheuristics and
+                  compare the performance of these heuristics, namely,
+                  simulated annealing, genetic algorithms, tabu search, and ant
+                  colony optimization. These are selected to represent various
+                  characteristics of metaheuristics: nature-inspired
+                  vs. artificially created, population-based vs. local search,
+                  etc. A set of problems is generated to compare the solution
+                  quality and computational efforts of these heuristics.
+                  Results of the experimentation show that tabu search and ant
+                  colony perform better for large problems whereas simulated
+                  annealing performs best in small-size problems. Some
+                  conclusions are also drawn on the interactions between
+                  various problem parameters and the performance of the
+                  heuristics.}
+}
+
+ +
+@article{HeYen2016many,
+  title = {Many-Objective Evolutionary Algorithm: Objective Space
+                  Reduction and Diversity Improvement},
+  author = {He, Zhenan and Yen, Gary G.},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2016,
+  number = 1,
+  pages = {145--160},
+  volume = 20
+}
+
+ +
+@article{HeZhaChu2021automl,
+  author = {Xin He and Kaiyong Zhao and Xiaowen Chu},
+  title = {{AutoML}: A survey of the state-of-the-art},
+  journal = {Knowledge-Based Systems},
+  volume = 212,
+  pages = 106622,
+  year = 2021,
+  issn = {0950-7051},
+  doi = {10.1016/j.knosys.2020.106622}
+}
+
+ +
+@article{HelBraMos2013bound,
+  author = {Helwig, Sabine and  J{\"u}rgen Branke  and  Mostaghim, Sanaz },
+  title = {Experimental Analysis of Bound Handling Techniques in
+                  Particle Swarm Optimization},
+  doi = {10.1109/tevc.2012.2189404},
+  year = 2013,
+  month = apr,
+  volume = 17,
+  number = 2,
+  pages = {259--271},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  keywors = {PSO; box-constraints; constraint handling; bounds}
+}
+
+ +
+@article{HelKar1970,
+  author = {Held, Michael and Karp, Richard M.},
+  title = {The Traveling-Salesman Problem and Minimum Spanning Trees},
+  journal = {Operations Research},
+  volume = 18,
+  number = 6,
+  pages = {1138--1162},
+  year = 1970
+}
+
+ +
+@article{HelRen1998mp,
+  author = {Helmberg, Christoph and Rendl, Franz},
+  title = {Solving quadratic (0,1)-problems by semidefinite programs and
+                  cutting planes},
+  journal = {Mathematical Programming},
+  year = 1998,
+  volume = 82,
+  number = 3,
+  pages = {291--315}
+}
+
+ +
+@article{Helsgaun00,
+  author = { Keld Helsgaun },
+  title = {An Effective Implementation of the {Lin}-{Kernighan}
+Traveling Salesman Heuristic},
+  journal = {European Journal of Operational Research},
+  year = 2000,
+  volume = 126,
+  pages = {106--130}
+}
+
+ +
+@article{Helsgaun09,
+  author = { Keld Helsgaun },
+  title = {General {\it k}-opt Submoves for the {Lin}-{Kernighan} {TSP}
+               Heuristic},
+  journal = {Mathematical Programming Computation},
+  year = 2009,
+  volume = 1,
+  number = {2--3},
+  pages = {119--163}
+}
+
+ +
+@article{Her2015toms,
+  author = {Michael A. Heroux},
+  title = {Editorial: {ACM} {TOMS} Replicated Computational Results
+                  Initiative},
+  doi = {10.1145/2743015},
+  year = 2015,
+  month = jun,
+  publisher = {Association for Computing Machinery ({ACM})},
+  volume = 41,
+  number = 3,
+  pages = {1--5},
+  journal = {ACM Transactions on Mathematical Software}
+}
+
+ +
+@article{HerBlu2009:si,
+  author = {H. Hern{\'a}ndez and  Christian Blum },
+  title = {Ant colony optimization for multicasting in static
+                  wireless ad-hoc networks},
+  journal = {Swarm Intelligence},
+  volume = 3,
+  number = 2,
+  pages = {125--148},
+  year = 2009
+}
+
+ +
+@article{HerColDua2021efficient,
+  title = {An efficient Variable Neighborhood Search for the Space-Free
+                  Multi-Row Facility Layout problem},
+  author = {Herr{\'a}n, Alberto and  Colmenar, J. Manuel  and  Duarte, Abraham },
+  journal = {European Journal of Operational Research},
+  year = 2021,
+  publisher = {Elsevier},
+  doi = {10.1016/j.ejor.2021.03.027}
+}
+
+ +
+@article{HerLozSan03:real-coded-crossover,
+  author = { Francisco Herrera  and  Manuel Lozano  and  A. M. S{\'a}nchez },
+  title = {A taxonomy for the crossover operator for real-coded genetic
+                  algorithms: An experimental study},
+  journal = {International Journal of Intelligent Systems},
+  year = 2003,
+  volume = 18,
+  number = 3,
+  pages = {309--338},
+  doi = {10.1002/int.10091}
+}
+
+ +
+@article{HerLozVer1998air,
+  author = { Francisco Herrera  and  Manuel Lozano  and Verdegay, J. L.},
+  title = {Tackling Real-Coded Genetic Algorithms: Operators
+                  and Tools for Behavioural Analysis},
+  journal = {Artificial Intelligence Review},
+  year = 1998,
+  volume = 12,
+  pages = {265--319},
+  keywords = {genetic algorithms, real coding, continuous search
+                  spaces, mutation, recombination}
+}
+
+ +
+@article{HerSch2022hausdorff,
+  title = {A Bounded Archiver for {Hausdorff} Approximations of the
+                  {Pareto} Front for Multi-Objective Evolutionary Algorithms},
+  author = { Hern{\'a}ndez Castellanos, Carlos Ignacio  and  Oliver Sch{\"u}tze },
+  journal = {Mathematical and Computational Applications},
+  volume = 27,
+  number = 3,
+  pages = 48,
+  year = 2022,
+  doi = {10.3390/mca27030048},
+  publisher = {Multidisciplinary Digital Publishing Institute}
+}
+
+ +
+@article{HerSchSun2020noneps,
+  author = { Hern{\'a}ndez Castellanos, Carlos Ignacio  and  Oliver Sch{\"u}tze  and Sun, J. Q. and  Ober-Bl\"obaum, S.},
+  title = {Non-epsilon dominated evolutionary algorithm for the set of
+                  approximate solutions},
+  journal = {Mathematical and Computational Applications},
+  year = 2020,
+  volume = 25,
+  number = 1,
+  pages = 3,
+  keywords = {archiving, multimodal}
+}
+
+ +
+@article{Hermet2006ecj,
+  author = { van Hemert, Jano I. },
+  title = {Evolving Combinatorial Problem Instances That Are Difficult
+                  to Solve},
+  journal = {Evolutionary Computation},
+  volume = 14,
+  number = 4,
+  pages = {433--462},
+  year = 2006,
+  doi = {10.1162/evco.2006.14.4.433},
+  abstract = {This paper demonstrates how evolutionary computation can be
+                  used to acquire difficult to solve combinatorial problem
+                  instances. As a result of this technique, the corresponding
+                  algorithms used to solve these instances are
+                  stress-tested. The technique is applied in three important
+                  domains of combinatorial optimisation, binary constraint
+                  satisfaction, Boolean satisfiability, and the travelling
+                  salesman problem. The problem instances acquired through this
+                  technique are more difficult than the ones found in popular
+                  benchmarks. In this paper, these evolved instances are
+                  analysed with the aim to explain their difficulty in terms of
+                  structural properties, thereby exposing the weaknesses of
+                  corresponding algorithms. }
+}
+
+ +
+@article{HeuNieKru2020publish,
+  author = {Robert Heum{\"u}ller and Sebastian Nielebock and Jacob
+                  Kr{\"u}ger and Frank Ortmeier},
+  title = {Publish or perish, but do not forget your software artifacts},
+  doi = {10.1007/s10664-020-09851-6},
+  year = 2020,
+  volume = 25,
+  number = 6,
+  pages = {4585--4616},
+  journal = {Empirical Software Engineering}
+}
+
+ +
+@article{Hic2006,
+  author = {Christian Hicks},
+  title = {A Genetic Algorithm tool for optimising cellular or
+                  functional layouts in the capital goods industry},
+  journal = {International Journal of Production Economics},
+  volume = 104,
+  number = 2,
+  pages = {598--614},
+  year = 2006,
+  doi = {10.1016/j.ijpe.2005.03.010}
+}
+
+ +
+@article{HieLiLiuPar2020,
+  title = {Many-objective test suite generation for software product
+                  lines},
+  author = {Hierons, Robert M. and  Li, Miqing  and Liu, Xiaohui and
+                  Parejo, Jose Antonio and Segura, Sergio and  Xin Yao },
+  journal = {{ACM} Transactions on Software Engineering and Methodology},
+  year = 2020,
+  number = 1,
+  pages = {1--46},
+  volume = 29
+}
+
+ +
+@article{HinSal2006autoencoder,
+  title = {Reducing the dimensionality of data with neural networks},
+  author = {Hinton, Geoffrey E. and Salakhutdinov, Ruslan R.},
+  journal = {Science},
+  volume = 313,
+  number = 5786,
+  pages = {504--507},
+  year = 2006
+}
+
+ +
+@article{Hoeffding1963,
+  title = {Probability inequalities for sums of bounded random
+                  variables},
+  author = {Hoeffding, Wassily},
+  journal = {Journal of the American Statistical Association},
+  volume = 58,
+  number = 301,
+  pages = {13--30},
+  year = 1963,
+  alias = {hoeffding1963probability}
+}
+
+ +
+@article{HonKahMoo1997,
+  author = {I. Hong and A. B. Kahng and B. R. Moon},
+  title = {Improved large-step {Markov} chain variants for the symmetric {TSP}},
+  journal = {Journal of Heuristics},
+  year = 1997,
+  volume = 3,
+  number = 1,
+  pages = {63--81}
+}
+
+ +
+@article{Hoo1994or,
+  author = { John N. Hooker },
+  title = {Needed: An Empirical Science of Algorithms},
+  journal = {Operations Research},
+  year = 1994,
+  volume = 42,
+  number = 2,
+  pages = {201--212}
+}
+
+ +
+@article{Hoo1996joh,
+  author = { John N. Hooker },
+  title = {Testing Heuristics: We Have It All Wrong},
+  journal = {Journal of Heuristics},
+  year = 1996,
+  volume = 1,
+  number = 1,
+  pages = {33--42},
+  doi = {10.1007/BF02430364}
+}
+
+ +
+@article{Hoo2012,
+  title = {Generalized functional {ANOVA} diagnostics for
+                  high-dimensional functions of dependent variables},
+  author = {Hooker, Giles},
+  journal = {Journal of Computational and Graphical Statistics},
+  year = 2012,
+  volume = 16,
+  number = 3,
+  pages = {709--732},
+  doi = {10.1198/106186007X237892}
+}
+
+ +
+@article{HooLinSch2014claspfolio2,
+  title = {Claspfolio 2: Advances in Algorithm Selection for Answer Set
+                  Programming},
+  author = { Holger H. Hoos  and  Marius Thomas Lindauer  and Schaub, Torsten},
+  journal = {Theory and Practice of Logic Programming},
+  volume = 14,
+  number = {4-5},
+  pages = {560--585},
+  year = 2014,
+  publisher = {Cambridge University Press}
+}
+
+ +
+@article{HooStu2014,
+  author = { Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {On the Empirical Scaling of Run-time for Finding Optimal
+                  Solutions to the Traveling Salesman Problem},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 238,
+  number = 1,
+  pages = {87--94}
+}
+
+ +
+@article{HooStu2015:ol,
+  author = { Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {On the Empirical Time Complexity of Finding Optimal Solutions
+                  vs.\ Proving Optimality for {Euclidean} {TSP} Instances},
+  journal = {Optimization Letters},
+  year = 2015,
+  volume = 9,
+  number = 6,
+  pages = {1247--1254}
+}
+
+ +
+@article{Hoos:PbO,
+  author = { Holger H. Hoos },
+  title = {Programming by optimization},
+  journal = {Communications of the ACM},
+  volume = 55,
+  number = 2,
+  month = feb,
+  year = 2012,
+  pages = {70--80},
+  numpages = 11,
+  doi = {10.1145/2076450.2076469}
+}
+
+ +
+@article{HotTanTie2020deep,
+  author = {Hottung, Andr{\'e} and Tanaka, Shunji and  Kevin Tierney },
+  title = {Deep learning assisted heuristic tree search for the
+                  container pre-marshalling problem},
+  journal = {Computers \& Operations Research},
+  year = 2020,
+  volume = 113,
+  pages = 104781,
+  doi = {10.1016/j.cor.2019.104781}
+}
+
+ +
+@article{HotTie2022nlns,
+  author = {Hottung, Andr{\'e} and  Kevin Tierney },
+  title = {Neural large neighborhood search for routing problems},
+  journal = {Artificial Intelligence},
+  year = 2022,
+  volume = 313,
+  pages = 103786,
+  month = dec,
+  doi = {10.1016/j.artint.2022.103786}
+}
+
+ +
+@article{HozDjuHoz2005,
+  title = {Estimating the mean and variance from the median, range, and
+                  the size of a sample},
+  author = {Hozo, Stela Pudar and Djulbegovic, Benjamin and Hozo, Iztok},
+  journal = {BMC Medical Research Methodology},
+  volume = 5,
+  number = 1,
+  pages = 13,
+  year = 2005
+}
+
+ +
+@article{HuKahTsa95,
+  author = {T. C. Hu and A. B. Kahng and C.-W. A. Tsao},
+  title = {Old Bachelor Acceptance: A New Class of Non-Monotone
+                  Threshold Accepting Methods},
+  journal = {ORSA Journal on Computing},
+  year = 1995,
+  volume = 7,
+  number = 4,
+  pages = {417--425}
+}
+
+ +
+@article{HuWanQiu2018nca,
+  author = {Hu, Wenbin and Wang, Huan and Qiu, Zhenyu and Nie, Cong and
+                  Yan, Liping},
+  doi = {10.1007/s00521-016-2508-0},
+  journal = {Neural Computing \& Applications},
+  keywords = {BML,Optimization,Simulation,Traffic congestion,Updating
+                  rules},
+  title = {A quantum particle swarm optimization driven urban traffic
+                  light scheduling model},
+  year = 2018,
+  alias = {Hu2018}
+}
+
+ +
+@article{HuYanWang2017,
+  author = {Hu, Wenbin and Yan, Liping and Wang, Huan and Du, Bo and Tao,
+                  Dacheng},
+  journal = {Information Sciences},
+  keywords = {BML model,Prediction,Real-time,Traffic jam,Urban traffic
+                  network},
+  title = {Real-time traffic jams prediction inspired by {Biham},
+                  {Middleton} and {Levine} (BML) model},
+  year = 2017,
+  alias = {Hu2017}
+}
+
+ +
+@article{HuaAllNot2006global,
+  title = {Global Optimization of Stochastic Black-Box Systems via Sequential Kriging Meta-Models},
+  author = {Huang, Deng and Allen, Theodore T. and Notz, William I. and
+                  Zeng, Ning},
+  journal = {Journal of Global Optimization},
+  volume = 34,
+  number = 3,
+  pages = {441--466},
+  year = 2006,
+  publisher = {Springer},
+  doi = {10.1007/s10898-005-2454-3}
+}
+
+ +
+@article{HuaLiYao2020algconf,
+  author = {Changwu Huang and Yuanxiang Li and  Xin Yao },
+  title = {A Survey of Automatic Parameter Tuning Methods for
+                  Metaheuristics},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 24,
+  number = 2,
+  pages = {201--216},
+  year = 2020,
+  doi = {10.1109/TEVC.2019.2921598}
+}
+
+ +
+@article{HubHinBar2006tec,
+  title = {A Review of Multiobjective Test Problems and a Scalable Test
+                  Problem Toolkit},
+  volume = 10,
+  doi = {10.1109/TEVC.2005.861417},
+  number = 5,
+  journal = {IEEE Transactions on Evolutionary Computation},
+  author = {Huband, S. and Hingston, P. and Barone, L. and While, L.},
+  year = 2006,
+  pages = {477--506},
+  annote = {Proposed the WFG benchmark suite}
+}
+
+ +
+@article{HubLukHog1997,
+  author = {B. Huberman and R. Lukose and T. Hogg},
+  title = {An Economic Approach to Hard Computational Problems},
+  journal = {Science},
+  year = 1997,
+  volume = 275,
+  pages = {51--54}
+}
+
+ +
+@article{HueRioRui2017,
+  author = {D. L. Huerta-Mu{\~n}oz and R. Z. R{\'i}os-Mercado and  Rub{\'e}n Ruiz },
+  title = {An Iterated Greedy Heuristic for a Market Segmentation Problem with Multiple Attributes},
+  journal = {European Journal of Operational Research},
+  year = 2017,
+  volume = 261,
+  number = 1,
+  pages = {75--87}
+}
+
+ +
+@article{HumLieTal2013joh,
+  author = { J{\'e}r{\'e}mie Humeau  and  Arnaud Liefooghe  and  Talbi, El-Ghazali  and  Verel, S{\'e}bastien },
+  title = {{ParadisEO-MO}: From Fitness Landscape Analysis to
+                  Efficient Local Search Algorithms},
+  volume = 19,
+  doi = {10.1007/s10732-013-9228-8},
+  number = 6,
+  journal = {Journal of Heuristics},
+  month = jun,
+  year = 2013,
+  pages = {881--915}
+}
+
+ +
+@article{HunJosMel2009dace,
+  title = {Design and Analysis of Computer Experiments With Branching
+                  and Nested Factors},
+  author = {Hung, Ying and Joseph, V. Roshan and Melkote, Shreyes N.},
+  journal = {Technometrics},
+  year = 2009,
+  number = 4,
+  pages = {354--365},
+  volume = 51,
+  alias = {Hung2009},
+  doi = {10.1198/TECH.2009.07097}
+}
+
+ +
+@article{HurMau2010,
+  author = {M. Hurtgen and J.-C. Maun},
+  title = {Optimal {PMU} placement using Iterated Local Search},
+  journal = {International Journal of Electrical Power \& Energy Systems},
+  year = 2010,
+  volume = 32,
+  number = 8,
+  pages = {857--860}
+}
+
+ +
+@article{Hurlbert1984,
+  author = {S. H. Hurlbert},
+  title = {Pseudoreplication and the Design of Ecological Field
+                  Experiments},
+  journal = {Ecological Monographs},
+  volume = 54,
+  number = 2,
+  pages = {187--211},
+  year = 1984
+}
+
+ +
+@article{HusStu2014cor,
+  author = {Mohamed Saifullah Hussin and  Thomas St{\"u}tzle },
+  title = {Tabu Search vs. Simulated Annealing for Solving
+                  Large Quadratic Assignment Instances},
+  journal = {Computers \& Operations Research},
+  volume = 43,
+  pages = {286--291},
+  year = 2014
+}
+
+ +
+@article{HutHooLey2010amai,
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Tradeoffs in the Empirical Evaluation of Competing Algorithm Designs},
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  year = 2010,
+  volume = 60,
+  number = {1--2},
+  pages = {65--89}
+}
+
+ +
+@article{HutHooLey2013arxiv,
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Bayesian Optimization With Censored Response Data},
+  journal = {Arxiv preprint arXiv:1310.1947},
+  year = 2013,
+  url = {http://arxiv.org/abs/1310.1947}
+}
+
+ +
+@article{HutHooLeyStu2009jair,
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown  and  Thomas St{\"u}tzle },
+  title = {{\softwarepackage{ParamILS}:} An Automatic Algorithm Configuration Framework},
+  journal = {Journal of Artificial Intelligence Research},
+  year = 2009,
+  volume = 36,
+  pages = {267--306},
+  month = oct,
+  doi = {10.1613/jair.2861}
+}
+
+ +
+@article{HutLinBal2017aij,
+  author = { Frank Hutter  and  Marius Thomas Lindauer  and Adrian Balint and Sam Bayless and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {The Configurable {SAT} Solver Challenge {(CSSC)}},
+  journal = {Artificial Intelligence},
+  year = 2017,
+  volume = 243,
+  pages = {1--25},
+  alias = {HutLinBal++2017:aij},
+  doi = {10.1016/j.artint.2016.09.006}
+}
+
+ +
+@article{HutXuHooLey2014,
+  author = { Frank Hutter  and  Lin Xu  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Algorithm Runtime Prediction: Methods \& evaluation},
+  journal = {Artificial Intelligence},
+  year = 2014,
+  volume = 206,
+  pages = {79--111},
+  doi = {10.1016/j.artint.2013.10.003},
+  abstract = {Perhaps surprisingly, it is possible to predict how long an
+                  algorithm will take to run on a previously unseen input,
+                  using machine learning techniques to build a model of the
+                  algorithm's runtime as a function of problem-specific
+                  instance features. Such models have important applications to
+                  algorithm analysis, portfolio-based algorithm selection, and
+                  the automatic configuration of parameterized algorithms. Over
+                  the past decade, a wide variety of techniques have been
+                  studied for building such models. Here, we describe
+                  extensions and improvements of existing models, new families
+                  of models, and---perhaps most importantly---a much more thorough
+                  treatment of algorithm parameters as model inputs. We also
+                  comprehensively describe new and existing features for
+                  predicting algorithm runtime for propositional satisfiability
+                  (SAT), travelling salesperson (TSP) and mixed integer
+                  programming (MIP) problems. We evaluate these innovations
+                  through the largest empirical analysis of its kind, comparing
+                  to a wide range of runtime modelling techniques from the
+                  literature. Our experiments consider 11 algorithms and 35
+                  instance distributions; they also span a very wide range of
+                  SAT, MIP, and TSP instances, with the least structured having
+                  been generated uniformly at random and the most structured
+                  having emerged from real industrial applications. Overall, we
+                  demonstrate that our new models yield substantially better
+                  runtime predictions than previous approaches in terms of
+                  their generalization to new problem instances, to new
+                  algorithms from a parameterized space, and to both
+                  simultaneously.},
+  keywords = {Empirical performance models; Mixed integer programming; SAT}
+}
+
+ +
+@article{IOHanalyzer,
+  author = {Hao Wang and  Diederick Vermetten  and Furong Ye and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {{IOHanalyzer}: Detailed Performance Analyses for Iterative
+                  Optimization Heuristics},
+  journal = {ACM Transactions on Evolutionary Learning and Optimization},
+  year = 2022,
+  volume = 2,
+  number = 1,
+  pages = {3:1--3:29},
+  doi = {10.1145/3510426}
+}
+
+ +
+@article{IOHexperimenter2021,
+  author = {Jacob de Nobel and Furong Ye and  Diederick Vermetten  and Wang,
+                  Hao and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {{IOHexperimenter}: Benchmarking Platform for Iterative
+                  Optimization Heuristics},
+  journal = {Arxiv preprint arXiv:2111.04077},
+  year = 2021,
+  annote = {Published in ECJ~\cite{IOHexperimenter2024}},
+  doi = {10.48550/arXiv.2111.04077}
+}
+
+ +
+@article{IOHexperimenter2024,
+  author = {Jacob de Nobel and Furong Ye and  Diederick Vermetten  and Wang,
+                  Hao and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {{IOHexperimenter}: Benchmarking Platform for Iterative
+                  Optimization Heuristics},
+  journal = {Evolutionary Computation},
+  year = 2024,
+  pages = {1--6},
+  doi = {10.1162/evco_a_00342},
+  eprint = {https://direct.mit.edu/evco/article-pdf/doi/10.1162/evco\_a\_00342/2335957/evco\_a\_00342.pdf}
+}
+
+ +
+@article{IOHprofiler,
+  author = { Carola Doerr  and Hao Wang and Furong Ye and van Rijn,
+                  Sander and  Thomas B{\"a}ck },
+  title = {{IOHprofiler}: A Benchmarking and Profiling Tool for
+                  Iterative Optimization Heuristics},
+  journal = {Arxiv preprint arXiv:1806.07555},
+  year = 2018,
+  month = oct,
+  keywords = {Benchmarking; Heuristics},
+  doi = {10.48550/arXiv.1810.05281}
+}
+
+ +
+@article{IacPal2012:si,
+  author = {Claudio Iacopino and Phil Palmer},
+  title = {The Dynamics of Ant Colony Optimization Algorithms Applied to Binary Chains},
+  journal = {Swarm Intelligence},
+  year = 2012,
+  volume = 6,
+  number = 4,
+  pages = {343--377}
+}
+
+ +
+@article{IacPalPol2014ants,
+  journal = {Acta Futura},
+  title = {How Ants Can Manage Your Satellites},
+  doi = {10.2420/AF09.2014.59},
+  volume = 9,
+  author = {Iacopino, Claudio and Palmer, Phil and Policella, N. and
+                  Donati, A. and Brewer, A.},
+  pages = {59--72},
+  year = 2014,
+  keywords = {ACO, Space}
+}
+
+ +
+@article{IbaImaNon++2008:dam,
+  author = {Toshihide Ibaraki and Shinji Imahori and Koji Nonobe and Kensuke Sobue and Takeaki Uno and  Mutsunori Yagiura },
+  title = {An Iterated Local Search Algorithm for the Vehicle Routing Problem
+               with Convex Time Penalty Functions},
+  journal = {Discrete Applied Mathematics},
+  year = 2008,
+  volume = 156,
+  number = 11,
+  pages = {2050--2069}
+}
+
+ +
+@article{Ibaraki2010:itor,
+  author = {Toshihide Ibaraki},
+  title = {A Personal Perspective on Problem Solving by General Purpose Solvers},
+  journal = {International Transactions in Operational Research},
+  year = 2010,
+  volume = 17,
+  number = 3,
+  pages = {303--315}
+}
+
+ +
+@article{IdeSch2016,
+  author = {Ide, Jonas and  Sch{\"o}bel, Anita },
+  title = {Robustness for uncertain multi-objective optimization: a
+                  survey and analysis of different concepts},
+  journal = {OR Spectrum},
+  year = 2016,
+  volume = 38,
+  number = 1,
+  pages = {235--271},
+  abstract = {In this paper, we discuss various concepts of robustness for
+                  uncertain multi-objective optimization problems. We extend
+                  the concepts of flimsily, highly, and lightly robust
+                  efficiency and we collect different versions of minmax robust
+                  efficiency and concepts based on set order relations from the
+                  literature. Altogether, we compare and analyze ten different
+                  concepts and point out their relations to each
+                  other. Furthermore, we present reduction results for the
+                  class of objective-wise uncertain multi-objective
+                  optimization problems.},
+  doi = {10.1007/s00291-015-0418-7}
+}
+
+ +
+@article{IgeHanRot2007ec,
+  author = { Christian Igel  and  Nikolaus Hansen  and S. Roth},
+  title = {Covariance Matrix Adaptation for Multi-objective
+                  Optimization},
+  journal = {Evolutionary Computation},
+  year = 2007,
+  volume = 15,
+  pages = {1--28},
+  number = 1
+}
+
+ +
+@article{IgeHeiGla2008jmlr,
+  author = { Christian Igel  and V. Heidrich-Meisner and  T. Glasmachers },
+  title = {Shark},
+  journal = {Journal of Machine Learning Research},
+  year = 2008,
+  volume = 9,
+  month = jun,
+  pages = {993--996},
+  url = {http://www.jmlr.org/papers/volume9/igel08a/igel08a.pdf},
+  alias = {Igel:2008ut}
+}
+
+ +
+@article{Ilich98,
+  author = { Nesa Ilich  and  Slobodan P. Simonovic },
+  title = {Evolutionary Algorithm for minimization of pumping
+                  cost},
+  journal = {Journal of Computing in Civil Engineering, {ASCE}},
+  year = 1998,
+  volume = 12,
+  number = 4,
+  pages = {232--240},
+  month = oct,
+  abstract = {This paper deals with minimizing the total cost of
+                  pumping in a liquid pipeline. Previous experience
+                  with the most common solution procedures in pipeline
+                  optimization is discussed along with their strengths
+                  and weaknesses. The proposed method is an
+                  evolutionary algorithm with two distinct features:
+                  (1) The search is restricted to feasible region
+                  only; and (2) it utilizes a floating point decision
+                  variable rather than integer or binary as is the
+                  case with most other similar approaches. A numerical
+                  example is presented as a basis for verification of
+                  the proposed method and its comparison with the
+                  existing solver that utilizes the nonlinear
+                  Newtonian search. The proposed method provides
+                  promising improvements in terms of optimality when
+                  compared to the widespread gradient search methods
+                  because it does not involve evaluation of the
+                  gradient of the objective function. It also provides
+                  potential to improve the performance of previous
+                  evolutionary programs because it restricts the
+                  search to the feasible region, thus eliminating
+                  large overhead associated with generation and
+                  inspection of solutions that are
+                  infeasible. Comparison of the two solutions revealed
+                  improvement of the solution in favor of the proposed
+                  algorithm, which ranged up to 6\% depending on the
+                  initial values of the decision variables in the
+                  Newtonian search. The proposed method was not
+                  sensitive to the starting value of the decision
+                  variables.}
+}
+
+ +
+@article{ImaYagNag2009:ejor,
+  author = {Takashi Imamichi and  Mutsunori Yagiura  and Hiroshi Nagamochi},
+  title = {An Iterated Local Search Algorithm Based on Nonlinear
+                  Programming for the Irregular Strip Packing Problem},
+  journal = {Discrete Optimization},
+  year = 2009,
+  volume = 6,
+  number = 4,
+  pages = {345--361}
+}
+
+ +
+@article{Inselberg1985pc,
+  title = {The Plane with Parallel Coordinates},
+  author = {Inselberg, Alfred},
+  journal = {The {Visual} {Computer}},
+  volume = 1,
+  number = 2,
+  pages = {69--91},
+  year = 1985
+}
+
+ +
+@article{Ioa2005why,
+  title = {Why Most Published Research Findings Are False},
+  author = { John P. A. Ioannidis },
+  journal = {PLoS Medicine},
+  volume = 2,
+  number = 8,
+  doi = {10.1371/journal.pmed.0020124},
+  pages = {e124},
+  year = 2005
+}
+
+ +
+@article{Irnich2008,
+  author = {Stefan Irnich},
+  title = {A Unified Modeling and Solution Framework for Vehicle Routing and
+               Local Search-Based Metaheuristics},
+  journal = {INFORMS Journal on Computing},
+  year = 2008,
+  volume = 20,
+  number = 2,
+  pages = {270--287}
+}
+
+ +
+@article{IruCalLoz2016mallows,
+  author = { Irurozki, Ekhine  and Borja Calvo and  Jos{\'e} A. Lozano },
+  title = {Sampling and Learning {Mallows} and Generalized {Mallows}
+                  Models Under the {Cayley} Distance},
+  doi = {10.1007/s11009-016-9506-7},
+  year = 2016,
+  month = jun,
+  volume = 20,
+  number = 1,
+  pages = {1--35},
+  journal = {Methodology and Computing in Applied Probability}
+}
+
+ +
+@article{IruCalLoz2016permallows,
+  title = {{\rpackage{PerMallows}}: An {\proglang{R}} Package for Mallows
+                  and Generalized Mallows Models},
+  author = { Irurozki, Ekhine  and Calvo, Borja and  Jos{\'e} A. Lozano },
+  abstract = {In this paper we present the R package PerMallows, which is a
+                  complete toolbox to work with permutations, distances and
+                  some of the most popular probability models for permutations:
+                  Mallows and the Generalized Mallows models. The Mallows model
+                  is an exponential location model, considered as analogous to
+                  the Gaussian distribution. It is based on the definition of a
+                  distance between permutations. The Generalized Mallows model
+                  is its best-known extension. The package includes functions
+                  for making inference, sampling and learning such
+                  distributions. The distances considered in PerMallows are
+                  Kendall's $\tau$, Cayley, Hamming and Ulam.},
+  doi = {10.18637/jss.v071.i12},
+  issn = 15487660,
+  journal = {Journal of Statistical Software},
+  keywords = {Cayley,Generalized Mallows,Hamming,Kendall's
+                  $\tau$,Learning,Mallows,Permutation,R,Ranking,Sampling,Ulam},
+  volume = 71,
+  year = 2019
+}
+
+ +
+@article{IruLobPer2020arxiv,
+  title = {Rank aggregation for non-stationary data streams},
+  author = { Irurozki, Ekhine  and Lobo, Jesus and Perez, Aritz and Del Ser,
+                  Javier},
+  keywords = {uborda},
+  journal = {Arxiv preprint arXiv:1910.08795 [stat.ML]},
+  year = 2020,
+  url = {https://arxiv.org/abs/1910.08795}
+}
+
+ +
+@article{Ish1998,
+  author = { Ishibuchi, Hisao  and  Murata, T. },
+  title = {A multi-objective genetic local search algorithm and its
+                  application to flowshop scheduling},
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics -- Part C},
+  number = 3,
+  pages = {392--403},
+  volume = 28,
+  year = 1998
+}
+
+ +
+@article{IshAkeNoj2015tec,
+  author = { Ishibuchi, Hisao  and N. Akedo and Y. Nojima},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Behavior of Multiobjective Evolutionary Algorithms on
+                  Many-Objective Knapsack Problems},
+  year = 2015,
+  volume = 19,
+  number = 2,
+  pages = {264--283},
+  doi = {10.1109/TEVC.2014.2315442}
+}
+
+ +
+@article{IshImaSet2018refpoint,
+  title = {How to specify a reference point in hypervolume calculation
+                  for fair performance comparison},
+  author = { Ishibuchi, Hisao  and Imada, Ryo and Setoguchi, Yu and
+                  Nojima, Yusuke},
+  journal = {Evolutionary Computation},
+  year = 2018,
+  number = 3,
+  pages = {411--440},
+  volume = 26
+}
+
+ +
+@article{IshMisTan1995,
+  title = {Modified simulated annealing algorithms for the flow shop
+                  sequencing problem},
+  author = { Ishibuchi, Hisao  and Misaki, Shinta and Tanaka, Hideo},
+  journal = {European Journal of Operational Research},
+  volume = 81,
+  number = 2,
+  pages = {388--398},
+  year = 1995
+}
+
+ +
+@article{IshSetMas2017shape,
+  title = {Performance of decomposition-based many-objective algorithms
+                  strongly depends on {Pareto} front shapes},
+  author = { Ishibuchi, Hisao  and Setoguchi, Yu and Masuda, Hiroyuki and
+                  Nojima, Yusuke},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2017,
+  number = 2,
+  pages = {169--190},
+  volume = 21,
+  publisher = {IEEE}
+}
+
+ +
+@article{IviTha2019repro,
+  author = {Peter Ivie and Douglas Thain},
+  title = {Reproducibility in Scientific Computing},
+  doi = {10.1145/3186266},
+  year = 2019,
+  publisher = {Association for Computing Machinery ({ACM})},
+  volume = 51,
+  number = 3,
+  pages = {1--36},
+  journal = {{ACM} Computing Surveys}
+}
+
+ +
+@article{IyeSax2004,
+  author = {Srikanth K. Iyer and Barkha Saxena},
+  title = {Improved genetic algorithm for the permutation flowshop
+                  scheduling problem},
+  journal = {Computers \& Operations Research},
+  volume = 31,
+  number = 4,
+  pages = {593--606},
+  year = 2004,
+  doi = {10.1016/S0305-0548(03)00016-9}
+}
+
+ +
+@article{Jac2011jss,
+  title = {Multi-State Models for Panel Data: The {\rpackage{msm}} Package for \proglang{R}},
+  author = {Christopher H. Jackson},
+  journal = {Journal of Statistical Software},
+  year = 2011,
+  volume = 38,
+  number = 8,
+  pages = {1--29},
+  url = {http://www.jstatsoft.org/v38/i08/}
+}
+
+ +
+@article{JacBogNas1991,
+  author = {Richard H. F. Jackson and Paul T. Boggs and Stephen G. Nash and Susan Powell},
+  title = {Guidelines for Reporting Results of Computational Experiments. Report of the Ad
+		  Hoc Committee},
+  journal = {Mathematical Programming},
+  year = 1991,
+  volume = 49,
+  number = 3,
+  pages = {413--425}
+}
+
+ +
+@article{JacBru1995:ig,
+  author = {Jacobs, Larry W. and Brusco, Michael J.},
+  title = {A Local Search Heuristic for Large Set-Covering
+                  Problems},
+  journal = {Naval Research Logistics},
+  year = 1995,
+  volume = 42,
+  number = 7,
+  pages = {1129--1140}
+}
+
+ +
+@article{JacKah1995anchoring,
+  author = {Jacowitz, Karen E. and  Kahneman, Daniel },
+  title = {Measures of {Anchoring} in {Estimation} {Tasks}},
+  journal = {Personality and Social Psychology Bulletin},
+  year = 1995,
+  volume = 21,
+  number = 11,
+  pages = {1161--1166},
+  month = nov,
+  issn = {0146-1672},
+  doi = {10.1177/01461672952111004},
+  abstract = {The authors describe a method for the quantitative study of
+                  anchoring effects in estimation tasks. A calibration group
+                  provides estimates of a set of uncertain quantities. Subjects
+                  in the anchored condition first judge whether a specified
+                  number (the anchor) is higher or lower than the true value
+                  before estimating each quantity. The anchors are set at
+                  predetermined percentiles of the distribution of estimates in
+                  the calibration group (15th and 85th percentiles in this
+                  study). This procedure permits the transformation of anchored
+                  estimates into percentiles in the calibration group, allows
+                  pooling of results across problems, and provides a natural
+                  measure of the size of the effect. The authors illustrate the
+                  method by a demonstration that the initial judgment of the
+                  anchor is susceptible to an anchoring-like bias and by an
+                  analysis of the relation between anchoring and subjective
+                  confidence.},
+  language = {en}
+}
+
+ +
+@article{JacOzcJoh2018,
+  author = {Warren G. Jackson and  Ender {\"O}zcan  and Robert I. John},
+  title = {Move acceptance in local search metaheuristics for cross-domain search},
+  journal = {Expert Systems with Applications},
+  volume = 109,
+  pages = {131--151},
+  year = 2018
+}
+
+ +
+@article{JaeParKip2008ejor,
+  title = {The development of a multi-objective Tabu Search algorithm for continuous optimisation problems},
+  author = {Daniel M Jaeggi and Geoffrey T Parks and  Timoleon Kipouros and P John Clarkson},
+  journal = {European Journal of Operational Research},
+  volume = 185,
+  number = 3,
+  pages = {1192--1212},
+  year = 2008
+}
+
+ +
+@article{JajMinHarPro1992,
+  title = {{CLASS}: computerized layout solutions using simulated annealing},
+  author = {Jajodia, Satish and Minis, Ioannis and Harhalakis, George and Proth, Jean-Marie},
+  journal = {International Journal of Production Research},
+  volume = 30,
+  number = 1,
+  pages = {95--108},
+  year = 1992,
+  publisher = {Taylor \& Francis}
+}
+
+ +
+@article{Jas2002,
+  title = {Genetic local search for multi-objective combinatorial optimization},
+  author = { Andrzej Jaszkiewicz },
+  journal = {European Journal of Operational Research},
+  volume = 137,
+  number = 1,
+  pages = {50--71},
+  year = 2002
+}
+
+ +
+@article{Jas2018ejor,
+  title = {Many-Objective {Pareto} Local Search},
+  author = { Andrzej Jaszkiewicz },
+  journal = {European Journal of Operational Research},
+  year = 2018,
+  volume = 271,
+  number = 3,
+  pages = {1001--1013},
+  doi = {10.1016/j.ejor.2018.06.009}
+}
+
+ +
+@article{JasLus2018ndtree,
+  title = {{ND}-tree-based update: a fast algorithm for the dynamic
+                  nondominance problem},
+  author = { Andrzej Jaszkiewicz  and  Thibaut Lust },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  number = 5,
+  pages = {778--791},
+  volume = 22,
+  publisher = {IEEE}
+}
+
+ +
+@article{Jasz02mogls,
+  author = { Andrzej Jaszkiewicz },
+  title = {On the performance of multiple-objective genetic
+                  local search on the 0/1 knapsack problem -- A
+                  comparative experiment},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002,
+  volume = 6,
+  number = 4,
+  pages = {402--412}
+}
+
+ +
+@article{Jen03,
+  title = {Reducing the run-time complexity of multiobjective
+                  {EA}s: The {NSGA-II} and other algorithms},
+  author = {M. T. Jensen},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 7,
+  number = 5,
+  pages = {503--515},
+  year = 2003
+}
+
+ +
+@article{Jensen2004,
+  author = {M. T. Jensen},
+  title = {Helper-Objectives: {Using} Multi-Objective Evolutionary
+                  Algorithms for Single-Objective Optimisation},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2004,
+  volume = 3,
+  number = 4,
+  pages = {323--347},
+  keywords = {multi-objectivization}
+}
+
+ +
+@article{JerSor1998dam,
+  title = {The {Metropolis} algorithm for graph bisection},
+  author = { Mark Jerrum  and Sorkin, Gregory},
+  journal = {Discrete Applied Mathematics},
+  volume = 82,
+  number = 1,
+  pages = {155--175},
+  year = 1998,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Jerrum1992,
+  title = {Large cliques elude the {Metropolis} process},
+  author = { Mark Jerrum },
+  journal = {Random Structures \& Algorithms},
+  volume = 3,
+  number = 4,
+  pages = {347--359},
+  year = 1992,
+  publisher = {John Wiley \& Sons}
+}
+
+ +
+@article{JiaOngZhaFen2014metrics,
+  author = {S. Jiang and Y. S. Ong and J. Zhang and L. Feng},
+  journal = {IEEE Transactions on Cybernetics},
+  title = {Consistencies and Contradictions of Performance Metrics in
+                  Multiobjective Optimization},
+  year = 2014,
+  volume = 44,
+  number = 12,
+  pages = {2391--2404}
+}
+
+ +
+@article{JiaZouYanYao2022dynamic,
+  author = {Jiang, Shouyong and Zou, Juan and Yang, Shengxiang and  Xin Yao },
+  title = {Evolutionary Dynamic Multi-Objective Optimisation: A Survey},
+  year = 2022,
+  volume = 55,
+  number = 4,
+  doi = {10.1145/3524495},
+  journal = {{ACM} Computing Surveys},
+  month = nov,
+  articleno = 76,
+  numpages = 47,
+  keywords = {evolutionary algorithm, evolutionary dynamic multi-objective
+                  optimisation, dynamic environment, Multi-objective
+                  optimisation}
+}
+
+ +
+@article{Jin2005fitness,
+  author = { Yaochu Jin },
+  title = {A Comprehensive Survey of Fitness Approximation in
+                  Evolutionary Computation},
+  journal = {Soft Computing},
+  year = 2005,
+  volume = 9,
+  number = 1,
+  pages = {3--12}
+}
+
+ +
+@article{Jin2011surrogate,
+  author = { Yaochu Jin },
+  title = {Surrogate-Assisted Evolutionary Computation: Recent
+                  Advances and Future Challenges},
+  shorttitle = {Surrogate-Assisted Evolutionary Computation},
+  year = 2011,
+  month = jun,
+  journal = {Swarm and Evolutionary Computation},
+  volume = 1,
+  number = 2,
+  pages = {61--70},
+  issn = {2210-6502},
+  doi = {10.1016/j.swevo.2011.05.001},
+  abstract = {Surrogate-assisted, or meta-model based evolutionary
+                  computation uses efficient computational models, often known
+                  as surrogates or meta-models, for approximating the fitness
+                  function in evolutionary algorithms. Research on
+                  surrogate-assisted evolutionary computation began over a
+                  decade ago and has received considerably increasing interest
+                  in recent years. Very interestingly, surrogate-assisted
+                  evolutionary computation has found successful applications
+                  not only in solving computationally expensive single- or
+                  multi-objective optimization problems, but also in addressing
+                  dynamic optimization problems, constrained optimization
+                  problems and multi-modal optimization problems. This paper
+                  provides a concise overview of the history and recent
+                  developments in surrogate-assisted evolutionary computation
+                  and suggests a few future trends in this research area.},
+  langid = {english},
+  keywords = {Evolutionary computation,Expensive optimization
+                  problems,Machine learning,Meta-models,Model
+                  management,Surrogates}
+}
+
+ +
+@article{JinWanChu2019data,
+  author = { Yaochu Jin  and Handing Wang and  Tinkle Chugh  and Dan
+                  Guo and  Kaisa Miettinen },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Data-Driven Evolutionary Optimization: An Overview and Case
+                  Studies},
+  year = 2019,
+  month = jun,
+  number = 3,
+  pages = {442--458},
+  volume = 23,
+  doi = {10.1109/tevc.2018.2869001}
+}
+
+ +
+@article{JinWon2010adapt,
+  title = {Adaptive, convergent, and diversified archiving strategy for
+                  multiobjective evolutionary algorithms},
+  author = {Jin, Huidong and Wong, Man-Leung},
+  journal = {Expert Systems with Applications},
+  year = 2010,
+  number = 12,
+  pages = {8462--8470},
+  volume = 37,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Joh54,
+  author = {David S. Johnson},
+  title = {Optimal Two- and Three-stage Production Scheduling
+                  with Setup Times Included},
+  journal = {Naval Research Logistics Quarterly},
+  year = 1954,
+  volume = 1,
+  pages = {61--68}
+}
+
+ +
+@article{JohAraMcGSch1989,
+  author = {David S. Johnson and Cecilia R. Aragon and  Lyle A. McGeoch  and Catherine Schevon},
+  title = {Optimization by Simulated Annealing: An Experimental
+                  Evaluation: Part {I}, Graph Partitioning},
+  journal = {Operations Research},
+  year = 1989,
+  volume = 37,
+  number = 6,
+  pages = {865--892},
+  doi = {10.1287/opre.39.3.378}
+}
+
+ +
+@article{JohAraMcGSch1991,
+  author = {David S. Johnson and Cecilia R. Aragon and  Lyle A. McGeoch  and Catherine Schevon},
+  title = {Optimization by Simulated Annealing: An Experimental
+                  Evaluation: Part {II}, Graph Coloring and Number Partitioning},
+  journal = {Operations Research},
+  year = 1991,
+  volume = 39,
+  number = 3,
+  pages = {378--406}
+}
+
+ +
+@article{JohJac2002:dam,
+  title = {On the Convergence of Generalized Hill Climbing Algorithms},
+  author = { Alan W. Johnson  and  Sheldon H. Jacobson },
+  journal = {Discrete Applied Mathematics},
+  volume = 119,
+  number = 1,
+  pages = {37--57},
+  year = 2002,
+  publisher = {Elsevier}
+}
+
+ +
+@article{JohMooYlv1990,
+  title = {Minimax and maximin distance designs},
+  author = {Johnson, Mark E. and Moore, Leslie M. and Ylvisaker, Donald},
+  journal = {Journal of Statistical Planning and Inference},
+  year = 1990,
+  number = 2,
+  pages = {131--148},
+  volume = 26,
+  keywords = {Bayesian design},
+  alias = {Johnson1990b}
+}
+
+ +
+@article{JohPapYan1988,
+  author = {David S. Johnson and  Christos H. Papadimitriou  and  Mihalis Yannakakis },
+  title = {How Easy is Local Search?},
+  journal = {Journal of Computer System Science},
+  year = 1988,
+  volume = 37,
+  number = 1,
+  pages = {79--100}
+}
+
+ +
+@article{JonKriMicSch2023generalized,
+  title = {Generalized Relax-and-Fix Heuristic},
+  author = {Joncour, C. and Kritter, J. and Michel, S. and Schepler, X.},
+  journal = {Computers \& Operations Research},
+  volume = 149,
+  pages = 106038,
+  year = 2023,
+  publisher = {Elsevier}
+}
+
+ +
+@article{JonSchWel98ego,
+  author = {Donald R. Jones and Matthias Schonlau and William J. Welch},
+  title = {Efficient Global Optimization of Expensive Black-Box
+                  Functions},
+  journal = {Journal of Global Optimization},
+  year = 1998,
+  volume = 13,
+  number = 4,
+  pages = {455--492},
+  keywords = {EGO},
+  alias = {JonSchWel98go},
+  annote = {Proposed EGO algorithm}
+}
+
+ +
+@article{JonSpe1992formal,
+  title = {A formal analysis of the role of multi-point crossover in
+                  genetic algorithms},
+  author = { De Jong, Kenneth A.  and Spears, William M.},
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  volume = 5,
+  number = 1,
+  pages = {1--26},
+  year = 1992,
+  publisher = {Springer}
+}
+
+ +
+@article{JooLeyDec2022knapsack,
+  author = {Jorik Jooken and Pieter Leyman and Patrick {De Causmaecker}},
+  title = {A new class of hard problem instances for the 0--1 knapsack
+                  problem},
+  journal = {European Journal of Operational Research},
+  year = 2022,
+  volume = 301,
+  number = 3,
+  pages = {841--854}
+}
+
+ +
+@article{JooLeyWau2023exploring,
+  author = {Jorik Jooken and Pieter Leyman and Tony Wauters and Patrick {De Causmaecker}},
+  title = {Exploring search space trees using an adapted version of
+                  {Monte} {Carlo} tree search for combinatorial optimization
+                  problems},
+  journal = {Computers \& Operations Research},
+  year = 2023,
+  volume = 150,
+  pages = 106070,
+  doi = {10.1016/j.cor.2022.106070}
+}
+
+ +
+@article{JosCle1999:jair,
+  author = {D. E. Joslin and D. P. Clements},
+  title = {Squeaky Wheel Optimization},
+  journal = {Journal of Artificial Intelligence Research},
+  year = 1999,
+  volume = 10,
+  pages = {353--373}
+}
+
+ +
+@article{Jowitt92,
+  author = { P. W. Jowitt  and  G. Germanopoulos },
+  title = {Optimal pump scheduling in water supply networks},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 1992,
+  volume = 118,
+  number = 4,
+  pages = {406--422},
+  note = {},
+  abstract = {The electricity cost of pumping accounts for a large
+                  part of the total operating cost for water-supply
+                  networks. This study presents a method based on
+                  linear programming for determining an optimal
+                  (minimum cost) schedule of pumping on a 24-hr
+                  basis. Both unit and maximum demand electricity
+                  charges are considered. Account is taken of the
+                  relative efficiencies of the available pumps, the
+                  structure of the electricity tariff, the
+                  consumer-demand profile, and the hydraulic
+                  characteristics and operational constraints of the
+                  network. The use of extended-period simulation of
+                  the network operation in determining the parameters
+                  of the linearized network equations and constraints
+                  and in studying the optimized network operation is
+                  described. An application of the method to an
+                  existing network in the United Kingdom is presented,
+                  showing that considerable savings are possible. The
+                  method was found to be robust and with low
+                  computation-time requirements, and is therefore
+                  suitable for real-time implementation.}
+}
+
+ +
+@article{JuaFauGras2015orp,
+  author = {Angel A. Juan and Javier Faulin and Scott E. Grasman and
+                  Markus Rabe and Gon{\c c}alo Figueira},
+  title = {A review of simheuristics: Extending metaheuristics to deal
+                  with stochastic combinatorial optimization problems},
+  journal = {Operations Research Perspectives},
+  volume = 2,
+  pages = {62--72},
+  year = 2015,
+  doi = {10.1016/j.orp.2015.03.001},
+  keywords = {Metaheuristics; Simulation; Combinatorial optimization;
+                  Stochastic problems}
+}
+
+ +
+@article{JuaLouMatLuoCas2014,
+  author = {Angel A. Juan and  Helena R. {Louren{\c c}o}  and Manuel Mateo and Rachel Luo and Quim Castell{\`{a}}},
+  title = {Using Iterated Local Search for Solving the Flow-shop Problem: Parallelization,
+               Parametrization, and Randomization Issues},
+  journal = {International Transactions in Operational Research},
+  year = 2014,
+  volume = 21,
+  number = 1,
+  pages = {103--126}
+}
+
+ +
+@article{JunReiThi1994,
+  author = {M. J{\"u}nger and  Gerhard Reinelt  and S. Thienel},
+  title = {Provably Good Solutions for the Traveling Salesman Problem},
+  journal = {Zeitschrift f{\"u}r Operations Research},
+  year = 1994,
+  volume = 40,
+  number = 2,
+  pages = {183--217}
+}
+
+ +
+@article{KabColKorLop2017jacryst,
+  author = { Kabova, Elena A.  and  Cole, Jason C.  and  Oliver Korb  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Williams, Adrian C.  and  Shankland, Kenneth },
+  title = {Improved performance of crystal structure solution from
+                  powder diffraction data through parameter tuning of a
+                  simulated annealing algorithm},
+  journal = {Journal of Applied Crystallography},
+  year = 2017,
+  volume = 50,
+  number = 5,
+  pages = {1411--1420},
+  month = oct,
+  doi = {10.1107/S1600576717012602},
+  abstract = {Significant gains in the performance of the simulated
+                  annealing algorithm in the {\it DASH} software package have
+                  been realized by using the {\it irace} automatic
+                  configuration tool to optimize the values of three key
+                  simulated annealing parameters. Specifically, the success
+                  rate in finding the global minimum in intensity $\chi^2$
+                  space is improved by up to an order of magnitude. The general
+                  applicability of these revised simulated annealing parameters
+                  is demonstrated using the crystal structure determinations of
+                  over 100 powder diffraction datasets.},
+  keywords = {crystal structure determination, powder diffraction,
+                  simulated annealing, parameter tuning, irace}
+}
+
+ +
+@article{KahTve1979prospect,
+  title = {Prospect theory: {An} analysis of decision under risk},
+  author = { Kahneman, Daniel  and  Tversky, Amos },
+  journal = {Econometrica},
+  pages = {263--291},
+  volume = 47,
+  number = 2,
+  year = 1979,
+  doi = {10.2307/1914185}
+}
+
+ +
+@article{Kahneman2003maps,
+  title = {Maps of bounded rationality: Psychology for behavioral
+                  economics},
+  author = { Kahneman, Daniel },
+  journal = {The American Economic Review},
+  volume = 93,
+  number = 5,
+  pages = {1449--1475},
+  year = 2003
+}
+
+ +
+@article{KalHasHemSor2023deeprl,
+  author = {Kallestad, Jakob and Hasibi, Ramin and Hemmati, Ahmad and  Kenneth S{\"o}rensen },
+  title = {A general deep reinforcement learning hyperheuristic
+                  framework for solving combinatorial optimization problems},
+  journal = {European Journal of Operational Research},
+  year = 2023,
+  volume = 309,
+  number = 1,
+  pages = {446--468},
+  month = aug,
+  doi = {10.1016/j.ejor.2023.01.017},
+  keywords = {Deep RL, hyper-heuristic, ALNS}
+}
+
+ +
+@article{KanHeWei2013,
+  author = {Qinma Kang and Hong He and Jun Wei},
+  title = {An Effective Iterated Greedy Algorithm for
+                  Reliability-oriented Task Allocation in Distributed Computing
+                  Systems},
+  journal = {Journal of Parallel and Distributed Computing},
+  year = 2013,
+  volume = 73,
+  number = 8,
+  pages = {1106--1115}
+}
+
+ +
+@article{Kar2016,
+  author = {Korhan Karabulut},
+  title = {A hybrid iterated greedy algorithm for total tardiness
+                  minimization in permutation flowshops},
+  journal = {Computers and Industrial Engineering},
+  year = 2016,
+  volume = 98,
+  number = {Supplement C},
+  pages = {300 -- 307}
+}
+
+ +
+@article{KarAka2009:airev,
+  author = { Dervis Karaboga  and  Bahriye Akay },
+  title = {A Survey: Algorithms Simulating Bee Swarm Intelligence},
+  journal = {Artificial Intelligence Review},
+  year = 2009,
+  volume = 31,
+  number = {1--4},
+  pages = {61--85}
+}
+
+ +
+@article{KarHooEib2015:tec,
+  author = {Giorgos Karafotias and Mark Hoogendoorn and  Agoston E. Eiben },
+  title = {Parameter Control in Evolutionary Algorithms: Trends and Challenges},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  volume = 19,
+  number = 2,
+  pages = {167--187}
+}
+
+ +
+@article{KarKok2010tdea,
+  title = {A territory defining multiobjective evolutionary algorithms
+                  and preference incorporation},
+  author = { Karahan, {\.I}brahim  and  Murat K{\"o}ksalan },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 14,
+  number = 4,
+  pages = {636--664},
+  year = 2010,
+  keywords = {TDEA},
+  doi = {10.1109/TEVC.2009.2033586}
+}
+
+ +
+@article{KarMohMey2022ml,
+  author = {Maryam Karimi-Mamaghan and Mehrdad Mohammadi and Patrick
+                  Meyer and Amir Mohammad Karimi-Mamaghan and  Talbi, El-Ghazali },
+  title = {Machine learning at the service of meta-heuristics for
+                  solving combinatorial optimization problems: A
+                  state-of-the-art},
+  journal = {European Journal of Operational Research},
+  year = 2022,
+  volume = 296,
+  number = 2,
+  pages = {393--422},
+  doi = {10.1016/j.ejor.2021.04.032},
+  keywords = {Meta-heuristics, Machine learning, Combinatorial optimization
+                  problems, State-of-the-art},
+  abstract = {In recent years, there has been a growing research interest
+                  in integrating machine learning techniques into
+                  meta-heuristics for solving combinatorial optimization
+                  problems. This integration aims to lead meta-heuristics
+                  toward an efficient, effective, and robust search and improve
+                  their performance in terms of solution quality, convergence
+                  rate, and robustness. Since various integration methods with
+                  different purposes have been developed, there is a need to
+                  review the recent advances in using machine learning
+                  techniques to improve meta-heuristics. To the best of our
+                  knowledge, the literature is deprived of having a
+                  comprehensive yet technical review. To fill this gap, this
+                  paper provides such a review on the use of machine learning
+                  techniques in the design of different elements of
+                  meta-heuristics for different purposes including algorithm
+                  selection, fitness evaluation, initialization, evolution,
+                  parameter setting, and cooperation. First, we describe the
+                  key concepts and preliminaries of each of these ways of
+                  integration. Then, the recent advances in each way of
+                  integration are reviewed and classified based on a proposed
+                  unified taxonomy. Finally, we provide a technical discussion
+                  on the advantages, limitations, requirements, and challenges
+                  of implementing each of these integration ways, followed by
+                  promising future research directions.}
+}
+
+ +
+@article{KarShiDai05:medicine,
+  title = {Prediction of {MHC} class {II} binders using the ant colony
+                  search strategy},
+  author = {Karpenko, Oleksiy and Shi, Jianming and Dai, Yang},
+  journal = {Artificial Intelligence in Medicine},
+  volume = 35,
+  number = 1,
+  pages = {147--156},
+  year = 2005
+}
+
+ +
+@article{KarTas2014,
+  author = {Korhan Karabulut and Fatih M. Tasgetiren},
+  title = {A Variable Iterated Greedy Algorithm for the Traveling Salesman Problem with Time Windows},
+  journal = {Information Sciences},
+  year = 2014,
+  volume = 279,
+  pages = {383--395}
+}
+
+ +
+@article{KasNatRee2017ems,
+  title = {Many objective robust decision making for complex
+                  environmental systems undergoing change},
+  author = { Kasprzyk, Joseph R.  and Nataraj, Shanthi and  Patrick M. Reed  and Lempert,
+                  Robert J.},
+  journal = {Environmental Modelling \& Software},
+  volume = 42,
+  pages = {55--71},
+  year = 2013,
+  keywords = {scenario-based}
+}
+
+ +
+@article{KasReeCha2012ems,
+  title = {Many-objective de {Novo} water supply portfolio planning
+                  under deep uncertainty},
+  author = { Kasprzyk, Joseph R.  and  Patrick M. Reed  and Characklis, Gregory W. and
+                  Kirsch, Brian R.},
+  journal = {Environmental Modelling \& Software},
+  volume = 34,
+  pages = {87--104},
+  year = 2012,
+  keywords = {scenario-based}
+}
+
+ +
+@article{KazCohJea2020,
+  author = {Artem Kaznatcheev and David A. Cohen and Peter Jeavons},
+  title = {Representing Fitness Landscapes by Valued Constraints to Understand
+                  the Complexity of Local Search},
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 69,
+  pages = {1077--1102},
+  year = 2020,
+  doi = {10.1613/jair.1.12156}
+}
+
+ +
+@article{KeArcFen08,
+  author = {Liangjun Ke and Claudia Archetti and Zuren Feng},
+  title = {Ants can solve the team orienteering problem},
+  volume = 54,
+  number = 3,
+  journal = {Computers and Industrial Engineering},
+  year = 2008,
+  pages = {648--665},
+  doi = {10.1016/j.cie.2007.10.001},
+  abstract = {The team orienteering problem {(TOP)} involves
+                  finding a set of paths from the starting point to
+                  the ending point such that the total collected
+                  reward received from visiting a subset of locations
+                  is maximized and the length of each path is
+                  restricted by a pre-specified limit. In this paper,
+                  an ant colony optimization {(ACO)} approach is
+                  proposed for the team orienteering problem. Four
+                  methods, i.e., the sequential,
+                  deterministic-concurrent and random-concurrent and
+                  simultaneous methods, are proposed to construct
+                  candidate solutions in the framework of {ACO}. We
+                  compare these methods according to the results
+                  obtained on well-known problems from the
+                  literature. Finally, we compare the algorithm with
+                  several existing algorithms. The results show that
+                  our algorithm is promising.},
+  keywords = {Ant colony optimization, Ant system, Heuristics,
+                  Team orienteering problem}
+}
+
+ +
+@article{Kee1981or,
+  author = {R. L. Keeney},
+  title = {Analysis of preference dependencies among objectives},
+  journal = {Operations Research},
+  year = 1981,
+  volume = 29,
+  pages = {1105--1120}
+}
+
+ +
+@article{KenBaiBla2016good,
+  author = { Graham Kendall  and Ruibin Bai and Jacek B{\l}azewicz and Patrick {De Causmaecker} and  Michel Gendreau  and Robert John and Jiawei Li and  Barry McCollum  and Erwin Pesch and  Rong Qu  and Nasser Sabar and  Vanden Berghe, Greet   and Angelina Yee},
+  title = {Good Laboratory Practice for Optimization Research},
+  year = 2016,
+  volume = 67,
+  number = 4,
+  pages = {676--689},
+  journal = {Journal of the Operational Research Society},
+  doi = {10.1057/jors.2015.77},
+  alias = {Ken++2016:jors}
+}
+
+ +
+@article{KerHooNeuTra2019,
+  author = { Pascal Kerschke  and  Holger H. Hoos  and  Frank Neumann  and  Heike Trautmann },
+  title = {Automated Algorithm Selection: Survey and Perspectives},
+  journal = {Evolutionary Computation},
+  volume = 27,
+  number = 1,
+  pages = {3--45},
+  year = 2019,
+  doi = {10.1162/evco_a_00242},
+  month = mar
+}
+
+ +
+@article{KerLin70,
+  author = {B. W. Kernighan and S. Lin},
+  title = {An Efficient Heuristic Procedure for Partitioning
+                  Graphs},
+  journal = {Bell Systems Technology Journal},
+  year = 1970,
+  volume = 49,
+  number = 2,
+  pages = {213--219}
+}
+
+ +
+@article{KerTra2019,
+  author = { Pascal Kerschke  and  Heike Trautmann },
+  title = {Automated Algorithm Selection on Continuous Black-Box
+                  Problems by Combining Exploratory Landscape Analysis and
+                  Machine Learning},
+  journal = {Evolutionary Computation},
+  volume = 27,
+  number = 1,
+  pages = {99--127},
+  year = 2019,
+  doi = {10.1162/evco_a_00236},
+  abstract = {In this article, we build upon previous work on designing
+                  informative and efficient Exploratory Landscape Analysis
+                  features for characterizing problems' landscapes and show
+                  their effectiveness in automatically constructing algorithm
+                  selection models in continuous black-box optimization
+                  problems. Focusing on algorithm performance results of the
+                  COCO platform of several years, we construct a representative
+                  set of high-performing complementary solvers and present an
+                  algorithm selection model that, compared to the portfolio's
+                  single best solver, on average requires less than half of the
+                  resources for solving a given problem. Therefore, there is a
+                  huge gain in efficiency compared to classical ensemble
+                  methods combined with an increased insight into problem
+                  characteristics and algorithm properties by using informative
+                  features. The model acts on the assumption that the function
+                  set of the Black-Box Optimization Benchmark is representative
+                  enough for practical applications. The model allows for
+                  selecting the best suited optimization algorithm within the
+                  considered set for unseen problems prior to the optimization
+                  itself based on a small sample of function evaluations. Note
+                  that such a sample can even be reused for the initial
+                  population of an evolutionary (optimization) algorithm so
+                  that even the feature costs become negligible. }
+}
+
+ +
+@article{KerWanPreuGrim2019search,
+  doi = {10.1162/evco_a_00234},
+  year = 2019,
+  publisher = {MIT Press},
+  volume = 27,
+  number = 4,
+  pages = {577--609},
+  author = { Pascal Kerschke  and Hao Wang and  Mike Preuss  and Christian
+                  Grimme and   Andr{\'{e}} H. Deutz  and  Heike Trautmann  and  Emmerich, Michael T. M. },
+  title = {Search Dynamics on Multimodal Multiobjective Problems},
+  journal = {Evolutionary Computation}
+}
+
+ +
+@article{Kerr1998harking,
+  doi = {10.1207/s15327957pspr0203_4},
+  year = 1998,
+  month = aug,
+  publisher = {{SAGE} Publications},
+  volume = 2,
+  number = 3,
+  pages = {196--217},
+  author = {Norbert L. Kerr},
+  title = {{HARKing}: Hypothesizing After the Results are Known},
+  journal = {Personality and Social Psychology Review}
+}
+
+ +
+@article{KhuXuHooLey16:aij,
+  author = { KhudaBukhsh, A. R.  and  Lin Xu  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {{SATenstein}: Automatically Building Local Search {SAT}
+                  {Solvers} from {Components}},
+  journal = {Artificial Intelligence},
+  year = 2016,
+  volume = 232,
+  pages = {20--42},
+  doi = {10.1016/j.artint.2015.11.002}
+}
+
+ +
+@article{KilUrl2015constr,
+  author = {Philip Kilby  and Tommaso Urli},
+  title = {Fleet design optimisation from historical data using constraint programming and large neighbourhood search},
+  journal = {Constraints},
+  year = 2015,
+  pages = {1--20},
+  publisher = {Springer, US},
+  doi = {10.1007/s10601-015-9203-0},
+  keywords = {F-race}
+}
+
+ +
+@article{Kim1993,
+  author = {Kim, Yeong-Dae},
+  title = {Heuristics for Flowshop Scheduling Problems Minimizing Mean
+                  Tardiness},
+  journal = {Journal of the Operational Research Society},
+  year = 1993,
+  volume = 44,
+  number = 1,
+  pages = {19--28},
+  doi = {10.1057/jors.1993.3}
+}
+
+ +
+@article{KimAllLop2020arxiv,
+  author = { Kim, Youngmin  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Safe Learning and Optimization Techniques: Towards a Survey
+                  of the State of the Art},
+  journal = {Arxiv preprint arXiv:2101.09505 [cs.LG]},
+  year = 2020,
+  url = {https://arxiv.org/abs/2101.09505},
+  abstract = {Safe learning and optimization deals with learning and
+                  optimization problems that avoid, as much as possible, the
+                  evaluation of non-safe input points, which are solutions,
+                  policies, or strategies that cause an irrecoverable loss
+                  (e.g., breakage of a machine or equipment, or life
+                  threat). Although a comprehensive survey of safe
+                  reinforcement learning algorithms was published in 2015, a
+                  number of new algorithms have been proposed thereafter, and
+                  related works in active learning and in optimization were not
+                  considered. This paper reviews those algorithms from a number
+                  of domains including reinforcement learning, Gaussian process
+                  regression and classification, evolutionary algorithms, and
+                  active learning. We provide the fundamental concepts on which
+                  the reviewed algorithms are based and a characterization of
+                  the individual algorithms. We conclude by explaining how the
+                  algorithms are connected and suggestions for future
+                  research. }
+}
+
+ +
+@article{KimCouYou2021set,
+  title = {Bayesian Optimization with Approximate Set Kernels},
+  author = {Jungtaek Kim and Michael McCourt and Tackgeun You and Saehoon
+                  Kim and Seungjin Choi},
+  abstract = {We propose a practical Bayesian optimization method over
+                  sets, to minimize a black-box function that takes a set as a
+                  single input. Because set inputs are permutation-invariant,
+                  traditional Gaussian process-based Bayesian optimization
+                  strategies which assume vector inputs can fall short. To
+                  address this, we develop a Bayesian optimization method with
+                  \emph{set kernel} that is used to build surrogate
+                  functions. This kernel accumulates similarity over set
+                  elements to enforce permutation-invariance, but this comes at
+                  a greater computational cost. To reduce this burden, we
+                  propose two key components: (i) a more efficient approximate
+                  set kernel which is still positive-definite and is an
+                  unbiased estimator of the true set kernel with upper-bounded
+                  variance in terms of the number of subsamples, (ii) a
+                  constrained acquisition function optimization over sets,
+                  which uses symmetry of the feasible region that defines a set
+                  input. Finally, we present several numerical experiments
+                  which demonstrate that our method outperforms other methods.},
+  journal = {Machine Learning},
+  year = 2021,
+  doi = {10.1007/s10994-021-05949-0}
+}
+
+ +
+@article{KimParLee2017,
+  author = {Kim, J.-S. and Park, J.-H. and Lee, D.-H.},
+  title = {Iterated Greedy Algorithms to Minimize the Total Family Flow
+                  Time for Job-shop Scheduling with Job Families and
+                  Sequence-dependent Set-ups},
+  journal = {Engineering Optimization},
+  year = 2017,
+  volume = 49,
+  number = 10,
+  pages = {1719--1732}
+}
+
+ +
+@article{KinBa2014adam,
+  title = {Adam: A method for stochastic optimization},
+  author = {Kingma, Diederik P. and Ba, Jimmy},
+  journal = {Arxiv preprint arXiv:1412.6980 [cs.LG]},
+  year = 2014,
+  url = {https://arxiv.org/abs/1412.6980},
+  annote = {Published as a conference paper at the 3rd International
+                  Conference for Learning Representations, San Diego, 2015~\cite{KinBa2015adam}}
+}
+
+ +
+@article{KirTou1985,
+  author = { Scott Kirkpatrick  and G. Toulouse},
+  title = {Configuration Space Analysis of Travelling Salesman Problems},
+  journal = {Journal de Physique},
+  year = 1985,
+  volume = 46,
+  number = 8,
+  pages = {1277--1292}
+}
+
+ +
+@article{Kirkpatrick1984,
+  author = { Scott Kirkpatrick },
+  title = {Optimization by Simulated Annealing: Quantitative Studies},
+  journal = {Journal of Statistical Physics},
+  year = 1984,
+  volume = 34,
+  number = {5-6},
+  pages = {975--986}
+}
+
+ +
+@article{Kirkpatrick83,
+  author = { Scott Kirkpatrick  and C. D. Gelatt and M. P. Vecchi},
+  title = {Optimization by Simulated Annealing},
+  journal = {Science},
+  year = 1983,
+  volume = 220,
+  number = 4598,
+  pages = {671--680},
+  annote = {Proposed Simulated Annealing},
+  doi = {10.1126/science.220.4598.671}
+}
+
+ +
+@article{KlaMosNau2017iwoven,
+  author = { Kathrin Klamroth  and  Mostaghim, Sanaz  and  Boris Naujoks  and Silvia
+                  Poles and  Robin C. Purshouse  and  G{\"u}nther Rudolph  and Ruzika,
+                  Stefan and Serpil Say{\i}n and  Margaret M. Wiecek  and  Xin Yao },
+  title = {Multiobjective optimization for interwoven systems},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  year = 2017,
+  volume = 24,
+  number = {1-2},
+  pages = {71--81},
+  doi = {10.1002/mcda.1598}
+}
+
+ +
+@article{KleShaHom2002,
+  author = {Anton J. Kleywegt and Alexander Shapiro and Tito Homem{-}de{-}Mello},
+  title = {The Sample Average Approximation Method for Stochastic Discrete Optimization},
+  journal = {SIAM Journal on Optimization},
+  year = 2002,
+  volume = 12,
+  number = 2,
+  pages = {479--502}
+}
+
+ +
+@article{Kno2005tec,
+  author = { Joshua D. Knowles },
+  title = {{ParEGO}: A hybrid algorithm with on-line landscape
+                  approximation for expensive multiobjective optimization
+                  problems},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2006,
+  volume = 10,
+  number = 1,
+  pages = {50--66},
+  doi = {10.1109/TEVC.2005.851274},
+  keywords = {ParEGO, online, metamodel}
+}
+
+ +
+@article{Kno2009closed,
+  author = { Joshua D. Knowles },
+  title = {Closed-loop evolutionary multiobjective optimization},
+  journal = {IEEE Computational Intelligence Magazine},
+  volume = 4,
+  issue = 3,
+  pages = {77--91},
+  doi = {10.1109/MCI.2009.933095},
+  year = 2009,
+  abstract = {Artificial evolution has been used for more than 50 years as a method of optimization in engineering, operations research and computational intelligence. In closed-loop evolution (a term used by the statistician, George Box) or, equivalently, evolutionary experimentation (Ingo Rechenberg's terminology), the ``phenotypes'' are evaluated in the real world by conducting a physical experiment, whilst selection and breeding is simulated. Well-known early work on artificial evolution\textemdash design engineering problems in fluid dynamics, and chemical plant process optimization\textemdash was carried out in this experimental mode. More recently, the closed-loop approach has been successfully used in much evolvable hardware and evolutionary robotics research, and in some microbiology and biochemistry applications. In this article, several further new targets for closed-loop evolutionary and multiobjective optimization are considered. Four case studies from my own collaborative work are described: (i) instrument optimization in analytical biochemistry; (ii) finding effective drug combinations in vitro; (iii) onchip synthetic biomolecule design; and (iv) improving chocolate production processes. Accurate simulation in these applications is not possible due to complexity or a lack of adequate analytical models. In these and other applications discussed, optimizing experimentally brings with it several challenges: noise; nuisance factors; ephemeral resource constraints; expensive evaluations, and evaluations that must be done in (large) batches. Evolutionary algorithms (EAs) are largely equal to these vagaries, whilst modern multiobjective EAs also enable tradeoffs among conflicting optimization goals to be explored. Nevertheless, principles from other disciplines, such as statistics, Design of Experiments, machine learning and global optimization are also relevant to aspects of the closed-loop problem, and may inspire futher development of multiobjective EAs.},
+  langid = {english}
+}
+
+ +
+@article{KnoCor00paes,
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {Approximating the Nondominated Front Using the
+                  {Pareto} Archived Evolution Strategy},
+  journal = {Evolutionary Computation},
+  volume = 8,
+  number = 2,
+  pages = {149--172},
+  year = 2000,
+  doi = {10.1162/106365600568167},
+  annote = {Proposed PAES}
+}
+
+ +
+@article{KnoCor2003tec,
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {Properties of an Adaptive Archiving Algorithm for Storing
+                  Nondominated Vectors},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2003,
+  volume = 7,
+  number = 2,
+  pages = {100--116},
+  month = apr,
+  keywords = {S-metric, hypervolume},
+  annote = {Proposed to use S-metric (hypervolume metric) for
+                  environmental selection}
+}
+
+ +
+@article{KnoVanGro2011,
+  author = {Knol, Mirjam J. and VanderWeele, Tyler J. and Groenwold, Rolf H. H.
+                  and Klungel, Olaf H. and Rovers, Maroeska M. and Grobbee, Diederick E.},
+  title = {Estimating measures of interaction on an additive scale for preventive exposures},
+  journal = {European Journal of Epidemiology},
+  year = 2011,
+  volume = 26,
+  number = 6,
+  pages = {433--438}
+}
+
+ +
+@article{KocGloAli2004ors,
+  author = { Gary A. Kochenberger  and  Fred Glover  and Alidaee, Bahram and Rego,
+                  Cesar},
+  title = {A unified modeling and solution framework for combinatorial
+                  optimization problems},
+  journal = {OR Spektrum},
+  year = 2004,
+  volume = 26,
+  number = 2,
+  pages = {237--250}
+}
+
+ +
+@article{KocHaoGlo2014bqap,
+  title = {The unconstrained binary quadratic programming problem: a
+                  survey},
+  author = { Gary A. Kochenberger  and  Jin-Kao Hao  and  Fred Glover  and Lewis, Mark and L{\"u}, Zhipeng and Wang, Haibo and Wang, Yang},
+  journal = {Journal of Combinatorial Optimization},
+  volume = 28,
+  number = 1,
+  pages = {58--81},
+  year = 2014,
+  doi = {10.1007/s10878-014-9734-0}
+}
+
+ +
+@article{Koe2009jmcda,
+  author = { Murat K{\"o}ksalan },
+  title = {Multiobjective Combinatorial Optimization: Some
+                  Approaches},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  year = 2009,
+  volume = 15,
+  pages = {69--78},
+  doi = {10.1002/mcda.425}
+}
+
+ +
+@article{KokKar2010itdea,
+  title = {An Interactive Territory Defining Evolutionary Algorithm:
+                  {iTDEA}},
+  volume = 14,
+  doi = {10.1109/TEVC.2010.2070070},
+  number = 5,
+  journal = {IEEE Transactions on Evolutionary Computation},
+  author = { Murat K{\"o}ksalan  and  Karahan, {\.I}brahim },
+  month = oct,
+  year = 2010,
+  pages = {702--722}
+}
+
+ +
+@article{KolHar2006ejor,
+  author = {Kolisch, Rainer and Hartmann, S{\"o}nke},
+  title = {Experimental investigation of heuristics for
+                  resource-constrained project scheduling: An update},
+  volume = 174,
+  doi = {10.1016/j.ejor.2005.01.065},
+  abstract = {This paper considers heuristics for the well-known
+                  resource-constrained project scheduling problem
+                  ({RCPSP).} It provides an update of our survey which
+                  was published in 2000. We summarize and categorize a
+                  large number of heuristics that have recently been
+                  proposed in the literature. Most of these heuristics
+                  are then evaluated in a computational study and
+                  compared on the basis of our standardized
+                  experimental design. Based on the computational
+                  results we discuss features of good heuristics. The
+                  paper closes with some remarks on our test design
+                  and a summary of the recent developments in research
+                  on heuristics for the {RCPSP}.},
+  number = 1,
+  journal = {European Journal of Operational Research},
+  month = oct,
+  year = 2006,
+  keywords = {Computational evaluation, Heuristics, Project
+                  scheduling, Resource constraints},
+  pages = {23--37}
+}
+
+ +
+@article{KolPap2007approx,
+  title = {Approximately dominating representatives},
+  author = {Koltun, Vladlen and  Christos H. Papadimitriou },
+  journal = {Theoretical Computer Science},
+  year = 2007,
+  number = 3,
+  pages = {148--154},
+  volume = 371,
+  publisher = {Elsevier}
+}
+
+ +
+@article{KolPes1994,
+  author = {A. Kolen and  Erwin Pesch },
+  title = {Genetic Local Search in Combinatorial Optimization},
+  journal = {Discrete Applied Mathematics},
+  year = 1994,
+  volume = 48,
+  number = 3,
+  pages = {273--284}
+}
+
+ +
+@article{KolRee2007video,
+  title = {A framework for visually interactive decision-making and
+                  design using evolutionary multi-objective optimization
+                  ({VIDEO})},
+  author = { Kollat, Joshua B.  and  Patrick M. Reed },
+  journal = {Environmental Modelling \& Software},
+  volume = 22,
+  number = 12,
+  pages = {1691--1704},
+  year = 2007,
+  keywords = {glyph plot}
+}
+
+ +
+@article{KooBec57,
+  author = {Tjalling C. Koopmans and Martin J. Beckmann},
+  title = {Assignment Problems and the Location of Economic Activities},
+  journal = {Econometrica},
+  volume = 25,
+  pages = {53--76},
+  year = 1957,
+  annote = {Introduced the Quadratic Assignment Problem (QAP)}
+}
+
+ +
+@article{Kor1985omega,
+  author = {Kornbluth, Jsh},
+  title = {Sequential multi-criterion decision making},
+  doi = {10.1016/0305-0483(85)90045-3},
+  abstract = {In this paper we consider a simple sequential
+                  multicriterion decision making problem in which a
+                  decision maker has to accept or reject a series of
+                  multi-attributed outcomes. We show that using very
+                  simple programming techniques, a great deal of the
+                  decision making can be automated. The method might
+                  be applicable to situations in which a dealer is
+                  having to consider sequential offers in a trading
+                  market.},
+  number = 6,
+  volume = 13,
+  journal = {Omega},
+  year = 1985,
+  keywords = {machine decision making},
+  pages = {569--574}
+}
+
+ +
+@article{KorMosWal1990choice,
+  author = { Pekka Korhonen  and Moskowitz, Herbert and  Wallenius, Jyrki },
+  title = {Choice Behavior in Interactive Multiple-Criteria Decision
+                  Making},
+  journal = {Annals of Operations Research},
+  year = 1990,
+  volume = 23,
+  number = 1,
+  pages = {161--179},
+  month = dec,
+  doi = {10.1007/BF02204844},
+  abstract = {Choice behavior in an interactive multiple-criteria decision
+                  making environment is examined experimentally. A ``free
+                  search'' discrete visual interactive reference direction
+                  approach was used on a microcomputer by management students
+                  to solve two realistic and relevant multiple-criteria
+                  decision problems. The results revealed persistent patterns
+                  of intransitive choice behavior, and an unexpectedly rapid
+                  degree of convergence of the reference direction approach on
+                  a preferred solution. The results can be explained using
+                  Tversky' additive utility difference model and
+                  Kahneman-Tversky's prospect theory. The implications of the
+                  results for the design of interactive multiple-criteria
+                  decision procedures are discussed.}
+}
+
+ +
+@article{KorPagFal2001,
+  title = {On the ``dimensionality curse'' and the ``self-similarity
+                  blessing''},
+  author = {Korn, Flip and Pagel, B.-U. and Faloutsos, Christos},
+  journal = {IEEE Transactions on Knowledge and Data Engineering},
+  volume = 13,
+  number = 1,
+  pages = {96--111},
+  year = 2001,
+  doi = {10.1109/69.908983},
+  abstract = {Spatial queries in high-dimensional spaces have been studied
+                  extensively. Among them, nearest neighbor queries are
+                  important in many settings, including spatial databases (Find
+                  the k closest cities) and multimedia databases (Find the k
+                  most similar images). Previous analyses have concluded that
+                  nearest-neighbor search is hopeless in high dimensions due to
+                  the notorious "curse of dimensionality". We show that this
+                  may be overpessimistic. We show that what determines the
+                  search performance (at least for R-tree-like structures) is
+                  the intrinsic dimensionality of the data set and not the
+                  dimensionality of the address space (referred to as the
+                  embedding dimensionality). The typical (and often implicit)
+                  assumption in many previous studies is that the data is
+                  uniformly distributed, with independence between
+                  attributes. However, real data sets overwhelmingly disobey
+                  these assumptions; rather, they typically are skewed and
+                  exhibit intrinsic ("fractal") dimensionalities that are much
+                  lower than their embedding dimension, e.g. due to subtle
+                  dependencies between attributes. We show how the Hausdorff
+                  and Correlation fractal dimensions of a data set can yield
+                  extremely accurate formulas that can predict the I/O
+                  performance to within one standard deviation on multiple real
+                  and synthetic data sets.}
+}
+
+ +
+@article{KorSilRob04:ml-aco,
+  author = { P. Koro{\v s}ec  and  Jurij {\v S}ilc  and B. Robi{\v c}},
+  title = {Solving the mesh-partitioning problem with an
+                  ant-colony algorithm},
+  journal = {Parallel Computing},
+  year = 2004,
+  volume = 30,
+  pages = {785--801}
+}
+
+ +
+@article{KorSilWalOor2012linear,
+  author = { Pekka Korhonen  and Silvennoinen, Kari and  Wallenius, Jyrki  and {\"O}{\"o}rni, Anssi},
+  title = {Can a linear value function explain choices? {An}
+                  experimental study},
+  journal = {European Journal of Operational Research},
+  year = 2012,
+  volume = 219,
+  number = 2,
+  pages = {360--367},
+  month = jun,
+  shorttitle = {Can a linear value function explain choices?},
+  doi = {10.1016/j.ejor.2011.12.040},
+  abstract = {We investigate in a simple bi-criteria experimental study,
+                  whether subjects are consistent with a linear value function
+                  while making binary choices. Many inconsistencies appeared in
+                  our experiment. However, the impact of inconsistencies on the
+                  linearity vs. non-linearity of the value function was
+                  minor. Moreover, a linear value function seems to predict
+                  choices for bi-criteria problems quite well. This ability to
+                  predict is independent of whether the value function is
+                  diagnosed linear or not. Inconsistencies in responses did not
+                  necessarily change the original diagnosis of the form of the
+                  value function. Our findings have implications for the design
+                  and development of decision support tools for Multiple
+                  Criteria Decision Making problems.},
+  language = {en},
+  keywords = {Binary choices, Inconsistency, Linear value function,
+                  Multiple criteria, Weights}
+}
+
+ +
+@article{KorStuExn07:si,
+  author = { Oliver Korb  and  Thomas St{\"u}tzle  and  Thomas E. Exner },
+  title = {An Ant Colony Optimization Approach to Flexible
+                  Protein--Ligand Docking},
+  journal = {Swarm Intelligence},
+  year = 2007,
+  volume = 1,
+  number = 2,
+  pages = {115--134}
+}
+
+ +
+@article{KorStuExn2009jcim,
+  author = { Oliver Korb  and  Thomas St{\"u}tzle  and  Thomas E. Exner },
+  title = {Empirical Scoring Functions for Advanced Protein-Ligand Docking with {PLANTS}},
+  journal = {Journal of Chemical Information and Modeling},
+  year = 2009,
+  volume = 49,
+  number = 2,
+  pages = {84--96}
+}
+
+ +
+@article{KorStuExn2010jcim,
+  author = { Oliver Korb  and Peter Monecke and Gerhard Hessler and  Thomas St{\"u}tzle  and  Thomas E. Exner },
+  title = {pharm{ACO}phore: Multiple Flexible Ligand Alignment Based on Ant Colony Optimization},
+  journal = {Journal of Chemical Information and Modeling},
+  year = 2010,
+  volume = 50,
+  number = 9,
+  pages = {1669--1681}
+}
+
+ +
+@article{KorWal1998paretorace,
+  author = { Pekka Korhonen  and  Wallenius, Jyrki },
+  title = {A pareto race},
+  journal = {Naval Research Logistics},
+  year = 1988,
+  volume = 35,
+  number = 6,
+  pages = {615--623},
+  doi = {10.1002/1520-6750(198812)35:6<615::AID-NAV3220350608>3.0.CO;2-K},
+  abstract = {A dynamic and visual ``free-search'' type of interactive
+                  procedure for multiple-objective linear programming is
+                  presented. The method enables a decision maker to freely
+                  search any part of the efficient frontier by controlling the
+                  speed and direction of motion. The objective function values
+                  are represented in numeric form and as bar graphs on a
+                  display. The method is implemented on an IBM PC/1
+                  microcomputer and is illustrated using a multiple-objective
+                  linear-programming model for managing disposal of sewage
+                  sludge in the New York Bight. Some other applications are
+                  also briefly discussed.}
+}
+
+ +
+@article{Kot2014:aim,
+  author = {Kotthoff, Lars},
+  title = {Algorithm Selection for Combinatorial Search Problems: {A} Survey},
+  journal = {{AI} Magazine},
+  year = 2014,
+  volume = 35,
+  number = 3,
+  pages = {48--60}
+}
+
+ +
+@article{KotNeuRogWit2012swarm,
+  author = {K{\"o}tzing, Timo  and  Frank Neumann  and R{\"o}glin, Heiko and  Carsten Witt },
+  title = {Theoretical Analysis of Two {ACO} Approaches for the
+                  Traveling Salesman Problem},
+  journal = {Swarm Intelligence},
+  year = 2012,
+  volume = 6,
+  number = 1,
+  pages = {1--21},
+  abstract = {Bioinspired algorithms, such as evolutionary algorithms and
+                  ant colony optimization, are widely used for different
+                  combinatorial optimization problems. These algorithms rely
+                  heavily on the use of randomness and are hard to understand
+                  from a theoretical point of view. This paper contributes to
+                  the theoretical analysis of ant colony optimization and
+                  studies this type of algorithm on one of the most prominent
+                  combinatorial optimization problems, namely the traveling
+                  salesperson problem (TSP). We present a new construction
+                  graph and show that it has a stronger local property than one
+                  commonly used for constructing solutions of the TSP. The
+                  rigorous runtime analysis for two ant colony optimization
+                  algorithms, based on these two construction procedures, shows
+                  that they lead to good approximation in expected polynomial
+                  time on random instances. Furthermore, we point out in which
+                  situations our algorithms get trapped in local optima and
+                  show where the use of the right amount of heuristic
+                  information is provably beneficial.},
+  doi = {10.1007/s11721-011-0059-7}
+}
+
+ +
+@article{KotThoHooHutLey2016autoweka,
+  title = {{Auto-WEKA} 2.0: Automatic model selection and hyperparameter
+                  optimization in {WEKA}},
+  author = {Kotthoff, Lars and Thornton, Chris and  Holger H. Hoos  and  Frank Hutter  and  Kevin Leyton-Brown },
+  journal = {Journal of Machine Learning Research},
+  volume = 17,
+  pages = {1--5},
+  year = 2016
+}
+
+ +
+@article{KowStaMad2009sustainable,
+  title = {Sustainable energy futures: Methodological challenges in
+                  combining scenarios and participatory multi-criteria
+                  analysis},
+  author = {Kowalski, Katharina and Stagl, Sigrid and Madlener, Reinhard
+                  and Omann, Ines},
+  journal = {European Journal of Operational Research},
+  volume = 197,
+  number = 3,
+  pages = {1063--1074},
+  year = 2009,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Kra2010,
+  author = {Oliver Kramer},
+  title = {Iterated Local Search with {Powell}'s Method: A Memetic
+                  Algorithm for Continuous Global Optimization},
+  journal = {Memetic Computing},
+  year = 2010,
+  volume = 2,
+  number = 1,
+  pages = {69--83},
+  doi = {10.1007/s12293-010-0032-9},
+  alias = {DBLP:journals/memetic/Kramer10}
+}
+
+ +
+@article{KraErdBeh2012sumo,
+  title = {Recent development and applications of {SUMO} - {Simulation}
+                  of {Urban} {MO}bility},
+  author = { Krajzewicz, Daniel  and Erdmann, Jakob and Behrisch, Michael
+                  and Bieker, Laura},
+  journal = {International Journal On Advances in Systems and
+                  Measurements},
+  volume = 5,
+  number = {3-4},
+  year = 2012,
+  pages = {128--138}
+}
+
+ +
+@article{Kreipl00:js,
+  author = {S. Kreipl},
+  title = {A Large Step Random Walk for Minimizing Total Weighted Tardiness in a Job Shop},
+  journal = {Journal of Scheduling},
+  year = 2000,
+  volume = 3,
+  number = 3,
+  pages = {125--138}
+}
+
+ +
+@article{KriTriDoe2017,
+  author = {Stefanie Kritzinger and Fabien Tricoire and  Karl F. Doerner  and  Richard F. Hartl  and  Thomas St{\"u}tzle },
+  title = {A Unified Framework for Routing Problems with a Fixed Fleet Size},
+  journal = {International Journal of Metaheuristics},
+  year = 2017,
+  volume = 6,
+  number = 3,
+  pages = {160--209}
+}
+
+ +
+@article{Kruskal1956,
+  title = {On the shortest spanning subtree of a graph and the traveling salesman problem},
+  author = {Kruskal, Joseph B},
+  journal = {Proceedings of the American Mathematical Society},
+  volume = 7,
+  number = 1,
+  pages = {48--50},
+  year = 1956
+}
+
+ +
+@article{KuhBie2016:cor,
+  author = {Kuhpfahl, J and Christian Bierwirth},
+  title = {A Study on Local Search Neighborhoods for the Job Shop Scheduling Problem with Total Weighted Tardiness Objective},
+  journal = {Computers \& Operations Research},
+  year = 2016,
+  volume = 66,
+  pages = {44--57}
+}
+
+ +
+@article{KuhFonPaqRuz2016hvsubset,
+  title = {Hypervolume subset selection in two dimensions: Formulations
+                  and algorithms},
+  author = {Kuhn, Tobias and  Carlos M. Fonseca  and  Lu{\'i}s Paquete  and Ruzika,
+                  Stefan and Duarte, Miguel M. and  Jos{\'e} Rui Figueira },
+  journal = {Evolutionary Computation},
+  year = 2016,
+  number = 3,
+  pages = {411--425},
+  volume = 24
+}
+
+ +
+@article{Kuhn1955,
+  title = {The hungarian method for the assignment problem},
+  author = {Kuhn, Harold W.},
+  journal = {Naval Research Logistics Quarterly},
+  volume = 2,
+  number = {1--2},
+  pages = {83--97},
+  year = 1955
+}
+
+ +
+@article{Kuhn2008:jss,
+  author = {Kuhn, Max},
+  title = {Building Predictive Models in \proglang{R} Using the {\rpackage{caret}} Package},
+  journal = {Journal of Statistical Software},
+  volume = 28,
+  number = 5,
+  year = 2008,
+  pages = {1--26}
+}
+
+ +
+@article{KumSin2007sci,
+  author = {Kumar, R. and Singh, P. K.},
+  title = {{Pareto} Evolutionary Algorithm Hybridized with
+                  Local Search for Biobjective {TSP}},
+  journal = {Studies in Computational Intelligence},
+  year = 2007,
+  volume = 75,
+  pages = {361--398}
+}
+
+ +
+@article{KunLucPre1975jacm,
+  author = {H. T. Kung and F. Luccio and F. P. Preparata},
+  title = {On Finding the Maxima of a Set of Vectors},
+  pages = {469--476},
+  journal = {Journal of the ACM},
+  year = 1975,
+  volume = 22,
+  number = 4,
+  alias = {kung:maxima}
+}
+
+ +
+@article{KurDav1982ms,
+  author = {Kurtulus, I. and Davis, E. W.},
+  title = {Multi-Project Scheduling: Categorization of
+                  Heuristic Rules Performance},
+  volume = 28,
+  doi = {10.1287/mnsc.28.2.161},
+  abstract = {Application of heuristic solution procedures to the
+                  practical problem of project scheduling has
+                  previously been studied by numerous
+                  researchers. However, there is little consensus
+                  about their findings, and the practicing manager is
+                  currently at a loss as to which scheduling rule to
+                  use. Furthermore, since no categorization process
+                  was developed, it is assumed that once a rule is
+                  selected it must be used throughout the whole
+                  project. This research breaks away from this
+                  tradition by providing a categorization process
+                  based on two powerful project summary measures. The
+                  first measure identifies the location of the peak of
+                  total resource requirements and the second measure
+                  identifies the rate of utilization of each resource
+                  type. The performance of the rules are classified
+                  according to values of these two measures, and it is
+                  shown that a rule introduced by this research
+                  performs significantly better on most categories of
+                  projects.},
+  number = 2,
+  journal = {Management Science},
+  year = 1982,
+  keywords = {project management, research and development},
+  pages = {161--172}
+}
+
+ +
+@article{Kushner1964,
+  author = {Kushner, H. J.},
+  journal = {Journal of Basic Engineering},
+  title = {A New Method of Locating the Maximum Point of an Arbitrary
+                  Multipeak Curve in the Presence of Noise},
+  year = 1964,
+  issn = {0021-9223},
+  month = mar,
+  number = 1,
+  pages = {97--106},
+  volume = 86,
+  abstract = {A versatile and practical method of searching a parameter
+                  space is presented. Theoretical and experimental results
+                  illustrate the usefulness of the method for such problems as
+                  the experimental optimization of the performance of a system
+                  with a very general multipeak performance function when the
+                  only available information is noise-distributed samples of
+                  the function. At present, its usefulness is restricted to
+                  optimization with respect to one system parameter. The
+                  observations are taken sequentially; but, as opposed to the
+                  gradient method, the observation may be located anywhere on
+                  the parameter interval. A sequence of estimates of the
+                  location of the curve maximum is generated. The location of
+                  the next observation may be interpreted as the location of
+                  the most likely competitor (with the current best estimate)
+                  for the location of the curve maximum. A Brownian motion
+                  stochastic process is selected as a model for the unknown
+                  function, and the observations are interpreted with respect
+                  to the model. The model gives the results a simple intuitive
+                  interpretation and allows the use of simple but efficient
+                  sampling procedures. The resulting process possesses some
+                  powerful convergence properties in the presence of noise; it
+                  is nonparametric and, despite its generality, is efficient in
+                  the use of observations. The approach seems quite promising
+                  as a solution to many of the problems of experimental system
+                  optimization.},
+  doi = {10.1115/1.3653121},
+  epub = {https://asmedigitalcollection.asme.org/fluidsengineering/article-pdf/86/1/97/5763745/97_1.pdf}
+}
+
+ +
+@article{Kwa2017ems,
+  title = {The Exploratory Modeling Workbench: An open source toolkit
+                  for exploratory modeling, scenario discovery, and
+                  (multi-objective) robust decision making},
+  author = { Kwakkel, Jan H. },
+  journal = {Environmental Modelling \& Software},
+  volume = 96,
+  pages = {239--250},
+  year = 2017
+}
+
+ +
+@article{LTDZ2002b,
+  author = { Marco Laumanns  and  Lothar Thiele  and  Kalyanmoy Deb  and  Eckart Zitzler },
+  title = {Combining Convergence and Diversity in Evolutionary
+                  Multiobjective Optimization},
+  journal = {Evolutionary Computation},
+  year = 2002,
+  volume = 10,
+  number = 3,
+  pages = {263--282},
+  doi = {10.1162/106365602760234108},
+  keywords = {archiving, $\epsilon$-dominance, $\epsilon$-approximation,
+                  $\epsilon$-Pareto},
+  annote = {Proposed $\epsilon$-approx and $\epsilon$-Pareto archivers}
+}
+
+ +
+@article{LaTMuePen11:soco,
+  author = {LaTorre, Antonio and Muelas, Santiago and Pe{\~n}a,
+                  Jos{\'e}-Mar{\'i}a},
+  title = {A {MOS}-based dynamic memetic differential evolution
+                  algorithm for continuous optimization: a scalability test},
+  journal = {Soft Computing},
+  year = 2011,
+  volume = 15,
+  number = 11,
+  pages = {2187--2199}
+}
+
+ +
+@article{LaaAarLen1992:or,
+  author = { Peter J. M. van Laarhoven and  Emile H. L. Aarts  and  Jan Karel Lenstra },
+  title = {Job Shop Scheduling by Simulated Annealing},
+  journal = {Operations Research},
+  year = 1992,
+  volume = 40,
+  number = 1,
+  pages = {113--125}
+}
+
+ +
+@article{LabMarSav1998:npp,
+  author = {Martine Labb{\'e} and Patrice Marcotte and Gilles Savard},
+  title = {A Bilevel Model of Taxation and Its Application to Optimal Highway Pricing},
+  journal = {Management Science},
+  volume = 44,
+  number = 12,
+  year = 1998,
+  pages = {1608--1622},
+  doi = {10.1287/mnsc.44.12.1608},
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{LabVio2013:npp,
+  author = {Martine Labb{\'e} and Alessia Violin},
+  title = {Bilevel programming and price setting problems},
+  journal = {{4OR}: A Quarterly Journal of Operations Research},
+  year = 2013,
+  volume = 11,
+  number = 1,
+  pages = {1--30},
+  doi = {10.1007/s10288-012-0213-0}
+}
+
+ +
+@article{LacMolHer2014is,
+  author = {Benjamin Lacroix and  Daniel Molina  and  Francisco Herrera },
+  title = {Region based memetic algorithm for real-parameter
+                  optimisation},
+  journal = {Information Sciences},
+  year = 2014,
+  volume = 262,
+  pages = {15--31},
+  doi = {10.1016/j.ins.2013.11.032},
+  keywords = {irace}
+}
+
+ +
+@article{Laguna2016:editornote,
+  author = { Manuel Laguna },
+  title = {Editor's Note on the {MIC} 2013 Special Issue of the
+                  Journal of Heuristics (Volume 22, Issue 4, August 2016)},
+  journal = {Journal of Heuristics},
+  year = 2016,
+  volume = 22,
+  number = 5,
+  pages = {665--666}
+}
+
+ +
+@article{LaiHao2016,
+  author = {Xiangjing Lai and  Jin-Kao Hao },
+  title = {Iterated Maxima Search for the Maximally Diverse Grouping Problem},
+  journal = {European Journal of Operational Research},
+  year = 2016,
+  volume = 254,
+  number = 3,
+  pages = {780--800}
+}
+
+ +
+@article{LanDoi1960,
+  author = {A. H. Land and A. G. Doig},
+  journal = {Econometrica},
+  number = 3,
+  pages = {497--520},
+  publisher = {Wiley, Econometric Society},
+  title = {An Automatic Method of Solving Discrete Programming Problems},
+  volume = 28,
+  year = 1960
+}
+
+ +
+@article{LanHar2015,
+  author = { Langdon, William B.  and Mark Harman},
+  title = {Optimising Software with Genetic Programming},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  volume = 19,
+  number = 1,
+  pages = {118--135}
+}
+
+ +
+@article{LanKotMar2014jscs,
+  author = {M. Lang and H. Kotthaus and P. Marwedel and C. Weihs and
+                  J. Rahnenf{\"u}hrer and  Bernd Bischl },
+  title = {Automatic Model Selection for High-Dimensional Survival
+                  Analysis},
+  journal = {Journal of Statistical Computation and Simulation},
+  year = 2014,
+  volume = 85,
+  number = 1,
+  pages = {62--76},
+  doi = {10.1080/00949655.2014.929131}
+}
+
+ +
+@article{LanSouDes1990,
+  title = {Classification of travelling salesman problem formulations},
+  author = { A. Langevin  and  F. Soumis  and  J. Desrosiers },
+  journal = {Operations Research Letters},
+  volume = 9,
+  number = 2,
+  pages = {127--132},
+  year = 1990,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Langevin93tsptw,
+  author = { A. Langevin  and  M. Desrochers  and  J. Desrosiers  and  Sylvie G{\'e}linas  and  F. Soumis },
+  title = {A Two-Commodity Flow Formulation for the Traveling Salesman
+                  and Makespan Problems with Time Windows},
+  journal = {Networks},
+  year = 1993,
+  volume = 23,
+  number = 7,
+  pages = {631--640}
+}
+
+ +
+@article{Lansey94,
+  author = { Kevin E. Lansey  and  K. Awumah },
+  title = {Optimal Pump Operations Considering Pump Switches},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 120,
+  number = 1,
+  pages = {17--35},
+  date = {1994-01/1994-02},
+  year = 1994,
+  month = jan # { / } # feb
+}
+
+ +
+@article{Laporte2009,
+  author = { Gilbert Laporte },
+  title = {Fifty Years of Vehicle Routing},
+  journal = {Transportation Science},
+  year = 2009,
+  volume = 43,
+  number = 4,
+  pages = {408--416}
+}
+
+ +
+@article{Lau2007arxiv,
+  title = {Stochastic convergence of random search to fixed size
+                  {Pareto} set approximations},
+  author = { Marco Laumanns },
+  journal = {Arxiv preprint arXiv:0711.2949},
+  year = 2007,
+  alias = {laumanns2007stochastic},
+  url = {https://arxiv.org/abs/0711.2949}
+}
+
+ +
+@article{LauHao2009:cie,
+  author = {Beno{\^i}t Laurent and  Jin-Kao Hao },
+  title = {Iterated Local Search for the Multiple Depot Vehicle Scheduling Problem},
+  journal = {Computers and Industrial Engineering},
+  year = 2009,
+  volume = 57,
+  number = 1,
+  pages = {277--286}
+}
+
+ +
+@article{LauThiZit2004,
+  title = {Running time analysis of multiobjective evolutionary
+                  algorithms on pseudo-boolean functions},
+  author = { Marco Laumanns  and  Lothar Thiele  and  Eckart Zitzler },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 8,
+  number = 2,
+  pages = {170--182},
+  year = 2004
+}
+
+ +
+@article{LauThiZit2004nc,
+  author = { Marco Laumanns  and  Lothar Thiele  and  Eckart Zitzler },
+  title = {Running time analysis of evolutionary algorithms on a
+                  simplified multiobjective knapsack problem},
+  volume = 3,
+  journal = {Natural Computing},
+  number = 1,
+  pages = {37--51},
+  year = 2004
+}
+
+ +
+@article{LauZen2011ejor,
+  author = { Marco Laumanns  and Zenklusen, Rico},
+  title = {Stochastic convergence of random search methods to
+                  fixed size {Pareto} front approximations},
+  journal = {European Journal of Operational Research},
+  volume = 213,
+  number = 2,
+  pages = {414--421},
+  year = 2011,
+  doi = {10.1016/j.ejor.2011.03.039}
+}
+
+ +
+@article{LawWoo1966,
+  author = {Lawler, E. L. and Wood, D. E.},
+  title = {Branch-and-Bound Methods: A Survey},
+  journal = {Operations Research},
+  volume = 14,
+  number = 4,
+  year = 1966,
+  pages = {699--719},
+  numpages = 21,
+  doi = {10.1287/opre.14.4.699},
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{Lazic2004,
+  author = {S. E. Lazic},
+  title = {The problem of pseudoreplication in neuroscientific studies:
+                  is it affecting your analysis?},
+  journal = {{BMC} Neuroscience},
+  volume = 11,
+  number = 5,
+  pages = {397--407},
+  year = 2004,
+  doi = {10.1186/1471-2202-11-5}
+}
+
+ +
+@article{LeCBen1995convnet,
+  title = {Convolutional networks for images, speech, and time series},
+  author = {LeCun, Yann and  Bengio, Yoshua  and others},
+  journal = {The handbook of brain theory and neural networks},
+  volume = 3361,
+  number = 10,
+  pages = {255--258},
+  year = 1995
+}
+
+ +
+@article{LeCBenHin2015nature,
+  title = {Deep learning},
+  author = {LeCun, Yann and  Bengio, Yoshua  and Hinton, Geoffrey},
+  journal = {Nature},
+  volume = 521,
+  number = 7553,
+  pages = {436--444},
+  year = 2015,
+  publisher = {Nature Research}
+}
+
+ +
+@article{LeaTavMouMac2016:itor,
+  author = {Vin{\'i}cius {Leal do Forte} and Fl{\'a}vio Marcelo {Tavares
+                  Montenegro} and Jos{\'e} Andr{\'e} {de Moura Brito} and
+                  Nelson Maculan},
+  title = {Iterated Local Search Algorithms for the {Euclidean}
+                  {Steiner} Tree Problem in \emph{n} Dimensions},
+  journal = {International Transactions in Operational Research},
+  year = 2016,
+  volume = 23,
+  number = 6,
+  pages = {1185--1199}
+}
+
+ +
+@article{LehWit2012,
+  title = {Black-box search by unbiased variation},
+  author = {Lehre, Per Kristian and  Carsten Witt },
+  journal = {Algorithmica},
+  volume = 64,
+  number = 4,
+  pages = {623--642},
+  year = 2012,
+  publisher = {Springer}
+}
+
+ +
+@article{Leighton1979,
+  author = {Frank Thomson Leighton},
+  title = {A Graph Coloring Algorithm for Large Scheduling Problems},
+  journal = {Journal of Research of the National Bureau of Standards},
+  year = 1979,
+  volume = 84,
+  number = 6,
+  pages = {489--506}
+}
+
+ +
+@article{LemGroPop2006general,
+  title = {A general analytic method for generating robust strategies
+                  and narrative scenarios},
+  author = {Lempert, Robert J. and Groves, David G. and Popper, Steven
+                  W. and  Bankes, Steven C. },
+  journal = {Management Science},
+  volume = 52,
+  number = 4,
+  pages = {514--528},
+  year = 2006,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{Leon00,
+  author = {C. Leon and S. Martin and J. M. Elena and J. Luque},
+  title = {{EXPLORE}: Hybrid expert system for water networks
+                  management},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2000,
+  volume = 126,
+  number = 2,
+  pages = {65--74}
+}
+
+ +
+@article{Levin1973,
+  title = {Universal'nyie perebornyie zadachi},
+  author = {Levin, Leonid},
+  journal = {Problemy Peredachi Informatsii},
+  volume = 9,
+  pages = {265--266},
+  year = 1973
+}
+
+ +
+@article{LewKurJoe2009gen,
+  author = {Lewandowski, Daniel and Kurowicka, Dorota and Joe, Harry},
+  title = {Generating Random Correlation Matrices Based on Vines and
+                  Extended Onion Method},
+  journal = {Journal of Multivariate Analysis},
+  year = 2009,
+  volume = 100,
+  number = 9,
+  pages = {1989--2001},
+  doi = {10.1016/j.jmva.2009.04.008},
+  abstract = {We extend and improve two existing methods of generating
+                  random correlation matrices, the onion method of Ghosh and
+                  Henderson [S. Ghosh, S.G. Henderson, Behavior of the norta
+                  method for correlated random vector generation as the
+                  dimension increases, ACM Transactions on Modeling and
+                  Computer Simulation (TOMACS) 13 (3) (2003) 276-294] and the
+                  recently proposed method of Joe [H. Joe, Generating random
+                  correlation matrices based on partial correlations, Journal
+                  of Multivariate Analysis 97 (2006) 2177-2189] based on
+                  partial correlations. The latter is based on the so-called
+                  D-vine. We extend the methodology to any regular vine and
+                  study the relationship between the multiple correlation and
+                  partial correlations on a regular vine. We explain the onion
+                  method in terms of elliptical distributions and extend it to
+                  allow generating random correlation matrices from the same
+                  joint distribution as the vine method. The methods are
+                  compared in terms of time necessary to generate 5000 random
+                  correlation matrices of given dimensions.},
+  keywords = {Correlation matrix; Dependence vines; Onion method; Partial
+                  correlation; LKJ}
+}
+
+ +
+@article{Li2008two,
+  title = {A two-step rejection procedure for testing multiple
+                  hypotheses},
+  author = {Li, Jianjun David},
+  journal = {Journal of Statistical Planning and Inference},
+  volume = 138,
+  number = 6,
+  pages = {1521--1527},
+  year = 2008
+}
+
+ +
+@article{Li2021telo,
+  title = {Is Our Archiving Reliable? Multiobjective Archiving Methods
+                  on ``Simple'' Artificial Input Sequences},
+  author = { Li, Miqing },
+  journal = {ACM Transactions on Evolutionary Learning and Optimization},
+  year = 2021,
+  number = 3,
+  pages = {1--19},
+  volume = 1,
+  doi = {10.1145/3465335}
+}
+
+ +
+@article{LiCheFuYao2018twoarch,
+  title = {Two-archive evolutionary algorithm for constrained
+                  multiobjective optimization},
+  author = {Li, Ke and Chen, Renzhi and Fu, Guangtao and  Xin Yao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  number = 2,
+  pages = {303--315},
+  volume = 23,
+  publisher = {IEEE}
+}
+
+ +
+@article{LiCheYao2020ieeese,
+  author = { Li, Miqing  and Chen, Tao and  Xin Yao },
+  title = {How to evaluate solutions in {Pareto}-based search-based
+                  software engineering? {A} critical review and methodological
+                  guidance},
+  journal = {IEEE Transactions on Software Engineering},
+  year = 2020,
+  volume = 48,
+  number = 5,
+  pages = {1771--1799},
+  doi = {10.1109/TSE.2020.3036108}
+}
+
+ +
+@article{LiGroYanLiu2018multi,
+  title = {Multi-line distance minimization: A visualized many-objective
+                  test problem suite},
+  author = { Li, Miqing  and Grosan, Crina and Yang, Shengxiang and
+                  Liu, Xiaohui and  Xin Yao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  number = 1,
+  pages = {61--78},
+  volume = 22,
+  annote = {highly degenerate Pareto fronts}
+}
+
+ +
+@article{LiShaBah2016traffic,
+  author = {Li, Zhiyi and Shahidehpour, Mohammad and Bahramirad, Shay and
+                  Khodaei, Amin},
+  doi = {10.1109/TSG.2016.2526032},
+  alias = {Li2016},
+  journal = {IEEE Transactions on Smart Grid},
+  number = 4,
+  pages = {1--1},
+  title = {Optimizing Traffic Signal Settings in Smart Cities},
+  volume = 3053,
+  year = 2016,
+  abstract = {Traffic signals play a critical role in smart cities for
+                  mitigating traffic congestions and reducing the emission in
+                  metropolitan areas. This paper proposes a bi-level
+                  optimization framework to settle the optimal traffic signal
+                  setting problem. The upper-level problem determines the
+                  traffic signal settings to minimize the drivers' average
+                  travel time, while the lower-level problem aims for achieving
+                  the network equilibrium using the settings calculated at the
+                  upper level. Genetic algorithm is employed with the
+                  integration of microscopic-traffic-simulation based dynamic
+                  traffic assignment (DTA) to decouple the complex bi-level
+                  problem into tractable single-level problems which are solved
+                  sequentially. Case studies on a synthetic traffic network and
+                  a real-world traffic subnetwork are conducted to examine the
+                  effectiveness of the proposed model and relevant solution
+                  methods. Additional strategies are provided for the extension
+                  of the proposed model and the acceleration solution process
+                  in large-area traffic network applications.}
+}
+
+ +
+@article{LiChenXuGupta2015,
+  author = {Xiaoping Li and Long Chen and Haiyan Xu and Jatinder N.D. Gupta},
+  title = {Trajectory Scheduling Methods for Minimizing Total Tardiness in a Flowshop},
+  journal = {Operations Research Perspectives},
+  volume = 2,
+  pages = {13--23},
+  year = 2015,
+  issn = {2214--7160},
+  doi = {10.1016/j.orp.2014.12.001}
+}
+
+ +
+@article{LiJamSal2018hyperband,
+  author = {Lisha Li and Kevin Jamieson and Giulia DeSalvo and Afshin
+                  Rostamizadeh and Ameet Talwalkar},
+  title = {Hyperband: A Novel Bandit-Based Approach to Hyperparameter
+                  Optimization},
+  journal = {Journal of Machine Learning Research},
+  year = 2018,
+  volume = 18,
+  number = 185,
+  pages = {1--52},
+  epub = {http://jmlr.org/papers/v18/16-558.html},
+  abstract = {Performance of machine learning algorithms depends critically on identifying a good set of hyperparameters. While recent approaches use Bayesian optimization to adaptively select configurations, we focus on speeding up random search through adaptive resource allocation and early-stopping. We formulate hyperparameter optimization as a pure-exploration non-stochastic infinite-armed bandit problem where a predefined resource like iterations, data samples, or features is allocated to randomly sampled configurations. We introduce a novel algorithm, our algorithm, for this framework and analyze its theoretical properties, providing several desirable guarantees. Furthermore, we compare our algorithm with popular Bayesian optimization methods on a suite of hyperparameter optimization problems. We observe that our algorithm can provide over an order-of-magnitude speedup over our competitor set on a variety of deep-learning and kernel-based learning problems.},
+  keywords = {racing}
+}
+
+ +
+@article{LiLi07,
+  author = {Y. Li and W. Li},
+  title = {Adaptive Ant Colony Optimization Algorithm Based on
+                  Information Entropy: Foundation and Application},
+  journal = {Fundamenta Informaticae},
+  volume = 77,
+  number = 3,
+  year = 2007,
+  pages = {229--242},
+  publisher = {IOS Press},
+  address = {Amsterdam, The Netherlands}
+}
+
+ +
+@article{LiLiTanYao2015many,
+  author = {Li, Bingdong and Li, Jinlong and Tang, Ke and  Xin Yao },
+  title = {Many-Objective Evolutionary Algorithms: A Survey},
+  journal = {{ACM} Computing Surveys},
+  volume = 48,
+  number = 1,
+  year = 2015,
+  pages = {1--35},
+  doi = {10.1145/2792984},
+  numpages = 35
+}
+
+ +
+@article{LiLopYao2023archiving,
+  author = { Li, Miqing  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Xin Yao },
+  title = {Multi-Objective Archiving},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2023,
+  doi = {10.1109/TEVC.2023.3314152},
+  abstract = {Most multi-objective optimisation algorithms maintain an
+                  archive explicitly or implicitly during their search. Such an
+                  archive can be solely used to store high-quality solutions
+                  presented to the decision maker, but in many cases may
+                  participate in the search process (e.g., as the population in
+                  evolutionary computation). Over the last two decades,
+                  archiving, the process of comparing new solutions with
+                  previous ones and deciding how to update the
+                  archive/population, stands as an important issue in
+                  evolutionary multi-objective optimisation (EMO). This is
+                  evidenced by constant efforts from the community on
+                  developing various effective archiving methods, ranging from
+                  conventional Pareto-based methods to more recent
+                  indicator-based and decomposition-based ones. However, the
+                  focus of these efforts is on empirical performance comparison
+                  in terms of specific quality indicators; there is lack of
+                  systematic study of archiving methods from a general
+                  theoretical perspective. In this paper, we attempt to conduct
+                  a systematic overview of multi-objective archiving, in the
+                  hope of paving the way to understand archiving algorithms
+                  from a holistic perspective of theory and practice, and more
+                  importantly providing a guidance on how to design
+                  theoretically desirable and practically useful archiving
+                  algorithms. In doing so, we also present that archiving
+                  algorithms based on weakly Pareto compliant indicators (e.g.,
+                  $\epsilon$-indicator), as long as designed properly, can
+                  achieve the same theoretical desirables as archivers based on
+                  Pareto compliant indicators (e.g., hypervolume
+                  indicator). Such desirables include the property
+                  limit-optimal, the limit form of the possible optimal
+                  property that a bounded archiving algorithm can have with
+                  respect to the most general form of superiority between
+                  solution sets.}
+}
+
+ +
+@article{LiTanLiYao2016stochastic,
+  title = {Stochastic ranking algorithm for many-objective optimization
+                  based on multiple indicators},
+  author = {Li, Bingdong and Tang, Ke and Li, Jinlong and  Xin Yao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2016,
+  number = 6,
+  pages = {924--938},
+  volume = 20,
+  publisher = {IEEE}
+}
+
+ +
+@article{LiYanLiu2014shift,
+  title = {Shift-based density estimation for {Pareto}-based algorithms
+                  in many-objective optimization},
+  author = { Li, Miqing  and Yang, Shengxiang and Liu, Xiaohui},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2014,
+  number = 3,
+  pages = {348--365},
+  volume = 18,
+  publisher = {IEEE},
+  annote = {Proposed SDE indicator algorithm}
+}
+
+ +
+@article{LiYanLiu2016tec,
+  title = {{Pareto} or non-{Pareto}: {Bi}-criterion evolution in
+                  multiobjective optimization},
+  author = { Li, Miqing  and Yang, Shengxiang and Liu, Xiaohui},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2016,
+  number = 5,
+  pages = {645--665},
+  volume = 20
+}
+
+ +
+@article{LiYao2019qual,
+  title = {Quality Evaluation of Solution Sets in Multiobjective
+                  Optimisation: A Survey},
+  author = { Li, Miqing  and  Xin Yao },
+  journal = {{ACM} Computing Surveys},
+  year = 2019,
+  number = 2,
+  volume = 52,
+  pages = {1--38},
+  doi = {10.1145/3300148},
+  publisher = {ACM}
+}
+
+ +
+@article{LiYao2017arxiv,
+  title = {Dominance Move: A Measure of Comparing Solution Sets in
+                  Multiobjective Optimization},
+  author = { Li, Miqing  and  Xin Yao },
+  journal = {arXiv preprint arXiv:1702.00477},
+  year = 2017
+}
+
+ +
+@article{LiYao2020ec,
+  title = {What weights work for you? Adapting weights for any {Pareto}
+                  front shape in decomposition-based evolutionary
+                  multiobjective optimisation},
+  author = { Li, Miqing  and  Xin Yao },
+  journal = {Evolutionary Computation},
+  year = 2020,
+  number = 2,
+  pages = {227--253},
+  volume = 28
+}
+
+ +
+@article{LiZha2009:moead-de,
+  title = {Multiobjective Optimization Problems with Complicated
+                  {Pareto} sets, {MOEA/D} and {NSGA-II}},
+  author = {Li, Hui and  Zhang, Qingfu },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 13,
+  number = 2,
+  pages = {284--302},
+  year = 2009
+}
+
+ +
+@article{LiZouYan2021twoarch,
+  title = {A two-archive algorithm with decomposition and fitness
+                  allocation for multi-modal multi-objective optimization},
+  author = {Li, Zhipan and Zou, Juan and Yang, Shengxiang and Zheng,
+                  Jinhua},
+  journal = {Information Sciences},
+  year = 2021,
+  pages = {413--430},
+  volume = 574,
+  publisher = {Elsevier}
+}
+
+ +
+@article{LiaAydStu13,
+  author = {Liao, Tianjun  and  Do\v{g}an Ayd{\i}n  and  Thomas St{\"u}tzle },
+  title = {Artificial Bee Colonies for Continuous Optimization: Experimental Analysis and Improvements},
+  journal = {Swarm Intelligence},
+  year = 2013,
+  volume = 7,
+  number = 4,
+  pages = {327--356}
+}
+
+ +
+@article{LiaMolMonStu2014,
+  author = {Liao, Tianjun  and  Daniel Molina  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle },
+  title = {A Note on the Effects of Enforcing Bound Constraints on
+Algorithm Comparisons using the {IEEE} {CEC'05} Benchmark Function Suite},
+  journal = {Evolutionary Computation},
+  year = 2014,
+  volume = 22,
+  number = 2,
+  pages = {351--359}
+}
+
+ +
+@article{LiaMolStu2015,
+  author = {Liao, Tianjun  and  Daniel Molina  and  Thomas St{\"u}tzle },
+  title = {Performance Evaluation of Automatically Tuned Continuous
+  Optimizers on Different Benchmark Sets},
+  journal = {Applied Soft Computing},
+  year = 2015,
+  volume = 27,
+  pages = {490--503}
+}
+
+ +
+@article{LiaMonStu13:soco,
+  author = {Liao, Tianjun  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle },
+  title = {Computational results for an automatically tuned {CMA-ES}
+                  with increasing population size on the {CEC'05} benchmark
+                  set},
+  journal = {Soft Computing},
+  pages = {1031--1046},
+  volume = 17,
+  number = 6,
+  year = 2013,
+  doi = {0.1007/s00500-012-0946-x}
+}
+
+ +
+@article{LiaSocMonStuDor2014,
+  author = {Liao, Tianjun  and  Krzysztof Socha  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Ant Colony Optimization for Mixed-Variable Optimization
+Problems},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2014,
+  volume = 18,
+  number = 4,
+  pages = {503--518},
+  keywords = {ACOR}
+}
+
+ +
+@article{LiaStuMonDor2014,
+  author = {Liao, Tianjun  and  Thomas St{\"u}tzle  and  Marco A. {Montes de Oca}  and  Marco Dorigo },
+  title = {A Unified Ant Colony Optimization Algorithm for Continuous
+Optimization},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 234,
+  number = 3,
+  pages = {597--609}
+}
+
+ +
+@article{LiaTseLua07,
+  author = { C.-J. Liao  and  C.-T. Tseng  and  P. Luarn },
+  title = {A Discrete Version of Particle Swarm Optimization
+                  for Flowshop Scheduling Problems},
+  journal = {Computers \& Operations Research},
+  volume = 34,
+  number = 10,
+  pages = {3099--3111},
+  year = 2007
+}
+
+ +
+@article{LieDaoVerDer2019tec,
+  author = { Arnaud Liefooghe  and Fabio Daolio and  Bilel Derbel  and  Verel, S{\'e}bastien  and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  number = 6,
+  pages = {1063--1077},
+  title = {Landscape-Aware Performance Prediction for Evolutionary
+                  Multi-objective Optimization},
+  volume = 24,
+  year = 2020
+}
+
+ +
+@article{LieHumMes2011,
+  author = { Arnaud Liefooghe  and  J{\'e}r{\'e}mie Humeau  and  Salma Mesmoudi  and  Laetitia Jourdan  and  Talbi, El-Ghazali },
+  title = {On dominance-based multiobjective local search: design,
+                  implementation and experimental analysis on scheduling and
+                  traveling salesman problems},
+  journal = {Journal of Heuristics},
+  volume = 18,
+  number = 2,
+  pages = {317--352},
+  year = 2012,
+  abstract = {This paper discusses simple local search approaches for
+                  approximating the efficient set of multiobjective
+                  combinatorial optimization problems. We focus on algorithms
+                  defined by a neighborhood structure and a dominance relation
+                  that iteratively improve an archive of nondominated
+                  solutions. Such methods are referred to as dominance-based
+                  multiobjective local search. We first provide a concise
+                  overview of existing algorithms, and we propose a model
+                  trying to unify them through a fine-grained
+                  decomposition. The main problem-independent search components
+                  of dominance relation, solution selection, neighborhood
+                  exploration and archiving are largely discussed. Then, a
+                  number of state-of-the-art and original strategies are
+                  experimented on solving a permutation flowshop scheduling
+                  problem and a traveling salesman problem, both on a two- and
+                  a three-objective formulation. Experimental results and a
+                  statistical comparison are reported in the paper, and some
+                  directions for future research are highlighted.},
+  doi = {10.1007/s10732-011-9181-3}
+}
+
+ +
+@article{LieJouTal2011paradiseo,
+  title = {A Software Framework Based on a Conceptual Unified
+                  Model for Evolutionary Multiobjective Optimization:
+                  {ParadisEO}-{MOEO}},
+  author = { Arnaud Liefooghe  and  Laetitia Jourdan  and  Talbi, El-Ghazali },
+  journal = {European Journal of Operational Research},
+  volume = 209,
+  number = 2,
+  pages = {104--112},
+  year = 2011
+}
+
+ +
+@article{LieVerHao2014hybrid,
+  author = { Arnaud Liefooghe  and  Verel, S{\'e}bastien  and  Jin-Kao Hao },
+  title = {A hybrid metaheuristic for multiobjective unconstrained
+                  binary quadratic programming},
+  journal = {Applied Soft Computing},
+  year = 2014,
+  volume = 16,
+  pages = {10--19},
+  publisher = {Elsevier}
+}
+
+ +
+@article{LikKoc2007predictive,
+  title = {Predictive control of a gas--liquid separation plant based on
+                  a {Gaussian} process model},
+  author = {Likar, Bojan and Kocijan, Ju{\v{s}}},
+  journal = {Computers \& Chemical Engineering},
+  volume = 31,
+  number = 3,
+  pages = {142--152},
+  year = 2007,
+  publisher = {Elsevier},
+  doi = {10.1016/j.compchemeng.2006.05.011}
+}
+
+ +
+@article{LinEggFeu2022smac3,
+  author = { Marius Thomas Lindauer  and  Katharina Eggensperger  and  Matthias Feurer  and  Biedenkapp, Andr{\'e}  and Difan Deng and Carolin Benjamins and Tim
+                  Ruhkopf and René Sass and  Frank Hutter },
+  title = {{SMAC3}: A Versatile Bayesian Optimization Package for
+                  Hyperparameter Optimization},
+  journal = {Journal of Machine Learning Research},
+  year = 2022,
+  volume = 23,
+  pages = {1--9},
+  epub = {http://jmlr.org/papers/v23/21-0888.html}
+}
+
+ +
+@article{LinHooHutSch2015autofolio,
+  title = {{AutoFolio}: An Automatically Configured Algorithm Selector},
+  author = { Marius Thomas Lindauer  and  Holger H. Hoos  and  Frank Hutter  and Schaub, Torsten},
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 53,
+  pages = {745--778},
+  year = 2015
+}
+
+ +
+@article{LinKer73,
+  author = {S. Lin and B. W. Kernighan},
+  title = {An Effective Heuristic Algorithm for the Traveling
+                  Salesman Problem},
+  journal = {Operations Research},
+  year = 1973,
+  volume = 21,
+  number = 2,
+  pages = {498--516}
+}
+
+ +
+@article{LinVanKot2019,
+  title = {The algorithm selection competitions 2015 and 2017},
+  author = { Marius Thomas Lindauer  and  van Rijn, Jan N.  and Kotthoff, Lars},
+  journal = {Artificial Intelligence},
+  volume = 272,
+  pages = {86--100},
+  year = 2019
+}
+
+ +
+@article{LisWit2015tcs,
+  title = {Runtime Analysis of Ant Colony Optimization on Dynamic
+                  Shortest Path Problems},
+  journal = {Theoretical Computer Science},
+  volume = 561,
+  number = {Part A},
+  pages = {73--85},
+  year = 2015,
+  doi = {10.1016/j.tcs.2014.06.035},
+  author = { Andrei Lissovoi  and  Carsten Witt },
+  abstract = {A simple ACO algorithm called $\lambda$-MMAS for dynamic
+                  variants of the single-destination shortest paths problem is
+                  studied by rigorous runtime analyses. Building upon previous
+                  results for the special case of 1-MMAS, it is studied to what
+                  extent an enlarged colony using $\lambda$ ants per vertex
+                  helps in tracking an oscillating optimum. It is shown that
+                  easy cases of oscillations can be tracked by a constant
+                  number of ants. However, the paper also identifies more
+                  involved oscillations that with overwhelming probability
+                  cannot be tracked with any polynomial-size colony. Finally,
+                  parameters of dynamic shortest-path problems which make the
+                  optimum difficult to track are discussed. Experiments
+                  illustrate theoretical findings and conjectures. }
+}
+
+ +
+@article{LitMurSweKar63,
+  author = {J. D. C. Little and K. G. Murty and D. W. Sweeney and C. Karel},
+  title = {An Algorithm for the Traveling Salesman Problem},
+  journal = {Operations Research},
+  year = 1963,
+  volume = 11,
+  pages = {972--989}
+}
+
+ +
+@article{LiuMalWanBre2017vis,
+  author = { Shusen Liu  and  Dan Maljovec  and  Bei Wang  and 
+                  Peer-Timo Bremer  and  Valerio Pascucci},
+  title = {Visualizing High-Dimensional Data: Advances in the Past
+                  Decade},
+  doi = {10.1109/TVCG.2016.2640960},
+  year = 2017,
+  journal = { IEEE Transactions on Visualization and Computer Graphics },
+  volume = 23,
+  number = 3
+}
+
+ +
+@article{LiuRee2001,
+  author = {Jiyin Liu and  Colin R. Reeves },
+  title = {Constructive and Composite Heuristic Solutions to the
+                  {P//$\Sigma$Ci} Scheduling Problem},
+  journal = {European Journal of Operational Research},
+  volume = 132,
+  number = 2,
+  pages = {439--452},
+  year = 2001,
+  doi = {10.1016/S0377-2217(00)00137-5}
+}
+
+ +
+@article{LiuYenGon2018twoarch,
+  title = {A multimodal multiobjective evolutionary algorithm using
+                  two-archive and recombination strategies},
+  author = {Liu, Yiping and Yen, Gary G. and Gong, Dunwei},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  number = 4,
+  pages = {660--674},
+  volume = 23
+}
+
+ +
+@article{LocSch1999mlsl,
+  author = {Locatelli, Marco and Schoen, Fabio},
+  title = {Random Linkage: a family of acceptance/rejection algorithms
+                  for global optimisation.},
+  journal = {Mathematical Programming},
+  year = 1999,
+  volume = 85,
+  number = 2,
+  keywords = {Multi-Level Single-Linkage (MLSL)}
+}
+
+ +
+@article{LodMarMon2002,
+  title = {Two-dimensional packing problems: A survey},
+  author = { Andrea Lodi  and  Silvano Martello  and  Monaci, Michele },
+  journal = {European Journal of Operational Research},
+  volume = 141,
+  number = 2,
+  pages = {241--252},
+  year = 2002,
+  doi = {10.1016/S0377-2217(02)00123-6}
+}
+
+ +
+@article{LodMarVig1999binpack,
+  title = {Heuristic and metaheuristic approaches for a class of
+                  two-dimensional bin packing problems},
+  author = { Andrea Lodi  and  Silvano Martello  and  Vigo, Daniele },
+  journal = {INFORMS Journal on Computing},
+  volume = 11,
+  number = 4,
+  pages = {345--357},
+  year = 1999,
+  publisher = {{INFORMS}},
+  doi = {10.1287/ijoc.11.4.345}
+}
+
+ +
+@article{LodMarVig2004tspack,
+  title = {{TSpack}: a unified tabu search code for multi-dimensional bin
+                  packing problems},
+  author = { Andrea Lodi  and  Silvano Martello  and  Vigo, Daniele },
+  journal = {Annals of Operations Research},
+  volume = 131,
+  number = {1-4},
+  pages = {203--213},
+  year = 2004,
+  publisher = {Springer},
+  doi = {10.1023/B:ANOR.0000039519.03572.08}
+}
+
+ +
+@article{LodZar2017learning,
+  title = {On Learning and Branching: A Survey},
+  author = { Andrea Lodi  and Zarpellon, Giulia},
+  journal = {TOP},
+  volume = 25,
+  pages = {207--236},
+  year = 2017,
+  publisher = {Springer}
+}
+
+ +
+@article{LohHorLin2008antennas,
+  author = {Lohn, Jason D. and Hornby, Gregory S. and Linden, Derek S.},
+  title = {Human-competitive Evolved Antennas},
+  journal = {Artificial Intelligence for Engineering Design, Analysis and Manufacturing},
+  volume = 22,
+  number = 3,
+  year = 2008,
+  pages = {235--247},
+  doi = {10.1017/s0890060408000164},
+  publisher = {Cambridge University Press},
+  annote = {Evolutionary optimization of antennas for NASA}
+}
+
+ +
+@article{LopBlu2010cor,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Christian Blum },
+  title = {Beam-{ACO} for the travelling salesman problem with
+                  time windows},
+  journal = {Computers \& Operations Research},
+  year = 2010,
+  doi = {10.1016/j.cor.2009.11.015},
+  volume = 37,
+  number = 9,
+  pages = {1570--1583},
+  keywords = {Ant colony optimization, Travelling salesman problem with
+                  time windows, Hybridization},
+  alias = {LopBlu09tsptw},
+  abstract = {The travelling salesman problem with time windows is
+                  a difficult optimization problem that arises, for
+                  example, in logistics. This paper deals with the
+                  minimization of the travel-cost. For solving this
+                  problem, this paper proposes a Beam-ACO algorithm,
+                  which is a hybrid method combining ant colony
+                  optimization with beam search.  In general, Beam-ACO
+                  algorithms heavily rely on accurate and
+                  computationally inexpensive bounding information for
+                  differentiating between partial solutions. This work
+                  uses stochastic sampling as a useful alternative. An
+                  extensive experimental evaluation on seven benchmark
+                  sets from the literature shows that the proposed
+                  Beam-ACO algorithm is currently a state-of-the-art
+                  technique for the travelling salesman problem with
+                  time windows when travel-cost optimization is
+                  concerned.}
+}
+
+ +
+@article{LopBlu2013asoc,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Christian Blum  and  Jeffrey W. Ohlmann  and  Barrett W. Thomas },
+  title = {The Travelling Salesman Problem with Time Windows:
+                  Adapting Algorithms from Travel-time to Makespan
+                  Optimization},
+  journal = {Applied Soft Computing},
+  year = 2013,
+  volume = 13,
+  number = 9,
+  pages = {3806--3815},
+  doi = {10.1016/j.asoc.2013.05.009},
+  epub = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2013-011.pdf}
+}
+
+ +
+@article{LopBraPaq2021arxiv,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\"u}rgen Branke  and  Lu{\'i}s Paquete },
+  title = {Reproducibility in Evolutionary Computation},
+  journal = {Arxiv preprint arXiv:20102.03380 [cs.AI]},
+  year = 2021,
+  url = {https://arxiv.org/abs/2102.03380},
+  abstract = {Experimental studies are prevalent in Evolutionary
+                  Computation (EC), and concerns about the reproducibility and
+                  replicability of such studies have increased in recent times,
+                  reflecting similar concerns in other scientific fields. In
+                  this article, we suggest a classification of different types
+                  of reproducibility that refines the badge system of the
+                  Association of Computing Machinery (ACM) adopted by TELO. We
+                  discuss, within the context of EC, the different types of
+                  reproducibility as well as the concepts of artifact and
+                  measurement, which are crucial for claiming
+                  reproducibility. We identify cultural and technical obstacles
+                  to reproducibility in the EC field. Finally, we provide
+                  guidelines and suggest tools that may help to overcome some
+                  of these reproducibility obstacles.},
+  keywords = {Evolutionary Computation, Reproducibility, Empirical study,
+                  Benchmarking}
+}
+
+ +
+@article{LopBraPaq2021telo,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\"u}rgen Branke  and  Lu{\'i}s Paquete },
+  title = {Reproducibility in Evolutionary Computation},
+  journal = {ACM Transactions on Evolutionary Learning and Optimization},
+  year = 2021,
+  volume = 1,
+  number = 4,
+  pages = {1--21},
+  doi = {10.1145/3466624},
+  epub = {https://arxiv.org/abs/2102.03380},
+  abstract = {Experimental studies are prevalent in Evolutionary
+                  Computation (EC), and concerns about the reproducibility and
+                  replicability of such studies have increased in recent times,
+                  reflecting similar concerns in other scientific fields. In
+                  this article, we suggest a classification of different types
+                  of reproducibility that refines the badge system of the
+                  Association of Computing Machinery (ACM) adopted by TELO. We
+                  discuss, within the context of EC, the different types of
+                  reproducibility as well as the concepts of artifact and
+                  measurement, which are crucial for claiming
+                  reproducibility. We identify cultural and technical obstacles
+                  to reproducibility in the EC field. Finally, we provide
+                  guidelines and suggest tools that may help to overcome some
+                  of these reproducibility obstacles.},
+  keywords = {Evolutionary Computation, Reproducibility, Empirical study,
+                  Benchmarking}
+}
+
+ +
+@article{LopDubPerStuBir2016irace,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and   P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {The {\rpackage{irace}} Package: Iterated Racing for Automatic
+                  Algorithm Configuration},
+  journal = {Operations Research Perspectives},
+  year = 2016,
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2016-003/},
+  doi = {10.1016/j.orp.2016.09.002},
+  volume = 3,
+  pages = {43--58}
+}
+
+ +
+@article{LopKesStu2017:cim,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Marie-El{\'e}onore Kessaci  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Metaheuristics from Algorithmic Components},
+  journal = {Submitted},
+  year = 2017,
+  optvolume = {},
+  optnumber = {},
+  optpages = {}
+}
+
+ +
+@article{LopPaqStu05:jmma,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Hybrid Population-based Algorithms for the
+                  Bi-objective Quadratic Assignment Problem},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2006,
+  volume = 5,
+  number = 1,
+  pages = {111--137},
+  doi = {10.1007/s10852-005-9034-x},
+  alias = {LopPaqStu06:jmma},
+  abstract = {We present variants of an ant colony optimization
+                  (MO-ACO) algorithm and of an evolutionary algorithm
+                  (SPEA2) for tackling multi-objective combinatorial
+                  optimization problems, hybridized with an iterative
+                  improvement algorithm and the robust tabu search
+                  algorithm. The performance of the resulting hybrid
+                  stochastic local search (SLS) algorithms is
+                  experimentally investigated for the bi-objective
+                  quadratic assignment problem (bQAP) and compared
+                  against repeated applications of the underlying
+                  local search algorithms for several
+                  scalarizations. The experiments consider structured
+                  and unstructured bQAP instances with various degrees
+                  of correlation between the flow matrices. We do a
+                  systematic experimental analysis of the algorithms
+                  using outperformance relations and the attainment
+                  functions methodology to asses differences in the
+                  performance of the algorithms. The experimental
+                  results show the usefulness of the hybrid algorithms
+                  if the available computation time is not too limited
+                  and identify SPEA2 hybridized with very short tabu
+                  search runs as the most promising variant.}
+}
+
+ +
+@article{LopPerStu2020ifors,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and   P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle },
+  title = {{irace}: A Tool for the Automatic Configuration of
+                  Algorithms},
+  journal = {International Federation of Operational Research Societies
+                  (IFORS) News},
+  year = 2020,
+  volume = 14,
+  number = 2,
+  pages = {30--32},
+  month = jun,
+  url = {https://www.ifors.org/newsletter/ifors-news-june2020.pdf}
+}
+
+ +
+@article{LopPraPae08aco,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Ant Colony Optimisation for the Optimal Control of
+                  Pumps in Water Distribution Networks},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2008,
+  volume = 134,
+  number = 4,
+  pages = {337--346},
+  publisher = {{ASCE}},
+  epub = {http://link.aip.org/link/?QWR/134/337/1},
+  doi = {10.1061/(ASCE)0733-9496(2008)134:4(337)},
+  abstract = {Reducing energy consumption of water distribution
+  networks has never had more significance than today. The greatest
+  energy savings can be obtained by careful scheduling of operation of
+  pumps. Schedules can be defined either implicitly, in terms of other
+  elements of the network such as tank levels, or explicitly by
+  specifying the time during which each pump is on/off. The
+  traditional representation of explicit schedules is a string of
+  binary values with each bit representing pump on/off status during a
+  particular time interval. In this paper a new explicit
+  representation is presented. It is based on time controlled
+  triggers, where the maximum number of pump switches is specified
+  beforehand. In this representation a pump schedule is divided into a
+  series of integers with each integer representing the number of
+  hours for which a pump is active/inactive. This reduces the number
+  of potential schedules (search space) compared to the binary
+  representation. Ant colony optimization (ACO) is a stochastic
+  meta-heuristic for combinatorial optimization problems that is
+  inspired by the foraging behavior of some species of ants. In this
+  paper, an application of the ACO framework was developed for the
+  optimal scheduling of pumps. The proposed representation was adapted
+  to an ant colony Optimization framework and solved for the optimal
+  pump schedules. Minimization of electrical cost was considered as
+  the objective, while satisfying system constraints. Instead of using
+  a penalty function approach for constraint violations, constraint
+  violations were ordered according to their importance and solutions
+  were ranked based on this order. The proposed approach was tested on
+  a small test network and on a large real-world network. Results are
+  compared with those obtained using a simple genetic algorithm based
+  on binary representation and a hybrid genetic algorithm that uses
+  level-based triggers.}
+}
+
+ +
+@article{LopPraPae2011ec,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Representations and Evolutionary Operators for the
+                  Scheduling of Pump Operations in Water Distribution
+                  Networks},
+  journal = {Evolutionary Computation},
+  year = 2011,
+  doi = {10.1162/EVCO_a_00035},
+  volume = 19,
+  number = 3,
+  pages = {429--467},
+  abstract = {Reducing the energy consumption of water
+                  distribution networks has never had more
+                  significance. The greatest energy savings can be
+                  obtained by carefully scheduling the operations of
+                  pumps. Schedules can be defined either implicitly,
+                  in terms of other elements of the network such as
+                  tank levels, or explicitly by specifying the time
+                  during which each pump is on/off.  The traditional
+                  representation of explicit schedules is a string of
+                  binary values with each bit representing pump on/off
+                  status during a particular time interval.  In this
+                  paper, we formally define and analyze two new
+                  explicit representations based on time-controlled
+                  triggers, where the maximum number of pump switches
+                  is established beforehand and the schedule may
+                  contain less switches than the maximum. In these
+                  representations, a pump schedule is divided into a
+                  series of integers with each integer representing
+                  the number of hours for which a pump is
+                  active/inactive.  This reduces the number of
+                  potential schedules compared to the binary
+                  representation, and allows the algorithm to operate
+                  on the feasible region of the search space.  We
+                  propose evolutionary operators for these two new
+                  representations. The new representations and their
+                  corresponding operations are compared with the two
+                  most-used representations in pump scheduling,
+                  namely, binary representation and level-controlled
+                  triggers. A detailed statistical analysis of the
+                  results indicates which parameters have the greatest
+                  effect on the performance of evolutionary
+                  algorithms. The empirical results show that an
+                  evolutionary algorithm using the proposed
+                  representations improves over the results obtained
+                  by a recent state-of-the-art Hybrid Genetic
+                  Algorithm for pump scheduling using level-controlled
+                  triggers.}
+}
+
+ +
+@article{LopStu2012swarm,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An experimental analysis of design choices of multi-objective ant colony optimization algorithms},
+  journal = {Swarm Intelligence},
+  year = 2012,
+  number = 3,
+  volume = 6,
+  pages = {207--232},
+  doi = {10.1007/s11721-012-0070-7},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2012-006/}
+}
+
+ +
+@article{LopStu2012tec,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {The Automatic Design of Multi-Objective Ant Colony
+                  Optimization Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2012,
+  volume = 16,
+  number = 6,
+  pages = {861--875},
+  doi = {10.1109/TEVC.2011.2182651},
+  abstract = {
+Multi-objective optimization problems are problems with several,
+typically conflicting criteria for evaluating solutions. Without
+any a priori preference information, the Pareto optimality
+principle establishes a partial order among solutions, and the
+output of the algorithm becomes a set of nondominated solutions
+rather than a single one. Various ant colony optimization (ACO)
+algorithms have been proposed in recent years for solving such
+problems. These multi-objective ACO (MOACO) algorithms exhibit
+different design choices for dealing with the particularities of
+the multi-objective context. This paper proposes a formulation of
+algorithmic components that suffices to describe most MOACO
+algorithms proposed so far. This formulation also shows that
+existing MOACO algorithms often share equivalent design choices
+but they are described in different terms. Moreover, this
+formulation is synthesized into a flexible algorithmic framework,
+from which not only existing MOACO algorithms may be
+instantiated, but also combinations of components that were never
+studied in the literature. In this sense, this paper goes beyond
+proposing a new MOACO algorithm, but it rather introduces a
+family of MOACO algorithms. The flexibility of the proposed MOACO
+framework facilitates the application of automatic algorithm
+configuration techniques. The experimental results presented in
+this paper show that the automatically configured MOACO framework
+outperforms the MOACO algorithms that inspired the framework
+itself. This paper is also among the first to apply automatic
+algorithm configuration techniques to multi-objective algorithms.}
+}
+
+ +
+@article{LopStu2013ejor,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Improving the Anytime Behaviour of Optimisation
+                  Algorithms},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 235,
+  number = 3,
+  pages = {569--582},
+  doi = {10.1016/j.ejor.2013.10.043},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2012-011/},
+  abstract = {Optimisation algorithms with good anytime behaviour try to
+                  return as high-quality solutions as possible independently of
+                  the computation time allowed. Designing algorithms with good
+                  anytime behaviour is a difficult task, because performance is
+                  often evaluated subjectively, by plotting the trade-off curve
+                  between computation time and solution quality. Yet, the
+                  trade-off curve may be modelled also as a set of mutually
+                  nondominated, bi-objective points. Using this model, we
+                  propose to combine an automatic configuration tool and the
+                  hypervolume measure, which assigns a single quality measure
+                  to a nondominated set. This allows us to improve the anytime
+                  behaviour of optimisation algorithms by means of
+                  automatically finding algorithmic configurations that produce
+                  the best nondominated sets. Moreover, the recently proposed
+                  weighted hypervolume measure is used here to incorporate the
+                  decision-maker's preferences into the automatic tuning
+                  procedure. We report on the improvements reached when
+                  applying the proposed method to two relevant scenarios: (i)
+                  the design of parameter variation strategies for MAX-MIN Ant
+                  System and (ii) the tuning of the anytime behaviour of SCIP,
+                  an open-source mixed integer programming solver with more
+                  than 200 parameters.}
+}
+
+ +
+@article{LopTerRos2014esa,
+  author = {Eunice López-Camacho and Hugo Terashima-Marin and  Peter Ross  and  Gabriela Ochoa },
+  title = {A unified hyper-heuristic framework for solving bin packing
+                  problems},
+  journal = {Expert Systems with Applications},
+  volume = 41,
+  number = 15,
+  pages = {6876--6889},
+  year = 2014,
+  doi = {10.1016/j.eswa.2014.04.043}
+}
+
+ +
+@article{LouBoi2008vns_anytime,
+  author = { Samir Loudni  and  Patrice Boizumault },
+  title = {Combining {VNS} with constraint programming for
+                  solving anytime optimization problems},
+  journal = {European Journal of Operational Research},
+  year = 2008,
+  volume = 191,
+  pages = {705--735},
+  doi = {10.1016/j.ejor.2006.12.062}
+}
+
+ +
+@article{Lourenco1995,
+  author = { Helena R. {Louren{\c c}o} },
+  title = {Job-Shop Scheduling: Computational Study of Local
+  Search and Large-Step Optimization Methods},
+  journal = {European Journal of Operational Research},
+  year = 1995,
+  volume = 83,
+  number = 2,
+  pages = {347--364}
+}
+
+ +
+@article{LovTor2001aor,
+  author = {Lova, Antonio and Tormos, Pilar},
+  title = {Analysis of Scheduling Schemes and Heuristic Rules
+                  Performance in Resource-Constrained Multiproject
+                  Scheduling},
+  volume = 102,
+  doi = {10.1023/A:1010966401888},
+  abstract = {Frequently, the availability of resources assigned
+                  to a project is limited and not sufficient to
+                  execute all the concurrent activities. In this
+                  situation, decision making about their schedule is
+                  necessary. Many times this schedule supposes an
+                  increase in the project completion
+                  time. Additionally, companies commonly manage
+                  various projects simultaneously, sharing a pool of
+                  renewable resources. Given these resource
+                  constraints, we often can only apply heuristic
+                  methods to solve the scheduling problem. In this
+                  work the effect of the schedule generation schemes -
+                  serial or parallel - and priority rules - {MINLFT},
+                  {MINSLK}, {MAXTWK}, {SASP} or {FCFS} - with two
+                  approaches - multi-project and single-project - are
+                  analysed. The time criteria considered are the mean
+                  project delay and the multiproject duration
+                  increase. Through an extensive computational study,
+                  results show that with the parallel scheduling
+                  generation scheme and the multi-project approach the
+                  project manager can obtain a good multiproject
+                  schedule with the time criterion selected:
+                  minimising mean project delay or minimising
+                  multiproject duration increase. New heuristics -
+                  based on priority rules with a two-phase approach -
+                  that outperform classical ones are proposed to
+                  minimise mean project delay with a multi-project
+                  approach. Finally, the best heuristics analysed are
+                  evaluated together with a representative sample of
+                  commercial project management software.},
+  number = {1-4},
+  journal = {Annals of Operations Research},
+  month = feb,
+  year = 2001,
+  keywords = {Combinatorics, heuristic based on priority rules,
+                  Multiproject scheduling, Operation
+                  {Research/Decision} Theory, Project management,
+                  project management software, Resource allocation,
+                  Theory of Computation},
+  pages = {263--286}
+}
+
+ +
+@article{LovTorCer2009ijpe,
+  author = {Lova, Antonio and Tormos, Pilar and Cervantes,
+                  Mariamar and Barber, Federico},
+  title = {An efficient hybrid genetic algorithm for scheduling
+                  projects with resource constraints and multiple
+                  execution modes},
+  volume = 117,
+  doi = {10.1016/j.ijpe.2008.11.002},
+  abstract = {Multi-mode Resource Constrained Project Scheduling
+                  Problem ({MRCPSP)} aims at finding the start times
+                  and execution modes for the activities of a project
+                  that optimize a given objective function while
+                  verifying a set of precedence and resource
+                  constraints. In this paper, we focus on this problem
+                  and develop a hybrid Genetic Algorithm ({MM-HGA)} to
+                  solve it. Its main contributions are the mode
+                  assignment procedure, the fitness function and the
+                  use of a very efficient improving method. Its
+                  performance is demonstrated by extensive
+                  computational results obtained on a set of standard
+                  instances and against the best currently available
+                  algorithms.},
+  number = 2,
+  journal = {International Journal of Production Economics},
+  year = 2009,
+  keywords = {genetic algorithm, multi-mode resource-constrained
+                  project scheduling},
+  pages = {302--316}
+}
+
+ +
+@article{LozGloGarRodMar2014,
+  author = { Manuel Lozano  and  Fred Glover  and  Carlos Garc{\'i}a-Mart{\'i}nez  and Francisco J. Rodr{\'i}guez and  Rafael Mart{\'i} },
+  title = {Tabu Search with Strategic Oscillation for the Quadratic Minimum Spanning Tree},
+  journal = {IIE Transactions},
+  year = 2014,
+  volume = 46,
+  number = 4,
+  pages = {414--428}
+}
+
+ +
+@article{LozMolGar2011,
+  author = { Manuel Lozano  and  Daniel Molina  and  Carlos Garc{\'i}a-Mart{\'i}nez },
+  title = {Iterated Greedy for the Maximum Diversity Problem},
+  journal = {European Journal of Operational Research},
+  year = 2011,
+  volume = 214,
+  number = 1,
+  pages = {31--38}
+}
+
+ +
+@article{LuGloHao2010ejor,
+  author = { L{\"u}, Zhipeng  and  Fred Glover  and  Jin-Kao Hao },
+  title = {A hybrid metaheuristic approach to solving the
+                  {UBQP} problem},
+  journal = {European Journal of Operational Research},
+  volume = 207,
+  number = 3,
+  pages = {1254--1262},
+  year = 2010,
+  doi = {10.1016/j.ejor.2010.06.039}
+}
+
+ +
+@article{Lucas2014ising,
+  title = {Ising formulations of many {NP} problems},
+  author = {Lucas, Andrew},
+  journal = { Frontiers in Physics },
+  volume = 2,
+  pages = 5,
+  year = 2014,
+  publisher = {Frontiers},
+  doi = {10.3389/fphy.2014.00005}
+}
+
+ +
+@article{LundyMees1986,
+  title = {Convergence of an Annealing Algorithm},
+  author = { M. Lundy  and  A. Mees },
+  journal = {Mathematical Programming},
+  volume = 34,
+  number = 1,
+  pages = {111--124},
+  year = 1986
+}
+
+ +
+@article{LusTeg2009tpls,
+  author = { Thibaut Lust  and  Jacques Teghem },
+  title = {Two-phase {Pareto} local search for the biobjective traveling
+                  salesman problem},
+  doi = {10.1007/s10732-009-9103-9},
+  abstract = {In this work, we present a method, called {Two-Phase}
+                  {Pareto} Local Search, to find a good approximation of the
+                  efficient set of the biobjective traveling salesman
+                  problem. In the first phase of the method, an initial
+                  population composed of a good approximation of the extreme
+                  supported efficient solutions is generated. We use as second
+                  phase a {Pareto} Local Search method applied to each solution
+                  of the initial population. We show that using the combination
+                  of these two techniques: good initial population generation
+                  plus {Pareto} Local Search gives better results than
+                  state-of-the-art algorithms. Two other points are introduced:
+                  the notion of ideal set and a simple way to produce
+                  near-efficient solutions of multiobjective problems, by using
+                  an efficient single-objective solver with a data perturbation
+                  technique.  },
+  journal = {Journal of Heuristics},
+  volume = 16,
+  number = 3,
+  pages = {475--510},
+  year = 2010,
+  alias = {Lust09}
+}
+
+ +
+@article{LusTeg2010arxiv,
+  author = { Thibaut Lust  and  Jacques Teghem },
+  title = {The multiobjective multidimensional knapsack
+                  problem: a survey and a new approach},
+  journal = {Arxiv preprint arXiv:1007.4063},
+  year = 2010,
+  note = {Published as~\cite{LusTeg2012itor}}
+}
+
+ +
+@article{LusTeg2012itor,
+  title = {The multiobjective multidimensional knapsack
+                  problem: a survey and a new approach},
+  author = { Thibaut Lust  and  Jacques Teghem },
+  journal = {International Transactions in Operational Research},
+  volume = 19,
+  number = 4,
+  pages = {495--520},
+  year = 2012,
+  doi = {10.1111/j.1475-3995.2011.00840.x}
+}
+
+ +
+@article{LustJasz09btsp,
+  author = { Thibaut Lust  and  Andrzej Jaszkiewicz },
+  title = {Speed-up techniques for solving large-scale biobjective
+                  {TSP}},
+  journal = {Computers \& Operations Research},
+  year = 2010,
+  doi = {10.1016/j.cor.2009.01.005},
+  pages = {521--533},
+  volume = 37,
+  number = 3,
+  keywords = {Multiobjective combinatorial optimization, Hybrid
+                  metaheuristics, TSP, Local search, Speed-up techniques}
+}
+
+ +
+@article{LuvBarBri2014,
+  title = {A survey on multi-objective evolutionary algorithms
+                  for many-objective problems},
+  author = { C. von L{\"u}cken  and  Benjam{\'i}n Bar{\'a}n  and  Brizuela, Carlos},
+  pages = {707--756},
+  year = 2014,
+  journal = {Computational Optimization and Applications},
+  volume = 58,
+  number = 3
+}
+
+ +
+@article{MaaHin2008tsne,
+  author = {Laurens van der Maaten and Geoffrey Hinton},
+  title = {Visualizing Data using t-{SNE}},
+  journal = {Journal of Machine Learning Research},
+  year = 2008,
+  volume = 9,
+  number = 86,
+  pages = {2579--2605},
+  epub = {http://jmlr.org/papers/v9/vandermaaten08a.html}
+}
+
+ +
+@article{MachBelTal2018ale,
+  author = {Machado, Marlos C. and Bellemare, Marc G. and Talvitie, Erik
+                  and Veness, Joel and Hausknecht, Matthew and Bowling,
+                  Michael},
+  title = {Revisiting the {Arcade} {Learning} {Environment}: Evaluation
+                  Protocols and Open Problems for General Agents},
+  year = 2018,
+  publisher = {AI Access Foundation},
+  address = {El Segundo, CA, USA},
+  volume = 61,
+  number = 1,
+  issn = {1076-9757},
+  abstract = {The Arcade Learning Environment (ALE) is an evaluation
+                  platform that poses the challenge of building AI agents with
+                  general competency across dozens of Atari 2600 games. It
+                  supports a variety of different problem settings and it has
+                  been receiving increasing attention from the scientific
+                  community, leading to some high-pro_le success stories such
+                  as the much publicized Deep Q-Networks (DQN). In this article
+                  we take a big picture look at how the ALE is being used by
+                  the research community. We show how diverse the evaluation
+                  methodologies in the ALE have become with time, and highlight
+                  some key concerns when evaluating agents in the ALE. We use
+                  this discussion to present some methodological best practices
+                  and provide new benchmark results using these best
+                  practices. To further the progress in the field, we introduce
+                  a new version of the ALE that supports multiple game modes
+                  and provides a form of stochasticity we call sticky
+                  actions. We conclude this big picture look by revisiting
+                  challenges posed when the ALE was introduced, summarizing the
+                  state-of-the-art in various problems and highlighting
+                  problems that remain open.},
+  journal = {Journal of Artificial Intelligence Research},
+  month = jan,
+  pages = {523--562},
+  numpages = 40
+}
+
+ +
+@article{Madden2012,
+  title = {From Databases to Big Data},
+  author = {Madden, Sam},
+  journal = {IEEE Internet Computing},
+  volume = 16,
+  number = 3,
+  year = 2012
+}
+
+ +
+@article{MahFesDam2007harmony,
+  author = {M. Mahdavi and M. Fesanghary and E. Damangir},
+  title = {An improved harmony search algorithm for solving optimization
+                  problems},
+  journal = {Applied Mathematics and Computation},
+  volume = 188,
+  number = 2,
+  pages = {1567--1579},
+  year = 2007,
+  doi = {10.1016/j.amc.2006.11.033},
+  keywords = {Global optimization, Heuristics, Harmony search algorithm,
+                  Mathematical programming},
+  abstract = {This paper develops an Improved harmony search (IHS)
+                  algorithm for solving optimization problems. IHS employs a
+                  novel method for generating new solution vectors that
+                  enhances accuracy and convergence rate of harmony search (HS)
+                  algorithm. In this paper the impacts of constant parameters
+                  on harmony search algorithm are discussed and a strategy for
+                  tuning these parameters is presented. The IHS algorithm has
+                  been successfully applied to various benchmarking and
+                  standard engineering optimization problems. Numerical results
+                  reveal that the proposed algorithm can find better solutions
+                  when compared to HS and other heuristic or deterministic
+                  methods and is a powerful search algorithm for various
+                  engineering optimization problems.}
+}
+
+ +
+@article{MaiRon2012,
+  title = {New heuristics for total tardiness minimization in
+                  a flexible flowshop},
+  author = {Mainieri, Guilherme B. and Ronconi, D{\'e}bora P.},
+  journal = {Optimization Letters},
+  pages = {1--20},
+  year = 2012
+}
+
+ +
+@article{MaieSimp03:ACODesignWDN,
+  author = { Holger R. Maier  and  Angus R. Simpson  and  Aaron C. Zecchin  and  Wai Kuan Foong  and  Kuang Yeow Phang  and  Hsin Yeow Seah  and  Tan, Chan Lim },
+  title = {Ant Colony Optimization for Design of Water Distribution
+                  Systems},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 129,
+  number = 3,
+  pages = {200--209},
+  date = {2003-05/2003-06},
+  year = 2003,
+  month = may # { / } # jun
+}
+
+ +
+@article{MMalDas2022twiarch,
+  title = {A twin-archive guided decomposition based
+                  multi/many-objective evolutionary algorithm},
+  author = {M, Sri Srinivasa Raju and Mallipeddi, Rammohan and Das, Kedar Nath},
+  journal = {Swarm and Evolutionary Computation},
+  volume = 71,
+  pages = 101082,
+  year = 2022,
+  doi = {10.1016/j.swevo.2022.101082},
+  publisher = {Elsevier}
+}
+
+ +
+@article{MalEng2013survey,
+  author = {Katherine M. Malan and  Andries Engelbrecht },
+  title = {A survey of techniques for characterising fitness landscapes
+                  and some possible ways forward},
+  journal = {Information Sciences},
+  volume = 241,
+  pages = {148--163},
+  year = 2013,
+  issn = {0020-0255},
+  doi = {10.1016/j.ins.2013.04.015}
+}
+
+ +
+@article{Males85,
+  author = { R. M. Males  and  R. M. Clark  and  P. J. Wehrman  and  W. E. Gateset },
+  title = {Algorithm for mixing problems in water systems},
+  journal = { Journal of Hydraulic Engineering, {ASCE}},
+  year = 1985,
+  volume = 111,
+  number = 2,
+  pages = {206--219}
+}
+
+ +
+@article{Man1999:joc,
+  author = { Vittorio Maniezzo },
+  title = {Exact and Approximate Nondeterministic Tree-Search Procedures
+                  for the Quadratic Assignment Problem},
+  journal = {INFORMS Journal on Computing},
+  year = 1999,
+  volume = 11,
+  number = 4,
+  pages = {358--369},
+  alias = {Man99:informs}
+}
+
+ +
+@article{ManCar2000:fgcs,
+  author = { Vittorio Maniezzo  and  A. Carbonaro },
+  title = {An {ANTS} Heuristic for the Frequency Assignment Problem},
+  journal = {Future Generation Computer Systems},
+  year = 2000,
+  volume = 16,
+  number = 8,
+  pages = {927--935},
+  alias = {ManCar00}
+}
+
+ +
+@article{ManCol99,
+  author = { Vittorio Maniezzo  and  Alberto Colorni },
+  title = {The {Ant} {System} Applied to the Quadratic
+                  Assignment Problem},
+  journal = {IEEE Transactions on Knowledge and Data Engineering},
+  year = 1999,
+  volume = 11,
+  number = 5,
+  pages = {769--778}
+}
+
+ +
+@article{Mar84,
+  title = {On a multicritera shortest path problem},
+  journal = {European Journal of Operational Research},
+  volume = 16,
+  pages = {236--245},
+  year = 1984,
+  author = {E. Q. V. Martins}
+}
+
+ +
+@article{MarAro2004smo,
+  author = {Marler, R. T. and Arora, J. S.},
+  title = {Survey of multi-objective optimization methods for
+                  engineering},
+  journal = {Structural and Multidisciplinary Optimization},
+  year = 2004,
+  volume = 26,
+  number = 6,
+  pages = {369--395},
+  month = apr,
+  doi = {10.1007/s00158-003-0368-6},
+  annote = {Discusses a priori (scalarized) methods.}
+}
+
+ +
+@article{MarCavHer2023repr,
+  author = { Raul Mart{\'i}n-Santamar{\'i}a  and Cavero, Sergio and Herrán, Alberto and  Duarte, Abraham  and  Colmenar, J. Manuel },
+  title = {A Practical Methodology for Reproducible Experimentation: An
+                  Application to the Double-Row Facility Layout Problem},
+  journal = {Evolutionary Computation},
+  year = 2023,
+  pages = {1--35},
+  month = 11,
+  issn = {1063-6560},
+  doi = {10.1162/evco_a_00317},
+  publisher = {MIT Press},
+  keywords = {irace}
+}
+
+ +
+@article{MarDeBHaeVanSnoBae07,
+  author = {D. Martens and M. De Backer and R. Haesen and
+                  J. Vanthienen and M. Snoeck and B. Baesens},
+  title = {Classification With Ant Colony Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2007,
+  volume = 11,
+  number = 5,
+  pages = {651--665}
+}
+
+ +
+@article{MarMarWeiWol2002,
+  title = {Cutting planes in integer and mixed integer programming},
+  author = {Marchand, Hugues and Martin, Alexander and Weismantel, Robert and Wolsey, Laurence},
+  journal = {Discrete Applied Mathematics},
+  volume = 123,
+  number = {1--3},
+  pages = {397--446},
+  year = 2002,
+  publisher = {Elsevier}
+}
+
+ +
+@article{MarMoo1997air,
+  author = {O. Maron and A. W. Moore},
+  title = {The Racing Algorithm: {Model} Selection for Lazy Learners},
+  journal = {Artificial Intelligence Research},
+  year = 1997,
+  volume = 11,
+  number = {1--5},
+  pages = {193--225}
+}
+
+ +
+@article{MarOtt1995,
+  author = { Olivier Martin  and S. W. Otto},
+  title = {Partitioning of Unstructured Meshes for Load
+                  Balancing},
+  journal = {Concurrency: Practice and Experience},
+  year = 1995,
+  volume = 7,
+  number = 4,
+  pages = {303--314}
+}
+
+ +
+@article{MarOtt1996:aor,
+  author = { Olivier Martin  and S. W. Otto},
+  title = {Combining Simulated Annealing with Local Search
+                  Heuristics},
+  journal = {Annals of Operations Research},
+  year = 1996,
+  volume = 63,
+  pages = {57--75}
+}
+
+ +
+@article{MarOttFel91:cs,
+  author = { Olivier Martin  and S. W. Otto and E. W. Felten},
+  title = {Large-Step {Markov} Chains for the Traveling
+                  Salesman Problem},
+  journal = {Complex Systems},
+  year = 1991,
+  volume = 5,
+  number = 3,
+  pages = {299--326}
+}
+
+ +
+@article{MarOttFel92:orl,
+  author = { Olivier Martin  and S. W. Otto and E. W. Felten},
+  title = {Large-step {Markov} Chains for the {TSP}
+                  Incorporating Local Search Heuristics},
+  journal = {Operations Research Letters},
+  year = 1992,
+  volume = 11,
+  number = 4,
+  pages = {219--224}
+}
+
+ +
+@article{MarReiDua2012,
+  author = { Rafael Mart{\'i}  and  Gerhard Reinelt  and  Duarte, Abraham },
+  title = {A Benchmark Library and a Comparison of Heuristic Methods for the Linear Ordering Problem},
+  journal = {Computational Optimization and Applications},
+  year = 2012,
+  volume = 51,
+  number = 3,
+  pages = {1297--1317}
+}
+
+ +
+@article{MarSanPer2022strategic,
+  author = { Raul Mart{\'i}n-Santamar{\'i}a  and   Jes{\'u}s S{\'a}nchez-Oro  and S. P\'{e}rez-Pel\'{o} and  Duarte, Abraham },
+  title = {Strategic oscillation for the balanced minimum sum-of-squares
+                  clustering problem},
+  journal = {Information Sciences},
+  year = 2022,
+  volume = 585,
+  pages = {529--542},
+  doi = {10.1016/j.ins.2021.11.048}
+}
+
+ +
+@article{MarTot1990dam,
+  author = { Silvano Martello  and  Paolo Toth },
+  title = {Lower bounds and reduction procedures for the bin
+                  packing problem},
+  journal = {Discrete Applied Mathematics},
+  volume = 28,
+  number = 1,
+  year = 1990,
+  pages = {59--70},
+  doi = {10.1016/0166-218X(90)90094-S}
+}
+
+ +
+@article{MarVig1998exact,
+  title = {Exact solution of the two-dimensional finite bin packing
+                  problem},
+  author = { Silvano Martello  and  Vigo, Daniele },
+  journal = {Management Science},
+  volume = 44,
+  number = 3,
+  pages = {388--399},
+  year = 1998,
+  publisher = {{INFORMS}},
+  doi = {10.1287/mnsc.44.3.388}
+}
+
+ +
+@article{MasHwa1981isgp,
+  author = {Masud, Abu S. and Hwang, C. L.},
+  title = {Interactive Sequential Goal Programming},
+  journal = {Journal of the Operational Research Society},
+  year = 1981,
+  volume = 32,
+  number = 5,
+  pages = {391--400},
+  month = may,
+  issn = {1476-9360},
+  doi = {10.1057/jors.1981.76},
+  abstract = {This paper introduces a new solution method based on Goal
+                  Programming for Multiple Objective Decision Making (MODM)
+                  problems. The method, called Interactive Sequential Goal
+                  Programming (ISGP), combines and extends the attractive
+                  features of both Goal Programming and interactive solution
+                  approaches for MODM problems. ISGP is applicable to both
+                  linear and non-linear problems. It uses existing single
+                  objective optimization techniques and, hence, available
+                  computer codes utilizing these techniques can be adapted for
+                  use in ISGP. The non-dominance of the "best-compromise"
+                  solution is assured. The information required from the
+                  decision maker in each iteration is simple. The proposed
+                  method is illustrated by solving a nutrition problem.}
+}
+
+ +
+@article{MasLopDubStu2014cor,
+  author = { Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle },
+  title = {Grammar-Based Generation of Stochastic Local Search
+                  Heuristics through Automatic Algorithm Configuration Tools},
+  journal = {Computers \& Operations Research},
+  year = 2014,
+  doi = {10.1016/j.cor.2014.05.020},
+  volume = 51,
+  pages = {190--199}
+}
+
+ +
+@article{MasPelStuBir2014itor,
+  author = { Franco Mascia  and  Paola Pellegrini  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {An Analysis of Parameter Adaptation in Reactive Tabu Search},
+  journal = {International Transactions in Operational Research},
+  year = 2014,
+  volume = 21,
+  number = 1,
+  pages = {127--152}
+}
+
+ +
+@article{MasVidMic++2013,
+  author = {Renaud Masson and  Thibaut Vidal  and Julien Michallet and Puca Huachi {Vaz Penna} and Vinicius Petrucci and  Anand Subramanian  and Hugues Dubedout},
+  title = {An Iterated Local Search Heuristic for Multi-capacity Bin Packing and Machine Reassignment Problems},
+  journal = {Expert Systems with Applications},
+  year = 2013,
+  volume = 40,
+  number = 13,
+  pages = {5266--5275}
+}
+
+ +
+@article{MatDauLah2011:ejor,
+  author = {Yazid Mati and  St{\'e}phane Dauz{\`e}re-P{\`e}r{\'e}s  and Chams Lahlou},
+  title = {A General Approach for Optimizing Regular Criteria in the Job-shop Scheduling Problem},
+  journal = {European Journal of Operational Research},
+  year = 2011,
+  volume = 212,
+  number = 1,
+  pages = {33--42}
+}
+
+ +
+@article{MazLopChuMie2023tgp,
+  author = { Atanu Mazumdar  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Tinkle Chugh  and  Kaisa Miettinen },
+  title = {Treed {Gaussian} Process Regression for Solving Offline
+                  Data-Driven Continuous Multiobjective Optimization Problems},
+  journal = {Evolutionary Computation},
+  year = 2023,
+  volume = 31,
+  number = 4,
+  pages = {375--399},
+  doi = {10.1162/evco_a_00329},
+  abstract = {For offline data-driven multiobjective optimization problems
+                  (MOPs), no new data is available during the optimization
+                  process. Approximation models (or surrogates) are first built
+                  using the provided offline data and an optimizer, e.g. a
+                  multiobjective evolutionary algorithm, can then be utilized
+                  to find Pareto optimal solutions to the problem with
+                  surrogates as objective functions. In contrast to online
+                  data-driven MOPs, these surrogates cannot be updated with new
+                  data and, hence, the approximation accuracy cannot be
+                  improved by considering new data during the optimization
+                  process. Gaussian process regression (GPR) models are widely
+                  used as surrogates because of their ability to provide
+                  uncertainty information. However, building GPRs becomes
+                  computationally expensive when the size of the dataset is
+                  large. Using sparse GPRs reduces the computational cost of
+                  building the surrogates. However, sparse GPRs are not
+                  tailored to solve offline data-driven MOPs, where good
+                  accuracy of the surrogates is needed near Pareto optimal
+                  solutions. Treed GPR (TGPR-MO) surrogates for offline
+                  data-driven MOPs with continuous decision variables are
+                  proposed in this paper. The proposed surrogates first split
+                  the decision space into subregions using regression trees and
+                  build GPRs sequentially in regions close to Pareto optimal
+                  solutions in the decision space to accurately approximate
+                  tradeoffs between the objective functions. TGPR-MO surrogates
+                  are computationally inexpensive because GPRs are built only
+                  in a smaller region of the decision space utilizing a subset
+                  of the data. The TGPR-MO surrogates were tested on
+                  distance-based visualizable problems with various data sizes,
+                  sampling strategies, numbers of objective functions, and
+                  decision variables. Experimental results showed that the
+                  TGPR-MO surrogates are computationally cheaper and can handle
+                  datasets of large size. Furthermore, TGPR-MO surrogates
+                  produced solutions closer to Pareto optimal solutions
+                  compared to full GPRs and sparse GPRs.},
+  keywords = {Gaussian processes, Kriging, Regression trees, Metamodelling,
+                  Surrogate, Pareto optimality}
+}
+
+ +
+@article{McConMehNah2011certifying,
+  title = {Certifying algorithms},
+  journal = {Computer Science Review},
+  volume = 5,
+  number = 2,
+  pages = {119--161},
+  year = 2011,
+  issn = {1574-0137},
+  doi = {10.1016/j.cosrev.2010.09.009},
+  author = {Ross M. McConnell and Kurt Mehlhorn and Stefan N{\"a}her and
+                  Pascal Schweitzer},
+  keywords = {Algorithms, Software reliability, Certification},
+  abstract = {A certifying algorithm is an algorithm that produces, with
+                  each output, a certificate or witness (easy-to-verify proof)
+                  that the particular output has not been compromised by a
+                  bug. A user of a certifying algorithm inputs x, receives the
+                  output y and the certificate w, and then checks, either
+                  manually or by use of a program, that w proves that y is a
+                  correct output for input x. In this way, he/she can be sure
+                  of the correctness of the output without having to trust the
+                  algorithm. We put forward the thesis that certifying
+                  algorithms are much superior to non-certifying algorithms,
+                  and that for complex algorithmic tasks, only certifying
+                  algorithms are satisfactory. Acceptance of this thesis would
+                  lead to a change of how algorithms are taught and how
+                  algorithms are researched. The widespread use of certifying
+                  algorithms would greatly enhance the reliability of
+                  algorithmic software. We survey the state of the art in
+                  certifying algorithms and add to it. In particular, we start
+                  a theory of certifying algorithms and prove that the concept
+                  is universal.}
+}
+
+ +
+@article{McCorPow03demand,
+  title = {Optimal Pump Scheduling in Water Supply Systems with
+                  Maximum Demand Charges},
+  author = { G. McCormick  and  R. S. Powell },
+  publisher = {ASCE},
+  year = 2003,
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 129,
+  number = 5,
+  pages = {372--379},
+  keywords = {water supply; pumps; Markov processes; cost optimal
+                  control},
+  epub = {http://link.aip.org/link/?QWR/129/372/1},
+  doi = {10.1061/(ASCE)0733-9496(2003)129:5(372)}
+}
+
+ +
+@article{McCormick04,
+  author = { G. McCormick  and  R. S. Powell },
+  title = {Derivation of near-optimal pump schedules for water
+                  distribution by simulated annealing},
+  journal = {Journal of the Operational Research Society},
+  year = 2004,
+  volume = 55,
+  number = 7,
+  pages = {728--736},
+  month = jul,
+  doi = {10.1057/palgrave.jors.2601718},
+  abstract = {The scheduling of pumps for clean water distribution is a
+                  partially discrete non-linear problem with many
+                  variables. The scheduling method described in this paper
+                  typically produces costs within 1\% of a linear program-based
+                  solution, and can incorporate realistic non-linear costs that
+                  may be hard to incorporate in linear programming
+                  formulations. These costs include pump switching and maximum
+                  demand charges. A simplified model is derived from a standard
+                  hydraulic simulator. An initial schedule is produced by a
+                  descent method. Two-stage simulated annealing then produces
+                  solutions in a few minutes. Iterative recalibration ensures
+                  that the solution agrees closely with the results from a full
+                  hydraulic simulation.}
+}
+
+ +
+@article{McDermott2020nfl,
+  author = {James McDermott},
+  title = {When and Why Metaheuristics Researchers can Ignore "No Free
+                  Lunch" Theorems},
+  journal = {{SN} Computer Science},
+  volume = 1,
+  number = 60,
+  pages = {1--18},
+  year = 2020,
+  doi = {10.1007/s42979-020-0063-3}
+}
+
+ +
+@article{McG1992vrt,
+  author = { Catherine C. McGeoch },
+  title = {Analyzing Algorithms by Simulation: Variance Reduction
+                  Techniques and Simulation Speedups},
+  abstract = {Although experimental studies have been widely applied to the
+                  investigation of algorithm performance, very little attention
+                  has been given to experimental method in this area. This is
+                  unfortunate, since much can be done to improve the quality of
+                  the data obtained; often, much improvement may be needed for
+                  the data to be useful. This paper gives a tutorial discussion
+                  of two aspects of good experimental technique: the use of
+                  variance reduction techniques and simulation speedups in
+                  algorithm studies.  In an illustrative study, application of
+                  variance reduction techniques produces a decrease in variance
+                  by a factor 1000 in one case, giving a dramatic improvement
+                  in the precision of experimental results. Furthermore, the
+                  complexity of the simulation program is improved from
+                  $\Theta(m n/H_n)$ to $\Theta(m + n \log n)$ (where $m$ is
+                  typically much larger than $n$), giving a much faster
+                  simulation program and therefore more data per unit of
+                  computation time. The general application of variance
+                  reduction techniques is also discussed for a variety of
+                  algorithm problem domains.},
+  volume = 24,
+  doi = {10.1145/130844.130853},
+  number = 2,
+  journal = {{ACM} Computing Surveys},
+  year = 1992,
+  keywords = {experimental analysis of algorithms, move-to-front rule,
+                  self-organizing sequential search, statistical analysis of
+                  algorithms, transpose rule, variance reduction techniques},
+  pages = {195--212}
+}
+
+ +
+@article{McG1998joc,
+  author = { Catherine C. McGeoch },
+  title = {Toward an Experimental Method for Algorithm Simulation},
+  journal = {INFORMS Journal on Computing},
+  year = 1996,
+  volume = 8,
+  number = 1,
+  pages = {1--15},
+  doi = {10.1287/ijoc.8.1.1}
+}
+
+ +
+@article{MckBecCon1979lhs,
+  title = {A Comparison of Three Methods for Selecting Values of Input
+                  Variables in the Analysis of Output from a Computer Code},
+  author = {Michael D. McKay and Richard J. Beckman and  W. J. Conover },
+  journal = {Technometrics},
+  year = 1979,
+  number = 2,
+  pages = {239--245},
+  volume = 21,
+  alias = {McKay1979},
+  abstract = {Two types of sampling plans are examined as alternatives to
+                  simple random sampling in Monte Carlo studies. These plans
+                  are shown to be improvements over simple random sampling with
+                  respect to variance for a class of estimators which includes
+                  the sample mean and the empirical distribution function.},
+  publisher = {American Statistical Association and American Society for
+                  Quality},
+  doi = {10.2307/1268522}
+}
+
+ +
+@article{MckBerMaiFic2018combining,
+  title = {Combining local preferences with multi-criteria decision
+                  analysis and linear optimization to develop feasible energy
+                  concepts in small communities},
+  author = {McKenna, Russell and Bertsch, Valentin and Mainzer, Kai and
+                  Fichtner, Wolf},
+  journal = {European Journal of Operational Research},
+  volume = 268,
+  number = 3,
+  pages = {1092--1110},
+  year = 2018
+}
+
+ +
+@article{Mckay2010,
+  author = {Mckay, Robert I. and Hoai, Nguyen Xuan and Whigham,
+                  Peter Alexander and Shan, Yin and  O'Neill, Michael },
+  title = {Grammar-based Genetic Programming: A Survey},
+  journal = {Genetic Programming and Evolvable Machines},
+  volume = 11,
+  number = {3-4},
+  month = sep,
+  year = 2010,
+  pages = {365--396},
+  doi = {10.1007/s10710-010-9109-y}
+}
+
+ +
+@article{Meer2007,
+  author = {Klaus Meer},
+  title = {Simulated annealing versus {Metropolis} for a {TSP} instance},
+  journal = {Information Processing Letters},
+  volume = 104,
+  number = 6,
+  year = 2007,
+  pages = {216--219}
+}
+
+ +
+@article{MelDyeBlu2017neural,
+  author = {G{\'{a}}bor Melis and Chris Dyer and Phil Blunsom},
+  title = {On the State of the Art of Evaluation in Neural Language
+                  Models},
+  journal = {Arxiv preprint arXiv:1807.02811},
+  year = 2017,
+  url = {http://arxiv.org/abs/1707.05589}
+}
+
+ +
+@article{MelNicSal2009facility,
+  title = {Facility location and supply chain management: {A} review},
+  author = {Melo, M. T. and Nickel, S. and  Saldanha-da-Gama, F. },
+  year = 2009,
+  journal = {European Journal of Operational Research},
+  volume = 196,
+  number = 2,
+  pages = {401--412},
+  doi = {10.1016/j.ejor.2008.05.007}
+}
+
+ +
+@article{Men2008,
+  author = {Ole J. Mengshoel},
+  title = {Understanding the role of noise in stochastic local search:
+                  Analysis and experiments},
+  journal = {Artificial Intelligence},
+  volume = 172,
+  number = 8,
+  pages = {955--990},
+  year = 2008
+}
+
+ +
+@article{MerCot2006sigevo,
+  author = { Juan-Juli{\'a}n Merelo  and  Carlos Cotta },
+  title = {Building bridges: the role of subfields in metaheuristics},
+  journal = { {SIGEVO}lution },
+  year = 2006,
+  volume = 1,
+  number = 4,
+  pages = {9--15},
+  doi = {10.1145/1229735.1229737}
+}
+
+ +
+@article{MerFre02:cs,
+  author = { Peter Merz  and  Bernd Freisleben },
+  title = {Memetic Algorithms for the Traveling Salesman
+                  Problem},
+  journal = {Complex Systems},
+  year = 2001,
+  volume = 13,
+  number = 4,
+  pages = {297--345}
+}
+
+ +
+@article{MerFre2000:tec,
+  author = { Peter Merz  and  Bernd Freisleben },
+  title = {Fitness Landscape Analysis and Memetic Algorithms for the Quadratic Assignment Problem},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2000,
+  volume = 4,
+  number = 4,
+  pages = {337--352}
+}
+
+ +
+@article{MerKat2004ubqp,
+  author = { Peter Merz  and Kengo Katayama},
+  title = {Memetic algorithms for the unconstrained binary quadratic
+                  programming problem},
+  journal = {BioSystems},
+  volume = 78,
+  number = 1,
+  pages = {99--118},
+  year = 2004,
+  doi = {10.1016/j.biosystems.2004.08.002}
+}
+
+ +
+@article{MerMid2002:appi,
+  author = { D. Merkle  and  Martin Middendorf },
+  title = {Ant Colony Optimization with Global Pheromone
+                  Evaluation for Scheduling a Single Machine},
+  journal = {Applied Intelligence},
+  year = 2003,
+  volume = 18,
+  number = 1,
+  pages = {105--111},
+  alias = {MerMid02:ai}
+}
+
+ +
+@article{MerMid2002:ec,
+  author = { D. Merkle  and  Martin Middendorf },
+  title = {Modeling the Dynamics of Ant Colony Optimization},
+  journal = {Evolutionary Computation},
+  volume = 10,
+  number = 3,
+  pages = {235--262},
+  year = 2002
+}
+
+ +
+@article{MerMidSch02:tec,
+  author = { D. Merkle  and  Martin Middendorf  and  Hartmut Schmeck },
+  title = {Ant Colony Optimization for Resource-Constrained
+                  Project Scheduling},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002,
+  volume = 6,
+  number = 4,
+  pages = {333--346}
+}
+
+ +
+@article{Merz2002joh,
+  author = { Peter Merz  and  Bernd Freisleben },
+  title = {Greedy and Local Search Heuristics for Unconstrained
+                  Binary Quadratic Programming},
+  year = 2002,
+  journal = {Journal of Heuristics},
+  volume = 8,
+  number = 2,
+  doi = {10.1023/A:1017912624016},
+  pages = {197--213}
+}
+
+ +
+@article{MesSilMelMir20152esa,
+  author = {Rafael G. Mesquita and Ricardo M. A. Silva and Carlos A. B. Mello and P\'{e}ricles B. C. Miranda},
+  title = {Parameter tuning for document image binarization using a racing algorithm},
+  journal = {Expert Systems with Applications},
+  volume = 42,
+  number = 5,
+  pages = {2593--2603},
+  year = 2015,
+  doi = {10.1016/j.eswa.2014.10.039},
+  keywords = {irace}
+}
+
+ +
+@article{MetRosRosTel53,
+  author = {N. Metropolis and A. W. Rosenbluth and M. N. Rosenbluth and A. Teller and E. Teller},
+  title = {Equation of State Calculations by Fast Computing Machines},
+  journal = {Journal of Chemical Physics},
+  year = 1953,
+  volume = 21,
+  pages = {1087--1092}
+}
+
+ +
+@article{MeuDor2002:al,
+  author = {Nicolas Meuleau and  Marco Dorigo },
+  title = {Ant Colony Optimization and Stochastic Gradient Descent},
+  volume = 8,
+  number = 2,
+  pages = {103--121},
+  journal = {Artificial Life},
+  year = 2002
+}
+
+ +
+@article{MeuRakWon2020iclrbb,
+  author = {Laurent Meunier and Herilalaina Rakotoarison and Pak{-}Kan
+                  Wong and Baptiste Rozi{\`{e}}re and J{\'{e}}r{\'{e}}my Rapin and Olivier Teytaud  and Antoine Moreau and  Carola Doerr },
+  title = {Black-Box Optimization Revisited: Improving Algorithm
+                  Selection Wizards through Massive Benchmarking},
+  journal = {Arxiv preprint arXiv:2010.04542},
+  year = 2020,
+  doi = {10.48550/arXiv.2010.04542},
+  keywords = {Nevergrad, NGOpt}
+}
+
+ +
+@article{MeuRakWon2022ngopt,
+  author = {Laurent Meunier and Herilalaina Rakotoarison and Pak{-}Kan
+                  Wong and Baptiste Rozi{\`{e}}re and J{\'{e}}r{\'{e}}my Rapin and Olivier Teytaud  and Antoine Moreau and  Carola Doerr },
+  title = {Black-Box Optimization Revisited: Improving Algorithm
+                  Selection Wizards Through Massive Benchmarking},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2022,
+  volume = 26,
+  number = 3,
+  pages = {490--500},
+  doi = {10.1109/TEVC.2021.3108185},
+  keywords = {nevergrad, NGOpt}
+}
+
+ +
+@article{Mhallah2014,
+  author = {R. {M'Hallah}},
+  title = {An iterated local search variable neighborhood descent hybrid heuristic for the total earliness tardiness permutation flow shop},
+  journal = {International Journal of Production Research},
+  year = 2014,
+  volume = 52,
+  number = 13,
+  pages = {3802--3819}
+}
+
+ +
+@article{MicDasRic1996cie,
+  author = { Zbigniew Michalewicz  and Dipankar Dasgupta and Rodolphe
+                  G. Le Riche and  Marc Schoenauer },
+  title = {Evolutionary algorithms for constrained engineering
+                  problems},
+  journal = {Computers and Industrial Engineering},
+  volume = 30,
+  number = 4,
+  pages = {851--870},
+  year = 1996,
+  doi = {10.1016/0360-8352(96)00037-X}
+}
+
+ +
+@article{MicPriAmoYal2014:cor,
+  author = {Julien Michallet and  Christian Prins  and Farouk Yalaoui and Gr{\'{e}}goire Vitry},
+  title = {Multi-start Iterated Local Search for the Periodic Vehicle Routing
+               Problem with Time Windows and Time Spread Constraints on Services},
+  journal = {Computers \& Operations Research},
+  year = 2014,
+  volume = 41,
+  pages = {196--207}
+}
+
+ +
+@article{Mie2014or,
+  title = {Survey of methods to visualize alternatives in multiple
+                  criteria decision making problems},
+  author = { Kaisa Miettinen },
+  journal = {OR Spectrum},
+  volume = 36,
+  number = 1,
+  pages = {3--37},
+  year = 2014,
+  doi = {10.1007/s00291-012-0297-0}
+}
+
+ +
+@article{MieMusSte2014nimbus,
+  title = {Interactive multiobjective optimization with {NIMBUS} for
+                  decision making under uncertainty},
+  author = { Kaisa Miettinen  and Mustajoki, Jyri and  T. J. Stewart },
+  journal = {OR Spectrum},
+  volume = 36,
+  number = 1,
+  pages = {39--56},
+  year = 2014
+}
+
+ +
+@article{MilAnd2005repl,
+  author = {R. B. {Millar} and M. J. {Anderson}},
+  title = {Remedies for pseudoreplication},
+  journal = { Fisheries Research },
+  volume = 70,
+  number = {2--3},
+  pages = {397--407},
+  year = 2004,
+  doi = {10.1016/j.fishres.2004.08.016}
+}
+
+ +
+@article{Miller1956,
+  title = {The magical number seven, plus or minus two: Some limits on
+                  our capacity for processing information},
+  author = {Miller, George A.},
+  journal = {Psychological Review},
+  volume = 63,
+  number = 2,
+  pages = {81--97},
+  year = 1956,
+  publisher = {American Psychological Association},
+  doi = {10.1037/h0043158}
+}
+
+ +
+@article{Min96auto,
+  author = {Steven Minton},
+  title = {Automatically configuring constraint satisfaction programs: A
+                  case study},
+  journal = {Constraints},
+  year = 1996,
+  volume = 1,
+  number = 1,
+  pages = {7--43},
+  doi = {10.1007/BF00143877}
+}
+
+ +
+@article{MinRuiCia08,
+  author = { Gerardo Minella  and  Rub{\'e}n Ruiz  and  M. Ciavotta },
+  title = {A Review and Evaluation of Multiobjective Algorithms
+                  for the Flowshop Scheduling Problem},
+  journal = {INFORMS Journal on Computing},
+  volume = 20,
+  number = 3,
+  pages = {451--471},
+  year = 2008,
+  alias = {GerardoMinella,MinRuiCia08mfsp}
+}
+
+ +
+@article{MisAfsLar2022rximo,
+  author = {Misitano, Giovanni and Afsar, Bekir and Larraga, Giomara and  Kaisa Miettinen },
+  title = {Towards explainable interactive multiobjective optimization:
+                  {R-XIMO}},
+  journal = {Autonomous Agents and Multi-Agent Systems},
+  year = 2022,
+  volume = 36,
+  number = 42,
+  doi = {10.1007/s10458-022-09577-3}
+}
+
+ +
+@article{MisKuz2018,
+  title = {Investigating some strategies for construction of
+                  initial populations in genetic algorithms},
+  author = { Misevi{\v{c}}ius, Alfonsas  and Kuznecovait{\.e}, Dovil{\.e}},
+  journal = {Computational Science and Techniques},
+  volume = 5,
+  number = 1,
+  pages = {560--573},
+  year = 2018
+}
+
+ +
+@article{Misevicius2003,
+  author = { Misevi{\v{c}}ius, Alfonsas },
+  title = {Genetic Algorithm Hybridized with Ruin and Recreate Procedure: Application to the Quadratic Assignment Problem},
+  journal = {Knowledge-Based Systems},
+  year = 2003,
+  volume = 16,
+  number = {5--6},
+  pages = {261--268}
+}
+
+ +
+@article{Misevicius2003:inf,
+  title = {A modified simulated annealing algorithm for the quadratic assignment problem},
+  author = { Misevi{\v{c}}ius, Alfonsas },
+  journal = {Informatica},
+  volume = 14,
+  number = 4,
+  pages = {497--514},
+  year = 2003,
+  publisher = {IOS Press, Amsterdam, The Netherlands}
+}
+
+ +
+@article{MitMurPal2002unsupervised,
+  author = {P. Mitra and C. A. Murthy and S. K. Pal},
+  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
+  title = {Unsupervised feature selection using feature similarity},
+  year = 2002,
+  volume = 24,
+  number = 3,
+  pages = {301--312},
+  doi = {10.1109/34.990133}
+}
+
+ +
+@article{MizKuzPla2018,
+  author = { Misevi{\v{c}}ius, Alfonsas  and Kuznecovait{\.e}, Dovil{\.e} and
+                  J{\={u}}rat{\.e} Platu{\v{z}}ien{\.e}},
+  title = {Some Further Experiments with Crossover Operators for Genetic Algorithms},
+  journal = {Informatica},
+  volume = 29,
+  number = 3,
+  year = 2018,
+  pages = {499--516}
+}
+
+ +
+@article{MlaHan1997:cor,
+  author = { Nenad Mladenovi{\'c}  and  Pierre Hansen },
+  title = {Variable Neighborhood Search},
+  journal = {Computers \& Operations Research},
+  year = 1997,
+  volume = 24,
+  number = 11,
+  pages = {1097--1100}
+}
+
+ +
+@article{MniKavSil2015human,
+  title = {Human-level control through deep reinforcement learning},
+  author = {Mnih, Volodymyr and Kavukcuoglu, Koray and Silver, David and
+                  Rusu, Andrei A. and Veness, Joel and Bellemare, Marc G. and
+                  Graves, Alex and Riedmiller, Martin and Fidjeland, Andreas K.
+                  and Ostrovski, Georg and others},
+  journal = {Nature},
+  volume = 518,
+  number = 7540,
+  pages = 529,
+  year = 2015,
+  publisher = {Nature Publishing Group}
+}
+
+ +
+@article{MocTieZil1978,
+  author = { Jonas Mo{\v{c}}kus  and Tiesis, Vytautas and Zilinskas, Antanas},
+  title = {The application of bayesian methods for seeking the extremum},
+  journal = {Towards global optimization},
+  year = 1978,
+  pages = {117--129},
+  annote = {Proposed Bayesian optimization (but later than
+                  \cite{Mockus1975})}
+}
+
+ +
+@article{MolSanHer2009gdominance,
+  author = { Molina, Juli{\'a}n  and Luis V. Santana and Alfredo
+                  G. Hern\'{a}ndez-D{\'i}az and  Carlos A. {Coello Coello}  and Rafael Caballero},
+  title = {{g}-Dominance: Reference point based dominance for
+                  Multiobjective Metaheuristics},
+  journal = {European Journal of Operational Research},
+  volume = 197,
+  number = 2,
+  pages = {685--692},
+  month = sep,
+  year = 2009,
+  doi = {10.1016/j.ejor.2008.07.015},
+  annote = {Proposed $g$-NSGA-II}
+}
+
+ +
+@article{MonAydStu2011soco,
+  author = { Marco A. {Montes de Oca}  and  Do\v{g}an Ayd{\i}n  and  Thomas St{\"u}tzle },
+  title = {An Incremental Particle Swarm for Large-Scale
+                  Continuous Optimization Problems: An Example of
+                  Tuning-in-the-loop (Re)Design of Optimization
+                  Algorithms},
+  journal = {Soft Computing},
+  year = 2011,
+  volume = 15,
+  number = 11,
+  pages = {2233--2255},
+  doi = {10.1007/s00500-010-0649-0}
+}
+
+ +
+@article{MonFraLiu2018bridge,
+  title = {Multi-criteria robust optimization framework for bridge
+                  adaptation under climate change},
+  author = {Mondoro, Alysson and Frangopol, Dan M. and Liu, Liang},
+  journal = {Structural Safety},
+  volume = 74,
+  pages = {14--23},
+  year = 2018
+}
+
+ +
+@article{MonGamRiz2005,
+  author = { Roberto Montemanni  and  L. M. Gambardella  and A. E. Rizzoli and
+                  A. V. Donati},
+  title = {Ant colony system for a dynamic vehicle routing problem},
+  journal = {Journal of Combinatorial Optimization},
+  year = 2005,
+  volume = 10,
+  pages = {327--343},
+  alias = {Mon++05}
+}
+
+ +
+@article{MonRandHen08:bias,
+  author = { James Montgomery  and  Marcus Randall  and  Tim Hendtlass },
+  title = {Solution bias in ant colony optimisation: {Lessons}
+                  for selecting pheromone models},
+  journal = {Computers \& Operations Research},
+  volume = 35,
+  number = 9,
+  year = 2008,
+  pages = {2728--2749},
+  doi = {10.1016/j.cor.2006.12.014}
+}
+
+ +
+@article{MonRifNev2014asoc,
+  author = { Elizabeth Montero  and  Mar{\'i}a-Cristina Riff  and Neveu, Bertrand},
+  title = {A Beginner's Buide to Tuning Methods},
+  journal = {Applied Soft Computing},
+  volume = 17,
+  pages = {39--51},
+  year = 2014,
+  doi = {10.1016/j.asoc.2013.12.017}
+}
+
+ +
+@article{MonStuBirDor2009tec,
+  author = { Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Marco Dorigo },
+  title = {Frankenstein's {PSO}: A Composite Particle Swarm
+                  Optimization Algorithm},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 13,
+  number = 5,
+  year = 2009,
+  pages = {1120--1132},
+  doi = {10.1109/TEVC.2009.2021465}
+}
+
+ +
+@article{MonVenSli2000:fgcs,
+  author = { Nicolas Monmarch{\'e}  and G. Venturini and M. Slimane},
+  title = {On how {\em pachycondyla apicalis} ants suggest a
+                  new search algorithm},
+  journal = {Future Generation Computer Systems},
+  year = 2000,
+  volume = 16,
+  number = 8,
+  pages = {937--946}
+}
+
+ +
+@article{Mor1993sim,
+  title = {Simulation of an adaptive behavior mechanism in an
+                  expert decision-maker},
+  volume = 23,
+  number = 1,
+  journal = {IEEE Transactions on Systems, Man, and Cybernetics},
+  author = {Morgan, Peter D.},
+  year = 1993,
+  pages = {65--76}
+}
+
+ +
+@article{Mor80,
+  title = {Reducing the size of the nondominated set: Pruning
+                  by clustering},
+  author = {J. N. Morse},
+  journal = {Computers \& Operations Research},
+  volume = 7,
+  number = {1-2},
+  pages = {55--66},
+  year = 1980
+}
+
+ +
+@article{MorDesLod2021machine,
+  title = {Machine-learning--based column selection for column
+                  generation},
+  author = {Morabit, Mouad and Desaulniers, Guy and  Andrea Lodi },
+  journal = {Transportation Science},
+  volume = 55,
+  number = 4,
+  pages = {815--831},
+  year = 2021,
+  publisher = {INFORMS},
+  keywords = {graph neural networks}
+}
+
+ +
+@article{MorGagGra09:ejor,
+  author = {Sara Morin and  Caroline Gagn{\'e}  and Marc Gravel},
+  title = {Ant colony optimization with a specialized pheromone
+                  trail for the car-sequencing problem},
+  volume = 197,
+  doi = {10.1016/j.ejor.2008.03.033},
+  abstract = { This paper studies the learning process in an ant
+                  colony optimization algorithm designed to solve the
+                  problem of ordering cars on an assembly line
+                  (car-sequencing problem). This problem has been
+                  shown to be {NP-hard} and evokes a great deal of
+                  interest among practitioners. Learning in an ant
+                  algorithm is achieved by using an artificial
+                  pheromone trail, which is a central element of this
+                  metaheuristic. Many versions of the algorithm are
+                  found in literature, the main distinction among them
+                  being the management of the pheromone
+                  trail. Nevertheless, few of them seek to perfect
+                  learning by modifying the internal structure of the
+                  trail. In this paper, a new pheromone trail
+                  structure is proposed that is specifically adapted
+                  to the type of constraints in the car-sequencing
+                  problem. The quality of the results obtained when
+                  solving three sets of benchmark problems is superior
+                  to that of the best solutions found in literature
+                  and shows the efficiency of the specialized trail.},
+  number = 3,
+  journal = {European Journal of Operational Research},
+  year = 2009,
+  keywords = {Ant colony {optimization,Car-sequencing}
+                  {problem,Pheromone} {trail,Scheduling}},
+  pages = {1185--1191}
+}
+
+ +
+@article{MorMerLarMilTor09,
+  author = {Mora, A. M. and  Juan-Juli{\'a}n Merelo  and  Jim{\'e}nez Laredo, Juan Luis  and Millan, C. and Torrecillas, J.},
+  title = {{CHAC}, a {MOACO} algorithm for computation of bi-criteria
+                  military unit path in the battlefield: Presentation and first
+                  results},
+  journal = {International Journal of Intelligent Systems},
+  volume = 24,
+  number = 7,
+  publisher = {Wiley Subscription Services, Inc., A Wiley Company},
+  pages = {818--843},
+  year = 2009
+}
+
+ +
+@article{MorMit1995,
+  title = {Exploratory designs for computational experiments},
+  author = {Morris, Max D. and Mitchell, Toby J.},
+  journal = {Journal of Statistical Planning and Inference},
+  year = 1995,
+  number = 3,
+  pages = {381--402},
+  volume = 43,
+  doi = {10.1016/0378-3758(94)00035-T},
+  keywords = {Bayesian prediction},
+  alias = {Morris1995}
+}
+
+ +
+@article{MosFon1990,
+  title = {Stochastic Versus Deterministic Update in Simulated Annealing},
+  author = { Pablo Moscato  and  Jos{\'e} F. Fontanari },
+  journal = {Physics Letters A},
+  volume = 146,
+  number = 4,
+  pages = {204--208},
+  year = 1990,
+  publisher = {Elsevier}
+}
+
+ +
+@article{MotMurOls1991,
+  author = {John Mote and Ishwar Murthy and David L. Olson},
+  title = {A parametric approach to solving bicriterion shortest path
+                  problems},
+  journal = {European Journal of Operational Research},
+  volume = 53,
+  number = 1,
+  pages = {81--92},
+  year = 1991,
+  doi = {10.1016/0377-2217(91)90094-C}
+}
+
+ +
+@article{MotOlsVen1988moprog,
+  author = {John Mote and David L. Olson and M. A. Venkataramanan},
+  doi = {10.1016/0895-7177(88)90085-4},
+  year = 1988,
+  publisher = {Elsevier {BV}},
+  volume = 10,
+  number = 10,
+  pages = {719--729},
+  title = {A comparative multiobjective programming study},
+  journal = {Mathematical and Computer Modelling},
+  keywords = {artificial DM, interactive},
+  annote = {The purpose of this study was to systematically evaluate a
+                  number of multiobjective programming concepts relative to
+                  reflection of utility, assurance of nondominated solutions
+                  and practicality for larger problems using conventional
+                  software. In the problem used, the nonlinear simulated DM
+                  utility function applied resulted in a nonextreme point
+                  solution. Very often, the preferred solution could end up
+                  being an extreme point solution, in which case the techniques
+                  relying upon LP concepts would work as well if not better
+                  than utilizing constrained objective attainments. The point
+                  is that there is no reason to expect linear or near linear
+                  utility.}
+}
+
+ +
+@article{MouDevHen2012,
+  author = {S\'ebastien Mouthuy and  Yves Deville  and van Hentenryck, Pascal },
+  title = {Constraint-based Very Large-Scale Neighborhood Search},
+  journal = {Constraints},
+  year = 2012,
+  volume = 17,
+  number = 2,
+  pages = {87--122},
+  doi = {10.1007/s10601-011-9114-7}
+}
+
+ +
+@article{MouKesDha2019,
+  author = {Mousin, Lucien and  Marie-El{\'e}onore Kessaci  and  Dhaenens, Clarisse },
+  title = {Exploiting Promising Sub-Sequences of Jobs to solve the No-Wait Flowshop Scheduling Problem},
+  journal = {Arxiv preprint arXiv:1903.09035},
+  year = 2019,
+  url = {http://arxiv.org/abs/1903.09035}
+}
+
+ +
+@article{MouPetMcC2014mopso,
+  author = {Al Moubayed, Noura and Petrovski, Andrei and  McCall, John },
+  title = {{$D^2MOPSO$}: {MOPSO} based on decomposition and dominance
+                  with archiving using crowding distance in objective and
+                  solution spaces},
+  journal = {Evolutionary Computation},
+  year = 2014,
+  volume = 22,
+  number = 1,
+  pages = {47--77}
+}
+
+ +
+@article{MouSlo1998:jgo,
+  author = { Vincent Mousseau  and  Roman S{\l}owi{\'n}ski },
+  year = 1998,
+  title = {Inferring an {ELECTRE TRI} model from assignment examples},
+  journal = {Journal of Global Optimization},
+  volume = 12,
+  number = 2,
+  pages = {157--174}
+}
+
+ +
+@article{MueSba2012ecj,
+  title = {Energy Landscapes of Atomic Clusters as Black Box
+                  Optimization Benchmarks},
+  author = {M\"uller, Christian L. and Sbalzarini, Ivos F.},
+  journal = {Evolutionary Computation},
+  year = 2012,
+  number = 4,
+  pages = {543--573},
+  volume = 20,
+  doi = {10.1162/EVCO_a_00086},
+  publisher = {MIT Press},
+  alias = {Mueller2012}
+}
+
+ +
+@article{Muhlenbein93,
+  author = { H. M{\"u}hlenbein  and  D. Schlierkamp-Voosen },
+  title = {Predictive models for the breeder genetic algorithm},
+  journal = {Evolutionary Computation},
+  year = 1993,
+  volume = 1,
+  number = 1,
+  pages = {25--49},
+  note = {},
+  keywords = {crossover, intermediate, line}
+}
+
+ +
+@article{MunSmi2020ec,
+  author = { Mario A. Mu{\~{n}}oz  and  Kate Smith{-}Miles },
+  title = {Generating New Space-Filling Test Instances for Continuous
+                  Black-Box Optimization},
+  doi = {10.1162/evco_a_00262},
+  year = 2020,
+  month = sep,
+  publisher = {{MIT} Press},
+  volume = 28,
+  number = 3,
+  pages = {379--404},
+  journal = {Evolutionary Computation}
+}
+
+ +
+@article{MunSunKirHal2015sel,
+  title = {Algorithm selection for black-box continuous optimization
+                  problems: a survey on methods and challenges},
+  author = { Mario A. Mu{\~{n}}oz  and Sun, Yuan and Kirley, Michael and
+                  Halgamuge, Saman K.},
+  journal = {Information Sciences},
+  volume = 317,
+  pages = {224--245},
+  year = 2015
+}
+
+ +
+@article{MunVilBaaSmi2018ismlc,
+  author = { Mario A. Mu{\~{n}}oz  and Villanova, Laura and Baatar, Davaatseren and  Kate Smith{-}Miles },
+  title = {Instance Spaces for Machine Learning Classification},
+  journal = {Machine Learning},
+  year = 2018,
+  volume = 107,
+  number = 1,
+  pages = {109--147},
+  doi = {10.1007/s10994-017-5629-5}
+}
+
+ +
+@article{NagKob2013,
+  author = {Yuichi Nagata and Shigenobu Kobayashi},
+  title = {A Powerful Genetic Algorithm Using Edge Assembly Crossover
+                  for the Traveling Salesman Problem},
+  journal = {INFORMS Journal on Computing},
+  year = 2013,
+  volume = 25,
+  number = 2,
+  pages = {346--363},
+  doi = {10.1287/ijoc.1120.0506},
+  keywords = {TSP, EAX, evolutionary algorithms},
+  abstract = {This paper presents a genetic algorithm (GA) for solving the
+                  traveling salesman problem (TSP). To construct a powerful GA,
+                  we use edge assembly crossover (EAX) and make substantial
+                  enhancements to it: (i) localization of EAX together with its
+                  efficient implementation and (ii) the use of a local search
+                  procedure in EAX to determine good combinations of building
+                  blocks of parent solutions for generating even better
+                  offspring solutions from very high-quality parent
+                  solutions. In addition, we develop (iii) an innovative
+                  selection model for maintaining population diversity at a
+                  negligible computational cost. Experimental results on
+                  well-studied TSP benchmarks demonstrate that the proposed GA
+                  outperforms state-of-the-art heuristic algorithms in finding
+                  very high-quality solutions on instances with up to 200,000
+                  cities. In contrast to the state-of-the-art TSP heuristics,
+                  which are all based on the Lin-Kernighan (LK) algorithm, our
+                  GA achieves top performance without using an LK-based
+                  algorithm.}
+}
+
+ +
+@article{NagRosMar2019:eo,
+  title = {High-performing heuristics to minimize flowtime in no-idle permutation flowshop},
+  author = {Marcelo S. Nagano and  Fernando L. Rossi and  N{\'a}dia J. Martarelli},
+  journal = {Engineering Optimization},
+  volume = 51,
+  number = 2,
+  pages = {185--198},
+  year = 2019
+}
+
+ +
+@article{NagSol2012,
+  author = {Yuichi Nagata and David Soler},
+  title = {A New Genetic Algorithm for the Asymmetric {TSP}},
+  journal = {Expert Systems with Applications},
+  year = 2012,
+  volume = 39,
+  number = 10,
+  pages = {8947--8953}
+}
+
+ +
+@article{NalOliHerSud2019,
+  title = {On the Analysis of Trajectory-Based Search Algorithms:
+                  When is it Beneficial to Reject Improvements?},
+  author = { Samadhi Nallaperuma  and  Oliveto, Pietro S.  and Heredia, Jorge P{\'e}rez and  Dirk Sudholt },
+  journal = {Algorithmica},
+  volume = 81,
+  number = 2,
+  pages = {858--885},
+  year = 2019,
+  publisher = {Springer}
+}
+
+ +
+@article{NanShaIsh2020reverse,
+  title = {Reverse strategy for non-dominated archiving},
+  author = {Nan, Yang and Shang, Ke and  Ishibuchi, Hisao  and He, Linjun},
+  journal = {{IEEE} Access},
+  year = 2020,
+  pages = {119458--119469},
+  volume = 8,
+  publisher = {IEEE}
+}
+
+ +
+@article{NarSetTan2016sc,
+  author = {Narukawa, Kaname and Setoguchi, Yu and Tanigaki, Yuki and
+                  Olhofer, Markus and Sendhoff, Bernhard and  Ishibuchi, Hisao },
+  title = {Preference representation using {Gaussian} functions on a
+                  hyperplane in evolutionary multi-objective optimization},
+  journal = {Soft Computing},
+  year = 2016,
+  volume = 20,
+  number = 7,
+  pages = {2733--2757},
+  month = jul,
+  doi = {10.1007/s00500-015-1674-9}
+}
+
+ +
+@article{NasVar2011optimx,
+  title = {Unifying Optimization Algorithms to Aid Software System
+                  Users: \rpackage{optimx} for \proglang{R}},
+  author = {Nash, John and Varadhan, Ravi},
+  journal = {Journal of Statistical Software},
+  year = 2011,
+  number = 9,
+  pages = {1--14},
+  volume = 43,
+  alias = {Nash2011}
+}
+
+ +
+@article{NawEnsHam83,
+  author = { M. Nawaz  and  Enscore, Jr, E.  and I. Ham},
+  title = {A Heuristic Algorithm for the $m$-Machine, $n$-Job
+                  Flow-Shop Sequencing Problem},
+  journal = {Omega},
+  year = 1983,
+  volume = 11,
+  number = 1,
+  pages = {91--95},
+  keywords = {NEH heuristic}
+}
+
+ +
+@article{NebLopGarCoe2023automopso,
+  author = { Nebro, Antonio J.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Jos{\'e} Garc{\'i}a-Nieto  and  Carlos A. {Coello Coello} },
+  title = {On the automatic design of multi-objective particle swarm
+                  optimizers: experimentation and analysis},
+  journal = {Swarm Intelligence},
+  year = 2023,
+  doi = {10.1007/s11721-023-00227-2},
+  abstract = {Research in multi-objective particle swarm optimizers
+                  (MOPSOs) progresses by proposing one new MOPSO at a time. In
+                  spite of the commonalities among different MOPSOs, it is
+                  often unclear which algorithmic components are crucial for
+                  explaining the performance of a particular MOPSO
+                  design. Moreover, it is expected that different designs may
+                  perform best on different problem families and identifying a
+                  best overall MOPSO is a challenging task. We tackle this
+                  challenge here by: (1) proposing AutoMOPSO, a flexible
+                  algorithmic template for designing MOPSOs with a design space
+                  that can instantiate thousands of potential MOPSOs; and (2)
+                  searching for good-performing MOPSO designs given a family of
+                  training problems by means of an automatic configuration tool
+                  (irace). We apply this automatic design methodology to
+                  generate a MOPSO that significantly outperforms two
+                  state-of-the-art MOPSOs on four well-known bi-objective
+                  problem families. We also identify the key design choices and
+                  parameters of the winning MOPSO by means of
+                  ablation. AutoMOPSO is publicly available as part of the
+                  jMetal framework.}
+}
+
+ +
+@article{NebLunAlb2008abyss,
+  author = { Nebro, Antonio J.  and F. Luna and  Alba, Enrique  and  Bernab{\'e} Dorronsoro  and  Durillo, Juan J.  and A. Beham},
+  title = {{AbYSS}: Adapting Scatter Search to Multiobjective Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 12,
+  number = 4,
+  month = aug,
+  year = 2008,
+  pages = {439--457},
+  doi = {10.1109/TEVC.2007.913109}
+}
+
+ +
+@article{NerCot2012,
+  author = {F. Nerri and  Carlos Cotta },
+  title = {Memetic algorithms and memetic computing optimization:
+                  A literature review},
+  journal = {Swarm and Evolutionary Computation},
+  year = 2012,
+  volume = 2,
+  pages = {1--14},
+  doi = {10.1016/j.swevo.2011.11.003}
+}
+
+ +
+@article{NeuSudWit2009:swarm,
+  author = { Frank Neumann  and  Dirk Sudholt  and  Carsten Witt },
+  title = {Analysis of different {MMAS} {ACO} algorithms on
+                  unimodal functions and plateaus},
+  journal = {Swarm Intelligence},
+  year = 2009,
+  volume = 3,
+  number = 1,
+  pages = {35--68}
+}
+
+ +
+@article{NeuWit2006:eccc,
+  author = { Frank Neumann  and  Carsten Witt },
+  title = {Runtime Analysis of a Simple Ant Colony Optimization
+                  Algorithm},
+  journal = {Electronic Colloquium on Computational Complexity (ECCC)},
+  volume = 13,
+  number = 084,
+  year = 2006
+}
+
+ +
+@article{NewSim1976cacm,
+  author = {Newell, Allen and Simon, Herbert A.},
+  title = {Computer Science as Empirical Inquiry: Symbols and Search},
+  year = 1976,
+  volume = 19,
+  number = 3,
+  issn = {0001-0782},
+  doi = {10.1145/360018.360022},
+  abstract = {Computer science is the study of the phenomena surrounding
+                  computers. The founders of this society understood this very
+                  well when they called themselves the Association for
+                  Computing Machinery. The machine-not just the hardware, but
+                  the programmed, living machine-is the organism we study.},
+  journal = {Communications of the ACM},
+  month = mar,
+  pages = {113--126},
+  numpages = 14,
+  keywords = {cognition, Turing, search, problem solving, symbols,
+                  heuristics, list processing, computer science, artificial
+                  intelligence, science, empirical}
+}
+
+ +
+@article{NguPriPro2012:eaai,
+  author = {Viet-Phuong Nguyen and  Christian Prins  and Caroline Prodhon},
+  title = {A Multi-start Iterated Local Search with Tabu List and Path Relinking
+               for the Two-echelon Location-routing Problem},
+  journal = {Engineering Applications of Artificial Intelligence},
+  year = 2012,
+  volume = 25,
+  number = 1,
+  pages = {56--71}
+}
+
+ +
+@article{NguReiRig2014,
+  title = {A review on simulation-based optimization methods applied to
+                  building performance analysis},
+  journal = {Applied Energy},
+  volume = 113,
+  pages = {1043--1058},
+  year = 2014,
+  doi = {10.1016/j.apenergy.2013.08.061},
+  author = {Anh-Tuan Nguyen and Sigrid Reiter and Philippe Rigo}
+}
+
+ +
+@article{NguYanBra2012:swec,
+  author = {Trung Thanh Nguyen and Shengxiang Yang and  J{\"u}rgen Branke },
+  title = {Evolutionary Dynamic Optimization: A Survey of the State of the Art},
+  journal = {Swarm and Evolutionary Computation},
+  year = 2012,
+  volume = 6,
+  pages = {1--24}
+}
+
+ +
+@article{NguZhaJohChe14:ec,
+  author = {Su Nguyen and  Zhang, Mengjie  and Mark Johnston and  Tan, Kay Chen },
+  title = {Genetic Programming for Evolving Due-Date Assignment Models in Job Shop Environments},
+  journal = {Evolutionary Computation},
+  year = 2014,
+  volume = 22,
+  number = 1,
+  pages = {105--138}
+}
+
+ +
+@article{NguZhaJohChe14:tec,
+  author = {Su Nguyen and  Zhang, Mengjie  and Mark Johnston and  Tan, Kay Chen },
+  title = {Automatic Design of Scheduling Policies for Dynamic Multi-objective Job Shop Scheduling via Cooperative Coevolution Genetic Programming},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2014,
+  volume = 18,
+  number = 2,
+  pages = {193--208}
+}
+
+ +
+@article{NigAkgGen++2017:aij,
+  author = {Peter Nightingale and \"Ozgu\"ur Akg\"un and  Ian P. Gent  and Christopher Jefferson and Ian Miguel and Patrick Spracklen},
+  title = {Automatically Improving Constraint Models in {Savile} {Row}},
+  journal = {Artificial Intelligence},
+  year = 2017,
+  volume = 251,
+  pages = {35--61}
+}
+
+ +
+@article{NinYou2019optunc,
+  author = {Chao Ning and Fengqi You},
+  title = {Optimization under uncertainty in the era of big data and
+                  deep learning: When machine learning meets mathematical
+                  programming},
+  journal = {Computers \& Chemical Engineering},
+  year = 2019,
+  volume = 125,
+  pages = {434--448},
+  doi = {10.1016/j.compchemeng.2019.03.034},
+  keywords = {Data-driven optimization, Decision making under uncertainty,
+                  Big data, Machine learning, Deep learning},
+  abstract = {This paper reviews recent advances in the field of
+                  optimization under uncertainty via a modern data lens,
+                  highlights key research challenges and promise of data-driven
+                  optimization that organically integrates machine learning and
+                  mathematical programming for decision-making under
+                  uncertainty, and identifies potential research
+                  opportunities. A brief review of classical mathematical
+                  programming techniques for hedging against uncertainty is
+                  first presented, along with their wide spectrum of
+                  applications in Process Systems Engineering. A comprehensive
+                  review and classification of the relevant publications on
+                  data-driven distributionally robust optimization, data-driven
+                  chance constrained program, data-driven robust optimization,
+                  and data-driven scenario-based optimization is then
+                  presented. This paper also identifies fertile avenues for
+                  future research that focuses on a closed-loop data-driven
+                  optimization framework, which allows the feedback from
+                  mathematical programming to machine learning, as well as
+                  scenario-based optimization leveraging the power of deep
+                  learning techniques. Perspectives on online learning-based
+                  data-driven multistage optimization with a
+                  learning-while-optimizing scheme are presented.}
+}
+
+ +
+@article{NisTanSugMiy2019item,
+  title = {Item listing optimization for e-commerce websites based on
+                  diversity},
+  author = {Nishimura, Naoki and Tanahashi, Kotaro and Suganuma, Koji and
+                  Miyama, Masamichi J. and Ohzeki, Masayuki},
+  journal = {Frontiers in Computer Science},
+  volume = 1,
+  pages = 2,
+  year = 2019,
+  publisher = {Frontiers},
+  keywords = {Quantum Annealing}
+}
+
+ +
+@article{Nitivatta96,
+  author = { Vilas Nitivattananon  and  Elaine C. Sadowski  and  Rafael G. Quimpo },
+  title = {Optimization of Water Supply System Operation},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 122,
+  number = 5,
+  pages = {374--384},
+  date = {1996-09/1996-10},
+  year = 1996,
+  month = sep # { / } # oct
+}
+
+ +
+@article{NogPinSub2017:ol,
+  author = {Bruno Nogueira and Rian G. S. Pinheiro and Anand Subramanian},
+  title = {A Hybrid Iterated Local Search Heuristic for the Maximum
+                  Weight Independent Set Problem},
+  journal = {Optimization Letters},
+  year = 2018,
+  volume = 12,
+  number = 3,
+  pages = {567--583},
+  doi = {10.1007/s11590-017-1128-7}
+}
+
+ +
+@article{NosAltBank2015promoting,
+  doi = {10.1126/science.aab2374},
+  year = 2015,
+  month = jun,
+  publisher = {American Association for the Advancement of Science ({AAAS})},
+  volume = 348,
+  number = 6242,
+  pages = {1422--1425},
+  author = {B. A. Nosek and G. Alter and G. C. Banks and D. Borsboom and
+                  S. D. Bowman and S. J. Breckler and S. Buck and
+                  C. D. Chambers and G. Chin and G. Christensen and
+                  M. Contestabile and A. Dafoe and E. Eich and J. Freese and
+                  R. Glennerster and D. Goroff and D. P. Green and B. Hesse and
+                  M. Humphreys and J. Ishiyama and D. Karlan and A. Kraut and
+                  A. Lupia and P. Mabry and T. Madon and N. Malhotra and
+                  E. Mayo-Wilson and M. McNutt and E. Miguel and E. L. Paluck
+                  and U. Simonsohn and C. Soderberg and B. A. Spellman and
+                  J. Turitto and G. VandenBos and S. Vazire and
+                  E. J. Wagenmakers and R. Wilson and T. Yarkoni},
+  title = {Promoting an open research culture},
+  journal = {Science}
+}
+
+ +
+@article{NosEbeHav2018preregistration,
+  author = {Nosek, Brian A. and Ebersole, Charles R. and DeHaven,
+                  Alexander C. and Mellor, David T.},
+  title = {The Preregistration Revolution},
+  volume = 115,
+  issn = {0027-8424, 1091-6490},
+  doi = {10.1073/pnas.1708274114},
+  language = {en},
+  number = 11,
+  journal = {Proceedings of the National Academy of Sciences},
+  month = mar,
+  year = 2018,
+  pages = {2600--2606},
+  abstract = {Progress in science relies in part on generating hypotheses
+                  with existing observations and testing hypotheses with new
+                  observations. This distinction between postdiction and
+                  prediction is appreciated conceptually but is not respected
+                  in practice. Mistaking generation of postdictions with
+                  testing of predictions reduces the credibility of research
+                  findings. However, ordinary biases in human reasoning, such
+                  as hindsight bias, make it hard to avoid this mistake. An
+                  effective solution is to define the research questions and
+                  analysis plan before observing the research outcomes--a
+                  process called preregistration. Preregistration distinguishes
+                  analyses and outcomes that result from predictions from those
+                  that result from postdictions. A variety of practical
+                  strategies are available to make the best possible use of
+                  preregistration in circumstances that fall short of the ideal
+                  application, such as when the data are preexisting. Services
+                  are now available for preregistration across all disciplines,
+                  facilitating a rapid increase in the practice. Widespread
+                  adoption of preregistration will increase distinctiveness
+                  between hypothesis generation and hypothesis testing and will
+                  improve the credibility of research findings.}
+}
+
+ +
+@article{NouAnd1998,
+  title = {A Comparison of Simulated Annealing Cooling Strategies},
+  author = { Yaghout Nourani  and  Bjarne Andresen },
+  journal = {Journal of Physics A},
+  volume = 31,
+  number = 41,
+  pages = {8373--8385},
+  year = 1998,
+  publisher = {IOP Publishing}
+}
+
+ +
+@article{NowSmu1996:ms,
+  author = {Nowicki, Eugeniusz and Smutnicki, Czeslaw},
+  title = {A Fast Taboo Search Algorithm for the Job Shop Problem},
+  journal = {Management Science},
+  year = 1996,
+  volume = 42,
+  number = 6,
+  pages = {797--813}
+}
+
+ +
+@article{NowSmu1996ejor,
+  title = {A fast tabu search algorithm for the permutation flow-shop problem},
+  author = {Nowicki, Eugeniusz and Smutnicki, Czes{\l}aw},
+  journal = {European Journal of Operational Research},
+  volume = 91,
+  number = 1,
+  pages = {160--175},
+  year = 1996
+}
+
+ +
+@article{OSC2015estimating,
+  title = {Estimating the reproducibility of psychological science },
+  author = {{Open Science Collaboration}},
+  journal = {Science},
+  volume = 349,
+  number = 6251,
+  numpages = 7,
+  pages = {aac4716},
+  year = 2015,
+  doi = {10.1126/science.aac4716}
+}
+
+ +
+@article{OchVer2018,
+  title = {Mapping the global structure of {TSP} fitness landscapes},
+  author = { Gabriela Ochoa  and  Veerapen, Nadarajen },
+  journal = {Journal of Heuristics},
+  volume = 24,
+  number = 3,
+  pages = {265--294},
+  year = 2018,
+  publisher = {Springer}
+}
+
+ +
+@article{OddCesPolSmi2008,
+  author = { Angelo Oddi  and  Amadeo Cesta  and  Nicola Policella  and  Stephen F. Smith },
+  title = {Combining Variants of Iterative Flattening Search},
+  journal = {Engineering Applications of Artificial Intelligence},
+  year = 2008,
+  volume = 21,
+  number = 5,
+  pages = {683--690}
+}
+
+ +
+@article{OddCesPolSmi2010,
+  author = { Angelo Oddi  and  Amadeo Cesta  and  Nicola Policella  and  Stephen F. Smith },
+  title = {Iterative Flattening Search for Resource Constrained Scheduling},
+  journal = {Journal of Intelligent Manufacturing},
+  year = 2010,
+  volume = 21,
+  number = 1,
+  pages = {17--30}
+}
+
+ +
+@article{OgbuSmith1990,
+  author = { F. A. Ogbu  and  David K. Smith },
+  title = {The Application of the Simulated Annealing Algorithm to the Solution of the {n/m/C} Max Flowshop Problem},
+  journal = {Computers \& Operations Research},
+  volume = 17,
+  number = 3,
+  pages = {243--253},
+  year = 1990,
+  publisher = {Elsevier}
+}
+
+ +
+@article{OhlTho07tsptw_sa,
+  author = { Jeffrey W. Ohlmann  and  Barrett W. Thomas },
+  title = {A Compressed-Annealing Heuristic for the Traveling
+                  Salesman Problem with Time Windows},
+  journal = {INFORMS Journal on Computing},
+  volume = 19,
+  number = 1,
+  year = 2007,
+  issn = {1526-5528},
+  pages = {80--90},
+  doi = {10.1287/ijoc.1050.0145},
+  publisher = {{INFORMS}},
+  address = {Institute for Operations Research and the Management
+                  Sciences (INFORMS), Linthicum, Maryland, USA}
+}
+
+ +
+@article{OliHeYao2007,
+  title = {Time complexity of evolutionary algorithms for combinatorial
+                  optimization: A decade of results},
+  author = { Oliveto, Pietro S.  and He, Jun and  Xin Yao },
+  journal = {International Journal of Automation and Computing},
+  volume = 4,
+  number = 3,
+  pages = {281--293},
+  year = 2007,
+  publisher = {Springer}
+}
+
+ +
+@article{OliWit2015,
+  author = { Oliveto, Pietro S.  and  Carsten Witt },
+  title = {Improved time complexity analysis of the Simple Genetic Algorithm},
+  journal = {Theoretical Computer Science},
+  volume = 605,
+  pages = {21--41},
+  year = 2015,
+  doi = {10.1016/j.tcs.2015.01.002}
+}
+
+ +
+@article{Ols1992review,
+  author = {Olson, David L.},
+  title = {Review of Empirical Studies in Multiobjective Mathematical
+                  Programming: Subject Reflection of Nonlinear Utility and
+                  Learning},
+  journal = {Decision Sciences},
+  volume = 23,
+  number = 1,
+  pages = {1--20},
+  year = 1992,
+  keywords = {Decision Analysis, Human Information Processing, Linear
+                  Programming},
+  doi = {10.1111/j.1540-5915.1992.tb00374.x},
+  abstract = {Multiple objective programming provides a means of
+                  aiding decision makers facing complex decisions where
+                  trade-offs among conflicting objectives must be
+                  reconciled. Interactive multiobjective programming provides a
+                  means for decision makers to learn what these trade-offs
+                  involve, while the mathematical program generates solutions
+                  that seek improvement of the implied utility of the decision
+                  maker. A variety of multiobjective programming techniques
+                  have been presented in the multicriteria decision-making
+                  literature. This study reviews published studies with human
+                  subjects where some of these techniques were applied. While
+                  all of the techniques have the ability to support decision
+                  makers under conditions of multiple objectives, a number of
+                  features in applying these systems have been tested by these
+                  studies. A general evolution of techniques is traced,
+                  starting with methods relying upon linear combinations of
+                  value, to more recent methods capable of reflecting nonlinear
+                  trade-offs of value. Support of nonlinear utility and
+                  enhancing decision-maker learning are considered.}
+}
+
+ +
+@article{OlsLok2013joh,
+  author = {Roland Olsson and  Arne L{\o}kketangen },
+  title = {Using Automatic Programming to Generate State-of-the-art
+                  Algorithms for Random 3-{SAT}},
+  journal = {Journal of Heuristics},
+  volume = 19,
+  number = 5,
+  year = 2013,
+  pages = {819--844},
+  doi = {10.1007/s10732-013-9226-x},
+  annote = {Uses evolution but it is not genetic programming, nor
+                  grammatical evolution.}
+}
+
+ +
+@article{Oltean05,
+  author = {Mihai Oltean},
+  title = {Evolving Evolutionary Algorithms Using Linear Genetic
+                  Programming},
+  journal = {Evolutionary Computation},
+  year = 2005,
+  volume = 13,
+  number = 3,
+  pages = {387--410},
+  doi = {10.1162/1063656054794815}
+}
+
+ +
+@article{OneRya2001tec,
+  title = {Grammatical Evolution},
+  author = { O'Neill, Michael  and Ryan, Conor},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 5,
+  number = 4,
+  pages = {349--358},
+  year = 2001
+}
+
+ +
+@article{Ormsbee89,
+  author = { Lindell E. Ormsbee  and  Thomas M. Walski  and  Donald V. Chase  and  W. W. Sharp },
+  title = {Methodology for improving pump operation efficiency},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 1989,
+  volume = 115,
+  number = 2,
+  pages = {148--164},
+  note = {}
+}
+
+ +
+@article{Ormsbee94,
+  author = { Lindell E. Ormsbee  and  Kevin E. Lansey },
+  title = {Optimal Control of Water Supply Pumping Systems},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 1994,
+  volume = 120,
+  number = 2,
+  pages = {237--252}
+}
+
+ +
+@article{Ormsbee95,
+  author = { Lindell E. Ormsbee  and  Srinivasa L. Reddy },
+  title = {Nonlinear Heuristic for Pump Operations},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 121,
+  number = 4,
+  pages = {302--309},
+  date = {1995-07/1995-08},
+  year = 1995,
+  month = jul # { / } # aug
+}
+
+ +
+@article{OroJac2002,
+  title = {Analysis of Static Simulated Annealing Algorithms},
+  author = { Jeffrey E. Orosz  and  Sheldon H. Jacobson },
+  journal = {Journal of Optimization Theory and Applications},
+  volume = 115,
+  number = 1,
+  pages = {165--182},
+  year = 2002,
+  publisher = {Springer}
+}
+
+ +
+@article{OsmPot89,
+  author = { Ibrahim H. Osman  and  Chris N. Potts },
+  title = {Simulated Annealing for Permutation Flow-Shop
+                  Scheduling},
+  journal = {Omega},
+  year = 1989,
+  volume = 17,
+  number = 6,
+  pages = {551--557}
+}
+
+ +
+@article{OwMor1988:ijpr,
+  author = {Ow, P. S. and Morton, T. E.},
+  title = {Filtered Beam Search in Scheduling},
+  journal = {International Journal of Production Research},
+  year = 1988,
+  volume = 26,
+  pages = {297--307},
+  alias = {OwMor88}
+}
+
+ +
+@article{OzeKar2007:jors,
+  author = { G{\"u}l {\"O}zerol  and  Esra Karasakal },
+  title = {Interactive outranking approaches for multicriteria
+                  decision-making problems with imprecise information},
+  journal = {Journal of the Operational Research Society},
+  year = 2007,
+  volume = 59,
+  pages = {1253--1268}
+}
+
+ +
+@article{PadRin1991siamr,
+  title = {A branch-and-cut algorithm for the resolution of large-scale symmetric traveling salesman problems},
+  author = {Padberg, Manfred and Rinaldi, Giovanni},
+  journal = {SIAM Review},
+  volume = 33,
+  number = 1,
+  pages = {60--100},
+  year = 1991
+}
+
+ +
+@article{PagStu2016,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Speeding up Local Search for the Insert Neighborhood in the Weighted Tardiness Permutation Flowshop Problem},
+  journal = {Optimization Letters},
+  year = 2017,
+  volume = 11,
+  pages = {1283--1292},
+  doi = {10.1007/s11590-016-1086-5}
+}
+
+ +
+@article{PagStu2019:ejor,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Stochastic Local Search Algorithms
+                  for Permutation Flowshop Problems},
+  journal = {European Journal of Operational Research},
+  year = 2019,
+  volume = 276,
+  pages = {409--421},
+  issue = 2,
+  doi = {10.1016/j.ejor.2019.01.018},
+  keywords = {EMILI},
+  abstract = {Stochastic local search methods are at the core of many
+                  effective heuristics for tackling different permutation
+                  flowshop problems (PFSPs). Usually, such algorithms require a
+                  careful, manual algorithm engineering effort to reach high
+                  performance. An alternative to the manual algorithm
+                  engineering is the automated design of effective SLS
+                  algorithms through building flexible algorithm frameworks and
+                  using automatic algorithm configuration techniques to
+                  instantiate high-performing algorithms. In this paper, we
+                  automatically generate new high-performing algorithms for
+                  some of the most widely studied variants of the PFSP. More in
+                  detail, we (i) developed a new algorithm framework, EMILI,
+                  that implements algorithm-specific and problem-specific
+                  building blocks; (ii) define the rules of how to compose
+                  algorithms from the building blocks; and (iii) employ an
+                  automatic algorithm configuration tool to search for high
+                  performing algorithm configurations. With these ingredients,
+                  we automatically generate algorithms for the PFSP with the
+                  objectives makespan, total completion time and total
+                  tardiness, which outperform the best algorithms obtained by a
+                  manual algorithm engineering process.}
+}
+
+ +
+@article{PagStu2020:itor,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Evaluating the impact of grammar complexity in automatic algorithm design},
+  journal = {International Transactions in Operational Research},
+  pages = {1--26},
+  doi = {10.1111/itor.12902},
+  year = 2020
+}
+
+ +
+@article{PagStu2021:orp,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Automatic design of hybrid stochastic local search algorithms
+                  for permutation flowshop problems with additional
+                  constraints},
+  journal = {Operations Research Perspectives},
+  year = 2021,
+  volume = 8,
+  pages = 100180,
+  doi = {10.1016/j.orp.2021.100180},
+  abstract = {Automatic design of stochastic local search algorithms has
+                  been shown to be very effective in generating algorithms for
+                  the permutation flowshop problem for the most studied
+                  objectives including makespan, flowtime and total
+                  tardiness. The automatic design system uses a configuration
+                  tool to combine algorithmic components following a set of
+                  rules defined as a context-free grammar. In this paper we use
+                  the same system to tackle two of the most studied additional
+                  constraints for these objectives: sequence dependent setup
+                  times and no-idle constraint. Additional components have been
+                  added to adapt the system to the new problems while keeping
+                  intact the grammar structure and the experimental setup. The
+                  experiments show that the generated algorithms outperform the
+                  state of the art in each case.}
+}
+
+ +
+@article{PajBlaHerMar2021archiving,
+  title = {A Comparison of Archiving Strategies for Characterization of
+                  Nearly Optimal Solutions under Multi-Objective Optimization},
+  author = {Pajares, Alberto and Blasco, Xavier and Herrero, Juan Manuel
+                  and Mart{\'i}nez, Miguel A.},
+  journal = {Mathematics},
+  year = 2021,
+  doi = {10.3390/math9090999},
+  volume = 9,
+  number = 9,
+  pages = {999},
+  abstract = {In a multi-objective optimization problem, in addition to
+                  optimal solutions, multimodal and/or nearly optimal
+                  alternatives can also provide additional useful information
+                  for the decision maker. However, obtaining all nearly optimal
+                  solutions entails an excessive number of
+                  alternatives. Therefore, to consider the nearly optimal
+                  solutions, it is convenient to obtain a reduced set, putting
+                  the focus on the potentially useful alternatives. These
+                  solutions are the alternatives that are close to the optimal
+                  solutions in objective space, but which differ significantly
+                  in the decision space. To characterize this set, it is
+                  essential to simultaneously analyze the decision and
+                  objective spaces. One of the crucial points in an
+                  evolutionary multi-objective optimization algorithm is the
+                  archiving strategy. This is in charge of keeping the solution
+                  set, called the archive, updated during the optimization
+                  process. The motivation of this work is to analyze the three
+                  existing archiving strategies proposed in the literature
+                  (ArchiveUpdate$P_{Q,\epsilon}D_{xy}$, Archive\_nevMOGA, and
+                  targetSelect) that aim to characterize the potentially useful
+                  solutions. The archivers are evaluated on two benchmarks and
+                  in a real engineering example. The contribution clearly shows
+                  the main differences between the three archivers. This
+                  analysis is useful for the design of evolutionary algorithms
+                  that consider nearly optimal solutions.},
+  keywords = {multi-objective optimization; nearly optimal solutions;
+                  non-epsilon dominance; multimodality; decision space
+                  diversity; archiving strategy; evolutionary algorithm;
+                  non-linear parametric identification}
+}
+
+ +
+@article{PalGooSor2014,
+  author = {Daniel {Palhazi Cuervo} and Peter Goos and  Kenneth S{\"o}rensen  and Emely Arr{\'{a}}iz},
+  title = {An Iterated Local Search Algorithm for the Vehicle Routing Problem
+               with Backhauls},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 237,
+  number = 2,
+  pages = {454--464}
+}
+
+ +
+@article{Palubeckis2006,
+  title = {Iterated tabu search for the unconstrained binary quadratic
+                  optimization problem},
+  author = {Palubeckis, Gintaras},
+  journal = {Informatica},
+  volume = 17,
+  number = 2,
+  pages = {279--296},
+  year = 2006,
+  publisher = {Institute of Mathematics and Informatics},
+  doi = {10.15388/Informatica.2006.138}
+}
+
+ +
+@article{PanRui2012ejor,
+  title = {Local Search Methods for the Flowshop Scheduling Problem with Flowtime Minimization},
+  author = {Pan, Quan-Ke and  Rub{\'e}n Ruiz },
+  journal = {European Journal of Operational Research},
+  volume = 222,
+  number = 1,
+  pages = {31--43},
+  year = 2012
+}
+
+ +
+@article{PanRui2013cor,
+  title = {A Comprehensive Review and Evaluation of Permutation
+                  Flowshop Heuristics to Minimize Flowtime},
+  author = {Pan, Quan-Ke and  Rub{\'e}n Ruiz },
+  journal = {Computers \& Operations Research},
+  volume = 40,
+  number = 1,
+  pages = {117--128},
+  year = 2013,
+  alias = {PanRui2012}
+}
+
+ +
+@article{PanRuiAlf2017:cor,
+  author = { Quan-Ke Pan  and  Rub{\'e}n Ruiz  and  Pedro Alfaro-Fern{\'a}ndez },
+  title = {Iterated Search Methods for Earliness and Tardiness Minimization in Hybrid Flowshops with Due Windows},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  volume = 80,
+  pages = {50--60}
+}
+
+ +
+@article{PanTasLia2008,
+  title = {A Discrete Differential Evolution Algorithm for the
+                  Permutation Flowshop Scheduling Problem },
+  journal = {Computers and Industrial Engineering},
+  volume = 55,
+  number = 4,
+  pages = {795 -- 816},
+  year = 2008,
+  author = {Quan-Ke Pan and Mehmet Fatih Tasgetiren and Yun-Chia
+                  Liang}
+}
+
+ +
+@article{PanWanZha2008,
+  year = 2008,
+  journal = {International Journal of Advanced Manufacturing Technology},
+  volume = 38,
+  number = {7-8},
+  title = {An improved iterated greedy algorithm for the
+                  no-wait flow shop scheduling problem with makespan
+                  criterion},
+  author = {Pan, Quan-Ke and Wang, Ling and Zhao, Bao-Hua},
+  pages = {778--786}
+}
+
+ +
+@article{PanYang2009,
+  title = {A survey on transfer learning},
+  author = {Pan, Sinno Jialin and Yang, Qiang},
+  journal = {IEEE Transactions on Knowledge and Data Engineering},
+  volume = 22,
+  number = 10,
+  pages = {1345--1359},
+  year = 2009
+}
+
+ +
+@article{PaqSchStu07:aor,
+  author = { Lu{\'i}s Paquete  and  Tommaso Schiavinotto  and  Thomas St{\"u}tzle },
+  title = {On Local Optima in Multiobjective Combinatorial
+                  Optimization Problems},
+  journal = {Annals of Operations Research},
+  year = 2007,
+  volume = 156,
+  pages = {83--97},
+  doi = {10.1007/s10479-007-0230-0},
+  keywords = {Pareto local search, PLS},
+  abstract = {In this article, local optimality in multiobjective
+                  combinatorial optimization is used as a baseline for
+                  the design and analysis of two iterative improvement
+                  algorithms. Both algorithms search in a neighborhood
+                  that is defined on a collection of sets of feasible
+                  solutions and their acceptance criterion is based on
+                  outperformance relations. Proofs of the soundness
+                  and completeness of these algorithms are given.}
+}
+
+ +
+@article{PaqStu06:mqap,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {A study of stochastic local search algorithms for
+                  the biobjective {QAP} with correlated flow matrices},
+  journal = {European Journal of Operational Research},
+  year = 2006,
+  volume = 169,
+  number = 3,
+  alias = {lpaquete:14},
+  pages = {943--959}
+}
+
+ +
+@article{PaqStu09:cor,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Design and analysis of stochastic local search for
+                  the multiobjective traveling salesman problem},
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  volume = 36,
+  number = 9,
+  pages = {2619--2631},
+  doi = {10.1016/j.cor.2008.11.013}
+}
+
+ +
+@article{ParDoeHar2009,
+  author = {Parragh, S. N. and  Karl F. Doerner  and  Richard F. Hartl  and  Xavier Gandibleux },
+  title = {A heuristic two-phase solution approach for the
+                  multi-objective dial-a-ride problem},
+  journal = {Networks},
+  volume = 54,
+  number = 4,
+  pages = {227--242},
+  year = 2009
+}
+
+ +
+@article{ParJoh1997,
+  author = {Rebecca Parsons and Mark Johnson},
+  title = {A Case Study in Experimental Design Applied to Genetic Algorithms with Applications to {DNA} Sequence Assembly},
+  journal = {American Journal of Mathematical and Management Sciences},
+  publisher = {Taylor \& Francis},
+  volume = 17,
+  number = {3-4},
+  pages = {369--396},
+  year = 1997,
+  doi = {10.1080/01966324.1997.10737444}
+}
+
+ +
+@article{ParKim98,
+  author = {Moon-Won Park and Yeong-Dae Kim},
+  title = {A systematic procedure for setting parameters in simulated annealing algorithms},
+  journal = {Computers \& Operations Research},
+  volume = 25,
+  number = 3,
+  pages = {207--217},
+  year = 1998,
+  doi = {10.1016/S0305-0548(97)00054-3}
+}
+
+ +
+@article{ParLopFre2002:ieee-tec,
+  author = {R. S. Parpinelli and H. S. Lopes and A. A. Freitas},
+  title = {Data Mining with an Ant Colony Optimization Algorithm},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002,
+  volume = 6,
+  number = 4,
+  pages = {321--332}
+}
+
+ +
+@article{ParQuo1995antlr,
+  author = {Parr, Terence J. and Quong, Russell W.},
+  title = {{ANTLR}: A predicated-{LL} (k) parser generator},
+  journal = {Software --- Practice \& Experience},
+  year = 1995,
+  volume = 25,
+  number = 7,
+  pages = {789--810}
+}
+
+ +
+@article{ParVas2007:ejor,
+  author = { R. O. Parreiras  and  J. A. Vascocelos },
+  title = {A multiplicative version of {PROMETHEE II} applied to
+                   multiobjective optimization problems},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 183,
+  pages = {729--740}
+}
+
+ +
+@article{Paul2010,
+  title = {Comparative performance of tabu search and simulated annealing heuristics for the quadratic assignment problem},
+  author = { Paul, Gerald },
+  journal = {Operations Research Letters},
+  volume = 38,
+  number = 6,
+  pages = {577--581},
+  year = 2010,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Pearl2019:cacm,
+  title = {The seven tools of causal inference, with reflections on machine learning},
+  author = { Judea Pearl },
+  journal = {Communications of the ACM},
+  volume = 62,
+  number = 3,
+  pages = {54--60},
+  year = 2019
+}
+
+ +
+@article{PedNesCan2011:asc,
+  author = {Mart{\'i}n Pedemonte and Sergio Nesmachnow and H{\'e}ctor Cancela},
+  title = {A survey on parallel ant colony optimization},
+  journal = {Applied Soft Computing},
+  volume = 11,
+  number = 8,
+  pages = {5181--5197},
+  year = 2011
+}
+
+ +
+@article{PelBirStu2012:si,
+  author = { Paola Pellegrini  and  Mauro Birattari  and  Thomas St{\"u}tzle },
+  title = {A Critical Analysis of Parameter Adaptation in Ant Colony Optimization},
+  journal = {Swarm Intelligence},
+  year = 2012,
+  volume = 6,
+  number = 1,
+  pages = {23--48},
+  doi = {10.1007/s11721-011-0061-0}
+}
+
+ +
+@article{PelCasPes2012its,
+  author = { Paola Pellegrini  and L. Castelli and R. Pesenti},
+  title = {Metaheuristic algorithms for the simultaneous slot
+                  allocation problem},
+  journal = {IET Intelligent Transport Systems},
+  year = 2012,
+  volume = 6,
+  number = 4,
+  pages = {453--462},
+  month = dec,
+  doi = {10.1049/iet-its.2011.0179}
+}
+
+ +
+@article{PelMasBirStu14,
+  author = { Paola Pellegrini  and  Franco Mascia  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {On the Sensitivity of Reactive Tabu Search to its
+                  Meta-parameters},
+  journal = {Soft Computing},
+  year = 2014,
+  volume = 18,
+  number = 11,
+  pages = {2177--2190},
+  doi = {10.1007/s00500-013-1192-6}
+}
+
+ +
+@article{PenSubOch2013:joh,
+  author = {Puca Huachi {Vaz Penna} and  Anand Subramanian  and  Luiz Satoru Ochi },
+  title = {An Iterated Local Search Heuristic for the Heterogeneous Fleet Vehicle Routing Problem},
+  journal = {Journal of Heuristics},
+  year = 2013,
+  volume = 19,
+  number = 2,
+  pages = {201--232}
+}
+
+ +
+@article{Per2020challenge,
+  title = {Challenge to scientists: does your ten-year-old code still
+                  run?},
+  author = {Jeffrey M. Perkel},
+  journal = {Nature},
+  volume = 584,
+  pages = {556--658},
+  year = 2020,
+  doi = {10.1038/d41586-020-02462-7},
+  keywords = {reproducibility; software engineering; ReScience C; Ten Years
+                  Reproducibility Challenge; code reusability}
+}
+
+ +
+@article{PerLopStu2015si,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Ant colony optimization on a limited budget of evaluations},
+  journal = {Swarm Intelligence},
+  year = 2015,
+  doi = {10.1007/s11721-015-0106-x},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2015-004},
+  volume = 9,
+  number = {2-3},
+  pages = {103--124}
+}
+
+ +
+@article{PerRuiNes2018,
+  author = {P{\'{e}}res, Matias and Ruiz, Germ{\'{a}}n and Nesmachnow,
+                  Sergio and Olivera, Ana C.},
+  journal = {Applied Soft Computing},
+  keywords = {Multiobjective evolutionary
+                  algorithms,Pollution,Simulation,Traffic flow},
+  title = {Multiobjective evolutionary optimization of traffic flow and
+                  pollution in {Montevideo}, {Uruguay}},
+  year = 2018,
+  alias = {Peres2018},
+  volume = 70,
+  pages = {472--485},
+  publisher = {Elsevier}
+}
+
+ +
+@article{PesUchAraRod2010,
+  author = {A. Pessoa and E. Uchoa and M. Arag{\~a}o and R. Rodrigues},
+  title = {Exact Algorithm over an Arc-time-indexed formulation for Parallel Machine Scheduling Problems},
+  journal = {Mathematical Programming Computation},
+  year = 2010,
+  volume = 2,
+  number = {3--4},
+  pages = {259--290}
+}
+
+ +
+@article{Pesant98tsptw,
+  author = { Gilles Pesant  and  Michel Gendreau  and  Jean-Yves Potvin  and J.-M. Rousseau },
+  title = {An Exact Constraint Logic Programming Algorithm
+  for the Traveling Salesman Problem with Time Windows},
+  journal = {Transportation Science},
+  year = 1998,
+  volume = 32,
+  pages = {12--29}
+}
+
+ +
+@article{Pet1998boeing,
+  author = {Charles W. Petit},
+  title = {Touched by nature: putting evolution to work on the assembly
+                  line},
+  journal = {U.S. News \& World Report},
+  volume = 125,
+  number = 4,
+  month = jul,
+  year = 1998,
+  pages = {43--45},
+  url = {http://dynamics.org/~altenber/IN_THE_MEDIA/US_NEWS.7-27-98/index.html},
+  annote = {Evolutionary optimization of turbine design of the
+                  Boeing~777~GE}
+}
+
+ +
+@article{PetHarHarLan2018gi,
+  author = {Petke, Justyna and Haraldsson, Saemundur O. and Harman, Mark and  Langdon, William B.  and White, David R. and John R. Woodward},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Genetic Improvement of Software: A Comprehensive Survey},
+  year = 2018,
+  volume = 22,
+  number = 3,
+  pages = {415--432},
+  doi = {10.1109/TEVC.2017.2693219}
+}
+
+ +
+@article{PetZil2006parallel,
+  author = {Petrik, Marek and  Shlomo Zilberstein },
+  journal = {Annals of Mathematics and Artificial Intelligence},
+  number = 1,
+  pages = {85--106},
+  title = {Learning parallel portfolios of algorithms},
+  volume = 48,
+  year = 2006,
+  keywords = {algorithm selection}
+}
+
+ +
+@article{Pezeshk96,
+  author = { S. Pezeshk  and  O. J. Helweg },
+  title = {Adaptative Search Optimisation in reducing pump operation
+                  costs},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 122,
+  number = 1,
+  pages = {57--63},
+  date = {1996-01/1996-02},
+  year = 1996,
+  month = jan # { / } # feb
+}
+
+ +
+@article{PheKok2003interactive,
+  title = {An interactive evolutionary metaheuristic for multiobjective
+                  combinatorial optimization},
+  author = {Phelps, Selcen and  Murat K{\"o}ksalan },
+  journal = {Management Science},
+  volume = 49,
+  number = 12,
+  pages = {1726--1738},
+  year = 2003
+}
+
+ +
+@article{PieKhuSav2007investigation,
+  title = {An investigation on preference order ranking scheme for
+                  multiobjective evolutionary optimization},
+  author = {di Pierro, Francesco and Khu, Soon-Thiam and  Dragan A. Savic },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 11,
+  number = 1,
+  pages = {17--45},
+  year = 2007
+}
+
+ +
+@article{PinVinSin2020improving,
+  title = {Improving Reproducibility in Machine Learning Research (A
+                  Report from the {NeurIPS} 2019 Reproducibility Program)},
+  author = {Joelle Pineau and Philippe Vincent-Lamarre and Koustuv Sinha
+                  and Vincent Larivière and Alina Beygelzimer and Florence
+                  d'Alché-Buc and Emily Fox and Hugo Larochelle},
+  year = 2020,
+  journal = {Arxiv preprint arXiv:2003.12206 [cs.LG]},
+  url = {https://arxiv.org/abs/2003.12206}
+}
+
+ +
+@article{Pis2005where,
+  author = { David Pisinger },
+  journal = {Computers \& Operations Research},
+  number = 9,
+  pages = {2271--2284},
+  title = {Where are the hard knapsack problems?},
+  volume = 32,
+  year = 2005
+}
+
+ +
+@article{PisRop2007:cor,
+  author = { David Pisinger  and  Stefan Ropke },
+  title = {A General Heuristic for Vehicle Routing Problems},
+  journal = {Computers \& Operations Research},
+  year = 2007,
+  volume = 34,
+  number = 8,
+  pages = {2403--2435}
+}
+
+ +
+@article{PitAlmDoe06,
+  author = {Rapeepan Pitakaso and  Christian Almeder  and  Karl F. Doerner  and  Richard F. Hartl },
+  title = {Combining exact and population-based methods for the
+                  Constrained Multilevel Lot Sizing Problem},
+  journal = {International Journal of Production Research},
+  year = 2006,
+  volume = 44,
+  number = 22,
+  pages = {4755--4771}
+}
+
+ +
+@article{PitAlmDoe07,
+  author = {Rapeepan Pitakaso and  Christian Almeder  and  Karl F. Doerner  and  Richard F. Hartl },
+  title = {A {\MaxMinAntSystem} for unconstrained multi-level
+                  lot-sizing problems},
+  volume = 34,
+  number = 9,
+  journal = {Computers \& Operations Research},
+  year = 2007,
+  keywords = {Ant colony optimization, Material requirements
+                  planning, Multi-level lot-sizing, {Wagner-Whitin}
+                  algorithm},
+  pages = {2533--2552},
+  doi = {10.1016/j.cor.2005.09.022},
+  abstract = { In this paper, we present an ant-based algorithm
+                  for solving unconstrained multi-level lot-sizing
+                  problems called ant system for multi-level
+                  lot-sizing algorithm {(ASMLLS).} We apply a hybrid
+                  approach where we use ant colony optimization in
+                  order to find a good lot-sizing sequence, i.e. a
+                  sequence of the different items in the product
+                  structure in which we apply a modified
+                  {Wagner-Whitin} algorithm for each item
+                  separately. Based on the setup costs each ant
+                  generates a sequence of items. Afterwards a simple
+                  single-stage lot-sizing rule is applied with
+                  modified setup costs. This modification of the setup
+                  costs depends on the position of the item in the
+                  lot-sizing sequence, on the items which have been
+                  lot-sized before, and on two further parameters,
+                  which are tried to be improved by a systematic
+                  search. For small-sized problems {ASMLLS} is among
+                  the best algorithms, but for most medium- and
+                  large-sized problems it outperforms all other
+                  approaches regarding solution quality as well as
+                  computational time.}
+}
+
+ +
+@article{Ples2018repro,
+  doi = {10.3389/fninf.2017.00076},
+  year = 2018,
+  month = jan,
+  publisher = {Frontiers Media {SA}},
+  volume = 11,
+  author = {Hans E. Plesser},
+  title = {Reproducibility vs.\ Replicability: A Brief History of a
+                  Confused Terminology},
+  journal = {Frontiers in Neuroinformatics}
+}
+
+ +
+@article{PorGonAllHsu2017:ejor,
+  author = {Daniel Porumbel and Gilles Goncalves and Hamid Allaoui and Tient{\'e} Hsu},
+  title = {Iterated Local Search and Column Generation to solve Arc-Routing as a Permutation Set-Covering Problem},
+  journal = {European Journal of Operational Research},
+  year = 2017,
+  volume = 256,
+  number = 2,
+  pages = {349--367}
+}
+
+ +
+@article{PorParDoa++2013,
+  author = {Juan Porta and Jorge Parapar and Ram{\'o}n Doallo and Vasco Barbosa and In{\'e}s Sant{\'e} and Rafael Crecente and Carlos D{\'i}az},
+  title = {A Population-based Iterated Greedy Algorithm for the Delimitation and Zoning of Rural Settlements},
+  journal = {Computers, Environment and Urban Systems},
+  year = 2013,
+  volume = 39,
+  pages = {12--26}
+}
+
+ +
+@article{PotBen96vrptw_gene,
+  author = { Jean-Yves Potvin  and  S. Bengio },
+  title = {The Vehicle Routing Problem with Time Windows Part
+                  {II}: Genetic Search},
+  journal = {INFORMS Journal on Computing},
+  year = 1996,
+  volume = 8,
+  pages = {165--172}
+}
+
+ +
+@article{Pra2009jwrpm,
+  author = { T. Devi Prasad },
+  title = {Design of pumped water distribution networks with
+                  storage},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2009,
+  volume = 136,
+  number = 4,
+  pages = {129--136}
+}
+
+ +
+@article{PraPac2015:joh,
+  author = {Marco Pranzo and D. Pacciarelli},
+  title = {An Iterated Greedy Metaheuristic for the Blocking Job Shop Scheduling Problem},
+  journal = {Journal of Heuristics},
+  year = 2016,
+  volume = 22,
+  number = 4,
+  pages = {587--611},
+  doi = {10.1007/s10732-014-9279-5}
+}
+
+ +
+@article{PraAveLem2019learning,
+  author = {Prates, Marcelo and Avelar, Pedro H. C. and Lemos, Henrique
+                  and Lamb, Luis C. and Vardi, Moshe Y.},
+  title = {Learning to Solve {NP-Complete} Problems: A Graph Neural
+                  Network for Decision {TSP}},
+  journal = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  year = 2019,
+  volume = 33,
+  number = 01,
+  pages = {4731--4738},
+  month = jul,
+  issn = {2159-5399},
+  doi = {10.1609/aaai.v33i01.33014731},
+  publisher = {Association for the Advancement of Artificial Intelligence
+                  (AAAI)}
+}
+
+ +
+@article{PriKumSug2023trial,
+  author = {Kenneth V. Price and Abhishek Kumar and  Ponnuthurai N. Suganthan },
+  title = {Trial-based dominance for comparing both the speed and
+                  accuracy of stochastic optimizers with standard
+                  non-parametric tests},
+  journal = {Swarm and Evolutionary Computation},
+  year = 2023,
+  volume = 78,
+  pages = 101287,
+  doi = {10.1016/j.swevo.2023.101287},
+  keywords = {Benchmarking, Two-variable non-parametric tests, Evolutionary
+                  algorithms, Dominance, Stochastic optimization, Numerical
+                  optimization, Mann-Whitney test}
+}
+
+ +
+@article{Prim1957,
+  title = {Shortest connection networks and some generalizations},
+  author = {Prim, Robert Clay},
+  journal = {{Bell System Technical Journal}},
+  volume = 36,
+  number = 6,
+  pages = {1389--1401},
+  year = 1957
+}
+
+ +
+@article{ProBisBou2018tuna,
+  title = {Tunability: Importance of Hyperparameters of Machine Learning
+                  Algorithms},
+  author = {Probst, Philipp and  Bernd Bischl  and Boulesteix,
+                  Anne-Laure},
+  journal = {Arxiv preprint arXiv:1802.09596},
+  year = 2018,
+  url = {https://arxiv.org/abs/1802.09596},
+  keywords = {parameter importance}
+}
+
+ +
+@article{ProBisBou2019tuna,
+  author = {Probst, Philipp and  Bernd Bischl  and Boulesteix,
+                  Anne-Laure},
+  title = {Tunability: Importance of Hyperparameters of Machine Learning
+                  Algorithms},
+  journal = {Journal of Machine Learning Research},
+  year = 2019,
+  volume = 20,
+  number = 53,
+  pages = {1--32}
+}
+
+ +
+@article{ProMue2012,
+  title = {Design of computer experiments: space filling and beyond},
+  author = {Pronzato, Luc and M\"uller, Werner G.},
+  journal = {Statistics and Computing},
+  year = 2012,
+  number = 3,
+  pages = {681--701},
+  volume = 22,
+  keywords = {Kriging; Entropy; Design of experiments; Space-filling;
+                  Sphere packing; Maximin design; Minimax design},
+  alias = {Pronzato2012}
+}
+
+ +
+@article{Psaraftis1995:aor,
+  author = {Harilaos N. Psaraftis},
+  title = {Dynamic Vehicle Routing: Status and Prospects},
+  journal = {Annals of Operations Research},
+  year = 1995,
+  volume = 61,
+  pages = {143--164}
+}
+
+ +
+@article{PukHei2006,
+  title = {Optimizing heuristic search in forest planning},
+  author = {Pukkala, Timo and Heinonen, Tero},
+  journal = {Nonlinear Analysis: Real World Applications},
+  volume = 7,
+  number = 5,
+  year = 2006,
+  pages = {1284--1297},
+  publisher = {Elsevier}
+}
+
+ +
+@article{PulTac2009aqme,
+  title = {A self-adaptive multi-engine solver for quantified {Boolean}
+                  formulas},
+  author = {Pulina, Luca and Tacchella, Armando},
+  journal = {Constraints},
+  volume = 14,
+  number = 1,
+  pages = {80--116},
+  year = 2009
+}
+
+ +
+@article{PurFle2007tec,
+  title = {On the Evolutionary Optimization of Many Conflicting
+                  Objectives},
+  author = { Robin C. Purshouse  and  Peter J. Fleming },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2007,
+  doi = {10.1109/TEVC.2007.910138},
+  volume = 11,
+  number = 6,
+  pages = {770--784}
+}
+
+ +
+@article{QiMaLiu2014moead,
+  title = {{MOEA/D} with adaptive weight adjustment},
+  author = {Qi, Yutao and Ma, Xiaoliang and Liu, Fang and Jiao, Licheng
+                  and Sun, Jianyong and Wu, Jianshe},
+  journal = {Evolutionary Computation},
+  year = 2014,
+  number = 2,
+  pages = {231--264},
+  volume = 22,
+  publisher = {MIT Press},
+  annote = {Uses an external population},
+  doi = {10.1162/EVCO_a_00109}
+}
+
+ +
+@article{QuiReeKel2017direct,
+  title = {Direct policy search for robust multi-objective management of
+                  deeply uncertain socio-ecological tipping points},
+  author = {Quinn, Julianne D. and  Patrick M. Reed  and Keller, Klaus},
+  journal = {Environmental Modelling \& Software},
+  volume = 92,
+  pages = {125--141},
+  year = 2017,
+  publisher = {Elsevier}
+}
+
+ +
+@article{RadRuiBor2009,
+  title = {New High Performing Heuristics for Minimizing Makespan in Permutation Flowshops},
+  author = { Shahriar Farahmand Rad  and  Rub{\'e}n Ruiz  and  Naser Boroojerdian },
+  journal = {Omega},
+  year = 2009,
+  number = 2,
+  pages = {331--345},
+  volume = 37
+}
+
+ +
+@article{Raj93,
+  title = {Heuristic algorithm for scheduling in a flowshop to
+                  minimize total flowtime},
+  author = {C. Rajendran},
+  journal = {International Journal of Production Economics},
+  volume = 29,
+  number = 1,
+  pages = {65--73},
+  year = 1993
+}
+
+ +
+@article{RajZie04,
+  author = {C. Rajendran and  H. Ziegler },
+  title = {Ant-colony algorithms for permutation flowshop
+                  scheduling to minimize makespan/total flowtime of
+                  jobs},
+  journal = {European Journal of Operational Research},
+  volume = 155,
+  number = 2,
+  pages = {426--438},
+  year = 2004
+}
+
+ +
+@article{RajZie1997,
+  author = {C. Rajendran and  H. Ziegler },
+  title = {An efficient heuristic for scheduling in a flowshop to
+                  minimize total weighted flowtime of jobs},
+  journal = {European Journal of Operational Research},
+  volume = 103,
+  number = 1,
+  pages = {129--138},
+  year = 1997,
+  issn = {0377 -- 2217},
+  doi = {10.1016/S0377-2217(96)00273-1}
+}
+
+ +
+@article{RamBir2020appsci,
+  title = {Automatic Design of Collective Behaviors for Robots that Can Display and Perceive Colors},
+  author = { David Garz{\'o}n Ramos  and  Mauro Birattari },
+  journal = {Applied Sciences},
+  volume = 10,
+  number = 13,
+  pages = {4654},
+  year = 2020,
+  publisher = {Multidisciplinary Digital Publishing Institute}
+}
+
+ +
+@article{RamMirSer2021foodplan,
+  author = {Ramos-Pérez, Juan-Manuel and Miranda, Gara and Segredo,
+                  Eduardo and León, Coromoto and Rodríguez-León, Casiano},
+  title = {Application of Multi-Objective Evolutionary Algorithms for
+                  Planning Healthy and Balanced School Lunches},
+  journal = {Mathematics},
+  month = dec,
+  publisher = {{MDPI} {AG}},
+  volume = 9,
+  year = 2021,
+  number = 1,
+  pages = 80,
+  doi = {10.3390/math9010080},
+  abstract = {A multi-objective formulation of the Menu Planning Problem,
+                  which is termed the Multi-objective Menu Planning Problem, is
+                  presented herein. Menu planning is of great interest in the
+                  health field due to the importance of proper nutrition in
+                  today's society, and particularly, in school
+                  canteens. In addition to considering the cost of the meal
+                  plan as the classic objective to be minimized, we also
+                  introduce a second objective aimed at minimizing the degree
+                  of repetition of courses and food groups that a particular
+                  meal plan consists of. The motivation behind this particular
+                  multi-objective formulation is to offer a meal plan that is
+                  not only affordable but also varied and balanced from a
+                  nutritional standpoint. The plan is designed for a given
+                  number of days and ensures that the specific nutritional
+                  requirements of school-age children are satisfied. The main
+                  goal of the current work is to demonstrate the
+                  multi-objective nature of the said formulation, through a
+                  comprehensive experimental assessment carried out over a set
+                  of multi-objective evolutionary algorithms applied to
+                  different instances. At the same time, we are also interested
+                  in validating the multi-objective formulation by performing
+                  quantitative and qualitative analyses of the solutions
+                  attained when solving it. Computational results show the
+                  multi-objective nature of the said formulation, as well as
+                  that it allows suitable meal plans to be obtained.}
+}
+
+ +
+@article{RamMonMor2011jors,
+  title = {Extending the use of scenario planning and {MCDA} for the
+                  evaluation of strategic options},
+  author = {Ram, Camelia and Montibeller, Gilberto and Morton, Alec},
+  journal = {Journal of the Operational Research Society},
+  volume = 62,
+  number = 5,
+  pages = {817--829},
+  year = 2011
+}
+
+ +
+@article{RaoSal2007hydroinf,
+  author = { Zhengfu Rao  and Elad Salomons},
+  title = {Development of a real-time, near-optimal control
+                  process for water-distribution networks},
+  journal = { Journal of Hydroinformatics },
+  year = 2007,
+  volume = 9,
+  number = 1,
+  doi = {10.2166/hydro.2006.015},
+  pages = {25--37}
+}
+
+ +
+@article{RarUzs2001:joh,
+  author = {Ronald L. Rardin and Reha Uzsoy},
+  title = {Experimental Evaluation of Heuristic Optimization Algorithms: A Tutorial},
+  journal = {Journal of Heuristics},
+  year = 2001,
+  volume = 7,
+  number = 3,
+  pages = {261--304}
+}
+
+ +
+@article{RasMusKar2019vrp,
+  author = {Jussi Rasku and  Musliu, Nysret  and Tommi K{\"a}rkk{\"a}inen},
+  title = {On automatic algorithm configuration of vehicle routing
+                  problem solvers},
+  journal = {Journal on Vehicle Routing Algorithms},
+  year = 2019,
+  month = feb,
+  volume = 2,
+  number = {1-4},
+  pages = {1--22},
+  doi = {10.1007/s41604-019-00010-9},
+  keywords = {irace, SMAC, GGA, REVAC, VRP}
+}
+
+ +
+@article{Rech2000case,
+  title = {Case studies in evolutionary experimentation and computation},
+  author = { Rechenberg, Ingo },
+  journal = {Computer Methods in Applied Mechanics and Engineering},
+  volume = 186,
+  number = {2-4},
+  pages = {125--140},
+  year = 2000,
+  doi = {10.1016/S0045-7825(99)00381-3},
+  publisher = {Elsevier}
+}
+
+ +
+@article{ReeEre2004jors,
+  author = { Colin R. Reeves  and Eremeev, A. V.},
+  title = {Statistical analysis of local search landscapes},
+  journal = {Journal of the Operational Research Society},
+  pages = {687--693},
+  volume = 55,
+  number = 7,
+  year = 2004,
+  epub = {http://www.jstor.org/stable/4102015}
+}
+
+ +
+@article{ReeGon1989imcdm,
+  title = {A comparison of two interactive {MCDM} procedures},
+  author = {Reeves, Gary R. and Gonzalez, Juan J.},
+  journal = {European Journal of Operational Research},
+  volume = 41,
+  number = 2,
+  pages = {203--209},
+  year = 1989,
+  doi = {10.1016/0377-2217(89)90385-8},
+  publisher = {Elsevier},
+  keywords = {artificial DM, interactive}
+}
+
+ +
+@article{ReeHadHerKas2013water,
+  title = {Evolutionary multiobjective optimization in water resources:
+                  The past, present, and future},
+  author = { Patrick M. Reed  and  David Hadka  and Herman, Jonathan D. and  Kasprzyk, Joseph R.  and  Kollat, Joshua B. },
+  journal = {Advances in Water Resources},
+  volume = 51,
+  pages = {438--456},
+  year = 2013
+}
+
+ +
+@article{CheLiYao2019,
+  title = {Standing on the shoulders of giants: Seeding search-based
+                  multi-objective optimization with prior knowledge for
+                  software service composition},
+  author = {Chen, Tao and  Li, Miqing  and  Xin Yao },
+  journal = {Information and Software Technology},
+  year = 2019,
+  pages = {155--175},
+  volume = 114,
+  publisher = {Elsevier},
+  annote = {Example of deteroriation in archiving}
+}
+
+ +
+@article{RehZaeFisRud2022bench,
+  author = {Rehbach, Frederik and  Martin Zaefferer  and  Andreas Fischbach  and  G{\"u}nther Rudolph  and  Thomas Bartz-Beielstein },
+  title = {Benchmark-Driven Configuration of a Parallel Model-Based
+                  Optimization Algorithm},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2022,
+  volume = 26,
+  number = 6,
+  pages = {1365--1379},
+  doi = {10.1109/TEVC.2022.3163843}
+}
+
+ +
+@article{Rei91a,
+  author = { Gerhard Reinelt },
+  title = {{TSPLIB} --- A Traveling Salesman Problem Library},
+  journal = {ORSA Journal on Computing},
+  year = 1991,
+  volume = 3,
+  number = 4,
+  pages = {376--384}
+}
+
+ +
+@article{ReiDoeHar04:cor,
+  author = { Marc Reimann  and  Karl F. Doerner  and  Richard F. Hartl },
+  title = {{D}-ants: {Savings} based ants divide and conquer
+                  the vehicle routing problems},
+  journal = {Computers \& Operations Research},
+  year = 2004,
+  volume = 31,
+  number = 4,
+  pages = {563--591}
+}
+
+ +
+@article{ReiLau06,
+  author = { Marc Reimann  and  Marco Laumanns },
+  title = {Savings based ant colony optimization for the
+                  capacitated minimum spanning tree problem},
+  volume = 33,
+  number = 6,
+  journal = {Computers \& Operations Research},
+  year = 2006,
+  keywords = {Ant colony Optimization, Capacitated minimum
+                  spanning tree problem},
+  pages = {1794--1822},
+  doi = {10.1016/j.cor.2004.11.019},
+  abstract = { The problem of connecting a set of client nodes
+                  with known demands to a root node through a minimum
+                  cost tree network, subject to capacity constraints
+                  on all links is known as the capacitated minimum
+                  spanning tree {(CMST)} problem. As the problem is
+                  {NP-hard,} we propose a hybrid ant colony
+                  optimization {(ACO)} algorithm to tackle it
+                  heuristically. The algorithm exploits two important
+                  problem characteristics: (i) the {CMST} problem is
+                  closely related to the capacitated vehicle routing
+                  problem {(CVRP),} and (ii) given a clustering of
+                  client nodes that satisfies capacity constraints,
+                  the solution is to find a {MST} for each cluster,
+                  which can be done exactly in polynomial time. Our
+                  {ACO} exploits these two characteristics of the
+                  {CMST} by a solution construction originally
+                  developed for the {CVRP.} Given the {CVRP} solution,
+                  we then apply an implementation of Prim's algorithm
+                  to each cluster to obtain a feasible {CMST}
+                  solution. Results from a comprehensive computational
+                  study indicate the efficiency and effectiveness of
+                  the proposed approach.}
+}
+
+ +
+@article{RenFenKeZha2010:cie,
+  author = {Zhi-Gang Ren and Zu-Ren Feng and Liang-Jun Ke and Zhao-Jun Zhang},
+  title = {New Ideas for Applying Ant Colony Optimization to the Set Covering Problem},
+  journal = {Computers and Industrial Engineering},
+  year = 2010,
+  volume = 58,
+  number = 4,
+  pages = {774--784}
+}
+
+ +
+@article{ReyCoe06,
+  author = {M. Reyes-Sierra  and  Carlos A. {Coello Coello} },
+  title = {Multi-objective particle swarm optimizers: A survey of the
+                  state-of-the-art},
+  year = 2006,
+  journal = {International Journal of Computational Intelligence Research},
+  volume = 2,
+  number = 3,
+  pages = {287--308}
+}
+
+ +
+@article{Reynolds1987,
+  author = {Craig W. Reynolds},
+  title = {Flocks, Herds, and Schools: A Distributed Behavioral Model},
+  journal = {{ACM} Computer Graphics},
+  volume = 21,
+  number = 4,
+  pages = {25--34},
+  year = 1987,
+  alias = {Reynolds1987}
+}
+
+ +
+@article{RezAraMeh2022anchoring,
+  title = {Analyzing anchoring bias in attribute weight elicitation of
+                  {SMART}, {Swing}, and best-worst method},
+  doi = {10.1111/itor.13171},
+  journal = {International Transactions in Operational Research},
+  author = {Rezaei, Jafar and Arab, Alireza and Mehregan, Mohammadreza},
+  year = 2022,
+  keywords = {anchoring bias, best-worst method, cognitive bias, MADM,
+                  multi-attribute weighting, SMART, Swing}
+}
+
+ +
+@article{RezHej2005,
+  title = {Flowshop-scheduling Problems with Makespan Criterion: A Review},
+  author = {S. Reza Hejazi and S. Saghafian},
+  journal = {International Journal of Production Research},
+  year = 2005,
+  number = 14,
+  pages = {2895--2929},
+  volume = 43
+}
+
+ +
+@article{RibComTor2011:omega,
+  title = {An iterated greedy algorithm for the flowshop
+                  scheduling problem with blocking},
+  journal = {Omega},
+  volume = 39,
+  number = 3,
+  pages = {293 -- 301},
+  year = 2011,
+  author = {Imma Ribas and Ramon Companys and Xavier
+                  Tort-Martorell}
+}
+
+ +
+@article{RibComTor2013,
+  author = {Imma Ribas and Ramon Companys and Xavier Tort-Martorell},
+  title = {An Efficient Iterated Local Search Algorithm for the Total Tardiness Blocking Flow Shop Problem},
+  journal = {International Journal of Production Research},
+  year = 2013,
+  volume = 51,
+  number = 17,
+  pages = {5238--5252}
+}
+
+ +
+@article{RibUrr2007,
+  author = { Celso C. Ribeiro  and Sebasti{\'a}n Urrutia},
+  title = {Heuristics for the Mirrored Traveling Tournament Problem},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 179,
+  number = 3,
+  pages = {775--787}
+}
+
+ +
+@article{RicBea2004:joh,
+  author = {A. J. Richmond and  John E. Beasley },
+  title = {An Iterative Construction Heuristic for the Ore Selection Problem},
+  journal = {Journal of Heuristics},
+  year = 2004,
+  volume = 10,
+  number = 2,
+  pages = {153--167}
+}
+
+ +
+@article{Rice1976,
+  author = {John R. Rice},
+  title = {The Algorithm Selection Problem},
+  journal = {Advances in Computers},
+  year = 1976,
+  volume = 15,
+  pages = {65--118},
+  doi = {10.1016/S0065-2458(08)60520-3},
+  abstract = {The problem of selecting an effective algorithm arises in a
+                  wide variety of situations. This chapter starts with a
+                  discussion on abstract models: the basic model and associated
+                  problems, the model with selection based on features, and the
+                  model with variable performance criteria. One objective of
+                  this chapter is to explore the applicability of the
+                  approximation theory to the algorithm selection
+                  problem. There is an intimate relationship here and that the
+                  approximation theory forms an appropriate base upon which to
+                  develop a theory of algorithm selection methods. The
+                  approximation theory currently lacks much of the necessary
+                  machinery for the algorithm selection problem. There is a
+                  need to develop new results and apply known techniques to
+                  these new circumstances. The final pages of this chapter form
+                  a sort of appendix, which lists 15 specific open problems and
+                  questions in this area. There is a close relationship between
+                  the algorithm selection problem and the general optimization
+                  theory. This is not surprising since the approximation
+                  problem is a special form of the optimization problem. Most
+                  realistic algorithm selection problems are of moderate to
+                  high dimensionality and thus one should expect them to be
+                  quite complex. One consequence of this is that most
+                  straightforward approaches (even well-conceived ones) are
+                  likely to lead to enormous computations for the best
+                  selection. The single most important part of the solution of
+                  a selection problem is the appropriate choice of the form for
+                  selection mapping. It is here that theories give the least
+                  guidance and that the art of problem solving is most
+                  crucial.}
+}
+
+ +
+@article{RivAfrPri2015:coa,
+  author = {Juan Carlos Rivera and H. Murat Afsar and  Christian Prins },
+  title = {A Multistart Iterated Local Search for the Multitrip Cumulative Capacitated
+               Vehicle Routing Problem},
+  journal = {Computational Optimization and Applications},
+  year = 2015,
+  volume = 61,
+  number = 1,
+  pages = {159--187}
+}
+
+ +
+@article{RivYanLop2021tweet,
+  author = { Rivadeneira, Luc{\'i}a  and  Yang, Jian-Bo  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Predicting tweet impact using a novel evidential reasoning
+                  prediction method},
+  journal = {Expert Systems with Applications},
+  year = 2021,
+  volume = 169,
+  pages = 114400,
+  month = may,
+  doi = {10.1016/j.eswa.2020.114400},
+  abstract = {This study presents a novel evidential reasoning (ER)
+                  prediction model called MAKER-RIMER to examine how different
+                  features embedded in Twitter posts (tweets) can predict the
+                  number of retweets achieved during an electoral campaign. The
+                  tweets posted by the two most voted candidates during the
+                  official campaign for the 2017 Ecuadorian Presidential
+                  election were used for this research. For each tweet, five
+                  features including type of tweet, emotion, URL, hashtag, and
+                  date are identified and coded to predict if tweets are of
+                  either high or low impact. The main contributions of the new
+                  proposed model include its suitability to analyse tweet
+                  datasets based on likelihood analysis of data. The model is
+                  interpretable, and the prediction process relies only on the
+                  use of available data. The experimental results show that
+                  MAKER-RIMER performed better, in terms of misclassification
+                  error, when compared against other predictive machine
+                  learning approaches. In addition, the model allows observing
+                  which features of the candidates' tweets are linked to high
+                  and low impact. Tweets containing allusions to the contender
+                  candidate, either with positive or negative connotations,
+                  without hashtags, and written towards the end of the
+                  campaign, were persistently those with the highest
+                  impact. URLs, on the other hand, is the only variable that
+                  performs differently for the two candidates in terms of
+                  achieving high impact. MAKER-RIMER can provide campaigners of
+                  political parties or candidates with a tool to measure how
+                  features of tweets are predictors of their impact, which can
+                  be useful to tailor Twitter content during electoral
+                  campaigns.},
+  keywords = {Evidential reasoning rule,Belief rule-based inference,Maximum
+                  likelihood data analysis,Twitter,Retweet,Prediction}
+}
+
+ +
+@article{Rob1995statcomp,
+  author = {Robert, C. P.},
+  title = {Simulation of truncated normal variables},
+  journal = {Statistics and Computing},
+  year = 1995,
+  volume = 5,
+  number = 2,
+  pages = {121--125},
+  month = jun
+}
+
+ +
+@article{RomKraArn2012protein,
+  author = {P. A. Romero and A. Krause and F. H. Arnold},
+  title = {Navigating the Protein Fitness Landscape with {Gaussian}
+                  Processes},
+  doi = {10.1073/pnas.1215251110},
+  year = 2012,
+  month = dec,
+  volume = 110,
+  number = 3,
+  pages = {E193--E201},
+  journal = {Proceedings of the National Academy of Sciences},
+  keywords = {Combinatorial Black-box Expensive}
+}
+
+ +
+@article{RomSan1991,
+  title = {A Theoretical Framework for Simulated Annealing},
+  author = { Fabio Romeo  and  Alberto Sangiovanni-Vincentelli },
+  journal = {Algorithmica},
+  volume = 6,
+  number = {1-6},
+  pages = {302--345},
+  year = 1991,
+  publisher = {Springer}
+}
+
+ +
+@article{Roos2001science,
+  title = {Bioinformatics--trying to swim in a sea of data},
+  author = {Roos, David S.},
+  journal = {Science},
+  volume = 291,
+  number = 5507,
+  pages = {1260--1261},
+  year = 2001,
+  publisher = {American Association for the Advancement of Science}
+}
+
+ +
+@article{RopPis2006:ejor,
+  author = { Stefan Ropke  and  David Pisinger },
+  title = {A Unified Heuristic for a Large Class of Vehicle Routing Problems
+               with Backhauls},
+  journal = {European Journal of Operational Research},
+  year = 2006,
+  volume = 171,
+  number = 3,
+  pages = {750--775}
+}
+
+ +
+@article{RopPis2006:ts,
+  author = { Stefan Ropke  and  David Pisinger },
+  title = {An Adaptive Large Neighborhood Search Heuristic for the Pickup and Delivery Problme with Time Windows},
+  journal = {Transportation Science},
+  year = 2006,
+  volume = 40,
+  number = 4,
+  pages = {455--472}
+}
+
+ +
+@article{Ros2014mutual,
+  title = {Mutual Information between Discrete and Continuous Data Sets},
+  author = {Ross, Brian C.},
+  doi = {10.1371/journal.pone.0087357},
+  journal = {PLoS One},
+  publisher = {Public Library of Science},
+  year = 2014,
+  month = feb,
+  volume = 9,
+  pages = {1--5},
+  number = 2,
+  abstract = {Mutual information (MI) is a powerful method for detecting
+                  relationships between data sets. There are accurate methods
+                  for estimating MI that avoid problems with ``binning'' when
+                  both data sets are discrete or when both data sets are
+                  continuous. We present an accurate, non-binning MI estimator
+                  for the case of one discrete data set and one continuous data
+                  set. This case applies when measuring, for example, the
+                  relationship between base sequence and gene expression level,
+                  or the effect of a cancer drug on patient survival time. We
+                  also show how our method can be adapted to calculate the
+                  Jensen-Shannon divergence of two or more data sets.}
+}
+
+ +
+@article{RosKleWol1990,
+  title = {Temperature measurement and equilibrium dynamics of simulated annealing placements},
+  author = {Rose, Jonathan and Klebsch, Wolfgang and Wolf, J{\"u}rgen},
+  journal = {IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems},
+  volume = 9,
+  number = 3,
+  pages = {253--259},
+  year = 1990
+}
+
+ +
+@article{Rothberg2007,
+  title = {An evolutionary algorithm for polishing mixed integer programming solutions},
+  author = {Rothberg, Edward},
+  journal = {INFORMS Journal on Computing},
+  volume = 19,
+  number = 4,
+  pages = {534--541},
+  year = 2007,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{Rothman1985,
+  title = {Nonlinear inversion, statistical mechanics, and residual statics estimation},
+  author = {Rothman, Daniel H.},
+  journal = {Geophysics},
+  volume = 50,
+  number = 12,
+  pages = {2784--2796},
+  year = 1985,
+  publisher = {Society of Exploration Geophysicists}
+}
+
+ +
+@article{Rothman1986,
+  title = {Automatic estimation of large residual statics corrections},
+  author = {Rothman, Daniel H.},
+  journal = {Geophysics},
+  volume = 51,
+  number = 2,
+  pages = {332--346},
+  year = 1986,
+  publisher = {Society of Exploration Geophysicists}
+}
+
+ +
+@article{Roy2010:ejor,
+  author = { Bernard Roy },
+  title = {Robustness in operational research and decision
+                  aiding: A multi-faceted issue},
+  journal = {European Journal of Operational Research},
+  volume = 200,
+  number = 3,
+  pages = {629--638},
+  year = 2010,
+  doi = {10.1016/j.ejor.2008.12.036}
+}
+
+ +
+@article{RudSchGri2016coa,
+  author = { G{\"u}nther Rudolph  and  Oliver Sch{\"u}tze  and Christian Grimme and Christian Dom{\'i}nguez-Medina and  Heike Trautmann },
+  title = {Optimal averaged {Hausdorff} archives for bi-objective
+                  problems: theoretical and numerical results},
+  journal = {Computational Optimization and Applications},
+  volume = 64,
+  number = 2,
+  year = 2016,
+  pages = {589--618}
+}
+
+ +
+@article{Rudolph1994:tnn,
+  title = {Convergence analysis of canonical genetic algorithms},
+  author = { G{\"u}nther Rudolph },
+  journal = {IEEE Transactions on Neural Networks},
+  volume = 5,
+  number = 1,
+  pages = {96--101},
+  year = 1994
+}
+
+ +
+@article{RuiMar05,
+  author = { Rub{\'e}n Ruiz  and  C. Maroto },
+  title = {A Comprehensive Review and Evaluation of Permutation
+                  Flow\-shop Heuristics},
+  journal = {European Journal of Operational Research},
+  year = 2005,
+  volume = 165,
+  number = 2,
+  pages = {479--494}
+}
+
+ +
+@article{RuiMar06,
+  author = { Rub{\'e}n Ruiz  and  C. Maroto  and Javier Alcaraz},
+  title = {Two new robust genetic algorithms for the flowshop scheduling problem},
+  journal = {Omega},
+  publisher = {Elsevier},
+  volume = 34,
+  number = 5,
+  pages = {461--476},
+  year = 2006,
+  doi = {10.1016/j.omega.2004.12.006}
+}
+
+ +
+@article{RuiSabLuq2015wasfga,
+  author = { Ruiz, Ana Bel{\'e}n  and  Rub{\'{e}}n Saborido  and  Mariano Luque },
+  title = {A preference-based evolutionary algorithm for multiobjective
+                  optimization: the weighting achievement scalarizing function
+                  genetic algorithm},
+  volume = 62,
+  doi = {10.1007/s10898-014-0214-y},
+  abstract = {When solving multiobjective optimization problems,
+                  preference-based evolutionary multiobjective optimization
+                  (EMO) algorithms introduce preference information into an
+                  evolutionary algorithm in order to focus the search for
+                  objective vectors towards the region of interest of the
+                  Pareto optimal front. In this paper, we suggest a
+                  preference-based EMO algorithm called weighting achievement
+                  scalarizing function genetic algorithm (WASF-GA), which
+                  considers the preferences of the decision maker (DM)
+                  expressed by means of a reference point. The main purpose of
+                  WASF-GA is to approximate the region of interest of the
+                  Pareto optimal front determined by the reference point, which
+                  contains the Pareto optimal objective vectors that obey the
+                  preferences expressed by the DM in the best possible way. The
+                  proposed approach is based on the use of an achievement
+                  scalarizing function (ASF) and on the classification of the
+                  individuals into several fronts. At each generation of
+                  WASF-GA, this classification is done according to the values
+                  that each solution takes on the ASF for the reference point
+                  and using different weight vectors. These vectors of weights
+                  are selected so that the vectors formed by their inverse
+                  components constitute a well-distributed representation of
+                  the weight vectors space. The efficiency and usefulness of
+                  WASF-GA is shown in several test problems in comparison to
+                  other preference-based EMO algorithms. Regarding a metric
+                  based on the hypervolume, we can say that WASF-GA has
+                  outperformed the other algorithms considered in most of the
+                  problems.},
+  number = 1,
+  journal = {Journal of Global Optimization},
+  month = may,
+  year = 2015,
+  pages = {101--129},
+  annote = {Proposed WASF-GA}
+}
+
+ +
+@article{RuiStu04:ejor,
+  author = { Rub{\'e}n Ruiz  and  Thomas St{\"u}tzle },
+  title = {A Simple and Effective Iterated Greedy Algorithm for
+                  the Permutation Flowshop Scheduling Problem},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 177,
+  number = 3,
+  pages = {2033--2049}
+}
+
+ +
+@article{RuiStu2008:ejor,
+  author = { Rub{\'e}n Ruiz  and  Thomas St{\"u}tzle },
+  title = {An {Iterated} {Greedy} heuristic for the sequence
+                  dependent setup times flowshop problem with makespan
+                  and weighted tardiness objectives },
+  journal = {European Journal of Operational Research},
+  volume = 187,
+  number = 3,
+  pages = {1143 -- 1159},
+  year = 2008
+}
+
+ +
+@article{Russell95,
+  author = {Robert A. Russell},
+  title = {Hybrid Heuristics for the Vehicle Routing Problem with Time Windows},
+  journal = {Transportation Science},
+  year = 1995,
+  volume = 29,
+  number = 2,
+  pages = {156--166}
+}
+
+ +
+@article{SabAyoKen2013tec,
+  author = {N. R. {Sabar} and M. {Ayob} and  Graham Kendall  and R. {Qu}},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Grammatical Evolution Hyper-Heuristic for Combinatorial Optimization Problems},
+  year = 2013,
+  volume = 17,
+  number = 6,
+  pages = {840--861}
+}
+
+ +
+@article{SabAyoKen2015tcyb,
+  author = {N. R. {Sabar} and M. {Ayob} and  Graham Kendall  and R. {Qu}},
+  journal = {IEEE Transactions on Cybernetics},
+  title = {A Dynamic Multiarmed Bandit-Gene Expression Programming Hyper-Heuristic for Combinatorial Optimization Problems},
+  year = 2015,
+  volume = 45,
+  number = 2,
+  pages = {217--228}
+}
+
+ +
+@article{SabAyoKen2015tec,
+  author = {N. R. Sabar and M. Ayob and  Graham Kendall  and R. Qu},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Automatic Design of a Hyper-Heuristic Framework With Gene Expression Programming for Combinatorial Optimization Problems},
+  year = 2015,
+  volume = 19,
+  number = 3,
+  pages = {309--325}
+}
+
+ +
+@article{SacDuvMai2018ego-ls-svm,
+  title = {A classification approach to efficient global optimization in
+                  presence of non-computable domains},
+  author = {Sacher, Matthieu and Duvigneau, R{\'e}gis and Le Maitre,
+                  Olivier and Durand, Mathieu and Berrini, Elisa and Hauville,
+                  Fr{\'e}d{\'e}ric and Astolfi, Jacques-Andr{\'e}},
+  journal = {Structural and Multidisciplinary Optimization},
+  volume = 58,
+  number = 4,
+  pages = {1537--1557},
+  year = 2018,
+  publisher = {Springer},
+  doi = {10.1007/s00158-018-1981-8},
+  keywords = {Safe optimization; CMA-ES, Gaussian processes; Least-Squares
+                  Support Vector Machine},
+  annote = {Proposed EGO-LS-SVM}
+}
+
+ +
+@article{SadFow2012nosql,
+  title = {{NoSQL} distilled},
+  author = {Sadalage, Pramod J. and Fowler, Martin},
+  journal = {AddisonWesley Professional},
+  year = 2012
+}
+
+ +
+@article{SakarWQ00,
+  author = { A. Burcu Altan Sakarya  and  Larry W. Mays },
+  title = {Optimal Operation of Water Distribution Pumps Considering
+                  Water Quality},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 126,
+  number = 4,
+  pages = {210--220},
+  date = {2000-07/2000-08},
+  year = 2000,
+  month = jul # { / } # aug
+}
+
+ +
+@article{SamPelDarRodPac2016ant,
+  author = {Marcela Sam\`a and  Paola Pellegrini  and Andrea D'Ariano and Joaquin Rodriguez and Dario Pacciarelli},
+  title = {Ant colony optimization for the real-time train routing selection problem},
+  journal = {Transportation Research Part B: Methodological},
+  volume = 85,
+  pages = {89--108},
+  year = 2016,
+  doi = {10.1016/j.trb.2016.01.005},
+  keywords = {irace}
+}
+
+ +
+@article{Sambridge1999,
+  title = {Geophysical inversion with a neighbourhood algorithm--I. Searching a parameter space},
+  author = {Sambridge, Malcolm},
+  journal = {Geophysical Journal International},
+  volume = 138,
+  number = 2,
+  pages = {479--494},
+  year = 1999
+}
+
+ +
+@article{SanDorNeb2019fame,
+  title = {A novel multi-objective evolutionary algorithm with fuzzy
+                  logic based adaptive selection of operators: {FAME}},
+  journal = {Information Sciences},
+  volume = 471,
+  pages = {233--251},
+  year = 2019,
+  issn = {0020-0255},
+  doi = {10.1016/j.ins.2018.09.005},
+  author = {Alejandro Santiago and  Bernab{\'e} Dorronsoro  and  Nebro, Antonio J.  and  Durillo, Juan J.  and Oscar Castillo and H{\'e}ctor
+                  J. Fraire},
+  keywords = {Multi-objective optimization, density estimation,
+                  evolutionary algorithm, adaptive algorithm, fuzzy logic, spatial spread deviation}
+}
+
+ +
+@article{SanGalRub2008tec,
+  author = {S\'anchez, Javier and Gal\'an, Manuel and Rubio, Enrique},
+  title = {Applying a traffic lights evolutionary optimization technique
+                  to a real case: ``{Las} {Ramblas}'' area in {Santa} {Cruz} de
+                  {Tenerife}},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2008,
+  volume = 12,
+  number = 1,
+  pages = {25--40},
+  keywords = {Cellular automata, Combinatorial optimization, Genetic
+                  algorithms, Microscopic traffic simulator, Traffic lights
+                  optimization}
+}
+
+ +
+@article{SanGalRub2010tec,
+  author = {J. J. S\'anchez-Medina and M. J. Gal\'an-Moreno and
+                  E. Rubio-Royo},
+  title = {Traffic Signal Optimization in ``{La} {Almozara}'' District
+                  in {Saragossa} Under Congestion Conditions, Using Genetic
+                  Algorithms, Traffic Microsimulation, and Cluster Computing},
+  journal = {IEEE Transactions on Intelligent Transportation Systems},
+  year = 2010,
+  volume = 11,
+  number = 1,
+  pages = {132--141},
+  month = mar,
+  keywords = {cellular automata; genetic algorithms; road traffic;traffic
+                  light programming;urban traffic congestion},
+  doi = {10.1109/TITS.2009.2034383},
+  issn = {1524-9050}
+}
+
+ +
+@article{SanOst2018wrr,
+  title = {Stochastic Scenario Evaluation in Evolutionary Algorithms
+                  Used for Robust Scenario-Based Optimization},
+  author = {Sankary, Nathan and  Avi Ostfeld },
+  journal = {Water Resources Research},
+  volume = 54,
+  number = 4,
+  pages = {2813--2833},
+  year = 2018
+}
+
+ +
+@article{SanRopHva2018joh,
+  author = {Alberto Santini and Stefan Ropke and Hvattum, Lars Magnus},
+  title = {A comparison of acceptance criteria for the adaptive large neighbourhood search metaheuristic},
+  journal = {Journal of Heuristics},
+  volume = 24,
+  pages = {783--815},
+  year = 2018,
+  doi = {10.1007/s10732-018-9377-x}
+}
+
+ +
+@article{Sandgren1988,
+  author = {Sandgren, E.},
+  title = {Nonlinear integer and discrete programming in mechanical
+                  design optimization},
+  volume = 112,
+  number = 2,
+  journal = {Journal of Mechanical Design},
+  pages = {223--229},
+  year = 1990,
+  doi = {10.1115/1.2912596}
+}
+
+ +
+@article{SasBerBie2022deepcave,
+  author = {Sass, René and Bergman, Eddie and  Biedenkapp, Andr{\'e}  and  Frank Hutter  and  Marius Thomas Lindauer },
+  title = {{DeepCAVE}: An Interactive Analysis Tool for Automated
+                  Machine Learning},
+  journal = {Arxiv preprint arXiv:2206.03493 [cs.LG]},
+  year = 2022,
+  doi = {10.48550/arXiv.2206.03493}
+}
+
+ +
+@article{Savelsbergh85tw,
+  title = {Local search in routing problems with time windows},
+  volume = 4,
+  doi = {10.1007/BF02022044},
+  abstract = {We develop local search algorithms for routing
+                  problems with time windows. The presented algorithms
+                  are based on thek-interchange concept. The presence
+                  of time windows introduces feasibility constraints,
+                  the checking of which normally requires O(N)
+                  time. Our method reduces this checking effort to
+                  O(1) time. We also consider the problem of finding
+                  initial solutions. A complexity result is given and
+                  an insertion heuristic is described.},
+  number = 1,
+  journal = {Annals of Operations Research},
+  author = { Martin W. P. Savelsbergh },
+  month = dec,
+  year = 1985,
+  pages = {285--305}
+}
+
+ +
+@article{SaxDurDebZha2013pca,
+  title = {Objective Reduction in Many-Objective Optimization: Linear
+                  and Nonlinear Algorithms},
+  author = { Saxena, Dhish Kumar  and  Jo{\~a}o A. Duro  and Tiwari, Anish and  Kalyanmoy Deb  and  Zhang, Qingfu },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 17,
+  number = 1,
+  pages = {77--99},
+  year = 2013,
+  doi = {10.1109/TEVC.2012.2185847}
+}
+
+ +
+@article{SchDoeHar09,
+  author = { Michael Schilde  and  Karl F. Doerner  and  Richard F. Hartl  and Guenter
+                  Kiechle},
+  title = {Metaheuristics for the bi-objective orienteering
+                  problem},
+  number = 3,
+  journal = {Swarm Intelligence},
+  year = 2009,
+  pages = {179--201},
+  volume = 3,
+  doi = {10.1007/s11721-009-0029-5},
+  abstract = {In this paper, heuristic solution
+                  techniques for the multi-objective orienteering
+                  problem are developed. The motivation stems from the
+                  problem of planning individual tourist routes in a
+                  city. Each point of interest in a city provides
+                  different benefits for different categories (e.g.,
+                  culture, shopping). Each tourist has different
+                  preferences for the different categories when
+                  selecting and visiting the points of interests
+                  (e.g., museums, churches). Hence, a multi-objective
+                  decision situation arises.  To determine all the
+                  Pareto optimal solutions, two metaheuristic search
+                  techniques are developed and applied. We use the
+                  Pareto ant colony optimization algorithm and extend
+                  the design of the variable neighborhood search
+                  method to the multi-objective case. Both methods are
+                  hybridized with path relinking procedures. The
+                  performances of the two algorithms are tested on
+                  several benchmark instances as well as on real world
+                  instances from different Austrian regions and the
+                  cities of Vienna and Padua.  The computational
+                  results show that both implemented methods are well
+                  performing algorithms to solve the multi-objective
+                  orienteering problem.}
+}
+
+ +
+@article{SchEgeBan2009aco,
+  title = {Extended ant colony optimization for non-convex mixed integer
+                  nonlinear programming},
+  author = {Schl{\"u}ter, Martin and Egea, Jose A. and  Banga, Julio R. },
+  journal = {Computers \& Operations Research},
+  volume = 36,
+  number = 7,
+  pages = {2217--2229},
+  year = 2009,
+  doi = {10.1016/j.cor.2008.08.015}
+}
+
+ +
+@article{SchEsqLarCoe2012tec,
+  author = { Oliver Sch{\"u}tze  and X. Esquivel and A. Lara and  Carlos A. {Coello Coello} },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {Using the Averaged {Hausdorff} Distance as a Performance
+                  Measure in Evolutionary Multiobjective Optimization},
+  year = 2012,
+  volume = 16,
+  number = 4,
+  pages = {504--522}
+}
+
+ +
+@article{SchHan79,
+  author = {Josef Schmee and Gerald J. Hahn},
+  title = {A Simple Method for Regression Analysis with Censored Data},
+  journal = {Technometrics},
+  year = 1979,
+  volume = 21,
+  number = 4,
+  pages = {417--432},
+  publisher = {Taylor \& Francis},
+  doi = {10.2307/1268280}
+}
+
+ +
+@article{SchHarSka2017safe,
+  title = {Safe active learning and safe {Bayesian} optimization for
+                  tuning a {PI}-controller},
+  author = {Schillinger, Mark and Hartmann, Benjamin and Skalecki, Patric
+                  and Meister, Mona and Nguyen-Tuong, Duy and Nelles, Oliver},
+  journal = {{IFAC}-{PapersOnLine}},
+  volume = 50,
+  number = 1,
+  pages = {5967--5972},
+  year = 2017,
+  doi = {10.1016/j.ifacol.2017.08.1258},
+  publisher = {Elsevier}
+}
+
+ +
+@article{SchHenSie2004hiv,
+  author = {Julie R. Schames and Richard H. Henchman and Jay S. Siegel
+                  and Christoph A. Sotriffer and Haihong Ni and J. Andrew
+                  McCammon},
+  title = {Discovery of a Novel Binding Trench in {HIV} Integrase},
+  journal = {Journal of Medicinal Chemistry},
+  volume = 47,
+  number = 8,
+  pages = {1879--1881},
+  year = 2004,
+  doi = {10.1021/jm0341913},
+  annote = {Evolutionary optimization of the first clinically approved
+                  anti-viral drug for HIV}
+}
+
+ +
+@article{SchHerTal2019archiver,
+  title = {Archivers for the representation of the set of approximate
+                  solutions for {MOPs}},
+  author = { Oliver Sch{\"u}tze  and  Carlos Hern{\'a}ndez  and  Talbi, El-Ghazali  and Sun, Jian-Qiao
+                  and Naranjani, Yousef and Xiong, F-R},
+  journal = {Journal of Heuristics},
+  year = 2019,
+  pages = {71--105},
+  volume = 25,
+  doi = {10.1007/s10732-018-9383-z},
+  keywords = {archiving, nearly optimality, epsilon-dominance, epsilon-approximation, hausdorff convergence}
+}
+
+ +
+@article{SchKoe2009,
+  year = 2009,
+  volume = 123,
+  number = 4,
+  pages = {421--433},
+  author = {Jeffrey C. Schank and Thomas J. Koehnle},
+  title = {Pseudoreplication is a pseudoproblem},
+  journal = {Journal of Comparative Psychology}
+}
+
+ +
+@article{SchLarCoe2011tec,
+  author = { Oliver Sch{\"u}tze  and A. Lara and  Carlos A. {Coello Coello} },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {On the Influence of the Number of Objectives on the Hardness
+                  of a Multiobjective Optimization Problem},
+  year = 2011,
+  volume = 15,
+  number = 4,
+  pages = {444--455}
+}
+
+ +
+@article{SchLauCoeDelTal2008,
+  title = {Convergence of stochastic search algorithms to finite size
+                  {Pareto} set approximations},
+  author = { Oliver Sch{\"u}tze  and  Marco Laumanns  and  Carlos A. {Coello Coello}  and Dellnitz,
+                  Michael and  Talbi, El-Ghazali },
+  journal = {Journal of Global Optimization},
+  volume = 41,
+  number = 4,
+  pages = {559--577},
+  year = 2008
+}
+
+ +
+@article{SchLauTanCoeTal2010,
+  title = {Computing gap free {Pareto} front approximations with
+                  stochastic search algorithms},
+  author = { Oliver Sch{\"u}tze  and  Marco Laumanns  and Emilia Tantar and  Carlos A. {Coello Coello}  and  Talbi, El-Ghazali },
+  journal = {Evolutionary Computation},
+  volume = 18,
+  number = 1,
+  pages = {65--96},
+  year = 2010
+}
+
+ +
+@article{SchMar1999,
+  author = {G. R. Schreiber and  Olivier Martin },
+  title = {Cut Size Statistics of Graph Bisection Heuristics},
+  journal = {SIAM Journal on Optimization},
+  year = 1999,
+  volume = 10,
+  number = 1,
+  pages = {231--251}
+}
+
+ +
+@article{SchSchStaDue2000,
+  title = {Record Breaking Optimization Results Using the Ruin
+                  and Recreate Principle},
+  journal = {Journal of Computational Physics},
+  volume = 159,
+  number = 2,
+  pages = {139--171},
+  year = 2000,
+  author = {Gerhard Schrimpf and Schneider, Johannes and Hermann
+                  Stamm-Wilbrandt and Gunter Dueck}
+}
+
+ +
+@article{SchSchoTho2019min,
+  title = {Min-ordering and max-ordering scalarization methods for
+                  multi-objective robust optimization},
+  author = {Schmidt, Marie and  Sch{\"o}bel, Anita  and Thom, Lisa},
+  journal = {European Journal of Operational Research},
+  volume = 275,
+  number = 2,
+  pages = {446--459},
+  year = 2019,
+  publisher = {Elsevier}
+}
+
+ +
+@article{SchSpeKra2018gptut,
+  author = {Schulz, Eric and Speekenbrink, Maarten and Krause, Andreas},
+  year = 2018,
+  month = aug,
+  pages = {1--16},
+  title = {A tutorial on {Gaussian} process regression: {Modelling},
+                  exploring, and exploiting functions},
+  volume = 85,
+  journal = {Journal of Mathematical Psychology},
+  doi = {10.1016/j.jmp.2018.03.001}
+}
+
+ +
+@article{SchStu2004:jmma,
+  author = { Tommaso Schiavinotto  and  Thomas St{\"u}tzle },
+  title = {The Linear Ordering Problem: Instances, Search Space Analysis
+  and Algorithms},
+  journal = {Journal of Mathematical Modelling and Algorithms},
+  year = 2004,
+  volume = 3,
+  number = 4,
+  pages = {367--402}
+}
+
+ +
+@article{SchStu2007:cor,
+  author = { Tommaso Schiavinotto  and  Thomas St{\"u}tzle },
+  title = {A Review of Metrics on Permutations for Search Space Analysis},
+  journal = {Computers \& Operations Research},
+  year = 2007,
+  volume = 34,
+  number = 10,
+  pages = {3143--3153}
+}
+
+ +
+@article{SchTacWuiSamStu2013,
+  author = {Tom Schrijvers and Guido Tack and Pieter Wuille and Horst Samulowitz and Peter J. Stuckey},
+  title = {Search Combinators},
+  journal = {Constraints},
+  year = 2013,
+  volume = 18,
+  number = 2,
+  pages = {269--305}
+}
+
+ +
+@article{SchVasCoe2011space,
+  author = { Oliver Sch{\"u}tze  and Massimiliano Vasile and  Carlos A. {Coello Coello} },
+  title = {Computing the Set of Epsilon-Efficient Solutions in
+                  Multiobjective Space Mission Design},
+  doi = {10.2514/1.46478},
+  year = 2011,
+  publisher = {American Institute of Aeronautics and Astronautics ({AIAA})},
+  volume = 8,
+  number = 3,
+  pages = {53--70},
+  journal = {Journal of Aerospace Computing, Information, and
+                  Communication}
+}
+
+ +
+@article{SchWelJon1998,
+  author = {Matthias Schonlau and William J. Welch and Donald R. Jones},
+  title = {Global versus Local Search in Constrained Optimization of
+                  Computer Models},
+  journal = {Lecture Notes-Monograph Series},
+  volume = 34,
+  year = 1998,
+  pages = {11--25},
+  editor = {Nancy Flournoy and William F. Rosenberger and Weng Kee Wong},
+  publisher = {Institute of Mathematical Statistics},
+  doi = {10.2307/4356058}
+}
+
+ +
+@article{ScheBraTor2022jair,
+  author = {Schede, Elias and Brandt, Jasmin and Tornede, Alexander and
+                  Wever, Marcel and Bengs, Viktor and  Eyke H{\"u}llermeier  and  Kevin Tierney },
+  title = {A survey of methods for automated algorithm configuration},
+  journal = {Journal of Artificial Intelligence Research},
+  year = 2022,
+  volume = 75,
+  pages = {425--487},
+  doi = {10.1613/jair.1.13676}
+}
+
+ +
+@article{Scipy2020natmet,
+  fullauthor = {Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and
+                  Haberland, Matt and Reddy, Tyler and Cournapeau, David and
+                  Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren
+                  and Bright, Jonathan and {van der Walt}, St{\'e}fan J. and
+                  Brett, Matthew and Wilson, Joshua and Millman, K. Jarrod and
+                  Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and
+                  Kern, Robert and Larson, Eric and Carey, C J and Polat,
+                  {\.I}lhan and Feng, Yu and Moore, Eric W. and {VanderPlas},
+                  Jake and Laxalde, Denis and Perktold, Josef and Cimrman,
+                  Robert and Henriksen, Ian and Quintero, E. A. and Harris,
+                  Charles R. and Archibald, Anne M. and Ribeiro, Ant{\^o}nio
+                  H. and Pedregosa, Fabian and {van Mulbregt}, Paul and {SciPy
+                  1.0 Contributors}},
+  author = {Virtanen, Pauli and others},
+  title = {{SciPy} 1.0: Fundamental Algorithms for Scientific Computing
+                  in {Python}},
+  journal = {Nature Methods},
+  year = 2020,
+  volume = 17,
+  pages = {261--272},
+  epub = {https://rdcu.be/b08Wh},
+  doi = {10.1038/s41592-019-0686-2}
+}
+
+ +
+@article{Sha1970bfgs,
+  author = {David F. Shanno},
+  title = {Conditioning of Quasi-Newton Methods for Function
+                  Minimization},
+  journal = {Mathematics of Computation},
+  year = 1970,
+  volume = 24,
+  number = 111,
+  pages = {647--656},
+  annote = {One of the four papers that proposed BFGS.},
+  publisher = {American Mathematical Society},
+  issn = {00255718, 10886842},
+  eprint = {http://www.jstor.org/stable/2004840},
+  keywords = {BFGS}
+}
+
+ +
+@article{ShaLopAlm2023hidden,
+  author = { Shavarani, Seyed Mahdi  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Allmendinger, Richard },
+  title = {Detecting Hidden and Irrelevant Objectives in Interactive
+                  Multi-Objective Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2023,
+  doi = {10.1109/TEVC.2023.3269348},
+  abstract = {Evolutionary multi-objective optimization algorithms (EMOAs)
+                  typically assume that all objectives that are relevant to the
+                  decision-maker (DM) are optimized by the EMOA. In some
+                  scenarios, however, there are irrelevant objectives that are
+                  optimized by the EMOA but ignored by the DM, as well as,
+                  hidden objectives that the DM considers when judging the
+                  utility of solutions but are not optimized. This discrepancy
+                  between the EMOA and the DM's preferences may impede the
+                  search for the most-preferred solution and waste resources
+                  evaluating irrelevant objectives. Research on objective
+                  reduction has focused so far on the structure of the problem
+                  and correlations between objectives and neglected the role of
+                  the DM. We formally define here the concepts of irrelevant
+                  and hidden objectives and propose methods for detecting them,
+                  based on uni-variate feature selection and recursive feature
+                  elimination, that use the preferences already elicited when a
+                  DM interacts with a ranking-based interactive EMOA
+                  (iEMOA). We incorporate the detection methods into an iEMOA
+                  capable of dynamically switching the objectives being
+                  optimized. Our experiments show that this approach can
+                  efficiently identify which objectives are relevant to the DM
+                  and reduce the number of objectives being optimized, while
+                  keeping and often improving the utility, according to the DM,
+                  of the best solution found.}
+}
+
+ +
+@article{ShaLopKno2023bench,
+  author = { Shavarani, Seyed Mahdi  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles },
+  title = {On Benchmarking Interactive Evolutionary Multi-Objective
+                  Algorithms},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2023,
+  doi = {10.1109/TEVC.2023.3289872},
+  abstract = {We carry out a detailed performance assessment of two
+                  interactive evolutionary multi-objective algorithms (EMOAs)
+                  using a machine decision maker that enables us to repeat
+                  experiments and study specific behaviours modeled after human
+                  decision makers (DMs). Using the same set of benchmark test
+                  problems as in the original papers on these interactive EMOAs
+                  (in up to 10 objectives), we bring to light interesting
+                  effects when we use a machine DM based on sigmoidal utility
+                  functions that have support from the psychology literature
+                  (replacing the simpler utility functions used in the original
+                  papers). Our machine DM enables us to go further and simulate
+                  human biases and inconsistencies as well. Our results from
+                  this study, which is the most comprehensive assessment of
+                  multiple interactive EMOAs so far conducted, suggest that
+                  current well-known algorithms have shortcomings that need
+                  addressing. These results further demonstrate the value of
+                  improving the benchmarking of interactive EMOAs}
+}
+
+ +
+@article{ShaLopMie2021visual,
+  title = {Visualizations for Decision Support in Scenario-based
+                  Multiobjective Optimization},
+  author = { Shavazipour, Babooshka  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kaisa Miettinen },
+  journal = {Information Sciences},
+  volume = 578,
+  pages = {1--21},
+  year = 2021,
+  abstract = {We address challenges of decision problems when managers need
+                  to optimize several conflicting objectives simultaneously
+                  under uncertainty. We propose visualization tools to support
+                  the solution of such scenario-based multiobjective
+                  optimization problems. Suitable graphical visualizations are
+                  necessary to support managers in understanding, evaluating,
+                  and comparing the performances of management decisions
+                  according to all objectives in all plausible scenarios. To
+                  date, no appropriate visualization has been suggested. This
+                  paper fills this gap by proposing two visualization methods:
+                  a novel extension of empirical attainment functions for
+                  scenarios and an adapted version of heatmaps. They help a
+                  decision-maker in gaining insight into realizations of
+                  trade-offs and comparisons between objective functions in
+                  different scenarios. Some fundamental questions that a
+                  decision-maker may wish to answer with the help of
+                  visualizations are also identified. Several examples are
+                  utilized to illustrate how the proposed visualizations
+                  support a decision-maker in evaluating and comparing
+                  solutions to be able to make a robust decision by answering
+                  the questions. Finally, we validate the usefulness of the
+                  proposed visualizations in a real-world problem with a real
+                  decision-maker. We conclude with guidelines regarding which
+                  of the proposed visualizations are best suited for different
+                  problem classes.},
+  doi = {10.1016/j.ins.2021.07.025},
+  supplement = {https://doi.org/10.5281/zenodo.5040421}
+}
+
+ +
+@article{ShaPiSha2017:asoco,
+  author = { Weishi Shao  and  Dechang Pi  and  Zhongshi Shao },
+  title = {Memetic algorithm with node and edge histogram for no-idle flow shop scheduling problem to minimize the makespan criterion},
+  journal = {Applied Soft Computing},
+  volume = 54,
+  pages = {164--182},
+  year = 2017
+}
+
+ +
+@article{ShaPiSha2018:cor,
+  author = { Weishi Shao  and  Dechang Pi  and  Zhongshi Shao },
+  title = {A hybrid discrete teaching-learning based meta-heuristic for solving no-idle flow shop scheduling problem with total tardiness criterion},
+  journal = {Computers \& Operations Research},
+  volume = 94,
+  pages = {89--105},
+  year = 2018
+}
+
+ +
+@article{ShaShuIsh2023is,
+  doi = {10.1016/j.ins.2022.11.155},
+  year = 2023,
+  volume = 622,
+  pages = {755--770},
+  author = {Ke Shang and Tianye Shu and  Ishibuchi, Hisao  and Yang Nan and
+                  Lie Meng Pang},
+  title = {Benchmarking large-scale subset selection in evolutionary
+                  multi-objective optimization},
+  journal = {Information Sciences}
+}
+
+ +
+@article{ShaSte2019deep,
+  author = { Shavazipour, Babooshka  and  T. J. Stewart },
+  title = {Multi-objective optimisation under deep uncertainty},
+  journal = {Operational Research},
+  year = 2019,
+  month = sep,
+  abstract = {This paper presents a scenario-based Multi-Objective
+                  structure to handle decision problems under deep
+                  uncertainty. Most of the decisions in real-life problems need
+                  to be made in the absence of complete knowledge about the
+                  consequences of the decision and/or are characterised by
+                  uncertainties about the future which is unpredictable. These
+                  uncertainties are almost impossible to reduce by gathering
+                  more information and are not statistical in
+                  nature. Therefore, classical probability-based approaches,
+                  such as stochastic programming, do not address these
+                  problems; as they require a correctly-defined complete sample
+                  space, strong assumptions (e.g. normality), or both. The
+                  proposed method extends the concept of two-stage stochastic
+                  programming with recourse to address the capability of
+                  dealing with deep uncertainty through the use of scenario
+                  planning rather than statistical expectation. In this
+                  research, scenarios are used as a dimension of preference to
+                  avoid problems relating to the assessment and use of
+                  probabilities under deep uncertainty. Such scenario-based
+                  thinking involved a multi-objective representation of
+                  performance under different future conditions as an
+                  alternative to expectation. To the best of our knowledge,
+                  this is the first attempt of performing a multi-criteria
+                  evaluation under deep uncertainty through a structured
+                  optimisation model. The proposed structure replacing
+                  probabilities (in dynamic systems with deep uncertainties) by
+                  aspirations within a goal programming structure. In fact,
+                  this paper also proposes an extension of the goal programming
+                  paradigm to deal with deep uncertainty. Furthermore, we will
+                  explain how this structure can be modelled, implemented, and
+                  solved by Goal Programming using some simple, but not
+                  trivial, examples. Further discussion and comparisons with
+                  some popular existing methods will also provided to highlight
+                  the superiorities of the proposed structure.},
+  doi = {10.1007/s12351-019-00512-1}
+}
+
+ +
+@article{ShaStrSte2020cce,
+  author = { Shavazipour, Babooshka  and Jonas Stray and  T. J. Stewart },
+  title = {Sustainable planning in sugar-bioethanol supply chain under
+                  deep uncertainty: A case study of {South} {African} sugarcane
+                  industry},
+  journal = {Computers \& Chemical Engineering},
+  volume = 143,
+  pages = 107091,
+  year = 2020,
+  doi = {10.1016/j.compchemeng.2020.107091},
+  keywords = {Supply chain management, Multi-objective optimisation, Deep
+                  uncertainty, Scenario planning, Renewable energy,},
+  abstract = {In this paper, the strategic planning of sugar-bioethanol
+                  supply chains (SCs) under deep uncertainty has been addressed
+                  by applying a two-stage scenario-based multiobjective
+                  optimisation methodology. In practice, the depth of
+                  uncertainty is very high, potential outcomes are not
+                  precisely enumerable, and probabilities of outcomes are not
+                  properly definable. To date, no appropriate framework has
+                  been suggested for dealing with deep uncertainty in supply
+                  chain management and energy-related problems. This study is
+                  the first try to fills this gap. Particularly, the
+                  sustainability of the whole infrastructure of the
+                  sugar-bioethanol SCs is analysed in such a way that the final
+                  solutions are sustainable, robust and adaptable for a broad
+                  range of plausible futures. Three objectives are considered
+                  in this problem under six uncertain parameters. A case study
+                  of South African sugarcane industry is utilised to study and
+                  examine the proposed model. The results prove the economic
+                  profitability and sustainability of the project.}
+}
+
+ +
+@article{ShaSweWan2016taking,
+  author = {Shahriari, B. and Swersky, K. and Wang, Z. and Adams, R. P. and  Nando de Freitas },
+  journal = {Proceedings of the IEEE},
+  title = {Taking the human out of the loop: A review of {Bayesian}
+                  optimization},
+  year = 2016,
+  number = 1,
+  pages = {148--175},
+  volume = 104,
+  publisher = {IEEE}
+}
+
+ +
+@article{ShaSweWanAdaFre2016,
+  author = {Bobak Shahriari and Kevin Swersky and Ziyu Wang and Ryan P. Adams and  Nando de Freitas },
+  title = {Taking the Human Out of the Loop: {A} Review of {Bayesian} Optimization},
+  journal = {Proceedings of the IEEE},
+  year = 2016,
+  volume = 104,
+  number = 1,
+  pages = {148--175}
+}
+
+ +
+@article{ShiBac2009niching,
+  title = {Niching with derandomized evolution strategies in artificial
+                  and real-world landscapes},
+  author = { Shir, Ofer M.  and  Thomas B{\"a}ck },
+  journal = {Natural Computing},
+  volume = 8,
+  number = 1,
+  pages = {171--196},
+  year = 2009,
+  doi = {10.1007/s11047-007-9065-5},
+  publisher = {Springer}
+}
+
+ +
+@article{ShiMarDud2008stat,
+  author = {David Shilane and Jarno Martikainen and Sandrine Dudoit and
+                  Seppo J. Ovaska},
+  title = {A general framework for statistical performance comparison of
+                  evolutionary computation algorithms},
+  journal = {Information Sciences},
+  volume = 178,
+  number = 14,
+  pages = {2870--2879},
+  year = 2008,
+  doi = {10.1016/j.ins.2008.03.007}
+}
+
+ +
+@article{ShiZha2016,
+  title = {The generalization of {Latin} hypercube sampling},
+  author = {Shields, Michael D. and Zhang, Jiaxin},
+  journal = {Reliability Engineering \& System Safety},
+  year = 2016,
+  pages = {96--108},
+  volume = 148,
+  alias = {Shields2016}
+}
+
+ +
+@article{ShmHoo05:bmc,
+  author = { A. Shmygelska  and  Holger H. Hoos },
+  title = {An Ant Colony Optimisation Algorithm for the {2D}
+                  and {3D} Hydrophobic Polar Protein Folding Problem},
+  journal = {BMC Bioinformatics},
+  year = 2005,
+  volume = 6,
+  pages = 30,
+  doi = {10.1186/1471-2105-6-30},
+  alias = {ShmHoo05:protein}
+}
+
+ +
+@article{SilFraBer2021,
+  author = { Silva-Mu\~noz, Mois\'es  and  Alberto Franzin  and  Hughes Bersini },
+  title = {Automatic configuration of the {Cassandra} database using irace},
+  year = 2021,
+  journal = {{PeerJ} Computer Science},
+  volume = 7,
+  pages = {e634},
+  doi = {10.7717/peerj-cs.634}
+}
+
+ +
+@article{SilRit2017:cor,
+  author = {Paulo Vitor Silvestrin and  Marcus Ritt},
+  title = {An Iterated Tabu Search for the Multi-compartment Vehicle Routing Problem},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  volume = 81,
+  pages = {192--202}
+}
+
+ +
+@article{SilSubOch2015,
+  author = {Marcos {Melo Silva} and  Anand Subramanian  and  Luiz Satoru Ochi },
+  title = {An Iterated Local Search Heuristic for the Split Delivery Vehicle
+               Routing Problem},
+  journal = {Computers \& Operations Research},
+  year = 2015,
+  volume = 53,
+  pages = {234--249}
+}
+
+ +
+@article{SimChaThi2014:swarm,
+  author = {Olivier Simonin and  Fran{\c{c}}ois Charpillet and Eric Thierry},
+  title = {Revisiting wavefront construction with collective agents: an approach to foraging},
+  journal = {Swarm Intelligence},
+  year = 2014,
+  volume = 9,
+  number = 2,
+  pages = {113--138},
+  doi = {10.1007/s11721-014-0093-3},
+  keywords = {irace}
+}
+
+ +
+@article{SimHarPae2015ecj,
+  author = {Kevin Sim and  Emma Hart  and  Ben Paechter },
+  title = {A Lifelong Learning Hyper-heuristic Method for Bin Packing},
+  volume = 23,
+  number = 1,
+  pages = {37--67},
+  year = 2015,
+  doi = {10.1162/EVCO_a_00121},
+  journal = {Evolutionary Computation}
+}
+
+ +
+@article{SimNelSim2011phack,
+  author = {Simmons, Joseph P. and Nelson, Leif D. and Simonsohn, Uri},
+  title = {False-Positive Psychology: Undisclosed Flexibility in Data
+                  Collection and Analysis Allows Presenting Anything as
+                  Significant},
+  journal = {Psychological Science},
+  year = 2011,
+  url = {https://ssrn.com/abstract=1850704},
+  annote = {Proposed the term p-hacking}
+}
+
+ +
+@article{SimNew1958heur,
+  title = {Heuristic Problem Solving: The Next Advance in Operations
+                  Research},
+  volume = 6,
+  doi = {10.1287/opre.6.1.1},
+  number = 1,
+  journal = {Operations Research},
+  author = { Simon, Herbert A.  and Newell, Allen},
+  year = 1958,
+  pages = {1--10}
+}
+
+ +
+@article{SimLebNel2010anchoring,
+  author = {Simmons, Joseph P. and LeBoeuf, Robyn A. and Nelson, Leif D.},
+  title = {The effect of accuracy motivation on anchoring and
+                  adjustment: {Do} people adjust from provided anchors?},
+  journal = {Journal of Personality and Social Psychology},
+  year = 2010,
+  volume = 99,
+  number = 6,
+  pages = {917--932},
+  issn = {1939-1315, 0022-3514},
+  shorttitle = {The effect of accuracy motivation on anchoring and
+                  adjustment},
+  doi = {10.1037/a0021540},
+  abstract = {Increasing accuracy motivation (e.g., by providing monetary
+                  incentives for accuracy) often fails to increase adjustment
+                  away from provided anchors, a result that has led researchers
+                  to conclude that people do not effortfully adjust away from
+                  such anchors. We challenge this conclusion. First, we show
+                  that people are typically uncertain about which way to adjust
+                  from provided anchors and that this uncertainty often causes
+                  people to believe that they have initially adjusted too far
+                  away from such anchors (Studies 1a and 1b). Then, we show
+                  that although accuracy motivation fails to increase the gap
+                  between anchors and final estimates when people are uncertain
+                  about the direction of adjustment, accuracy motivation does
+                  increase anchor-estimate gaps when people are certain about
+                  the direction of adjustment, and that this is true regardless
+                  of whether the anchors are provided or self-generated
+                  (Studies 2, 3a, 3b, and 5). These results suggest that people
+                  do effortfully adjust away from provided anchors but that
+                  uncertainty about the direction of adjustment makes that
+                  adjustment harder to detect than previously assumed. This
+                  conclusion has important theoretical implications, suggesting
+                  that currently emphasized distinctions between anchor types
+                  (self-generated vs. provided) are not fundamental and that
+                  ostensibly competing theories of anchoring (selective
+                  accessibility and anchoring-and-adjustment) are
+                  complementary.},
+  language = {en}
+}
+
+ +
+@article{Simon1955,
+  author = { Simon, Herbert A. },
+  title = {A Behavioral Model of Rational Choice},
+  journal = {The Quarterly Journal of Economics},
+  volume = 69,
+  number = 1,
+  pages = {99--118},
+  year = 1955,
+  epub = {http://www.jstor.org/stable/1884852}
+}
+
+ +
+@article{SinIsaTap2011pareto,
+  author = {Singh, Hemant Kumar and Isaacs, Amitay and  Ray, Tapabrata },
+  title = {A {Pareto} Corner Search Evolutionary Algorithm and
+                  Dimensionality Reduction in Many-Objective Optimization
+                  Problems},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2011,
+  volume = 15,
+  number = 4,
+  pages = {539--556},
+  abstract = {Many-objective optimization refers to the optimization
+                  problems containing large number of objectives, typically
+                  more than four.  Non-dominance is an inadequate strategy for
+                  convergence to the Pareto front for such problems, as almost
+                  all solutions in the population become non-dominated,
+                  resulting in loss of convergence pressure.  However, for some
+                  problems, it may be possible to generate the Pareto front
+                  using only a few of the objectives, rendering the rest of the
+                  objectives redundant. Such problems may be reducible to a
+                  manageable number of relevant objectives, which can be
+                  optimized using conventional multiobjective evolutionary
+                  algorithms (MOEAs). For dimensionality reduction, most
+                  proposals in the paper rely on analysis of a representative
+                  set of solutions obtained by running a conventional MOEA for
+                  a large number of generations, which is computationally
+                  overbearing.  A novel algorithm, Pareto corner search
+                  evolutionary algorithm (PCSEA), is introduced in this paper,
+                  which searches for the corners of the Pareto front instead of
+                  searching for the complete Pareto front. The solutions
+                  obtained using PCSEA are then used for dimensionality
+                  reduction to identify the relevant objectives. The potential
+                  of the proposed approach is demonstrated by studying its
+                  performance on a set of benchmark test problems and two
+                  engineering examples. While the preliminary results obtained
+                  using PCSEA are promising, there are a number of areas that
+                  need further investigation. This paper provides a number of
+                  useful insights into dimensionality reduction and, in
+                  particular, highlights some of the roadblocks that need to be
+                  cleared for future development of algorithms attempting to
+                  use few selected solutions for identifying relevant
+                  objectives},
+  doi = {10.1109/TEVC.2010.2093579}
+}
+
+ +
+@article{SinPin1998:IIE,
+  author = {Marcos Singer and Michael L. Pinedo},
+  title = {A Computational Study of Branch and Bound Techniques for Minimizing the Total Weighted Tardiness in Job Shops},
+  journal = {IIE Transactions},
+  year = 1998,
+  volume = 30,
+  number = 2,
+  pages = {109--118}
+}
+
+ +
+@article{SinSaxDeb2013asc,
+  author = {Ankur Sinha and  Saxena, Dhish Kumar  and  Kalyanmoy Deb  and  Ashutosh Tiwari },
+  title = {Using objective reduction and interactive procedure to handle
+                  many-objective optimization problems},
+  journal = {Applied Soft Computing},
+  volume = 13,
+  number = 1,
+  pages = {415--427},
+  year = 2013,
+  doi = {10.1016/j.asoc.2012.08.030},
+  keywords = {Evolutionary algorithms, Evolutionary multi- and
+                  many-objective optimization, Multi-criteria decision making,
+                  Machine learning, Interactive optimization},
+  abstract = {A number of practical optimization problems are posed as
+                  many-objective (more than three objectives) problems. Most of
+                  the existing evolutionary multi-objective optimization
+                  algorithms, which target the entire Pareto-front are not
+                  equipped to handle many-objective problems. Though there have
+                  been copious efforts to overcome the challenges posed by such
+                  problems, there does not exist a generic procedure to
+                  effectively handle them. This paper presents a simplify and
+                  solve framework for handling many-objective optimization
+                  problems. In that, a given problem is simplified by
+                  identification and elimination of the redundant objectives,
+                  before interactively engaging the decision maker to converge
+                  to the most preferred solution on the Pareto-optimal
+                  front. The merit of performing objective reduction before
+                  interacting with the decision maker is two fold. Firstly, the
+                  revelation that certain objectives are redundant,
+                  significantly reduces the complexity of the optimization
+                  problem, implying lower computational cost and higher search
+                  efficiency. Secondly, it is well known that human beings are
+                  not efficient in handling several factors (objectives in the
+                  current context) at a time. Hence, simplifying the problem a
+                  priori addresses the fundamental issue of cognitive overload
+                  for the decision maker, which may help avoid inconsistent
+                  preferences during the different stages of interactive
+                  engagement. The implementation of the proposed framework is
+                  first demonstrated on a three-objective problem, followed by
+                  its application on two real-world engineering problems.}
+}
+
+ +
+@article{SinBahRay2019distance,
+  title = {Distance-based subset selection for benchmarking in
+                  evolutionary multi/many-objective optimization},
+  author = {Singh, Hemant Kumar and Bhattacharjee, Kalyan Shankar and  Ray, Tapabrata },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2019,
+  number = 5,
+  pages = {904--912},
+  volume = 23,
+  publisher = {IEEE}
+}
+
+ +
+@article{SioGag2018:ejor,
+  author = {Sioud, Aymen and  Caroline Gagn{\'e} },
+  title = {Enhanced migrating birds optimization algorithm for the permutation flow shop problem with sequence dependent setup times},
+  journal = {European Journal of Operational Research},
+  volume = 264,
+  number = 1,
+  pages = {66--73},
+  year = 2018
+}
+
+ +
+@article{SmaMcCAll2011efficient,
+  title = {Efficient discovery of anti-inflammatory small-molecule
+                  combinations using evolutionary computing},
+  author = {Small, Ben G. and McColl, Barry W. and  Allmendinger, Richard  and Pahle, J{\"u}rgen and L{\'o}pez-Castej{\'o}n, Gloria and
+                  Rothwell, Nancy J. and  Joshua D. Knowles  and Mendes, Pedro and
+                  Brough, David and Kell, Douglas B.},
+  journal = {Nature Chemical Biology},
+  volume = 7,
+  number = 12,
+  pages = {902--908},
+  year = 2011,
+  publisher = {Nature Publishing Group}
+}
+
+ +
+@article{SmiBaaWreLew2014isa,
+  author = { Kate Smith{-}Miles  and Baatar, Davaatseren and Wreford, Brendan and  Lewis, Rhyd M. R. },
+  title = {Towards Objective Measures of Algorithm Performance across
+                  Instance Space},
+  journal = {Computers \& Operations Research},
+  year = 2014,
+  volume = 45,
+  pages = {12--24},
+  doi = {10.1016/j.cor.2013.11.015},
+  abstract = {This paper tackles the difficult but important task of
+                  objective algorithm performance assessment for
+                  optimization. Rather than reporting average performance of
+                  algorithms across a set of chosen instances, which may bias
+                  conclusions, we propose a methodology to enable the strengths
+                  and weaknesses of different optimization algorithms to be
+                  compared across a broader instance space. The results
+                  reported in a recent Computers and Operations Research paper
+                  comparing the performance of graph coloring heuristics are
+                  revisited with this new methodology to demonstrate (i) how
+                  pockets of the instance space can be found where algorithm
+                  performance varies significantly from the average performance
+                  of an algorithm; (ii) how the properties of the instances can
+                  be used to predict algorithm performance on previously unseen
+                  instances with high accuracy; and (iii) how the relative
+                  strengths and weaknesses of each algorithm can be visualized
+                  and measured objectively.},
+  keywords = {Algorithm selection; Instance Space Analysis; Graph coloring;
+                  Heuristics; Performance prediction}
+}
+
+ +
+@article{SmiBow2015:cor,
+  author = { Kate Smith{-}Miles  and Simon Bowly},
+  title = {Generating New Test Instances by Evolving in Instance Space},
+  journal = {Computers \& Operations Research},
+  year = 2015,
+  volume = 63,
+  pages = {102--113},
+  doi = {10.1016/j.cor.2015.04.022},
+  abstract = {Our confidence in the future performance of any algorithm,
+                  including optimization algorithms, depends on how carefully
+                  we select test instances so that the generalization of
+                  algorithm performance on future instances can be inferred. In
+                  recent work, we have established a methodology to generate a
+                  2-d representation of the instance space, comprising a set of
+                  known test instances. This instance space shows the
+                  similarities and differences between the instances using
+                  measurable features or properties, and enables the
+                  performance of algorithms to be viewed across the instance
+                  space, where generalizations can be inferred. The power of
+                  this methodology is the insights that can be generated into
+                  algorithm strengths and weaknesses by examining the regions
+                  in instance space where strong performance can be
+                  expected. The representation of the instance space is
+                  dependent on the choice of test instances however. In this
+                  paper we present a methodology for generating new test
+                  instances with controllable properties, by filling observed
+                  gaps in the instance space. This enables the generation of
+                  rich new sets of test instances to support better the
+                  understanding of algorithm strengths and weaknesses. The
+                  methodology is demonstrated on graph colouring as a case
+                  study.},
+  keywords = {Benchmarking; Evolving instances; Graph colouring; Instance
+                  space; Test instances}
+}
+
+ +
+@article{SmiChrMun2021where,
+  author = { Kate Smith{-}Miles  and Jeffrey Christiansen and  Mario A. Mu{\~{n}}oz },
+  title = {Revisiting Where Are the Hard Knapsack Problems? Via
+                  {Instance} {Space} {Analysis}},
+  journal = {Computers \& Operations Research},
+  year = 2021,
+  volume = 128,
+  pages = 105184,
+  doi = {10.1016/j.cor.2020.105184},
+  keywords = {0-1 Knapsack problem; Algorithm portfolios; Algorithm
+                  selection; Instance difficulty; Instance generation; Instance
+                  Space Analysis; Performance evaluation}
+}
+
+ +
+@article{SmiLop2012:cor,
+  author = { Kate Smith{-}Miles  and Lopes, Leo},
+  title = {Measuring instance difficulty for combinatorial optimization
+                  problems},
+  journal = {Computers \& Operations Research},
+  year = 2012,
+  volume = 39,
+  pages = {875--889}
+}
+
+ +
+@article{SmiMun2023isa,
+  author = { Kate Smith{-}Miles  and  Mario A. Mu{\~{n}}oz },
+  title = {Instance Space Analysis for Algorithm Testing: Methodology
+                  and Software Tools},
+  journal = {{ACM} Computing Surveys},
+  year = 2023,
+  volume = 55,
+  number = 12,
+  month = mar,
+  issue_date = {December 2023},
+  doi = {10.1145/3572895},
+  abstract = {Instance Space Analysis (ISA) is a recently developed
+                  methodology to (a) support objective testing of algorithms
+                  and (b) assess the diversity of test instances. Representing
+                  test instances as feature vectors, the ISA methodology
+                  extends Rice's 1976 Algorithm Selection Problem framework to
+                  enable visualization of the entire space of possible test
+                  instances, and gain insights into how algorithm performance
+                  is affected by instance properties. Rather than reporting
+                  algorithm performance on average across a chosen set of test
+                  problems, as is standard practice, the ISA methodology offers
+                  a more nuanced understanding of the unique strengths and
+                  weaknesses of algorithms across different regions of the
+                  instance space that may otherwise be hidden on average. It
+                  also facilitates objective assessment of any bias in the
+                  chosen test instances and provides guidance about the
+                  adequacy of benchmark test suites. This article is a
+                  comprehensive tutorial on the ISA methodology that has been
+                  evolving over several years, and includes details of all
+                  algorithms and software tools that are enabling its worldwide
+                  adoption in many disciplines. A case study comparing
+                  algorithms for university timetabling is presented to
+                  illustrate the methodology and tools.},
+  articleno = 255,
+  numpages = 31,
+  keywords = {test instance diversity, benchmarking, timetabling, Algorithm
+                  footprints, MATLAB, software as a service, meta-heuristics,
+                  algorithm selection, meta-learning}
+}
+
+ +
+@article{Smith-Miles2008,
+  author = { Kate Smith{-}Miles },
+  title = {Cross-disciplinary Perspectives on Meta-learning for Algorithm Selection},
+  journal = {{ACM} Computing Surveys},
+  year = 2008,
+  volume = 41,
+  number = 1,
+  pages = {1--25}
+}
+
+ +
+@article{SocBlu07,
+  author = { Krzysztof Socha  and  Christian Blum },
+  title = {An ant colony optimization algorithm for continuous
+                  optimization: An application to feed-forward neural
+                  network training},
+  journal = {Neural Computing \& Applications},
+  year = 2007,
+  volume = 16,
+  number = 3,
+  pages = {235--247},
+  alias = {SocBlu2007:nca}
+}
+
+ +
+@article{SocDor2008:ejor,
+  author = { Krzysztof Socha  and  Marco Dorigo },
+  title = {Ant Colony Optimization for Continuous Domains},
+  year = 2008,
+  journal = {European Journal of Operational Research},
+  volume = 185,
+  number = 3,
+  pages = {1155--1173},
+  doi = {10.1016/j.ejor.2006.06.046},
+  annote = {Proposed ACOR (ACO$_\mathbb{R}$)},
+  keywords = {ACOR}
+}
+
+ +
+@article{Sol2002:tec,
+  author = { Christine Solnon },
+  title = {Ants Can Solve Constraint Satisfaction Problems},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002,
+  volume = 6,
+  number = 4,
+  pages = {347--357}
+}
+
+ +
+@article{SolMarMic2008,
+  author = {D. Soler and E. Mart{\'i}nez and J. C. Mic\'o},
+  title = {A Transformation for the Mixed General Routing Problem with Turn Penalties},
+  journal = {Journal of the Operational Research Society},
+  year = 2008,
+  volume = 59,
+  pages = {540--547}
+}
+
+ +
+@article{Solomon87vrptw,
+  author = { M. M. Solomon },
+  title = {Algorithms for the Vehicle Routing and Scheduling
+                  Problems with Time Windows},
+  journal = {Operations Research},
+  year = 1987,
+  volume = 35,
+  pages = {254--265}
+}
+
+ +
+@article{SonWanHeJin2021kriging,
+  title = {A {Kriging}-assisted two-archive evolutionary algorithm for
+                  expensive many-objective optimization},
+  author = {Song, Zhenshou and Wang, Handing and He, Cheng and  Yaochu Jin },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 25,
+  number = 6,
+  pages = {1013--1027},
+  year = 2021,
+  publisher = {IEEE}
+}
+
+ +
+@article{Sor2013,
+  title = {Metaheuristics---the metaphor exposed},
+  author = { Kenneth S{\"o}rensen },
+  journal = {International Transactions in Operational Research},
+  year = 2015,
+  volume = 22,
+  number = 1,
+  pages = {3--18},
+  doi = {10.1111/itor.12001}
+}
+
+ +
+@article{SorArnPal2017,
+  author = { Kenneth S{\"o}rensen  and Florian Arnold and Daniel {Palhazi Cuervo}},
+  title = {A critical analysis of the ``improved {Clarke} and {Wright}
+                  savings algorithm''},
+  journal = {International Transactions in Operational Research},
+  volume = 26,
+  number = 1,
+  pages = {54--63},
+  year = 2017,
+  alias = {SorArnPal2019},
+  doi = {10.1111/itor.12443},
+  keywords = {reproducibility, vehicle routing}
+}
+
+ +
+@article{SorOchSotBur2017:ejor,
+  author = {Jorge A. Soria-Alcaraz and  Gabriela Ochoa  and Marco A. Sotelo-Figeroa and  Edmund K. Burke },
+  title = {A Methodology for Determining an Effective Subset of Heuristics in Selection Hyper-heuristics},
+  journal = {European Journal of Operational Research},
+  year = 2017,
+  volume = 260,
+  pages = {972--983}
+}
+
+ +
+@article{SouRitLop2021cap,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Capping Methods for the Automatic Configuration of
+                  Optimization Algorithms},
+  journal = {Computers \& Operations Research},
+  doi = {10.1016/j.cor.2021.105615},
+  year = 2022,
+  volume = 139,
+  pages = 105615,
+  supplement = {https://github.com/souzamarcelo/supp-cor-capopt},
+  abstract = {Automatic configuration techniques are widely and
+                  successfully used to find good parameter settings for
+                  optimization algorithms. Configuration is costly, because it
+                  is necessary to evaluate many configurations on different
+                  instances. For decision problems, when the objective is to
+                  minimize the running time of the algorithm, many
+                  configurators implement capping methods to discard poor
+                  configurations early. Such methods are not directly
+                  applicable to optimization problems, when the objective is to
+                  optimize the cost of the best solution found, given a
+                  predefined running time limit. We propose new capping methods
+                  for the automatic configuration of optimization
+                  algorithms. They use the previous executions to determine a
+                  performance envelope, which is used to evaluate new
+                  executions and cap those that do not satisfy the envelope
+                  conditions. We integrate the capping methods into the irace
+                  configurator and evaluate them on different optimization
+                  scenarios. Our results show that the proposed methods can
+                  save from about 5\% to 78\% of the configuration effort,
+                  while finding configurations of the same quality. Based on
+                  the computational analysis, we identify two conservative and
+                  two aggressive methods, that save an average of about 20\%
+                  and 45\% of the configuration effort, respectively. We also
+                  provide evidence that capping can help to better use the
+                  available budget in scenarios with a configuration time
+                  limit.}
+}
+
+ +
+@article{Souilah1995,
+  title = {Simulated annealing for manufacturing systems layout design},
+  author = {Souilah, Abdelghani},
+  journal = {European Journal of Operational Research},
+  volume = 82,
+  number = 3,
+  pages = {592--614},
+  year = 1995,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Spearman1904,
+  title = {The proof and measurement of association between two things},
+  author = {Charles Spearman},
+  journal = {The American journal of psychology},
+  volume = 15,
+  number = 1,
+  pages = {72--101},
+  year = 1904,
+  publisher = {University of Illinois Press}
+}
+
+ +
+@article{SpecCPU2000,
+  author = {J. L. Henning},
+  journal = {Computer},
+  title = {{SPEC} {CPU2000}: measuring {CPU} performance in the New Millennium},
+  publisher = {IEEE Computer Society Press},
+  year = 2000,
+  volume = 33,
+  number = 7,
+  pages = {28--35},
+  doi = {10.1109/2.869367}
+}
+
+ +
+@article{SpiTen2004,
+  title = {Smoothed analysis of algorithms: Why the simplex algorithm
+                  usually takes polynomial time},
+  author = {Spielman, Daniel A. and Teng, Shang-Hua},
+  journal = {Journal of the ACM},
+  volume = 51,
+  number = 3,
+  pages = {385--463},
+  year = 2004,
+  publisher = {ACM Press}
+}
+
+ +
+@article{SprHarDre1997ors,
+  author = {Sprecher, Arno and Hartmann, S{\"o}nke and Drexl, Andreas},
+  title = {An exact algorithm for project scheduling with
+                  multiple modes},
+  volume = 19,
+  doi = {10.1007/BF01545587},
+  number = 3,
+  journal = {OR Spektrum},
+  year = 1997,
+  keywords = {branch-and-bound, multi-mode resource-constrained
+                  project scheduling, project scheduling},
+  pages = {195--203}
+}
+
+ +
+@article{SprKolDre1995ejor,
+  author = {Sprecher, Arno and Kolisch, Rainer and Drexl,
+                  Andreas},
+  title = {Semi-active, active, and non-delay schedules for the
+                  resource-constrained project scheduling problem},
+  volume = 80,
+  doi = {10.1016/0377-2217(93)E0294-8},
+  abstract = {We consider the resource-constrained project
+                  scheduling problem ({RCPSP).} The focus of the paper
+                  is on a formal definition of semi-active, active,
+                  and non-delay schedules. Traditionally these
+                  schedules establish basic concepts within the job
+                  shop scheduling literature. There they are usually
+                  defined in a rather informal way which does not
+                  create any substantial problems. Using these
+                  concepts in the more general {RCPSP} without giving
+                  a formal definition may cause serious
+                  problems. After providing a formal definition of
+                  semi-active, active, and non-delay schedules for the
+                  {RCPSP} we outline some of these problems occurring
+                  within the disjunctive arc concept.},
+  number = 1,
+  journal = {European Journal of Operational Research},
+  year = 1995,
+  keywords = {active schedules, Branch-and-bound methods,
+                  non-delay schedules, Resource-constrained project
+                  scheduling, Semi-active schedules},
+  pages = {94--102}
+}
+
+ +
+@article{SriDeb1994:nsga,
+  author = {N. Srinivas and  Kalyanmoy Deb },
+  title = {Multiobjective Optimization Using Nondominated
+                  Sorting in Genetic Algorithms},
+  journal = {Evolutionary Computation},
+  year = 1994,
+  volume = 2,
+  pages = {221--248},
+  number = 3
+}
+
+ +
+@article{Ste1996jmcda,
+  author = { T. J. Stewart },
+  title = {Robustness of Additive Value Function Methods in {MCDM}},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  year = 1996,
+  volume = 5,
+  number = 4,
+  pages = {301--309},
+  keywords = {machine decision-making}
+}
+
+ +
+@article{Ste1999ejor,
+  author = { T. J. Stewart },
+  title = {Evaluation and refinement of aspiration-based methods in {MCDM}},
+  journal = {European Journal of Operational Research},
+  year = 1999,
+  volume = 113,
+  number = 3,
+  pages = {643--652},
+  keywords = {machine decision-making}
+}
+
+ +
+@article{Ste2005jors,
+  author = { T. J. Stewart },
+  title = {Goal programming and cognitive biases in decision-making},
+  volume = 56,
+  doi = {10.1057/palgrave.jors.2601948},
+  number = 10,
+  journal = {Journal of the Operational Research Society},
+  year = 2005,
+  keywords = {machine decision making},
+  pages = {1166--1175}
+}
+
+ +
+@article{SteFreRio2013omg,
+  title = {Integrating multicriteria decision analysis and scenario
+                  planning: Review and extension},
+  author = { T. J. Stewart  and Simon French and Jesus Rios},
+  journal = {Omega},
+  year = 2013,
+  number = 4,
+  pages = {679--688},
+  volume = 41,
+  doi = {10.1016/j.omega.2012.09.003},
+  keywords = {Multicriteria decision analysis}
+}
+
+ +
+@article{SteHeiHah2020,
+  author = {Stegherr, Helena and Heider, Michael and H\"ahner, J\"org},
+  title = {Classifying Metaheuristics: Towards a unified multi-level classification system},
+  journal = {Natural Computing},
+  year = 2020,
+  doi = {10.1007/s11047-020-09824-0}
+}
+
+ +
+@article{SteRad2008computing,
+  title = {Computing all efficient solutions of the biobjective minimum
+                  spanning tree problem},
+  author = {Steiner, Sarah and Radzik, Tomasz},
+  journal = {Computers \& Operations Research},
+  year = 2008,
+  number = 1,
+  pages = {198--211},
+  volume = 35
+}
+
+ +
+@article{Sto2014edge,
+  author = {Stodden, Victoria},
+  title = {What scientific idea is ready for retirement?
+                  Reproducibility},
+  journal = {Edge},
+  year = 2014,
+  url = {https://www.edge.org/annual-question/2014/response/25340},
+  annote = {Introduces computational reproducibility, empirical
+                  reproducibility and statistical reproducibility}
+}
+
+ +
+@article{StoAlb2014,
+  author = {Stolfi, Daniel H. and  Alba, Enrique },
+  doi = {10.1016/j.asoc.2014.07.014},
+  journal = {Applied Soft Computing},
+  keywords = {Evolutionary algorithm,Road traffic,Smart city,Smart
+                  mobility,Traffic light,WiFi connections},
+  pages = {181--195},
+  title = {Red Swarm: Reducing travel times in smart cities by using
+                  bio-inspired algorithms},
+  volume = 24,
+  year = 2014,
+  abstract = {This article presents an innovative approach to solve one of
+                  the most relevant problems related to smart mobility: the
+                  reduction of vehicles' travel time. Our original approach,
+                  called Red Swarm, suggests a potentially customized route to
+                  each vehicle by using several spots located at traffic lights
+                  in order to avoid traffic jams by using \{V2I\}
+                  communications. That is quite different from other existing
+                  proposals, as it deals with real maps and actual streets, as
+                  well as several road traffic distributions. We propose an
+                  evolutionary algorithm (later efficiently parallelized) to
+                  optimize our case studies which have been imported from
+                  OpenStreetMap into \{SUMO\} as they belong to a real city. We
+                  have also developed a Rerouting Algorithm which accesses the
+                  configuration of the Red Swarm and communicates the route
+                  chosen to vehicles, using the spots (via WiFi
+                  link). Moreover, we have developed three competing algorithms
+                  in order to compare their results to those of Red Swarm and
+                  have observed that Red Swarm not only achieved the best
+                  results, but also outperformed the experts' solutions in a
+                  total of 60 scenarios tested, with up to 19\% shorter travel
+                  times.}
+}
+
+ +
+@article{StoMcNutBai2016enhancing,
+  doi = {10.1126/science.aah6168},
+  year = 2016,
+  month = dec,
+  publisher = {American Association for the Advancement of Science ({AAAS})},
+  volume = 354,
+  number = 6317,
+  pages = {1240--1241},
+  author = {Stodden, Victoria and Marcia McNutt and David H. Bailey and Ewa Deelman and
+                  Yolanda Gil and Brooks Hanson and Michael A. Heroux and  John P. A. Ioannidis  and Michela Taufer},
+  title = {Enhancing reproducibility for computational methods},
+  journal = {Science}
+}
+
+ +
+@article{StoPri1997:de,
+  author = {Storn, Rainer and Price, Kenneth},
+  title = {Differential Evolution -- A Simple and Efficient Heuristic
+                  for Global Optimization over Continuous Spaces},
+  journal = {Journal of Global Optimization},
+  year = 1997,
+  volume = 11,
+  number = 4,
+  pages = {341--359},
+  annote = {Proposed differential evolution},
+  doi = {10.1023/A:1008202821328}
+}
+
+ +
+@article{StoSeiZha2018pnas,
+  author = {Stodden, Victoria and Jennifer Seiler and Zhaokun Ma},
+  title = {An empirical analysis of journal policy effectiveness for
+                  computational reproducibility},
+  doi = {10.1073/pnas.1708290115},
+  year = 2018,
+  month = mar,
+  journal = {Proceedings of the National Academy of Sciences},
+  volume = 115,
+  number = 11,
+  pages = {2584--2589}
+}
+
+ +
+@article{StrKir1991,
+  title = {Analysis of Finite Length Annealing Schedules},
+  author = { Philip N. Strenski  and  Scott Kirkpatrick },
+  journal = {Algorithmica},
+  volume = 6,
+  number = {1-6},
+  pages = {346--366},
+  year = 1991,
+  publisher = {Springer}
+}
+
+ +
+@article{StrLopBroLee2020,
+  title = {General Northern English: Exploring regional variation in the
+                  North of England with machine learning},
+  author = { Strycharczuk, Patrycja  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Brown, Georgina  and  Adrian Leemann },
+  journal = { Frontiers in Artificial Intelligence },
+  year = 2020,
+  volume = 3,
+  number = 48,
+  keywords = {vowels, accent features, dialect leveling, Random forest
+                  (bagging), Feature selecion},
+  doi = {10.3389/frai.2020.00048},
+  abstract = {In this paper, we present a novel computational approach to
+                  the analysis of accent variation. The case study is dialect
+                  leveling in the North of England, manifested as reduction of
+                  accent variation across the North and emergence of General
+                  Northern English (GNE), a pan-regional standard accent
+                  associated with middle-class speakers. We investigated this
+                  instance of dialect leveling using random forest
+                  classification, with audio data from a crowd-sourced corpus
+                  of 105 urban, mostly highly-educated speakers from five
+                  northern UK cities: Leeds, Liverpool, Manchester, Newcastle
+                  upon Tyne, and Sheffield. We trained random forest models to
+                  identify individual northern cities from a sample of other
+                  northern accents, based on first two formant measurements of
+                  full vowel systems. We tested the models using unseen
+                  data. We relied on undersampling, bagging (bootstrap
+                  aggregation) and leave-one-out cross-validation to address
+                  some challenges associated with the data set, such as
+                  unbalanced data and relatively small sample size. The
+                  accuracy of classification provides us with a measure of
+                  relative similarity between different pairs of cities, while
+                  calculating conditional feature importance allows us to
+                  identify which input features (which vowels and which
+                  formants) have the largest influence in the prediction. We do
+                  find a considerable degree of leveling, especially between
+                  Manchester, Leeds and Sheffield, although some differences
+                  persist. The features that contribute to these differences
+                  most systematically are typically not the ones discussed in
+                  previous dialect descriptions. We propose that the most
+                  systematic regional features are also not salient, and as
+                  such, they serve as sociolinguistic regional indicators. We
+                  supplement the random forest results with a more traditional
+                  variationist description of by-city vowel systems, and we use
+                  both sources of evidence to inform a description of the
+                  vowels of General Northern English.}
+}
+
+ +
+@article{Stu06:ejor,
+  author = { Thomas St{\"u}tzle },
+  title = {Iterated Local Search for the Quadratic Assignment Problem},
+  journal = {European Journal of Operational Research},
+  year = 2006,
+  volume = 174,
+  number = 3,
+  pages = {1519--1539}
+}
+
+ +
+@article{StuDor2002:tec,
+  author = { Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {A Short Convergence Proof for a Class of {ACO}
+                  Algorithms},
+  volume = 6,
+  number = 4,
+  pages = {358--365},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2002
+}
+
+ +
+@article{StuHoo2000:fgcs,
+  author = { Thomas St{\"u}tzle  and  Holger H. Hoos },
+  title = {{\MaxMinAntSystem}},
+  journal = {Future Generation Computer Systems},
+  year = 2000,
+  volume = 16,
+  number = 8,
+  pages = {889--914},
+  alias = {StuHoo00:fgcs}
+}
+
+ +
+@article{SuZhaYue2021enhanced,
+  title = {Enhanced Constraint Handling for Reliability-Constrained
+                  Multiobjective Testing Resource Allocation},
+  author = {Su, Zhaopin and Zhang, Guofu and Yue, Feng and Zhan, Dezhi and  Li, Miqing  and Li, Bin and  Xin Yao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2021,
+  number = 3,
+  pages = {537--551},
+  volume = 25,
+  publisher = {IEEE}
+}
+
+ +
+@article{SubBat2013,
+  author = { Anand Subramanian  and Maria Battarra},
+  title = {An Iterated Local Search Algorithm for the Travelling Salesman Problem with Pickups and Deliveries},
+  journal = {Journal of the Operational Research Society},
+  year = 2013,
+  volume = 64,
+  number = 3,
+  pages = {402--409}
+}
+
+ +
+@article{SubBatPot2014,
+  author = { Anand Subramanian  and Maria Battarra and  Chris N. Potts },
+  title = {An Iterated Local Search Heuristic for the Single Machine Total Weighted Tardiness Scheduling Problem with Sequence-dependent Setup Times},
+  journal = {International Journal of Production Research},
+  year = 2014,
+  volume = 52,
+  number = 9,
+  pages = {2729--2742}
+}
+
+ +
+@article{SuiZhuBur2018arxiv,
+  author = {Sui, Yanan and Zhuang, Vincent and Burdick, Joel W. and Yue,
+                  Yisong},
+  title = {Stagewise Safe {Bayesian} Optimization with {Gaussian}
+                  Processes},
+  journal = {Arxiv preprint arXiv:1806.07555},
+  year = 2018,
+  note = {Published as \cite{SuiZhuBur2018stageopt}},
+  url = {https://arxiv.org/abs/1806.07555},
+  abstract = {Enforcing safety is a key aspect of many problems pertaining
+                  to sequential decision making under uncertainty, which
+                  require the decisions made at every step to be both
+                  informative of the optimal decision and also safe. For
+                  example, we value both efficacy and comfort in medical
+                  therapy, and efficiency and safety in robotic control. We
+                  consider this problem of optimizing an unknown utility
+                  function with absolute feedback or preference feedback
+                  subject to unknown safety constraints. We develop an
+                  efficient safe Bayesian optimization algorithm, StageOpt,
+                  that separates safe region expansion and utility function
+                  maximization into two distinct stages. Compared to existing
+                  approaches which interleave between expansion and
+                  optimization, we show that StageOpt is more efficient and
+                  naturally applicable to a broader class of problems. We
+                  provide theoretical guarantees for both the satisfaction of
+                  safety constraints as well as convergence to the optimal
+                  utility value. We evaluate StageOpt on both a variety of
+                  synthetic experiments, as well as in clinical practice. We
+                  demonstrate that StageOpt is more effective than existing
+                  safe optimization approaches, and is able to safely and
+                  effectively optimize spinal cord stimulation therapy in our
+                  clinical experiments.},
+  keywords = {Safe Optimization, StageOpt}
+}
+
+ +
+@article{SunYenYi2019igd,
+  title = {{IGD} Indicator-based Evolutionary Algorithm for
+                  Many-objective Optimization Problems},
+  author = {Sun, Yanan and Yen, Gary G. and Yi, Zhang},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  publisher = {IEEE},
+  volume = 23,
+  number = 2,
+  pages = {173--187},
+  year = 2019,
+  doi = {10.1109/TEVC.2018.2791283}
+}
+
+ +
+@article{SupSefPar2000,
+  title = {A simulated annealing algorithm for multiobjective optimization},
+  author = {Suppapitnarm, A. and Seffen, K. A. and Parks, G. T. and Clarkson, P. J.},
+  journal = {Engineering Optimization},
+  volume = 33,
+  number = 1,
+  pages = {59--85},
+  year = 2000
+}
+
+ +
+@article{SuyVan1999least,
+  title = {Least Squares Support Vector Machine Classifiers},
+  author = {Suykens, Johan A. K. and Vandewalle, Joos},
+  journal = {Neural Processing Letters},
+  volume = 9,
+  number = 3,
+  pages = {293--300},
+  year = 1999,
+  publisher = {Springer},
+  doi = {10.1023/A:1018628609742},
+  keywords = {LS-SVM}
+}
+
+ +
+@article{SwaAdrBar2019extending,
+  author = { Jerry Swan  and  Steven Adriaensen  and Barwell, Adam D. and Hammond,
+                  Kevin and White, David R.},
+  title = {Extending the ``Open-Closed Principle'' to Automated
+                  Algorithm Configuration},
+  journal = {Evolutionary Computation},
+  year = 2019,
+  volume = 27,
+  number = 1,
+  pages = {173--193},
+  doi = {10.1162/evco_a_00245}
+}
+
+ +
+@article{SwaAdrBro2022metah,
+  author = { Jerry Swan  and  Steven Adriaensen  and Alexander
+                  E. I. Brownlee and Kevin Hammond and Colin
+                  G. Johnson and Ahmed Kheiri and Faustyna Krawiec and  Juan-Juli{\'a}n Merelo  and Leandro L. Minku and  Ender {\"O}zcan  and  Gisele Pappa  and  Pablo Garc{\'i}a-S{\'a}nchez  and  Kenneth S{\"o}rensen  and  Stefan Vo{\ss}  and  Markus Wagner  and David R. White},
+  doi = {10.1016/j.ejor.2021.05.042},
+  year = 2022,
+  month = mar,
+  volume = 297,
+  number = 2,
+  pages = {393--406},
+  title = {Metaheuristics ``In the Large''},
+  journal = {European Journal of Operational Research}
+}
+
+ +
+@article{SwaWooOzc2014cogcomp,
+  author = { Jerry Swan  and John R. Woodward and  Ender {\"O}zcan  and  Graham Kendall  and  Edmund K. Burke },
+  title = {Searching the Hyper-heuristic Design Space},
+  journal = {Cognitive Computation},
+  year = 2014,
+  volume = 6,
+  number = 1,
+  pages = {66--73},
+  month = mar,
+  doi = {10.1007/s12559-013-9201-8}
+}
+
+ +
+@article{SzuHar1987,
+  title = {Fast Simulated Annealing},
+  author = { Harold Szu  and  Ralph Hartley },
+  journal = {Physics Letters A},
+  volume = 122,
+  number = 3,
+  pages = {157--162},
+  year = 1987,
+  publisher = {Elsevier}
+}
+
+ +
+@article{Tai90,
+  author = { {\'E}ric D. Taillard },
+  title = {Some Efficient Heuristic Methods for the Flow Shop
+                  Sequencing Problem},
+  journal = {European Journal of Operational Research},
+  year = 1990,
+  volume = 47,
+  number = 1,
+  pages = {65--74}
+}
+
+ +
+@article{Tai91,
+  author = { {\'E}ric D. Taillard },
+  title = {Robust Taboo Search for the Quadratic Assignment Problem},
+  journal = {Parallel Computing},
+  year = 1991,
+  volume = 17,
+  number = {4-5},
+  pages = {443--455},
+  annote = {faster 2-exchange delta evaluation in QAP}
+}
+
+ +
+@article{Tai93,
+  author = { {\'E}ric D. Taillard },
+  title = {Benchmarks for Basic Scheduling Problems},
+  journal = {European Journal of Operational Research},
+  year = 1993,
+  volume = 64,
+  number = 2,
+  pages = {278--285}
+}
+
+ +
+@article{Tai95,
+  author = { {\'E}ric D. Taillard },
+  title = {Comparison of Iterative Searches for the Quadratic
+                  Assignment Problem},
+  journal = {Location Science},
+  year = 1995,
+  volume = 3,
+  number = 2,
+  pages = {87--105}
+}
+
+ +
+@article{Talbi02,
+  author = { Talbi, El-Ghazali },
+  title = {A Taxonomy of Hybrid Metaheuristics},
+  journal = {Journal of Heuristics},
+  year = 2002,
+  volume = 8,
+  number = 5,
+  pages = {541--564}
+}
+
+ +
+@article{Tam1992,
+  title = {A Simulated Annealing Algorithm for Allocating Space to Manufacturing Cells},
+  author = { Tam, Kar Yan },
+  journal = {International Journal of Production Research},
+  volume = 30,
+  number = 1,
+  pages = {63--87},
+  year = 1992,
+  publisher = {Taylor \& Francis}
+}
+
+ +
+@article{TamJonEld1995,
+  author = {Tamiz, M. and Jones, D. F. and El-Darzi, E.},
+  title = {A review of {Goal} {Programming} and its applications},
+  journal = {Annals of Operations Research},
+  year = 1995,
+  volume = 58,
+  number = 1,
+  pages = {39--53},
+  month = jan,
+  doi = {10.1007/BF02032309},
+  abstract = {This paper presents a review of the current literature on the
+                  branch of multi-criteria decision modelling known as Goal
+                  Programming (GP). The result of our indepth investigations of
+                  the two main GP methods, lexicographic and weighted GP
+                  together with their distinct application areas is
+                  reported. Some guidelines to the scope of GP as an
+                  application tool are given and methods of determining which
+                  problem areas are best suited to the different GP approaches
+                  are proposed. The correlation between the method of assigning
+                  weights and priorities and the standard of the results is
+                  also ascertained.},
+  language = {en},
+  keywords = {Goal Programming, lexicographic, weighted}
+}
+
+ +
+@article{TanAra2013,
+  author = {Shunji Tanaka and Mituhiko Araki},
+  title = {An Exact Algorithm for the Single-machine Total Weighted Tardiness
+               Problem with Sequence-dependent Setup Times},
+  journal = {Computers \& Operations Research},
+  year = 2013,
+  volume = 40,
+  number = 1,
+  pages = {344--352}
+}
+
+ +
+@article{TanIsh2020re,
+  author = {Ryoji Tanabe and  Ishibuchi, Hisao },
+  title = {An easy-to-use real-world multi-objective optimization
+                  problem suite},
+  journal = {Applied Soft Computing},
+  volume = 89,
+  pages = 106078,
+  year = 2020,
+  annote = {Proposed the RE benchmark suite}
+}
+
+ +
+@article{TanIshOya2017,
+  author = {Ryoji Tanabe and  Ishibuchi, Hisao  and Akira Oyama},
+  journal = {{IEEE} Access},
+  title = {Benchmarking Multi- and Many-Objective Evolutionary
+                  Algorithms Under Two Optimization Scenarios},
+  year = 2017,
+  volume = 5,
+  pages = {19597--19619},
+  annote = {compared a number of MOEAs using a wide range of numbers of
+                  objectives and stopping criteria, with and without archivers; unbounded archive}
+}
+
+ +
+@article{TanWan2006,
+  author = {Lixin Tang and Xianpeng Wang},
+  title = {Iterated local search algorithm based on very large-scale
+                  neighborhood for prize-collecting vehicle routing problem},
+  journal = {International Journal of Advanced Manufacturing Technology},
+  year = 2006,
+  volume = 29,
+  number = 11,
+  pages = {1246--1258}
+}
+
+ +
+@article{Tarquin89,
+  author = { A. J. Tarquin  and  J. Dowdy },
+  title = {Optimal pump operation in water distribution},
+  journal = { Journal of Hydraulic Engineering, {ASCE}},
+  year = 1989,
+  volume = 115,
+  number = 2,
+  pages = {158--169 or 496--501},
+  month = feb,
+  note = {}
+}
+
+ +
+@article{TasKizPanSug2017,
+  author = {M. F. Tasgetiren and D. Kizilay and  Quan-Ke Pan  and  Ponnuthurai N. Suganthan },
+  title = {Iterated Greedy Algorithms for the Blocking Flowshop
+                  Scheduling Problem with Makespan Criterion},
+  journal = {Computers \& Operations Research},
+  year = 2017,
+  volume = 77,
+  pages = {111--126}
+}
+
+ +
+@article{TasLiaSevGen2007,
+  author = {M. Fatih Tasgetiren and Yun-Chia Liang and Mehmet Sevkli and
+                  Gunes Gencyilmaz},
+  title = {A particle swarm optimization algorithm for makespan and
+                  total flowtime minimization in the permutation flowshop
+                  sequencing problem},
+  journal = {European Journal of Operational Research},
+  year = 2007,
+  volume = 177,
+  number = 3,
+  pages = {1930--1947},
+  doi = {10.1016/j.ejor.2005.12.024}
+}
+
+ +
+@article{TasPanSugBuy2013:cor,
+  author = {M. Fatih Tasgetiren and  Quan-Ke Pan  and  Ponnuthurai N. Suganthan  and Ozge
+                  Buyukdagli},
+  title = {A variable iterated greedy algorithm with differential
+                  evolution for the no-idle permutation flowshop scheduling
+                  problem},
+  journal = {Computers \& Operations Research},
+  year = 2013,
+  volume = 40,
+  number = 7,
+  pages = {1729--1743}
+}
+
+ +
+@article{TayHo2008cie,
+  author = {Joc Cing Tay and Nhu Binh Ho},
+  title = {Evolving dispatching rules using genetic programming for
+                  solving multi-objective flexible job-shop problems},
+  journal = {Computers and Industrial Engineering},
+  volume = 54,
+  number = 3,
+  pages = {453 -- 473},
+  year = 2008,
+  doi = {10.1016/j.cie.2007.08.008}
+}
+
+ +
+@article{TeiCovStuCun11itor,
+  author = {Cristina Teixeira and Jos{\'e} Covas and  Thomas St{\"u}tzle  and  Ant{\'o}nio Gaspar{-}Cunha },
+  title = {Engineering an Efficient Two-Phase Local Search for the
+                  Co-Rotating Twin-Screw Configuration Problem},
+  journal = {International Transactions in Operational Research},
+  year = 2011,
+  volume = 18,
+  number = 2,
+  pages = {271--291}
+}
+
+ +
+@article{TeiCovStuCun12eo,
+  author = {Cristina Teixeira and Jos{\'e} Covas and  Thomas St{\"u}tzle  and  Ant{\'o}nio Gaspar{-}Cunha },
+  title = {Multi-Objective Ant Colony Optimization for Solving the
+Twin-Screw Extrusion Configuration Problem},
+  journal = {Engineering Optimization},
+  year = 2012,
+  volume = 44,
+  number = 3,
+  pages = {351--371}
+}
+
+ +
+@article{TeiCovStuCun14asc,
+  author = {Cristina Teixeira and Jos{\'e} Covas and  Thomas St{\"u}tzle  and  Ant{\'o}nio Gaspar{-}Cunha },
+  title = {Hybrid Algorithms for the Twin-Screw Extrusion Configuration
+  Problem},
+  journal = {Applied Soft Computing},
+  year = 2014,
+  volume = 23,
+  pages = {298--307}
+}
+
+ +
+@article{Teklu2007,
+  author = {Teklu, Fitsum and Sumalee, Agachai and Watling, David},
+  doi = {10.1111/j.1467-8667.2006.00468.x},
+  journal = {Computer-Aided Civil and Infrastructure Engineering},
+  month = jan,
+  number = 1,
+  pages = {31--43},
+  title = {A Genetic Algorithm Approach for Optimizing Traffic Control
+                  Signals Considering Routing},
+  volume = 22,
+  year = 2007
+}
+
+ +
+@article{TenSilLan2000,
+  author = {J. B. Tenenbaum and V. D. Silva and J. C. Langford},
+  title = {A global geometric framework for nonlinear dimensionality
+                  reduction},
+  journal = {Science},
+  volume = 290,
+  number = 5500,
+  year = 2000,
+  pages = {2319--2323}
+}
+
+ +
+@article{TeoAbb2004ec,
+  author = {J. Teo and  Abbass, Hussein A. },
+  title = {Automatic generation of controllers for embodied
+                  legged organisms: A {Pareto} evolutionary
+                  multi-objective approach},
+  journal = {Evolutionary Computation},
+  year = 2004,
+  volume = 12,
+  number = 3,
+  pages = {355--394},
+  doi = {10.1162/1063656041774974},
+  alias = {Teo:2004vg}
+}
+
+ +
+@article{TerSumTam2021auto,
+  title = {Black-Box Optimization for Automated Discovery},
+  author = {Terayama, Kei and Sumita, Masato and Tamura, Ryo and Tsuda,
+                  Koji},
+  year = 2021,
+  month = mar,
+  journal = {Accounts of Chemical Research},
+  volume = 54,
+  number = 6,
+  pages = {1334--1346},
+  publisher = {American Chemical Society},
+  doi = {10.1021/acs.accounts.0c00713},
+  abstract = {In chemistry and materials science, researchers and engineers
+                  discover, design, and optimize chemical compounds or
+                  materials with their professional knowledge and
+                  techniques. At the highest level of abstraction, this process
+                  is formulated as black-box optimization. For instance, the
+                  trial-and-error process of synthesizing various molecules for
+                  better material properties can be regarded as optimizing a
+                  black-box function describing the relation between a chemical
+                  formula and its properties. Various black-box optimization
+                  algorithms have been developed in the machine learning and
+                  statistics communities. Recently, a number of researchers
+                  have reported successful applications of such algorithms to
+                  chemistry. They include the design of photofunctional
+                  molecules and medical drugs, optimization of thermal emission
+                  materials and high Li-ion conductive solid electrolytes, and
+                  discovery of a new phase in inorganic thin films for solar
+                  cells.There are a wide variety of algorithms available for
+                  black-box optimization, such as Bayesian optimization,
+                  reinforcement learning, and active learning. Practitioners
+                  need to select an appropriate algorithm or, in some cases,
+                  develop novel algorithms to meet their demands. It is also
+                  necessary to determine how to best combine machine learning
+                  techniques with quantum mechanics- and molecular
+                  mechanics-based simulations, and experiments. In this
+                  Account, we give an overview of recent studies regarding
+                  automated discovery, design, and optimization based on
+                  black-box optimization. The Account covers the following
+                  algorithms: Bayesian optimization to optimize the chemical or
+                  physical properties, an optimization method using a quantum
+                  annealer, best-arm identification, gray-box optimization, and
+                  reinforcement learning. In addition, we introduce active
+                  learning and boundless objective-free exploration, which may
+                  not fall into the category of black-box optimization.Data
+                  quality and quantity are key for the success of these
+                  automated discovery techniques. As laboratory automation and
+                  robotics are put forward, automated discovery algorithms
+                  would be able to match human performance at least in some
+                  domains in the near future.}
+}
+
+ +
+@article{Thi2011cw,
+  author = {Patrick Thibodeau},
+  title = {Machine-based decision-making is coming},
+  journal = {Computer World},
+  year = 2011,
+  month = nov,
+  url = {http://www.computerworld.com/s/article/359630/Machine_Based_Decision_Making_Is_Coming},
+  note = {Last accessed: 15 January 2014}
+}
+
+ +
+@article{ThiMieKorMol2009ec,
+  author = { Lothar Thiele  and  Kaisa Miettinen  and  Pekka Korhonen  and  Molina, Juli{\'a}n },
+  title = {A Preference-Based Evolutionary Algorithm for Multi-Objective
+                  Optimization},
+  journal = {Evolutionary Computation},
+  volume = 17,
+  number = 3,
+  pages = {411--436},
+  year = 2009,
+  doi = {10.1162/evco.2009.17.3.411},
+  abstract = { Abstract In this paper, we discuss the idea of incorporating
+                  preference information into evolutionary multi-objective
+                  optimization and propose a preference-based evolutionary
+                  approach that can be used as an integral part of an
+                  interactive algorithm. One algorithm is proposed in the
+                  paper. At each iteration, the decision maker is asked to give
+                  preference information in terms of his or her reference point
+                  consisting of desirable aspiration levels for objective
+                  functions. The information is used in an evolutionary
+                  algorithm to generate a new population by combining the
+                  fitness function and an achievement scalarizing function. In
+                  multi-objective optimization, achievement scalarizing
+                  functions are widely used to project a given reference point
+                  into the Pareto optimal set. In our approach, the next
+                  population is thus more concentrated in the area where more
+                  preferred alternatives are assumed to lie and the whole
+                  Pareto optimal set does not have to be generated with equal
+                  accuracy. The approach is demonstrated by numerical
+                  examples. }
+}
+
+ +
+@article{TiaCheZha2017indicator,
+  title = {An Indicator-Based Multiobjective Evolutionary Algorithm With
+                  Reference Point Adaptation for Better Versatility},
+  author = {Ye Tian and Ran Cheng and Xingyi Zhang and Fan Cheng and  Yaochu Jin },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2018,
+  volume = 22,
+  number = 4,
+  pages = {609--622},
+  doi = {10.1109/TEVC.2017.2749619},
+  annote = {IGD-based archiver}
+}
+
+ +
+@article{TinRaoLoo2003joh,
+  author = { Tiew-On Ting  and  M. V. C. Rao  and  C. K. Loo  and  S. S. Ngu },
+  title = {Solving Unit Commitment Problem Using Hybrid
+                  Particle Swarm Optimization},
+  journal = {Journal of Heuristics},
+  year = 2003,
+  volume = 9,
+  number = 6,
+  pages = {507--520},
+  doi = {10.1023/B:HEUR.0000012449.84567.1a}
+}
+
+ +
+@article{TiwFadDeb2011amga2,
+  title = {{AMGA2}: Improving the performance of the archive-based
+                  micro-genetic algorithm for multi-objective optimization},
+  author = {Tiwari, Santosh and Fadel, Georges and  Kalyanmoy Deb },
+  journal = {Engineering Optimization},
+  year = 2011,
+  number = 4,
+  pages = {377--401},
+  volume = 43
+}
+
+ +
+@article{TkiMonTer2002ejor,
+  author = { V. {T'Kindt}  and  Nicolas Monmarch{\'e}  and F. Tercinet and D. La{\"u}gt},
+  title = {An ant colony optimization algorithm to solve a
+                  2-machine bicriteria flowshop scheduling problem},
+  journal = {European Journal of Operational Research},
+  year = 2002,
+  volume = 142,
+  number = 2,
+  pages = {250--257}
+}
+
+ +
+@article{TomKad2019decomposition,
+  title = {Decomposition-based interactive evolutionary algorithm for
+                  multiple objective optimization},
+  author = { Tomczyk, Micha{\l} K  and  Kadzi{\'n}ski, Mi{\l}osz  },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 24,
+  number = 2,
+  pages = {320--334},
+  year = 2019,
+  publisher = {IEEE},
+  doi = {10.1109/TEVC.2019.2915767},
+  abstract = {We propose a decomposition-based interactive evolutionary
+                  algorithm (EA) for multiple objective optimization. During an
+                  evolutionary search, a decision maker (DM) is asked to
+                  compare pairwise solutions from the current population. Using
+                  the Monte Carlo simulation, the proposed algorithm generates
+                  from a uniform distribution a set of instances of the
+                  preference model compatible with such an indirect preference
+                  information. These instances are incorporated as the search
+                  directions with the aim of systematically converging a
+                  population toward the DMs most preferred region of the Pareto
+                  front. The experimental comparison proves that the proposed
+                  decomposition-based method outperforms the state-of-the-art
+                  interactive counterparts of the dominance-based EAs. We also
+                  show that the quality of constructed solutions is highly
+                  affected by the form of the incorporated preference model.},
+  keywords = {interactive multi-objective; decision-making}
+}
+
+ +
+@article{TomKad2019emosor,
+  author = { Tomczyk, Micha{\l} K  and  Kadzi{\'n}ski, Mi{\l}osz  },
+  title = {{EMOSOR}: Evolutionary multiple objective optimization guided
+                  by interactive stochastic ordinal regression},
+  journal = {Computers \& Operations Research},
+  volume = 108,
+  pages = {134--154},
+  year = 2019,
+  doi = {10.1016/j.cor.2019.04.008},
+  keywords = {Multiple objective optimization, Interactive evolutionary
+                  hybrids, Stochastic ordinal regression, Preference
+                  disaggregation, Pairwise comparisons, Active learning},
+  abstract = {We propose a family of algorithms, called EMOSOR, combining
+                  Evolutionary Multiple Objective Optimization with Stochastic
+                  Ordinal Regression. The proposed methods ask the Decision
+                  Maker (DM) to holistically compare, at regular intervals, a
+                  pair of solutions, and use the Monte Carlo simulation to
+                  construct a set of preference model instances compatible with
+                  such indirect and incomplete information. The specific
+                  variants of EMOSOR are distinguished by the following three
+                  aspects. Firstly, they make use of two different preference
+                  models, i.e., either an additive value function or a
+                  Chebyshev function. Secondly, they aggregate the
+                  acceptability indices derived from the stochastic analysis in
+                  various ways, and use thus constructed indicators or
+                  relations to sort the solutions obtained in each
+                  generation. Thirdly, they incorporate different active
+                  learning strategies for selecting pairs of solutions to be
+                  critically judged by the DM. The extensive computational
+                  experiments performed on a set of benchmark optimization
+                  problems reveal that EMOSOR is able to bias an evolutionary
+                  search towards a part of the Pareto front being the most
+                  relevant to the DM, outperforming in this regard the
+                  state-of-the-art interactive evolutionary hybrids. Moreover,
+                  we demonstrate that the performance of EMOSOR improves in
+                  case the forms of a preference model used by the method and
+                  the DM's value system align. Furthermore, we discuss how
+                  vastly incorporation of different indicators based on the
+                  stochastic acceptability indices influences the quality of
+                  both the best constructed solution and an entire
+                  population. Finally, we demonstrate that our novel
+                  questioning strategies allow to reduce a number of
+                  interactions with the DM until a high-quality solution is
+                  constructed or, alternatively, to discover a better solution
+                  after the same number of interactions.}
+}
+
+ +
+@article{TomKad2021ciemod,
+  author = { Tomczyk, Micha{\l} K  and  Kadzi{\'n}ski, Mi{\l}osz  },
+  title = {Decomposition-based co-evolutionary algorithm for interactive
+                  multiple objective optimization},
+  journal = {Information Sciences},
+  volume = 549,
+  pages = {178--199},
+  year = 2021,
+  doi = {10.1016/j.ins.2020.11.030},
+  keywords = {Evolutionary multiple objective optimization, Co-evolution,
+                  Decomposition, Indirect preference information, Preference
+                  learning},
+  abstract = {We propose a novel co-evolutionary algorithm for interactive
+                  multiple objective optimization, named CIEMO/D. It aims at
+                  finding a region in the Pareto front that is highly relevant
+                  to the Decision Maker (DM). For this reason, CIEMO/D asks the
+                  DM, at regular intervals, to compare pairs of solutions from
+                  the current population and uses such preference information
+                  to bias the evolutionary search. Unlike the existing
+                  interactive evolutionary algorithms dealing with just a
+                  single population, CIEMO/D co-evolves a pool of
+                  subpopulations in a steady-state decomposition-based
+                  evolutionary framework. The evolution of each subpopulation
+                  is driven by the use of a different preference model. In this
+                  way, the algorithm explores various regions in the objective
+                  space, thus increasing the chances of finding DM's most
+                  preferred solution. To improve the pace of the evolutionary
+                  search, CIEMO/D allows for the migration of solutions between
+                  different subpopulations. It also dynamically alters the
+                  subpopulations' size based on compatibility between the
+                  incorporated preference models and the decision examples
+                  supplied by the DM. The extensive experimental evaluation
+                  reveals that CIEMO/D can successfully adjust to different
+                  DM's decision policies. We also compare CIEMO/D with selected
+                  state-of-the-art interactive evolutionary hybrids that make
+                  use of the DM's pairwise comparisons, demonstrating its high
+                  competitiveness.}
+}
+
+ +
+@article{TorRosKefetal2010:si,
+  author = {C. E. Torres and L. F. Rossi and J. Keffer and K. Li and
+                  C.-C. Shen},
+  title = {Modeling, analysis and simulation of ant-based network
+                  routing protocols},
+  journal = {Swarm Intelligence},
+  volume = 4,
+  number = 3,
+  pages = {221--244},
+  year = 2010
+}
+
+ +
+@article{TraMeh2009eo,
+  author = { Heike Trautmann  and J\"{o}rn Mehnen},
+  title = {Preference-based {Pareto} optimization in certain and
+                  noisy environments},
+  journal = {Engineering Optimization},
+  volume = 41,
+  number = 1,
+  pages = {23--38},
+  month = jan,
+  year = 2009
+}
+
+ +
+@article{TriLop2015plos,
+  author = { Vito Trianni  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Advantages of Task-Specific Multi-Objective Optimisation in
+                  Evolutionary Robotics},
+  journal = {PLoS One},
+  year = 2015,
+  volume = 10,
+  number = 8,
+  pages = {e0136406},
+  doi = {10.1371/journal.pone.0136406},
+  abstract = {The application of multi-objective optimisation to
+                  evolutionary robotics is receiving increasing attention. A
+                  survey of the literature reveals the different possibilities
+                  it offers to improve the automatic design of efficient and
+                  adaptive robotic systems, and points to the successful
+                  demonstrations available for both task-specific and
+                  task-agnostic approaches (i.e., with or without reference to
+                  the specific design problem to be tackled). However, the
+                  advantages of multi-objective approaches over
+                  single-objective ones have not been clearly spelled out and
+                  experimentally demonstrated. This paper fills this gap for
+                  task-specific approaches: starting from well-known results in
+                  multi-objective optimisation, we discuss how to tackle
+                  commonly recognised problems in evolutionary robotics. In
+                  particular, we show that multi-objective optimisation (i)
+                  allows evolving a more varied set of behaviours by exploring
+                  multiple trade-offs of the objectives to optimise, (ii)
+                  supports the evolution of the desired behaviour through the
+                  introduction of objectives as proxies, (iii) avoids the
+                  premature convergence to local optima possibly introduced by
+                  multi-component fitness functions, and (iv) solves the
+                  bootstrap problem exploiting ancillary objectives to guide
+                  evolution in the early phases. We present an experimental
+                  demonstration of these benefits in three different case
+                  studies: maze navigation in a single robot domain, flocking
+                  in a swarm robotics context, and a strictly collaborative
+                  task in collective robotics.}
+}
+
+ +
+@article{TriNol2011al,
+  author = { Vito Trianni  and Nolfi, S.},
+  title = {Engineering the evolution of self-organizing
+                  behaviors in swarm robotics: A case study},
+  journal = {Artificial Life},
+  year = 2011,
+  volume = 17,
+  number = 3,
+  pages = {183--202},
+  alias = {Trianni:2011wl}
+}
+
+ +
+@article{TriSriSanGho2016survey,
+  title = {A survey of multiobjective evolutionary algorithms based on
+                  decomposition},
+  author = {Trivedi, Anupam and Srinivasan, Dipti and Sanyal, Krishnendu
+                  and Ghosh, Abhiroop},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2016,
+  number = 3,
+  pages = {440--462},
+  volume = 21,
+  publisher = {IEEE}
+}
+
+ +
+@article{TseLin09,
+  author = { L.-Y. Tseng  and  Y.-T. Lin },
+  title = {A hybrid genetic local search algorithm for the permutation
+                  flowshop scheduling problem},
+  journal = {European Journal of Operational Research},
+  volume = 198,
+  number = 1,
+  pages = {84--92},
+  year = 2009
+}
+
+ +
+@article{Tsu2007tjsai,
+  author = {Tsutsui, S.},
+  title = {Ant Colony Optimization with Cunning Ants},
+  journal = {Transactions of the Japanese Society for Artificial Intelligence},
+  keywords = {ant colony optimization, traveling salesman problem, cunning
+                  ant, donor ant, local search},
+  year = 2007,
+  volume = 22,
+  pages = {29--36},
+  doi = {10.1527/tjsai.22.29}
+}
+
+ +
+@article{TugThrFil2017jbe,
+  title = {Conceptual Design of Modular Bridges Including Layout
+                  Optimization and Component Reusability},
+  author = {Tugilimana, Alexis and Thrall, Ashley P. and Filomeno Coelho,
+                  Rajan},
+  journal = {Journal of Bridge Engineering},
+  volume = 22,
+  number = 11,
+  pages = 04017094,
+  year = 2017,
+  keywords = {scenario-based},
+  doi = {10.1061/(ASCE)BE.1943-5592.0001138},
+  publisher = {American Society of Civil Engineers}
+}
+
+ +
+@article{TurSorHva2021meta,
+  author = {Renata Turkeš and  Kenneth S{\"o}rensen  and Lars Magnus Hvattum},
+  title = {Meta-analysis of metaheuristics: Quantifying the effect of
+                  adaptiveness in adaptive large neighborhood search},
+  journal = {European Journal of Operational Research},
+  volume = 292,
+  number = 2,
+  pages = {423--42},
+  year = 2021,
+  doi = {10.1016/j.ejor.2020.10.045},
+  keywords = {Metaheuristics, Meta-analysis, Adaptive large neighborhood
+                  search}
+}
+
+ +
+@article{TusFil2014mpe,
+  author = { Tea Tu{\v s}ar  and Bogdan Filipi{\v c}},
+  title = {Visualizing Exact and Approximated {3D} Empirical Attainment
+                  Functions},
+  journal = {Mathematical Problems in Engineering},
+  volume = 2014,
+  note = {Article ID 569346, 18 pages},
+  year = 2014,
+  doi = {10.1155/2014/569346},
+  comment = {This is how Tea Tusar suggests to encode this entry in
+                  bibtex.}
+}
+
+ +
+@article{TusFil2015tec,
+  title = {Visualization of {Pareto} front approximations in
+                  evolutionary multiobjective optimization: A critical review
+                  and the prosection method},
+  author = { Tea Tu{\v s}ar  and Bogdan Filipi{\v c}},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  volume = 19,
+  number = 2,
+  pages = {225--245},
+  doi = {10.1109/TEVC.2014.2313407}
+}
+
+ +
+@article{Tuyttens00,
+  author = {D. Tuyttens and  Jacques Teghem  and  Philippe Fortemps  and K. Van Nieuwenhuyze},
+  title = {Performance of the {MOSA} Method for the Bicriteria Assignment Problem},
+  journal = {Journal of Heuristics},
+  year = 2000,
+  volume = 6,
+  pages = {295--310}
+}
+
+ +
+@article{TveKah1974judgment,
+  title = {Judgment under uncertainty: Heuristics and biases},
+  author = { Tversky, Amos  and  Kahneman, Daniel },
+  journal = {Science},
+  volume = 185,
+  number = 4157,
+  pages = {1124--1131},
+  year = 1974,
+  publisher = {American association for the advancement of science}
+}
+
+ +
+@article{TveKah1991,
+  author = { Tversky, Amos  and  Kahneman, Daniel },
+  year = 1991,
+  title = {Loss aversion in riskless choice: a reference-dependent
+                  model},
+  journal = {The Quarterly Journal of Economics},
+  volume = 106,
+  number = 4,
+  pages = {1039--1061}
+}
+
+ +
+@article{Tversky1972,
+  author = { Tversky, Amos },
+  year = 1972,
+  title = {Choice by elimination},
+  journal = {Journal of Mathematical Psychology},
+  volume = 9,
+  number = 4,
+  pages = {341--367}
+}
+
+ +
+@article{TwoStuDorManBir2010is,
+  author = {Colin Twomey and  Thomas St{\"u}tzle  and  Marco Dorigo  and Max Manfrin and  Mauro Birattari },
+  title = {An Analysis of Communication Policies for Homogeneous Multi-colony {ACO} Algorithms},
+  journal = {Information Sciences},
+  year = 2010,
+  volume = 180,
+  number = 12,
+  pages = {2390--2404},
+  doi = {10.1016/j.ins.2010.02.017}
+}
+
+ +
+@article{UluTeg1995,
+  author = { E. Ulungu  and  Jacques Teghem },
+  title = {The two phases method: An efficient procedure
+                  to solve bi-objective combinatorial optimization
+                  problems},
+  journal = {Foundations of Computing and Decision Sciences},
+  volume = 20,
+  number = 2,
+  pages = {149--165},
+  year = 1995
+}
+
+ +
+@article{UluTegFor1999,
+  author = { E. Ulungu  and  Jacques Teghem  and  Fortemps, P. H. and Tuyttens, D.},
+  title = {{MOSA} method: a tool for solving multiobjective
+                  combinatorial optimization problems},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  number = 4,
+  pages = {221--236},
+  volume = 8,
+  year = 1999
+}
+
+ +
+@article{UrlRuiSef2010,
+  author = {Thijs Urlings and  Rub{\'e}n Ruiz  and F. Sivrikaya-\c{S}erifo\u{g}lu},
+  title = {Genetic Algorithms for Complex Hybrid Flexible Flow Line Problems},
+  journal = {International Journal of Metaheuristics},
+  year = 2010,
+  volume = 1,
+  number = 1,
+  pages = {30--54}
+}
+
+ +
+@article{UrlRuiStu2010ejor,
+  author = {Thijs Urlings and  Rub{\'e}n Ruiz  and  Thomas St{\"u}tzle },
+  title = {Shifting Representation Search for Hybrid Flexible Flowline Problems},
+  journal = {European Journal of Operational Research},
+  year = 2010,
+  volume = 207,
+  number = 2,
+  pages = {1086--1095},
+  doi = {10.1016/j.ejor.2010.05.041}
+}
+
+ +
+@article{VaeAarLen1998cor,
+  author = { Rob J. M. Vaessens  and  Emile H. L. Aarts  and  Jan Karel Lenstra },
+  title = {A Local Search Template},
+  journal = {Computers \& Operations Research},
+  volume = 25,
+  number = 11,
+  year = 1998,
+  pages = {969--979},
+  doi = {10.1016/S0305-0548(97)00093-2}
+}
+
+ +
+@article{ValMarWanDeb2021mipdom,
+  title = {Analyzing dominance move ({MIP}-{DoM}) indicator for
+                  multi-and many-objective optimization},
+  author = {do Val Lopes, Claudio Lucio and Martins, Fl{\'a}vio
+                  Vin{\'i}cius Cruzeiro and  Wanner, Elizabeth F.  and  Kalyanmoy Deb },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2021,
+  publisher = {IEEE}
+}
+
+ +
+@article{ValRui2010,
+  author = { Eva Vallada  and  Rub{\'e}n Ruiz },
+  title = {Genetic algorithms with path relinking for the minimum tardiness permutation flowshop problem},
+  journal = {Omega},
+  volume = 38,
+  number = {1--2},
+  pages = {57--67},
+  year = 2010,
+  doi = {10.1016/j.omega.2009.04.002}
+}
+
+ +
+@article{ValRuiFra2015,
+  author = { Eva Vallada  and  Rub{\'e}n Ruiz  and  Jose M. Frami{\~n}{\'a}n },
+  title = {New hard benchmark for flowshop scheduling problems minimising makespan },
+  journal = {European Journal of Operational Research},
+  volume = 240,
+  number = 3,
+  pages = {666--677},
+  year = 2015,
+  doi = {10.1016/j.ejor.2014.07.033}
+}
+
+ +
+@article{ValRuiMin08,
+  author = { Eva Vallada  and  Rub{\'e}n Ruiz  and  Gerardo Minella },
+  title = {Minimising total tardiness in the m-machine flowshop
+                  problem: A review and evaluation of heuristics and
+                  metaheuristics},
+  journal = {Computers \& Operations Research},
+  volume = 35,
+  number = 4,
+  pages = {1350--1373},
+  year = 2008
+}
+
+ +
+@article{VanMou2014:ejor,
+  author = { Pieter Vansteenwegen  and Manuel Mateo},
+  title = {An Iterated Local Search Algorithm for the Single-vehicle Cyclic Inventory
+               Routing Problem},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 237,
+  number = 3,
+  pages = {802--813}
+}
+
+ +
+@article{VanSouBer2009:cor,
+  author = { Pieter Vansteenwegen  and Wouter Souffriau and  Vanden Berghe, Greet   and  Dirk Van Oudheusden },
+  title = {Iterated Local Search for the Team Orienteering Problem with Time
+               Tindows},
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  volume = 36,
+  number = 12,
+  pages = {3281--3290}
+}
+
+ +
+@article{Vanschoren2014openml,
+  author = { Joaquin Vanschoren  and  van Rijn, Jan N.  and  Bernd Bischl  and Luis Torgo},
+  title = {{OpenML}: Networked Science in Machine Learning},
+  journal = {{ACM} {SIGKDD} Explorations Newsletter},
+  year = 2014,
+  volume = 15,
+  number = 2,
+  pages = {49--60},
+  month = jun,
+  doi = {10.1145/2641190.2641198}
+}
+
+ +
+@article{VarDel2000effsize,
+  author = {Vargha, A. and Delaney, H. D.},
+  title = {A critique and improvement of the {CL} common language effect
+                  size statistics of {McGraw} and {Wong}},
+  journal = {Journal of Educational and Behavioral Statistics},
+  year = 2000,
+  volume = 25,
+  number = 2,
+  pages = {101--132},
+  keywords = {effect size test, A12 test}
+}
+
+ +
+@article{VarRaj05,
+  author = { T. K. Varadharajan  and C. Rajendran},
+  title = {A multi-objective simulated-annealing algorithm for
+                  scheduling in flowshops to minimize the makespan and
+                  total flowtime of jobs},
+  journal = {European Journal of Operational Research},
+  volume = 167,
+  number = 3,
+  pages = {772--795},
+  year = 2005,
+  alias = {Varadharajan05}
+}
+
+ +
+@article{VasShaPar2017transformer,
+  title = {Attention Is All You Need},
+  author = {Ashish Vaswani and Noam Shazeer and Niki Parmar and Jakob
+                  Uszkoreit and Llion Jones and Aidan N. Gomez and Lukasz
+                  Kaiser and Illia Polosukhin},
+  journal = {Arxiv preprint arXiv:1706.03762},
+  year = 2017,
+  url = {http://arxiv.org/abs/1706.03762},
+  abstract = {The dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 English-to-German translation task, improving over the existing best results, including ensembles by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.8 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data.}
+}
+
+ +
+@article{VasSim2010jwrpm,
+  author = {A. Vasan and  Slobodan P. Simonovic },
+  title = {Optimization of Water Distribution Network Design
+                  Using Differential Evolution},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2010,
+  volume = 136,
+  number = 2,
+  pages = {279--287}
+}
+
+ +
+@article{VasYan2005effic,
+  title = {Efficiently computing succinct trade-off curves},
+  author = {Vassilvitskii, Sergei and  Mihalis Yannakakis },
+  journal = {Theoretical Computer Science},
+  year = 2005,
+  number = {2-3},
+  pages = {334--356},
+  volume = 348,
+  publisher = {Elsevier}
+}
+
+ +
+@article{VazOch2010,
+  author = { J. A. V{\'a}zquez-Rodr{\'i}guez  and  Gabriela Ochoa },
+  title = {On the Automatic Discovery of Variants of the {NEH} Procedure
+                  for Flow Shop Scheduling Using Genetic Programming},
+  journal = {Journal of the Operational Research Society},
+  year = 2010,
+  volume = 62,
+  number = 2,
+  pages = {381--396}
+}
+
+ +
+@article{VazPaqFonKla2015rep,
+  author = {Daniel Vaz and  Lu{\'i}s Paquete  and  Carlos M. Fonseca  and  Kathrin Klamroth  and Michael Stiglmayr},
+  title = {Representation of the non-dominated set in biobjective
+                  discrete optimization},
+  journal = {Computers \& Operations Research},
+  volume = 63,
+  pages = {172--186},
+  year = 2015,
+  issn = {0305-0548},
+  doi = {10.1016/j.cor.2015.05.003}
+}
+
+ +
+@article{VelLam2000ec,
+  author = { David A. {Van Veldhuizen}  and  Gary B. Lamont },
+  title = {Multiobjective Evolutionary Algorithms: {Analyzing}
+                  the State-of-the-art},
+  journal = {Evolutionary Computation},
+  year = 2000,
+  volume = 8,
+  number = 2,
+  pages = {125--147},
+  doi = {10.1162/106365600568158}
+}
+
+ +
+@article{VerLew2020penalty,
+  author = {Amit Verma and Mark Lewis},
+  title = {Penalty and partitioning techniques to improve performance of
+                  {QUBO} solvers},
+  journal = {Discrete Optimization},
+  pages = 100594,
+  year = 2020,
+  issn = {1572-5286},
+  doi = {10.1016/j.disopt.2020.100594},
+  keywords = {Quadratic Unconstrained Binary Optimization, Nonlinear
+                  optimization, Pseudo-Boolean optimization, Equality
+                  constraint, Inequality constraint}
+}
+
+ +
+@article{VerLieJou2013ejor,
+  title = {On the Structure of Multiobjective Combinatorial
+                  Search Space: {MNK}-landscapes with Correlated
+                  Objectives},
+  author = { Verel, S{\'e}bastien  and  Arnaud Liefooghe  and  Laetitia Jourdan  and  Dhaenens, Clarisse },
+  journal = {European Journal of Operational Research},
+  volume = 227,
+  number = 2,
+  year = 2013,
+  pages = {331--342},
+  doi = {10.1016/j.ejor.2012.12.019},
+  alias = {verel2012}
+}
+
+ +
+@article{ViaFalPu2006pref,
+  title = {Preference-based Search using Example-Critiquing with
+                  Suggestions},
+  author = {Viappiani, Paolo and Faltings, Boi and Pu, Pearl},
+  journal = {Journal of Artificial Intelligence Research},
+  volume = 27,
+  pages = {465--503},
+  year = 2006
+}
+
+ +
+@article{ViaPuFal2008pref,
+  title = {Preference-based Search with Adaptive Recommendations},
+  author = {Viappiani, Paolo and Pu, Pearl and Faltings, Boi},
+  journal = {AI Communications},
+  volume = 21,
+  number = 2,
+  pages = {155--175},
+  year = 2008
+}
+
+ +
+@article{VidCraGenPri2013ejor,
+  author = { Thibaut Vidal  and  Teodor Gabriel Crainic  and  Michel Gendreau  and  Christian Prins },
+  title = {Heuristics for Multi-attribute Vehicle Routing Problems: A Survey and Synthesis},
+  journal = {European Journal of Operational Research},
+  year = 2013,
+  volume = 231,
+  number = 1,
+  pages = {1--21}
+}
+
+ +
+@article{VidCraGenPri2014ejor,
+  author = { Thibaut Vidal  and  Teodor Gabriel Crainic  and  Michel Gendreau  and  Christian Prins },
+  title = {A Unified Solution Framework for Multi-attribute Vehicle Routing Problems},
+  journal = {European Journal of Operational Research},
+  year = 2014,
+  volume = 234,
+  number = 3,
+  pages = {658--673}
+}
+
+ +
+@article{VilSanMer2021optimal,
+  title = {Optimal decision trees for the algorithm selection problem:
+                  integer programming based approaches},
+  author = {Vilas Boas, Matheus Guedes and Santos, Haroldo Gambini and Merschmann, Luiz Henrique de Campos and  Vanden Berghe, Greet  },
+  journal = {International Transactions in Operational Research},
+  volume = 28,
+  number = 5,
+  pages = {2759--2781},
+  year = 2021,
+  doi = {10.1111/itor.12724},
+  publisher = {Wiley Online Library}
+}
+
+ +
+@article{VouTsa99:ejor,
+  author = { Christos Voudouris  and  Edward P. K. Tsang },
+  title = {Guided Local Search and its Application to the
+                  Travelling Salesman Problem},
+  journal = {European Journal of Operational Research},
+  year = 1999,
+  volume = 113,
+  number = 2,
+  pages = {469--499}
+}
+
+ +
+@article{Wal1975ms,
+  title = {Comparative Evaluation of Some Interactive Approaches to Multicriterion Optimization},
+  author = { Wallenius, Jyrki },
+  journal = {Management Science},
+  volume = 21,
+  number = 12,
+  pages = {1387--1396},
+  year = 1975
+}
+
+ +
+@article{WalCro2000:siam-jsc,
+  author = {C. Walshaw and M. Cross},
+  title = {Mesh Partitioning: {A} Multilevel Balancing and
+                  Refinement Algorithm},
+  journal = {SIAM Journal on Scientific Computing},
+  year = 2000,
+  volume = 22,
+  number = 1,
+  pages = {63--80},
+  doi = {10.1137/S1064827598337373}
+}
+
+ +
+@article{WalEveFie2012visualizing,
+  title = {Visualizing mutually nondominating solution sets in
+                  many-objective optimization},
+  author = {Walker, David J. and  Everson, Richard M.  and  Jonathan E. Fieldsend },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 17,
+  number = 2,
+  pages = {165--184},
+  year = 2012
+}
+
+ +
+@article{WanChePro1997,
+  author = {Chengen Wang and Chengbin Chu and Jean-Marie Proth},
+  title = {Heuristic Approaches for {n/m/F/$\Sigma$Ci} Scheduling
+                  Problems},
+  journal = {European Journal of Operational Research},
+  volume = 96,
+  number = 3,
+  pages = {636--644},
+  year = 1997,
+  doi = {10.1016/0377-2217(95)00347-9}
+}
+
+ +
+@article{WanJiaYao2015twoarch2,
+  title = {{TwoArch2}: An improved two-archive algorithm for
+                  many-objective optimization},
+  author = {Wang, Handing and Jiao, Licheng and  Xin Yao },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2015,
+  number = 4,
+  pages = {524--541},
+  volume = 19
+}
+
+ +
+@article{WanJinSchOlh2021transfer,
+  author = {Xilu Wang and  Yaochu Jin  and Sebastian Schmitt and
+                  Markus Olhofer and  Allmendinger, Richard },
+  title = {Transfer learning based surrogate assisted evolutionary
+                  bi-objective optimization for objectives with different
+                  evaluation times},
+  journal = {Knowledge-Based Systems},
+  year = 2021,
+  volume = 227,
+  pages = 107190,
+  doi = {10.1016/j.knosys.2021.107190}
+}
+
+ +
+@article{WanLuGlo2013cor,
+  author = {Wang, Yang and  L{\"u}, Zhipeng  and  Fred Glover  and  Jin-Kao Hao },
+  title = {Probabilistic {GRASP}-Tabu Search algorithms for the {UBQP} problem},
+  journal = {Computers \& Operations Research},
+  volume = 40,
+  number = 12,
+  pages = {3100--3107},
+  year = 2013,
+  doi = {10.1016/j.cor.2011.12.006}
+}
+
+ +
+@article{WanLuGlov2013joh,
+  author = {Wang, Yang and  L{\"u}, Zhipeng  and  Fred Glover  and  Jin-Kao Hao },
+  title = {Backbone Guided Tabu Search for Solving the {UBQP} Problem},
+  year = 2013,
+  journal = {Journal of Heuristics},
+  volume = 19,
+  number = 4,
+  doi = {10.1007/s10732-011-9164-4},
+  pages = {679--695}
+}
+
+ +
+@article{WanPurFle2013tec,
+  title = {Preference-Inspired Coevolutionary Algorithms for
+                  Many-Objective Optimization},
+  author = {Wang, Rui and  Robin C. Purshouse  and  Peter J. Fleming },
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 17,
+  number = 4,
+  pages = {474--494},
+  year = 2013
+}
+
+ +
+@article{WanXioHeGao2020multi,
+  author = {Wang, Rui and Xiong, Jian and He, Min-fan and Gao, Liang and
+                  Wang, Ling},
+  title = {Multi-objective optimal design of hybrid renewable energy
+                  system under multiple scenarios},
+  journal = {Renewable Energy},
+  volume = 151,
+  pages = {226--237},
+  year = 2020,
+  publisher = {Elsevier},
+  doi = {10.1016/j.renene.2019.11.015}
+}
+
+ +
+@article{WangEtAl2012ejor,
+  author = {Wang, Yang and L{\"u}, Zhipeng and  Fred Glover  and  Jin-Kao Hao },
+  title = {Path relinking for unconstrained binary quadratic
+                  programming},
+  journal = {European Journal of Operational Research},
+  year = 2012,
+  volume = 223,
+  number = 3,
+  pages = {595--604},
+  publisher = {Elsevier},
+  doi = {10.1016/j.ejor.2012.07.012}
+}
+
+ +
+@article{WatBarWhiHow02,
+  author = { Jean-Paul Watson  and L. Barbulescu and  Darrell Whitley  and  Adele E. Howe },
+  title = {Contrasting Structured and Random Permutation
+                  Flow-Shop Scheduling Problems: Search Space Topology
+                  and Algorithm Performance},
+  journal = {INFORMS Journal on Computing},
+  year = 2002,
+  volume = 14,
+  number = 2,
+  pages = {98--123}
+}
+
+ +
+@article{WatBecHowWhi03,
+  author = { Jean-Paul Watson  and J. C. Beck and A. E. Howe and  Darrell Whitley },
+  title = {Problem Difficulty for Tabu Search in Job-Shop
+                  Scheduling},
+  journal = {Artificial Intelligence},
+  year = 2003,
+  volume = 143,
+  number = 2,
+  pages = {189--217}
+}
+
+ +
+@article{WatHowWhi2006,
+  title = {Deconstructing {Nowicki} and {Smutnicki}'s {i-TSAB} tabu search
+                  algorithm for the job-shop scheduling problem},
+  author = { Jean-Paul Watson  and Howe, Adele E and  Darrell Whitley },
+  journal = {Computers \& Operations Research},
+  volume = 33,
+  number = 9,
+  pages = {2623--2644},
+  year = 2006,
+  publisher = {Elsevier}
+}
+
+ +
+@article{WatKas2017ems,
+  title = {Incorporating deeply uncertain factors into the many
+                  objective search process},
+  author = {Watson, Abigail A. and  Kasprzyk, Joseph R. },
+  journal = {Environmental Modelling \& Software},
+  volume = 89,
+  pages = {159--171},
+  year = 2017,
+  keywords = {scenario-based}
+}
+
+ +
+@article{Wegman1990pc,
+  title = {Hyperdimensional data analysis using parallel coordinates},
+  author = {Wegman, Edward J.},
+  journal = {Journal of the American Statistical Association},
+  volume = 85,
+  number = 411,
+  pages = {664--675},
+  year = 1990
+}
+
+ +
+@article{Weinberger1991,
+  author = {Weinberger, Edward D.},
+  title = {Local properties of {Kauffman}'s {N-k} model: A tunably
+                  rugged energy landscape},
+  journal = {Physical Review A},
+  year = 1991,
+  volume = 44,
+  number = 10,
+  pages = 6399,
+  publisher = {APS}
+}
+
+ +
+@article{WeiKhoWan2016jbd,
+  title = {A survey of transfer learning},
+  author = {Weiss, Karl and Khoshgoftaar, Taghi M. and Wang, DingDing},
+  journal = {Journal of Big Data},
+  volume = 3,
+  number = 1,
+  pages = {1--40},
+  year = 2016,
+  publisher = {SpringerOpen}
+}
+
+ +
+@article{Welch1938significance,
+  title = {The significance of the difference between two means when the
+                  population variances are unequal},
+  author = {Welch, Bernard L.},
+  journal = {Biometrika},
+  volume = 29,
+  number = {3/4},
+  pages = {350--362},
+  year = 1938
+}
+
+ +
+@article{WesLop2018ecj,
+  title = {Latin Hypercube Designs with Branching and Nested Factors for
+                  Initialization of Automatic Algorithm Configuration},
+  author = { Simon Wessing  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  doi = {10.1162/evco_a_00241},
+  journal = {Evolutionary Computation},
+  year = 2018,
+  volume = 27,
+  number = 1,
+  pages = {129--145}
+}
+
+ +
+@article{Weyland2010,
+  title = {A Rigorous Analysis of the Harmony Search Algorithm: How the Research Community can be misled by a ``novel'' Methodology},
+  author = { Dennis Weyland },
+  journal = {International Journal of Applied Metaheuristic Computing},
+  volume = 12,
+  number = 2,
+  pages = {50--60},
+  year = 2010
+}
+
+ +
+@article{Weyland2015,
+  title = {A critical analysis of the harmony search algorithm: How not
+                  to solve {Sudoku}},
+  author = { Dennis Weyland },
+  journal = {Operations Research Perspectives},
+  volume = 2,
+  pages = {97--105},
+  year = 2015
+}
+
+ +
+@article{WhiArcCla2011:tec,
+  author = {D. R. White and A. Arcuri and J. A. Clark},
+  title = {Evolutionary Improvement of Programs},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2011,
+  volume = 15,
+  number = 4,
+  pages = {515--538}
+}
+
+ +
+@article{WhiBraBar2012tec,
+  author = {While, L. and Bradstreet, L. and Barone, L.},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {A Fast Way of Calculating Exact Hypervolumes},
+  year = 2012,
+  volume = 16,
+  number = 1,
+  pages = {86--95}
+}
+
+ +
+@article{WhiRanDzuMat96,
+  author = { Darrell Whitley  and Soraya Rana and John Dzubera and Keith E. Mathias},
+  title = {Evaluating Evolutionary Algorithms},
+  journal = {Artificial Intelligence},
+  year = 1996,
+  volume = 85,
+  pages = {245--296}
+}
+
+ +
+@article{Wil92,
+  author = {R. J. Williams},
+  title = {Simple Statistical Gradient-Following Algorithms for
+                  Connectionist Reinforcement Learning},
+  journal = {Machine Learning},
+  year = 1992,
+  pages = {229--256},
+  volume = 8,
+  number = 3
+}
+
+ +
+@article{Winkler1985order,
+  author = {P. Winkler},
+  title = {Random Orders},
+  journal = {Order},
+  year = 1985,
+  volume = 1,
+  pages = {317--331},
+  annote = {Showed that fraction of Pareto-optimal increases with number
+                  of objectives}
+}
+
+ +
+@article{Witt2012,
+  author = { Carsten Witt },
+  title = {Analysis of an Iterated Local Search Algorithm for Vertex Cover in Sparse Random Graphs},
+  journal = {Theoretical Computer Science},
+  year = 2012,
+  volume = 425,
+  pages = {117--125}
+}
+
+ +
+@article{WolMac97:ieee-tec,
+  author = {D. H. Wolpert and W. G. Macready},
+  title = {No Free Lunch Theorems for Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 1997,
+  volume = 1,
+  number = 1,
+  pages = {67--82},
+  doi = {10.1109/4235.585893}
+}
+
+ +
+@article{WooReeSim2013many,
+  title = {Many objective visual analytics: rethinking the design of
+                  complex engineered systems},
+  author = { Matthew J. Woodruff  and  Patrick M. Reed  and Simpson, Timothy W.},
+  journal = {Structural and Multidisciplinary Optimization},
+  volume = 48,
+  number = 1,
+  pages = {201--219},
+  year = 2013,
+  doi = {10.1007/s00158-013-0891-z},
+  publisher = {Springer}
+}
+
+ +
+@article{WooRitOpp2011ijmh,
+  author = { David L. Woodruff  and Ulrike Ritzinger and Johan
+                  Oppen},
+  title = {Research Note: The Point of Diminishing Returns in
+                  Heuristic Search},
+  journal = {International Journal of Metaheuristics},
+  year = 2011,
+  volume = 1,
+  number = 3,
+  pages = {222--231},
+  doi = {10.1504/IJMHeur.2011.041195},
+  keywords = {anytime}
+}
+
+ +
+@article{WooYim98,
+  author = { H. S. Woo  and  D. S. Yim },
+  title = {A Heuristic Algorithm for Mean Flowtime Objective in
+                  Flowshop Scheduling},
+  journal = {Computers \& Operations Research},
+  volume = 25,
+  number = 3,
+  pages = {175--182},
+  year = 1998
+}
+
+ +
+@article{WuSchCheLeNorMacKriCaoGaoMac2016gnmt,
+  title = {Google's neural machine translation system: Bridging the gap
+                  between human and machine translation},
+  author = {Wu, Yonghui and Schuster, Mike and Chen, Zhifeng and Le, Quoc
+                  V and Norouzi, Mohammad and Macherey, Wolfgang and Krikun,
+                  Maxim and Cao, Yuan and Gao, Qin and Macherey, Klaus and
+                  others},
+  journal = {Arxiv preprint arXiv:1609.08144 [cs.CL]},
+  url = {https://arxiv.org/abs/1609.08144},
+  year = 2016
+}
+
+ +
+@article{WuZhuWeDin2014mining,
+  title = {Data mining with big data},
+  author = {Wu, Xindong and Zhu, Xingquan and Wu, Gong-Qing and Ding,
+                  Wei},
+  journal = {IEEE Transactions on Knowledge and Data Engineering},
+  volume = 26,
+  number = 1,
+  pages = {97--107},
+  year = 2014,
+  publisher = {IEEE}
+}
+
+ +
+@article{XavXav2011mssc,
+  title = {Solving the minimum sum-of-squares clustering problem by
+                  hyperbolic smoothing and partition into boundary and
+                  gravitational regions},
+  journal = {Pattern Recognition},
+  volume = 44,
+  number = 1,
+  pages = {70--77},
+  year = 2011,
+  issn = {0031-3203},
+  doi = {10.1016/j.patcog.2010.07.004},
+  author = {Xavier, Adilson Elias and Xavier, Vinicius Layter },
+  keywords = {Cluster analysis, Min-sum-min problems, Nondifferentiable
+                  programming, Smoothing}
+}
+
+ +
+@article{XinCheChe2018review,
+  author = {Xin, B. and Chen, L. and Chen, J. and  Ishibuchi, Hisao  and Hirota, K. and Liu, B.},
+  journal = {{IEEE} Access},
+  title = {Interactive Multiobjective Optimization: A Review of the
+                  State-of-the-Art},
+  year = 2018,
+  volume = 6,
+  pages = {41256--41279},
+  doi = {10.1109/ACCESS.2018.2856832},
+  keywords = {Decision making, Evolutionary computation, Pareto
+                  optimization, Evolutionary multiobjective optimization,
+                  interactive multiobjective optimization, multiple criteria
+                  decision making, preference information, preference models},
+  abstract = {Interactive multiobjective optimization (IMO) aims at finding
+                  the most preferred solution of a decision maker with the
+                  guidance of his/her preferences which are provided
+                  progressively. During the process, the decision maker can
+                  adjust his/her preferences and explore only interested
+                  regions of the search space. In recent decades, IMO has
+                  gradually become a common interest of two distinct
+                  communities, namely, the multiple criteria decision making
+                  (MCDM) and the evolutionary multiobjective optimization
+                  (EMO). The IMO methods developed by the MCDM community
+                  usually use the mathematical programming methodology to
+                  search for a single preferred Pareto optimal solution, while
+                  those which are rooted in EMO often employ evolutionary
+                  algorithms to generate a representative set of solutions in
+                  the decision maker's preferred region. This paper aims to
+                  give a review of IMO research from both MCDM and EMO
+                  perspectives. Taking into account four classification
+                  criteria including the interaction pattern, preference
+                  information, preference model, and search engine (i.e.,
+                  optimization algorithm), a taxonomy is established to
+                  identify important IMO factors and differentiate various IMO
+                  methods. According to the taxonomy, state-of-the-art IMO
+                  methods are categorized and reviewed and the design ideas
+                  behind them are summarized. A collection of important issues,
+                  e.g., the burdens, cognitive biases and preference
+                  inconsistency of decision makers, and the performance
+                  measures and metrics for evaluating IMO methods, are
+                  highlighted and discussed. Several promising directions
+                  worthy of future research are also presented.}
+}
+
+ +
+@article{XuChiGlo98,
+  author = {Jiefeng Xu and Steve Y. Chiu and  Fred Glover },
+  title = {Fine-tuning a tabu search algorithm with statistical tests},
+  journal = {International Transactions in Operational Research},
+  volume = 5,
+  number = 3,
+  pages = {233--244},
+  year = 1998,
+  doi = {10.1111/j.1475-3995.1998.tb00117.x}
+}
+
+ +
+@article{XuHutHooLey2008jair,
+  author = { Lin Xu  and  Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {{SATzilla:} Portfolio-based Algorithm Selection for {SAT}},
+  journal = {Journal of Artificial Intelligence Research},
+  year = 2008,
+  volume = 32,
+  pages = {565--606},
+  month = jun,
+  epub = {http://www.cs.ubc.ca/~hutter/papers/SATzilla.pdf},
+  doi = {10.1613/jair.2490}
+}
+
+ +
+@article{XuLuChe2013,
+  author = {Hongyun Xu and Zhipeng L{\"u} and T. C. E. Cheng},
+  title = {Iterated Local Search for Single-machine Scheduling with Sequence-dependent Setup Times to Minimize Total Weighted Tardiness},
+  journal = {Journal of Scheduling},
+  year = 2014,
+  volume = 17,
+  number = 3,
+  pages = {271--287}
+}
+
+ +
+@article{XuYan2003ids,
+  author = { Dong-Ling Xu  and  Yang, Jian-Bo },
+  title = {Intelligent Decision System for Self-Assessment},
+  journal = {Journal of Multi-Criteria Decision Analysis},
+  year = 2003,
+  volume = 12,
+  number = 1,
+  pages = {43--60},
+  doi = {10.1002/mcda.343},
+  abstract = {Many small and medium enterprises (SMEs) in the UK use the
+                  beta (Business Excellence Through Action) approach to the
+                  EFQM Excellence Model to conduct business excellence
+                  self-assessment, which is in essence a multiple criteria
+                  decision analysis (MCDA) problem. This paper introduces a
+                  decision support software package called Intelligent Decision
+                  System (IDS) to implement the beta approach. It is
+                  demonstrated in the paper that the IDS-beta package can
+                  provide not only average scores but also the following
+                  numerical results and graphical displays on: Distributed
+                  assessment results to demonstrate the diversity of company
+                  performances The performance range to cater for incomplete
+                  assessment information Comparisons between current
+                  performances and past performances, among different companies
+                  among different action plans. Strengths and weaknesses The
+                  IDS-beta package also provides a structured knowledge base to
+                  help assessors to make judgements more objectively. The
+                  knowledge base contains guidelines provided by the developers
+                  of the beta approach, best practices gathered from research
+                  on award winning organizations, evidence collected from
+                  companies being assessed and comments provided by assessors
+                  to record the reasons why a specific criterion is assessed to
+                  a certain grade for a company. Four small UK companies, the
+                  industry partners of the research project, have carried out
+                  the preliminary self-assessment using the package. The
+                  results and experience of the application are discussed at
+                  the end of the paper.},
+  keywords = {decision support system, business excellence, MCDA, quality
+                  award, self-assessment, the evidential reasoning approach}
+}
+
+ +
+@article{YagKisIba2006,
+  author = { Mutsunori Yagiura  and M. Kishida and  Toshihide Ibaraki },
+  title = {A 3-Flip Neighborhood Local Search for the Set Covering Problem},
+  journal = {European Journal of Operational Research},
+  year = 2006,
+  volume = 172,
+  number = 2,
+  pages = {472--499}
+}
+
+ +
+@article{Yam2018parking,
+  author = {Yuki Yamada},
+  title = {How to Crack Pre-registration: Toward Transparent and Open
+                  Science},
+  doi = {10.3389/fpsyg.2018.01831},
+  year = 2018,
+  month = sep,
+  publisher = {Frontiers Media {SA}},
+  volume = 9,
+  journal = {Frontiers in Psychology},
+  keywords = {HARKing; PARKing}
+}
+
+ +
+@article{YanEmmDeu2019bayesian,
+  author = {Kaifeng Yang and  Emmerich, Michael T. M.  and   Andr{\'{e}} H. Deutz  and  Thomas B{\"a}ck },
+  journal = {Swarm and Evolutionary Computation},
+  title = {Multi-Objective {Bayesian} Global Optimization using Expected
+                  Hypervolume Improvement Gradient},
+  year = 2019,
+  month = feb,
+  pages = {945--956},
+  volume = 44,
+  doi = {10.1016/j.swevo.2018.10.007},
+  keywords = {Bayesian Optimisation with preferences},
+  publisher = {Elsevier {BV}}
+}
+
+ +
+@article{YanKrePin00:js,
+  author = {Y. Yang and S. Kreipl and M. L. Pinedo},
+  title = {Heuristics for Minimizing Total Weighted Tardiness in Flexible Flow Shops},
+  journal = {Journal of Scheduling},
+  year = 2000,
+  volume = 3,
+  number = 2,
+  pages = {89--108}
+}
+
+ +
+@article{YanLiLiuZhe2013,
+  author = {S. Yang and  Li, Miqing  and X. Liu and J. Zheng},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  title = {A Grid-Based Evolutionary Algorithm for Many-Objective
+                  Optimization},
+  year = 2013,
+  volume = 17,
+  number = 5,
+  pages = {721--736},
+  doi = {10.1109/TEVC.2012.2227145},
+  annote = {epsilon-grid}
+}
+
+ +
+@article{YeDoeWang2022ac,
+  author = {Furong Ye and  Carola Doerr  and Hao Wang and  Thomas B{\"a}ck },
+  title = {Automated Configuration of Genetic Algorithms by Tuning for
+                  Anytime Performance},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2022,
+  volume = 26,
+  number = 6,
+  pages = {1526--1538},
+  doi = {10.1109/TEVC.2022.3159087}
+}
+
+ +
+@article{YuLin2015,
+  author = {Vincent F. Yu and Shih-Wei Lin},
+  title = {Iterated Greedy Heuristic for the Time-dependent
+                  Prize-collecting Arc Routing Problem},
+  journal = {Computers and Industrial Engineering},
+  year = 2015,
+  volume = 90,
+  pages = {54--66}
+}
+
+ +
+@article{YuPowSte94,
+  author = { G. Yu  and  R. S. Powell  and  M. J. H. Sterling },
+  title = {Optimized Pump Scheduling in Water Distribution
+                  Systems},
+  journal = {Journal of Optimization Theory and Applications},
+  year = 1994,
+  volume = 83,
+  number = 3,
+  pages = {463--488}
+}
+
+ +
+@article{YuaMonStuBir12:si,
+  author = { Zhi Yuan  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {Continuous Optimization Algorithms for Tuning Real and
+Integer Algorithm Parameters of Swarm Intelligence Algorithms},
+  journal = {Swarm Intelligence},
+  year = 2012,
+  volume = 6,
+  number = 1,
+  pages = {49--75}
+}
+
+ +
+@article{ZenYan2009cor,
+  author = {Zeng, Q. and Yang, Z.},
+  title = {Integrating Simulation and Optimization to Schedule
+                  Loading Operations in Container Terminals},
+  journal = {Computers \& Operations Research},
+  year = 2009,
+  volume = 36,
+  number = 6,
+  pages = {1935--1944},
+  doi = {10.1016/j.cor.2008.06.010}
+}
+
+ +
+@article{ZhaGeoAna2016,
+  author = {Tiantian Zhang and Michael Georgiopoulos and Georgios C. Anagnostopoulos},
+  title = {Multi-Objective Model Selection via Racing},
+  journal = {IEEE Transactions on Cybernetics},
+  year = 2016,
+  volume = 46,
+  number = 8,
+  pages = {1863--1876}
+}
+
+ +
+@article{ZhaLi07:moead,
+  author = { Zhang, Qingfu  and Hui Li},
+  title = {{MOEA/D}: A Multiobjective Evolutionary Algorithm Based on
+                  Decomposition},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2007,
+  volume = 11,
+  number = 6,
+  pages = {712--731},
+  doi = {10.1109/TEVC.2007.892759},
+  annote = {Introduces penalty-based boundary intersection (PBI)
+                  function}
+}
+
+ +
+@article{ZhaSan2009jade,
+  title = {{JADE}: Adaptive differential evolution with optional
+                  external archive},
+  author = {Zhang, Jingqiao and Sanderson, Arthur C.},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 13,
+  number = 5,
+  pages = {945--958},
+  year = 2009,
+  doi = {10.1109/TEVC.2009.2014613}
+}
+
+ +
+@article{ZhaSud2004dts,
+  author = {Zhao, H. and Sudha Ram},
+  journal = {IEEE Transactions on Knowledge and Data Engineering},
+  title = {Constrained cascade generalization of decision trees},
+  year = 2004,
+  volume = 16,
+  number = 6,
+  pages = {727--739},
+  doi = {10.1109/TKDE.2004.3}
+}
+
+ +
+@article{ZheCha2012berth,
+  title = {A bi-objective model for robust berth allocation scheduling},
+  author = {Zhen, Lu and Chang, Dao-Fang},
+  journal = {Computers and Industrial Engineering},
+  volume = 63,
+  number = 1,
+  pages = {262--273},
+  year = 2012
+}
+
+ +
+@article{ZhoZhaJin2009igdx,
+  author = {Zhou, A. and  Zhang, Qingfu  and  Yaochu Jin },
+  title = {Approximating the set of {Pareto}-optimal solutions in both
+                  the decision and objective spaces by an estimation of
+                  distribution algorithm},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2009,
+  volume = 13,
+  number = 5,
+  pages = {1167--1189},
+  doi = {10.1109/TEVC.2009.2021467},
+  keywords = {multi-modal, IGDX}
+}
+
+ +
+@article{Zilberstein96:anytime,
+  author = { Shlomo Zilberstein },
+  title = {Using Anytime Algorithms in Intelligent Systems},
+  journal = {{AI} Magazine},
+  year = 1996,
+  volume = 17,
+  number = 3,
+  pages = {73--83},
+  epub = {http://www.aaai.org/ojs/index.php/aimagazine/article/viewArticle/1232},
+  keywords = {performance profiles},
+  doi = {10.1609/aimag.v17i3.1232},
+  abstract = {Anytime algorithms give intelligent systems the capability to trade deliberation time for quality of results. This capability is essential for successful operation in domains such as signal interpretation, real-time diagnosis and repair, and mobile robot control. What characterizes these domains is that it is not feasible (computationally) or desirable (economically) to compute the optimal answer. This article surveys the main control problems that arise when a system is composed of several anytime algorithms. These problems relate to optimal management of uncertainty and precision. After a brief introduction to anytime computation, I outline a wide range of existing solutions to the metalevel control problem and describe current work that is aimed at increasing the applicability of anytime computation.}
+}
+
+ +
+@article{ZioWal1983interactive,
+  title = {An interactive multiple objective linear programming method
+                  for a class of underlying nonlinear utility functions},
+  author = {Zionts, Stanley and  Wallenius, Jyrki },
+  journal = {Management Science},
+  volume = 29,
+  number = 5,
+  pages = {519--529},
+  year = 1983,
+  publisher = {{INFORMS}}
+}
+
+ +
+@article{ZitThi99:spea,
+  author = { Eckart Zitzler  and  Lothar Thiele },
+  title = {Multiobjective Evolutionary Algorithms: {A} Comparative Case
+                  Study and the Strength {Pareto} Approach},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  volume = 3,
+  number = 4,
+  pages = {257--271},
+  year = 1999,
+  doi = {10.1109/4235.797969},
+  annote = {Proposed SPEA,
+                  \url{http://www.tik.ee.ethz.ch/sop/publicationListFiles/zt1999a.pdf}}
+}
+
+ +
+@article{ZitThiBad2010tec,
+  author = { Eckart Zitzler  and  Lothar Thiele  and  Johannes Bader },
+  title = {On Set-Based Multiobjective Optimization},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2010,
+  volume = 14,
+  number = 1,
+  pages = {58--79},
+  doi = {10.1109/TEVC.2009.2016569},
+  keywords = {Performance assessment; Preference articulation; refinement;
+                  Set Partitioning; Set-preference},
+  annote = {Proposed SPAM and explores combination of quality indicators}
+}
+
+ +
+@article{ZitThiDeb2000ec,
+  author = { Eckart Zitzler  and  Lothar Thiele  and  Kalyanmoy Deb },
+  title = {Comparison of Multiobjective Evolutionary
+                  Algorithms: Empirical Results},
+  journal = {Evolutionary Computation},
+  year = 2000,
+  volume = 8,
+  number = 2,
+  pages = {173--195},
+  doi = {10.1162/106365600568202},
+  keywords = {ZDT benchmark}
+}
+
+ +
+@article{ZitThiLauFon2003:tec,
+  author = { Eckart Zitzler  and  Lothar Thiele  and  Marco Laumanns  and  Carlos M. Fonseca  and  Viviane {Grunert da Fonseca} },
+  title = {Performance Assessment of Multiobjective Optimizers: an
+                  Analysis and Review},
+  journal = {IEEE Transactions on Evolutionary Computation},
+  year = 2003,
+  volume = 7,
+  number = 2,
+  amonth = apr,
+  pages = {117--132},
+  alias = {perfassess},
+  doi = {10.1109/TEVC.2003.810758},
+  annote = {Proposed the combination of quality indicators; proposed epsilon-indicator}
+}
+
+ +
+@article{ZloBirMeuDor2004:aor,
+  author = {M. Zlochin and  Mauro Birattari  and  N. Meuleau  and  Marco Dorigo },
+  journal = {Annals of Operations Research},
+  number = {1--4},
+  pages = {373--395},
+  title = {Model-Based Search for Combinatorial Optimization: A
+                  Critical Survey},
+  volume = 131,
+  year = 2004
+}
+
+ +
+@article{mlrMBO,
+  title = {{mlrMBO}: A Modular Framework for Model-Based Optimization of
+                  Expensive Black-Box Functions},
+  author = { Bernd Bischl  and Jakob Richter and  Jakob Bossek  and Daniel Horn and Janek Thomas and Michel Lang},
+  year = 2017,
+  journal = {Arxiv preprint arXiv:1703.03373 [stat.ML]},
+  url = {http://arxiv.org/abs/1703.03373}
+}
+
+ +
+@article{msc:special-issue,
+  author = { Oscar Cord{\'o}n  and  Francisco Herrera  and  Thomas St{\"u}tzle },
+  title = {Special Issue on Ant Colony Optimization: Models and
+                  Applications},
+  journal = {Mathware \& Soft Computing},
+  year = 2002,
+  volume = 9,
+  number = 3,
+  pages = {137--268}
+}
+
+ +
+@article{powel03:demand,
+  author = { G. McCormick  and  R. S. Powell },
+  title = {Optimal Pump Scheduling in Water Supply Systems with Maximum
+                  Demand Charges},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  volume = 129,
+  number = 5,
+  pages = {372--379},
+  date = {2003-09/2003-10},
+  year = 2003,
+  month = sep # { / } # oct
+}
+
+ +
+@article{QuaGreLiuHu2007searching,
+  title = {Searching for multiobjective preventive maintenance
+                  schedules: Combining preferences with evolutionary
+                  algorithms},
+  journal = {European Journal of Operational Research},
+  volume = 177,
+  number = 3,
+  pages = {1969--1984},
+  year = 2007,
+  doi = {10.1016/j.ejor.2005.12.015},
+  author = {Gang Quan and Garrison W. Greenwood and Donglin Liu and
+                  Sharon Hu},
+  keywords = {Evolutionary computations, Scheduling, Utility theory,
+                  Preventive maintenance, Multi-objective optimization,
+                  ranking-based, interactive},
+  abstract = {Heavy industry maintenance facilities at aircraft service
+                  centers or railroad yards must contend with scheduling
+                  preventive maintenance tasks to ensure critical equipment
+                  remains available. The workforce that performs these tasks
+                  are often high-paid, which means the task scheduling should
+                  minimize worker idle time. Idle time can always be minimized
+                  by reducing the workforce. However, all preventive
+                  maintenance tasks should be completed as quickly as possible
+                  to make equipment available. This means the completion time
+                  should be also minimized. Unfortunately, a small workforce
+                  cannot complete many maintenance tasks per hour. Hence, there
+                  is a tradeoff: should the workforce be small to reduce idle
+                  time or should it be large so more maintenance can be
+                  performed each hour? A cost effective schedule should strike
+                  some balance between a minimum schedule and a minimum size
+                  workforce. This paper uses evolutionary algorithms to solve
+                  this multiobjective problem. However, rather than conducting
+                  a conventional dominance-based Pareto search, we introduce a
+                  form of utility theory to find Pareto optimal solutions. The
+                  advantage of this method is the user can target specific
+                  subsets of the Pareto front by merely ranking a small set of
+                  initial solutions. A large example problem is used to
+                  demonstrate our method.}
+}
+
+ +
+@article{ranger2015,
+  author = {Marvin N. Wright and Andreas Ziegler},
+  title = {{\rpackage{ranger}}: A Fast Implementation of Random Forests for High
+                  Dimensional Data in {\proglang{C++}} and {\proglang{R}}},
+  journal = {Arxiv preprint arXiv:1508.04409 [stat.ML]},
+  url = {https://arxiv.org/abs/1508.04409},
+  year = 2015
+}
+
+ +
+@article{ranger2017:jss,
+  author = {Marvin N. Wright and Andreas Ziegler},
+  title = {{\rpackage{ranger}}: A Fast Implementation of Random Forests for High
+                  Dimensional Data in {\proglang{C++}} and {\proglang{R}}},
+  journal = {Journal of Statistical Software},
+  year = 2017,
+  volume = 77,
+  number = 1,
+  pages = {1--17},
+  doi = {10.18637/jss.v077.i01}
+}
+
+ +
+@article{scikit-learn2011,
+  title = {Scikit-learn: Machine learning in {\proglang{Python}}},
+  author = {Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel,
+                  V.  and Thirion, B. and Grisel, O. and Blondel, M. and
+                  Prettenhofer, P.  and Weiss, R. and Dubourg, V. and
+                  Vanderplas, J. and Passos, A. and Cournapeau, D. and Brucher,
+                  M. and Perrot, M. and Duchesnay, E.},
+  journal = {Journal of Machine Learning Research},
+  volume = 12,
+  pages = {2825--2830},
+  year = 2011
+}
+
+ +
+@article{vanZyl04,
+  author = { Jakobus E. van Zyl  and  Dragan A. Savic  and  Godfrey A. Walters },
+  title = {Operational Optimization of Water Distribution
+                  Systems using a Hybrid Genetic Algorithm},
+  journal = {Journal of Water Resources Planning and Management, {ASCE}},
+  year = 2004,
+  volume = 130,
+  number = 2,
+  pages = {160--170},
+  month = mar
+}
+
+ +
+@misc{AAAI2021checklist,
+  author = {{AAAI}},
+  title = {35th AAAI Conference on Artificial Intelligence:
+                  Reproducibility Checklist},
+  howpublished = {\url{https://aaai.org/Conferences/AAAI-21/reproducibility-checklist/}},
+  year = 2021,
+  note = {Last accessed: June 6th, 2021}
+}
+
+ +
+@misc{ACM2020badging_v1_1,
+  author = {{ACM}},
+  title = {Artifact Review and Badging Version 1.1},
+  howpublished = {\url{https://www.acm.org/publications/policies/artifact-review-and-badging-current}},
+  year = 2020,
+  month = aug
+}
+
+ +
+@incollection{AarKorMic2005,
+  year = 2005,
+  address = {Boston, MA},
+  publisher = {Springer},
+  doi = {10.1007/0-387-28356-0},
+  editor = { Edmund K. Burke  and  Graham Kendall },
+  booktitle = {Search Methodologies},
+  title = {Simulated Annealing},
+  author = { Emile H. L. Aarts  and  Jan H. M. Korst  and  Wil Michiels },
+  pages = {187--210}
+}
+
+ +
+@inproceedings{Abb2002selfpde,
+  year = 2002,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2002 Congress on Evolutionary Computation (CEC'02)},
+  key = {IEEE CEC},
+  title = {The self-adaptive {Pareto} differential evolution algorithm},
+  author = { Abbass, Hussein A. },
+  pages = {831--836}
+}
+
+ +
+@inproceedings{LimPoz2017automopso,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = {de Lima, Ricardo Henrique Remes and Pozo, Aurora Trinidad
+                  Ramirez},
+  title = {A study on auto-configuration of Multi-Objective Particle
+                  Swarm Optimization Algorithm},
+  pages = {718--725},
+  doi = {10.1109/CEC.2017.7969381}
+}
+
+ +
+@inproceedings{AbbSarNew2001pde,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2001,
+  booktitle = {Proceedings of  the 2001 Congress on Evolutionary Computation (CEC'01)},
+  key = {IEEE CEC},
+  title = {{PDE}: a {Pareto}-frontier differential evolution approach
+                  for multi-objective optimization problems},
+  author = { Abbass, Hussein A.  and Sarker, Ruhul and Newton, Charles},
+  pages = {971--978}
+}
+
+ +
+@inproceedings{AbdKriCha1997,
+  year = 1997,
+  booktitle = {Proceedings of MIC 1997, the 2nd Metaheuristics International
+                  Conference},
+  editor = { Mauricio G. C. Resende  and Pinho de Souza, Jorge},
+  title = {A hybrid heuristic for multiobjective knapsack problems},
+  author = {Ben Abdelaziz, F. and Krichen, S. and Chaouachi, J.},
+  pages = {205--212},
+  doi = {10.1007/978-1-4615-5775-3_14}
+}
+
+ +
+@incollection{Aca2004memaco,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = {Acan, A.},
+  title = {An external memory implementation in ant colony optimization},
+  pages = {73--84},
+  keywords = {memory-based ACO}
+}
+
+ +
+@incollection{Aca2005evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2005,
+  editor = { G{\"u}nther R. Raidl  and Gottlieb, Jens},
+  volume = 3448,
+  booktitle = {Proceedings of EvoCOP 2005 -- 5th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = {Acan, A.},
+  title = {An external partial permutations memory for ant colony
+                  optimization},
+  pages = {1--11},
+  keywords = {memory-based ACO}
+}
+
+ +
+@incollection{AguZapLieVer2016many,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  year = 2016,
+  volume = 9554,
+  fulleditor = {St\'ephane Bonnevay and Pierrick Legrand and  Nicolas Monmarch{\'e}  and Evelyne Lutton and  Marc Schoenauer },
+  editor = {St\'ephane Bonnevay and others},
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Artificial Evolution: 12th International Conference, Evolution Artificielle, EA, 2015},
+  title = {Approaches for Many-Objective Optimization: Analysis and
+                  Comparison on {MNK}-Landscapes},
+  author = { Aguirre, Hern\'{a}n E.  and  Zapotecas, Sa{\'{u}}l  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien  and  Tanaka, Kiyoshi },
+  pages = {14--28},
+  doi = {10.1007/978-3-319-31471-6_2}
+}
+
+ +
+@inproceedings{CheHuhHul2009dt,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  editor = {Andrea Pohoreckyj Danyluk and L{\'{e}}on Bottou and Michael
+                  L. Littman},
+  year = 2009,
+  booktitle = {Proceedings of  the 26th International Conference on Machine Learning, {ICML} 2009},
+  author = {Cheng, Weiwei and H\"{u}hn, Jens and  Eyke H{\"u}llermeier },
+  title = {Decision Tree and Instance-Based Learning for Label Ranking},
+  doi = {10.1145/1553374.1553395},
+  pages = {161--168},
+  numpages = 8
+}
+
+ +
+@incollection{AguTan2009:space,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  title = {Many-Objective Optimization by Space Partitioning and
+                  Adaptive $\epsilon$-Ranking on {MNK}-Landscapes},
+  author = { Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  pages = {407--422}
+}
+
+ +
+@incollection{Aguirre2013,
+  year = 2013,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2013},
+  editor = { Christian Blum  and  Alba, Enrique },
+  author = { Aguirre, Hern\'{a}n E. },
+  title = {Advances on Many-objective Evolutionary Optimization},
+  pages = {641--666},
+  keywords = {many-objective evolutionary optimization}
+}
+
+ +
+@book{AhoHopUll83:data-structures,
+  author = { A. Aho  and  J. Hopcroft  and  J. Ullman },
+  title = {Data structures and algorithms},
+  publisher = {Addison-Wesley, Reading, MA},
+  year = 1983
+}
+
+ +
+@book{AhujMagOrl1993netflows,
+  author = { R. K. Ahuja   and T. Magnanti and   J. B. Orlin },
+  title = {Network Flows: Theory, Algorithms and Applications},
+  publisher = {Prentice-Hall},
+  year = 1993
+}
+
+ +
+@incollection{AikBurLi2006,
+  year = 2006,
+  volume = 4193,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Runarsson, Thomas Philip and   Hans-Georg Beyer  and  Edmund K. Burke  and  Juan-Juli{\'a}n Merelo  and  Darrell Whitley  and  Xin Yao },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IX}},
+  author = {Uwe Aickelin and  Edmund K. Burke  and Jingpeng Li},
+  title = {Improved Squeaky Wheel Optimisation for Driver Scheduling},
+  pages = {182--191}
+}
+
+ +
+@incollection{AisRoy2010:isorms,
+  year = 2010,
+  volume = 142,
+  publisher = {Springer, US},
+  editor = { Matthias Ehrgott  and  Jos{\'e} Rui Figueira  and  Salvatore Greco },
+  series = {International Series in Operations Research \& Management Science},
+  booktitle = {Trends in Multiple Criteria Decision Analysis},
+  author = { Hassene Aissi  and  Bernard Roy },
+  title = {Robustness in Multi-criteria Decision Aiding},
+  chapter = 4,
+  pages = {87--121}
+}
+
+ +
+@incollection{AkiSanYan2019optuna,
+  key = {SIGKDD},
+  month = jul,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = {Teredesai and others},
+  booktitle = {25th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  doi = {10.1145/3292500.3330701},
+  author = {Takuya Akiba and Shotaro Sano and Toshihiko Yanase and Takeru
+                  Ohta and Masanori Koyama},
+  title = {Optuna: A Next-generation Hyperparameter Optimization Framework},
+  pages = {2623--2631}
+}
+
+ +
+@techreport{AktAtaGur2007conic,
+  author = {S. M. Akt{\"u}rk and  Alper Atamt{\"u}rk  and S. G{\"u}rel},
+  title = {A Strong Conic Quadratic Reformulation for Machine-Job
+Assignment with Controllable Processing Times},
+  institution = {University of California-Berkeley},
+  year = 2007,
+  type = {Research Report},
+  number = {BCOL.07.01}
+}
+
+ +
+@incollection{AlaSolGhe07,
+  author = {I. Alaya and  Christine Solnon  and  Khaled Gh{\'e}dira},
+  title = {Ant Colony Optimization for Multi-Objective
+                  Optimization Problems},
+  booktitle = {19th IEEE International Conference on Tools with
+                  Artificial Intelligence (ICTAI 2007)},
+  year = 2007,
+  volume = 1,
+  publisher = {IEEE Computer Society Press},
+  address = {Los Alamitos, CA},
+  pages = {450--457}
+}
+
+ +
+@inproceedings{AlaSolGhe2004:bioma,
+  url = {https://books.google.be/books?id=0ZLsAAAACAAJ},
+  editor = {Bogdan Filipi{\v c} and  Jurij {\v S}ilc },
+  year = 2004,
+  booktitle = {International Conference on Bioinspired Optimization Methods
+                  and their Applications (BIOMA 2004)},
+  author = {I. Alaya and  Christine Solnon  and  Khaled Gh{\'e}dira},
+  title = {Ant algorithm for the multi-dimensional knapsack
+                  problem},
+  pages = {63--72}
+}
+
+ +
+@incollection{AlbChi2007gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = { Alba, Enrique  and  Chicano, Francisco },
+  title = {{ACOhg}: dealing with huge graphs},
+  pages = {10--17},
+  doi = {10.1145/1276958.1276961}
+}
+
+ +
+@incollection{AliSimHar2019,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  doi = {10.1145/3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = {Alissa, Mohamad and Sim, Kevin and  Emma Hart },
+  title = {Algorithm Selection Using Deep Learning without Feature Extraction},
+  pages = {198--206}
+}
+
+ +
+@incollection{AllBurHyd2009reusable,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = {Allen, Sam and  Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall },
+  title = {Evolving reusable 3d packing heuristics with genetic
+                  programming},
+  pages = {931--938},
+  doi = {10.1145/1569901.1570029},
+  keywords = {hyper-heuristic}
+}
+
+ +
+@incollection{AllKno2010variables,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  author = { Allmendinger, Richard  and  Joshua D. Knowles },
+  title = {Evolutionary Optimization on Problems Subject to Changes of
+                  Variables},
+  editor = {Schaefer, Robert and  Carlos Cotta  and Ko{\l}odziej, Joanna and  G{\"u}nther Rudolph },
+  pages = {151--160},
+  abstract = {Motivated by an experimental problem involving the
+                  identification of effective drug combinations drawn from a
+                  non-static drug library, this paper examines evolutionary
+                  algorithm strategies for dealing with changes of
+                  variables. We consider four standard techniques from dynamic
+                  optimization, and propose one new technique. The results show
+                  that only little additional diversity needs to be introduced
+                  into the population when changing a small number of
+                  variables, while changing many variables or optimizing a
+                  rugged landscape requires often a restart of the optimization
+                  process}
+}
+
+ +
+@inproceedings{AllKno2011ecta,
+  author = { Allmendinger, Richard  and  Joshua D. Knowles },
+  title = {Evolutionary Search in Lethal Environments},
+  booktitle = {International Conference on Evolutionary Computation Theory
+                  and Applications},
+  year = 2011,
+  pages = {63--72},
+  publisher = {SciTePress},
+  doi = {10.5220/0003673000630072},
+  epub = {https://www.scitepress.org/papers/2011/36730/36730.pdf}
+}
+
+ +
+@incollection{AllKno2011policy,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { Allmendinger, Richard  and  Joshua D. Knowles },
+  title = {Policy Learning in Resource-Constrained Optimization},
+  pages = {1971--1979},
+  doi = {10.1145/2001576.2001841},
+  abstract = {We consider an optimization scenario in which resources are
+                  required in the evaluation process of candidate
+                  solutions. The challenge we are focussing on is that certain
+                  resources have to be committed to for some period of time
+                  whenever they are used by an optimizer. This has the effect
+                  that certain solutions may be temporarily non-evaluable
+                  during the optimization. Previous analysis revealed that
+                  evolutionary algorithms (EAs) can be effective against this
+                  resourcing issue when augmented with static strategies for
+                  dealing with non-evaluable solutions, such as repairing,
+                  waiting, or penalty methods. Moreover, it is possible to
+                  select a suitable strategy for resource-constrained problems
+                  offline if the resourcing issue is known in advance. In this
+                  paper we demonstrate that an EA that uses a reinforcement
+                  learning (RL) agent, here Sarsa({$\lambda$}), to learn
+                  offline when to switch between static strategies, can be more
+                  effective than any of the static strategies themselves. We
+                  also show that learning the same task as the RL agent but
+                  online using an adaptive strategy selection method, here
+                  D-MAB, is not as effective; nevertheless, online learning is
+                  an alternative to static strategies.},
+  isbn = {978-1-4503-0557-0},
+  langid = {english}
+}
+
+ +
+@inproceedings{AllMouLiu2019human,
+  year = 2019,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the Thirty-Second International Florida Artificial
+                  Intelligence Research Society Conference},
+  editor = {Roman Bart{\'{a}}k and Keith W. Brawner},
+  author = {Joseph Allen and Ahmed Moussa and Xudong Liu},
+  title = {Human-in-the-Loop Learning of Qualitative Preference Models},
+  pages = {108--111},
+  doi = {10.48550/arXiv.1909.09064}
+}
+
+ +
+@phdthesis{Allmendinger2012phd,
+  author = { Allmendinger, Richard },
+  title = {Tuning Evolutionary Search for Closed-Loop Optimization},
+  school = {The University of Manchester, UK},
+  year = 2012,
+  month = jan
+}
+
+ +
+@inproceedings{AlsTsa2009,
+  address = {Hamburg, Germany},
+  publisher = {University of Hamburg},
+  editor = {M. Caserta and  Stefan Vo{\ss} },
+  year = 2010,
+  booktitle = {Proceedings of MIC 2009, the 8th Metaheuristics International Conference},
+  title = {Guided {Pareto} local search and its application to
+                  the 0/1 multi-objective knapsack problems},
+  author = {Alsheddy, A. and Tsang, E.}
+}
+
+ +
+@inproceedings{AmaAliThr2019nips,
+  year = 2019,
+  editor = {Hanna M. Wallach and Hugo Larochelle and Alina Beygelzimer
+                  and Florence d'Alch{\'{e}}{-}Buc and Emily B. Fox and Roman
+                  Garnett},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS 32)},
+  title = {Linear Stochastic Bandits Under Safety Constraints},
+  author = {Amani, Sanae and Alizadeh, Mahnoosh and Thrampoulidis,
+                  Christos},
+  pages = {9256--9266},
+  epub = {http://papers.nips.cc/paper/9124-linear-stochastic-bandits-under-safety-constraints.pdf}
+}
+
+ +
+@incollection{AndVidIve1993,
+  publisher = {Springer},
+  year = 1993,
+  editor = { Vidal, Ren{\'e} Victor Valqui  },
+  booktitle = {Applied Simulated Annealing},
+  title = {Design of a Teleprocessing Communication Network Using Simulated Annealing},
+  author = { Klaus Andersen  and  Vidal, Ren{\'e} Victor Valqui   and  Villy B{\ae}k Iversen },
+  pages = {201--215}
+}
+
+ +
+@incollection{Andersen99,
+  author = { J. H. Andersen  and  R. S. Powell },
+  title = {The Use of Continuous Decision Variables in an
+                  Optimising Fixed Speed Pump Scheduling Algorithm},
+  booktitle = {Computing and Control for the Water Industry},
+  pages = {119--128},
+  publisher = { Research Studies Press Ltd. },
+  year = 1999,
+  editor = { R. S. Powell  and  K. S. Hindi }
+}
+
+ +
+@incollection{AngBocPaoVec08,
+  year = 2008,
+  volume = 5361,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {X. Li and others},
+  fulleditor = {X. Li and M. Kirley and M. Zhang and D. G. Green and
+                  V. Ciesielski and  Abbass, Hussein A.  and Z. Michalewicz and
+                  T. Hendtlass and  Kalyanmoy Deb  and  Tan, Kay Chen  and  J{\"u}rgen Branke  and Y. Shi},
+  booktitle = {Simulated Evolution and Learning, 7th International
+                  Conference, SEAL 2008},
+  title = {Performance Evaluation of an Adaptive Ant Colony
+                  Optimization Applied to Single Machine Scheduling},
+  author = {D. Anghinolfi and A. Boccalatte and M. Paolucci and
+                  C. Vecchiola},
+  pages = {411--420}
+}
+
+ +
+@incollection{Angus2007,
+  editor = {Marcus Randall and  Abbass, Hussein A.  and Janet Wiles},
+  volume = 4828,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2007,
+  booktitle = {Progress in Artificial Life (ACAL)},
+  author = { Daniel Angus },
+  title = {Population-Based Ant Colony Optimisation for
+                  Multi-objective Function Optimisation},
+  pages = {232--244},
+  doi = {10.1007/978-3-540-76931-6_21}
+}
+
+ +
+@inproceedings{AnsKamVeeRag2014open,
+  year = 2014,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of the 23rd International Conference on Parallel
+                  Architectures and Compilation},
+  key = {PACT},
+  author = {J. Ansel and S. Kamil and K. Veeramachaneni and J. Ragan-Kelley and J. Bosboom and  Una-May O'Reilly  and S. Amarasinghe},
+  title = {{OpenTuner}: An extensible framework for program autotuning},
+  pages = {303--315},
+  doi = {10.1145/2628071.2628092}
+}
+
+ +
+@inproceedings{AnsMalSamSelTie2015:ijcai,
+  publisher = {IJCAI/AAAI Press, Menlo Park, CA},
+  editor = {Qiang Yang and Michael Wooldridge},
+  year = 2015,
+  booktitle = {Proceedings of  the 24th International Joint Conference on Artificial Intelligence (IJCAI-15)},
+  author = { Carlos Ans{\'o}tegui  and  Yuri Malitsky  and Horst Samulowitz and  Meinolf Sellmann  and  Kevin Tierney },
+  title = {Model-Based Genetic Algorithms for Algorithm Configuration},
+  pages = {733--739},
+  keywords = {GGA++},
+  epub = {https://www.ijcai.org/Abstract/15/109}
+}
+
+ +
+@inproceedings{AnsMalSel2014isacpp,
+  year = 2014,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {David Stracuzzi and others},
+  author = { Carlos Ans{\'o}tegui  and  Yuri Malitsky  and  Meinolf Sellmann },
+  title = {{MaxSAT} by Improved Instance-Specific Algorithm
+                  Configuration},
+  pages = {2594--2600}
+}
+
+ +
+@incollection{AnsSelTie2009cp,
+  year = 2009,
+  volume = 5732,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Principles and Practice of Constraint Programming,
+                  CP 2009},
+  editor = { Ian P. Gent },
+  author = { Carlos Ans{\'o}tegui  and  Meinolf Sellmann  and  Kevin Tierney },
+  title = {A Gender-Based Genetic Algorithm for the Automatic
+                  Configuration of Algorithms},
+  pages = {142--157},
+  doi = {10.1007/978-3-642-04244-7_14},
+  alias = {Ansotegui2009},
+  keywords = {GGA}
+}
+
+ +
+@techreport{AppBixChvCoo95:tr,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {Finding Cuts in the {TSP}},
+  institution = {DIMACS Center, Rutgers University, Piscataway, NJ, USA},
+  year = 1995,
+  number = {95--05},
+  month = mar
+}
+
+ +
+@techreport{AppBixChvCoo99:tr,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {Finding Tours in the {TSP}},
+  institution = {Forschungsinstitut f{\"u}r Diskrete Mathematik, University of Bonn, Germany},
+  year = 1999,
+  number = 99885
+}
+
+ +
+@book{AppEtAl06,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {The Traveling Salesman Problem: A Computational Study},
+  publisher = {Princeton University Press, Princeton, NJ},
+  year = 2006
+}
+
+ +
+@inproceedings{AprGloKel2003,
+  volume = 1,
+  month = dec,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  editor = {Stephen E. Chick and Paul J. Sanchez and David M. Ferrin and Douglas J. Morrice},
+  year = 2003,
+  booktitle = {Proceedings of the 35th Winter Simulation Conference: Driving Innovation},
+  author = { Jay April  and  Fred Glover  and  James P. Kelly  and  Manuel Laguna },
+  title = {Simulation-based optimization: Practical introduction to simulation optimization},
+  pages = {71--78},
+  doi = {10.1109/WSC.2003.1261410}
+}
+
+ +
+@book{AroBar2009,
+  title = {Computational complexity: a modern approach},
+  author = {Arora, Sanjeev and Barak, Boaz},
+  year = 2009,
+  publisher = {Cambridge University Press}
+}
+
+ +
+@incollection{ArzCebPer2019qap,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = {Etor Arza and  Josu Ceberio  and Aritz P{\'{e}}rez and  Irurozki, Ekhine },
+  title = {Approaching the quadratic assignment problem with kernels of
+                  mallows models under the hamming distance},
+  doi = {10.1145/3319619.3321976},
+  keywords = {QAP, EDA, Mallows}
+}
+
+ +
+@incollection{AsaIwaMiy96,
+  series = {{DIMACS} Series on Discrete Mathematics and Theoretical Computer Science},
+  volume = 26,
+  year = 1996,
+  address = { Providence, RI},
+  publisher = {American Mathematical Society},
+  booktitle = {Cliques, Coloring, and Satisfiability: Second {DIMACS}
+                  Implementation Challenge},
+  editor = {David S. Johnson and  Michael A. Trick },
+  author = {Y. Asahiro and K. Iwama and E. Miyano},
+  title = {Random Generation of Test Instances with Controlled
+                  Attributes},
+  pages = {377--393}
+}
+
+ +
+@phdthesis{Asch95PhD,
+  author = { N. Ascheuer },
+  title = {Hamiltonian Path Problems in the On-line
+                  Optimization of Flexible Manufacturing Systems},
+  school = {Technische Universit{\"a}t Berlin},
+  year = 1995,
+  address = {Berlin, Germany}
+}
+
+ +
+@incollection{Atkinson00,
+  author = { R. Atkinson  and  Jakobus E. van Zyl  and  Godfrey A. Walters  and  Dragan A. Savic },
+  title = {Genetic algorithm optimisation of level-controlled
+                  pumping station operation},
+  booktitle = {Water network modelling for optimal design and
+                  management},
+  pages = {79--90},
+  publisher = {Centre for Water Systems, Exeter, UK},
+  year = 2000
+}
+
+ +
+@incollection{AudDanOrb10,
+  editor = {K. Naono and K. Teranishi and J. Cavazos and R. Suda},
+  year = 2010,
+  publisher = {Springer},
+  booktitle = {Software Automatic Tuning: From Concepts to State-of-the-Art Results},
+  author = { Charles Audet  and  Cong-Kien Dang  and  Dominique Orban },
+  title = {Algorithmic Parameter Optimization of the {DFO} Method with
+                  the {OPAL} Framework},
+  pages = {255--274}
+}
+
+ +
+@incollection{AugBadBroZit2009gecco,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = { Anne Auger  and  Johannes Bader  and  Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Articulating User Preferences in Many-Objective
+                  Problems by Sampling the Weighted Hypervolume},
+  pages = {555--562}
+}
+
+ +
+@incollection{AugBadBroZit2009gecco2,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = { Anne Auger  and  Johannes Bader  and  Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Investigating and Exploiting the Bias of the
+                  Weighted Hypervolume to Articulate User Preferences},
+  pages = {563--570}
+}
+
+ +
+@incollection{AugBadBroZit2009hv,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  title = {Theory of the hypervolume indicator: optimal
+                  $\mu$-distributions and the choice of the reference point},
+  author = { Anne Auger  and  Johannes Bader  and  Dimo Brockhoff  and  Eckart Zitzler },
+  pages = {87--102}
+}
+
+ +
+@incollection{AugBroLop2012dagstuhl,
+  doi = {10.4230/DagRep.2.1.50},
+  series = {Dagstuhl Reports},
+  volume = {2(1)},
+  year = 2012,
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  booktitle = {Learning in Multiobjective Optimization (Dagstuhl Seminar
+                  12041)},
+  editor = { Salvatore Greco  and  Joshua D. Knowles  and  Kaisa Miettinen  and  Eckart Zitzler },
+  author = { Anne Auger  and  Dimo Brockhoff  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kaisa Miettinen  and  Boris Naujoks  and  G{\"u}nther Rudolph },
+  title = {Which questions should be asked to find the most appropriate
+                  method for decision making and problem solving? ({Working}
+                  {Group} ``{Algorithm} {Design} {Methods}'')},
+  pages = {92--93}
+}
+
+ +
+@book{AugDoe2011,
+  editor = { Anne Auger  and  Benjamin Doerr },
+  title = {Theory of Randomized Search Heuristics: Foundations and Recent Developments},
+  series = {Series on Theoretical Computer Science},
+  volume = 1,
+  publisher = {World Scientific Publishing Co., Singapore},
+  year = 2011
+}
+
+ +
+@inproceedings{AugHan2005cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2005,
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  key = {IEEE CEC},
+  author = { Anne Auger  and  Nikolaus Hansen },
+  title = {A restart {CMA} evolution strategy with increasing population
+                  size},
+  pages = {1769--1776},
+  doi = {10.1109/CEC.2005.1554902},
+  keywords = {IPOP-CMA-ES}
+}
+
+ +
+@inproceedings{AugHan2005lrcmaes,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2005,
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  key = {IEEE CEC},
+  author = { Anne Auger  and  Nikolaus Hansen },
+  title = {Performance evaluation of an advanced local search
+                  evolutionary algorithm},
+  pages = {1777--1784},
+  keywords = {LR-CMAES}
+}
+
+ +
+@incollection{AvrAllLop2021evo,
+  volume = {12694},
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  year = 2021,
+  editor = {Pedro Castillo and  Jim{\'e}nez Laredo, Juan Luis },
+  author = { Andreea Avramescu  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {A Multi-Objective Multi-Type Facility Location Problem for
+                  the Delivery of Personalised Medicine},
+  pages = {388--403},
+  doi = {10.1007/978-3-030-72699-7_25},
+  abstract = {Advances in personalised medicine targeting specific
+                  sub-populations and individuals pose a challenge to the
+                  traditional pharmaceutical industry. With a higher level of
+                  personalisation, an already critical supply chain is facing
+                  additional demands added by the very sensitive nature of its
+                  products. Nevertheless, studies concerned with the efficient
+                  development and delivery of these products are scarce. Thus,
+                  this paper presents the case of personalised medicine and the
+                  challenges imposed by its mass delivery. We propose a
+                  multi-objective mathematical model for the
+                  location-allocation problem with two interdependent facility
+                  types in the case of personalised medicine products. We show
+                  its practical application through a cell and gene therapy
+                  case study. A multi-objective genetic algorithm with a novel
+                  population initialisation procedure is used as solution
+                  method.},
+  supplement = {https://doi.org/10.5281/zenodo.4495162},
+  keywords = {Personalised medicine, Biopharmaceuticals Supply chain,
+                  Facility location-allocation, Evolutionary multi-objective
+                  optimisation}
+}
+
+ +
+@incollection{AydYavOzyYasStu2017,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2017},
+  author = { Do\v{g}an Ayd{\i}n  and  G{\"{u}}rcan Yavuz  and Serdar \"Ozy\"on and Celal Yasar and  Thomas St{\"u}tzle },
+  title = {Artificial Bee Colony Framework to Non-convex Economic
+                  Dispatch Problem with Valve Point Effects: A Case Study},
+  pages = {1311--1318}
+}
+
+ +
+@incollection{AyoAllLop2023gecco,
+  location = {Lisbon, Portugal},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2023},
+  annote = {ISBN: 979-8-4007-0120-7},
+  address = { New York, NY},
+  year = 2023,
+  publisher = {ACM Press},
+  editor = {Silva, Sara and  Lu{\'i}s Paquete },
+  author = { Ayodele, Mayowa  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Parizy, Matthieu  and  Arnaud Liefooghe },
+  title = {Applying {Ising} Machines to Multi-Objective {QUBOs}},
+  pages = {2166--2174},
+  doi = {10.1145/3583133.3596312},
+  abstract = {Multi-objective optimisation problems involve finding
+                  solutions with varying trade-offs between multiple and often
+                  conflicting objectives. Ising machines are physical devices
+                  that aim to find the absolute or approximate ground states of
+                  an Ising model. To apply Ising machines to multi-objective
+                  problems, a weighted sum objective function is used to
+                  convert multi-objective into single-objective
+                  problems. However, deriving scalarisation weights that
+                  archives evenly distributed solutions across the Pareto front
+                  is not trivial. Previous work has shown that adaptive weights
+                  based on dichotomic search, and one based on averages of
+                  previously explored weights can explore the Pareto front
+                  quicker than uniformly generated weights. However, these
+                  adaptive methods have only been applied to bi-objective
+                  problems in the past. In this work, we extend the adaptive
+                  method based on averages in two ways: (i) we extend the
+                  adaptive method of deriving scalarisation weights for
+                  problems with two or more objectives, and (ii) we use an
+                  alternative measure of distance to improve performance. We
+                  compare the proposed method with existing ones and show that
+                  it leads to the best performance on multi-objective
+                  Unconstrained Binary Quadratic Programming (mUBQP) instances
+                  with 3 and 4 objectives and that it is competitive with the
+                  best one for instances with 2 objectives.},
+  numpages = 9,
+  keywords = {digital annealer, multi-objective, bi-objective QAP, QUBO}
+}
+
+ +
+@incollection{AyoAllLop2022gecco,
+  location = {Boston, Massachusetts},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Ayodele, Mayowa  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Parizy, Matthieu },
+  title = {Multi-Objective {QUBO} Solver: Bi-Objective Quadratic
+                  Assignment Problem},
+  pages = {467--475},
+  doi = {10.1145/3512290.3528698},
+  abstract = {Quantum and quantum-inspired optimisation algorithms are
+                  designed to solve problems represented in binary, quadratic
+                  and unconstrained form. Combinatorial optimisation problems
+                  are therefore often formulated as Quadratic Unconstrained
+                  Binary Optimisation Problems (QUBO) to solve them with these
+                  algorithms. Moreover, these QUBO solvers are often
+                  implemented using specialised hardware to achieve enormous
+                  speedups, e.g. Fujitsu's Digital Annealer (DA) and D-Wave's
+                  Quantum Annealer. However, these are single-objective
+                  solvers, while many real-world problems feature multiple
+                  conflicting objectives. Thus, a common practice when using
+                  these QUBO solvers is to scalarise such multi-objective
+                  problems into a sequence of single-objective problems. Due to
+                  design trade-offs of these solvers, formulating each
+                  scalarisation may require more time than finding a local
+                  optimum. We present the first attempt to extend the algorithm
+                  supporting a commercial QUBO solver as a multi-objective
+                  solver that is not based on scalarisation. The proposed
+                  multi-objective DA algorithm is validated on the bi-objective
+                  Quadratic Assignment Problem. We observe that algorithm
+                  performance significantly depends on the archiving strategy
+                  adopted, and that combining DA with non-scalarisation methods
+                  to optimise multiple objectives outperforms the current
+                  scalarised version of the DA in terms of final solution
+                  quality.},
+  numpages = 9,
+  keywords = {digital annealer, multi-objective, bi-objective QAP, QUBO}
+}
+
+ +
+@incollection{AyoAllLop2022or,
+  booktitle = {Operations Research Proceedings 2022, OR 2022},
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Operations Research},
+  year = 2022,
+  publisher = {Springer},
+  editor = {Oliver Grothe and Stefan Nickel and Steffen Rebennack and
+                  Oliver Stein},
+  author = { Ayodele, Mayowa  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Parizy, Matthieu },
+  title = {A Study of Scalarisation Techniques for Multi-objective
+                  {QUBO} Solving},
+  pages = {393--399},
+  doi = {10.1007/978-3-031-24907-5_47}
+}
+
+ +
+@incollection{Ayodele2022penalty,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2022,
+  booktitle = {Proceedings of EvoCOP 2022 -- 22nd European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = {  P{\'e}rez C{\'a}ceres, Leslie and  Verel, S{\'e}bastien },
+  title = {Penalty Weights in {QUBO} Formulations: Permutation Problems},
+  author = { Ayodele, Mayowa },
+  pages = {159--174}
+}
+
+ +
+@incollection{AziDoeDre2021,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  author = {Aziz-Alaoui, Amine and  Carola Doerr  and  Johann Dr{\'e}o },
+  title = {Towards Large Scale Automated Algorithm Design by Integrating Modular Benchmarking Frameworks},
+  pages = {1365--1374},
+  doi = {10.1145/3449726.3463155}
+}
+
+ +
+@misc{BBCOMP2017,
+  title = {Black Box Optimization Competition},
+  author = {Ilya Loshchilov  and  T. Glasmachers },
+  year = 2017,
+  url = {https://bbcomp.ini.rub.de/},
+  alias = {Loshchilov2017}
+}
+
+ +
+@misc{BBOB2016bi,
+  author = { Anne Auger  and  Dimo Brockhoff  and  Nikolaus Hansen  and Dejan Tusar and  Tea Tu{\v s}ar  and  Tobias Wagner },
+  title = {{GECCO} Workshop on Real-Parameter Black-Box Optimization
+                  Benchmarking ({BBOB} 2016): Focus on multi-objective
+                  problems},
+  howpublished = {\url{https://numbbo.github.io/workshops/BBOB-2016/}},
+  year = 2016
+}
+
+ +
+@incollection{ZitLauBleu2004tutorial,
+  year = 2004,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  volume = 535,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  editor = { Xavier Gandibleux  and Marc Sevaux and  Kenneth S{\"o}rensen  and  V. {T'Kindt} },
+  booktitle = {Metaheuristics for Multiobjective Optimisation},
+  title = {A tutorial on evolutionary multiobjective optimization},
+  author = { Eckart Zitzler  and  Marco Laumanns  and  S. Bleuler },
+  pages = {3--37}
+}
+
+ +
+@incollection{BLTZ2003a,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { S. Bleuler  and  Marco Laumanns  and  Lothar Thiele  and  Eckart Zitzler },
+  title = {{PISA} -- A Platform and Programming Language
+                  Independent Interface for Search Algorithms },
+  pages = {494--508}
+}
+
+ +
+@misc{Bab2008spear,
+  author = { Domagoj Babi{\'c} },
+  title = {Spear theorem prover},
+  howpublished = {\url{https://www.domagoj-babic.com/index.php/ResearchProjects/Spear}},
+  year = 2008
+}
+
+ +
+@inproceedings{BabHu2007cav,
+  author = { Domagoj Babi{\'c}  and  Alan J. Hu},
+  title = {Structural Abstraction of Software Verification
+                  Conditions},
+  booktitle = {Computer Aided Verification: 19th International
+                  Conference, CAV 2007},
+  year = 2007,
+  pages = {366--378},
+  annote = {Spear-swv instances,
+                  \url{http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/benchmark_instances/SpearSWV/SWV-scrambled-first302.tar.gz},
+                  \url{http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/benchmark_instances/SpearSWV/SWV-scrambled-last302.tar.gz}}
+}
+
+ +
+@inproceedings{BabHut2008spear,
+  author = { Domagoj Babi{\'c}  and  Frank Hutter },
+  title = {Spear Theorem Prover},
+  booktitle = {SAT'08: Proceedings of the SAT 2008 Race},
+  year = 2008,
+  annote = {Unreviewed paper},
+  epub = {https://www.domagoj-babic.com/index.php/Pubs/SAT08},
+  supplement = {https://www.domagoj-babic.com/index.php/ResearchProjects/Spear}
+}
+
+ +
+@book{BacFogMic1997,
+  title = {Handbook of evolutionary computation},
+  author = { Thomas B{\"a}ck  and  David B. Fogel  and  Zbigniew Michalewicz },
+  year = 1997,
+  publisher = {IOP Publishing}
+}
+
+ +
+@techreport{BacSteWot1994tr,
+  author = {Achim Bachem and Barthel Steckemetz and Michael
+                  Wottawa},
+  title = {An efficient parallel cluster-heuristic for large
+                  Traveling Salesman Problems},
+  year = 1994,
+  institution = {University of Koln, Germany},
+  number = {94-150},
+  keywords = {Genetic Edge Recombination (ERX)}
+}
+
+ +
+@book{Back1996evolutionary,
+  author = { Thomas B{\"a}ck },
+  title = {Evolutionary algorithms in theory and practice: evolution
+                  strategies, evolutionary programming, genetic algorithms},
+  year = 1996,
+  publisher = {Oxford University Press}
+}
+
+ +
+@incollection{BalBirStu06,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Incremental local search in ant colony optimization:
+                  Why it fails for the quadratic assignment problem},
+  pages = {156--166}
+}
+
+ +
+@incollection{BalBirStu07,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4771,
+  editor = { Thomas Bartz-Beielstein  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Boris Naujoks  and  Andrea Roli  and  G{\"u}nther Rudolph  and  M. Sampels },
+  year = 2007,
+  booktitle = {Hybrid Metaheuristics},
+  author = {  Prasanna Balaprakash  and  Mauro Birattari  and  Thomas St{\"u}tzle },
+  title = {Improvement Strategies for the {F}-Race Algorithm:
+                  Sampling Design and Iterative Refinement},
+  pages = {108--122},
+  keywords = {Iterated Race},
+  doi = {10.1007/978-3-540-75514-2_9}
+}
+
+ +
+@incollection{BalHo1980,
+  author = { Egon Balas  and Andrew Ho},
+  title = {Set Covering Algorithms Using Cutting Planes, Heuristics, and
+                  Subgradient Optimization: A Computational Study},
+  booktitle = {Combinatorial optimization},
+  series = {Mathematical Programming Studies},
+  year = 1980,
+  volume = 12,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  pages = {37--60},
+  editor = {Padberg, M. W.},
+  doi = {10.1007/BFb0120886}
+}
+
+ +
+@inproceedings{BapHgu1997,
+  author = {P. Baptiste and L. K. Hguny},
+  title = {A branch and bound algorithm for the F$/$no\_idle$/C_\text{max}$},
+  booktitle = {Proceedings of the international conference on industrial engineering and production management, IEPM'97},
+  year = 1997,
+  address = {Lyon},
+  pages = {429--438}
+}
+
+ +
+@book{Bar2006newexp,
+  author = { Thomas Bartz-Beielstein },
+  title = {Experimental Research in Evolutionary Computation:
+                  The New Experimentalism},
+  publisher = {Springer},
+  year = 2006,
+  address = { Berlin, Germany},
+  keywords = {SPO}
+}
+
+ +
+@incollection{Bar2015genera,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  editor = {Kacprzyk, Janusz and Pedrycz, Witold},
+  booktitle = {Springer Handbook of Computational Intelligence},
+  year = 2015,
+  author = { Thomas Bartz-Beielstein },
+  title = {How to Create Generalizable Results},
+  pages = {1127--1142},
+  keywords = {Mixed-effects models, random-effects model, problem instance
+                  generation}
+}
+
+ +
+@inproceedings{BarFlaKocKon2010spot,
+  title = {{SPOT}: A Toolbox for Interactive and Automatic Tuning in the
+                  \proglang{R} Environment},
+  author = { Thomas Bartz-Beielstein  and Flasch, Oliver and Koch, Patrick
+                  and Konen, Wolfgang},
+  booktitle = {Proceedings 20. Workshop Computational Intelligence},
+  year = 2010,
+  address = {Karlsruhe},
+  publisher = {KIT Scientific Publishing},
+  alias = {Bartz-Beielstein2010},
+  pages = {264--273}
+}
+
+ +
+@inproceedings{BarLasPre2005cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2005,
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  key = {IEEE CEC},
+  author = { Thomas Bartz-Beielstein  and  C. Lasarczyk  and  Mike Preuss },
+  title = {Sequential Parameter Optimization},
+  pages = {773--780}
+}
+
+ +
+@incollection{BarLasPre2010emaoa,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = { Thomas Bartz-Beielstein  and  C. Lasarczyk  and  Mike Preuss },
+  title = {The Sequential Parameter Optimization Toolbox},
+  pages = {337--360},
+  keywords = {SPOT},
+  doi = {10.1007/978-3-642-02538-9_14}
+}
+
+ +
+@inproceedings{BarMar2004,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2004,
+  booktitle = {Proceedings of the 2004 Congress on Evolutionary
+                  Computation (CEC 2004)},
+  key = {IEEE CEC},
+  title = {Tuning search algorithms for real-world applications: A
+                  regression tree based approach},
+  author = { Thomas Bartz-Beielstein  and Markon, Sandor},
+  pages = {1111--1118}
+}
+
+ +
+@inproceedings{BarPea2012aaai,
+  year = 2012,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Jorg Hoffmann and Bart Selman},
+  title = {Transportability of causal effects: Completeness results},
+  author = { Elias Bareinboim  and  Judea Pearl },
+  pages = {698,704}
+}
+
+ +
+@inproceedings{BarPre2005em,
+  address = {Reykjavik, Iceland},
+  editor = { Lu{\'i}s Paquete  and  Marco Chiarandini  and Dario Basso},
+  year = {2006},
+  booktitle = {Empirical Methods for the Analysis of Algorithms, Workshop EMAA 2006, Proceedings},
+  author = { Thomas Bartz-Beielstein  and  Mike Preuss },
+  title = {Considerations of budget allocation for sequential parameter
+                  optimization ({SPO})},
+  pages = {35--40}
+}
+
+ +
+@incollection{BarPre2014experimental,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  year = 2014,
+  series = {Natural Computing Series},
+  booktitle = {Theory and Principled Methods for the Design of
+                  Metaheuristics},
+  editor = {Borenstein, Yossi and  A. Moraglio },
+  author = { Thomas Bartz-Beielstein  and  Mike Preuss },
+  title = {Experimental Analysis of Optimization Algorithms: Tuning and
+                  Beyond},
+  doi = {10.1007/978-3-642-33206-7_10},
+  pages = {205--245}
+}
+
+ +
+@inproceedings{BarSch03,
+  author = { Benjam{\'i}n Bar{\'a}n  and  Matilde Schaerer },
+  title = {A multiobjective ant colony system for vehicle
+                  routing problem with time windows},
+  booktitle = {Proceedings of the Twenty-first IASTED International
+                  Conference on Applied Informatics},
+  pages = {97--102},
+  year = 2003,
+  address = {Insbruck, Austria}
+}
+
+ +
+@incollection{BasGoeLie2013gecco,
+  isbn = {978-1-4503-1963-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = { Christian Blum  and  Alba, Enrique },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  author = {Basseur, Matthieu and   Go{\"e}ffon, Adrien  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien },
+  title = {On Set-based Local Search for Multiobjective Combinatorial
+                  Optimization},
+  pages = {471--478},
+  doi = {10.1145/2463372.2463430},
+  acmid = 2463430
+}
+
+ +
+@incollection{BasYevDeuEmm2017,
+  booktitle = {{EVOLVE} - A Bridge between Probability, Set Oriented
+                  Numerics, and Evolutionary Computation {VII}},
+  address = { Cham, Switzerland},
+  series = {Studies in Computational Intelligence},
+  volume = 662,
+  year = 2017,
+  publisher = {Springer},
+  author = {Basto-Fernandes, Vitor and Yevseyeva, Iryna and   Andr{\'{e}} Deutz  and  Emmerich, Michael T. M. },
+  title = {A survey of diversity oriented optimization: Problems,
+                  indicators, and algorithms},
+  pages = {3--23},
+  doi = {10.1007/978-3-319-49325-1_1}
+}
+
+ +
+@book{BatBruMas08:book,
+  author = { Roberto Battiti  and M. Brunato and  Franco Mascia },
+  title = {Reactive Search and Intelligent Optimization},
+  publisher = {Springer},
+  address = { New York, NY},
+  year = 2008,
+  series = {Operations Research/Computer Science Interfaces},
+  volume = 45,
+  doi = {10.1007/978-0-387-09624-7}
+}
+
+ +
+@inproceedings{BatCam2009reactive,
+  address = {Hamburg, Germany},
+  publisher = {University of Hamburg},
+  editor = {M. Caserta and  Stefan Vo{\ss} },
+  year = 2010,
+  booktitle = {Proceedings of MIC 2009, the 8th Metaheuristics International Conference},
+  title = {Reactive search optimization: Learning while optimizing. An
+                  experiment in interactive multi-objective optimization},
+  author = { Roberto Battiti  and  Paolo Campigotto }
+}
+
+ +
+@inproceedings{BatSchUrl2014patat,
+  publisher = {PATAT},
+  year = 2014,
+  editor = { Ender {\"O}zcan  and Edmund K. Burke and Barry McCollum},
+  booktitle = {PATAT 2014: Proceedings of the 10th International Conference of the Practice and Theory of Automated Timetabling},
+  author = {Michele Battistutta and Andrea Schaerf and  Tommaso Urli },
+  title = {Feature-based tuning of single-stage simulated annealing for examination timetabling},
+  pages = {53--61},
+  keywords = {F-race}
+}
+
+ +
+@unpublished{Bau1986a,
+  author = {E. B. Baum},
+  title = {Iterated Descent: A Better Algorithm for Local
+                  Search in Combinatorial Optimization Problems},
+  note = {Manuscript},
+  year = 1986
+}
+
+ +
+@inproceedings{Bau1986b,
+  author = {E. B. Baum},
+  title = {Towards Practical ``Neural'' Computation for Combinatorial
+                  Optimization Problems},
+  booktitle = {Neural Networks for Computing, AIP Conference Proceedings},
+  year = 1986,
+  pages = {53--64}
+}
+
+ +
+@inproceedings{BayDerSab2005cie,
+  author = {A. Baykasoglu and T. Dereli and I. Sabuncu},
+  title = {A multiple objective ant colony optimization
+                  approach to assembly line balancing problems},
+  booktitle = {35th International Conference on Computers and
+                  Industrial Engineering (CIE35)},
+  pages = {263--268},
+  year = 2005,
+  address = {Istanbul, Turkey}
+}
+
+ +
+@inproceedings{BeaGra2002,
+  title = {Improved Distributed Hypercube Sampling},
+  author = {Beachkofski, Brian and Grandhi, Ramana},
+  booktitle = {Proceedings of the 43rd AIAA/ASME/ASCE/AHS/ASC Structures,
+                  Structural Dynamics, and Materials Conference},
+  year = 2002,
+  organization = {AIAA paper 2002-1274},
+  publisher = {American Institute of Aeronautics and Astronautics},
+  alias = {Beachkofski2002}
+}
+
+ +
+@techreport{Beasley1998,
+  title = {Heuristic algorithms for the unconstrained binary quadratic
+                  programming problem},
+  author = { John E. Beasley },
+  address = {London, England},
+  year = 1998,
+  institution = {The Management School, Imperial College},
+  epub = {http://people.brunel.ac.uk/~mastjjb/jeb/bqp.pdf}
+}
+
+ +
+@incollection{BecGotStu05,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3871,
+  editor = { Talbi, El-Ghazali  and Pierre Liardet and Pierre
+                  Collet and Evelyne Lutton and Marc Schoenauer},
+  year = 2005,
+  booktitle = {Artificial Evolution},
+  author = {S. Becker and J. Gottlieb and  Thomas St{\"u}tzle },
+  title = {Applications of Racing Algorithms: An Industrial
+                  Perspective},
+  pages = {271--283}
+}
+
+ +
+@book{BedBai1982,
+  title = {Integrated Production Control Systems: Management, Analysis,
+                  Design},
+  author = {Bedworth, David D. and Bailey, James E.},
+  volume = 2,
+  year = 1982,
+  publisher = {John Wiley \& Sons},
+  address = { New York, NY}
+}
+
+ +
+@incollection{BehAffWag2017,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2017},
+  title = {Instance-based algorithm selection on quadratic assignment problem landscapes},
+  author = {Beham, Andreas and Affenzeller, Michael and Wagner, Stefan},
+  pages = {1471--1478}
+}
+
+ +
+@incollection{BelBraEsk2088moointevo,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  author = {Belton, Valerie and  J{\"u}rgen Branke  and Eskelinen,
+                  Petri and  Salvatore Greco  and  Molina, Juli{\'a}n  and  Francisco Ruiz  and  Roman S{\l}owi{\'n}ski },
+  title = {Interactive Multiobjective Optimization from a
+                  Learning Perspective},
+  pages = {405--433},
+  doi = {10.1007/978-3-540-88908-3_15}
+}
+
+ +
+@incollection{BelDreSavSch2016:ppsn,
+  isbn = {978-3-319-45822-9},
+  doi = {10.1007/978-3-319-45823-6},
+  year = 2016,
+  volume = 9921,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Julia Handl  and  Emma Hart  and  Lewis, P. R.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Gabriela Ochoa  and  Ben Paechter },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIV}},
+  author = {Nacim Belkhir and  Johann Dr{\'e}o  and Pierre Sav{\'{e}}ant and  Marc Schoenauer },
+  title = {Feature Based Algorithm Configuration: A Case Study with Differential Evolution},
+  pages = {156--166}
+}
+
+ +
+@incollection{BelDreSavSch2017:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  author = {Nacim Belkhir and  Johann Dr{\'e}o  and Pierre Sav{\'{e}}ant and  Marc Schoenauer },
+  title = {Per Instance Algorithm Configuration of {CMA-ES} with Limited Budget},
+  pages = {681--688}
+}
+
+ +
+@inproceedings{Ben90,
+  author = { Jon Louis Bentley },
+  title = {Experiments on Traveling Salesman Heuristics},
+  booktitle = {Proceedings of the First Annual ACM-SIAM Symposium on
+                  Discrete Algorithms},
+  year = 1990,
+  editor = {David S. Johnson},
+  pages = {91--99},
+  publisher = {Society for Industrial and Applied Mathematics, Philadelphia, PA, USA}
+}
+
+ +
+@inproceedings{BenCasLus2020interactive,
+  publisher = {{AAAI} Press},
+  isbn = {978-1-57735-823-7},
+  year = 2020,
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  key = {AAAI2020},
+  author = {Nawal Benabbou and Cassandre Leroy and  Thibaut Lust },
+  title = {An Interactive Regret-Based Genetic Algorithm for Solving
+                  Multi-Objective Combinatorial Optimization Problems},
+  pages = {2335--2342},
+  doi = {10.1609/aaai.v34i03.5612},
+  keywords = {interactive, multi-objective, decision-makers}
+}
+
+ +
+@inproceedings{BenCorManZaf2015,
+  epub = {http://jmlr.org/proceedings/papers/v37/},
+  publisher = {{PMLR}},
+  year = 2015,
+  volume = 37,
+  booktitle = {Proceedings of  the 32nd International Conference on Machine Learning, {ICML} 2015},
+  editor = {Francis Bach and David Blei},
+  author = {Alessio Benavoli and Giorgio Corani and Francesca Mangili and
+                  Marco Zaffalon},
+  title = {A {Bayesian} nonparametric procedure for comparing
+                  algorithms},
+  pages = {1264--1272},
+  keywords = {racing}
+}
+
+ +
+@incollection{BenRit2015:icaps,
+  year = 2015,
+  address = { Menlo Park, CA},
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of the Twenty-Fifth International Conference on
+                  Automated Planning and Scheduling, {ICAPS} 2015},
+  editor = { Ronen I. Brafman  and Carmel Domshlak and Patrik Haslum and  Shlomo Zilberstein },
+  author = {Alexander Javier Benavides and  Marcus Ritt},
+  title = {Iterated Local Search Heuristics for Minimizing Total
+                  Completion Time in Permutation and Non-permutation Flow
+                  Shops},
+  pages = {34--41},
+  keywords = {irace}
+}
+
+ +
+@incollection{BenRolBlu2010,
+  doi = {10.1007/978-3-642-13800-3},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Roberto Battiti },
+  series = {Lecture Notes in Computer Science},
+  volume = 6073,
+  year = 2010,
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  author = {Stefano Benedettini and  Andrea Roli  and  Christian Blum },
+  title = {A Randomized Iterated Greedy Algorithm for the Founder Sequence Reconstruction
+               Problem},
+  pages = {37--51}
+}
+
+ +
+@incollection{BenRolDig2008ants,
+  volume = 5217,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Christian Blum  and  Clerc, Maurice  and  Thomas St{\"u}tzle  and A. F. T. Winfield},
+  year = 2008,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 6th
+                  International Conference, ANTS 2008},
+  author = {Stefano Benedettini and  Andrea Roli  and Luca {Di Gaspero}},
+  title = {Two-level {ACO} for Haplotype Inference under Pure Parsimony},
+  pages = {179--190},
+  alias = {ants2008-benedigaroli}
+}
+
+ +
+@book{Ber95,
+  author = {D. Bertsekas},
+  title = {Dynamic Programming and Optimal Control},
+  year = 1995,
+  publisher = {Athena Scientific, Belmont, MA}
+}
+
+ +
+@book{Ber98,
+  author = {D. Bertsekas},
+  title = {Network Optimization: Continuous and Discrete Models},
+  year = 1998,
+  publisher = {Athena Scientific, Belmont, MA}
+}
+
+ +
+@incollection{BerBarBen2011nips,
+  publisher = {Curran Associates, Red Hook, NY},
+  year = 2011,
+  editor = {J. Shawe-Taylor and R. S. Zemel and P. L. Bartlett and
+                  F. Pereira and K. Q. Weinberger},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 24)},
+  title = {Algorithms for Hyper-Parameter Optimization},
+  author = { James S. Bergstra  and Bardenet, R\'{e}mi and  Bengio, Yoshua  and Bal\'{a}zs K\'{e}gl},
+  pages = {2546--2554},
+  url = {http://papers.nips.cc/paper/4443-algorithms-for-hyper-parameter-optimization.pdf}
+}
+
+ +
+@incollection{BerDorLanSerGam1996:icec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1996,
+  editor = { Thomas B{\"a}ck  and  T. Fukuda  and  Zbigniew Michalewicz },
+  booktitle = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  author = { Hughes Bersini  and  Marco Dorigo  and  S. Langerman  and  G. Seront  and  L. M. Gambardella },
+  title = {Results of the First International Contest on
+                  Evolutionary Optimisation},
+  pages = {611--615}
+}
+
+ +
+@inproceedings{BerSchKra2016safe,
+  author = {Berkenkamp, Felix and Schoellig, Angela P. and Krause,
+                  Andreas},
+  title = {Safe controller optimization for quadrotors with {Gaussian}
+                  processes},
+  booktitle = {2016 IEEE International Conference on Robotics and Automation
+                  (ICRA)},
+  year = 2016,
+  pages = {491--496},
+  organization = {IEEE},
+  doi = {10.1109/ICRA.2016.7487170},
+  keywords = {Safe Optimization, SafeOpt}
+}
+
+ +
+@inproceedings{BerYasCox2013,
+  url = {http://jmlr.org/proceedings/papers/v28/},
+  year = 2013,
+  volume = 28,
+  booktitle = {Proceedings of  the 30th International Conference on Machine Learning, {ICML} 2013},
+  editor = {Dasgupta, Sanjoy and McAllester, David},
+  title = {Making a science of model search: Hyperparameter optimization
+                  in hundreds of dimensions for vision architectures},
+  author = { James S. Bergstra  and Yasmin, Daniel and Cox, David},
+  pages = {115--123}
+}
+
+ +
+@incollection{BesStuDor00:ppsn,
+  anote = {IC.29},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marc Schoenauer  and others},
+  aeditor = { Marc Schoenauer  and  Kalyanmoy Deb  and  G{\"u}nther Rudolph  and  Xin Yao  and E. Lutton and  Juan-Juli{\'a}n Merelo  and  Hans-Paul Schwefel },
+  year = 2000,
+  volume = 1917,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VI}},
+  author = { Matthijs L. {den Besten}  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Ant Colony Optimization for the Total Weighted
+                  Tardiness Problem},
+  pages = {611--620}
+}
+
+ +
+@incollection{BesStuDor2001:evoworkshops,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2037,
+  editor = {E. J. W. Boers and others},
+  aeditor = {E. J. W. Boers and J. Gottlieb and P. L. Lanzi and R. E. Smith
+                 and S. Cagnoni and E. Hart and G. R. Raidl and H. Tijink},
+  year = 2001,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2001},
+  author = { Matthijs L. {den Besten}  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Design of Iterated Local Search Algorithms: An
+                  Example Application to the Single Machine Total
+                  Weighted Tardiness Problem},
+  pages = {441--452}
+}
+
+ +
+@inproceedings{BeuRud06:hypervolume,
+  author = { Nicola Beume  and  G{\"u}nther Rudolph },
+  title = {Faster {S}-Metric Calculation by Considering
+                  Dominated Hypervolume as {Klee's} Measure Problem},
+  booktitle = {Proceedings of the Second IASTED Conference on
+                  Computational Intelligence},
+  editor = {B. Kovalerchuk},
+  pages = {231--236},
+  publisher = {ACTA Press, Anaheim},
+  year = 2006
+}
+
+ +
+@incollection{BezLopStu2012:ants,
+  volume = 7461,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  year = 2012,
+  booktitle = {Swarm Intelligence, 8th International Conference, ANTS 2012},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Generation of Multi-Objective {ACO}
+                  Algorithms for the Biobjective Knapsack},
+  pages = {37--48},
+  doi = {10.1007/978-3-642-32650-9_4},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2012-008/}
+}
+
+ +
+@misc{BezLopStu2012:ants-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Generation of {MOACO} Algorithms for the Biobjective Bidimensional Knapsack Problem: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2012-008/}},
+  year = 2012
+}
+
+ +
+@misc{BezLopStu2013:evocop-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Analysis of Local Search for the Bi-objective Bidimensional Knapsack: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2012-016/}},
+  year = 2013
+}
+
+ +
+@misc{BezLopStu2013:lion-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Deconstructing Multi-Objective Evolutionary Algorithms: An Iterative Analysis on the Permutation Flowshop: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2013-010/}},
+  year = 2013
+}
+
+ +
+@incollection{BezLopStu2013evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  volume = 7832,
+  booktitle = {Proceedings of EvoCOP 2013 -- 13th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Martin Middendorf  and  Christian Blum },
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Analysis of Local Search for the Bi-objective
+                  Bidimensional Knapsack Problem},
+  pages = {85--96},
+  doi = {10.1007/978-3-642-37198-1_8}
+}
+
+ +
+@techreport{BezLopStu2014:automoeaTR,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Com\-ponent-Wise Design of Multi-Objective
+                  Evolutionary Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2014,
+  number = {TR/IRIDIA/2014-012},
+  month = aug
+}
+
+ +
+@incollection{BezLopStu2014:lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8426,
+  editor = { Panos M. Pardalos  and  Mauricio G. C. Resende  and Chrysafis Vogiatzis and Jose
+                  L. Walteros},
+  series = {Lecture Notes in Computer Science},
+  year = 2014,
+  booktitle = {Learning and Intelligent Optimization, 8th International Conference, LION 8},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Deconstructing Multi-Objective Evolutionary Algorithms: An
+                  Iterative Analysis on the Permutation Flowshop},
+  pages = {57--172},
+  doi = {10.1007/978-3-319-09584-4_16},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2013-010/}
+}
+
+ +
+@incollection{BezLopStu2014:ppsn,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Evolutionary Algorithms for
+                  Multi-Objective Combinatorial Optimization},
+  doi = {10.1007/978-3-319-10762-2_50},
+  pages = {508--517}
+}
+
+ +
+@misc{BezLopStu2014:ppsn-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Evolutionary Algorithms for Multi-Objective Combinatorial Optimization},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2014-007/}},
+  year = 2014
+}
+
+ +
+@misc{BezLopStu2015:supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Component-Wise Design of Multi-Objective Evolutionary Algorithms},
+  howpublished = {\url{https://github.com/iridia-ulb/automoea-tevc-2016}},
+  year = 2015
+}
+
+ +
+@misc{BezLopStu2015emoDEsupp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {To {DE} or Not to {DE}? {Multi}-objective Differential
+                  Evolution Revisited from a Component-Wise Perspective: {Supplementary} material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2015-001/}},
+  year = 2015
+}
+
+ +
+@incollection{BezLopStu2015emode,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {To {DE} or Not to {DE}? {Multi}-objective Differential
+                  Evolution Revisited from a Component-Wise Perspective},
+  pages = {48--63},
+  doi = {10.1007/978-3-319-15934-8_4}
+}
+
+ +
+@incollection{BezLopStu2015moead,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Comparing De\-com\-po\-sition-Based and Automatically
+                  Component-Wise Designed Multi-Objective Evolutionary
+                  Algorithms},
+  pages = {396--410},
+  doi = {10.1007/978-3-319-15934-8_27}
+}
+
+ +
+@misc{BezLopStu2016supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Large-Scale Experimental Evaluation of High-Performing Multi- and Many-Objective Evolutionary Algorithms},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2015-007/}},
+  year = 2017
+}
+
+ +
+@techreport{BezLopStu2017:techreport-005,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Large-Scale Experimental Evaluation of High-Performing
+                  Multi- and Many-Objective Evolutionary Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2017,
+  number = {TR/IRIDIA/2017-005},
+  month = feb
+}
+
+ +
+@incollection{BezLopStu2017emo,
+  editor = {Heike Trautmann and G{\"{u}}nter Rudolph and Kathrin Klamroth
+                  and Oliver Sch{\"{u}}tze and Margaret M. Wiecek and Yaochu
+                  Jin and Christian Grimme},
+  year = 2017,
+  volume = 10173,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2017},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Empirical Assessment of the Properties of Inverted
+                  Generational Distance Indicators on Multi- and Many-objective
+                  Optimization},
+  pages = {31--45},
+  doi = {10.1007/978-3-319-54157-0_3}
+}
+
+ +
+@misc{BezLopStu2019ec-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Designing State-of-the-Art Multi- and
+                  Many-Objective Evolutionary Algorithms: Supplementary
+                  material},
+  howpublished = {\url{https://github.com/iridia-ulb/automoea-ecj-2020}},
+  year = 2019,
+  alias = {BezLopStu2017tec-supp}
+}
+
+ +
+@inproceedings{WanSunJin2019multi,
+  year = 2019,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2019 Congress on Evolutionary Computation (CEC 2019)},
+  key = {IEEE CEC},
+  title = {A Multi-indicator based Selection Strategy for Evolutionary
+                  Many-objective Optimization},
+  author = {Wang, Hao and Sun, Chaoli and  Yaochu Jin  and Qin, Shufen and
+                  Yu, Haibo},
+  pages = {2042--2049},
+  annote = {unbounded archive}
+}
+
+ +
+@incollection{BezLopStu2019gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Archiver Effects on the Performance of State-of-the-art
+                  Multi- and Many-objective Evolutionary Algorithms},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2019-004/},
+  doi = {10.1145/3321707.3321789}
+}
+
+ +
+@misc{BezLopStu2019gecco-supp,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Archiver Effects on the Performance of State-of-the-art Multi- and Many-objective Evolutionary Algorithms: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2019-004/}},
+  year = 2019
+}
+
+ +
+@incollection{BezLopStu2020chapter,
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  editor = { Thomas Bartz-Beielstein  and Bogdan Filipi{\v c} and  P. Koro{\v s}ec  and  Talbi, El-Ghazali },
+  year = 2020,
+  booktitle = {High-Performance Simulation-Based Optimization},
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Configuration of Multi-objective Optimizers and
+                  Multi-objective Configuration},
+  pages = {69--92},
+  doi = {10.1007/978-3-030-18764-4_4},
+  abstract = {Heuristic optimizers are an important tool in academia and industry, and their performance-optimizing configuration requires a significant amount of expertise. As the proper configuration of algorithms is a crucial aspect in the engineering of heuristic algorithms, a significant research effort has been dedicated over the last years towards moving this step to the computer and, thus, make it automatic. These research efforts go way beyond tuning only numerical parameters of already fully defined algorithms, but exploit automatic configuration as a means for automatic algorithm design. In this chapter, we review two main aspects where the research on automatic configuration and multi-objective optimization intersect. The first is the automatic configuration of multi-objective optimizers, where we discuss means and specific approaches. In addition, we detail a case study that shows how these approaches can be used to design new, high-performing multi-objective evolutionary algorithms. The second aspect is the research on multi-objective configuration, that is, the possibility of using multiple performance metrics for the evaluation of algorithm configurations. We highlight some few examples in this direction.}
+}
+
+ +
+@phdthesis{Bezerra2016PhD,
+  author = { Leonardo C. T. Bezerra },
+  title = {A component-wise approach to multi-objective evolutionary
+                  algorithms: from flexible frameworks to automatic design},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2016,
+  annote = {Supervised by  Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez }
+}
+
+ +
+@incollection{BiaGamDor02:ppsn,
+  anote = {IC.34},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Juan-Juli{\'a}n Merelo  and others},
+  aeditor = { Juan-Juli{\'a}n Merelo  and P. Adamidis and   Hans-Georg Beyer  and J.-L. Fern\'{a}ndez-Villacanas and  Hans-Paul Schwefel },
+  volume = 2439,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VII}},
+  author = { Leonora Bianchi  and  L. M. Gambardella  and  Marco Dorigo },
+  title = {An Ant Colony Optimization Approach to the
+                  Probabilistic Traveling Salesman Problem},
+  pages = {883--892}
+}
+
+ +
+@inproceedings{Bie14:sat,
+  publisher = {University of Helsinki},
+  series = {Science Series of Publications B},
+  volume = {B-2014-2},
+  year = 2014,
+  editor = {A. Belov and D. Diepold and M. Heule and M. J\"{a}rvisalo},
+  booktitle = {Proceedings of SAT Competition 2014: Solver and Benchmark Descriptions},
+  title = {Yet another Local Search Solver and {Lingeling} and Friends Entering the {SAT} Competition 2014},
+  author = {Armin Biere},
+  pages = {39--40}
+}
+
+ +
+@incollection{BieBozEim2020dynaac,
+  publisher = {IOS Press},
+  editor = {Giuseppe De Giacomo and Alejandro Catala and Bistra Dilkina
+                  and Michela Milano and Senén Barro and Alberto Bugarín and
+                  Jérôme Lang},
+  series = {Frontiers in Artificial Intelligence and Applications},
+  volume = 325,
+  year = 2020,
+  booktitle = {Proceedings of the 24th European Conference on Artificial Intelligence (ECAI)},
+  author = { Biedenkapp, Andr{\'e}  and Bozkurt, H. Furkan and Eimer, Theresa and  Frank Hutter  and  Marius Thomas Lindauer },
+  title = {Dynamic Algorithm Configuration: Foundation of a New
+                  Meta-Algorithmic Framework},
+  epub = {https://ecai2020.eu/papers/1237_paper.pdf},
+  pages = {427--434}
+}
+
+ +
+@incollection{BieLinEggFraFawHoo2017,
+  publisher = {{AAAI} Press},
+  month = feb,
+  year = 2017,
+  editor = {Satinder P. Singh and Shaul Markovitch},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  author = { Biedenkapp, Andr{\'e}  and  Marius Thomas Lindauer  and  Katharina Eggensperger  and  Frank Hutter  and  Chris Fawcett  and  Holger H. Hoos },
+  title = {Efficient Parameter Importance Analysis via Ablation with
+                  Surrogates},
+  doi = {10.1609/aaai.v31i1.10657}
+}
+
+ +
+@incollection{BieMarLinHut2018lion,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 11353,
+  editor = { Roberto Battiti  and Mauro Brunato and Ilias Kotsireas and  Panos M. Pardalos },
+  year = 2018,
+  booktitle = {Learning and Intelligent Optimization, 12th International
+                  Conference, LION 12},
+  author = { Biedenkapp, Andr{\'e}  and Marben, Joshua and  Marius Thomas Lindauer  and  Frank Hutter },
+  title = {{CAVE}: Configuration assessment, visualization and
+                  evaluation},
+  pages = {115--130},
+  doi = {10.1007/978-3-030-05348-2_10}
+}
+
+ +
+@incollection{BilPar1995:aisb,
+  booktitle = {Evolutionary Computing, AISB Workshop},
+  address = { Berlin, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 993,
+  year = 1995,
+  publisher = {Springer},
+  editor = {T. C. Fogarty},
+  title = {The Ant Colony Metaphor for Searching Continuous Design
+                  Spaces},
+  author = {George Bilchev and Ian C. Parmee},
+  pages = {25--39},
+  doi = {10.1007/3-540-60469-3_22},
+  alias = {BilPar1995}
+}
+
+ +
+@incollection{BirBalDor06,
+  address = { New York, NY},
+  series = {Operations Research/Computer Science Interfaces Series},
+  volume = 39,
+  editor = {K. F. Doerner and M. Gendreau and P. Greistorfer and
+                  W. J. Gutjahr and R. F. Hartl and M. Reimann},
+  year = 2006,
+  publisher = {Springer},
+  booktitle = {Metaheuristics -- Progress in Complex Systems Optimization},
+  author = { Mauro Birattari  and   Prasanna Balaprakash  and  Marco Dorigo },
+  title = {The {ACO/F-RACE} algorithm for combinatorial optimization
+                  under uncertainty},
+  pages = {189--203}
+}
+
+ +
+@incollection{BirChiSaeStu2011,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = {T. Berthold and A. M. Gleixner and S. Heinz and T. Koch},
+  year = 2011,
+  booktitle = {Integration of AI and OR Techniques in Contraint Programming
+                  for Combinatorial Optimization Problems},
+  author = { Mauro Birattari  and  Marco Chiarandini  and  Marco Saerens  and  Thomas St{\"u}tzle },
+  title = {Learning Graphical Models for Algorithm Configuration}
+}
+
+ +
+@incollection{BirDicDor02:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { Mauro Birattari  and  Gianni A. {Di Caro}  and  Marco Dorigo },
+  title = {Toward the formal foundation of Ant Programming},
+  pages = {188--201}
+}
+
+ +
+@book{BirKleLop2009nltk,
+  title = {Natural language processing with {Python}: analyzing text with
+                  the natural language toolkit},
+  author = {Bird, Steven and Klein, Ewan and Loper, Edward},
+  year = 2009,
+  publisher = {O'Reilly Media, Inc.}
+}
+
+ +
+@incollection{BirStuPaqVar02:gecco,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = { Langdon, William B.  and others},
+  year = 2002,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2002},
+  author = { Mauro Birattari  and  Thomas St{\"u}tzle  and  Lu{\'i}s Paquete  and  Klaus Varrentrapp },
+  title = {A Racing Algorithm for Configuring Metaheuristics},
+  pages = {11--18},
+  keywords = {F-race},
+  epub = {https://dl.acm.org/doi/pdf/10.5555/2955491.2955494}
+}
+
+ +
+@incollection{BirYuaBal2010:emaoa,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = { Mauro Birattari  and  Zhi Yuan  and   Prasanna Balaprakash  and  Thomas St{\"u}tzle },
+  title = {{F}-Race and Iterated {F}-Race: An Overview},
+  pages = {311--336},
+  keywords = {F-race, iterated F-race, irace, tuning},
+  doi = {10.1007/978-3-642-02538-9_13}
+}
+
+ +
+@inproceedings{BirYuaBalStu2010:mic,
+  address = {Hamburg, Germany},
+  publisher = {University of Hamburg},
+  editor = {M. Caserta and  Stefan Vo{\ss} },
+  year = 2010,
+  booktitle = {Proceedings of MIC 2009, the 8th Metaheuristics International Conference},
+  author = { Mauro Birattari  and  Zhi Yuan  and   Prasanna Balaprakash  and  Thomas St{\"u}tzle },
+  title = {Parameter Adaptation in Ant Colony Optimization},
+  alias = {adaptive_techreport}
+}
+
+ +
+@book{Birattari09tuning,
+  title = {Tuning Metaheuristics: A Machine Learning
+                  Perspective},
+  doi = {10.1007/978-3-642-00483-4},
+  author = { Mauro Birattari },
+  year = 2009,
+  volume = 197,
+  series = {Studies in Computational Intelligence},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  alias = {Bir09:book},
+  annote = {Based on the PhD thesis~\cite{Birattari2004PhD}}
+}
+
+ +
+@phdthesis{Birattari2004PhD,
+  author = { Mauro Birattari },
+  title = {The Problem of Tuning Metaheuristics as Seen from a
+                  Machine Learning Perspective},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2004,
+  annote = {Supervised by Marco Dorigo}
+}
+
+ +
+@inproceedings{BisIzzYam2010:pagmo,
+  title = {A Global Optimisation Toolbox for Massively Parallel
+                  Engineering Optimisation},
+  author = {Biscani, Francesco and Izzo, Dario and Yam, Chit Hong},
+  booktitle = {Astrodynamics Tools and Techniques (ICATT 2010), 4th
+                  International Conference on},
+  year = 2010,
+  url = {http://arxiv.org/abs/1004.3824},
+  keywords = {PaGMO}
+}
+
+ +
+@incollection{BisMerTraPre12:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2012,
+  editor = {Terence Soule and Jason H. Moore},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  author = { Bernd Bischl  and  Olaf Mersmann  and  Heike Trautmann  and  Mike Preuss },
+  title = {Algorithm Selection Based on Exploratory Landscape Analysis and Cost-sensitive Learning},
+  pages = {313--320},
+  keywords = {continuous optimization, landscape analysis, algorithm selection}
+}
+
+ +
+@book{Bishop2006,
+  title = {Pattern recognition and machine learning},
+  author = {Bishop, Christopher M.},
+  year = 2006,
+  publisher = {Springer}
+}
+
+ +
+@inproceedings{BiyMarAli2019acc,
+  year = 2019,
+  publisher = {{IEEE}},
+  booktitle = {2019 American Control Conference ({ACC})},
+  key = {ACC2019},
+  author = {Erdem B{\i }y{\i }k and Jonathan Margoliash and Shahrouz Ryan Alimo
+                  and Dorsa Sadigh},
+  title = {Efficient and Safe Exploration in Deterministic {Markov}
+                  Decision Processes with Unknown Transition Models},
+  pages = {1792--1799},
+  doi = {10.23919/ACC.2019.8815276}
+}
+
+ +
+@incollection{BleBlu04:disjoint,
+  aeditor = { G{\"u}nther R. Raidl  and S. Cagnoni and  J{\"u}rgen Branke  and D. W. Corne and
+                  R. Drechsler and Y. Jin and C. G. Johnson and  Penousal Machado  and E. Marchiori and R. Rothlauf and G. D. Smith and
+                  G. Squillero},
+  booktitle = {Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 3005,
+  year = 2004,
+  publisher = {Springer},
+  editor = { G{\"u}nther R. Raidl  and others},
+  author = { Mar{\'i}a J. Blesa  and  Christian Blum },
+  title = {Ant Colony Optimization for the Maximum
+                  Edge-Disjoint Paths Problem},
+  pages = {160--169}
+}
+
+ +
+@inproceedings{BliMcDPer2006emnlp,
+  year = 2006,
+  editor = {Jurafsky, Dan and Gaussier, Eric},
+  series = {Empirical Methods in Natural Language Processing},
+  booktitle = {Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP2006},
+  title = {Domain adaptation with structural correspondence learning},
+  author = {Blitzer, John and McDonald, Ryan and Pereira, Fernando},
+  pages = {120--128}
+}
+
+ +
+@incollection{BloHooJouKesTra2016:lion,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 10079,
+  editor = {Paola Festa and  Meinolf Sellmann  and  Joaquin Vanschoren },
+  series = {Lecture Notes in Computer Science},
+  year = 2016,
+  booktitle = {Learning and Intelligent Optimization, 10th International
+                  Conference, LION 10},
+  author = { Aymeric Blot  and  Holger H. Hoos  and  Laetitia Jourdan  and  Marie-El{\'e}onore Kessaci-Marmion  and  Heike Trautmann },
+  title = {{MO-ParamILS}: A Multi-objective Automatic Algorithm Configuration Framework},
+  pages = {32--47}
+}
+
+ +
+@incollection{BloJouKess2017gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  author = { Aymeric Blot  and  Laetitia Jourdan  and  Marie-El{\'e}onore Kessaci-Marmion },
+  title = {Automatic design of multi-objective local search algorithms:
+                  case study on a bi-objective permutation flowshop scheduling
+                  problem},
+  pages = {227--234},
+  doi = {10.1145/3071178.3071323}
+}
+
+ +
+@incollection{BloLopKesJou2018ppsn,
+  volume = 11101,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Aymeric Blot  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Marie-El{\'e}onore Kessaci-Marmion  and  Laetitia Jourdan },
+  title = {New Initialisation Techniques for Multi-Objective Local
+                  Search: Application to the Bi-objective Permutation Flowshop},
+  doi = {10.1007/978-3-319-99253-2_26},
+  pages = {323--334}
+}
+
+ +
+@incollection{BloPerJouKesHoo2017emo,
+  editor = {Heike Trautmann and G{\"{u}}nter Rudolph and Kathrin Klamroth
+                  and Oliver Sch{\"{u}}tze and Margaret M. Wiecek and Yaochu
+                  Jin and Christian Grimme},
+  year = 2017,
+  volume = 10173,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2017},
+  author = { Aymeric Blot  and Alexis Pernet and  Laetitia Jourdan  and  Marie-El{\'e}onore Kessaci-Marmion  and  Holger H. Hoos },
+  title = {Automatically Configuring Multi-objective Local Search Using
+                  Multi-objective Optimisation},
+  pages = {61--76}
+}
+
+ +
+@incollection{BluBauPer06:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  author = { Christian Blum  and  J. Bautista  and  J. Pereira },
+  title = {{Beam-ACO} applied to assembly line balancing},
+  pages = {96--107},
+  doi = {10.1007/11839088_9}
+}
+
+ +
+@techreport{BluBleLop08:lcs,
+  author = { Christian Blum  and  Mar{\'i}a J. Blesa  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Beam Search for the Longest Common Subsequence
+                  Problem},
+  institution = {Department LSI, Universitat Polit{\`e}cnica de Catalunya},
+  year = 2008,
+  number = {LSI-08-29},
+  note = {Published in Computers \& Operations Research~\cite{BluBleLop09-BeamSearch-LCS}}
+}
+
+ +
+@incollection{BluCotFerGal07:evocop,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4446,
+  year = 2007,
+  editor = { Carlos Cotta  and  others },
+  booktitle = {Proceedings of EvoCOP 2007 -- Seventh European Conference on
+                  Evolutionary Computation in Combinatorial Optimisation},
+  author = { Christian Blum  and  Carlos Cotta  and  Antonio J. Fern{\'a}ndez  and  J. E. Gallardo },
+  title = {A probabilistic beam search algorithm for the
+                  shortest common supersequence problem},
+  pages = {36--47}
+}
+
+ +
+@incollection{BluLop2011ieh,
+  author = { Christian Blum  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  booktitle = {The Industrial Electronics Handbook: Intelligent Systems},
+  title = {Ant Colony Optimization},
+  publisher = {CRC Press},
+  year = 2011,
+  edition = {2nd},
+  isbn = 9781439802830,
+  url = {http://www.crcpress.com/product/isbn/9781439802830},
+  annnote = {http://www.eng.auburn.edu/~wilambm/ieh/}
+}
+
+ +
+@incollection{BluMas2007hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4771,
+  editor = { Thomas Bartz-Beielstein  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Boris Naujoks  and  Andrea Roli  and  G{\"u}nther Rudolph  and  M. Sampels },
+  year = 2007,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Christian Blum  and M. Mastrolilli},
+  title = {Using Branch {\&} Bound Concepts in
+                  Construction-Based Metaheuristics: {Exploiting} the
+                  Dual Problem Knowledge},
+  pages = {123--139}
+}
+
+ +
+@book{BluMer08:si-book,
+  title = {Swarm Intelligence--Introduction and Applications},
+  year = 2008,
+  editor = { Christian Blum  and  D. Merkle },
+  series = {Natural Computing Series},
+  publisher = {Springer Verlag, Berlin, Germany}
+}
+
+ +
+@book{BluRai2016:book,
+  author = { Christian Blum  and  G{\"u}nther R. Raidl },
+  title = {Hybrid Metaheuristics---Powerful Tools for Optimization},
+  publisher = {Springer},
+  year = 2016,
+  series = {Artificial Intelligence: Foundations, Theory, and Algorithms},
+  address = { Berlin, Germany}
+}
+
+ +
+@incollection{BluRol2008hybrid,
+  alias = {BluEtAl08:hm-book},
+  series = {Studies in Computational Intelligence},
+  volume = 114,
+  year = 2008,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Mar{\'i}a J. Blesa  and  Andrea Roli  and  M. Sampels },
+  booktitle = {Hybrid Metaheuristics: An emergent approach for optimization},
+  title = {Hybrid metaheuristics: an introduction},
+  author = { Christian Blum  and  Andrea Roli },
+  pages = {1--30}
+}
+
+ +
+@incollection{BluYab06:hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4030,
+  editor = {F. Almeida and others},
+  aeditor = {F. Almeida and M. Blesa and C. Blum and J. M. Moreno
+                  and M. P{\'e}rez and A. Roli and  M. Sampels },
+  year = 2006,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Christian Blum  and  M. {Y{\'a}bar Vall{\`e}s} },
+  title = {Multi-level ant colony optimization for {DNA}
+                  sequencing by hybridization},
+  pages = {94--109},
+  doi = {10.1007/11890584}
+}
+
+ +
+@phdthesis{Boese1996,
+  author = {K. D. Boese},
+  title = {Models for Iterative Global Optimization},
+  school = {University of California, Computer Science Department,
+Los Angeles, CA},
+  year = 1996
+}
+
+ +
+@book{Bollobas2001,
+  author = {B{\'e}la Bollob{\'a}s},
+  title = {Random Graphs},
+  publisher = {Cambridge University Press},
+  address = { New York, NY},
+  year = 2001,
+  edition = {2nd}
+}
+
+ +
+@book{BooRumJac2005,
+  author = {Grady Booch and James E. Rumbaugh and Ivar Jacobson},
+  title = {The Unified Modeling Language User Guide},
+  publisher = {Addison-Wesley},
+  year = 2005,
+  edition = {2nd}
+}
+
+ +
+@techreport{Bor1998,
+  author = {Borges, P. C. and  Michael Pilegaard Hansen },
+  title = {A basis for future successes in multiobjective
+                  combinatorial optimization},
+  year = 1998,
+  institution = {Institute of Mathematical Modelling, Technical
+                  University of Denmark},
+  number = {IMM-REP-1998-8},
+  address = {Lyngby, Denmark},
+  alias = {borges1998}
+}
+
+ +
+@book{BorEly1998online,
+  author = {Borodin, Allan and El-Yaniv, Ran},
+  title = {Online computation and competitive analysis},
+  year = 1998,
+  isbn = {0-521-56392-5},
+  publisher = {Cambridge University Press},
+  address = { New York, NY}
+}
+
+ +
+@book{BorHedHigRot2009metanalysis,
+  title = {Introduction to Meta-Analysis},
+  author = {Michael Borenstein and Larry V. Hedges and Julian P. T. Higgins and Hannah R. Rothstein},
+  year = 2009,
+  publisher = {Wiley}
+}
+
+ +
+@inproceedings{BosGuyVap1992,
+  year = 1992,
+  publisher = {ACM Press},
+  booktitle = {COLT'92},
+  editor = {David Haussler},
+  author = {Bernhard E. Boser and Isabelle Guyon and Vladimir Vapnik},
+  title = {A Training Algorithm for Optimal Margin Classifiers},
+  pages = {144--152},
+  doi = {10.1145/130385.130401},
+  annote = {Proposed SVM}
+}
+
+ +
+@incollection{BosKerNeu2019,
+  publisher = {{ACM}},
+  editor = { Tobias Friedrich  and  Carola Doerr  and Arnold, Dirk V.},
+  year = 2019,
+  booktitle = {Proceedings of the 15th {ACM}/{SIGEVO} Conference on Foundations of Genetic Algorithms},
+  author = { Jakob Bossek  and  Pascal Kerschke  and Neumann, Aneta and  Markus Wagner  and  Frank Neumann  and  Heike Trautmann },
+  title = {Evolving Diverse TSP Instances by Means of Novel and Creative Mutation Operators},
+  pages = {58--71}
+}
+
+ +
+@inproceedings{Boulos01,
+  author = { Paul F. Boulos  and  Chun Hou Orr  and  Werner de Schaetzen  and  J. G. Chatila  and  Michael Moore  and  Paul Hsiung  and  Devan Thomas },
+  title = {Optimal pump operation of water distribution systems
+                  using genetic algorithms},
+  booktitle = {AWWA Distribution System Symp.},
+  year = 2001,
+  address = {Denver, USA},
+  publisher = {American Water Works Association}
+}
+
+ +
+@incollection{Bow1976,
+  author = {Bowman, V. and Joseph, Jr.},
+  title = {On the Relationship of the {Tchebycheff} Norm and the Efficient
+                   Frontier of Multiple-Criteria Objectives},
+  year = 1976,
+  booktitle = {Multiple Criteria Decision Making},
+  volume = 130,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  pages = {76--86},
+  editor = {Thiriez, Herv\'e and Zionts, Stanley},
+  doi = {10.1007/978-3-642-87563-2_5},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg}
+}
+
+ +
+@book{BoxDra2007rsm,
+  title = {Response surfaces, mixtures, and ridge analyses},
+  author = {Box, George E. P. and Draper, Norman R.},
+  year = 2007,
+  publisher = {John Wiley \& Sons}
+}
+
+ +
+@book{BoxHunHun1978stat,
+  title = {Statistics for experimenters: an introduction to design, data
+                  analysis, and model building},
+  author = {Box, G. E. P. and Hunter, W. G. and Hunter, J. S.},
+  year = 1978,
+  publisher = {John Wiley \& Sons},
+  address = { New York, NY}
+}
+
+ +
+@incollection{Bra88:lnpam,
+  author = {A. Brandt},
+  title = {Multilevel Computations: {Review} and Recent Developments},
+  booktitle = {Multigrid Methods: Theory, Applications, and Supercomputing, Proceedings of the 3rd Copper Mountain Conference on Multigrid Methods},
+  pages = {35--62},
+  year = 1988,
+  editor = {S. F. McCormick},
+  volume = 110,
+  series = {Lecture Notes in Pure and Applied Mathematics},
+  publisher = {Marcel Dekker},
+  address = { New York, NY}
+}
+
+ +
+@inproceedings{BraBarWhiHubHin2007wfg,
+  author = {L. Bradstreet and L. Barone and L. While and S. Huband and
+                  P. Hingston},
+  booktitle = {{IEEE} Symposium on Computational Intelligence in
+                  Multicriteria Decision-Making, {IEEE} {MCDM}},
+  title = {Use of the {WFG} Toolkit and {PISA} for Comparison of
+                  {MOEAs}},
+  year = 2007,
+  pages = {382--389}
+}
+
+ +
+@incollection{BarOjaGar2018ppsn,
+  volume = 11101,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Barba-Gonz{\'a}lez, Crist{\'o}bal  and Vesa Ojalehto and  Jos{\'e} Garc{\'i}a-Nieto  and  Nebro, Antonio J.  and  Kaisa Miettinen  and Jos{\'{e}} F. Aldana-Montes},
+  title = {Artificial Decision Maker Driven by {PSO}: An Approach for
+                  Testing Reference Point Based Interactive Methods},
+  doi = {10.1007/978-3-319-99253-2_22},
+  pages = {274--285},
+  keywords = {machine decision-maker}
+}
+
+ +
+@incollection{BraCorGre2015dagstuhl,
+  keywords = {multiple criteria decision making, evolutionary
+                  multiobjective optimization},
+  doi = {10.4230/DagRep.5.1.96},
+  volume = {5(1)},
+  year = 2015,
+  series = {Dagstuhl Reports},
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  booktitle = {Understanding Complexity in Multiobjective Optimization
+                  (Dagstuhl Seminar 15031)},
+  editor = { Salvatore Greco  and  Kathrin Klamroth  and  Joshua D. Knowles  and  G{\"u}nther Rudolph },
+  author = { J{\"u}rgen Branke  and  Salvatore Corrente  and  Salvatore Greco  and  Kadzi{\'n}ski, Mi{\l}osz   and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Vincent Mousseau  and  Mauro Munerato  and  Roman S{\l}owi{\'n}ski },
+  title = {Behavior-Realistic Artificial Decision-Makers to Test
+                  Preference-Based Multi-objective Optimization Method
+                  ({Working} {Group} ``{Machine} {Decision}-{Making}'')},
+  pages = {110--116}
+}
+
+ +
+@incollection{BraDeb2055integrating,
+  author = { J{\"u}rgen Branke  and  Kalyanmoy Deb },
+  title = {Integrating User Preferences into Evolutionary
+                  Multi-Objective Optimization},
+  booktitle = {Knowledge Incorporation in Evolutionary Computation},
+  publisher = {Springer},
+  year = 2005,
+  editor = { Yaochu Jin },
+  pages = {461--477},
+  address = {Berlin\slash Heidelberg},
+  abstract = {Many real-world optimization problems involve multiple,
+                  typically conflicting objectives. Often, it is very difficult
+                  to weigh the different criteria exactly before alternatives
+                  are known. Evolutionary multi-objective optimization usually
+                  solves this predicament by searching for the whole
+                  Pareto-optimal front of solutions. However, often the user
+                  has at least a vague idea about what kind of solutions might
+                  be preferred. In this chapter, we argue that such knowledge
+                  should be used to focus the search on the most interesting
+                  (from a user's perspective) areas of the Paretooptimal
+                  front. To this end, we present and compare two methods which
+                  allow to integrate vague user preferences into evolutionary
+                  multi-objective algorithms. As we show, such methods may
+                  speed up the search and yield a more fine-grained selection
+                  of alternatives in the most relevant parts of the
+                  Pareto-optimal front.},
+  doi = {10.1007/978-3-540-44511-1_21}
+}
+
+ +
+@incollection{BraFerLuq2016lncs,
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  year = 2016,
+  booktitle = {Smart Cities (Smart-CT 2016)},
+  editor = { Alba, Enrique  and  Chicano, Francisco  and  Gabriel J. Luque },
+  author = {Bravo, Yesnier and  Javier Ferrer  and  Gabriel J. Luque  and  Alba, Enrique },
+  title = {Smart Mobility by Optimizing the Traffic Lights: A New Tool
+                  for Traffic Control Centers},
+  pages = {147--156},
+  abstract = {Urban traffic planning is a fertile area of Smart Cities to
+                  improve efficiency, environmental care, and safety, since the
+                  traffic jams and congestion are one of the biggest sources of
+                  pollution and noise. Traffic lights play an important role in
+                  solving these problems since they control the flow of the
+                  vehicular network at the city. However, the increasing number
+                  of vehicles makes necessary to go from a local control at one
+                  single intersection to a holistic approach considering a
+                  large urban area, only possible using advanced computational
+                  resources and techniques. Here we propose HITUL, a system
+                  that supports the decisions of the traffic control managers
+                  in a large urban area. HITUL takes the real traffic
+                  conditions and compute optimal traffic lights plans using
+                  bio-inspired techniques and micro-simulations. We compare our
+                  system against plans provided by experts. Our solutions not
+                  only enable continuous traffic flows but reduce the
+                  pollution. A case study of M{\'a}laga city allows us to
+                  validate the approach and show its benefits for other cities
+                  as well.},
+  doi = {10.1007/978-3-319-39595-1_15},
+  keywords = {Multi-objective optimization, Smart mobility, Traffic lights
+                  planning}
+}
+
+ +
+@book{BraMar2002:book,
+  author = { Jean-Pierre Brans  and  Bertrand Mareschal },
+  title = {{PROMETHEE-GAIA}. Une m{\'e}thode d'aide {\`a} la d{\'e}cision en pr{\'e}sence de crit{\`e}res multiples},
+  year = 2002,
+  isbn = {2-7298-1253-9},
+  publisher = {Editions Ellipses},
+  address = { Paris, France}
+}
+
+ +
+@incollection{BraMar2005:mcda,
+  editor = { Jos{\'e} Rui Figueira  and  Salvatore Greco  and  Matthias Ehrgott },
+  year = 2005,
+  publisher = {Springer},
+  booktitle = {Multiple Criteria Decision Analysis, State of the
+                  Art Surveys},
+  author = { Jean-Pierre Brans  and  Bertrand Mareschal },
+  title = {{PROMETHEE} Methods},
+  chapter = 5,
+  pages = {163--195}
+}
+
+ +
+@incollection{BraSchSch2001gecco,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = {Erik D. Goodman},
+  year = 2001,
+  booktitle = {Proceedings of the 3rd Annual Conference on Genetic and
+                  Evolutionary Computation, GECCO 2001},
+  author = { J{\"u}rgen Branke  and C. Schmidt and H. Schmeck},
+  title = {Efficient fitness estimation in noisy environments},
+  pages = {243--250}
+}
+
+ +
+@techreport{BranCorrGreSlow2014,
+  author = { J{\"u}rgen Branke  and  Salvatore Corrente  and  Salvatore Greco  and  Roman S{\l}owi{\'n}ski  and Zielniewicz, P.},
+  title = {Using {Choquet} integral as preference model in interactive
+                  evolutionary multiobjective optimization},
+  institution = {WBS, University of Warwick},
+  year = 2014
+}
+
+ +
+@incollection{BranElo2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { J{\"u}rgen Branke  and  Jawad Elomari },
+  title = {Simultaneous tuning of metaheuristic parameters for
+                  various computing budgets},
+  pages = {263--264},
+  doi = {10.1145/2001858.2002006},
+  keywords = {meta-optimization, offline parameter optimization}
+}
+
+ +
+@incollection{BranElo2013lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7997,
+  editor = { Panos M. Pardalos  and G. Nicosia},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  booktitle = {Learning and Intelligent Optimization, 7th
+                  International Conference, LION 7},
+  author = { J{\"u}rgen Branke  and  Jawad Elomari },
+  title = {Racing with a Fixed Budget and a Self-Adaptive
+                  Significance Level}
+}
+
+ +
+@book{BreFriSto1984trees,
+  title = {Classification and regression trees},
+  author = {Breiman, Leo and Friedman, Jerome and Stone, Charles J. and
+                  Olshen, Richard A.},
+  year = 1984,
+  publisher = {CRC Press}
+}
+
+ +
+@incollection{BreSch2011ea,
+  booktitle = {Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7401,
+  year = 2012,
+  publisher = {Springer},
+  editor = { Jin-Kao Hao  and Legrand, Pierrick and Collet, Pierre and
+                  Monmarch{\'e}, Nicolas and Lutton, Evelyne and Schoenauer,
+                  Marc},
+  author = {M\'aty\'as Brendel and  Marc Schoenauer },
+  title = {Learn-and-Optimize: A Parameter Tuning Framework for Evolutionary {AI} Planning},
+  pages = {145--155},
+  doi = {10.1007/978-3-642-35533-2_13}
+}
+
+ +
+@incollection{BreSch2011gecco,
+  year = 2011,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2011},
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  author = {M{\'a}ty{\'a}s Brendel and  Marc Schoenauer },
+  title = {Instance-based Parameter Tuning for Evolutionary {AI} Planning},
+  pages = {591--598},
+  doi = {10.1145/2001858.2002053}
+}
+
+ +
+@incollection{BriFri2009emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  title = {Approximating the Least Hypervolume Contributor:
+                  {NP}-Hard in General, But Fast in Practice},
+  pages = {6--20},
+  annote = {Extended version published in \cite{BriFri2012tcs}}
+}
+
+ +
+@incollection{BriFri2010gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  title = {The Maximum Hypervolume Set Yields Near-optimal
+                  Approximation},
+  pages = {511--518},
+  annote = {Proved that hypervolume approximates the additive
+                  $\epsilon$-indicator, converging quickly as $N$ increases,
+                  that is, sets that maximize hypervolume are near optimal on
+                  additive $\epsilon$ too, with the gap diminishing as quickly
+                  as O(1/N).}
+}
+
+ +
+@incollection{BriFri2010ppsn,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  title = {Tight bounds for the approximation ratio of the hypervolume
+                  indicator},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  pages = {607--616}
+}
+
+ +
+@incollection{BriFri2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  title = {Convergence of Hypervolume-Based Archiving Algorithms~{I}:
+                  Effectiveness},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  pages = {745--752},
+  doi = {10.1145/2001576.2001678},
+  annote = {Extended version published as \cite{BriFri2014convergence}}
+}
+
+ +
+@incollection{BriFri2012gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2012,
+  editor = {Terence Soule and Jason H. Moore},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  title = {Convergence of Hypervolume-Based Archiving Algorithms~{II}:
+                  Competitiveness},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  pages = {457--464},
+  doi = {10.1145/2330163.2330229},
+  annote = {Extended version published as \cite{BriFri2014convergence}}
+}
+
+ +
+@incollection{BriFriKli2014generic,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  title = {Generic postprocessing via subset selection for hypervolume
+                  and epsilon-indicator},
+  author = { Karl Bringmann  and  Tobias Friedrich  and Patrick Klitzke},
+  pages = {518--527}
+}
+
+ +
+@incollection{BriFriKli2014subset,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2014,
+  editor = {Christian Igel and Dirk V. Arnold},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2014},
+  title = {Two-dimensional subset selection for hypervolume and
+                  epsilon-indicator},
+  author = { Karl Bringmann  and  Tobias Friedrich  and Patrick Klitzke},
+  doi = {10.1145/2576768.2598276}
+}
+
+ +
+@inproceedings{BriPoz2012cec,
+  year = 2012,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2012 Congress on Evolutionary Computation (CEC 2012)},
+  key = {IEEE CEC},
+  title = {Using archiving methods to control convergence and diversity
+                  for many-objective problems in particle swarm optimization},
+  author = {Britto, Andre and Pozo, Aurora},
+  pages = {1--8},
+  doi = {10.1109/CEC.2012.6256149}
+}
+
+ +
+@incollection{BrigFri2009foga,
+  isbn = {978-1-60558-414-0},
+  publisher = {{ACM}},
+  editor = {Ivan I. Garibay and Thomas Jansen and R. Paul Wiegand and
+                  Annie S. Wu},
+  year = 2009,
+  booktitle = {Proceedings of the Tenth ACM SIGEVO Workshop on Foundations of Genetic Algorithms (FOGA)},
+  author = { Karl Bringmann  and  Tobias Friedrich },
+  title = {Don't be greedy when calculating hypervolume contributions},
+  pages = {103--112},
+  alias = {bringmann2009don},
+  annote = {Extended version published in \cite{BriFri2010eff}}
+}
+
+ +
+@inproceedings{BrinFriNeuWag2011,
+  publisher = {IJCAI/AAAI Press, Menlo Park, CA},
+  editor = {Toby Walsh},
+  year = 2011,
+  booktitle = {Proceedings of  the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11)},
+  author = { Karl Bringmann  and  Tobias Friedrich  and  Frank Neumann  and  Markus Wagner },
+  title = {Approximation-guided Evolutionary Multi-objective
+                  Optimization},
+  pages = {1198--1203}
+}
+
+ +
+@incollection{Bro2015emo,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  title = {A Bug in the Multiobjective Optimizer {IBEA}: Salutary
+                  Lessons for Code Release and a Performance Re-Assessment},
+  author = { Dimo Brockhoff },
+  doi = {10.1007/978-3-319-15934-8_13},
+  pages = {187--201}
+}
+
+ +
+@incollection{BroCalLop2018dagstuhl,
+  keywords = {multiple criteria decision making, evolutionary
+                  multiobjective optimization},
+  doi = {10.4230/DagRep.8.1.33},
+  volume = {8(1)},
+  year = 2018,
+  series = {Dagstuhl Reports},
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  booktitle = {Personalized Multiobjective Optimization: An Analytics
+                  Perspective (Dagstuhl Seminar 18031)},
+  editor = { Kathrin Klamroth  and  Joshua D. Knowles  and  G{\"u}nther Rudolph  and  Margaret M. Wiecek },
+  author = { Dimo Brockhoff  and  Roberto Calandra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Frank Neumann  and Selvakumar Ulaganathan},
+  title = {Meta-modeling for (interactive) multi-objective optimization
+                  (WG5)},
+  pages = {85--94}
+}
+
+ +
+@incollection{BroFriHebKle2007gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = { Dimo Brockhoff  and  Tobias Friedrich  and N. Hebbinghaus and
+                  C. Klein and  Frank Neumann  and  Eckart Zitzler },
+  title = {Do Additional Objectives Make a Problem Harder?},
+  pages = {765--772},
+  doi = {10.1145/1276958.1277114}
+}
+
+ +
+@incollection{BroLopNau2012ppsn,
+  volume = 7491,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {I}},
+  author = { Dimo Brockhoff  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Boris Naujoks  and  G{\"u}nther Rudolph },
+  title = {Runtime Analysis of Simple Interactive Evolutionary
+                  Biobjective Optimization Algorithms},
+  pages = {123--132},
+  doi = {10.1007/978-3-642-32937-1_13},
+  abstract = {Development and deployment of interactive evolutionary
+                  multiobjective optimization algorithms (EMOAs) have recently
+                  gained broad interest. In this study, first steps towards a
+                  theory of interactive EMOAs are made by deriving bounds on
+                  the expected number of function evaluations and queries to a
+                  decision maker. We analyze randomized local search and the
+                  (1+1)-EA on the biobjective problems LOTZ and COCZ under the
+                  scenario that the decision maker interacts with these
+                  algorithms by providing a subjective preference whenever
+                  solutions are incomparable. It is assumed that this decision
+                  is based on the decision maker's internal utility
+                  function. We show that the performance of the interactive
+                  EMOAs may dramatically worsen if the utility function is
+                  non-linear instead of linear.}
+}
+
+ +
+@incollection{BroSaxDeb2008handling,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  series = {Natural Computing Series},
+  year = 2008,
+  booktitle = {Multiobjective Problem Solving from Nature},
+  author = { Dimo Brockhoff  and  Saxena, Dhish Kumar  and  Kalyanmoy Deb  and  Eckart Zitzler },
+  editor = { Joshua D. Knowles  and  David Corne  and  Kalyanmoy Deb  and Chair, Deva Raj},
+  title = {On Handling a Large Number of Objectives A Posteriori and
+                  During Optimization},
+  pages = {377--403},
+  abstract = {Dimensionality reduction methods are used routinely in
+                  statistics, pattern recognition, data mining, and machine
+                  learning to cope with high-dimensional spaces. Also in the
+                  case of high-dimensional multiobjective optimization
+                  problems, a reduction of the objective space can be
+                  beneficial both for search and decision making. New questions
+                  arise in this context, e.g., how to select a subset of
+                  objectives while preserving most of the problem structure. In
+                  this chapter, two different approaches to the task of
+                  objective reduction are developed, one based on assessing
+                  explicit conflicts, the other based on principal component
+                  analysis (PCA). Although both methods use different
+                  principles and preserve different properties of the
+                  underlying optimization problems, they can be effectively
+                  utilized either in an a posteriori scenario or during
+                  search. Here, we demonstrate the usability of the
+                  conflict-based approach in a decision-making scenario after
+                  the search and show how the principal-component-based
+                  approach can be integrated into an evolutionary
+                  multicriterion optimization (EMO) procedure.},
+  doi = {10.1007/978-3-540-72964-8_18}
+}
+
+ +
+@incollection{BroTus2019bench,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  doi = {10.1145/3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  title = {Benchmarking algorithms from the platypus framework on the
+                  biobjective bbob-biobj testbed},
+  author = { Dimo Brockhoff  and  Tea Tu{\v s}ar },
+  pages = {1905--1911},
+  keywords = {unbounded archive}
+}
+
+ +
+@incollection{BroWagTrau2012r2,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2012,
+  editor = {Terence Soule and Jason H. Moore},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  title = {On the properties of the {R2} indicator},
+  author = { Dimo Brockhoff  and  Tobias Wagner  and  Heike Trautmann },
+  pages = {465--472},
+  annote = {Proof that R2 is weakly Pareto compliant.}
+}
+
+ +
+@incollection{BroZit2006allobjectives,
+  year = 2006,
+  volume = 4193,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Runarsson, Thomas Philip and   Hans-Georg Beyer  and  Edmund K. Burke  and  Juan-Juli{\'a}n Merelo  and  Darrell Whitley  and  Xin Yao },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IX}},
+  author = { Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Are All Objectives Necessary? {On} Dimensionality Reduction in
+                  Evolutionary Multiobjective Optimization},
+  pages = {533--542},
+  abstract = {Most of the available multiobjective evolutionary algorithms
+                  (MOEA) for approximating the Pareto set have been designed
+                  for and tested on low dimensional problems ($\leq$3
+                  objectives). However, it is known that problems with a high
+                  number of objectives cause additional difficulties in terms
+                  of the quality of the Pareto set approximation and running
+                  time. Furthermore, the decision making process becomes the
+                  harder the more objectives are involved. In this context, the
+                  question arises whether all objectives are necessary to
+                  preserve the problem characteristics. One may also ask under
+                  which conditions such an objective reduction is feasible, and
+                  how a minimum set of objectives can be computed. In this
+                  paper, we propose a general mathematical framework, suited to
+                  answer these three questions, and corresponding algorithms,
+                  exact and heuristic ones. The heuristic variants are geared
+                  towards direct integration into the evolutionary search
+                  process. Moreover, extensive experiments for four well-known
+                  test problems show that substantial dimensionality reductions
+                  are possible on the basis of the proposed methodology.}
+}
+
+ +
+@incollection{BroZit2006dimensionality,
+  author = { Dimo Brockhoff  and  Eckart Zitzler },
+  editor = {Waldmann, Karl-Heinz and Stocker, Ulrike M.},
+  title = {Dimensionality Reduction in Multiobjective Optimization: The
+                  Minimum Objective Subset Problem},
+  booktitle = {Operations Research Proceedings 2006},
+  year = 2007,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  pages = {423--429},
+  abstract = {The number of objectives in a multiobjective optimization
+                  problem strongly influences both the performance of
+                  generating methods and the decision making process in
+                  general. On the one hand, with more objectives, more
+                  incomparable solutions can arise, the number of which affects
+                  the generating method's performance. On the other hand, the
+                  more objectives are involved the more complex is the choice
+                  of an appropriate solution for a (human) decision maker. In
+                  this context, the question arises whether all objectives are
+                  actually necessary and whether some of the objectives may be
+                  omitted; this question in turn is closely linked to the
+                  fundamental issue of conflicting and non-conflicting
+                  optimization criteria. Besides a general definition of
+                  conflicts between objective sets, we here introduce the
+                  NP-hard problem of computing a minimum subset of objectives
+                  without losing information (MOSS). Furthermore, we present
+                  for MOSS both an approximation algorithm with optimum
+                  approximation ratio and an exact algorithm which works well
+                  for small input instances. We conclude with experimental
+                  results for a random problem and the multiobjective
+                  0/1-knapsack problem},
+  doi = {10.1007/978-3-540-69995-8_68},
+  keywords = {objective reduction}
+}
+
+ +
+@inproceedings{BroZit2007hypervolumeReduction,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  author = { Dimo Brockhoff  and  Eckart Zitzler },
+  title = {Improving hypervolume-based multiobjective evolutionary
+                  algorithms by using objective reduction methods},
+  pages = {2086--2093},
+  doi = {10.1109/CEC.2007.4424730},
+  keywords = {objective reduction}
+}
+
+ +
+@inproceedings{BruRit2018cec,
+  year = 2018,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2018 Congress on Evolutionary Computation (CEC 2018)},
+  key = {IEEE CEC},
+  author = { Artur Brum and  Marcus Ritt},
+  title = {Automatic Design of Heuristics for Minimizing the Makespan in
+                  Permutation Flow Shops},
+  pages = {1--8},
+  doi = {10.1109/CEC.2018.8477787}
+}
+
+ +
+@incollection{BruRit2018evo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 10782,
+  series = {Lecture Notes in Computer Science},
+  year = 2018,
+  booktitle = {Proceedings of EvoCOP 2018 -- 18th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Arnaud Liefooghe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  author = { Artur Brum and  Marcus Ritt},
+  title = {Automatic Algorithm Configuration for the Permutation Flow
+                  Shop Scheduling Problem Minimizing Total Completion Time},
+  pages = {85--100},
+  doi = {10.1007/978-3-319-77449-7_6}
+}
+
+ +
+@incollection{BuiRiz04:gecco,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3102,
+  editor = { Kalyanmoy Deb and others},
+  year = 2004,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2004, Part I},
+  author = {T. N. Bui  and  Rizzo, Jr, J. R. },
+  title = {Finding Maximum Cliques with Distributed Ants},
+  pages = {24--35}
+}
+
+ +
+@techreport{BurByk2012,
+  author = { Edmund K. Burke  and  Yuri Bykov },
+  title = {The Late Acceptance Hill-Climbing Heuristic},
+  institution = {University of Stirling},
+  number = {CSM-192},
+  year = 2012
+}
+
+ +
+@incollection{BurHydKen2007gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = { Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall  and John R. Woodward},
+  title = {Automatic Heuristic Generation with Genetic Programming: Evolving a Jack-of-all-trades or a Master of One},
+  pages = {1559--1565},
+  doi = {10.1145/1276958.1277273}
+}
+
+ +
+@incollection{BurHydKen2019hb,
+  publisher = {Springer},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 272,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2019,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Edmund K. Burke  and  Matthew R. Hyde  and  Graham Kendall  and  Gabriela Ochoa  and  Ender {\"O}zcan  and John R. Woodward},
+  title = {A Classification of Hyper-Heuristic Approaches: Revisited},
+  chapter = 14,
+  pages = {453--477},
+  doi = {10.1007/978-3-319-91086-4_14}
+}
+
+ +
+@incollection{Burkard:QAP,
+  volume = 2,
+  editor = { Panos M. Pardalos  and  D.-Z. Du },
+  year = 1998,
+  publisher = {Kluwer Academic Publishers},
+  booktitle = {Handbook of Combinatorial Optimization},
+  author = { Burkard, Rainer E.  and  Eranda {\c C}ela  and  Panos M. Pardalos  and  L. S. Pitsoulis },
+  title = {The quadratic assignment problem},
+  pages = {241--338}
+}
+
+ +
+@incollection{Buz2019signif,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = {Maxim Buzdalov},
+  title = {Towards better estimation of statistical significance when
+                  comparing evolutionary algorithms},
+  pages = {1782--1788},
+  doi = {10.1145/3319619.3326899}
+}
+
+ +
+@techreport{CI-235-07,
+  author = { Nicola Beume  and  Carlos M. Fonseca  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Jan Vahrenhold },
+  title = {On the Complexity of Computing the Hypervolume
+                  Indicator},
+  institution = {University of Dortmund},
+  year = 2007,
+  number = {CI-235/07},
+  month = dec,
+  note = {Published in IEEE Transactions on Evolutionary Computation~\cite{BeuFonLopPaqVah09:tec}}
+}
+
+ +
+@misc{COSEAL,
+  title = {COnfiguration and SElection of ALgorithms},
+  key = {COSEAL},
+  howpublished = {http://www.coseal.net},
+  year = 2017
+}
+
+ +
+@misc{CPLEX,
+  author = {IBM},
+  title = {{ILOG} {CPLEX} Optimizer},
+  year = 2017,
+  howpublished = {\url{http://www.ibm.com/software/integration/optimization/cplex-optimizer/}}
+}
+
+ +
+@incollection{CalShiCebDoe2019bayesian,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  doi = {10.1145/3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = {Calvo, Borja and  Shir, Ofer M.  and  Josu Ceberio  and  Carola Doerr  and Wang, Hao and  Thomas B{\"a}ck  and  Jos{\'e} A. Lozano },
+  title = {Bayesian Performance Analysis for Black-box Optimization
+                  Benchmarking},
+  pages = {1789--1797},
+  numpages = 9,
+  acmid = 3326888,
+  keywords = {bayesian inference, benchmarking, black-box optimization,
+                  evolutionary algorithms, performance measures, plackett-luce
+                  model}
+}
+
+ +
+@incollection{CamDorStu2018ants,
+  volume = 11172,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and  Mauro Birattari  and Christensen, Anders L. and Reina, Andreagiovanni and  Vito Trianni },
+  year = 2018,
+  booktitle = {Swarm Intelligence, 11th International Conference, ANTS 2018},
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {Why the Intelligent Water Drops Cannot Be Considered as a Novel Algorithm},
+  pages = {302--314}
+}
+
+ +
+@incollection{CamPas2010lion,
+  doi = {10.1007/978-3-642-13800-3},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Roberto Battiti },
+  series = {Lecture Notes in Computer Science},
+  volume = 6073,
+  year = 2010,
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  title = {Adapting to a realistic decision maker: experiments
+                  towards a reactive multi-objective optimizer},
+  author = { Paolo Campigotto  and  Andrea Passerini },
+  pages = {338--341}
+}
+
+ +
+@incollection{CamStuDor2020ants,
+  volume = 12421,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and  Thomas St{\"u}tzle  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Heiko Hamann  and Heinrich, Mary Katherine},
+  year = 2020,
+  booktitle = {Swarm Intelligence, 12th International Conference, ANTS 2020},
+  title = {Grey Wolf, Firefly and Bat Algorithms: Three Widespread Algorithms that Do Not Contain Any Novelty},
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  pages = {121--133}
+}
+
+ +
+@unpublished{CamTriLop2017pseudo,
+  author = {Felipe Campelo and \'Athila R. Trindade and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Pseudoreplication in Racing Methods for Tuning Metaheuristics},
+  note = {In preparation},
+  year = 2017
+}
+
+ +
+@book{Can00:book,
+  author = {E. Cant{\'u}-Paz},
+  title = {Efficient and Accurate Parallel Genetic Algorithms},
+  publisher = {Kluwer Academic Publishers, Boston, MA},
+  year = 2000
+}
+
+ +
+@inproceedings{CarJesMar2003,
+  author = {P. Cardoso and M. Jesus and A. Marquez},
+  title = {{MONACO}: multi-objective network optimisation based on an {ACO}},
+  booktitle = {Proc. X  Encuentros de Geometr\'ia Computacional},
+  year = 2003,
+  address = {Seville, Spain}
+}
+
+ +
+@incollection{CarPinOli2017recipe,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  isbn = {978-3-319-55695-6},
+  year = 2017,
+  volume = 10196,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Proceedings of  the 20th European Conference on Genetic Programming, EuroGP 2017},
+  editor = {James McDermott and Mauro Castelli and Luk{\'{a}}s Sekanina
+                  and Evert Haasdijk and  Pablo Garc{\'i}a-S{\'a}nchez },
+  author = {de S{\'{a}}, Alex Guimar{\~{a}}es Cardoso  and Pinto, Walter
+                  Jos{\'{e}} G. S. and Oliveira, Luiz Ot{\'{a}}vio Vilas Boas and  Gisele Pappa },
+  title = {{RECIPE:} {A} Grammar-Based Framework for Automatically
+                  Evolving Classification Pipelines},
+  pages = {246--261},
+  doi = {10.1007/978-3-319-55696-3_16}
+}
+
+ +
+@incollection{CarProSha2013votes,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = {Michael J. Kearns and R. Preston McAfee and {\'{E}}va Tardos},
+  booktitle = {Proceedings of the Fourteenth ACM Conference on Electronic
+                  Commerce},
+  title = {When Do Noisy Votes Reveal the Truth?},
+  author = {Ioannis Caragiannis and Ariel D. Procaccia and Nisarg Shah},
+  doi = {10.1145/2482540.2482570},
+  keywords = {computer social choice, mallows model, sample complexity},
+  pages = {143--160},
+  abstract = {A well-studied approach to the design of voting rules views
+                  them as maximum likelihood estimators; given votes that are
+                  seen as noisy estimates of a true ranking of the
+                  alternatives, the rule must reconstruct the most likely true
+                  ranking. We argue that this is too stringent a requirement,
+                  and instead ask: How many votes does a voting rule need to
+                  reconstruct the true ranking? We define the family of
+                  pairwise-majority consistent rules, and show that for all
+                  rules in this family the number of samples required from the
+                  Mallows noise model is logarithmic in the number of
+                  alternatives, and that no rule can do asymptotically better
+                  (while some rules like plurality do much worse). Taking a
+                  more normative point of view, we consider voting rules that
+                  surely return the true ranking as the number of samples tends
+                  to infinity (we call this property accuracy in the limit);
+                  this allows us to move to a higher level of abstraction. We
+                  study families of noise models that are parametrized by
+                  distance functions, and find voting rules that are accurate
+                  in the limit for all noise models in such general
+                  families. We characterize the distance functions that induce
+                  noise models for which pairwise-majority consistent rules are
+                  accurate in the limit, and provide a similar result for
+                  another novel family of position-dominance consistent
+                  rules. These characterizations capture three well-known
+                  distance functions.}
+}
+
+ +
+@incollection{CebMenLoz2015mallows,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  title = {Kernels of {Mallows} Models for Solving Permutation-based
+                  Problems},
+  author = { Josu Ceberio  and  Alexander Mendiburu  and  Jos{\'e} A. Lozano },
+  pages = {505--512}
+}
+
+ +
+@book{Cela:QAP,
+  author = { Eranda {\c C}ela },
+  title = {The Quadratic Assignment Problem: Theory and
+                  Algorithms},
+  publisher = {Kluwer Academic Publishers},
+  year = 1998,
+  address = {Dordrecht, The Netherlands}
+}
+
+ +
+@inproceedings{CesOddSmi2000:aaai,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 2000,
+  booktitle = {Proceedings of AAAI 2000 -- Seventeenth National Conference
+                  on Artificial Intelligence},
+  editor = {Henry A. Kautz and Bruce W. Porter},
+  author = { Amadeo Cesta  and  Angelo Oddi  and  Stephen F. Smith },
+  title = {Iterative Flattening: A Scalable Method for Solving Multi-Capacity Scheduling Problems},
+  pages = {742--747}
+}
+
+ +
+@mastersthesis{Chang99,
+  author = { S. T. H. Chang },
+  title = {Optimizing the Real Time Operation of a Pumping
+                  Station at a Water Filtration Plant using Genetic
+                  Algorithms},
+  school = {Department of Civil and Environmental Engineering,
+                  The University of Adelaide},
+  year = 1999,
+  type = {Honors Thesis}
+}
+
+ +
+@inproceedings{Chase89,
+  author = { Donald V. Chase  and  Lindell E. Ormsbee },
+  title = {Optimal pump operation of water distribution systems
+                  with multiple storage tanks},
+  booktitle = {Proceedings of American Water Works Association
+                  Computer Specialty Conference},
+  pages = {205--214},
+  year = 1989,
+  address = {Denver, USA},
+  organization = {AWWA}
+}
+
+ +
+@inproceedings{Chase91,
+  author = { Donald V. Chase  and  Lindell E. Ormsbee },
+  title = {An alternate formulation of time as a decision
+                  variable to facilitate real-time operation of water
+                  supply systems},
+  booktitle = {Proceedings of the 18th Annual Conference of Water
+                  Resources Planning and Management},
+  pages = {923--927},
+  year = 1991,
+  address = { New York, NY},
+  organization = {ASCE}
+}
+
+ +
+@incollection{CheBuzDoeDan2023aac,
+  publisher = {{ACM}},
+  editor = { Chicano, Francisco  and  Tobias Friedrich  and K{\"o}tzing, Timo  and  Franz Rothlauf },
+  year = 2023,
+  booktitle = {Proceedings of the 17th {ACM}/{SIGEVO} Conference on Foundations of Genetic Algorithms},
+  author = {Chen, Deyao and Buzdalov, Maxim and  Carola Doerr  and Nguyen Dang},
+  title = {Using Automated Algorithm Configuration for Parameter
+                  Control},
+  pages = {38--49},
+  doi = {10.1145/3594805.3607127}
+}
+
+ +
+@inproceedings{CheGaoChen2005scga,
+  title = {{SCGA}: Controlling genetic algorithms with {Sarsa}(0)},
+  author = {Chen, Fei and Gao, Yang and Chen, Zhao-qian and Chen, Shi-fu},
+  booktitle = {Computational Intelligence for Modelling, Control and
+                  Automation, 2005 and International Conference on Intelligent
+                  Agents, Web Technologies and Internet Commerce, International
+                  Conference on},
+  volume = 1,
+  pages = {1177--1183},
+  year = 2005,
+  publisher = {IEEE},
+  doi = {10.1109/CIMCA.2005.1631422}
+}
+
+ +
+@incollection{CheGinBecMol2013moda,
+  address = { Heidelberg, Germany},
+  publisher = {Springer International Publishing},
+  booktitle = {mODa 10--Advances in Model-Oriented Design and Analysis},
+  year = 2013,
+  editor = {Ucinski, Dariusz and Atkinson, Anthony C.  and Patan, Maciej},
+  author = {Chevalier, Cl{\'e}ment and Ginsbourger, David and Bect,
+                  Julien and Molchanov, Ilya},
+  title = {Estimating and Quantifying Uncertainties on Level Sets Using
+                  the {Vorob}'ev Expectation and Deviation with {Gaussian}
+                  Process Models},
+  pages = {35--43},
+  abstract = {Several methods based on Kriging have recently been proposed
+                  for calculating a probability of failure involving
+                  costly-to-evaluate functions. A closely related problem is to
+                  estimate the set of inputs leading to a response exceeding a
+                  given threshold. Now, estimating such a level set---and not
+                  solely its volume---and quantifying uncertainties on it are
+                  not straightforward. Here we use notions from random set
+                  theory to obtain an estimate of the level set, together with
+                  a quantification of estimation uncertainty. We give explicit
+                  formulae in the Gaussian process set-up and provide a
+                  consistency result. We then illustrate how space-filling
+                  versus adaptive design strategies may sequentially reduce
+                  level set estimation uncertainty.},
+  doi = {10.1007/978-3-319-00218-7_5}
+}
+
+ +
+@inproceedings{CheIshSha2021clustering,
+  title = {Clustering-Based Subset Selection in Evolutionary
+                  Multiobjective Optimization},
+  author = {Chen, Weiyu and  Ishibuchi, Hisao  and Shang, Ke},
+  booktitle = {2021 IEEE International Conference on Systems, Man, and
+                  Cybernetics},
+  year = 2021,
+  organization = {IEEE},
+  pages = {468--475}
+}
+
+ +
+@incollection{CheKanTay1991ijcai,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = {Mylopoulos, John and Reiter, Raymond},
+  year = 1995,
+  booktitle = {Proceedings of  the 12th International Joint Conference on Artificial Intelligence (IJCAI-91)},
+  author = {Cheeseman, Peter C. and Kanefsky, Bob and Taylor, William M.},
+  title = {Where the Really Hard Problems Are},
+  pages = {331--340}
+}
+
+ +
+@inproceedings{CheXuChe04,
+  publisher = {IEEE Press},
+  year = 2004,
+  booktitle = {Proceedings of the International Conference on
+                  Machine Learning and Cybernetics},
+  editor = {Cloete, Ian and Wong, Kit-Po and Berthold, Michael},
+  author = {L. Chen and X. H. Xu and Y. X. Chen},
+  title = {An adaptive ant colony clustering algorithm},
+  pages = {1387--1392}
+}
+
+ +
+@inproceedings{CheIshSha2020subset,
+  year = 2020,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2020 Congress on Evolutionary Computation (CEC 2020)},
+  key = {IEEE CEC},
+  title = {Modified Distance-based Subset Selection for Evolutionary
+                  Multi-objective Optimization Algorithms},
+  author = {Chen, Weiyu and  Ishibuchi, Hisao  and Shang, Ke},
+  pages = {1--8},
+  keywords = {IGD+}
+}
+
+ +
+@inproceedings{CheXinChe2017vdmlibrary,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = {Chen, Lu and Xin, Bin and Chen, Jie and Juan Li},
+  title = {A virtual-decision-maker library considering personalities
+                  and dynamically changing preference structures for
+                  interactive multiobjective optimization},
+  pages = {636--641},
+  doi = {10.1109/CEC.2017.7969370},
+  keywords = {machine DM, interactive EMOA}
+}
+
+ +
+@incollection{ChiGoe2010,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  title = {Mixed Models for the Analysis of Optimization Algorithms},
+  author = { Marco Chiarandini  and Yuri Goegebeur},
+  annote = {Preliminary version available as \emph{Tech.\
+                  Rep.} MF-2009-07-001 at the The Danish Mathematical Society},
+  pages = {225--264},
+  doi = {10.1007/978-3-642-02538-9}
+}
+
+ +
+@phdthesis{ChiarandiniPhD,
+  author = { Marco Chiarandini },
+  title = {Stochastic Local Search Methods for Highly
+                  Constrained Combinatorial Optimisation Problems},
+  school = {FB Informatik, TU Darmstadt, Germany},
+  year = 2005
+}
+
+ +
+@misc{Chieng2014,
+  author = {Tsung-Che Chiang},
+  title = {nsga3cpp: A {C++} implementation of {NSGA-III}},
+  howpublished = {\url{http://web.ntnu.edu.tw/~tcchiang/publications/nsga3cpp/nsga3cpp.htm}},
+  year = 2014
+}
+
+ +
+@inproceedings{ChrSchBur2011patus,
+  publisher = {IEEE Computer Society},
+  year = 2011,
+  series = {IPDPS '11},
+  booktitle = {Proceedings of the 2011 IEEE International Parallel \&
+                  Distributed Processing Symposium},
+  editor = {Frank Mueller},
+  author = {Matthias Christen and Olaf Schenk and Helmar Burkhart},
+  title = {{PATUS:} A Code Generation and Autotuning Framework for
+                  Parallel Iterative Stencil Computations on Modern
+                  Microarchitectures},
+  pages = {676--687},
+  doi = {10.1109/IPDPS.2011.70}
+}
+
+ +
+@techreport{ChrVan2018,
+  author = {Jan Christiaens and Greet Vanden Berghe},
+  title = {Slack Induction by String Removals for Vehicle Routing Problems},
+  institution = {Department of Computing Science, KU Leuven, Gent, Belgium},
+  year = 2018,
+  number = {7-05-2018}
+}
+
+ +
+@techreport{Christofides1976,
+  title = {Worst-case analysis of a new heuristic for the travelling salesman problem},
+  author = { Christofides, Nicos },
+  year = 1976,
+  number = 388,
+  institution = {Graduate School of Industrial Administration, Carnegie-Mellon University, Pittsburgh, PA}
+}
+
+ +
+@incollection{ChuLop2021gecco,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  author = { Tinkle Chugh  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Maximising Hypervolume and Minimising $\epsilon$-Indicators
+                  using Bayesian Optimisation over Sets},
+  doi = {10.1145/3449726.3463178},
+  keywords = {multi-objective, surrogate models, epsilon, hypervolume},
+  supplement = {https://doi.org/10.5281/zenodo.4675569},
+  pages = {1326--1334}
+}
+
+ +
+@incollection{ChuNuaJanPho06,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2006,
+  editor = {M. Cattolico and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2006},
+  author = {S. Chusanapiputt and D. Nualhong and S. Jantarang
+                  and S. Phoomvuthisarn},
+  title = {Selective self-adaptive approach to ant system for
+                  solving unit commitment problem},
+  pages = {1729--1736}
+}
+
+ +
+@phdthesis{Chugh2017phd,
+  author = { Tinkle Chugh },
+  title = {Handling expensive multiobjective optimization problems with
+                  evolutionary algorithms},
+  school = {University of Jyv{\"a}skyl{\"a}},
+  year = 2017
+}
+
+ +
+@inproceedings{Chugh2020scalar,
+  year = 2020,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2020 Congress on Evolutionary Computation (CEC 2020)},
+  key = {IEEE CEC},
+  author = { Tinkle Chugh },
+  title = {Scalarizing Functions in Bayesian Multiobjective
+                  Optimization},
+  pages = {1--8},
+  doi = {10.1109/CEC48606.2020.9185706}
+}
+
+ +
+@incollection{CinFerLopAl2021evocop,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 12692,
+  series = {Lecture Notes in Computer Science},
+  year = 2021,
+  booktitle = {Proceedings of EvoCOP 2021 -- 21th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Christine Zarges  and  Verel, S{\'e}bastien },
+  author = { Christian Cintrano  and  Javier Ferrer  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Alba, Enrique },
+  title = {Hybridization of Racing Methods with Evolutionary Operators
+                  for Simulation Optimization of Traffic Lights Programs},
+  abstract = {In many real-world optimization problems, like the traffic
+                  light scheduling problem tackled here, the evaluation of
+                  candidate solutions requires the simulation of a process
+                  under various scenarios. Thus, good solutions should not only
+                  achieve good objective function values, but they must be
+                  robust (low variance) across all different scenarios.
+                  Previous work has revealed the effectiveness of IRACE for
+                  this task.  However, the operators used by IRACE to generate
+                  new solutions were designed for configuring algorithmic
+                  parameters, that have various data types (categorical,
+                  numerical, etc.). Meanwhile, evolutionary algorithms have
+                  powerful operators for numerical optimization, which could
+                  help to sample new solutions from the best ones found in the
+                  search. Therefore, in this work, we propose a hybridization
+                  of the elitist iterated racing mechanism of IRACE with
+                  evolutionary operators from differential evo- lution and
+                  genetic algorithms. We consider a realistic case study
+                  derived from the traffic network of Malaga (Spain) with 275
+                  traffic lights that should be scheduled optimally. After a
+                  meticulous study, we discovered that the hybrid algorithm
+                  comprising IRACE plus differential evolution offers
+                  statistically better results than conventional algorithms and
+                  also improves travel times and reduces pollution.},
+  keywords = {Hybrid algorithms, Evolutionary algorithms, Simulation
+                  optimization, Uncertainty, Traffic light planning},
+  pages = {17--33},
+  doi = {10.1007/978-3-030-72904-2_2},
+  annote = {Extended version published in Evolutionary Computation journal~\cite{CinFerLopAlb2022irace}.}
+}
+
+ +
+@inproceedings{CirJohMcGZha2001,
+  author = {Jill Cirasella and David S. Johnson and  Lyle A. McGeoch  and Weixiong Zhang},
+  title = {The Asymmetric Traveling Salesman Problem: Algorithms,
+                  Instance Generators, and Tests},
+  booktitle = {Algorithm Engineering and Experimentation, Third
+                  International Workshop, {ALENEX} 2001, Washington, DC, USA,
+                  January 5-6, 2001, Revised Papers},
+  pages = {32--59},
+  series = {Lecture Notes in Computer Science},
+  volume = 2153,
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2001,
+  doi = {10.1007/3-540-44808-X_3},
+  editor = {Adam L. Buchsbaum and Jack Snoeyink}
+}
+
+ +
+@inproceedings{ClaKar1992electronic,
+  author = {Jon Claerbout and Martin Karrenbach},
+  year = 1992,
+  title = {Electronic documents give reproducible research a new
+                  meaning},
+  booktitle = {SEG Technical Program Expanded Abstracts 1992},
+  publisher = {Society of Exploration Geophysicists},
+  pages = {601--604},
+  doi = {10.1190/1.1822162},
+  annote = {Proposed a reproducibility taxonomy, defined reproducibility
+                  and taxonomy}
+}
+
+ +
+@misc{CleKen2011spso,
+  author = { Clerc, Maurice  and  J. Kennedy },
+  title = {Standard {PSO} 2011},
+  howpublished = {Particle Swarm Central},
+  year = 2011,
+  url = {http://www.particleswarm.info/},
+  alias = {pso:central}
+}
+
+ +
+@unpublished{Clerc2012spso,
+  title = {Standard {Particle} {Swarm} {Optimisation}},
+  author = { Clerc, Maurice },
+  url = {https://hal.archives-ouvertes.fr/hal-00764996},
+  numpages = 15,
+  year = 2012,
+  month = sep,
+  hal_id = {hal-00764996},
+  hal_version = {v1},
+  keywords = {particle swarm optimisation},
+  abstract = {Since 2006, three successive standard PSO versions have been
+                  put on line on the Particle Swarm Central
+                  (\url{http://particleswarm.info}), namely SPSO 2006, 2007,
+                  and 2011. The basic principles of all three versions can be
+                  informally described the same way, and in general, this
+                  statement holds for almost all PSO variants. However, the
+                  exact formulae are slightly different, because they took
+                  advantage of latest theoretical analysis available at the
+                  time they were designed.},
+  note = {hal-00764996}
+}
+
+ +
+@incollection{Coe2015multi,
+  title = {Multi-objective Evolutionary Algorithms in Real-World
+                  Applications: Some Recent Results and Current Challenges},
+  author = { Carlos A. {Coello Coello} },
+  booktitle = {Advances in Evolutionary and Deterministic Methods for
+                  Design, Optimization and Control in Engineering and Sciences},
+  pages = {3--18},
+  year = 2015,
+  doi = {10.1007/978-3-319-11541-2_1},
+  publisher = {Springer}
+}
+
+ +
+@book{CoeLamVVe2007:book,
+  author = { Carlos A. {Coello Coello}  and  Gary B. Lamont  and  David A. {Van Veldhuizen} },
+  title = {Evolutionary Algorithms for Solving Multi-Objective Problems},
+  year = 2007,
+  publisher = {Springer},
+  address = { New York, NY},
+  edition = {2nd},
+  doi = {10.1007/978-0-387-36797-2}
+}
+
+ +
+@incollection{CoeSie2004igd,
+  year = 2004,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2972,
+  series = {Lecture Notes in Artificial Intelligence},
+  booktitle = {Proceedings of MICAI},
+  editor = {Monroy, Ra{\'u}l and Arroyo-Figueroa, Gustavo and Sucar, Luis
+                  Enrique and Sossa, Humberto},
+  author = { Carlos A. {Coello Coello}  and Reyes-Sierra, Margarita},
+  title = {A Study of the Parallelization of a Coevolutionary
+                  Multi-objective Evolutionary Algorithm},
+  pages = {688--697},
+  keywords = {IGD},
+  annote = {Introduces Inverted Generational Distance (IGD)}
+}
+
+ +
+@inproceedings{Coello2000cec,
+  month = jul,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2000,
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  key = {IEEE CEC},
+  author = { Carlos A. {Coello Coello} },
+  title = {Handling Preferences in Evolutionary Multiobjective
+                  Optimization: A Survey},
+  pages = {30--37},
+  alias = {Coe2000cec}
+}
+
+ +
+@incollection{Coello2017results,
+  volume = 10687,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  booktitle = {Theory and Practice of Natural Computing - 6th International Conference,
+               {TPNC} 2017},
+  editor = {Carlos Mart{\'i}n{-}Vide and Roman Neruda and Miguel A. Vega{-}Rodr{\'i}guez},
+  author = { Carlos A. {Coello Coello} },
+  title = {Recent Results and Open Problems in Evolutionary Multiobjective Optimization},
+  pages = {3--21}
+}
+
+ +
+@book{Cohen1995ai,
+  author = {Paul R. Cohen},
+  title = {Empirical Methods for Artificial Intelligence},
+  publisher = {MIT Press},
+  address = {Cambridge, MA},
+  year = 1995,
+  alias = {Coh95}
+}
+
+ +
+@incollection{Cohen82,
+  author = { G. Cohen },
+  title = {Optimal Control of Water Supply Networks},
+  booktitle = {Optimization and Control of Dynamic Operational
+                  Research Models},
+  pages = {251--276},
+  publisher = {North-Holland Publishing Company},
+  year = 1982,
+  editor = { S. G. Tzafestas },
+  volume = 4,
+  chapter = 8,
+  address = {Amsterdam}
+}
+
+ +
+@inproceedings{ColDorMan92:ecal,
+  publisher = {MIT Press, Cambridge, MA},
+  editor = {F. J. Varela and P. Bourgine},
+  year = 1992,
+  booktitle = {Proceedings of  the First European Conference on
+                  Artificial Life},
+  author = { Alberto Colorni  and  Marco Dorigo  and  Vittorio Maniezzo },
+  title = {Distributed Optimization by Ant Colonies},
+  pages = {134--142}
+}
+
+ +
+@incollection{ColMonGauSli07,
+  volume = 4926,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2008,
+  doi = {10.1007/978-3-540-79305-2},
+  shorteditor = {Monmarch{\'e}, Nicolas and others},
+  editor = {Monmarch{\'e}, Nicolas and  Talbi, El-Ghazali  and Collet, Pierre and  Marc Schoenauer  and Lutton, Evelyne},
+  booktitle = {Artificial Evolution},
+  author = {Sonia Colas and  Nicolas Monmarch{\'e}  and Pierre
+                  Gaucher and Mohamed Slimane},
+  pages = {87--99},
+  title = {Artificial Ants for the Optimization of Virtual
+                  Keyboard Arrangement for Disabled People}
+}
+
+ +
+@book{ConSchVic2009,
+  author = {Andrew R. Conn and Katya Scheinberg and Luis N. Vicente},
+  title = {Introduction to Derivative-Free Optimization},
+  publisher = {Society for Industrial and Applied Mathematics, Philadelphia, PA, USA},
+  year = 2009,
+  series = {MPS--SIAM Series on Optimization}
+}
+
+ +
+@misc{ConcordeSolver,
+  author = { David Applegate  and  Robert E. Bixby  and  Va{\v{s}}ek Chv{\'a}tal  and  William J. Cook },
+  title = {Concorde {TSP} Solver},
+  howpublished = {\url{http://www.math.uwaterloo.ca/tsp/concorde.html}},
+  note = {Version visited last on 15 April 2014},
+  year = 2014
+}
+
+ +
+@book{Conover99:pns,
+  author = { W. J. Conover },
+  title = {Practical Nonparametric Statistics},
+  publisher = {John Wiley \& Sons},
+  address = { New York, NY},
+  year = 1999,
+  edition = {3rd}
+}
+
+ +
+@inproceedings{Cook1971,
+  author = {Cook, Stephen A.},
+  title = {The Complexity of Theorem-proving Procedures},
+  booktitle = {Proceedings of the Third Annual ACM Symposium on Theory of
+                  Computing},
+  series = {STOC '71},
+  year = 1971,
+  location = {Shaker Heights, Ohio, USA},
+  pages = {151--158},
+  numpages = 8,
+  doi = {10.1145/800157.805047},
+  acmid = 805047,
+  publisher = {ACM}
+}
+
+ +
+@book{Cook2012,
+  author = { William J. Cook },
+  title = {In Pursuit of the Traveling Salesman},
+  publisher = {Princeton University Press, Princeton, NJ},
+  year = 2012
+}
+
+ +
+@incollection{Cook2019,
+  year = 2019,
+  editor = {Bernhard Steffen and Gerhard Woeginger},
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 10000,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Computing and Software Science: State of the Art and Perspectives},
+  title = {Computing in Combinatorial Optimization},
+  author = { William J. Cook },
+  pages = {27--47},
+  doi = {10.1007/978-3-319-91908-9_3}
+}
+
+ +
+@incollection{CorKno2001pesa2,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = {Erik D. Goodman},
+  year = 2001,
+  booktitle = {Proceedings of the 3rd Annual Conference on Genetic and
+                  Evolutionary Computation, GECCO 2001},
+  author = { David Corne  and Jerram, Nick R. and  Joshua D. Knowles  and Oates,
+                  Martin J.},
+  title = {{PESA-II}: Region-Based Selection in Evolutionary
+                  Multiobjective Optimization},
+  pages = {283--290},
+  numpages = 8,
+  doi = {10.5555/2955239.2955289}
+}
+
+ +
+@inproceedings{CorKno2003cec,
+  year = 2003,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = dec,
+  booktitle = {Proceedings of  the 2003 Congress on Evolutionary Computation (CEC'03)},
+  key = {IEEE CEC},
+  author = { David Corne  and  Joshua D. Knowles },
+  title = {Some Multiobjective Optimizers are Better than Others},
+  pages = {2506--2512}
+}
+
+ +
+@incollection{CorKno2003nfl,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { David Corne  and  Joshua D. Knowles },
+  title = {No free lunch and free leftovers theorems for multiobjective
+                  optimisation problems},
+  pages = {327--341},
+  doi = {10.1007/3-540-36970-8_23}
+}
+
+ +
+@incollection{CorKnoOat2000ppsn,
+  anote = {IC.29},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marc Schoenauer  and others},
+  aeditor = { Marc Schoenauer  and  Kalyanmoy Deb  and  G{\"u}nther Rudolph  and  Xin Yao  and E. Lutton and  Juan-Juli{\'a}n Merelo  and  Hans-Paul Schwefel },
+  year = 2000,
+  volume = 1917,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VI}},
+  author = { David Corne  and  Joshua D. Knowles  and M. J. Oates},
+  title = {The {Pareto} Envelope-Based Selection Algorithm for
+                  Multiobjective Optimization},
+  pages = {839--848}
+}
+
+ +
+@book{CorLeiRiv2009,
+  title = {Introduction to algorithms},
+  author = {Cormen, Thomas H. and Leiserson, Charles E. and Rivest, Ronald L. and Stein, Clifford},
+  year = 2009,
+  publisher = {MIT Press},
+  address = {Cambridge, MA}
+}
+
+ +
+@incollection{CorRey2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { David Corne  and Reynolds, Alan},
+  title = {Evaluating optimization algorithms: bounds on the performance
+                  of optimizers on unseen problems},
+  pages = {707--710},
+  doi = {10.1145/2001858.2002073},
+  supplement = {http://is.gd/evalopt}
+}
+
+ +
+@inproceedings{CorViaHerMor2000bwas,
+  month = sep # { 7--9},
+  year = 2000,
+  date = {2000-09-07/2000-09-09},
+  organization = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Martin Middendorf  and  Thomas St{\"u}tzle },
+  booktitle = {Abstract proceedings of ANTS 2000 -- From Ant
+                  Colonies to Artificial Ants: Second International
+                  Workshop on Ant Algorithms},
+  author = { Oscar Cord{\'o}n  and I. Fern{\'a}ndez de Viana and  Francisco Herrera  and L. Moreno},
+  title = {A New {ACO} Model Integrating Evolutionary Computation
+                  Concepts: The Best-Worst Ant System},
+  pages = {22--29}
+}
+
+ +
+@incollection{CowKenSou2000hyper,
+  publisher = {Springer},
+  volume = 2079,
+  series = {Lecture Notes in Computer Science},
+  year = 2000,
+  editor = {Edmund K. Burke and Wilhelm Erben},
+  booktitle = {PATAT 2000: Proceedings of the 3rd International Conference
+                  of the Practice and Theory of Automated Timetabling},
+  author = {Peter I. Cowling and  Graham Kendall  and Eric Soubeiga},
+  title = {A Hyperheuristic Approach to Scheduling a Sales Summit},
+  pages = {176--190},
+  doi = {10.1007/3-540-44629-X_11},
+  annote = {First mention of the term hyper-heuristic.}
+}
+
+ +
+@book{Crawley2012rbook,
+  author = {M. J. Crawley},
+  title = {The \proglang{R} Book},
+  publisher = {Wiley},
+  year = 2012,
+  edition = {2nd}
+}
+
+ +
+@techreport{CroGloThoTra1963,
+  author = {W. B. Crowston and F. Glover and G. L. Thompson and
+                  J. D. Trawick},
+  title = {Probabilistic and Parametric Learning Combinations of Local
+                  Job Shop Scheduling Rules},
+  institution = {GSIA, Carnegie-Mellon University, Pittsburgh, PA, USA},
+  year = 1963,
+  number = {No.\ 117},
+  type = {ONR Research Memorandum}
+}
+
+ +
+@techreport{Cul92,
+  author = { Joseph C. Culberson },
+  title = {Iterated Greedy Graph Coloring and the Difficulty
+                  Landscape},
+  institution = {Department of Computing Science, The University of
+                  Alberta, Edmonton, Alberta, Canada},
+  year = 1992,
+  number = {92-07}
+}
+
+ +
+@inproceedings{CulBeaPap95,
+  author = { Joseph C. Culberson  and A. Beacham and D. Papp},
+  title = {Hiding our Colors},
+  booktitle = {Proceedings of the CP'95 Workshop on Studying and Solving
+                  Really Hard Problems},
+  pages = {31--42},
+  year = 1995,
+  address = {Cassis, France},
+  month = sep
+}
+
+ +
+@incollection{CulLuo1996,
+  series = {{DIMACS} Series on Discrete Mathematics and Theoretical Computer Science},
+  volume = 26,
+  year = 1996,
+  address = { Providence, RI},
+  publisher = {American Mathematical Society},
+  booktitle = {Cliques, Coloring, and Satisfiability: Second {DIMACS}
+                  Implementation Challenge},
+  editor = {David S. Johnson and  Michael A. Trick },
+  author = { Joseph C. Culberson  and F. Luo},
+  title = {Exploring the $k$-colorable Landscape with Iterated Greedy},
+  pages = {245--284},
+  alias = {CulLuo96}
+}
+
+ +
+@book{Cumming2012,
+  author = {Jeff Cumming},
+  title = {Understanding the New Statistics -- Effect Sizes, Confidence Intervals, and Meta-analysis},
+  publisher = {Taylor \& Francis},
+  year = 2012
+}
+
+ +
+@incollection{DanDeC2014,
+  year = 2014,
+  publisher = {SciTePress},
+  booktitle = {{ICORES} 2014 - Proceedings of the 3rd International Conference on
+               Operations Research and Enterprise Systems},
+  editor = {Bego{\~{n}}a Vitoriano and Eric Pinson and Fernando Valente},
+  author = {Nguyen {Dang Thi Thanh} and Patrick {De Causmaecker}},
+  title = {Motivations for the Development of a Multi-objective
+                  Algorithm Configurator},
+  pages = {328--333}
+}
+
+ +
+@incollection{DanDec2016neighborhood,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 10079,
+  editor = {Paola Festa and  Meinolf Sellmann  and  Joaquin Vanschoren },
+  series = {Lecture Notes in Computer Science},
+  year = 2016,
+  booktitle = {Learning and Intelligent Optimization, 10th International
+                  Conference, LION 10},
+  title = {Characterization of Neighborhood Behaviours in a
+                  Multi-neighborhood Local Search Algorithm},
+  author = {Nguyen {Dang Thi Thanh} and Patrick {De Causmaecker}},
+  pages = {234--239}
+}
+
+ +
+@incollection{DanDec2019analysis,
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = 11968,
+  publisher = {Springer},
+  year = 2019,
+  editor = {Nikolaos F. Matsatsinis and Yannis Marinakis and  Panos M. Pardalos },
+  booktitle = {Learning and Intelligent Optimization, 13th International
+                  Conference, LION 13},
+  author = {Nguyen Dang and Patrick {De Causmaecker}},
+  title = {Analysis of Algorithm Components and Parameters: Some Case
+                  Studies},
+  pages = {288--303},
+  abstract = {Modern high-performing algorithms are usually highly
+                  parameterised, and can be configured either manually or by an
+                  automatic algorithm configurator. The algorithm performance
+                  dataset obtained after the configuration step can be used to
+                  gain insights into how different algorithm parameters
+                  influence algorithm performance. This can be done by a number
+                  of analysis methods that exploit the idea of learning
+                  prediction models from an algorithm performance dataset and
+                  then using them for the data analysis on the importance of
+                  variables. In this paper, we demonstrate the complementary
+                  usage of three methods along this line, namely forward
+                  selection, fANOVA and ablation analysis with surrogates on
+                  three case studies, each of which represents some special
+                  situations that the analyses can fall into. By these
+                  examples, we illustrate how to interpret analysis results and
+                  discuss the advantage of combining different analysis
+                  methods.},
+  doi = {10.1007/978-3-030-05348-2_25}
+}
+
+ +
+@incollection{DanDoe2019gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = {Nguyen Dang and  Carola Doerr },
+  title = {Hyper-parameter tuning for the ({1 + (\(\lambda\), \(\lambda\))}) {GA}},
+  pages = {889--897},
+  doi = {10.1145/3321707.3321725},
+  keywords = {irace; theory}
+}
+
+ +
+@incollection{DanPerCauStu2017:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  author = {Nguyen {Dang Thi Thanh} and   P{\'e}rez C{\'a}ceres, Leslie and Patrick {De Causmaecker} and  Thomas St{\"u}tzle },
+  title = {Configuring {\rpackage{irace}} Using Surrogate Configuration Benchmarks},
+  pages = {243--250},
+  keywords = {irace},
+  doi = {10.1145/3071178.3071238}
+}
+
+ +
+@inproceedings{DanPoz2018,
+  year = 2018,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2018 Congress on Evolutionary Computation (CEC 2018)},
+  key = {IEEE CEC},
+  title = {A Meta-Learning Algorithm Selection Approach for the Quadratic Assignment Problem},
+  author = {Dantas, Augusto Lopez and Pozo, Aurora Trinidad Ramirez},
+  pages = {1--8}
+}
+
+ +
+@incollection{Dandy03,
+  author = { Graeme C. Dandy  and  Matthew S. Gibbs },
+  editor = {Paul Bizier and Paul DeBarry},
+  title = {Optimizing System Operations and Water Quality},
+  publisher = {ASCE},
+  year = 2003,
+  booktitle = {Proceedings of World Water and Environmental
+                  Resources Congress},
+  address = {Philadelphia, USA},
+  doi = {10.1061/40685(2003)127},
+  note = {on CD-ROM}
+}
+
+ +
+@phdthesis{Dang2018PhD,
+  title = {Data analytics for algorithm design},
+  school = {KU Leuven, Belgium},
+  author = {Nguyen {Dang Thi Thanh}},
+  year = 2018,
+  annote = {Supervised by Patrick {De Causmaecker}}
+}
+
+ +
+@incollection{DaoVerOchTom2012:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2012,
+  editor = {Terence Soule and Jason H. Moore},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  author = {Fabio Daolio and  Verel, S{\'e}bastien  and  Gabriela Ochoa  and Marco Tomassini},
+  title = {Local Optima Networks and the Performance of Iterated Local
+                  Search},
+  pages = {369--376}
+}
+
+ +
+@inproceedings{DauBalBak2020different,
+  year = 2020,
+  editor = {Hugo Larochelle and Marc'Aurelio Ranzato and Raia Hadsell and
+                  Maria{-}Florina Balcan and Hsuan{-}Tien Lin},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS
+                  33)},
+  author = {Daulton, Samuel and Balandat, Maximilian and Bakshy, Eytan},
+  title = {Differentiable Expected Hypervolume Improvement for Parallel
+                  Multi-Objective {Bayesian} Optimization},
+  pages = {9851--9864},
+  epub = {https://proceedings.neurips.cc/paper/2020/file/6fec24eac8f18ed793f5eaad3dd7977c-Paper.pdf}
+}
+
+ +
+@inproceedings{DeSchaetzen98,
+  author = { Werner de Schaetzen  and  Dragan A. Savic  and  Godfrey A. Walters },
+  title = {A genetic algorithm approach to pump scheduling in
+                  water supply},
+  booktitle = {Hydroinformatics '98},
+  pages = {897--899},
+  year = 1998,
+  editor = { V. Babovic  and  L. C. Larsen },
+  address = {Rotterdam, Balkema}
+}
+
+ +
+@inproceedings{DeaBod1988aaai,
+  year = 1988,
+  booktitle = {Proceedings of the 7th National Conference on Artificial
+                  Intelligence, AAAI-88},
+  url = {http://www.aaai.org/Conferences/AAAI/aaai88.php},
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  editor = {Howard E. Shrobe and Tom M. Mitchell and Reid G. Smith},
+  author = {Thomas Dean and Mark S. Boddy},
+  title = {An Analysis of Time-Dependent Planning},
+  pages = {49--54},
+  keywords = {anytime, performance profiles}
+}
+
+ +
+@book{DeanVoss99:DAE,
+  author = { Angela Dean  and  Daniel Voss },
+  title = {Design and Analysis of Experiments},
+  publisher = {Springer},
+  address = { London, UK },
+  doi = {10.1007/b97673},
+  year = 1999
+}
+
+ +
+@incollection{Deb2008introduction,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  title = {Introduction to evolutionary multiobjective optimization},
+  author = { Kalyanmoy Deb },
+  abstract = {In its current state, evolutionary multiobjective
+                  optimization (EMO) is an established field of research and
+                  application with more than 150 PhD theses, more than ten
+                  dedicated texts and edited books, commercial softwares and
+                  numerous freely downloadable codes, a biannual conference
+                  series running successfully since 2001, special sessions and
+                  workshops held at all major evolutionary computing
+                  conferences, and full-time researchers from universities and
+                  industries from all around the globe. In this chapter, we
+                  provide a brief introduction to EMO principles, illustrate
+                  some EMO algorithms with simulated results, and outline the
+                  current research and application potential of EMO. For
+                  solving multiobjective optimization problems, EMO procedures
+                  attempt to find a set of well-distributed Pareto-optimal
+                  points, so that an idea of the extent and shape of the
+                  Pareto-optimal front can be obtained. Although this task was
+                  the early motivation of EMO research, EMO principles are now
+                  being found to be useful in various other problem solving
+                  tasks, enabling one to treat problems naturally as they
+                  are. One of the major current research thrusts is to combine
+                  EMO procedures with other multiple criterion decision making
+                  (MCDM)  tools so as to develop hybrid and interactive
+                  multiobjective optimization algorithms for finding a set of
+                  trade-off optimal solutions and then choose a preferred
+                  solution for implementation. This chapter provides the
+                  background of EMO principles and their potential to launch
+                  such collaborative studies with MCDM researchers in the
+                  coming years.},
+  doi = {10.1007/978-3-540-88908-3_3},
+  pages = {59--96}
+}
+
+ +
+@incollection{Deb2005,
+  year = 2005,
+  address = {Boston, MA},
+  publisher = {Springer},
+  editor = { Edmund K. Burke  and  Graham Kendall },
+  booktitle = {Search Methodologies},
+  title = {Multi-objective optimization},
+  author = { Kalyanmoy Deb },
+  pages = {273--316},
+  doi = {10.1007/0-387-28356-0_10}
+}
+
+ +
+@book{Deb:MOEA,
+  author = { Kalyanmoy Deb },
+  title = {Multi-Objective Optimization Using Evolutionary
+                  Algorithms},
+  year = 2001,
+  publisher = {Wiley},
+  address = {Chichester, UK}
+}
+
+ +
+@incollection{DebAg1999polymut,
+  doi = {10.1007/978-3-7091-6384-9},
+  booktitle = {Artificial Neural Nets and Genetic Algorithms (ICANNGA-99)},
+  key = {ICANNGA},
+  year = 1999,
+  publisher = {Springer Verlag},
+  editor = {Andrej Dobnikar and Nigel C. Steele and David
+                  W. Pearson and Rudolf F. Albrecht},
+  author = { Kalyanmoy Deb  and S. Agrawal},
+  title = {A Niched-Penalty Approach for Constraint Handling in Genetic
+                  Algorithms},
+  pages = {235--243},
+  keywords = {polynomial mutation}
+}
+
+ +
+@incollection{DebAgrPra2000ppsn,
+  anote = {IC.29},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marc Schoenauer  and others},
+  aeditor = { Marc Schoenauer  and  Kalyanmoy Deb  and  G{\"u}nther Rudolph  and  Xin Yao  and E. Lutton and  Juan-Juli{\'a}n Merelo  and  Hans-Paul Schwefel },
+  year = 2000,
+  volume = 1917,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VI}},
+  title = {A fast elitist non-dominated sorting genetic algorithm
+             for multi-objective optimization: {NSGA-II}},
+  author = { Kalyanmoy Deb  and S. Agarwal and A. Pratap and T. Meyarivan},
+  pages = {849--858}
+}
+
+ +
+@techreport{DebJain02,
+  author = { Kalyanmoy Deb  and  Sachin Jain },
+  title = {Multi-Speed Gearbox Design Using Multi-Objective
+                  Evolutionary Algorithms},
+  institution = {KanGAL},
+  year = 2002,
+  number = 2002001,
+  month = feb
+}
+
+ +
+@incollection{DebMyb2016gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2016},
+  title = {Breaking the billion-variable barrier in real-world
+                  optimization using a customized evolutionary algorithm},
+  author = { Kalyanmoy Deb  and Myburgh, Christie},
+  pages = {653--660}
+}
+
+ +
+@incollection{DebSin2009emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  title = {Solving Bilevel Multi-Objective Optimization Problems Using
+                  Evolutionary Algorithms},
+  author = { Kalyanmoy Deb  and Sinha, Ankur},
+  pages = {110--124},
+  alias = {Deb2009}
+}
+
+ +
+@incollection{DebSun2006gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2006,
+  editor = {M. Cattolico and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2006},
+  author = { Kalyanmoy Deb  and Sundar, J.},
+  title = {Reference point based multi-objective optimization using
+                  evolutionary algorithms},
+  pages = {635--642},
+  doi = {10.1145/1143997.1144112},
+  annote = {Proposed R-NSGA-II}
+}
+
+ +
+@inproceedings{DebTewDixDut2007finding,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  title = {Finding trade-off solutions close to {KKT} points using
+                  evolutionary multi-objective optimization},
+  author = { Kalyanmoy Deb  and Tewari, Rahul and Dixit, Mayur and Dutta, Joydeep},
+  pages = {2109--2116}
+}
+
+ +
+@techreport{DebThiLau2001dtlz,
+  author = { Kalyanmoy Deb  and  Lothar Thiele  and  Marco Laumanns  and  Eckart Zitzler },
+  institution = {Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Z{\"u}rich, Switzerland},
+  number = 112,
+  title = {Scalable Test Problems for Evolutionary
+                  Multi-Objective Optimization},
+  year = 2001,
+  keywords = {DTLZ benchmark},
+  note = {Do not cite this TR! It is incorrect and it is superseeded by~\cite{DebThiLau2005dtlz}}
+}
+
+ +
+@incollection{DebThiLau2005dtlz,
+  address = { London, UK },
+  year = 2005,
+  month = jan,
+  editor = {Abraham, Ajith and Jain, Lakhmi and Goldberg, Robert},
+  series = {Advanced Information and Knowledge Processing},
+  publisher = {Springer},
+  booktitle = {Evolutionary Multiobjective Optimization},
+  author = { Kalyanmoy Deb  and  Lothar Thiele  and  Marco Laumanns  and  Eckart Zitzler },
+  title = {Scalable Test Problems for Evolutionary
+                  Multiobjective Optimization},
+  keywords = {DTLZ benchmark},
+  pages = {105--145},
+  doi = {10.1007/1-84628-137-7_6}
+}
+
+ +
+@incollection{DeeKar1982,
+  author = {William A. {Dees, Jr.} and Patrick G. Karger},
+  title = {Automated Rip-up and Reroute Techniques},
+  booktitle = {DAC'82, Proceedings of the 19th Design Automation Workshop},
+  publisher = {IEEE Press},
+  year = 1982,
+  pages = {432--439}
+}
+
+ +
+@phdthesis{DenBestenPhD,
+  author = { Matthijs L. {den Besten} },
+  title = {Simple Metaheuristics for Scheduling},
+  school = {FB Informatik, TU Darmstadt, Germany},
+  year = 2004,
+  url = {http://tuprints.ulb.tu-darmstadt.de/516/}
+}
+
+ +
+@incollection{DenCosEsp2013igd,
+  isbn = {978-1-4503-1963-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = { Christian Blum  and  Alba, Enrique },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  title = {Many-objective optimization using differential evolution with
+                  variable-wise mutation restriction},
+  author = {Denysiuk, Roman and Costa, Lino and Esp{\'i}rito Santo,
+                  Isabel},
+  pages = {591--598}
+}
+
+ +
+@inproceedings{DenDonScoLiLiFei2009imagenet,
+  title = {Imagenet: A large-scale hierarchical image database},
+  author = {Deng, Jia and Dong, Wei and Socher, Richard and Li, Li-Jia
+                  and Li, Kai and Fei-Fei, Li},
+  booktitle = {Computer Vision and Pattern Recognition, 2009. CVPR
+                  2009. IEEE Conference on},
+  pages = {248--255},
+  year = 2009,
+  organization = {IEEE}
+}
+
+ +
+@inproceedings{DesRit2018cec,
+  year = 2018,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2018 Congress on Evolutionary Computation (CEC 2018)},
+  key = {IEEE CEC},
+  author = { Marcelo {De Souza}  and  Marcus Ritt},
+  title = {An Automatically Designed Recombination Heuristic for the
+                  Test-Assignment Problem},
+  pages = {1--8},
+  doi = {10.1109/CEC.2018.8477801}
+}
+
+ +
+@incollection{DesRit2018evo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 10782,
+  series = {Lecture Notes in Computer Science},
+  year = 2018,
+  booktitle = {Proceedings of EvoCOP 2018 -- 18th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Arnaud Liefooghe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  author = { Marcelo {De Souza}  and  Marcus Ritt},
+  title = {Automatic Grammar-Based Design of Heuristic Algorithms for
+                  Unconstrained Binary Quadratic Programming},
+  pages = {67--84},
+  doi = {10.1007/978-3-319-77449-7_5}
+}
+
+ +
+@misc{DesRit2018hhbqp,
+  author = { Marcelo {De Souza}  and  Marcus Ritt},
+  title = {Hybrid Heuristic for Unconstrained Binary Quadratic
+                  Programming -- Source Code of {HHBQP}},
+  howpublished = {\url{https://github.com/souzamarcelo/hhbqp}},
+  year = 2018
+}
+
+ +
+@misc{DesRitLop2020acviz,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and   P{\'e}rez C{\'a}ceres, Leslie},
+  title = {{\softwarepackage{ACVIZ}}: A Tool for the Visual Analysis of
+                  the Configuration of Algorithms with {\rpackage{irace}} --
+                  Source Code},
+  howpublished = {\url{https://github.com/souzamarcelo/acviz}},
+  year = 2020
+}
+
+ +
+@misc{DesRitLopPer2020zenodo,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and   P{\'e}rez C{\'a}ceres, Leslie},
+  title = {{\softwarepackage{ACVIZ}}: Algorithm Configuration
+                  Visualizations for {\rpackage{irace}}: Supplementary material},
+  month = sep,
+  year = 2020,
+  publisher = {Zenodo},
+  howpublished = {\url{http://doi.org/10.5281/zenodo.4714582}}
+}
+
+ +
+@phdthesis{Dewez2004PhD,
+  author = {Sophie Dewez},
+  title = {On the toll setting problem},
+  school = {Facult\'{e} de Sciences, Universit\'{e} Libre de Bruxelles},
+  year = 2014,
+  annote = {Supervised by Dr. Martine Labb\'{e}}
+}
+
+ +
+@inproceedings{DiaYan2008succint,
+  title = {Succinct approximate convex {Pareto} curves},
+  author = {Diakonikolas, Ilias and  Mihalis Yannakakis },
+  booktitle = {Proceedings of the nineteenth annual ACM-SIAM symposium on
+                  Discrete algorithms},
+  year = 2008,
+  organization = {Society for Industrial and Applied Mathematics},
+  pages = {74--83}
+}
+
+ +
+@incollection{DieValAreRodSua2014ants,
+  volume = 8667,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  year = 2014,
+  booktitle = {Swarm Intelligence, 9th International Conference, ANTS 2014},
+  author = {Diego D{\'i}az and Pablo Valledor and Paula Areces and Jorge Rodil and Montserrat Su{\'a}rez},
+  title = {An {ACO} Algorithm to Solve an Extended Cutting Stock Problem for Scrap Minimization in a Bar Mill},
+  pages = {13--24}
+}
+
+ +
+@incollection{DigChiSch2006,
+  publisher = {IOS Press},
+  year = 2006,
+  booktitle = {Proceedings of the 17th European Conference on Artificial Intelligence,
+                  {ECAI} 2006, Riva del Garda, Italy, August29 - September 1, 2006},
+  editor = {Brewka, Gerhard and Coradeschi, Silvia and Perini, Anna and Traverso, Paolo},
+  author = {Luca {Di Gaspero} and  Marco Chiarandini  and Andrea Schaerf},
+  title = {A Study on the Short-Term Prohibition Mechanisms in Tabu Search},
+  pages = {83--87}
+}
+
+ +
+@incollection{DigRenUrl2013cp,
+  year = 2013,
+  volume = 8124,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Principles and Practice of Constraint Programming},
+  editor = {Christian Schulte},
+  title = {Constraint-Based Approaches for Balancing Bike Sharing
+                  Systems},
+  author = {Luca {Di Gaspero} and  Andrea Rendl  and  Tommaso Urli },
+  pages = {758--773},
+  doi = {10.1007/978-3-642-40627-0_56},
+  keywords = {F-race}
+}
+
+ +
+@incollection{DigRenUrl2013hyme,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7919,
+  series = {Lecture Notes in Computer Science},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Paola Festa and  Andrea Roli  and  M. Sampels },
+  isbn = {978-3-642-38515-5},
+  year = 2013,
+  booktitle = {Hybrid Metaheuristics},
+  title = {A Hybrid {ACO+CP} for Balancing Bicycle Sharing Systems},
+  author = {Luca {Di Gaspero} and  Andrea Rendl  and  Tommaso Urli },
+  pages = {198--212},
+  keywords = {F-race},
+  doi = {10.1007/978-3-642-38516-2_16}
+}
+
+ +
+@incollection{DobNebLop2022ants,
+  volume = 13491,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and  Heiko Hamann  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Jos{\'e} Garc{\'i}a-Nieto  and  Andries Engelbrecht  and  Carlo Pinciroli  and  Volker Strobel  and Camacho-Villal\'{o}n, Christian Leonardo},
+  year = 2022,
+  booktitle = {Swarm Intelligence, 13th International Conference, ANTS 2022},
+  author = {Doblas, Daniel and  Nebro, Antonio J.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Jos{\'e} Garc{\'i}a-Nieto  and  Carlos A. {Coello Coello} },
+  title = {Automatic Design of Multi-objective Particle Swarm
+                  Optimizers},
+  doi = {10.1007/978-3-031-20176-9_3},
+  pages = {28--40}
+}
+
+ +
+@incollection{DomHul2000,
+  year = 2000,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {The 6th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining, {KDD} 2000},
+  epub = {http://dl.acm.org/citation.cfm?id=347090},
+  editor = {Raghu Ramakrishnan and Salvatore J. Stolfo and Roberto
+                  J. Bayardo and Ismail Parsa},
+  key = {SIGKDD},
+  author = {Domingos, Pedro and Hulten, Geoff},
+  title = {Mining high-speed data streams},
+  pages = {71--80}
+}
+
+ +
+@incollection{DorDic99:nio,
+  alias = {CorDorGlo99},
+  address = {London, UK},
+  year = 1999,
+  publisher = {McGraw Hill},
+  editor = { David Corne  and  Marco Dorigo  and  Fred Glover },
+  booktitle = {New Ideas in Optimization},
+  author = { Marco Dorigo  and  Gianni A. {Di Caro} },
+  title = {The {Ant} {Colony} {Optimization} Meta-Heuristic},
+  pages = {11--32},
+  anote = {Also available as Technical Report IRIDIA/99-1,
+                  Universit{\'e} Libre de Bruxelles, Belgium}
+}
+
+ +
+@techreport{DorGam1996:iridia,
+  author = { Marco Dorigo  and  L. M. Gambardella },
+  title = {Ant {Colony} {System}},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 1996,
+  number = {IRIDIA/96-05}
+}
+
+ +
+@techreport{DorManCol1991:tr16Revised,
+  author = { Marco Dorigo  and  Vittorio Maniezzo  and  Alberto Colorni },
+  title = {The {Ant} {System}: {An} autocatalytic optimizing process},
+  institution = {Dipartimento di Elettronica, Politecnico di Milano, Italy},
+  year = 1991,
+  number = {91-016 Revised}
+}
+
+ +
+@techreport{DorManCol91:tr16,
+  author = { Marco Dorigo  and  Vittorio Maniezzo  and  Alberto Colorni },
+  title = {Positive Feedback as a Search Strategy},
+  institution = {Dipartimento di Elettronica, Politecnico di Milano, Italy},
+  year = 1991,
+  number = {91-016}
+}
+
+ +
+@incollection{DorMonOliStu2011eorms,
+  doi = {10.1002/9780470400531},
+  year = 2011,
+  publisher = {John Wiley \& Sons},
+  editor = {J. J. Cochran},
+  booktitle = {Wiley Encyclopedia of Operations Research and
+                  Management Science},
+  author = { Marco Dorigo  and  Marco A. {Montes de Oca}  and  Sabrina Oliveira  and  Thomas St{\"u}tzle },
+  title = {Ant Colony Optimization},
+  pages = {114--125},
+  volume = 1
+}
+
+ +
+@incollection{DorStu02:mh,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {The Ant Colony Optimization Metaheuristic: Algorithms, Applications and Advances},
+  pages = {251--285}
+}
+
+ +
+@book{DorStu2004:book,
+  author = { Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {Ant Colony Optimization},
+  publisher = {MIT Press},
+  address = {Cambridge, MA},
+  year = 2004,
+  pagination = 305,
+  anote = {305 p},
+  alias = {DorStu04:AcoBook},
+  alias = {DorStu04:book}
+}
+
+ +
+@phdthesis{DorigoPhD,
+  author = { Marco Dorigo },
+  title = {Optimization, Learning and Natural Algorithms},
+  school = {Dipartimento di Elettronica, Politecnico di Milano, Italy},
+  year = 1992,
+  atype = {{Ph.D.} thesis},
+  note = {In Italian},
+  alias = {Dor92:thesis}
+}
+
+ +
+@incollection{Dre2009gecco,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = { Johann Dr{\'e}o },
+  title = {Using performance fronts for parameter setting of
+                  stochastic metaheuristics},
+  pages = {2197--2200},
+  doi = {10.1145/1570256.1570301}
+}
+
+ +
+@incollection{DreDoeSem2019,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  doi = {10.1145/3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = { Johann Dr{\'e}o  and  Carola Doerr  and Semet, Yann},
+  title = {Coupling the design of benchmark with algorithm in landscape-aware solver design},
+  pages = {1419--1420}
+}
+
+ +
+@incollection{DreLieVer2021paradiseo,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  title = {Paradiseo: from a modular framework for evolutionary
+                  computation to the automated design of metaheuristics: 22
+                  years of {Paradiseo}},
+  doi = {10.1145/3449726.3463276},
+  author = { Johann Dr{\'e}o  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien  and  Marc Schoenauer  and  Juan-Juli{\'a}n Merelo  and Quemy, Alexandre and Bouvier, Benjamin and Gmys, Jan},
+  pages = {1522--1530},
+  numpages = 9,
+  keywords = {metaheuristics, evolutionary computation, software framework,
+                  automated algorithm design}
+}
+
+ +
+@incollection{DreSia02:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { Johann Dr{\'e}o  and P. Siarry},
+  title = {A New Ant Colony Algorithm Using the Heterarchical Concept Aimed at Optimization of Multiminima Continuous Functions},
+  pages = {216--221}
+}
+
+ +
+@phdthesis{Dreo2003phd,
+  author = { Johann Dr{\'e}o },
+  title = {Adaptation de la m{\'e}taheuristique des colonies de fourmis
+                  pour l'optimisation difficile en variables continues:
+                  Application en g{\'e}nie biologique et m{\'e}dical},
+  school = {Universit{\'e} Paris XII Val de Marne},
+  year = 2003,
+  month = dec,
+  hal_id = {tel-00093143},
+  hal_version = {v1},
+  keywords = {metaheuristic ; continuous optimization ; global optimization
+                  ; imagery ; registration ; ant colony algorithm ; estimation
+                  of distribution algorithm ; evolutionary computation ;
+                  m{\'e}taheuristique ; optimisation continue ; optimisation
+                  globale ; imagerie ; biom{\'e}dical ; recalage ; algorithme
+                  de colonie de fourmis ; algorithme {\`a} estimation de
+                  distribution ; algorithme {\'e}volutionnaire},
+  url = {https://tel.archives-ouvertes.fr/tel-00093143}
+}
+
+ +
+@incollection{DroJanWeg2002,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = { De Jong, Kenneth A.  and Poli, Riccardo and Rowe, Jonathan E.},
+  year = 2002,
+  booktitle = {Proceedings of the Seventh Workshop on Foundations of Genetic Algorithms (FOGA)},
+  title = {A new framework for the valuation of algorithms for black-box-optimization},
+  author = {Droste, Stefan and Jansen, Thomas and  Ingo Wegener },
+  pages = {253--270}
+}
+
+ +
+@incollection{IshPanSha2020unbounded,
+  publisher = {IOS Press},
+  editor = {Giuseppe De Giacomo and Alejandro Catala and Bistra Dilkina
+                  and Michela Milano and Senén Barro and Alberto Bugarín and
+                  Jérôme Lang},
+  series = {Frontiers in Artificial Intelligence and Applications},
+  volume = 325,
+  year = 2020,
+  booktitle = {Proceedings of the 24th European Conference on Artificial Intelligence (ECAI)},
+  title = {A new framework of evolutionary multi-objective algorithms
+                  with an unbounded external archive},
+  author = { Ishibuchi, Hisao  and Pang, Lie Meng and Shang, Ke}
+}
+
+ +
+@inproceedings{Dru2009replicability,
+  author = {Chris Drummond},
+  title = {Replicability is not Reproducibility: Nor is it Good Science},
+  booktitle = {Proceedings of the Evaluation Methods for Machine Learning
+                  Workshop at the 26th ICML},
+  address = {Montreal, Canada},
+  url = {http://www.site.uottawa.ca/~cdrummon/pubs/ICMLws09.pdf},
+  year = 2009
+}
+
+ +
+@incollection{DruThi2010,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  author = { M{\u{a}}d{\u{a}}lina M. Drugan  and  Dirk Thierens },
+  title = {Path-Guided Mutation for Stochastic {Pareto} Local Search Algorithms},
+  pages = {485--495}
+}
+
+ +
+@incollection{DuaSanMla2018vnd,
+  isbn = {978-3-319-07125-1},
+  publisher = {Springer International Publishing},
+  year = 2018,
+  booktitle = {Handbook of Heuristics},
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  author = { Duarte, Abraham  and   Jes{\'u}s S{\'a}nchez-Oro  and  Nenad Mladenovi{\'c}  and Todosijevi{\'c}, Raca},
+  title = {Variable Neighborhood Descent},
+  pages = {341--367},
+  doi = {10.1007/978-3-319-07124-4_9}
+}
+
+ +
+@techreport{Dub2009:sls-ds,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste },
+  title = {Weight Setting Strategies for Two-Phase Local Search:
+                  A Study on Biobjective Permutation Flowshop Scheduling},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2009,
+  number = {TR/IRIDIA/2009-024}
+}
+
+ +
+@incollection{DubHooStu2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {On the Empirical Scaling Behaviour of State-of-the-art Local
+                  Search Algorithms for the Euclidean {TSP}},
+  pages = {377--384},
+  doi = {10.1145/2739480.2754747}
+}
+
+ +
+@incollection{DubLopStu09:hm-bfsp,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 5818,
+  series = {Lecture Notes in Computer Science},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Luca {Di Gaspero} and  Andrea Roli  and  M. Sampels  and Andrea Schaerf},
+  year = 2009,
+  booktitle = {Hybrid Metaheuristics},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Effective Hybrid Stochastic Local Search Algorithms for
+                  Biobjective Permutation Flowshop Scheduling},
+  pages = {100--114},
+  doi = {10.1007/978-3-642-04918-7_8},
+  alias = {DuboisHM09}
+}
+
+ +
+@misc{DubLopStu10:journal-anytime-supp,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {{Supplementary material: Improving the Anytime Behavior of Two-Phase Local Search}},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2010-012}},
+  year = 2010
+}
+
+ +
+@misc{DubLopStu10:journal-bfsp-supp,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {{Supplementary material: A Hybrid TP+PLS Algorithm for Bi-objective Flow-shop Scheduling Problems}},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2010-001}},
+  year = 2010
+}
+
+ +
+@incollection{DubLopStu10:lion-bfsp,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Roberto Battiti },
+  series = {Lecture Notes in Computer Science},
+  volume = 6073,
+  year = 2010,
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Adaptive ``Anytime'' Two-Phase Local Search},
+  pages = {52--67},
+  doi = {10.1007/978-3-642-13800-3_5}
+}
+
+ +
+@incollection{DubLopStu2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Configuration of State-of-the-art Multi-Objective
+                  Optimizers Using the {TP+PLS} Framework},
+  pages = {2019--2026},
+  doi = {10.1145/2001576.2001847}
+}
+
+ +
+@incollection{DubLopStu2012evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 7245,
+  year = 2012,
+  editor = { Jin-Kao Hao  and  Martin Middendorf },
+  booktitle = {Proceedings of EvoCOP 2012 -- 12th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {{Pareto} Local Search Algorithms for Anytime
+                  Bi-objective Optimization},
+  pages = {206--217},
+  doi = {10.1007/978-3-642-29124-1_18},
+  alias = {DubLopStu12:evocop}
+}
+
+ +
+@incollection{DubLopStu2013hm,
+  url = {http://www.springer.com/engineering/computational+intelligence+and+complexity/book/978-3-642-30670-9},
+  year = 2013,
+  volume = 434,
+  series = {Studies in Computational Intelligence},
+  editor = { Talbi, El-Ghazali },
+  publisher = {Springer Verlag},
+  booktitle = {Hybrid Metaheuristics},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Combining Two Search Paradigms for Multi-objective
+                  Optimization: {Two}-{Phase} and {Pareto} Local Search},
+  pages = {97--117},
+  doi = {10.1007/978-3-642-30671-6_3},
+  alias = {DubLopStu2012hm}
+}
+
+ +
+@misc{DubPagStu2017:flowshop-makespan-supp,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Supplementary material: {An} iterated greedy algorithm with optimization of partial
+  solutions for the permutation flowshop problem},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2013-006}},
+  year = 2017
+}
+
+ +
+@inproceedings{DubStu2017cec,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle },
+  title = {Tuning of a Stigmergy-based Traffic Light Controller as a Dynamic Optimization Problem},
+  pages = {1--8}
+}
+
+ +
+@mastersthesis{Dubois2009,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste },
+  title = {A study of {Pareto} and Two-Phase Local Search
+                  Algorithms for Biobjective Permutation Flowshop
+                  Scheduling},
+  school = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2009
+}
+
+ +
+@mastersthesis{Dubois2010,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste },
+  title = {Effective Stochastic Local Search Algorithms For
+                  Bi-Objective Permutation Flowshop Scheduling},
+  school = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  type = {Rapport d'avancement de recherches pr\'esent\'e pour
+                  la Formation Doctorale en sciences de l'Ing\'enieur},
+  year = 2010
+}
+
+ +
+@phdthesis{DuboisPhD,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste },
+  title = {Anytime Local Search for Multi-Objective Combinatorial
+                  Optimization: Design, Analysis and Automatic Configuration},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2014,
+  annote = {Supervised by Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez }
+}
+
+ +
+@misc{Duecketal1999:patent,
+  author = { Gunter Dueck  and Martin Maehler and Johannes Schneider and Gerhard Schrimpf and Hermann Stamm-Wilbrandt},
+  title = {Optimization with Ruin Recreate},
+  howpublished = {US Patent 6,418,398 B1},
+  month = jul,
+  year = 2002,
+  note = {Filed on October 1, 1999 and granted on July 9, 2002;
+                  Assignee is IBM Corporation, Armonk, NY (US)}
+}
+
+ +
+@incollection{DumStu2003,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2003,
+  editor = { G{\"u}nther R. Raidl  and Gottlieb, Jens},
+  volume = 2611,
+  booktitle = {Proceedings of EvoCOP 2003 -- 3rd European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = {Irina Dumitrescu and  Thomas St{\"u}tzle },
+  title = {Combinations of Local Search and Exact Algorithms},
+  pages = {211--223},
+  doi = {10.1007/3-540-36605-9_20}
+}
+
+ +
+@incollection{DumStu2009,
+  address = { New York, NY},
+  series = {Annals of Information Systems},
+  volume = 10,
+  year = 2009,
+  publisher = {Springer},
+  booktitle = {Matheuristics---Hybridizing Metaheuristics and Mathematical
+                  Programming},
+  editor = { Vittorio Maniezzo  and  Thomas St{\"u}tzle  and  Stefan Vo{\ss} },
+  author = {Irina Dumitrescu and  Thomas St{\"u}tzle },
+  title = {Usage of Exact Algorithms to Enhance Stochastic Local Search Algorithms},
+  pages = {103--134}
+}
+
+ +
+@incollection{DurGarNebCoe2009emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  author = { Durillo, Juan J.  and  Jos{\'e} Garc{\'i}a-Nieto  and  Nebro, Antonio J.  and  Carlos A. {Coello Coello}  and Luna, Francisco and  Alba, Enrique },
+  title = {Multi-Objective Particle Swarm Optimizers: An Experimental
+                  Comparison},
+  pages = {495--509},
+  abstract = {Particle Swarm Optimization (PSO) has received increasing
+                  attention in the optimization research community since its
+                  first appearance in the mid-1990s. Regarding multi-objective
+                  optimization, a considerable number of algorithms based on
+                  Multi-Objective Particle Swarm Optimizers (MOPSOs) can be
+                  found in the specialized literature. Unfortunately, no
+                  experimental comparisons have been made in order to clarify
+                  which MOPSO version shows the best performance. In this
+                  paper, we use a benchmark composed of three well-known
+                  problem families (ZDT, DTLZ, and WFG) with the aim of
+                  analyzing the search capabilities of six representative
+                  state-of-the-art MOPSOs, namely, NSPSO, SigmaMOPSO, OMOPSO,
+                  AMOPSO, MOPSOpd, and CLMOPSO. We additionally propose a new
+                  MOPSO algorithm, called SMPSO, characterized by including a
+                  velocity constraint mechanism, obtaining promising results
+                  where the rest perform inadequately.},
+  isbn = {978-3-642-01020-0}
+}
+
+ +
+@incollection{DurNebLunAlb2009,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  title = {On the Effect of the Steady-State Selection Scheme in
+                  Multi-Objective Genetic Algorithms},
+  author = { Durillo, Juan J.  and  Nebro, Antonio J.  and Luna, Francisco and  Alba, Enrique },
+  pages = {183--197}
+}
+
+ +
+@incollection{DwoKumNao2001rank,
+  isbn = {1-58113-348-0},
+  year = 2001,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of the Tenth International World Wide Web
+                  Conference, {WWW} 10},
+  editor = {Vincent Y. Shen and Nobuo Saito and Michael R. Lyu and Mary
+                  Ellen Zurko},
+  title = {Rank aggregation methods for the Web},
+  author = {Dwork, Cynthia and Kumar, Ravi and Naor, Moni and Sivakumar,
+                  D.},
+  doi = {10.1145/371920.372165},
+  keywords = {Kemeny ranking,multi-word queries,rank aggregation,ranking
+                  functions,spam},
+  pages = {613--622}
+}
+
+ +
+@manual{EPANET2Manual,
+  title = {{EPANET} 2 Users Manual},
+  author = { L. A. Rossman },
+  organization = {U.S. Environmental Protection Agency},
+  address = {Cincinnati, USA},
+  year = 2000
+}
+
+ +
+@manual{EPANET94,
+  title = {{EPANET} User's Guide},
+  author = { L. A. Rossman },
+  organization = {Risk Reduction Engineering Laboratory, Office of
+                  Research and Development, U.S. Environmental
+                  Protection Agency},
+  address = {Cincinnati, USA},
+  year = 1994
+}
+
+ +
+@inproceedings{EPANET_Toolkit,
+  author = { L. A. Rossman },
+  title = {The {EPANET} {Programmer's} {Toolkit} for Analysis
+                  of Water Distribution Systems},
+  booktitle = {Proceedings of the Annual Water Resources Planning
+                  and Management Conference},
+  year = 1999,
+  address = {Reston, USA},
+  publisher = {ASCE},
+  anote = {CD-ROM}
+}
+
+ +
+@inproceedings{EbeKen1995:pso,
+  author = { Eberhart, Russell C.  and  J. Kennedy },
+  booktitle = {Proceedings of the Sixth International Symposium on
+                  Micro Machine and Human Science},
+  title = {A New Optimizer Using Particle Swarm Theory},
+  year = 1995,
+  pages = {39--43}
+}
+
+ +
+@inproceedings{EggHutHooLey2015,
+  year = 2015,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Blai Bonet and Sven Koenig},
+  author = { Katharina Eggensperger  and  Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Efficient Benchmarking of Hyperparameter Optimizers via
+                  Surrogates},
+  pages = {1114--1120},
+  doi = {10.1609/aaai.v29i1.9375}
+}
+
+ +
+@incollection{Ehm2016,
+  author = {Werner Ehm},
+  title = {Reproducibility from the perspective of meta-analysis},
+  editor = {Harald Atmanspacher and Sabine Maasen},
+  booktitle = {Reproducibility -- Principles, problems, practices and prospects},
+  publisher = {Wiley},
+  year = 2016,
+  pages = {141--168}
+}
+
+ +
+@incollection{EhrGan08hybrid,
+  alias = {BluEtAl08:hm-book},
+  series = {Studies in Computational Intelligence},
+  volume = 114,
+  year = 2008,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Mar{\'i}a J. Blesa  and  Andrea Roli  and  M. Sampels },
+  booktitle = {Hybrid Metaheuristics: An emergent approach for optimization},
+  author = { Matthias Ehrgott  and  Xavier Gandibleux },
+  title = {Hybrid Metaheuristics for Multi-objective
+                  Combinatorial Optimization},
+  doi = {10.1007/978-3-540-78295-7_8},
+  abstract = {Many real-world optimization problems can be
+                  modelled as combinatorial optimization
+                  problems. Often, these problems are characterized by
+                  their large size and the presence of multiple,
+                  conflicting objectives. Despite progress in solving
+                  multi-objective combinatorial optimization problems
+                  exactly, the large size often means that heuristics
+                  are required for their solution in acceptable time.
+                  Since the middle of the nineties the trend is
+                  towards heuristics that ``pick and choose'' elements
+                  from several of the established metaheuristic
+                  schemes. Such hybrid approximation techniques may
+                  even combine exact and heuristic approaches. In this
+                  chapter we give an overview over approximation
+                  methods in multi-objective combinatorial
+                  optimization. We briefly summarize ``classical''
+                  metaheuristics and focus on recent approaches, where
+                  metaheuristics are hybridized and/or combined with
+                  exact methods.  },
+  pages = {221--259}
+}
+
+ +
+@book{Ehrgott00:multicriteria,
+  author = { Matthias Ehrgott },
+  title = {Multicriteria Optimization},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2000,
+  volume = 491,
+  series = {Lecture Notes in Economics and Mathematical Systems}
+}
+
+ +
+@book{Ehrgott2005multicrit,
+  author = { Matthias Ehrgott },
+  title = {Multicriteria Optimization},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2005,
+  edition = {2nd},
+  doi = {10.1007/3-540-27659-9}
+}
+
+ +
+@incollection{EibHorKow2006rl,
+  title = {Reinforcement learning for online control of evolutionary
+                  algorithms},
+  author = { Agoston E. Eiben  and Horvath, Mark and Kowalczyk, Wojtek and Schut, Martijn C.},
+  booktitle = {International Workshop on Engineering Self-Organising
+                  Applications},
+  pages = {151--160},
+  year = 2006,
+  publisher = {Springer}
+}
+
+ +
+@inproceedings{EibJel2002critical,
+  year = 2002,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2002 Congress on Evolutionary Computation (CEC'02)},
+  key = {IEEE CEC},
+  author = { Agoston E. Eiben  and M. Jelasity},
+  title = {A critical note on experimental research methodology in {EC}},
+  annote = {Discusses reproducibility, generalizability and separation
+                  between training (for tuning and experimentation) and testing
+                  instances (for comparisons).},
+  doi = {10.1109/cec.2002.1006991},
+  pages = {582--587}
+}
+
+ +
+@incollection{EibMichSChSmi07,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  year = 2007,
+  booktitle = {Parameter Setting in Evolutionary Algorithms},
+  editor = {F. Lobo and C. F. Lima and  Zbigniew Michalewicz },
+  author = { Agoston E. Eiben  and  Zbigniew Michalewicz  and  Marc Schoenauer  and James E. Smith},
+  title = {Parameter Control in Evolutionary Algorithms},
+  pages = {19--46}
+}
+
+ +
+@book{EibSmi2003,
+  title = {Introduction to Evolutionary Computing},
+  author = { Agoston E. Eiben  and  Smith, James E. },
+  publisher = {Springer},
+  year = 2003,
+  isbn = 3540401849,
+  alias = {Eiben2003}
+}
+
+ +
+@book{EibSmi2007,
+  author = { Agoston E. Eiben  and  Smith, James E. },
+  title = {Introduction to Evolutionary Computing},
+  publisher = {Springer},
+  year = 2007,
+  series = {Natural Computing Series},
+  edition = {2nd},
+  alias = {EibSmi07}
+}
+
+ +
+@incollection{Ela2011:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = {Mohammed El-Abd},
+  title = {Opposition-based Artificial Bee Colony Algorithm},
+  pages = {109--116}
+}
+
+ +
+@techreport{ElsKhaTor2017financial,
+  author = {Elsokkary, Nada and Khan, Faisal Shah and La Torre, Davide
+                  and Humble, Travis S. and Gottlieb, Joel},
+  title = {Financial Portfolio Management using {D-Wave}'s Quantum
+                  Optimizer: The Case of {Abu} {Dhabi} Securities Exchange},
+  institution = {Oak Ridge National Lab, Oak Ridge, TN, USA},
+  year = 2017,
+  url = {https://www.osti.gov/biblio/1423041}
+}
+
+ +
+@inproceedings{EmmDeuKli2011ehvi,
+  year = 2011,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2011 Congress on Evolutionary Computation (CEC 2011)},
+  key = {IEEE CEC},
+  author = { Emmerich, Michael T. M.  and   Andr{\'{e}} H. Deutz  and J. W. Klinkenberg},
+  title = {Hypervolume-based expected improvement: Monotonicity
+                  properties and exact computation},
+  pages = {2147--2154},
+  doi = {10.1109/CEC.2011.5949880},
+  annote = {Proposed Expected Hypervolume Improvement (EHVI)}
+}
+
+ +
+@incollection{EmmFon2011emo,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  title = {Computing Hypervolume Contributions in Low
+                  Dimensions: Asymptotically Optimal Algorithm and
+                  Complexity Results},
+  doi = {10.1007/978-3-642-19893-9_9},
+  abstract = {Given a finite set $Y \subset \mathbb{R}^d$ of n mutually
+                  non-dominated vectors in $d \geq 2 dimensions$, the
+                  hypervolume contribution of a point $y \in Y$ is the
+                  difference between the hypervolume indicator of $Y$
+                  and the hypervolume indicator of $Y \setminus \{y\}$. In
+                  multi-objective metaheuristics, hypervolume
+                  contributions are computed in several selection and
+                  bounded-size archiving procedures. This paper
+                  presents new results on the (time) complexity of
+                  computing all hypervolume contributions. It is
+                  proved that for $d = 2$ and 3 the problem has time
+                  complexity $\Theta(n logn)$, and, for $d > 3$,
+                  the time complexity is bounded below by $\Omega(n
+                  logn)$. Moreover, complexity bounds are derived for
+                  computing a single hypervolume contribution. A
+                  dimension sweep algorithm with time complexity
+                  $\mathcal{O} (n logn)$ and space
+                  complexity $\mathcal{O}(n)$ is
+                  proposed for computing all hypervolume contributions
+                  in three dimensions. It improves the complexity of
+                  the best known algorithm for $d = 3$ by a factor of
+                  $\sqrt{n}$ . Theoretical results
+                  are complemented by performance tests on randomly
+                  generated test-problems.},
+  author = { Emmerich, Michael T. M.  and  Carlos M. Fonseca },
+  pages = {121--135}
+}
+
+ +
+@incollection{EppDeSStu2011:adt,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Ronen I. Brafman  and  F. Roberts  and  Alexis Tsouki{\`a}s },
+  volume = 6992,
+  series = {Lecture Notes in Artificial Intelligence},
+  year = 2011,
+  booktitle = {Algorithmic Decision Theory, Third International Conference,
+                  {ADT} 2011},
+  author = { Stefan Eppe  and  Yves {De Smet}  and  Thomas St{\"u}tzle },
+  title = {A bi-objective optimization model to eliciting decision maker's preferences for the {PROMETHEE II} method},
+  pages = {56--66}
+}
+
+ +
+@inproceedings{EppLopStuDeS2011:cec,
+  year = 2011,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2011 Congress on Evolutionary Computation (CEC 2011)},
+  key = {IEEE CEC},
+  author = { Stefan Eppe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle  and  Yves {De Smet} },
+  title = {An Experimental Study of Preference Model Integration into
+                  Multi-Objective Optimization Heuristics},
+  pages = {2751--2758},
+  doi = {10.1109/CEC.2011.5949963}
+}
+
+ +
+@inproceedings{EriPeaGar2019scalable,
+  year = 2019,
+  editor = {Hanna M. Wallach and Hugo Larochelle and Alina Beygelzimer
+                  and Florence d'Alch{\'{e}}{-}Buc and Emily B. Fox and Roman
+                  Garnett},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS 32)},
+  title = {Scalable Global Optimization via Local {Bayesian}
+                  Optimization},
+  author = {Eriksson, David and Pearce, Michael and Gardner, Jacob and
+                  Turner, Ryan D. and Poloczek, Matthias},
+  pages = {5496--5507},
+  epub = {http://papers.nips.cc/paper/8788-scalable-global-optimization-via-local-bayesian-optimization.pdf},
+  annote = {Arxiv preprint arXiv: \url{https://arxiv.org/abs/1910.01739}}
+}
+
+ +
+@inproceedings{Ertin01,
+  author = { Emre Ertin  and  Anthony N. Dean  and  Mathew L. Moore  and  Kevin L. Priddy },
+  title = {Dynamic Optimization for Optimal Control of Water
+                  Distribution Systems},
+  booktitle = {Applications and Science of Computational
+                  Intelligence {IV}, Proceedings of {SPIE}},
+  year = 2001,
+  month = mar,
+  pages = {142--149},
+  editor = { Kevin L. Priddy  and  Paul E. Keller  and  Peter J. Angeline },
+  volume = 4390
+}
+
+ +
+@inproceedings{Esat94,
+  author = { V. Esat  and  M. Hall },
+  title = {Water resources system optimization using genetic algorithms},
+  booktitle = {Hydroinformatics'94},
+  pages = {225--231},
+  year = 1994,
+  editor = { A. Verwey  and  A. Minns  and  V. Babovic  and  C. Maksimovi{\'c} },
+  address = {Balkema, Rotterdam, The Netherlands},
+  note = {}
+}
+
+ +
+@incollection{EshSch1992,
+  isbn = {1-55860-263-1},
+  year = 1993,
+  publisher = {Morgan Kaufmann Publishers},
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  editor = { Darrell Whitley },
+  author = { Larry J. Eshelman  and  J. David Schaffer },
+  title = {Real-Coded Genetic Algorithms and Interval-Schemata},
+  pages = {187--202},
+  alias = {DBLP:conf/foga/EshelmanS92}
+}
+
+ +
+@inproceedings{Eshelman89crossoverbiases,
+  publisher = {Morgan Kaufmann Publishers, San Mateo, CA},
+  editor = { J. David Schaffer },
+  year = 1989,
+  booktitle = {Proceedings of  the Third International Conference on Genetic Algorithms (ICGA'89)},
+  author = { Larry J. Eshelman  and  A. Caruana  and  J. David Schaffer },
+  title = {Biases in the Crossover Landscape},
+  pages = {86--91}
+}
+
+ +
+@incollection{EveFieSin2002full,
+  title = {Full Elite Sets for Multi-objective Optimisation},
+  author = { Everson, Richard M.  and  Jonathan E. Fieldsend  and Singh, Sameer},
+  booktitle = {Adaptive Computing in Design and Manufacture {V}},
+  publisher = {Springer},
+  address = { London, UK },
+  year = 2002,
+  pages = {343--354},
+  doi = {10.1007/978-0-85729-345-9_29},
+  annote = {Extended version published as \cite{FieEveSing2003tec}}
+}
+
+ +
+@incollection{EycSno02,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = {C. J. Eyckelhof and M. Snoek},
+  title = {Ant Systems for a Dynamic {TSP}: {Ants} Caught in a
+                  Traffic Jam},
+  pages = {88--99}
+}
+
+ +
+@incollection{FalLinHut2015spysmac,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  editor = {Heule, Marijn and Weaver, Sean},
+  volume = 9340,
+  series = {Lecture Notes in Computer Science},
+  year = 2015,
+  booktitle = {Theory and Applications of Satisfiability Testing -- {SAT}
+                  2015},
+  title = {{SpySMAC}: Automated configuration and performance analysis
+                  of {SAT} solvers},
+  author = {Falkner, Stefan and  Marius Thomas Lindauer  and  Frank Hutter },
+  doi = {10.1007/978-3-319-24318-4_16},
+  pages = {215--222}
+}
+
+ +
+@incollection{FalZapGar2021gecco,
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  author = { Falc{\'{o}}n-Cardona, Jes{\'{u}}s Guillermo  and  Zapotecas-Mart{\'{i}}nez, Sa{\'{u}}l  and Abel
+                  Garc{\'{i}}a-N{\'{a}}jera},
+  title = {Pareto compliance from a practical point of view},
+  doi = {10.1145/3449639.3459276},
+  pages = {395--402}
+}
+
+ +
+@inproceedings{FarAma2002nafips,
+  publisher = {IEEE Service Center},
+  month = jun,
+  address = {Piscataway, New Jersey},
+  year = 2002,
+  booktitle = {Proceedings of the NAFIPS-FLINT International
+                  Conference'2002},
+  key = {NAFIPS},
+  author = {M. Farina and P. Amato},
+  title = {On the Optimal Solution Definition for Many-criteria
+                  Optimization Problems},
+  pages = {233--238},
+  doi = {10.1109/nafips.2002.1018061},
+  annote = {First to mention exponential number of nondominated solutions
+                  with respect to many objectives?}
+}
+
+ +
+@incollection{FavMorPel09:sls,
+  volume = 5752,
+  series = {Lecture Notes in Computer Science},
+  year = 2009,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Engineering Stochastic Local Search
+                  Algorithms. Designing, Implementing and Analyzing
+                  Effective Heuristics. SLS~2009},
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  author = { D. Favaretto  and  E. Moretti  and  Paola Pellegrini },
+  title = {On the explorative behavior of {\MaxMinAntSystem}},
+  pages = {115--119}
+}
+
+ +
+@inproceedings{FawHelHooKar2011icaps,
+  year = 2011,
+  booktitle = {Proceedings of ICAPS-PAL11},
+  editor = {Karpas, Erez and Jim{\'e}nez Celorrio, Sergio and Kambhampati, Subbarao},
+  author = { Chris Fawcett  and Malte Helmert and  Holger H. Hoos  and Erez Karpas
+                  and Gabriele R\"{o}ger and Jendrik Seipp},
+  title = {{FD-Autotune}: Domain-Specific Configuration using
+                  Fast-Downward}
+}
+
+ +
+@inproceedings{FawHoos2013mic,
+  year = 2013,
+  booktitle = {Proceedings of MIC 2013, the 10th Metaheuristics
+                  International Conference},
+  key = {MIC},
+  author = { Chris Fawcett  and  Holger H. Hoos },
+  title = {Analysing Differences between Algorithm
+                  Configurations through Ablation},
+  pages = {123--132},
+  epub = {http://www.cs.ubc.ca/~hoos/Publ/FawHoo13.pdf}
+}
+
+ +
+@incollection{FerAlvDiaIglEna2014ants,
+  volume = 8667,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  year = 2014,
+  booktitle = {Swarm Intelligence, 9th International Conference, ANTS 2014},
+  author = { Fern\'{a}ndez, Silvino  and  \'{A}lvarez, Segundo  and  Diego D{\'i}az  and Miguel Iglesias and Borja Ena},
+  title = {Scheduling a Galvanizing Line by Ant Colony Optimization},
+  doi = {10.1007/978-3-319-09952-1_13},
+  pages = {146--157}
+}
+
+ +
+@incollection{FerAlvMalValDia2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2015},
+  author = { Fern\'{a}ndez, Silvino  and  \'{A}lvarez, Segundo  and Malatsetxebarria, Eneko and Valledor, Pablo and  Diego D{\'i}az },
+  title = {Performance Comparison of Ant Colony Algorithms for the
+                  Scheduling of Steel Production Lines},
+  doi = {10.1145/2739482.2764658},
+  keywords = {irace}
+}
+
+ +
+@incollection{FerFonGas2007:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = { Jos{\'e} C. Ferreira  and  Carlos M. Fonseca  and  Ant{\'o}nio Gaspar{-}Cunha },
+  title = {Methodology to select solutions from the {Pareto}-optimal set: a comparative study},
+  pages = {789--796}
+}
+
+ +
+@incollection{FerPudHat1994,
+  doi = {10.1016/b978-0-444-81892-8.50040-7},
+  year = 1994,
+  pages = {403--413},
+  author = {F. J. Ferri and P. Pudil and M. Hatef and J. Kittler},
+  title = {Comparative study of techniques for large-scale feature
+                  selection},
+  editor = {Edzard S. Gelsema and Laveen S. Kanal},
+  series = {Machine Intelligence and Pattern Recognition},
+  publisher = {North-Holland},
+  volume = 16,
+  booktitle = {Pattern Recognition in Practice IV},
+  abstract = {The combinatorial search problem arising in feature selection
+                  in high dimensional spaces is considered. Recently developed
+                  techniques based on the classical sequential methods and the
+                  (l, r) search called Floating search algorithms are compared
+                  against the Genetic approach to feature subset search. Both
+                  approaches have been designed with the view to give a good
+                  compromise between efficiency and effectiveness for large
+                  problems. The purpose of this paper is to investigate the
+                  applicability of these techniques to high dimensional
+                  problems of feature selection. The aim is to establish
+                  whether the properties inferred for these techniques from
+                  medium scale experiments involving up to a few tens of
+                  dimensions extend to dimensionalities of one order of
+                  magnitude higher. Further, relative merits of these
+                  techniques vis-a-vis such high dimensional problems are
+                  explored and the possibility of exploiting the best aspects
+                  of these methods to create a composite feature selection
+                  procedure with superior properties is considered.},
+  annote = {Describes sequential forward / backward selection}
+}
+
+ +
+@incollection{FerValDia2016gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2016},
+  author = { Fern\'{a}ndez, Silvino  and Valledor, Pablo and  Diego D{\'i}az  and Malatsetxebarria, Eneko and Iglesias, Miguel},
+  title = {Criticality of Response Time in the usage of Metaheuristics
+                  in Industry},
+  pages = {937--940}
+}
+
+ +
+@inproceedings{FeuKleEggSprBluHut2015autosklearn,
+  url = {http://papers.nips.cc/book/advances-in-neural-information-processing-systems-28-2015},
+  year = 2015,
+  booktitle = {Advances in Neural Information Processing Systems (NIPS
+                  28)},
+  editor = {Corinna Cortes and Neil D. Lawrence and Daniel D. Lee and
+                  Masashi Sugiyama and Roman Garnett},
+  author = { Matthias Feurer  and Klein, Aaron and  Katharina Eggensperger  and Springenberg, Jost and Blum, Manuel and  Frank Hutter },
+  title = {Efficient and robust automated machine learning},
+  pages = {2962--2970},
+  epub = {http://papers.nips.cc/paper/5872-efficient-and-robust-automated-machine-learning.pdf}
+}
+
+ +
+@incollection{FeuKleEggSprBluHut2019autosklearn,
+  booktitle = {Automated Machine Learning},
+  doi = {10.1007/978-3-030-05318-5},
+  year = 2019,
+  publisher = {Springer},
+  epub = {http://automl.org/book},
+  editor = { Frank Hutter  and Kotthoff, Lars and  Joaquin Vanschoren },
+  author = { Matthias Feurer  and Klein, Aaron and  Katharina Eggensperger  and Springenberg, Jost and Blum, Manuel and  Frank Hutter },
+  title = {Auto-sklearn: Efficient and Robust Automated Machine
+                  Learning},
+  pages = {113--134},
+  abstract = {The success of machine learning in a broad range of
+                  applications has led to an ever-growing demand for machine
+                  learning systems that can be used off the shelf by
+                  non-experts. To be effective in practice, such systems need
+                  to automatically choose a good algorithm and feature
+                  preprocessing steps for a new dataset at hand, and also set
+                  their respective hyperparameters. Recent work has started to
+                  tackle this automated machine learning (AutoML) problem with
+                  the help of efficient Bayesian optimization methods. Building
+                  on this, we introduce a robust new AutoML system based on the
+                  Python machine learning package scikit-learn (using 15
+                  classifiers, 14 feature preprocessing methods, and 4 data
+                  preprocessing methods, giving rise to a structured hypothesis
+                  space with 110 hyperparameters). This system, which we dub
+                  Auto-sklearn, improves on existing AutoML methods by
+                  automatically taking into account past performance on similar
+                  datasets, and by constructing ensembles from the models
+                  evaluated during the optimization. Our system won six out of
+                  ten phases of the first ChaLearn AutoML challenge, and our
+                  comprehensive analysis on over 100 diverse datasets shows
+                  that it substantially outperforms the previous state of the
+                  art in AutoML. We also demonstrate the performance gains due
+                  to each of our contributions and derive insights into the
+                  effectiveness of the individual components of Auto-sklearn.}
+}
+
+ +
+@incollection{FiaRosScho2010comp,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  title = {Comparison-based adaptive strategy selection with bandits in
+                  differential evolution},
+  author = { {\'A}lvaro Fialho  and Ros, Raymond and  Marc Schoenauer  and  Mich{\`e}le Sebag },
+  pages = {194--203}
+}
+
+ +
+@incollection{FiaSchoSeb2010fauc,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Fitness-{AUC} bandit adaptive strategy selection vs.\ the
+                  probability matching one within differential evolution: an
+                  empirical comparison on the {BBOB-2010} noiseless testbed},
+  author = { {\'A}lvaro Fialho  and  Marc Schoenauer  and  Mich{\`e}le Sebag },
+  pages = {1535--1542}
+}
+
+ +
+@incollection{FiaSchoSeb2010toward,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Toward comparison-based adaptive operator selection},
+  author = { {\'A}lvaro Fialho  and  Marc Schoenauer  and  Mich{\`e}le Sebag },
+  pages = {767--774},
+  annote = {Proposed F-AUC and F-SR}
+}
+
+ +
+@phdthesis{Fialho2010PhD,
+  title = {Adaptive operator selection for optimization},
+  author = { {\'A}lvaro Fialho },
+  year = 2010,
+  school = {Universit{\'e} Paris Sud-Paris XI}
+}
+
+ +
+@incollection{Fie2017gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  author = { Jonathan E. Fieldsend },
+  title = {University staff teaching allocation: formulating and
+                  optimising a many-objective problem},
+  pages = {1097--1104},
+  doi = {10.1145/3071178.3071230},
+  annote = {Example of NSGA-III deteriorating.}
+}
+
+ +
+@incollection{FieEve2013visualising,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  title = {Visualising high-dimensional {Pareto} relationships in
+                  two-dimensional scatterplots},
+  author = { Jonathan E. Fieldsend  and  Everson, Richard M. },
+  pages = {558--572},
+  doi = {10.1007/978-3-642-37140-0_42}
+}
+
+ +
+@incollection{Fieldsend2020data,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  doi = {10.1145/3377930},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  title = {Data structures for non-dominated sets: implementations and
+                  empirical assessment of two decades of advances},
+  author = { Jonathan E. Fieldsend },
+  booktitle = {Proceedings of the 2020 Genetic and Evolutionary Computation
+                  Conference},
+  pages = {489--497},
+  annote = {unbounded archives}
+}
+
+ +
+@incollection{FinVos2002,
+  year = 2002,
+  publisher = {Kluwer Academic Publishers, Boston, MA},
+  editor = { Stefan Vo{\ss}  and  David L. Woodruff },
+  booktitle = {Optimization Software Class Libraries},
+  author = {Andreas Fink and  Stefan Vo{\ss} },
+  title = {{HotFrame}: A Heuristic Optimization Framework},
+  pages = {81--154}
+}
+
+ +
+@incollection{FisDhaJou2015lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8994,
+  editor = {Clarisse Dhaenens and  Laetitia Jourdan  and  Marie-El{\'e}onore Marmion },
+  series = {Lecture Notes in Computer Science},
+  year = 2015,
+  booktitle = {Learning and Intelligent Optimization, 9th International
+                  Conference, LION 9},
+  author = {Benjamin Fisset and Clarisse Dhaenens and  Laetitia Jourdan },
+  title = {{MO-Mine$_\text{clust}$}: A Framework for Multi-Objective
+                  Clustering},
+  pages = {293--305},
+  keywords = {irace}
+}
+
+ +
+@incollection{FlePurLyg2005,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 3410,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  year = 2005,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  title = {Many-objective optimization: An engineering design
+                  perspective},
+  author = { Peter J. Fleming  and  Robin C. Purshouse  and Lygoe, Robert J.},
+  pages = {14--32}
+}
+
+ +
+@incollection{PurJalFle2011pref,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  title = {Preference-Driven Co-Evolutionary Algorithms Show Promise for
+                  Many-Objective optimisation},
+  author = { Robin C. Purshouse  and Jalb{\u{a}}, Cezar and  Peter J. Fleming },
+  pages = {136--150}
+}
+
+ +
+@incollection{Fleischer2003emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = {M. Fleischer},
+  title = {The Measure of {Pareto} Optima. Applications to
+                  Multi-objective Metaheuristics},
+  pages = {519--533}
+}
+
+ +
+@book{Fletcher1987,
+  author = {Fletcher, R.},
+  publisher = {John Wiley \& Sons},
+  title = {Practical methods of optimization},
+  year = 1987,
+  address = { New York, NY},
+  annote = {BFGS}
+}
+
+ +
+@incollection{FloMon1994automatic,
+  address = {Cambridge, MA},
+  year = 1994,
+  publisher = {MIT Press},
+  booktitle = {Proceedings of the third international conference on
+                  Simulation of adaptive behavior: From Animals to Animats 3},
+  editor = {Cliff, D. and Husbands, P. and Meyer, J.-A. and Wilson, S.},
+  title = {Automatic creation of an autonomous agent: Genetic evolution
+                  of a neural network driven robot},
+  author = {Floreano, Dario and Mondada, Francesco},
+  annote = {LIS-CONF-1994-003},
+  pages = {421--430}
+}
+
+ +
+@incollection{FocLabLod2002,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Filippo Focacci  and Fran{\c{c}}ois Laburthe and  Andrea Lodi },
+  title = {Local Search and Constraint Programming},
+  pages = {369--403}
+}
+
+ +
+@book{FogOweWal1966,
+  title = {Artificial Intelligence Through Simulated Evolution},
+  author = { David B. Fogel  and  Owens, Alvin J.  and  Walsh, Michael J. },
+  year = 1966,
+  publisher = {John Wiley \& Sons}
+}
+
+ +
+@book{Fogel95:EvolutionaryComputation,
+  author = { David B. Fogel },
+  title = {Evolutionary Computation. Toward a New Philosophy of
+                  Machine Intelligence},
+  journal = {Evolutionary Computation},
+  year = 1995,
+  publisher = {IEEE Press}
+}
+
+ +
+@inproceedings{FonFle1993:moga,
+  isbn = {1-55860-299-2},
+  year = 1993,
+  publisher = {Morgan Kaufmann Publishers},
+  booktitle = {Proceedings of  the Fifth International Conference on Genetic Algorithms (ICGA'93)},
+  editor = {Stephanie Forrest},
+  author = { Carlos M. Fonseca  and  Peter J. Fleming },
+  title = {Genetic Algorithms for Multiobjective Optimization:
+                  Formulation, Discussion and Generalization},
+  pages = {416--423},
+  epub = {http://eden.dei.uc.pt/~cmfonsec/fonseca-ga93-reprint.pdf},
+  annote = {Proposes MOGA and P-MOGA}
+}
+
+ +
+@incollection{FonFle1996:ppsn,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 1141,
+  editor = {H.-M. Voigt and others},
+  aeditor = {H.-M. Voigt and W. Ebeling and  Rechenberg, Ingo  and  Hans-Paul Schwefel },
+  year = 1996,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IV}},
+  author = { Carlos M. Fonseca  and  Peter J. Fleming },
+  title = {On the Performance Assessment and Comparison of
+                  Stochastic Multiobjective Optimizers},
+  pages = {584--593},
+  alias = {Fonseca96}
+}
+
+ +
+@incollection{FonFon2012,
+  booktitle = {Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7401,
+  year = 2012,
+  publisher = {Springer},
+  editor = { Jin-Kao Hao  and Legrand, Pierrick and Collet, Pierre and
+                  Monmarch{\'e}, Nicolas and Lutton, Evelyne and Schoenauer,
+                  Marc},
+  author = { Viviane {Grunert da Fonseca}  and  Carlos M. Fonseca },
+  title = {The Relationship between the Covered Fraction, Completeness
+                  and Hypervolume Indicators},
+  pages = {25--36}
+}
+
+ +
+@incollection{FonGruPaq2005:emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 3410,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  year = 2005,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  author = { Carlos M. Fonseca  and  Viviane {Grunert da Fonseca}  and  Lu{\'i}s Paquete },
+  title = {Exploring the Performance of Stochastic
+                  Multiobjective Optimisers with the Second-Order
+                  Attainment Function},
+  pages = {250--264},
+  alias = {lpaquete:15},
+  doi = {10.1007/978-3-540-31880-4_18},
+  abstract = {The attainment function has been proposed as a
+                  measure of the statistical performance of stochastic
+                  multiobjective optimisers which encompasses both the
+                  quality of individual non-dominated solutions in
+                  objective space and their spread along the trade-off
+                  surface. It has also been related to results from
+                  random closed-set theory, and cast as a mean-like,
+                  first-order moment measure of the outcomes of
+                  multiobjective optimisers. In this work, the use of
+                  more informative, second-order moment measures for
+                  the evaluation and comparison of multiobjective
+                  optimiser performance is explored experimentally,
+                  with emphasis on the interpretability of the
+                  results.}
+}
+
+ +
+@incollection{FonGueLopPaq2011emo,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  author = { Carlos M. Fonseca  and  Andreia P. Guerreiro  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete },
+  title = {On the Computation of the Empirical Attainment Function},
+  doi = {10.1007/978-3-642-19893-9_8},
+  pages = {106--120},
+  abstract = {The attainment function provides a description of the
+                  location of the distribution of a random non-dominated point
+                  set. This function can be estimated from experimental data
+                  via its empirical counterpart, the empirical attainment
+                  function (EAF). However, computation of the EAF in more than
+                  two dimensions is a non-trivial task. In this article, the
+                  problem of computing the empirical attainment function is
+                  formalised, and upper and lower bounds on the corresponding
+                  number of output points are presented. In addition, efficient
+                  algorithms for the two and three-dimensional cases are
+                  proposed, and their time complexities are related to lower
+                  bounds derived for each case.}
+}
+
+ +
+@inproceedings{FonPaqLop06:hypervolume,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = jul,
+  year = 2006,
+  booktitle = {Proceedings of  the 2006 Congress on Evolutionary Computation (CEC 2006)},
+  key = {IEEE CEC},
+  author = { Carlos M. Fonseca  and  Lu{\'i}s Paquete  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {An improved dimension-\hspace{0pt}sweep
+                  algorithm for the hypervolume indicator},
+  pages = {1157--1163},
+  doi = {10.1109/CEC.2006.1688440},
+  abstract = {This paper presents a recursive, dimension-sweep
+                  algorithm for computing the hypervolume indicator of
+                  the quality of a set of $n$ non-dominated points in
+                  $d>2$ dimensions. It improves upon the existing HSO
+                  (Hypervolume by Slicing Objectives) algorithm by
+                  pruning the recursion tree to avoid repeated
+                  dominance checks and the recalculation of partial
+                  hypervolumes. Additionally, it incorporates a recent
+                  result for the three-dimensional special case.  The
+                  proposed algorithm achieves $O(n^{d-2} \log n)$ time
+                  and linear space complexity in the worst-case, but
+                  experimental results show that the pruning
+                  techniques used may reduce the time complexity
+                  exponent even further.}
+}
+
+ +
+@incollection{FonTag2020repro,
+  author = {Fonseca Cacho, Jorge Ram{\'o}n and Taghva, Kazem},
+  title = {The State of Reproducible Research in Computer Science},
+  doi = {10.1007/978-3-030-43020-7_68},
+  series = {Advances in Intelligent Systems and Computing},
+  booktitle = {17th {International} {Conference} on {Information}
+                  {Technology}-{New} {Generations} ({ITNG} 2020)},
+  abstract = {Reproducible research is the cornerstone of cumulative
+                  science and yet is one of the most serious crisis that we
+                  face today in all fields. This paper aims to describe the
+                  ongoing reproducible research crisis along with
+                  counter-arguments of whether it really is a crisis, suggest
+                  solutions to problems limiting reproducible research along
+                  with the tools to implement such solutions by covering the
+                  latest publications involving reproducible research.},
+  language = {en},
+  publisher = {Springer International Publishing},
+  editor = {Latifi, Shahram},
+  year = 2020,
+  keywords = {Docker, Improving transparency, OCR, Open science,
+                  Replicability, Reproducibility},
+  pages = {519--524}
+}
+
+ +
+@incollection{FosBickHardKok2007,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = {Manuel F\"{o}rster and Bettina Bickel and Bernd
+                  Hardung and Gabriella K\'{o}kai},
+  title = {Self-adaptive ant colony optimisation applied to
+                  function allocation in vehicle networks},
+  pages = {1991--1998}
+}
+
+ +
+@incollection{FosHugObr2020,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  doi = {10.1145/3377930},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  title = {Do sophisticated evolutionary algorithms perform better than simple ones?},
+  author = {Foster, Michael and Hughes, Matthew and O'Brien, George and  Oliveto, Pietro S.  and Pyle, James and  Dirk Sudholt  and Williams, James},
+  pages = {184--192}
+}
+
+ +
+@book{FouGayKer2002,
+  author = {Robert Fourer and David M. Gay and Brian W. Kernighan},
+  title = {{AMPL}: A Modeling Language for Mathematical Programming},
+  publisher = {Duxbury},
+  year = 2002,
+  edition = {2nd}
+}
+
+ +
+@inproceedings{Fox1992uniting,
+  author = { Bennett L. Fox },
+  title = {Uniting probabilistic methods for optimization},
+  booktitle = {Proceedings of the 24th conference on Winter simulation},
+  pages = {500--505},
+  year = 1992,
+  organization = {ACM}
+}
+
+ +
+@incollection{Fox1995simulated,
+  author = { Bennett L. Fox },
+  title = {Simulated annealing: folklore, facts, and directions},
+  booktitle = {Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing},
+  pages = {17--48},
+  year = 1995,
+  publisher = {Springer}
+}
+
+ +
+@phdthesis{Fra2021:phd,
+  author = { Alberto Franzin },
+  title = {Empirical Analysis of Stochastic Local Search Behaviour: Connecting Structure, Components and Landscape},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2021
+}
+
+ +
+@phdthesis{Fra95:phd,
+  author = { C. B. Fraser },
+  title = {Subsequences and Supersequences of Strings},
+  school = {University of Glasgow},
+  year = 1995
+}
+
+ +
+@inproceedings{FraGyoNad2020,
+  url = {https://bnaic.liacs.leidenuniv.nl/wordpress/wp-content/uploads/bnaic2020proceedings.pdf},
+  year = 2020,
+  editor = {Cao, Lu and Kosters, Walter and Lijffijt, Jefrey},
+  booktitle = {Proceedings of the 32nd Benelux Conference on Artificial Intelligence,
+                  BNAIC 2020, Leiden, The Netherlands, 19-20 November 2020},
+  author = { Alberto Franzin  and Gyory, Rapha\"el and Nad\'e, Jean-Charles and
+                 Aubert, Guillaume and Klenkle, Georges and  Hughes Bersini },
+  title = {Phil\'{e}as: Anomaly Detection for {IoT} Monitoring},
+  pages = {56--70}
+}
+
+ +
+@book{FraLeiRui2014,
+  title = {Manufacturing Scheduling Systems: An Integrated View on
+                  Models, Methods, and Tools},
+  author = { Jose M. Frami{\~n}{\'a}n  and  Rainer Leisten  and  Rub{\'e}n Ruiz },
+  publisher = {Springer},
+  address = { New York, NY},
+  year = 2014
+}
+
+ +
+@incollection{FraStu2016:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2016},
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Exploration of Metaheuristics through Automatic Algorithm Configuration Techniques and Algorithmic Frameworks},
+  pages = {1341--1347}
+}
+
+ +
+@incollection{FraStu2017:EA,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Lutton, Evelyne and Legrand, Pierrick and Parrend, Pierre and  Nicolas Monmarch{\'e}  and  Marc Schoenauer },
+  volume = 10764,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  booktitle = {EA 2017: Artificial Evolution},
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Comparison of Acceptance Criteria in Randomized Local Searches},
+  pages = {16--29}
+}
+
+ +
+@misc{FraStu2018-supp,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Revisiting Simulated Annealing: a Component-Based Analysis: {Supplementaty} Material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2018-001}},
+  year = 2018
+}
+
+ +
+@inproceedings{FraStu2020:lmca,
+  year = 2020,
+  editor = {Vlastelica, Marin and Song, Jialin and Ferber, Aaron and Amos, Brandon and
+                  Martius, Georg and Dilkina, Bistra and Yue, Yisong},
+  booktitle = {Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020,
+                  {LMCA} 2020, Vancouver, Canada, December 12, 2020},
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Towards transferring algorithm configurations across problems},
+  pages = {1--6}
+}
+
+ +
+@incollection{FraStu2020:tailor,
+  year = 2021,
+  volume = 12641,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Trustworthy AI -- Integrating Learning, Optimization and
+                  Reasoning. TAILOR 2020},
+  editor = {Fredrik Heintz and Michela Milano and   O'Sullivan, Barry },
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {A causal framework for understanding optimisation algorithms},
+  pages = {140--145}
+}
+
+ +
+@misc{FraStu2021-supp,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {A Landscape-based Analysis of Fixed Temperature and Simulated Annealing: {Supplementaty} Material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2021-002}},
+  year = 2021
+}
+
+ +
+@inproceedings{FreFleGui2013nonparametric,
+  year = 2013,
+  publisher = {IEEE Press},
+  booktitle = {2013 IEEE International Conference on Systems, Man, and
+                  Cybernetics},
+  key = {SMC},
+  author = {A. R. R. {Freitas} and  Peter J. Fleming  and Frederico
+                  G. Guimar{\~{a}}es},
+  title = {A Non-parametric Harmony-Based Objective Reduction Method for
+                  Many-Objective Optimization},
+  pages = {651--656},
+  doi = {10.1109/SMC.2013.116}
+}
+
+ +
+@incollection{FreMer1996icec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1996,
+  editor = { Thomas B{\"a}ck  and  T. Fukuda  and  Zbigniew Michalewicz },
+  booktitle = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  author = {B. Freisleben and P. Merz},
+  title = {A Genetic Local Search Algorithm for Solving
+                  Symmetric and Asymmetric Traveling Salesman
+                  Problems},
+  pages = {616--621}
+}
+
+ +
+@incollection{FriGobQuiWag2018ppsn,
+  volume = 11101,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Tobias Friedrich  and G{\"o}bel, Andreas and Quinzan, Francesco and  Markus Wagner },
+  title = {Heavy-Tailed Mutation Operators in Single-Objective
+                  Combinatorial Optimization},
+  pages = {134--145},
+  abstract = {A core feature of evolutionary algorithms is their mutation
+                  operator. Recently, much attention has been devoted to the
+                  study of mutation operators with dynamic and non-uniform
+                  mutation rates. Following up on this line of work, we propose
+                  a new mutation operator and analyze its performance on the
+                  (1+1) Evolutionary Algorithm (EA). Our analyses show that
+                  this mutation operator competes with pre-existing ones, when
+                  used by the (1+1)-EA on classes of problems for which
+                  results on the other mutation operators are available. We
+                  present a ``jump'' function for which the performance of the
+                  (1+1)-EA using any static uniform mutation and any restart
+                  strategy can be worse than the performance of the (1+1)-EA
+                  using our mutation operator with no restarts. We show that
+                  the (1+1)-EA using our mutation operator finds a
+                  (1/3)-approximation ratio on any non-negative submodular
+                  function in polynomial time. This performance matches that of
+                  combinatorial local search algorithms specifically designed
+                  to solve this problem.}
+}
+
+ +
+@incollection{FriKotKre2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = { Tobias Friedrich  and K{\"o}tzing, Timo  and Krejca, Martin S. and  Andrew M. Sutton },
+  title = {Robustness of Ant Colony Optimization to Noise},
+  pages = {17--24},
+  numpages = 8,
+  doi = {10.1145/2739480.2754723},
+  keywords = {ant colony optimization, noisy fitness, run time analysis,
+                  theory}
+}
+
+ +
+@incollection{FriKotWag2017:aaai,
+  publisher = {{AAAI} Press},
+  month = feb,
+  year = 2017,
+  editor = {Satinder P. Singh and Shaul Markovitch},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  author = { Tobias Friedrich  and K{\"o}tzing, Timo  and  Markus Wagner },
+  title = {A Generic Bet-and-Run Strategy for Speeding Up Stochastic Local Search},
+  pages = {801--807}
+}
+
+ +
+@incollection{FriQuiWag2018mutation,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2018,
+  editor = { Aguirre, Hern\'{a}n E.  and Keiki Takadama},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2018},
+  author = { Tobias Friedrich  and Quinzan, Francesco and  Markus Wagner },
+  title = {Escaping Large Deceptive Basins of Attraction with
+                  Heavy-tailed Mutation Operators},
+  pages = {293--300},
+  numpages = {8},
+  doi = {10.1145/3205455.3205515},
+  acmid = {3205515},
+  keywords = {combinatorial optimization, heavy-tailed mutation,
+                  single-objective optimization, experiments-motivated theory,
+                  irace}
+}
+
+ +
+@inproceedings{Friendly1991stat,
+  title = {Statistical graphics for multivariate data},
+  author = {Friendly, Michael},
+  year = 1991,
+  booktitle = {SAS Conference Proceedings: SAS Users Group International 16
+                  (SUGI 16)},
+  annote = {February 17-20, 1991, New Orleans, Louisiana, 297 papers}
+}
+
+ +
+@book{FudTir83,
+  author = {Fudenberg, D. and Tirole, J.},
+  year = 1983,
+  title = {Game Theory},
+  publisher = {MIT Press, Cambridge, MA}
+}
+
+ +
+@incollection{FujNan2021solving,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  title = {Solving {QUBO} with {GPU} parallel {MOPSO}},
+  author = {Fujimoto, Noriyuki and Nanai, Kouki},
+  pages = {1788--1794}
+}
+
+ +
+@incollection{Fuk2004gecco,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3103,
+  editor = { Kalyanmoy Deb and others},
+  year = 2004,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2004, Part II},
+  title = {Evolving Local Search Heuristics for {SAT} Using
+                  Genetic Programming},
+  author = { Fukunaga, Alex S. },
+  abstract = {Satisfiability testing ({SAT)} is a very active area
+                  of research today, with numerous real-world
+                  applications. We describe {CLASS2.0}, a genetic
+                  programming system for semi-automatically designing
+                  {SAT} local search heuristics. An empirical
+                  comparison shows that that the heuristics generated
+                  by our {GP} system outperform the state of the art
+                  human-designed local search algorithms, as well as
+                  previously proposed evolutionary approaches, with
+                  respect to both runtime as well as search efficiency
+                  (number of variable flips to solve a problem).},
+  pages = {483--494}
+}
+
+ +
+@book{FurLovLov2000stats,
+  author = {Nancy E. Furlong and Eugene A. Lovelace and Kristin
+                  L. Lovelace},
+  title = {Research Methods and Statistics: An Integrated
+                  Approach},
+  publisher = {Harcourt College Publishers},
+  year = 2000,
+  alias = {furlong00:stats}
+}
+
+ +
+@incollection{GaeCla04,
+  isbn = {1-932415-66-1},
+  year = 2005,
+  publisher = {CSREA Press},
+  booktitle = {Proceedings of the 2005 International Conference on Artificial Intelligence, ICAI 2005},
+  editor = {Hamid R. Arabnia and Rose Joshua},
+  author = {D. Gaertner and K. Clark},
+  title = {On Optimal Parameters for Ant Colony Optimization
+                  Algorithms},
+  pages = {83--89}
+}
+
+ +
+@incollection{GagLeg2010emaa,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = { Matteo Gagliolo  and Catherine Legrand},
+  title = {Algorithm Survival Analysis},
+  pages = {161--184},
+  doi = {10.1007/978-3-642-02538-9_7},
+  abstract = {Algorithm selection is typically based on models of
+                  algorithm performance,learned during a separate
+                  offline training sequence, which can be
+                  prohibitively expensive. In recent work, we adopted
+                  an online approach, in which models of the runtime
+                  distributions of the available algorithms are
+                  iteratively updated and used to guide the allocation
+                  of computational resources, while solving a sequence
+                  of problem instances. The models are estimated using
+                  survival analysis techniques, which allow us to
+                  reduce computation time, censoring the runtimes of
+                  the slower algorithms. Here, we review the
+                  statistical aspects of our online selection method,
+                  discussing the bias induced in the runtime
+                  distributions (RTD) models by the competition of
+                  different algorithms on the same problem instances.}
+}
+
+ +
+@incollection{GamDor95:ml,
+  publisher = {Morgan Kaufmann Publishers, Palo Alto, CA},
+  editor = {A. Prieditis and S. Russell},
+  booktitle = {Proceedings of the Twelfth International Conference
+                  on Machine Learning (ML-95)},
+  year = 1995,
+  author = { L. M. Gambardella  and  Marco Dorigo },
+  title = {Ant-{Q}: A Reinforcement Learning Approach to the
+                  Traveling Salesman Problem},
+  pages = {252--260}
+}
+
+ +
+@incollection{GamDor96:icec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1996,
+  editor = { Thomas B{\"a}ck  and  T. Fukuda  and  Zbigniew Michalewicz },
+  booktitle = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  author = { L. M. Gambardella  and  Marco Dorigo },
+  title = {Solving Symmetric and Asymmetric {TSP}s by Ant
+                  Colonies},
+  pages = {622--627},
+  anote = {IC.18}
+}
+
+ +
+@incollection{GamTaiAga99,
+  address = {London, UK},
+  year = 1999,
+  publisher = {McGraw Hill},
+  editor = { David Corne  and  Marco Dorigo  and  Fred Glover },
+  booktitle = {New Ideas in Optimization},
+  author = { L. M. Gambardella  and  {\'E}ric D. Taillard  and  G. Agazzi },
+  title = {{MACS-VRPTW}: A Multiple Ant Colony System for Vehicle
+                  Routing Problems with Time Windows},
+  pages = {63--76},
+  alias = {GamTaiAga99:newideas}
+}
+
+ +
+@incollection{GanDelKin04:ants2004,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Xavier Gandibleux  and  X. Delorme  and  V. {T'Kindt} },
+  title = {An Ant Colony Optimisation Algorithm for the Set
+                  Packing Problem},
+  pages = {49--60}
+}
+
+ +
+@incollection{GanMezFre1997,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  volume = 455,
+  editor = {R. Caballero and  Francisco Ruiz  and R. Steuer},
+  year = 1997,
+  booktitle = {Advances in Multiple Objective and Goal Programming},
+  title = {A tabu search procedure to solve multiobjective
+                  combinatorial optimization problem},
+  author = { Xavier Gandibleux  and Mezdaoui, N. and Fr{\'e}ville, A.},
+  pages = {291--300}
+}
+
+ +
+@incollection{GanMorKat2003,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  title = {Use of a genetic heritage for solving the assignment
+                  problem with two objectives},
+  author = { Xavier Gandibleux  and  H. Morita  and  Katoh, N. },
+  pages = {43--57}
+}
+
+ +
+@inproceedings{GaoNieLi2019vis,
+  year = 2019,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2019 Congress on Evolutionary Computation (CEC 2019)},
+  key = {IEEE CEC},
+  author = {Gao, Huiru and Nie, Haifeng and Li, Ke},
+  title = {Visualisation of {Pareto} Front Approximation: A Short Survey
+                  and Empirical Comparisons},
+  pages = {1750--1757},
+  doi = {10.1109/CEC.2019.8790298}
+}
+
+ +
+@incollection{GarDas2008,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2008,
+  editor = { Vittorio Maniezzo  and  Roberto Battiti  and Jean-Paul Watson},
+  volume = 5313,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Learning and Intelligent Optimization, Second International Conference, LION 2},
+  title = {Multiobjective landscape analysis and the generalized assignment problem},
+  author = {Garrett, Deon and Dasgupta, Dipankar},
+  pages = {110--124}
+}
+
+ +
+@book{GarJoh1979,
+  title = {Computers and Intractability: A Guide to the Theory
+                  of {NP}-Completeness},
+  author = {Garey, M. R. and David S. Johnson},
+  publisher = {Freeman \& Co, San Francisco, CA},
+  year = 1979
+}
+
+ +
+@incollection{GarLopGod2016pso,
+  volume = 9882,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and  Mauro Birattari  and  Li, Xiaodong  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kazuhiro Ohkura  and  Carlo Pinciroli  and  Thomas St{\"u}tzle },
+  year = 2016,
+  booktitle = {Swarm Intelligence, 10th International Conference, ANTS 2016},
+  title = {A study of archiving strategies in multi-objective {PSO} for
+                  molecular docking},
+  author = { Jos{\'e} Garc{\'i}a-Nieto  and L{\'o}pez-Camacho, Esteban and Godoy
+                  Garc{\'i}a, Mar{\'i}a Jes{\'u}s and  Nebro, Antonio J.  and  Durillo, Juan J.  and Aldana-Montes, Jos{\'e} F.},
+  pages = {40--52},
+  doi = {10.1007/978-3-319-44427-7_4}
+}
+
+ +
+@inproceedings{GarSosVaz07,
+  author = {Beatriz A. Garro and Humberto Sossa and Roberto
+                  A. Vazquez},
+  title = {Evolving ant colony system for optimizing path
+                  planning in mobile robots},
+  booktitle = {Electronics, Robotics and Automotive Mechanics
+                  Conference},
+  year = 2007,
+  pages = {444--449},
+  doi = {10.1109/CERMA.2007.60},
+  publisher = {IEEE Computer Society},
+  address = {Los Alamitos, CA}
+}
+
+ +
+@incollection{GasScha07:easysyn,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4638,
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  year = 2007,
+  booktitle = {Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS~2007},
+  author = {Luca {Di Gaspero} and Andrea Schaerf},
+  title = {Easysyn++: A tool for automatic synthesis of stochastic local
+                  search algorithms},
+  pages = {177--181}
+}
+
+ +
+@incollection{GebKamKauSchSchZil2013claspfolio,
+  editor = {Pedro Calabar and Tran Cao Son},
+  volume = 8148,
+  year = 2013,
+  series = {Lecture Notes in Artificial Intelligence},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Logic Programming and Nonmonotonic Reasoning},
+  title = {A portfolio solver for answer set programming: Preliminary
+                  report},
+  author = {Gebser, Martin and Kaminski, Roland and Kaufmann, Benjamin
+                  and Schaub, Torsten and Schneider, Marius Thomas and Ziller,
+                  Stefan},
+  pages = {352--357}
+}
+
+ +
+@inproceedings{Gei2006ecml,
+  isbn = {978-3-540-46056-5},
+  volume = 4212,
+  series = {Lecture Notes in Computer Science},
+  year = 2006,
+  booktitle = {Machine Learning: ECML 2006},
+  editor = {F{\"u}rnkranz, Johannes and Scheffer, Tobias and
+                  Spiliopoulou, Myra},
+  author = {Geibel, Peter},
+  title = {Reinforcement Learning for {MDPs} with Constraints},
+  pages = {646--653},
+  doi = {10.1007/11871842_63},
+  abstract = {In this article, I will consider Markov Decision Processes
+                  with two criteria, each defined as the expected value of an
+                  infinite horizon cumulative return. The second criterion is
+                  either itself subject to an inequality constraint, or there
+                  is maximum allowable probability that the single returns
+                  violate the constraint. I describe and discuss three new
+                  reinforcement learning approaches for solving such control
+                  problems.},
+  keywords = {Safe RL}
+}
+
+ +
+@techreport{GenGraMac1997hownotto,
+  author = { Ian P. Gent  and Stuart A. Grant and Ewen MacIntyre and Patrick
+                  Prosser and Paul Shaw and Barbara M. Smith and Toby Walsh},
+  title = {How Not To Do It},
+  institution = {School of Computer Studies, University of Leeds},
+  year = 1997,
+  number = {97.27},
+  month = may,
+  abstract = {We give some dos and don'ts for those analysing algorithms
+                  experimentally. We illustrate these with many examples from
+                  our own research on the study of algorithms for NP-complete
+                  problems such as satisfiability and constraint
+                  satisfaction. Where we have not followed these maxims, we
+                  have suffered as a result.}
+}
+
+ +
+@inproceedings{GenHooProWal99,
+  author = { Ian P. Gent  and  Holger H. Hoos  and P. Prosser and T. Walsh},
+  title = {Morphing: Combining Structure and Randomness},
+  booktitle = {Proceedings of  the Sixteenth National Conference on Artificial Intelligence},
+  pages = {654--660},
+  year = 1999
+}
+
+ +
+@incollection{GenPot2010:handbook,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Michel Gendreau  and  Jean-Yves Potvin },
+  title = {Tabu Search},
+  pages = {41--59}
+}
+
+ +
+@incollection{GesHutKotMal2014lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8426,
+  editor = { Panos M. Pardalos  and  Mauricio G. C. Resende  and Chrysafis Vogiatzis and Jose
+                  L. Walteros},
+  series = {Lecture Notes in Computer Science},
+  year = 2014,
+  booktitle = {Learning and Intelligent Optimization, 8th International Conference, LION 8},
+  author = {Daniel Geschwender and  Frank Hutter  and Kotthoff, Lars and  Yuri Malitsky  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Algorithm Configuration in the Cloud: A Feasibility Study},
+  pages = {41--46},
+  doi = {10.1007/978-3-319-09584-4_5}
+}
+
+ +
+@inproceedings{Gibbs05:cal,
+  author = { Matthew S. Gibbs  and  Graeme C. Dandy  and  Holger R. Maier  and  John B. Nixon },
+  title = {Calibrating genetic algorithms for water
+                  distribution system optimisation},
+  booktitle = {7th Annual Symposium on Water Distribution Systems
+                  Analysis},
+  year = 2005,
+  month = may,
+  organization = {ASCE}
+}
+
+ +
+@incollection{Gla2017fast,
+  editor = {Heike Trautmann and G{\"{u}}nter Rudolph and Kathrin Klamroth
+                  and Oliver Sch{\"{u}}tze and Margaret M. Wiecek and Yaochu
+                  Jin and Christian Grimme},
+  year = 2017,
+  volume = 10173,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2017},
+  author = { T. Glasmachers },
+  title = {A fast incremental {BSP} tree archive for non-dominated
+                  points},
+  pages = {252--266},
+  keywords = {archiving}
+}
+
+ +
+@incollection{Glo98,
+  volume = 1363,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  shorteditor = { Jin-Kao Hao  and others},
+  editor = { Jin-Kao Hao  and Evelyne Lutton and Edmund M. A. Ronald and  Marc Schoenauer  and Dominique Snyers},
+  booktitle = {Artificial Evolution},
+  author = { Fred Glover },
+  title = {A Template for Scatter Search and Path Relinking},
+  doi = {10.1007/BFb0026589},
+  pages = {1--51},
+  year = 1998
+}
+
+ +
+@inproceedings{GloBen2010,
+  title = {Understanding the difficulty of training deep feedforward
+                  neural networks},
+  author = {Glorot, Xavier and  Bengio, Yoshua },
+  booktitle = {Proceedings of the Thirteenth International Conference on
+                  Artificial Intelligence and Statistics},
+  pages = {249--256},
+  year = 2010
+}
+
+ +
+@incollection{GloKoc1996:mic,
+  author = { Fred Glover  and  Gary A. Kochenberger },
+  title = {Critical Even Tabu Search for Multidimensional Knapsack
+                  Problems},
+  booktitle = {Metaheuristics: Theory \& Applications},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 1996,
+  editor = { Ibrahim H. Osman  and James P. Kelly},
+  pages = {407--427}
+}
+
+ +
+@book{GloLag97,
+  author = { Fred Glover  and  Manuel Laguna },
+  title = {Tabu Search},
+  publisher = {Kluwer Academic Publishers},
+  address = {Boston, MA, USA},
+  year = 1997
+}
+
+ +
+@incollection{GloLagMar2002:mh,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Fred Glover  and  Manuel Laguna  and  Rafael Mart{\'i} },
+  title = {Scatter Search and Path Relinking: Advances and Applications},
+  pages = {1--35}
+}
+
+ +
+@incollection{GolSolMoi2017gvizier,
+  key = {SIGKDD},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = {Stan Matwin and Shipeng Yu and Faisal Farooq},
+  booktitle = {23rd {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  author = {Daniel Golovin and Benjamin Solnik and Subhodeep Moitra and
+                  Greg Kochanski and John Karro and D. Sculley},
+  title = {{Google} {Vizier}: {A} Service for Black-Box Optimization},
+  pages = {1487--1495},
+  doi = {10.1145/3097983.3098043}
+}
+
+ +
+@incollection{GolSouGol2006:pso,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2006,
+  editor = {Gottlieb, Jens and  G{\"u}nther R. Raidl },
+  volume = 3906,
+  booktitle = {Proceedings of EvoCOP 2006 -- 6th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  title = {Particle Swarm for the Traveling Salesman Problem},
+  author = { Goldbarg, Elizabeth Ferreira Gouv{\^e}a  and Souza, Givanaldo R. and  Goldbarg, Marco Cesar },
+  pages = {99--110}
+}
+
+ +
+@book{Goldberg89,
+  author = { David E. Goldberg },
+  title = {Genetic Algorithms in Search, Optimization and
+                  Machine Learning},
+  publisher = {Addison-Wesley},
+  address = { Boston, MA},
+  year = 1989
+}
+
+ +
+@inproceedings{GoldmanMays,
+  author = { Fred E. Goldman  and  Larry W. Mays },
+  title = {The Application of Simulated Annealing to the
+                  Optimal Operation of Water Systems},
+  booktitle = {Proceedings of 26th Annual Water Resources Planning
+                  and Management Conference},
+  year = 2000,
+  address = {Tempe, USA},
+  month = jun,
+  organization = {ASCE}
+}
+
+ +
+@incollection{Gomory1963,
+  year = 1963,
+  publisher = {McGraw Hill,  New York, NY},
+  editor = {Graves, R.L. and Wolfe, P.},
+  booktitle = {Recent Advances in Mathematical Programming},
+  title = {An algorithm for integer solutions to linear programs},
+  author = {Gomory, Ralph E.},
+  pages = {260--302}
+}
+
+ +
+@incollection{GonFiaCai2010adaptive,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Adaptive strategy selection in differential evolution},
+  author = {Gong, Wenyin and  {\'A}lvaro Fialho  and Cai, Zhihua},
+  pages = {409--416},
+  doi = {10.1145/1830483.1830559}
+}
+
+ +
+@incollection{Gor1997,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1997,
+  editor = { Thomas B{\"a}ck  and  Zbigniew Michalewicz  and  Xin Yao },
+  booktitle = {Proceedings of the 1997 IEEE International
+                  Conference on Evolutionary Computation (ICEC'97)},
+  author = {M. Gorges-Schleuter},
+  title = {Asparagos96 and the {Travelling} {Salesman} {Problem}},
+  pages = {171--174}
+}
+
+ +
+@incollection{GotPucSol03:evoworkshops,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2611,
+  editor = {S. Cagnoni and others},
+  aeditor = {S. Cagnoni and J. J. {Romero Cardalda} and D. W. Corne
+                  and J. Gottlieb and A. Guillot and E. Hart and
+                  C. G. Johnson and E. Marchiori and J.-A. Meyer and
+                  M. Middendorf and G. R. Raidl},
+  year = 2003,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of EvoWorkshops 2003},
+  author = {J. Gottlieb and M. Puchta and  Christine Solnon },
+  title = {A Study of Greedy, Local Search, and Ant Colony
+                  Optimization Approaches for Car Sequencing Problems},
+  pages = {246--257}
+}
+
+ +
+@inproceedings{GraDej1992composer,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  editor = {William R. Swartout},
+  year = 1992,
+  booktitle = {Proceedings of  the 10th National Conference on Artificial Intelligence},
+  author = {Jonathan Gratch and DeJong, Gerald},
+  title = {{COMPOSER}: {A} probabilistic solution to the utility problem
+                  in speed-up learning},
+  pages = {235--240},
+  annote = {Eearliest hyper-heuristic?}
+}
+
+ +
+@inproceedings{GraMohHin2013speech,
+  title = {Speech recognition with deep recurrent neural networks},
+  author = {Graves, Alex and Mohamed, Abdel-rahman and Hinton, Geoffrey},
+  booktitle = {Acoustics, speech and signal processing (icassp), 2013 ieee
+                  international conference on},
+  pages = {6645--6649},
+  year = 2013,
+  organization = {IEEE}
+}
+
+ +
+@incollection{GreHuDam1996foga,
+  publisher = {Morgan Kaufmann Publishers},
+  year = 1996,
+  editor = {Richard K. Belew and Michael D. Vose},
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  title = {Fitness functions for multiple objective optimization
+                  problems: Combining preferences with {Pareto} rankings},
+  author = {Greenwood, Garrison W. and Hu, Xiaobo and D'Ambrosio,
+                  Joseph G.},
+  pages = {437--455}
+}
+
+ +
+@inproceedings{GreMatSlo2010cec,
+  year = 2010,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2010 Congress on Evolutionary Computation (CEC 2010)},
+  editor = { Ishibuchi, Hisao  and others},
+  key = {IEEE CEC},
+  title = {Interactive evolutionary multiobjective optimization using
+                  dominance-based rough set approach},
+  author = { Salvatore Greco  and Matarazzo, Benedetto and  Roman S{\l}owi{\'n}ski },
+  pages = {1--8}
+}
+
+ +
+@techreport{GruFon2004tr,
+  author = { Viviane {Grunert da Fonseca}  and  Carlos M. Fonseca },
+  year = 2004,
+  title = {A characterization of the outcomes of stochastic
+                  multiobjective optimizers through a reduction of the hitting
+                  function test sets},
+  institution = {CSI, Universidade do Algarve},
+  keywords = {high-order EAF}
+}
+
+ +
+@incollection{GruFon2009:emaa,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = { Viviane {Grunert da Fonseca}  and  Carlos M. Fonseca },
+  title = {The Attainment-Function Approach to Stochastic
+                  Multiobjective Optimizer Assessment and Comparison},
+  pages = {103--130}
+}
+
+ +
+@incollection{Grunert01,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2001},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 1993,
+  year = 2001,
+  publisher = {Springer},
+  editor = { Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele  and  Carlos A. {Coello Coello}  and  David Corne },
+  author = { Viviane {Grunert da Fonseca}  and  Carlos M. Fonseca  and  Andreia O. Hall },
+  key = {Fonseca et al., 2001},
+  title = {Inferential Performance Assessment of Stochastic Optimisers
+                  and the Attainment Function},
+  pages = {213--225},
+  alias = {Fonseca01},
+  doi = {10.1007/3-540-44719-9_15},
+  annote = {Proposed looking at anytime behavior as a multi-objective
+                  problem},
+  keywords = {EAF},
+  abstract = {The performance of stochastic optimisers can be assessed
+                  experimentally on given problems by performing multiple
+                  optimisation runs, and analysing the results. Since an
+                  optimiser may be viewed as an estimator for the (Pareto)
+                  minimum of a (vector) function, stochastic optimiser
+                  performance is discussed in the light of the criteria
+                  applicable to more usual statistical
+                  estimators. Multiobjective optimisers are shown to deviate
+                  considerably from standard point estimators, and to require
+                  special statistical methodology. The attainment function is
+                  formulated, and related results from random closed-set theory
+                  are presented, which cast the attainment function as a
+                  mean-like measure for the outcomes of multiobjective
+                  optimisers. Finally, a covariance-measure is defined, which
+                  should bring additional insight into the stochastic behaviour
+                  of multiobjective optimisers. Computational issues and
+                  directions for further work are discussed at the end of the
+                  paper.}
+}
+
+ +
+@incollection{GueMonSli04:ants2004,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = {C. Gu{\'e}ret  and  Nicolas Monmarch{\'e}  and M. Slimane},
+  title = {Ants Can Play Music},
+  pages = {310--317}
+}
+
+ +
+@incollection{GunBra2003:evoworkshops,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2611,
+  editor = {S. Cagnoni and others},
+  aeditor = {S. Cagnoni and J. J. {Romero Cardalda} and D. W. Corne
+                  and J. Gottlieb and A. Guillot and E. Hart and
+                  C. G. Johnson and E. Marchiori and J.-A. Meyer and
+                  M. Middendorf and G. R. Raidl},
+  year = 2003,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of EvoWorkshops 2003},
+  author = { M. Guntsch and  J{\"u}rgen Branke },
+  title = {New Ideas for Applying Ant Colony Optimization to the Probabilistic TSP},
+  pages = {165--175}
+}
+
+ +
+@incollection{GunMid01:evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2037,
+  editor = {E. J. W. Boers and others},
+  aeditor = {E. J. W. Boers and J. Gottlieb and P. L. Lanzi and R. E. Smith
+                 and S. Cagnoni and E. Hart and G. R. Raidl and H. Tijink},
+  year = 2001,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2001},
+  author = { M. Guntsch and  Martin Middendorf },
+  title = {Pheromone Modification Strategies for Ant Algorithms
+                  Applied to Dynamic {TSP}},
+  pages = {213--222},
+  anote = {Also available as Tech. Rep. AIDA-00-07,
+                  Intellectics Group, Darmstadt University of
+                  Technology, Germany.},
+  alias = {GunMid01:evoworkshops}
+}
+
+ +
+@incollection{GunMid02:EvoWorkshops,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2279,
+  editor = {S. Cagnoni and others},
+  aeditor = {S. Cagnoni and J. Gottlieb and E. Hart and
+                 M. Middendorf and G. R. Raidl},
+  year = 2002,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2002},
+  author = { M. Guntsch and  Martin Middendorf },
+  title = {A Population Based Approach for {ACO}},
+  pages = {71--80}
+}
+
+ +
+@incollection{GunMid03:emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { M. Guntsch and  Martin Middendorf },
+  title = {Solving Multi-Objective Permutation Problems with
+                  Population Based {ACO}},
+  pages = {464--478}
+}
+
+ +
+@incollection{GunMid2002:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { M. Guntsch and  Martin Middendorf },
+  title = {Applying Population Based {ACO} to Dynamic Optimization
+                  Problems},
+  pages = {111--122}
+}
+
+ +
+@misc{Gurobi,
+  author = {Gurobi},
+  title = {Gurobi Optimizer},
+  howpublished = {\url{http://www.gurobi.com/products/gurobi-optimizer}},
+  year = 2017
+}
+
+ +
+@incollection{Gus97:sequence-algorithms,
+  author = { D. Gusfield },
+  title = {Algorithms on Strings, Trees, and Sequences},
+  booktitle = {Computer Science and Computational Biology},
+  publisher = {Cambridge University Press},
+  year = 1997
+}
+
+ +
+@incollection{Gut04:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Gutjahr, Walter J. },
+  title = {{S-ACO}: An Ant-Based Approach to Combinatorial
+                  Optimization Under Uncertainty},
+  pages = {238--249}
+}
+
+ +
+@inproceedings{Gut2003:saga,
+  doi = {10.1007/b13596},
+  series = {Lecture Notes in Computer Science},
+  year = 2003,
+  volume = 2827,
+  publisher = {Springer Verlag},
+  editor = {Andreas Albrecht and Kathleen Steinh\"{o}fel},
+  booktitle = {Stochastic Algorithms: Foundations and Applications},
+  author = { Gutjahr, Walter J. },
+  title = {A converging {ACO} algorithm for stochastic combinatorial optimization},
+  pages = {10--25}
+}
+
+ +
+@incollection{HaasAttEib2011racing,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  title = {Racing to improve on-line, on-board evolutionary robotics},
+  author = {Haasdijk, Evert and Atta-ul-Qayyum, Arif and  Agoston E. Eiben },
+  pages = {187--194}
+}
+
+ +
+@incollection{HacFisZecTei08:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  editor = {Conor Ryan},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  author = { S. H{\"a}ckel  and  M. Fischer  and  D. Zechel  and  T. Teich },
+  title = {A multi-objective ant colony approach for {Pareto}-optimization
+                 using dynamic programming},
+  pages = {33--40}
+}
+
+ +
+@inproceedings{HadReeSim2012ec,
+  year = 2012,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2012 Congress on Evolutionary Computation (CEC 2012)},
+  key = {IEEE CEC},
+  author = { David Hadka  and  Patrick M. Reed  and T. W. Simpson},
+  title = {Diagnostic assessment of the {Borg} {MOEA} for many-objective
+                  product family design problems},
+  pages = {1--10}
+}
+
+ +
+@misc{Hadoop,
+  author = {{Apache Software Foundation}},
+  title = {Hadoop},
+  url = {https://hadoop.apache.org},
+  year = 2008
+}
+
+ +
+@book{HaestadBook03,
+  author = { Thomas M. Walski  and  Donald V. Chase  and  Dragan A. Savic  and  Walter Grayman  and  Stephen Beckwith  and  Edmundo Koelle },
+  title = {Advanced Water Distribution Modeling and Management},
+  publisher = {Haestad Methods, Inc., Haestad Press},
+  year = 2003,
+  edition = {1st}
+}
+
+ +
+@incollection{HalOliSud2019cutoff,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = { George T. Hall  and  Oliveto, Pietro S.  and  Dirk Sudholt },
+  title = {On the impact of the cutoff time on the performance of
+                  algorithm configurators},
+  pages = {907--915},
+  doi = {10.1145/3321707.3321879},
+  keywords = {theory, automatic configuration, capping}
+}
+
+ +
+@incollection{HalOliSud2020fast,
+  volume = 12269,
+  year = 2020,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas B{\"a}ck  and  Mike Preuss  and Deutz, Andr{\'e} and Wang, Hao and  Carola Doerr  and  Emmerich, Michael T. M.  and  Heike Trautmann },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVI}},
+  title = {Fast Perturbative Algorithm Configurators},
+  author = { George T. Hall  and  Oliveto, Pietro S.  and  Dirk Sudholt },
+  doi = {10.1007/978-3-030-58112-1_2},
+  pages = {19--32}
+}
+
+ +
+@incollection{HalOliSud2020gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  author = {George T. Hall and  Oliveto, Pietro S.  and  Dirk Sudholt },
+  title = {Analysis of the performance of algorithm configurators for
+                  search heuristics with global mutation operators},
+  pages = {823--831},
+  doi = {10.1145/3377930.3390218}
+}
+
+ +
+@incollection{HamElk2003gmeans,
+  publisher = {MIT Press},
+  editor = {S. Thrun and L. Saul and B. Sch\"{o}lkopf},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 16)},
+  year = 2003,
+  title = {Learning the k in k-means},
+  author = {Hamerly, Greg and Elkan, Charles},
+  epub = {https://proceedings.neurips.cc/paper/2003/file/234833147b97bb6aed53a8f4f1c7a7d8-Paper.pdf}
+}
+
+ +
+@incollection{HamStu2017,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  doi = {10.1007/978-3-319-55453-2},
+  volume = 10197,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  booktitle = {Proceedings of EvoCOP 2017 -- 17th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Bin Hu  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  author = { Hayfa Hammami  and  Thomas St{\"u}tzle },
+  title = {A Computational Study of Neighborhood Operators for Job-Shop
+                  Scheduling Problems with Regular Objectives},
+  pages = {1--17}
+}
+
+ +
+@incollection{Han1997mots,
+  publisher = {Springer Verlag},
+  editor = {J. Climaco},
+  year = 1997,
+  booktitle = {Proceedings of the 13th International Conference on
+                  Multiple Criteria Decision Making (MCDM'97)},
+  title = {Tabu search for multiobjective optimization: {MOTS}},
+  author = { Michael Pilegaard Hansen },
+  pages = {574--586},
+  alias = {Han1997}
+}
+
+ +
+@misc{HanAkiBau2019pycma,
+  author = { Nikolaus Hansen  and Youhei Akimoto and Petr Baudis},
+  title = {{CMA-ES/pycma} on {Github}},
+  howpublished = {Zenodo},
+  month = feb,
+  year = 2019,
+  doi = {10.5281/zenodo.2559634}
+}
+
+ +
+@techreport{HanAugFin2009bbob_setup,
+  author = { Nikolaus Hansen  and  Anne Auger  and Finck, S. and Ros, R.},
+  title = {Real-Parameter Black-Box Optimization Benchmarking 2009:
+                  Experimental setup},
+  institution = {INRIA, France},
+  year = 2009,
+  number = {RR-6828},
+  supplement = {http://coco.gforge.inria.fr/bbob2012-downloads}
+}
+
+ +
+@incollection{HanAugRosFin2010comparing,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Comparing Results of 31 Algorithms from the Black-Box
+                  Optimization Benchmarking {BBOB-2009}},
+  author = { Nikolaus Hansen  and  Anne Auger  and Ros, Raymond and Finck,
+                  Steffen and  Po{\v{s}}{\'i}k, Petr },
+  pages = {1689--1696},
+  doi = {10.1145/1830761.1830790},
+  abstract = {This paper presents results of the BBOB-2009 benchmarking of
+                  31 search algorithms on 24 noiseless functions in a black-box
+                  optimization scenario in continuous domain. The runtime of
+                  the algorithms, measured in number of function evaluations,
+                  is investigated and a connection between a single convergence
+                  graph and the runtime distribution is uncovered. Performance
+                  is investigated for different dimensions up to 40-D, for
+                  different target precision values, and in different subgroups
+                  of functions. Searching in larger dimension and multi-modal
+                  functions appears to be more difficult. The choice of the
+                  best algorithm also depends remarkably on the available
+                  budget of function evaluations.},
+  keywords = {benchmarking, black-box optimization}
+}
+
+ +
+@techreport{HanFinRosAug2009bbob,
+  author = { Nikolaus Hansen  and Finck, Steffen and Ros, Raymond and  Anne Auger },
+  title = {Real-Parameter Black-Box Optimization Benchmarking 2009:
+                  Noiseless Functions Definitions},
+  institution = {INRIA, France},
+  year = 2009,
+  number = {RR-6829},
+  note = {Updated February 2010},
+  annote = {\url{http://coco.gforge.inria.fr/bbob2012-downloads}},
+  epub = {https://hal.inria.fr/inria-00362633/document}
+}
+
+ +
+@techreport{HanJas1998,
+  author = { Michael Pilegaard Hansen  and  Andrzej Jaszkiewicz },
+  title = {Evaluating the quality of approximations to the non-dominated
+                  set},
+  institution = {Institute of Mathematical Modelling, Technical University of
+                  Denmark},
+  year = 1998,
+  number = {IMM-REP-1998-7},
+  address = {Lyngby, Denmark},
+  annote = {Proposed R2 indicator}
+}
+
+ +
+@incollection{HanKno2008mpsn,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  series = {Natural Computing Series},
+  editor = { Joshua D. Knowles  and  David Corne  and  Kalyanmoy Deb  and Chair, Deva Raj},
+  year = 2008,
+  booktitle = {Multiobjective Problem Solving from Nature},
+  author = { Julia Handl  and  Joshua D. Knowles },
+  title = {Modes of Problem Solving with Multiple Objectives:
+                  Implications for Interpreting the {Pareto} Set and
+                  for Decision Making},
+  doi = {10.1007/978-3-540-72964-8_7},
+  pages = {131--151}
+}
+
+ +
+@incollection{HanMla02:mh,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Pierre Hansen  and  Nenad Mladenovi{\'c} },
+  title = {Variable Neighborhood Search},
+  pages = {145--184}
+}
+
+ +
+@incollection{HanMlaBriPer2010:handbook,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  title = {Variable {Neighborhood} {Search}},
+  author = { Pierre Hansen  and  Nenad Mladenovi{\'c}  and Jack Brimberg and Jos{\'e} A. Moreno P{\'e}rez},
+  pages = {61--86}
+}
+
+ +
+@incollection{HanOst1996cma,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1996,
+  editor = { Thomas B{\"a}ck  and  T. Fukuda  and  Zbigniew Michalewicz },
+  booktitle = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  author = { Nikolaus Hansen  and Ostermeier, Andreas},
+  title = {Adapting Arbitrary Normal Mutation Distributions in Evolution
+                  Strategies: The Covariance Matrix Adaptation},
+  pages = {312--317},
+  annote = {Proposed CMA-ES},
+  shorttitle = {Adapting Arbitrary Normal Mutation Distributions in Evolution
+                  Strategies},
+  doi = {10.1109/ICEC.1996.542381},
+  abstract = {A new formulation for coordinate system independent
+                  adaptation of arbitrary normal mutation distributions with
+                  zero mean is presented. This enables the evolution strategy
+                  (ES) to adapt the correct scaling of a given problem and also
+                  ensures invariance with respect to any rotation of the
+                  fitness function (or the coordinate system). Especially
+                  rotation invariance, here resulting directly from the
+                  coordinate system independent adaptation of the mutation
+                  distribution, is an essential feature of the ES with regard
+                  to its general applicability to complex fitness
+                  functions. Compared to previous work on this subject, the
+                  introduced formulation facilitates an interpretation of the
+                  resulting mutation distribution, making sensible manipulation
+                  by the user possible (if desired). Furthermore it enables a
+                  more effective control of the overall mutation variance
+                  (expected step length)},
+  keywords = {Evolution strategies, Evolutionary algorithms,
+                  self-adaptation, stochastic processes, Covariance matrix,
+                  matrix algebra, derandomised adaptation, mutation
+                  distribution, rotation invariance, electronic switching
+                  systems}
+}
+
+ +
+@incollection{Hanne2001emo,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2001},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 1993,
+  year = 2001,
+  publisher = {Springer},
+  editor = { Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele  and  Carlos A. {Coello Coello}  and  David Corne },
+  title = {Global Multiobjective Optimization with Evolutionary
+                  Algorithms: Selection Mechanisms and Mutation Control},
+  author = {Hanne, Thomas},
+  pages = {197--212}
+}
+
+ +
+@phdthesis{Hansen1998PhD,
+  author = { Michael Pilegaard Hansen },
+  title = {Metaheuristics for multiple objective combinatorial
+                  optimization},
+  school = {Institute of Mathematical Modelling, Technical
+                  University of Denmark},
+  month = mar,
+  year = 1998
+}
+
+ +
+@incollection{Hansen2006cma,
+  title = {The {CMA} evolution strategy: a comparing review},
+  author = { Nikolaus Hansen },
+  booktitle = {Towards a new evolutionary computation},
+  pages = {75--102},
+  year = 2006,
+  publisher = {Springer}
+}
+
+ +
+@incollection{Hansen2009bpopcma,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = { Nikolaus Hansen },
+  title = {Benchmarking a {BI}-population {CMA-ES} on the {BBOB}-2009
+                  function testbed},
+  pages = {2389--2396},
+  keywords = {bipop-cma-es}
+}
+
+ +
+@inproceedings{HaoCaiHua2006,
+  publisher = {IEEE Press},
+  year = 2006,
+  booktitle = {Proceedings of the International Conference on
+                  Machine Learning and Cybernetics},
+  key = {ICMLC},
+  author = {Zhifeng Hao and Ruichu Cai and Han Huang},
+  title = {An Adaptive Parameter Control Strategy for {ACO}},
+  pages = {203--206}
+}
+
+ +
+@incollection{HaoHuaQinCai2007,
+  year = 2007,
+  volume = 4490,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Yong Shi and G. Dick van Albada and Jack Dongarra
+                  and Peter M. A. Sloot},
+  booktitle = {Computational Science -- ICCS 2007, 7th International Conference, Proceedings, Part IV},
+  author = {Zhifeng Hao and Han Huang and Yong Qin and Ruichu
+                  Cai},
+  title = {An {ACO} Algorithm with Adaptive Volatility Rate of
+                  Pheromone Trail},
+  pages = {1167--1170}
+}
+
+ +
+@inproceedings{HaoPan1998,
+  year = 1998,
+  booktitle = {Fifth International Symposium on Artificial Intelligence and Mathematics,
+                  {AIM} 1998, Fort Lauderdale, Florida, USA, January 4-6, 1998},
+  editor = {Martin C. Golumbic and others},
+  author = { Jin-Kao Hao  and Pannier, J{\^{e}}rome},
+  title = {Simulated Annealing and Tabu Search for Constraint Solving},
+  pages = {1--15}
+}
+
+ +
+@incollection{HarGin1995,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = {Chris S. Mellish},
+  year = 1995,
+  booktitle = {Proceedings of  the 14th International Joint Conference on Artificial Intelligence (IJCAI-95)},
+  author = {William D. Harvey and Matthew L. Ginsberg},
+  title = {Limited Discrepancy Search},
+  pages = {607--615}
+}
+
+ +
+@incollection{HassGueSil2016deep,
+  epub = {http://www.aaai.org/Library/AAAI/aaai16contents.php},
+  year = 2016,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Dale Schuurmans and Michael P. Wellman},
+  title = {Deep Reinforcement Learning with Double {Q}-Learning},
+  author = {van Hasselt, Hado and Guez, Arthur and Silver, David}
+}
+
+ +
+@inproceedings{HeiIge2009aicml,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  editor = {Andrea Pohoreckyj Danyluk and L{\'{e}}on Bottou and Michael
+                  L. Littman},
+  year = 2009,
+  booktitle = {Proceedings of  the 26th International Conference on Machine Learning, {ICML} 2009},
+  title = {Hoeffding and {Bernstein} races for selecting policies in
+                  evolutionary direct policy search},
+  author = { Heidrich-Meisner, Verena  and  Christian Igel },
+  keywords = {automated algorithm configuration, CMA-ES, racing},
+  pages = {401--408},
+  doi = {10.1145/1553374.1553426}
+}
+
+ +
+@misc{Hel2018lkh,
+  author = { Keld Helsgaun },
+  title = {Source Code of the {Lin}-{Kernighan}-{Helsgaun} Traveling
+                  Salesman Heuristic},
+  howpublished = {\url{http://webhotel4.ruc.dk/~keld/research/LKH/}},
+  year = 2018
+}
+
+ +
+@incollection{Hel2018ppsn,
+  volume = 11101,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Keld Helsgaun },
+  title = {Efficient Recombination in the {Lin}-{Kernighan}-{Helsgaun}
+                  Traveling Salesman Heuristic},
+  pages = {95--107},
+  doi = {10.1007/978-3-319-99253-2_8}
+}
+
+ +
+@book{Hen1999:mit,
+  author = {van Hentenryck, Pascal },
+  title = {The {OPL} optimization programming language},
+  publisher = {MIT Press},
+  year = 1999,
+  address = {Cambridge, MA}
+}
+
+ +
+@incollection{HenJacJoh2003,
+  doi = {10.1007/b101874},
+  address = { Boston, MA},
+  publisher = {Springer},
+  year = 2003,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Darrall Henderson  and  Sheldon H. Jacobson  and  Alan W. Johnson },
+  title = {The Theory and Practice of Simulated Annealing},
+  pages = {287--319}
+}
+
+ +
+@book{HenMich05:mit,
+  author = {van Hentenryck, Pascal  and  Laurent D. Michel },
+  title = {Constraint-based Local Search},
+  publisher = {MIT Press},
+  year = 2005,
+  address = {Cambridge, MA}
+}
+
+ +
+@inproceedings{HenMich07synthesis,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 2007,
+  editor = {Robert C. Holte and Adele Howe},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  author = {van Hentenryck, Pascal  and  Laurent D. Michel },
+  title = {Synthesis of constraint-based local search
+                  algorithms from high-level models},
+  pages = {273--278}
+}
+
+ +
+@inproceedings{HerGraObe1999svmrank,
+  key = {ICANN},
+  booktitle = {ICANN'99: Proceedings of the 9th International Conference on
+                  Artificial Neural Networks},
+  year = 1999,
+  title = {Support vector learning for ordinal regression},
+  author = {R. Herbrich and T. Graepel and K. Obermayer},
+  keywords = {support vector machine;metric regression;support vector
+                  learning;ordinal regression;information retrieval;risk
+                  functional;machine learning;pattern classification;},
+  abstract = {We investigate the problem of predicting variables of ordinal
+                  scale. This task is referred to as ordinal regression and is
+                  complementary to the standard machine learning tasks of
+                  classification and metric regression. In contrast to
+                  statistical models we present a distribution independent
+                  formulation of the problem together with uniform bounds of
+                  the risk functional. The approach presented is based on a
+                  mapping from objects to scalar utility values. Similar to
+                  support vector methods we derive a new learning algorithm for
+                  the task of ordinal regression based on large margin rank
+                  boundaries. We give experimental results for an information
+                  retrieval task: learning the order of documents with respect
+                  to an initial query. Experimental results indicate that the
+                  presented algorithm outperforms more naive approaches to
+                  ordinal regression such as support vector classification and
+                  support vector regression in the case of more than two
+                  ranks.},
+  doi = {10.1049/cp:19991091},
+  pages = {97--102},
+  annote = {Proposed the pairwise transform for learning-to-rank}
+}
+
+ +
+@misc{HerLozMol2010test,
+  author = { Francisco Herrera  and  Manuel Lozano  and  Daniel Molina },
+  title = {Test suite for the special issue of {Soft}
+                  {Computing} on scalability of evolutionary
+                  algorithms and other metaheuristics for large scale
+                  continuous optimization problems},
+  year = 2010,
+  howpublished = {\url{http://sci2s.ugr.es/eamhco/}},
+  keywords = {SOCO benchmark}
+}
+
+ +
+@incollection{HerSch2022archive,
+  location = {Boston, Massachusetts},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Carlos Hern{\'a}ndez  and  Oliver Sch{\"u}tze },
+  title = {A bounded archive based for bi-objective problems based on
+                  distance and e-dominance to avoid cyclic behavior},
+  pages = {583--591},
+  doi = {10.1145/3512290.3528840}
+}
+
+ +
+@book{Heyman2003,
+  title = {Stochastic models in operations research: stochastic
+                  optimization},
+  author = {Heyman, Daniel P and Sobel, Matthew J},
+  volume = 2,
+  year = 2003,
+  publisher = {Courier Corporation}
+}
+
+ +
+@book{Hol75,
+  author = { J. Holland },
+  title = {Adaptation in Natural and Artificial Systems},
+  publisher = {University of Michigan Press},
+  year = 1975
+}
+
+ +
+@book{HolWol73:nonparam_stats,
+  author = {Myle Hollander and Douglas A. Wolfe},
+  title = {Nonparametric statistical inference},
+  publisher = {John Wiley \& Sons},
+  address = { New York, NY},
+  year = 1973,
+  note = {Second edition (1999)}
+}
+
+ +
+@incollection{Hoo2004discover,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2004,
+  editor = {Won Kim and Ronny Kohavi and Johannes Gehrke and William
+                  DuMouchel},
+  booktitle = {Proceedings of the tenth ACM SIGKDD international conference
+                  on Knowledge discovery and data mining, {KDD'04}},
+  author = {Hooker, Giles},
+  title = {Discovering Additive Structure in Black Box Functions},
+  pages = {575--580},
+  doi = {10.1145/1014052.1014122},
+  abstract = {Many automated learning procedures lack interpretability,
+                  operating effectively as a black box: providing a prediction
+                  tool but no explanation of the underlying dynamics that drive
+                  it. A common approach to interpretation is to plot the
+                  dependence of a learned function on one or two predictors. We
+                  present a method that seeks not to display the behavior of a
+                  function, but to evaluate the importance of non-additive
+                  interactions within any set of variables. Should the function
+                  be close to a sum of low dimensional components, these
+                  components can be viewed and even modeled
+                  parametrically. Alternatively, the work here provides an
+                  indication of where intrinsically high-dimensional behavior
+                  takes place.The calculations used in this paper correspond
+                  closely with the functional ANOVA decomposition; a
+                  well-developed construction in Statistics. In particular, the
+                  proposed score of interaction importance measures the loss
+                  associated with the projection of the prediction function
+                  onto a space of additive models. The algorithm runs in linear
+                  time and we present displays of the output as a graphical
+                  model of the function for interpretation purposes.},
+  numpages = 6,
+  keywords = {diagnostics, functional ANOVA, feature selection,
+                  interpretation, visualization, additive models, draphical
+                  models}
+}
+
+ +
+@incollection{HooHutLey2021acsat,
+  author = { Holger H. Hoos  and  Frank Hutter  and  Kevin Leyton-Brown },
+  title = {Automated Configuration and Selection of {SAT} Solvers},
+  booktitle = {Handbook of Satisfiability},
+  publisher = {IOS Press},
+  year = 2021,
+  pages = {481--507},
+  month = feb,
+  doi = {10.3233/faia200995}
+}
+
+ +
+@book{HooStu04:sls-elsevier,
+  author = { Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {Stochastic Local Search: Foundations and Applications},
+  publisher = {Elsevier},
+  address = {Amsterdam, The Netherlands},
+  year = 2004,
+  anote = {superseed by~\cite{{HooStu05sls-mk}}}
+}
+
+ +
+@book{HooStu05sls-mk,
+  author = { Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {Stochastic Local Search---Foundations and Applications},
+  publisher = {Morgan Kaufmann Publishers},
+  address = { San Francisco, CA},
+  year = 2005,
+  alias = {HooStu04:book}
+}
+
+ +
+@inproceedings{HooStu1998uai,
+  author = { Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {Evaluating {Las} {Vegas} Algorithms --- Pitfalls and
+                  Remedies},
+  booktitle = {Proceedings of the Fourteenth Conference on Uncertainty in
+                  Artificial Intelligence},
+  editor = {Gregory F. Cooper and Seraf{\'i}n Moral},
+  year = 1998,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  pages = {238--245}
+}
+
+ +
+@inproceedings{Hoos2011mic,
+  author = { Holger H. Hoos },
+  title = {Programming by Optimisation: Towards a new Paradigm
+                  for Developing High-Performance Software},
+  booktitle = {MIC 2011, the 9th Metaheuristics International
+                  Conference},
+  year = 2011,
+  note = {{Plenary talk}},
+  url = {http://mic2011.diegm.uniud.it/uploads/plenaries/Hoos-MIC2011.pdf}
+}
+
+ +
+@incollection{Hoos2012autsea,
+  year = 2012,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Autonomous Search},
+  editor = { Youssef Hamadi  and E. Monfroy and F. Saubion},
+  author = { Holger H. Hoos },
+  title = {Automated Algorithm Configuration and Parameter
+                  Tuning},
+  pages = {37--71},
+  doi = {10.1007/978-3-642-21434-9_3}
+}
+
+ +
+@incollection{HorNeu2008,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  editor = {Conor Ryan},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  author = {Horoba, Christian and  Frank Neumann },
+  title = {Benefits and drawbacks for the use of epsilon-dominance in
+                  evolutionary multi-objective optimization},
+  pages = {641--648},
+  annote = {Proposed $\epsilon$-box},
+  alias = {Hor08}
+}
+
+ +
+@inproceedings{HorNafGol1994npga,
+  month = jun,
+  year = 1994,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 1994 World Congress on Computational Intelligence (WCCI 1994)},
+  key = {WCCI},
+  author = {Horn, J. and Nafpliotis, N. and  David E. Goldberg },
+  title = {A niched {Pareto} genetic algorithm for multiobjective
+                  optimization},
+  pages = {82--87},
+  doi = {10.1109/ICEC.1994.350037}
+}
+
+ +
+@inproceedings{HosEec2008cole,
+  series = {CGO '08},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  booktitle = {Proceedings of the 6th Annual IEEE/ACM International
+                  Symposium on Code Generation and Optimization},
+  editor = {Soffa, Mary Lou and Duesterwald, Evelyn},
+  author = {Kenneth Hoste and Lieven Eeckhout},
+  title = {Cole: Compiler Optimization Level Exploration},
+  pages = {165--174},
+  doi = {10.1145/1356058.1356080}
+}
+
+ +
+@incollection{HuaYaHaoCai2006,
+  volume = 4115,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  editor = {De-Shuang Huang and Kang Li and George W. Irwin},
+  booktitle = {International Conference on Computational Science
+                  (3)},
+  author = {Han Huang and Xiaowei Yang and Zhifeng Hao and
+                  Ruichu Cai},
+  title = {A Novel {ACO} Algorithm with Adaptive Parameter},
+  pages = {12--21}
+}
+
+ +
+@inproceedings{HuaYanTse04:ics,
+  author = { Kuo-Si Huang  and  Chang-Biau Yang  and  Kuo-tsung Tseng },
+  title = {Fast algorithms for finding the common subsequences
+                  of multiple sequences},
+  booktitle = {Proceedings of the International Computer Symposium},
+  pages = {1006--1011},
+  year = 2004,
+  publisher = {IEEE Press}
+}
+
+ +
+@inproceedings{Hughes2003msops,
+  year = 2003,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = dec,
+  booktitle = {Proceedings of  the 2003 Congress on Evolutionary Computation (CEC'03)},
+  key = {IEEE CEC},
+  title = {Multiple single objective {Pareto} sampling},
+  author = { Hughes, Evan J. },
+  pages = {2678--2684}
+}
+
+ +
+@inproceedings{Hughes2007msops,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  title = {{MSOPS-II}: A general-purpose many-objective optimiser},
+  author = { Hughes, Evan J. },
+  pages = {3944--3951}
+}
+
+ +
+@incollection{Hughes2011models,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  title = {Many-objective directed evolutionary line search},
+  author = { Hughes, Evan J. },
+  pages = {761--768}
+}
+
+ +
+@inproceedings{HunLop2019turing,
+  isbn = {978-1-5262-0820-0},
+  organization = {Alan Turing Institute},
+  month = nov # { 21--22},
+  year = 2019,
+  date = {2019-11-21/2019-11-22},
+  address = {London, UK},
+  editor = {Iv{\'a}n Palomares},
+  booktitle = {International Alan Turing Conference on Decision Support and
+                  Recommender systems},
+  author = {Maura Hunt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Modeling a Decision-Maker in Goal Programming by means of
+                  Computational Rationality},
+  pages = {17--20},
+  abstract = {This paper extends a simulation of cognitive mechanisms in
+                  the context of multi-criteria decision-making by using ideas
+                  from computational rationality. Specifically, this paper
+                  improves the simulation of a human decision-maker (DM) by
+                  considering how resource constraints impact their evaluation
+                  process in an interactive Goal Programming problem. Our
+                  analysis confirms and emphasizes a previous simulation study
+                  by showing key areas that could be effected by cognitive
+                  mechanisms. While the results are promising, the effects
+                  should be validated by future experiments with human DMs.},
+  epub = {https://dsrs.blogs.bristol.ac.uk/files/2020/01/DSRS-Turing_19.pdf#page=24}
+}
+
+ +
+@incollection{HusStu2009hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 5818,
+  series = {Lecture Notes in Computer Science},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Luca {Di Gaspero} and  Andrea Roli  and  M. Sampels  and Andrea Schaerf},
+  year = 2009,
+  booktitle = {Hybrid Metaheuristics},
+  author = {Mohamed Saifullah Hussin and  Thomas St{\"u}tzle },
+  title = {Hierarchical Iterated Local Search for the Quadratic
+                  Assignment Problem},
+  pages = {115--129},
+  doi = {10.1007/978-3-642-04918-7_9}
+}
+
+ +
+@inproceedings{HutBabHooHu2007fmcad,
+  address = {Austin, Texas, USA},
+  year = 2007,
+  publisher = {IEEE Computer Society, Washington, DC, USA},
+  booktitle = {{FMCAD'07}: Proceedings of  the 7th International Conference
+                  Formal Methods in Computer Aided Design},
+  editor = {Jason Baumgartner and Mary Sheeran},
+  author = { Frank Hutter  and  Domagoj Babi{\'c}  and  Holger H. Hoos  and Alan J. Hu},
+  title = {Boosting Verification by Automatic Tuning of
+                  Decision Procedures},
+  pages = {27--34}
+}
+
+ +
+@incollection{HutHooLey2009gecco,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009},
+  address = { New York, NY},
+  year = 2009,
+  publisher = {ACM Press},
+  editor = { Franz Rothlauf },
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown  and Kevin P. Murphy},
+  title = {An experimental investigation of model-based
+                  parameter optimisation: {SPO} and beyond},
+  pages = {271--278},
+  doi = {10.1145/1569901.1569940}
+}
+
+ +
+@incollection{HutHooLey2010:cpaior,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2010,
+  editor = { Andrea Lodi  and  Michela Milano  and  Paolo Toth },
+  volume = 6140,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 7th International Conference, CPAIOR 2010},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Automated Configuration of Mixed Integer Programming Solvers},
+  pages = {186--202},
+  keywords = {MIP, ParamILS},
+  alias = {HutHooLey10-mipconfig},
+  doi = {10.1007/978-3-642-13520-0_23}
+}
+
+ +
+@incollection{HutHooLey2011lion,
+  volume = 6683,
+  editor = { Carlos A. {Coello Coello} },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2011,
+  booktitle = {Learning and Intelligent Optimization, 5th
+                  International Conference, LION 5},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Sequential Model-Based Optimization for General
+                  Algorithm Configuration},
+  pages = {507--523},
+  keywords = {SMAC,ROAR},
+  doi = {10.1007/978-3-642-25566-3_40}
+}
+
+ +
+@incollection{HutHooLey2012lion,
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7219,
+  publisher = {Springer},
+  year = 2012,
+  editor = { Youssef Hamadi  and  Marc Schoenauer },
+  booktitle = {Learning and Intelligent Optimization, 6th International
+                  Conference, LION 6},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Parallel Algorithm Configuration},
+  pages = {55--70}
+}
+
+ +
+@incollection{HutHooLey2013lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7997,
+  editor = { Panos M. Pardalos  and G. Nicosia},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  booktitle = {Learning and Intelligent Optimization, 7th
+                  International Conference, LION 7},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Identifying Key Algorithm Parameters and Instance Features
+                  using Forward Selection},
+  pages = {364--381},
+  doi = {10.1007/978-3-642-44973-4_40},
+  abstract = {Most state-of-the-art algorithms for large-scale optimization
+                  problems expose free parameters, giving rise to combinatorial
+                  spaces of possible configurations. Typically, these spaces
+                  are hard for humans to understand. In this work, we study a
+                  model-based approach for identifying a small set of both
+                  algorithm parameters and instance features that suffices for
+                  predicting empirical algorithm performance well. Our
+                  empirical analyses on a wide variety of hard combinatorial
+                  problem benchmarks spanning SAT, MIP, and TSP show that--for
+                  parameter configurations sampled uniformly at random--very
+                  good performance predictions can typically be obtained based
+                  on just two key parameters, and that similarly, few instance
+                  features and algorithm parameters suffice to predict the most
+                  salient algorithm performance characteristics in the combined
+                  configuration/feature space. We also use these models to
+                  identify settings of these key parameters that are predicted
+                  to achieve the best overall performance, both on average
+                  across instances and in an instance-specific way. This serves
+                  as a further way of evaluating model quality and also
+                  provides a tool for further understanding the parameter
+                  space. We provide software for carrying out this analysis on
+                  arbitrary problem domains and hope that it will help
+                  algorithm developers gain insights into the key parameters of
+                  their algorithms, the key features of their instances, and
+                  their interactions.},
+  keywords = {parameter importance}
+}
+
+ +
+@inproceedings{HutHooLey2014icml,
+  publisher = {{PMLR}},
+  year = 2014,
+  volume = 32,
+  booktitle = {Proceedings of  the 31st International Conference on Machine Learning, {ICML} 2014},
+  editor = {Xing, Eric P. and Jebara, Tony},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {An Efficient Approach for Assessing Hyperparameter
+                  Importance},
+  pages = {754--762},
+  url = {https://proceedings.mlr.press/v32/hutter14.html},
+  keywords = {fANOVA, parameter importance}
+}
+
+ +
+@incollection{HutHooLeyMur2010lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Roberto Battiti },
+  series = {Lecture Notes in Computer Science},
+  volume = 6073,
+  year = 2010,
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown  and Kevin Murphy},
+  title = {Time-Bounded Sequential Parameter Optimization},
+  pages = {281--298},
+  doi = {10.1007/978-3-642-13800-3_30}
+}
+
+ +
+@inproceedings{HutHooStu07aaai,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 2007,
+  editor = {Robert C. Holte and Adele Howe},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  author = { Frank Hutter  and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {Automatic Algorithm Configuration Based on Local Search},
+  pages = {1152--1157},
+  alias = {HutHooStu07}
+}
+
+ +
+@incollection{HutLopFaw2014lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8426,
+  editor = { Panos M. Pardalos  and  Mauricio G. C. Resende  and Chrysafis Vogiatzis and Jose
+                  L. Walteros},
+  series = {Lecture Notes in Computer Science},
+  year = 2014,
+  booktitle = {Learning and Intelligent Optimization, 8th International Conference, LION 8},
+  author = { Frank Hutter  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Chris Fawcett  and  Marius Thomas Lindauer  and  Holger H. Hoos  and  Kevin Leyton-Brown  and  Thomas St{\"u}tzle },
+  title = {{AClib}: A Benchmark Library for Algorithm Configuration},
+  pages = {36--40},
+  doi = {10.1007/978-3-319-09584-4_4}
+}
+
+ +
+@misc{Hutter2007satbench,
+  author = { Frank Hutter },
+  title = {{SAT} benchmarks used in automated algorithm configuration},
+  howpublished = {\url{http://www.cs.ubc.ca/labs/beta/Projects/AAC/SAT-benchmarks.html}},
+  year = 2007
+}
+
+ +
+@phdthesis{HutterPhD,
+  author = { Frank Hutter },
+  title = {Automated Configuration of Algorithms for Solving
+                  Hard Computational Problems},
+  school = {University of British Columbia, Department of
+                  Computer Science},
+  address = {Vancouver, Canada},
+  year = 2009,
+  month = oct
+}
+
+ +
+@misc{IJCAI2021checklist,
+  author = {Zhiyuan Liu and Jian Tang},
+  title = {IJCAI 2021 Reproducibility Guidelines, 35th International
+                  Joint Conference on Artificial Intelligence},
+  year = 2021,
+  howpublished = {\url{https://ijcai-21.org/wp-content/uploads/2020/12/20201226-IJCAI-Reproducibility.pdf}}
+}
+
+ +
+@techreport{INRIA-RR-7871,
+  author = { J{\'e}r{\'e}mie Humeau  and  Arnaud Liefooghe  and  Talbi, El-Ghazali  and  Verel, S{\'e}bastien },
+  title = {{ParadisEO-MO}: From Fitness Landscape Analysis to Efficient
+                  Local Search Algorithms},
+  institution = {INRIA, France},
+  year = 2012,
+  type = {Rapport de recherche},
+  number = {RR-7871},
+  epub = {http://hal.inria.fr/hal-00665421/PDF/RR-7871.pdf}
+}
+
+ +
+@techreport{IRIDIA-2003-037,
+  author = { Mauro Birattari },
+  title = {The {\rpackage{race}} Package for~\proglang{R}: {Racing}
+                  Methods for the Selection of the Best},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2003,
+  number = {TR/IRIDIA/2003-037}
+}
+
+ +
+@techreport{IRIDIA-2004-001,
+  author = { Mauro Birattari },
+  title = {On the Estimation of the Expected Performance of a Metaheuristic on a Class of Instances. How Many Instances, How Many Runs?},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2004,
+  number = {TR/IRIDIA/2004-001}
+}
+
+ +
+@techreport{IRIDIA-2007-019,
+  author = { Krzysztof Socha  and  Marco Dorigo },
+  title = {Ant Colony Optimization for Mixed-Variable Optimization
+                  Problems},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2007,
+  number = {TR/IRIDIA/2007-019},
+  month = oct,
+  alias = {SocDor97:iridia019}
+}
+
+ +
+@techreport{IRIDIA-2009-015,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Exploratory Analysis of Stochastic Local Search Algorithms in Biobjective Optimization},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2009,
+  number = {TR/IRIDIA/2009-015},
+  month = may,
+  note = {Published as a book chapter~\cite{LopPaqStu09emaa}}
+}
+
+ +
+@techreport{IRIDIA-2009-019,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Analysis of Algorithmic Components for
+                  Multiobjective Ant Colony Optimization: A Case Study
+                  on the Biobjective {TSP}},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2009-019},
+  year = 2009,
+  month = jun,
+  note = {Published in the proceedings of Evolution Artificielle, 2009~\cite{LopStu09ea}}
+}
+
+ +
+@techreport{IRIDIA-2009-020,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Effective Hybrid Stochastic Local Search Algorithms
+                  for Biobjective Permutation Flowshop Scheduling},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2009-020},
+  year = 2009,
+  month = jun,
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2009-020.pdf},
+  note = {Published in the proceedings of Hybrid Metaheuristics 2009~\cite{DubLopStu09:hm-bfsp}}
+}
+
+ +
+@techreport{IRIDIA-2009-026,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Adaptive ``Anytime'' Two-Phase Local Search},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2010,
+  number = {TR/IRIDIA/2009-026},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2009-026.pdf},
+  note = {Published in the proceedings of LION 4~\cite{DubLopStu10:lion-bfsp}}
+}
+
+ +
+@techreport{IRIDIA-2010-002,
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Paola Pellegrini  and  Michael Maur  and  Marco A. {Montes de Oca}  and  Mauro Birattari  and  Marco Dorigo },
+  title = {Parameter Adaptation in Ant Colony Optimization},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2010-002},
+  year = 2010,
+  month = jan,
+  note = {Published as a book chapter~\cite{StuLopPel2011autsea}}
+}
+
+ +
+@techreport{IRIDIA-2010-019,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Hybrid {TP+PLS} Algorithm for Bi-objective
+                  Flow-Shop Scheduling Problems},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2010,
+  number = {TR/IRIDIA/2010-019},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2010-019.pdf},
+  note = {Published in Computers \& Operations Research~\cite{DubLopStu2011cor}}
+}
+
+ +
+@techreport{IRIDIA-2010-020,
+  author = {M. S. Hussin and  Thomas St{\"u}tzle },
+  title = {Tabu Search vs. Simulated Annealing for Solving
+                  Large Quadratic Assignment Instances},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2010,
+  number = {TR/IRIDIA/2010-020}
+}
+
+ +
+@techreport{IRIDIA-2010-022,
+  author = { J{\'e}r{\'e}mie Dubois-Lacoste  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Improving the Anytime Behavior of Two-Phase Local
+                  Search},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2010,
+  number = {TR/IRIDIA/2010-022},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2010-022.pdf},
+  note = {Published in Annals of Mathematics and Artificial Intelligence~\cite{DubLopStu2011amai}}
+}
+
+ +
+@techreport{IRIDIA-2011-001,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles  and  Marco Laumanns },
+  title = {On Sequential Online Archiving of Objective Vectors},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  number = {TR/IRIDIA/2011-001},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-001.pdf},
+  note = {This is a revised version of the paper published in EMO 2011~\cite{LopKnoLau2011emo}}
+}
+
+ +
+@techreport{IRIDIA-2011-002,
+  author = { Mauro Birattari  and  Marco Chiarandini  and  Marco Saerens  and  Thomas St{\"u}tzle },
+  year = 2011,
+  title = {Learning graphical models for parameter tuning},
+  number = {TR/IRIDIA/2011-002},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-002.pdf}
+}
+
+ +
+@techreport{IRIDIA-2011-003,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {The Automatic Design of Multi-Objective Ant Colony
+                  Optimization Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  number = {TR/IRIDIA/2011-003},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-003.pdf},
+  note = {Published in IEEE Transactions on Evolutionary
+                  Computation~\cite{LopStu2012tec}}
+}
+
+ +
+@techreport{IRIDIA-2011-010,
+  author = {Liao, Tianjun  and  Daniel Molina  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle },
+  title = {A Note on the Effects of Enforcing Bound Constraints on
+                  Algorithm Comparisons using the {IEEE CEC'05} Benchmark
+                  Function Suite},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  number = {TR/IRIDIA/2011-010},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-010.pdf},
+  note = {Published in Evolutionary Computation~\cite{LiaMolMonStu2014}}
+}
+
+ +
+@techreport{IRIDIA-2011-022,
+  author = {Liao, Tianjun  and  Daniel Molina  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle },
+  title = {Computational Results for an Automatically Tuned {IPOP-CMA-ES} on the {CEC'05} Benchmark Set},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  number = {TR/IRIDIA/2011-022}
+}
+
+ +
+@techreport{IRIDIA-2012-012,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Improving the Anytime Behaviour of
+                  Optimisation Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2012,
+  number = {TR/IRIDIA/2012-012},
+  month = may,
+  note = {Published in European Journal of Operational Research~\cite{LopStu2013ejor}}
+}
+
+ +
+@techreport{IRIDIA-2012-019,
+  author = { Andreea Radulescu  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Improving the Anytime Behaviour of
+                  Multiobjective Evolutionary Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2012,
+  number = {TR/IRIDIA/2012-019},
+  note = {Published in the proceedings of EMO 2013~\cite{RadLopStu2013emo}}
+}
+
+ +
+@techreport{IRIDIA-2013-002,
+  author = {Liao, Tianjun  and  Thomas St{\"u}tzle  and  Marco A. {Montes de Oca}  and  Marco Dorigo },
+  title = {A Unified Ant Colony Optimization Algorithm for Continuous Optimization},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2013,
+  number = {TR/IRIDIA/2013-002}
+}
+
+ +
+@techreport{IRIDIA-2013-015,
+  author = { Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle },
+  year = 2013,
+  title = {Grammar-based generation of stochastic local search
+                  heuristics through automatic algorithm configuration
+                  tools},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2013-015}
+}
+
+ +
+@techreport{IRIDIA-2014-009,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien },
+  title = {Local Optimal Sets and Bounded Archiving on
+                  Multi-objective {NK}-Landscapes with Correlated
+                  Objectives},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2014,
+  number = {TR/IRIDIA/2014-009}
+}
+
+ +
+@techreport{IRIDIA-2014-014,
+  author = { Vito Trianni  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Advantages of Multi-Objective Optimisation in Evolutionary
+                  Robotics: Survey and Case Studies},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2014,
+  number = {TR/IRIDIA/2014-014},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2014-014.pdf}
+}
+
+ +
+@techreport{IRIDIA-2017-005,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {A Large-Scale Experimental Evaluation of High-Performing
+                  Multi- and Many-Objective Evolutionary Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2017,
+  number = {TR/IRIDIA/2017-005},
+  month = nov
+}
+
+ +
+@techreport{IRIDIA-2017-006,
+  author = { Alberto Franzin  and   P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle },
+  title = {Effect of Transformations of Numerical Parameters
+                 in Automatic Algorithm Configuration},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2017-006},
+  year = 2017,
+  month = mar,
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2017-006.pdf}
+}
+
+ +
+@techreport{IRIDIA-2017-011,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Configuration of Multi-objective Optimizers and
+                  Multi-objective Configuration},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2017,
+  number = {TR/IRIDIA/2017-011},
+  month = nov,
+  alias = {BezLopStu2017:techreport-011},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2017-011.pdf},
+  note = {Published as a book chapter~\cite{BezLopStu2020chapter}}
+}
+
+ +
+@techreport{IRIDIA-2017-012,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Marie-El{\'e}onore Kessaci  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Metaheuristics from Algorithmic
+                  Components},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2017,
+  number = {TR/IRIDIA/2017-012},
+  month = dec,
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2017-012.pdf}
+}
+
+ +
+@techreport{IRIDIA-2018-001,
+  author = { Leonardo C. T. Bezerra  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Designing State-of-the-Art Multi- and
+                  Many-Objective Evolutionary Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2018,
+  number = {TR/IRIDIA/2018-001},
+  month = jan,
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2018-001.pdf},
+  note = {Published in Evolutionary Computation journal~\cite{BezLopStu2019ec}}
+}
+
+ +
+@techreport{IRIDIA-2018-010,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Revisiting Simulated Annealing: a Component-Based Analysis},
+  year = 2018,
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2018-010},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2018-010.pdf}
+}
+
+ +
+@techreport{IRIDIA-2021-002,
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {{PSO-X}: A Component-Based Framework for the Automatic Design
+                  of Particle Swarm Optimization Algorithms},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2021,
+  number = {TR/IRIDIA/2021-002},
+  annote = {Published as \cite{CamStuDor2021psox}},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2021-002.pdf}
+}
+
+ +
+@techreport{IRIDIA-2021-005,
+  author = { Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {A Landscape-based Analysis of Fixed Temperature and Simulated Annealing},
+  year = 2021,
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2021-005},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2021-005.pdf}
+}
+
+ +
+@techreport{IRIDIA-2021-006,
+  author = {Camacho-Villal\'{o}n, Christian Leonardo and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Cuckoo Search {$\equiv (\mu + \lambda$)}-Evolution Strategy --
+                  A Rigorous Analysis of an Algorithm That Has Been Misleading
+                  the Research Community for More Than 10 Years and Nobody Seems to Have Noticed},
+  year = 2021,
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2021-006},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2021-006.pdf}
+}
+
+ +
+@incollection{Ige2005mosvm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 3410,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  year = 2005,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  author = { Christian Igel },
+  title = {Multi-objective Model Selection for Support Vector Machines},
+  annote = {Early work on multi-objective hyper-parameter optimization
+                  (AutoML)},
+  pages = {534--546},
+  doi = {10.1007/978-3-540-31880-4_37}
+}
+
+ +
+@inproceedings{IkeKitShi2001cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2001,
+  booktitle = {Proceedings of  the 2001 Congress on Evolutionary Computation (CEC'01)},
+  key = {IEEE CEC},
+  author = {Ikeda, Kokolo and Hajime Kita and Shigenobu Kobayashi},
+  title = {Failure of {Pareto}-based {MOEA}s: {Does} non-dominated really
+                  mean near to optimal?},
+  pages = {957--962},
+  keywords = {dominance resistance}
+}
+
+ +
+@book{IllPenSto2008,
+  title = {Statistical Analysis and Modelling of Spatial Point Patterns},
+  author = {Illian, Janine and Penttinen, Antti and Stoyan, Helga and
+                  Stoyan, Dietrich},
+  publisher = {Wiley},
+  year = 2008,
+  alias = {Illian2008}
+}
+
+ +
+@incollection{IreMerMid2001,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2001},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 1993,
+  year = 2001,
+  publisher = {Springer},
+  editor = { Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele  and  Carlos A. {Coello Coello}  and  David Corne },
+  author = { S. Iredi  and  D. Merkle  and  Martin Middendorf },
+  title = {Bi-Criterion Optimization with Multi Colony Ant
+                  Algorithms},
+  pages = {359--372},
+  alias = {IrediMid01,IreMerMid01},
+  keywords = {BicriterionAnt}
+}
+
+ +
+@misc{IridiaSupp2012-011,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {{Automatically Improving the Anytime Behaviour of Optimisation Algorithms: Supplementary material}},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2012-011/}},
+  year = 2012
+}
+
+ +
+@incollection{IruLop2021gecco,
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  author = { Irurozki, Ekhine  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Unbalanced Mallows Models for Optimizing Expensive Black-Box
+                  Permutation Problems},
+  doi = {10.1145/3449639.3459366},
+  pages = {225--233},
+  supplement = {https://zenodo.org/record/4500974},
+  abstract = {Expensive black-box combinatorial optimization problems arise
+                  in practice when the objective function is evaluated by means
+                  of a simulator or a real-world experiment. Since each fitness
+                  evaluation is expensive in terms of time or resources, only a
+                  limited number of evaluations is possible, typically several
+                  orders of magnitude smaller than in non-expensive
+                  problems. In this scenario, classical optimization methods
+                  such as mixed-integer programming and local search are not
+                  useful.  In the continuous case, Bayesian optimization, in
+                  particular using Gaussian processes, has proven very
+                  effective under these conditions. Much less research is
+                  available in the combinatorial case. In this paper, we
+                  propose and analyze UMM, an estimation-of-distribution (EDA)
+                  algorithm based on a Mallows probabilistic model and
+                  unbalanced rank aggregation (uBorda). Experimental results on
+                  black-box versions of LOP and PFSP show that UMM is able to
+                  match, and sometimes surpass, the solutions obtained by CEGO,
+                  a Bayesian optimization algorithm for combinatorial
+                  optimization. Moreover, the computational complexity of UMM
+                  increases linearly with both the number of function
+                  evaluations and the permutation size.},
+  keywords = {UMM, Permutation, Expensive, Black-box}
+}
+
+ +
+@incollection{IshMasNoj2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = { Ishibuchi, Hisao  and Masuda, Hiroyuki and Nojima, Yusuke},
+  title = {A Study on Performance Evaluation Ability of a Modified
+                  Inverted Generational Distance Indicator},
+  pages = {695--702}
+}
+
+ +
+@incollection{IshMasTanNoj2015igd,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = { Ishibuchi, Hisao  and Masuda, Hiroyuki and Tanigaki, Yuki and
+                  Nojima, Yusuke},
+  title = {Modified Distance Calculation in Generational Distance and
+                  Inverted Generational Distance},
+  pages = {110--125},
+  annote = {Proposed IGD+},
+  keywords = {Performance metrics, multi-objective, IGD, IGD+}
+}
+
+ +
+@inproceedings{IshTsuNoj2008,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2008,
+  booktitle = {Proceedings of  the 2008 Congress on Evolutionary Computation (CEC 2008)},
+  key = {IEEE CEC},
+  author = { Ishibuchi, Hisao  and Tsukamoto, N. and Nojima, Y.},
+  title = {Evolutionary many-objective optimization: {A} short review},
+  doi = {10.1109/CEC.2008.4631121},
+  pages = {2419--2426}
+}
+
+ +
+@incollection{JacJouTal2014evapp,
+  year = 2014,
+  volume = 8602,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  editor = { Anna I. Esparcia{-}Alc{\'{a}}zar  and Antonio M. Mora},
+  author = {Sophie Jacquin and  Laetitia Jourdan  and  Talbi, El-Ghazali },
+  title = {Dynamic Programming Based Metaheuristic for Energy Planning Problems},
+  pages = {165--176},
+  doi = {10.1007/978-3-662-45523-4_14},
+  keywords = {irace}
+}
+
+ +
+@incollection{JaiCoeCha2008,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  editor = {Conor Ryan},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  title = {Objective reduction using a feature selection technique},
+  author = { L{\'o}pez Jaimes, Antonio  and  Carlos A. {Coello Coello}  and Chakraborty, Debrup},
+  pages = {673--680}
+}
+
+ +
+@incollection{JaiCoeUri2009onlinered,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  title = {Online Objective Reduction to Deal with Many-Objective
+                  Problems},
+  author = { L{\'o}pez Jaimes, Antonio  and  Carlos A. {Coello Coello}  and Ur{\'i}as Barrientos, Jes{\'u}s E.},
+  pages = {423--437},
+  abstract = {In this paper, we propose and analyze two schemes to
+                  integrate an objective reduction technique into a
+                  multi-objective evolutionary algorithm (moea) in order to
+                  cope with many-objective problems. One scheme reduces
+                  periodically the number objectives during the search until
+                  the required objective subset size has been reached and,
+                  towards the end of the search, the original objective set is
+                  used again. The second approach is a more conservative scheme
+                  that alternately uses the reduced and the entire set of
+                  objectives to carry out the search. Besides improving
+                  computational efficiency by removing some objectives, the
+                  experimental results showed that both objective reduction
+                  schemes also considerably improve the convergence of a moea
+                  in many-objective problems.}
+}
+
+ +
+@inproceedings{JamTal2016,
+  publisher = {{JMLR}.org},
+  volume = 51,
+  series = {{JMLR} Workshop and Conference Proceedings},
+  year = 2016,
+  booktitle = {Proceedings of the 19th International Conference on Artificial Intelligence
+                  and Statistics, {AISTATS} 2016, Cadiz, Spain, May 9-11, 2016},
+  editor = {Arthur Gretton and Christian C. Robert},
+  author = {Jamieson, Kevin G. and Talwalkar, Ameet},
+  title = {Non-stochastic Best Arm Identification and Hyperparameter Optimization},
+  pages = {240--248},
+  url = {http://proceedings.mlr.press/v51/jamieson16.html}
+}
+
+ +
+@incollection{JasBran2008,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  author = { Andrzej Jaszkiewicz  and  J{\"u}rgen Branke },
+  title = {Interactive Multiobjective Evolutionary Algorithms},
+  pages = {179--193},
+  doi = {10.1007/978-3-540-88908-3_7}
+}
+
+ +
+@incollection{JasIshZha2012,
+  series = {Studies in Computational Intelligence},
+  publisher = {Springer},
+  year = 2011,
+  volume = 379,
+  editor = {Neri, Ferrante and  Carlos Cotta  and  Pablo Moscato },
+  booktitle = {Handbook of Memetic Algorithms},
+  title = {Multiobjective memetic algorithms},
+  author = { Andrzej Jaszkiewicz  and  Ishibuchi, Hisao  and  Zhang, Qingfu },
+  pages = {201--217}
+}
+
+ +
+@manual{Java:smac,
+  title = {Manual for {SMAC}},
+  author = { Frank Hutter  and Steve Ramage},
+  year = 2015,
+  note = {SMAC version 2.10.03},
+  organization = {University of British Columbia},
+  url = {http://www.cs.ubc.ca/labs/beta/Projects/SMAC/v2.10.03/manual.pdf}
+}
+
+ +
+@incollection{JerSin1996,
+  publisher = {PWS Publishing Co.},
+  year = 1996,
+  editor = {Hochbaum, Dorit S.},
+  booktitle = {Approximation Algorithms For {NP}-hard Problems},
+  title = {The {Markov} chain {Monte} {Carlo} method: an approach
+                  to approximate counting and integration},
+  author = { Mark Jerrum  and  Alistair Sinclair },
+  pages = {482--520}
+}
+
+ +
+@incollection{JesLieDerPaq2020gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  doi = {10.1145/3377930},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  author = {Jesus, Alexandre D. and  Arnaud Liefooghe  and  Bilel Derbel  and  Lu{\'i}s Paquete },
+  pages = {850---858},
+  title = {Algorithm Selection of Anytime Algorithms}
+}
+
+ +
+@misc{JoH2015:policy,
+  key = {{Journal of Heuristics}},
+  title = {{Journal of Heuristics. Policies on Heuristic Search Research}},
+  howpublished = {\url{http://www.springer.com/journal/10732}},
+  year = 2015,
+  note = {Version visited last on June 10, 2015}
+}
+
+ +
+@incollection{JohGutMcG++02:atsp,
+  editor = {G. Gutin and A. Punnen},
+  year = 2002,
+  publisher = {Kluwer Academic Publishers, Dordrecht, The Netherlands},
+  booktitle = {The Traveling Salesman Problem and its Variations},
+  author = {David S. Johnson and G. Gutin and  Lyle A. McGeoch  and A. Yeo and W. Zhang and A. Zverovitch},
+  title = {Experimental Analysis of Heuristics for the {ATSP}},
+  pages = {445--487}
+}
+
+ +
+@incollection{JohMcG02:stsp,
+  editor = {G. Gutin and A. Punnen},
+  year = 2002,
+  publisher = {Kluwer Academic Publishers, Dordrecht, The Netherlands},
+  booktitle = {The Traveling Salesman Problem and its Variations},
+  author = {David S. Johnson and  Lyle A. McGeoch },
+  title = {Experimental Analysis of Heuristics for the {STSP}},
+  pages = {369--443}
+}
+
+ +
+@incollection{JohMcG97,
+  editor = { Emile H. L. Aarts  and  Jan Karel Lenstra },
+  year = 1997,
+  address = { Chichester, UK},
+  publisher = {John Wiley \& Sons},
+  booktitle = {Local Search in Combinatorial Optimization},
+  author = {David S. Johnson and  Lyle A. McGeoch },
+  title = {The Traveling Salesman Problem: A Case Study in Local
+                  Optimization},
+  pages = {215--310}
+}
+
+ +
+@inproceedings{Johnson1990,
+  author = {David S. Johnson},
+  title = {Local Optimization and the Traveling Salesman
+                  Problem},
+  booktitle = {Automata, Languages and Programming, 17th
+                  International Colloquium},
+  volume = 443,
+  series = {Lecture Notes in Computer Science},
+  year = 1990,
+  editor = {M. Paterson},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  pages = {446--461}
+}
+
+ +
+@misc{Johnson2001,
+  author = {David S. Johnson and  Lyle A. McGeoch  and Rego, C. and  Fred Glover },
+  title = {8th {DIMACS} Implementation Challenge: The Traveling Salesman Problem},
+  year = 2001,
+  howpublished = {\url{http://dimacs.rutgers.edu/archive/Challenges/TSP}},
+  keywords = {TSP Challenge, RUE, RCE, generators}
+}
+
+ +
+@incollection{Johnson2002,
+  editor = {Michael H. Goldwasser and David S. Johnson and  Catherine C. McGeoch },
+  year = 2002,
+  address = { Providence, RI},
+  publisher = {American Mathematical Society},
+  volume = 59,
+  series = {{DIMACS} Series in Discrete Mathematics and Theoretical
+                  Computer Science},
+  booktitle = {Data Structures, Near Neighbor Searches, and Methodology:
+                  Fifth and Sixth {DIMACS} Implementation Challenges},
+  author = {David S. Johnson},
+  title = {A Theoretician's Guide to the Experimental Analysis of
+                  Algorithms},
+  pages = {215--250},
+  doi = {10.1090/dimacs/059/11}
+}
+
+ +
+@book{Jon2006,
+  author = { De Jong, Kenneth A. },
+  title = {Evolutionary computation: a unified approach},
+  publisher = {MIT Press},
+  address = {Cambridge, MA},
+  year = 2006
+}
+
+ +
+@incollection{JonFor1995fdc,
+  year = 1995,
+  address = { Pittsburgh, PA},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  booktitle = {Proceedings of  the Sixth International Conference on Genetic Algorithms (ICGA'95)},
+  editor = {Larry J. Eshelman},
+  author = {Jones, Terry and  Stephanie Forrest },
+  title = {Fitness Distance Correlation as a Measure of Problem
+                  Difficulty for Genetic Algorithms},
+  pages = {184--192}
+}
+
+ +
+@book{JonPev2004,
+  title = {An introduction to bioinformatics algorithms},
+  author = {Jones, Neil C. and Pevzner, Pavel A.},
+  year = 2004,
+  publisher = {MIT Press},
+  address = {Cambridge, MA}
+}
+
+ +
+@inproceedings{JuiPol98:aaai,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 1998,
+  booktitle = {Proceedings of AAAI 1998 -- Fifteenth National Conference on
+                  Artificial Intelligence},
+  editor = {Jack Mostow and Chuck Rich},
+  author = {H. Juill{\'e} and J. B. Pollack},
+  title = {A Sampling-Based Heuristic for Tree Search Applied
+                  to Grammar Induction},
+  pages = {776--783}
+}
+
+ +
+@incollection{Julstrom1995,
+  year = 1995,
+  address = { Pittsburgh, PA},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  booktitle = {Proceedings of  the Sixth International Conference on Genetic Algorithms (ICGA'95)},
+  editor = {Larry J. Eshelman},
+  author = {Bryant A. Julstrom},
+  title = {What Have You Done for Me Lately? Adapting Operator Probabilities in a Steady-State Genetic Algorithm},
+  pages = {81--87}
+}
+
+ +
+@incollection{KadMalSelTie2010isac,
+  publisher = {IOS Press},
+  year = 2010,
+  booktitle = {Proceedings of the 19th European Conference on Artificial Intelligence},
+  editor = {Coelho, H. and Studer, R. and Wooldridge, M.},
+  author = {Kadioglu, Serdar and  Yuri Malitsky  and  Meinolf Sellmann  and  Kevin Tierney },
+  title = {{ISAC}: Instance-Specific Algorithm Configuration},
+  pages = {751--756}
+}
+
+ +
+@inproceedings{KajIkeHaj2009cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2009,
+  booktitle = {Proceedings of  the 2009 Congress on Evolutionary Computation (CEC 2009)},
+  key = {IEEE CEC},
+  author = {H. Kaji and Ikeda, Kokolo and Hajime Kita},
+  title = {Avoidance of constraint violation for experiment-based
+                  evolutionary multi-objective optimization},
+  pages = {2756--2763},
+  keywords = {Safe Optimization, evolutionary computation, constraint
+                  violation, experiment-based evolutionary multiobjective
+                  optimization, evolutionary algorithm, risky-constraint
+                  violation, Constraint optimization, Diesel engines,
+                  Calibration, Evolutionary computation, Electric breakdown,
+                  Optimization methods, Uncertainty, Computational fluid
+                  dynamics},
+  doi = {10.1109/CEC.2009.4983288}
+}
+
+ +
+@incollection{KarEibHoo2014generic,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2014,
+  editor = {Christian Igel and Dirk V. Arnold},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2014},
+  author = {Karafotias, Giorgos and  Agoston E. Eiben  and Mark Hoogendoorn},
+  title = {Generic parameter control with reinforcement learning},
+  pages = {1319--1326}
+}
+
+ +
+@incollection{KarHooEib2015eval,
+  year = 2015,
+  volume = 9028,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  editor = {Antonio M. Mora and Squillero, Giovanni},
+  title = {Evaluating reward definitions for parameter control},
+  author = {Karafotias, Giorgos and Hoogendoorn, Mark and  Agoston E. Eiben },
+  pages = {667--680},
+  doi = {10.1007/978-3-319-16549-3_54}
+}
+
+ +
+@inproceedings{KarKorSom2013,
+  url = {http://jmlr.org/proceedings/papers/v28/},
+  year = 2013,
+  volume = 28,
+  booktitle = {Proceedings of  the 30th International Conference on Machine Learning, {ICML} 2013},
+  editor = {Dasgupta, Sanjoy and McAllester, David},
+  title = {Almost optimal exploration in multi-armed bandits},
+  author = {Karnin, Zohar and Koren, Tomer and Somekh, Oren},
+  pages = {1238--1246},
+  annote = {Sequential Halving, Successive Halving}
+}
+
+ +
+@incollection{KarParStu2018lion,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 11353,
+  editor = { Roberto Battiti  and Mauro Brunato and Ilias Kotsireas and  Panos M. Pardalos },
+  year = 2018,
+  booktitle = {Learning and Intelligent Optimization, 12th International
+                  Conference, LION 12},
+  title = {Algorithm Configuration: Learning policies for the quick
+                  termination of poor performers},
+  author = {Karapetyan, Daniel and  Andrew J. Parkes  and  Thomas St{\"u}tzle },
+  pages = {220--224},
+  doi = {10.1007/978-3-030-05348-2_20}
+}
+
+ +
+@incollection{KarSmiEib2012generic,
+  year = 2012,
+  volume = 7248,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  editor = {Di Chio, Cecillia and others},
+  title = {A generic approach to parameter control},
+  author = {Karafotias, Giorgos and  Smit, Selmar K.  and  Agoston E. Eiben },
+  pages = {366--375},
+  doi = {10.1007/978-3-642-29178-4_37}
+}
+
+ +
+@inproceedings{Karmarkar1984,
+  publisher = {ACM Press},
+  year = 1984,
+  editor = {DeMillo, Richard A.},
+  booktitle = {Proceedings of the sixteenth annual {ACM} Symposium on Theory of Computing},
+  title = {A new polynomial-time algorithm for linear programming},
+  author = {Karmarkar, Narendra},
+  pages = {302--311}
+}
+
+ +
+@inproceedings{Karp1972,
+  publisher = {Springer},
+  series = {The IBM Research Symposia Series},
+  editor = {Miller, Raymond E. and Thatcher, James, W.},
+  year = 1972,
+  booktitle = {Proceedings of a symposium on the Complexity of Computer Computations,
+                  held March 20-22, 1972, at the {IBM} {T}homas {J}. {W}atson Research Center,
+                  Yorktown Heights, New York, USA},
+  title = {Reducibility among combinatorial problems},
+  author = {Karp, Richard M.},
+  pages = {85--103}
+}
+
+ +
+@incollection{KatNar1999:gecco,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  year = 1999,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 1999},
+  shorteditor = {Wolfgang Banzhaf and others},
+  editor = {Wolfgang Banzhaf and Jason M. Daida and A. E. Eiben
+                  and Max H. Garzon and Vasant Honavar and Mark
+                  J. Jakiela and Robert E. Smith},
+  author = {K. Katayama and H. Narihisa},
+  title = {Iterated Local Search Approach using Genetic Transformation to the
+  Traveling Salesman Problem},
+  volume = 1,
+  pages = {321--328}
+}
+
+ +
+@book{Kau1993order,
+  author = { Kauffman, S. A. },
+  publisher = {Oxford University Press},
+  title = {The Origins of Order},
+  year = 1993,
+  alias = {kauffman1993}
+}
+
+ +
+@inproceedings{Kazantzis02,
+  author = { Michael D. Kazantzis  and  Angus R. Simpson  and  David Kwong  and  Tan, Shyh Min },
+  title = {A new methodology for optimizing the daily
+                  operations of a pumping plant},
+  year = 2002,
+  booktitle = {Proceedings of 2002 Conference on Water Resources
+                  Planning},
+  address = {Roanoke, USA},
+  month = may,
+  organization = {ASCE}
+}
+
+ +
+@inproceedings{KeFenXuShaWan10,
+  author = {Liangjun Ke and Zuren Feng and Zongben Xu and Ke Shang and
+                  Yonggang Wang},
+  booktitle = {Circuits, Communications and System (PACCS), 2010 Second
+                  Pacific-Asia Conference on},
+  title = {A multiobjective {ACO} algorithm for rough feature selection},
+  year = 2010,
+  volume = 1,
+  pages = {207--210}
+}
+
+ +
+@incollection{KeeAirCyr2001adaptive,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = {Erik D. Goodman},
+  year = 2001,
+  booktitle = {Proceedings of the 3rd Annual Conference on Genetic and
+                  Evolutionary Computation, GECCO 2001},
+  title = {An adaptive genetic algorithm},
+  author = {Kee, Eric and Airey, Sarah and Cyre, Walling},
+  pages = {391--397}
+}
+
+ +
+@book{KelPfePis04,
+  author = {Kellerer, Hans and Ulrich Pferschy and  David Pisinger },
+  title = {Knapsack problems},
+  publisher = {Springer},
+  year = 2004
+}
+
+ +
+@inproceedings{KelPol2007cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  author = {Robert E. Keller and Riccardo Poli},
+  title = {Linear genetic programming of parsimonious metaheuristics},
+  pages = {4508--4515},
+  doi = {10.1109/CEC.2007.4425062}
+}
+
+ +
+@incollection{KelPol2008ae,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Lutton, Evelyne and Legrand, Pierrick and Parrend, Pierre and  Nicolas Monmarch{\'e}  and  Marc Schoenauer },
+  volume = 10764,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  booktitle = {EA 2017: Artificial Evolution},
+  author = {Robert E. Keller and Riccardo Poli},
+  title = {Cost-Benefit Investigation of a Genetic-Programming
+                  Hyperheuristic},
+  pages = {13--24}
+}
+
+ +
+@inproceedings{KenEbe1995pso,
+  author = { J. Kennedy  and  Eberhart, Russell C. },
+  title = {Particle Swarm Optimization},
+  booktitle = {Proceedings of International Conference on Neural Networks
+                  (ICNN'95)},
+  year = 1995,
+  pages = {1942--1948},
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  annote = {Proposed PSO},
+  doi = {10.1109/ICNN.1995.488968},
+  alias = {Kennedy95}
+}
+
+ +
+@inproceedings{KenEbe1997binpso,
+  author = { J. Kennedy  and  Eberhart, Russell C. },
+  title = {A discrete binary version of the particle swarm algorithm},
+  booktitle = {Proceedings of the 1997 IEEE International Conference on
+                  Systems, Man, and Cybernetics},
+  year = 1997,
+  pages = {4104--4108},
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  alias = {Kennedy97a}
+}
+
+ +
+@book{KenEbeShi01,
+  author = { J. Kennedy  and  Eberhart, Russell C.  and  Shi, Yuhui },
+  title = {Swarm Intelligence},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  year = 2001
+}
+
+ +
+@book{Kendall1948,
+  author = {Maurice G. Kendall},
+  title = {Rank correlation methods},
+  publisher = {Griffin},
+  address = {London},
+  year = 1948
+}
+
+ +
+@inproceedings{KerTra2016cec,
+  year = 2016,
+  isbn = {978-1-5090-0623-6},
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2016 Congress on Evolutionary Computation (CEC 2016)},
+  key = {IEEE CEC},
+  author = { Pascal Kerschke  and  Heike Trautmann },
+  title = {The \proglang{R}-package {FLACCO} for exploratory landscape
+                  analysis with applications to multi-objective optimization
+                  problems},
+  pages = {5262--5269},
+  doi = {10.1109/CEC.2016.7748359}
+}
+
+ +
+@incollection{KerWanPre2016ppsn,
+  isbn = {978-3-319-45822-9},
+  year = 2016,
+  volume = 9921,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Julia Handl  and  Emma Hart  and  Lewis, P. R.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Gabriela Ochoa  and  Ben Paechter },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIV}},
+  author = { Pascal Kerschke  and Hao Wang and  Mike Preuss  and Christian
+                  Grimme and   Andr{\'{e}} H. Deutz  and  Heike Trautmann  and  Emmerich, Michael T. M. },
+  title = {Towards Analyzing Multimodality of Continuous Multiobjective
+                  Landscapes},
+  pages = {962--972},
+  doi = {10.1007/978-3-319-45823-6_90}
+}
+
+ +
+@misc{Keras,
+  author = {Chollet, Fran\c{c}ois and others},
+  title = {Keras},
+  howpublished = {\url{https://keras.io}},
+  year = 2015
+}
+
+ +
+@incollection{Kerrisk05:POSIX-threads,
+  author = {M. Kerrisk},
+  title = {pthreads - {POSIX} Threads},
+  booktitle = {Linux Programmer's Manual},
+  publisher = {\url{https://man7.org/linux/man-pages/man7/pthreads.7.html}},
+  year = 2021,
+  type = {{Section}},
+  chapter = 7,
+  note = {(Last accessed Feb 22 2023)}
+}
+
+ +
+@incollection{KhaYaoDeb2003,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = {Khare, V. and  Xin Yao  and  Kalyanmoy Deb },
+  title = {Performance Scaling of Multi-objective Evolutionary Algorithms},
+  pages = {376--390}
+}
+
+ +
+@incollection{KhiAlbSol08,
+  volume = 5217,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Christian Blum  and  Clerc, Maurice  and  Thomas St{\"u}tzle  and A. F. T. Winfield},
+  year = 2008,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 6th
+                  International Conference, ANTS 2008},
+  author = {M. Khichane and P. Albert and  Christine Solnon },
+  title = {Integration of {ACO} in a Constraint Programming
+                  Language},
+  pages = {84--95}
+}
+
+ +
+@incollection{KhiAlbSol09,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  editor = { Thomas St{\"u}tzle },
+  volume = 5851,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Learning and Intelligent Optimization, Third International Conference, LION 3},
+  author = {M. Khichane and P. Albert and  Christine Solnon },
+  title = {An {ACO}-Based Reactive Framework for Ant Colony
+                  Optimization: First Experiments on Constraint
+                  Satisfaction Problems},
+  doi = {10.1007/978-3-642-11169-3_9},
+  pages = {119--133}
+}
+
+ +
+@inproceedings{KhuXuHooLey2009:satenstein,
+  publisher = {AAAI Press, Menlo Park, CA},
+  editor = {Craig Boutilier},
+  year = 2009,
+  booktitle = {Proceedings of  the 21st International Joint Conference on Artificial Intelligence (IJCAI-09)},
+  author = { KhudaBukhsh, A. R.  and  Lin Xu  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {{SATenstein}: Automatically Building Local Search
+                  {SAT} Solvers from Components},
+  pages = {517--524},
+  epub = {http://ijcai.org/papers09/Papers/IJCAI09-093.pdf}
+}
+
+ +
+@incollection{KimAllLop2020safe,
+  year = 2021,
+  volume = 12641,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Trustworthy AI -- Integrating Learning, Optimization and
+                  Reasoning. TAILOR 2020},
+  editor = {Fredrik Heintz and Michela Milano and   O'Sullivan, Barry },
+  author = { Kim, Youngmin  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Safe Learning and Optimization Techniques: Towards a Survey
+                  of the State of the Art},
+  pages = {123--139},
+  doi = {10.1007/978-3-030-73959-1_12},
+  abstract = {Safe learning and optimization deals with learning and
+                  optimization problems that avoid, as much as possible, the
+                  evaluation of non-safe input points, which are solutions,
+                  policies, or strategies that cause an irrecoverable loss
+                  (e.g., breakage of a machine or equipment, or life
+                  threat). Although a comprehensive survey of safe
+                  reinforcement learning algorithms was published in 2015, a
+                  number of new algorithms have been proposed thereafter, and
+                  related works in active learning and in optimization were not
+                  considered. This paper reviews those algorithms from a number
+                  of domains including reinforcement learning, Gaussian process
+                  regression and classification, evolutionary computing, and
+                  active learning. We provide the fundamental concepts on which
+                  the reviewed algorithms are based and a characterization of
+                  the individual algorithms. We conclude by explaining how the
+                  algorithms are connected and suggestions for future
+                  research.}
+}
+
+ +
+@incollection{KimAllLop2022easafe,
+  location = {Boston, Massachusetts},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Kim, Youngmin  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Are Evolutionary Algorithms Safe Optimizers?},
+  doi = {10.1145/3512290.3528818},
+  abstract = {We consider a type of constrained optimization problem, where
+                  the violation of a constraint leads to an irrevocable loss,
+                  such as breakage of a valuable experimental resource/platform
+                  or loss of human life. Such problems are referred to as safe
+                  optimization problems (SafeOPs). While SafeOPs have received
+                  attention in the machine learning community in recent years,
+                  there was little interest in the evolutionary computation
+                  (EC) community despite some early attempts between 2009 and
+                  2011. Moreover, there is a lack of acceptable guidelines on
+                  how to benchmark different algorithms for SafeOPs, an area
+                  where the EC community has significant experience in. Driven
+                  by the need for more eficient algorithms and benchmark
+                  guidelines for SafeOPs, the objective of this paper is to
+                  reignite the interest of the EC community in this problem
+                  class. To achieve this we (i) provide a formal definition of
+                  SafeOPs and contrast it to other types of optimization
+                  problems that the EC community is familiar with, (ii)
+                  investigate the impact of key SafeOP parameters on the
+                  performance of selected safe optimization algorithms, (iii)
+                  benchmark EC against state-of-the-art safe optimization
+                  algorithms from the machine learning community, and (iv)
+                  provide an open-source Python framework to replicate and
+                  extend our work.},
+  pages = {814--822},
+  numpages = 9,
+  keywords = {Bayesian optimization, constrained optimization,
+                  benchmarking, safety constraints, safe optimization}
+}
+
+ +
+@inproceedings{KimParKim2021learning,
+  editor = {M. Ranzato and A. Beygelzimer and Y. Dauphin and P.S. Liang
+                  and J. Wortman Vaughan},
+  booktitle = {Advances in Neural Information Processing Systems 34 (NeurIPS
+                  2021)},
+  year = 2021,
+  author = {Kim, Minsu and Park, Jinkyoo and Kim, Joungho},
+  title = {Learning Collaborative Policies to Solve {NP-hard} Routing
+                  Problems},
+  epub = {https://papers.nips.cc/paper_files/paper/2021/hash/564127c03caab942e503ee6f810f54fd-Abstract.html},
+  keywords = {Deep RL, TSP, prize collecting, PCTSP, CVRP, routing,
+                  attention model}
+}
+
+ +
+@inproceedings{KinBa2015adam,
+  editor = { Bengio, Yoshua  and Yann {LeCun}},
+  booktitle = {3rd International Conference on Learning Representations,
+                  {ICLR} 2015, San Diego, CA, USA, May 7-9, 2015, Conference
+                  Track Proceedings},
+  year = 2015,
+  author = {Diederik P. Kingma and Jimmy Ba},
+  title = {Adam: {A} Method for Stochastic Optimization}
+}
+
+ +
+@inproceedings{Kno2005:isda,
+  year = 2005,
+  booktitle = {Proceedings of the 5th International Conference on
+                  Intelligent Systems Design and Applications},
+  editor = {Abraham, Ajith and Paprzycki, Marcin},
+  author = { Joshua D. Knowles },
+  title = {A summary-attainment-surface plotting method for visualizing
+                  the performance of stochastic multiobjective optimizers},
+  pages = {552--557},
+  supplement = {https://www.cs.bham.ac.uk/~jdk/plot_attainments/},
+  doi = {10.1109/ISDA.2005.15},
+  abstract = {When evaluating the performance of a stochastic optimizer it
+                  is sometimes desirable to express performance in terms of the
+                  quality attained in a certain fraction of sample runs. For
+                  example, the sample median quality is the best estimator of
+                  what one would expect to achieve in 50\% of runs, and
+                  similarly for other quantiles. In multiobjective
+                  optimization, the notion still applies but the outcome of a
+                  run is measured not as a scalar (i.e. the cost of the best
+                  solution), but as an attainment surface in $k$-dimensional
+                  space (where $k$ is the number of objectives). In this paper
+                  we report an algorithm that can be conveniently used to plot
+                  summary attainment surfaces in any number of dimensions
+                  (though it is particularly suited for three). A summary
+                  attainment surface is defined as the union of all tightest
+                  goals that have been attained (independently) in precisely
+                  $s$ of the runs of a sample of $n$ runs, for any $s \in
+                  1\ldots n$, and for any $k$. We also discuss the
+                  computational complexity of the algorithm and give some
+                  examples of its use. C code for the algorithm is available
+                  from the author.}
+}
+
+ +
+@inproceedings{KnoCor1999cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1999,
+  booktitle = {Proceedings of the 1999 Congress on Evolutionary Computation
+                  (CEC 1999)},
+  key = {IEEE CEC},
+  author = { Joshua D. Knowles  and  David Corne },
+  pages = {98--105},
+  title = {The {Pareto} Archived Evolution Strategy: A New
+                  Baseline Algorithm for Multiobjective Optimisation},
+  annote = {first mention of Adaptive Grid Archiving}
+}
+
+ +
+@inproceedings{KnoCor2000mpaes,
+  month = jul,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2000,
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  key = {IEEE CEC},
+  title = {{M-PAES}: A memetic algorithm for multiobjective
+                  optimization},
+  author = { Joshua D. Knowles  and  David Corne },
+  pages = {325--332}
+}
+
+ +
+@inproceedings{KnoCor2002cec,
+  year = 2002,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2002 Congress on Evolutionary Computation (CEC'02)},
+  key = {IEEE CEC},
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {On Metrics for Comparing Non-Dominated Sets},
+  pages = {711--716}
+}
+
+ +
+@incollection{KnoCor2003emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {Instance Generators and Test Suites for the
+                  Multiobjective Quadratic Assignment Problem},
+  pages = {295--310}
+}
+
+ +
+@incollection{KnoCor2004lnems,
+  year = 2004,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  volume = 535,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  editor = { Xavier Gandibleux  and Marc Sevaux and  Kenneth S{\"o}rensen  and  V. {T'Kindt} },
+  booktitle = {Metaheuristics for Multiobjective Optimisation},
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {Bounded {Pareto} Archiving: {Theory} and Practice},
+  pages = {39--64},
+  alias = {Knowles+CorneLNEMS535},
+  doi = {10.1007/978-3-642-17144-4_2}
+}
+
+ +
+@incollection{KnoCor2005mem,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  series = {Studies in Fuzziness and Soft Computing},
+  volume = 166,
+  year = 2005,
+  editor = {Hart W. E. and Smith J. E. and Krasnogor N.},
+  booktitle = {Recent Advances in Memetic Algorithms},
+  title = {Memetic algorithms for multiobjective optimization: issues,
+                  methods and prospects},
+  author = { Joshua D. Knowles  and  David Corne },
+  pages = {313--352},
+  doi = {10.1007/3-540-32363-5_14}
+}
+
+ +
+@incollection{KnoCor2007emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  author = { Joshua D. Knowles  and  David Corne },
+  title = {Quantifying the Effects of Objective Space Dimension in
+                  Evolutionary Multiobjective Optimization},
+  pages = {757--771},
+  abstract = {The scalability of EMO algorithms is an issue of significant
+                  concern for both algorithm developers and users. A key aspect
+                  of the issue is scalability to objective space dimension,
+                  other things being equal. Here, we make some observations
+                  about the efficiency of search in discrete spaces as a
+                  function of the number of objectives, considering both
+                  uncorrelated and correlated objective values. Efficiency is
+                  expressed in terms of a cardinality-based
+                  (scaling-independent) performance indicator. Considering
+                  random sampling of the search space, we measure, empirically,
+                  the fraction of the true PF covered after p iterations, as
+                  the number of objectives grows, and for different
+                  correlations. A general analytical expression for the
+                  expected performance of random search is derived, and is
+                  shown to agree with the empirical results. We postulate that
+                  for even moderately large numbers of objectives, random
+                  search will be competitive with an EMO algorithm and show
+                  that this is the case empirically: on a function where each
+                  objective is relatively easy for an EA to optimize (an
+                  NK-landscape with K=2), random search compares favourably to
+                  a well-known EMO algorithm when objective space dimension is
+                  ten, for a range of inter-objective correlation values. The
+                  analytical methods presented here may be useful for
+                  benchmarking of other EMO algorithms.}
+}
+
+ +
+@incollection{KnoCorDeb2008,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  series = {Natural Computing Series},
+  editor = { Joshua D. Knowles  and  David Corne  and  Kalyanmoy Deb  and Chair, Deva Raj},
+  year = 2008,
+  booktitle = {Multiobjective Problem Solving from Nature},
+  title = {Introduction: {Problem} solving, {EC} and {EMO}},
+  author = { Joshua D. Knowles  and  David Corne  and  Kalyanmoy Deb },
+  pages = {1--28},
+  doi = {10.1007/978-3-540-72964-8_1}
+}
+
+ +
+@inproceedings{KnoCorFle2003,
+  year = 2003,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = dec,
+  booktitle = {Proceedings of  the 2003 Congress on Evolutionary Computation (CEC'03)},
+  key = {IEEE CEC},
+  title = {Bounded archiving using the {Lebesgue} measure},
+  author = { Joshua D. Knowles  and  David Corne  and Fleischer, Mark},
+  pages = {2490--2497}
+}
+
+ +
+@incollection{KnoCorRey09emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5467,
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  author = { Joshua D. Knowles  and  David Corne  and Alan P. Reynolds},
+  title = {Noisy Multiobjective Optimization on a Budget of 250 Evaluations},
+  pages = {36--50}
+}
+
+ +
+@techreport{KnoThiZit06:tutorial,
+  author = { Joshua D. Knowles  and  Lothar Thiele  and  Eckart Zitzler },
+  title = {A tutorial on the performance assessment of stochastic
+                  multiobjective optimizers},
+  institution = {Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Z{\"u}rich, Switzerland},
+  year = 2006,
+  type = {TIK-Report},
+  number = 214,
+  month = feb,
+  note = {Revised version},
+  epub = {https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/23822/2/KTZ2006a.pdf}
+}
+
+ +
+@incollection{KnoWatCor2001reducing,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2001},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 1993,
+  year = 2001,
+  publisher = {Springer},
+  editor = { Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele  and  Carlos A. {Coello Coello}  and  David Corne },
+  author = { Joshua D. Knowles  and Watson, Richard A. and  David Corne },
+  title = {Reducing Local Optima in Single-Objective Problems by
+                  Multi-objectivization},
+  pages = {269--283},
+  doi = {10.1007/3-540-44719-9_19},
+  annote = {Proposed multi-objectivization}
+}
+
+ +
+@phdthesis{Knowles2002PhD,
+  author = { Joshua D. Knowles },
+  title = {Local-Search and Hybrid Evolutionary Algorithms for
+                  {Pareto} Optimization},
+  school = {University of Reading, UK},
+  annote = {(Examiners: Prof. K. Deb and Prof. K. Warwick)},
+  year = 2002
+}
+
+ +
+@book{KolFri2009,
+  title = {Probabilistic graphical models: principles and techniques},
+  author = {Koller, Daphne and Friedman, Nir},
+  year = {2009},
+  publisher = {MIT Press}
+}
+
+ +
+@inproceedings{KopYos2007visualization,
+  title = {Visualization of {Pareto}-sets in evolutionary
+                  multi-objective optimization},
+  author = {Koppen, Mario and Yoshida, Kaori},
+  booktitle = {7th International Conference on Hybrid Intelligent Systems
+                  (HIS 2007)},
+  pages = {156--161},
+  year = 2007,
+  organization = {IEEE}
+}
+
+ +
+@inproceedings{KorSilOblKos07:cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  author = { P. Koro{\v s}ec  and  Jurij {\v S}ilc  and K. Oblak and F. Kosel},
+  title = {The differential ant-stigmergy algorithm: an
+                  experimental evaluation and a real-world
+                  application},
+  pages = {157--164}
+}
+
+ +
+@incollection{KorSilRob04:ants2004,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { P. Koro{\v s}ec  and  Jurij {\v S}ilc  and B. Robi{\v c}},
+  title = {Mesh-Partitioning with the Multiple Ant-Colony
+                  Algorithm},
+  pages = {430--431}
+}
+
+ +
+@incollection{KorStuExn06,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  author = { Oliver Korb  and  Thomas St{\"u}tzle  and  Thomas E. Exner },
+  title = {{PLANTS}: {Application} of ant colony optimization
+                  to structure-based drug design},
+  pages = {247--258},
+  doi = {10.1007/11839088_22}
+}
+
+ +
+@incollection{KorWall1997behavioral,
+  author = { Pekka Korhonen  and  Wallenius, Jyrki },
+  title = {Behavioral Issues in {MCDM}: {Neglected} Research Questions},
+  booktitle = {Multicriteria Analysis},
+  publisher = {Springer},
+  year = 1997,
+  editor = { Jo{\~ao} Cl{\'i}maco },
+  pages = {412--422},
+  address = {Berlin\slash Heidelberg},
+  isbn = {978-3-642-60667-0},
+  shorttitle = {Behavioral Issues in {MCDM}},
+  doi = {10.1007/978-3-642-60667-0_39},
+  abstract = {Behavior decision theorists have studied human decision
+                  making in great detail. Since the late 1960's, Einhorn,
+                  Edwards, Kahneman, Roy, Trevsky, and others have developed
+                  new thoeries to explain choice and decision behavior. Thus
+                  far this behavior research has had little impact on multiple
+                  criteria decision making (MCDM). Only a handful of
+                  MCDM-research have critically examined the behavioral
+                  underpinnings of our field. To improve the success of
+                  decision tools in practice, MCDM-research should pay more
+                  attention to the behavioral realities of decision making. In
+                  this paper, we discuss various behavioral issues relevent for
+                  MCDM based on our personal observations and experiments with
+                  human subjects. The spirit of our paper is to pose questions
+                  rather than provide definite answers.},
+  language = {en},
+  keywords = {Aspiration Level, Decision Tool, Nondominated Solution,
+                  Prefer Solution, Prospect Theory}
+}
+
+ +
+@incollection{KosVerDoe2021,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  author = {Kostovska, Ana and  Diederick Vermetten  and  Carola Doerr  and D\v{z}eroski, Sa\v{s}o and Panov, Pan\v{c}e and  Tome Eftimov },
+  title = {{OPTION}: optimization algorithm benchmarking ontology},
+  pages = {239--240}
+}
+
+ +
+@incollection{KosVerDze2023:evoapp,
+  volume = 13989,
+  series = {Lecture Notes in Computer Science},
+  address = {Switzerland},
+  publisher = {Springer Nature},
+  booktitle = {EvoApplications 2023: Applications of Evolutionary Computation},
+  year = 2023,
+  editor = {Correia, Jo\~{a}o and Smith, Stephen and Qaddoura, Raneem},
+  title = {Using Knowledge Graphs for Performance Prediction of Modular Optimization Algorithms},
+  author = {Kostovska, Ana and  Diederick Vermetten  and D{\v{z}}eroski, Sa{\v{s}}o and Panov, Pan{\v{c}}e and  Tome Eftimov  and  Carola Doerr },
+  pages = {253--268}
+}
+
+ +
+@book{KouYu1997:robustopt,
+  author = { P. Kouvelis  and  G. Yu },
+  title = {Robust discrete optimization and its applications},
+  publisher = {Kluwer Academic Publishers, Dordrecht, The Netherlands},
+  series = {Nonconvex optimization and its applications},
+  year = 1997
+}
+
+ +
+@incollection{KovSkr08,
+  adoi = {10.1007/978-3-540-87536-9},
+  booktitle = {ICANN'08: Proceedings of the 18th International Conference on
+                  Artificial Neural Networks, Part I},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 5163,
+  year = 2008,
+  publisher = {Springer},
+  editor = {Kurkova-Pohlova, Vera and Koutnik, Jan},
+  author = {O. Kov\'{a}\v{r}{\'i}k and M. Skrbek},
+  title = {Ant Colony Optimization with Castes},
+  pages = {435--442}
+}
+
+ +
+@incollection{KozEchLei2011,
+  year = 2011,
+  editor = {Slawomir Koziel and Xin-She Yang},
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  volume = 356,
+  series = {Studies in Computational Intelligence},
+  booktitle = {Computational Optimization, Methods and Algorithms},
+  author = {Slawomir Koziel and David Echeverr{\'i}a Ciaurri and Leifur Leifsson},
+  title = {Surrogate-Based Methods},
+  pages = {33--59}
+}
+
+ +
+@book{Koza1992gp,
+  author = {J. Koza},
+  title = {Genetic Programming: On the Programming of Computers
+                  By the Means of Natural Selection},
+  publisher = {MIT Press},
+  address = {Cambridge, MA},
+  year = 1992
+}
+
+ +
+@incollection{KraGlaHan2016unbounded,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2016},
+  title = {Unbounded population {MO-CMA-ES} for the bi-objective {BBOB}
+                  test suite},
+  author = {Krause, Oswin and  T. Glasmachers  and  Nikolaus Hansen  and  Christian Igel },
+  pages = {1177--1184},
+  keywords = {archiving}
+}
+
+ +
+@incollection{KraGloGoe2007,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = {Kramer, Oliver and Gloger, Bartek and Goebels, Andreas},
+  title = {An Experimental Analysis of Evolution Strategies and
+                  Particle Swarm Optimisers Using Design of Experiments},
+  pages = {674--681}
+}
+
+ +
+@inproceedings{KraHeiMil++2013,
+  author = { Krajzewicz, Daniel  and Marek Heinrich and Michela Milano and Paolo Bellavista and  Thomas St{\"u}tzle  and J{\'e}r{\^o}me H{\"a}rri and Thrasyvoulos Spyropoulos and Robbin Blokpoel and Stefan Hausberger and Martin Fellendorf},
+  title = {{COLOMBO}: Investigating the Potential of {V2X} for Traffic Management Purposes assuming low penetration Rates},
+  booktitle = {Proceedings of ITS Europe 2013},
+  year = 2013,
+  address = {Dublin, Ireland}
+}
+
+ +
+@incollection{KraLeiBloMilStu2016,
+  author = { Krajzewicz, Daniel  and Andreas Leich and Robbin Blokpoel and Michela Milano and  Thomas St{\"u}tzle },
+  title = {{COLOMBO}: Exploiting Vehicular Communications at Low Equipment Rates for Traffic Management Purposes},
+  booktitle = {Advanced Microsystems for Automotive Applications 2015: Smart Systems for Green and Automated Driving},
+  publisher = {Springer International Publishing},
+  year = 2016,
+  editor = {Tim Schulze and Beate M{\"u}ller and Gereon Meyer},
+  pages = {117--130},
+  address = {Cham, Switzerland}
+}
+
+ +
+@incollection{KraPru1978,
+  author = {Jakob Krarup and Peter Mark Pruzan},
+  title = {Computer-aided Layout Design},
+  booktitle = {Mathematical Programming in Use},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  year = 1978,
+  editor = {M. L. Balinski and C. Lemarechal},
+  volume = 9,
+  series = {Mathematical Programming Studies},
+  pages = {75--94}
+}
+
+ +
+@incollection{KreBraHofBer2009:aisc,
+  year = 2009,
+  volume = 58,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  series = {Advances in Intelligent and Soft Computing},
+  editor = { J{\"o}rn Mehnen  and  Mario K{\"o}ppen  and  Ashraf Saad  and  Ashutosh Tiwari },
+  booktitle = {Applications of Soft Computing},
+  author = { Johannes Krettek  and  Jan Braun  and  Frank Hoffmann  and  Torsten Bertram },
+  title = {Interactive Incorporation of User Preferences in Multiobjective Evolutionary Algorithms},
+  pages = {379--388}
+}
+
+ +
+@incollection{KreBraHofBer2010:ipmu,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2010,
+  volume = 6178,
+  series = {Lecture Notes in Artificial Intelligence},
+  editor = { Eyke H{\"u}llermeier  and  Rudolf Kruse  and  Frank Hoffmann },
+  booktitle = {Information Processing and Management of Uncertainty, 13th International
+                  Conference, {IPMU2010}},
+  author = { Johannes Krettek  and  Jan Braun  and  Frank Hoffmann  and  Torsten Bertram },
+  title = {Preference Modeling and Model Management for Interactive Multi-objective Evolutionary Optimization},
+  pages = {574--583}
+}
+
+ +
+@book{KruTan1978,
+  author = {William H. Kruskal and Judith M. Tanur},
+  year = 1978,
+  title = {Linear Hypotheses},
+  publisher = {Free Press},
+  volume = 1
+}
+
+ +
+@inproceedings{KukLam2005:gde3,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2005,
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  key = {IEEE CEC},
+  author = {Kukkonen, S. and Lampinen, J.},
+  title = {{GDE3}: the third evolution step of generalized differential
+                  evolution},
+  pages = {443--450}
+}
+
+ +
+@incollection{KumVas2010:www,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  editor = { Michael Rappa  and  Paul Jones  and  Juliana Freire  and  Soumen Chakrabarti },
+  booktitle = {Proceedings of the 19th International Conference on World Wide Web, WWW 2010},
+  author = { Ravi Kumar  and  Sergei Vassilvitskii },
+  title = {Generalized Distances between Rankings}
+}
+
+ +
+@incollection{Kur1990variant,
+  address = {Berlin\slash Heidelberg},
+  aseries = {Lecture Notes in Computer Science},
+  avolume = 496,
+  publisher = {Springer},
+  editor = { Hans-Paul Schwefel  and R. M{\"a}nner},
+  year = 1991,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {I}},
+  title = {A variant of evolution strategies for vector optimization},
+  author = {Kursawe, Frank},
+  pages = {193--197},
+  doi = {10.1007/BFb0029752},
+  annote = {Proposed KUR benchmark}
+}
+
+ +
+@inproceedings{LacMolHer2013cec,
+  year = 2013,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  key = {IEEE CEC},
+  author = {Benjamin Lacroix and  Daniel Molina  and  Francisco Herrera },
+  title = {Dynamically updated region based memetic algorithm
+                  for the 2013 {CEC} Special Session and Competition
+                  on Real Parameter Single Objective Optimization},
+  pages = {1945--1951}
+}
+
+ +
+@inproceedings{LarLap2014kriging,
+  author = {Lark, R. M. and Lapworth, D. J.},
+  title = {A new statistic to express the uncertainty of kriging
+                  predictions for purposes of survey planning},
+  booktitle = {EGU General Assembly Conference Abstracts},
+  year = 2014,
+  month = may,
+  eid = 2183,
+  url = {https://ui.adsabs.harvard.edu/abs/2014EGUGA..16.2183L}
+}
+
+ +
+@book{LarLoz2002eda,
+  author = {Larra{\~n}aga, Pedro and  Jos{\'e} A. Lozano },
+  title = {Estimation of Distribution Algorithms: A New Tool for
+                  Evolutionary Computation},
+  publisher = {Kluwer Academic Publishers},
+  address = { Boston, MA},
+  year = 2002
+}
+
+ +
+@book{Larman2005,
+  author = {Craig Larman},
+  title = {Applying {UML} and Patterns: An Introduction to Object-Oriented Analysis and Design and Iterative Development},
+  publisher = {Prentice Hall, Englewood Cliffs, NJ},
+  year = 2004,
+  edition = {3rd}
+}
+
+ +
+@incollection{LauThiZitDeb2002archiving,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = { Langdon, William B.  and others},
+  year = 2002,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2002},
+  title = {Archiving with guaranteed convergence and diversity in
+                  multi-objective optimization},
+  author = { Marco Laumanns  and  Lothar Thiele  and  Eckart Zitzler  and  Kalyanmoy Deb },
+  pages = {439--447}
+}
+
+ +
+@unpublished{LauZen2010prep,
+  author = { Marco Laumanns  and Zenklusen, Rico},
+  title = {Stochastic convergence of random search methods to fixed size
+                  {Pareto} front approximations},
+  note = {(submitted)},
+  month = nov,
+  year = 2010,
+  annote = {Published as~\cite{LauZen2011ejor}. Keep this reference for historical reasons.}
+}
+
+ +
+@inproceedings{LauZitThi2000:elitism,
+  month = jul,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2000,
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  key = {IEEE CEC},
+  author = { Marco Laumanns  and  Eckart Zitzler  and  Lothar Thiele },
+  title = {A unified model for multi-objective evolutionary algorithms
+                  with elitism},
+  pages = {46--53}
+}
+
+ +
+@book{LawLenRinShm85,
+  author = {E. L. Lawler and J. K. Lenstra and A. H. G. {Rinnooy Kan} and
+                  D. B. Shmoys},
+  title = {The Traveling Salesman Problem},
+  publisher = {John Wiley \& Sons},
+  address = { Chichester, UK},
+  year = 1985
+}
+
+ +
+@incollection{LegAlb2013acodyn,
+  author = { Leguizam\'{o}n, Guillermo  and  Alba, Enrique },
+  year = 2013,
+  booktitle = {Metaheuristics for Dynamic Optimization},
+  volume = 433,
+  series = {Studies in Computational Intelligence},
+  editor = { Alba, Enrique  and Nakib, Amir and Siarry, Patrick},
+  doi = {10.1007/978-3-642-30665-5_9},
+  title = {Ant Colony Based Algorithms for Dynamic Optimization
+                  Problems},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  pages = {189--210},
+  language = {English}
+}
+
+ +
+@inproceedings{LegMic1999:cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1999,
+  booktitle = {Proceedings of the 1999 Congress on Evolutionary Computation
+                  (CEC 1999)},
+  key = {IEEE CEC},
+  author = { Leguizam\'{o}n, Guillermo  and  Zbigniew Michalewicz },
+  title = {A New Version of {Ant} {System} for Subset Problems},
+  pages = {1459--1464},
+  alias = {LegMic99}
+}
+
+ +
+@book{LemPopBan2003,
+  title = {Shaping the Next One Hundred Years: New Methods for
+                  Quantitative, Long Term Policy Analysis},
+  author = {Lempert, R. J. and Popper, S. and  Bankes, Steven C. },
+  publisher = {RAND},
+  year = 2003
+}
+
+ +
+@incollection{LesDumStu04:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = {L. Lessing and Irina Dumitrescu and  Thomas St{\"u}tzle },
+  title = {A Comparison Between {ACO} Algorithms for the Set
+                  Covering Problem},
+  pages = {1--12}
+}
+
+ +
+@book{Lewis2016,
+  title = {A Guide to Graph Colouring: Algorithms and Applications},
+  author = { Lewis, Rhyd M. R. },
+  year = 2016,
+  doi = {10.1007/978-3-319-25730-3},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  annote = {Supplementary material available at \cite{Lewis2016sup}}
+}
+
+ +
+@misc{Lewis2016sup,
+  author = { Lewis, Rhyd M. R. },
+  title = {Suite of Graph Colouring Algorithms -- Supplementary Material
+                  to the Book ``{A} Guide to Graph Colouring: Algorithms and
+                  Applications''},
+  howpublished = {\url{http://rhydlewis.eu/resources/gCol.zip}},
+  year = 2016
+}
+
+ +
+@inproceedings{LeyNudAnd2003ijcai,
+  booktitle = {Proceedings of  the 18th International Joint Conference on Artificial Intelligence (IJCAI-03)},
+  epub = {http://ijcai.org/proceedings/2003},
+  year = 2003,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = {Georg Gottlob and Toby Walsh},
+  author = { Kevin Leyton-Brown  and Nudelman, Eugene and Andrew, Galen and
+                  McFadden, Jim and Shoham, Yoav},
+  title = {A Portfolio Approach to Algorithm Selection},
+  pages = {1542--1543},
+  annote = {First example of modern algorithm selection for
+                  optimisation?}
+}
+
+ +
+@incollection{LeyNudSho2002,
+  year = 2002,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Principles and Practice of Constraint Programming, CP 2002},
+  editor = {van Hentenryck, Pascal },
+  title = {Learning the Empirical Hardness of Optimization Problems: The
+                  Case of Combinatorial Auctions},
+  author = { Kevin Leyton-Brown  and Nudelman, Eugene and Shoham, Yoav},
+  pages = {556--572}
+}
+
+ +
+@incollection{LeyPeaSho2000acmec,
+  editor = {Anant Jhingran and others},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2000,
+  booktitle = {ACM Conference on Electronic Commerce (EC-00)},
+  author = { Kevin Leyton-Brown  and  M. Pearson and Y. Shoham},
+  title = {Towards a Universal Test Suite for Combinatorial Auction
+                  Algorithms},
+  pages = {66--76},
+  doi = {10.1145/352871.352879},
+  annote = {CPLEX-regions200 benchmark set,
+                  \url{http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/results.html}}
+}
+
+ +
+@inproceedings{LiLiTanYao2014taa,
+  year = 2014,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2014 Congress on Evolutionary Computation (CEC 2014)},
+  key = {IEEE CEC},
+  title = {An Improved Two Archive Algorithm for Many-Objective
+                  Optimization},
+  author = {Li, Bingdong and Li, Jinlong and Tang, Ke and  Xin Yao },
+  pages = {2869--2876}
+}
+
+ +
+@inproceedings{LiWanYuZhaLi08,
+  author = {Z. Li and Y. Wang and J. Yu and Y. Zhang and X. Li},
+  title = {A Novel Cloud-Based Fuzzy Self-Adaptive Ant Colony
+                  System},
+  booktitle = {ICNC'08: Proceedings of the 2008 Fourth
+                  International Conference on Natural Computation},
+  volume = 7,
+  year = 2008,
+  pages = {460--465},
+  publisher = {IEEE Computer Society},
+  address = {Washington, DC}
+}
+
+ +
+@incollection{LiYanLiu2015pci,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  title = {A performance comparison indicator for {Pareto} front
+                  approximations in many-objective optimization},
+  author = { Li, Miqing  and Yang, Shengxiang and Liu, Xiaohui},
+  pages = {703--710},
+  annote = {Proposed PCI indicator}
+}
+
+ +
+@incollection{LiYanLiuShe2013many,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  title = {A Comparative Study on Evolutionary Algorithms for
+                  Many-Objective Optimization},
+  author = { Li, Miqing  and Yang, Shengxiang and Liu, Xiaohui and Shen,
+                  Ruimin},
+  pages = {261--275}
+}
+
+ +
+@incollection{LiYao2019emo,
+  isbn = {978-3-030-12597-4},
+  year = 2019,
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  volume = 11411,
+  series = {Lecture Notes in Computer Science},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2019},
+  editor = { Kalyanmoy Deb  and Erik D. Goodman and  Carlos A. {Coello Coello}  and Kathrin
+                  Klamroth and  Kaisa Miettinen  and Sanaz Mostaghim and Patrick
+                  Reed},
+  author = { Li, Miqing  and  Xin Yao },
+  title = {An Empirical Investigation of the Optimality and Monotonicity
+                  Properties of Multiobjective Archiving Methods},
+  pages = {15--26},
+  doi = {10.1007/978-3-030-12598-1_2}
+}
+
+ +
+@incollection{LiYevBas2017,
+  editor = {Heike Trautmann and G{\"{u}}nter Rudolph and Kathrin Klamroth
+                  and Oliver Sch{\"{u}}tze and Margaret M. Wiecek and Yaochu
+                  Jin and Christian Grimme},
+  year = 2017,
+  volume = 10173,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2017},
+  author = {Li, Longmei and Yevseyeva, Iryna and Basto-Fernandes,
+                  Vitor and  Heike Trautmann  and Jing, Ning and  Emmerich, Michael T. M. },
+  title = {Building and using an ontology of preference-based
+                  multiobjective evolutionary algorithms},
+  pages = {406--421}
+}
+
+ +
+@incollection{LiaMonDogStuDor2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = {Liao, Tianjun  and  Marco A. {Montes de Oca}  and  Do\v{g}an Ayd{\i}n  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {An Incremental Ant Colony Algorithm with Local Search for Continuous Optimization},
+  pages = {125--132}
+}
+
+ +
+@incollection{LiaMonStu2011gecco,
+  year = 2011,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2011},
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  author = {Liao, Tianjun  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle },
+  title = {Tuning Parameters across Mixed Dimensional
+                  Instances: A Performance Scalability Study of
+                  {Sep-G-CMA-ES}},
+  annote = {Workshop on Scaling Behaviours of Landscapes,
+                  Parameters and Algorithms},
+  pages = {703--706}
+}
+
+ +
+@inproceedings{LiaStu2013cec,
+  year = 2013,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  key = {IEEE CEC},
+  author = {Liao, Tianjun  and  Thomas St{\"u}tzle },
+  title = {Benchmark results for a simple hybrid algorithm on
+                  the {CEC} 2013 benchmark set for real-parameter
+                  optimization},
+  pages = {1938--1944}
+}
+
+ +
+@phdthesis{Liao2013,
+  author = {Liao, Tianjun },
+  title = {Population-based Heuristic Algorithms for Continuous and Mixed Discrete-Continuous Optimization Problem},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2013
+}
+
+ +
+@incollection{LieDerVerAguTan2017evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 10197,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  booktitle = {Proceedings of EvoCOP 2017 -- 17th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Bin Hu  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  author = { Arnaud Liefooghe  and  Bilel Derbel  and  Verel, S{\'e}bastien  and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  title = {Towards Landscape-Aware Automatic Algorithm Configuration:
+                  Preliminary Experiments on Neutral and Rugged Landscapes},
+  pages = {215--232},
+  doi = {10.1007/978-3-319-55453-2_15}
+}
+
+ +
+@incollection{LieDerVerLop2018ppsn,
+  volume = 11102,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Arnaud Liefooghe  and  Bilel Derbel  and  Verel, S{\'e}bastien  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  title = {On {Pareto} Local Optimal Solutions Networks},
+  pages = {232--244},
+  doi = {10.1007/978-3-319-99259-4_19}
+}
+
+ +
+@incollection{LieLop2023many,
+  location = {Lisbon, Portugal},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2023},
+  annote = {ISBN: 9798400701191},
+  address = { New York, NY},
+  year = 2023,
+  publisher = {ACM Press},
+  editor = {Silva, Sara and  Lu{\'i}s Paquete },
+  author = { Arnaud Liefooghe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Many-objective (Combinatorial) Optimization is Easy},
+  pages = {704--712},
+  doi = {10.1145/3583131.3590475},
+  abstract = {It is a common held assumption that problems with many
+                  objectives are harder to optimize than problems with two or
+                  three objectives. In this paper, we challenge this assumption
+                  and provide empirical evidence that increasing the number of
+                  objectives tends to reduce the difficulty of the landscape
+                  being optimized. Of course, increasing the number of
+                  objectives brings about other challenges, such as an increase
+                  in the computational effort of many operations, or the memory
+                  requirements for storing non-dominated solutions. More
+                  precisely, we consider a broad range of multi- and
+                  many-objective combinatorial benchmark problems, and we
+                  measure how the number of objectives impacts the dominance
+                  relation among solutions, the connectedness of the Pareto
+                  set, and the landscape multimodality in terms of local
+                  optimal solutions and sets. Our analysis shows the limit
+                  behavior of various landscape features when adding more
+                  objectives to a problem. Our conclusions do not contradict
+                  previous observations about the inability of
+                  Pareto-optimality to drive search, but we explain these
+                  observations from a different perspective. Our findings have
+                  important implications for the design and analysis of
+                  many-objective optimization algorithms.}
+}
+
+ +
+@incollection{LieLopPaqVer2018gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2018,
+  editor = { Aguirre, Hern\'{a}n E.  and Keiki Takadama},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2018},
+  author = { Arnaud Liefooghe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Verel, S{\'e}bastien },
+  title = {Dominance, Epsilon, and Hypervolume Local Optimal Sets in
+                  Multi-objective Optimization, and How to Tell the Difference},
+  pages = {324--331},
+  doi = {10.1145/3205455.3205572}
+}
+
+ +
+@incollection{LieMesHum2009sls,
+  volume = 5752,
+  series = {Lecture Notes in Computer Science},
+  year = 2009,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Engineering Stochastic Local Search
+                  Algorithms. Designing, Implementing and Analyzing
+                  Effective Heuristics. SLS~2009},
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  title = {A Study on Dominance-based Local Search Approaches
+                  for Multiobjective Combinatorial Optimization},
+  author = { Arnaud Liefooghe  and  Salma Mesmoudi  and  J{\'e}r{\'e}mie Humeau  and  Laetitia Jourdan  and  Talbi, El-Ghazali },
+  pages = {120--124}
+}
+
+ +
+@incollection{LiePaqSim2011,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 6622,
+  year = 2011,
+  editor = { Peter Merz  and  Jin-Kao Hao },
+  booktitle = {Proceedings of EvoCOP 2011 -- 11th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = { Arnaud Liefooghe  and  Lu{\'i}s Paquete  and Sim{\={o}}es, Marco and  Jos{\'e} Rui Figueira },
+  title = {Connectedness and Local Search for Bicriteria Knapsack Problems},
+  pages = {48--59},
+  doi = {10.1007/978-3-642-20364-0_5}
+}
+
+ +
+@incollection{LieVerAguTan2013ea,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2013,
+  volume = 8752,
+  fulleditor = {Pierrick Legrand and Marc{-}Michel Corsini and  Jin-Kao Hao  and Nicolas Monmarch{\'{e}} and Evelyne Lutton and Marc
+                  Schoenauer},
+  editor = {Pierrick Legrand and others},
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Artificial Evolution: 11th International Conference, Evolution Artificielle, EA, 2013},
+  title = {What Makes an Instance Difficult for Black-box 0--1
+                  Evolutionary Multiobjective Optimizers?},
+  author = { Arnaud Liefooghe  and  Bilel Derbel  and  Verel, S{\'e}bastien  and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  pages = {3--15},
+  doi = {10.1007/978-3-319-11683-9_1}
+}
+
+ +
+@incollection{LieVerLac2021gecco,
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  author = { Arnaud Liefooghe  and  Verel, S{\'e}bastien  and Benjamin Lacroix and Alexandru{-}Ciprian Zavoianu and  McCall, John },
+  doi = {10.1145/3449639.3459353},
+  pages = {421--429},
+  title = {Landscape features and automated algorithm selection for
+                  multi-objective interpolated continuous optimisation
+                  problems}
+}
+
+ +
+@incollection{LieVerPaqHao2015bubqp,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = { Arnaud Liefooghe  and  Verel, S{\'e}bastien  and  Lu{\'i}s Paquete  and  Jin-Kao Hao },
+  title = {Experiments on Local Search for Bi-objective Unconstrained
+                  Binary Quadratic Programming},
+  pages = {171--186},
+  abstract = {This article reports an experimental analysis on stochastic
+                  local search for approximating the Pareto set of bi-objective
+                  unconstrained binary quadratic programming problems. First,
+                  we investigate two scalarizing strategies that iteratively
+                  identify a high-quality solution for a sequence of
+                  sub-problems. Each sub-problem is based on a static or
+                  adaptive definition of weighted-sum aggregation coefficients,
+                  and is addressed by means of a state-of-the-art
+                  single-objective tabu search procedure. Next, we design a
+                  Pareto local search that iteratively improves a set of
+                  solutions based on a neighborhood structure and on the Pareto
+                  dominance relation. At last, we hybridize both classes of
+                  algorithms by combining a scalarizing and a Pareto local
+                  search in a sequential way. A comprehensive experimental
+                  analysis reveals the high performance of the proposed
+                  approaches, which substantially improve upon previous
+                  best-known solutions. Moreover, the obtained results show the
+                  superiority of the hybrid algorithm over non-hybrid ones in
+                  terms of solution quality, while requiring a competitive
+                  computational cost. In addition, a number of structural
+                  properties of the problem instances allow us to explain the
+                  main difficulties that the different classes of local search
+                  algorithms have to face.}
+}
+
+ +
+@book{Lilja2000measuring,
+  author = {Lilja, David J.},
+  title = {Measuring Computer Performance: A Practitioner's Guide},
+  doi = {10.1017/CBO9780511612398},
+  publisher = {Cambridge University Press},
+  year = 2000,
+  abstract = {Measuring Computer Performance sets out the fundamental
+                  techniques used in analyzing and understanding the
+                  performance of computer systems. Throughout the book, the
+                  emphasis is on practical methods of measurement, simulation,
+                  and analytical modeling. The author discusses performance
+                  metrics and provides detailed coverage of the strategies used
+                  in benchmark programmes. He gives intuitive explanations of
+                  the key statistical tools needed to interpret measured
+                  performance data. He also describes the general 'design of
+                  experiments' technique, and shows how the maximum amount of
+                  information can be obtained for the minimum effort. The book
+                  closes with a chapter on the technique of queueing
+                  analysis. Appendices listing common probability distributions
+                  and statistical tables are included, along with a glossary of
+                  important technical terms. This practically-oriented book
+                  will be of great interest to anyone who wants a detailed, yet
+                  intuitive, understanding of computer systems performance
+                  analysis.}
+}
+
+ +
+@inproceedings{LinHooHutSch2015aaai,
+  year = 2015,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Blai Bonet and Sven Koenig},
+  title = {{AutoFolio}: Algorithm Configuration for Algorithm Selection},
+  author = { Marius Thomas Lindauer  and  Holger H. Hoos  and  Frank Hutter  and Schaub, Torsten}
+}
+
+ +
+@inproceedings{LinLuo07,
+  author = {W. Ling and H. Luo},
+  title = {An Adaptive Parameter Control Strategy for Ant
+                  Colony Optimization},
+  booktitle = {CIS'07: Proceedings of the 2007 International
+                  Conference on Computational Intelligence and
+                  Security},
+  year = 2007,
+  pages = {142--146},
+  publisher = {IEEE Computer Society},
+  address = {Washington, DC}
+}
+
+ +
+@misc{LocalSolver,
+  author = {{Innovation 24}},
+  title = {{LocalSolver}},
+  howpublished = {\url{http://www.localsolver.com}},
+  note = {Last visited, August 15, 2016},
+  year = 2016
+}
+
+ +
+@incollection{LodTra2013,
+  year = 2013,
+  editor = {Topaluglu, Huseyin},
+  publisher = {{INFORMS}},
+  booktitle = {Theory Driven by Influential Applications},
+  author = { Andrea Lodi  and Tramontani, Andrea},
+  title = {Performance Variability in Mixed-Integer Programming},
+  pages = {1--12}
+}
+
+ +
+@misc{LodiEtAl2004sup,
+  author = { Andrea Lodi  and  Silvano Martello  and  Vigo, Daniele },
+  title = {Two- and Three-Dimensional Bin Packing -- Source Code of
+                  {TSpack}},
+  howpublished = {\url{https://site.unibo.it/operations-research/en/research/library-of-codes-and-instances-1/tspack-tar.gz/@@download/file/TSpack.tar.gz}},
+  year = 2004
+}
+
+ +
+@incollection{LohNow2013faster,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  doi = {10.1007/978-3-642-40935-6},
+  year = 2013,
+  volume = 8139,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Proceedings of Algorithmic Learning Theory},
+  editor = {Sanjay Jain and R{\'{e}}mi Munos and Frank Stephan and Thomas
+                  Zeugmann},
+  title = {Faster {Hoeffding} Racing: {Bernstein} Races via Jackknife
+                  Estimates},
+  author = {Loh, Po-Ling and Nowozin, Sebastian},
+  pages = {203--217}
+}
+
+ +
+@incollection{Lop07:HPC_ACO,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  editor = {Paola Alberigo and Giovanni Erbacci and Francesca
+                  Garofalo and Silvia Monfardini},
+  booktitle = {Science and Sumpercomputing in Europe},
+  title = {High Performance Ant Colony Optimisation of the Pump
+                  Scheduling Problem},
+  publisher = {CINECA},
+  year = 2007,
+  pages = {371--375},
+  isbn = {978-88-86037-21-1}
+}
+
+ +
+@techreport{LopBlu08:tsptw,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Christian Blum },
+  title = {Beam-{ACO} Based on Stochastic Sampling: {A} Case Study on
+                  the {TSP} with Time Windows},
+  institution = {Department LSI, Universitat Polit{\`e}cnica de Catalunya},
+  year = 2008,
+  number = {LSI-08-28},
+  note = {Extended version published in Computers \& Operations Research~\cite{LopBlu2010cor}}
+}
+
+ +
+@incollection{LopBlu09:evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5482,
+  year = 2009,
+  editor = { Carlos Cotta  and P. Cowling},
+  booktitle = {Proceedings of EvoCOP 2009 -- 9th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Christian Blum  and  Dhananjay Thiruvady  and  Andreas T. Ernst  and  Bernd Meyer },
+  title = {Beam-{ACO} based on stochastic sampling for makespan
+                  optimization concerning the {TSP} with time windows},
+  pages = {97--108},
+  doi = {10.1007/978-3-642-01009-5_9},
+  alias = {Lop++09}
+}
+
+ +
+@incollection{LopBlu09:lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2009,
+  editor = { Thomas St{\"u}tzle },
+  volume = 5851,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Learning and Intelligent Optimization, Third International Conference, LION 3},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Christian Blum },
+  title = {Beam-{ACO} Based on Stochastic Sampling: {A} Case
+                  Study on the {TSP} with Time Windows},
+  pages = {59--73},
+  doi = {10.1007/978-3-642-11169-3_5}
+}
+
+ +
+@incollection{LopChiGil2022evo,
+  volume = 13224,
+  series = {Lecture Notes in Computer Science},
+  address = {Switzerland},
+  publisher = {Springer Nature},
+  booktitle = {EvoApplications 2022: Applications of Evolutionary Computation},
+  year = 2022,
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and others},
+  fulleditor = { Jim{\'e}nez Laredo, Juan Luis  and Hidalgo Perez, J. Ignacio  and Oluwatoyin Babaagba, Kehinde},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Chicano, Francisco  and Rodrigo Gil-Merino},
+  title = {The Asteroid Routing Problem: A Benchmark for Expensive
+                  Black-Box Permutation Optimization},
+  pages = {124--140},
+  abstract = {Inspired by the recent 11th Global Trajectory Optimisation
+                  Competition, this paper presents the asteroid routing problem
+                  (ARP) as a realistic benchmark of algorithms for expensive
+                  bound-constrained black-box optimization in permutation
+                  space. Given a set of asteroids' orbits and a departure
+                  epoch, the goal of the ARP is to find the optimal sequence
+                  for visiting the asteroids, starting from Earth's orbit, in
+                  order to minimize both the cost, measured as the sum of the
+                  magnitude of velocity changes required to complete the trip,
+                  and the time, measured as the time elapsed from the departure
+                  epoch until visiting the last asteroid. We provide
+                  open-source code for generating instances of arbitrary sizes
+                  and evaluating solutions to the problem.  As a preliminary
+                  analysis, we compare the results of two methods for expensive
+                  black-box optimization in permutation spaces, namely,
+                  Combinatorial Efficient Global Optimization (CEGO), a
+                  Bayesian optimizer based on Gaussian processes, and
+                  Unbalanced Mallows Model (UMM), an estimation-of-distribution
+                  algorithm based on probabilistic Mallows models. We
+                  investigate the best permutation representation for each
+                  algorithm, either rank-based or order-based. Moreover, we
+                  analyze the effect of providing a good initial solution,
+                  generated by a greedy nearest neighbor heuristic, on the
+                  performance of the algorithms. The results suggest directions
+                  for improvements in the algorithms being compared.},
+  keywords = {Spacecraft Trajectory Optimization, Unbalanced Mallows Model,
+                  Combinatorial Efficient Global Optimization, Estimation of
+                  Distribution Algorithms, Bayesian Optimization},
+  supplement = {https://doi.org/10.5281/zenodo.5725837},
+  doi = {10.1007/978-3-031-02462-7_9},
+  epub = {https://arxiv.org/abs/2203.15708}
+}
+
+ +
+@misc{LopDubPerStuBir2016irace-supp,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and   P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {The {\rpackage{irace}} Package: Iterated Racing for Automatic Algorithm
+                  Configuration (Supplementary Material)},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2016-003}},
+  year = 2016
+}
+
+ +
+@techreport{LopDubStu2011irace,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {The {\rpackage{irace}} package, Iterated Race for Automatic
+                  Algorithm Configuration},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  number = {TR/IRIDIA/2011-004},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-004.pdf},
+  note = {Published in Operations Research Perspectives~\cite{LopDubPerStuBir2016irace}}
+}
+
+ +
+@incollection{LopKno2015emo,
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and  Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  volume = 9019,
+  year = 2015,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {II}},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles },
+  title = {Machine Decision Makers as a Laboratory for Interactive {EMO}},
+  pages = {295--309},
+  abstract = {A key challenge, perhaps the central challenge, of
+                  multi-objective optimization is how to deal with candidate
+                  solutions that are ultimately evaluated by the hidden or
+                  unknown preferences of a human decision maker (DM) who
+                  understands and cares about the optimization problem.
+                  Alternative ways of addressing this challenge exist but
+                  perhaps the favoured one currently is the interactive
+                  approach (proposed in various forms). Here, an evolutionary
+                  multi-objective optimization algorithm (EMOA) is controlled
+                  by a series of interactions with the DM so that preferences
+                  can be elicited and the direction of search controlled. MCDM
+                  has a key role to play in designing and evaluating these
+                  approaches, particularly in testing them with real DMs, but
+                  so far quantitative assessment of interactive EMOAs has been
+                  limited.  In this paper, we propose a conceptual framework
+                  for this problem of quantitative assessment, based on the
+                  definition of machine decision makers (machine DMs), made
+                  somewhat realistic by the incorporation of various
+                  non-idealities. The machine DM proposed here draws from
+                  earlier models of DM biases and inconsistencies in the MCDM
+                  literature.  As a practical illustration of our approach, we
+                  use the proposed machine DM to study the performance of an
+                  interactive EMOA, and discuss how this framework could help
+                  in the evaluation and development of better interactive
+                  EMOAs.},
+  doi = {10.1007/978-3-319-15892-1_20}
+}
+
+ +
+@incollection{LopKnoLau2011emo,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles  and  Marco Laumanns },
+  title = {On Sequential Online Archiving of Objective Vectors},
+  pages = {46--60},
+  doi = {10.1007/978-3-642-19893-9_4},
+  abstract = {In this paper, we examine the problem of maintaining
+                  an approximation of the set of nondominated points
+                  visited during a multiobjective optimization, a
+                  problem commonly known as archiving. Most of the
+                  currently available archiving algorithms are
+                  reviewed, and what is known about their convergence
+                  and approximation properties is summarized. The main
+                  scenario considered is the restricted case where the
+                  archive must be updated online as points are
+                  generated one by one, and at most a fixed number of
+                  points are to be stored in the archive at any one
+                  time. In this scenario, the better-monotonicity of
+                  an archiving algorithm is proposed as a weaker, but
+                  more practical, property than negative efficiency
+                  preservation. This paper shows that
+                  hypervolume-based archivers and a recently proposed
+                  multi-level grid archiver have this property. On the
+                  other hand, the archiving methods used by SPEA2 and
+                  NSGA-II do not, and they may better-deteriorate with
+                  time. The better-monotonicity property has meaning
+                  on any input sequence of points. We also classify
+                  archivers according to limit properties,
+                  i.e. convergence and approximation properties of the
+                  archiver in the limit of infinite (input) samples
+                  from a finite space with strictly positive
+                  generation probabilities for all points. This paper
+                  establishes a number of research questions, and
+                  provides the initial framework and analysis for
+                  answering them.},
+  annote = {Revised version available at \url{http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2011-001.pdf}}
+}
+
+ +
+@incollection{LopLiaStu2012ppsn,
+  volume = 7491,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {I}},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and Liao, Tianjun  and  Thomas St{\"u}tzle },
+  title = {On the anytime behavior of {IPOP-CMA-ES}},
+  pages = {357--366},
+  doi = {10.1007/978-3-642-32937-1_36}
+}
+
+ +
+@misc{LopLiaStu2012ppsn-supp,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and Liao, Tianjun  and  Thomas St{\"u}tzle },
+  title = {On the anytime behavior of {IPOP-CMA-ES}: Supplementary material},
+  howpublished = {\url{https://iridia.ulb.ac.be/supp/IridiaSupp2012-010/IridiaSupp2012-010.pdf}},
+  year = 2012
+}
+
+ +
+@incollection{LopLieVer2014ppsn,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien },
+  doi = {10.1007/978-3-319-10762-2_61},
+  title = {Local Optimal Sets and Bounded Archiving on Multi-objective
+                  {NK}-Landscapes with Correlated Objectives},
+  pages = {621--630},
+  abstract = {The properties of local optimal solutions in multi-objective
+                  combinatorial optimization problems are crucial for the
+                  effectiveness of local search algorithms, particularly when
+                  these algorithms are based on Pareto dominance. Such local
+                  search algorithms typically return a set of mutually
+                  nondominated Pareto local optimal (PLO) solutions, that is, a
+                  PLO-set. This paper investigates two aspects of PLO-sets by
+                  means of experiments with Pareto local search (PLS). First,
+                  we examine the impact of several problem characteristics on
+                  the properties of PLO-sets for multi-objective NK-landscapes
+                  with correlated objectives. In particular, we report that
+                  either increasing the number of objectives or decreasing the
+                  correlation between objectives leads to an exponential
+                  increment on the size of PLO-sets, whereas the variable
+                  correlation has only a minor effect. Second, we study the
+                  running time and the quality reached when using bounding
+                  archiving methods to limit the size of the archive handled by
+                  PLS, and thus, the maximum size of the PLO-set found. We
+                  argue that there is a clear relationship between the running
+                  time of PLS and the difficulty of a problem instance.}
+}
+
+ +
+@inproceedings{LopMasMarStu2013mista,
+  year = 2013,
+  editor = { Graham Kendall  and  Vanden Berghe, Greet   and Barry McCollum},
+  address = {Gent, Belgium},
+  booktitle = {Multidisciplinary International Conference on Scheduling:
+                  Theory and Applications (MISTA 2013)},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Franco Mascia  and  Marie-El{\'e}onore Marmion  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of a Hybrid Iterated Local Search for the
+                  Multi-Mode Resource-Constrained Multi-Project Scheduling
+                  Problem},
+  pages = {1--6},
+  epub = {https://hal.inria.fr/hal-01094681}
+}
+
+ +
+@incollection{LopPaqStu04:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {On the Design of {ACO} for the Biobjective Quadratic
+                  Assignment Problem},
+  pages = {214--225},
+  doi = {10.1007/978-3-540-28646-2_19}
+}
+
+ +
+@techreport{LopPaqStu04:hybrid,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Hybrid Population-based Algorithms for the
+                  Bi-objective Quadratic Assignment Problem},
+  institution = {FG Intellektik, FB Informatik, TU Darmstadt},
+  year = 2004,
+  number = {AIDA--04--11},
+  month = dec,
+  note = {Published in Journal of Mathematical Modelling and Algorithms~\cite{LopPaqStu05:jmma}}
+}
+
+ +
+@incollection{LopPaqStu09emaa,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Exploratory Analysis of Stochastic Local Search
+                  Algorithms in Biobjective Optimization},
+  pages = {209--222},
+  doi = {10.1007/978-3-642-02538-9_9},
+  abstract = {This chapter introduces two Perl programs that
+                  implement graphical tools for exploring the
+                  performance of stochastic local search algorithms
+                  for biobjective optimization problems. These tools
+                  are based on the concept of the empirical attainment
+                  function (EAF), which describes the probabilistic
+                  distribution of the outcomes obtained by a
+                  stochastic algorithm in the objective space. In
+                  particular, we consider the visualization of
+                  attainment surfaces and differences between the
+                  first-order EAFs of the outcomes of two
+                  algorithms. This visualization allows us to identify
+                  certain algorithmic behaviors in a graphical way.
+                  We explain the use of these visualization tools and
+                  illustrate them with examples arising from
+                  practice.}
+}
+
+ +
+@misc{LopPaqStu2010:eaftools,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {{EAF} Graphical Tools},
+  year = 2010,
+  howpublished = {\url{http://lopez-ibanez.eu/eaftools}},
+  note = {These tools are described in the book chapter
+                  ``\emph{Exploratory analysis of stochastic local search
+                  algorithms in biobjective
+                  optimization}''~\cite{LopPaqStu09emaa}.},
+  annote = {Please cite the book chapter, not this.}
+}
+
+ +
+@techreport{LopPerDubStuBir2016iraceguide,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and   P{\'e}rez C{\'a}ceres, Leslie and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {The {\rpackage{irace}} package: User Guide},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2016,
+  number = {TR/IRIDIA/2016-004},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2016-004.pdf}
+}
+
+ +
+@inproceedings{LopPraPae08:WDSA,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Parallel Optimisation Of Pump Schedules With A
+                  Thread-Safe Variant Of {EPANET} Toolkit},
+  booktitle = {Proceedings of the 10th Annual Water Distribution
+                  Systems Analysis Conference (WDSA 2008)},
+  year = 2008,
+  editor = { Jakobus E. van Zyl  and  A. A. Ilemobade  and  H. E. Jacobs },
+  month = aug,
+  doi = {10.1061/41024(340)40},
+  publisher = {ASCE}
+}
+
+ +
+@incollection{LopPraPae:gecco07,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2007,
+  editor = {Dirk Thierens and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Solving Optimal Pump Control Problem using
+                  {\MaxMinAntSystem}},
+  volume = 1,
+  pages = 176,
+  doi = {10.1145/1276958.1276990}
+}
+
+ +
+@inproceedings{LopPraPaech05:cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = sep,
+  year = 2005,
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  key = {IEEE CEC},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Multi-objective Optimisation of the Pump Scheduling
+                  Problem using {SPEA2}},
+  pages = {435--442},
+  volume = 1,
+  doi = {10.1109/CEC.2005.1554716}
+}
+
+ +
+@inproceedings{LopPraPaech:ccwi2005,
+  month = sep,
+  address = {University of Exeter, UK},
+  volume = 1,
+  editor = { Dragan A. Savic  and  Godfrey A. Walters  and  Roger King  and  Soon Thiam-Khu },
+  year = 2005,
+  booktitle = {Proceedings of the Eighth International Conference on
+                  Computing and Control for the Water Industry (CCWI 2005)},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  T. Devi Prasad  and  Ben Paechter },
+  title = {Optimal Pump Scheduling: Representation and Multiple
+                  Objectives},
+  pages = {117--122}
+}
+
+ +
+@incollection{LopStu09ea,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Pierre Collet and Nicolas Monmarch{\'e} and Pierrick
+                  Legrand and Marc Schoenauer and Evelyne Lutton},
+  shorteditor = {Pierre Collet and others},
+  volume = 5975,
+  series = {Lecture Notes in Computer Science},
+  year = 2010,
+  booktitle = {Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Analysis of Algorithmic Components for
+                  Multiobjective Ant Colony Optimization: {A} Case
+                  Study on the Biobjective {TSP}},
+  pages = {134--145},
+  doi = {10.1007/978-3-642-14156-0_12}
+}
+
+ +
+@incollection{LopStu2010:ants,
+  volume = 6234,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Gianni A. {Di Caro}  and Doursat, R. and Engelbrecht, A. P. and Floreano,
+                  D. and Gambardella, L. M. and Gro\ss, R. and Sahin,
+                  E. and  Thomas St{\"u}tzle  and Sayama, H.},
+  editor = { Marco Dorigo  and others },
+  year = 2010,
+  booktitle = {Swarm Intelligence, 7th International Conference, ANTS 2010},
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Configuration of Multi-Objective {ACO}
+                  Algorithms},
+  pages = {95--106},
+  doi = {10.1007/978-3-642-15461-4_9},
+  abstract = {In the last few years a significant number of ant
+                  colony optimization (ACO) algorithms have been
+                  proposed for tackling multi-objective optimization
+                  problems. In this paper, we propose a software
+                  framework that allows to instantiate the most
+                  prominent multi-objective ACO (MOACO)
+                  algorithms. More importantly, the flexibility of
+                  this MOACO framework allows the application of
+                  automatic algorithm configuration techniques. The
+                  experimental results presented in this paper show
+                  that such an automatic configuration of MOACO
+                  algorithms is highly desirable, given that our
+                  automatically configured algorithms clearly
+                  outperform the best performing MOACO algorithms that
+                  have been proposed in the literature. As far as we
+                  are aware, this paper is also the first to apply
+                  automatic algorithm configuration techniques to
+                  multi-objective stochastic local search algorithms.}
+}
+
+ +
+@incollection{LopStu2010:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {The impact of design choices of multi-objective ant
+                  colony optimization algorithms on performance: An
+                  experimental study on the biobjective {TSP}},
+  doi = {10.1145/1830483.1830494},
+  pages = {71--78},
+  abstract = {Over the last few years, there have been a number of
+                  proposals of ant colony optimization (ACO)
+                  algorithms for tackling multiobjective combinatorial
+                  optimization problems. These proposals adapt ACO
+                  concepts in various ways, for example, some use
+                  multiple pheromone matrices and multiple heuristic
+                  matrices and others use multiple ant colonies.\\ In
+                  this article, we carefully examine several of the
+                  most prominent of these proposals. In particular, we
+                  identify commonalities among the approaches by
+                  recasting the original formulation of the algorithms
+                  in different terms. For example, several proposals
+                  described in terms of multiple colonies can be cast
+                  equivalently using a single ant colony, where ants
+                  use different weights for aggregating the pheromone
+                  and/or the heuristic information. We study
+                  algorithmic choices for the various proposals and we
+                  identify previously undetected trade-offs in their
+                  performance.}
+}
+
+ +
+@misc{LopStu2010:gecco-supp,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {The impact of design choices of multi-objective ant
+                  colony optimization algorithms on performance: An
+                  experimental study on the biobjective {TSP}},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2010-003/}},
+  year = 2010,
+  note = {Supplementary material of \cite{LopStu2010:gecco}}
+}
+
+ +
+@misc{LopStu2011moaco-supp,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {The Automatic Design of Multi-Objective Ant Colony
+                  Optimization Algorithms: {Supplementary} material},
+  url = {http://iridia.ulb.ac.be/supp/IridiaSupp2011-007/Iridia-2011-007.pdf},
+  year = 2011
+}
+
+ +
+@misc{LopStu2012si-supp,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An experimental analysis of design choices of
+                  multi-objective ant colony optimization algorithms:
+                  Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2012-006/}},
+  year = 2012
+}
+
+ +
+@incollection{LopStuDor2017aco,
+  isbn = {978-3-319-07125-1},
+  publisher = {Springer International Publishing},
+  year = 2018,
+  booktitle = {Handbook of Heuristics},
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {Ant Colony Optimization: A Component-Wise Overview},
+  doi = {10.1007/978-3-319-07124-4_21},
+  pages = {371--407},
+  supplement = {http://iridia.ulb.ac.be/aco-tsp-qap/},
+  annote = {Proposed ACOTSPQAP software}
+}
+
+ +
+@phdthesis{LopezDiploma,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Multi-objective Ant Colony Optimization},
+  school = {Intellectics Group, Computer Science Department, Technische
+                  Universit{\"a}t Darmstadt, Germany},
+  year = 2004,
+  type = {Diploma thesis}
+}
+
+ +
+@phdthesis{LopezIbanezPhD,
+  author = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Operational Optimisation of Water Distribution
+                  Networks},
+  school = {School of Engineering and the Built Environment},
+  year = 2009,
+  address = {Edinburgh Napier University, UK},
+  url = {https://researchrepository.napier.ac.uk/id/eprint/3044}
+}
+
+ +
+@incollection{LosSchSeb2012ppsn,
+  volume = 7491,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {I}},
+  author = {Ilya Loshchilov  and  Marc Schoenauer  and  Mich{\`e}le Sebag },
+  title = {Alternative Restart Strategies for {CMA-ES}},
+  pages = {296--305},
+  doi = {10.1007/978-3-642-32937-1_30}
+}
+
+ +
+@incollection{LotMie2008vis,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  author = {Lotov, A. V.  and  Kaisa Miettinen },
+  title = {Visualizing the {Pareto} Frontier},
+  pages = {213--243}
+}
+
+ +
+@incollection{LouMarStu01,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Helena R. {Louren{\c c}o}  and  Olivier Martin  and  Thomas St{\"u}tzle },
+  title = {Iterated Local Search},
+  pages = {321--353},
+  doi = {10.1007/0-306-48056-5_11}
+}
+
+ +
+@incollection{LouMarStu2010:mh,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Helena R. {Louren{\c c}o}  and  Olivier Martin  and  Thomas St{\"u}tzle },
+  title = {Iterated Local Search: Framework and Applications},
+  chapter = 9,
+  pages = {363--397},
+  doi = {10.1007/978-1-4419-1665-5_12}
+}
+
+ +
+@incollection{LouMarStu2019hb,
+  publisher = {Springer},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 272,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2019,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Helena R. {Louren{\c c}o}  and  Olivier Martin  and  Thomas St{\"u}tzle },
+  title = {Iterated Local Search: Framework and Applications},
+  chapter = 5,
+  pages = {129--168},
+  doi = {10.1007/978-3-319-91086-4_5}
+}
+
+ +
+@inproceedings{LunLee2017shap,
+  year = 2016,
+  editor = {Isabelle Guyon and Ulrike von Luxburg and Samy Bengio and
+                  Hanna M. Wallach and Rob Fergus and S. V. N. Vishwanathan and
+                  Roman Garnett},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 30)},
+  author = {Lundberg, Scott M. and Lee, Su{-}In},
+  title = {A unified approach to interpreting model predictions},
+  keywords = {SHAP, interpretable AI},
+  pages = {4765--4774},
+  epub = {https://proceedings.neurips.cc/paper/2017/hash/8a20a8621978632d76c43dfd28b67767-Abstract.html}
+}
+
+ +
+@incollection{LuoBos2012elitist,
+  volume = 7492,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {II}},
+  author = {Hoang N. Luong and  Peter A. N. Bosman },
+  title = {Elitist Archiving for Multi-Objective Evolutionary
+                  Algorithms: To Adapt or Not to Adapt},
+  pages = {72--81},
+  doi = {10.1007/978-3-642-32964-7_8}
+}
+
+ +
+@incollection{LusTeg2010,
+  publisher = {Springer},
+  editor = { Carlos A. {Coello Coello}  and  Dhaenens, Clarisse  and  Laetitia Jourdan },
+  volume = 272,
+  year = 2010,
+  series = {Studies in Computational Intelligence},
+  booktitle = {Advances in Multi-Objective Nature Inspired Computing},
+  author = { Thibaut Lust  and  Jacques Teghem },
+  title = {The multiobjective traveling salesman problem:
+                  A survey and a new approach},
+  pages = {119--141}
+}
+
+ +
+@inproceedings{LwiQuZhe2013moss,
+  author = {Khin Lwin and Rong Qu and Jianhua Zheng},
+  title = {Multi-objective Scatter Search with External Archive for
+                  Portfolio Optimization},
+  booktitle = {Proceedings of the 5th International Joint Conference on
+                  Computational Intelligence - ECTA (IJCCI 2013)},
+  year = 2013,
+  pages = {111--119},
+  publisher = {SciTePress},
+  doi = {10.5220/0004552501110119},
+  annote = {Crowding archive}
+}
+
+ +
+@phdthesis{Lygoe2010phd,
+  title = {Complexity reduction in high-dimensional multi-objective
+                  optimisation},
+  author = {Lygoe, Robert John},
+  year = 2010,
+  school = {University of Sheffield Sheffield, UK}
+}
+
+ +
+@misc{MATILDA,
+  author = { Kate Smith{-}Miles  and  Mario A. Mu{\~{n}}oz  and Neelofar},
+  title = {Melbourne Algorithm Test Instance Library with Data Analytics
+                  ({MATILDA})},
+  year = 2020,
+  url = {https://matilda.unimelb.edu.au/}
+}
+
+ +
+@inproceedings{Mackle95,
+  author = { Gunther M{\"a}ckle  and  Dragan A. Savic  and  Godfrey A. Walters },
+  title = {Application of Genetic Algorithms to Pump Scheduling for
+                  Water Supply},
+  booktitle = {Genetic Algorithms in Engineering Systems: Innovations and
+                  Applications, GALESIA'95},
+  pages = {400--405},
+  year = 1995,
+  month = sep,
+  volume = 414,
+  address = {Sheffield, UK},
+  publisher = {{IEE} Conference Publication},
+  abstract = { A simple Genetic Algorithm has been applied to the
+                  scheduling of multiple pumping units in a water supply system
+                  with the objective of minimising the overall cost of the
+                  pumping operation, taking advantage of storage capacity in
+                  the system and the availability of off peak electricity
+                  tariffs. A simple example shows that the method is easy to
+                  apply and has produced encouraging preliminary results},
+  doi = {10.1049/cp:19951082}
+}
+
+ +
+@inproceedings{Mad2002pdea,
+  year = 2002,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2002 World Congress on Computational Intelligence (WCCI 2002)},
+  key = {WCCI},
+  editor = { David B. Fogel  and others},
+  title = {Multiobjective optimization using a {Pareto} differential
+                  evolution approach},
+  author = {Madavan, Nateri K.},
+  pages = {1145--1150}
+}
+
+ +
+@inproceedings{Maie04:ann_ga,
+  author = { D. R. Broad  and  Graeme C. Dandy  and  Holger R. Maier },
+  title = {A Metamodeling Approach to Water Distribution System
+                  Optimization},
+  booktitle = {6th Annual Symposium on Water Distribution Systems
+                  Analysis},
+  year = 2004,
+  month = jun,
+  organization = {ASCE}
+}
+
+ +
+@incollection{MalSel2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 7298,
+  editor = {Nicolas Beldiceanu and Narendra Jussien and Eric Pinson},
+  isbn = {978-3-642-29827-1},
+  year = 2012,
+  booktitle = {Integration of AI and OR Techniques in Contraint Programming
+                  for Combinatorial Optimization Problems},
+  author = { Yuri Malitsky  and  Meinolf Sellmann },
+  title = {Instance-specific algorithm configuration as a method for
+                  non-model-based portfolio generation},
+  pages = {244--259},
+  doi = {10.1007/978-3-642-29828-8_16}
+}
+
+ +
+@incollection{MalitskyEtAl2013tuning,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2013,
+  editor = {Gomes, C. and  Meinolf Sellmann },
+  volume = 7874,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Integration of AI and OR Techniques in Constraint Programming
+                  for Combinatorial Optimization Problems, CPAIOR 2010},
+  author = { Yuri Malitsky  and Mehta, Deepak and   O'Sullivan, Barry  and Simonis, Helmut},
+  title = {Tuning parameters of large neighborhood search for the
+                  machine reassignment problem},
+  pages = {176--192},
+  doi = {10.1007/978-3-642-38171-3_12}
+}
+
+ +
+@incollection{ManBosJel04:ants2004,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Vittorio Maniezzo  and  M. Boschetti and M. Jelasity},
+  title = {An Ant Approach to Membership Overlay Design},
+  pages = {37--48}
+}
+
+ +
+@incollection{ManMil2002:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { Vittorio Maniezzo  and M. Milandri},
+  title = {An Ant-Based Framework for Very Strongly Constrained
+                  Problems},
+  pages = {222--227}
+}
+
+ +
+@inproceedings{ManSurBauFinBetMcC2014,
+  title = {The {Stanford} {CoreNLP} Natural Language Processing Toolkit},
+  author = {Manning, Christopher D. and Surdeanu, Mihai and Bauer, John
+                  and Finkel, Jenny Rose and Bethard, Steven J. and McClosky,
+                  David},
+  booktitle = {Association for Computational Linguistics (ACL) System
+                  Demonstrations},
+  pages = {55--60},
+  year = 2014,
+  annote = {\url{http://www.aclweb.org/anthology/P/P14/P14-5010}}
+}
+
+ +
+@inproceedings{MarBouHer05:ANNforWDN:ccwi,
+  month = sep,
+  address = {University of Exeter, UK},
+  volume = 1,
+  editor = { Dragan A. Savic  and  Godfrey A. Walters  and  Roger King  and  Soon Thiam-Khu },
+  year = 2005,
+  booktitle = {Proceedings of the Eighth International Conference on
+                  Computing and Control for the Water Industry (CCWI 2005)},
+  author = { F. Mart{\'i}nez  and  V. Bou  and  V. Hern{\'a}ndez  and  F. Alvarruiz  and  J. M. Alonso },
+  title = {{ANN} Architectures for Simulating Water
+                  Distribution Networks},
+  pages = {251--256}
+}
+
+ +
+@incollection{MarDhaJouLieVer2011:evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 6622,
+  year = 2011,
+  editor = { Peter Merz  and  Jin-Kao Hao },
+  booktitle = {Proceedings of EvoCOP 2011 -- 11th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = { Marie-El{\'e}onore Marmion  and  Dhaenens, Clarisse  and  Laetitia Jourdan  and  Arnaud Liefooghe  and  Verel, S{\'e}bastien },
+  title = {{NILS:} {A} Neutrality-Based Iterated Local Search and Its Application to Flowshop Scheduling},
+  pages = {191--202}
+}
+
+ +
+@incollection{MarMasLop2013hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7919,
+  series = {Lecture Notes in Computer Science},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Paola Festa and  Andrea Roli  and  M. Sampels },
+  isbn = {978-3-642-38515-5},
+  year = 2013,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Marie-El{\'e}onore Marmion  and  Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Stochastic Local Search
+                  Algorithms},
+  pages = {144--158},
+  doi = {10.1007/978-3-642-38516-2_12}
+}
+
+ +
+@incollection{MarMoo1994hoeffding,
+  address = { San Francisco, CA},
+  publisher = {Morgan Kaufmann Publishers},
+  year = 1994,
+  editor = {J. D. Cowan and G. Tesauro and J. Alspector},
+  volume = 6,
+  booktitle = {Advances in Neural Information Processing Systems},
+  title = {{Hoeffding} races: {Accelerating} model selection search for
+                  classification and function approximation},
+  author = {Maron, Oded and Moore, Andrew W.},
+  pages = {59--66},
+  alias = {marmoo1994hoeffding}
+}
+
+ +
+@incollection{MarMor99,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  year = 1999,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 1999},
+  shorteditor = {Wolfgang Banzhaf and others},
+  editor = {Wolfgang Banzhaf and Jason M. Daida and A. E. Eiben
+                  and Max H. Garzon and Vasant Honavar and Mark
+                  J. Jakiela and Robert E. Smith},
+  author = { C. E. Mariano  and  E. Morales },
+  title = {{MOAQ}: An {Ant}-{Q} Algorithm for Multiple
+                  Objective Optimization Problems},
+  pages = {894--901}
+}
+
+ +
+@inproceedings{MarSte1998,
+  publisher = {Max-Planck-Institut f{\"{u}}r Informatik, Saarbr\"ucken,
+                  Germany},
+  editor = {Kurt Mehlhorn},
+  booktitle = {Algorithm Engineering, 2nd International Workshop, {WAE}'92},
+  year = 1998,
+  author = { Elena Marchiori  and Adri G. Steenbeek},
+  title = {An Iterated Heuristic Algorithm for the Set Covering Problem},
+  pages = {155--166}
+}
+
+ +
+@incollection{MarSte2000,
+  year = 2000,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 1803,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Real-World Applications of Evolutionary Computing, EvoWorkshops 2000},
+  editor = {Stefano Cagnoni and others},
+  fulleditor = {Stefano Cagnoni and Riccardo Poli and Yun Li and George
+                  D. Smith and David Corne and Martin J. Oates and Emma Hart
+                  and Pier Luca Lanzi and Egbert J. W. Boers and Ben Paechter
+                  and Terence C. Fogarty},
+  author = { Elena Marchiori  and Adri G. Steenbeek},
+  title = {An Evolutionary Algorithm for Large Scale Set Covering
+                  Problems with Application to Airline Crew Scheduling},
+  pages = {367--381}
+}
+
+ +
+@book{MarStu98:cp,
+  author = {K. Marriott and P. Stuckey},
+  title = {Programming With Constraints},
+  publisher = {MIT Press, Cambridge, MA},
+  year = 1998
+}
+
+ +
+@book{MarTot1990knapsack,
+  author = { Silvano Martello  and  Paolo Toth },
+  title = {Knapsack Problems: Algorithms and Computer Implementations},
+  publisher = {John Wiley \& Sons},
+  address = { Chichester, UK},
+  year = 1990,
+  keywords = {bin packing}
+}
+
+ +
+@mastersthesis{Maron1994hoeffding,
+  title = {{Hoeffding} Races: {Model} selection for {MRI}
+                  classification},
+  author = {Maron, Oded},
+  year = 1994,
+  school = {Massachusetts Institute of Technology}
+}
+
+ +
+@incollection{MasBirStu2013lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7997,
+  editor = { Panos M. Pardalos  and G. Nicosia},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  booktitle = {Learning and Intelligent Optimization, 7th
+                  International Conference, LION 7},
+  author = { Franco Mascia  and  Mauro Birattari  and  Thomas St{\"u}tzle },
+  title = {Tuning Algorithms for Tackling Large Instances: An
+                  Experimental Protocol},
+  pages = {410--422},
+  doi = {10.1007/978-3-642-44973-4_44}
+}
+
+ +
+@incollection{MasDevHen2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 7298,
+  editor = {Nicolas Beldiceanu and Narendra Jussien and Eric Pinson},
+  isbn = {978-3-642-29827-1},
+  year = 2012,
+  booktitle = {Integration of AI and OR Techniques in Contraint Programming
+                  for Combinatorial Optimization Problems},
+  author = { Florence Massen  and  Yves Deville  and van Hentenryck, Pascal },
+  title = {Pheromone-Based Heuristic Column Generation for Vehicle
+                  Routing Problems with Black Box Feasibility},
+  pages = {260--274},
+  doi = {10.1007/978-3-642-29828-8_17}
+}
+
+ +
+@misc{MasLopDub2014-supp,
+  title = {Grammar-based generation of stochastic local search
+                  heuristics through automatic algorithm configuration
+                  tools: Supplementary material},
+  author = { Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle },
+  year = 2013,
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2013-009/}}
+}
+
+ +
+@incollection{MasLopDub2014hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 8457,
+  isbn = {978-3-319-07643-0},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and  Stefan Vo{\ss} },
+  year = 2014,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Marie-El{\'e}onore Marmion  and  Thomas St{\"u}tzle },
+  title = {Algorithm Comparison by Automatically Configurable Stochastic
+                  Local Search Frameworks: A Case Study Using Flow-Shop
+                  Scheduling Problems},
+  pages = {30--44},
+  doi = {10.1007/978-3-319-07644-7_3}
+}
+
+ +
+@incollection{MasLopDubStu2013lion,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7997,
+  editor = { Panos M. Pardalos  and G. Nicosia},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  booktitle = {Learning and Intelligent Optimization, 7th
+                  International Conference, LION 7},
+  author = { Franco Mascia  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle },
+  title = {From Grammars to Parameters: Automatic Iterated Greedy Design
+                  for the Permutation Flow-shop Problem with Weighted
+                  Tardiness},
+  pages = {321--334},
+  doi = {10.1007/978-3-642-44973-4_36}
+}
+
+ +
+@incollection{MasLopStu2013hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7919,
+  series = {Lecture Notes in Computer Science},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Paola Festa and  Andrea Roli  and  M. Sampels },
+  isbn = {978-3-642-38515-5},
+  year = 2013,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Florence Massen  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle  and  Yves Deville },
+  title = {Experimental Analysis of Pheromone-Based Heuristic
+                  Column Generation Using irace},
+  pages = {92--106},
+  doi = {10.1007/978-3-642-38516-2_8}
+}
+
+ +
+@inproceedings{MasNesDor2020,
+  year = 2020,
+  editor = {Vlastelica, Marin and Song, Jialin and Ferber, Aaron and Amos, Brandon and
+                  Martius, Georg and Dilkina, Bistra and Yue, Yisong},
+  booktitle = {Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020,
+                  {LMCA} 2020, Vancouver, Canada, December 12, 2020},
+  author = {Massobrio, Renzo and Nesmachnow, Sergio and  Bernab{\'e} Dorronsoro },
+  title = {Virtual {Savant}: learning for optimization},
+  pages = {1--5}
+}
+
+ +
+@inproceedings{MatTakMiyShi2020digital,
+  author = {Matsubara, Satoshi and Takatsu, Motomu and Miyazawa,
+                  Toshiyuki and Shibasaki, Takayuki and Watanabe, Yasuhiro and
+                  Takemoto, Kazuya and Tamura, Hirotaka},
+  title = {Digital Annealer for High-Speed Solving of Combinatorial
+                  optimization Problems and Its Applications},
+  booktitle = {2020 25th Asia and South Pacific Design Automation Conference
+                  (ASP-DAC)},
+  year = 2020,
+  pages = {667--672},
+  organization = {IEEE},
+  doi = {10.1109/ASP-DAC47756.2020.9045100},
+  abstract = {A Digital Annealer (DA) is a dedicated architecture for
+                  high-speed solving of combinatorial optimization problems
+                  mapped to an Ising model. With fully coupled bit connectivity
+                  and high coupling resolution as a major feature, it can be
+                  used to express a wide variety of combinatorial optimization
+                  problems. The DA uses Markov Chain Monte Carlo as a basic
+                  search mechanism, accelerated by the hardware implementation
+                  of multiple speed-enhancement techniques such as parallel
+                  search, escape from a local solution, and replica
+                  exchange. It is currently being offered as a cloud service
+                  using a second-generation chip operating on a scale of 8,192
+                  bits. This paper presents an overview of the DA, its
+                  performance against benchmarks, and application examples.}
+}
+
+ +
+@inproceedings{MauLopStu2010:cec,
+  year = 2010,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2010 Congress on Evolutionary Computation (CEC 2010)},
+  editor = { Ishibuchi, Hisao  and others},
+  key = {IEEE CEC},
+  author = { Michael Maur  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Pre-scheduled and adaptive parameter variation in
+                  {\MaxMinAntSystem}},
+  pages = {3823--3830},
+  doi = {10.1109/CEC.2010.5586332}
+}
+
+ +
+@incollection{MazChuMietLop2019emo,
+  isbn = {978-3-030-12597-4},
+  year = 2019,
+  address = { Cham, Switzerland},
+  publisher = {Springer International Publishing},
+  volume = 11411,
+  series = {Lecture Notes in Computer Science},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2019},
+  editor = { Kalyanmoy Deb  and Erik D. Goodman and  Carlos A. {Coello Coello}  and Kathrin
+                  Klamroth and  Kaisa Miettinen  and Sanaz Mostaghim and Patrick
+                  Reed},
+  author = { Atanu Mazumdar  and  Tinkle Chugh  and  Kaisa Miettinen  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {On Dealing with Uncertainties from Kriging Models in Offline
+                  Data-Driven Evolutionary Multiobjective Optimization},
+  pages = {463--474},
+  doi = {10.1007/978-3-030-12598-1_37}
+}
+
+ +
+@incollection{McCormick03,
+  publisher = {CRC Press},
+  year = 2003,
+  booktitle = {Advances in Water Supply Management},
+  editor = { C. Maksimovi{\'c}  and  David Butler  and  Fayyaz Ali Memon },
+  author = { G. McCormick  and  R. S. Powell },
+  title = {A progressive mixed integer-programming method for
+                  pump scheduling},
+  pages = {307--313}
+}
+
+ +
+@book{McG2012guide,
+  author = { Catherine C. McGeoch },
+  title = {A Guide to Experimental Algorithmics},
+  year = 2012,
+  publisher = {Cambridge University Press}
+}
+
+ +
+@techreport{McGFar2020dwave,
+  author = { Catherine C. McGeoch  and Farr{\'e}, Pau},
+  title = {The {D-Wave} Advantage System: An Overview},
+  institution = {D-Wave Systems Inc., Burnaby, BC, Canada},
+  year = 2020,
+  url = {https://www.dwavesys.com/media/s3qbjp3s/14-1049a-a_the_d-wave_advantage_system_an_overview.pdf}
+}
+
+ +
+@inproceedings{MedGolGol2014bracis,
+  author = {Medeiros, Hudson Geovane de and  Goldbarg, Elizabeth Ferreira Gouv{\^e}a  and  Goldbarg, Marco Cesar },
+  booktitle = {2014 Brazilian Conference on Intelligent Systems},
+  title = {Analyzing Limited Size Archivers of Multi-objective
+                  Optimizers},
+  year = 2014,
+  pages = {85--90},
+  doi = {10.1109/BRACIS.2014.26},
+  keywords = {archiving}
+}
+
+ +
+@misc{MeiCla2014prep,
+  title = {A versatile heuristic approach for generalized hub location
+                  problems},
+  author = {J. Fabian Meier and Uwe Clausen},
+  howpublished = {Preprint, Provided upon personal request},
+  year = 2014,
+  url = {https://optimization-online.org/wp-content/uploads/2014/12/4690.pdf},
+  keywords = {irace}
+}
+
+ +
+@incollection{MelPerCos2009:ea,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Pierre Collet and Nicolas Monmarch{\'e} and Pierrick
+                  Legrand and Marc Schoenauer and Evelyne Lutton},
+  shorteditor = {Pierre Collet and others},
+  volume = 5975,
+  series = {Lecture Notes in Computer Science},
+  year = 2010,
+  booktitle = {Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009},
+  title = {{MC-ANT}: a Multi-colony Ant Algorithm},
+  author = {Melo, L. and Pereira, F. and Costa, E.}
+}
+
+ +
+@incollection{MenCoe2015gd,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = {Menchaca-Mendez, Adriana and  Carlos A. {Coello Coello} },
+  title = {{GD-MOEA}: A New Multi-Objective Evolutionary Algorithm Based
+                  on the Generational Distance Indicator},
+  pages = {156--170}
+}
+
+ +
+@inproceedings{MenCoe2015gde,
+  year = 2015,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2015 Congress on Evolutionary Computation (CEC 2015)},
+  key = {IEEE CEC},
+  author = {Menchaca-Mendez, Adriana and  Carlos A. {Coello Coello} },
+  title = {{GDE-MOEA}: A New {MOEA} based on the generational distance
+                  indicator and $\epsilon$-dominance},
+  pages = {947--955}
+}
+
+ +
+@inproceedings{MenKleFeuSprHut2016autoNN,
+  author = {Mendoza, Hector and Klein, Aaron and  Matthias Feurer  and Springenberg, Jost Tobias and  Frank Hutter },
+  title = {Towards automatically-tuned neural networks},
+  booktitle = {Workshop on Automatic Machine Learning},
+  year = 2016,
+  pages = {58--65}
+}
+
+ +
+@incollection{MerBisTraPreuWeiRud11:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { Olaf Mersmann  and  Bernd Bischl  and  Heike Trautmann  and  Mike Preuss  and  Claus Weihs  and  G{\"u}nther Rudolph },
+  title = {Exploratory Landscape Analysis},
+  pages = {829--836},
+  keywords = {continuous optimization, landscape analysis, instance features}
+}
+
+ +
+@incollection{MerHuh2008:ppsn,
+  year = 2008,
+  volume = 5199,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { G{\"u}nther Rudolph  and others},
+  aeditor = { G{\"u}nther Rudolph  and Thomas Jansen and Simon Lucas and
+                  Carlo Poloni and Nicola Beume},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {X}},
+  author = { Peter Merz  and Jutta Huhse},
+  title = {An Iterated Local Search Approach for Finding Provably Good Solutions
+  for Very Large {TSP} Instances},
+  pages = {929--939}
+}
+
+ +
+@incollection{MerMid01,
+  author = { D. Merkle  and  Martin Middendorf },
+  title = {Prospects for Dynamic Algorithm Control: Lessons
+                  from the Phase Structure of Ant Scheduling
+                  Algorithms},
+  booktitle = {Proceedings of the 2001 Genetic and Evolutionary
+                  Computation Conference -- Workshop Program. Workshop
+                  ``The Next Ten Years of Scheduling Research''},
+  pages = {121--126},
+  year = 2001,
+  editor = {R. B. Heckendorn},
+  address = {San Francisco, CA},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@incollection{MerMidSch00:gecco,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = { Darrell Whitley  and others},
+  fulleditor = { Darrell Whitley  and  David E. Goldberg  and E. Cantu-Paz and L. Spector and
+                  I. Parmee and   Hans-Georg Beyer },
+  year = 2000,
+  booktitle = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2000},
+  author = { D. Merkle  and  Martin Middendorf  and  Hartmut Schmeck },
+  title = {Ant Colony Optimization for Resource-Constrained Project Scheduling},
+  pages = {893--900}
+}
+
+ +
+@inproceedings{MerTraNau2010cec,
+  year = 2010,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2010 Congress on Evolutionary Computation (CEC 2010)},
+  editor = { Ishibuchi, Hisao  and others},
+  key = {IEEE CEC},
+  author = { Olaf Mersmann  and  Heike Trautmann  and  Boris Naujoks  and  Claus Weihs },
+  title = {Benchmarking Evolutionary Multiobjective Optimization Algorithms},
+  pages = {1--8},
+  annote = {TR: \url{http://hdl.handle.net/2003/26671}}
+}
+
+ +
+@inproceedings{Mey2004:gecco,
+  author = { Bernd Meyer },
+  title = {Convergence control in {ACO}},
+  note = {Late-breaking paper available on CD},
+  year = 2004,
+  booktitle = {Genetic and Evolutionary Computation Conference
+                  (GECCO)},
+  address = {Seattle, WA}
+}
+
+ +
+@incollection{MeyErn04:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Bernd Meyer  and  Andreas T. Ernst },
+  title = {Integrating {ACO} and Constraint Propagation},
+  pages = {166--177},
+  year = 2004
+}
+
+ +
+@incollection{MezReyCoe2008,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Uday K. Chakraborty},
+  year = 2008,
+  booktitle = {Advances in differential evolution},
+  author = { Efr{\'e}n Mezura-Montes  and Reyes-Sierra, M. and  Carlos A. {Coello Coello} },
+  title = {Multi-objective optimization using differential
+                  evolution: a survey of the state-of-the-art},
+  pages = {173--196},
+  doi = {10.1007/978-3-540-68830-3_7}
+}
+
+ +
+@incollection{MezVelCoe2006,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2006,
+  editor = {M. Cattolico and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2006},
+  title = {A comparative study of differential evolution variants for
+                  global optimization},
+  author = { Efr{\'e}n Mezura-Montes  and Vel{\'a}zquez-Reyes, Jes{\'u}s and  Carlos A. {Coello Coello} },
+  pages = {485--492},
+  doi = {10.1145/1143997.1144086}
+}
+
+ +
+@book{MicFog04:howtosolveit,
+  author = { Zbigniew Michalewicz  and  David B. Fogel },
+  title = {How to Solve It: Modern Heuristics},
+  publisher = {Springer},
+  year = 2004,
+  edition = {2nd}
+}
+
+ +
+@incollection{MicHen2004,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 2004,
+  editor = { Shlomo Zilberstein  and J. Koehler and S. Koenig},
+  booktitle = {Proceedings of the Fourteenth International Conference on
+                  Automated Planning and Scheduling (ICAPS 2004)},
+  author = { Laurent D. Michel  and van Hentenryck, Pascal },
+  title = {Iterative Relaxations for Iterative Flattening in Cumulative
+                  Scheduling},
+  pages = {200--208}
+}
+
+ +
+@incollection{MicMid98,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Agoston E. Eiben  and  Thomas B{\"a}ck  and  Marc Schoenauer  and  Hans-Paul Schwefel },
+  volume = 1498,
+  series = {Lecture Notes in Computer Science},
+  year = 1998,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {V}},
+  author = {R. Michel and M. Middendorf},
+  title = {An Island Model based {Ant} {System} with Lookahead
+                  for the Shortest Supersequence Problem},
+  pages = {692--701}
+}
+
+ +
+@book{Michalewicz1996,
+  author = { Zbigniew Michalewicz },
+  title = {Genetic Algorithms + Data Structures = Evolution Programs},
+  edition = {3rd},
+  year = 1996,
+  address = { Berlin, Germany},
+  publisher = {Springer}
+}
+
+ +
+@book{Mie99,
+  author = { Kaisa Miettinen },
+  title = {Nonlinear Multiobjective Optimization},
+  publisher = {Kluwer Academic Publishers},
+  address = { Boston, MA},
+  year = 1999,
+  abstract = {Nonlinear Multiobjective Optimization provides an extensive,
+                  up-to-date, self-contained and consistent survey and review
+                  of the literature and of the state of the art on nonlinear
+                  (deterministic) multiobjective optimization, its methods, its
+                  theory and its background. This book is intended for both
+                  researchers and students in the areas of (applied)
+                  mathematics, engineering, economics, operations research and
+                  management science; it is meant for both professionals and
+                  practitioners in many different fields of application. The
+                  intention is to provide a consistent summary that may help in
+                  selecting an appropriate method for the problem to be
+                  solved. The extensive bibliography will be of value to
+                  researchers.},
+  numpages = 298
+}
+
+ +
+@incollection{MieRuiWie08interactive,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  author = { Kaisa Miettinen  and  Francisco Ruiz  and  Andrzej P. Wierzbicki },
+  title = {Introduction to Multiobjective Optimization:
+                  Interactive Approaches},
+  doi = {10.1007/978-3-540-88908-3_2},
+  abstract = {We give an overview of interactive methods developed
+                  for solving nonlinear multiobjective optimization
+                  problems. In interactive methods, a decision maker
+                  plays an important part and the idea is to support
+                  her/him in the search for the most preferred
+                  solution. In interactive methods, steps of an
+                  iterative solution algorithm are repeated and the
+                  decision maker progressively provides preference
+                  information so that the most preferred solution can
+                  be found. We identify three types of specifying
+                  preference information in interactive methods and
+                  give some examples of methods representing each
+                  type. The types are methods based on trade-off
+                  information, reference points and classification of
+                  objective functions.}
+}
+
+ +
+@inproceedings{MirSilPru2014esann,
+  epub = {https://www.esann.org/proceedings/2014},
+  year = 2014,
+  booktitle = {European Symposium on Artificial Neural Networks, ESSAN},
+  key = {ESANN},
+  author = {P{\'e}ricles Miranda and Ricardo M. Silva and Ricardo
+                  B. Prud{\^e}ncio},
+  title = {Fine-Tuning of Support Vector Machine Parameters Using Racing
+                  Algorithms},
+  pages = {325--330},
+  keywords = {irace}
+}
+
+ +
+@inproceedings{MirSilPru2015:esann,
+  epub = {https://www.esann.org/proceedings/2015},
+  year = 2015,
+  booktitle = {European Symposium on Artificial Neural Networks, ESSAN},
+  key = {ESANN},
+  author = {P{\'e}ricles Miranda and Ricardo M. Silva and Ricardo
+                  B. Prud{\^e}ncio},
+  title = {{I/S-Race}: An Iterative Multi-objective Racing Algorithm for
+                  the {SVM} Parameter Selection Problem},
+  pages = {573--578}
+}
+
+ +
+@incollection{Misevicius2003:gecco,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2723,
+  editor = {E. Cant\'u-Paz and others},
+  year = 2003,
+  booktitle = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2003, Part I},
+  author = { Misevi{\v{c}}ius, Alfonsas },
+  title = {Ruin and Recreate Principle Based Approach for the Quadratic
+                  Assignment Problem},
+  pages = {598--609}
+}
+
+ +
+@inproceedings{MitRomSan1985,
+  title = {Convergence and Finite-Time Behavior of Simulated Annealing},
+  author = { Debasis Mitra  and  Fabio Romeo  and  Alberto Sangiovanni-Vincentelli },
+  booktitle = {Decision and Control, 1985 24th IEEE Conference on},
+  pages = {761--767},
+  year = 1985,
+  organization = {IEEE}
+}
+
+ +
+@inproceedings{MitSelLev1992,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  editor = {William R. Swartout},
+  year = 1992,
+  booktitle = {Proceedings of  the 10th National Conference on Artificial Intelligence},
+  author = { David G. Mitchell  and  Bart Selman  and  Hector J. Levesque },
+  title = {Hard and Easy Distributions of {SAT} Problems},
+  pages = {459--465}
+}
+
+ +
+@inproceedings{MniSzeAud2008,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  booktitle = {Proceedings of  the 25th International Conference on Machine Learning, {ICML} 2008},
+  editor = {William W. Cohen and Andrew McCallum and Sam T. Roweis},
+  title = {Empirical {Bernstein} stopping},
+  author = {Mnih, Volodymyr and Szepesv{\'a}ri, Csaba and Audibert,
+                  Jean-Yves},
+  pages = {672--679}
+}
+
+ +
+@incollection{Mockus1975,
+  author = { Jonas Mo{\v{c}}kus },
+  booktitle = {Optimization Techniques IFIP Technical Conference
+                  Novosibirsk, July 1--7, 1974},
+  title = {On Bayesian Methods for Seeking the Extremum},
+  year = 1975,
+  editor = {Marchuk, G. I.},
+  pages = {400--404},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 27,
+  doi = {10.1007/3-540-07165-2_55},
+  annote = {Proposed Bayesian optimization}
+}
+
+ +
+@book{Mockus1989,
+  author = { Jonas Mo{\v{c}}kus },
+  title = {Bayesian Approach to Global Optimization: Theory and
+                  Applications},
+  publisher = {Kluwer Academic Publishers},
+  year = 1989
+}
+
+ +
+@misc{ModCMA,
+  author = {Sander van Rijn},
+  title = {Modular {CMA-ES} framework
+                  from~\cite{RijWanLeeBac2016ssci}, v0.3.0},
+  year = 2018,
+  howpublished = {\url{https://github.com/sjvrijn/ModEA}},
+  note = {Available also as {\softwarepackage{pypi}} package at
+                  \url{https://pypi.org/project/ModEA/0.3.0/}}
+}
+
+ +
+@incollection{MogAteYalAmo11:lorenz,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  title = {{Lorenz} versus {Pareto} Dominance in a Single Machine
+                  Scheduling Problem with Rejection},
+  author = {Moghaddam, Atefeh and Yalaoui, Farouk and Amodeo,
+                  Lionel},
+  pages = {520--534}
+}
+
+ +
+@inproceedings{MolTeo2012safe,
+  year = 2012,
+  publisher = {Omnipress},
+  booktitle = {Proceedings of  the 29th International Conference on Machine Learning, {ICML} 2012},
+  editor = {John Langford and Joelle Pineau},
+  author = {Moldovan, Teodor Mihai and Abbeel, Pieter},
+  title = {Safe Exploration in {Markov} Decision Processes},
+  pages = {1451--1458},
+  numpages = 8,
+  epub = {http://icml.cc/2012/papers/838.pdf}
+}
+
+ +
+@book{Molchanov2005theory,
+  author = {Molchanov, Ilya},
+  title = {Theory of Random Sets},
+  publisher = {Springer},
+  year = 2005,
+  keywords = {Vorob'ev expectation}
+}
+
+ +
+@incollection{MonDevHen2009,
+  author = {Jean-No\"el Monette and  Yves Deville  and van Hentenryck, Pascal },
+  title = {Aeon: Synthesizing Scheduling Algorithms from High-Level Models},
+  booktitle = {Operations Research and Cyber-Infrastructure},
+  publisher = {Springer},
+  year = 2009,
+  editor = {John W. Chinneck and Bjarni Kristjansson and Matthew J. Saltzman},
+  volume = 47,
+  series = {Operations Research/Computer Science Interfaces},
+  pages = {43--59},
+  address = { New York, NY}
+}
+
+ +
+@incollection{MonPerRiffCoe12dummy,
+  volume = 7491,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {I}},
+  author = { Elizabeth Montero  and   P{\'e}rez C{\'a}ceres, Leslie and  Mar{\'i}a-Cristina Riff  and  Carlos A. {Coello Coello} },
+  title = {Are State-of-the-Art Fine-Tuning Algorithms Able to
+                  Detect a Dummy Parameter?},
+  pages = {306--315},
+  doi = {10.1007/978-3-642-32937-1_31}
+}
+
+ +
+@inproceedings{MonRif2013gecco,
+  year = 2013,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  key = {IEEE CEC},
+  title = {A new algorithm for reducing metaheuristic design effort},
+  author = { Mar{\'i}a-Cristina Riff  and  Elizabeth Montero },
+  pages = {3283--3290},
+  doi = {10.1109/CEC.2013.6557972}
+}
+
+ +
+@incollection{MonRif2014ppsn,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  author = { Elizabeth Montero  and  Mar{\'i}a-Cristina Riff },
+  title = {Towards a Method for Automatic Algorithm Configuration: A Design Evaluation Using Tuners},
+  pages = {90--99},
+  doi = {10.1007/978-3-319-10762-2_9}
+}
+
+ +
+@incollection{MonRifNev2010,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  author = { Elizabeth Montero  and  Mar{\'i}a-Cristina Riff  and Neveu, Bertrand},
+  title = {An Evaluation of Off-line Calibration Techniques for
+                  Evolutionary Algorithms},
+  pages = {299--300}
+}
+
+ +
+@incollection{MonYos11,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011},
+  address = {Berlin\slash Heidelberg},
+  series = {Lecture Notes in Computer Science},
+  volume = 6576,
+  year = 2011,
+  publisher = {Springer},
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  title = {A Framework for Locating Logistic Facilities with
+                  Multi-Criteria Decision Analysis},
+  author = {Montibeller, Gilberto and Yoshizaki, Hugo},
+  pages = {505--519}
+}
+
+ +
+@phdthesis{Montes2011PhD,
+  author = { Marco A. {Montes de Oca} },
+  title = {Incremental Social Learning in Swarm Intelligence Systems},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2011,
+  annote = {Supervised by Marco Dorigo}
+}
+
+ +
+@phdthesis{Montgomery2005phd,
+  author = { James Montgomery },
+  title = {Solution Biases and Pheromone Representation
+                  Selection in Ant Colony Optimisation},
+  school = {School of Information Technology, Bond University},
+  year = 2005,
+  address = {Australia},
+  alias = {montgomery05:phd}
+}
+
+ +
+@book{Montgomery2012,
+  author = {Douglas C. Montgomery},
+  title = {Design and Analysis of Experiments},
+  publisher = {John Wiley \& Sons},
+  year = 2012,
+  edition = {8th},
+  address = { New York, NY}
+}
+
+ +
+@inproceedings{MooLee1994efficient,
+  address = { San Francisco, CA},
+  publisher = {Morgan Kaufmann Publishers},
+  year = 1994,
+  booktitle = {Proceedings of  the 11th International Conference on Machine Learning, {ICML} 1994},
+  editor = {William W. Cohen and Haym Hirsh},
+  author = {Moore, Andrew W. and Lee, Mary S.},
+  title = {Efficient Algorithms for Minimizing Cross Validation Error},
+  pages = {190--198}
+}
+
+ +
+@incollection{MorKat2011:evo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 6622,
+  year = 2011,
+  editor = { Peter Merz  and  Jin-Kao Hao },
+  booktitle = {Proceedings of EvoCOP 2011 -- 11th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = { A. Moraglio  and  A. Kattan },
+  title = {Geometric Generalisation of Surrogate Model Based
+                  Optimization to Combinatorial Spaces},
+  pages = {142--154}
+}
+
+ +
+@incollection{MorKat2011:gecco,
+  year = 2011,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2011},
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  author = { A. Moraglio  and Yong{-}Hyuk Kim and Yourim Yoon},
+  title = {Geometric Surrogate-based Optimisation for Permutation-based
+                  Problems},
+  pages = {133--134}
+}
+
+ +
+@inproceedings{Morris1993,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  editor = {Richard Fikes and Wendy G. Lehnert},
+  year = 1993,
+  booktitle = {Proceedings of  the 11th National Conference on Artificial Intelligence},
+  author = {Pail Morris},
+  title = {The Breakout Method for Escaping from Local Minima},
+  pages = {40--45}
+}
+
+ +
+@incollection{MosJoh03:bioinf,
+  booktitle = {Artificial Neural Networks and Genetic Algorithms},
+  key = {ICANNGA},
+  year = 2003,
+  publisher = {Springer Verlag},
+  editor = {D. W. Pearson and N. C. Steele and R. F. Albrecht},
+  author = {J. D. Moss and C. G. Johnson},
+  title = {An ant colony algorithm for multiple sequence alignment in
+                  bioinformatics},
+  pages = {182--186}
+}
+
+ +
+@techreport{Moscato1989,
+  title = {On Evolution, Search, Optimization, Genetic Algorithms
+                 and Martial Arts: Towards Memetic Algorithms},
+  author = { Pablo Moscato },
+  institution = {Caltech},
+  type = {Caltech Concurrent Computation Program, C3P Report},
+  number = 826,
+  year = 1989
+}
+
+ +
+@incollection{Moscato99,
+  alias = {CorDorGlo99},
+  address = {London, UK},
+  year = 1999,
+  publisher = {McGraw Hill},
+  editor = { David Corne  and  Marco Dorigo  and  Fred Glover },
+  booktitle = {New Ideas in Optimization},
+  author = { Pablo Moscato },
+  title = {Memetic algorithms: a short introduction},
+  pages = {219--234}
+}
+
+ +
+@phdthesis{Mou2003:PhD,
+  author = { Vincent Mousseau },
+  title = {Elicitation des pr{\'e}f{\'e}rences pour l'aide multicrit{\`e}re {\`a} la d{\'e}cision},
+  school = {Universit{\'e} Paris-Dauphine, Paris, France},
+  year = 2003
+}
+
+ +
+@misc{Moy94:opsf,
+  author = {Moy, J.},
+  title = {{RFC 1583}: Open shortest path first protocol},
+  online = {yes},
+  year = 1994
+}
+
+ +
+@misc{MuDubHooStu2017:scaling-supp,
+  author = {Zongxu Mu and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {On the Empirical Scaling of Running Time for Finding Optimal
+                  Solutions to the {TSP}: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2017-010/}},
+  year = 2017
+}
+
+ +
+@incollection{MuHooStu2016:lion,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 10079,
+  editor = {Paola Festa and  Meinolf Sellmann  and  Joaquin Vanschoren },
+  series = {Lecture Notes in Computer Science},
+  year = 2016,
+  booktitle = {Learning and Intelligent Optimization, 10th International
+                  Conference, LION 10},
+  author = {Zongxu Mu and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {The Impact of Automated Algorithm Configuration on the
+                  Scaling Behaviour of State-of-the-Art Inexact {TSP} Solvers},
+  pages = {157--172},
+  doi = {10.1007/978-3-319-50349-3_11}
+}
+
+ +
+@misc{MudKomLopKaz2019gecco-supp,
+  author = { Mudita Sharma  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Dimitar Kazakov },
+  title = {Deep Reinforcement Learning Based Parameter Control in
+                  Differential Evolution: Supplementary material},
+  howpublished = {\url{https://github.com/mudita11/DE-DDQN}},
+  doi = {10.5281/zenodo.2628228},
+  year = 2019
+}
+
+ +
+@inproceedings{MuhZim2000,
+  month = jul,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2000,
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  key = {IEEE CEC},
+  title = {Size of neighborhood more important than temperature for stochastic local search},
+  author = { H. M{\"u}hlenbein  and Zimmermann, J{\"o}rg},
+  pages = {1017--1024}
+}
+
+ +
+@incollection{Muhlenbein1991,
+  year = 1991,
+  publisher = {Morgan Kaufmann Publishers, San Mateo, CA},
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  editor = {G. Rawlins},
+  author = { H. M{\"u}hlenbein },
+  title = {Evolution in Time and Space---The Parallel Genetic Algorithm},
+  pages = {316--337}
+}
+
+ +
+@book{Mulenthaler2015fair,
+  title = {Fairness in Academic Course Timetabling},
+  author = {Moritz M{\"u}hlenthaler},
+  year = 2015,
+  publisher = {Springer},
+  doi = {10.1007/978-3-319-12799-6},
+  keywords = {irace}
+}
+
+ +
+@inproceedings{Murphy94,
+  author = { L. J. Murphy  and  Graeme C. Dandy  and  Angus R. Simpson },
+  title = {Optimum Design and Operation of Pumped Water
+                  Distribution Systems},
+  booktitle = {1994 International Conference on Hydraulics and
+                  Civil Engineering, Hidraulic working with the
+                  Environment},
+  pages = {149--155},
+  year = 1994,
+  address = {Brisbane, Australia},
+  month = feb,
+  organization = {The Institution of Engineers},
+  note = {}
+}
+
+ +
+@incollection{NagKob1997,
+  year = 1997,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  booktitle = {ICGA},
+  editor = { Thomas B{\"a}ck },
+  author = {Yuichi Nagata and Shigenobu Kobayashi},
+  title = {Edge Assembly Crossover: A High-power Genetic
+                  Algorithm for the Traveling Salesman Problem},
+  pages = {450--457}
+}
+
+ +
+@inproceedings{NagKob1999analysis,
+  year = 1999,
+  editor = {Koji Ito and Fumio Harashima and Kazuo Tanie},
+  publisher = {IEEE Press},
+  booktitle = {{IEEE} {SMC}'99 Conference Proceedings, 1999 {IEEE}
+                  International Conference on Systems, Man, and Cybernetics},
+  key = {SMC},
+  author = {Yuichi Nagata and Shigenobu Kobayashi},
+  title = {An analysis of edge assembly crossover for the traveling
+                  salesman problem},
+  doi = {10.1109/icsmc.1999.823285},
+  pages = {628--633}
+}
+
+ +
+@inproceedings{NagSucDum2012,
+  author = {R. Nagy and M. Suciu and D. Dumitrescu},
+  booktitle = {Symbolic and Numeric Algorithms for Scientific Computing
+                  (SYNASC), 2012 14th International Symposium on},
+  title = {Exploring {Lorenz} Dominance},
+  year = 2012,
+  pages = {254--259}
+}
+
+ +
+@inproceedings{NaiHin2010rectified,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the 27th International Conference on Machine Learning, {ICML} 2010},
+  editor = {Johannes F{\"u}rnkranz and Thorsten Joachims},
+  author = {Nair, Vinod and Hinton, Geoffrey E.},
+  title = {Rectified linear units improve restricted boltzmann machines},
+  pages = {807--814}
+}
+
+ +
+@incollection{NalWagNeu2014ppsn,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  author = { Samadhi Nallaperuma  and  Markus Wagner  and  Frank Neumann },
+  title = {Parameter Prediction Based on Features of Evolved Instances
+                  for Ant Colony Optimization and the Traveling Salesperson
+                  Problem},
+  pages = {100--109},
+  doi = {10.1007/978-3-319-10762-2_10}
+}
+
+ +
+@incollection{NanEib2006gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2006,
+  editor = {M. Cattolico and others},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2006},
+  author = {V. Nannen and  Agoston E. Eiben },
+  title = {A Method for Parameter Calibration and Relevance Estimation
+                  in Evolutionary Algorithms},
+  pages = {183--190},
+  keywords = {REVAC},
+  doi = {10.1145/1143997.1144029}
+}
+
+ +
+@inproceedings{NanEib2007ijcai,
+  publisher = {AAAI Press, Menlo Park, CA},
+  editor = {Manuela M. Veloso},
+  year = 2007,
+  booktitle = {Proceedings of  the 20th International Joint Conference on Artificial Intelligence (IJCAI-07)},
+  author = {V. Nannen and  Agoston E. Eiben },
+  title = {Relevance Estimation and Value Calibration of Evolutionary
+                  Algorithm Parameters},
+  pages = {975--980},
+  keywords = {REVAC}
+}
+
+ +
+@incollection{NasMesUgo2012,
+  volume = 7492,
+  year = 2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  editor = { Carlos A. {Coello Coello}  and others},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {II}},
+  author = {Youssef S. G. Nashed and Pablo Mesejo and Roberto Ugolotti and  J{\'e}r{\'e}mie Dubois-Lacoste  and Stefano Cagnoni},
+  title = {A Comparative Study of Three {GPU}-Based Metaheuristics},
+  pages = {398--407},
+  doi = {10.1007/978-3-642-32964-7_40}
+}
+
+ +
+@inproceedings{NebDurCoe2013cec,
+  year = 2013,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  key = {IEEE CEC},
+  author = { Nebro, Antonio J.  and  Durillo, Juan J.  and  Carlos A. {Coello Coello} },
+  title = {Analysis of leader selection strategies in a multi-objective
+                  {Particle} {Swarm} {Optimizer}},
+  pages = {3153--3160},
+  doi = {10.1109/CEC.2013.6557955}
+}
+
+ +
+@inproceedings{NebDurGar2009smpso,
+  author = { Nebro, Antonio J.  and  Durillo, Juan J.  and  Jos{\'e} Garc{\'i}a-Nieto  and  Carlos A. {Coello Coello}  and F. Luna and  Alba, Enrique },
+  doi = {10.1109/MCDM.2009.4938830},
+  isbn = 9781424427642,
+  booktitle = {2009 IEEE Symposium on Computational Intelligence in
+                  Multi-Criteria Decision-Making (MCDM)},
+  pages = {66--73},
+  title = {{SMPSO}: A new {PSO}-based metaheuristic for multi-objective
+                  optimization},
+  year = 2009
+}
+
+ +
+@incollection{NebDurVer2015jmetal,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2015},
+  author = { Nebro, Antonio J.  and  Durillo, Juan J.  and Vergne, Matthieu},
+  title = {Redesigning the {jMetal} Multi-Objective Optimization
+                  Framework},
+  keywords = {JMetal, Multi-objective metaheuristics, Open source,
+                  Optimization framework},
+  pages = {1093--1100},
+  numpages = 8,
+  doi = {10.1145/2739482.2768462}
+}
+
+ +
+@incollection{NebLopBarGar2019gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = { Nebro, Antonio J.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Barba-Gonz{\'a}lez, Crist{\'o}bal  and  Jos{\'e} Garc{\'i}a-Nieto },
+  title = {Automatic Configuration of {NSGA-II} with {jMetal} and irace},
+  pages = {1374--1381},
+  doi = {10.1145/3319619.3326832}
+}
+
+ +
+@book{NemWol88,
+  author = {G. L. Nemhauser and L. A. Wolsey},
+  title = {Integer and Combinatorial Optimization},
+  publisher = {John Wiley \& Sons},
+  address = { New York, NY},
+  year = 1988
+}
+
+ +
+@incollection{NikJac2010,
+  address = { New York, NY},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  title = {Simulated Annealing},
+  author = { Alexander G. Nikolaev  and  Sheldon H. Jacobson },
+  pages = {1--39},
+  publisher = {Springer}
+}
+
+ +
+@inproceedings{NikMarJan2009,
+  title = {Instance-based selection of policies for {SAT} solvers},
+  author = {Nikoli{\'c}, Mladen and Mari{\'c}, Filip and Jani{\v{c}}i{\'c}, Predrag},
+  booktitle = {International Conference on Theory and Applications of Satisfiability Testing},
+  pages = {326--340},
+  year = 2009,
+  organization = {Springer}
+}
+
+ +
+@incollection{NisOyaAkiAguTan2014:space,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8426,
+  editor = { Panos M. Pardalos  and  Mauricio G. C. Resende  and Chrysafis Vogiatzis and Jose
+                  L. Walteros},
+  series = {Lecture Notes in Computer Science},
+  year = 2014,
+  booktitle = {Learning and Intelligent Optimization, 8th International Conference, LION 8},
+  author = {Y. Nishio and A. Oyama and Y. Akimoto and  Aguirre, Hern\'{a}n E.  and  Tanaka, Kiyoshi },
+  title = {Many-objective Optimization of Trajectory Design for
+                  {DESTINY} Mission}
+}
+
+ +
+@book{NixAgu2012,
+  title = {Feature extraction \& image processing for computer vision},
+  author = {Nixon, Mark S. and Aguado, Alberto S.},
+  year = 2012,
+  publisher = {Academic Press},
+  address = { New York, NY}
+}
+
+ +
+@incollection{NobVerWan2021gecco,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  author = {Jacob de Nobel and  Diederick Vermetten  and Hao Wang and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {Tuning as a means of assessing the benefits of new ideas in
+                  interplay with existing algorithmic modules},
+  pages = {1375--1384},
+  doi = {10.1145/3449726.3463167},
+  supplement = {https://doi.org/10.5281/zenodo.4524959}
+}
+
+ +
+@misc{NobVerWan2021gecco-supp,
+  author = {Jacob de Nobel and  Diederick Vermetten  and Hao Wang and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {Data and Code from Tuning as a means of assessing the
+                  benefits of new ideas in interplay with existing algorithmic
+                  modules},
+  month = feb,
+  year = 2021,
+  publisher = {Zenodo},
+  version = {1.0},
+  doi = {10.5281/zenodo.4524959}
+}
+
+ +
+@book{NocWri2006,
+  author = {Jorge Nocedal and Stephen J. Wright},
+  title = {Numerical Optimization},
+  publisher = {Springer},
+  year = 2006,
+  series = {Springer Series in Operations Research and Financial Engineering},
+  edition = {2nd}
+}
+
+ +
+@incollection{NouDriGhe2016,
+  volume = 464,
+  series = {Advances in Intelligent Systems and Computing},
+  publisher = {Springer International Publishing},
+  editor = {Silhavy, Radek and Senkerik, Roman and Oplatkova, Zuzana Kominkova and Silhavy, Petr and Prokopova, Zdenka},
+  year = 2016,
+  booktitle = {Artificial Intelligence Perspectives in Intelligent Systems},
+  title = {A Classification Schema for the Job Shop Scheduling Problem with Transportation Resources: State-of-the-Art Review},
+  author = {Nouri, Houssem Eddine and Driss, Olfa Belkahla and Gh{\'e}dira, Khaled},
+  pages = {1--11}
+}
+
+ +
+@incollection{Now2014hv,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  title = {Empirical Performance of the Approximation of the Least
+                  Hypervolume Contributor},
+  author = {Nowak, Krzysztof and M{\"a}rtens, Marcus and Izzo, Dario},
+  pages = {662--671}
+}
+
+ +
+@inproceedings{OMaHebHolNugOSu2008cphydra,
+  title = {Using case-based reasoning in an algorithm portfolio for
+                  constraint solving},
+  author = {O'Mahony, Eoin and Hebrard, Emmanuel and Holland, Alan and
+                  Nugent, Conor and   O'Sullivan, Barry },
+  booktitle = {Irish Conference on Artificial Intelligence and Cognitive
+                  Science},
+  pages = {210--216},
+  editor = {Bridge and others},
+  year = 2008
+}
+
+ +
+@incollection{ObaSas2003visualization,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  title = {Visualization and data mining of {Pareto} solutions using
+                  self-organizing map},
+  author = {Obayashi, Shigeru and Sasaki, Daisuke},
+  pages = {796--809},
+  keywords = {objective reduction}
+}
+
+ +
+@incollection{OchHydCur2012evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 7245,
+  year = 2012,
+  editor = { Jin-Kao Hao  and  Martin Middendorf },
+  booktitle = {Proceedings of EvoCOP 2012 -- 12th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  title = {Hyflex: A benchmark framework for cross-domain heuristic
+                  search},
+  author = { Gabriela Ochoa  and Matthew Hyde and Tim Curtois and Jose
+                  A. Vazquez-Rodriguez and James Walker and  Michel Gendreau  and  Graham Kendall  and  Barry McCollum  and  Andrew J. Parkes  and Sanja Petrovic and  Edmund K. Burke },
+  pages = {136--147}
+}
+
+ +
+@incollection{OchTomVerDar2008,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  editor = {Conor Ryan},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  author = { Gabriela Ochoa  and Tomassini, Marco and  Verel, S{\'e}bastien  and Darabos, Christian},
+  title = {A Study of {NK} Landscapes' Basins and Local Optima Networks},
+  pages = {555--562}
+}
+
+ +
+@inproceedings{OddRasCesSmi2011,
+  publisher = {IJCAI/AAAI Press, Menlo Park, CA},
+  editor = {Toby Walsh},
+  year = 2011,
+  booktitle = {Proceedings of  the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11)},
+  author = { Angelo Oddi  and Riccardo Rasconi and  Amadeo Cesta  and  Stephen F. Smith },
+  title = {Iterative Flattening Search for the Flexible Job Shop Scheduling Problem},
+  pages = {1991--1996}
+}
+
+ +
+@incollection{OjaPodMie2016ppsn,
+  isbn = {978-3-319-45822-9},
+  year = 2016,
+  volume = 9921,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Julia Handl  and  Emma Hart  and  Lewis, P. R.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Gabriela Ochoa  and  Ben Paechter },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIV}},
+  author = {Vesa Ojalehto and Dmitry Podkopaev and  Kaisa Miettinen },
+  title = {Towards Automatic Testing of Reference Point Based
+                  Interactive Methods},
+  pages = {483--492},
+  doi = {10.1007/978-3-319-45823-6_45},
+  keywords = {artificial DMs},
+  abstract = {In this research, we proposed to build an automated framework
+                  for testing interactive multiobjective optimization methods,
+                  without utilizing a value function to represent the DM's
+                  preferences. This was achieved by replacing the human DM with
+                  an artificial DM constructed from two distinct parts: the
+                  steady part and the current context. With the steady part the
+                  artificial DM tries to maintain the search towards its
+                  preferences, while at the same time the current context
+                  allows changing the direction as well as ending the solution
+                  process prematurely, mimicking actions of a human DM.}
+}
+
+ +
+@inproceedings{OliHusRolDorStu2017cec,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = {Sabrina M. Oliveira and Mohamed Saifullah Hussin and  Andrea Roli  and  Marco Dorigo  and  Thomas St{\"u}tzle },
+  title = {Analysis of the Population-based Ant Colony Optimization Algorithm for the {TSP} and the {QAP}},
+  pages = {1734--1741}
+}
+
+ +
+@incollection{OlsBarUrb2016gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2016},
+  author = {Olson, Randal S. and Bartley, Nathan and Urbanowicz, Ryan
+                  J. and Moore, Jason H.},
+  title = {Evaluation of a Tree-based Pipeline Optimization Tool for
+                  Automating Data Science},
+  pages = {485--492},
+  numpages = 8,
+  doi = {10.1145/2908812.2908918},
+  acmid = 2908918,
+  keywords = {TPOT}
+}
+
+ +
+@incollection{OlsUrbAnd2016evobio,
+  volume = 9597,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  year = 2016,
+  editor = {Squillero, Giovanni and Burelli, Paolo},
+  author = {Olson, Randal S. and Urbanowicz, Ryan J. and Andrews, Peter
+                  C. and Lavender, Nicole A. and Kidd, La Creis and Moore,
+                  Jason H.},
+  title = {Automating Biomedical Data Science Through Tree-Based
+                  Pipeline Optimization},
+  pages = {123--137},
+  doi = {10.1007/978-3-319-31204-0_9},
+  keywords = {TPOT}
+}
+
+ +
+@inproceedings{OstSal04,
+  author = { Avi Ostfeld  and Elad Salomons},
+  title = {Optimal Scheduling of Pumping and Chlorine
+                  Injections under Unsteady Hydraulics},
+  booktitle = {Critical Transitions In Water And Environmental
+                  Resources Management},
+  pages = {1--9},
+  year = 2004,
+  editor = {Gerald Sehlke and Donald F. Hayes and David
+                  K. Stevens},
+  month = jul,
+  abstract = {This paper describes the methodology and application
+                  of a genetic algorithm (GA) scheme, tailor-made to
+                  EPANET for simultaneously optimizing the scheduling
+                  of existing pumping and booster disinfection units,
+                  as well as the design of new disinfection booster
+                  chlorination stations, under unsteady
+                  hydraulics. The objective is to minimize the total
+                  cost of operating the pumping units and the chlorine
+                  booster operation and design for a selected
+                  operational time horizon, while delivering the
+                  consumers required water quantities, at acceptable
+                  pressures and chlorine residual concentrations. The
+                  decision variables, for each of the time steps that
+                  encompass the total operational time horizon,
+                  include: the scheduling of the pumping units,
+                  settings of the water distribution system control
+                  valves, and the mass injection rates at each of the
+                  booster chlorination stations. The constraints are
+                  domain heads and chlorine concentrations at the
+                  consumer nodes, maximum injection rates at the
+                  chlorine injection stations, maximum allowable
+                  amounts of water withdraws at the sources, and
+                  returning at the end of the operational time horizon
+                  to a prescribed total volume in the tanks. The model
+                  is demonstrated through an example application.}
+}
+
+ +
+@incollection{OztTsoVin2005:mcda,
+  editor = { Jos{\'e} Rui Figueira  and  Salvatore Greco  and  Matthias Ehrgott },
+  year = 2005,
+  publisher = {Springer},
+  booktitle = {Multiple Criteria Decision Analysis, State of the
+                  Art Surveys},
+  author = { Meltem {\"O}zt{\"u}rk  and  Alexis Tsouki{\`a}s  and  Philippe Vincke },
+  title = {Preference Modelling},
+  chapter = 2,
+  pages = {27--72}
+}
+
+ +
+@techreport{PagStu2018:pfsp,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Stochastic Local Search Algorithms
+                 for Permutation Flowshop Problems},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  number = {TR/IRIDIA/2018-005},
+  year = 2018,
+  month = apr,
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2018-005.pdf}
+}
+
+ +
+@misc{PagStu2018:pfsp-supp,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Automatic Design of Hybrid Stochastic Local Search Algorithms
+  for Permutation Flowshop Problems: Supplementary Material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2018-002/}},
+  year = 2018
+}
+
+ +
+@misc{PagStu2019:pfsp-supp,
+  author = { Federico Pagnozzi  and  Thomas St{\"u}tzle },
+  title = {Automatic design of hybrid stochastic local search algorithms for permutation flowshop problems with additional constraints},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2018-002/}},
+  year = 2019
+}
+
+ +
+@phdthesis{Pagnozzi2019PhD,
+  author = { Federico Pagnozzi },
+  title = {Automatic Design of Hybrid Stochastic Local Search Algorithms},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2019,
+  annote = {Supervised by Thomas St\"utzle}
+}
+
+ +
+@inproceedings{PanIshSha2020,
+  title = {Algorithm configurations of {MOEA/D} with an unbounded
+                  external archive},
+  author = {Pang, Lie Meng and  Ishibuchi, Hisao  and Shang, Ke},
+  booktitle = {2020 IEEE International Conference on Systems, Man, and
+                  Cybernetics (SMC)},
+  year = 2020,
+  organization = {IEEE},
+  pages = {1087--1094}
+}
+
+ +
+@book{PapSte1982:ph,
+  author = { Christos H. Papadimitriou  and Steiglitz, K.},
+  title = {Combinatorial Optimization -- Algorithms and
+                  Complexity},
+  publisher = {Prentice Hall, Englewood Cliffs, NJ},
+  year = 1982
+}
+
+ +
+@inproceedings{PapYan2000focs,
+  publisher = {IEEE Computer Society Press},
+  year = 2000,
+  booktitle = {41st Annual Symposium on Foundations of Computer Science},
+  editor = {Avrim Blum},
+  author = { Christos H. Papadimitriou  and  Mihalis Yannakakis },
+  title = {On the Approximability of Trade-offs and Optimal
+                  Access of Web Sources},
+  pages = {86--92},
+  doi = {10.1109/SFCS.2000.892068}
+}
+
+ +
+@mastersthesis{Paq2001:Msc,
+  author = { Lu{\'i}s Paquete },
+  title = {Algoritmos Evolutivos Multiobjectivo para
+                  Afecta\c{c}\~ao de Recursos e sua Aplica\c{c}\~ao
+                  \`a Gera\c{c}\~ao de Hor\'arios em Universidades
+                  ({Multiobjective} Evolutionary Algorithms for
+                  Resource Allocation and their Application to
+                  University Timetabling)},
+  school = {University of Algarve},
+  year = 2001,
+  note = {In Portuguese},
+  abstract = {The aim of this study is the application of
+                  multiobjective evolutionary algorithms to resource
+                  allocation problems, such as university examination
+                  timetabling and course timetabling
+                  problems. Usually, these problems are characterized
+                  by multiple conflicting objectives. A multiobjective
+                  formalization of these problems is presented, based
+                  on goals and priorities. Various aspects of
+                  evolutionary algorithms are proposed and studied for
+                  these problems, particulary, selection methods and
+                  types and parameters of mutation operator. The
+                  choice of both representation and operators is made
+                  so as not to favour excessively certain objectives
+                  with respect to others at the level of the
+                  exploration mechanism. A comparative study of
+                  performance is presented for the proposed algorithms
+                  by means of statistical inference, based on real
+                  problems of the University of Algarve. The notion of
+                  attainment functions is used as a base for the
+                  assessment of performance of multiobjective
+                  evolutionary algorithms. Finally, the evolution of
+                  the solution cost during the runs is analysed by
+                  means of attainment functions, as well.}
+}
+
+ +
+@phdthesis{Paq2005:PhD,
+  author = { Lu{\'i}s Paquete },
+  title = {Stochastic Local Search Algorithms for
+                  Multiobjective Combinatorial Optimization: Methods
+                  and Analysis},
+  school = {FB Informatik, TU Darmstadt, Germany},
+  year = 2005
+}
+
+ +
+@incollection{PaqChiStu2004mmo,
+  year = 2004,
+  address = {Berlin\slash Heidelberg},
+  publisher = {Springer},
+  volume = 535,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  editor = { Xavier Gandibleux  and Marc Sevaux and  Kenneth S{\"o}rensen  and  V. {T'Kindt} },
+  booktitle = {Metaheuristics for Multiobjective Optimisation},
+  author = { Lu{\'i}s Paquete  and  Marco Chiarandini  and  Thomas St{\"u}tzle },
+  title = {{Pareto} Local Optimum Sets in the Biobjective Traveling
+                  Salesman Problem: An Experimental Study},
+  pages = {177--199},
+  alias = {lpaquete:10},
+  abstract = {In this article, we study {Pareto} local optimum sets for the
+                  biobjective Traveling Salesman Problem applying
+                  straightforward extensions of local search algorithms for the
+                  single objective case. The performance of the local search
+                  algorithms is illustrated by experimental results obtained
+                  for well known benchmark instances and comparisons to methods
+                  from literature.  In fact, a 3-opt local search is able to
+                  compete with the best performing metaheuristics in terms of
+                  solution quality. Finally, we also present an empirical study
+                  of the features of the solutions found by 3-opt on a set of
+                  randomly generated instances. The results indicate the
+                  existence of several clusters of near-optimal solutions that
+                  are separated by only a few edges.},
+  keywords = {Pareto local search, PLS},
+  doi = {10.1007/978-3-642-17144-4_7}
+}
+
+ +
+@techreport{PaqFonLop06-CSI-klee,
+  author = { Lu{\'i}s Paquete  and  Carlos M. Fonseca  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {An optimal algorithm for a special case of {Klee}'s
+                  measure problem in three dimensions},
+  institution = {CSI, Universidade do Algarve},
+  year = 2006,
+  number = {CSI-RT-I-01/2006},
+  abstract = {The measure of the region dominated by (the maxima
+                  of) a set of $n$ points in the positive $d$-orthant
+                  has been proposed as an indicator of performance in
+                  multiobjective optimization, known as the
+                  hypervolume indicator, and the problem of computing
+                  it efficiently is attracting increasing
+                  attention. In this report, this problem is
+                  formulated as a special case of Klee's measure
+                  problem in $d$ dimensions, which immediately
+                  establishes $O(n^{d/2}\log n)$ as a, possibly
+                  conservative, upper bound on the required
+                  computation time. Then, an $O(n log n)$ algorithm
+                  for the 3-dimensional version of this special case
+                  is constructed, based on an existing dimension-sweep
+                  algorithm for the related maxima problem. Finally,
+                  $O(n log n)$ is shown to remain a lower bound on the
+                  time required by the hypervolume indicator for
+                  $d>1$, which attests the optimality of the algorithm
+                  proposed.},
+  note = {Superseded by paper in IEEE Transactions on Evolutionary Computation~\cite{BeuFonLopPaqVah09:tec}},
+  annote = {Proof of Theorem 3.1 is incorrect}
+}
+
+ +
+@incollection{PaqStu08:lnems,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Clusters of non-dominated solutions in multiobjective
+                  combinatorial optimization: An experimental analysis},
+  booktitle = {Multiobjective Programming and Goal Programming: Theoretical
+                  Results and Practical Applications},
+  pages = {69--77},
+  year = 2009,
+  volume = 618,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  editor = {V. Barichard and M. Ehrgott and  Xavier Gandibleux  and V. T'Kindt},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  doi = {10.1007/978-3-540-85646-7}
+}
+
+ +
+@incollection{PaqStu2002:evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2279,
+  editor = {S. Cagnoni and others},
+  aeditor = {S. Cagnoni and J. Gottlieb and E. Hart and
+                 M. Middendorf and G. R. Raidl},
+  year = 2002,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2002},
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {An Experimental Investigation of Iterated Local Search for Coloring Graphs},
+  pages = {122--131}
+}
+
+ +
+@incollection{PaqStu2003tpls,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {A Two-Phase Local Search for the Biobjective
+                  Traveling Salesman Problem},
+  pages = {479--493},
+  alias = {lpaquete:8}
+}
+
+ +
+@incollection{PaqStu2018handbook,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle },
+  title = {Stochastic Local Search Algorithms for Multiobjective
+                  Combinatorial Optimization: {A} Review},
+  booktitle = {Handbook of Approximation Algorithms and Metaheuristics},
+  pages = {411--425},
+  publisher = {Chapman \& Hall/CRC},
+  address = { Boca Raton, FL},
+  doi = {10.1201/9781351236423-24},
+  year = 2018,
+  editor = {Teofilo F. Gonzalez},
+  alias = {paquete18}
+}
+
+ +
+@techreport{PaqStuLop-IRIDIA-2005-029,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {On the design and analysis of {SLS} algorithms for
+                  multiobjective combinatorial optimization problems},
+  institution = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2005,
+  number = {TR/IRIDIA/2005-029},
+  abstract = {Effective Stochastic Local Search (SLS) algorithms
+                  can be seen as being composed of several algorithmic
+                  components, each of which plays some specific role
+                  with respect to overall performance. In this
+                  article, we explore the application of experimental
+                  design techniques to analyze the effect of different
+                  choices for these algorithmic components on SLS
+                  algorithms applied to Multiobjective Combinatorial
+                  Optimization Problems that are solved in terms of
+                  {Pareto} optimality. This analysis is done using the
+                  example application of SLS algorithms to the
+                  biobjective Quadratic Assignment Problem and we show
+                  also that the same choices for algorithmic
+                  components can lead to different behavior in
+                  dependence of various instance features, such as the
+                  structure of input data and the correlation between
+                  objectives.},
+  url = {http://iridia.ulb.ac.be/IridiaTrSeries/link/IridiaTr2005-029.pdf}
+}
+
+ +
+@inproceedings{PaqStuLop05mic,
+  address = {Vienna, Austria},
+  year = 2005,
+  booktitle = {6th Metaheuristics International Conference (MIC 2005)},
+  editor = { Karl F. Doerner  and  Michel Gendreau  and Peter Greistorfer and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Marc Reimann },
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Towards the Empirical Analysis of {SLS} Algorithms
+                  for Multiobjective Combinatorial Optimization
+                  Problems through Experimental Design},
+  pages = {739--746},
+  abstract = { Stochastic Local Search (SLS) algorithms for
+                  Multiobjective Combinatorial Optimization Problems
+                  (MCOPs) typically involve the selection and
+                  parameterization of many algorithm components whose
+                  role with respect to their overall performance and
+                  relation to certain instance features is often not
+                  clear. In this abstract, we use a modular approach
+                  for the design of SLS algorithms for MCOPs defined
+                  in terms of {Pareto} optimality and we present an
+                  extensive analysis of SLS algorithms through
+                  experimental design techniques, where each algorithm
+                  component is considered a factor. The experimental
+                  analysis is based on a sound experimental
+                  methodology for analyzing the output of algorithms
+                  for MCOPs. We show that different choices for
+                  algorithm components can lead to different behavior
+                  in dependence of various instance features.}
+}
+
+ +
+@incollection{PaqStuLop07metaheuristics,
+  author = { Lu{\'i}s Paquete  and  Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Using experimental design to analyze stochastic local search
+                  algorithms for multiobjective problems},
+  booktitle = {Metaheuristics: Progress in Complex Systems Optimization},
+  pages = {325--344},
+  year = 2007,
+  doi = {10.1007/978-0-387-71921-4_17},
+  volume = 39,
+  series = {Operations Research / Computer Science Interfaces},
+  publisher = {Springer},
+  address = { New York, NY},
+  annote = {Post-Conference Proceedings of the 6th Metaheuristics
+                  International Conference (MIC 2005)},
+  editor = {Karl F. Doerner and Michel Gendreau and Peter Greistorfer and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Marc Reimann },
+  abstract = {Stochastic Local Search (SLS) algorithms can be seen as being
+                  composed of several algorithmic components, each playing some
+                  specific role with respect to overall performance. This
+                  article explores the application of experimental design
+                  techniques to analyze the effect of components of SLS
+                  algorithms for Multiobjective Combinatorial Optimization
+                  problems, in particular for the Biobjective Quadratic
+                  Assignment Problem. The analysis shows that there exists a
+                  strong dependence between the choices for these components
+                  and various instance features, such as the structure of the
+                  input data and the correlation between the objectives.}
+}
+
+ +
+@incollection{Paulli1993,
+  publisher = {Springer},
+  year = 1993,
+  editor = { Vidal, Ren{\'e} Victor Valqui  },
+  booktitle = {Applied Simulated Annealing},
+  title = {A computational comparison of simulated annealing and tabu search applied to the quadratic assignment problem},
+  author = {Paulli, J},
+  pages = {85--102}
+}
+
+ +
+@incollection{PavDelKes2019,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  doi = {10.1145/3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = {Pavelski, Lucas Marcondes and Delgado, Myriam Regattieri and  Marie-El{\'e}onore Kessaci },
+  title = {Meta-Learning on Flowshop Using Fitness Landscape Analysis},
+  pages = {925--933}
+}
+
+ +
+@book{Pea84,
+  author = { Judea Pearl },
+  title = {Heuristics: Intelligent Search Strategies for Computer
+                  Problem Solving},
+  publisher = {Addison-Wesley},
+  address = { Reading, MA},
+  year = 1984
+}
+
+ +
+@book{Pea93,
+  author = {Glen S. Peace},
+  title = {Taguchi Methods: A Hands-On Approach},
+  publisher = {Addison-Wesley},
+  year = 1993
+}
+
+ +
+@inproceedings{Pea2012uai,
+  year = 2013,
+  publisher = {AUAI Press},
+  editor = { Nando de Freitas  and Murphy, Kevin},
+  booktitle = {Proceedings of the Twenty-Eighth Conference on Uncertainty in Artificial Intelligence
+                  (UAI'12), Catalina Island, CA August 14-18 2012},
+  title = {The do-calculus revisited},
+  author = { Judea Pearl },
+  pages = {4--11}
+}
+
+ +
+@inproceedings{PeaBar2011aaai,
+  year = 2011,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Wolfram Burgard and Dan Roth},
+  title = {Transportability of causal and statistical relations: A formal approach},
+  author = { Judea Pearl  and  Elias Bareinboim },
+  pages = {247--254}
+}
+
+ +
+@book{PeaMac2018,
+  title = {The book of why: the new science of cause and effect},
+  author = { Judea Pearl  and Mackenzie, Dana},
+  year = 2018,
+  publisher = {Basic books}
+}
+
+ +
+@book{Pearl2009causality,
+  author = { Judea Pearl },
+  title = {Causality: Models, Reasoning and Inference},
+  year = 2009,
+  publisher = {Cambridge University Press},
+  edition = {2nd}
+}
+
+ +
+@incollection{PedCarCanVen2014,
+  author = {Juan A. Pedraza and  Carlos Garc{\'i}a-Mart{\'i}nez  and Alberto Cano and Sebasti\'an Ventura},
+  title = {Classification Rule Mining with Iterated Greedy},
+  booktitle = {Hybrid Artificial Intelligence Systems - 9th International Conference,
+               {HAIS} 2014, Salamanca, Spain, June 11-13, 2014. Proceedings},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2014,
+  editor = {Marios M. Polycarpou and Andr{\'{e}} Carlos Ponce Leon Ferreira de Carvalho and Jeng{-}Shyang Pan and Michal Wozniak and H{\'{e}}ctor Quinti{\'{a}}n and Emilio Corchado},
+  volume = 8480,
+  series = {Lecture Notes in Computer Science},
+  pages = {585--596}
+}
+
+ +
+@incollection{PedTak2013emco,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  author = {Pedro, Luciana R. and  Takahashi, R. H. C. },
+  title = {Decision-Maker Preference Modeling in Interactive
+                  Multiobjective Optimization},
+  pages = {811--824},
+  doi = {10.1007/978-3-642-37140-0_60},
+  keywords = {decision-maker, interactive, neural networks}
+}
+
+ +
+@incollection{PelBir2007:slse,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4638,
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  year = 2007,
+  booktitle = {Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS~2007},
+  author = { Paola Pellegrini  and  Mauro Birattari },
+  title = {Implementation Effort and Performance},
+  pages = {31--45}
+}
+
+ +
+@incollection{PelFavMor06,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  author = { Paola Pellegrini  and  D. Favaretto  and  E. Moretti },
+  title = {On {\MaxMinAntSystem}'s Parameters},
+  pages = {203--214}
+}
+
+ +
+@incollection{PelFavMor09:NICSO,
+  doi = {10.1007/978-3-642-03211-0},
+  editor = {Natalio Krasnogor and Belén Melián-Batista and José
+                  Andrés Moreno-Pérez and J. Marcos Moreno-Vega and David Alejandro Pelta},
+  address = { Berlin, Germany},
+  volume = 236,
+  series = {Studies in Computational Intelligence},
+  year = 2009,
+  publisher = {Springer},
+  booktitle = {Nature Inspired Cooperative Strategies for Optimization
+                  (NICSO 2008)},
+  author = { Paola Pellegrini  and  D. Favaretto  and  E. Moretti },
+  title = {Exploration in stochastic algorithms: An application
+                  on {\MaxMinAntSystem}},
+  pages = {1--13}
+}
+
+ +
+@incollection{PelStuBir2010,
+  volume = 6234,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Gianni A. {Di Caro}  and Doursat, R. and Engelbrecht, A. P. and Floreano,
+                  D. and Gambardella, L. M. and Gro\ss, R. and Sahin,
+                  E. and  Thomas St{\"u}tzle  and Sayama, H.},
+  editor = { Marco Dorigo  and others },
+  year = 2010,
+  booktitle = {Swarm Intelligence, 7th International Conference, ANTS 2010},
+  author = { Paola Pellegrini  and  Thomas St{\"u}tzle  and  Mauro Birattari },
+  title = {Off-line vs. On-line Tuning: A Study on {\MaxMinAntSystem}
+                  for the {TSP}},
+  pages = {239--250},
+  doi = {10.1007/978-3-642-15461-4_21}
+}
+
+ +
+@incollection{PerBisStu2017rf,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2017},
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Bernd Bischl  and  Thomas St{\"u}tzle },
+  title = {Evaluating random forest models for irace},
+  pages = {1146--1153},
+  doi = {10.1145/3067695.3082057}
+}
+
+ +
+@incollection{PerLopHooStu2017:lion,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 10556,
+  editor = { Roberto Battiti  and Dmitri E. Kvasov and Yaroslav D. Sergeyev},
+  year = 2017,
+  booktitle = {Learning and Intelligent Optimization, 11th International
+                  Conference, LION 11},
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {An Experimental Study of Adaptive Capping in {\rpackage{irace}}},
+  pages = {235--250},
+  doi = {10.1007/978-3-319-69404-7_17},
+  supplement = {http://iridia.ulb.ac.be/supp/IridiaSupp2016-007/}
+}
+
+ +
+@misc{PerLopHooStu2017:lion-supp,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Holger H. Hoos  and  Thomas St{\"u}tzle },
+  title = {An experimental study of adaptive capping in irace: Supplementary material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2016-007/}},
+  year = 2017
+}
+
+ +
+@incollection{PerLopStu2014ants,
+  volume = 8667,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  year = 2014,
+  booktitle = {Swarm Intelligence, 9th International Conference, ANTS 2014},
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Ant Colony Optimization on a Budget of 1000},
+  doi = {10.1007/978-3-319-09952-1_5},
+  pages = {50--61}
+}
+
+ +
+@incollection{PerLopStu2014evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 8600,
+  series = {Lecture Notes in Computer Science},
+  year = 2014,
+  booktitle = {Proceedings of EvoCOP 2014 -- 14th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Christian Blum  and  Gabriela Ochoa },
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {An Analysis of Parameters of irace},
+  doi = {10.1007/978-3-662-44320-0_4},
+  pages = {37--48}
+}
+
+ +
+@misc{PerLopStu2015budget-supp,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Ant Colony Optimization on a Budget of 1000: {Supplementary} material},
+  url = {http://iridia.ulb.ac.be/supp/IridiaSupp2015-004},
+  year = 2015
+}
+
+ +
+@incollection{PerPagFraStu2017gcc,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Lutton, Evelyne and Legrand, Pierrick and Parrend, Pierre and  Nicolas Monmarch{\'e}  and  Marc Schoenauer },
+  volume = 10764,
+  series = {Lecture Notes in Computer Science},
+  year = 2017,
+  booktitle = {EA 2017: Artificial Evolution},
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Federico Pagnozzi  and  Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Automatic Configuration of {GCC} Using {\rpackage{irace}}},
+  pages = {202--216},
+  abstract = {Automatic algorithm configuration techniques have proved to
+                  be successful in finding performance-optimizing parameter
+                  settings of many search-based decision and optimization
+                  algorithms. A recurrent, important step in software
+                  development is the compilation of source code written in some
+                  programming language into machine-executable code. The
+                  generation of performance-optimized machine code itself is a
+                  difficult task that can be parametrized in many different
+                  possible ways. While modern compilers usually offer different
+                  levels of optimization as possible defaults, they have a
+                  larger number of other flags and numerical parameters that
+                  impact properties of the generated machine-code. While the
+                  generation of performance-optimized machine code has received
+                  large attention and is dealt with in the research area of
+                  auto-tuning, the usage of standard automatic algorithm
+                  configuration software has not been explored, even though, as
+                  we show in this article, the performance of the compiled code
+                  has significant stochasticity, just as standard optimization
+                  algorithms. As a practical case study, we consider the
+                  configuration of the well-known GNU compiler collection (GCC)
+                  for minimizing the run-time of machine code for various
+                  heuristic search methods. Our experimental results show that,
+                  depending on the specific code to be optimized, improvements
+                  of up to 40{\%} of execution time when compared to the -O2
+                  and -O3 optimization flags is possible.},
+  doi = {10.1007/978-3-319-78133-4_15}
+}
+
+ +
+@misc{PerPagFraStu2017gccsup,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Federico Pagnozzi  and  Alberto Franzin  and  Thomas St{\"u}tzle },
+  title = {Automatic configuration of {GCC} using irace: Supplementary
+                  material},
+  howpublished = {\url{http://iridia.ulb.ac.be/supp/IridiaSupp2017-009/}},
+  year = 2017
+}
+
+ +
+@phdthesis{Perez-Caceres2017phd,
+  author = {  P{\'e}rez C{\'a}ceres, Leslie and  Thomas St{\"u}tzle },
+  title = {Automatic Algorithm Configuration: Analysis, Improvements and
+                  Applications},
+  school = {IRIDIA, {\'E}cole polytechnique, Universit{\'e} Libre de Bruxelles, Belgium},
+  year = 2017,
+  annote = {Supervised by  Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  epub = {http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/262048}
+}
+
+ +
+@incollection{PetEve2002control,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  editor = { Langdon, William B.  and others},
+  year = 2002,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2002},
+  title = {Controlling genetic algorithms with reinforcement learning},
+  author = {Pettinger, James E. and  Everson, Richard M. },
+  pages = {692--692}
+}
+
+ +
+@book{PetJanSch2017,
+  title = {Elements of causal inference: foundations and learning algorithms},
+  author = {Peters, Jonas and Janzing, Dominik and Sch{\"o}lkopf, Bernhard},
+  year = 2017,
+  publisher = {MIT Press}
+}
+
+ +
+@incollection{PhiBhaPas2021portfolio,
+  author = {Phillipson, Frank and Bhatia, Harshil Singh},
+  title = {Portfolio Optimisation Using the {D-Wave} Quantum Annealer},
+  booktitle = {Computational Science -- ICCS 2021},
+  publisher = {Springer International Publishing},
+  year = 2021,
+  editor = {Paszynski, Maciej and Kranzlm{\"u}ller, Dieter and
+                  Krzhizhanovskaya, Valeria V.  and Dongarra, Jack J.  and
+                  Sloot, Peter M. A.},
+  pages = {45--59},
+  address = { Cham, Switzerland}
+}
+
+ +
+@inproceedings{PihMus2014,
+  publisher = {IEEE Press},
+  year = 2014,
+  editor = {Papadopoulos, George Angelos},
+  booktitle = {26th {IEEE} International Conference on Tools with Artificial Intelligence,
+                  {ICTAI} 2014, Limassol, Cyprus, November 10-12, 2014},
+  title = {Application of Machine Learning to Algorithm Selection for {TSP}},
+  author = {Pihera, Josef and  Musliu, Nysret },
+  pages = {47--54}
+}
+
+ +
+@incollection{PilWhi02,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = {M. L. Pilat and T. White},
+  title = {Using Genetic Algorithms to optimize {ACS-TSP}},
+  pages = {282--287}
+}
+
+ +
+@book{Pin2012,
+  author = {Michael L. Pinedo},
+  title = {Scheduling: Theory, Algorithms, and Systems},
+  publisher = {Springer},
+  year = 2012,
+  edition = {4th},
+  address = { New York, NY}
+}
+
+ +
+@incollection{PinRunSoup07:maxsat,
+  author = {Pedro Pinto and Thomas Runkler and Jo{\~a}o Sousa},
+  title = {Ant Colony Optimization and its Application to
+                  Regular and Dynamic {MAX-SAT} Problems},
+  booktitle = {Advances in Biologically Inspired Information
+                  Systems},
+  year = 2007,
+  volume = 69,
+  pages = {285--304},
+  series = {Studies in Computational Intelligence},
+  doi = {10.1007/978-3-540-72693-7_15},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  abstract = {In this chapter we discuss the ant colony
+                  optimization meta-heuristic {(ACO)} and its
+                  application to static and dynamic constraint
+                  satisfaction optimization problems, in particular
+                  the static and dynamic maximum satisfiability
+                  problems {(MAX-SAT).} In the first part of the
+                  chapter we give an introduction to meta-heuristics
+                  in general and ant colony optimization in
+                  particular, followed by an introduction to
+                  constraint satisfaction and static and dynamic
+                  constraint satisfaction optimization problems.
+                  Then, we describe how to apply the {ACO} algorithm
+                  to the problems, and do an analysis of the results
+                  obtained for several benchmarks.  The adapted ant
+                  colony optimization accomplishes very well the task
+                  of dealing with systematic changes of dynamic
+                  {MAX-SAT} instances derived from static problems.  }
+}
+
+ +
+@misc{PinSin2020checklist,
+  author = {Joelle Pineau and Koustuv Sinha},
+  title = {The Machine Learning Reproducibility Checklist (v2.0)},
+  howpublished = {\url{https://www.cs.mcgill.ca/~jpineau/ReproducibilityChecklist-v2.0.pdf}},
+  year = 2020,
+  annote = {Used in NeurIPS 2020}
+}
+
+ +
+@incollection{PisRop2010:handbook,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  title = {Large Neighborhood Search},
+  author = { David Pisinger  and  Stefan Ropke },
+  pages = {399--419}
+}
+
+ +
+@incollection{PitBehAff2013,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2013,
+  volume = 7832,
+  booktitle = {Proceedings of EvoCOP 2013 -- 13th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Martin Middendorf  and  Christian Blum },
+  author = {Pitzer, Erik and Beham, Andreas and Affenzeller, Michael},
+  title = {Automatic Algorithm Selection for the Quadratic Assignment
+                   Problem Using Fitness Landscape Analysis},
+  pages = {109--120}
+}
+
+ +
+@inproceedings{PloMelVarBucZhuLee2013tact,
+  series = {Procedia Computer Science},
+  year = 2013,
+  volume = 18,
+  publisher = {Elsevier},
+  editor = {Vassil Alexandrov and Michael Lees and Valeria Krzhizhanovskaya and Jack Dongarra and Peter M.A. Sloot},
+  booktitle = {2013 International Conference on Computational Science},
+  author = {Dmitry Plotnikov and Dmitry Melnik and Mamikon Vardanyan and
+                  Ruben Buchatskiy and Roman Zhuykov and Je-Hyung Lee},
+  title = {Automatic Tuning of Compiler Optimizations and Analysis of
+                  their Impact},
+  pages = {1312--1321},
+  doi = {10.1016/j.procs.2013.05.298}
+}
+
+ +
+@book{Plotnick2009book,
+  author = {Robert Plotnick},
+  title = {The Genie in the Machine: How Computer-Automated Inventing Is
+                  Revolutionizing Law and Business},
+  publisher = {Stanford Law Books},
+  year = 2009,
+  annote = {Mentions evolutionary optimization of Oral-B toothbrushes}
+}
+
+ +
+@techreport{Pow2009bobyqa,
+  author = { Powell, Michael J. D.},
+  title = {The {BOBYQA} algorithm for bound constrained optimization
+                  without derivatives},
+  institution = {University of Cambridge, UK},
+  year = 2009,
+  number = {Cambridge NA Report NA2009/06},
+  epub = {http://www6.cityu.edu.hk/rcms/publications/preprint26.pdf}
+}
+
+ +
+@incollection{Powell1994cobyla,
+  author = { Powell, Michael J. D.},
+  title = {A Direct Search Optimization Method That Models the Objective
+                  and Constraint Functions by Linear Interpolation},
+  booktitle = {Advances in Optimization and Numerical Analysis},
+  publisher = {Springer},
+  year = 1994,
+  pages = {51--67},
+  annote = {Proposed COBYLA},
+  isbn = 9789401583305,
+  doi = {10.1007/978-94-015-8330-5_4}
+}
+
+ +
+@inproceedings{PraTraWanBaeKer2020,
+  publisher = {IEEE Press},
+  year = 2020,
+  booktitle = {2020 {IEEE} Symposium Series on Computational Intelligence, {SSCI}
+                  2020, Canberra, Australia, December 1-4, 2020},
+  editor = { Carlos A. {Coello Coello} },
+  title = {Per-Instance Configuration of the Modularized {CMA-ES}
+                  by Means of Classifier Chains and Exploratory Landscape Analysis},
+  author = {Prager, Raphael Patrick and  Heike Trautmann  and Wang, Hao and  Thomas B{\"a}ck  and  Pascal Kerschke },
+  pages = {996--1003}
+}
+
+ +
+@inproceedings{PraYao2006taa,
+  title = {A new multi-objective evolutionary optimisation algorithm:
+                  the two-archive algorithm},
+  author = {Praditwong, Kata and  Xin Yao },
+  booktitle = {International Conference on Computational Intelligence and
+                  Security},
+  volume = 1,
+  pages = {286--291},
+  year = 2006,
+  organization = {IEEE}
+}
+
+ +
+@incollection{Prasad03,
+  publisher = {CRC Press},
+  year = 2003,
+  booktitle = {Advances in Water Supply Management},
+  editor = { C. Maksimovi{\'c}  and  David Butler  and  Fayyaz Ali Memon },
+  author = { T. Devi Prasad  and  Godfrey A. Walters },
+  title = {Optimal rerouting to minimise residence times in
+                  water distribution networks},
+  pages = {299--306}
+}
+
+ +
+@book{PrepSha1988:compgeom,
+  author = {F. P. Preparata and M. I. Shamos},
+  title = {Computational Geometry. An Introduction},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 1988,
+  edition = {2nd},
+  alias = {preparata:computational}
+}
+
+ +
+@incollection{PriAllLop2022gecco,
+  location = {Boston, Massachusetts},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Pricopie, Stefan  and  Allmendinger, Richard  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and Fare,
+                  Clyde and Benatan, Matt and  Joshua D. Knowles },
+  title = {Expensive Optimization with Production-Graph Resource
+                  Constraints: A First Look at a New Problem Class},
+  doi = {10.1145/3512290.3528741},
+  abstract = {We consider a new class of expensive, resource-constrained
+                  optimization problems (here arising from molecular discovery)
+                  where costs are associated with the experiments (or
+                  evaluations) to be carried out during the optimization
+                  process. In the molecular discovery problem, candidate
+                  compounds to be optimized must be synthesized in an iterative
+                  process that starts from a set of purchasable items and
+                  builds up to larger molecules. To produce target molecules,
+                  their required resources are either used from
+                  already-synthesized items in storage or produced themselves
+                  on-demand at an additional cost. Any remaining resources from
+                  the production process are stored for reuse for the next
+                  evaluations. We model these resource dependencies with a
+                  directed acyclic production graph describing the development
+                  process from granular purchasable items to evaluable target
+                  compounds. Moreover, we develop several resource-eficient
+                  algorithms to address this problem. In particular, we develop
+                  resource-aware variants of Random Search heuristics and of
+                  Bayesian Optimization and analyze their performance in terms
+                  of anytime behavior. The experimental results were obtained
+                  from a real-world molecular optimization problem. Our results
+                  suggest that algorithms that encourage exploitation by
+                  reusing existing resources achieve satisfactory results while
+                  using fewer resources overall.},
+  pages = {840--848},
+  numpages = 9,
+  keywords = {molecular discovery, resource constraints, expensive
+                  optimization, production costs}
+}
+
+ +
+@book{PriStoLam2005:book,
+  title = {Differential Evolution: A Practical Approach to Global
+                  Optimization},
+  author = {Price, Kenneth and Storn, Rainer M. and Lampinen, Jouni A.},
+  year = 2005,
+  publisher = {Springer},
+  address = { New York, NY},
+  doi = {10.1007/3-540-31306-0}
+}
+
+ +
+@incollection{PryMosNaz2007heatmap,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  title = {Heatmap visualization of population based multi objective
+                  algorithms},
+  author = {Pryke, Andy and  Mostaghim, Sanaz  and Nazemi, Alireza},
+  pages = {361--375}
+}
+
+ +
+@incollection{PulCoe2003emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 2632,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  year = 2003,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  author = { Gregorio {Toscano Pulido}  and  Carlos A. {Coello Coello} },
+  title = {The Micro Genetic Algorithm 2: Towards Online
+                  Adaptation in Evolutionary Multiobjective
+                  Optimization},
+  pages = {252--266},
+  doi = {10.1007/3-540-36970-8_18}
+}
+
+ +
+@techreport{PurDebMan2014coin,
+  title = {A review of hybrid evolutionary multiple criteria decision
+                  making methods},
+  author = { Robin C. Purshouse  and  Kalyanmoy Deb  and Mansor, Maszatul M. and  Mostaghim, Sanaz  and Wang, Rui},
+  year = 2014,
+  institution = {Computational Optimization and Innovation (COIN) Laboratory, University of Michigan, USA},
+  type = {COIN Report},
+  number = 2014005,
+  month = jan
+}
+
+ +
+@inproceedings{PurFle2003cec,
+  year = 2003,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = dec,
+  booktitle = {Proceedings of  the 2003 Congress on Evolutionary Computation (CEC'03)},
+  key = {IEEE CEC},
+  title = {Evolutionary many-objective optimisation: an exploratory
+                  analysis},
+  author = { Robin C. Purshouse  and  Peter J. Fleming },
+  doi = {10.1109/CEC.2003.1299927},
+  pages = {2066--2073},
+  annote = {First to mention NSGA-II failure to deal with
+                  many-objectives. Mentions exponential number of nondominated
+                  solutions with respect to many objectives (but
+                  \cite{FarAma2002nafips} already did).}
+}
+
+ +
+@incollection{PusFraVor2011spiral,
+  doi = {10.1007/978-0-387-09766-4_244},
+  publisher = {Springer, US},
+  year = 2011,
+  editor = {David Padua},
+  booktitle = {Encyclopedia of Parallel Computing},
+  author = {Markus P\"{u}schel and Franz Franchetti and Yevgen Voronenko},
+  title = {Spiral},
+  pages = {1920--1933}
+}
+
+ +
+@incollection{PusHoo2018ppsn,
+  volume = 11101,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = {Yasha Pushak and  Holger H. Hoos },
+  title = {Algorithm Configuration Landscapes: More Benign Than
+                  Expected?},
+  pages = {271--283},
+  doi = {10.1007/978-3-319-99259-4_22},
+  supplement = {http://www.cs.ubc.ca/labs/beta/Projects/ACLandscapes/},
+  annote = {Best paper award at PPSN2018}
+}
+
+ +
+@incollection{PusHoo2020golden,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  doi = {10.1145/3377930},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  author = {Yasha Pushak and  Holger H. Hoos },
+  title = {Golden parameter search: exploiting structure to quickly
+                  configure parameters in parallel},
+  pages = {245--253},
+  keywords = {algorithm configuration}
+}
+
+ +
+@manual{R:ParamHelpers,
+  title = {{\rpackage{ParamHelpers}} : Helpers for Parameters in
+                  Black-Box Optimization, Tuning and Machine Learning},
+  author = { Bernd Bischl  and Michel Lang and  Jakob Bossek  and Daniel Horn
+                  and Karin Schork and Jakob Richter and  Pascal Kerschke },
+  note = {\proglang{R} package version 1.10},
+  url = {https://cran.r-project.org/package=ParamHelpers},
+  year = 2017,
+  alias = {Bischl2017}
+}
+
+ +
+@manual{R:Rmpi,
+  title = {{\rpackage{Rmpi}}: Interface (Wrapper) to MPI (Message-Passing
+                  Interface)},
+  author = {Hao Yu},
+  year = 2010,
+  note = {\proglang{R} package version 0.5-8},
+  url = {http://cran.r-project.org/package=Rmpi}
+}
+
+ +
+@manual{R:SPOT,
+  title = {{\rpackage{SPOT}}: Sequential Parameter
+                  Optimization},
+  author = { Thomas Bartz-Beielstein  and J. Ziegenhirt and W. Konen
+                  and O. Flasch and P. Koch and  Martin Zaefferer },
+  year = 2011,
+  note = {\proglang{R} package},
+  url = {http://cran.r-project.org/package=SPOT}
+}
+
+ +
+@manual{R:cmaes,
+  title = {{\rpackage{cmaes}}: Covariance Matrix Adapting Evolutionary Strategy},
+  author = { Heike Trautmann  and  Olaf Mersmann  and  David Arnu },
+  note = {\proglang{R} package},
+  url = {http://cran.r-project.org/package=cmaes},
+  year = 2011
+}
+
+ +
+@manual{R:lhs,
+  title = {{\rpackage{lhs}}: Latin Hypercube Samples},
+  author = {Carnell, Rob},
+  note = {\proglang{R} package version 0.14},
+  url = {http://r-forge.r-project.org/projects/lhs/},
+  year = 2016,
+  alias = {Carnell2016}
+}
+
+ +
+@manual{R:mco,
+  title = {{\rpackage{mco}}: Multiple Criteria Optimization Algorithms and Related Functions},
+  author = { Olaf Mersmann },
+  year = 2014,
+  note = {\proglang{R} package version 1.0-15.1},
+  url = {http://CRAN.R-project.org/package=mco}
+}
+
+ +
+@manual{R:mlr,
+  title = {{\rpackage{mlr}}: Machine Learning in \proglang{R}},
+  author = { Bernd Bischl  and Michel Lang and  Jakob Bossek  and Leonard Judt and Jakob Richter and Tobias Kuehn and Erich Studerus},
+  year = 2013,
+  note = {\proglang{R} package},
+  url = {http://cran.r-project.org/package=mlr}
+}
+
+ +
+@manual{R:multicore,
+  title = {{\rpackage{multicore}}: Parallel Processing of \proglang{R} Code on Machines
+                  with Multiple Cores or CPUs},
+  author = {Simon Urbanek},
+  year = 2010,
+  note = {\proglang{R} package version 0.1-3},
+  url = {http://www.rforge.net/multicore/}
+}
+
+ +
+@manual{R:smoof,
+  title = {{\rpackage{smoof}}: Single and Multi-Objective Optimization Test Functions},
+  author = { Jakob Bossek },
+  year = 2016,
+  note = {\proglang{R} package version 1.2},
+  url = {http://CRAN.R-project.org/package=smoof}
+}
+
+ +
+@inproceedings{RachSri2006,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  month = jul,
+  year = 2006,
+  booktitle = {Proceedings of  the 2006 Congress on Evolutionary Computation (CEC 2006)},
+  key = {IEEE CEC},
+  author = {L. Rachmawati and D. Srinivasan},
+  title = {Preference incorporation in multiobjective evolutionary
+                  algorithms: A survey},
+  pages = {3385--3391}
+}
+
+ +
+@incollection{RadLopStu2013emo,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  author = { Andreea Radulescu  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Thomas St{\"u}tzle },
+  title = {Automatically Improving the Anytime Behaviour of
+                  Multiobjective Evolutionary Algorithms},
+  pages = {825--840},
+  doi = {10.1007/978-3-642-37140-0_61}
+}
+
+ +
+@incollection{RahEveFie2017infill,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  author = {Rahat, Alma A. M. and  Everson, Richard M.  and  Jonathan E. Fieldsend },
+  title = {Alternative infill strategies for expensive multi-objective
+                  optimisation},
+  pages = {873--880}
+}
+
+ +
+@incollection{Ran04,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  title = {Near Parameter Free Ant Colony Optimisation},
+  author = { Marcus Randall },
+  pages = {374--381}
+}
+
+ +
+@incollection{RanMon2002:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { Marcus Randall  and  James Montgomery },
+  title = {Candidate Set Strategies for Ant Colony
+                  Optimisation},
+  pages = {243--249}
+}
+
+ +
+@inproceedings{Rao05,
+  month = sep,
+  address = {University of Exeter, UK},
+  volume = 1,
+  editor = { Dragan A. Savic  and  Godfrey A. Walters  and  Roger King  and  Soon Thiam-Khu },
+  year = 2005,
+  booktitle = {Proceedings of the Eighth International Conference on
+                  Computing and Control for the Water Industry (CCWI 2005)},
+  author = { Zhengfu Rao  and  Jon Wicks and Sue West},
+  title = {{ENCOMS} - An Energy Cost Minimisation System for Real-Time,
+                  Operational Control of Water Distribution Networks},
+  pages = {85--90}
+}
+
+ +
+@incollection{RapWanBuj2020bopca,
+  volume = 12269,
+  year = 2020,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas B{\"a}ck  and  Mike Preuss  and Deutz, Andr{\'e} and Wang, Hao and  Carola Doerr  and  Emmerich, Michael T. M.  and  Heike Trautmann },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVI}},
+  author = {Elena Raponi and Hao Wang and Mariusz Bujny and Simonetta
+                  Boria and  Carola Doerr },
+  title = {High Dimensional Bayesian Optimization Assisted by Principal
+                  Component Analysis},
+  pages = {169--183},
+  doi = {10.1007/978-3-030-58112-1_12}
+}
+
+ +
+@incollection{RasMusKar2014MSOST,
+  year = 2014,
+  volume = 34,
+  series = {Computational Methods in Applied Sciences},
+  publisher = {Springer},
+  booktitle = {Modeling, Simulation and Optimization for Science and Technology},
+  editor = {William Fitzgibbon and
+                  Yuri A. Kuznetsov and
+                  Pekka Neittaanm{\"a}ki and
+                  Olivier Pironneau},
+  author = {Jussi Rasku and  Musliu, Nysret  and Tommi K{\"a}rkk{\"a}inen},
+  title = {Automating the Parameter Selection in {VRP}: An Off-line Parameter Tuning Tool Comparison},
+  pages = {191--209},
+  doi = {10.1007/978-94-017-9054-3_11},
+  keywords = {irace}
+}
+
+ +
+@book{RasWil2006gp,
+  title = {Gaussian Processes for Machine Learning},
+  author = {Rasmussen, Carl Edward and Williams, Christopher K. I.},
+  year = 2006,
+  keywords = {Gaussian processes, data processing},
+  language = {English},
+  publisher = {MIT Press},
+  address = {Cambridge, MA},
+  alias = {Rasmussen2006},
+  isbn = {026218253X}
+}
+
+ +
+@techreport{Ray2011gart,
+  author = {Rayner, N.},
+  title = {Maverick Research: Judgment Day, or Why We Should Let Machines Automate Decision Making},
+  year = 2011,
+  month = oct,
+  institution = {Gartner, Inc},
+  type = {Gartner Research Note}
+}
+
+ +
+@phdthesis{Rec1971PhD,
+  author = { Rechenberg, Ingo },
+  title = {Evolutionsstrategie: {Optimierung} technischer {Systeme}
+                  nach {Prinzipien} der biologischen {Evolution}},
+  school = {Department of Process Engineering, Technical
+                  University of Berlin},
+  year = 1971
+}
+
+ +
+@book{Rec1973,
+  author = { Rechenberg, Ingo },
+  title = {Evolutionsstrategie: {Optimierung} technischer {Systeme}
+                  nach {Prinzipien} der biologischen {Evolution}},
+  publisher = {Frommann-Holzboog, Stuttgart, Germany},
+  year = 1973
+}
+
+ +
+@incollection{Ree2010:ga,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Colin R. Reeves },
+  title = {Genetic algorithms},
+  chapter = 5,
+  pages = {109--140}
+}
+
+ +
+@incollection{Ree2013:many,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  title = {Many-Objective Visual Analytics: Rethinking the Design of
+                  Complex Engineered Systems},
+  author = { Patrick M. Reed },
+  pages = {1--1}
+}
+
+ +
+@incollection{Rei2007:hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4771,
+  editor = { Thomas Bartz-Beielstein  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Boris Naujoks  and  Andrea Roli  and  G{\"u}nther Rudolph  and  M. Sampels },
+  year = 2007,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Marc Reimann },
+  title = {Guiding {ACO} by Problem Relaxation: {A} Case Study
+                  on the Symmetric {TSP}},
+  pages = {45--56}
+}
+
+ +
+@book{Rei94,
+  author = { Gerhard Reinelt },
+  title = {The Traveling Salesman: Computational Solutions for
+                  {TSP} Applications},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 1994,
+  volume = 840,
+  series = {Lecture Notes in Computer Science}
+}
+
+ +
+@incollection{ResRib2002,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Mauricio G. C. Resende  and  Celso C. Ribeiro },
+  title = {Greedy Randomized Adaptive Search Procedures},
+  pages = {219--249}
+}
+
+ +
+@incollection{ResRib2010,
+  address = { New York, NY},
+  publisher = {Springer},
+  edition = {2nd},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 146,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Mauricio G. C. Resende  and  Celso C. Ribeiro },
+  title = {Greedy Randomized Adaptive Search Procedures: Advances, Hybridizations, and Applications},
+  pages = {283--319}
+}
+
+ +
+@incollection{ReyCoe2005omopso,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 3410,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  year = 2005,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  author = {Reyes Sierra, Margarita and  Carlos A. {Coello Coello} },
+  title = {Improving {PSO}-Based Multi-objective Optimization Using
+                  Crowding, Mutation and $\epsilon$-Dominance},
+  pages = {505--519},
+  keywords = {OMOPSO}
+}
+
+ +
+@incollection{RiaDanEkeLar2009:adt,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Francesca Rossi  and  Alexis Tsouki{\`a}s },
+  volume = 5783,
+  series = {Lecture Notes in Computer Science},
+  year = 2009,
+  booktitle = {Algorithmic Decision Theory, First International
+                  Conference, {ADT} 2009},
+  author = { Mona Riabacke  and  Mats Danielson  and  Love Ekenberg  and  Aron Larsson },
+  title = {A Prescriptive Approach for Eliciting Imprecise Weight Statements in an {MCDA} Process},
+  pages = {168--179}
+}
+
+ +
+@incollection{RidKud07,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4638,
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  year = 2007,
+  booktitle = {Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS~2007},
+  author = {Enda Ridge and Daniel Kudenko},
+  title = {Tuning the Performance of the {MMAS} Heuristic},
+  pages = {46--60}
+}
+
+ +
+@incollection{RidKud2010,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss },
+  year = 2010,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  author = {Enda Ridge and Daniel Kudenko},
+  title = {Tuning an Algorithm Using Design of Experiments},
+  pages = {265--286}
+}
+
+ +
+@inproceedings{RijWanLeeBac2016ssci,
+  year = 2016,
+  booktitle = {Computational Intelligence (SSCI), 2016 IEEE Symposium Series
+                  on},
+  editor = {Chen, Xuewen and Stafylopatis, Andreas},
+  title = {Evolving the structure of {Evolution} {Strategies}},
+  author = {van Rijn, Sander and Wang, Hao and van Leeuwen,
+                  Matthijs and  Thomas B{\"a}ck },
+  pages = {1--8},
+  doi = {10.1109/SSCI.2016.7850138},
+  keywords = {automated design, automatic configuration, CMA-ES, Gaussian
+                  distribution}
+}
+
+ +
+@manual{Rmanual,
+  title = {\proglang{R}: A Language and Environment for Statistical
+                  Computing},
+  author = {{\proglang{R} Development Core Team}},
+  organization = {\proglang{R} Foundation for Statistical Computing},
+  address = {Vienna, Austria},
+  year = 2008,
+  isbn = {3-900051-07-0},
+  url = {http://www.R-project.org},
+  alias = {R2008lang}
+}
+
+ +
+@incollection{RobFil05demo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 3410,
+  series = {Lecture Notes in Computer Science},
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  year = 2005,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  author = {Tea Robi{\v c} and Bogdan Filipi{\v c}},
+  title = {{DEMO}: Differential Evolution for Multiobjective Optimization},
+  pages = {520--533}
+}
+
+ +
+@incollection{RodBLuLozGar2012,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 7245,
+  year = 2012,
+  editor = { Jin-Kao Hao  and  Martin Middendorf },
+  booktitle = {Proceedings of EvoCOP 2012 -- 12th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = {Francisco J. Rodr{\'i}guez and  Christian Blum  and  Manuel Lozano  and  Carlos Garc{\'i}a-Mart{\'i}nez },
+  title = {Iterated Greedy Algorithms for the Maximal Covering Location Problem},
+  pages = {172--181}
+}
+
+ +
+@incollection{RodCoe2012new,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2012,
+  editor = {Terence Soule and Jason H. Moore},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  title = {A new multi-objective evolutionary algorithm based on a
+                  performance assessment indicator},
+  author = {Rodr{\'i}guez Villalobos, Cynthia A. and  Carlos A. {Coello Coello} },
+  pages = {505--512}
+}
+
+ +
+@incollection{Ros2005hyper,
+  year = 2005,
+  address = {Boston, MA},
+  publisher = {Springer},
+  editor = { Edmund K. Burke  and  Graham Kendall },
+  booktitle = {Search Methodologies},
+  author = { Peter Ross },
+  title = {Hyper-Heuristics},
+  pages = {529--556},
+  doi = {10.1007/0-387-28356-0_17}
+}
+
+ +
+@incollection{Rubin1974,
+  author = {Frank Rubin},
+  title = {An Iterative Technique for Printed Wire Routing},
+  booktitle = {DAC'74, Proceedings of the 11th Design Automation Workshop},
+  publisher = {IEEE Press},
+  year = 1974,
+  pages = {308--313}
+}
+
+ +
+@inproceedings{RudAga2000cec,
+  month = jul,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2000,
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  key = {IEEE CEC},
+  author = { G{\"u}nther Rudolph  and Alexandru Agapie},
+  title = {Convergence Properties of Some Multi-Objective
+                  Evolutionary Algorithms},
+  pages = {1010--1016},
+  volume = 2
+}
+
+ +
+@incollection{RudTraSen2013evenly,
+  isbn = {978-3-642-37139-4},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7811,
+  year = 2013,
+  publisher = {Springer},
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  title = {Evenly spaced {Pareto} front approximations for tricriteria
+                  problems based on triangulation},
+  author = { G{\"u}nther Rudolph  and  Heike Trautmann  and Sengupta, Soumyadip and  Oliver Sch{\"u}tze },
+  pages = {443--458},
+  annote = {unbounded archiver, AA$_{\Delta_1}$}
+}
+
+ +
+@mastersthesis{Rudolph1990diploma,
+  author = { G{\"u}nther Rudolph },
+  title = {Globale Optimierung mit parallelen Evolutionsstrategien},
+  school = {Department of Computer Science, University of Dortmund},
+  year = 1990,
+  type = {Diplomarbeit},
+  month = jul,
+  annote = {Proposed the generalized Rastrigin function}
+}
+
+ +
+@incollection{Rudolph1992,
+  year = 1992,
+  publisher = {Elsevier},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {II}},
+  editor = {Reinhard M{\"a}nner and Bernard Manderick},
+  author = { G{\"u}nther Rudolph },
+  title = {On Correlated Mutations in Evolution Strategies},
+  pages = {107--116}
+}
+
+ +
+@incollection{Rudolph1994:icec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1994,
+  editor = { Zbigniew Michalewicz },
+  booktitle = {Proceedings of the First IEEE International Conference on
+                  Evolutionary Computation (ICEC'94)},
+  title = {Convergence of non-elitist strategies},
+  author = { G{\"u}nther Rudolph },
+  pages = {63--66}
+}
+
+ +
+@incollection{Rudolph1998ep,
+  year = 1998,
+  publisher = {Springer},
+  volume = 1447,
+  series = {Lecture Notes in Computer Science},
+  booktitle = {International Conference on Evolutionary Programming},
+  editor = {V. William Porto and N. Saravanan and Donald E. Waagen and  Agoston E. Eiben },
+  author = { G{\"u}nther Rudolph },
+  title = {Evolutionary Search for Minimal Elements in Partially Ordered
+                  Finite Sets},
+  pages = {345--353},
+  doi = {10.1007/BFb0040787}
+}
+
+ +
+@incollection{RuiLuqMieSab2015,
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and  Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  volume = 9019,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {II}},
+  doi = {10.1007/978-3-319-15892-1_17},
+  year = 2015,
+  pages = {249--263},
+  author = { Ruiz, Ana Bel{\'e}n  and  Mariano Luque  and  Kaisa Miettinen  and  Rub{\'{e}}n Saborido },
+  title = {An Interactive Evolutionary Multiobjective Optimization
+                  Method: Interactive {WASF}-{GA}}
+}
+
+ +
+@incollection{RuiValFer2009,
+  title = {Scheduling in flowshops with no-idle machines},
+  author = { Rub{\'e}n Ruiz  and  Eva Vallada  and Fern{\'a}ndez-Mart{\'i}nez, Carlos},
+  booktitle = {Computational intelligence in flow shop and job shop scheduling},
+  pages = {21--51},
+  year = 2009,
+  publisher = {Springer}
+}
+
+ +
+@inproceedings{Rum01:ijcai,
+  publisher = {IEEE Press},
+  year = 2001,
+  booktitle = {Proceedings of  the 17th International Joint Conference on Artificial Intelligence (IJCAI-01)},
+  editor = {Bernhard Nebel},
+  author = {W. Ruml},
+  title = {Incomplete Tree Search using Adaptive Probing},
+  pages = {235--241}
+}
+
+ +
+@book{RusNor2003,
+  title = {Artificial Intelligence: A Modern Approach},
+  author = {Russell, Stuart J. and Norvig, Peter},
+  volume = 2,
+  year = 2003,
+  publisher = {Prentice Hall, Englewood Cliffs, NJ}
+}
+
+ +
+@incollection{Rust1994mdp,
+  title = {Structural estimation of {Markov} decision processes},
+  author = {Rust, John},
+  booktitle = {Handbook of Econometrics},
+  volume = 4,
+  pages = {3081--3143},
+  year = 1994,
+  publisher = {Elsevier},
+  doi = {10.1016/S1573-4412(05)80020-0}
+}
+
+ +
+@inproceedings{SUMO2011,
+  author = {Behrisch, Michael and Bieker, Laura and Erdmann, Jakob and  Krajzewicz, Daniel },
+  title = {{SUMO} - {Simulation} of {Urban} {MO}bility: An Overview},
+  booktitle = {SIMUL 2011, The Third International Conference on Advances in
+                  System Simulation},
+  year = 2011,
+  pages = {63--68},
+  address = {Barcelona, Spain},
+  organization = {ThinkMind}
+}
+
+ +
+@incollection{SaiLopMie2019gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3319619},
+  isbn = {978-1-4503-6748-6},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  author = { Saini, Bhupinder Singh  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kaisa Miettinen },
+  title = {Automatic Surrogate Modelling Technique Selection based on
+                  Features of Optimization Problems},
+  doi = {10.1145/3319619.3326890},
+  pages = {1765--1772},
+  abstract = {A typical scenario when solving industrial single or
+                  multiobjective optimization problems is that no explicit
+                  formulation of the problem is available. Instead, a dataset
+                  containing vectors of decision variables together with their
+                  objective function value(s) is given and a surrogate model
+                  (or metamodel) is build from the data and used for
+                  optimization and decision-making. This data-driven
+                  optimization process strongly depends on the ability of the
+                  surrogate model to predict the objective value of decision
+                  variables not present in the original dataset. Therefore, the
+                  choice of surrogate modelling technique is crucial. While
+                  many surrogate modelling techniques have been discussed in
+                  the literature, there is no standard procedure that will
+                  select the best technique for a given problem. In this work,
+                  we propose the automatic selection of a surrogate modelling
+                  technique based on exploratory landscape features of the
+                  optimization problem that underlies the given dataset. The
+                  overall idea is to learn offline from a large pool of
+                  benchmark problems, on which we can evaluate a large number
+                  of surrogate modelling techniques. When given a new dataset,
+                  features are used to select the most appropriate surrogate
+                  modelling technique. The preliminary experiments reported
+                  here suggest that the proposed automatic selector is able to
+                  identify high-accuracy surrogate models as long as an
+                  appropriate classifier is used for selection.}
+}
+
+ +
+@incollection{SakTakKaw2010,
+  title = {A method to control parameters of evolutionary algorithms by
+                  using reinforcement learning},
+  author = {Sakurai, Yoshitaka and Takada, Kouhei and Kawabe, Takashi and
+                  Tsuruta, Setsuo},
+  booktitle = {2010 Sixth International Conference on Signal-Image
+                  Technology and Internet Based Systems},
+  pages = {74--79},
+  year = 2010,
+  publisher = {IEEE}
+}
+
+ +
+@incollection{Sakarya99,
+  address = {Baldock, United Kingdom},
+  publisher = { Research Studies Press Ltd. },
+  volume = 2,
+  year = 1999,
+  booktitle = {Water Industry Systems: Modelling and Optimization
+                  Applications},
+  editor = { Dragan A. Savic  and  Godfrey A. Walters },
+  author = { A. Burcu Altan Sakarya  and  Fred E. Goldman  and  Larry W. Mays },
+  title = {Models for the optimal scheduling of pumps to meet
+                  water quality},
+  pages = {379--391},
+  note = {}
+}
+
+ +
+@inproceedings{SamDiCFra2015,
+  publisher = {IEEE Press},
+  year = 2015,
+  editor = {Lovell, Nigel and Mainardi, Luca},
+  series = {Annual International Conference of the {IEEE} Engineering in Medicine and Biology},
+  booktitle = {37th Annual International Conference of the {IEEE} Engineering
+                  in Medicine and Biology Society, EMBC 2015, Proceedings},
+  author = {Sambo, Francesco and Di Camillo, Barbara and  Alberto Franzin  and Facchinetti, Andrea
+                  and Hakaste, Liisa and Kravic, Jasmina and Fico, Giuseppe and Tuomilehto, Jaakko
+                  and Groop, Leif and Gabriel, Rafael and Tuomi, Tiinamaija and Cobelli, Claudio},
+  title = {A Bayesian Network analysis of the probabilistic relations between
+                  risk factors in the predisposition to type 2 diabetes},
+  pages = {2119--2122}
+}
+
+ +
+@inproceedings{SanBai2022permutations,
+  year = 2022,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2022 World Congress on Computational Intelligence (WCCI 2022)},
+  key = {WCCI},
+  title = {A Fast Randomized Local Search for Low Budget Optimization in
+                  Black-Box Permutation Problems},
+  author = {Valentino Santucci and Marco Baioletti},
+  keywords = {UMM, CEGO}
+}
+
+ +
+@book{SanWilNot2003,
+  author = {Thomas J. Santner and Brian J. Williams and William I. Notz},
+  title = {The Design and Analysis of Computer Experiments},
+  publisher = {Springer Verlag},
+  address = { New York, NY},
+  year = 2003,
+  doi = {10.1007/978-1-4757-3799-8},
+  pages = {2083},
+  alias = {SatWilNot2003;SanWilNot03}
+}
+
+ +
+@misc{SatYouPat2017tpu,
+  author = {Sato, Kaz and Young, Cliff},
+  title = {An in-depth look at Google's first Tensor Processing Unit (TPU)},
+  year = 2017,
+  howpublished = {\url{https://cloud.google.com/blog/products/ai-machine-learning/an-in-depth-look-at-googles-first-tensor-processing-unit-tpu}}
+}
+
+ +
+@inproceedings{Savic97,
+  author = { Dragan A. Savic  and  Godfrey A. Walters  and  Martin Schwab },
+  title = {Multiobjective Genetic Algorithms for Pump
+                  Scheduling in Water Supply},
+  booktitle = {Evolutionary Computing Workshop, {AISB}'97},
+  pages = {227--236},
+  year = 1997,
+  editor = { David Corne  and  J. L. Shapiro },
+  volume = 1305,
+  series = {Lecture Notes in Computer Science},
+  publisher = { Berlin, Germany},
+  postscript = {Savic97 - Multiobjective GA for Pump Scheduling.ps}
+}
+
+ +
+@book{SawNakTan1985theory,
+  author = {Sawaragi, Y. and Nakayama, H. and Tanino, T.},
+  title = {Theory of multiobjective optimization},
+  publisher = {Elsevier},
+  year = 1985
+}
+
+ +
+@incollection{SaxDeb2007nonlinear,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  author = { Saxena, Dhish Kumar  and  Kalyanmoy Deb },
+  title = {Non-linear Dimensionality Reduction Procedures for Certain
+                  Large-Dimensional Multi-objective Optimization Problems:
+                  Employing Correntropy and a Novel Maximum Variance Unfolding},
+  doi = {10.1007/978-3-540-70928-2_58},
+  pages = {772--787},
+  abstract = {In our recent publication, we began with an understanding
+                  that many real-world applications of multi-objective
+                  optimization involve a large number (10 or more) of
+                  objectives but then, existing evolutionary multi-objective
+                  optimization (EMO) methods have primarily been applied to
+                  problems having smaller number of objectives (5 or
+                  less). After highlighting the major impediments in handling
+                  large number of objectives, we proposed a principal component
+                  analysis (PCA) based EMO procedure, for dimensionality
+                  reduction, whose efficacy was demonstrated by solving upto
+                  50-objective optimization problems. Here, we are addressing
+                  the fact that, when the data points live on a non-linear
+                  manifold or that the data structure is non-gaussian, PCA
+                  which yields a smaller dimensional 'linear' subspace may be
+                  ineffective in revealing the underlying dimensionality. To
+                  overcome this, we propose two new non-linear dimensionality
+                  reduction algorithms for evolutionary multi-objective
+                  optimization, namely C-PCA-NSGA-II and MVU-PCA-NSGA-II. While
+                  the former is based on the newly introduced correntropy PCA
+                  [2], the later implements maximum variance unfolding
+                  principle [3,4,5] in a novel way. We also establish the
+                  superiority of these new EMO procedures over the earlier
+                  PCA-based procedure, both in terms of accuracy and
+                  computational time, by solving upto 50-objective optimization
+                  problems.}
+}
+
+ +
+@inproceedings{SaxDeb2007trading,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2007,
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  key = {IEEE CEC},
+  author = { Saxena, Dhish Kumar  and  Kalyanmoy Deb },
+  title = {Trading on infeasibility by exploiting constraint's
+                  criticality through multi-ob\-jec\-ti\-vi\-za\-tion: A system
+                  design perspective},
+  pages = {919--926},
+  doi = {10.1109/CEC.2007.4424568},
+  keywords = {multi-objectivization}
+}
+
+ +
+@inproceedings{SaxDeb2008dimensionality,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2008,
+  booktitle = {Proceedings of  the 2008 Congress on Evolutionary Computation (CEC 2008)},
+  key = {IEEE CEC},
+  author = { Saxena, Dhish Kumar  and  Kalyanmoy Deb },
+  title = {Dimensionality reduction of objectives and constraints in
+                  multi-objective optimization problems: A system design
+                  perspective},
+  pages = {3204--3211},
+  doi = {10.1109/CEC.2008.4631232}
+}
+
+ +
+@inproceedings{Sch1985,
+  isbn = {0-8058-0426-9},
+  year = 1985,
+  publisher = {Lawrence Erlbaum Associates},
+  booktitle = {Proceedings of  the First International Conference on Genetic Algorithms (ICGA'85)},
+  editor = {John J. Grefenstette},
+  author = { J. David Schaffer },
+  title = {Multiple Objective Optimization with Vector
+                  Evaluated Genetic Algorithms},
+  pages = {93--100},
+  keywords = {VEGA}
+}
+
+ +
+@incollection{Sch1996exploiting,
+  year = 1996,
+  publisher = {MIT Press},
+  editor = {Michael Mozer and Michael I. Jordan and Thomas Petsche},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 9)},
+  title = {Exploiting model uncertainty estimates for safe dynamic
+                  control learning},
+  author = { Schneider, Jeff G. },
+  pages = {1047--1053},
+  epub = {http://papers.nips.cc/paper/1317-exploiting-model-uncertainty-estimates-for-safe-dynamic-control-learning}
+}
+
+ +
+@incollection{SchEsqLarCoe2010hausdorff,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  author = { Oliver Sch{\"u}tze  and X. Esquivel and A. Lara and  Carlos A. {Coello Coello} },
+  title = {Some Comments on {GD} and {IGD} and Relations to the {Hausdorff} Distance},
+  pages = {1971--1974}
+}
+
+ +
+@book{SchHer2021archiving,
+  title = {Archiving Strategies for Evolutionary Multi-objective
+                  Optimization Algorithms},
+  author = { Oliver Sch{\"u}tze  and  Carlos Hern{\'a}ndez },
+  publisher = {Springer},
+  year = 2021
+}
+
+ +
+@incollection{SchHoo2012quanti,
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7219,
+  publisher = {Springer},
+  year = 2012,
+  editor = { Youssef Hamadi  and  Marc Schoenauer },
+  booktitle = {Learning and Intelligent Optimization, 6th International
+                  Conference, LION 6},
+  author = { Marius Schneider  and  Holger H. Hoos },
+  title = {Quantifying Homogeneity of Instance Sets for Algorithm
+                  Configuration},
+  pages = {190--204},
+  keywords = {Quantifying Homogeneity; Empirical Analysis; Parameter
+                  Optimization; Algorithm Configuration},
+  doi = {10.1007/978-3-642-34413-8_14}
+}
+
+ +
+@inproceedings{SchKalPhi2015facenet,
+  title = {Facenet: A unified embedding for face recognition and
+                  clustering},
+  author = {Schroff, Florian and Kalenichenko, Dmitry and Philbin, James},
+  booktitle = {Proceedings of the IEEE Conference on Computer Vision and
+                  Pattern Recognition},
+  pages = {815--823},
+  year = 2015
+}
+
+ +
+@incollection{SchLip2011age,
+  title = {Age-Fitness {Pareto} Optimization},
+  author = {Schmidt, Michael and Lipson, Hod},
+  booktitle = {Genetic Programming Theory and Practice VIII. Genetic and
+                  Evolutionary Computation},
+  pages = {129--146},
+  publisher = {Springer},
+  year = 2011,
+  doi = {10.1007/978-1-4419-7747-2_8}
+}
+
+ +
+@inproceedings{SchNguEbe2015sal,
+  year = 2015,
+  publisher = {Springer},
+  volume = 9286,
+  series = {Lecture Notes in Computer Science},
+  fulleditor = {Albert Bifet and Michael May and Bianca Zadrozny and Ricard
+                  Gavald{\`{a}} and Dino Pedreschi and Francesco Bonchi and
+                  Jaime S. Cardoso and Myra Spiliopoulou},
+  booktitle = {Machine Learning and Knowledge Discovery in Databases, ECML
+                  PKDD 2015},
+  key = {ECML PKDD},
+  title = {Safe Exploration for Active Learning with {Gaussian}
+                  Processes},
+  author = {Schreiter, Jens and Nguyen-Tuong, Duy and Eberts, Mona and
+                  Bischoff, Bastian and Markert, Heiner and Toussaint, Marc},
+  pages = {133--149},
+  doi = {10.1007/978-3-319-23461-8_9},
+  annote = {Proposed Safe Active Learning (SAL) algorithm}
+}
+
+ +
+@inproceedings{SchOrtHart2018safe,
+  title = {Safe active learning of a high pressure fuel supply system},
+  author = {Schillinger, Mark and Ortelt, Benedikt and Hartmann, Benjamin
+                  and Schreiter, Jens and Meister, Mona and Nguyen-Tuong, Duy
+                  and Nelles, Oliver},
+  booktitle = {Proceedings of the 9th {EUROSIM} Congress on Modelling and
+                  Simulation, {EUROSIM} 2016 and the 57th {SIMS} Conference on
+                  Simulation and Modelling {SIMS} 2016},
+  numpages = 7,
+  pages = {286--292},
+  year = 2018,
+  doi = {10.3384/ecp17142286},
+  organization = {Link{\"o}ping University Electronic Press}
+}
+
+ +
+@incollection{Schaerf97,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = {Martha E. Pollack},
+  year = 1997,
+  booktitle = {Proceedings of  the 15th International Joint Conference on Artificial Intelligence (IJCAI-97)},
+  author = {Andrea Schaerf},
+  title = {Combining Local Search and Look-Ahead for Scheduling
+                  and Constraint Satisfaction Problems},
+  pages = {1254--1259},
+  volume = 2
+}
+
+ +
+@book{Scheffe1959anova,
+  title = {The Analysis of Variance},
+  author = {Scheffe, Henry},
+  publisher = {John Wiley \& Sons},
+  edition = {1st},
+  address = { New York, NY},
+  year = 1959
+}
+
+ +
+@book{Schwefel1977,
+  author = { Hans-Paul Schwefel },
+  title = {Numerische {Optimierung} von {Computer}--{Modellen}
+			mittels der {Evolutionsstrategie}},
+  publisher = {Birkh{\"a}user, Basel, Switzerland},
+  year = 1977
+}
+
+ +
+@inproceedings{ScoMat1999,
+  title = {Feature engineering for text classification},
+  author = {Scott, Sam and Matwin, Stan},
+  booktitle = {ICML},
+  volume = 99,
+  pages = {379--388},
+  year = 1999
+}
+
+ +
+@inproceedings{ScuSnoRahWil2018,
+  publisher = {OpenReview.net},
+  year = 2018,
+  editor = {Murray, Iain and Ranzato, Marc'{A}urelio and Vinyals, Oriol},
+  booktitle = {6th International Conference on Learning Representations, {ICLR} 2018,
+                  Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings},
+  title = {Winner's Curse? On Pace, Progress and Empirical Rigor},
+  author = {Sculley, D. and  Jasper Snoek  and Rahimi, Ali and Wiltschko, Alexander B.},
+  pages = {1--4},
+  url = {https://openreview.net/pdf?id=rJWF0Fywf}
+}
+
+ +
+@incollection{SeaDeb2015unsga3,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9018,
+  year = 2015,
+  publisher = {Springer},
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  author = {Seada, Haitham and  Kalyanmoy Deb },
+  title = {{U-NSGA-III}: A Unified Evolutionary Optimization Procedure
+                  for Single, Multiple, and Many Objectives: Proof-of-Principle
+                  Results},
+  pages = {34--49}
+}
+
+ +
+@incollection{SeiPohBosKerTra2020,
+  volume = 12269,
+  year = 2020,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas B{\"a}ck  and  Mike Preuss  and Deutz, Andr{\'e} and Wang, Hao and  Carola Doerr  and  Emmerich, Michael T. M.  and  Heike Trautmann },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVI}},
+  author = {Seiler, Moritz and Pohl, Janina and  Jakob Bossek  and  Pascal Kerschke  and  Heike Trautmann },
+  title = {Deep Learning as a Competitive Feature-Free Approach
+                  for Automated Algorithm Selection on the Traveling Salesperson Problem},
+  pages = {48--64}
+}
+
+ +
+@inproceedings{SeiSieHelHut2015cedalion,
+  year = 2015,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Blai Bonet and Sven Koenig},
+  title = {Automatic Configuration of Sequential Planning Portfolios},
+  author = {Seipp, Jendrik and Sievers, Silvan and Helmert, Malte and  Frank Hutter },
+  pages = {3364--3370}
+}
+
+ +
+@incollection{Ser1992,
+  publisher = {Springer Verlag},
+  editor = {G. H. Tzeng and P. L. Yu},
+  year = 1992,
+  booktitle = {Proceedings of the 10th International Conference on
+                  Multiple Criteria Decision Making (MCDM'91)},
+  title = {Simulated annealing for multiple objective
+                  optimization problems},
+  author = {Serafini, P.},
+  volume = 1,
+  pages = {87--96}
+}
+
+ +
+@incollection{Serafini86,
+  author = {P. Serafini},
+  title = {Some Considerations About Computational Complexity for
+                  Multiobjective Combinatorial Problems},
+  booktitle = {Recent Advances and Historical Development of Vector
+                  Optimization},
+  editor = {J. Jahn and W. Krabs},
+  pages = {222--231},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  volume = 294,
+  year = 1986,
+  series = {Lecture Notes in Economics and Mathematical Systems}
+}
+
+ +
+@inproceedings{ShaFonNor1999cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1999,
+  booktitle = {Proceedings of the 1999 Congress on Evolutionary Computation
+                  (CEC 1999)},
+  key = {IEEE CEC},
+  author = {K. J. Shaw and  Carlos M. Fonseca  and A. L. Nortcliffe and
+                  M. Thompson and J. Love and  Peter J. Fleming },
+  title = {Assessing the performance of multiobjective genetic
+                  algorithms for optimization of a batch process scheduling
+                  problem},
+  alias = {Shaw99},
+  pages = {34--75},
+  volume = 1
+}
+
+ +
+@incollection{ShaIshChe2021greedy,
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  author = {Shang, Ke and  Ishibuchi, Hisao  and Chen, Weiyu},
+  title = {Greedy approximated hypervolume subset selection for
+                  many-objective optimization},
+  year = 2021,
+  pages = {448--456},
+  doi = {10.1145/3449639.3459390}
+}
+
+ +
+@incollection{ShaIshNan2021subset,
+  doi = {10.1145/3449639.3459373},
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  title = {Distance-based subset selection revisited},
+  author = {Shang, Ke and  Ishibuchi, Hisao  and Nan, Yang},
+  pages = {439--447}
+}
+
+ +
+@incollection{ShaKomLopKaz2019gecco,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = { Mudita Sharma  and Alexandros Komninos  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Dimitar Kazakov },
+  title = {Deep Reinforcement Learning-Based Parameter Control in
+                  Differential Evolution},
+  pages = {709--717},
+  supplement = {https://dx.doi.org/10.5281/zenodo.2628228},
+  doi = {10.1145/3321707.3321813},
+  keywords = {DE-DDQN}
+}
+
+ +
+@incollection{ShaLopKaz2018ppsn,
+  volume = 11102,
+  year = 2018,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  author = { Mudita Sharma  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Dimitar Kazakov },
+  title = {Performance Assessment of Recursive Probability Matching for
+                  Adaptive Operator Selection in Differential Evolution},
+  supplement = {https://github.com/mudita11/AOS-comparisons},
+  doi = {10.1007/978-3-319-99259-4_26},
+  pages = {321--333},
+  keywords = {Rec-PM}
+}
+
+ +
+@misc{ShaLopKaz2018ppsn-supp,
+  author = { Mudita Sharma  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Dimitar Kazakov },
+  title = {Performance Assessment of Recursive Probability Matching for
+                  Adaptive Operator Selection in Differential Evolution:
+                  Supplementary material},
+  howpublished = {\url{https://github.com/mudita11/AOS-comparisons}},
+  doi = {10.5281/zenodo.1257672},
+  year = 2018
+}
+
+ +
+@incollection{ShaLopKno2021gecco,
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  author = { Shavarani, Seyed Mahdi  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles },
+  title = {Realistic Utility Functions Prove Difficult for
+                  State-of-the-Art Interactive Multiobjective Optimization
+                  Algorithms},
+  pages = {457--465},
+  doi = {10.1145/3449639.3459373}
+}
+
+ +
+@misc{ShaLopKno2023bench-supp,
+  author = { Shavarani, Seyed Mahdi  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Joshua D. Knowles },
+  title = {On Benchmarking Interactive Evolutionary Multi-Objective
+                  Algorithms: Supplementary material},
+  howpublished = {\url{https://doi.org/10.5281/zenodo.7863301}},
+  year = 2023
+}
+
+ +
+@phdthesis{Shavazipour2018PhD,
+  title = {Multi-Objective Optimisation under Deep Uncertainty},
+  author = { Shavazipour, Babooshka },
+  year = 2018,
+  school = {UCT Statistical sciences},
+  address = {South Africa},
+  epub = {https://open.uct.ac.za/bitstream/handle/11427/28122/thesis_sci_2018_shavazipour_babooshka.pdf?sequence=1}
+}
+
+ +
+@incollection{Shaw1998:lns,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Maher, Michael and Puget, Jean-Francois},
+  series = {Lecture Notes in Computer Science},
+  volume = 1520,
+  booktitle = {Principles and Practice of Constraint Programming, CP98},
+  year = 1998,
+  title = {Using Constraint Programming and Local Search Methods to
+                  Solve Vehicle Routing Problems},
+  author = {Shaw, Paul},
+  pages = {417--431}
+}
+
+ +
+@book{She00Handbook,
+  author = { David J. Sheskin },
+  title = {Handbook of Parametric and Nonparametric Statistical
+                  Procedures},
+  publisher = {Chapman \& Hall/CRC},
+  year = 2000,
+  edition = {2nd}
+}
+
+ +
+@book{Sheskin2011,
+  author = { David J. Sheskin },
+  title = {Handbook of Parametric and Nonparametric Statistical
+                  Procedures},
+  publisher = {Chapman \& Hall/CRC},
+  year = 2011,
+  edition = {5th}
+}
+
+ +
+@inproceedings{ShiEbe1998,
+  doi = {10.1007/BFb0040753},
+  year = 1998,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 1447,
+  series = {Lecture Notes in Computer Science},
+  editor = {V. W. Porto and N. Saravanan and D. Waagen and  Agoston E. Eiben },
+  booktitle = {Evolutionary Programming VII},
+  author = { Shi, Yuhui  and  Eberhart, Russell C. },
+  title = {Parameter selection in particle swarm optimization},
+  pages = {591--600}
+}
+
+ +
+@book{Shipley2000,
+  author = {B. Shipley},
+  title = {Cause and Correlation in Biology: a User's Guide to Path
+                  Analysis, Structural Equations and Causal Inference},
+  publisher = {Cambridge University Press},
+  year = 2000,
+  edition = {1st}
+}
+
+ +
+@incollection{ShmAguHoo2002:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = {A. Shmygelska and R. Aguirre-Hern{\'a}ndez and  Holger H. Hoos },
+  title = {An Ant Colony Optimization Algorithm for the {2D HP} Protein
+                  Folding Problem},
+  pages = {40--52}
+}
+
+ +
+@book{Sid1982optimal,
+  title = {Optimal Engineering Design: Principles and Applications},
+  author = {Siddall, James N.},
+  year = 1982,
+  publisher = {Marcel Dekker Inc.},
+  address = { New York, NY}
+}
+
+ +
+@book{SieCas1988,
+  author = { Sydney Siegel  and  Castellan, Jr, N. John },
+  title = {Non Parametric Statistics for the Behavioral Sciences},
+  publisher = {McGraw Hill},
+  year = 1988,
+  edition = {2nd},
+  address = { New York, NY}
+}
+
+ +
+@incollection{SilCalFraBer2021,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021},
+  address = { New York, NY},
+  year = 2021,
+  publisher = {ACM Press},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  author = { Silva-Mu\~noz, Mois\'es  and Calderon, Gonzalo and  Alberto Franzin  and  Hughes Bersini },
+  title = {Determining a consistent experimental setup for benchmarking
+                  and optimizing databases},
+  pages = {1614--1621},
+  doi = {10.1145/3449726.3463180}
+}
+
+ +
+@incollection{SilRunSouPal2002:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = {C. A. Silva and T. A. Runkler and J. M. Sousa and R. Palm},
+  title = {Ant Colonies as Logistic Processes Optimizers},
+  pages = {76--87}
+}
+
+ +
+@incollection{Simpson99,
+  address = {Baldock, United Kingdom},
+  publisher = { Research Studies Press Ltd. },
+  volume = 2,
+  year = 1999,
+  booktitle = {Water Industry Systems: Modelling and Optimization
+                  Applications},
+  editor = { Dragan A. Savic  and  Godfrey A. Walters },
+  author = { Angus R. Simpson  and  D. C. Sutton  and  D. S. Keane  and  S. J. Sherriff },
+  title = {Optimal control of pumping at a water filtration
+                  plant using genetic algorithms}
+}
+
+ +
+@misc{Slo11emo,
+  title = {Inducing preference models from pairwise
+                  comparisons: implications for preference-guided
+                  {EMO}},
+  year = 2011,
+  author = { Roman S{\l}owi{\'n}ski },
+  howpublished = {Evolutionary Multi-Criterion Optimization, EMO 2011},
+  note = {Keynote talk}
+}
+
+ +
+@inproceedings{SmiEib2009cec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2009,
+  booktitle = {Proceedings of  the 2009 Congress on Evolutionary Computation (CEC 2009)},
+  key = {IEEE CEC},
+  author = { Smit, Selmar K.  and  Agoston E. Eiben },
+  title = {Comparing Parameter Tuning Methods for Evolutionary
+                  Algorithms},
+  pages = {399--406}
+}
+
+ +
+@inproceedings{SmiEib2010cec,
+  year = 2010,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2010 Congress on Evolutionary Computation (CEC 2010)},
+  editor = { Ishibuchi, Hisao  and others},
+  key = {IEEE CEC},
+  author = { Smit, Selmar K.  and  Agoston E. Eiben },
+  title = {Beating the 'world champion' evolutionary algorithm
+                  via {REVAC} tuning},
+  pages = {1--8},
+  doi = {10.1109/CEC.2010.5586026}
+}
+
+ +
+@incollection{SmiEib2010evoapp,
+  year = 2010,
+  volume = 6024,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Applications of Evolutionary Computation},
+  editor = {Cecilia Di Chio and Stefano Cagnoni and  Carlos Cotta  and Marc
+                  Ebner and Anik{\'o} Ek{\'a}rt and  Anna I. Esparcia{-}Alc{\'{a}}zar  and Chi Keong Goh and  Juan-Juli{\'a}n Merelo  and Ferrante Neri and  Mike Preuss  and Julian Togelius and Georgios N. Yannakakis},
+  author = { Smit, Selmar K.  and  Agoston E. Eiben },
+  title = {Parameter Tuning of Evolutionary Algorithms: Generalist
+               vs. Specialist},
+  pages = {542--551},
+  doi = {10.1007/978-3-642-12239-2_56}
+}
+
+ +
+@incollection{SmiEib2011ae,
+  booktitle = {Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7401,
+  year = 2012,
+  publisher = {Springer},
+  editor = { Jin-Kao Hao  and Legrand, Pierrick and Collet, Pierre and
+                  Monmarch{\'e}, Nicolas and Lutton, Evelyne and Schoenauer,
+                  Marc},
+  author = { Smit, Selmar K.  and  Agoston E. Eiben },
+  title = {Multi-Problem Parameter Tuning using {BONESA}},
+  pages = {222--233},
+  annote = {For some reason, this was not actually published in the LNCS
+                  Proceedings of EA}
+}
+
+ +
+@inproceedings{SmiEibSzl2010ijcci,
+  year = 2010,
+  publisher = {SciTePress},
+  booktitle = {Proceedings of the International Joint Conference on
+                  Computational Intelligence (IJCCI-2010)},
+  editor = { Filipe, J.  and  J. Kacprzyk },
+  author = { Smit, Selmar K.  and  Agoston E. Eiben  and  Szl\'{a}vik, Z. },
+  title = {An {MOEA}-based Method to Tune {EA} Parameters on
+                  Multiple Objective Functions},
+  pages = {261--268}
+}
+
+ +
+@inproceedings{SmiSet92,
+  author = {Tobiah E. Smith and Dorothy E. Setliff},
+  booktitle = {Proceedings of the Seventh Knowledge-Based Software Engineering Conference},
+  title = {Knowledge-based constraint-driven software synthesis},
+  year = 1992,
+  pages = {18--27},
+  publisher = {IEEE},
+  doi = {10.1109/KBSE.1992.252912}
+}
+
+ +
+@incollection{SmiStoSer2016explo,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2016,
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2016},
+  doi = {10.1145/2908812.2908854},
+  author = {Jim Smith and Christopher Stone and Martin Serpell},
+  title = {Exploiting Diverse Distance Metrics for Surrogate-Based
+                  Optimisation of Ordering Problems},
+  pages = {701--708}
+}
+
+ +
+@incollection{SmiVanLim2010,
+  doi = {10.1007/978-3-642-13800-3},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Christian Blum  and  Roberto Battiti },
+  series = {Lecture Notes in Computer Science},
+  volume = 6073,
+  year = 2010,
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  author = { Kate Smith{-}Miles  and  van Hemert, Jano I.  and Lim, Xin Yu},
+  title = {Understanding {TSP} difficulty by Learning from evolved instances},
+  pages = {266--280}
+}
+
+ +
+@inproceedings{Smith-Miles2008ijcnn,
+  publisher = {IEEE Press},
+  year = 2008,
+  editor = {Liu, Derong and others},
+  booktitle = {Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008),
+                  Hong Kong, China, June 1-6, 2008},
+  key = {IJCNN},
+  author = { Kate Smith{-}Miles },
+  title = {Towards insightful algorithm selection for optimisation using meta-learning concepts},
+  pages = {4118--4124}
+}
+
+ +
+@book{SneCoc1967stat,
+  title = {Statistical Methods},
+  author = {Snedecor, George W. and Cochran, William G.},
+  edition = {6th},
+  year = 1967,
+  publisher = {Iowa State University Press},
+  address = {Ames, IA, USA}
+}
+
+ +
+@incollection{SnoLarAda2012nips,
+  publisher = {Curran Associates, Red Hook, NY},
+  year = 2012,
+  editor = {Peter L. Bartlett and Fernando C. N. Pereira and Christopher
+                  J. C. Burges and L{\'{e}}on Bottou and Kilian Q. Weinberger},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 25)},
+  author = { Jasper Snoek  and Hugo Larochelle and  Ryan P. Adams },
+  title = {Practical {Bayesian} Optimization of Machine Learning Algorithms},
+  pages = {2960--2968}
+}
+
+ +
+@inproceedings{SnoSweZemAda2014icml,
+  url = {http://jmlr.org/proceedings/papers/v32/},
+  publisher = {{PMLR}},
+  year = 2014,
+  volume = 32,
+  booktitle = {Proceedings of  the 31st International Conference on Machine Learning, {ICML} 2014},
+  editor = {Xing, Eric P. and Jebara, Tony},
+  author = { Jasper Snoek  and  Kevin Swersky  and  Richard Zemel  and  Ryan P. Adams },
+  title = {Input Warping for {Bayesian} Optimization of Non-Stationary Functions},
+  pages = {1674--1682}
+}
+
+ +
+@incollection{SocKnoSam02:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  volume = 2463,
+  series = {Lecture Notes in Computer Science},
+  year = 2002,
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  author = { Krzysztof Socha  and  Joshua D. Knowles  and  M. Sampels },
+  title = {A {\MaxMinAntSystem} for the University Course Timetabling Problem},
+  pages = {1--13},
+  alias = {antsystimtabl:ants2002}
+}
+
+ +
+@incollection{SocSamMan03:evoworkshops,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 2611,
+  editor = {S. Cagnoni and others},
+  aeditor = {S. Cagnoni and J. J. {Romero Cardalda} and D. W. Corne
+                  and J. Gottlieb and A. Guillot and E. Hart and
+                  C. G. Johnson and E. Marchiori and J.-A. Meyer and
+                  M. Middendorf and G. R. Raidl},
+  year = 2003,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of EvoWorkshops 2003},
+  author = { Krzysztof Socha  and  M. Sampels  and M. Manfrin},
+  title = {Ant algorithms for the university course timetabling
+                  problem with regard to the state-of-the-art},
+  pages = {334--345}
+}
+
+ +
+@incollection{Socha04:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3172,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  year = 2004,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  author = { Krzysztof Socha },
+  title = {{ACO} for Continuous and Mixed-Variable Optimization},
+  pages = {25--36}
+}
+
+ +
+@book{Solnon2010,
+  author = { Christine Solnon },
+  title = {Ant Colony Optimization and Constraint Programming},
+  publisher = {Wiley},
+  year = 2010,
+  doi = {10.1002/9781118557563}
+}
+
+ +
+@incollection{SorSevGlo2017,
+  isbn = {978-3-319-07125-1},
+  publisher = {Springer International Publishing},
+  year = 2018,
+  booktitle = {Handbook of Heuristics},
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  title = {A history of metaheuristics},
+  author = { Kenneth S{\"o}rensen  and  Marc Sevaux  and  Fred Glover },
+  pages = {1--27}
+}
+
+ +
+@inproceedings{SotBasDol20016_ES,
+  alias = {Sotelo_ES},
+  year = 2001,
+  author = { Aldo Sotelo  and  Julio Basulado  and  Pedro Dold{\'a}n  and  Benjam{\'i}n Bar{\'a}n },
+  title = {Algoritmos Evolutivos Multiobjetivo Combinados para la
+                  Optimizaci{\'o}n de la Programaci{\'o}n de Bombeo en Sistemas
+                  de Suministro de Agua},
+  booktitle = {Congreso Internacional de Tecnolog{\'i}as y Aplicaciones
+                  Inform\'aticas, JIT-CITA 2001, Asunci\'on, Paraguay},
+  note = {(In Spanish)}
+}
+
+ +
+@inproceedings{Sotelo02,
+  author = { Aldo Sotelo  and  C. von L{\"u}cken  and  Benjam{\'i}n Bar{\'a}n },
+  title = {Multiobjective Evolutionary Algorithms in Pump
+                  Scheduling Optimisation},
+  booktitle = {Proceedings of the Third International Conference on
+                  Engineering Computational Technology},
+  publisher = {Civil-Comp Press, Stirling, Scotland},
+  year = 2002,
+  editor = { Barry H. V. Topping  and  Zden{\'e}ek Bittnar },
+  abstract = {Operation of pumping stations represents high costs
+                  to water supply companies. Therefore, reducing such
+                  costs through an optimal pump scheduling becomes an
+                  important issue. This work presents the use of
+                  Multiobjective Evolutionary Algorithms (MOEAs) to
+                  solve an optimal pump-scheduling problem. For the
+                  first time, six different approaches were
+                  implemented and compared. These algorithms aim to
+                  minimise four objectives: electric energy cost,
+                  pumps' maintenance cost, maximum power peak, and
+                  level variation in the reservoir. In order to
+                  consider hydraulic and technical constrains, a
+                  heuristic constrain algorithm was developed and
+                  combined with each MOEA utilised. Evaluation of
+                  experimental results of a set of metrics shows that
+                  the Strength {Pareto} Evolutionary Algorithm (SPEA)
+                  achieves the best performance for this
+                  problem. Moreover, SPEA's set of solutions provide
+                  pumping station operation engineers with a wide
+                  range of optimal pump schedules to chose from.}
+}
+
+ +
+@misc{SouRitLop2020capopt,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {{CAPOPT}: Capping Methods for the Automatic Configuration of Optimization Algorithms},
+  howpublished = {\url{https://github.com/souzamarcelo/capopt}},
+  year = 2020
+}
+
+ +
+@misc{SouRitLop2021cap-supp,
+  author = { Marcelo {De Souza}  and  Marcus Ritt and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Capping Methods for the Automatic Configuration of
+                  Optimization Algorithms -- Supplementary Material},
+  howpublished = {\url{https://github.com/souzamarcelo/supp-cor-capopt}},
+  year = 2021
+}
+
+ +
+@misc{Spark,
+  author = {{Apache Software Foundation}},
+  title = {Spark},
+  url = {https://spark.apache.org},
+  year = 2012
+}
+
+ +
+@book{SraNowWri2012,
+  title = {Optimization for machine learning},
+  author = {Sra, Suvrit and Nowozin, Sebastian and Wright, Stephen J.},
+  year = 2012,
+  publisher = {MIT Press},
+  address = {Cambridge, MA}
+}
+
+ +
+@incollection{Stadler1995landscapes,
+  author = {P. F. Stadler},
+  title = {Toward a theory of landscapes},
+  booktitle = {Complex Systems and Binary Networks},
+  year = 1995,
+  editor = {R. L{\'o}pez-Pe{\~n}a and R. Capovilla and
+                  R. Garc{\'i}a-Pelayo and H. Waelbroeck and F. Zertruche},
+  pages = {77--163},
+  publisher = {Springer}
+}
+
+ +
+@book{Starr1963,
+  title = {Product design and decision theory},
+  author = {Starr, Martin Kenneth},
+  year = 1963,
+  series = {Prentice-Hall Series in Engineering Design, Fundamentals of
+                  Engineering Design},
+  publisher = {Prentice-Hall},
+  address = {Englewood, Cliffs, NJ}
+}
+
+ +
+@incollection{SteAggBurGonRes2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = {Fernando Stefanello and Vaneet Aggarwal and Luciana Salete Buriol and Jos{\'e} Fernando Gon\c{c}alves and  Mauricio G. C. Resende },
+  title = {A Biased Random-key Genetic Algorithm for Placement of Virtual Machines Across Geo-Separated Data Centers},
+  pages = {919--926},
+  doi = {10.1145/2739480.2754768},
+  keywords = {irace}
+}
+
+ +
+@incollection{SteGar1991computational,
+  author = { R. E. Steuer  and Gardiner, Lorraine},
+  editor = {Fandel, G{\"u}nter and Gehring, Hermann},
+  title = {On the Computational Testing of Procedures for Interactive
+                  Multiple Objective Linear Programming},
+  booktitle = {Operations Research},
+  year = 1991,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  pages = {121--131},
+  isbn = {978-3-642-76537-7},
+  doi = {10.1007/978-3-642-76537-7_8},
+  annote = {Proposed difference between ad hoc and non-ad hoc interactive
+                  multi-objective optimization methods}
+}
+
+ +
+@incollection{SteSmiJan2019autostat,
+  booktitle = {Automated Machine Learning},
+  year = 2019,
+  publisher = {Springer},
+  epub = {http://automl.org/book},
+  editor = { Frank Hutter  and Kotthoff, Lars and  Joaquin Vanschoren },
+  title = {The Automatic Statistician},
+  author = {Christian Steinruecken and Emma Smith and David Janz and
+                  James Lloyd and Zoubin Ghahramani},
+  doi = {10.1007/978-3-030-05318-5_9},
+  pages = {161--173}
+}
+
+ +
+@book{Steuer1986,
+  author = { R. E. Steuer },
+  title = {Multiple Criteria Optimization: Theory, Computation and
+                  Application},
+  publisher = {John Wiley \& Sons},
+  year = 1986,
+  series = {Wiley Series in Probability and Mathematical Statistics},
+  address = { New York, NY},
+  keywords = {Maximally dispersed weights}
+}
+
+ +
+@incollection{Stolfi2015,
+  volume = 9422,
+  series = {Lecture Notes in Computer Science},
+  editor = {Puerta, Jos{\'e} M.  and G{\'a}mez, Jos{\'e} A.  and
+                  Dorronsoro, Bernabe and Barrenechea, Edurne and Troncoso,
+                  Alicia and Baruque, Bruno and Galar, Mikel},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2015,
+  booktitle = {Advances in Artificial Intelligence, CAEPIA 2015},
+  abstract = {In this article we present a strategy based on an evolution-
+                  ary algorithm to calculate the real vehicle flows in cities
+                  according to data from sensors placed in the streets. We have
+                  worked with a map imported from OpenStreetMap into the SUMO
+                  traffic simulator so that the resulting scenarios can be used
+                  to perform different optimizations with the confidence of
+                  being able to work with a traffic distribution close to
+                  reality. We have compared the results of our algorithm to
+                  other competitors and achieved results that replicate the
+                  real traffic distribution with a precision higher than
+                  {90\%}.},
+  author = {Stolfi, Daniel H. and  Alba, Enrique },
+  title = {An Evolutionary Algorithm to Generate Real Urban Traffic
+                  Flows},
+  pages = {332--343},
+  doi = {10.1007/978-3-319-24598-0_30},
+  keywords = {Evolutionary algorithm,SUMO,Smart city,Smart mobility,Traffic
+                  simulation}
+}
+
+ +
+@techreport{Stu1998:pfsp,
+  author = { Thomas St{\"u}tzle },
+  title = {Applying Iterated Local Search to the Permutation Flow Shop Problem},
+  institution = {FG Intellektik, FB Informatik, TU Darmstadt, Germany},
+  year = 1998,
+  number = {AIDA--98--04},
+  month = aug
+}
+
+ +
+@misc{Stu2002,
+  author = { Thomas St{\"u}tzle },
+  title = {{\softwarepackage{ACOTSP}}: A Software Package of Various Ant Colony
+                  Optimization Algorithms Applied to the Symmetric
+                  Traveling Salesman Problem},
+  url = {http://www.aco-metaheuristic.org/aco-code},
+  annote = {\url{http://www.aco-metaheuristic.org/aco-code}},
+  year = 2002
+}
+
+ +
+@inproceedings{Stu2009:eume,
+  editor = {Ana Viana and others},
+  year = 2009,
+  booktitle = {Proceedings of the EU/MEeting 2009: Debating the future: new
+                  areas of application and innovative approaches},
+  author = { Thomas St{\"u}tzle },
+  title = {Some Thoughts on Engineering Stochastic Local Search
+                  Algorithms},
+  pages = {47--52}
+}
+
+ +
+@techreport{Stu97:qap,
+  author = { Thomas St{\"u}tzle },
+  title = {{\MaxMinAntSystem} for the Quadratic Assignment
+                  Problem},
+  institution = {FG Intellektik, FB Informatik, TU Darmstadt, Germany},
+  year = 1997,
+  number = {AIDA--97--4},
+  month = jul
+}
+
+ +
+@inproceedings{Stu98:eufit,
+  author = { Thomas St{\"u}tzle },
+  title = {An Ant Approach to the Flow Shop Problem},
+  booktitle = {Proceedings of the 6th European Congress on
+                  Intelligent Techniques {\&} Soft Computing
+                  (EUFIT'98)},
+  pages = {1560--1564},
+  year = 1998,
+  volume = 3,
+  publisher = {Verlag Mainz, Aachen, Germany}
+}
+
+ +
+@incollection{StuDor99:nio,
+  alias = {CorDorGlo99},
+  address = {London, UK},
+  year = 1999,
+  publisher = {McGraw Hill},
+  editor = { David Corne  and  Marco Dorigo  and  Fred Glover },
+  booktitle = {New Ideas in Optimization},
+  author = { Thomas St{\"u}tzle  and  Marco Dorigo },
+  title = {{ACO} Algorithms for the Quadratic Assignment Problem},
+  pages = {33--50},
+  anote = {Also available as Technical Report IRIDIA/99-2,
+                  Universit{\'e} Libre de Bruxelles, Belgium}
+}
+
+ +
+@incollection{StuFer2004qap,
+  booktitle = {Proceedings of EvoCOP 2004 -- 4th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 3004,
+  year = 2004,
+  publisher = {Springer},
+  editor = {Gottlieb, Jens and  G{\"u}nther R. Raidl },
+  author = { Thomas St{\"u}tzle  and Fernandes, Susana},
+  title = {New Benchmark Instances for the QAP and the Experimental
+                  Analysis of Algorithms},
+  pages = {199--209},
+  doi = {10.1007/978-3-540-24652-7_20},
+  abstract = {The quadratic assignment problem arises in a variety of
+                  practical settings. It is known to be among the hardest
+                  combinatorial problems for exact algorithms. Therefore, a
+                  large number of heuristic approaches have been proposed for
+                  its solution. In this article we introduce a new, large set
+                  of QAP instances that is intended to allow the systematic
+                  study of the performance of metaheuristics in dependence of
+                  QAP instance characteristics. Additionally, we give
+                  computational results with several high performing algorithms
+                  known from literature and give exemplary results on the
+                  influence of instance characteristics on the performance of
+                  these algorithms.}
+}
+
+ +
+@incollection{StuHoo01:mic,
+  author = { Thomas St{\"u}tzle  and  Holger H. Hoos },
+  title = {Analysing the Run-time Behaviour of Iterated Local
+                  Search for the Travelling Salesman Problem},
+  booktitle = {Essays and Surveys on Metaheuristics},
+  pages = {589--611},
+  publisher = {Kluwer Academic Publishers, Boston, MA},
+  year = 2001,
+  editor = {P. Hansen and C. Ribeiro},
+  series = {Operations Research/Computer Science Interfaces
+                  Series}
+}
+
+ +
+@techreport{StuHoo1996:aida,
+  author = { Thomas St{\"u}tzle  and  Holger H. Hoos },
+  title = {Improving the {Ant} {System}: A Detailed Report on
+                  the {\MaxMinAntSystem}},
+  institution = {FG Intellektik, FB Informatik, TU Darmstadt, Germany},
+  year = 1996,
+  number = {AIDA--96--12},
+  month = aug
+}
+
+ +
+@incollection{StuHoo97:icec,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 1997,
+  editor = { Thomas B{\"a}ck  and  Zbigniew Michalewicz  and  Xin Yao },
+  booktitle = {Proceedings of the 1997 IEEE International
+                  Conference on Evolutionary Computation (ICEC'97)},
+  author = { Thomas St{\"u}tzle  and  Holger H. Hoos },
+  title = {The {\MaxMinAntSystem} and Local Search for the
+                  Traveling Salesman Problem},
+  pages = {309--314}
+}
+
+ +
+@incollection{StuHoo99:mic,
+  author = { Thomas St{\"u}tzle  and  Holger H. Hoos },
+  title = {{\MaxMinAntSystem} and Local Search for Combinatorial
+                  Optimization Problems},
+  booktitle = {Meta-Heuristics: Advances and Trends in Local Search
+                  Paradigms for Optimization},
+  publisher = {Kluwer Academic Publishers, Dordrecht, The Netherlands},
+  year = 1999,
+  editor = { Stefan Vo{\ss}  and  Silvano Martello  and  Ibrahim H. Osman  and C. Roucairol},
+  pages = {137--154}
+}
+
+ +
+@incollection{StuLop2015gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2015},
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Automatic (Offline) Configuration of Algorithms},
+  pages = {681--702},
+  doi = {10.1145/2739482.2756581}
+}
+
+ +
+@incollection{StuLop2017gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2017},
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Automated Offline Design of Algorithms},
+  pages = {1038--1065},
+  doi = {10.1145/3067695.3067722}
+}
+
+ +
+@incollection{StuLop2019hb,
+  publisher = {Springer},
+  series = {International Series in Operations Research \& Management
+                  Science},
+  volume = 272,
+  booktitle = {Handbook of Metaheuristics},
+  year = 2019,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Automated Design of Metaheuristic Algorithms},
+  pages = {541--579},
+  doi = {10.1007/978-3-319-91086-4_17},
+  keywords = {automatic design, automatic configuration}
+}
+
+ +
+@incollection{StuLopDor2011eorms,
+  year = 2011,
+  publisher = {John Wiley \& Sons},
+  editor = {J. J. Cochran},
+  booktitle = {Wiley Encyclopedia of Operations Research and
+                  Management Science},
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Marco Dorigo },
+  title = {A Concise Overview of Applications of Ant Colony
+                  Optimization},
+  pages = {896--911},
+  volume = 2,
+  doi = {10.1002/9780470400531.eorms0001}
+}
+
+ +
+@incollection{StuLopPel2011autsea,
+  year = 2012,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  booktitle = {Autonomous Search},
+  editor = { Youssef Hamadi  and E. Monfroy and F. Saubion},
+  author = { Thomas St{\"u}tzle  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Paola Pellegrini  and  Michael Maur  and  Marco A. {Montes de Oca}  and  Mauro Birattari  and  Marco Dorigo },
+  title = {Parameter Adaptation in Ant Colony Optimization},
+  doi = {10.1007/978-3-642-21434-9_8},
+  pages = {191--215}
+}
+
+ +
+@incollection{StuRui2017ig,
+  isbn = {978-3-319-07125-1},
+  publisher = {Springer International Publishing},
+  year = 2018,
+  booktitle = {Handbook of Heuristics},
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  author = { Thomas St{\"u}tzle  and  Rub{\'e}n Ruiz },
+  title = {Iterated Greedy},
+  doi = {10.1007/978-3-319-07153-4_10-1},
+  pages = {1--31}
+}
+
+ +
+@incollection{StuRui2017ils,
+  isbn = {978-3-319-07125-1},
+  publisher = {Springer International Publishing},
+  year = 2018,
+  booktitle = {Handbook of Heuristics},
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  author = { Thomas St{\"u}tzle  and  Rub{\'e}n Ruiz },
+  title = {Iterated Local Search},
+  doi = {10.1007/978-3-319-07153-4_8-1},
+  pages = {1--27}
+}
+
+ +
+@phdthesis{StuetzlePhD,
+  author = { Thomas St{\"u}tzle },
+  title = {Local Search Algorithms for Combinatorial Problems --- Analysis, Improvements, and New Applications},
+  school = {FB Informatik, TU Darmstadt, Germany},
+  year = 1998
+}
+
+ +
+@incollection{StyHoo2013gecco,
+  isbn = {978-1-4503-1963-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = { Christian Blum  and  Alba, Enrique },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  author = {James Styles and  Holger H. Hoos },
+  title = {Ordered racing protocols for automatically configuring algorithms
+               for scaling performance},
+  pages = {551--558},
+  doi = {10.1145/2463372.2463438}
+}
+
+ +
+@incollection{StyHooMul2012:lion,
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7219,
+  publisher = {Springer},
+  year = 2012,
+  editor = { Youssef Hamadi  and  Marc Schoenauer },
+  booktitle = {Learning and Intelligent Optimization, 6th International
+                  Conference, LION 6},
+  author = {James Styles and  Holger H. Hoos  and Martin M{\"{u}}ller},
+  title = {Automatically Configuring Algorithms for Scaling Performance},
+  pages = {205--219}
+}
+
+ +
+@techreport{SugHanLia2005cec,
+  author = { Ponnuthurai N. Suganthan  and  Nikolaus Hansen  and J. J. Liang and  Kalyanmoy Deb  and Y. P. Chen and  Anne Auger  and S. Tiwari},
+  title = {Problem definitions and evaluation criteria for the
+                  {CEC 2005} special session on real-parameter
+                  optimization},
+  institution = {Nanyang Technological University, Singapore},
+  year = 2005,
+  keywords = {CEC'05 benchmark},
+  annote = {Also known as KanGAL Report Number 2005005 (Kanpur Genetic Algorithms
+                  Laboratory, IIT Kanpur)}
+}
+
+ +
+@inproceedings{SuiGotBur2015icml,
+  publisher = {{PMLR}},
+  year = 2015,
+  volume = 37,
+  booktitle = {Proceedings of  the 32nd International Conference on Machine Learning, {ICML} 2015},
+  editor = {Francis Bach and David Blei},
+  author = {Sui, Yanan and Alkis Gotovos and Burdick, Joel W. and Andreas
+                  Krause},
+  title = {Safe Exploration for Optimization with {Gaussian} Processes},
+  pages = {997--1005},
+  epub = {http://proceedings.mlr.press/v37/sui15.html},
+  abstract = {We consider sequential decision problems under uncertainty,
+                  where we seek to optimize an unknown function from noisy
+                  samples. This requires balancing exploration (learning about
+                  the objective) and exploitation (localizing the maximum), a
+                  problem well-studied in the multi-armed bandit literature. In
+                  many applications, however, we require that the sampled
+                  function values exceed some prespecified "safety" threshold,
+                  a requirement that existing algorithms fail to meet. Examples
+                  include medical applications where patient comfort must be
+                  guaranteed, recommender systems aiming to avoid user
+                  dissatisfaction, and robotic control, where one seeks to
+                  avoid controls causing physical harm to the platform. We
+                  tackle this novel, yet rich, set of problems under the
+                  assumption that the unknown function satisfies regularity
+                  conditions expressed via a Gaussian process prior. We develop
+                  an efficient algorithm called SafeOpt, and theoretically
+                  guarantee its convergence to a natural notion of optimum
+                  reachable under safety constraints. We evaluate SafeOpt on
+                  synthetic data, as well as two real applications: movie
+                  recommendation, and therapeutic spinal cord stimulation.},
+  keywords = {Safe Optimization, SafeOpt}
+}
+
+ +
+@inproceedings{SuiZhuBur2018stageopt,
+  year = 2018,
+  publisher = {{PMLR}},
+  volume = 80,
+  series = {Proceedings of Machine Learning Research},
+  booktitle = {Proceedings of  the 35th International Conference on Machine Learning, {ICML} 2018},
+  editor = {Jennifer G. Dy and Andreas Krause},
+  author = {Sui, Yanan and Zhuang, Vincent and Burdick, Joel W. and Yue,
+                  Yisong},
+  title = {Stagewise Safe {Bayesian} Optimization with {Gaussian}
+                  Processes},
+  pages = {4788--4796},
+  epub = {http://proceedings.mlr.press/v80/sui18a.html},
+  keywords = {StageOpt}
+}
+
+ +
+@inproceedings{SunHan2010:ccie,
+  address = {Los Alamitos, CA},
+  publisher = {IEEE Computer Society Press},
+  year = 2010,
+  booktitle = {Proceedings of the 2010 International Conference on
+                  Computing, Control and Industrial Engineering},
+  key = {CCIE},
+  author = { Zhaoxu Sun  and  Min Han },
+  title = {Multi-criteria Decision Making Based on {PROMETHEE} Method},
+  pages = {416--418},
+  alias = {ZhaMin2010:ccie}
+}
+
+ +
+@book{SutBar1998reinf,
+  author = {Richard S. Sutton and Andrew G. Barto},
+  title = {Reinforcement Learning: An Introduction},
+  publisher = {MIT Press, Cambridge, MA},
+  year = 1998
+}
+
+ +
+@book{SutBar2018reinf,
+  author = {Richard S. Sutton and Andrew G. Barto},
+  title = {Reinforcement Learning: An Introduction},
+  publisher = {MIT Press, Cambridge, MA},
+  edition = {2nd},
+  year = 2018
+}
+
+ +
+@mastersthesis{Sutton98,
+  author = { D. C. Sutton  and  D. S. Keane  and  S. J. Sherriff },
+  title = {Optimizing the Real Time Operation of a Pumping
+                  Station at a Water Filtration Plant using Genetic
+                  Algorithms},
+  school = {Department of Civil and Environmental Engineering,
+                  The University of Adelaide},
+  year = 1998,
+  type = {Honors Thesis}
+}
+
+ +
+@incollection{SwaOzcKen2011lion,
+  volume = 6683,
+  editor = { Carlos A. {Coello Coello} },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  year = 2011,
+  booktitle = {Learning and Intelligent Optimization, 5th
+                  International Conference, LION 5},
+  author = { Jerry Swan  and  Ender {\"O}zcan  and  Graham Kendall },
+  title = {Hyperion: A Recursive Hyper-heuristic Framework},
+  pages = {616--630}
+}
+
+ +
+@inproceedings{Swan2015,
+  year = 2015,
+  booktitle = {Proceedings of MIC 2015, the 11th Metaheuristics
+                  International Conference},
+  editor = { Talbi, El-Ghazali },
+  author = { Jerry Swan  and others},
+  anauthor = { Jerry Swan  and Steven Adriaensen and Mohamed Bishr and  Edmund K. Burke  and John A. Clark and Patrick {De Causmaecker} and  Durillo, Juan J.  and Kevin Hammond and Emma Hart and Colin
+                  G. Johnson and Zoltan A. Kocsis and Ben Kovitz and  Krzysztof Krawiec  and Simon Martin and J. J. Merelo and Leandro
+                  L. Minku and  Ender {\"O}zcan  and  Gisele Pappa  and Erwin
+                  Pesch and  Pablo Garc{\'i}a-S{\'a}nchez  and Andrea Schaerf and Kevin
+                  Sim and Jim E. Smith and  Thomas St{\"u}tzle  and  Stefan Vo{\ss}  and Stefan Wagner and  Xin Yao },
+  title = {A Research Agenda for Metaheuristic Standardization}
+}
+
+ +
+@inproceedings{Syswerda89,
+  publisher = {Morgan Kaufmann Publishers, San Mateo, CA},
+  editor = { J. David Schaffer },
+  year = 1989,
+  booktitle = {Proceedings of  the Third International Conference on Genetic Algorithms (ICGA'89)},
+  author = { Gilbert Syswerda },
+  title = {Uniform Crossover in Genetic Algorithms},
+  pages = {2--9},
+  keywords = {uniform crossover}
+}
+
+ +
+@incollection{TagShiNak2011ibde,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = {Tagawa, Kiyoharu and Shimizu, Hidehito and Nakamura,
+                  Hiroyuki},
+  title = {Indicator-based Differential Evolution Using Exclusive
+                  Hypervolume Approximation and Parallelization for Multi-core
+                  Processors},
+  pages = {657--664}
+}
+
+ +
+@inproceedings{TaiYanRanWol2014deepface,
+  title = {Deepface: Closing the gap to human-level performance in face
+                  verification},
+  author = {Taigman, Yaniv and Yang, Ming and Ranzato, Marc'Aurelio and
+                  Wolf, Lior},
+  booktitle = {Proceedings of the IEEE conference on computer vision and
+                  pattern recognition},
+  pages = {1701--1708},
+  year = 2014
+}
+
+ +
+@incollection{TanOya2017gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2017,
+  editor = { Peter A. N. Bosman },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  title = {Benchmarking {MOEAs} for multi-and many-objective
+                  optimization using an unbounded external archive},
+  author = {Tanabe, Ryoji and Oyama, Akira},
+  pages = {633--640}
+}
+
+ +
+@inproceedings{TasBuyPanSug2013,
+  publisher = {Springer International Publishing},
+  series = {Theoretical Computer Science and General Issues},
+  volume = 8298,
+  year = 2013,
+  editor = {B. K. Panigrahi and P. N. Suganthan and S. Das and S. S. Dash},
+  booktitle = {Swarm, Evolutionary, and Memetic Computing},
+  author = {M. Fatih Tasgetiren and Ozge Buyukdagli and  Quan-Ke Pan  and  Ponnuthurai N. Suganthan },
+  title = {A general variable neighborhood search algorithm for the
+                  no-idle permutation flowshop scheduling problem},
+  pages = {24--34}
+}
+
+ +
+@incollection{TavPer2012eurogp,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  volume = 7244,
+  series = {Lecture Notes in Computer Science},
+  editor = { A. Moraglio  and Sara Silva and  Krzysztof Krawiec  and  Penousal Machado  and  Carlos Cotta },
+  year = 2012,
+  booktitle = {Proceedings of  the 15th European Conference on Genetic Programming, EuroGP 2012},
+  author = { Jorge Tavares  and  Francisco B. Pereira },
+  title = {Automatic Design of Ant Algorithms with Grammatical
+                  Evolution },
+  pages = {206--217}
+}
+
+ +
+@inproceedings{TeiCovStuGas2009:eume,
+  editor = {Ana Viana and others},
+  year = 2009,
+  booktitle = {Proceedings of the EU/MEeting 2009: Debating the future: new
+                  areas of application and innovative approaches},
+  author = {Cristina Teixeira and Jos\'e Covas and  Thomas St{\"u}tzle  and  Ant{\'o}nio Gaspar{-}Cunha },
+  title = {Application of {Pareto} Local Search and Multi-Objective Ant
+                  Colony Algorithms to the Optimization of Co-Rotating Twin
+                  Screw Extruders},
+  pages = {115--120}
+}
+
+ +
+@misc{TensorFlow,
+  author = {Google},
+  title = {TensorFlow},
+  year = 2017,
+  howpublished = {\url{https://www.tensorflow.org}}
+}
+
+ +
+@inproceedings{Teo2010,
+  author = {Teo, K. T. K. and Kow, W. Y. and Chin, Y. K.},
+  booktitle = {Proceedings - 2nd International Conference on Computational
+                  Intelligence, Modelling and Simulation, CIMSim 2010},
+  keywords = {Genetic algorithm,T-junction,Traffic control system,Traffic
+                  flows},
+  pages = {172--177},
+  title = {Optimization of traffic flow within an urban traffic light
+                  intersection with genetic algorithm},
+  year = 2010,
+  organization = {IEEE},
+  publisher = {IEEE Press}
+}
+
+ +
+@incollection{TerRosVal99gecco,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  year = 1999,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 1999},
+  shorteditor = {Wolfgang Banzhaf and others},
+  editor = {Wolfgang Banzhaf and Jason M. Daida and A. E. Eiben
+                  and Max H. Garzon and Vasant Honavar and Mark
+                  J. Jakiela and Robert E. Smith},
+  author = {Hugo Terashima-Mar\'{i}n and  Peter Ross  and Manuel
+                  Valenzuela-Rend\'{o}n},
+  title = {Evolution of Constraint Satisfaction Strategies in
+                  Examination Timetabling},
+  pages = {635--642}
+}
+
+ +
+@incollection{Thie2007adaptive,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  year = 2007,
+  booktitle = {Parameter Setting in Evolutionary Algorithms},
+  editor = {F. Lobo and C. F. Lima and  Zbigniew Michalewicz },
+  title = {Adaptive strategies for operator allocation},
+  author = { Dirk Thierens },
+  pages = {77--90}
+}
+
+ +
+@incollection{Thie2009adaptive,
+  volume = 5752,
+  series = {Lecture Notes in Computer Science},
+  year = 2009,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Engineering Stochastic Local Search
+                  Algorithms. Designing, Implementing and Analyzing
+                  Effective Heuristics. SLS~2009},
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  title = {Adaptive operator selection for iterated local search},
+  author = { Dirk Thierens },
+  pages = {140--144}
+}
+
+ +
+@incollection{Thierens2004:gecco,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 3103,
+  editor = { Kalyanmoy Deb and others},
+  year = 2004,
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2004, Part II},
+  author = { Dirk Thierens },
+  title = {Population-based Iterated Local Search: Restricting the Neighborhood
+  Search by Crossover},
+  pages = {234--245}
+}
+
+ +
+@incollection{Thierens2005:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2005,
+  editor = {  Hans-Georg Beyer  and  Una-May O'Reilly },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2005},
+  author = { Dirk Thierens },
+  title = {An Adaptive Pursuit Strategy for Allocating Operator Probabilities},
+  pages = {1539--1546}
+}
+
+ +
+@incollection{ThoHutHooLey2013:kdd,
+  editor = {Inderjit S. Dhillon and Yehuda Koren and Rayid Ghani and Ted
+                  E. Senator and Paul Bradley and Rajesh Parekh and Jingrui He
+                  and Robert L. Grossman and Ramasamy Uthurusamy},
+  year = 2013,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {The 19th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining, {KDD} 2013},
+  author = {Chris Thornton and  Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {{Auto-WEKA}: Combined Selection and Hyperparameter
+                  Optimization of Classification Algorithms},
+  pages = {847--855}
+}
+
+ +
+@book{ThrunPratt1998,
+  title = {Learning to learn},
+  author = {Thrun, Sebastian and Pratt, Lorien},
+  year = 1998,
+  publisher = {Springer}
+}
+
+ +
+@incollection{TinWhiOch2014:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2014,
+  editor = {Christian Igel and Dirk V. Arnold},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2014},
+  author = {Renato Tin\'os and  Darrell Whitley  and  Gabriela Ochoa },
+  title = {Generalized Asymmetric Partition Crossover ({GAPX}) for the Asymmetric {TSP}},
+  pages = {501--508}
+}
+
+ +
+@incollection{TomKad2019iemoi,
+  epub = {https://dl.acm.org/citation.cfm?id=3321707},
+  isbn = {978-1-4503-6111-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2019,
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  author = { Tomczyk, Micha{\l} K  and  Kadzi{\'n}ski, Mi{\l}osz  },
+  title = {Robust Indicator-Based Algorithm for Interactive Evolutionary
+                  Multiple Objective Optimization},
+  doi = {10.1145/3321707.3321742},
+  abstract = {We propose a novel robust indicator-based algorithm, called
+                  IEMO/I, for interactive evolutionary multiple objective
+                  optimization. During the optimization run, IEMO/I selects at
+                  regular intervals a pair of solutions from the current
+                  population to be compared by the Decision Maker. The
+                  successively provided holistic judgements are employed to
+                  divide the population into fronts of potential
+                  optimality. These fronts are, in turn, used to bias the
+                  evolutionary search toward a subset of Pareto-optimal
+                  solutions being most relevant to the Decision Maker. To
+                  ensure a fine approximation of such a subset, IEMO/I employs
+                  a hypervolume metric within a steady-state indicator-based
+                  evolutionary framework. The extensive experimental evaluation
+                  involving a number of benchmark problems confirms that IEMO/I
+                  is able to construct solutions being highly preferred by the
+                  Decision Maker after a reasonable number of interactions. We
+                  also compare IEMO/I with some selected state-of-the-art
+                  interactive evolutionary hybrids incorporating preference
+                  information in form of pairwise comparisons, proving its
+                  competitiveness.},
+  pages = {629--637},
+  numpages = 9,
+  keywords = {preference learning, indicator-based algorithms, interactive
+                  algorithms, multiple objective optimization, pairwise
+                  comparisons, evolutionary algorithms}
+}
+
+ +
+@book{TotVig2002vrp,
+  title = {The vehicle routing problem},
+  author = { Paolo Toth  and  Vigo, Daniele },
+  year = 2002,
+  publisher = {Society for Industrial and Applied Mathematics, Philadelphia, PA, USA}
+}
+
+ +
+@incollection{ToyShoMorMiy2012,
+  author = {Toyama, F. and Shoji, K. and Mori, H. and Miyamichi, J.},
+  title = {An Iterated Greedy Algorithm for the Binary Quadratic
+                  Programming Problem},
+  booktitle = {Joint 6th International Conference on Soft Computing and
+                  Intelligent Systems (SCIS) and 13th International Symposium
+                  on Advanced Intelligent Systems (ISIS), 2012},
+  publisher = {IEEE Press},
+  year = 2012,
+  pages = {2183--2188}
+}
+
+ +
+@incollection{TraNikCen2022ngopt,
+  volume = 13398,
+  year = 2022,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { G{\"u}nther Rudolph  and  Anna V. Kononova  and  Aguirre, Hern\'{a}n E.  and  Pascal Kerschke  and  Gabriela Ochoa  and  Tea Tu{\v s}ar },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVII}},
+  author = {Trajanov, Risto and Nikolikj, Ana and Cenikj, Gjorgjina and
+                  Teytaud, Fabien and Videau, Mathurin and Olivier Teytaud  and  Tome Eftimov  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Carola Doerr },
+  title = {Improving {Nevergrad}'s Algorithm Selection Wizard {NGOpt}
+                  Through Automated Algorithm Configuration},
+  pages = {18--31},
+  doi = {10.1007/978-3-031-14714-2_2},
+  abstract = {Algorithm selection wizards are effective and versatile tools
+                  that automatically select an optimization algorithm given
+                  high-level information about the problem and available
+                  computational resources, such as number and type of decision
+                  variables, maximal number of evaluations, possibility to
+                  parallelize evaluations, etc. State-of-the-art algorithm
+                  selection wizards are complex and difficult to improve. We
+                  propose in this work the use of automated configuration
+                  methods for improving their performance by finding better
+                  configurations of the algorithms that compose them. In
+                  particular, we use elitist iterated racing (irace) to find
+                  CMA configurations for specific artificial benchmarks that
+                  replace the hand-crafted CMA configurations currently used in
+                  the NGOpt wizard provided by the Nevergrad platform. We
+                  discuss in detail the setup of irace for the purpose of
+                  generating configurations that work well over the diverse set
+                  of problem instances within each benchmark. Our approach
+                  improves the performance of the NGOpt wizard, even on
+                  benchmark suites that were not part of the tuning by irace.}
+}
+
+ +
+@inproceedings{TreWag2019msr,
+  author = {Treude, Christoph and  Markus Wagner },
+  title = {Predicting Good Configurations for GitHub and Stack Overflow
+                  Topic Models},
+  booktitle = {Proceedings of the 16th International Conference on Mining
+                  Software Repositories},
+  year = 2019,
+  series = {MSR '19},
+  pages = {84--95},
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  location = {Montreal, Quebec, Canada},
+  numpages = 12,
+  doi = {10.1109/MSR.2019.00022},
+  acmid = 3341897,
+  keywords = {algorithm portfolio, corpus features, topic modelling}
+}
+
+ +
+@misc{Trick2018sup,
+  author = { Michael A. Trick },
+  title = {Graph Coloring Instances},
+  howpublished = {\url{https://mat.gsia.cmu.edu/COLOR/instances.html}},
+  year = 2018
+}
+
+ +
+@incollection{Tsu06,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  title = {An Enhanced Aggregation Pheromone System for Real-Parameter
+                  Optimization in the {ACO} Metaphor},
+  author = {S. Tsutsui},
+  pages = {60--71}
+}
+
+ +
+@incollection{Tsutsui06:ppsn,
+  year = 2006,
+  volume = 4193,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Runarsson, Thomas Philip and   Hans-Georg Beyer  and  Edmund K. Burke  and  Juan-Juli{\'a}n Merelo  and  Darrell Whitley  and  Xin Yao },
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IX}},
+  author = {S. Tsutsui},
+  title = {{cAS}: Ant Colony Optimization with Cunning Ants},
+  pages = {162--171}
+}
+
+ +
+@book{Tufte1982vis,
+  author = {Tufte, Edward R.},
+  title = {The Visual Display of Quantitative Information},
+  publisher = {Graphics Press},
+  abstract = {The classic book on statistical graphics, charts,
+                  tables. Theory and practice in the design of data graphics,
+                  250 illustrations of the best (and a few of the worst)
+                  statistical graphics, with detailed analysis of how to
+                  display data for precise, effective, quick analysis. Design
+                  of the high-resolution displays, small multiples. Editing and
+                  improving graphics. The data-ink ratio. Time-series,
+                  relational graphics, data maps, multivariate
+                  designs. Detection of graphical deception: design variation
+                  vs. data variation. Sources of deception. Aesthetics and data
+                  graphical displays. This new edition provides excellent color
+                  reproductions of the many graphics of William Playfair, adds
+                  color to other images, and includes all the changes and
+                  corrections accumulated during 17 printings of the first
+                  edition.},
+  address = {Cheshire, CT},
+  edition = {2nd},
+  isbn = {0-9613921-4-2},
+  year = 2001,
+  origyear = 1982,
+  keywords = {data visualization, information graphics, cognitive science}
+}
+
+ +
+@inproceedings{TurBerKra2016safemdp,
+  year = 2016,
+  editor = {D. D. Lee and M. Sugiyama and U. V. Luxburg and I. Guyon and
+                  R. Garnett},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 29)},
+  title = {Safe Exploration in Finite {Markov} Decision Processes with
+                  {Gaussian} Processes},
+  author = {Turchetta, Matteo and Berkenkamp, Felix and Krause, Andreas},
+  pages = {4312--4320},
+  keywords = {SafeMDP},
+  doi = {10.1109/TEVC.2014.2313407},
+  epub = {http://papers.nips.cc/paper/6357-safe-exploration-in-finite-markov-decision-processes-with-gaussian-processes}
+}
+
+ +
+@inproceedings{TurBerKra2019safe,
+  epub = {http://papers.nips.cc/book/advances-in-neural-information-processing-systems-32-2019},
+  year = 2019,
+  editor = {Hanna M. Wallach and Hugo Larochelle and Alina Beygelzimer
+                  and Florence d'Alch{\'{e}}{-}Buc and Emily B. Fox and Roman
+                  Garnett},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS 32)},
+  title = {Safe Exploration for Interactive Machine Learning},
+  author = {Turchetta, Matteo and Berkenkamp, Felix and Krause, Andreas},
+  pages = {2887--2897},
+  keywords = {Reinforcement Learning; Markov Decision Process; SafeML}
+}
+
+ +
+@book{TuringWay2019,
+  key = {TW2019},
+  author = {{The Turing Way Community} and Becky Arnold and Louise Bowler
+                  and Sarah Gibson and Patricia Herterich and Rosie Higman and
+                  Anna Krystalli and Alexander Morley and Martin O'Reilly and
+                  Kirstie Whitaker},
+  title = {The {Turing} {Way}: A Handbook for Reproducible Data Science},
+  month = mar,
+  year = 2019,
+  annote = {Available from \url{https://the-turing-way.netlify.app}. This work was supported by The UKRI Strategic Priorities Fund
+                  under the EPSRC Grant EP/T001569/1, particularly the "Tools,
+                  Practices and Systems" theme within that grant, and by The
+                  Alan Turing Institute under the EPSRC grant EP/N510129/1.},
+  publisher = {Zenodo},
+  version = {v0.0.4},
+  doi = {10.5281/zenodo.3233986}
+}
+
+ +
+@incollection{TusFil2007,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  author = { Tea Tu{\v s}ar  and Bogdan Filipi{\v c}},
+  title = {Differential Evolution versus Genetic Algorithms in
+                  Multiobjective Optimization},
+  pages = {257--271}
+}
+
+ +
+@incollection{TusFil2011vis4d,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  title = {Visualizing {4D} approximation sets of multiobjective
+                  optimizers with prosections},
+  author = { Tea Tu{\v s}ar  and Bogdan Filipi{\v c}},
+  pages = {737--744}
+}
+
+ +
+@mastersthesis{Tusar07master,
+  author = { Tea Tu{\v s}ar },
+  title = {Design of an Algorithm for Multiobjective Optimization with
+                  Differential Evolution},
+  school = {Faculty of Computer and Information Science, University of
+                  Ljubljana},
+  year = 2007,
+  type = {M.Sc. Thesis}
+}
+
+ +
+@incollection{UldAarBan1991,
+  doi = {10.1007/BFb0029723},
+  address = {Berlin\slash Heidelberg},
+  aseries = {Lecture Notes in Computer Science},
+  avolume = 496,
+  publisher = {Springer},
+  editor = { Hans-Paul Schwefel  and R. M{\"a}nner},
+  year = 1991,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {I}},
+  author = {N. L. J. Ulder and  Emile H. L. Aarts  and H.-J. Bandelt and  Peter J. M. van Laarhoven and  Erwin Pesch },
+  title = {Genetic Local Search Algorithms for the Travelling Salesman
+                  Problem},
+  pages = {109--116}
+}
+
+ +
+@incollection{UlrBadThi2010ppsn,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  author = {Tamara Ulrich and  Johannes Bader  and  Lothar Thiele },
+  title = {Defining and Optimizing Indicator-Based Diversity Measures in
+                  Multiobjective Search},
+  doi = {10.1007/978-3-642-15844-5_71},
+  pages = {707--717},
+  annote = {Two archive; two populations; decision space diversity}
+}
+
+ +
+@inproceedings{ValDubStu13:cec,
+  year = 2013,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  key = {IEEE CEC},
+  author = {Andrea Valsecchi and  J{\'e}r{\'e}mie Dubois-Lacoste  and  Thomas St{\"u}tzle  and Sergio Damas and Jos{\'e} Santamar{\'i}a and Linda Marrakchi-Kacem},
+  title = {Evolutionary Medical Image Registration using Automatic Parameter Tuning},
+  pages = {1326--1333}
+}
+
+ +
+@inproceedings{ValFawGerHoo2011icaps,
+  year = 2011,
+  booktitle = {Proceedings of ICAPS-PAL11},
+  editor = {Karpas, Erez and Jim{\'e}nez Celorrio, Sergio and Kambhampati, Subbarao},
+  author = {Mauro Vallati and  Chris Fawcett  and Alfonso E. Gerevini and  Holger H. Hoos  and Alessandro Saetti},
+  title = {Generating Fast Domain-Optimized Planners by Automatically
+                  Configuring a Generic Parameterised Planner}
+}
+
+ +
+@book{VanAar1987,
+  title = {Simulated Annealing: Theory and Applications},
+  author = { Peter J. M. van Laarhoven and  Emile H. L. Aarts },
+  volume = 37,
+  year = 1987,
+  publisher = {Springer}
+}
+
+ +
+@incollection{VanHut2018hyper,
+  key = {SIGKDD},
+  month = jul,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2018,
+  editor = {Yike Guo and Faisal Farooq},
+  booktitle = {24th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  author = { van Rijn, Jan N.  and  Frank Hutter },
+  title = {Hyperparameter Importance Across Datasets},
+  pages = {2367--2376},
+  doi = {10.1145/3219819.3220058},
+  abstract = {With the advent of automated machine learning, automated
+                  hyperparameter optimization methods are by now routinely used
+                  in data mining. However, this progress is not yet matched by
+                  equal progress on automatic analyses that yield information
+                  beyond performance-optimizing hyperparameter settings. In
+                  this work, we aim to answer the following two questions:
+                  Given an algorithm, what are generally its most important
+                  hyperparameters, and what are typically good values for
+                  these? We present methodology and a framework to answer these
+                  questions based on meta-learning across many datasets. We
+                  apply this methodology using the experimental meta-data
+                  available on OpenML to determine the most important
+                  hyperparameters of support vector machines, random forests
+                  and Adaboost, and to infer priors for all their
+                  hyperparameters. The results, obtained fully automatically,
+                  provide a quantitative basis to focus efforts in both manual
+                  algorithm design and in automated hyperparameter
+                  optimization. The conducted experiments confirm that the
+                  hyperparameters selected by the proposed method are indeed
+                  the most important ones and that the obtained priors also
+                  lead to statistically significant improvements in
+                  hyperparameter optimization.},
+  numpages = 10,
+  keywords = {hyperparameter optimization, meta-learning, hyperparameter
+                  importance}
+}
+
+ +
+@incollection{VanKorBlo2018,
+  publisher = {{AAAI} Press},
+  month = feb,
+  year = 2018,
+  editor = {Sheila A. McIlraith and Kilian Q. Weinberger},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  title = {{MERCS}: multi-directional ensembles of regression and classification trees},
+  author = {Van Wolputte, Elia and Korneva, Evgeniya and Blockeel, Hendrik},
+  pages = {4276--4283}
+}
+
+ +
+@inproceedings{VedFul2010vlfeat,
+  title = {{VLFeat}: An open and portable library of computer vision
+                  algorithms},
+  author = {Vedaldi, Andrea and Fulkerson, Brian},
+  booktitle = {Proceedings of the 18th ACM international conference on
+                  Multimedia},
+  pages = {1469--1472},
+  year = 2010,
+  organization = {ACM}
+}
+
+ +
+@inproceedings{VelLam1998gp,
+  year = 1998,
+  publisher = {Stanford University Bookstore},
+  address = {Stanford University, California},
+  month = jul,
+  editor = {John R. Koza},
+  booktitle = {Late Breaking Papers at the Genetic Programming 1998
+                  Conference},
+  alias = {Veldhuizen98a},
+  key = {Van Veldhuizen and Lamont, 1998a},
+  title = {Evolutionary Computation and Convergence to a
+                  {Pareto} Front},
+  author = { David A. {Van Veldhuizen}  and  Gary B. Lamont },
+  pages = {221--228},
+  keywords = {generational distance}
+}
+
+ +
+@phdthesis{Veldhuizen1999phd,
+  title = {Multiobjective evolutionary algorithms: Classifications,
+                  analyses, and new innovations},
+  author = { David A. {Van Veldhuizen} },
+  school = {Department of Electrical and Computer Engineering, Graduate
+                  School of Engineering, Air Force Institute of Technology},
+  year = 1999,
+  address = {Wright-Patterson AFB, Ohio}
+}
+
+ +
+@incollection{VerCarSte2022bias,
+  isbn = 9781450392686,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Diederick Vermetten  and Caraffini, Fabio and van Stein, Bas and  Anna V. Kononova },
+  title = {Using Structural Bias to Analyse the Behaviour of Modular
+                  {CMA-ES}},
+  pages = {1674--1682},
+  doi = {10.1145/3520304.3534035},
+  location = {Boston, Massachusetts}
+}
+
+ +
+@incollection{VerLieDha2011gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2011,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  author = { Verel, S{\'e}bastien  and  Arnaud Liefooghe  and  Dhaenens, Clarisse },
+  title = {Set-based Multiobjective Fitness Landscapes: A Preliminary
+                  Study},
+  pages = {769--776},
+  doi = {10.1145/2001576.2001681},
+  acmid = 2001681
+}
+
+ +
+@incollection{VerWanDoeBac2020cash,
+  epub = {https://dl.acm.org/citation.cfm?id=3377930},
+  location = {Canc{\'u}n, Mexico},
+  isbn = {978-1-4503-7128-5},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2020,
+  editor = { Carlos A. {Coello Coello} },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  author = { Diederick Vermetten  and Hao Wang and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {Integrated vs. Sequential Approaches for Selecting and Tuning
+                  {CMA-ES} Variants},
+  doi = {10.1145/3377930.3389831}
+}
+
+ +
+@incollection{VerWanLopDoe2022undersampling,
+  location = {Boston, Massachusetts},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  address = { New York, NY},
+  year = 2022,
+  publisher = {ACM Press},
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  author = { Diederick Vermetten  and Hao Wang and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Carola Doerr  and  Thomas B{\"a}ck },
+  title = {Analyzing the Impact of Undersampling on the Benchmarking and
+                  Configuration of Evolutionary Algorithms},
+  doi = {10.1145/3512290.3528799},
+  abstract = {The stochastic nature of iterative optimization heuristics
+                  leads to inherently noisy performance measurements. Since
+                  these measurements are often gathered once and then used
+                  repeatedly, the number of collected samples will have a
+                  significant impact on the reliability of algorithm
+                  comparisons. We show that care should be taken when making
+                  decisions based on limited data. Particularly, we show that
+                  the number of runs used in many benchmarking studies, e.g.,
+                  the default value of 15 suggested by the COCO environment,
+                  can be insufficient to reliably rank algorithms on well-known
+                  numerical optimization benchmarks.Additionally, methods for
+                  automated algorithm configuration are sensitive to
+                  insufficient sample sizes. This may result in the
+                  configurator choosing a "lucky" but poor-performing
+                  configuration despite exploring better ones. We show that
+                  relying on mean performance values, as many configurators do,
+                  can require a large number of runs to provide accurate
+                  comparisons between the considered configurations. Common
+                  statistical tests can greatly improve the situation in most
+                  cases but not always. We show examples of performance losses
+                  of more than 20\%, even when using statistical races to
+                  dynamically adjust the number of runs, as done by irace. Our
+                  results underline the importance of appropriately considering
+                  the statistical distribution of performance values.},
+  pages = {867--875},
+  numpages = 9,
+  keywords = {parameter tuning, evolution strategies, algorithm
+                  configuration, performance measures}
+}
+
+ +
+@incollection{VidLeiTey2022eurogp,
+  address = { Cham, Switzerland},
+  publisher = {Springer Nature},
+  year = 2022,
+  series = {Lecture Notes in Computer Science},
+  editor = {Eric Medvet and  Gisele Pappa  and Bing Xue},
+  booktitle = {Proceedings of  the 25th European Conference on Genetic Programming, EuroGP 2022},
+  author = {Mathurin Videau and Alessandro Leite and Olivier Teytaud  and  Marc Schoenauer },
+  title = {Multi-Objective Genetic Programming for Explainable
+                  Reinforcement Learning},
+  pages = {256--281},
+  keywords = {genetic algorithms, genetic programming: Poster}
+}
+
+ +
+@inproceedings{Vieira2021nas,
+  year = 2021,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2021 Congress on Evolutionary Computation (CEC 2021)},
+  key = {IEEE CEC},
+  author = {Vieira, Carlos and   P{\'e}rez C{\'a}ceres, Leslie and  Leonardo C. T. Bezerra },
+  title = {Evaluating Anytime Performance on {NAS}-{Bench}-101},
+  pages = {1249--1256},
+  doi = {10.1109/CEC45853.2021.9504902}
+}
+
+ +
+@phdthesis{Violin2014PhD,
+  author = {Alessia Violin},
+  title = {Mathematical Programming Approaches to Pricing Problems},
+  school = {Facult\'{e} de Sciences, Universit\'{e} Libre de Bruxelles and Dipartimento di Ingegneria e Architettura, Universit\`{a} degli studi di Trieste},
+  year = 2014,
+  annote = {Supervised by Dr. Martine Labb\'{e} and Dr. Lorenzo Castelli}
+}
+
+ +
+@incollection{VosHanIgel2010gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2010,
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Improved Step Size Adaptation for the {MO-CMA-ES}},
+  author = { Vo{\ss}, Thomas  and  Nikolaus Hansen  and  Christian Igel },
+  pages = {487--494}
+}
+
+ +
+@incollection{VouTsa2003,
+  alias = {GloKoc02:handbook-mh},
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  year = 2002,
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  booktitle = {Handbook of Metaheuristics},
+  author = { Christos Voudouris  and  Edward P. K. Tsang },
+  title = {Guided Local Search},
+  pages = {185--218}
+}
+
+ +
+@book{WIS:MOA_vol2,
+  editor = { Dragan A. Savic  and  Godfrey A. Walters },
+  title = {Water Industry Systems: Modelling and Optimization
+                  Applications},
+  booktitle = {Water Industry Systems: Modelling and Optimization
+                  Applications},
+  year = 1999,
+  volume = 2,
+  publisher = { Research Studies Press Ltd. },
+  address = {Baldock, United Kingdom}
+}
+
+ +
+@incollection{WacSuiYueOno2018aaai,
+  publisher = {{AAAI} Press},
+  month = feb,
+  year = 2018,
+  editor = {Sheila A. McIlraith and Kilian Q. Weinberger},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  author = {Akifumi Wachi and Yanan Sui and Yisong Yue and Masahiro Ono},
+  title = {Safe Exploration and Optimization of Constrained {MDP}s Using
+                  {Gaussian} Processes},
+  pages = {6548--6556},
+  keywords = {Markov Decision Process, Gaussian Processes},
+  abstract = {We present a reinforcement learning approach to explore and
+                  optimize a safety-constrained Markov Decision
+                  Process(MDP). In this setting, the agent must maximize
+                  discounted cumulative reward while constraining the
+                  probability of entering unsafe states, defined using a safety
+                  function being within some tolerance. The safety values of
+                  all states are not known a priori, and we probabilistically
+                  model them via a Gaussian Process (GP) prior. As such,
+                  properly behaving in such an environment requires balancing a
+                  three-way trade-off of exploring the safety function,
+                  exploring the reward function, and exploiting acquired
+                  knowledge to maximize reward. We propose a novel approach to
+                  balance this trade-off. Specifically, our approach explores
+                  unvisited states selectively; that is, it prioritizes the
+                  exploration of a state if visiting that state significantly
+                  improves the knowledge on the achievable cumulative
+                  reward. Our approach relies on a novel information gain
+                  criterion based on Gaussian Process representations of the
+                  reward and safety functions. We demonstrate the effectiveness
+                  of our approach on a range of experiments, including a
+                  simulation using the real Martian terrain data.},
+  doi = {10.1609/aaai.v32i1.12103}
+}
+
+ +
+@incollection{WagBeuNau2007:many,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  author = { Tobias Wagner  and Nicola Beume and  Boris Naujoks },
+  title = {{Pareto}-, Aggregation-, and Indicator-Based Methods in Many-Objective
+               Optimization},
+  pages = {742--756}
+}
+
+ +
+@inproceedings{WagFriLin2017vertex,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = { Markus Wagner  and  Tobias Friedrich  and  Marius Thomas Lindauer },
+  title = {Improving local search in a minimum vertex cover solver for
+                  classes of networks},
+  pages = {1704--1711},
+  keywords = {graph theory;search problems;local search;minimum vertex
+                  cover solver;network classes;straightforward alternative
+                  approach;benchmark sets;graphs;algorithm portfolio;single
+                  integrated approach;Training;Portfolios;Algorithm design and
+                  analysis;Prediction algorithms;Machine learning
+                  algorithms;Optimization;Benchmark testing,smac,paramils},
+  doi = {10.1109/CEC.2017.7969507}
+}
+
+ +
+@incollection{WagNeu2013,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  title = {A Fast Approximation-guided Evolutionary Multi-objective
+                  Algorithm},
+  author = { Markus Wagner  and  Frank Neumann },
+  pages = {687--694}
+}
+
+ +
+@incollection{WahChe2000cp,
+  year = 2000,
+  volume = 1894,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Principles and Practice of Constraint Programming,
+                  CP 2000},
+  editor = {Rina Dechter},
+  author = { Benjamin W. Wah  and  Yi Xin Chen },
+  title = {Optimal Anytime Constrained Simulated Annealing for
+                  Constrained Global Optimization},
+  pages = {425--440},
+  doi = {10.1007/3-540-45349-0_31}
+}
+
+ +
+@inproceedings{Wal97,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 1997,
+  booktitle = {Proceedings of AAAI 1997 -- Fourteenth National Conference on
+                  Artificial Intelligence},
+  editor = {Benjamin Kuipers and Bonnie L. Webber},
+  author = {J. P. Walser},
+  title = {Solving Linear Pseudo-Boolean Constraint Problems with Local Search},
+  pages = {269--274}
+}
+
+ +
+@book{Wal98:phd,
+  author = {J. P. Walser},
+  title = {Integer Optimization by Local Search: A
+                  Domain-Independent Approach},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 1999,
+  volume = 1637,
+  series = {Lecture Notes in Computer Science}
+}
+
+ +
+@inproceedings{WalIyeVen98,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 1998,
+  booktitle = {Proceedings of AAAI 1998 -- Fifteenth National Conference on
+                  Artificial Intelligence},
+  editor = {Jack Mostow and Chuck Rich},
+  author = {J. P. Walser and R. Iyer and N. Venkatasubramanyan},
+  title = {An Integer Local Search Method with Application to
+                  Capacitated Production Planning},
+  pages = {373--379}
+}
+
+ +
+@incollection{Walsh1995:ijcai,
+  publisher = {Morgan Kaufmann Publishers},
+  editor = {Martha E. Pollack},
+  year = 1997,
+  booktitle = {Proceedings of  the 15th International Joint Conference on Artificial Intelligence (IJCAI-97)},
+  author = {Toby Walsh},
+  title = {Depth-bounded Discrepancy Search},
+  pages = {1388--1395}
+}
+
+ +
+@inproceedings{WanDohJin2018cec,
+  year = 2018,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2018 Congress on Evolutionary Computation (CEC 2018)},
+  key = {IEEE CEC},
+  title = {Hierarchical surrogate-assisted evolutionary multi-scenario
+                  airfoil shape optimization},
+  author = {Wang, Handing and Doherty, John and  Yaochu Jin },
+  pages = {1--8},
+  keywords = {scenario-based}
+}
+
+ +
+@incollection{WanDonChenLin2014,
+  author = {Wang, Yanqi and Dong, Xingye and Chen, Ping and Lin, Youfang},
+  title = {Iterated local search algorithms for the sequence-dependent
+                  setup times flow shop scheduling problem minimizing makespan},
+  booktitle = {Foundations of Intelligent Systems},
+  pages = {329--338},
+  year = 2014,
+  publisher = {Springer}
+}
+
+ +
+@incollection{WanMeiZha2021mogp,
+  doi = {10.1145/3449639.3459373},
+  location = {Lille, France},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2021,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  title = {Two-stage multi-objective genetic programming with archive
+                  for uncertain capacitated arc routing problem},
+  author = {Wang, Shaolin and Mei, Yi and  Zhang, Mengjie },
+  pages = {287--295}
+}
+
+ +
+@incollection{Ward2008multivariate,
+  title = {Multivariate data glyphs: Principles and practice},
+  author = {Ward, Matthew O.},
+  booktitle = {Handbook of Data Visualization},
+  editor = {Chen, Chun-houh and H{\"a}rdle, Wolfgang Karl and Unwin, Antony},
+  pages = {179--198},
+  year = 2008,
+  publisher = {Springer}
+}
+
+ +
+@inproceedings{Wau2017eternity,
+  author = {Wauters, Tony},
+  title = {10 years of Eternity II--from \$2 million puzzle to
+                  challenging optimization problem},
+  booktitle = {International Workshop on Cutting, Packing and Related
+                  Topics},
+  address = {Gent, Belgium},
+  year = 2017,
+  url = {https://lirias.kuleuven.be/1675982?limo=0},
+  abstract = {The Eternity II (EII) puzzle is a commercial edge matching
+                  puzzle in which 256 square tiles with four coloured edges
+                  must be arranged on a 16 by 16 grid such that all tile edges
+                  are matched. In addition, a complete solution requires that
+                  the `grey' patterns, which appear only on a subset of the
+                  tiles, should be matched to the outer edges of the grid. The
+                  puzzle belongs to the more general class of Edge Matching
+                  Puzzles, which have been shown to be NP-complete. In July
+                  2007, toy distributor Tomy UK Ltd. released this challenging
+                  edge matching puzzle with a \$2 million prize. However, to
+                  the best of our knowledge, no complete solution has ever been
+                  found. Meanwhile, the final scrutiny date for the cash price,
+                  31 December 2010, has passed, leaving the large money prize
+                  unclaimed. In its 10 years of existence many people tried to
+                  solve EII and some are still trying. Many approaches to Edge
+                  Matching Puzzles are reported in the literature. Among these
+                  approaches are constraint programming and backtracking,
+                  metaheuristics, and evolutionary methods. Other approaches
+                  translate the problem into SAT, MILP or max-clique and then
+                  solve it with appropriate state of the art solvers. Some
+                  approaches have also been implemented on parallel computing
+                  or dedicated hardware.}
+}
+
+ +
+@incollection{Wegener2005,
+  year = 2005,
+  volume = 3580,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  booktitle = {Proceedings of the 32nd International Colloquium on Automata, Languages and Programming, {ICALP} 2005},
+  editor = {Lu{\'i}s Caires and Giuseppe F. Italiano and Lu{\'i}s Monteiro and Catuscia Palamidessi and Moti Yung},
+  title = {Simulated annealing beats metropolis in combinatorial optimization},
+  author = { Ingo Wegener },
+  pages = {589--601}
+}
+
+ +
+@inproceedings{Wegley00,
+  author = { Chad Wegley  and  Muzaffar Eusuff  and  Kevin E. Lansey },
+  title = {Determining Pump Operations Using Particle Swarm
+                  Optimization},
+  booktitle = {Building Partnerships, Proceedings of the Joint
+                  Conference on Water Resources Engineering and Water
+                  Resources Planning and Management},
+  year = 2000,
+  editor = {Rollin H. Hotchkiss and Michael Glade},
+  address = {Minneapolis, USA}
+}
+
+ +
+@inproceedings{Wegner76research,
+  title = {Research paradigms in computer science},
+  author = {Peter Wegner},
+  booktitle = {ICSE'76: Proceedings of the 2nd international conference on
+                  Software engineering},
+  month = oct,
+  year = 1976,
+  pages = {322--330}
+}
+
+ +
+@inproceedings{WeiSau2006introduction,
+  editor = {Anthony Cohn},
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  year = 2006,
+  volume = 6,
+  booktitle = {Proceedings of  the 21st National Conference on Artificial Intelligence},
+  author = {Weinberger, Kilian Q. and Saul, Lawrence K.},
+  title = {An Introduction to Nonlinear Dimensionality Reduction by
+                  Maximum Variance Unfolding},
+  pages = {1683--1686}
+}
+
+ +
+@inproceedings{WeiShaSau2004dimension,
+  year = 2004,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  booktitle = {Proceedings of  the 21st International Conference on Machine Learning, {ICML} 2004},
+  editor = {Carla E. Brodley},
+  author = {Kilian Q. Weinberger and Fei Sha and Lawrence K. Saul},
+  title = {Learning a kernel matrix for nonlinear dimensionality
+                  reduction},
+  doi = {10.1145/1015330.1015345},
+  abstract = {We investigate how to learn a kernel matrix for high
+                  dimensional data that lies on or near a low dimensional
+                  manifold. Noting that the kernel matrix implicitly maps the
+                  data into a nonlinear feature space, we show how to discover
+                  a mapping that "unfolds" the underlying manifold from which
+                  the data was sampled. The kernel matrix is constructed by
+                  maximizing the variance in feature space subject to local
+                  constraints that preserve the angles and distances between
+                  nearest neighbors. The main optimization involves an instance
+                  of semidefinite programming---a fundamentally different
+                  computation than previous algorithms for manifold learning,
+                  such as Isomap and locally linear embedding. The optimized
+                  kernels perform better than polynomial and Gaussian kernels
+                  for problems in manifold learning, but worse for problems in
+                  large margin classification. We explain these results in
+                  terms of the geometric properties of different kernels and
+                  comment on various interpretations of other manifold learning
+                  algorithms as kernel methods.}
+}
+
+ +
+@incollection{WesBeuRud2010ppsn,
+  volume = 6238,
+  year = 2010,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  series = {Lecture Notes in Computer Science},
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  author = { Simon Wessing  and  Nicola Beume  and  G{\"u}nther Rudolph  and  Boris Naujoks },
+  title = {Parameter Tuning Boosts Performance of Variation
+                  Operators in Multiobjective Optimization},
+  pages = {728--737},
+  doi = {10.1007/978-3-642-15844-5_73}
+}
+
+ +
+@incollection{Whaley2011atlas,
+  doi = {10.1007/978-0-387-09766-4_244},
+  publisher = {Springer, US},
+  year = 2011,
+  editor = {David Padua},
+  booktitle = {Encyclopedia of Parallel Computing},
+  author = {Clint R. Whaley},
+  title = {{ATLAS}: Automatically Tuned Linear Algebra Software},
+  pages = {95--101}
+}
+
+ +
+@inproceedings{WhiBra2012cec,
+  year = 2012,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2012 Congress on Evolutionary Computation (CEC 2012)},
+  key = {IEEE CEC},
+  author = {While, L. and Bradstreet, L.},
+  title = {Applying the {WFG} Algorithm to Calculate Incremental
+                  Hypervolumes},
+  pages = {1--8}
+}
+
+ +
+@incollection{WhiPagOpp98,
+  publisher = {CSREA Press},
+  year = 1998,
+  editor = {H. R. Arabnia},
+  booktitle = {Proceedings of the International Conference on
+                  Parallel and Distributed Processing Techniques and
+                  Applications (PDPTA'98)},
+  author = {T. White and B. Pagurek and F. Oppacher},
+  title = {Connection Management Using Adaptive Mobile Agents},
+  pages = {802--809}
+}
+
+ +
+@incollection{WieStu2006:ants,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2006,
+  volume = 4150,
+  series = {Lecture Notes in Computer Science},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  author = {W. Wiesemann and  Thomas St{\"u}tzle },
+  title = {Iterated Ants: An Experimental Study for the
+                  Quadratic Assignment Problem},
+  pages = {179--190}
+}
+
+ +
+@techreport{Wiegele2007biqmac,
+  title = {{Biq} {Mac} {Library} -- A collection of {Max}-{Cut} and
+                  quadratic 0-1 programming instances of medium size},
+  author = {Wiegele, Angelika},
+  institution = {Institut f{\"u}r Mathematik, Alpen-Adria-Universit{\"a}t
+                  Klagenfurt},
+  year = 2007,
+  url = {http://biqmac.aau.at/biqmaclib.pdf}
+}
+
+ +
+@misc{Wiegele2007sup,
+  author = {Wiegele, Angelika},
+  title = {{Biq} {Mac} {Library} -- Binary Quadratic and Max Cut
+                  Library},
+  howpublished = {\url{http://biqmac.aau.at/biqmaclib.html}},
+  year = 2007
+}
+
+ +
+@incollection{Wierzbicki1980mcdmta,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  number = 177,
+  editor = {Fandel, G. and Gal, T.},
+  year = 1980,
+  booktitle = {Multiple Criteria Decision Making Theory and Application},
+  author = { Andrzej P. Wierzbicki },
+  title = {The Use of Reference Objectives in Multiobjective
+                  Optimisation},
+  pages = {468--486},
+  doi = {10.1007/978-3-642-48782-8_32}
+}
+
+ +
+@book{WilShm2011,
+  title = {The design of approximation algorithms},
+  author = {Williamson, David P. and Shmoys, David B.},
+  year = 2011,
+  publisher = {Cambridge University Press}
+}
+
+ +
+@incollection{WolMer2009:evocop,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5482,
+  year = 2009,
+  editor = { Carlos Cotta  and P. Cowling},
+  booktitle = {Proceedings of EvoCOP 2009 -- 9th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  author = {Steffen Wolf and  Peter Merz },
+  title = {Iterated Local Search for Minimum Power Symmetric Connectivity in Wireless Networks},
+  pages = {192--203}
+}
+
+ +
+@inproceedings{XuHooLey2010aaai,
+  year = 2010,
+  publisher = {{AAAI} Press},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Maria Fox and David Poole},
+  author = { Lin Xu  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  title = {Hydra: Automatically Configuring Algorithms for
+                  Portfolio-Based Selection},
+  keywords = {automated algorithm design; portfolio-based algorithm
+                  selection; automated algorithm configuration; SAT; stochastic
+                  local search},
+  doi = {10.1609/aaai.v24i1.7565}
+}
+
+ +
+@techreport{XuHutHoo2011tr01,
+  title = {{Hydra-MIP}: Automated Algorithm Configuration and
+                  Selection for Mixed Integer Programming},
+  author = { Lin Xu  and  Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  institution = {Department of Computer Science, University of British Columbia, Canada},
+  year = 2011,
+  number = {TR-2011-01},
+  url = {https://www.cs.ubc.ca/tr/2011/tr-2011-01}
+}
+
+ +
+@incollection{XuKhHo2016,
+  address = { Cham, Switzerland},
+  publisher = {Springer},
+  volume = 10079,
+  editor = {Paola Festa and  Meinolf Sellmann  and  Joaquin Vanschoren },
+  series = {Lecture Notes in Computer Science},
+  year = 2016,
+  booktitle = {Learning and Intelligent Optimization, 10th International
+                  Conference, LION 10},
+  title = {Quantifying the similarity of algorithm configurations},
+  author = { Lin Xu  and  KhudaBukhsh, A. R.  and  Holger H. Hoos  and  Kevin Leyton-Brown },
+  pages = {203--217}
+}
+
+ +
+@inproceedings{XuPokPai2006correntropyPCA,
+  publisher = {{IEEE}},
+  year = 2006,
+  booktitle = {Proceedings of the International Joint Conference on Neural
+                  Networks, {IJCNN} 2006},
+  key = {IJCNN},
+  author = {Jian{-}Wu Xu and Puskal P. Pokharel and Ant{\'{o}}nio
+                  R. C. Paiva and Jos{\'{e}} C. Pr{\'i}ncipe},
+  title = {Nonlinear Component Analysis Based on Correntropy},
+  pages = {1851--1855},
+  doi = {10.1109/IJCNN.2006.246905}
+}
+
+ +
+@incollection{YamHalColIac2017,
+  doi = {10.1007/978-3-319-55849-3},
+  booktitle = {Applications of Evolutionary Computation},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 10199,
+  year = 2017,
+  publisher = {Springer},
+  editor = {Squillero, Giovanni and Sim, Kevin},
+  author = {Yaman, Anil and Hallawa, Ahmed and Coler, Matt and Iacca, Giovanni},
+  title = {Presenting the {ECO}: evolutionary computation ontology},
+  pages = {603--619}
+}
+
+ +
+@book{Yao1999ecbook,
+  author = { Xin Yao },
+  title = {Evolutionary Computation: Theory and Applications},
+  isbn = 9810223064,
+  publisher = {World Scientific Singapore},
+  address = {River Edge, NJ},
+  numpages = 360,
+  year = 1999,
+  language = {English},
+  keywords = {Evolutionary programming (Computer science); Neural networks
+                  (Computer science); Evolutionary computation}
+}
+
+ +
+@inproceedings{YarAstOzcPar2014eals,
+  publisher = {IEEE},
+  year = 2014,
+  booktitle = {Evolving and Autonomous Learning Systems (EALS), 2014 IEEE
+                  Symposium on},
+  editor = {Angelov, Plamen and others},
+  author = {A. Yarimcam and S. Asta and  Ender {\"O}zcan  and  Andrew J. Parkes },
+  title = {Heuristic Generation via Parameter Tuning for Online Bin Packing},
+  pages = {102--108},
+  doi = {10.1109/EALS.2014.7009510},
+  keywords = {irace}
+}
+
+ +
+@incollection{YasAraMei2019comparison,
+  author = {Yasojima, Carlos and Ara{\'u}jo, Tiago and Meiguins, Bianchi
+                  and Neto, Nelson and Morais, Jefferson},
+  booktitle = {Progress in Artificial Intelligence},
+  title = {A Comparison of Genetic Algorithms and Particle Swarm
+                  Optimization to Estimate Cluster-Based Kriging Parameters},
+  year = 2019,
+  address = { Cham, Switzerland},
+  editor = {Moura Oliveira, Paulo and Novais, Paulo and Reis, Lu{\'i}s
+                  Paulo },
+  pages = {750--761},
+  publisher = {Springer International Publishing},
+  abstract = {Kriging is one of the most used spatial estimation methods in
+                  real-world applications. Some kriging parameters must be
+                  estimated in order to reach a good accuracy in the
+                  interpolation process, however, this task remains a
+                  challenge. Various optimization methods have been tested to
+                  find good parameters of the kriging process. In recent years,
+                  many authors are using bio-inspired techniques and achieving
+                  good results in estimating these parameters in comparison
+                  with traditional techniques. This paper presents a comparison
+                  between well known bio-inspired techniques such as Genetic
+                  Algorithms and Particle Swarm Optimization in the estimation
+                  of the essential kriging parameters: nugget, sill, range,
+                  angle, and factor. In order to perform the tests, we proposed
+                  a methodology based on the cluster-based kriging
+                  method. Considering the Friedman test, the results showed no
+                  statistical difference between the evaluated algorithms in
+                  optimizing kriging parameters. On the other hand, the
+                  Particle Swarm Optimization approach presented a faster
+                  convergence, which is important in this high-cost
+                  computational problem.},
+  isbn = {978-3-030-30241-2}
+}
+
+ +
+@inproceedings{YavAydStu2016,
+  year = 2016,
+  isbn = {978-1-5090-0623-6},
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2016 Congress on Evolutionary Computation (CEC 2016)},
+  key = {IEEE CEC},
+  author = { G{\"{u}}rcan Yavuz  and  Do\v{g}an Ayd{\i}n  and  Thomas St{\"u}tzle },
+  title = {Self-adaptive Search Equation-based Artificial Bee Colony Algorithm on the {CEC} 2014 Benchmark Functions},
+  pages = {1173--1180}
+}
+
+ +
+@inproceedings{YouJohKArSmi1997,
+  author = {Cliff Young and David S. Johnson and David R. Karger and Michael D. Smith},
+  title = {Near-optimal Intraprocedural Branch Alignment},
+  booktitle = {Proceedings of the {ACM} {SIGPLAN}'97 Conference on Programming Language
+               Design and Implementation (PLDI), Las Vegas, Nevada},
+  pages = {183--193},
+  editor = {Marina C. Chen and Ron K. Cytron and A. Michael Berman},
+  publisher = {ACM Press},
+  year = 1997
+}
+
+ +
+@incollection{YuWanLee2011dt,
+  isbn = {978-3-642-14125-6},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  year = 2011,
+  booktitle = {Preference Learning},
+  editor = {F{\"u}rnkranz, Johannes and  Eyke H{\"u}llermeier },
+  author = {Yu, Philip L. H.  and Wan, Wai Ming and Lee, Paul H.},
+  title = {Decision Tree Modeling for Ranking Data},
+  pages = {83--106},
+  doi = {10.1007/978-3-642-14125-6_5}
+}
+
+ +
+@incollection{YuaFug2008:hm,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5296,
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and  Carlos Cotta  and  Antonio J. Fern{\'a}ndez  and Jos\'e E. Gallardo and  Andrea Roli  and  M. Sampels },
+  year = 2008,
+  booktitle = {Hybrid Metaheuristics},
+  author = { Zhi Yuan  and Armin F\"ugenschuh and Henning Homfeld and   Prasanna Balaprakash  and  Thomas St{\"u}tzle  and Michael Schoch},
+  title = {Iterated Greedy Algorithms for a Real-World Cyclic Train
+                  Scheduling Problem},
+  pages = {102--116}
+}
+
+ +
+@incollection{YuaGal2004racing,
+  year = 2004,
+  volume = 3242,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Xin Yao  and others},
+  aeditor = { Xin Yao  and  Edmund K. Burke  and  Jos{\'e} A. Lozano  and Smith,
+                  Jim and Merelo-Guerv{\'o}s, Juan Juli{\'a}n and Bullinaria,
+                  John A.  and Rowe, Jonathan E.  and Ti{\v{n}}o, Peter and
+                  Kab{\'a}n, Ata and Schwefel, Hans-Paul},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VIII}},
+  author = {Yuan, Bo and Gallagher, Marcus},
+  title = {Statistical Racing Techniques for Improved Empirical
+                  Evaluation of Evolutionary Algorithms},
+  pages = {172--181}
+}
+
+ +
+@incollection{YuaGal2007meta,
+  address = { Berlin, Germany},
+  publisher = {Springer},
+  year = 2007,
+  booktitle = {Parameter Setting in Evolutionary Algorithms},
+  editor = {F. Lobo and C. F. Lima and  Zbigniew Michalewicz },
+  author = {Yuan, Bo and Gallagher, Marcus},
+  title = {Combining {Meta}-{EAs} and racing for difficult {EA}
+                  parameter tuning tasks},
+  pages = {121--142}
+}
+
+ +
+@incollection{YuaStuMonLauBir13,
+  isbn = {978-1-4503-1963-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = { Christian Blum  and  Alba, Enrique },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  author = { Zhi Yuan  and  Marco A. {Montes de Oca}  and  Thomas St{\"u}tzle  and  Hoong Chuin Lau  and  Mauro Birattari },
+  title = {An Analysis of Post-selection in Automatic Configuration},
+  pages = {1557--1564}
+}
+
+ +
+@inproceedings{YueDuStu2017cec,
+  year = 2017,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  key = {IEEE CEC},
+  author = {Lin Yuefeng and Wenli Du and  Thomas St{\"u}tzle },
+  title = {Three {L-SHADE} Based Algorithms on Mixed-variables Optimization Problems},
+  pages = {2274--2281}
+}
+
+ +
+@inproceedings{YueGaoWag2012cec,
+  year = 2012,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  booktitle = {Proceedings of  the 2012 Congress on Evolutionary Computation (CEC 2012)},
+  key = {IEEE CEC},
+  title = {An adaptive data structure for evolutionary multi-objective
+                  algorithms with unbounded archives},
+  author = {Yuen, Joseph and Gao, Sophia and  Markus Wagner  and  Frank Neumann },
+  pages = {1--8}
+}
+
+ +
+@incollection{YunEps2012learn,
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7219,
+  publisher = {Springer},
+  year = 2012,
+  editor = { Youssef Hamadi  and  Marc Schoenauer },
+  booktitle = {Learning and Intelligent Optimization, 6th International
+                  Conference, LION 6},
+  author = {Yun, Xi and Epstein, Susan L.},
+  title = {Learning Algorithm Portfolios for Parallel Execution},
+  pages = {323--338},
+  doi = {10.1007/978-3-642-34413-8_23}
+}
+
+ +
+@incollection{ZaeStoBar2014:ppsn,
+  year = 2014,
+  volume = 8672,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  author = { Martin Zaefferer  and  J. Stork  and  Thomas Bartz-Beielstein },
+  title = {Distance Measures for Permutations in Combinatorial Efficient
+                  Global Optimization},
+  pages = {373--383},
+  doi = {10.1007/978-3-319-10762-2_37},
+  keywords = {CEGO, Bayesian optimization}
+}
+
+ +
+@incollection{ZaeStoFriFisNauBar2014,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2014,
+  editor = {Christian Igel and Dirk V. Arnold},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2014},
+  author = { Martin Zaefferer  and  J. Stork  and  M. Friese  and  Andreas Fischbach  and  Boris Naujoks  and  Thomas Bartz-Beielstein },
+  title = {Efficient Global Optimization for Combinatorial Problems},
+  pages = {871--878},
+  doi = {10.1145/2576768.2598282},
+  keywords = {CEGO, Bayesian optimization},
+  annote = {Proposed CEGO algorithm}
+}
+
+ +
+@inproceedings{Zar2005,
+  year = 2005,
+  volume = 3569,
+  editor = {Bacchus, Fahiem and Walsh, Toby},
+  booktitle = {International Conference on Theory and Applications of Satisfiability Testing},
+  author = {Zarpas, Emmanuel},
+  title = {Benchmarking {SAT} solvers for bounded model checking},
+  pages = {340--354}
+}
+
+ +
+@incollection{ZhaGeoAna2013:gecco,
+  isbn = {978-1-4503-1963-8},
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2013,
+  editor = { Christian Blum  and  Alba, Enrique },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  author = {Tiantian Zhang and Michael Georgiopoulos and Georgios C. Anagnostopoulos},
+  title = {{S-Race}: A Multi-Objective Racing Algorithm},
+  pages = {1565--1572}
+}
+
+ +
+@incollection{ZhaGeoAna2015sprint,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = {Zhang, Tiantian and Georgiopoulos, Michael and
+                  Anagnostopoulos, Georgios C.},
+  title = {{SPRINT}: Multi-Objective Model Racing},
+  pages = {1383--1390},
+  numpages = 8,
+  doi = {10.1145/2739480.2754791},
+  keywords = {model selection, multi-objective optimization, racing
+                  algorithm, sequential probability ratio test},
+  annote = {Extended version published as \cite{ZhaGeoAna2016}}
+}
+
+ +
+@inproceedings{ZhaLiuLi2009moead,
+  address = {Piscataway, NJ},
+  publisher = {IEEE Press},
+  year = 2009,
+  booktitle = {Proceedings of  the 2009 Congress on Evolutionary Computation (CEC 2009)},
+  key = {IEEE CEC},
+  author = { Zhang, Qingfu  and Wudong Liu and Hui Li},
+  title = {The Performance of a New Version of {MOEA/D} on {CEC09}
+                  Unconstrained {MOP} Test Instances},
+  pages = {203--208}
+}
+
+ +
+@techreport{ZhaZhoZha2009cec,
+  author = { Zhang, Qingfu  and A. Zhou and S. Zhao and  Ponnuthurai N. Suganthan  and W. Liu and S. Tiwari},
+  title = {Multiobjective Optimization Test Instances for the {CEC} 2009
+                  Special Session and Competition},
+  institution = {School of Computer Science and Electronic Engieering,
+                  University of Essex},
+  type = {Working Report},
+  year = 2009,
+  number = {CES-487},
+  month = apr,
+  annote = {Proposed UF benchmark}
+}
+
+ +
+@misc{Zhang09moeacomp,
+  author = { Zhang, Qingfu  and  Ponnuthurai N. Suganthan },
+  title = {Special Session on Performance Assessment of
+                  Multiobjective Optimization Algorithms/{CEC}'09
+                  {MOEA} Competition},
+  howpublished = {\url{https://www3.ntu.edu.sg/home/epnsugan/index_files/CEC09-MOEA/CEC09-MOEA.htm}},
+  annote = {Previously available at \url{http://dces.essex.ac.uk/staff/qzhang/moeacompetition09.htm}},
+  year = 2009
+}
+
+ +
+@inproceedings{ZilIshMos2016mut,
+  year = 2016,
+  booktitle = {Computational Intelligence (SSCI), 2016 IEEE Symposium Series
+                  on},
+  editor = {Chen, Xuewen and Stafylopatis, Andreas},
+  author = {Zille, Heiner and  Ishibuchi, Hisao  and  Mostaghim, Sanaz  and Nojima, Yusuke},
+  title = {Mutation operators based on variable grouping for
+                  multi-objective large-scale optimization},
+  pages = {1--8},
+  doi = {10.1109/SSCI.2016.7850214},
+  annote = {linked polynomial mutation}
+}
+
+ +
+@incollection{ZitBroThi2007emo,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 4403,
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  editor = {S. Obayashi and others},
+  year = 2007,
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  author = { Eckart Zitzler  and  Dimo Brockhoff  and  Lothar Thiele },
+  title = {The Hypervolume Indicator Revisited: On the Design
+                  of {Pareto}-compliant Indicators Via Weighted
+                  Integration},
+  pages = {862--876},
+  doi = {10.1007/978-3-540-70928-2_64},
+  supplement = {https://sop.tik.ee.ethz.ch/download/supplementary/weightedHypervolume/}
+}
+
+ +
+@incollection{ZitKnoThi2008quality,
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski },
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  volume = 5252,
+  year = 2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  author = { Eckart Zitzler  and  Joshua D. Knowles  and  Lothar Thiele },
+  title = {Quality Assessment of {Pareto} Set Approximations},
+  pages = {373--404},
+  alias = {ZKT2008},
+  doi = {10.1109/TEVC.2009.2016569}
+}
+
+ +
+@incollection{ZitKun2004ppsn,
+  year = 2004,
+  volume = 3242,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Xin Yao  and others},
+  aeditor = { Xin Yao  and  Edmund K. Burke  and  Jos{\'e} A. Lozano  and Smith,
+                  Jim and Merelo-Guerv{\'o}s, Juan Juli{\'a}n and Bullinaria,
+                  John A.  and Rowe, Jonathan E.  and Ti{\v{n}}o, Peter and
+                  Kab{\'a}n, Ata and Schwefel, Hans-Paul},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VIII}},
+  author = { Eckart Zitzler  and  Simon K{\"u}nzli},
+  title = {Indicator-based Selection in Multiobjective Search},
+  pages = {832--842},
+  keywords = {IBEA}
+}
+
+ +
+@techreport{ZitLauThi2001tr,
+  author = { Eckart Zitzler  and  Marco Laumanns  and  Lothar Thiele },
+  title = {{SPEA2}: Improving the Strength {Pareto} Evolutionary
+                  Algorithm},
+  institution = {Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Z{\"u}rich, Switzerland},
+  year = 2001,
+  number = 103,
+  annote = {Published as \cite{ZitLauThie2002spea2}}
+}
+
+ +
+@incollection{ZitLauThie2002spea2,
+  isbn = {84-89925-97-6},
+  shorteditor = {K. C. Giannakoglou and others},
+  booktitle = {Evolutionary Methods for Design, Optimisation and Control},
+  year = 2002,
+  publisher = {CIMNE, Barcelona, Spain},
+  editor = {K. C. Giannakoglou and D. T. Tsahalis and J. Periaux and
+                  K. D. Papaliliou and T. Fogarty},
+  author = { Eckart Zitzler  and  Marco Laumanns  and  Lothar Thiele },
+  title = {{SPEA2}: Improving the Strength {Pareto} Evolutionary
+                  Algorithm for Multiobjective Optimization},
+  pages = {95--100},
+  annote = {Proposed SPEA2.},
+  alias = {ZLT2002a}
+}
+
+ +
+@incollection{ZitThi1998ppsn,
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Agoston E. Eiben  and  Thomas B{\"a}ck  and  Marc Schoenauer  and  Hans-Paul Schwefel },
+  volume = 1498,
+  series = {Lecture Notes in Computer Science},
+  year = 1998,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {V}},
+  author = { Eckart Zitzler  and  Lothar Thiele },
+  title = {Multiobjective Optimization Using Evolutionary Algorithms -
+                  {A} Comparative Case Study},
+  pages = {292--301},
+  alias = {Zitzler98},
+  doi = {10.1007/BFb0056872},
+  annote = {Proposed hypervolume measure}
+}
+
+ +
+@incollection{ZitThiBad2008ppsn,
+  year = 2008,
+  volume = 5199,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { G{\"u}nther Rudolph  and others},
+  aeditor = { G{\"u}nther Rudolph  and Thomas Jansen and Simon Lucas and
+                  Carlo Poloni and Nicola Beume},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {X}},
+  author = { Eckart Zitzler  and  Lothar Thiele  and  Johannes Bader },
+  title = {{SPAM}: {Set} Preference Algorithm for
+                  Multiobjective Optimization},
+  pages = {847--858}
+}
+
+ +
+@incollection{TiwKochFad2008amga,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2008,
+  editor = {Conor Ryan},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  title = {{AMGA}: An archive-based micro genetic algorithm for
+                  multi-objective optimization},
+  author = {Tiwari, Santosh and Koch, Patrick and Fadel, Georges and  Kalyanmoy Deb },
+  pages = {729--736},
+  doi = {10.1145/1389095.1389235}
+}
+
+ +
+@phdthesis{ZitzlerPhD,
+  title = {Evolutionary Algorithms for Multiobjective Optimization:
+                  Methods and Applications},
+  author = { Eckart Zitzler },
+  school = {ETH Z{\"u}rich, Switzerland},
+  year = 1999,
+  atype = {{Ph.D.} thesis}
+}
+
+ +
+@inproceedings{ZujEid2011newdm,
+  title = {New decision maker model for multiobjective optimization
+                  interactive methods},
+  author = {Zujevs, Andrejs and Eiduks, Janis},
+  booktitle = {17th International Conference on Information and Software
+                  Technologies, Kaunas, Lithuania},
+  year = 2011,
+  keywords = {Machine Decision Maker},
+  pages = {51--58},
+  epub = {https://isd.ktu.lt/it2011/material/menu/proceedings.html}
+}
+
+ +
+@incollection{ants2008-cminer,
+  volume = 5217,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  publisher = {Springer},
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Christian Blum  and  Clerc, Maurice  and  Thomas St{\"u}tzle  and A. F. T. Winfield},
+  year = 2008,
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 6th
+                  International Conference, ANTS 2008},
+  author = {F. E. B. Otero and A. A. Freitas and C. G. Johnson},
+  title = {{cAnt-Miner}: An Ant Colony Classification Algorithm
+                  to Cope with Continuous Attributes},
+  pages = {48--59}
+}
+
+ +
+@incollection{dePDoeDoe2015:gecco,
+  address = { New York, NY},
+  publisher = {ACM Press},
+  year = 2015,
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  author = {Axel {de Perthuis de Laillevault} and  Benjamin Doerr  and  Carola Doerr },
+  title = {Money for Nothing: Speeding Up Evolutionary Algorithms
+                  Through Better Initialization},
+  pages = {815--822}
+}
+
+ +
+@misc{nevergrad,
+  author = {Jeremy Rapin and Olivier Teytaud },
+  title = {Nevergrad: A gradient-free optimization platform},
+  year = 2018,
+  publisher = {GitHub},
+  journal = {GitHub repository},
+  howpublished = {\url{https://github.com/FacebookResearch/Nevergrad}}
+}
+
+ +
+@misc{oscar,
+  author = {{OscaR Team}},
+  title = {{OscaR}: {Scala} in {OR}},
+  year = 2012,
+  note = {Available from \url{https://bitbucket.org/oscarlib/oscar}}
+}
+
+ +
+@misc{poliastro,
+  author = {Cano Rodr{\'i}guez, Juan Luis and others},
+  title = {poliastro: Astrodynamics in {Python}},
+  year = 2015,
+  howpublished = {Zenodo},
+  doi = {10.5281/zenodo.593610}
+}
+
+ +
+@misc{url:ACOVEA,
+  author = {Scott Robert Ladd},
+  title = {{ACOVEA} ({Analysis} of Compiler Options via Evolutionary Algorithm)},
+  howpublished = {\url{https://github.com/Acovea/libacovea}},
+  year = 2000
+}
+
+ +
+@misc{url:GCC,
+  author = {{GNU Project, Free Software Foundation}},
+  title = {{GCC}, the {GNU} Compiler Collection},
+  howpublished = {\url{https://gcc.gnu.org}},
+  year = 1987
+}
+
+ +
+@misc{url:GGA,
+  author = { Carlos Ans{\'o}tegui  and  Meinolf Sellmann  and  Kevin Tierney },
+  title = {{GGA}: Gender-based Genetic Algorithm Configurator},
+  note = {Version visited last on July 2017},
+  howpublished = {\url{https://bitbucket.org/gga_ac/}},
+  year = 2017
+}
+
+ +
+@misc{url:ISAT,
+  author = {{Intel}},
+  title = {{Intel} Software Autotuning Tool},
+  howpublished = {\url{https://software.intel.com/en-us/articles/intel-software-autotuning-tool/}},
+  year = 2010
+}
+
+ +
+@misc{url:MOEAD,
+  author = { Zhang, Qingfu },
+  title = {{MOEA/D} homepage},
+  howpublished = {\url{https://sites.google.com/view/moead/}},
+  year = 2007,
+  annote = {Previous URL was at
+                  \url{https://dces.essex.ac.uk/staff/zhang/webofmoead.htm}.},
+  alias = {MOEAD2007}
+}
+
+ +
+@misc{url:SMAC3,
+  title = {{SMAC} v3 Project},
+  author = {{ML4AAD Group}},
+  howpublished = {\url{https://github.com/automl/SMAC3}},
+  note = {Version visited last on August 2017},
+  year = 2017
+}
+
+ +
+@misc{url:TSPLIB,
+  author = { Gerhard Reinelt },
+  title = {{TSPLIB}},
+  howpublished = {\url{http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/}},
+  note = {Version visited last on 24 February 2023},
+  year = 1995
+}
+
+ +
+@misc{url:TSPsite,
+  author = { William J. Cook },
+  title = {The Traveling Salesman Problem},
+  howpublished = {\url{http://www.math.uwaterloo.ca/tsp}},
+  note = {Version visited last on 15 April 2014},
+  year = 2010
+}
+
+ +
+@misc{url:paramils,
+  author = { Frank Hutter  and  Holger H. Hoos  and  Kevin Leyton-Brown  and  Thomas St{\"u}tzle },
+  title = {{ParamILS}},
+  howpublished = {\url{http://www.cs.ubc.ca/labs/beta/Projects/ParamILS/}},
+  note = {Version visited last on July 2017},
+  year = 2017
+}
+
+ +
+@phdthesis{vanZyl:PhD,
+  author = { Jakobus E. van Zyl },
+  title = {A Methodology for Improved Operational Optimization of Water
+                  Distribution Systems},
+  school = {School of Engineering and Computer Science, University of
+                  Exeter, UK},
+  year = 2001
+}
+
+ +
+@proceedings{AAAI1988,
+  editor = {Howard E. Shrobe and Tom M. Mitchell and Reid G. Smith},
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  url = {http://www.aaai.org/Conferences/AAAI/aaai88.php},
+  title = {Proceedings of the 7th National Conference on Artificial
+                  Intelligence, St. Paul, MN, August 21-26, AAAI-88},
+  booktitle = {Proceedings of the 7th National Conference on Artificial
+                  Intelligence, AAAI-88},
+  year = 1988
+}
+
+ +
+@proceedings{AAAI1992,
+  title = {Proceedings of  the 10th National Conference on Artificial Intelligence},
+  booktitle = {Proceedings of  the 10th National Conference on Artificial Intelligence},
+  year = 1992,
+  editor = {William R. Swartout},
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI1993,
+  title = {Proceedings of  the 11th National Conference on Artificial Intelligence},
+  booktitle = {Proceedings of  the 11th National Conference on Artificial Intelligence},
+  year = 1993,
+  editor = {Richard Fikes and Wendy G. Lehnert},
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI1997,
+  editor = {Benjamin Kuipers and Bonnie L. Webber},
+  title = {Proceedings of the Fourteenth National Conference on
+                  Artificial Intelligence and Ninth Innovative Applications of
+                  Artificial Intelligence Conference, {AAAI} 97, {IAAI} 97,
+                  July 27-31, 1997, Providence, Rhode Island},
+  booktitle = {Proceedings of AAAI 1997 -- Fourteenth National Conference on
+                  Artificial Intelligence},
+  year = 1997,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI1998,
+  editor = {Jack Mostow and Chuck Rich},
+  title = {Proceedings of the Fifteenth National Conference on
+                  Artificial Intelligence and Tenth Innovative Applications of
+                  Artificial Intelligence Conference, {AAAI} 98, {IAAI} 98,
+                  July 26-30, 1998, Madison, Wisconsin, {USA}},
+  booktitle = {Proceedings of AAAI 1998 -- Fifteenth National Conference on
+                  Artificial Intelligence},
+  year = 1998,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI2000,
+  editor = {Henry A. Kautz and Bruce W. Porter},
+  title = {Proceedings of the Seventeenth National Conference on
+                  Artificial Intelligence and Twelfth Conference on on
+                  Innovative Applications of Artificial Intelligence, July 30 --
+                  August 3, 2000, Austin, Texas, {USA}},
+  booktitle = {Proceedings of AAAI 2000 -- Seventeenth National Conference
+                  on Artificial Intelligence},
+  year = 2000,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI2006,
+  title = {Proceedings, The Twenty-First National Conference on
+                  Artificial Intelligence and the Eighteenth Innovative
+                  Applications of Artificial Intelligence Conference, July
+                  16-20, 2006, Boston, Massachusetts, {USA}},
+  booktitle = {Proceedings of  the 21st National Conference on Artificial Intelligence},
+  volume = 6,
+  year = 2006,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA},
+  editor = {Anthony Cohn}
+}
+
+ +
+@proceedings{AAAI2007,
+  title = {Proceedings of the Twenty-Second {AAAI} Conference on
+                  Artificial Intelligence, July 22-26, 2007, Vancouver, British
+                  Columbia, Canada},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Robert C. Holte and Adele Howe},
+  year = 2007,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{AAAI2010,
+  editor = {Maria Fox and David Poole},
+  title = {Proceedings of the Twenty-Fourth AAAI Conference on Artificial
+               Intelligence, AAAI 2010, Atlanta, Georgia, USA, July 11-15,
+               2010},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2010
+}
+
+ +
+@proceedings{AAAI2011,
+  editor = {Wolfram Burgard and Dan Roth},
+  title = {Proceedings of the Twenty-Fifth AAAI Conference on Artificial
+               Intelligence, AAAI 2011, San Francisco, California, USA, August 07-11,
+               2011},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2011
+}
+
+ +
+@proceedings{AAAI2012,
+  editor = {Jorg Hoffmann and Bart Selman},
+  title = {Proceedings of the Twenty-Sixth AAAI Conference on Artificial
+               Intelligence, AAAI 2012, Toronto, Ontario, Canada, July 22-26,
+               2012},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2012
+}
+
+ +
+@proceedings{AAAI2014,
+  editor = {David Stracuzzi and others},
+  title = {Proceedings of the Twenty-Eighth AAAI Conference on
+                  Artificial Intelligence, AAAI 2014, Qu{\'e}bec City,
+                  Qu{\'e}bec, Canada, July 27-31, 2014},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2014
+}
+
+ +
+@proceedings{AAAI2015,
+  editor = {Blai Bonet and Sven Koenig},
+  title = {Proceedings of the Twenty-Ninth AAAI Conference on Artificial
+               Intelligence, AAAI 2015, Austin, Texas, USA, January 25-30,
+               2015},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2015
+}
+
+ +
+@book{AAAI2016,
+  editor = {Dale Schuurmans and Michael P. Wellman},
+  title = {Proceedings of the Thirtieth {AAAI} Conference on Artificial
+                  Intelligence, AAAI 2016, February 12-17, 2016, Phoenix,
+                  Arizona, {USA.}},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  publisher = {{AAAI} Press},
+  year = 2016,
+  epub = {http://www.aaai.org/Library/AAAI/aaai16contents.php}
+}
+
+ +
+@book{AAAI2017,
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Satinder P. Singh and Shaul Markovitch},
+  title = {Proceedings of the Thirty-First {AAAI} Conference on
+                  Artificial Intelligence, February 4-9, 2017, San Francisco,
+                  California, {USA}},
+  year = 2017,
+  month = feb,
+  publisher = {{AAAI} Press}
+}
+
+ +
+@book{AAAI2018,
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  editor = {Sheila A. McIlraith and Kilian Q. Weinberger},
+  title = {Proceedings of the Thirty-Second {AAAI} Conference on
+                  Artificial Intelligence, February 2-7, 2018, New Orleans,
+                  Louisiana, {USA}},
+  year = 2018,
+  month = feb,
+  publisher = {{AAAI} Press}
+}
+
+ +
+@proceedings{AAAI2020,
+  key = {AAAI2020},
+  booktitle = {Proceedings of  the {AAAI} Conference on Artificial Intelligence},
+  year = 2020,
+  title = {The Thirty-Fourth {AAAI} Conference on Artificial
+                  Intelligence, {AAAI} 2020, The Thirty-Second Innovative
+                  Applications of Artificial Intelligence Conference, {IAAI}
+                  2020, The Tenth {AAAI} Symposium on Educational Advances in
+                  Artificial Intelligence, {EAAI} 2020, New York, NY, USA,
+                  February 7-12, 2020},
+  isbn = {978-1-57735-823-7},
+  publisher = {{AAAI} Press}
+}
+
+ +
+@book{ACAL2007,
+  booktitle = {Progress in Artificial Life (ACAL)},
+  title = {Progress in Artificial Life, Third Australian
+                  Conference, {ACAL} 2007},
+  year = 2007,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 4828,
+  editor = {Marcus Randall and  Abbass, Hussein A.  and Janet Wiles}
+}
+
+ +
+@proceedings{ACC2019,
+  key = {ACC2019},
+  booktitle = {2019 American Control Conference ({ACC})},
+  title = {2019 American Control Conference, {ACC} 2019, Philadelphia,
+                  PA, USA, July 10-12, 2019},
+  publisher = {{IEEE}},
+  year = 2019
+}
+
+ +
+@book{ADT2009,
+  title = {Algorithmic Decision Theory, First International
+                  Conference, {ADT} 2009, Venice, Italy, October
+                  20-23, 2009},
+  booktitle = {Algorithmic Decision Theory, First International
+                  Conference, {ADT} 2009},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5783,
+  editor = { Francesca Rossi  and  Alexis Tsouki{\`a}s },
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{ADT2011,
+  title = {Algorithmic Decision Theory, Third International Conference,
+                  {ADT} 2011, Piscataway, New Jersey, USA, October 26-28, 2011},
+  booktitle = {Algorithmic Decision Theory, Third International Conference,
+                  {ADT} 2011},
+  year = 2011,
+  series = {Lecture Notes in Artificial Intelligence},
+  volume = 6992,
+  editor = { Ronen I. Brafman  and  F. Roberts  and  Alexis Tsouki{\`a}s },
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@proceedings{AIM1998,
+  editor = {Martin C. Golumbic and others},
+  booktitle = {Fifth International Symposium on Artificial Intelligence and Mathematics,
+                  {AIM} 1998, Fort Lauderdale, Florida, USA, January 4-6, 1998},
+  title = {Fifth International Symposium on Artificial Intelligence and Mathematics,
+                  {AIM} 1998, Fort Lauderdale, Florida, USA, January 4-6, 1998},
+  year = 1998
+}
+
+ +
+@book{AIPIS2016,
+  title = {Artificial Intelligence Perspectives in Intelligent Systems},
+  booktitle = {Artificial Intelligence Perspectives in Intelligent Systems},
+  year = 2016,
+  editor = {Silhavy, Radek and Senkerik, Roman and Oplatkova, Zuzana Kominkova and Silhavy, Petr and Prokopova, Zdenka},
+  publisher = {Springer International Publishing},
+  series = {Advances in Intelligent Systems and Computing},
+  volume = 464
+}
+
+ +
+@book{AISB1995,
+  editor = {T. C. Fogarty},
+  title = {Evolutionary Computing, {AISB} Workshop, Sheffield, UK, April
+                  3-4, 1995, Selected Papers},
+  publisher = {Springer},
+  year = 1995,
+  volume = 993,
+  series = {Lecture Notes in Computer Science},
+  address = { Berlin, Germany},
+  booktitle = {Evolutionary Computing, AISB Workshop}
+}
+
+ +
+@proceedings{AISTATS2016,
+  editor = {Arthur Gretton and Christian C. Robert},
+  title = {Proceedings of the 19th International Conference on Artificial Intelligence
+                  and Statistics, {AISTATS} 2016, Cadiz, Spain, May 9-11, 2016},
+  booktitle = {Proceedings of the 19th International Conference on Artificial Intelligence
+                  and Statistics, {AISTATS} 2016, Cadiz, Spain, May 9-11, 2016},
+  year = 2016,
+  series = {{JMLR} Workshop and Conference Proceedings},
+  volume = 51,
+  publisher = {{JMLR}.org}
+}
+
+ +
+@book{ALT2013,
+  editor = {Sanjay Jain and R{\'{e}}mi Munos and Frank Stephan and Thomas
+                  Zeugmann},
+  title = {Algorithmic Learning Theory - 24th International Conference,
+                  {ALT} 2013, Singapore, October 6-9, 2013. Proceedings},
+  booktitle = {Proceedings of Algorithmic Learning Theory},
+  series = {Lecture Notes in Computer Science},
+  volume = 8139,
+  year = 2013,
+  doi = {10.1007/978-3-642-40935-6},
+  publisher = {Springer},
+  address = { Berlin, Germany}
+}
+
+ +
+@book{AMONIC2010,
+  booktitle = {Advances in Multi-Objective Nature Inspired Computing},
+  title = {Advances in Multi-Objective Nature Inspired Computing},
+  series = {Studies in Computational Intelligence},
+  year = 2010,
+  volume = 272,
+  editor = { Carlos A. {Coello Coello}  and  Dhaenens, Clarisse  and  Laetitia Jourdan },
+  publisher = {Springer}
+}
+
+ +
+@book{ANIMATS1994,
+  editor = {Cliff, D. and Husbands, P. and Meyer, J.-A. and Wilson, S.},
+  title = {Proceedings of the third international conference on
+                  Simulation of adaptive behavior: From Animals to Animats 3},
+  booktitle = {Proceedings of the third international conference on
+                  Simulation of adaptive behavior: From Animals to Animats 3},
+  publisher = {MIT Press},
+  year = 1994,
+  address = {Cambridge, MA}
+}
+
+ +
+@proceedings{ANTS2000,
+  title = {Abstract proceedings of ANTS 2000 -- From Ant
+                  Colonies to Artificial Ants: Second International
+                  Workshop on Ant Algorithms},
+  booktitle = {Abstract proceedings of ANTS 2000 -- From Ant
+                  Colonies to Artificial Ants: Second International
+                  Workshop on Ant Algorithms},
+  fulleditor = { Marco Dorigo  and  Martin Middendorf  and  Thomas St{\"u}tzle },
+  editor = { Marco Dorigo  and others },
+  organization = {IRIDIA, Universit{\'e} Libre de Bruxelles, Belgium},
+  date = {2000-09-07/2000-09-09},
+  year = 2000,
+  month = sep # { 7--9}
+}
+
+ +
+@book{ANTS2002,
+  title = {Ant Algorithms, Third International Workshop, ANTS
+                  2002, Brussels, Belgium, September 12-14, 2002,
+                  Proceedings},
+  booktitle = {Ant Algorithms, Third International Workshop, ANTS
+                  2002},
+  year = 2002,
+  series = {Lecture Notes in Computer Science},
+  volume = 2463,
+  fulleditor = { Marco Dorigo  and  Gianni A. {Di Caro}  and  M. Sampels },
+  editor = { Marco Dorigo  and others },
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{ANTS2004,
+  title = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004},
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 4th
+                  International Workshop, ANTS 2004 },
+  year = 2004,
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Francesco Mondada  and  Thomas St{\"u}tzle  and  Mauro Birattari  and  Christian Blum },
+  editor = { Marco Dorigo  and others },
+  volume = 3172,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{ANTS2006,
+  title = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 5th
+                  International Workshop, ANTS 2006},
+  fulleditor = { Marco Dorigo  and  L. M. Gambardella  and  Mauro Birattari  and 
+                  Martinoli, A. and  Poli, R.  and  Thomas St{\"u}tzle },
+  editor = { Marco Dorigo  and others },
+  series = {Lecture Notes in Computer Science},
+  volume = 4150,
+  year = 2006,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{ANTS2008,
+  title = {Ant Colony Optimization and Swarm Intelligence, 6th
+                  International Conference, ANTS 2008},
+  booktitle = {Ant Colony Optimization and Swarm Intelligence, 6th
+                  International Conference, ANTS 2008},
+  year = 2008,
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Christian Blum  and  Clerc, Maurice  and  Thomas St{\"u}tzle  and A. F. T. Winfield},
+  editor = { Marco Dorigo  and others },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 5217
+}
+
+ +
+@book{ANTS2010,
+  title = {Ant Colony Optimization and Swarm Intelligence, 7th
+                  International Conference, ANTS 2010},
+  booktitle = {Swarm Intelligence, 7th International Conference, ANTS 2010},
+  year = 2010,
+  editor = { Marco Dorigo  and others },
+  fulleditor = { Marco Dorigo  and  Mauro Birattari  and  Gianni A. {Di Caro}  and Doursat, R. and Engelbrecht, A. P. and Floreano,
+                  D. and Gambardella, L. M. and Gro\ss, R. and Sahin,
+                  E. and  Thomas St{\"u}tzle  and Sayama, H.},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 6234
+}
+
+ +
+@book{ANTS2012,
+  title = {Swarm Intelligence, 8th
+                  International Conference, ANTS 2012},
+  booktitle = {Swarm Intelligence, 8th International Conference, ANTS 2012},
+  year = 2012,
+  editor = { Marco Dorigo  and others },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7461
+}
+
+ +
+@book{ANTS2014,
+  title = {Swarm Intelligence, 9th
+                  International Conference, ANTS 2014},
+  booktitle = {Swarm Intelligence, 9th International Conference, ANTS 2014},
+  year = 2014,
+  editor = { Marco Dorigo  and others },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 8667
+}
+
+ +
+@book{ANTS2016,
+  title = {Swarm Intelligence, 10th International Conference, ANTS 2016,
+                  Brussels, Belgium, September 7-9, 2016, Proceedings},
+  booktitle = {Swarm Intelligence, 10th International Conference, ANTS 2016},
+  year = 2016,
+  editor = { Marco Dorigo  and  Mauro Birattari  and  Li, Xiaodong  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Kazuhiro Ohkura  and  Carlo Pinciroli  and  Thomas St{\"u}tzle },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9882,
+  doi = {10.1007/978-3-319-44427-7}
+}
+
+ +
+@book{ANTS2018,
+  title = {Swarm Intelligence, 11th International Conference, ANTS 2018,
+                  Rome, Italy, October 29-31, 2018, Proceedings},
+  booktitle = {Swarm Intelligence, 11th International Conference, ANTS 2018},
+  year = 2018,
+  editor = { Marco Dorigo  and  Mauro Birattari  and Christensen, Anders L. and Reina, Andreagiovanni and  Vito Trianni },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 11172
+}
+
+ +
+@book{ANTS2020,
+  title = {Swarm Intelligence, 12th International Conference, ANTS 2020,
+                  Barcelona, Spain, October 26-28, 2020, Proceedings},
+  booktitle = {Swarm Intelligence, 12th International Conference, ANTS 2020},
+  year = 2020,
+  editor = { Marco Dorigo  and  Thomas St{\"u}tzle  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Heiko Hamann  and Heinrich, Mary Katherine},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 12421
+}
+
+ +
+@book{ANTS2022,
+  title = {Swarm Intelligence, 13th International Conference, ANTS 2022,
+                  M\'alaga, Spain, November 2-4, 2022, Proceedings},
+  booktitle = {Swarm Intelligence, 13th International Conference, ANTS 2022},
+  year = 2022,
+  editor = { Marco Dorigo  and  Heiko Hamann  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Jos{\'e} Garc{\'i}a-Nieto  and  Andries Engelbrecht  and  Carlo Pinciroli  and  Volker Strobel  and Camacho-Villal\'{o}n, Christian Leonardo},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = 13491,
+  doi = {10.1007/978-3-031-20176-9}
+}
+
+ +
+@book{AUTSEA2011,
+  editor = { Youssef Hamadi  and E. Monfroy and F. Saubion},
+  title = {Autonomous Search},
+  booktitle = {Autonomous Search},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2012
+}
+
+ +
+@book{AWSM03,
+  editor = { C. Maksimovi{\'c}  and  David Butler  and  Fayyaz Ali Memon },
+  title = {Advances in Water Supply Management: Proceedings of the CCWI
+                  '03 Conference, London, 15-17 September 2003},
+  booktitle = {Advances in Water Supply Management},
+  year = 2003,
+  publisher = {CRC Press}
+}
+
+ +
+@book{AarLen97,
+  title = {Local Search in Combinatorial Optimization},
+  booktitle = {Local Search in Combinatorial Optimization},
+  publisher = {John Wiley \& Sons},
+  address = { Chichester, UK},
+  year = 1997,
+  editor = { Emile H. L. Aarts  and  Jan Karel Lenstra }
+}
+
+ +
+@book{AbrJaiGol2005emo,
+  title = {Evolutionary Multiobjective Optimization},
+  booktitle = {Evolutionary Multiobjective Optimization},
+  publisher = {Springer},
+  series = {Advanced Information and Knowledge Processing},
+  editor = {Abraham, Ajith and Jain, Lakhmi and Goldberg, Robert},
+  month = jan,
+  year = 2005,
+  address = { London, UK }
+}
+
+ +
+@book{AdvDE2008,
+  title = {Advances in differential evolution},
+  booktitle = {Advances in differential evolution},
+  year = 2008,
+  editor = {Uday K. Chakraborty},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@proceedings{BIOMA2004,
+  title = {Bioinspired optimization methods and their applications:
+                  Proceedings of the International Conference on Bioinspired
+                  Optimization Methods and their Applications - BIOMA 2004,
+                  11-12 October 2004, Ljubljana, Slovenia},
+  booktitle = {International Conference on Bioinspired Optimization Methods
+                  and their Applications (BIOMA 2004)},
+  year = 2004,
+  editor = {Bogdan Filipi{\v c} and  Jurij {\v S}ilc },
+  url = {https://books.google.be/books?id=0ZLsAAAACAAJ}
+}
+
+ +
+@proceedings{BNAIC2020,
+  title = {Proceedings of the 32nd Benelux Conference on Artificial Intelligence,
+                  BNAIC 2020, Leiden, The Netherlands, 19-20 November 2020},
+  booktitle = {Proceedings of the 32nd Benelux Conference on Artificial Intelligence,
+                  BNAIC 2020, Leiden, The Netherlands, 19-20 November 2020},
+  editor = {Cao, Lu and Kosters, Walter and Lijffijt, Jefrey},
+  year = 2020,
+  url = {https://bnaic.liacs.leidenuniv.nl/wordpress/wp-content/uploads/bnaic2020proceedings.pdf}
+}
+
+ +
+@book{BarChiPaqPre2010emaoa,
+  title = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  booktitle = {Experimental Methods for the Analysis of
+                  Optimization Algorithms},
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2010,
+  editor = { Thomas Bartz-Beielstein  and  Marco Chiarandini  and  Lu{\'i}s Paquete  and  Mike Preuss }
+}
+
+ +
+@book{BarFilKorTal2020high,
+  booktitle = {High-Performance Simulation-Based Optimization},
+  title = {High-Performance Simulation-Based Optimization},
+  year = 2020,
+  editor = { Thomas Bartz-Beielstein  and Bogdan Filipi{\v c} and  P. Koro{\v s}ec  and  Talbi, El-Ghazali },
+  publisher = {Springer International Publishing},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{BluBleRol2008hybrid,
+  title = {Hybrid Metaheuristics: An emergent approach for optimization},
+  booktitle = {Hybrid Metaheuristics: An emergent approach for optimization},
+  editor = { Christian Blum  and  Mar{\'i}a J. Blesa  and  Andrea Roli  and  M. Sampels },
+  publisher = {Springer},
+  address = { Berlin, Germany},
+  year = 2008,
+  volume = 114,
+  series = {Studies in Computational Intelligence},
+  alias = {BluEtAl08:hm-book}
+}
+
+ +
+@book{BorMor2014theory,
+  editor = {Borenstein, Yossi and  A. Moraglio },
+  title = {Theory and Principled Methods for the Design of
+                  Metaheuristics},
+  booktitle = {Theory and Principled Methods for the Design of
+                  Metaheuristics},
+  series = {Natural Computing Series},
+  year = 2014,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg}
+}
+
+ +
+@book{CAEPIA2015,
+  title = {Advances in Artificial Intelligence: 16th Conference of the
+                  Spanish Association for Artificial Intelligence, CAEPIA 2015
+                  Albacete, Spain, November 9-12, 2015 Proceedings},
+  booktitle = {Advances in Artificial Intelligence, CAEPIA 2015},
+  year = 2015,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  editor = {Puerta, Jos{\'e} M.  and G{\'a}mez, Jos{\'e} A.  and
+                  Dorronsoro, Bernabe and Barrenechea, Edurne and Troncoso,
+                  Alicia and Baruque, Bruno and Galar, Mikel},
+  series = {Lecture Notes in Computer Science},
+  volume = 9422
+}
+
+ +
+@proceedings{CCC1972,
+  title = {Complexity of Computer Computations},
+  booktitle = {Proceedings of a symposium on the Complexity of Computer Computations,
+                  held March 20-22, 1972, at the {IBM} {T}homas {J}. {W}atson Research Center,
+                  Yorktown Heights, New York, USA},
+  year = 1972,
+  editor = {Miller, Raymond E. and Thatcher, James, W.},
+  series = {The IBM Research Symposia Series},
+  publisher = {Springer}
+}
+
+ +
+@proceedings{CCIE2010,
+  key = {CCIE},
+  title = {Proceedings of the 2010 International Conference on
+                  Computing, Control and Industrial Engineering},
+  booktitle = {Proceedings of the 2010 International Conference on
+                  Computing, Control and Industrial Engineering},
+  year = 2010,
+  publisher = {IEEE Computer Society Press},
+  address = {Los Alamitos, CA}
+}
+
+ +
+@proceedings{CCWI2005,
+  title = {Proceedings of the Eighth International Conference on
+                  Computing and Control for the Water Industry (CCWI 2005)},
+  booktitle = {Proceedings of the Eighth International Conference on
+                  Computing and Control for the Water Industry (CCWI 2005)},
+  year = 2005,
+  editor = { Dragan A. Savic  and  Godfrey A. Walters  and  Roger King  and  Soon Thiam-Khu },
+  volume = 1,
+  address = {University of Exeter, UK},
+  month = sep
+}
+
+ +
+@proceedings{CEC1999,
+  key = {IEEE CEC},
+  title = {Proceedings of the 1999 Congress on Evolutionary Computation
+                  (CEC 1999)},
+  booktitle = {Proceedings of the 1999 Congress on Evolutionary Computation
+                  (CEC 1999)},
+  year = 1999,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2000,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2000 Congress on Evolutionary
+                  Computation (CEC 2000)},
+  booktitle = {Proceedings of  the 2000 Congress on Evolutionary Computation (CEC'00)},
+  year = 2000,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  month = jul
+}
+
+ +
+@proceedings{CEC2001,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2001 Congress on Evolutionary
+                  Computation (CEC 2001)},
+  booktitle = {Proceedings of  the 2001 Congress on Evolutionary Computation (CEC'01)},
+  year = 2001,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2002,
+  key = {IEEE CEC},
+  booktitle = {Proceedings of  the 2002 Congress on Evolutionary Computation (CEC'02)},
+  title = {Proceedings of  the 2002 Congress on Evolutionary Computation (CEC'02)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2002
+}
+
+ +
+@proceedings{CEC2003,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2003 Congress on Evolutionary
+                  Computation (CEC 2003), Canberra, ACT, Australia},
+  booktitle = {Proceedings of  the 2003 Congress on Evolutionary Computation (CEC'03)},
+  month = dec,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2003
+}
+
+ +
+@proceedings{CEC2004,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2004 Congress on Evolutionary
+                  Computation (CEC 2004)},
+  booktitle = {Proceedings of the 2004 Congress on Evolutionary
+                  Computation (CEC 2004)},
+  year = 2004,
+  month = sep,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2005,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2005 Congress on Evolutionary
+                  Computation (CEC 2005)},
+  booktitle = {Proceedings of  the 2005 Congress on Evolutionary Computation (CEC 2005)},
+  year = 2005,
+  month = sep,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2006,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2006 Congress on Evolutionary
+                  Computation (CEC 2006)},
+  booktitle = {Proceedings of  the 2006 Congress on Evolutionary Computation (CEC 2006)},
+  year = 2006,
+  month = jul,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2007,
+  key = {IEEE CEC},
+  title = {Proceedings of the {IEEE} Congress on Evolutionary
+                  Computation, {CEC} 2007, 25-28 September 2007, Singapore},
+  booktitle = {Proceedings of  the 2007 Congress on Evolutionary Computation (CEC 2007)},
+  year = 2007,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2008,
+  key = {IEEE CEC},
+  title = {Proceedings of the {IEEE} Congress on Evolutionary
+                  Computation, {CEC} 2008, June 1-6, 2008, Hong Kong, China},
+  booktitle = {Proceedings of  the 2008 Congress on Evolutionary Computation (CEC 2008)},
+  year = 2008,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2009,
+  key = {IEEE CEC},
+  title = {Proceedings of  the 2009 Congress on Evolutionary Computation (CEC 2009)},
+  booktitle = {Proceedings of  the 2009 Congress on Evolutionary Computation (CEC 2009)},
+  year = 2009,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{CEC2010,
+  key = {IEEE CEC},
+  editor = { Ishibuchi, Hisao  and others},
+  title = {Proceedings of the 2010 Congress on Evolutionary
+                  Computation (CEC 2010)},
+  booktitle = {Proceedings of  the 2010 Congress on Evolutionary Computation (CEC 2010)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2010
+}
+
+ +
+@proceedings{CEC2011,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2011 Congress on Evolutionary
+                  Computation (CEC 2011), New Orleans, LA, USA},
+  booktitle = {Proceedings of  the 2011 Congress on Evolutionary Computation (CEC 2011)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2011
+}
+
+ +
+@proceedings{CEC2012,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2012 Congress on Evolutionary
+                  Computation (CEC 2012)},
+  booktitle = {Proceedings of  the 2012 Congress on Evolutionary Computation (CEC 2012)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2012
+}
+
+ +
+@proceedings{CEC2013,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2013 Congress on Evolutionary
+                  Computation (CEC 2013)},
+  booktitle = {Proceedings of  the 2013 Congress on Evolutionary Computation (CEC 2013)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2013
+}
+
+ +
+@proceedings{CEC2014,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2014 Congress on Evolutionary
+                  Computation (CEC 2014)},
+  booktitle = {Proceedings of  the 2014 Congress on Evolutionary Computation (CEC 2014)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2014
+}
+
+ +
+@proceedings{CEC2015,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2015 Congress on Evolutionary
+                  Computation (CEC 2015)},
+  booktitle = {Proceedings of  the 2015 Congress on Evolutionary Computation (CEC 2015)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2015
+}
+
+ +
+@proceedings{CEC2016,
+  key = {IEEE CEC},
+  title = {{IEEE} Congress on Evolutionary Computation, {CEC} 2016,
+                  Vancouver, BC, Canada, July 24-29, 2016},
+  booktitle = {Proceedings of  the 2016 Congress on Evolutionary Computation (CEC 2016)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  isbn = {978-1-5090-0623-6},
+  year = 2016
+}
+
+ +
+@proceedings{CEC2017,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2017 Congress on Evolutionary
+                  Computation (CEC 2017)},
+  booktitle = {Proceedings of  the 2017 Congress on Evolutionary Computation (CEC 2017)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2017
+}
+
+ +
+@proceedings{CEC2018,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2018 Congress on Evolutionary
+                  Computation (CEC 2018)},
+  booktitle = {Proceedings of  the 2018 Congress on Evolutionary Computation (CEC 2018)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2018
+}
+
+ +
+@proceedings{CEC2019,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2019 Congress on Evolutionary
+                  Computation (CEC 2019)},
+  booktitle = {Proceedings of  the 2019 Congress on Evolutionary Computation (CEC 2019)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2019
+}
+
+ +
+@proceedings{CEC2020,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2020 Congress on Evolutionary
+                  Computation (CEC 2020)},
+  booktitle = {Proceedings of  the 2020 Congress on Evolutionary Computation (CEC 2020)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2020
+}
+
+ +
+@proceedings{CEC2021,
+  key = {IEEE CEC},
+  title = {Proceedings of the 2021 Congress on Evolutionary
+                  Computation (CEC 2021)},
+  booktitle = {Proceedings of  the 2021 Congress on Evolutionary Computation (CEC 2021)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2021
+}
+
+ +
+@proceedings{CGO2008,
+  editor = {Soffa, Mary Lou and Duesterwald, Evelyn},
+  title = {Proceedings of the 6th Annual IEEE/ACM International
+                  Symposium on Code Generation and Optimization},
+  booktitle = {Proceedings of the 6th Annual IEEE/ACM International
+                  Symposium on Code Generation and Optimization},
+  year = 2008,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  series = {CGO '08}
+}
+
+ +
+@book{CO:MA2011,
+  title = {Computational Optimization, Methods and Algorithms},
+  booktitle = {Computational Optimization, Methods and Algorithms},
+  series = {Studies in Computational Intelligence},
+  volume = 356,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  editor = {Slawomir Koziel and Xin-She Yang},
+  year = 2011
+}
+
+ +
+@proceedings{COLT1992,
+  editor = {David Haussler},
+  title = {Proceedings of the Fifth Annual {ACM} Conference on
+                  Computational Learning Theory, {COLT} 1992, Pittsburgh, PA,
+                  USA, July 27-29, 1992},
+  booktitle = {COLT'92},
+  publisher = {ACM Press},
+  year = 1992
+}
+
+ +
+@book{CP1998,
+  year = 1998,
+  title = {Principles and Practice of Constraint Programming, CP98},
+  booktitle = {Principles and Practice of Constraint Programming, CP98},
+  volume = 1520,
+  series = {Lecture Notes in Computer Science},
+  editor = {Maher, Michael and Puget, Jean-Francois},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{CP2000,
+  editor = {Rina Dechter},
+  title = {Principles and Practice of Constraint Programming,
+                  CP 2000, 6th International Conference, Singapore,
+                  September 18-21, 2000, Proceedings},
+  booktitle = {Principles and Practice of Constraint Programming,
+                  CP 2000},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 1894,
+  year = 2000
+}
+
+ +
+@book{CP2002,
+  editor = {van Hentenryck, Pascal },
+  title = {Principles and Practice of Constraint Programming, CP 2002},
+  booktitle = {Principles and Practice of Constraint Programming, CP 2002},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  year = 2002
+}
+
+ +
+@book{CP2009,
+  editor = { Ian P. Gent },
+  title = {Principles and Practice of Constraint Programming --
+                  CP 2009, 15th International Conference, CP 2009,
+                  Lisbon, Portugal, September 20-24, 2009,
+                  Proceedings},
+  booktitle = {Principles and Practice of Constraint Programming,
+                  CP 2009},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 5732,
+  year = 2009,
+  doi = {10.1007/978-3-642-04244-7}
+}
+
+ +
+@book{CP2013,
+  editor = {Christian Schulte},
+  title = {Principles and Practice of Constraint Programming -- CP 2013,
+                 19th International Conference, CP 2013,
+                 Uppsala, Sweden, September 16-20, 2013, Proceedings},
+  booktitle = {Principles and Practice of Constraint Programming},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 8124,
+  year = 2013,
+  doi = {10.1007/978-3-642-40627-0}
+}
+
+ +
+@book{CPAIOR2010,
+  title = {Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 7th International Conference, CPAIOR 2010},
+  booktitle = {Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, 7th International Conference, CPAIOR 2010},
+  series = {Lecture Notes in Computer Science},
+  volume = 6140,
+  editor = { Andrea Lodi  and  Michela Milano  and  Paolo Toth },
+  year = 2010,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{CPAIOR2011,
+  title = {Integration of {AI} and {OR} Techniques in Contraint
+                  Programming for Combinatorial Optimization Problems -- 8th
+                  International Conference, {CPAIOR} 2011, Berlin, Germany, May
+                  23 -- 27, 2011. Proceedings},
+  booktitle = {Integration of AI and OR Techniques in Contraint Programming
+                  for Combinatorial Optimization Problems},
+  year = 2011,
+  editor = {T. Berthold and A. M. Gleixner and S. Heinz and T. Koch},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{CPAIOR2012,
+  title = {Integration of {AI} and {OR} Techniques in Contraint
+                  Programming for Combinatorial Optimization Problems -- 9th
+                  International Conference, {CPAIOR} 2012, Nantes, France, May
+                  28 -- June 1, 2012. Proceedings},
+  booktitle = {Integration of AI and OR Techniques in Contraint Programming
+                  for Combinatorial Optimization Problems},
+  year = 2012,
+  isbn = {978-3-642-29827-1},
+  editor = {Nicolas Beldiceanu and Narendra Jussien and Eric Pinson},
+  volume = 7298,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{CPAIOR2013,
+  title = {Integration of AI and OR Techniques in Constraint Programming
+                  for Combinatorial Optimization Problems, 10th International
+                  Conference, CPAIOR 2013, Yorktown Heights, NY, USA, May
+                  18-22, 2013. Proceedings},
+  booktitle = {Integration of AI and OR Techniques in Constraint Programming
+                  for Combinatorial Optimization Problems, CPAIOR 2010},
+  series = {Lecture Notes in Computer Science},
+  volume = 7874,
+  editor = {Gomes, C. and  Meinolf Sellmann },
+  year = 2013,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{DIMACS2002,
+  booktitle = {Data Structures, Near Neighbor Searches, and Methodology:
+                  Fifth and Sixth {DIMACS} Implementation Challenges},
+  title = {Data Structures, Near Neighbor Searches, and Methodology:
+                  Fifth and Sixth {DIMACS} Implementation Challenges,
+                  Proceedings of a {DIMACS} Workshop, USA, 1999},
+  series = {{DIMACS} Series in Discrete Mathematics and Theoretical
+                  Computer Science},
+  volume = 59,
+  publisher = {American Mathematical Society},
+  address = { Providence, RI},
+  year = 2002,
+  editor = {Michael H. Goldwasser and David S. Johnson and  Catherine C. McGeoch }
+}
+
+ +
+@proceedings{DSRSturing2019,
+  title = {International Alan Turing Conference on Decision Support and
+                  Recommender systems (DSRC-Turing'19)},
+  booktitle = {International Alan Turing Conference on Decision Support and
+                  Recommender systems},
+  editor = {Iv{\'a}n Palomares},
+  address = {London, UK},
+  date = {2019-11-21/2019-11-22},
+  year = 2019,
+  month = nov # { 21--22},
+  organization = {Alan Turing Institute},
+  isbn = {978-1-5262-0820-0}
+}
+
+ +
+@book{Dagstuhl12041,
+  editor = { Salvatore Greco  and  Joshua D. Knowles  and  Kaisa Miettinen  and  Eckart Zitzler },
+  title = {Learning in Multiobjective Optimization (Dagstuhl Seminar
+                  12041)},
+  booktitle = {Learning in Multiobjective Optimization (Dagstuhl Seminar
+                  12041)},
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  year = 2012,
+  volume = {2(1)},
+  series = {Dagstuhl Reports},
+  pages = {50--99},
+  doi = {10.4230/DagRep.2.1.50}
+}
+
+ +
+@book{Dagstuhl15031,
+  editor = { Salvatore Greco  and  Kathrin Klamroth  and  Joshua D. Knowles  and  G{\"u}nther Rudolph },
+  title = {Understanding Complexity in Multiobjective Optimization
+                  (Dagstuhl Seminar 15031)},
+  booktitle = {Understanding Complexity in Multiobjective Optimization
+                  (Dagstuhl Seminar 15031)},
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  series = {Dagstuhl Reports},
+  pages = {96--163},
+  year = 2015,
+  volume = {5(1)},
+  doi = {10.4230/DagRep.5.1.96},
+  keywords = {multiple criteria decision making, evolutionary
+                  multiobjective optimization}
+}
+
+ +
+@book{Dagstuhl18031,
+  editor = { Kathrin Klamroth  and  Joshua D. Knowles  and  G{\"u}nther Rudolph  and  Margaret M. Wiecek },
+  title = {Personalized Multiobjective Optimization: An Analytics
+                  Perspective (Dagstuhl Seminar 18031)},
+  booktitle = {Personalized Multiobjective Optimization: An Analytics
+                  Perspective (Dagstuhl Seminar 18031)},
+  publisher = {Schloss Dagstuhl--Leibniz-Zentrum f{\"u}r Informatik,
+                  Germany},
+  series = {Dagstuhl Reports},
+  pages = {33--99},
+  year = 2018,
+  volume = {8(1)},
+  doi = {10.4230/DagRep.8.1.33},
+  keywords = {multiple criteria decision making, evolutionary
+                  multiobjective optimization}
+}
+
+ +
+@book{DoeGenGrei2006metaheuristics,
+  booktitle = {Metaheuristics -- Progress in Complex Systems Optimization},
+  title = {Metaheuristics -- Progress in Complex Systems Optimization},
+  publisher = {Springer},
+  year = 2006,
+  editor = {K. F. Doerner and M. Gendreau and P. Greistorfer and
+                  W. J. Gutjahr and R. F. Hartl and M. Reimann},
+  volume = 39,
+  series = {Operations Research/Computer Science Interfaces Series},
+  address = { New York, NY}
+}
+
+ +
+@book{DoeNeu2020theory,
+  title = {Theory of Evolutionary Computation},
+  editor = { Benjamin Doerr  and  Frank Neumann },
+  year = 2020,
+  doi = {10.1007/978-3-030-29414-4},
+  publisher = {Springer International Publishing}
+}
+
+ +
+@book{EA1997,
+  title = {Artificial Evolution, Third European Conference, AE'97,
+                  N{\^i}mes, France, 22-24 October 1997, Selected Papers},
+  booktitle = {Artificial Evolution},
+  editor = { Jin-Kao Hao  and Evelyne Lutton and Edmund M. A. Ronald and  Marc Schoenauer  and Dominique Snyers},
+  shorteditor = { Jin-Kao Hao  and others},
+  doi = {10.1007/BFb0026589},
+  year = 1998,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 1363
+}
+
+ +
+@book{EA2005,
+  title = {Artificial Evolution: 7th International Conference,
+                  Evolution Artificielle, EA 2005, Lille, France},
+  booktitle = {Artificial Evolution},
+  year = 2005,
+  editor = { Talbi, El-Ghazali  and Pierre Liardet and Pierre
+                  Collet and Evelyne Lutton and Marc Schoenauer},
+  volume = 3871,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EA2007,
+  title = {Artificial Evolution, 8th International Conference,
+                  Evolution Artificielle, EA 2007, Tours, France,
+                  October 29-31, 2007 Revised Selected Papers},
+  booktitle = {Artificial Evolution},
+  editor = {Monmarch{\'e}, Nicolas and  Talbi, El-Ghazali  and Collet, Pierre and  Marc Schoenauer  and Lutton, Evelyne},
+  shorteditor = {Monmarch{\'e}, Nicolas and others},
+  doi = {10.1007/978-3-540-79305-2},
+  year = 2008,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 4926
+}
+
+ +
+@book{EA2009,
+  title = {Artificial Evolution: 9th International Conference,
+                  Evolution Artificielle, EA, 2009, Strasbourg,
+                  France, October 26-28, 2009. Revised Selected
+                  Papers},
+  booktitle = {Artificial Evolution: 9th International Conference, Evolution Artificielle, EA, 2009},
+  year = 2010,
+  series = {Lecture Notes in Computer Science},
+  volume = 5975,
+  shorteditor = {Pierre Collet and others},
+  editor = {Pierre Collet and Nicolas Monmarch{\'e} and Pierrick
+                  Legrand and Marc Schoenauer and Evelyne Lutton},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EA2011,
+  editor = { Jin-Kao Hao  and Legrand, Pierrick and Collet, Pierre and
+                  Monmarch{\'e}, Nicolas and Lutton, Evelyne and Schoenauer,
+                  Marc},
+  title = {Artificial Evolution: 10th International Conference,
+                  Evolution Artificielle, EA, 2011, Angers, France, October
+                  24-26, 2011. Revised Selected Papers},
+  publisher = {Springer},
+  year = 2012,
+  volume = 7401,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {Artificial Evolution: 10th International Conference, Evolution Artificielle, EA, 2011}
+}
+
+ +
+@book{EA2013,
+  title = {Artificial Evolution: 11th International Conference,
+                  Evolution Artificielle, {EA} 2013, Bordeaux, France, October
+                  21-23, 2013. Revised Selected Papers},
+  booktitle = {Artificial Evolution: 11th International Conference, Evolution Artificielle, EA, 2013},
+  series = {Lecture Notes in Computer Science},
+  editor = {Pierrick Legrand and others},
+  fulleditor = {Pierrick Legrand and Marc{-}Michel Corsini and  Jin-Kao Hao  and Nicolas Monmarch{\'{e}} and Evelyne Lutton and Marc
+                  Schoenauer},
+  volume = 8752,
+  year = 2013,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EA2015,
+  title = {Artificial Evolution: 12th International Conference,
+                  Evolution Artificielle, {EA} 2015, Lyon, France, October
+                  26-28, 2015. Revised Selected Papers},
+  booktitle = {Artificial Evolution: 12th International Conference, Evolution Artificielle, EA, 2015},
+  series = {Lecture Notes in Computer Science},
+  editor = {St\'ephane Bonnevay and others},
+  fulleditor = {St\'ephane Bonnevay and Pierrick Legrand and  Nicolas Monmarch{\'e}  and Evelyne Lutton and  Marc Schoenauer },
+  volume = 9554,
+  year = 2016,
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{EA2017,
+  title = {Artificial Evolution: 13th International Conference,
+                  {\'E}volution Artificielle, EA 2017, Paris, France, October
+                  25-27, 2017, Revised Selected},
+  booktitle = {EA 2017: Artificial Evolution},
+  year = 2017,
+  series = {Lecture Notes in Computer Science},
+  volume = 10764,
+  editor = {Lutton, Evelyne and Legrand, Pierrick and Parrend, Pierre and  Nicolas Monmarch{\'e}  and  Marc Schoenauer },
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@proceedings{EALS2014,
+  editor = {Angelov, Plamen and others},
+  booktitle = {Evolving and Autonomous Learning Systems (EALS), 2014 IEEE
+                  Symposium on},
+  title = {Evolving and Autonomous Learning Systems (EALS), 2014 IEEE
+                  Symposium on},
+  year = 2014,
+  publisher = {IEEE}
+}
+
+ +
+@book{EC2000,
+  title = {ACM Conference on Electronic Commerce (EC-00)},
+  booktitle = {ACM Conference on Electronic Commerce (EC-00)},
+  year = 2000,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  editor = {Anant Jhingran and others}
+}
+
+ +
+@book{EC2013,
+  title = {Proceedings of the fourteenth {ACM} Conference on Electronic
+                  Commerce, {EC} 2013, Philadelphia, PA, USA, June 16-20, 2013},
+  booktitle = {Proceedings of the Fourteenth ACM Conference on Electronic
+                  Commerce},
+  editor = {Michael J. Kearns and R. Preston McAfee and {\'{E}}va Tardos},
+  year = 2013,
+  doi = {10.1145/2492002},
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{ECAI2006,
+  editor = {Brewka, Gerhard and Coradeschi, Silvia and Perini, Anna and Traverso, Paolo},
+  title = {Proceedings of the 17th European Conference on Artificial Intelligence,
+                  {ECAI} 2006, Riva del Garda, Italy, August29 - September 1, 2006},
+  booktitle = {Proceedings of the 17th European Conference on Artificial Intelligence,
+                  {ECAI} 2006, Riva del Garda, Italy, August29 - September 1, 2006},
+  year = 2006,
+  publisher = {IOS Press}
+}
+
+ +
+@book{ECAI2010,
+  editor = {Coelho, H. and Studer, R. and Wooldridge, M.},
+  title = {Proceedings of the 19th European Conference on Artificial Intelligence},
+  booktitle = {Proceedings of the 19th European Conference on Artificial Intelligence},
+  year = 2010,
+  publisher = {IOS Press}
+}
+
+ +
+@book{ECAI2020,
+  title = {Proceedings of the 24th European Conference on Artificial Intelligence (ECAI)},
+  booktitle = {Proceedings of the 24th European Conference on Artificial Intelligence (ECAI)},
+  year = 2020,
+  volume = 325,
+  series = {Frontiers in Artificial Intelligence and Applications},
+  editor = {Giuseppe De Giacomo and Alejandro Catala and Bistra Dilkina
+                  and Michela Milano and Senén Barro and Alberto Bugarín and
+                  Jérôme Lang},
+  publisher = {IOS Press}
+}
+
+ +
+@proceedings{ECAL1992,
+  title = {Proceedings of  the First European Conference on
+                  Artificial Life},
+  booktitle = {Proceedings of  the First European Conference on
+                  Artificial Life},
+  year = 1992,
+  editor = {F. J. Varela and P. Bourgine},
+  publisher = {MIT Press, Cambridge, MA}
+}
+
+ +
+@proceedings{ECML2006,
+  editor = {F{\"u}rnkranz, Johannes and Scheffer, Tobias and
+                  Spiliopoulou, Myra},
+  title = {17th European Conference on Machine Learning, Berlin,
+                  Germany, September 18-22, 2006 Proceedings},
+  booktitle = {Machine Learning: ECML 2006},
+  year = 2006,
+  series = {Lecture Notes in Computer Science},
+  volume = 4212,
+  isbn = {978-3-540-46056-5}
+}
+
+ +
+@proceedings{ECMLPKDD2015-3,
+  key = {ECML PKDD},
+  booktitle = {Machine Learning and Knowledge Discovery in Databases, ECML
+                  PKDD 2015},
+  fulleditor = {Albert Bifet and Michael May and Bianca Zadrozny and Ricard
+                  Gavald{\`{a}} and Dino Pedreschi and Francesco Bonchi and
+                  Jaime S. Cardoso and Myra Spiliopoulou},
+  title = {Machine Learning and Knowledge Discovery in Databases -
+                  European Conference, {ECML} {PKDD} 2015, Porto, Portugal,
+                  September 7-11, 2015, Proceedings, Part {III}},
+  series = {Lecture Notes in Computer Science},
+  volume = 9286,
+  publisher = {Springer},
+  year = 2015
+}
+
+ +
+@proceedings{EMAA2006,
+  title = {Empirical Methods for the Analysis of Algorithms, Workshop EMAA 2006, Proceedings},
+  booktitle = {Empirical Methods for the Analysis of Algorithms, Workshop EMAA 2006, Proceedings},
+  year = {2006},
+  editor = { Lu{\'i}s Paquete  and  Marco Chiarandini  and Dario Basso},
+  address = {Reykjavik, Iceland}
+}
+
+ +
+@proceedings{EMBC2015,
+  title = {37th Annual International Conference of the {IEEE} Engineering
+                  in Medicine and Biology Society, EMBC 2015, Proceedings},
+  booktitle = {37th Annual International Conference of the {IEEE} Engineering
+                  in Medicine and Biology Society, EMBC 2015, Proceedings},
+  series = {Annual International Conference of the {IEEE} Engineering in Medicine and Biology},
+  editor = {Lovell, Nigel and Mainardi, Luca},
+  year = 2015,
+  publisher = {IEEE Press}
+}
+
+ +
+@proceedings{EMNLP2006,
+  title = {Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP2006},
+  booktitle = {Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP2006},
+  series = {Empirical Methods in Natural Language Processing},
+  editor = {Jurafsky, Dan and Gaussier, Eric},
+  year = 2006
+}
+
+ +
+@book{EMO2001,
+  editor = { Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele  and  Carlos A. {Coello Coello}  and  David Corne },
+  title = {Evolutionary Multi-Criterion Optimization, First
+                  International Conference, {EMO} 2001, Zurich, Switzerland,
+                  March 7-9, 2001, Proceedings},
+  publisher = {Springer},
+  year = 2001,
+  volume = 1993,
+  series = {Lecture Notes in Computer Science},
+  address = {Berlin\slash Heidelberg},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2001}
+}
+
+ +
+@book{EMO2003,
+  title = {Evolutionary Multi-Criterion Optimization, Second
+                  International Conference, EMO 2003, Faro, Portugal,
+                  April 2003: proceedings},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2003},
+  year = 2003,
+  editor = { Carlos M. Fonseca  and  Peter J. Fleming  and  Eckart Zitzler  and  Kalyanmoy Deb  and  Lothar Thiele },
+  series = {Lecture Notes in Computer Science},
+  volume = 2632,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EMO2005,
+  title = {Evolutionary Multi-Criterion Optimization, Third
+                  International Conference, EMO 2005, Guanajuato, Mexico, March
+                  9-11, 2005. Proceedings},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2005},
+  year = 2005,
+  editor = { Carlos A. {Coello Coello}  and Hern{\'a}ndez Aguirre, Arturo and  Eckart Zitzler },
+  series = {Lecture Notes in Computer Science},
+  volume = 3410,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EMO2007,
+  title = {Evolutionary Multi-Criterion Optimization, 4th International
+                  Conference, {EMO} 2007, Matsushima, Japan, March 5-8, 2007,
+                  Proceedings},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2007},
+  year = 2007,
+  editor = {S. Obayashi and others},
+  fulleditor = {Obayashi, Shigeru and  Kalyanmoy Deb  and Poloni, Carlo and Hiroyasu, Tomoyuki and Murata, Tadahiko},
+  volume = 4403,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EMO2009,
+  title = {Evolutionary Multi-Criterion Optimization. 5th International
+                  Conference, EMO 2009},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2009},
+  editor = { Matthias Ehrgott  and  Carlos M. Fonseca  and  Xavier Gandibleux  and  Jin-Kao Hao  and  Marc Sevaux },
+  volume = 5467,
+  series = {Lecture Notes in Computer Science},
+  year = 2009,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EMO2011,
+  editor = { Takahashi, R. H. C.  and  Kalyanmoy Deb  and  Wanner, Elizabeth F.  and  Salvatore Greco },
+  title = {Evolutionary Multi-Criterion Optimization. 6th International
+                  Conference, EMO 2011, Ouro Preto, Brazil, April 5-8, 2011,
+                  Proceedings},
+  publisher = {Springer},
+  year = 2011,
+  volume = 6576,
+  series = {Lecture Notes in Computer Science},
+  address = {Berlin\slash Heidelberg},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2011}
+}
+
+ +
+@book{EMO2013,
+  editor = { Robin C. Purshouse  and  Peter J. Fleming  and  Carlos M. Fonseca  and  Salvatore Greco  and Jane Shaw},
+  title = {Evolutionary Multi-Criterion Optimization -- 7th
+                  International Conference, EMO 2013, Sheffield, UK, March
+                  19-22, 2013.  Proceedings},
+  publisher = {Springer},
+  year = 2013,
+  volume = 7811,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2013},
+  isbn = {978-3-642-37139-4}
+}
+
+ +
+@book{EMO2015_1,
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and Carlos Henggeler Antunes and  Carlos A. {Coello Coello} },
+  title = {Evolutionary Multi-Criterion Optimization -- 8th
+                  International Conference, EMO 2015, Guimar{\~{a}}es,
+                  Portugal, March 29 -- April 1, 2015.  Proceedings, Part {I}},
+  publisher = {Springer},
+  year = 2015,
+  volume = 9018,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {I}}
+}
+
+ +
+@book{EMO2015_2,
+  title = {Evolutionary Multi-Criterion Optimization -- 8th
+                  International Conference, EMO 2015, Guimar{\~{a}}es,
+                  Portugal, March 29 -- April 1, 2015.  Proceedings, Part {II}
+                  },
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2015 Part {II}},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  year = 2015,
+  volume = 9019,
+  editor = { Ant{\'o}nio Gaspar{-}Cunha  and  Carlos Henggeler Antunes and  Carlos A. {Coello Coello} }
+}
+
+ +
+@book{EMO2017,
+  title = {Evolutionary Multi-Criterion Optimization -- 9th
+                  International Conference, EMO 2017, M{\"u}nster, Germany,
+                  March 19 - 22, 2017.  Proceedings},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2017},
+  publisher = {Springer International Publishing},
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = 10173,
+  year = 2017,
+  editor = {Heike Trautmann and G{\"{u}}nter Rudolph and Kathrin Klamroth
+                  and Oliver Sch{\"{u}}tze and Margaret M. Wiecek and Yaochu
+                  Jin and Christian Grimme}
+}
+
+ +
+@book{EMO2019,
+  editor = { Kalyanmoy Deb  and Erik D. Goodman and  Carlos A. {Coello Coello}  and Kathrin
+                  Klamroth and  Kaisa Miettinen  and Sanaz Mostaghim and Patrick
+                  Reed},
+  title = {Evolutionary Multi-Criterion Optimization -- 10th
+                  International Conference, {EMO} 2019, East Lansing, MI, USA,
+                  March 10-13, 2019, Proceedings},
+  booktitle = { Evolutionary Multi-criterion Optimization, EMO 2019},
+  series = {Lecture Notes in Computer Science},
+  volume = 11411,
+  publisher = {Springer International Publishing},
+  address = { Cham, Switzerland},
+  year = 2019,
+  doi = {10.1007/978-3-030-12598-1},
+  isbn = {978-3-030-12597-4}
+}
+
+ +
+@book{EOP2011,
+  title = {Encyclopedia of Parallel Computing},
+  booktitle = {Encyclopedia of Parallel Computing},
+  editor = {David Padua},
+  year = 2011,
+  publisher = {Springer, US},
+  doi = {10.1007/978-0-387-09766-4_244}
+}
+
+ +
+@book{EORMS2011,
+  title = {Wiley Encyclopedia of Operations Research and Management
+                  Science},
+  booktitle = {Wiley Encyclopedia of Operations Research and
+                  Management Science},
+  editor = {J. J. Cochran},
+  publisher = {John Wiley \& Sons},
+  year = 2011,
+  doi = {10.1002/9780470400531}
+}
+
+ +
+@book{EP1998,
+  editor = {V. William Porto and N. Saravanan and Donald E. Waagen and  Agoston E. Eiben },
+  title = {Evolutionary Programming VII, 7th International Conference,
+                  EP98, San Diego, CA, USA, March 25-27, 1998, Proceedings},
+  booktitle = {International Conference on Evolutionary Programming},
+  series = {Lecture Notes in Computer Science},
+  volume = 1447,
+  publisher = {Springer},
+  year = 1998
+}
+
+ +
+@proceedings{ESANN2014,
+  key = {ESANN},
+  title = {Proceedings of 22th European Symposium on Artificial Neural
+                  Networks, {ESANN} 2014, Bruges, Belgium, April 23-25, 2014},
+  booktitle = {European Symposium on Artificial Neural Networks, ESSAN},
+  year = 2014,
+  epub = {https://www.esann.org/proceedings/2014}
+}
+
+ +
+@proceedings{ESANN2015,
+  key = {ESANN},
+  title = {Proceedings of 23rd European Symposium on Artificial Neural
+                  Networks, {ESANN} 2015, Bruges, Belgium, April 22-24, 2015},
+  booktitle = {European Symposium on Artificial Neural Networks, ESSAN},
+  year = 2015,
+  epub = {https://www.esann.org/proceedings/2015}
+}
+
+ +
+@proceedings{EUME2009,
+  title = {Proceedings of the EU/MEeting 2009: Debating the future: new
+                  areas of application and innovative approaches},
+  booktitle = {Proceedings of the EU/MEeting 2009: Debating the future: new
+                  areas of application and innovative approaches},
+  year = 2009,
+  editor = {Ana Viana and others}
+}
+
+ +
+@book{EUROGEN2001,
+  editor = {K. C. Giannakoglou and D. T. Tsahalis and J. Periaux and
+                  K. D. Papaliliou and T. Fogarty},
+  title = {Evolutionary Methods for Design, Optimisation and Control
+                  with Application to Industrial Problems. Proceedings of the
+                  EUROGEN 2001 Conference},
+  publisher = {CIMNE, Barcelona, Spain},
+  year = 2002,
+  booktitle = {Evolutionary Methods for Design, Optimisation and Control},
+  shorteditor = {K. C. Giannakoglou and others},
+  isbn = {84-89925-97-6}
+}
+
+ +
+@book{EUROGP2012,
+  title = {Genetic Programming, 15th European Conference on Genetic
+                  Programming, EuroGP 2012, Proceedings},
+  booktitle = {Proceedings of  the 15th European Conference on Genetic Programming, EuroGP 2012},
+  year = 2012,
+  editor = { A. Moraglio  and Sara Silva and  Krzysztof Krawiec  and  Penousal Machado  and  Carlos Cotta },
+  series = {Lecture Notes in Computer Science},
+  volume = 7244,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EUROGP2017,
+  editor = {James McDermott and Mauro Castelli and Luk{\'{a}}s Sekanina
+                  and Evert Haasdijk and  Pablo Garc{\'i}a-S{\'a}nchez },
+  booktitle = {Proceedings of  the 20th European Conference on Genetic Programming, EuroGP 2017},
+  title = {Genetic Programming, 20th European Conference, EuroGP 2017,
+                  Amsterdam, The Netherlands, April 19-21, 2017, Proceedings},
+  series = {Lecture Notes in Computer Science},
+  volume = 10196,
+  year = 2017,
+  doi = {10.1007/978-3-319-55696-3},
+  isbn = {978-3-319-55695-6},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EUROGP2022,
+  title = {Genetic Programming, 25th European Conference, EuroGP 2022,
+                  Held as Part of EvoStar 2022, Madrid, Spain, April 20-22,
+                  2022, Proceedings},
+  booktitle = {Proceedings of  the 25th European Conference on Genetic Programming, EuroGP 2022},
+  editor = {Eric Medvet and  Gisele Pappa  and Bing Xue},
+  series = {Lecture Notes in Computer Science},
+  year = 2022,
+  publisher = {Springer Nature},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{EVOAPP2010,
+  editor = {Cecilia Di Chio and Stefano Cagnoni and  Carlos Cotta  and Marc
+                  Ebner and Anik{\'o} Ek{\'a}rt and  Anna I. Esparcia{-}Alc{\'{a}}zar  and Chi Keong Goh and  Juan-Juli{\'a}n Merelo  and Ferrante Neri and  Mike Preuss  and Julian Togelius and Georgios N. Yannakakis},
+  title = {Applications of Evolutionary Computation, EvoApplicatons
+                  2010: EvoCOMPLEX, EvoGAMES, EvoIASP, EvoINTELLIGENCE, EvoNUM,
+                  and EvoSTOC, Istanbul, Turkey, April 7-9, 2010, Proceedings,
+                  Part I},
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 6024,
+  year = 2010,
+  doi = {10.1007/978-3-642-12239-2}
+}
+
+ +
+@book{EVOAPP2012,
+  editor = {Di Chio, Cecillia and others},
+  title = {EvoApplications 2012: EvoCOMNET, EvoCOMPLEX, EvoFIN,
+                  EvoGAMES, EvoHOT, EvoIASP, EvoNUM, EvoPAR, EvoRISK, EvoSTIM,
+                  and EvoSTOC, Málaga, Spain, April 11-13, 2012, Proceedings},
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 7248,
+  year = 2012
+}
+
+ +
+@book{EVOAPP2014,
+  editor = { Anna I. Esparcia{-}Alc{\'{a}}zar  and Antonio M. Mora},
+  title = {17th European Conference, EvoApplications 2014, Granada,
+                  Spain, April 23-25, 2014, Revised Selected Papers},
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 8602,
+  year = 2014
+}
+
+ +
+@book{EVOAPP2015,
+  editor = {Antonio M. Mora and Squillero, Giovanni},
+  title = {Applications of Evolutionary Computation - 18th European
+                  Conference, EvoApplications 2015, Copenhagen, Denmark, April
+                  8 -- 10, 2015, Proceedings},
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9028,
+  year = 2015
+}
+
+ +
+@book{EVOAPP2016_1,
+  editor = {Squillero, Giovanni and Burelli, Paolo},
+  title = {Applications of Evolutionary Computation: 19th European
+                  Conference, EvoApplications 2016, Porto, Portugal, March 30
+                  -- April 1, 2016, Proceedings, Part I},
+  year = 2016,
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 9597,
+  doi = {10.1007/978-3-319-31204-0}
+}
+
+ +
+@book{EVOAPP2017_1,
+  editor = {Squillero, Giovanni and Sim, Kevin},
+  title = {Applications of Evolutionary Computation: 20th European
+                  Conference, EvoApplications 2017, Amsterdam, The Netherlands,
+                  April 19-21, 2017, Proceedings, Part I},
+  publisher = {Springer},
+  year = 2017,
+  volume = 10199,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {Applications of Evolutionary Computation},
+  doi = {10.1007/978-3-319-55849-3}
+}
+
+ +
+@book{EVOAPP2021,
+  editor = {Pedro Castillo and  Jim{\'e}nez Laredo, Juan Luis },
+  title = {Applications of Evolutionary Computation -- 24th
+                  International Conference, EvoApplications 2021, Held as Part
+                  of EvoStar 2021, Virtual Event, April 7-9, 2021, Proceedings},
+  year = 2021,
+  booktitle = {Applications of Evolutionary Computation},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = {12694}
+}
+
+ +
+@book{EVOAPP2022,
+  fulleditor = { Jim{\'e}nez Laredo, Juan Luis  and Hidalgo Perez, J. Ignacio  and Oluwatoyin Babaagba, Kehinde},
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and others},
+  title = {Applications of Evolutionary Computation -- 25th European
+                  Conference, EvoApplications 2022, Held as Part of EvoStar
+                  2022, Madrid, Spain, April 20-22, 2022, Proceedings},
+  year = 2022,
+  booktitle = {EvoApplications 2022: Applications of Evolutionary Computation},
+  publisher = {Springer Nature},
+  address = {Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = 13224
+}
+
+ +
+@book{EVOAPP2023,
+  editor = {Correia, Jo\~{a}o and Smith, Stephen and Qaddoura, Raneem},
+  title = {Applications of Evolutionary Computation -- 26th European
+                  Conference, EvoApplications 2023, Held as Part of EvoStar
+                  2023, Brno, Czech Republic, April 12-14, 2023, Proceedings},
+  year = 2023,
+  booktitle = {EvoApplications 2023: Applications of Evolutionary Computation},
+  publisher = {Springer Nature},
+  address = {Switzerland},
+  series = {Lecture Notes in Computer Science},
+  volume = 13989
+}
+
+ +
+@book{EVOCOP2003,
+  title = {Proceedings of EvoCOP 2003 -- 3rd European Conference on Evolutionary Computation in Combinatorial Optimization },
+  booktitle = {Proceedings of EvoCOP 2003 -- 3rd European Conference on Evolutionary Computation in Combinatorial Optimization },
+  volume = 2611,
+  editor = { G{\"u}nther R. Raidl  and Gottlieb, Jens},
+  year = 2003,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2004,
+  editor = {Gottlieb, Jens and  G{\"u}nther R. Raidl },
+  title = {Proceedings of EvoCOP 2004 -- 4th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  publisher = {Springer},
+  year = 2004,
+  volume = 3004,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {Proceedings of EvoCOP 2004 -- 4th European Conference on Evolutionary Computation in Combinatorial Optimization }
+}
+
+ +
+@book{EVOCOP2005,
+  title = {Proceedings of EvoCOP 2005 -- 5th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  booktitle = {Proceedings of EvoCOP 2005 -- 5th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  volume = 3448,
+  editor = { G{\"u}nther R. Raidl  and Gottlieb, Jens},
+  year = 2005,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2006,
+  title = {Proceedings of EvoCOP 2006 -- 6th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  booktitle = {Proceedings of EvoCOP 2006 -- 6th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  volume = 3906,
+  editor = {Gottlieb, Jens and  G{\"u}nther R. Raidl },
+  year = 2006,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2007,
+  title = {Proceedings of EvoCOP 2007 -- Seventh European Conference on
+                  Evolutionary Computation in Combinatorial Optimisation},
+  booktitle = {Proceedings of EvoCOP 2007 -- Seventh European Conference on
+                  Evolutionary Computation in Combinatorial Optimisation},
+  editor = { Carlos Cotta  and  others },
+  year = 2007,
+  volume = 4446,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Berlin, Germany}
+}
+
+ +
+@book{EVOCOP2009,
+  title = {Proceedings of EvoCOP 2009 -- 9th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  booktitle = {Proceedings of EvoCOP 2009 -- 9th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Carlos Cotta  and P. Cowling},
+  year = 2009,
+  volume = 5482,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2011,
+  title = {Proceedings of EvoCOP 2011 -- 11th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  booktitle = {Proceedings of EvoCOP 2011 -- 11th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Peter Merz  and  Jin-Kao Hao },
+  year = 2011,
+  volume = 6622,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2012,
+  title = {Evolutionary Computation in Combinatorial Optimization --
+                  12th European Conference, EvoCOP 2012, M{\'a}laga, Spain,
+                  April 11-13, 2012, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2012 -- 12th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  editor = { Jin-Kao Hao  and  Martin Middendorf },
+  year = 2012,
+  volume = 7245,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2013,
+  editor = { Martin Middendorf  and  Christian Blum },
+  title = {Evolutionary Computation in Combinatorial
+                  Optimization -- 13th European Conference, EvoCOP
+                  2013, Vienna, Austria, April 3-5, 2013, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2013 -- 13th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  volume = 7832,
+  year = 2013,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2014,
+  editor = { Christian Blum  and  Gabriela Ochoa },
+  title = {Evolutionary Computation in Combinatorial Optimization -- 14th
+                  European Conference, EvoCOP 2014, Granada, Spain, April
+                  24-25, 2014, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2014 -- 14th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  year = 2014,
+  series = {Lecture Notes in Computer Science},
+  volume = 8600,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2017,
+  editor = { Bin Hu  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Evolutionary Computation in Combinatorial Optimization -- 17th
+                  European Conference, EvoCOP 2017, Amsterdam, The Netherlands,
+                  April 19-21, 2017, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2017 -- 17th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  year = 2017,
+  series = {Lecture Notes in Computer Science},
+  volume = 10197,
+  doi = {10.1007/978-3-319-55453-2},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2018,
+  editor = { Arnaud Liefooghe  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez },
+  title = {Evolutionary Computation in Combinatorial Optimization --
+                  18th European Conference, EvoCOP 2018, Parma, Italy, April
+                  4-6, 2018, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2018 -- 18th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  year = 2018,
+  series = {Lecture Notes in Computer Science},
+  volume = 10782,
+  doi = {10.1007/978-3-319-77449-7},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{EVOCOP2021,
+  editor = { Christine Zarges  and  Verel, S{\'e}bastien },
+  title = {Evolutionary Computation in Combinatorial Optimization --
+                  21st European Conference, EvoCOP 2021, Held as Part of
+                  EvoStar 2021, Virtual Event, April 7-9, 2021, Proceedings },
+  booktitle = {Proceedings of EvoCOP 2021 -- 21th European Conference on Evolutionary Computation in Combinatorial Optimization },
+  year = 2021,
+  series = {Lecture Notes in Computer Science},
+  volume = 12692,
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{EVOCOP2022,
+  editor = {  P{\'e}rez C{\'a}ceres, Leslie and  Verel, S{\'e}bastien },
+  title = {Evolutionary Computation in Combinatorial Optimization --
+                  22nd European Conference, EvoCOP 2022, Held as Part of
+                  EvoStar 2022, April 20-22, 2022, Proceedings},
+  booktitle = {Proceedings of EvoCOP 2022 -- 22nd European Conference on Evolutionary Computation in Combinatorial Optimization },
+  year = 2022,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{EVOLVE2017,
+  author = { Emmerich, Michael T. M.  and   Andr{\'{e}} Deutz  and  Oliver Sch{\"u}tze  and Legrand,
+                  Pierrick and Tantar, Emilia and Tantar,
+                  Alexandru-Adrian},
+  title = {{EVOLVE} - A Bridge between Probability, Set Oriented
+                  Numerics, and Evolutionary Computation {VII}},
+  publisher = {Springer},
+  year = 2017,
+  volume = 662,
+  series = {Studies in Computational Intelligence},
+  address = { Cham, Switzerland},
+  booktitle = {{EVOLVE} - A Bridge between Probability, Set Oriented
+                  Numerics, and Evolutionary Computation {VII}}
+}
+
+ +
+@proceedings{EVOPROG98,
+  booktitle = {Evolutionary Programming VII},
+  title = {7th International Conference, EP98 San Diego, California,
+                  USA, March 25--27, 1998 Proceedings},
+  editor = {V. W. Porto and N. Saravanan and D. Waagen and  Agoston E. Eiben },
+  series = {Lecture Notes in Computer Science},
+  volume = 1447,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 1998,
+  doi = {10.1007/BFb0040753}
+}
+
+ +
+@book{EhrFigGre2010:isorms,
+  booktitle = {Trends in Multiple Criteria Decision Analysis},
+  title = {Trends in Multiple Criteria Decision Analysis},
+  series = {International Series in Operations Research \& Management Science},
+  editor = { Matthias Ehrgott  and  Jos{\'e} Rui Figueira  and  Salvatore Greco },
+  publisher = {Springer, US},
+  volume = 142,
+  year = 2010
+}
+
+ +
+@proceedings{FLAIRS2019,
+  editor = {Roman Bart{\'{a}}k and Keith W. Brawner},
+  title = {Proceedings of the Thirty-Second International Florida
+                  Artificial Intelligence Research Society Conference,
+                  Sarasota, Florida, USA, May 19-22 2019},
+  booktitle = {Proceedings of  the Thirty-Second International Florida Artificial
+                  Intelligence Research Society Conference},
+  publisher = {{AAAI} Press},
+  year = 2019
+}
+
+ +
+@proceedings{FMCAD2007,
+  editor = {Jason Baumgartner and Mary Sheeran},
+  title = {{FMCAD'07}: Proceedings of the 7th International Conference
+                  Formal Methods in Computer Aided Design},
+  booktitle = {{FMCAD'07}: Proceedings of  the 7th International Conference
+                  Formal Methods in Computer Aided Design},
+  publisher = {IEEE Computer Society, Washington, DC, USA},
+  year = 2007,
+  address = {Austin, Texas, USA}
+}
+
+ +
+@proceedings{FOCS2000,
+  editor = {Avrim Blum},
+  booktitle = {41st Annual Symposium on Foundations of Computer Science},
+  title = {41st Annual Symposium on Foundations of Computer Science,
+                  FOCS 2000, 12-14 November 2000, Redondo Beach, California,
+                  USA},
+  year = 2000,
+  publisher = {IEEE Computer Society Press}
+}
+
+ +
+@book{FOGA1991,
+  editor = {G. Rawlins},
+  title = {Foundations of Genetic Algorithms},
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  publisher = {Morgan Kaufmann Publishers, San Mateo, CA},
+  year = 1991
+}
+
+ +
+@book{FOGA1992,
+  editor = { Darrell Whitley },
+  title = {Proceedings of the Second Workshop on Foundations of Genetic
+                  Algorithms},
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  publisher = {Morgan Kaufmann Publishers},
+  year = 1993,
+  isbn = {1-55860-263-1}
+}
+
+ +
+@book{FOGA1996,
+  booktitle = {Foundations of Genetic Algorithms (FOGA)},
+  editor = {Richard K. Belew and Michael D. Vose},
+  year = 1996,
+  title = {Proceedings of the 4th Workshop on Foundations of Genetic
+                  Algorithms, San Diego, CA, USA, August 5 1996},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@book{FOGA2002,
+  booktitle = {Proceedings of the Seventh Workshop on Foundations of Genetic Algorithms (FOGA)},
+  year = 2002,
+  editor = { De Jong, Kenneth A.  and Poli, Riccardo and Rowe, Jonathan E.},
+  title = {Foundations of Genetic Algorithms, 7th International Workshop,
+                  {FOGA} 2002, Torremolinos, Spain, September 2-4, 2002, Proceedings},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@book{FOGA2009,
+  booktitle = {Proceedings of the Tenth ACM SIGEVO Workshop on Foundations of Genetic Algorithms (FOGA)},
+  year = 2009,
+  editor = {Ivan I. Garibay and Thomas Jansen and R. Paul Wiegand and
+                  Annie S. Wu},
+  title = {Foundations of Genetic Algorithms, 10th {ACM} {SIGEVO}
+                  International Workshop, {FOGA} 2009, Orlando, Florida, USA,
+                  January 9-11, 2009, Proceedings},
+  publisher = {{ACM}},
+  isbn = {978-1-60558-414-0}
+}
+
+ +
+@book{FOGA2019,
+  booktitle = {Proceedings of the 15th {ACM}/{SIGEVO} Conference on Foundations of Genetic Algorithms},
+  year = 2019,
+  editor = { Tobias Friedrich  and  Carola Doerr  and Arnold, Dirk V.},
+  title = {Foundations of Genetic Algorithms, 15th {ACM}/{SIGEVO}
+                  International Workshop, {FOGA} 2019, Potsdam, Germany},
+  publisher = {{ACM}}
+}
+
+ +
+@book{FOGA2023,
+  booktitle = {Proceedings of the 17th {ACM}/{SIGEVO} Conference on Foundations of Genetic Algorithms},
+  year = 2023,
+  editor = { Chicano, Francisco  and  Tobias Friedrich  and K{\"o}tzing, Timo  and  Franz Rothlauf },
+  title = {Foundations of Genetic Algorithms, 17th {ACM}/{SIGEVO}
+                  International Workshop, {FOGA} 2023, Potsdam, Germany},
+  publisher = {{ACM}}
+}
+
+ +
+@book{FigGreEhr2005:mcda,
+  booktitle = {Multiple Criteria Decision Analysis, State of the
+                  Art Surveys},
+  title = {Multiple Criteria Decision Analysis, State of the
+                  Art Surveys},
+  publisher = {Springer},
+  year = 2005,
+  editor = { Jos{\'e} Rui Figueira  and  Salvatore Greco  and  Matthias Ehrgott }
+}
+
+ +
+@book{FurHul2011preflearn,
+  editor = {F{\"u}rnkranz, Johannes and  Eyke H{\"u}llermeier },
+  title = {Preference Learning},
+  booktitle = {Preference Learning},
+  year = 2011,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  isbn = {978-3-642-14125-6}
+}
+
+ +
+@book{GECCO1999,
+  editor = {Wolfgang Banzhaf and Jason M. Daida and A. E. Eiben
+                  and Max H. Garzon and Vasant Honavar and Mark
+                  J. Jakiela and Robert E. Smith},
+  shorteditor = {Wolfgang Banzhaf and others},
+  title = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 1999, 13-17 July 1999,
+                  Orlando, Florida, USA},
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 1999},
+  year = 1999,
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA}
+}
+
+ +
+@book{GECCO2000,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2000},
+  booktitle = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2000},
+  year = 2000,
+  fulleditor = { Darrell Whitley  and  David E. Goldberg  and E. Cantu-Paz and L. Spector and
+                  I. Parmee and   Hans-Georg Beyer },
+  editor = { Darrell Whitley  and others},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA}
+}
+
+ +
+@book{GECCO2001,
+  title = {Proceedings of the 3rd Annual Conference on Genetic and
+                  Evolutionary Computation, GECCO 2001},
+  booktitle = {Proceedings of the 3rd Annual Conference on Genetic and
+                  Evolutionary Computation, GECCO 2001},
+  year = 2001,
+  editor = {Erik D. Goodman},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA}
+}
+
+ +
+@book{GECCO2002,
+  title = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2002},
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2002},
+  year = 2002,
+  editor = { Langdon, William B.  and others},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA}
+}
+
+ +
+@book{GECCO2003_1,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2003, Part I},
+  booktitle = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2003, Part I},
+  year = 2003,
+  editor = {E. Cant\'u-Paz and others},
+  volume = 2723,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{GECCO2004_1,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2004, Seattle, WA, USA, June 26-30, 2004,
+                  Proceedings, Part I},
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2004, Part I},
+  year = 2004,
+  editor = { Kalyanmoy Deb and others},
+  volume = 3102,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{GECCO2004_2,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2004, Seattle, WA, USA, June 26-30, 2004,
+                  Proceedings, Part II},
+  booktitle = {Proceedings of the Genetic and Evolutionary
+                  Computation Conference, GECCO 2004, Part II},
+  year = 2004,
+  editor = { Kalyanmoy Deb and others},
+  volume = 3103,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{GECCO2005,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2005},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2005},
+  editor = {  Hans-Georg Beyer  and  Una-May O'Reilly },
+  year = 2005,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2006,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, GECCO 2006},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2006},
+  editor = {M. Cattolico and others},
+  year = 2006,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2007,
+  title = {Genetic and Evolutionary Computation Conference, {GECCO}
+                  2007, Proceedings, London, England, UK, July 7-11, 2007},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2007},
+  editor = {Dirk Thierens and others},
+  year = 2007,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2008,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2008, Proceedings, Atlanta, Georgia, USA
+                  July 12-16, 2008},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2008},
+  editor = {Conor Ryan},
+  year = 2008,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2009,
+  editor = { Franz Rothlauf },
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2009,
+                  Proceedings, Montreal, Qu{\'e}bec, Canada, July 8-12, 2009},
+  publisher = {ACM Press},
+  year = 2009,
+  address = { New York, NY},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2009}
+}
+
+ +
+@book{GECCO2009c,
+  editor = { Franz Rothlauf },
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2009,
+                  Proceedings, Montreal, Qu{\'e}bec, Canada, July 8-12, 2009,
+                  Companion Material},
+  publisher = {ACM Press},
+  year = 2009,
+  address = { New York, NY},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2009}
+}
+
+ +
+@book{GECCO2010,
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2010, Proceedings, Portland, Oregon, USA, July
+                  7-11, 2010},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2010},
+  year = 2010,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2010c,
+  editor = {Martin Pelikan and  J{\"u}rgen Branke },
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2010,
+                  Companion Material Proceedings, Portland, Oregon, USA, July
+                  7-11, 2010},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2010},
+  year = 2010,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2011,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2011, Proceedings, Dublin, Ireland, July
+                  12-16, 2011},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2011},
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  year = 2011,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2011c,
+  editor = {Natalio Krasnogor and Pier Luca Lanzi},
+  title = {13th Annual Genetic and Evolutionary Computation Conference,
+                  GECCO 2011, Companion Material Proceedings, Dublin, Ireland,
+                  July 12-16, 2011},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2011},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2011
+}
+
+ +
+@book{GECCO2012,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2012, Proceedings, Philadelphia, PA, USA, July
+                  7-11, 2012},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2012},
+  editor = {Terence Soule and Jason H. Moore},
+  year = 2012,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2012c,
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2012,
+                  Companion Material Proceedings, Philadelphia, PA, USA, July
+                  7-11, 2012},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2012},
+  editor = {Terence Soule and Jason H. Moore},
+  year = 2012,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2013,
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2013,
+                  Proceedings, Amsterdam, The Netherlands, July 6-10, 2013},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2013},
+  editor = { Christian Blum  and  Alba, Enrique },
+  year = 2013,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  isbn = {978-1-4503-1963-8}
+}
+
+ +
+@book{GECCO2013c,
+  editor = { Christian Blum  and  Alba, Enrique },
+  title = {Genetic and Evolutionary Computation
+                  Conference, GECCO 2013, Companion Material
+                  Proceedings, Amsterdam, The Netherlands, July 6-10, 2013},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2013},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2013
+}
+
+ +
+@book{GECCO2014,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2014, Proceedings, Vancouver, BC, Canada,
+                  July 12-16, 2014},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2014},
+  editor = {Christian Igel and Dirk V. Arnold},
+  year = 2014,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2015,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2015, Proceedings, Madrid, Spain,
+                  July 11-15, 2015},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2015},
+  editor = {Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  year = 2015,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2015c,
+  title = {Genetic and Evolutionary Computation Conference, {GECCO} 2015, Madrid,
+               Spain, July 11-15, 2015, Companion Material Proceedings},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2015},
+  editor = { Jim{\'e}nez Laredo, Juan Luis  and Sara Silva and  Anna I. Esparcia{-}Alc{\'{a}}zar },
+  year = 2015,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2016,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2016, Proceedings, Denver, CO, USA,
+                  July 20-24, 2016},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2016},
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  year = 2016,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2016c,
+  title = {Genetic and Evolutionary Computation Conference, {GECCO} 2016, Denver, CO, USA, July 20-24, 2016, Companion Material Proceedings},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2016},
+  editor = { Tobias Friedrich  and  Frank Neumann  and  Andrew M. Sutton },
+  year = 2016,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2017,
+  title = {Genetic and Evolutionary Computation Conference,
+                  GECCO 2017, Berlin, Germany, July 15-19, 2017},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2017},
+  editor = { Peter A. N. Bosman },
+  year = 2017,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2017c,
+  title = {Genetic and Evolutionary Computation Conference, {GECCO} 2017, Berlin,
+               Germany, July 15-19, 2017},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2017},
+  editor = { Peter A. N. Bosman },
+  year = 2017,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2018,
+  title = {Genetic and Evolutionary Computation Conference, GECCO 2018,
+                  Kyoto, Japan, July 15-19, 2018},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2018},
+  editor = { Aguirre, Hern\'{a}n E.  and Keiki Takadama},
+  doi = {10.1145/3205455},
+  year = 2018,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{GECCO2019,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, {GECCO} 2019, Prague, Czech Republic, July 13-17,
+                  2019},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2019},
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  year = 2019,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  isbn = {978-1-4503-6111-8},
+  doi = {10.1145/3321707},
+  epub = {https://dl.acm.org/citation.cfm?id=3321707}
+}
+
+ +
+@book{GECCO2019c,
+  title = {Genetic and Evolutionary Computation Conference Companion,
+                  {GECCO} 2019, Prague, Czech Republic, July 13-17, 2019},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2019},
+  editor = { Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Anne Auger  and  Thomas St{\"u}tzle },
+  year = 2019,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  isbn = {978-1-4503-6748-6},
+  doi = {10.1145/3319619},
+  epub = {https://dl.acm.org/citation.cfm?id=3319619}
+}
+
+ +
+@book{GECCO2020,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, {GECCO} 2020, Canc{\'u}n, Mexico, July 8-12,
+                  2020},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2020},
+  editor = { Carlos A. {Coello Coello} },
+  year = 2020,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  isbn = {978-1-4503-7128-5},
+  doi = {10.1145/3377930},
+  location = {Canc{\'u}n, Mexico},
+  epub = {https://dl.acm.org/citation.cfm?id=3377930}
+}
+
+ +
+@book{GECCO2021,
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, {GECCO} 2021, Lille, France, July 10-14, 2021},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2021},
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  year = 2021,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  location = {Lille, France},
+  doi = {10.1145/3449639.3459373}
+}
+
+ +
+@book{GECCO2021c,
+  editor = { Chicano, Francisco  and  Krzysztof Krawiec },
+  title = {Genetic and Evolutionary Computation Conference Companion,
+                  {GECCO} 2021, Lille, France, July 10-14, 2021},
+  publisher = {ACM Press},
+  year = 2021,
+  address = { New York, NY},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2021}
+}
+
+ +
+@book{GECCO2022,
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, {GECCO} 2022, Boston, Massachusetts, July 9-13,
+                  2022},
+  publisher = {ACM Press},
+  year = 2022,
+  address = { New York, NY},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2022},
+  location = {Boston, Massachusetts},
+  doi = {10.1145/3512290}
+}
+
+ +
+@book{GECCO2022c,
+  editor = { Jonathan E. Fieldsend  and  Markus Wagner },
+  title = {Genetic and Evolutionary Computation Conference Companion,
+                  {GECCO} 2022, Boston, Massachusetts, July 9-13, 2022},
+  publisher = {ACM Press},
+  year = 2022,
+  address = { New York, NY},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2022},
+  location = {Boston, Massachusetts},
+  doi = {10.1145/3520304},
+  isbn = 9781450392686
+}
+
+ +
+@book{GECCO2023,
+  editor = {Silva, Sara and  Lu{\'i}s Paquete },
+  title = {Proceedings of the Genetic and Evolutionary Computation
+                  Conference, {GECCO} 2023, Lisbon, Portugal, July 15-19, 2023},
+  publisher = {ACM Press},
+  year = 2023,
+  address = { New York, NY},
+  annote = {ISBN: 9798400701191},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO 2023},
+  location = {Lisbon, Portugal},
+  doi = {10.1145/3583131}
+}
+
+ +
+@book{GECCO2023c,
+  editor = {Silva, Sara and  Lu{\'i}s Paquete },
+  title = {Genetic and Evolutionary Computation Conference Companion,
+                  {GECCO} 2023, Lisbon, Portugal, July 15-19, 2023},
+  publisher = {ACM Press},
+  year = 2023,
+  address = { New York, NY},
+  annote = {ISBN: 979-8-4007-0120-7},
+  booktitle = {Proceedings of  the Genetic and Evolutionary Computation Conference, GECCO Companion 2023},
+  location = {Lisbon, Portugal},
+  doi = {10.1145/3583133}
+}
+
+ +
+@proceedings{GP1998,
+  title = {Genetic Programming 1998: Proceedings of the Third
+                  Annual Conference, Late Breaking Papers},
+  booktitle = {Late Breaking Papers at the Genetic Programming 1998
+                  Conference},
+  editor = {John R. Koza},
+  month = jul,
+  address = {Stanford University, California},
+  publisher = {Stanford University Bookstore},
+  year = 1998
+}
+
+ +
+@book{GraWol1963,
+  title = {Recent Advances in Mathematical Programming},
+  booktitle = {Recent Advances in Mathematical Programming},
+  editor = {Graves, R.L. and Wolfe, P.},
+  publisher = {McGraw Hill,  New York, NY},
+  year = 1963
+}
+
+ +
+@book{GutPun2002tsp,
+  title = {The Traveling Salesman Problem and its Variations},
+  booktitle = {The Traveling Salesman Problem and its Variations},
+  publisher = {Kluwer Academic Publishers, Dordrecht, The Netherlands},
+  year = 2002,
+  editor = {G. Gutin and A. Punnen}
+}
+
+ +
+@book{HM2006,
+  title = {Proceedings of HM 2006 -- 3rd International Workshop
+                  on Hybrid Metaheuristics},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2006,
+  aeditor = {F. Almeida and M. Blesa and C. Blum and J. M. Moreno
+                  and M. P{\'e}rez and A. Roli and  M. Sampels },
+  editor = {F. Almeida and others},
+  volume = 4030,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{HM2007,
+  title = {Hybrid Metaheuristics HM 2007, 4th International
+                  Workshop},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2007,
+  editor = { Thomas Bartz-Beielstein  and  Mar{\'i}a J. Blesa  and  Christian Blum  and  Boris Naujoks  and  Andrea Roli  and  G{\"u}nther Rudolph  and  M. Sampels },
+  volume = 4771,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{HM2008,
+  title = {Hybrid Metaheuristics HM 2008, 5th International
+                  Workshop},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2008,
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and  Carlos Cotta  and  Antonio J. Fern{\'a}ndez  and Jos\'e E. Gallardo and  Andrea Roli  and  M. Sampels },
+  volume = 5296,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{HM2009,
+  title = {Hybrid Metaheuristics, 6th International Workshop,
+                  HM 2009, Udine, Italy, October 16-17,
+                  2009. Proceedings},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2009,
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Luca {Di Gaspero} and  Andrea Roli  and  M. Sampels  and Andrea Schaerf},
+  series = {Lecture Notes in Computer Science},
+  volume = 5818,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{HM2013,
+  title = {Hybrid Metaheuristics, 8th International Workshop,
+                  HM 2013, Ischia, Italy, May 23-25,
+                  2013. Proceedings},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2013,
+  isbn = {978-3-642-38515-5},
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and Paola Festa and  Andrea Roli  and  M. Sampels },
+  series = {Lecture Notes in Computer Science},
+  volume = 7919,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{HM2014,
+  title = {Hybrid Metaheuristics, 9th International Workshop,
+                  HM 2014, Hamburg, Germany, June 11-13,
+                  2014. Proceedings},
+  booktitle = {Hybrid Metaheuristics},
+  year = 2014,
+  editor = { Mar{\'i}a J. Blesa  and  Christian Blum  and  Stefan Vo{\ss} },
+  isbn = {978-3-319-07643-0},
+  volume = 8457,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{Handbook2002,
+  title = {Handbook of Metaheuristics},
+  booktitle = {Handbook of Metaheuristics},
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  year = 2002,
+  publisher = {Kluwer Academic Publishers, Norwell, MA},
+  alias = {GloKoc02:handbook-mh}
+}
+
+ +
+@book{Handbook2003,
+  title = {Handbook of Metaheuristics},
+  booktitle = {Handbook of Metaheuristics},
+  editor = { Fred Glover  and Gary A. Kochenberger},
+  year = 2003,
+  publisher = {Springer},
+  address = { Boston, MA},
+  doi = {10.1007/b101874}
+}
+
+ +
+@book{Handbook2010,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  year = 2010,
+  title = {Handbook of Metaheuristics},
+  booktitle = {Handbook of Metaheuristics},
+  volume = 146,
+  series = {International Series in Operations Research \& Management
+                  Science},
+  edition = {2nd},
+  publisher = {Springer},
+  address = { New York, NY}
+}
+
+ +
+@book{Handbook2019,
+  editor = { Michel Gendreau  and  Jean-Yves Potvin },
+  year = 2019,
+  title = {Handbook of Metaheuristics},
+  booktitle = {Handbook of Metaheuristics},
+  volume = 272,
+  series = {International Series in Operations Research \& Management
+                  Science},
+  publisher = {Springer}
+}
+
+ +
+@book{HandbookCI2015,
+  year = 2015,
+  booktitle = {Springer Handbook of Computational Intelligence},
+  title = {Springer Handbook of Computational Intelligence},
+  editor = {Kacprzyk, Janusz and Pedrycz, Witold},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg}
+}
+
+ +
+@book{HandbookCO1998,
+  title = {Handbook of Combinatorial Optimization},
+  booktitle = {Handbook of Combinatorial Optimization},
+  publisher = {Kluwer Academic Publishers},
+  year = 1998,
+  editor = { Panos M. Pardalos  and  D.-Z. Du },
+  volume = 2
+}
+
+ +
+@book{HarSmiKra2005memetic,
+  title = {Recent Advances in Memetic Algorithms},
+  booktitle = {Recent Advances in Memetic Algorithms},
+  editor = {Hart W. E. and Smith J. E. and Krasnogor N.},
+  year = 2005,
+  volume = 166,
+  series = {Studies in Fuzziness and Soft Computing},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg}
+}
+
+ +
+@book{Heuristics2017,
+  editor = { Rafael Mart{\'i}  and  Panos M. Pardalos  and  Mauricio G. C. Resende },
+  title = {Handbook of Heuristics},
+  booktitle = {Handbook of Heuristics},
+  year = 2018,
+  publisher = {Springer International Publishing},
+  isbn = {978-3-319-07125-1}
+}
+
+ +
+@book{Hochbaum1996,
+  title = {Approximation Algorithms For {NP}-hard Problems},
+  booktitle = {Approximation Algorithms For {NP}-hard Problems},
+  editor = {Hochbaum, Dorit S.},
+  year = 1996,
+  publisher = {PWS Publishing Co.}
+}
+
+ +
+@book{HutKotVan2019automl,
+  editor = { Frank Hutter  and Kotthoff, Lars and  Joaquin Vanschoren },
+  epub = {http://automl.org/book},
+  publisher = {Springer},
+  year = 2019,
+  title = {Automated Machine Learning: Methods, Systems, Challenges},
+  doi = {10.1007/978-3-030-05318-5},
+  booktitle = {Automated Machine Learning}
+}
+
+ +
+@book{ICAI2005,
+  editor = {Hamid R. Arabnia and Rose Joshua},
+  title = {Proceedings of the 2005 International Conference on Artificial Intelligence, ICAI 2005},
+  booktitle = {Proceedings of the 2005 International Conference on Artificial Intelligence, ICAI 2005},
+  publisher = {CSREA Press},
+  year = 2005,
+  isbn = {1-932415-66-1}
+}
+
+ +
+@book{ICALP2005,
+  editor = {Lu{\'i}s Caires and Giuseppe F. Italiano and Lu{\'i}s Monteiro and Catuscia Palamidessi and Moti Yung},
+  title = {Proceedings of the 32nd International Colloquium on Automata, Languages and Programming, {ICALP} 2005},
+  booktitle = {Proceedings of the 32nd International Colloquium on Automata, Languages and Programming, {ICALP} 2005},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 3580,
+  year = 2005
+}
+
+ +
+@proceedings{ICANN1999,
+  title = {Proceedings of the 9th International Conference on Artificial
+                  Neural Networks: ICANN '99, Location: Edinburgh, UK, 7-10
+                  Sept. 1999},
+  year = 1999,
+  booktitle = {ICANN'99: Proceedings of the 9th International Conference on
+                  Artificial Neural Networks},
+  key = {ICANN}
+}
+
+ +
+@book{ICANN2008i,
+  editor = {Kurkova-Pohlova, Vera and Koutnik, Jan},
+  title = {ICANN'08: Proceedings of the 18th International Conference on
+                  Artificial Neural Networks, Part I},
+  publisher = {Springer},
+  year = 2008,
+  volume = 5163,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {ICANN'08: Proceedings of the 18th International Conference on
+                  Artificial Neural Networks, Part I},
+  adoi = {10.1007/978-3-540-87536-9}
+}
+
+ +
+@book{ICANN2008ii,
+  editor = {Kurkova-Pohlova, Vera and Koutnik, Jan},
+  title = {ICANN'08: Proceedings of the 18th International Conference on
+                  Artificial Neural Networks, Part II},
+  publisher = {Springer},
+  year = 2008,
+  volume = 5164,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {Artificial Neural Networks--ICANN 2008}
+}
+
+ +
+@book{ICANNGA1999,
+  editor = {Andrej Dobnikar and Nigel C. Steele and David
+                  W. Pearson and Rudolf F. Albrecht},
+  title = {Artificial Neural Nets and Genetic Algorithms (ICANNGA-99),
+                  Proceedings of the International Conference in Portorož,
+                  Slovenia, 1999},
+  publisher = {Springer Verlag},
+  year = 1999,
+  key = {ICANNGA},
+  booktitle = {Artificial Neural Nets and Genetic Algorithms (ICANNGA-99)},
+  doi = {10.1007/978-3-7091-6384-9}
+}
+
+ +
+@book{ICANNGA2003,
+  editor = {D. W. Pearson and N. C. Steele and R. F. Albrecht},
+  title = {Artificial Neural Networks and Genetic Algorithms},
+  publisher = {Springer Verlag},
+  year = 2003,
+  key = {ICANNGA},
+  booktitle = {Artificial Neural Networks and Genetic Algorithms}
+}
+
+ +
+@proceedings{ICAPS-PAL2011,
+  editor = {Karpas, Erez and Jim{\'e}nez Celorrio, Sergio and Kambhampati, Subbarao},
+  title = {Proceedings of the 3rd Workshop on Learning and Planning,
+                  collocated with the 21st International Conference on
+                  Automated Planning and Scheduling (ICAPS-PAL'11)},
+  booktitle = {Proceedings of ICAPS-PAL11},
+  year = 2011
+}
+
+ +
+@book{ICAPS2004,
+  title = {Proceedings of the Fourteenth International Conference on
+                  Automated Planning and Scheduling (ICAPS 2004)},
+  booktitle = {Proceedings of the Fourteenth International Conference on
+                  Automated Planning and Scheduling (ICAPS 2004)},
+  editor = { Shlomo Zilberstein  and J. Koehler and S. Koenig},
+  year = 2004,
+  publisher = {{AAAI} Press\slash {MIT} Press, Menlo Park, CA}
+}
+
+ +
+@book{ICAPS2015,
+  editor = { Ronen I. Brafman  and Carmel Domshlak and Patrik Haslum and  Shlomo Zilberstein },
+  title = {Proceedings of the Twenty-Fifth International Conference on
+                  Automated Planning and Scheduling, {ICAPS} 2015, Jerusalem,
+                  Israel, June 7-11, 2015},
+  booktitle = {Proceedings of the Twenty-Fifth International Conference on
+                  Automated Planning and Scheduling, {ICAPS} 2015},
+  publisher = {{AAAI} Press},
+  address = { Menlo Park, CA},
+  year = 2015
+}
+
+ +
+@book{ICEC1994,
+  title = {Proceedings of the First IEEE International Conference on
+                  Evolutionary Computation (ICEC'94)},
+  booktitle = {Proceedings of the First IEEE International Conference on
+                  Evolutionary Computation (ICEC'94)},
+  editor = { Zbigniew Michalewicz },
+  year = 1994,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@book{ICEC1996,
+  title = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  booktitle = {Proceedings of the 1996 IEEE International Conference on
+                  Evolutionary Computation (ICEC'96)},
+  editor = { Thomas B{\"a}ck  and  T. Fukuda  and  Zbigniew Michalewicz },
+  year = 1996,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@book{ICEC1997,
+  title = {Proceedings of the 1997 IEEE International
+                  Conference on Evolutionary Computation (ICEC'97)},
+  booktitle = {Proceedings of the 1997 IEEE International
+                  Conference on Evolutionary Computation (ICEC'97)},
+  editor = { Thomas B{\"a}ck  and  Zbigniew Michalewicz  and  Xin Yao },
+  year = 1997,
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ}
+}
+
+ +
+@proceedings{ICGA1985,
+  editor = {John J. Grefenstette},
+  title = {Proceedings of the 1st International Conference on
+                  Genetic Algorithms (ICGA), Pittsburgh, PA, USA, July 1985},
+  booktitle = {Proceedings of  the First International Conference on Genetic Algorithms (ICGA'85)},
+  publisher = {Lawrence Erlbaum Associates},
+  year = 1985,
+  isbn = {0-8058-0426-9}
+}
+
+ +
+@proceedings{ICGA1989,
+  title = {Proceedings of the 3rd International Conference on Genetic
+                  Algorithms (ICGA), George Mason University, Fairfax, Virginia, USA,
+                  June 1989},
+  booktitle = {Proceedings of  the Third International Conference on Genetic Algorithms (ICGA'89)},
+  year = 1989,
+  editor = { J. David Schaffer },
+  publisher = {Morgan Kaufmann Publishers, San Mateo, CA}
+}
+
+ +
+@proceedings{ICGA1993,
+  editor = {Stephanie Forrest},
+  title = {Proceedings of the 5th International Conference on Genetic
+                  Algorithms (ICGA), Urbana-Champaign, IL, USA, June 1993},
+  booktitle = {Proceedings of  the Fifth International Conference on Genetic Algorithms (ICGA'93)},
+  publisher = {Morgan Kaufmann Publishers},
+  year = 1993,
+  isbn = {1-55860-299-2}
+}
+
+ +
+@book{ICGA1995,
+  editor = {Larry J. Eshelman},
+  title = {Proceedings of the 6th International Conference on Genetic
+               Algorithms (ICGA), Pittsburgh, PA, USA, July 15-19, 1995},
+  booktitle = {Proceedings of  the Sixth International Conference on Genetic Algorithms (ICGA'95)},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  address = { Pittsburgh, PA},
+  year = 1995
+}
+
+ +
+@book{ICGA1997,
+  editor = { Thomas B{\"a}ck },
+  title = {Proceedings of the 7th International Conference on Genetic
+               Algorithms (ICGA), East Lansing, MI, USA, July 19-23, 1997},
+  booktitle = {ICGA},
+  publisher = {Morgan Kaufmann Publishers, San Francisco, CA},
+  year = 1997
+}
+
+ +
+@book{ICIC2006,
+  title = {International Conference on Computational Science
+                  (3)},
+  booktitle = {International Conference on Computational Science
+                  (3)},
+  editor = {De-Shuang Huang and Kang Li and George W. Irwin},
+  year = 2006,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 4115
+}
+
+ +
+@book{ICIC2007,
+  title = {Computational Science -- ICCS 2007, 7th International Conference, Proceedings, Part IV},
+  booktitle = {Computational Science -- ICCS 2007, 7th International Conference, Proceedings, Part IV},
+  editor = {Yong Shi and G. Dick van Albada and Jack Dongarra
+                  and Peter M. A. Sloot},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 4490,
+  year = 2007
+}
+
+ +
+@proceedings{ICLR2015,
+  title = {3rd International Conference on Learning Representations,
+                  {ICLR} 2015, San Diego, CA, USA, May 7-9, 2015, Conference
+                  Track Proceedings},
+  year = 2015,
+  booktitle = {3rd International Conference on Learning Representations,
+                  {ICLR} 2015, San Diego, CA, USA, May 7-9, 2015, Conference
+                  Track Proceedings},
+  editor = { Bengio, Yoshua  and Yann {LeCun}}
+}
+
+ +
+@proceedings{ICLR2018w,
+  title = {6th International Conference on Learning Representations, {ICLR} 2018,
+                  Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings},
+  booktitle = {6th International Conference on Learning Representations, {ICLR} 2018,
+                  Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings},
+  editor = {Murray, Iain and Ranzato, Marc'{A}urelio and Vinyals, Oriol},
+  year = 2018,
+  publisher = {OpenReview.net}
+}
+
+ +
+@proceedings{ICML1994,
+  editor = {William W. Cohen and Haym Hirsh},
+  title = {Proceedings of the 11th International Conference on Machine
+                  Learning, {ICML} 1994, New Brunswick, NJ, USA},
+  booktitle = {Proceedings of  the 11th International Conference on Machine Learning, {ICML} 1994},
+  year = 1994,
+  publisher = {Morgan Kaufmann Publishers},
+  address = { San Francisco, CA}
+}
+
+ +
+@proceedings{ICML2004,
+  editor = {Carla E. Brodley},
+  booktitle = {Proceedings of  the 21st International Conference on Machine Learning, {ICML} 2004},
+  title = {Machine Learning, Proceedings of the Twenty-first
+                  International Conference, {ICML} 2004, Banff, Alberta,
+                  Canada, July 4-8, 2004},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2004
+}
+
+ +
+@proceedings{ICML2008,
+  editor = {William W. Cohen and Andrew McCallum and Sam T. Roweis},
+  booktitle = {Proceedings of  the 25th International Conference on Machine Learning, {ICML} 2008},
+  title = {Proceedings of the 25th International Conference on Machine
+                  Learning, {ICML} 2008, Helsinki, Finland, July 05-09, 2008},
+  year = 2008,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@proceedings{ICML2009,
+  title = {Proceedings of the 26th Annual International Conference on
+                  Machine Learning, {ICML} 2009, Montreal, Quebec, Canada, June
+                  14-18, 2009},
+  booktitle = {Proceedings of  the 26th International Conference on Machine Learning, {ICML} 2009},
+  year = 2009,
+  editor = {Andrea Pohoreckyj Danyluk and L{\'{e}}on Bottou and Michael
+                  L. Littman},
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@proceedings{ICML2010,
+  editor = {Johannes F{\"u}rnkranz and Thorsten Joachims},
+  title = {Proceedings of the 27th international conference on machine
+                  learning, {ICML} 2010},
+  booktitle = {Proceedings of  the 27th International Conference on Machine Learning, {ICML} 2010},
+  year = 2010,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@proceedings{ICML2012,
+  editor = {John Langford and Joelle Pineau},
+  title = {Proceedings of the 29th International Conference on Machine
+                  Learning, {ICML} 2012, Edinburgh, Scotland, UK, June 26 -
+                  July 1, 2012},
+  booktitle = {Proceedings of  the 29th International Conference on Machine Learning, {ICML} 2012},
+  publisher = {Omnipress},
+  year = 2012
+}
+
+ +
+@proceedings{ICML2013,
+  editor = {Dasgupta, Sanjoy and McAllester, David},
+  title = {Proceedings of the 30th International Conference on Machine
+                  Learning, {ICML} 2013, Atlanta, GA, USA, 16-21 June 2013},
+  booktitle = {Proceedings of  the 30th International Conference on Machine Learning, {ICML} 2013},
+  volume = 28,
+  year = 2013,
+  url = {http://jmlr.org/proceedings/papers/v28/}
+}
+
+ +
+@proceedings{ICML2014,
+  editor = {Xing, Eric P. and Jebara, Tony},
+  title = {Proceedings of the 31st International Conference on Machine
+                  Learning, {ICML} 2014, Beijing, China, 21-26 June 2014},
+  booktitle = {Proceedings of  the 31st International Conference on Machine Learning, {ICML} 2014},
+  volume = 32,
+  year = 2014,
+  publisher = {{PMLR}},
+  url = {http://jmlr.org/proceedings/papers/v32/}
+}
+
+ +
+@proceedings{ICML2015,
+  editor = {Francis Bach and David Blei},
+  title = {Proceedings of the 32nd International Conference on Machine
+                  Learning, {ICML} 2015, Lille, France, 7-9 July 2015},
+  booktitle = {Proceedings of  the 32nd International Conference on Machine Learning, {ICML} 2015},
+  volume = 37,
+  year = 2015,
+  publisher = {{PMLR}},
+  epub = {http://jmlr.org/proceedings/papers/v37/}
+}
+
+ +
+@proceedings{ICML2018,
+  editor = {Jennifer G. Dy and Andreas Krause},
+  title = {Proceedings of the 35th International Conference on Machine
+                  Learning, {ICML} 2018, Stockholmsm{\"{a}}ssan, Stockholm,
+                  Sweden, July 10-15, 2018},
+  booktitle = {Proceedings of  the 35th International Conference on Machine Learning, {ICML} 2018},
+  series = {Proceedings of Machine Learning Research},
+  volume = 80,
+  publisher = {{PMLR}},
+  year = 2018
+}
+
+ +
+@proceedings{ICMLC2004,
+  editor = {Cloete, Ian and Wong, Kit-Po and Berthold, Michael},
+  title = {Proceedings of the 3rd International Conference on
+                  Machine Learning and Cybernetics},
+  booktitle = {Proceedings of the International Conference on
+                  Machine Learning and Cybernetics},
+  year = 2004,
+  publisher = {IEEE Press}
+}
+
+ +
+@proceedings{ICMLC2006,
+  key = {ICMLC},
+  title = {Proceedings of the International Conference on
+                  Machine Learning and Cybernetics},
+  booktitle = {Proceedings of the International Conference on
+                  Machine Learning and Cybernetics},
+  year = 2006,
+  publisher = {IEEE Press}
+}
+
+ +
+@book{ICORES2014,
+  editor = {Bego{\~{n}}a Vitoriano and Eric Pinson and Fernando Valente},
+  booktitle = {{ICORES} 2014 - Proceedings of the 3rd International Conference on
+               Operations Research and Enterprise Systems},
+  title = {{ICORES} 2014 - Proceedings of the 3rd International Conference on
+               Operations Research and Enterprise Systems, Angers, Loire Valley, France},
+  publisher = {SciTePress},
+  year = 2014
+}
+
+ +
+@proceedings{ICSMC1999,
+  key = {SMC},
+  title = {1999 IEEE International Conference on Systems, Man, and
+                  Cybernetics, October 12--15, 1999, Tokyo, Japan},
+  booktitle = {{IEEE} {SMC}'99 Conference Proceedings, 1999 {IEEE}
+                  International Conference on Systems, Man, and Cybernetics},
+  publisher = {IEEE Press},
+  editor = {Koji Ito and Fumio Harashima and Kazuo Tanie},
+  year = 1999
+}
+
+ +
+@proceedings{ICSMC2013,
+  key = {SMC},
+  title = {{IEEE} International Conference on Systems, Man, and
+                  Cybernetics, {SMC} 2013, Manchester, United Kingdom, October
+                  13-16, 2013},
+  booktitle = {2013 IEEE International Conference on Systems, Man, and
+                  Cybernetics},
+  publisher = {IEEE Press},
+  year = 2013
+}
+
+ +
+@proceedings{ICTAI2014,
+  title = {26th {IEEE} International Conference on Tools with Artificial Intelligence,
+                  {ICTAI} 2014, Limassol, Cyprus, November 10-12, 2014},
+  booktitle = {26th {IEEE} International Conference on Tools with Artificial Intelligence,
+                  {ICTAI} 2014, Limassol, Cyprus, November 10-12, 2014},
+  editor = {Papadopoulos, George Angelos},
+  year = 2014,
+  publisher = {IEEE Press}
+}
+
+ +
+@book{IJCAI1991,
+  booktitle = {Proceedings of  the 12th International Joint Conference on Artificial Intelligence (IJCAI-91)},
+  title = {Proceedings of the 12th International
+                  Joint Conference on Artificial Intelligence, IJCAI 91, Sydney, Australia, August
+                  24-30, 1991},
+  year = 1995,
+  editor = {Mylopoulos, John and Reiter, Raymond},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@book{IJCAI1995,
+  booktitle = {Proceedings of  the 14th International Joint Conference on Artificial Intelligence (IJCAI-95)},
+  title = {Proceedings of the 14th International
+                  Joint Conference on Artificial Intelligence, IJCAI 95, Montr{\'{e}}al Qu{\'{e}}bec, Canada, August
+               20-25, 1995, 2 Volumes},
+  year = 1995,
+  editor = {Chris S. Mellish},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@book{IJCAI1997,
+  booktitle = {Proceedings of  the 15th International Joint Conference on Artificial Intelligence (IJCAI-97)},
+  title = {IJCAI 1997, Proceedings of the 15th International
+                  Joint Conference on Artificial Intelligence, IJCAI 97, Nagoya, Japan, August
+                  23-29, 1997, 2 Volumes},
+  year = 1997,
+  editor = {Martha E. Pollack},
+  publisher = {Morgan Kaufmann Publishers}
+}
+
+ +
+@proceedings{IJCAI2001,
+  editor = {Bernhard Nebel},
+  title = {IJCAI 2001, Proceedings of the 17th International Joint
+                  Conference on Artificial Intelligence},
+  booktitle = {Proceedings of  the 17th International Joint Conference on Artificial Intelligence (IJCAI-01)},
+  year = 2001,
+  publisher = {IEEE Press}
+}
+
+ +
+@proceedings{IJCAI2003,
+  editor = {Georg Gottlob and Toby Walsh},
+  title = {IJCAI-03, Proceedings of the Eighteenth International Joint
+                  Conference on Artificial Intelligence, Acapulco, Mexico,
+                  August 9-15, 2003},
+  publisher = {Morgan Kaufmann Publishers},
+  year = 2003,
+  epub = {http://ijcai.org/proceedings/2003},
+  booktitle = {Proceedings of  the 18th International Joint Conference on Artificial Intelligence (IJCAI-03)}
+}
+
+ +
+@proceedings{IJCAI2007,
+  booktitle = {Proceedings of  the 20th International Joint Conference on Artificial Intelligence (IJCAI-07)},
+  title = {IJCAI 2007, Proceedings of the 20th International
+                  Joint Conference on Artificial Intelligence,
+                  Hyderabad, India, January 6-12, 2007},
+  year = 2007,
+  editor = {Manuela M. Veloso},
+  publisher = {AAAI Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{IJCAI2009,
+  booktitle = {Proceedings of  the 21st International Joint Conference on Artificial Intelligence (IJCAI-09)},
+  title = {IJCAI 2009, Proceedings of the 21st International
+                  Joint Conference on Artificial Intelligence,
+                  Pasadena, California, USA, July 11-17, 2009},
+  year = 2009,
+  editor = {Craig Boutilier},
+  publisher = {AAAI Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{IJCAI2011,
+  booktitle = {Proceedings of  the 22nd International Joint Conference on Artificial Intelligence (IJCAI-11)},
+  title = {IJCAI 2011, Proceedings of the 22nd International
+                  Joint Conference on Artificial Intelligence,
+                  Barcelona, Spain, July 16-22, 2011},
+  year = 2011,
+  editor = {Toby Walsh},
+  publisher = {IJCAI/AAAI Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{IJCAI2015,
+  booktitle = {Proceedings of  the 24th International Joint Conference on Artificial Intelligence (IJCAI-15)},
+  title = {IJCAI 2015, Proceedings of the 24th International
+                  Joint Conference on Artificial Intelligence,
+                  Buenos Aires, Argentina, July 25-31, 2015},
+  year = 2015,
+  editor = {Qiang Yang and Michael Wooldridge},
+  publisher = {IJCAI/AAAI Press, Menlo Park, CA}
+}
+
+ +
+@proceedings{IJCCI2010,
+  editor = { Filipe, J.  and  J. Kacprzyk },
+  title = {Proceedings of the International Joint Conference on
+                  Computational Intelligence (IJCCI-2010)},
+  booktitle = {Proceedings of the International Joint Conference on
+                  Computational Intelligence (IJCCI-2010)},
+  publisher = {SciTePress},
+  year = 2010
+}
+
+ +
+@proceedings{IJCNN2006,
+  key = {IJCNN},
+  booktitle = {Proceedings of the International Joint Conference on Neural
+                  Networks, {IJCNN} 2006},
+  title = {Proceedings of the International Joint Conference on Neural
+                  Networks, {IJCNN} 2006, part of the {IEEE} World Congress on
+                  Computational Intelligence, {WCCI} 2006, Vancouver, BC,
+                  Canada, 16-21 July 2006},
+  year = 2006,
+  publisher = {{IEEE}},
+  doi = {10.1109/IJCNN11286.2006}
+}
+
+ +
+@proceedings{IJCNN2008,
+  key = {IJCNN},
+  title = {Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008),
+                  Hong Kong, China, June 1-6, 2008},
+  booktitle = {Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008),
+                  Hong Kong, China, June 1-6, 2008},
+  editor = {Liu, Derong and others},
+  year = 2008,
+  publisher = {IEEE Press}
+}
+
+ +
+@book{IPMU2010,
+  title = {13th International Conference on Information Processing and Management
+                  of Uncertainty, IPMU 2010, Germany, June 28-July 2, 2010. Proceedings},
+  booktitle = {Information Processing and Management of Uncertainty, 13th International
+                  Conference, {IPMU2010}},
+  editor = { Eyke H{\"u}llermeier  and  Rudolf Kruse  and  Frank Hoffmann },
+  series = {Lecture Notes in Artificial Intelligence},
+  volume = 6178,
+  year = 2010,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@proceedings{ISDA2005,
+  editor = {Abraham, Ajith and Paprzycki, Marcin},
+  title = {Proceedings of the 5th International Conference on
+                  Intelligent Systems Design and Applications},
+  booktitle = {Proceedings of the 5th International Conference on
+                  Intelligent Systems Design and Applications},
+  year = 2005
+}
+
+ +
+@book{JohTri1996,
+  editor = {David S. Johnson and  Michael A. Trick },
+  title = {Cliques, Coloring, and Satisfiability: Second {DIMACS}
+                  Implementation Challenge},
+  booktitle = {Cliques, Coloring, and Satisfiability: Second {DIMACS}
+                  Implementation Challenge},
+  publisher = {American Mathematical Society},
+  address = { Providence, RI},
+  year = 1996,
+  volume = 26,
+  series = {{DIMACS} Series on Discrete Mathematics and Theoretical Computer Science}
+}
+
+ +
+@book{Kallrath2004,
+  editor = {Josef Kallrath},
+  title = {Modeling Languages in Mathematical Optimization},
+  publisher = {Kluwer Academic Publishers},
+  year = 2004,
+  volume = 88,
+  series = {Applied Optimization}
+}
+
+ +
+@book{LION2008,
+  title = {Learning and Intelligent Optimization, Second International
+                  Conference, LION 2007, Trento, Italy, December 8-12,
+                  2007. Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, Second International Conference, LION 2},
+  series = {Lecture Notes in Computer Science},
+  volume = 5313,
+  editor = { Vittorio Maniezzo  and  Roberto Battiti  and Jean-Paul Watson},
+  year = 2008,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2009,
+  title = {Third International Conference, LION 3, Trento,
+                  Italy, January 14-18, 2009. Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, Third International Conference, LION 3},
+  series = {Lecture Notes in Computer Science},
+  volume = 5851,
+  editor = { Thomas St{\"u}tzle },
+  year = 2009,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2010,
+  title = {4th International Conference, LION 4, Venice, Italy,
+                  January 18-22, 2010. Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 4th International Conference, LION 4},
+  year = 2010,
+  volume = 6073,
+  series = {Lecture Notes in Computer Science},
+  editor = { Christian Blum  and  Roberto Battiti },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  doi = {10.1007/978-3-642-13800-3}
+}
+
+ +
+@book{LION2011,
+  title = {5th International Conference, LION 5, Rome, Italy,
+                  January 17-21, 2011. Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 5th
+                  International Conference, LION 5},
+  year = 2011,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  editor = { Carlos A. {Coello Coello} },
+  volume = 6683
+}
+
+ +
+@book{LION2012,
+  booktitle = {Learning and Intelligent Optimization, 6th International
+                  Conference, LION 6},
+  editor = { Youssef Hamadi  and  Marc Schoenauer },
+  title = {6th International Conference, LION 6, Paris, France, January
+                  16-20, 2012. Selected Papers},
+  year = 2012,
+  publisher = {Springer},
+  volume = 7219,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2013,
+  title = {7th International Conference, LION 7, Catania,
+                  Italy, January 7-11, 2013. Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 7th
+                  International Conference, LION 7},
+  year = 2013,
+  series = {Lecture Notes in Computer Science},
+  editor = { Panos M. Pardalos  and G. Nicosia},
+  volume = 7997,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2014,
+  title = {8th International Conference, LION 8, Gainesville, Florida,
+                  USA, February 16-21, 2014. Revised Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 8th International Conference, LION 8},
+  year = 2014,
+  series = {Lecture Notes in Computer Science},
+  editor = { Panos M. Pardalos  and  Mauricio G. C. Resende  and Chrysafis Vogiatzis and Jose
+                  L. Walteros},
+  volume = 8426,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2015,
+  title = {9th International Conference, LION 9, Lille,
+                  France, January 12-15, 2015. Revised Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 9th International
+                  Conference, LION 9},
+  year = 2015,
+  series = {Lecture Notes in Computer Science},
+  editor = {Clarisse Dhaenens and  Laetitia Jourdan  and  Marie-El{\'e}onore Marmion },
+  volume = 8994,
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{LION2016,
+  title = {10th International Conference, LION 10, Ischia, Italy, May 29
+                  - June 1, 2016. Revised Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 10th International
+                  Conference, LION 10},
+  year = 2016,
+  series = {Lecture Notes in Computer Science},
+  editor = {Paola Festa and  Meinolf Sellmann  and  Joaquin Vanschoren },
+  volume = 10079,
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{LION2017,
+  title = {11th International Conference, LION 11, Nizhny Novgorod,
+                  Russia, June 19-21, 2017, Revised Selected Papers},
+  booktitle = {Learning and Intelligent Optimization, 11th International
+                  Conference, LION 11},
+  year = 2017,
+  editor = { Roberto Battiti  and Dmitri E. Kvasov and Yaroslav D. Sergeyev},
+  volume = 10556,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{LION2018,
+  title = {12th International Conference, LION 12, Kalamata, Greece, June 10-15, 2018},
+  booktitle = {Learning and Intelligent Optimization, 12th International
+                  Conference, LION 12},
+  year = 2018,
+  editor = { Roberto Battiti  and Mauro Brunato and Ilias Kotsireas and  Panos M. Pardalos },
+  volume = 11353,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{LION2019,
+  booktitle = {Learning and Intelligent Optimization, 13th International
+                  Conference, LION 13},
+  editor = {Nikolaos F. Matsatsinis and Yannis Marinakis and  Panos M. Pardalos },
+  title = {13th International Conference, LION 13, Chania, Crete,
+                  Greece, May 27-31, 2019, Revised Selected Papers},
+  year = 2019,
+  publisher = {Springer},
+  volume = 11968,
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland}
+}
+
+ +
+@proceedings{LMCA2020,
+  title = {Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020,
+                  {LMCA} 2020, Vancouver, Canada, December 12, 2020},
+  booktitle = {Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020,
+                  {LMCA} 2020, Vancouver, Canada, December 12, 2020},
+  editor = {Vlastelica, Marin and Song, Jialin and Ferber, Aaron and Amos, Brandon and
+                  Martius, Georg and Dilkina, Bistra and Yue, Yisong},
+  year = 2020
+}
+
+ +
+@book{LPNMR2013,
+  title = {12th International Conference, LPNMR 2013, Corunna, Spain, September 15-19, 2013. Proceedings},
+  booktitle = {Logic Programming and Nonmonotonic Reasoning},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Artificial Intelligence},
+  year = 2013,
+  volume = 8148,
+  editor = {Pedro Calabar and Tran Cao Son}
+}
+
+ +
+@book{LobLimMic07:book,
+  editor = {F. Lobo and C. F. Lima and  Zbigniew Michalewicz },
+  booktitle = {Parameter Setting in Evolutionary Algorithms},
+  title = {Parameter Setting in Evolutionary Algorithms},
+  year = 2007,
+  publisher = {Springer},
+  address = { Berlin, Germany}
+}
+
+ +
+@book{MCDM1991,
+  booktitle = {Proceedings of the 10th International Conference on
+                  Multiple Criteria Decision Making (MCDM'91)},
+  title = {Proceedings of the 10th International Conference on
+                  Multiple Criteria Decision Making (MCDM'91)},
+  year = 1992,
+  editor = {G. H. Tzeng and P. L. Yu},
+  publisher = {Springer Verlag}
+}
+
+ +
+@book{MCDM1997,
+  booktitle = {Proceedings of the 13th International Conference on
+                  Multiple Criteria Decision Making (MCDM'97)},
+  title = {Proceedings of the 13th International Conference on
+                  Multiple Criteria Decision Making (MCDM'97)},
+  year = 1997,
+  editor = {J. Climaco},
+  publisher = {Springer Verlag}
+}
+
+ +
+@book{MCDMTA1980,
+  booktitle = {Multiple Criteria Decision Making Theory and Application},
+  title = {Multiple Criteria Decision Making Theory and Application,
+                  Proceedings of the Third Conference Hagen/Königswinter, West
+                  Germany, August 20-24, 1979},
+  year = 1980,
+  editor = {Fandel, G. and Gal, T.},
+  number = 177,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@proceedings{MIC1997,
+  editor = { Mauricio G. C. Resende  and Pinho de Souza, Jorge},
+  booktitle = {Proceedings of MIC 1997, the 2nd Metaheuristics International
+                  Conference},
+  title = {Proceedings of MIC 1997, the 2nd Metaheuristics International
+                  Conference, Sophia-Antipolis, France, July 21-24, 1997},
+  year = 1997
+}
+
+ +
+@proceedings{MIC2005,
+  editor = { Karl F. Doerner  and  Michel Gendreau  and Peter Greistorfer and  Gutjahr, Walter J.  and  Richard F. Hartl  and  Marc Reimann },
+  title = {6th Metaheuristics International Conference (MIC 2005)},
+  booktitle = {6th Metaheuristics International Conference (MIC 2005)},
+  year = 2005,
+  address = {Vienna, Austria}
+}
+
+ +
+@proceedings{MIC2009,
+  title = {Proceedings of MIC 2009, the 8th Metaheuristics International Conference},
+  booktitle = {Proceedings of MIC 2009, the 8th Metaheuristics International Conference},
+  year = 2010,
+  editor = {M. Caserta and  Stefan Vo{\ss} },
+  publisher = {University of Hamburg},
+  address = {Hamburg, Germany}
+}
+
+ +
+@proceedings{MIC2011,
+  title = {Proceedings of MIC 2011, the 9th Metaheuristics International
+                  Conference},
+  booktitle = {MIC 2011, the 9th Metaheuristics International
+                  Conference},
+  editor = {Luca {Di Gaspero} and Andrea Schaerf and  Thomas St{\"u}tzle },
+  year = 2011
+}
+
+ +
+@proceedings{MIC2013,
+  key = {MIC},
+  title = {Proceedings of MIC 2013, the 10th Metaheuristics
+                  International Conference},
+  booktitle = {Proceedings of MIC 2013, the 10th Metaheuristics
+                  International Conference},
+  year = 2013
+}
+
+ +
+@proceedings{MIC2015,
+  editor = { Talbi, El-Ghazali },
+  title = {Proceedings of MIC 2015, the 11th Metaheuristics
+                  International Conference},
+  booktitle = {Proceedings of MIC 2015, the 11th Metaheuristics
+                  International Conference},
+  year = 2015
+}
+
+ +
+@book{MICAI2004,
+  editor = {Monroy, Ra{\'u}l and Arroyo-Figueroa, Gustavo and Sucar, Luis
+                  Enrique and Sossa, Humberto},
+  title = {MICAI 2004: Advances in Artificial Intelligence: Third
+                  Mexican International Conference on Artificial Intelligence,
+                  Mexico City, Mexico, April 26-30, 2004. Proceedings},
+  booktitle = {Proceedings of MICAI},
+  series = {Lecture Notes in Artificial Intelligence},
+  volume = 2972,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2004
+}
+
+ +
+@proceedings{MISTA2013,
+  title = {Multidisciplinary International Conference on Scheduling:
+                  Theory and Applications (MISTA 2013)},
+  booktitle = {Multidisciplinary International Conference on Scheduling:
+                  Theory and Applications (MISTA 2013)},
+  address = {Gent, Belgium},
+  editor = { Graham Kendall  and  Vanden Berghe, Greet   and Barry McCollum},
+  year = 2013
+}
+
+ +
+@book{ML1995,
+  year = 1995,
+  booktitle = {Proceedings of the Twelfth International Conference
+                  on Machine Learning (ML-95)},
+  title = {Proceedings of the Twelfth International Conference
+                  on Machine Learning (ML-95)},
+  editor = {A. Prieditis and S. Russell},
+  publisher = {Morgan Kaufmann Publishers, Palo Alto, CA}
+}
+
+ +
+@book{MMO2004,
+  title = {Metaheuristics for Multiobjective Optimisation},
+  booktitle = {Metaheuristics for Multiobjective Optimisation},
+  editor = { Xavier Gandibleux  and Marc Sevaux and  Kenneth S{\"o}rensen  and  V. {T'Kindt} },
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  volume = 535,
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  year = 2004
+}
+
+ +
+@book{MODA10,
+  title = {mODa 10 -- Advances in Model-Oriented Design and Analysis,
+                  Proceedings of the 10th International Workshop in
+                  Model-Oriented Design and Analysis Held in Łagów Lubuski,
+                  Poland, June 10-14, 2013},
+  editor = {Ucinski, Dariusz and Atkinson, Anthony C.  and Patan, Maciej},
+  year = 2013,
+  booktitle = {mODa 10--Advances in Model-Oriented Design and Analysis},
+  publisher = {Springer International Publishing},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{MOOINTEVO2008,
+  booktitle = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  title = {Multiobjective Optimization: Interactive and Evolutionary
+                  Approaches},
+  year = 2008,
+  volume = 5252,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  editor = { J{\"u}rgen Branke  and  Kalyanmoy Deb  and  Kaisa Miettinen  and  Roman S{\l}owi{\'n}ski }
+}
+
+ +
+@book{MOPGP1996,
+  title = {Advances in Multiple Objective and Goal Programming},
+  booktitle = {Advances in Multiple Objective and Goal Programming},
+  year = 1997,
+  editor = {R. Caballero and  Francisco Ruiz  and R. Steuer},
+  volume = 455,
+  series = {Lecture Notes in Economics and Mathematical Systems},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{MPSN2008,
+  booktitle = {Multiobjective Problem Solving from Nature},
+  title = {Multiobjective Problem Solving from Nature},
+  year = 2008,
+  editor = { Joshua D. Knowles  and  David Corne  and  Kalyanmoy Deb  and Chair, Deva Raj},
+  series = {Natural Computing Series},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg}
+}
+
+ +
+@book{MSOST,
+  editor = {William Fitzgibbon and
+                  Yuri A. Kuznetsov and
+                  Pekka Neittaanm{\"a}ki and
+                  Olivier Pironneau},
+  title = {Modeling, Simulation and Optimization for Science and Technology},
+  booktitle = {Modeling, Simulation and Optimization for Science and Technology},
+  publisher = {Springer},
+  series = {Computational Methods in Applied Sciences},
+  volume = 34,
+  year = 2014
+}
+
+ +
+@book{Matheuristics2009,
+  editor = { Vittorio Maniezzo  and  Thomas St{\"u}tzle  and  Stefan Vo{\ss} },
+  title = {Matheuristics---Hybridizing Metaheuristics and Mathematical
+                  Programming},
+  booktitle = {Matheuristics---Hybridizing Metaheuristics and Mathematical
+                  Programming},
+  publisher = {Springer},
+  year = 2009,
+  volume = 10,
+  series = {Annals of Information Systems},
+  address = { New York, NY}
+}
+
+ +
+@book{MehKoeSaaTiw2009:aisc,
+  title = {Applications of Soft Computing},
+  booktitle = {Applications of Soft Computing},
+  editor = { J{\"o}rn Mehnen  and  Mario K{\"o}ppen  and  Ashraf Saad  and  Ashutosh Tiwari },
+  series = {Advances in Intelligent and Soft Computing},
+  publisher = {Springer},
+  address = {Berlin\slash Heidelberg},
+  volume = 58,
+  year = 2009
+}
+
+ +
+@proceedings{NAFIPS2002,
+  key = {NAFIPS},
+  booktitle = {Proceedings of the NAFIPS-FLINT International
+                  Conference'2002},
+  title = {Proceedings of the NAFIPS-FLINT International
+                  Conference'2002},
+  year = 2002,
+  address = {Piscataway, New Jersey},
+  month = jun,
+  publisher = {IEEE Service Center}
+}
+
+ +
+@book{NICSO2009,
+  booktitle = {Nature Inspired Cooperative Strategies for Optimization
+                  (NICSO 2008)},
+  title = {Nature Inspired Cooperative Strategies for Optimization
+                  (NICSO 2008)},
+  publisher = {Springer},
+  year = 2009,
+  series = {Studies in Computational Intelligence},
+  volume = 236,
+  address = { Berlin, Germany},
+  editor = {Natalio Krasnogor and Belén Melián-Batista and José
+                  Andrés Moreno-Pérez and J. Marcos Moreno-Vega and David Alejandro Pelta},
+  doi = {10.1007/978-3-642-03211-0}
+}
+
+ +
+@book{NIO1999,
+  title = {New Ideas in Optimization},
+  booktitle = {New Ideas in Optimization},
+  editor = { David Corne  and  Marco Dorigo  and  Fred Glover },
+  publisher = {McGraw Hill},
+  year = 1999,
+  address = {London, UK},
+  alias = {CorDorGlo99}
+}
+
+ +
+@book{NIPS1994,
+  title = {Advances in Neural Information Processing Systems},
+  booktitle = {Advances in Neural Information Processing Systems},
+  volume = 6,
+  editor = {J. D. Cowan and G. Tesauro and J. Alspector},
+  year = 1994,
+  publisher = {Morgan Kaufmann Publishers},
+  address = { San Francisco, CA}
+}
+
+ +
+@book{NIPS1996,
+  title = {Advances in Neural Information Processing Systems 9, NIPS,
+                  Denver, CO, USA, December 2-5, 1996},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 9)},
+  editor = {Michael Mozer and Michael I. Jordan and Thomas Petsche},
+  publisher = {MIT Press},
+  year = 1996
+}
+
+ +
+@book{NIPS2003,
+  year = 2003,
+  title = {Proceedings of the 16th International Conference on Neural
+                  Information Processing Systems, NIPS},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 16)},
+  editor = {S. Thrun and L. Saul and B. Sch\"{o}lkopf},
+  publisher = {MIT Press}
+}
+
+ +
+@book{NIPS2011,
+  title = {Advances in Neural Information Processing Systems 24: Annual
+                  Conference on Neural Information Processing Systems 2011},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 24)},
+  editor = {J. Shawe-Taylor and R. S. Zemel and P. L. Bartlett and
+                  F. Pereira and K. Q. Weinberger},
+  year = 2011,
+  publisher = {Curran Associates, Red Hook, NY}
+}
+
+ +
+@book{NIPS2012,
+  title = {Advances in Neural Information Processing Systems 25: 26th
+                  Annual Conference on Neural Information Processing Systems
+                  2012},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 25)},
+  editor = {Peter L. Bartlett and Fernando C. N. Pereira and Christopher
+                  J. C. Burges and L{\'{e}}on Bottou and Kilian Q. Weinberger},
+  year = 2012,
+  publisher = {Curran Associates, Red Hook, NY}
+}
+
+ +
+@proceedings{NIPS2015,
+  editor = {Corinna Cortes and Neil D. Lawrence and Daniel D. Lee and
+                  Masashi Sugiyama and Roman Garnett},
+  title = {Advances in Neural Information Processing Systems 28: Annual
+                  Conference on Neural Information Processing Systems 2015,
+                  December 7-12, 2015, Montreal, Quebec, Canada},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS
+                  28)},
+  year = 2015,
+  url = {http://papers.nips.cc/book/advances-in-neural-information-processing-systems-28-2015}
+}
+
+ +
+@proceedings{NIPS2016,
+  title = {Advances in Neural Information Processing Systems 29: Annual
+                  Conference on Neural Information Processing Systems 2016,
+                  December 5-10, 2016, Barcelona, Spain},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 29)},
+  editor = {D. D. Lee and M. Sugiyama and U. V. Luxburg and I. Guyon and
+                  R. Garnett},
+  year = 2016
+}
+
+ +
+@proceedings{NIPS2017,
+  title = {Advances in Neural Information Processing Systems 30: Annual
+                  Conference on Neural Information Processing Systems 2017,
+                  December 4-9, 2017, Long Beach, CA, {USA}},
+  booktitle = {Advances in Neural Information Processing Systems (NIPS 30)},
+  editor = {Isabelle Guyon and Ulrike von Luxburg and Samy Bengio and
+                  Hanna M. Wallach and Rob Fergus and S. V. N. Vishwanathan and
+                  Roman Garnett},
+  year = 2016
+}
+
+ +
+@proceedings{NIPS2019,
+  title = {Advances in Neural Information Processing Systems 32: Annual
+                  Conference on Neural Information Processing Systems 2019,
+                  NeurIPS 2019, 8-14 December 2019, Vancouver, BC, Canada},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS 32)},
+  editor = {Hanna M. Wallach and Hugo Larochelle and Alina Beygelzimer
+                  and Florence d'Alch{\'{e}}{-}Buc and Emily B. Fox and Roman
+                  Garnett},
+  year = 2019,
+  epub = {http://papers.nips.cc/book/advances-in-neural-information-processing-systems-32-2019}
+}
+
+ +
+@proceedings{NIPS2020,
+  title = {Advances in Neural Information Processing Systems 33: Annual
+                  Conference on Neural Information Processing Systems 2020,
+                  NeurIPS 2020, December 6-12, 2020, Virtual},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS
+                  33)},
+  editor = {Hugo Larochelle and Marc'Aurelio Ranzato and Raia Hadsell and
+                  Maria{-}Florina Balcan and Hsuan{-}Tien Lin},
+  year = 2020,
+  epub = {https://proceedings.neurips.cc/paper/2020}
+}
+
+ +
+@proceedings{NIPS2021,
+  title = {Advances in Neural Information Processing Systems 34 (NeurIPS
+                  2021)},
+  year = 2021,
+  booktitle = {Advances in Neural Information Processing Systems 34 (NeurIPS
+                  2021)},
+  editor = {M. Ranzato and A. Beygelzimer and Y. Dauphin and P.S. Liang
+                  and J. Wortman Vaughan},
+  epub = {https://papers.nips.cc/paper/2021}
+}
+
+ +
+@book{NaoTerCav2010autotun,
+  title = {Software Automatic Tuning: From Concepts to State-of-the-Art Results},
+  booktitle = {Software Automatic Tuning: From Concepts to State-of-the-Art Results},
+  publisher = {Springer},
+  year = 2010,
+  editor = {K. Naono and K. Teranishi and J. Cavazos and R. Suda}
+}
+
+ +
+@book{Neri2011,
+  title = {Handbook of Memetic Algorithms},
+  booktitle = {Handbook of Memetic Algorithms},
+  editor = {Neri, Ferrante and  Carlos Cotta  and  Pablo Moscato },
+  volume = 379,
+  year = 2011,
+  publisher = {Springer},
+  series = {Studies in Computational Intelligence}
+}
+
+ +
+@book{OR2022,
+  editor = {Oliver Grothe and Stefan Nickel and Steffen Rebennack and
+                  Oliver Stein},
+  title = {Operations Research 2022, Selected Papers of the Annual
+                  International Conference of the German Operations Research
+                  Society (GOR), Karlsruhe, Germany, September 6-9, 2022},
+  publisher = {Springer},
+  year = 2022,
+  series = {Lecture Notes in Operations Research},
+  address = { Cham, Switzerland},
+  booktitle = {Operations Research Proceedings 2022, OR 2022}
+}
+
+ +
+@proceedings{PACT2014,
+  key = {PACT},
+  title = {Proceedings of the 23rd International Conference on Parallel
+                  Architectures and Compilation},
+  booktitle = {Proceedings of the 23rd International Conference on Parallel
+                  Architectures and Compilation},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2014
+}
+
+ +
+@book{PATAT2000,
+  title = {Practice and Theory of Automated Timetabling III, Third
+                  International Conference, {PATAT} 2000, Konstanz, Germany,
+                  August 16-18, 2000, Selected Papers},
+  booktitle = {PATAT 2000: Proceedings of the 3rd International Conference
+                  of the Practice and Theory of Automated Timetabling},
+  editor = {Edmund K. Burke and Wilhelm Erben},
+  year = 2000,
+  series = {Lecture Notes in Computer Science},
+  volume = 2079,
+  publisher = {Springer}
+}
+
+ +
+@proceedings{PATAT2014,
+  title = {PATAT 2014: Proceedings of the 10th International Conference of the Practice and Theory of Automated Timetabling},
+  booktitle = {PATAT 2014: Proceedings of the 10th International Conference of the Practice and Theory of Automated Timetabling},
+  editor = { Ender {\"O}zcan  and Edmund K. Burke and Barry McCollum},
+  year = 2014,
+  publisher = {PATAT}
+}
+
+ +
+@proceedings{PDP2011,
+  editor = {Frank Mueller},
+  title = {Proceedings of the 2011 IEEE International Parallel \&
+                  Distributed Processing Symposium},
+  booktitle = {Proceedings of the 2011 IEEE International Parallel \&
+                  Distributed Processing Symposium},
+  series = {IPDPS '11},
+  year = 2011,
+  publisher = {IEEE Computer Society}
+}
+
+ +
+@book{PDPTA1998,
+  title = {Proceedings of the International Conference on
+                  Parallel and Distributed Processing Techniques and
+                  Applications (PDPTA'98)},
+  booktitle = {Proceedings of the International Conference on
+                  Parallel and Distributed Processing Techniques and
+                  Applications (PDPTA'98)},
+  editor = {H. R. Arabnia},
+  year = 1998,
+  publisher = {CSREA Press}
+}
+
+ +
+@book{PPSN1991,
+  title = {Parallel Problem Solving from Nature, 1st Workshop, PPSN I
+                  Dortmund, FRG, October 1-3, 1990. Proceedings},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {I}},
+  year = 1991,
+  editor = { Hans-Paul Schwefel  and R. M{\"a}nner},
+  publisher = {Springer},
+  avolume = 496,
+  aseries = {Lecture Notes in Computer Science},
+  address = {Berlin\slash Heidelberg},
+  doi = {10.1007/BFb0029723}
+}
+
+ +
+@book{PPSN1992,
+  editor = {Reinhard M{\"a}nner and Bernard Manderick},
+  title = {Parallel Problem Solving from Nature 2, PPSN-II,
+                  Brussels, Belgium, September 28-30, 1992},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {II}},
+  publisher = {Elsevier},
+  year = 1992
+}
+
+ +
+@book{PPSN1996,
+  title = {The 4th International Conference on Parallel Problem
+                  Solving from Nature Berlin, Germany, September 22 -
+                  26, 1996. Proceedings},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IV}},
+  year = 1996,
+  aeditor = {H.-M. Voigt and W. Ebeling and  Rechenberg, Ingo  and  Hans-Paul Schwefel },
+  editor = {H.-M. Voigt and others},
+  volume = 1141,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{PPSN1998,
+  title = {Parallel Problem Solving from Nature -- PPSN V, 5th
+                  International Conference Amsterdam, The Netherlands September
+                  27-30, 1998 Proceedings},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {V}},
+  year = 1998,
+  series = {Lecture Notes in Computer Science},
+  volume = 1498,
+  editor = { Agoston E. Eiben  and  Thomas B{\"a}ck  and  Marc Schoenauer  and  Hans-Paul Schwefel },
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{PPSN2000,
+  title = {Parallel Problem Solving from Nature -- {PPSN} {VI}},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VI}},
+  series = {Lecture Notes in Computer Science},
+  volume = 1917,
+  year = 2000,
+  aeditor = { Marc Schoenauer  and  Kalyanmoy Deb  and  G{\"u}nther Rudolph  and  Xin Yao  and E. Lutton and  Juan-Juli{\'a}n Merelo  and  Hans-Paul Schwefel },
+  editor = { Marc Schoenauer  and others},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  anote = {IC.29}
+}
+
+ +
+@book{PPSN2002,
+  title = {Parallel Problem Solving from Nature -- {PPSN} {VII}},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VII}},
+  year = 2002,
+  series = {Lecture Notes in Computer Science},
+  volume = 2439,
+  aeditor = { Juan-Juli{\'a}n Merelo  and P. Adamidis and   Hans-Georg Beyer  and J.-L. Fern\'{a}ndez-Villacanas and  Hans-Paul Schwefel },
+  editor = { Juan-Juli{\'a}n Merelo  and others},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  anote = {IC.34}
+}
+
+ +
+@book{PPSN2004,
+  title = {Proceedings of PPSN-VIII, Eighth International Conference on
+                  Parallel Problem Solving from Nature, Birmingham, UK},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {VIII}},
+  aeditor = { Xin Yao  and  Edmund K. Burke  and  Jos{\'e} A. Lozano  and Smith,
+                  Jim and Merelo-Guerv{\'o}s, Juan Juli{\'a}n and Bullinaria,
+                  John A.  and Rowe, Jonathan E.  and Ti{\v{n}}o, Peter and
+                  Kab{\'a}n, Ata and Schwefel, Hans-Paul},
+  editor = { Xin Yao  and others},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 3242,
+  year = 2004
+}
+
+ +
+@book{PPSN2006,
+  title = {Proceedings of PPSN-IX, Ninth International
+                  Conference on Parallel Problem Solving from Nature},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {IX}},
+  editor = {Runarsson, Thomas Philip and   Hans-Georg Beyer  and  Edmund K. Burke  and  Juan-Juli{\'a}n Merelo  and  Darrell Whitley  and  Xin Yao },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 4193,
+  year = 2006
+}
+
+ +
+@book{PPSN2008,
+  title = {Proceedings of PPSN-X, Tenth International
+                  Conference on Parallel Problem Solving from Nature},
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {X}},
+  aeditor = { G{\"u}nther Rudolph  and Thomas Jansen and Simon Lucas and
+                  Carlo Poloni and Nicola Beume},
+  editor = { G{\"u}nther Rudolph  and others},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 5199,
+  year = 2008
+}
+
+ +
+@book{PPSN2010,
+  booktitle = {Parallel Problem Solving from Nature, PPSN XI},
+  title = {Parallel Problem Solving from Nature -- {PPSN} {XI}},
+  series = {Lecture Notes in Computer Science},
+  editor = {Schaefer, Robert and Cotta, Carlos and Kolodziej,
+                  Joanna and  G{\"u}nther Rudolph },
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2010,
+  volume = 6238
+}
+
+ +
+@book{PPSN2012-1,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {I}},
+  title = {Parallel Problem Solving from Nature, {PPSN} {XII}, 12th
+                  International Conference, Taormina, Italy, September 1-5,
+                  2012, Proceedings, Part {I}},
+  editor = { Carlos A. {Coello Coello}  and others},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2012,
+  volume = 7491
+}
+
+ +
+@book{PPSN2012-2,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XII}, Part {II}},
+  title = {Parallel Problem Solving from Nature - {PPSN} {XII} - 12th
+                  International Conference, Taormina, Italy, September 1-5,
+                  2012, Proceedings, Part {II}},
+  editor = { Carlos A. {Coello Coello}  and others},
+  fulleditor = { Carlos A. {Coello Coello}  and Vincenzo Cutello and  Kalyanmoy Deb  and Stephanie
+                  Forrest and Giuseppe Nicosia and Mario Pavone},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2012,
+  volume = 7492
+}
+
+ +
+@book{PPSN2014,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  title = {Parallel Problem Solving from Nature -- {PPSN} {XIII}},
+  editor = { Thomas Bartz-Beielstein  and  J{\"u}rgen Branke  and Bogdan Filipi{\v c} and Jim Smith},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  volume = 8672,
+  year = 2014
+}
+
+ +
+@book{PPSN2016,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XIV}},
+  title = {Parallel Problem Solving from Nature - PPSN XIV 14th
+                  International Conference, Edinburgh, UK, September 17-21,
+                  2016, Proceedings},
+  editor = { Julia Handl  and  Emma Hart  and  Lewis, P. R.  and  Manuel L{\'o}pez-Ib{\'a}{\~n}ez  and  Gabriela Ochoa  and  Ben Paechter },
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  volume = 9921,
+  year = 2016,
+  doi = {10.1007/978-3-319-45823-6},
+  isbn = {978-3-319-45822-9}
+}
+
+ +
+@book{PPSN2018,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  title = {Parallel Problem Solving from Nature - PPSN XV 15th
+                  International Conference, Coimbra, Portugal, September 8-12,
+                  2018, Proceedings, Part {I}},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  year = 2018,
+  volume = 11101
+}
+
+ +
+@book{PPSN2018_2,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XV}},
+  title = {Parallel Problem Solving from Nature - PPSN XV 15th
+                  International Conference, Coimbra, Portugal, September 8-12,
+                  2018, Proceedings, Part {II}},
+  editor = { Anne Auger  and  Carlos M. Fonseca  and Louren{\c c}o, N. and  Penousal Machado  and  Lu{\'i}s Paquete  and  Darrell Whitley },
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  year = 2018,
+  volume = 11102
+}
+
+ +
+@book{PPSN2020,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVI}},
+  title = {Parallel Problem Solving from Nature - PPSN XVI 16th
+                  International Conference, Leiden, The Netherlands, September
+                  5-9, 2020, Proceedings},
+  editor = { Thomas B{\"a}ck  and  Mike Preuss  and Deutz, Andr{\'e} and Wang, Hao and  Carola Doerr  and  Emmerich, Michael T. M.  and  Heike Trautmann },
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  year = 2020,
+  volume = 12269
+}
+
+ +
+@book{PPSN2022,
+  booktitle = {Parallel Problem Solving from Nature -- {PPSN} {XVII}},
+  title = {Parallel Problem Solving from Nature - PPSN XVII, 17th
+                  International Conference, PPSN 2022, Dortmund, Germany,
+                  September 10-14, 2022, Proceedings, Part I},
+  editor = { G{\"u}nther Rudolph  and  Anna V. Kononova  and  Aguirre, Hern\'{a}n E.  and  Pascal Kerschke  and  Gabriela Ochoa  and  Tea Tu{\v s}ar },
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  year = 2022,
+  volume = 13398
+}
+
+ +
+@proceedings{PROC2013,
+  booktitle = {2013 International Conference on Computational Science},
+  title = {2013 International Conference on Computational Science},
+  editor = {Vassil Alexandrov and Michael Lees and Valeria Krzhizhanovskaya and Jack Dongarra and Peter M.A. Sloot},
+  publisher = {Elsevier},
+  volume = 18,
+  year = 2013,
+  series = {Procedia Computer Science}
+}
+
+ +
+@proceedings{SAGA2003,
+  booktitle = {Stochastic Algorithms: Foundations and Applications},
+  title = {Second International Symposium, SAGA 2003, Hatfield, UK, September 22-23, 2003, Proceedings},
+  editor = {Andreas Albrecht and Kathleen Steinh\"{o}fel},
+  publisher = {Springer Verlag},
+  volume = 2827,
+  year = 2003,
+  series = {Lecture Notes in Computer Science},
+  doi = {10.1007/b13596}
+}
+
+ +
+@proceedings{SAT2005,
+  title = {International Conference on Theory and Applications of Satisfiability Testing},
+  booktitle = {International Conference on Theory and Applications of Satisfiability Testing},
+  editor = {Bacchus, Fahiem and Walsh, Toby},
+  volume = 3569,
+  year = 2005
+}
+
+ +
+@book{SAT2015,
+  booktitle = {Theory and Applications of Satisfiability Testing -- {SAT}
+                  2015},
+  title = {Theory and Applications of Satisfiability Testing -- {SAT}
+                  2015},
+  year = 2015,
+  series = {Lecture Notes in Computer Science},
+  volume = 9340,
+  editor = {Heule, Marijn and Weaver, Sean},
+  publisher = {Springer},
+  address = { Cham, Switzerland}
+}
+
+ +
+@proceedings{SATCOM2014,
+  booktitle = {Proceedings of SAT Competition 2014: Solver and Benchmark Descriptions},
+  title = {Proceedings of SAT Competition 2014: Solver and Benchmark Descriptions},
+  editor = {A. Belov and D. Diepold and M. Heule and M. J\"{a}rvisalo},
+  year = 2014,
+  volume = {B-2014-2},
+  series = {Science Series of Publications B},
+  publisher = {University of Helsinki}
+}
+
+ +
+@book{SEAL2008,
+  title = {Simulated Evolution and Learning, 7th International
+                  Conference, SEAL 2008},
+  booktitle = {Simulated Evolution and Learning, 7th International
+                  Conference, SEAL 2008},
+  fulleditor = {X. Li and M. Kirley and M. Zhang and D. G. Green and
+                  V. Ciesielski and  Abbass, Hussein A.  and Z. Michalewicz and
+                  T. Hendtlass and  Kalyanmoy Deb  and  Tan, Kay Chen  and  J{\"u}rgen Branke  and Y. Shi},
+  editor = {X. Li and others},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  series = {Lecture Notes in Computer Science},
+  volume = 5361,
+  year = 2008
+}
+
+ +
+@proceedings{SEMCCO2013,
+  booktitle = {Swarm, Evolutionary, and Memetic Computing},
+  title = {International Conference on Swarm, Evolutionary, and Memetic Computing},
+  editor = {B. K. Panigrahi and P. N. Suganthan and S. Das and S. S. Dash},
+  year = 2013,
+  volume = 8298,
+  series = {Theoretical Computer Science and General Issues},
+  publisher = {Springer International Publishing}
+}
+
+ +
+@book{SIGKDD2000,
+  key = {SIGKDD},
+  editor = {Raghu Ramakrishnan and Salvatore J. Stolfo and Roberto
+                  J. Bayardo and Ismail Parsa},
+  title = {Proceedings of the sixth {ACM} {SIGKDD} international
+                  conference on Knowledge discovery and data mining, Boston,
+                  MA, USA, August 20-23, 2000},
+  epub = {http://dl.acm.org/citation.cfm?id=347090},
+  booktitle = {The 6th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining, {KDD} 2000},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2000
+}
+
+ +
+@book{SIGKDD2004,
+  booktitle = {Proceedings of the tenth ACM SIGKDD international conference
+                  on Knowledge discovery and data mining, {KDD'04}},
+  editor = {Won Kim and Ronny Kohavi and Johannes Gehrke and William
+                  DuMouchel},
+  title = {KDD04: ACM SIGKDD International Conference on Knowledge
+                  Discovery and Data Mining, Seattle WA USA, August 22-25,
+                  2004},
+  year = 2004,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{SIGKDD2013,
+  booktitle = {The 19th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining, {KDD} 2013},
+  title = {The 19th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining, {KDD} 2013},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2013,
+  editor = {Inderjit S. Dhillon and Yehuda Koren and Rayid Ghani and Ted
+                  E. Senator and Paul Bradley and Rajesh Parekh and Jingrui He
+                  and Robert L. Grossman and Ramasamy Uthurusamy}
+}
+
+ +
+@book{SIGKDD2017,
+  booktitle = {23rd {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  editor = {Stan Matwin and Shipeng Yu and Faisal Farooq},
+  title = {KDD'17: The 23rd {ACM} {SIGKDD} International Conference on
+                  Knowledge Discovery and Data Mining, Halifax, NS, Canada,
+                  August 13-17, 2017},
+  year = 2017,
+  publisher = {ACM Press},
+  key = {SIGKDD}
+}
+
+ +
+@book{SIGKDD2018,
+  booktitle = {24th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  editor = {Yike Guo and Faisal Farooq},
+  title = {KDD'18: The 24th ACM SIGKDD International Conference on
+                  Knowledge Discovery and Data Mining, London United Kingdom,
+                  August 19-23, 2018},
+  year = 2018,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  month = jul,
+  key = {SIGKDD}
+}
+
+ +
+@book{SIGKDD2019,
+  booktitle = {25th {ACM} {SIGKDD} International Conference on Knowledge
+                  Discovery and Data Mining},
+  editor = {Teredesai and others},
+  title = {KDD'19: The 25th ACM SIGKDD International Conference on
+                  Knowledge Discovery and Data Mining, Anchorage, AK, USA,
+                  August 4-8, 2019},
+  year = 2019,
+  publisher = {ACM Press},
+  address = { New York, NY},
+  month = jul,
+  key = {SIGKDD}
+}
+
+ +
+@proceedings{SIMCONF2003,
+  booktitle = {Proceedings of the 35th Winter Simulation Conference: Driving Innovation},
+  title = {Proceedings of the 35th Winter Simulation Conference: Driving Innovation},
+  year = 2003,
+  editor = {Stephen E. Chick and Paul J. Sanchez and David M. Ferrin and Douglas J. Morrice},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  month = dec,
+  volume = 1
+}
+
+ +
+@book{SLS2007,
+  title = {Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS~2007},
+  booktitle = {Engineering Stochastic Local Search Algorithms. Designing, Implementing and Analyzing Effective Heuristics. SLS~2007},
+  year = 2007,
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  volume = 4638,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{SLS2009,
+  editor = { Thomas St{\"u}tzle  and  Mauro Birattari  and  Holger H. Hoos },
+  booktitle = {Engineering Stochastic Local Search
+                  Algorithms. Designing, Implementing and Analyzing
+                  Effective Heuristics. SLS~2009},
+  title = {Engineering Stochastic Local Search
+                  Algorithms. Designing, Implementing and Analyzing
+                  Effective Heuristics. SLS~2009},
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2009,
+  series = {Lecture Notes in Computer Science},
+  volume = 5752
+}
+
+ +
+@proceedings{SSCI2016,
+  editor = {Chen, Xuewen and Stafylopatis, Andreas},
+  title = {Computational Intelligence (SSCI), 2016 IEEE Symposium Series
+                  on},
+  booktitle = {Computational Intelligence (SSCI), 2016 IEEE Symposium Series
+                  on},
+  year = 2016
+}
+
+ +
+@proceedings{SSCI2020,
+  editor = { Carlos A. {Coello Coello} },
+  title = {2020 {IEEE} Symposium Series on Computational Intelligence, {SSCI}
+                  2020, Canberra, Australia, December 1-4, 2020},
+  booktitle = {2020 {IEEE} Symposium Series on Computational Intelligence, {SSCI}
+                  2020, Canberra, Australia, December 1-4, 2020},
+  year = 2020,
+  publisher = {IEEE Press}
+}
+
+ +
+@proceedings{STOC1984,
+  title = {Proceedings of the sixteenth annual {ACM} Symposium on Theory of Computing},
+  booktitle = {Proceedings of the sixteenth annual {ACM} Symposium on Theory of Computing},
+  editor = {DeMillo, Richard A.},
+  year = 1984,
+  publisher = {ACM Press}
+}
+
+ +
+@book{SearchMethod2005,
+  title = {Search Methodologies: Introductory Tutorials in Optimization
+                  and Decision Support Techniques},
+  booktitle = {Search Methodologies},
+  editor = { Edmund K. Burke  and  Graham Kendall },
+  doi = {10.1007/0-387-28356-0},
+  publisher = {Springer},
+  address = {Boston, MA},
+  year = 2005
+}
+
+ +
+@book{Smart-CT2016,
+  editor = { Alba, Enrique  and  Chicano, Francisco  and  Gabriel J. Luque },
+  booktitle = {Smart Cities (Smart-CT 2016)},
+  title = {Smart Cities: First International Conference, Smart-CT 2016,
+                  M{\'a}laga, Spain, June 15-17, 2016, Proceedings},
+  year = 2016,
+  publisher = {Springer},
+  series = {Lecture Notes in Computer Science},
+  address = { Cham, Switzerland}
+}
+
+ +
+@book{StaStu2009,
+  editor = {Steffen Staab and Rudi Studer},
+  title = {Handbook on Ontologies},
+  publisher = {Springer},
+  year = 2009,
+  series = {International Handbooks on Information Systems}
+}
+
+ +
+@book{SteWoe2019computing,
+  title = {Computing and Software Science: State of the Art and Perspectives},
+  booktitle = {Computing and Software Science: State of the Art and Perspectives},
+  series = {Lecture Notes in Computer Science},
+  volume = 10000,
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  editor = {Bernhard Steffen and Gerhard Woeginger},
+  year = 2019
+}
+
+ +
+@book{TAILOR2020,
+  editor = {Fredrik Heintz and Michela Milano and   O'Sullivan, Barry },
+  title = {Trustworthy AI - Integrating Learning, Optimization and
+                  Reasoning First International Workshop, TAILOR 2020, Virtual
+                  Event, September 4-5, 2020, Revised Selected Papers},
+  booktitle = {Trustworthy AI -- Integrating Learning, Optimization and
+                  Reasoning. TAILOR 2020},
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Cham, Switzerland},
+  volume = 12641,
+  year = 2021
+}
+
+ +
+@book{TPNC2017,
+  editor = {Carlos Mart{\'i}n{-}Vide and Roman Neruda and Miguel A. Vega{-}Rodr{\'i}guez},
+  title = {Theory and Practice of Natural Computing - 6th International Conference,
+               {TPNC} 2017},
+  booktitle = {Theory and Practice of Natural Computing - 6th International Conference,
+               {TPNC} 2017},
+  publisher = {Springer International Publishing},
+  address = { Cham, Switzerland},
+  year = 2017,
+  series = {Lecture Notes in Computer Science},
+  volume = 10687
+}
+
+ +
+@book{Tal2013hm,
+  title = {Hybrid Metaheuristics},
+  booktitle = {Hybrid Metaheuristics},
+  publisher = {Springer Verlag},
+  editor = { Talbi, El-Ghazali },
+  series = {Studies in Computational Intelligence},
+  volume = 434,
+  year = 2013,
+  url = {http://www.springer.com/engineering/computational+intelligence+and+complexity/book/978-3-642-30670-9}
+}
+
+ +
+@book{Top2013tdia,
+  title = {Theory Driven by Influential Applications},
+  booktitle = {Theory Driven by Influential Applications},
+  publisher = {{INFORMS}},
+  editor = {Topaluglu, Huseyin},
+  year = 2013
+}
+
+ +
+@proceedings{UAI2012,
+  title = {Proceedings of the Twenty-Eighth Conference on Uncertainty in Artificial Intelligence
+                  (UAI'12), Catalina Island, CA August 14-18 2012},
+  booktitle = {Proceedings of the Twenty-Eighth Conference on Uncertainty in Artificial Intelligence
+                  (UAI'12), Catalina Island, CA August 14-18 2012},
+  editor = { Nando de Freitas  and Murphy, Kevin},
+  publisher = {AUAI Press},
+  year = 2013
+}
+
+ +
+@book{Vidal1993,
+  booktitle = {Applied Simulated Annealing},
+  title = {Applied Simulated Annealing},
+  editor = { Vidal, Ren{\'e} Victor Valqui  },
+  year = 1993,
+  publisher = {Springer}
+}
+
+ +
+@book{VosWoo2002,
+  booktitle = {Optimization Software Class Libraries},
+  title = {Optimization Software Class Libraries},
+  editor = { Stefan Vo{\ss}  and  David L. Woodruff },
+  publisher = {Kluwer Academic Publishers, Boston, MA},
+  year = 2002
+}
+
+ +
+@proceedings{WCCI1994,
+  key = {WCCI},
+  booktitle = {Proceedings of  the 1994 World Congress on Computational Intelligence (WCCI 1994)},
+  title = {Proceedings of the First {IEEE} Conference on Evolutionary
+                  Computation, {IEEE} World Congress on Computational
+                  Intelligence, Orlando, Florida, USA, June 27-29, 1994},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 1994,
+  month = jun
+}
+
+ +
+@proceedings{WCCI2002,
+  editor = { David B. Fogel  and others},
+  key = {WCCI},
+  booktitle = {Proceedings of  the 2002 World Congress on Computational Intelligence (WCCI 2002)},
+  title = {Proceedings of  the 2002 World Congress on Computational Intelligence (WCCI 2002)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2002
+}
+
+ +
+@proceedings{WCCI2022,
+  key = {WCCI},
+  booktitle = {Proceedings of  the 2022 World Congress on Computational Intelligence (WCCI 2022)},
+  title = {Proceedings of  the 2022 World Congress on Computational Intelligence (WCCI 2022)},
+  publisher = {IEEE Press},
+  address = {Piscataway, NJ},
+  year = 2022
+}
+
+ +
+@book{WWW2001,
+  editor = {Vincent Y. Shen and Nobuo Saito and Michael R. Lyu and Mary
+                  Ellen Zurko},
+  title = {Proceedings of the Tenth International World Wide Web
+                  Conference, {WWW} 10, Hong Kong, China, May 1-5, 2001},
+  booktitle = {Proceedings of the Tenth International World Wide Web
+                  Conference, {WWW} 10},
+  publisher = {ACM Press},
+  address = { New York, NY},
+  year = 2001,
+  isbn = {1-58113-348-0}
+}
+
+ +
+@book{WWW2010,
+  title = {World Wide Web Conference,
+                  WWW 2010, Proceedings, Raleigh, North Carolina, USA, April
+                  26-30, 2010},
+  booktitle = {Proceedings of the 19th International Conference on World Wide Web, WWW 2010},
+  editor = { Michael Rappa  and  Paul Jones  and  Juliana Freire  and  Soumen Chakrabarti },
+  year = 2010,
+  publisher = {ACM Press},
+  address = { New York, NY}
+}
+
+ +
+@book{evoworkshops2000,
+  fulleditor = {Stefano Cagnoni and Riccardo Poli and Yun Li and George
+                  D. Smith and David Corne and Martin J. Oates and Emma Hart
+                  and Pier Luca Lanzi and Egbert J. W. Boers and Ben Paechter
+                  and Terence C. Fogarty},
+  editor = {Stefano Cagnoni and others},
+  title = {Real-World Applications of Evolutionary Computing,
+                  EvoWorkshops 2000: EvoIASP, EvoSCONDI, EvoTel, EvoSTIM,
+                  EvoROB, and EvoFlight, Edinburgh, Scotland, UK, April 17,
+                  2000, Proceedings},
+  booktitle = {Real-World Applications of Evolutionary Computing, EvoWorkshops 2000},
+  series = {Lecture Notes in Computer Science},
+  volume = 1803,
+  publisher = {Springer},
+  address = { Heidelberg, Germany},
+  year = 2000
+}
+
+ +
+@book{evoworkshops2001,
+  title = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2001},
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2001},
+  year = 2001,
+  aeditor = {E. J. W. Boers and J. Gottlieb and P. L. Lanzi and R. E. Smith
+                 and S. Cagnoni and E. Hart and G. R. Raidl and H. Tijink},
+  editor = {E. J. W. Boers and others},
+  volume = 2037,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{evoworkshops2002,
+  title = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2002},
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of  EvoWorkshops 2002},
+  year = 2002,
+  aeditor = {S. Cagnoni and J. Gottlieb and E. Hart and
+                 M. Middendorf and G. R. Raidl},
+  editor = {S. Cagnoni and others},
+  volume = 2279,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{evoworkshops2003,
+  booktitle = {Applications of Evolutionary Computing,
+                  Proceedings of EvoWorkshops 2003},
+  title = {Applications of Evolutionary Computing,
+                  Proceedings of EvoWorkshops 2003},
+  year = 2003,
+  aeditor = {S. Cagnoni and J. J. {Romero Cardalda} and D. W. Corne
+                  and J. Gottlieb and A. Guillot and E. Hart and
+                  C. G. Johnson and E. Marchiori and J.-A. Meyer and
+                  M. Middendorf and G. R. Raidl},
+  editor = {S. Cagnoni and others},
+  volume = 2611,
+  series = {Lecture Notes in Computer Science},
+  publisher = {Springer},
+  address = { Heidelberg, Germany}
+}
+
+ +
+@book{evoworkshops2004,
+  editor = { G{\"u}nther R. Raidl  and others},
+  title = {Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004},
+  publisher = {Springer},
+  year = 2004,
+  volume = 3005,
+  series = {Lecture Notes in Computer Science},
+  address = { Heidelberg, Germany},
+  booktitle = {Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004},
+  aeditor = { G{\"u}nther R. Raidl  and S. Cagnoni and  J{\"u}rgen Branke  and D. W. Corne and
+                  R. Drechsler and Y. Jin and C. G. Johnson and  Penousal Machado  and E. Marchiori and R. Rothlauf and G. D. Smith and
+                  G. Squillero}
+}
+
+ +
+@proceedings{wae1998,
+  title = {Algorithm Engineering, 2nd International Workshop, {WAE}'92},
+  year = 1998,
+  booktitle = {Algorithm Engineering, 2nd International Workshop, {WAE}'92},
+  editor = {Kurt Mehlhorn},
+  publisher = {Max-Planck-Institut f{\"{u}}r Informatik, Saarbr\"ucken,
+                  Germany}
+}
+
+ +

This file was generated by +bibtex2html 1.99.

+ + diff --git a/testbib.pdf b/testbib.pdf new file mode 100644 index 0000000000000000000000000000000000000000..cf8c6490671973581c977d6177d8f8cedf476b41 GIT binary patch literal 1377784 zcmb@t1C%DswkBM*jV^ZCR+nwtZ`pR2?W!)jx@_CFZQE9troQi-Irq$+yXMaN|H+jr z-klj+k$cC^74H*CDlaTb%SguxL)!I}`vJp9$UtaoU;)F;O-QfoZf8tLFK1w(SRX9#>&h{NH1z`7rmlwv- z$-x+44db?Yp`+!9%U%+Cp%J`NzsTZR*OpBRx%&@O=3?zKPlhf zN0)ckU{Er)CNcU*dPTS~3f8!E-i+BjxS4OReh^JPg)&hBofA?1VGv~?MtgOToh)W$ ztp80frlN#~UAsPwqArECXOBouB06|-5HdswRitC(c5C(Qs(*Sf*AR>itZ6${Gj+SN zxYtgU5PJDF{#6d|WVtW|L-RFUF`*#|q{*-w5E9OUy*5lhq){bOF+T;pA0<;Tkq`wZ zDQp-QX?!CP7~KpZ3n(^t^*o$JEGeOssS^>fhP|k@kSGXb`*0#eE=nvq5gD+C`*szO zhGVQSa4m*e28fZo07h}LrJ=L{?K}d>fDb}@EU?k{X>AcCgKAr>S2Oz>e@j`pu+M$0#HaT}W-V#XdemS@2RXa#n90;pCk zrzw`vtQOeFHj^IiAsq6S5X$2l7BPJaSR$Nr0)#cZ*(6OJcq!^Q=RKVV3{MeEMAj|{ ztLCZ92?-VucAvaI9q40aweL7j$|wj^_%$3MY?v!8A?z*%>~L?c0%184{=ADXO%EzK z4t2gTCJ+Y)I@~sp1`dPo_f2QNv2F&qbFuM)(6XK=LCS71+#n`1So>s$-@cGW6Z~5A zp+_@0{Zg8K^4lsotR7(b7^@&@MfN*07>L8cITrQeSm1FU{A|G~kYMP8icbLof%bwQ zQ(K?9%JgX>Jf)bd;oyVo_6i;4Q|X+8I?ZkmJEfGRLfi~lS(w>A_?w)~y4upF1ExMb znHoy;t0(*R9cJNE_@6Qi^}nzdK}{dN8-7N7zHoXRxxbgXE45o^u-T~bWI62`qC)b@Edn^NOV$x@0>S;?KMZDN8>e9mKS?DlvJ2-g#X~d=j=~ufj zSeK_`-L~`K(U7{pKG)=ZxWbH92(!4tKO15K&f-hR= zfPTXshCi5ZiMo?6tQWtXIxib1O*G*oV`qIzP&Vg^8kaK#JEWAx1`|~{S!#jg3#9t_FfjwwxKnoX}QA3&M zIJj?WMQ4vKYe7Ul63Y~3nlrB`N@811uuI+0-aF408{PkigwON-A?0 zl#yM&^eIRW$~SdaV#qS_+u}oy9R^B{eT+7rNjWrp_ zC5kEdl<~#k(w&JM0#Hrk`C3fyO3N%K&Z&&!d6L8Bzw;-tsq<*H1DlqDW`GUIX#;*HtT5#-Xqx0UG$_`tmvFF4d+M#XL z_}abiZYOT^7?u?i+qE|=U6}+}&5yGR@+~M$@a<+tNfSrFmW^u+KM8;K4~ddupT#4)2D7S+7}w?t9Lo zQu|9i9(!CfYwD-+qqEnF)lT|%H{gqFeFp?XoMigdmv7RjT-d?IVDvO;V^_oV>ZSp# z5T%XhrLXUQ4j*FVI(gtc``?1!B)A+Z*sT%Sjxsj`Iksxpof*uIwZIs& zKS?Hf!Hg$oy|CuYuR|akia25Md1ccZen=F@YDHNFEf$kVAE)^$N)o!hPv4~<|DreW ziwQC}vLE`*CX~`GpZAl6QeZcvn(%XY&XLVPWJ*zT4lS@_;_2E?OFsz{0?44DY!p;7 z>2T_tDG^&GO(cg*5ucy(N|E%1H<4O2TRcr$RQ&i}p_Q+<9ECl)c^&=z2UFr;7TGfO zb4{d&mTh~5|3z`iW%OB8lcShZ4a=tM>89J@tMxNEPUVqRuE7V#`<2S=?D3U7F|S z&XPuVO4L);S(>FXRff5+w7|Z=p4XVldgp>b`$TKBBH@f8kYyBGRl2m`y6`mZ!>7zA z2)T%?>%|q|Qg(Xq%1SqB%9^O7%Ub8ogRSzAix-o2TF;fa(5II+(9Ef||3-5>4PcGD zHmK;!$R7=vW7L6Y_))4V!*8?BzyFIa`ii~Jg{{4l8<+j6%p^=1l`;HJYs+tl zRi}DOb0VNDeV{MIyr`81E~*p6QGl#`X~z{&uB!7y^}W^w2OS+T?yBVGcg>En&AyBm zB-=ODcMRZ<#?zv#0?;`+$^jWKj#%E!petjLF<0l@)6@Q}=vD=CgL>mk@lmrf+4GW~ zK7y>Sms^tJjY=FR9agB%vLZ-hHAJeGJp6o#PL3JOcf?;uW^pqhTTA@+(?flRcat;( z<6HPuQaiA>6S+P~QvQI*gsioP?*nN>co`02%m?B16;z(nj?cgSbQNW&Qk0Cef4SQJ za$piYhLCCIOtUST9}wkO=FS=gN`U3Y`n_jp9z{Bj*X4*?zpx`BbB2@<=Lgi3WLI!D zEvt>XC!a-umL>I6lA*0+Z1r1~E^kp@l7M!+qVT}ymVT1@;zNuj4^e7shLRXdbHsWo zKvm~p|2)_sJQ){ZZikJwzS6O_@Fm5hB_87?WsrZxN=C{`0gM`+Keww0e9_1o?JSvO zIKC`d*!B9^CT>z~@5~)(W+Ed=H);1pHK&7WYWJsidP-I#f76cBA`SSxEe|IRK+-0; zZ*`z8Lr7fy%H^Q?a&#vvwp{@M)FP>G@ro)WMvu?y;lNP&=2S%tVQb$+awB=Tm7T!A zmJatSS_dST3JV}8M%b(&E`D8HdC7}4UJ;VYqdtfV-X4$hbq7w zndHbZ(N%fI80Op1lHsWNY-RHPOOEWw;#sbeUXF2g^=`yr)|3DlN$fVdaEZ<^%_Q!Q zM$uG$%`I6Qf!M)SJpD5EGO8&(xE`a7NXFnG=>%eX8h5KIy7T#lVcvBfdy9L*cFVwQ zWiv2f*&gT_3KX{7MeYP_F$REpkGF1W6TvI9Bu_QCGcLm5 zl{vr~YB{latX_H!NxN#sgEVrf-%uSi_p0-~fbD|5A>{pm8zy;UygJd;hdjWl4?zoC zpPgf01I=fgpljvzqu9Q2N9u|q&-E_<$n_ybt@8HrqekL`QH}22LoIZz66p$hA^j-w zVw4WGZPi$h?}zuz!kBu{npdht1kjVKZjq*|^bh8!%pWdi`q2R5GbSCdkuzT2Ve}0p z-JpzH3yoIY@8)n+QgTY#B!T(Tb%|`_@bX4RJ|dE~mEtE8z#lZ_q6@f<$E2B1gvd#_*+-U5*pIY8wE;vHDirI*m4O9Z>PQk0d68#H-a&O)*< zy-T2a$Jy;G!3Qz&=I!yV{E>>G(50S2yV0C?fAhmBa|MwoaPA091<&71L--?ayN5H} zd%6SkDf+{gN|SF2em^Nu1?eV*h|V^Ig`(Pn)XaKOseSu{51Jt!KL!OY+2=+5FfU!!&GUp;V$}e zfen0PVH^vO-OL20&BQ46?8Bn+%KWgBp-qALEb*uDa%I{}+f|JB^P|f3_9Ham$y5r| zidJgV&+!ziv>q?R7u}G}=sHA#Y-Ekkb+#Wn?Dq~EsX82MMlZvcc;J~DHy(2lt^{G= z*vD?{fs;=i&uF)egC;!q%QF#8O+)_0#x%nMq3&FirAtF%FAV`i@MG%tD+31hte;xo z_DW3$PA;-Ls-{>r@c`y31O$X=2!FM@L#n`#rlN~x9ZxJc`m`PTJ(ten4-|s}d>PX` zlcd}lb*NPs%yy{0i2-#sYb6&S#J#ksT;sgR>%tM z4tlYk_c6L(i zfyxg{F>9V?{A~Y^7O6*p^_IBapG3%#qBRhk$yqMx;<`+iu0n4qs=+8{_>P)yBkrrB z@pphLxzSf*a1*s5hBIMrOn<5&A-pwWzKe)wD|QS95nIVb!5c#ZMSkz_AvwS^B@7?< zPSNr!ng>BV+V83+VYJcLa0+SU8rLX_x7)QV0x5$$-SOu*nZ|3d2C_P=s2%J&Z|`W(VM+_FZEEU=b_oKR=fpkY9!!w z(iu}oY&>~?j-T5BaZn)&W|bFsa60| zf%{%)moB4XwP}us%Awo{$;XwbRkd^;@STdD(@nxsMUi7q$iSw#d~@-lTsUzSwuYd` z3;L(HgbGy}J?NkMl_15)cSwOp% zm!dXj=A{oJt*jxv_-#sj-ygsnrLSr3g;}_R&oCcu&M}i!CP#>t-ZY1;4&e`{az%;G zJY;tuxf(>8fv}0URGELU%eOJpv7x{f)_c@WRV&r(q|`QPzFL;j*j*r#Z+?}v*=SAM zQq#3HGT&vpwpu49hu`6SyVQqCAN}?RCF*Lx?*3hJVGE>wq9&s=s~%`gzU<+TALIs* zrVZQSyceLkYwJX`6Q5avc(mk06q>-m-dtfOUx(7ZX-_YAu2M?L_trFJHG5TuCRX#B$Q3rK`K5;w z4?mG~IEOt?5_eBd{hv?#J$XfP2wgyLhIPoo{T^`30R?SRfi_F(7u%V<=L7=^Rcxgj zADHZlLn*d#vel;S-7B=`eG7-JccS_udnWD0W%Q+{`nmDd!gX# zKNVnWF&nC90)8-uHWMWKjFne5*$_g;{q#&Y0KL1@C95K)#gQFGE>Jc(d6&>;dfj%M z?LvJma2*2hNOHM_B>FF=fK$t?H$8VJN%MrngGoV{z}5fB8f;S|%^d5OT5X{tBF9L-++1C1OA+(FZz>=m(+zRq5Sg}sh z;A$nywCbA!<~VsvcY2fIdMCs4h@Whi5Zla0wcV{Dv`VKux*SQ63Q0XT9laQ5Et)%_ zzq|G13NoXUOLu$!yg2?`j+r`we0FuSyI$B>ymV53>#e3_4QgRx(Jp}`sM1zfz8H;7 zdmk(VJ7v&=*H+ZLQV>n9PY%m3e0=K(di%CFkqW$L1*8`Pf3Aa;VDim;Q4XhgZnBt< ze-Ybrpfu-SN?5t1%=Ft zsF$<9I`pb@!a}?G!C#v=`KiLKf>&zPl zrO=~)2c=OFqNhtM(_qdXLO)A$!BWYmESnC0?y%+dt1ZmWyez?%iht}XL#IAlOpGOt z!;t8!nT*!F+b$$F-1dg)*8gT;!uK6BurRNlf`oAJBOq${I0)T@!0jGKf#>D&Iz9U= zn-Q#i!}hrM$i6$iSH%34>{_jNkUm%286gGtf@7xLvvKBweg9XQ!aUYfp6CL5>x*fv;j?I{iL%(KI=kNp5(SOLs zSnku(KN+8OqOjnyW69h|mkoP3u`{U|7?rH)=-L{UJYwdHOhkW>> zooe>P#r+vyp*<Vf5)+AQ|Jgfu(;n~$f8foCrG%|Z%%~G&f;J3;; z4K&HeeM>IQ#Ohm8`T2q!Pc%Q2_Acnv95i5Z>`8^eLO(IbkY=45#`9F7I>s^~ghStF zvEq%|6Di1ZqDJB*5uzC4EBL1ccG>1*PP!0=SATrxd%r$k_SkKlHz&XE-7%2jfcZ+K zw^P6&=7+KYVpadGz{v~JtEaRF?^!0T4+7yQ4*v(d?;9%63ntJlB9N7m^HT#}(48E6 z9ug1A_itRya3z?p6QD}`E!vah6u?=k4qY!`Fib4~%fY=z8b@0!84+ifV7Z8?l-mRW z1cc*Nx`lzVPjc|a7D1ZWk_-f*Q;j{Yj&s7EKU+eX=&_X=iaYDif}>HiHi7+$QU~Ti~vr6e}O-Y%EsdQB z4UA3AZFtD9+PcUI&5d}-)mUX1WbA~D&CJC;9E_DbWR(p)EDbr0$oY6-xcLBI_Kj)X z0Cuz{f8$7wgvO4BJj6nZ(xSwFF}QKLS=(8E87FkJwz6^La^oTYTZHS2{_8e9IpNN4_#PPVpImgY``Omv)djI@l*G=wG&0Bd7cTL()*Mmjd~FO?2PCR~cbqW@I* zb;Uz&=Hz6@MNjYQ>PqLzOlRw0O3%p2$w|+^M9;)T`^7=)=x*Z#aHF+x{QhO{Z%544 z|0QPQNcXo-bcVLp^ycdHjC2h2|I?80*Ki!n?VQYQZ3zD|3^1^DcH$u>{-+kZf3yC- zsQK#n(!>0vhx?1Bi~C1Y!3ZsIQdg$D96 zF)*>vGO*AxGAJ`}axt-RF|ufK)Bh)?uONR(34MWsU-;qwS?b@I{!Pl(LHR4{ywU(= zV|7AT^Dku5>??g89E?r=Ve4-1gxC?L0Fmpm6cpp02@;tayMEdV-tY0mD3mh-}z!?{!ij>e)|8=@kJT_6+Byq zf2dJ1w*Q~{{Z=6^5cUjV(fwJq)t^2es0JpnG* z`qL7z3=t=?LASZ~SOX}UpB;(Z?g(1U4jw6KSLf6@EhVP{Znku-qy_GL$JNC6eZQ`#^3z83KzsU z${0ovawKS?ZuRA!fQ0Y4f&i!0u%Md$c#NQ%{)D8UF#P;nXKGsfrPk`z?Tb~k#aXEa z+Hyz*#kOA2YWfEH48mP!3T*>y>DK;QjQn>8hBCFvdAj`=7| z!rJyEo&rW3r-&k%JBpYh>oS6>Szu>Ma`M^kaddQsF@Jhtw24;C7W)qx3ROYE3ni@ccdSdlKR*SOXg-M-@Q1gMq zE=g|r*KtuaL;EQ^zLR~3FG>8;CV3Id6)Y@EbISn;pxh+=;jzCu$#klU>1VY^id&aJ z5Qv8pNA@3BLmOdr!@(x*&@wxpQ364u6tJek`?w##1m~kVzB2Fmd^vUR(fh0~x3igD zn8@XOIMP_6{p@|YPm)PbV=aTF1CJMU!}r2E`NeEO63Jw;Kb8Md{(L*;PrzPg-az#M z&dmSmL-Aba1Y~f$t-^X@2Y14(2q7K{U_n>qfUwCJAi9TO&#D~K$vqb7aBC{CsWIbf zKlG8|l}3Wal9RO*HB^`ABhJ5yG&=6Zo89vFOHapD)sr2~3dyrys~D9SDgfuZVKmY@ z!LCQdP^^K90Kqdw#;|WPiaE1s&uTAg6@DBLaaWanMsi4umhN%GL97K-*OS+(0Q^mR zOFAqiU{T4ry#+OXgxW#27^Da_rG#jn{Z46MiB}J{?=a+?8a1Dz(P*I|uM5Y*5-c6; zo~x!8RTpi9tHVcSsUuaX`9ap8lT_v!laH2^K9lhqR2m#ADO{i*#}w3RPuH2qg4eu~ z4$~3gt{;dq*qqS{v|b%ooDJpCSM}$rV8n%Af*VO-a0Rmg%kU)d3xiDEFBKYbra6v6 z_f7M|_e%7a+)I-)#(-k?j?yA$PR*$yPnAryG7H9%RicEDZe?ktU(mh~`Vbh<`DP=5 zZ~RhE;U*RyL)5F}G%YLSRt+mf*gkb~b#|L-ESR>BjF7fj1Lr^#sZYTC#+#5d(m!!? zkcq5q>*m}++TPjrt+&F*ZHa!b*{W=wwzIlh-Zk3B>@U)^TKF|lj0f^yZ>b`^+B>EX zSmqh?cLUF9iVObK$fa&I=<{4)J45#%d$@*9lN#pK`PpJYR)p>h?(?z$y|1<5V&>;j zjrlNkoBK!n9YT8l!I&8^%OnpKz|Y}{gm4oM-8;|VioZ>hb$pTF(Mzai%x-v=UR$PmR|trYbG<-GvS%dnSSfEyL(C=6kMU)VWuC*z zdlonbQiu)Rpoa*QII@+?7LW#E>7HR$JXj=I;%1 z?acxQ0nsv%C209fcmR%V@#4&i|ZL03NOj(EmEj-2=20Ak=*C z3yNP&$duO$T%3L1FD6X_H%nk-lhD0x>jzFe$MC+gjzqFdNwW@mf}6M`3Iq}xAW4OY zWeC)L4;+Rl-E9Km?q8K* zpt~s0_Y=s95s&%^cw>Vc>5&)em7+TFs!(35YvQ-0seA!*@hrCT30$cYggI9L3m!g* zhKCxV)KH&ax%c%##ND0q8ZIJ5O>CdCdiQ?!CFeMIJ!w}jr&Nzjb*CP=IL;WrGB4i9 zk z`>17=`EkWZkv`;JRaE`h5O`H6#y{?6^Ar5*V1rRXr~GdT_dxBYqUS3)(^&xqMZ#{8 zuc_D2WHsfKOfZM-?%`;*Xo1_=weNo!i}GT4e}^$tTQWl9ucO)-xS^ln8m;&*GDkU`Uo?C&6eO5xy0 zVmrWm``LVu)JrstXQ;kzY1TtQ`>b{_P-Zc*UBvK%hAd7wjHwU=l!m0vlCl9-2)UQM zulh%yKCAw;El${Z*#3_a3_@$`M=L8fVm4gl^r+B@tf;Tki$UAU)mZs_vv(k$EaG9o zZN0jj*169%ZK2I&kw(kg^KIiM)IdH!F%Cx@M^-qz#bY$%(9Ir(;0j0KD&vp!D$YB} zZv{!k)(vtSeO(u6@0M0@IMOx6Xst>T4TpnWLUy1JoN8?VK)mGJkXwr5s2a6zlu#lx z6s>3$6u}F#aY(9gij5XlbFA`B4)|)&a#R#h{*R}w%g`n&@ejuavqJd3v5IdPcB`#X zsh3e}&;28tt&uoR9#p=5fFbn)O;p2qV6yjGYciIzHI_527c4bj8n}a7Hd({_llwU` zZ`^Oe>_`VXRN;b&38CvhJ%=O@s24V@EW9Gj_&aYmT(m9%SP+#LZs2YvK8i=8B+~!$u8jC`3#@W3RSicJGZ~ z3m?zR@&4sxYvMgi4Kc4(>e!m%Z;v&veL}*Q{$y_ z#R?PReUVh#Tm!!!v_!=F!ATz*wrA$-dAzx+y+dzt!?P}q^)H_v3$II8*)2aGlcyr| zwAY@lLMk>JQBUoBDUPLD-U>_UO*lvt;^a$0KF8&r9Y6fL_KV(FefqX;<6>R{<NB6^eqj7pUUr(a_t-h>sGe2^a2fn?iNET)yMH;)MR29YApQ(n8|Lk~BrS*L0WGjaJ=;agHKsjmcNN!P{E+3$kIl&ZWr z2SN^X8h!t21tv()u99yU9P@uDl{#s0pwWs7YdxU*k?kVGv|fbfb!g~D4{FPy4;6@` zYxx7(83PJ)R~8YqatA;f?spNz@ccziZP6%#>LD>NdzH#S0$M zf0D1}v927AICNm48+yox^knF1vf|W25icHHFoNK>gTB@DZtcIR#>$Zbhf9aqi;8GW z2HAVBt8?BEr!~(6`R1wrr#~wk@Kj(GpHwk~N(D$)vNObi`DTsFw$EU_v>mTlqAF5ec7e;3U zJK+(@ESyGo-GE?|)$UY!DWy1`RYbDf#*#_b>yz{>KZ}ugh@3PZB)`KQf6GKt9R5(^ zIDNN6qn)g`&+aMtdMOb>CJG`@1$|S7;VedTkf+@LtvD=D8XKTY1X+^nMzM_Mf^HRl z>s#RGH6j}dAW6W2Mn1W!W(9)LUEe?nKK7e^z)xbhhdnrqls5FzX~;A2(y3@IH9n$l zD{Hj z3-4#5?cDUzy#el;ru=wFH{HriGzIzGHB=?S%}GUGQwMP;st5WgiR&1UOPQawxYZ62 z0k=(P;^TK7D;+(9I?qeypB`R1MR+RR*xS^Z^M=%RrqbK2^aBb`e$^3;U5sFrnvQ-! z3lb+o+|rGZgn#xMIvs;&K^g){T6OzHQ296G|8XDXO)%MbS2^&AMalJ+O*}v z*JCC^L7^{2L~#NVvFR=71S!ri7&&1uXu@b%mr|o~u9X*Kdaw0@BpUIf*X*7rIz|Gj z)9dbhV$GJ9kB9R^w;53uO7!zAP@kns5-J=~8mcux*tw)?^Rwzih|EjjluZe;A?GU9 zZE5S0k1#t4q3NYQPEL`k`~XkyZtYcrx@km1fU&@h&zAJK7Hx(%p<%SZ?x|L_EAHH9 zOcR@-ZbXywE?tBA%(WRt_P)$65bv=9vM)g>+nnr}8Ar4rCXqu#L}dqQ;GM1imoIdb z?jT^`J)N@K16;fXJ{j<6%#xACaTtD2X_tQhpN@&qq_|CP&?m_X+(WTB<{lOjgmMstJ zMLv*xN$U!TXE+Br%Li27&1U(3TP`y(u>7~nWmZ=9|C{~n|HX3I5x42D<+5MQh!X;& zz90Cs1he7S-V7F59+$X6bEY&@=cHnJ5-G*#HN^J|FAOls1%)Esu>6Z?QD&}dH%JFBh1g+@P8UKT4L?QE(d+r5Nu84ZhsT^-9k`0XyEAnlDk!r?^oC~ z5RrgzR&d`tc*t)@h0#&XLcuXgFvU(Szl9F`;V^mxw;QHhEph3F9x6Zvj_xTM(oZEE zfSp$ri!$6Vy)4`kkN(4j33Om{kJ=W!Zu&fPm&w>O6k?u}fflAI1Ca`)71Lxuf4`!b z*wme@Bqh!>QH6=AN?*CQ0D%6d?{`lL8ZkaXz9L#9Sd}6|Z`beI2ZxXv4w5P1!~g~q zxCw>9UPn&jdr%V%!@$o-(*9C?r11fqDK|JNYj{X^gfg?Z@wl4EX3DKv8jlmwTy&!e zjvRDj%A3Fj${bT98zKoeIs-Uzk~Z~yVK+2Jh~R{WK`gpM8tNF;BzJ@Gq;MmFH=18G z0{vL~QuJ5>Fhv4Q1%Zi9>U?KWGji(KvOq*OzmzS{$5aafikDBt zhOJJw(pB3J#w&Zb2YUq5u zJvO$dl5<>kbALqZJa=Lh1h^FT&UfLxNLq!k^apk|hdNN#&Lkox!jkEZ(8tOKwpEQt z@a#^|8sicm^@qI5nLzrH`5T{~3f^~<`T!|4L=NtH90`#)Edz8WHl{cSciVdw%=0dT zC4Pf`CTY!aLr#z0v`Q7HQ0x+T>z1%EUl9|)%$TnFChzAi^z(9;Mh>J#39D4LYtcIbnqAb2?{YAdOGB90 zA{;W5x{tUZ1tEO@T@ujFwF$@xoGak7beY!l`*98wiBT-XNzM~=a4=ec1=fA|>iIRL z=XSoWU3%0U7>yHza37@+tj3-9Rks%@J47Kb>&x&T0f+dC#$B(X;vFeNO8@9kBnmMZ z_n~0Z`r|^6nQ_cI$4K2n#w$@ip#bmC+n4MvY!eUhbebNJzQb-CoT?5&P9YPQsP&EA ziDze`rXRYE2%aG+&j}DC97bbO!(gGoKU$F;njDKP)WB>>*1~?V1I~UCF3;|cZ{ZHT z$`{ugTZJmbwjLPFf7Xut!yJ(yYAJv^BOE zrD?Ljg6F-&uDlX|S)3(VjAQo8E9DUcovh|aKrk{VDHV1*6EuuDuNa2RNDxoe=+hC4 zrF3i78%$sKzI$#HGuZRf5#4>)01fQ|i?< z$jiyS8lI(|=k>zG z?LSWTRtR$4fh9bty`DQl6^*+`=&T_RXC#Pc8?mn9F z#CPMHLuqg|G}$8Ac6j(1AwIB7s-vDbEg)`thL$AQ)C}iGhCtNLFq6j);U+-{pvc?pBj`4ZwD&+|QD=E8m!WXssJRpOp4 z?GM$W54p&9_7MsG3XQs<`#~8yw%RA0i`(oRLSc@D{<#HV9p`ox5K3NDkUuF#!ylWF zJwvnU1Ojp!?O1+<89(3wiC8=hYFJn_=CV2QI#BB+Q-invp)~^jUD0pMyge|2=@$?t zf;#%MZ+fw6@1TkLfvkTqzf|UczetJUy=EzqP8?6u&sR5~rTDUqeGB%MRhdec{~+D* zD%~^H@3?9nF)e+*ns*r&wu{lpLA&MqYqqs*Bo1g{IH}s*#PiX~)C71X-{pPnNGleX zs~_fQao4JJ0QG{%U$l^la6GKFC3_i@v4jxIKyWTn&-7hyxmB^XxqDU)?lW2zF|PwB z4h_fUXO&GMgB|Jf0sHO-t4UK})@09r-nj%+qr>o+V;=4?~MT zpDmD(=!B;xdx+ zK0}~wFy-xn3Zkeir-*m9FCeE7k`JE zG8v7$%_kL6KHov4iVf9c$LX-Y`{WMR^tHp|`FIO z)(>R#c9Igf7R{A0(v?_bTP=e>9g9&MAZD3%W5Q&Srs|Sllyb#f4~tz*`kLA@(G_Vl(Dr%W?t|byKw+n23S_tfQR? z&CtW6B|=ky=+(~2cOPyeM2es#!Ive4Ac`{J`DrlX=r~^`Ryl{rma4D$?^k36x{-6R zktMiBCh3A9t+J`)gF5WOiXzXV;A7!DPr8mu#>(4fbPbS%Oni(@GM8*;hRF&>m`RoK ziA(9FCGE^K%8D=V<=Vl4=O!<$t6na2lq>wTg*p`~1$B3QY7d#~=Zv&|8(64f#Xcix z$@sx|-C&wG7&Fkkg#p!&TAogy4VXt_p2B5EIq>Eb zC`9gs83S&;35)|_FZa_jn>(c4-(@$(MiFD@YbRKEPyVk9wN+u7bJxi?Bd1m^boAx% z$$X6t;oE=k^iZB%U!)EKCq6w((L51eVad??+P?PBio-Qq%{Iiez7V|LoU9yvUnaqL zjek6A%%&fTUzu+#y2x+i;5}KD=+bV)Oui^>d#eWnFUTH5Ka^{oBhB#dZ3w}A%JEDF zKAwG*JEPBzB=4AqUi6B0xH`Mc_4(P7UD-O~cCYWjb#kNsYSpf7a5Lth&)ay{wW`ZsT9wQSTPr=AvTy_`#YEV;1nXf z|F#Xbe(Mj|7=juZf|X!oR;m(0Q;Dn_Vpolw>uyz^!OH<^2o!m_j|ovr^V4~zb)0E3 zPO&6vjE#yA8B@%#N1JtZ%*6YKV&?{;FFQS7ogvZHig;%4hf-k|%r4JOwN(}()T*%@ z|F@sjdX29;fg_Zi8Hg^AQm1)-*fv-fsxWXZ&3YJW^j#2yrd`|E0mzW0Wei3*Xa3(r za9l#k6T#Z<&H~$tz!D!0_~D-IvUyf?9{Z&hoHjjSi1*o#RQd_l|9D>y6(San-7I_+ zmnyb5l!g%c+8b>&@zTx+|E=6_?J&=Dy%7)lO_}=nKA;TE2Iv3Va*~;W^}k(CGIIR= z3G{#Wc{C^UfBHOH+dAeS%gLUY3rKs;KB5q~I5cebvX|T{htd1RHah}D}t^tvDnbatqqFowl<-SF) zYkxGwjzD;Ht}si#LU_2SL;g(X=y3!%ytA~C9{SLtNdBq1h0Cviy?LDdW(s_hNDM0q z(V~oV>UH#*Dm!+uwy3tC8KMsf{;FRQs1P;3%1~>5{N}IaK^+OB+6+Bc96rX^A;D<- zPAAc?#n>|M5xr7C{21B#EwV>l+r*RLM!OL*4=g#J=lgywnPv4DtNM2?_=POwqVi*n zI3gc zBPpoq3HgB>XAS2E|7<2S9%3ldXm`Dk?pPGyXsX_=!Y!5#yE%+ z0hyhiU4$JRMo+`19+YRgC~v5MH#6$SqAEC{*PQ-d&k&3ut^hJfAzA_!sZiX&AnI4| zb9Xaj^E|%tNxy!be5BX5v*x2mYj0blQ$g?A*q z2mey-!<-UE%*>g`9p(+7pZN9+L%LVW%db&w#iP0j7}2vR`g8g*h)ZWt8 zf!Awlqo3aAx5uD>i=T%)^&en$m|i%#VW?grOrapqFVJ82q=_Q9qbyP)gh_KBG^La# z%1-q3Q#eBL3<6+6yf8Rw-(XiB&7%@B?^(n_P`&~k^@MTvB|5?TUtNQgeO_ZPTOvta zPgo-Vyc^H(m-8HEwS&j#)UxjCazDX5W6w;-ZpJ=jUwcS@aNu2|+oChkH;hB~H+%?= z9rwV%^8#Xa2gOCq_IpQ@ujmA60lE5<-~tC+nV}tqk`PKpi3|ii9^;n`<2wFI3AB4X z`UZuCRn&bAHcg9!t_sXsw97gaPcdWiV$q_ZOP92yL;*Pq2yh46vP9~2fVWc*4ep(JaJ0l_+H8WaxLgY+{E0h*bd?SKD4a64hN1VsCYZFi3OfYR&Ze`ofm-vV zpR;gD^`7>J^{MR9@?>el3`6OnvLk!>HSC{F^V!gTbLTo8y#V}C`%m*{lQ8pV&-cm* zT!uBncH(O}nb4`jl(dR9SVQ-Y&K3Q_)kvS9OHH%I&Yz7M^tv?G7Qv=^-r(-^Th=oH z6UVamQ-@o$h-W{QqL`>n9SEYPsPKO!hV_~XgS80T=JF9a9q-r4`Xi|4U{DBcX-%u^ z+cVs8ZaZ?&-HKhac)o^odGAcj1DTd|OG^i2@N#BCx6g ziTo91<6vT7K+at9d;q0EJ%2^f1-o$Wx0tN%?rZ5psDQx`(kx@3lQc1&l=u+JY)zmO z4bfjD$t+{o{fFWe{lmYSg0||=Z!i)0o?JA7sgNl<4=ZmpU}PQ1h2Y~ICSybF%@ZV{ z8@JyX29P~cXaZ(9m+3#m`ns*JN1%A4knd>cv4}sq56iwb!0G+88qR0ffXF*n8T)Q@ zY)Z75l}*9ldP2$a{wXvE$+40KzRG2y_`6y61^M)Hvz>x^e2g|po?qR7i!DXGQ5=cv z15U;c$N*wko~l#xF$wS{eG12qw`(4daNGo)b7y&rw9$NqDom{L<-04`(ugfY3 z2~OEH5|Pld8HoX~>XvCB^xBHS@YjQ#A7wS;ImQ2jw08{7B?`0!W1C-W+jeqd+qQXP zJ2~-*ZQHhO+qRSQa%b+$yqP;y_ti|*s$IMPc6aUFs~0vyC@|7k@vxi$w_-Q-AR`Hw zh1ww_oxC2r(CV^_RWEJiE7%z9wbZ44nINA3|=uyXP>nw(0GmD8+At~{HIV`rnhF|v=t2gV1Xl_tfLoaRYJ?K-V; zlSQ-5y#75@^Ky)uW^zN<%Z&DAQiJ}tY3qC*vrZd6&46intFMT$Okx8FT|K>53+=akso? zjBS}E_JmVH|I^81t|WnBp@G~TM^F$bDglAM` z#g!OUKIufzhk+p{CWDZ?Ml7_0#=gw8vNVj#4e~+Ekbj^l&a2@9|8+T&m*tvvF!~=% z9cwO&bh0ys7aQ!{*P$Pc*iM6sZJ%YV$Tm)>s&aSxq<180g>W6Xcv>{$)_gY|=zi>` zIuQ(q%)MCy_$gI7FJN%~)s=gFrwT%okadhd977uK)>JmnZYzsuA_s z-mCaqQ($NfRhi8gAJr};MDPl`{X&Z~1X=e;fMPW`$sV-kWfXsy>R8$+)_+KWVg}Np zrQn%s6Lw5zLQkcw80=4PcJ@_Ed%hjTgY{8*Y+KyW^7GFX84-abd}*%FVfu(6Xl zUsLzEf2X*8FY@}B@A5g@?sl=)Z*O7P+DN&+l9umVbVylQ5%Ktv5`TP;n);l#H&f4e zvX-${KCF1#hkSe-%ZQq4Bqj8&T2R}x(B?N1pSIyiL+Coz3#ym1l+NrV!>i0EH~;tb zJT=5RK~BeE-w~X*eXYV|WJQITe7P|tl!{YPA0H}@gIaJO?Zn8Bcrlv8@judg;M2y*F;z9?h ztTWvYpbfmDMbnS@FAu0#p!C(l!5)f~T3EfILsA`+I|9n;*ShS25X zozh6u#$2Vu=h+MflQD@>1$Xi-yH@yj%2Jkn zbHxOm8lpBoYQ>qa_!8LE46#??zmTK?I&PA$U$bMN5=Im*86 zjk?iIjX%kdk8Gt2#$dw&1q6<-wVqvi{;01by;2Lj7$mF^-@>BuYm_Z`q!p`yX}QY9 zt(3_~W_SEOzYM5;Rif4<-i}`hN>M^qC%@i}5PTUmw64?E)MPXP!SMHJ+Jr4VO-LID zfmM(jJP&;NBIbKmgf(@U*XvVTWkkp~Upx5s`2w4pcQoot_qgG#P$8eQS1TLKJ)|_t z4dmHuWzQw9a>opu5PY3w*z-U1eGeu`PPQ2>ov6Ri#8&ca*O%vqfi)-sx$KZszaoB6 z_@@3BR>R83{@=117ABVeHIa<{e~JhEZ=TZslhr`OIC9}4=m(O9_|(niN*ZK7EH4KT zT^u&dcPt1cfWGX!V9^(KESe9}NN>;>ke(E&bsRcyK1{-f!O&xw2Tp|6>qYII$`YCf zKT45v^)F>)!dBai7DfMJJR~U~3gx6=%_oX@w_#2S^bL`VTB(V=E8Oxh?>q^Iz#<6T zz8e;=`MvrYy%&fWngweSZb%J4(`PYpzNyqnRSgKua>t1I)(o?8}TxCHYT?F~S08P#Mx1CTc#W8PxCg z5gIuu?US7=S&*a&11NT<_F+eRVaZIq(X*%&DZ-y9H7UZOlk#Q4+XI7DVRs3*FnnOF zPr;NCk)uxY=i#T3$#Z86u^g4#G-mG!nMuQv?H7rwrpr;187*6W2TiQeVbG~Tg%zH}X zXoVwK8VR8xQUiJe6iZm+TrUL}$zp{hMQ~#%@-U2Hq-739;t0V=kv>2|CVv)=dyu*GzuL4_C zShT(jjn8|@=)Z*BZz}gUHKHl{%r7kBAnIeA`!mqArySv$!}}4A1Hrj)@2nO4 zx>q|rzd#=K+@hBDZJ4q;yWQuiju}oW-j-~-@rB!@Lib6?E`ZRx(U7Vs<#vTU1p4Lk zFb_T9c~nOR>^EQ@HkyTRn&C&Ms;Ur%pLJd7ns54P{AXs@XT`0qOy2dHWt4@N%Ui2{ zKUuw73j|Jkc{HWz?@=Lc?EHSX!nN^g^tR zE*zWk4$roqU76fa;%sXAa#%64JM%dO!FLMtSn0{km*bb}U2lGk@9o1J;EEy_1jZ7f z$^gv(qO(**j8i8HEs^)V&(std_~R63VU5&vXhclv6KcW_1xyEqY=ssD#Q3U|GNFyX z$OkV6aesfYHC8lQEA!@DD{=Z!hCIB5!~Ax7-{tY*ZquXl8)8;N?#`b64$1D;(5Fvf z2$S$XPv!o{L|&0i=`8~Lo28%9{FMt*wM}HcOO*>?*!t%dUkUgzm>ZGvR=lNG9yIwDgGg$5)BC1Ael+jh3?x@>c-G!OAeTT&Esken5Vg6zY}B5IY&Ilh_yVL{phIte==;yP z<7flt)*G@$3O0^K{fw(UGcS$ndNyxL?Ey#Dlk4O|DThQL*%D~Jm&Vxb0DZLuRP9Yz zD7_JoKVM}#0ogC?mJld?%ChgswG4P$;$pS3n)MJ{uL1~>nrXJs z;i{S^&^HAI8?Rkwr%-zzVGG?+6jF^bzp2x8YSY67h_9ejbzhbqKV5unhZ5+51hbI; zLG^0;)1~hrQfVT(BJG+GV-cZp{d;nwo-gUr2|%-T;UH2NF5KrU@i~DK4d4cy!qcAP z8zD`OmGOzGc4>H+5{`%sYP-&3YKK%e4s$v17m|;-VcO&9#=Jzf?v2o(TjtY%^1~YY zZ9JagBW{cj`momvS7869;?ct$)Kz%1@BuJF^uuliOU)2pP6& znEEj~lwg-7y!()R!wT`s7qG>XEFLaVik8x@4(HE%3b7+TbElhKz&fLlZcaJHl5~tE z^#E1m5j@X3d3t>hcchSQSa_HJ-2x(yb*Sy3%2BBqTm*Ix2;HLE6~*mS`sQC6Cax@F zU;p{}Jgr^jw{|=_rcLE?Gq$6ugJ%DcUxwdI;=TFVU3ULcsT=pIb_Js2SNf^l1k3NG z9c6GX7bHl98Z5%`rNX(rl6u+5%3c*MOJf|hVze;5+Q`CwB}4aZ?Nym)`xTs=3`-Xb zD{I))d5T~g6E=UG^gMg~qKoQjEgKlb<#Ly0onNilt%)+TOt8kZDI86RGHWg4k+X?G zNSi84zbhT^G!B9u6UXIGZiN&mVIzC)-0hc{fMT-*9 zrmNiN72AX=U2#J-HI-{L38USL?c7apzO^mI&;VD_-%- zVwl#q70CAk50-DY29J+=n~>enI%C;-8nULXj5ND4+<(d%!CZJ~iXaB+lIk0vV_&*K ztyTQ^F8X(Ww{J8JV+J4ixcz%!;?2+T;$rI2h{?;r)t70+d8;Mr3A?hKrD7<2Yhmj1 zUc73jlkY{_h6jkB5fqTC)at4K+!qEG@$0Z{AS_rSA^+dbzKBzA2rZAdM0#v4ukg1q zi#81plC}l-PVS(x*m_)ra>-MFxE(EkD^xyAqx|E0E|x;79b*$riq!^^+Cx;iPn0u% z`frv9=C0?Um)r-v;@ME77UxY>vi1nF6QL@!VIJzf1q0a!+Dil|$iVP~EN*TDh2JJ8 zuSJ^|hwfMB>y(mPwCRi?QXm**(ct{b@^D5WR&r3RHvR2hqrK^`2^Gn!)?(U#m%g&; z_X60@u>$Iw@a#@eYzeR68{l3DfALp^=(mcUDd4U)2Ll4Dc27`pnzIE=>o|F>S*6f1 zXk3j^Ryf+s2HCdYS%b@!p5ytk^G+L_}MB}asFZv+#bWs7h)s68Z1){f`OSI(g zlP73&1$AkOwBMYs&f~re|A&YB%$q++p`b)}L6}wolhd8;4IA_0Z#ft$hh;Wf-oTV= z*W{wT__K;F4H*$(1(BN9u2lp1a(Nnl&@r*LhF0DD5JU?-zQuI*nm*0 z!BaS+NkatnmnLG#xCWVOuibITDC}R3ZL;IU`-=WyEXyZmN9zzzW7~ zH;XqY3I^pU%%ERE3hjWQbf@AJq2q%Podr`DyCd~JPqzVD3ISug3NMi(vGh{zFHC zj58!iT7U33$%S~+jycw8_rlJ#tmrV}nRza>vU|#ntFZ59{wPo?ZB=sq5R)XVHDdzz z&H;Ki%$aVW+?Bfdzbj?TbL!itywe#7)V4AOS~$OrRc}1-ZKm4l%oY> zyh}-PnkZx0iYmRDH$FMde?ue0fYENG@xD5u&f3__w*KmD*)un zareEcpyoTxENE756fDx&UI(H)Fl;!8IS_caQ)v?fX-ss|y0_CJf!SR!UmuV^f^pcF zP;`(8fJWIkb3g(Mk$Z-Nb1t$FBHe7b8k}r{4Do3=n}ulcV+lid>3NXUw?NtX zO-yY0`;T_L}WcuQvo6}VBP@&AGsk3-hpVfRYw{+R4$6k|R%96vn zWxkZ6ul7_h74k=NbH zwSBzPdu!Tn`<5+-L5(mGOMu-d;q+Z&-H806xzuh8!?AiJP@;&gw$4`_$oPx?n4493 z9idinJKL_ zs{69wWh}-`l)IrI3=l?^A`|S@8=zc6^`ehsYu^fb$^>3O>^BzK!|xxJYfO@)yfjC^ z7v_!qM*UQXbv5e!?X_|(`dL{941?LHrZ`HCF4??k4hD5=CAB2eTcs5w@d^nCn70(l zmkB4SbjSmp2^KgO(D;uPr;X->dufON?OVYx()#lx+KeUmx40ac;c`66@p-1^DCnU7Vg}3Dn}%ogAAfEzdU$=q7ujpT2sA)q`&l#!w!#WcA4F_ z97fG>&F*y;PCthSlzL_gWDiOvBh7RSx{<%~I@!*8*!z^xTEuJ>OwggNtrGUVD;0GB z{Ylom14HcEa zAh_MRKv#B2P1rbjOQb>0aNO--?g~-}N`P?y6H4lq@$?E*kow=h7L@%ONHfGziYRS1 zv98eC#Vl;uIwso`S5~z0rDGRSJxvQe$xLwRr6zwx>#eu(;$OG7YQK>je;%-6L}jU{ znhG%ubv*je1~KKj7TDisOd-qqr-GRt&-*PV1$4jg#AO&wT#6+bh$${n*-;3v?0C~v z|68yFvu7#gO-v2Ads_N%Gg^|JU2`um&t|BjZWcpfRjMiiCKNEgH(;U4vnfIyr zfmmuTyDILz%{~Shbg5>r#cP3<%_dPlsQwS+{3i?wA6g9awlFZbyFoGdv46|@cKN>D z5B)6j-$%D=V^Jzg^2doojafNIQexN7mK~RBVJ8P<331tn2gMn*8xu=LX%kLwWi=xg zVYv;~vap5?#XAchd0ExdV$(!LMv5KpIL`S0_Vmj#p6JTVg(h7lto!B_gtHq8y+qE* z^XrfZs|W6m)MdTgy1${IBOg!o zgFEpOGXBhPuL@%96>Mw$+7W`cn?Xn-noRU=~s2Xl?b7B<8 zU~{kaU3SAm%UJ^L9L!2ob-v?k{Ei0^Q273rHs5zApoja<|2tNoxzJ%D;T1!~i31M0 z;S}|1bPmI22*Vm~jc-$-XoaAR*>AJ-zL2j2u#&ciD>&c8@V)YPUL)c%hi zNT6^cK-KG5z42uu3M}{%UP#HrD}@-?E2kh=Hr^la+MdhxJS!v6IjBUf zMAj$C-WwL4exta;oUpcyf&`mrQaNQ|C8$0DO(ihJ$oB8wbJsEsqMVfdX99b{3klJ4 zP6SYr=)twP3(mTc41Iz*hQx6KgSd)I=ctqKyP*kcO%0%9@KJYzkxwi34~0o6MiU`$ zW>twE8etR(r6NhKvHt!WdUJeQ^HlzZ%5gGNqf#B0aZk-fVY3`Joif&9J77o{N3C6< zWBRd{_@lPmz~FUu%2r%JIE)u%4L=X}nYF1J;)De)c~Rn$7&>sB5)FxHyqvg$y_+7* zg-J-vuk=R=3M7Iq98%{13@Q#q6$m@|`v$bO8ld8>;<(-yg`P9+CpAz+Exu6=|1fHz zXX02ml*DDSvJ^5$XQ}12XXShoaT4Q#R*;nN$s?r!oLDQol?J+h|yljWrdP z!B*AIUsG+rFIgv#vw!MJE%(i|?I#rsRO$bBp*N4ZYrN18x|vVeEw>~!yMY+t#}-U_ z*GeuVxmcxecW2(dM|;B?Y12#z+CWm5O_d6u8-@>Gy^;!H*XdEb;$=VvzV4db69T}! z!k|zd4MF!T$3tH1_uR1EWGF0h{e3a%y*tvS%T79Od|0Pyli0TWTO)+{fDfk(TuYWq z?=Xhl8=CkisNmj_$sk_S=UR|VvigcMI}=1{D_U`)%QtES)u4%S7&v^{U7i06RG|MQ z{Ye)6Z#=rjS|A$Q0kw#uJw9eoL;Xj(QxMibq)v=`AHlipvx;7p@wzSNagEvhd4_9t$3J*Uc^_h-_x-1<%JM43BSEfrB zS)~M+I1-Z_vjP-`46uCyXBiwZNlpFcQ_o5J48)dlv?=@)Dz-y__<-Ax$o=8{+35GH zR6-D`51csSMUq#rLOXm(Y1O!L`Nq>< z*wlWKvi(rp(V~%Lq=TzA0d+=g+0dn&bN_)q29hi6kcz5R%EW)b4WTyECV?BWo}6m& zj6W=71ICjfWH?IGFIv7ldK+OwC^mPUX!IuugY>{H7c?_Jw`0G(!r5IGhtlUV=}TQm zV9!BRZz=ja(lst1tW~q5Necn!_UqW^K(Juq2_&t9DR!tF52v%Pntu*^ECU*vf5s{+ zRP{B}WoN!i{q3HDu%0LO%5$@&l>smtwYIKE`WKrLxSJ$X*(Lx(`5&*SIa zL!z%R0&UP=l?`|QJa=`V$lr*DWB3CJivn~#I&=E6ZAxwHPOeAi6HT4!zj}{;W3)yG zPT9idTWa?t_6}dM$mDr$R}Kmr?x(x7Mm3leu4vguVn<(c-&Hexv@mwB&5G)c36ae^ zZC3u<1AcqT#H}akY^J zGzZ7<6Hmwy0%Es61;KFv(d1j^Kc;`b^{RMjd~X3g7c! z`+z^8&WXuc+QyhH2{b+gd93K|xNmmrbx6$Fm-6$bEEAm5ge?T??EE{v>WJGd7vOaD z`uz`-P}IKiztBu}M$Z41W-@cI|Nm3)>If1MC&;*PyU{Eo$u0ds=nPircAM$pG+WGR z+7b~Grsdu*K6uf{c>1u$^hl4eGvPB9t$uUiPEW85v-x6!ul+n!X7pxOv0xPfjS5OIgmnN4ECC@ySsD~w`4=n-ru*+zB$ zoAL&O006C4>u1!f#|R3YCsTKbU8jd`MPG82^eA|zCqu$V*O>Hac6z={s@CAeW*Xa<1SIDLhrfvKpjY9YN$uwk&i(~y9vVfj{QMJdttr91 zJPR>nUl*QSI2^7zXmb|>w$@A1%E0xg)3<6+>#PCt2*x+WhkVGLLJ0$H+$wB@IyC4Q z18wZS2lyG9J~52lWAIsqZ#CBj07hd*op|y}af}?0b%Ur_#^D&-KVn>bU^G*qf&#PY zoFXp~G(8k1t8p~at7sp9=*}mO3=@VEnP@={Z)6CC3q4Y@4-hA-5Sg~m(~YR(q?HSg zZ0|Iu6bS>qjX6sHE%Yb^n~cln#H1Mw>{R!TVN6Q1NE#i_J4hmgrIRp-KP;gV5{%WF z)XQd-DGk1#PAEhlWZ&~~*q}dkJCD{$&!-lWnR0^=YA-3yXP^-Eyr&gUcYh~^Pv5MBO2)6cL`2+C4nHH z059H&wl{=2{L_xOJ!QtMN+KZ<{~ov2=G*1-QnS+I%Fz7=sO!X$Rg@$38v#MN@2l0O z10}e&?Mow0_P+;xU1Yc@LtZUL7Y{GjCt%e+aAV0o8i|C?eddX|8jMWE{>}nX7~Clx6TDa(@b>6Tv+3k>lTzUWH{@|VN&e)>}$%Y4RtEvg~Q5SPORc& zUZCyz49)g-`e|?Ho*I>ym_G}!*OQe-Y|4SB&h*F>TE8;=s2iAbov< zB~H|qv268MKInE+TT=(eFgtP9P4!dN-{sI{E9;Gq&ugEoMS_~u5r_Ox!<1^Rdc-_uX??yoIg6*dt>VQK_>^l zwL1eU+KvAuu3g+=O7YO;JTe~fsZ1Y#ze%T&-lh|BE{$LXv^>tP=?0b{;{x_HD~u|cH(h-oj+f-ZpYv_w;f~h zts=fE?~`t%<_A!zb7Eg079-06{_wi&zhEdW;RM7Z!3qU1Tf6*{r(x&AMhjGm! zWoB0)KVTD$-OpWQv_8**u0;?pdf)2!eQegeO@FL@aJ*&XV~a5{vBW4T%6zJ1o= zfelGrPNrIF5CXCPnjP_2oDQdI9}iUB-V^))djKg5CpT4%E^f9aPNTR&J_CQ<^-zhw zhTglRvY?D9ppS5!k?{!9CbjbJ`=22QWD+1!F<1bhd6s5?BKH?2f#Lq0!M=E29_#7K zwTY9-3i~Mq{JUh3FUn1zw-fCs#aU;ZJ?3R3O_wt^Zs3-s8rWVT^Ov%46L!KLW?EWc z25CBf*Pb-g_Z4XZsLAp?nHDG{Jr)=$ucA|Ir6Xpq%)^D$t#@749Rf>F=l8+>KH>|Z z4p0>znz8t6U!s6}yMW8NCSd$01};B36b?`Ee9nX1(sj5fEkpM7&Hn}&%SrwL>VGfh zUot;X>@(wz*k`D=t_>xZ_4Ac_iN7_$wJANp2k;-3|Y z11}SY;z(>T6@!ELYICGW>qnW~Bk?IL!jzYnI;-$w)~=jIF`^53Cb?TIh%o2MaYw;0 zu}+H{S3{oJ8ZiNRCd!5V07CN)4nIB^nOJGK!g2nvYsOnqulSob6MA)QhC*pDNVREG znE{ttvxku!D=ug7h2}|Hy}FqaAvw^jnvV|yF5R|6t?00IaW-IFeAM(e9VC*L-E8h5k0nc*Vs){ zyRnRJebKC>Y0c8sO)Dd7`DbdYV-bXYC#QiNYuZx4Q?msj?22U+zPv%YkWa!EK4qL$ zcnNQ4lzg%J5ghf$!-px_-dw#RIDc`ZB&v~_~Os$nmJ$5oWduTx8jqM zjIDE9N*O{p5wS50?AE~vNSMdFpv3_pVt1JPqnOimiN}*lxxP@YK1acBtqd_p$PSmQ z9+qF|kXsY`^oXUT>8_Qrp(EuQ_?}kj1eAObl*;|sL&i?(9ZuX;)kA7X9nrS9o@iqv z=U~UW4YXBb=Au9(8c*j0yY`$@jXM{$1>{t!2NA6#SdjSf>(r&0j{T&QV*}H}+pSg8 zyIZPmXVx6WbChWhuKh|wMYcBO#YKAHnY}^C$dp-~_m+M4j?Ku0?2U6Q_`GE(Ggfkm zffE;Q!Z9sQ-MJr$GjC;K2$~t^p%NV-`BilA$lyDZ$mD9Sw9PEpxo#XO9(=-}ehE48 z`@DeP*>zJ-LH&e_yXn5H2EV07?{BM;x1klr!)A@PoHK;Uk->CE#77Y*wpgi0WA|Id zEGtF<_SBqpa#brsk^ew@O8RvX27BQTcrFs}>s*+!>i8}7|Cs9-UgiAMYny5n4`Ls8 zp(G_N6>a)DXmo)g0pufry96=$;*fF|8{hF$VZj(GmmO#yEw%aOq7!eCsQ7K3(Cksz z%Kq{>r9}37Cx1qkFTwcxJN_(iGE8^lLKl!B;!J0Qz_??vPIIz=AA4)YywzY>2&5XoW?Z7bwHn z*l4pID+Pitds**3Ar#E!obAS!p=IjnPy`RW#t{vB2PnSyTFtW%@1XmeCc}r*31fG^ zMTeH8h}K@xpx-S4i8wZ6@Xd!1+D<`qT$|wX<8`}|%MWG@d#re{-)o*}jadp6*R?X~ zbhs3)a%aKt0gWWubni2nOBo$B^TT~{B)YDRXq@(O&!1B?B?8F1Sy-&BgBX_@Q z=^${Bs9o7G2_)=D;fDe#(h{d=5Hh)!ucB|>vDyR&J*_UjIqOa_8tw4)nVWY$=+cc^hKCsWkMSB#yQ-yCFlZC8;945?tXVnz{b88)iI?P zTzPOwH&8hx+sD6(*>8r>lxJ)K3{hwXD-Z$q6fv`eEIx^V^(^dDL*($Ii?>E0FefqL zlOQ$HQeUHu?!~c6Z$h3@LO`K5udc56J}$Qd^u~lS^vZ1Wl5*^BJJKfU|Rq_G{J0W#HWoQ z4j7IJUcL#bKmdC2d+a{JvPL#;e<_7DOWzc^%{hM)p$L* zw}*26iy~znd@2b9lBYuJpBoVz zBzR-0xnFSsIf6JsBEN<5{|IJJf!KnjNLi48;e0>NiHo@^-R_u+Jv}YVM4nfyTtKXt ziF9biq~Ply47?ANr}v;dsM4ryi4_$pD7{m^E2PhtY8drG`OREH4vi3CE)A%1*thTN zhsX%*U@jH71p3~NEts+IW>P}IswniqfLCdpfzc4}qp?TgrAStt8O$sG}G5$o~W*LTC|&S zS-(>ygb}*O4D5ysvMw@3T6mOG5(Tv0kiIJa`fYbWmBJM2seIhlNY&tDKW^jPmMQ41 z4B>5`4So-j;+zH}GgSviDpT}lULy`4&k(bJ#m6p$Z`2*`jdS_f+4_=ixL{g5F*Yf- zjX&#R;Xj6n`ML$b&t>8Ti?L7T%aW5ta-5P^TE~l6kfB$r=ie|$140FRLNZ~c_e5+F zE;O7Ywy*Q{fl+c;E-}|J#G;vOmx{H9(6D;X`&YDsg-&b*Q0-n$75Ve1R4J2mw}5uT;h2!UKd^E&aV-xZ>BI&(0ra39E50Rh^r{yfxA!@@TAGQHZE zM0~66!Sk`~L?0JV?U_~WPQOuG3Ju8!pL%7WHJa3-ayJdr0-cr+IkjQwnv9NtP}s!W z@tPzKe?cFwfiXzmqN-8p#CWOsvZG*hTjySSG(~l3PzZ`qNd-7a#>Y)lycKZ=mJ>bw zT3Rl%OrrH4>wO8D>bpxl3`dHI7dZ+iyY9>)Sx3XovOAo)z26?~Nh>i43_pczjU~+G zdJ&iA<}471g%X&eN67$^L~gXQ^lugfXrKgSU8(M;Fz83NSBs|!A#rSWadUCKRqo68 zPOhX}WN`yI4W-&ifMCx7AA~F%?>;=GC7DPMt~_@@GWos)Bq``(k!yCX2&1^jV{F?R zKVlP#uJ<6~twsm1t2rc#4aic(It!b)zjCT2lfNoK#&s!WQ?)5WA=H;nVK@$lXI?j+ zZoJ}jVqOy$!KVsSBPOHv`+WKYziu3IuKgcKbIO$Ak}JlgP>$({u!~BW>fSf44~tCs zr~PXzfmS3O<%p$W!o;PlWMk}GFX>>7ryJEr0DbjA@Y zsI?YOMmSa>nQu+v4&DcrqLJ#TWdY_#?rAt+XiA~I7aL&#Bd%nnb$Noe(Ym~A65x|4 zDvBaxn_7TJ@@efWaYtZ=`){6fFRd$M&L^G1XDB4#~FcWJ*hl(=(^UDQ38%e9dPBU>8S86Jqb$X_!|N*+W??*=gJ{F z&VmQ~ZMoDMY}MF3az?V-I3=CM9_GM)&^*;@dtm!&^Aa*6htI+4FO2%EVNJGyrj!CqPqb9LiJ|2J#tubZ210zAs~f zhg&CJ-QC$MsHGLOldtlFoKKJM`~7UCq{PqH?fo-8S_xyRWMa)~o1P)^ ze8bJ7{&FJ&w#0S?F_!)UA<@|8+i_v0-HorZz0sNh`raeUUP<9@in1QbiezH-x#h~c z)dX)ix4uc{4cZcMa%zSX>dc^TgV55nDC%%$O_f3q4pU%~$GB@Y$mJ@7uslfBA zrMRngGgBCx7ay@)zG9}LJzX`*O0^1e#b-&qM!^3#<8CkE#M-sJl);nTZ+E3MCk&@S zf&)wjqx$9JZeYP}4$xdMH#*P=cq!8Gsl4)zagm(2)E{TzvCY~x_7q7G#T~A_a0DTOe-F5jBT)23 z6)ctZDc+Qemx&8EY3`tsJ=6*BdDyovz=~a6aiSFr(H@B|DIH)Xt`KM3BcmFt5pr;f+R5P@PG78v+~d)BIEmRU58{F1(SPof z>;$^}@7|q%mq+*v<>KyEQfl2iXc@o-gD+Dg$R?qyAOg3m>k#tvB& zDw4?qs}EbkjuXbG7I%UwLjQOcGUYSu;bmsa99l3L;8%@-9PTk%Mi#~K*Hp-M5xJe(9)xA{&5k~ql9S8ju9SZ)aBZ}CJR+N@&@_eByJ@VB+{HhVq~)Y zDBGDMoHZVjk;voLqX^&fNI&uc@yo=_V}Xftv|*a5lj^kyF4_tEn9ufjc(krU`Oqy6Bc3 zy2<7ocr>onAPv9FZWl0C0txjH;E<9X=Q|FsTvhrB4CPPWV9Q(^`($8QbDa#N#)*rv zKBPqTkV(}d1a!QUo+ku#sHkK}C=g6YyF6%A8ZA|a?lLBUNy)ZLXne1B-a_nfUUshx zanZh0PGZ*%MlSD+pI@26psB+!HLr8+yzE(a8pCRZ=7{*dgclnsOlBI~O5Y%uyd0la z+`chI3~HRxH3y_aY+HNAlyLdiAt#5+b9AdQTwBaS9y+3}Qa&kI(7YNf7w&OF!nWF; zKTLOq-=$!u$I1W>G1V3dNj?SoGryQo`Rq?yE!ywkbE8rOht{6ZH+)0VtUB3MyYv{= zHJQt-`1};JCQs;c7NZ)z-#;T)Xw@bf+}04cT&>D-+AR#`5wwE~`+$6!8T=GjA?39$ zb}ODCMOO?=kpeI2dEl!^`t%!qIw=oF#;|;2h4J>P&<}jdn(lY_;?j4t6xWk@sxWWC zm<%gCr8~#cw!`pk%;7Dt{c1wggl<~`ammR09^GH}#!~$Js7e2U3YnK%(%QE}bzCy# z2|A1xPz6ZbqOE#m=h}ybu6$0bcOJKBapIR#subGm1s&?5z5i1p+xTMadGH>$riJgk z=&U{w@NF-B7G2{i&wGGHlb1363t8o4V*GE(DhtQ|N3yCLtKEX^b5pZx94Od%IO!mP zVHr6n=??_`AJrqVXs}h!o&_ddS`u0{4+%TlEb2*_#W;)V@rSN;Q2WwSUSoR>?#GjH zDiL$w>cm`9&n6W8?vcvW?&K+`du7$+73P=)q-c>*GUy@aM*TUMF!aa%zp2_asP%|h zv86-<@Zzaq_QPr3^Mb!&ut{B$Sl06{OR2Sfwcrk6SH`5>aYK$k<6)@$^jD14btwYw zL;Y7YGY2E-Il1=oa^;ljddB3OLV0yEllCgkQSz3E`^boBL{repDTV3(?xDJsnb8QM zE8cO3R%H_Al`aJPMVgdg{XwUatFRO)pP17NV`edSAQQ8Qs}iKQ?z@Q4rAG&qc@==5 zS*@)~u2j6B5wc@TlRYaQO(?GzkYH=Yrr}>crB)w}#Ud4#$!F@_%MIW^xL#K_VSiAv7>ca%xV8_4`K+&xO;9~4M~-ybO#*>LET1JpC`YuHbKr3`3kasq>ygnD&K5g3m^=zWHqo$Q*hK5xFB zZ%2;Wi;E%bugeW=Z2v!31vz zi2mj-ph42`ktE|lUIHf|_MiT&r+1s=mu)@lvWxgQs*;wYxN@eP7ZT4*9lO}0!gAIkzhx1de4E|3 z7E`S5k1rY9Ocll#s8qV#Hu|0_hOSW33CwP#SvlMr_c}@@hAutq%dfM2zsVj~dh`mw zGxGjW=g2X9>{t!}9ba!Se7_GWLmf&JBIB8MO7fv)mb(+C2K@_R>J14AZoY6#(a)_u zMj?f4JsWh4-XtEP0O<6Wg61Lvt^@k1ejzfvI2>*7>5}D49hp>ebEcqy+{VPxy0iCD z@a}0)5DR-Gj6fmggPV%~oS@wsPzZ9@izliOM!?g2OJjgO1kLDa)M<(I6b3feJgMv(xh-BSOWJD42r(4lpNDw!1$h zEx4=0ld*KnZL<|FII1Yfv?jzDXke6(3aq5lql7!Pq;- zh!#cJx@Ft8ZQHhO+ox>X<|*4YPT96?yKeP+_r9BSzoa{vnOQs8zjyXG$CzU&^Jls7 z&(Ni|((7wc?LB)h`8*wF)!wKT#7l-pm8Q#$=GU7@kU6oz zQ5%>*RURhu5IY!33KI-M^+J63fkYwuiyB+a-vj-Y(A8R>2v)iTFtX*OJk~?f$mieY zG%c_4|KCobHaE?pLC$ zR+D}oks$sa6Yfc*xoEei)Y+(&-q!XeOE;f5BA{P-utZT64-8|Jba)=i^{E3=uyQhf z&!;xGyh<0)Dvh$?aeUjW)9KLn2CRY3BROa)aB!-Yw{Ovpf;Xz~>;4T=4CzPUQ-9gY ztRGOJ zBceRE=0Vi)d_7@BYt!)N_e=8?bH7LkVpKr&w8}TZ#t!|HHW6sGrjp&i%Vv)5En+Y- z9PHg{g?H=cQpRR;t~H)Bw$UG7z2@IGEO4{u@8%|a3!BZrx<(#Vj0jx**LVX~6ngD! zB~%!YkkklfWCdtNiVU}mv{m9w_aehWzljj^ZLfL@p)G6J?qh>g{>xMsG>j@lT*ka4O4f*;xSrJkInH$7qj>CjqZE$Pg z&X6q9gxs^CKzhJGoMpe`x@|J`-@Hz-&t4|=btZ}bRPCnmP7@MGsR`q>#3|Zd?agWb zPn+<`Am>#%rd(MlndwJIRIbELjS#)Z%L=d}iP^RRqcR8|rvyf-NOg@hcSJeH5Lo|1 z0m@#N-=ad_-=I)74_R!U?~;AGR>o@*dJR2L0^**LbZO*x8{j=};6Q)oo`DeLgDKZ9 zbmCUhX29}q)#L4`fBO5~$#?7JS@PF=e`VI{?D;f@1O0w(G4n{#tLa*<=Ug}=L zd*di-cb+!fU%J(K1xN*++)(8U9$ShOxOm=iZ9W~q+Lc4M(A3;EJP;v`*JaPWB5b@8 zw{ym@4v%vrc0~JD5DXuuwOPjoyf2()|Fdfgz~ApqM^1criTE3qg-Y709km%JsxRka zy3qk@kgyn>_9M$A_|2&kSgyu@O-zTk^sP!u8GOYbhxEpK9y{l=)4eUxt!=Qw+b)_}hp} zTL;tEfG#SodBmp$gBAPqD!T-Yp*4wPYKQX2jv6;R5v2` z3J@gog>x#W;^@fWIouv$*`m&6>+ZphM-qvw3TysvuHtIc9@7Uw>Rj;7z<@N+r?e+f zQHEX@P`5szKUdC<6IZk8qLnjWGiTmw-=m_hb`Sq{WGnRV%_&?BOcW0F~u3b;Q2A)YUrYR6>YM2Pju zYhY3-nR3l7*0WicfQF0W07c!(ITt~+Q>k*t2O<^M6{o#L+Yxh*M-ouZU)8xj-*c6- z{LV4RFb+0!2dpXeyw>(2-)_6*<08x4X2y$tY8eJ(_Ln5?D@WpA>mAql5YogiW>Z;F!}(R#@@UntbLH%`ObW!|zErlH@!c4qVU zrv}@MUD7(Kp+&p2SyF2wCjjfF6SW!jH~aL@QCApaX_@^1ot-%XlV&&G|TcxmC=g5&?LvM_n>OTkb{Q8t@60g503mr&lY# z`{AS*=)l9~WJOH`Zbs9v3S=OAjGcH+2{#Fz!s#cM4D04g*No4ml{u^3Lfy}e>lDQi zu9Yy_4?32{MN)_5>)h3hHN&49TASmY>czmU{{m&4jEw(1lyR{Bk9?P`O#fLy{NKqb z_4pg`6Gt#%_M&>b9FXfofH35=^C%(piO+OSE?R=~!*YCwtuGm<>SNZw9_euvtE`sPnA zgDTq3J&9^cPSF-=>qH?j43}sAjOrBy#-V~!?MtneohVMII=vn(RC7z0%c0F#U-nP& z$a@(`o$h>MT9r<`Da0%o$jYq;=|r&rI>n>(W=Kj%!~JkPB_DZ{`fDbK51xVqdIG)!7B$S>UJp=Y30)))sg?Q>hND#3wzGOfj+7v zzJNAB-7(Lk{5c$+k%D|PnV&*G7l(#$%{eOa_B&8|Xhcw94v$ut z!k~vl*W`!C6}k>mbnflVP0?Yg?A-ngfx}<7s_SaH_VR zrzgVShr#oeWs;|Sf2b~fe!b~e9$(Dj)P@%HJk@i$$mJL}g{ z(sF)&p~`Qq?%KOXZ=cHFtR?<6*ify`yIp2}6}M%#R`YY=+^DG|dsY`U%@CiU!~(c{ zV#{EK#+l&{`GEI~!MsVrIvt*CWW(4TZQ49)Ah&*i+p%eox>%L^$`P8)`i^xMd z2*nKo3hcpQUQ`1d4?%j!j{Nxsz*+X*o*KZzlM0B*l&Z@8uOUz$@5V((8k6gwp6qGc zShmj#Zzl}hUzjg>4BzH8ym_(qWecm>fkxk| z=Q96L{(VI>1HcS8EHhXJOXaaDWdCJpb;)jJ$Bq$h(8D+O94wX5``Kn=ruSUsZmLn9 zKa8z&#cMJDz*XV0@wp?*2BB83n`x^?w-%gwJ3j^vskYV~mO+h75ccdfNCpJn29Ugd zSp>t)pOBm(Vgc*^CwYTk!Y&N*^kxWcSCoTR=zBb?--nF=ySX|kF58;CXPVvv%Tt7L zTrdaSxWLr`DjXIAvMr3-kNC~6vwZJ(!LC%6k{plbF3Zms+9+%cUg{z{f&l(vW9d9i ziL?#Zn+gGLVNjFJfDIuLOtm*dDGr$_84_b01g1u3LIlQ?C{&@QUPrk-6%uzMIM!t+ z_()lHb@U{&?gfG94z>v<#RT{PP$T%)h-U(XkUTtKg;NM`4Rl(2PaH!_XZTe!0#l`l zQ>>+@;E~=aF9u8$Ya;SS>5UwISM>0i^1u_;Nn>w4>eZIIc0O+7YTDAlq@B}i|J<%T z#fMJE{;?SB4gtWt0wG~Vt$EHtix+Vl6TTd<3>lE6v1x1q$M?|nAa7jm>Q!u$Qw`K4 z`iW){`6{$qCZG=AhCVbx+Ht#i~0I9f>^Uy>mjE-XKJd;*7v*S#(k=BW3A$?%jYSr zhn$_00#w~mfyMeY=1psIqyYYIG5O+fAWp3aMcf&@Xo(?ud%Wl$XU;#stn$yF7yg<$ z9`(<76n>JI1|H6oJj9-=y!Ij~HuD+Q?tIB|`UDdm%g}VjiR(J9y=A?`GhUfbVhwOM zD-;={WUE(;7&5Mzb8CSHPOd&$Yy=xfKG~pi6C@V8bVsfQ67i(!SOF&Tl+T2sZ!|H9 zy<8Fs7eTE@)x(n#Er;v$vr9AqlxBs{4OR9W91$~w?P^>~5A%q+k zYtp!Xnbp3zBfw5E+>*cuB}=Iesl+~q3nUP)?*9Qi1!;%_K%os+UtSV?^@BT&0e|7G z{e&=OtwzO4Bht)3A!-y4hqnb+$N}Nhg&4U2B0R|52YuYWes8zJJq3?sq3uKWpoX^e zL8`u{8|Zt@;pV^V4C#etz);q@gBE}apLK}XW?_FpIQ+Sf@?H2+`EgyY>CD>&U;m?i zV`u*^sVB{#o`e&^e;?v5@T{K*s>>EKc7gRb&P0c=cWeeuQWvotxplIXSrLUK>Mk}_{b?26o zkM51{=2d5w?o@ehrVqX{1keN2UYh{rE>6G^@7bY692}iqdjLZ^szj?>rchTOQ|d-qlc9VkI~cWgd@dHr5s)u z5}ahKYB*Pkh;Zpz3G}ey^4hxRRQiWyJP;u;Y$coAifJNy$Il?3m0ag?mGwcz!xBN) zbdzK%NVqY;GsQ_df#<{!&pJ?)qwlJUSs>jWv4GSyV;#4?!3#Iq#H^t%?L0L%w`Cjf z?~sy+UqNvi%|>Wpcze~^yi?T)-8h%5KBfgk07<{}Ak&%dBvNuWIjsKzewdL5Up33srFAGQGoM0 z`T+bG^0^og2#q*>O@UXWefMk8yC^Z6s^1tgl~<}ufM%p47?aHXc!gp3d11o|oxDXPd6&qyma=*vuNMr4%&(|8i3v$^KN@$Fo78z#4 zKf$TFR}i&5j60NfBg0pxYsm4bDepO~d&LsSDTt&`1L_V)UG*%S65_!SX^+iQhX@iU z8p$1v#GZN*A7jb)k;KQC7c!s0^2DC8Txb4#r`??P>yHanbF-~IJ>7z@=cp~@&zHxs zq+VWv6YYi)<7sq?De@Vj9G&O643Ox=kQj$}K&~csniE-mr%2h}HMdM}mo`_HsZs!Z zmuC3&6=YI7iH5e+;SlUULXtRw}24_Z{rC2?go+gztjb z?Uo$zu28RXrm*xWtX(R*Jg6M&@o+{MN z+-P6B%*s1qSO4-{naUb0uCCp8%3q6NJt!7^Iv5(KY@)i`OG%@m&Sl%0)gg;T#ZNyRnw@VVy?7FX`uoK|bmHxwaO=IenpD&!b!v+Y~ zH(E3Zoj#1>aWmH${LrX^Rx*Zj>ZvJ2p$!`wfa|>y7eQ|pGupp~)+u>n?D}qHgzMM( z-0(00XAS`#Wj-e|xk3AWhk&O1%K7k2$^FZ140|1Ls82Vojjgf(o>Az(( zMi!3$38OJG{Fl)F|Erum5a>Day&{5uWwQVRu?QHHiu#u@G&Nj~YDz;}uDHElc!eSz z{arLV8CWR}EDM~YQ4Tf}Oo9NF&6$*Ist%h4@yjAeI!R>HcL3$|GxaW^?qH;hg- zP?!8rrLs8?j0N3VbI5W^Rw=Det2a@{$a2LrjimWGQo7udYN_px|w?sRU1KK3>B(Hp{?3Mo07K1!m3wIy&*F>kXqHohOz@|?Yd7% z1#a2iNWD0QDT3BCX>q9`?J{vzt)rT$_}LaacJ-&c<-7XH2KK}zt$D5N%It%3q$Io< zqdjI)D|-A2h;`6%J?m`<%gP$O#T*(E@e4uGx;s-8tg~g8*;3m?8+2S2oEfsh7Ry#< z(-#_)o#Dn_*uE~!jn?yEF}1*idF`wDmF6nSAO^Y_Q1y}Q!?X!Cd68xzGhrU4p^&}Y z_IQeirR15GkNU!wjX;QvF-v=BnNDknK+E`)mer?-)~e6|+4wn|1PpmjTd5O94XXZ4 zxk<6uN(|>JG#Zl)k=6|EitRJGxM?KvkD&$aZZAwx{cykPFb*b>C-tHyvpI#%seGQ# zaVYt-FaQM)BPOmw+FX?jCT=lZ(VJ8Jp6}vK$cu6>dd^@0d;EsvdKT|FEGJPuh01A>U-15V)c{30gnh!@SMu*~Jv| zS_8av{jvLSY1gH7A7tg?@NLD1oxD7Li((`$%=7{EAXtBwd1@o7e;=g997QHm0u z%^%h4V_-p%_clX+lLwEp^f*G#s%Ag(A%U;LDdM(>XD1>qe}^F9qX|7ck8$)~?7%7@q%AK=(FMmDC6ng7nM*%T`( zH9{uT@^^YdB@~HDJjZts@QHxf8+7^$Cl`E+xPY?T9(=5x}#Uk{h6wEl> z+MI^)>|!-jnttk3F7y5!gJd+Nn9%Al5E(77@eF-Y1G3U?h06@gMCMO{m~lWqeRu_{ zq^JnWt~54m8Hw*I`x*VmR=oPOpHEgjt8W)a=Wk9vt$N*e7Z-c9$=z#o?wxyAQ#SQ4 z*?ZT=p3Htd8#`buukK=-aRnoRZ(?wG2b2-5NIZ~uBA{=%jv`_xDHCAV80WW-S@h9q z%zVX)oi+_a03in&A>GG&?kGC0>H($HZ;pZAL-wR@&_0p_pSGM7WJp1><~% zqId#EQ1Peb36E1ySCVPZf2FzOEx?hk1PWh(AoE})SY&%x6Oo0JJv>jGLD{D4DSVE)@-KIe1*-F`$2f#y zrbL0y^l{FF8-4H1aQ@s%!ZPHoy%dQVaFeLrl5tpqEryN)|inKY^Wt3HmbAX?VtF?x+|!_Z=FN=(}{b*2nAfbKg5^kh40Yg)9SgPlBa(a zszD1K$(f=&z6^FCBz`_jqi6m8`p7GHYnYd^JK$z=M}%Gd;(6rVs+sOXvSStCs^UtH z$i!libBCT!70v;Q7#2UeHrLl^^>hC8)Na#K9l zi3^V2%vD8%@Hn+3efXV*^UqN|Q2-Fe3F27A2s=%12BjSwI|9TNo^(>|4nv9CS>f4L zYe*T46q!5Hauxq6;Gaytv6DG%dt>URru20csrz!`&D8so4GyV%&Qn-^|GZhc1Eu4$ z9S#OrdM$42ajFdz0|;h{w8;k1(Z1t%C;gA_DmtN*1{SV*sW@Ch0=L5RniB*7gd{IX z5s%{jO#QS-{vZv1QW6E)X+g<>CL&!&7_%m!>*wZ0M9BU1FS{- zo#bx3@V~j#`Ync!!YcXfzz~ky%KQ48y83Zb4r{l;vBsQyD7WA^V=AK9=V1>l898TAny!xIKCxgxWPIjxvYs5V{&gf6&_b4feTE(`l zq>ubj2UMhoqSQQ`f4n9eaiOgtgFx59R=GrNs(D+M@wBbus+<2UYh6Rrl_+QRoX4ks5E&?5i zz*ArP=evJ*lEF;@2r_&}Q)?bhaO_BZkb~cY#HBr6#9a@uuv5!R(2zz6?-V=~GRU!w zpX$H*Jdi^2D?@%S{X|~Bfx3Ok$dW6KVfl-x>yxeTFF?0CqWRTs5igl;e0QWKrS*x- zX`jupuDh#^R=nee4I6zO`|a2eYifGQ)MqSYs`8xlra}BnNM~MIvg)<{rz*7aMy6*F z>7_-z$#{%xzFlW(ZS~q-K%Q^d2v&8`|H>_lzlr+{1olQ&P&_;Y^eTTGObO`ajjWVi zYzgRPOzq5FEC`sH|Hs_&pV!m0|HJX;`yYGXyc2O_U|G*24Qrr}WUvhMJLeBD1f zGGk{8gU!dj`=2xU1$l0g30W!DE=q?CGW#+_(q}}W!$PKZB;B0e$YH1l0bh_!V z!_{Zdkk>5z<}%un=slBgEN2pA3d@j&43tQ45qnC8o@|&T&pn8z@;4-Bi(N z1mIMa+{jAvk?PnrT@>qp$k0ny496lXiBBw`jy!2FA!-{(Y@nR4L}kWYZKB1w)XDP* zi=jK(^v)~1bQw~n9aFR&pwi9_&apV>@XqtL`us|>K$8Xtss+J3*dzY>k z&$LC>HLO=jr9QSq-;lVLVagZjG%MU}RZ5JM3v_PWHX$hUMwWusycZThDUl22;YIR| zqJKuXZuWSn9(rmk=EQpTp?k;eKlWGTGOpmcs3lcetKK$=Eb*uaE6bR;F~_&jWm^f4 zQ*oWa)G05Khk0qD!pLs+37&+jaC4qfjP9XGFGkFjI?pM_V?~uU=SP%E-ab)$?<&oa z$w6MwB-NywESHwcnxLzK>Bj4%p&y5vJszRw`C4`8V}E_jyZCwBX)pME9O%XwKU_I! zT*3LcbZ>3@%)3V6Um9tw8~QB$F|S9Or`3mbhBkFr22JB*%kweomH9+sqXvB@TSH<-*$iWWuqPP=GpG)gJ9E))#5Qt zF_u<)ypXf&eqp%%YvVh&S};@`z(}A(nW#hCx4;a=huQ!_CXZknl!&9c=T3rcKF6WK zaR~m^l{nHE=LcbI6phFGD}|TO*PFA^QTK)uwJ-d}QRnw{AIg>(G=YePn62 zUekZL_GZ3hF$N0TDJee~EnSx-b6q$fE%)<($p`i-FwdHI&9)3kSPwE7Sr59jH5-!t z*#T=)>O`-TyV83lOF!s|fqeR4%NDk6mQ)sG>j3iIk##Nn0BaVrh{@sL!W^{PsCA%5 zvDtZHSJ*l`8HE4flqS_|L)U1SawW@^XYAEp5L>hl{a`KWq}xzGwL3|>7(1y8PV1I) zc=?Q~Ww8E&*LEqmXvpe2S*ZLbx1wcf@j43}4fem=@j3i-#~(Z`li8#gq+SDib>INO z-fb{8n%D)7;0?u>xAaH|yy!{S>ps1qxg0K03=AV-6YPU>%g0n4N=(FtAt2?VG&oCEcE3{6XJqMH`GWD1eE}4 zT??%5S00oaN1n%pKkL&@`nV}0hh2BFPWB;g3+s!G1#^>~clOV5FaCHwVxOV$W~Cts zazA0LjC&mxqiGfJ0eud+g@D_DhyXR_r)+d>+mDmE zWeJQTbY-NwH$mZ6VnZp=4M(O~b)gv-16;%;llTnAJTVY2P|QcrCT<366&p;Pts`Nf ziHOk|7OQy{xQFWqE-H||IHyq=mqaI{?2wG!pZKfu>Uf8i;2Iw1?f^suWQDYg3VMhB z;RHs-R4WtA(?bqph=c5yS|AH~8;ly*jCO?@RU&XOL-cvI_1V#TrLHbjYM_%Y$8Lo{ zn6%WS3WPa{gxy0c2+xwk0^YM2-GCWkVe0>eIf}LhsMF#b4#@6z$?cr2Rckj*L2#< ze(-2!ho5ebjks4YemTA@Bv3x|!sfv0;;{h!qE62raY`FJn%?qxi#iTgKKl%J`Y+_Y zirjd<(SdaRUGeIx(j>fNfd>zriri?(cGn-{+Uu1;8%z0X)qn(c@ z3U~iG&^~_h^dhnpjscoR!%%)cTcz^EcFfhQwl1(gvW=zZbrLBh1PUr)sfkOriOz}~ zQzW1o<82uARLCru{ek{y$vI29;=fRmCvV{x3PIgGO3~cpY{jVlABB?H@| z%CC4Y7Ug1GZDA@vhgJY4_A5f&N1V!Id@+c%dZYao`{Hrv`ldn_ww1H6t=(iC_QOFc<-u}eQ*QAS4ZPX- ztAz%=`E)YsSQ3L{zbq0#9@)3i;&dAQoD%^1Hq?gv?hWXF#&@_qa&Jo_zUhSNv9Z^ks}gHN(4(^ z(SCgx8?b~BG7XTCaq~?R__)zBwnAj3rdaq+-^@+t$J~kSMY17-X7AWKkBBBEfTn(Y zua!%Esn*MiUVZu-HaEVT>k{{`j%6p;G-(HKD8{heqRGffGoCASN|^`Ku1DvLaU%KL z3s~c%q~st}3e7$mM54478DDVCA8}${6?v3CyJW3v#ogW@WjfdUrqO9_g~3GM@j)1vvJNmhk43 zl(+uwmKJ6;2E)w<0$2+AHxUSEArR0)jQ-X);WfSZ!FPr)+;y!QhH?qR-ai6}m4F`F zfpS)dyJ(1eh&6Z`meL%m8^YiyyU9c5z5KH(8wr#?L@0#9 zG^sE-9iypW`Cy+2Tp7rbn}Ig(jerS2>HQ6;8Uf7UN8Yrrz9ScP9K!2>iM=-Y(pvtH zt44a(IZceez~7!Y@Uk8KP#LZl+>|znw6jD?dY+H8%j72y`43~Xs3}?K<5SQ_$01G+ zLhXMF?hODf_%a_Tzmv0IRu-^3b*j^pz1pv>jQ3eyj}bbC8G3s|JPz2fA7zosuS-xs z<1L7nB4>-#b!Nie3~K2NgT6KxC(Q<7`Uzyw2D6r?7<8lO%}Bj2on*5V5q+fP%Wau{21E}az{mE z;``XtUf9`hP-sYNht;O3Is`dY$Pp%oknOkIE^o`}1*il#2MR8o5HVXnlBv@ziwUKN zC<=xdu10Se$mu*PwaxXv?=qm|4G{E-@x?_Fsf4QZ5SfIFSo0UN6Aw#Ui+w}aQ|F&~ zq?}obIWy;RWG!P$Sj3gbg};k&CVEf3a?D}v_Mfq0${7)quSf#Mjjs{+NCF~xVpY7P z3vlkwPiuRE{+v5F0hbR4B#jQGB>o#VA^{v2(Q;cnXA+z=9Ndn~E;9NZg{^+Jt#K3l zA4jDH;OM=@zxJB`~9`{vwpQli3Lj36NY#5~^ylQjStmbo7h0N>e#V@u)m zPOx{Eh<|VW@h%Nt`?q(gA5uU$R>orIvGKxiV@IU3CvRi;;u^p9BMX$qX+eJF;o|yY z=c@!8JInOcrs@e0!f|Nn<-1vnTZGTptPy3tJ0VIuCGI?nzoUKQ89lIM^ZqO-!uID& zr&V$+afY;mgAU2464P~PC>4Z7EMEYCW~$kecl%5_MzG5ydlaXt$40!G3u`$Op$l;k z%!%*K+gi~9lE&CCkC?%xu<2vGFtOcO~Z z|7%g_iKyRLiN*0#`S~?X>^A$)b(Z$aRQ1R4vXApcPy5re7v7@pOwjtxAB7KIc+cAx zp82`o=jk^uyk<67oy?{oGR5h+9B1dNJdss}h5yQ5OpGl5Eq}56&%AxVy3v2;um5w| z^KXLF`hTP<4z?1(?&=r~i6UXR+bQN#z~tEZY&ZVpJsoztzxaS59k+E9>^MoW0uBX? zsN6V$3mg2CiwS3&^E6IzVWYq%YAGW6dvPh{%u!lZ;e!Q7Q3EY(Wz|^m(Jf+PlZ57T z@$fhp9qndf_}`H0#KSWe)a@%5(^^L_RRT%t*!(xI8e^RUW(@kD4QBIg_$c}6iP=NmI z-v@_LJ->h+38QG{W90>Q!3x8UOmK5l65wi_#6Za~N!hDH6zZPjKqVBT%TpcF*bf@d z9XsWzCnP?e1V!>m`5_RGD12wCOxE)xs5Y1$_Lqs6;8*g2xVb`}YivARldTV@!&Z|Tc8wRG;?#eOtyz)x_Zqqttc+L=w$8drh(ON8tuO zB=e8X+8(7VWYRqKim!_!=B1w3x1K@!2V4HCXa29fCpWL3(w6Q&wf3QUzuv56tQgE^ zTCuod8Lu^cgo*lfkXHoO}1V4;I! zmKwm>{$lc_$sH6p)Nn0_|0E`}oBnNjHE2bHg;pI!;g(GKMIxy$`0H8=V%fU}b3X>t zb@P@Fjf6hSxw{W)aUWE%rJJIQV}{3M?nQE%<$%})&I$I^%S@IRv)5dyYg?LdOIf5f z3!oV?4%`@yuo%NC#d7r@$WlF+ZFk0OZ`HWAT6V78KY3e-k3o5%>5Jhs`Lq5jgG>kc zSIdj=(gYRzyW0?leIR4(bl(D5WRpd!a*h&@Y1R*QTg7K=ySo#;*&Db@QLWKE(W)`;8Lcslpyq_La{RlU09xYa3L%^oSRzS1N)`ALm-C(wc z4_dbUokjy;^tJ#tW?h)O!iQefl4w+{MBugQvtag2m>3_+#A zY}ze;{~z4?mzjs(EKz!~*h%?n1<$=VSd6_Yu5l}-o3)HU#%SrO-OS7ceqL?Zn<_b# zeV*S3is$9F{anc8qF(WgdFvKP8<4fxL5>EiFtD)=U}!#OSK^v=C}Z8jdeBhgKh~UO z(V)=-?o|*s?4Q^+&}oB3s7VdW+XXr36*6#e*mqftI5#WtFaMj5<8r|!V~SgjTEpE8 z!eV5WkeR4_Z-NJrTRcl6sD{gvX^l$z_iMLu_ex*)q7U27PJNihELpGKQ?*7{_|rFE z47RDb8M^VU7xr2)&3HyGBT+)L0PL|{l`2g%rc1`dBzfW3@&d9tYlJf-_BYa)%s3G) zf)Yp2V^chV_JDkdo<4-P$K=@|e-G)jn|SdEJR3$#nRUHGn4cJft%4iNqELE($4UT^ z0wXC=q6!HS)QC1BM0Jw=NRdwhq&X9bp%Y;u4;Cx^sdr2qq3JjnCk=CNZt}Zy(PUw+ zVvluE3#X6E{Z42Du-)J;MihKxC8^tsq-sLaJU*8ni@C~mPN6A?1*lx3D+V6Uxig=< z!_sK^4)0RML+lCJ(o@7!k-qXmfZ~Zj*tVtaCQH2_a^C@A5v84XLttZf{QrklW zI*qWYMnUz(X4yB)iW6wU0^w-G;7m3B6heKhAxXm1Q8(Xi^)I-)ugfQo=~a+8y{U)1 znyaiHigoGD%)!vuYc&A_G4?oKc4DB=aM2r{dz7DzG1yVBgQdd|ZPHz>xlT zJ>$F*s-KZNr}XE|s`lj4y7X$jFn?LmqHC1iT3uFk6?J#n7)2H#qXf_5*)bZe9m=HS zi^uDmla*+ZC;YT=r<2M~ITSE>i(n}t(Q@Vvnu<9AP+K8n7n>#bx+A1o!mBj^Pou4f z^+BEUyGAD<>l_ci>@zWYhmFNS#Ra0mQ&JMXZUtL|&uuN@J(x9apI!_*M9uu#mEoXy zvo5x>o>hBK*=T`a7qB*bya0? z^-4<`W! z6LXaA_TED88Ft1KsyUm+@h_A%rHH2V4*O6yf9>v4Ii6lk(FJSO6)!jBp7Lt1x$Ad>Yx29a#c zw3oslp%MHXoO8CDqPrHD=ktU?LhwK-5w&&B#O(V=oXLH zD4D94H&m*usa9N4sk)_5*-NLi7LRW@U2(4DI{!y|2%;R}z~r3x?&@U3XG4*Fjxd$v zFNtOqS>tmrhAXvkG$>AjKG}>0>yfxL!5el33M$I*KIF=`@jl1RfE?m4IWsX|F^JGC zJ?auS$N=MWhJb#^rOo?tSmGmtfCsCmuj-xlKQ%t5x?YIhmEbV^fC&?xP8-h=N^j(^ zL?qN1XDDK$5iDNC9_H#C!TF4Ns+*5 z&IU;qrZYP?N?fz!ra>2dy)g86z@V;mTK@Y$PIf$(Y$z%j8-EmhLDURM4%>>$y}?z zhyFAA!-wcg59jA%^}ihZejSnWtK%w4ZwdT~e4B3_ao1>lfvS9=s#rN}&MM-8c}J=2nRl4j`yvtkl2FQtV*&^cRB=jS!gr-l*VI{3 z{`5831yFeMiA&%2Bra7WeL3qi1^L&OjqbN4?c2>Gni^&XrCtWRiW#avot(}f-_c1xFR)gUU) zl(Q6m7Uw=O3Kv&)OSN`cR}3`k^(3tAI@9Cpj{!4?KwFXz`{wv-2@8@wNjM{k?_?hG zK=ANqP1Zq$wl}+^>J8dGmv$tzHkYFv8AF*Qu)5Q}68&QFRW;fBNie(URxX}aA}xhA z)n(Sy=j)Lj|8X?1Hn#n{J__G=Q{d(T^;$j1yK>CQcmgmj2lE&50B&7`zBv!o1KyvN z_c~JbXTCZ#t3aaX`WT^Bl*Y~prx`AXS?k-c(ErKeZyUeZ6|gZ#mF~Sc#dGoxLs7QO zuSewoh{i#PmPSFTe!0;obk;FXvxS^5H3ew+*{*TUeh|vO^_{n@5>{vTjr0I z&U-EHEDhxi%DhKEKGMlm5pLG2sDp-j5mpNb54 zOliemDNpj1Solb*lNIET-?q6fg)f_J96}QpGVeeo2~$QD?kXf8$X+W1 z43Hn8CFX)9o;FMZFMY|$l67g<>L#8VjnYM$g$^bqH%#q0>U}3nG(?*B4ca7BB4i(% zcW`7E#U&tx5v^#A3Qsjar^(PtJMx3ZOuuNjl18nu=dw^DpJ@{%882>(7$hs`%J@=L zq(zANmhF=ek(Q_;5$P&aC&iRmN-njehm(_42jY`yw4%OGbmsA)&fSM9V%!Fn-~$c3b!fCo z;qsEIfJ?p{I`341?t0wRyp*`15Emj8)fBZKk(#!mm%Z<_WW=LqnE+5;Dmmi&{SDTg`di2h?p0;f}^{gGYEtju9-v((Haa)gr zC4*I*QKAa_X~!3}vM*OoK5Mm8JN~rqH#a3-s`#_Hd|g-O+B}1FHlbT$Vun?QeF0B} zQB#_MBdmi6g`0#NUPl4fxyh%SCVTl%VhX3#9KGAzUI1+=vRj5Uy30AdUp)zl%; z==k^xuB|BXPkseBc(<0Wi{p#8BX6$u-mlr}k4oPydb2OH@O@70dC+h{w>~l&*VEdI z)j|L98RgGvt?rgPHB-n2D%n9(I}sN8=yqscd6#KN3Z$py;f^sq|3|3~J>Z@OmlyG0{cM=?VUL(Ez$XW-c9V;HqP7{yVjR;d4e~mbg#ko2vxwG zlT-MMy zn2F106n0V&$b;%a4VmC-@Cx7^#-E96A$}#HB6EdqSZ+WxXQkqL6>D{(4k3EEzO!uq zAGwc$dLIiLe5{SIk>~$&d`PnSUyQwTj3{B0_Sv>=+qP}nw!3fJwsqUKZQHhO8@IP- zzS+%YXOo%ar0S$9^=DP(dEe*!9(i(OFFx|kS_MmZTy%W)mm61G^-JHzbG(9EyAn6& zvPS<9J^Ncb@B}|VD+3^9GYoMXYNmelJmtqZO|H3AQ>BSo#%hhBz%zt(iqls;I~%`7 z9cz;W+L4)d!B}X~^>@8KQKOhx1Pcn%V}GjA1EQg%q^zhQZ$j~R-7@(-x4}n{LNp#= zqX7@_Ue8X!A`teeJ;mQFZBms%^-|EBnvzX~uu$Z|x8v&UnVX84Mot~EuYsAG8(;q3 zuWiH*<}$Jpr@8L=(^`RNGfF&;LOm$NF^WKXC~paR`YTQ_FDGMpyQ}dS-u8NO_LrkC z6!p+WM^j8rrWKq+$M+Txjm0N&&NZPm$Cwk-*;(35`R_rcanKPk_C;8!cEa@5j5&DB ztOBqDT1etA%Cdyn9LLQU_(oiMB{|c)8$rt~q(Jz7GaB)}F|^&7*e2q> zP_geO4mgMI4lGT^H*$u>qpLqz*S&{Z{Lojtq0SF6~Y z`oHnW&sU!R=G3VHKqk&PTrK4y^5w&B9+sRlMBw=MF2)p(v1?cZpPB0&0Ruh(K)|eN6ngZfT>sh;O-b4pf74a zi2&gi|LH>@J`g)o_dkN9!;RaQpRgDE4l_nhWcrNA2~;B^ zTtdwo$I&ezGHeQcpW-`c0ZNVX%bIrv4QLa^o#Z))9yymNO|mUKsHIqSx61s+e>!;I zYy&f4HMp|~y$TPwvY6EMKPtbKi-^(TC^hlw}_iff|+@ z{g`Di4^XHHh@UmDG;tAcb(E`mFEFzWgG2r-Fc!)|$v-z6u2U3%LYsom>luC?h=?S@ zh2u1_j%GZrtzOz6F!Ypn%30LmaL-O$@Xr?plS$mhAhY_v1W7@`qOj^ejdHn&W~s>V z-!XWg3fn)|?gFhDH6PCGz#2*O@n&WrIX01iZS zRL3I-EAo9J;v4`?hi#XHLxdx)Gn|34nF}pl(kQv1@XTue6K)8Fu49-6Q&cN=^c zXxrAVem^-Go8DRK@Rk&UMD8qU_jwV8Xk+7iYjHW|EqP*!GqC5WjDG6fOMDe+g6-oR zWh0S;x+dFpFZ;wEg7Pc!Dz5TRrJ^rFbATA8zhbD$&Xu2NTsYSHo8|!` z2IPE&kITpY9U##~VP`q5RnG+?Iyz(6s(RJ&p8j^GB98s6i>k&uDRAon7QEe_1W_>D z!S97O-k4-)dy+zgNM?=~)g9fVxs*nZTw;!;znOa9#_V?ObrV)XE`n@_oc6f?&sw+E ztzEwbwG*mbZ-w%yoJ%9u$@zXcq}0h_V=K%ue2>#5_PGwS>xAiihsu(_BYOT9s~52H zCUXBngYJI9-gqweshb0>^K6!O=lMOVzT4$$dLY1DIaz7g&ps%gmr=5U9)Yv41cn=AF$6Mr0HHqW(Iz<>^7i5`D{zu zlI6GQSXi3-1yUgo)pZiVeG`HanFGKuoqNJG0nPX_kv4!VekMSpLndxJ*XLNriSzp< z@P?o~{!8sAq3K)${`{;}wHPb`QlFx%FSbFe-kDy@=E}CotEV$<|LHT)y_g+)$6cA%NZqQp9#zW?|>hSWTcK;9_yzmJDBXMIRsy~{yRXBfK*d+v?v z4nn6Ib?2YbjUOOc2l|;C+GV590#t?=>EWriFi`6 zAR?*FHL;|eIRn`vbtYc|Hm)Jt``4`4NRUra&h$J^lIhBVr_8Z~Tc<&jW9|kx$AL_x zi)n6T8ki4NsVe#}Ebs-NK8o(kYW$&XGQsYW6I zDra3jd=AB%n2}vR7rPX=io@+Oqj)TI)9j3DP5jJRvu0I3`yNSod<&$KVCq_bu=m(m zelS$P>~MbqKoQ-az1YYbD3Cc1oa-{4|w#YQ-#E7Iao@? zC^_v;5>nM`P=S&?Mw7k5<;I*xXDj)ELePOWDTpXbx1&{2Cy~yxgBiG zisF^~V8fl3HT*d}{zy|1V@f1A6@D$Ei;P}lVi51%aQzr-E*5tEf~caN(pp?5IyzNO zq)Cj8KVAVgqB}oa`>5JNHZ;}9mo}eO>f!tW;IGYsCK*(FIyf1;yq2{r-sE=drMV(8 zb_}>iAugE)KxEsnW@ptgS7q982glzpfQoyq3X^Fj+vu$%EAJKFs7M3;_V5+&BOPEZ zXfAcjIW2|)47N7ZF>^@=>Z9{XC1VI9d7C*c2|B+h^n!!{)8D*>%>MxDt2!%;pGH-@ zpS(LRdw(yQhTW1?`ZDVKJ@0upci`Z$%659M_9{jlYcA%%#e~eHjCk*bEsx@>Papnf zUDHu9e)={DM+Of6d~q50J9P{E__wWF(-R?-YC`u;m? z=6HNJ3Q^n9*9+(NDJ_ZOo3w5Ulno;b(NM0ITou5PYvyEhjeauehB6Xiq*NrL(M(wL zmXb2-dpSb)$$kE>0B)A=hTIohA69PL)a^>U^=a~L2C)Z>6x1j-LCf^kkUq()NQTh9 zpI4->c8$!Ydz}d;ZD~z2S<0yARHiE?yNPtLmM-X$x?|mdAC(DzpGL|sD_v>D3<#z? z;=f5pVkbqi{V70cf%G8?Q1vJaB?28NE4Gm*zk%B9;@oV(F$DT93SWoAoiTxe08px; z<|@+p07@0gH;$?Qz4(}(8$3_nYN%IytyqtlNoo>`Bgh6d9FO4r% zp90k2CafSdb&}KoNbyv$89C$BfCmPJNS`fL7<-5iX)8ACJ$^Cd{s#U6k)Fw2>Hc8I z3*c%AqI)2pq1Pg{L^4m^ApgEgeu+MI8r87&#z?#0-(GXnCrJ!^|A5*~)En1j>v5jQ zj%M`?!8C4@74lz;&U}hHDgiZiVUqL!gfuHrc_Pca84d-oC;%%0a4GN6vcR?isMZq` z%bT^+Tc(FiHmL57DHcJwsq(Ek%rTH0sX#N0LB*j;x|EB7BU~Pfc)fOXsbo+@S)}fw zNn40uP`sfvqWX)17KqIp)~f5zmT-FAod7~CLq|qsc?0z@Yk2!%*TUgr_ zhBsU7mrWE(Pp?C)$S*ENPjYKZ`PtGMhcAkBt2cao3H!FAmAU{W{AFoy{zCksQ$wH# zSfLwNBV23VYx&<_9I8Fq+}YY3m6ys0^hm3?MIeJXL=LA9Y1V@2v0FvMlXv4#-u>mMu zTajnAzn(#?My4Jizy@_028EZx-^pe05r+zdP!Wu7$pi0aR?^RBmTydS0Dzs6_x|SH zgR}`8Rg4X;1*wid%nX-iWY>%*ik}a))%TPWfMO>3(7!E1;0f}e`RQ&XUCn?dUAP!6 zaI~TIssIg_DT&}G2BiT!Og*0RIc!1ebRuzd{1(t&4L}<;zzvjtC(z!F8bE0^Ajd6C zUlefjRcMW_+|1>;5!>a1a)oyGPZinev}G@r+GI_)#mwfr%L;2aE(mUPL>0C)K=%k+ zuHQE=!*1(^(}xwx4FFp;b4r*Vi{uHaGp8D0Puwtp&93^_?Vsli!h%WVoHasMGnegt z5_fqRLIqR?M_pX`kzrpl%DC~_5p;}&?3t2`s1?){DXo&p47Mdv<3CqSbQ?SgQ37Y)|=#Ug-%P ziNDa4&V;6PMN=Wk?eR?LN+;Tq>$Wk;-*|;o0sIaKq_~F>&2B`v;JpiyNMNaewjN}p zgl9DYvvteO@KySJ@uK?1`g|n5NBj?=_`JVEkV<|YceGzFF>iyHZ;T~>xxOywTmF5~ z84PYvX)+|$&66wFYl1U>Ms;_+@GX4=FSIQaD>cbUc;>F=6S9p_Vl*G-sTcWq7cm0R zak-&qb0@7LBiw?KM*QxM0M@nh`;THsYt6)H<3VVFlFP|}FhnjDCEB<_Kn#A}I9J|f zGyGM?NrQ*WIv3ZFRPsh!rF7Ese%iU#InH2NWkB(KCwQ0)le8XX-KMO^3Nq5X9%kUk zb^@W#q`!&y13z>DUGLcuL*(Iwa{eNn04z3&B?9qQ%ej2`;D(nZ$@V~Sk~9QuqdU%r zL>7ZL>b3aQZ0VZ^iPk%f{-3pb(fVGEX*&1GCS{G>Pr(|UXu~_f=s-uc-bB2k6&O}G z9t&_aZ@42o_6^vu`@~Z%3POQtN-Vw0#%=!TJVDSQFtFW){RS=wF?qtO4*t46;#IH# ztr#Mtk}u-_Uqv7Z(Yi0ThG#eQeK2xZs5~T3S=T-36nIsV^j!f`bhmL_4cv5CjKJ|{ zde}RiVqCpn5jRD*h=NWx2)a_~$YcM9^3I&s!j#R3_T=7L?qK4lB zJ4xmN4eZ9JZgU2?c@T68Lt}o%?x^N?b1O-`vC@~z<+0Is=g*|C?@H68D*xX<&&N-y z$UepZ5wMf4Mxzut>GSO$S@A5Z+uL>XKbH2;ewh0PVk&kV^jBE1R5pJ$=XgQC=s<@W zP?f?oIiV*BcL)uuC!X0;Kidqy`1ooCRgb%$ltj(2LL`EF0WhP=J zPG9PMo!!=Jvn6j+0=R##f!$?p!dG)@)&SR@ySMrfRvpC;(tiqNB|4LI&L%xS3NDy{ z6ev4{i2sElHKlS@1C?2c>1V)sJNma4h+fT%*tc=)^DZasI%Ox#y#?yVpxJf4VYZUn z($2!xD-Ix$)&#|X8o0wcu7}h3`K7lMvocOPOC#k*wGJ|Z(U}^@DT(K-%+$|k)|d~F z9rg(`p$EZ?Iv^9mpiDTW@}I3YHg8GwyzxWlR-4mur7KrcX%Lrvn*GMH%Adu`XGoH@ zTk>4r$MXQC27hZJUfkoA$qYvXrvTo`3wke7AH))pM%%l#U?;9KPhYk&54YwkdT`di zENR!Kts?SQV_hNYSae+ErCA~*a=?{_bXYv?ht62(Tn#_J@5hHxp#~ByqyR4H=L&d< zMv@fe^&d=0wD(ckEzB}Z0dra+V6iuJ0GBgivh43`wPwJd@xP9-5uh^U#5Bf74id)( zL|o%7Bz3X9VVAT6%{Xw)=lp@KS!mOJH0Gl0pxRKa)yp;og|q_AIDEM9%W<3eleg@C z+~azxO?#uuBn%@NGDkkoFLYQa$VKOUPT5J{$w}AUNlYiYpZJdKCz&tNCi!P3 z<4tWsBPeDDH!=oa9vW--x#`moWKo{;EHfpUZD%rj{<1!r&FioUKQ0b>^U(C_YO`$O zOm8&i2ZosBCtx8f>oh1Sk&pgX`9BS;%Mqy3f52T`WzN0DMs zfvQ1iLy?-sp7?;ZwYj!I%xoG3(QsYW)9(aXWxIW%KI<+_!^X=T)U zTcK0K{YQ~(JiKh)h*Jfz@^(E_Y+S|CcUSoot%P1cmjQ=zVIway~x-uNR*`5K*)p? zs07G}WTtvZQVQP4P##KDB;lby$uN%Py=KCR$?{a84QJhL!BT1y2I2jIv4MsaSxlI* zDbzfP$jkbL`#34oXvkshj)4H7+ztJx&z(jEDsqAZW(%5P_Cu=K85g1!}$)tY|f}9A}lNq+OWen#Pok}EBCf@RosJ4TZ*<~b2_l3H-&nuKGgy;S^ zw&|;&g8}U=p;ol+@9{3%6RP4j%+MU_&5HI;x2=+^hj9qJULdH0*1|_(f|rHSMO>Kf z1q7D$pk}8|kfM~zPjbXZ{RzjWsM}GyV=+?esGubIQ~Z3MEZhx%vhL9>yPP46nuTQU z`mD8l?x{)Rj#WAl6qzM$+CYH`U+?*gl;*-+VO29yBBvhZju0k8 z1hiKt%&5vpZUw;;ru?6($fc2v zJliUwUj>fT!p!QQ$v7ovk!B0?ZbjglS>NQEsDathF3|0lyq>2et&Z z1y&CWMnzRPyzJ=;iY2I8!*>+N9R5~0d>KQLUxJ3s(%9M#bvBMU)Wc?5(D)m70adA> z{)ioq4deWMKW^pd%cI=>D&O-DY#^&!FDRI9m0*?JeAaw(d`lcZ4~|}6x8};|xl2EX z*1YoaA3xu#9@iS{r_uL?Q#6|bVx~btQq-HtnsNP_(D1`?OZGbKOKc*GUz8Bm$w91Iu7YrfBKPfRvMOgym&fIP=D+wTA)uL9@cX)Z}F8@HS7_RCHFyN>%dox zDinU++qfQXrIT?ak5JFphva^3125n<^8gbK0-&-yf2}OFs~PDZBy;JzoAnO^L+DY= zTO&w8C?MwF#Zjk*zCtj;s5X_y>G66c(kX@SHDT|>b~KaAF>B=($II~v|FsOm=;(j^2F*e!?%W(8px_YA2NEi_Ql>uxhuCMLGvN2-xr!Vk;7 z^;Iz+rVZFoy}>HS#ja$vCjgCt(j`%&jU2N^L`?%X4Ui-_LyrU3$_>CYz&0w4znW%! z*(CNtA2^!@?Sjzwhs?Z;8@m~|Or1>a$=Db6g#y{R6*ye(c({Q0lw|5Qfcn=QX^s5IRK+c&1Sjxm+PLE?Vc#fVKDq{>tcb zJ|7YRgLA7l1q0_Lz>tF*fFH`Pn~j5;@LQc1L6`a_-^5d^szdCjuMlJ8o5Y(S>}tZ! z!jJk<5~jw=`(Tp$1vIMs0DEws`GzXWTTM60B2V^yCoGq(SCkFo1%i zNCt*SJIZmmeT@TtF9CLE7Dmw4qGXOnOdk%NI2t{2Ebzp##1X?@OHXP*?j8p%?0Cly z0hbh8X=O$!01`!p98e(3!Te@HAO)`HS+1Z|*Dl*+$!Fo?JrK`b9KT1cSFSHR^B~#~ zYz1*|Q1X*vF`fYS5xpSEI(~W_(zA4EmSjXVbWQEH2ll$Y3nl#I>cb`0N1^T4djWWT=X{@j9lx)40XL}q`{}%z0k7QMrJ+YRM_=te z8TP`zAdk*jQ}{5X7@UkiY^Jg9RuB36u-wEfFYKiFPja zsaow(u}+jL%EdU@B|8^p-A?75bPJYuNrOH2a9Sp`IxQd}xlgq^FN7mhSqy3g^z^l> zZvFmcM_J)GWXg&)!Qyz;xv+kt`nGd(S~Nq|ym%B6_i@{I2e?6W^H(UdU z`^%_@OF(nrPYl`TSF<2&MQdf@+P&Ah*!xFiVP}aMW;1pi)iR&p(sv>RxHdixT#a_k zE5GA2tc}UZjN;JMU$l2S22(A{HK}LamRkL1QRpjCn!Ac+6R|wThjyKoWj46C}In zwYpAv#^__5ER&@Rb+C>b@~d<|G6U>`=D^$vz8mP9 zYGr?NKS=q?d?UO#fVSr?A5~XpMsFP8`kul?S#Na?*fbLjlDdO~xLdoY$~9@P?N)uE zhA}`dd8k_d!KpAP?XQcI9M;bd$i_=hXrb zXSvMd6A-d`X)-I-`M%9C#>n!y+b-wBq)jx(qSRq(a3|pS!OTQOK5PGDulTvI+f~U0 zHp0cKh(T$SqqxgQ+E36$Z>Y1FRavjTqi$BOI0k_v#QEAdN|*7vMg`uP5i~KaT*@FE z1VfrL+pF?3>k9?$I(K}MfQ?-I5-{0Ybo8Jc;)5~{q-}nO)&NxgU1U}!rLp8FC0yOL8(&tTgCQ4Z^b-Lw^C-wwX(1mAOv$I}eh zDrgfVw$47$g9sC2;5DhUF?qb8N0;{E)E+EG&cc=N*!K5Qo*6$AXNr#rKNFo6eue1- zm;nH&|3h<&t~*Yt^u6Tu9)aGG`<~Ye{p|>_p4(&WutQEMuvN4O6him$&s*L!^j{{< zU2^yy9zIW(hdI)M8i#}=aEA8xqkQszGUOuXNBeC%yaJHdsDALp!;~kFYVC>z(~w}p zgmPOl(&%oL z5uzMP1NJ+D=>>8XepAOef0TmRG<Qk6(N4L29N%SUK|c)+dp4Em(I?mAw!hZbm#@URJI~F{cWcq3CNFgpm4YD@6XwLW zRN7iV{2adQfD=TP)%N%ee*FNPAFs{lPjU8VIqC*DE{}*!=mw#QCpI^R6aT<9myFr| zyNYA`RcQYs426T~e+O}7Vf_DiF#qpiyn6hkU(&uqsJ&nY5j`wIo+u1A?E>4dVCm-L zIIna?QB?N7557Uv;vrX{TAFdUfNYJE8WJZnjx3KHLUj}h_LlrTn?suwFjk@BeJuP$ zwl>L#CPLVC9lmR9BW5d#6pqA z6_R(7lS)jU5SjkXFj*sY^^(bWaoX0A0u3B06WBZ&SZnI@Bp2`sC$46UIkDMjQ(Np1m36LZR70=?V)fksnU@O;+k z!jJ`>(m7yZyN(&vey5BkY@Y9cjv0j^%S(l{!e0)s6dBmbK^608G8HqC7#g`x$c-3@ zzoM)~!A#AX*<7ge%Qib`^QyM;qEFHDJeBPQH9-12G{2K){1mcXZ!qX$%DvJZt_OSV zLmN0-RF;iX71a9JS4hp8I|UDs^995(D3eiV0trvUm2?jDT_VjoG}-2Va>^=+HC3sR z^pMhgAqtBRU=!{VBQ2v6lq1jKbA2~la@jvViq8MNyk(Yr-jtfgc`Z)cFyh6=(*Mpn+a=0UUQe}je(RXuiBYxwq z8PSS(MXVyG%Qn0^W}|T_r}%}Y*}R-r`dZU!mg%VqOnZv`8|R@TJ&doj_a%Lu9*w@$ z^?4opsDZ>piLt+M9BS7h)X<>q`M{3#dUd$7N6byQgu~zz!ppW()L;$Cm5UKXe@2u00NnU?+Dj?7X0YJP@v% zz{Y>gprzK7BdcgxB7xbwJ91F ziqe>Eh12Lv6BVYa76*uJVFhd!Xr#WKS* z&K8$`Llr5ZGMT;+168mkNS-kUCpL)qdv6nVUt80f+tkTxCGS8ov_fmzb@nYRTn7XT zq$zkKTb}SuwmkqsLjB;dO7$3Szz8)D;v`|KyE)_!C7$X|#j~8h)M#*Zlm@gCkzhQBGKA@rttQ3})twN9N8n7u8`H4DcHnFz z0If#-WPq<(N6b~{!&c?eTo`vI*4q^Og7v6I%#DVv&`mVaUx+uYSIdrT0eBYokX6H; zI?M)^JghW6ypkXUiZ@~*__v7W2XerYLF>DfV1+0MZ#}@ED@rI3Br84hSP@7qcR84j z97Q0_rElT64d!pb9ux2ixwwOFdyo+?Fu90DJ8q#-W9TV7-|FWK$5m|_HmI>NqZLSWEePb(!j z7APSn=`lB1A4NA)#Y?ao0i?DucI3J98-=F3i$6A9&)n8tXGNc7qbc9RQRrr_-5S+V zZ;$_mKU#KeK2+81T09GVT-k54{~S(A2Wko^+|vQJoZw!AQ!FIG!HXwsCli2Pu5<#U+m^65De%4zsw(GdXRxzh?t~){tuupYUy7D^4)!?e_QwYe%ZyM_JTp}6`jI6GJ$7o8du*i zs_`2m-oOaL6bVU&;hyrsU54R)G=(V-0o9;Jmn6spDPpt%yMF9nI_Vocw z_GPAZ5b_QTHyYra9DyWIHF|8~`M=?Bt>LiU%QZ<{%1v@qizK4MZICz>N~y z0#m|(544>GoC%~1yx$E9HmnX;S%~KXJIS$+g$>%?G@V-fRlcc7W9KiwGO&C*gs-?F z%1C_9TiYor3VAk8(`A99-H+tu@%ratd-A<))uNMm-lQ#;`dzW;(|u#s&~v^pv@Qu1 z8@~?z;AYrFHPBDC--&AG@ebbXwa%gA_A@kQ>eV#b0UO2myb*aTq^%_*X*bR2`r12 zq9tpndRd*stNk?b7P_7Q!4t0?r%|eyVbaO7Rx>H)N(-VkK6TR@A(KMtYCMV0-8e&U)cK?6A>s*1w|Ro;nKO-QzR^zxY5yN zV=9{e&NT5%dnVG~25>{N%7ZhKN`viVsxZn^hYCG15B{N@1!sgiA8vJp~JN*eBp|?K6E8-ADhfX~ZMXd7Edk8n^6#!gr73bdSSBTIq>$-!{l$= zX3my7-!@vJq?-o32azI+0te+XZs;`V>$Cd;rj^p;^Gi8$8+QzlZwJ0~J}o6+kIgUpd%ab#!g;r9+d0>{kz4z7CkD{UT_H@LX>W zlu0YMs&A}tz$k*BepCWb{iU9e#7D9aAGyAK`d2Fa)&%4k!)aLMFZio0bG1B)YkaHZjAS%yP&cv+6*?h)_}B2Qt=#mVNJW$;q`#c~5b9!E z{|W?)i2C!jBB=%X^V(ZxY1)($q0B-_H42cygT$`Q?CflxlfUbnp&~txx)LZI3PV^+ zj#Mjqa>m~x9qEm%O_S}uNS%Aib!rztO zlH>!=aMvUa&)@+h#yJ+naz16b$8y5ut0{7v+Orqj%NB=LXDsDDl}nb0mzwQSf|spx zDAL`+on<8@!IVQW-fbi&)41(};?J=Qg6t(hXWY>iyrXl)8muVg(wu))5<0Igt@h-} z%2y3Jhmy06uAHKu8<&m`tf}2Bmh`05^HiD=Z|fjh@HD!RhI?d}QdCx%9MvaIxL}Ti zb)B)dqiT0IE37NnNtppyr%82=qn5}_AoFV?X|409>MR)~SU8zO;p*wEMqmk@zfpe8 z#TR23F=er-j;K?Gu4*wf>j*D7Y!_2)X>7pg(IWxXyMeU^#B8I1{0xa3L#nN6Mqoh9 z^#Rag$ogRvZR%we9VE<08r;m8v0CIgC)LGQaB6Q=RKHL%6i|I)bXg z2$mVFl`pAQ&9WmSGOXirY|#2mDngTko*|?^rcZO-1|`k?W;^%g&39b}-T1dBIJLr%2)6@V#tqZs~JS{O+6F*F^pPHGLjL*}1o9 z=*g#T>vQY>d{<-3UH&3z>}N{Ut-HeHvH-ytIU7x(t|AQsLDZePAHKGu{-9pGHonJc z1D-w(V;y==z)+O#Uwi*K7Qx}x*um)M6h((^EpRg{QYtu{CgKFn`iATYJfb(hRKL}0 zs~p}mY{}A{CHwAVrw=`T3Oje=>dmMbk;rKzZu}RYCCHU>5c!u5{L~PmSL%+rQ3m^P zYNJlv%68MR+BS7(!_2YO)*a946)38F+d8(OsihBet0kbIc{$lEuxiS6ihmt_nSYy2 z$@*wz&Bby;V`_Dg(~flWwE5P)bnduucvL&vQZWphIhtNwmM2%B_p)Z%qC=}oTWI^w z`mboiPE+so?$Oc7mY3}xe5CIUT`9p|JsVgtE$ll@N?Vj~(%5va-L~m*I-+dzM@>>x zldgiJIwy$$-f=QQpXN3Tp~Z1&1ndYs|EwH7K3edCvrJw>P~LM5u9FRDUaOEZP#n0L zI>u45yIoXJx@nIS>Xwk)nk*i}cqc_fgGvsoI|g{YAnt>W3PAU{Td*hW=pid~nBf3b z9I6^1Ucd!DPq*8L!TD$*tezK$$A)kM7PKQ6J8mKDVIAl_fz!LzbA~l#n{6y_vbsRcX@Qh2B{%{#DEDBp8Zpo8!jWh@o z%?4snNn&R6Q@E$5Z_}bNEEmLC*-XMG_1@mt15FPGwN36euG6meBE*?@DKnYy>A@LC z7!y8+G6P5oJ9hw@6%?2PB@-D<&TRx65=lF>z_cNy8iEJyhY++`gCoeLa2boGV3S0ymy6XLL}LZ#uO$kO*e%U#G2Tly%qwjDa`)$Bks1Z5V2tV3j!2C!qVP*4ZEZ>2BxByLUpttjm*ycdM7fhsco4Ds3v0->~^TN;s*A) zG+xSGVj!wG6$j4Q&_*pz1|u|FATOFj>rppk9O-7|rkwMdqAqhf@bRIUM4Hgwd$+&a#EaWeW7C`?HeKA<>Q&5QxB5))M`fL$oUvf==` zsjDS{hXwqA0$Thmw(T)<*Y0UAb7$)A^Fm;mQk%>Nk%#AGXJ`33_i)zbDV}Se=m4`! zt|I9K1NaP2oF?BEPBigX=)j4lOaq~@U#@Si z^OOA|VbY1MvecyiIV|eimN0j9dNii%)Pgmd5sVYmzs&SeL zd=gKe#}bH^y3(mLuumS55ty$rm&vE%o8Mm2kkRkeCyvr7Pn)8bR+`khMFIE@eZm31 z5Zwv>!`2O(q~qY#3$evEx&jEIt5c{f@4VX&mi zC(ps|0N+&Y_||UTJzX8a-&FN#(&-sLafM`#FUKhi*zfAmx~}^nK?F;+ng=UEJG{sM z>g)D8){;G5vMaOa)__a@fk!dMN^@c@B3NrCI8vsLzhgCp0lK`Bj0!ryy?1!PG@wTv zf+FebRm=%N95DpCFiPl@ia09>4WoJ0tp_2gnGFQEyb_G0VMu*~Sm*_?^kNg46l{01 zqdZ$k+$;ukv+UEwa!3Q)`j7UjtLPI!=OuKndg7mZ@jbm+lq}f*)loM`A!*Kty|NAR z-QR1-H!@1^XsjOj+&(r;9DxE5HgkiTd^Ob*pVj&uc}R79vdp%fSkw0C$D(*7SF?44 ziuN7|Q2(%1jh{S9)bR?uqOVU)Zbh^d$3N6ig!xvDgbf{WNLjoxly_<_dpa$}Dm|YU zv!$8j!W?ff>dbTxkj)=4){0xY_uv$=`(JgYyI@EV`IHa3kP=AdArIiQgWxorZFqXa zr;g(^Qyf|wu(dl2*%yd6n}dYz@}pdxiuXQKe|=`tENKX%M!<~9i)E*-*x>CBeB4dI zN5_VnwMOUUa)l?WL(t+vPF60J%U?Nsw$m0{#dYH@o_NJ~&DA>Gl@5lLJs|EG(+}Vqmn)B-%vl_DG{n`=ZQ@a{ zVW1IRf_KYjCyklBp1I>h?_TnLXy10k>^ND*;9whsgRbTXxPRMMA=fQzlQ3f;w$go6 znbe>YU56RGQR#m-=R8;4!l&(S$!+C^_c6K_TZ05|7^d(AnBKryg%(nWRxJ?Q zYqi4hw&D%?r~|ISOD3bWoP1n8#v(NNoI`UsM6CQ0)9^}Zx$Isic)%_h2Y`mmGA=hqs zzb}*<{D20F75{%m?Kqei{zsj~_`AbJFJ@`uV(RpJwlQ=u6)`ooH~9?^kRxDZVEB)F zPR=gB?SG?@=-L~%BKrKFSO7gIvoXRb7STfz0e>*vs}e!}76uc|bqY=GP}0&&zg_0! za#UPP6=~QMTSfwv7^_+Y zPR_X5-q~fsRwkb$w2wugxMF-``k+aQIzb1rJ*Cg(N(?yeOrKgCM@n;6QP2`um8A_! zD~@3(MJp~@=XizKR9(4cUC6xRKOU&AM#COLa9dt0a%TBP1r@ehEH7g}vhtzpuNvT^ zSb_JpVO;|ex0*mj3EvW{6@bKupj1#Gt}0sg2dZCD3(7FO_O)gk*%0v-rObpwt?pOJ z2%jkb_2Y|jB>tJViRRQoMxE_iY%Kto>Z4*lG-`AUcTe5MP(?w2s@7n(q*eshN;2Uc zQV?3&;uE!ANUl)xfLah~N`+W~MeLP=2?`2NORf@T-6N$e0U+9tWMxM*Wxnvoum!Eh z{L{Rggsg4NCZbudM_ACj)l-RhZ2?n_;ajYt#n=?uZh}0YL3U}_+&^cerG8ntQX;2a zp+VyJc1FzpHN1Wa&HkG)jc)gLF`7xrr){HtdATi)@7=k*^W}&^xBE8vJW*?}56kG@ zR0}+t7_6yq#$EWB%;$pdcNHqX(wa7@{c4{V{r!s1p4AL~W(Pl;7DbyQIu2fyhB6K< z!yaRgd?~nU+H@rREV)?kb{QT-QT~oWN&GM*S0muCUmP;AW?gi5EJ_al5EvoLD za&v1}!Ry_z)l>pS#@u#l_cqG^14{D(;w7)-J+RQx-&wLPOv{%aoJP%uoOK@%vkJm;&y4Ncj3I&n|nd#&1r{SDoxkuPhNoqAg3+vWAs zutVp6+_p=PhW$6Tivd-5Fr05tT%o!L%gFf=EV5VX?j%&q&^%`bHn(h2fb@*%mdt%^ z_E`_COpzX1phR!kLinnIBle{M0K4p&++&sOiIAn4&9;3L-7fo3FI@M(g5k;`*)%RL zuI{l?6d_+a{Mv^!{2Tn7`Af>l;Mu{j_NB4Q&7GrvmY?wG*)KY^WyG_VXB4+$51Z(7 zw`$PoQ`rCqe-o!i%jkA&6#q(j;s*r`t1ggIz&0_=^ff^BF4bRCZ2b&0~* z3Lr5>;~2PG8ElvTtN&*N&x7F)XOR`5Ot{iu4K_Yi*fu%#^ z_DFd%D>O321GoTD1V6S*>vxP5h=j_>Y@vE+IgWpTAGe^FUp#EIL~SGeOM+`@gX@kj zMRn)7f4N2_%bAcYbxQ1r@66LZom4yruB>}GhNxutW2tUvV2qnNF$rJM^~n&@XZgHD zv12X`CmidsYQ2jWL;X7%dmYv&emO7Y|I0w#6m`I8HAUdX(j;Z`zA<9&tDag z%#tL0PZ&+mJI{x2RAvO(IJy(a2fB$7Zt8`K!_W$+N#zPeJddd;0&iu*TgJqVy%gLs z5-14i2LT|d@hyIeE?VAMu&AqeKE?3jBYOQ7xoQ3HGrluY+*w_-WLnO~ zXpE3(Hlh)$a7ye?IxTMC**7fD(nrO&G&$)08RrCx+_`yq3JL}wS?;LE*lcM9UjOIU z*hxU?>uJgk`Bl_=NFNN{gnHGDN~8F$@8P8fumpU-q|P(I+961{q9@;#wCp_(ohYI3`yu#)$Zc{B~O}Lt?@50-o9} zXn&c2dId_SElsibKHe|wPad4(_np$$5ifW7s*4>+X_-|H#-H@Q>onNADHjB@s2EZd zENa;U#l?;&33`G!*P&!H0NBG;IrCfkywU8VtCm08TzBBR+DSKpJKg0`B3Gp3U7Vn} zGPqTza;T2wP@PL7-xZMRFT6qc%8)!2PmkLF7G%gMb*zem-$iwLx_ZqB%VH8DjAF(eRhaC{+P4CUerM)5$+|L2j z&xQn0ScbbnzzU*mlg^fX**BPmYfe>X382dLROFltHIBJtrCXvkPE_Hc6h^Ey(8`5y zr57h7uqym@d*CDP;7e-@AFWDSf2mkxdVUVJ;maWXavX>6>t55B>Xa{Jnr`ng;VB+;;`) zJvRE|l&Srn*Ve9^#N{L}#)9M9wTw1ndl^$H!MTqB8_n;t=TQ z%VbW@aR{bZasmcl9Bc409}N^n=NK#s;-Epwj39*#iz9#t<}M`Ae%&|c(${3}M=iP8 z_1;F=*L-(%d@lHci|uhmjUw$>Z!`bv1zzIrGclD8y8;ArKFKH<0}(CefF5c@2zS{; zR1j&Y!*I)3`zK;dw3}S8<&Zx@ChyN={;+cdZ|y9DI2bnecXUnm$t@7vCa2iW_vpvjFrFFinOVK@wPbIf)3Z*=HeNF1hrr51T z>1Hibm(n{RvB+2vsk4;rZEvPBYq)i?^5}`Wal3N#=b@;QAWHhoX!#COfBFuNjsHxoM{l?G_Oe(Lvv?g^Cu!~3P&AR) zDc;qx#Q;}}Ex`Q-^iVC;_jFtMkkX#RwA&TE1UKg&fMKD*$yH{z z&}_fWJsM%^Aw7i zZ5)@^?buRO#@f9r7Z%4r($pk;^(cZk&ZBsd%4A1>@PI1-drSWh5C!`W|E1PmDCN1d z!s|UA$9Z?9jMWI~%b`D9j`{!8%N|e@!z?D|=swk@kgwrBX^?cnM-oor5t^Q*)53M7 z8``)xFztDt#a4%{yd7I#{n~9OYT2XSt}d2_cbrB5)2OGUWw!Go#ho8mq%ICdT z;6f(Lzvc+-{V6Fjv_Dbx?);ri_SXKpb&r#Y@qZL7Obnd=mxATL4Pe=0vm<@`^#TV+}}^&7{Jq8;{cdAH2d57n}XVrml%dz;;-sZp}nt zhYrV{zx+fYlkB@c8vKv782<`{7zoi&3nH)yGX^C8*mRA)#5?3_;+rQUiAWpT`|Ot? z+4{jnz zOf<{Sk*+EF(q+oj_+`m#H&K`+319I?E6d4w;|!nZfHQe2>&&Z>F=52M(_%ly#0nt+ z+g`NxC*w3Fhlpn*t1{&Ks8l_XvW!A5b+)1Xi9KfU+v|DlFS;s2wbL9?EVq z1#NaLBSl#pq+HD13zn|d_>l8y&Iua$)1DfzKFi7=Sz4WV*bqsf2*pQO(6Xz*C4z!S zu>vk~h0i4*vPW1D^?Nnmm=@+q1id1Bok)@vrdf34Hu0g17K{QEv3dBinc?d2u#7KY zJ#n2(WHA*aQ$WO)^8FpaO1j*loogjA6ZrS4Vsm5}ZR2SpSFGkaaLBsT)MwZXAM+Qw z7{9#C&sUUriT~Th63k1<#hX)o`_uTFm+w>A&CK?CRJ zq533slKb+vPv-^ycAcDQwd34d*7pcUnAADM9rR}-W) zlBfQptOZxqvx~R=Qo-(FqjoV?d>#Lr!f7sPU`-hIHzVQ|!&bnbYxy5*mfdL|Uq5H7 z(#c7`p3K}|?8o%5Wp8Hp-)gg{vrDh&{a+Mf#2)fv=v%VU(wCug78j zQy3yvOfV|hVApdvVIHd3ut2P*2|^)=l?`tMj{)X;=3ifnn@m}9fP4x2c_@49 zukc|5{HJV4X0CokQ6>&?hUem+h0w>%)_R7J-3Wz!6Za%R#a8)-3FoI5lD(Rc$!v8U1^|T4AS>VktJ`0s^>huOkEjt zLO2EJh$%>jtK@mS|2{eX-o^1rX&Su_6ISpFMm4}R`|^h-JajL;n&?6JnRC_fI+s?D zUR}YDFGqtEdijAuAK1^NF){Ict$bS5sD6*{bf<*Mvpyx9c`cKgpuyk=KKzXz(Ye?j zmX~^R13EDkEVek%IDz}tuF@9YgB=I{#!LM@sTMD{%$Mt_jxLr-5fIEVQvnER7iS#> z`KA&Ohb8?px1x1z81zoZ7GwmAl{j?1&k5gbh3~D5Wtk#|RqCKH|88#f)xtuVu89&I z3m-NnE_6g(xWk(5y?c5jK5mkab$UP-VdV&c)Xg#tMG>VX5zGY(n67^%E`exW^Cd8i zaaK|G%}{Nn#UsnWF)YkuLCP9nqDzhhubAh!sKj+}IfDT(cfgAHlL*U2;Ybj;>4XDsUER{8A1?53RG);(F36AUvjJo%WUVvc$CXMRSd9a`jmikVnIx77({Xs z?47Z?AAvDr{!HQVqpgD017cB>&M8ic?tlTlZeXhQY^wjIN

I=FOsi9A!v7@8kC`}~VNEuOZVH1DCHbJ)2p{VKk zSp3>u<&@3{3gp*Z?Xm|2(a+yYq#c_3;6tpBEKz59ZZivSGgl4NZR)7mR#CApBVpZ! z!8(nCb?@X>MZJGF;J%vabG7?FJUH$0og`&SPwt?Fh=F|tk`DNXue+y5L-p^i1aD#T zQtx}p4h00bH_7T|K!`z(gbM}kU?KU*P}UZuyw5GWnb(b^G7(K>`nIM6JnV4;P8 z(0%?|X*Yr2z@(z-NV??mxAdfWyluH~GNq)CQx5%h$y>WZ)r_ z6yDEA=IG&%X{sk@KD=|}myb_JxyR=In}Y6e72cRG_R!}q0#Amty!F*tf=0jJcv}+y zuPOLKr;I`tR=>Rngv`^LvN7+d$OwKBCZRLf;BA28?nt{d9}w;eoE}b-I0)s}@Jk!^ zr!Wv7!r6w*uoYe8swiK0%mHHA1~6Xb;J@WZNXnC>C8^R=wqnO@#LmFY?Sbo?0=9Of zCu&Z=i_Z#AYWMVk4&Ne=TR3aah9wg4#_mqwd?5?LrqhsQSMas4QaxQp@kS95_w9+X z7XdyARW(7c^iG(D4bPr`Ufm6- zkI zL8+SWlfvaMH*Z;?u#*n8+~UQSHi@uStXsd*EcEt=9tCuFrN8-AWQRgcko8g|T3k(? zYc{mn*}7Jvg?=ruZtKKD1F65UEx&F5pQp5~ zB=ebG?>7t141TUGecF}Q@iXx!#K=RF&CvGUS*&sdEaW|1h#a4;(iksv4RYSx_+hfr zv324pTm#x~hF6rO{EkBju!mvWu)=xpo9QilCXaVBlkljl1(Dgu6uw;&PH2NIij?q|GbAeS$MfQQ`xdSOY z5a9l&*X6qtfdPnQzgAP%*B_$R$#aOF4?b__c(Bcy%EF6_DAa-_c8W!23vfAWhwB*; z`F3}?K=O)`gbyR>s?t!1IqBBfdIGCC!nXTCs>Pex$@k|#*xH8}*!K}IU*q7OdcmF5 zJbRnjQnJnVI+PbN=r2L^R6kuAx|4+qtIb|ny1bt=nlkDzpIOi=Tg@vQH^Wb-aS2C7=Ji0*EqbjRU<)b9CYOV7TGzh6oPm?{@{XTNWZqZ_L|0}E4*{X6(LvR1n^dX ze;uXi(@dQ?y_@+t8kQz~x8m0xN39)#o$FMmn*Rbj&pwTMzMt;iY~B}!KfSiTe&H|M zX*dj`Msq8_kr68|0Y$?_*mcjgvg>8z=U1O6TtrC=7kbJD%gF7DmC!Iq0A0czcG+R* zkh9w}C704Wgp$>3e<>i?1NcuZgM`UaiT{>JT0KhjjF^_sH5S>c%%g(Di#m>iuwSL8Q%)j+}2#%WI#m>1YYxAfWC zytU<@^aPKyA2xfh>;V>qy8lh|FflMO{g0xDneBi3m*)S>EA_{*MIPPThMFXl)aA6n zT9t<8kXc^r*O6D>EVn0tRR44P_RY-~S(BibC6NwN#(~KepO0kD#{9=440ja$FhOCi zqrgT=DDkENp*Ex$UPhE+o;b1ni;`$&3%VE66ememO_j;dZpel{GF3pM+Lg!LIRH+p zR(K^4+1q7AaehPbM}lU~O=0g3h7$8Ufi_KnmkWiG^JKf%<=H=YU&!-;BT^+mwK2rh z0JRr2^Fa`z+mWDhpqep7O$;|=;Ab*5X^7Op%*ZCv7bB|;Z-^EWRFpa5R7y5}qoZyS z^@@z-O?yc+GMWQ49cCPJXQcfn_t&_xpqUAZ8&wxgbF~s9aVk4F(%RM_AYFo_FOPBb zUPooTgh@|n9NoK5CL)^$Jw~cN8mC&SeTtY1RLCSkPY_@a5m=_ZM8cE~Yrw>5>Znuk zh?U(Vh_wX+BlUOMkddSc?rCfVc?zy(09@$Ta+}W9f14*tL@}wQ&O?PbW)Om&J)lK* z{D-tkR&>Sz;Ub!nq>WRXC24rnN0I*%>>3Rk(vB|C0T}Ue7O;Qj=AkkWfQ{>u>TOyx{sE_(L@HPNd8FI0U9JH2&DBFkb|@-gp?wPyZ;@J zMbefKbp+^8*k=mj5GWcwqHD0Js$d^!vJTM&(0bd*JwT(oOlYKjd2*VW!bp3gkK4O;!Q{9@_XMtxqH(Dz2Q29BbDZz0 z>3Y|m)>+jhHsRq0y$lzs1hlBryb3(R=i$1JF3y#tH&V`fp{d$msNxXE-kpv`#CnPt4|EDzwF*`}we2a=I5lkJ z%8gXrOv(sS6Sfu-o2@<}GkQhEuf3G98c63bI7pid89#92wttUVA>qlrb$Va`irAibHHTyF^+K)3nT?T z*YK|QbXejy0u)d9;j0rh6k($#=C-Qy6n^)fvxF=R%p*^*hYdF8k1rYxu>=y+fmuIO z5XtT89QviWl)gG0cSJ2Ho<&&GBt&AO5H?pAw_19_bP8kgBjlxF0W2BRljU!A|BsU$EkEVlyjQ zhpLGS*}z@e{gx7Y|4=m%MgOj-aioRf`xED1acOu}wQ5EXS!Bkpu8!Y9Z%Ugr(j4%N z*e+~@1rVNJNd1^a-%E=XV_;V$|M6N+F)xiWG+|7E5&%X`zjYL+dZ|r^EQc-DT`s0D?Zg;%aA`J~Qu>cTRb?BQ*eO{K;Pp-L=Y+)^iVW*+ z$lkaJd8JqXMt~-Y7L<`to=q zL7=djmnV}&_7h~@vvun6_mi&*WOH9w$C2m$Lh>yeF1Zd-{+3QvKDow&D;BHv&yCFC zWsHV9Y6G&}pC-0g99mq~KuD)YMzdb+am<3Dvbs=2Q~#hwoVtShf8hVL7O1ymYodmd z3{{`Q9uJqEazoGXU*dI=u`H=QpVD^%4mXG~`*1k(lVL80!TR(@V>nWEq2nJ~8?UD@ z1paP4!g|n;-U@Iy4e7-7W7vCtj7N5>vNhD9ps|s6jnG7wctkdien0t8mMKv&5>#}% z?dwEmrU=WiI)TKMrOWgTCYSal>)a8kca5Rak3^yy4@b>yg~I)|H4GoU^9J?ROH$a9 zZF5qYU`|oekiEVwbB=KerA*F(<`mkY!-wc7%xW9?QO)*uvG8pfVmOUExdDh3!2);%MZ68;@32m6d_3tNCO<~nPP4YyP63sWBaP0= z;?-()8Cz@?C!baMTsp!iF0C4hVi9q<{fLtda==n@n%%%xBP3{x{rcipPcJt=6bcY6w?==Rhl5@bTay^mT9=4J zi^q3Q`u5@b0YuEdNllu3zWA2y?b2{t=&nFUX_U*n-`%{$$<+vRAR;j;(kIir!16YG zS@*f$r#t>SP7-Gm(csiwk+*5cpM7QS#DLd?PureEwiozb%m>dAUsXmt}Gq+Hxgf@B*btBgJUe3G(puPYmSZ z&>_Sh+0=m%@{Lzts>uMUbpF+m5t!F8@&WM7ezx5Cj#PKSXA0engks{$pz`^wAT5WI2V37*=LL@#xl?G@y=j8@(*@^5*e@(#%` z;`wA(OZjox?YvLU9?S}#u;IO`{%14E+&ud?BnzCXxq(eu8qbuk-B_>~8PZDrLcV`I z>=2E!OSz|-;9dG)Ps96V5d#_Wj-pM7frm@TMs^$&S{%A{2un# zep^gt>t_ZlpdygFQi(AWO5|xHI#5n1ZijO8)*FKJu*;I0)g9v+4veX|$m6>d^NBk8 zB7M6iH$cOH_Qam3^Yoyw?Ft=`h$@gZB3-l=vwJ~eT(+|S2&Ps#d5`yhCFyS(bttys z!;$i_479P%53Ohb!zZ5B7hKe}`tMMjH!bx=uGJ=*naVASkXaaxgNAksah491ak%2_ zxdSnZ8A;N&+;kdUYO9Q!Q+Ug(j;0oXPwP9m_c=iO>RJXMt*Q5@94{=v@M+Xm9R@TI zeLK@#i4&A5`4RGJM(z$yk#o(@P(TCf@cF;0t`4uB78{RG!w7WiO3ya%-^3Nf#W9|JiKXwt@vGTid_n)>(1e2)*ng@uLZSTYg3r^;nhRl2iDTWf z1G}(POc9qvX;h=iC1=tz1>@=()%!R&Z0Z}KXzKT(MRWgAtXslk&9JZP&M0L&XvzCI z2(O}5k=+n-lr%c1dNT}wjUWHgb6xY1OR-<6>dq>IH$_+DWq9j?A#RaB7M%%v)1~pW zvQu%$pS=hU#c9&3MfGLnSGvK|U+@oD%>V%YjWu;$lUO=RZq24O?sg-|Y_VO>F5gBj zbB!$L)TEcQexrN%o*6G3%gHr);GJEqD`byG+{yh#?ipf`;`!^n)!i@uJjr(E2X5{M zF!t$t@V|>OMkeO}QH-&0{$GkQ7KZ;HJYpl3_}4NwqyA|8Ict+RyiZz(Th*=C7nsD+ z`42QaSTJspdEp{KD}iNua^(bt^qFWJ!y&Ht-zsL`a)Z4|0?l78Rhn4VWz~*`5hlNE z6rZCeomB-#D5Y@Ka6PDUDM6?rYU3^*Vr9&GF{u`&%U}BOtpXLPddBGk?_!x|1*wjf zd2t97&e4mJX}6@(fczp+*kEu-i6RqfM!a+j(ZmyHML1{g2^Caw@kwMNDP=zVhy*sB ziiOa+L%kE76gHhIT^*09Zav#PJR52jiR~_+aYU) zU7_TBXq;+Mdas1);U*XuH(2U>P`Qy4kzk!F*OXNrRYHOybd6N@D12_QN81T?Ve`M? zfPF(0hD|a2P(l(E_yYvS8!n4>Y{ok-GtOIM2X1u6r&)po4x?*EQX(f4#WWuAS>oW4 z;w%huW2Lc_{AM0{iUL+?Pn%g?m%cG`$m+FtaqfIp>?EI1{4%5E=#b^HcOUd>#3M}! zKI_sk0NKz1Q(rOgeDijD$ZAeZDr?5TS*r4Wqai|06(G^3<=GxG({BH zG*e5sNZ@Xvx#I~2&vbcATp7g3?GyJoEeSI8IW3lk)LEuTH8QkTMHflFMm06_Df+x- z4ByxP#LhQ9d9B`5^x`WBs?)=lInabCg%`)ig3`nr_a6@MR~$bWcHb)f;JJatQvcm`U+%_HOg^x~$`bM;@g8kuim+K0_1ZR>dzHOD)fo?*=Cu0cE8r4M&F zUExN8vy9rhmI3pxBjo8|RP8n0i)KOX`;&X23ln^tfKQb!YCg?h3C-LDH9&<~+(Mw^ z{muIu02a#S^){`-tM@J8kdv~MB+>yrPzx;bYM$HMssZesF15jmPCL}wVM=4vc(=$` z23skBzBPb)VMW}s7$|`daXMd2AM?a{26c`0SsWqzE+Z*(^dD4#bgPr-ujiG_-Y_>3(iDx=OvSo8e!1H;qpU~7h+(2NrBGZd2-@Y*w1(AeV@(aEH#H5- zH?=BtD${s&XPRKqBPGlxc(rgHMw($9MMiOzDa2uKo7jcYMovq;tVZA)W58~W;JS1! zxeRaYCL#h>@Jb$t2H|QEprE2IHgjEf13B3K)qol4EWVP63)laJm3>o zz!e<%uHDqu3f-!P%P7B{32ZCJ=6BoP&%@O)@P6FEZQ(|N4#OKvUqG!=7GSNL*0eq= zmdg4GR(UKB=~pNk6kxi)fe&Jw2+mbR#ezS7-q|uQTu_~ahjMTgTvVN-m6GPM9sFEHhQ0Ss3R^Y zXcky3Q`X1&8o&40hoTe)A;G=Jk{&{iyk7F)<*UZl3$K|+3 z?oaNJ(b^f;sfn>IJqKUNJL4vf;hMMr(#)zwCMDX3!`-MuNt(nuMx<%FnAC}RO|Ji4 zJA!&V1P06VLJ(3whlNQltO1c3u32_}3+67+0$ad^}lPi@a2qm5o| z-=_={%e2+H@L2lZl?UODMrDH|xcSMr#Xln!U&hQH3|al?5_+)(^un{KIhbLKP^3^u zQW#_@46@`*;d0L3rEO+$eq6>j{2@cjzRYv$SsN4`M#ML>RbGp9sukT_IbgIOJ<$cy zW9sh~3?B3G^r9B|r@ittyx^pbfil(`;HeN-`Hmadnz&OR$WbD#Uy$qmCc3+w{SSg! zyA}FibssMS2@q&$J+{Y~6qD?(GersfAMOuql=_u;0SW^SUL}(`6Ym& zi)`yEKxcNd<9g5%2Ecf-QyKZ&pqpZ{FHCjUGjsHZ(JEeB#-}+MC9g#wkp@T=>6z4XSdJgdaujaN+0Nav00TRp~02*wahxFu&1a6%)$NX+@7j2Lx?Q8{?9cwzHyy@A1g^h zd3Z?o8om3d_pg*9ZJ7u*QHu5#NFjtfPDt<`5l*@y$`h*92aD6Ec%2n`U zymk^sCcMHsnzhj5Ap~BVj;EH7o}2A5*H(MBN$T;5&F|}4`o}1D`x?BKD1?;5vEQf% zNvy=!S9YGYR8AXdD#Lq~Xc;OoGTfNx?0zmZ8Ol|qww1A0WlT)Yc~<@Jf+7j-G#)ix zPB|7vQLN8@SsZqTCk(pj=I|fABJl7+2w+YeoZ%GOz>0D@1L4zY;NiTSP>~U z5bb5o@mfKGcoe6JYW*RYyjjUKrR%z@mSK(z%T7Qq={q^WJsKV6GdFr$v>(3KG7pz= zwN{&}ze13`ihM7RIex2D@fFGAgL8~8^j{YUEFnvTyrr{#6MUSuM?FV&+>%kZ2wvQ8 z^fCJ$`6OrXQL^)}I5vh9VX9Dlc8P|ZGEFIE>Vg7go6@cl1#6a4BjtZAw5%KdwLILV z-}+;mI4~Xf@c8oNI+96(K{)x15B-|7m&uRvE&7A3-(@&~k4fS$$^K>%|8$<(6k2Y( zUl?*i^Z_{I9U@0=WRlF`O7j6Y2g6QyaG#7<^1Lq$xXB?dmXGQ2+~x+J9@Gi%Bz_iy zz)PV3bqq}0;j~=6p+7yG_`eS}(LmM?9fejo!IWp`LRO6xjZ)rY+y_)hVT2`$gl@pApw^W*ps-}aamxk?JMI1}*B>{^Eusbm0@_E8A|JHtpTHYSrzuvRi+;Z?Vq%W@9;xZ=U`x*?8)V+A8Rcj|qy} zlKt@g=-KHz*4JrC8?5Vmt5HR|Tyy>DLSBA}$Vl~uG)P>~s%xc;$=9gjC8hhOZ#}?g z^EY`JKG26Pu)Y`2U>D!V`ZQJ!)?U$StbZD^4E8UjTp$HB285KX;_(f+4r7j2%Nmw# zJzri4LUZCjH7T(N643dxq{R~FGq!fkAB7hksB~gjIy4>p60%3WI#K4?^9qyAvX-8n z`h`Ku@NY5~Yoh(l$m6gUMe=Xr=UK959oBcfn_K3(h6l`B-k0Dt(uT*OQJf(xNoVf(AE8amxD~d6pH_eCCKtC8UJ?^RR1TeEHG}|B^tB< zA!epPgy}F6X-$HFG*VK|pUB*4&VCI0ra_$w1K&t3GTOYnx8W|KP$@{Pd^_xTGE|D} zwE!0ZuP`}zpg7#Ef0y)3++b73>IVnLv-E`#6G~LukZq`Zn>@5Yg$ipzQ2W?q_3d4k z?UG#{Wia`Pdvgk@nf)%*c$u68ny_dcVnbLe}A-iG$A7XGsc`Q?6#GmG)F|s#>T!f^u7NkN+o%R7^O0pC>k?i}D z6r%=YU!cL6p8M>sag`Hd(4;?2M&7Epoh0i#EaRO8$T&=9+aFvBoPmqufEj{P zrl8Xvx!KR!vsp8)1iRDV=&sqKt~863q0jk&>vo1qU3r%%k6_QqVL{tt?NYz~`r^dX zn6e6XmYJloPY3L=qW3bVmt;m&f;nt&Rbr=#d6ewLs`V>L$w#oIN)oR@nfk<2SsV_) zh;5N!R2M=|Lm7%x7_Z>vGh@7NNi|yXxb1>Mse)FFAim=ZD5nlS9HQ$sz2 z$!sjsae9lB)|wiwI++><1S*ym@hg@#5e~8Tuo1BYV^$f;_BE$g2618xIc?j$83j7X zqx4&pl8Ml|O`h4**SanV@0u-*anz!+NXhc#4xL$A8Hb)m(pYn)q^!(m%oN)X! z3n6pLV>aD5T*w$ZQfx!GUY^fdIh+u&+O8^b&}2(^@T8d5OQG;_2#VA5E?P+#VC z3GY+i;Fdy;Nd|soj$v$PG^!qmX@p~5fQGIOm?142N_`tXT<6nZLZfXjpZ})SOp_m( zEM4Fwy5<^#%kPo{;yECZA6^f$qhkE}_TbQ9kV}MYNA)oMU#@emKKur`Xk;4awvx!@ zd(I!xWj>#VpUQ@@#?4&KacmEva_g*ivq2KW?l0{De$cm%54xZZJ6utwyA_=2H#ec# zUk0YT9h~W>V&cVy#F|puo2QQG^i`gZqlR#2&;c7=NgxErcFb+XX4C)-fO&$vt{}ic zHGC8b8xA~@FCJaNg22etfWe3=Z&edpBjWlPoh5JQhY$l?vT>vF;`)918QdSz!^vpPuKOFdP_hi0po!#iLCzCd@W3=U z>pG+Jy`O(gB|T?51X^%~Xe37LG&}$(x7sL(1Jlewv_m#(gHypozhK#$yy$^Klq}^s zxd#-IS;*}`s_<>p9|Fy7!$bl$u^6&FEY+!@3QOO_m?vsZv3% zMlilMv<4poS2#}BsEZ?m!Y8*)A{m<~Gff|W0p+9^(J|HocZ=G#7$l`z@W`fbtOI za~wEnh5`|6zISwXWKOm$tT!iD10_9@f~Knv!DHdp_vXe4b? zN!X;3w8OFgs~&X4MsB9ZhMYEOtR+2#lEw`8SqT{+?M;$9h>My@zySZ$q|>IGB?&bE zT!6!4AzmCsr>Oo+bnjybW^1?BYuNm21Its=_O~E@I7wQWsiGJ47XC4`NzK~3!H|bG za`gbO+~C`HkUfG}_Jsn7f>>hyrhXRl7hP&T8#5LnWPGY(K)u{0GfL2z392A$yjyYc?t4vx6 z5WUj`I=&YCehIfveAg9ZHw7zI3Rb!}v{-3Ssob#Kg*+9?dr;|&Pvg8uANJ|(`faDj z^J_F|r`N;xo+#sb&8LGSYkEfZWy|KXuB0dH+Q$fexTODMnVzHHB;8a2{8{Av zRDRav+`W+BJ^AfpS^!~SZToqsR-olqApcYMcJy}XM3C}h26o}@e^wi+Ke<=w>d%?Q zmyxRvw>}uPW(7*9lFC{fnNS^8Y2Tc8@+H_!9$!y=RiQsaH{<2aoU03)R_eylS^p)= zPL&N(^uW*~83n{(6GM#q5vgR%lc@>qMaxiS4CprNTZHmtNdTYTIFR09(cEBOJJs9W zv3W>ZTW=C7xa7U0`?50VwDcuom^{nsP1SrabY)-D-?cIT5Ts1`2BCz%tyVt>c)^c{ zx`GhjT!Mq5+4-WNJf$YRhR626oeV`Iq$M?;B^XaI8Z1)Z#JYm3=x%z&=*h{Hv_ghY9)AGIL!_8IYWDsslDCY)Maf6R?8Bsp4m@5y_(Y0~)Z=#ZCel9pfqvwr_yc6sF8tk!bPEXNJfQhHFZ@Il;xPg-%T4K>rPz|by_CpLUE zEwby%gP*Cd6DP`v_%(}gODr@)*YdT8j9{IYS7uY<*b(uv8qs;k$IC%%jsovtS@V@o zXRTxOd}N%NZ`dg_F%F;@Wffp%&tX4UD)C6&Y4u@*H*-yut2Pb#?BCSWR-<-=$G-W% zCo+YXL*{5AaCE0zl#%MCBEd0R2B1gNCtl2tI59lOd;5KpoxEE6*%kIvY^+CE*mt{+ zH`G~KZ{Uu=STL6er!zmD=f-Kg4reI+Xgor)W0BEj}&A^{Lf(=KG~ zk@rS^U2+kvOT7^8>9$nDwj|`C7i{a5DJ3$1LB2 zYQ~Y>$-ElZ@S7i{vrd@#n2*mR=vMYV`Ty^B6{Y8AI{WhMhrZs!e>@2L-V4Zghxp&I z@J!51tpB5wV`cc?LI*cue`m{lcWQSJzyupLNutsS#);f;YyvPCN5+llHDT5>pEX*< zkQTA@pD$ieNGJ3LHQPq53=^YawU@T68h%r!xrU>v8U-;;b7CTuPzWi}yCF0kArP_^ zf6a_IqTZ&=h&*&st9?F9O6d{k*Vp6rzNHVF;)EnSmho@vFrG84{; zN~=`(!4^=YzI0rT>N0vRW@E((c230ng}z(eFoT29TjLp`VnTDE*@7Q;xbtUTc>oE!SaT#A7&ar-0=pl zE6B7*t@GrE5fre|hlYf&^Mu`lE*}5BdPG(an0OF~&5Z3t1#E7~A|kd^^B(nRj(xj9 z;p-fvlz`U){g)z(p(je9T&M>lQjjFV6eUVPR#9R&5+hKe7g6Es1@59Ewv6H89x(io zjMBEJjFFmNTiT1MWh@YQB9c1jL{`##l90E$(Z2Wo@AuM6eualWqZMCom)=X?k7ip3 z+cw<#wC=6npEjMk{&OwpwNJxVHR@TDsnMb0PS(Wlji)1RktvsGpDrp3Y(2p$f3CUNB%aF0V23Rty#8h+r}x| zJY^fFY}>YN+qP}nwsouf&cyUYPsfdzh#mXS_y1+D%$46#axyf4c3`cxOU#hIEgR1H zi&Nx7q-{G7Lc0q589e3}VwMsZv~XraBpQ2alddpr*s!2Jo6~hHKdbGb2V$$j?0)=6 zE6#2Sz!_y13bwl-QR{#K`e_II`cytX)bb@?9=~g~nD+1I``6u;*aSKoY}qIPp`}BQ zc2kZjZVHIxDh?Tqv4M1arvgIbm8pJYu&O*+v@Zz#ELbU6F1QW7t>jm(SFsyT0INmY z@`W_CG1!5=5%M+E+0o+O)dFlN4Xltro&^=oea_}8^tRP9^RDmImG&fb_IV?}8Eu>x z62n%pcAx_@$NT=w)mZJ*C4x9#)e+2lPWPd3&D`rlcd}u+BKK(Nv^34`K(V1kpWT@8 zZ4N(_JBHCC0l7sU$M(+^Zxcdq3~5mu`(1pNJE7I+`8d|hsDcKuW;~QTZ#_$>=TsatjLC1sx8^1b#J?X8Ccy=6u1<$g-H$vj2Vroe*5|b6ItTi+;$QK|v ziCF@C*57Pbz&32b7b4ZtP=L=fVP=5>0?= zuseM@@xV$XOd8WA1Tc~2viiIhfcCB-_B}lA^)|e%OE0At zCoqY}4hGZ$8NHGH$3j5Q*2=Ox(y|a+n*t*U4X{K6lppKZB~B(yac-=1z|w4E;aK}~ z2EDu#)cyzOyMlR#P&-_>9}BKJ-z!otsrWU-6e)dL`ZcUkS{y^X1ykitVTP?TdZr0+ z7-n@GAY@b+GI{6iShEkOW|qr|-nO@ruwyPI-nD{>NAGJV^SWl5eXYBJDm%*hNxM(? zdheP7kAD{C34?3ZZ&+wF@k8u1uqxs(K^ z%}%47N(K6jQ9M(Ic{=y_#83utun~!oaNDq=s^DeZ!l>rLDOSMT2nHntY^V+i8vyqa zw;N6COO^mZ({!Zbx^baUJp5WEpcGAKUt$rIo@Hl1!QWnm(m?K3s#o~DSROp6yFiM^ z9|JNVGC}|iA7t-S(i7==6Ksb3%F00^a2}DtoGA|lyy{@1-pC9^Xccnx4I%S2&?8v= zz_G}H$CM6ieR52F*69$e6QQXX!~cD7qwC;CPAlZm@eN*Q0i49Ja~ZR(efQzDGCf0O zpa{bYz+Kxaxx4XN-osandc^E02B>r5#68mDp3?gd&p^@uw17j3F%+Jn5P|_^@v8kc zVTH8ooJMK-2O%k%wk&72ESuYd@y(fmXJfJ1OCY6_2GjD`@Q=Iqu1-fhFP8Wa$npJd z>Ty9{@u2+-vQf4%Ic+|2yxZE<%*ii;DI@DU#Q6O;R(Ba25A>Nh_v066RX!QYCMyFh zrzLd6fLmc@7ZhaZ1k67wZJJm5M>15mu)8i>BHk9*+cj&%q7?j6=7q=+T!090KK&_IUw$CJl*IN1Mi~|^{a+dE}VNGov*9M^< zhhN3Kr$&0%VzRf6Xe1N-TQat0XomHd@vNm6Grv-vy?Ba`SLrUuK0j&_xT5UVRZs1NbJ zpPHFgWC_i==X6$pd7QCR=cOHC#u*Phd5!!GkLjXtuH$cZS9Z*qawMZN4Z!q{&1~27SFY>KL_hl#7&v8|k9|0HV*lQk!P6 zI{8scY3lI|q}->{r`)INSAkHDP!Y?EqA^e>Sb>A^XNzcB-{zP685?tm?crU3Ojid$ zvg!^I)FhQ|6oZ>%l3`uOKIPuv9=$Hbyu5V@3Ujw~S&n`3wA^NIOGr=yJPc&F`%D0e zGyc0l{kJeAX68AuH`UHG{QQx`g8_&$q!n6bvNh9P@H+v8?=m#_C}1L_cd-8DX;k72 z4uzE;F;AKAY~yAkDu*rRW2zH>9h8b=Q^>(Hz{lO|&T*59tKM=AAB|M3u!TfGcJdPq z+cC7;IHTo~J+edSuUYQc0^L}<_?1KNUcKh&)4w(1=T?BvuCdu$IzJNp;NZjCxlQjU z_~RBxZ_+XHpz`{t`a!N~2Zb0c_P zhuvw39&I+eqfVjLtMQ(#V_@`nq4wIgUkH-P{u*y{-4-u?ds{P_v}NYl-YE_9wK$Cx zMRvlGywl_QpKjtTa)K;-R}V4RG9=0Jz<8>>LP?7MEbVI)I4ir5EsM&VlRHYb!FCo^JHAnZR~Qe3B5Gh zysumg{rhWzm;P4o_D6TR#miyIN~cnX4gI63uQ<~NBiP$a2M{uS;@TYmq{rchjG3nC zDI~$mCv>@o2UooPGd1*Xl=sk72u95_v+>PX&eG+s9z@dow!UZE5n5bMHKQU6^EoPh z=xcAP_Npd$cumyxhl1C{$g~#0kTcmO7lcNE`fW#o7`~Erq@G!=xccygt8gyiN3{PZ zQ!5sG{S_Nm_z`X_)v^ThBjEiWS*=eL_dv|`*$0qOV3p**kNH@b*#3{&kco-?|L2%b+eY(0w4rX_$g@Bv zNeD9u`WJtCGru1i#J={WD!L!d4vD1Y81~psX^6s|p>-Q2$SBfNF{}8lZqj(dlX4W;H^m%1p+pNV#=-pdK_hW{1xjox zr6Pm;ik&)U#?aD3nZ*=AC^E@e(>#qyx(XB-nxbaI0p1m71B@CK%086JbbGc=RxtDd z(o^L9NDh^%L2Sq+SER^Ntk}`e*l3~rIIWMM{W0YVfu=(#s}$+{a^`mF%qDnBvZ8V? zd$OZ^FA|&jmvu2HvXK1uya^l!D|DIULK|62dlN5Rq`z8EMv}$7f2fTA8Z7am{vEuq z(O-=(_nC8}T;fSZ!b?B5&_livoGW4Oymo9v*mEX#DwOZ-kba+s87Z){@(nxks;r?! zDARxvy32M_8Qdqd7E_#w<0~ zmY3wR8HQ|&S3x3n#g(34&SLjjQ{}gu8dyba8$dKw$)25B06&6>MTDQiszgv7|2eVH zd>qy>rSo_k7TEHmCn$d)$Y-BaQ$g&qE01w|B!l~9i`eWpJRduODwJEzVQ2qd86-|5 z9*<-qrp$(GuZK(LFdiSm-QlOSXL^MBEhL(VT}71^k&N(O&UxYaQaf`w|1p`9i~RnX zboBB14mkV1-)kRnX_)eCTlIcgyO=tDpV4-?+t#5uac*;8ax>*o`S&vQa)R;eIpY1D z{u#I1Bv(Xu8-Ig(&gJ-IT$_IZ&^`@3csKW*^bY-jK}{I{LxUD?Hgslhbip9o$lbxY zqa@;;Nxyxaq(PI6{rFI(k&;Nx5{E&S?GTCV9rsD_MX5%EC5?H3(onw&qy(a-5uzVi z{@vknb9qkMcypD%lJREQcHTh<^>Ju9r>+z}M8%FW%5)6UN2G5V#mCl~avgQG zL!){MrOWZk0msx5vu&~qQrovIl6nU<6f8g7^1`#8!)ecLhx0I#ll1s-GTkPaz+eQy zPjnFwJZuTA1a_#lir;J6@QS?ZzHjwWS&?pAb68>)ptWpJ_KzVW~6DXP6 zcBu8Kuz>1TxSzHON3>Y!~LOFN@dp?VCB=~Z< zaihQX{RLJdzhf2kY587`kBZ7zwP%8bCVU6MuYkbCyo$g`2KEZS>OaNJMA6;WuBE(G zQ?l~)Z)+?0+PU2~UUh*IXYYQods|BzPH+&X5~OHj8)?@lgep`LqUv*Z$)*)(**Gco zU5^!B-`{2}Mh>ue;mrB|lY>uJ=6A=8TrT@__N+z6q+?01MpqZ!-${B9h9A74NsVF( zl;C{%t0pq<4G=SNDt*#`lVsKjH-UNtR&La_3mq%2=Ao?EwOM|W>n zE<{Y`E=VQ;8x;aJ>aHkgs{)|4lJ9?(sTl+^W+VX*JT!!Db80{-z-s+AA1&Um&AK+t zCTr38AHS~#b6-$Eh~!N7m5iaOqobNUA&}MtX+b2VOtQQoQ4R)$`Z0VPq?(=0>SG(U zExCz^n`{^Ymcx1^kMpZ-)-; zPVP@tdsc(dR<2S`W3!@bjT7_3z_PoIMPLlSh5_*4-STtaoZUbe_14giKmi ziGCgewUR=G!d=!_O#k4>au=o!MjyQe&jJ}iDTMl7K?C_s;4CvytC0`Jj~@7DhZMel zL%I#G-q1^00PvLP>{baJ-33InOeTM97x=)=cLsdffPx-@D`TDt>K*1N9SVNf0%d3d zi!(_Su7si3{A%3)UE0%g-51mmQ;JpQNvqzRv^cTJ&8%d}sXau;!%#thP5Q+T+LBca zB28dOFbqp3CE@gBa&9DhwxacK(kf(y^~DC)Ae;Fx>eAP+&+;v13DMov3IY>#UIRrGo3(=r z?C!_h-ILrgHMyo~bV}3e6r<7qjYugQnpoU7ZfDmCp4&I3D%o%-bz+t*LqK8~CtV=G zA=$+bi~A$_Pcs!VkAEg5k}Hj#?#j_wnKfsM{jZh2DSb_MwytLzb|1q&R6YOzB99G# zDe|sdY#Jm0D24Z@0qH!Fe_JT!pM8PyfZ|;{VnQ_@5FuX$1Q%kpa8G|G>JNliTK=F+ zNla()UtSU@*PONohIX$a5ic}v0}c_25`3)4X{g>(}bFww({Qv?aM%L zMYq@Yogg`kZ9C)c{E&LH38I)k05=_PV0k+sDEqYy0!kO@-LeUm;w3S7^wL9QGcf`W zC<4qaxoMiSleA{XXeFU?eaV#34lV^b&zZ5yAN3DJk-VBa0rrX>YXf(<)#KI&9 zE&xPN(fZsW{{6aNe>9y5`*Y{bwD$MLh5izA-}$=5TSfR0u;P6++gNip-!d&+ibYn^ zxE(S*AkfO!xh9l@QE4B%luJOqA71Y!c~q~h1i1v3%bd`M97Gp^B^_)%*F6ePUW6|$ z0Kj2v9l3}eeDRB`$vF$eX0wRY1{y7Fg4Dp_JVV>y2LeZQaBq-YXfK@;%ibv&q9rE6 zr-5GbemLk1I}ibw#4lYzH-q|9 z8Vf#lZtqCB!z6is#dh(L!L}#B#-fIRWBvkpDS0YoI{`)7OcceN_Py8_r=0&`zRbcykfuI2^a)*dA1;<#T29vhlC>| z1pjo88PW)Xz7eMYGvd?$T&!8Yp5Cb$3=-!dr+!}bUjV*?|w`?$eGm?))Jv z`Fq;vVeOv36cRv3Q`aDo(Ce%(kxBdQ_|e#H-OkA=?y?)6tpl~DKCju*x76W2(w)&V zY(!Yq{VH*XM1F@5xrqk-LSFaS8G(as%d7nIWvy@D-U0{Sd)vLe4bsauUwv0^E!sUP z+#@!+%D#N`<@G5cdLU=jcFLGUqv3rPy!9vk z{D|V#xl9Z{YvlrMj5j-zPsBLX>)@^X&U44LTTW(F#tc)~%z<;0D`nyL6S!b)M!c_2 zyD~aLAQ>VIEdwOj$wlRrkF$%+vZFfOSE$`J(Nh&kzLh|KM$(e4BdyDU9=vt9jYKeL z%%fQ3(@~%@(tv1NdHUL8cju#~%%@LEkP+Nd`ArE%BAncNsumGRmo4nWo|j2hO%f;J ziDXk;M|_6+ND`y)aNxgAE;2HkZ)i5h*m#PeArC`Kp$Hwm zIR5W{AIF<*NjzGedBztNZwoN^t}C2|ZQHnmoh>Ydlb9MC^UFCQy@Q>=_bxSLM%{jr zo?P2r{Ioe*_xXdr9ehj=Aup#sfNLE%RsUTEvHnV&|C8I1iG%HbltG;TX@vX_Ts-G5 zuY{Ku%E8gzSlho1NO0)OMKeQUJ~@UP#5G= zFdc&`2;FFq#i(MD(!t2^Xb`|af9ODb2UD3L(~m#k%|uG3>A4VBGkk?D335_5jT~}M z@G%XFax9lo!5Ji}!kwW|{@U}ssHHqa)sujKo6}(F+m$uS>F?jGK>RQrj zY!mbn)-+=Dj-dv8BO1~4m>4*~+5xTTKQF#3nJxQFmhP#-#sMA^jS}~EF5G!=}P}g87 zi5L&>o%{&51*Ptq>PVCr4bT3YRwcV}iiid}xetpRbdK5@kOwJfutwiLR1ptpn^eZn zlsb>yiY+If{pz9e0Cb-4z81-S!Te!kRHfA%rVBIRXp%lw-y$${TY9(Yk%5E>QUxBd zYt3>*N4;4mN956@$}mC6a$);;Wi}(q@~e`Mvj{fg6n42kN`HUUN++NQAvu7H>@gU0 z*BP?Qie`AA5|S*z$9KDP3>Vq`!7>YBE(6|8Q?GlBJcCcBvySEmO(+jw;Co9|H|#uZ zmv2VoX@#9$A$m99zneRDviV z#<9@{LDz@2+y<>eq5WTGw|ho>Nz!Pxh$o< z9SA(5BIDfYKyDkm#ywlTg2u?-nxbi08VRY8%y+o)m5PpMvI`wuN7Q3pgZ3)P0F#aW z~SQA{#W9 zjV1oXq|k6iO|G8%_TllT{7K~n+GT1QBeqawu?EC znpYEZot1^-7Hb_4%`zgv2ZX=#a}oo!lOAO(Knakwq`nQi#^GYlNrx3K_UbPOLy6j5 z{;X0(U}rPeh|V)*)% zYWsR66R-kDX=bIt5_P1YWm~q2b8+VTFS!40Mb!CHNlfXZVc>zxx`HK)1hKz@(*;}F z5+@+|P~_cw1KlF!{aDus0LY&V0wE)cK4PgHMRP$hbewB%0Znfdf)PvIn&RFbE`}|w z0);UXD5e27lx%Eha$X<48_rLl;J?3Ri{cFqAf`h7oD3Dfb0S0J^ zlh%mI`zc9)YpA(sPhwwh?i}{6fvrf3Iy{uyuI&U0{xn?fRBK<}y+BJv?F?h45|xkx z2Yc#iO!HP|)E{yfoSEZit1nU*WS%ivvhDVJAZJWiCI4d&^k*L`$2Lf=wVzyT54rjl zQuQ@Z6@E6f6i$?w8#hAq^WO_5@?WR6OYdxJdBSgImdOLoGgwp4sKdfsP-@07!m7Bq z{y7JF%E35HOi5yzM41A8j0rUgNVGkQ5(I19GxWsC(V+pMxe6Xk?OB#8479Ml14;#BsCE7B7YY#@Gw4}pAF25>bl4Dh0%{f z99j+n0ri1MD2|DK%KNqcx+I_POXXub>0Xr1nsw@c5T1V=fcSe597ov!KYz(CN%BTg zGLz{UM^i8krC{z!!QNiw2lEC*G9BwfAMUmnBI&dh?NT=uRNfN)QsTF_3dWtrIa^{xLz+yYVQCc6c8^`JQ3;$Xk3eRS5n+GN~=+?!EX zS9QsKu;9eV1n@=IE?Uw_1}U0c!k!n4)>6RC)7}K5Jt_cokL!w+ZKReCExS-DU*kh@ zxvCdeHftIn_r|!8m~kiI-;%9TNvnBup3Ii44k&gy1N&_pub9l7oGv3W$Q4}N zn|S>;WIC&ZblBc>VK+ly4JvIASUNg~~ipcE%C z`41;1<9z5-C9l*hGpM}sKdNSUIt;p`K@ZbUE?R4FvBhnoLzyszc1v+kbl1wCqnL0# zTtqAz4b3B!?{sBgR44Xz^-~}57hX$o6-hTSGMUN+x$)7&U&|X_#o4XpTbEEnlm-}- zyzWmoHHnzg?46u`IQ>-!swQbdrir^Cd`}SxLsy>7rR2&jI&`K}to}w_?``c?>u1V4 z(%nLazxe=`URz27PzRZZ{fSJ~^)$M#waybq{1m|4rjsB@aA|YH$g|u|?=d&NHhubj ziAn^dsNxc!ii&=TiFktX*LvdhFxP`i@)->rx&DnSQLT+t+FO=Oxv%e`JXf6i@>?z| zd#@__+ef0@{;AQ|b{E|lF0z{GTb&NLjp1h#sV;P~x1Mn|(w#(SQ+XBP>()d&FBY5^ z70om3WYrhl>P=6T{)W0fo_#U!jwS&re)Nj=xYjWIa*uXw&$YKFtH5n?v-@K|D=^jA zdzs{5#huJRSdCKe7G4@b%Kh)~1eT1T0J}|AS)suL5aGL&{}iv~Jtt zW@*a5V$&oVB2V|!`i29`czDK+W@>C~%m^^9Hk1+47GSn}+@>c&N!jH)q^MZ5~#B_yJ!;i-=DWg;n&Nute|y9DJ--TYEyJeRujtZG$SMncd zfoViiW7xc_1GG>VeX@DHq_I`;8n{nh6iYbktER+Rqrt()l0EC=YGh8MzRol{T8UyK zVw&3Z=5L>D5bFA~(dXnxt-5hJtN~NAW`#a}RWrp7N1tky?k+ zSz8r9ytq_Pz*dgLoJvi5&jzkXKf0Lursh)Ywu+s{%ETP zo!tf9r_J{XLdH8kNry%E1)+K*8J4cU4G?)lOy6IZ)irzCZkjY;>&3%+P{me57P{H5%Nr+s&!onz#hYG^ z<8_fSKg1zxR7oxc9;}KI*tKnV_=i-hxy*GMG2K0EPI!fC+5ARJY{zm zUH_N9$>2V7x;>-tRetxKkR7wWC&gV1Uw=dhGj~`5fSr)9U4BW%7$7WY%Y5VWKISxR zn#5?led|#z-#;UuIU4WNHW$sOD^LXo_stCc0=a!iUv88M$4AJs~65eAE^Rgm}*i1!b z;D+;H-hUS%$2mi(A0T)u@&%!bjN3!glguI)_B)sto*lL#BL-p!W-F=d3;T<$@ioV0 zWsmA8l7AnjR={p2^ZuBTTY{9Cn(m$Ov$8<}(hXaUkRbkc2GZ1@0jE_U#KD*Pxf)PI zBKCeR^K$6Q#pVlGZ8Ut~4ZBQLWi%=5G45=T!%Y_7UKpP}#6CtiQQ=UJZ+{T8I-o0> z&ku6hZJ}u~{3sOLGFs`CsYXz@cC4ba5BV;IKS5;4S?1bmINWM&Nrr1e#F3*oZw?C&<3ExZ6n1pLH1uyzgLs=Fc zhN!_>lFwUo@^zt(8vclHEXH&O0S9_%o+5qTQ#~CNu3q4*GLU{1 ze?XsW6){#CA4+eIE0;jvdt|NJtRp{64lsnTrya&H9|ZWt>(9T^G(*+kNTHEx2%zuH zBx2yoF&0IvU^w4~yp?d<>+yM_k*uA_KRp^yMvT`0N>*Z7e>KWo2rZAS5Idls!CIuc zXoU+5cBaFb0Pja{Pw85Y`YI^o!%DoiQ5jk#c@meEl@CAa-BY=EAS`mtKkN|HL={`m zl{bSxM7dr`%ObNr`)OI-Q7dWtJOJGmav!$J-aW#KOaiHND zGPs^NpyaZILg_FRXt#d{T3B9mSWW!&IrbK)8C4P^Max0+alhMk@gbrPFwY|*z~K4T zcWmiTa|Hpu9Iip^WFpTHbeq-+5pJ9_=Y3b-G5F4+_s8uQ*ajkoqJABrwH811!yJ&6 z4@$qzuE#xr9NF&{bwNmrv|fok2?-H5TUkaHpXX@~HD4b8tpy8%PbU%SbLGfx=T?)5 z-8Js1y~w!ukr|?&SWAp8SXCrOFkMz~4=2!h##w@MUy$4uS|6v$WrSm4)m^i`rz3}w z@l)kRrnb)w-P=BxM+#n41MTv1;|vr5XcFIN9@E*}+#C#H8qY5g)MWz)D^~^JA;u_F z;@w5WC{4{XzkMmmf;Y^?mi?GzX!=c21(~^Pq@a=#Sv)Dgjsgd7pUB{0B9sj_1uRD% zy`&%T5?$<{cR`DPMKaR_G?|&$?6^K{6FBEgK&`quS#PbTjjzaTo zQ5RO($2la%2gGk#+1h`o2G&}_p&#k5eG@TY$$}J}u#S6c%_(?8Z3{j~!N?+(S-_{( z-=;^1da*#;33FQP36u&sCsPKsS`xuo3W6SJ#@0mhN*~fp;x>pmUWXBw$Zw8f9szL; z_*`TvPm?7*mMcd!viHRZ12uo|jyYo~4Z{Mf#)JTzLkg$gzby%4f^yfHEX&FPeU<6G)ql6swd&7+!GllFand?69hQfYIJWA7Dk5#v$9=> zLjbmk#RTYhl6^1w?~4Bl(HU?Tg=a0Ja}6P7mNdAlbos&XVP*u@E(Fl{qfR)y)<61X z%m|{H@@l%RL%wf!8$(DycEn&( zOZ1pOO`Ru0xq$8wFR6U+kM}jvvlRMR>z!09xw32>B2(e#!L@vF2~I;^eZJTl0sxb& zesq`EInfC?5l#B^OMkTPS9Z9Hz+&ko@(t~%&iT&?O_Zh6LrTeEz zZ8%^RMtXjTQ|$7dGP4H2r@1VH;zEKAX8d9Dks$!!8E3AhPUobC=be9$X(%DleJ@Fj*&?4lS8VfL#L7`Li5MZ&{LKZ=Wa7cTNg?Og@sSh)5e{E6~=uXC;ernOf z5~sZ9DDU*%1BsH2)yO|*gCV=dY|41Gy91l)QvR|TO!1-YKo9GEgYw)1*9anZN{ybA zKE}Tp@Zk{Fs7p!CzPr3TTEhqLHP&S?-B&Jh8hTfv80E(JI-}E{z!{?OP%6A7BfgYw zep^*hC+-gyAqIX&q55;#B|VJ6xr1qrU&oV7t?&XkD{xv7;|a>(!YKj&#t9q|>5DBC z9FlIEpVjsvSNW)T4nfbqac9+|Jm~dnUj)X$a;%8=N9<78{MzRbdrXmvNGLYa^N+gH z(sl_UDK(`hh8mqZG&a}c%o%YiIDWIs3Z}0r*sf!sk;ta>B;Q7rV#W-HW%v!S6)^0Z z@T;&+mB4oPGC6^y2ZRUV<5LMPpg=LeDrkKKF8DBWPTOJq(#4#Wn=*4=6?ktU7E%bG z#YzGtQzukCHL1728vg-{tY_I~>j}>%x}8Ly)m@O_`LUFzvb5LS(BLL|GAZ4`Y>v}H0DZ`WD>eL!072>g zv4Sg#dG^I;#>n%pr9CJYIl+g@@Xo}jryBwPSy5K%_dWh0E^ByU!fE@F$2l+d2}rNI zx7{Qre(au~I3g%MTME3takkG#FhcfdgTh-+Q`IiYJZjNj}gB0;BdT zeQpvz*}2PKuM|r6OSJzkzwjU!4Z-@r+V{(c?$K{~<86ZpLdJM1!y=M~$uDcs|A7N5 zW0I0E7u@xW;tjU&-4MnOR1p{+qHD?NYha)a2YH9Y!&)TTJ07PEU$N zRN<~0(mjKX3i52T=56kH7Y7}jOzGU--g;*$tf4JkyJwJVWwzJa6rR%)Ac;eeKy>_A z0CWph)EOXQC(`nJ0*rH@H$bEWZVpMoB{^c>qEs?fpbKtilzwZz7UW4H6O!!lv1AekV$>1N~cy>Dl-oSqEjnsJA z8#DBA%4tDKZM?Th6Dr!Ny~sI^43^L-CfAj#Yz28wdXEsNo92*q5IOz=g9GN0JQyY^ z&dQyi;WxlqL{_^^W!I507t?s<@T3K5#-(D={`toGD5@s!;i#wIiiLVOB>Rk5r-Ol| zz=^{i`P^b3p*p#V%Ww><`wUw#VyqDJCrtwGzWY(u)eX?w4zveuefc2yO@n;qrEmFg zH5VM_zasHaSF*x@53cC2)PZQtKI-wxr9$kIu1_t}{&}R+iBJO@ zwj^B|xH1*08uhU)%)Kyg485hx=F&|WuSi{;FDGZq`j7>7@g+pe4D!&FKab9-;ZHE9 z!qk+enov9JFFrbDXKAJPOF7_Ds_s6+rBb=7%$%h`BPS4bOCRdwoToWXKqw_$CDzh+ z`W#U0c8Q}qcwo68mY!)U^c7TFqC?z|90Ym`UKd$Y%}8G4edeAvarrD0Sb{Zf{G+tmc`#YP z1c9G4e_P;IT_a>_Xii&)_-j_wZkyTjjOk#yv5^ulI?F(l=WewA-JNKM$5^)?GR;(j*%nuUyFM*-&xq6s_P&tPAT;fkYw}ei zAM}ULHqwycm+16zh`6Rj9(3zk1rnIOZ zmWT=Qmek-8Ic?g!Th$rqt~mHX!-wYaQxS(ioZ?ZXKzvYz8R8EMcIU%WeyDWzJ2y`*tbZ8lWCs9pC@8vqpo5GlSE zxjuoMgis&SLl7+<5u^;lphsjAC{jRfN0&Q)lIRvY_6MM-IKH)gbVHMh4_PtOIe5dIRzZ?BKRGt_0XG)>=Z zisEs9G-Sq5fO|WlhE`TeCdGs-d}O4Ulq6i_V5PX|8s%#x2P3$vk%I0I?H5lSiG!*0 z{yC3D!ZLKhb~@YFZ{YBxr^_-4OvIpxevMv59qa1F>K{vpCOQ!`*#=2?#5A5AG&A$-mL9*ds#tyFUgf zdn6md*Lds9mrMHiupgw3`erk?gL^4C6?k{Uq3dxMqT6oX&pmE+&~=DEaPGEv%uHs? z$g5;0R~zIAM1(?xz18RYU-H(J%t^{DDeN0BY2&%|P~s#b)P1E4)VS2d zoEB?s^~#9pmNL%X-o-{WWI-DS*yBjKgr`(c`Zd9y>bn3jdrh!3C=lAhFO3Ry0RA)}=OljntB2ZyS#x8%XoHIkU2old_Xk5^Ki)q7-^z`Ov(j z=94WzClv?BOk{2@fVh+*BQq5AhzH|w%#K7V;02#voF0CiitT+0RRxTYTDfkpp@(yX5APTX(M=XhL41EtG~@SlG7d|YBJ@E; zcLI|12+^?{MR}!l(eg9Cer1c?$Z8@!XjoSx+Mm6((v&O;70d+ZA_09WqnSd6y}nAE zCbCwMy@gb+p{037>c4j?-}nNdh=Rrve@}f}fvFtANn5du4y1Pr$?OC@bja)Lb@{?l zr+zWjl8gm!8U)Omv3$+}qln&pD4c4B8j1b=q{~{)#h!>Zq6WzV>G=tVj-xot7+{1R zFGZpp<%gPBYEZpQMi*yILx8m-M^TNql2^7oX@*Tt^{>OmU>Ulq zQ0F#q0nTWX{AdS|hkkuOM3me3xQc)k?Kma}ZBG0(J&+vzW1&0q5Jb@d9lwrxsy1tXu*~>YaFiVdC@tRFykO33 zUlbULn<1lDuNX!+TQmu}Eml2iqm}-n?RBg`Pdzpb_XHAs?2?-K|>S3yJD;4gPZ;g?IQbP-%WO?Z{HOehYTmCsoI zrDus^G?8@;@h9)H$}gyx6GgScb{d4VtNc~f0zZ7R$ZGj$#_{P%j;`{8-6a~kdEb|I zIw3b*Fr|np54Ufq+BB&3O?yu&P$=6j9;Zceh+^BN8f`*PZyS+nI4Z8J`}B@l&c&Qi zuKxZP+Xz&WC{TA_yE7fJWmZX@>4PZ)!Hr0RT34V`Pzds~F4GJYau|Pz702AoRf0Tf zV`1$h(GmtZ%6u$9*VbkBecm+gpp9amC+@#9RU1c<>`Dez(g;h+)oIqgi0OPwweIwt z&y|@I91J1~xhZXmdV-XaoD4;Gx3 zea^iPZrzUwb&D1`(Z3EJD{?R%yWm(e%SlvpGpNR(MhY@R^M8>#Wy)yO62R5gN*syP zkqiC25;@!D4gOIq$VIaFEZk3G+vM0Bqw?jzwD$Mu)*udUvE=;VcTf>c`!7c=YlN4h zXEZUeS?-q0FL#cHajFMk^MNk|sv(_UqR6Rz!WvF!PT6GYd3l2yEun0E*5Y=B9tZFj zasv|crwEVuoiDV;efSixHHkGuIbjpZ8_Ke@y`j?lPoteHMNM?P?m%;87?Ve?7|0R4qOr znqy<9ONrk%2uD9pkImJ}O^ey#Tf@Qx1-B>(uK1V4w1IQhPl*BAx<#eS%jQ^*PdHsa zcSBLhpD_zrjd8bZuICzkEUhq-QKGP*M%6q$#m)8Q!_-d2S2E)%Ci3oGvL(#6x90rD z3+dwDHsR%wN}A`ZZ8w!MlFcP-Mzbba%FvthW)V)H_DvWh1eNPYGsJH_)JP-Hmcldo z#~eOe4&$g+}I({jyG*blV?YUeZxJlp0QqL#eYkm2`yaq;9YR zPdY+UCn3EZH^baNxgY0c7N0F{e?Uqu=xxUlWFPxZ!j|6-JA@-|tXQf$0@}<9XaZ$) zt0&p4kWLhX+SY>9ZwX7^7q5a;iBV3NW90r-x3HyNSmm^NbV$l2u=x)p0kiL!Tp-;Le~zmoz5I&S2!e`w&-uw4(8fc*ei?LWrSpm&CbPKl$PSY`7b$#ly&D8% zJ2d4&=iz7r|BN*4Z&vMY`>pxYs9^06PvxZ3G0@U}$fd#or@(KMiRb4uBHZ1j^MHbi zQsl=6`Nc`YWXGEEii2xF=vEy891X({KzdE+kc~ACc7}P-UYJ*YDtz?{bHan>?8RH} z=~XwnIclDgZ%*d2l?#ZF6<^F#gds>^Fw(`RZkz#LrjbixgA-+D#;?~9Q?pY%5)bt% z0d^3IjZojTh{lFT{`C~Gffz_dR z_3^lFj9;nt(5+auVCzo>gCLBoF&PC^Zk0T>DYkDH=cgG95SFPi^CZIJ)}@p4p|~Tb z3k-EUToATXPl!w|7K5Hr;`qCm=UuZ&y!a0`*UK44{V#v0O2q$S?3{u#0k(D>+qP{R zUu@g9ZQIF26Wg|J+jb_lC;8`cpQ^p9&Q(|SeXrFE@B8c+;Zu2@lCqu6g3Q(0OEVvqa9*_TpRG0(`_+`V0~8Ii$6eG(Y;8Hfz$bYPaQ%zrIm@>}%1o&CNo z63F5EnaN33*45H?FbT;^;(+?O;tm;Q0;WDI-_gJfz@H6R4mF2W^h4{F(d>?z#3bPi z!2}E}4TRdfCLm^Tf?h&1;nEut(>}EhW*fH%%gLFOQa07e8j->_!FF&##3_N|G)5uc z-4gi)LU{@E9p!A}Vx5T|tYrv|;QO_Q32C))jeTLjV`@grz?T3%@DL1C%tge#&DAHX z_GbMHS>vKu3ozI=4{i>`Vm$TmgW*%;!s`kycOp5D8BW=cv*49R7nLc}aq zg^5j1Eo{W33*?~)g?vZiX=YJc$hDJKB{7b)_otO*8q=Xlr1bQ)4w^N~OHW0fb3&w^ z=ZtxI!7u=^H#zT79hJ6z2y_wX5-z=DKf)Y~QOeuh@q~&8hy&Yzzd8MZL-db~204%K zI`|R!;F+*Ei_hZt=Rwoszbs7I#`$+YI@4m9r)k`0$=u<2ll&&eZiHwTA_Iu<;Jv!B z;LSwVs?izdO5?7Wgt(nn!|oy?Elec>0A{FDY_?P}JvsijneoBSZ=XS0!KNI!tT;LnPIJWnybxh=WOTEEG!Ahw>tc4fUx%oDHjq3sWNXqt%2BpW zs@(z;s&U!mNu#fPY#qug>neYs#YER>sxivRX`?cbsuDt1>N{0$KdV4NNx6*!NUnu% zL>7y(Z3+?R5rd6McuXdQe~ZyXE^-FwcwvtFKnaF~KU2OiTol9FMK=a3|JwU`51JF~ zrNbH-E(93TSTg=<66ZZgr%5wpc}Pmhs;OxlLll zw9jqw{64rv-(DHP#-gVpoqcM1dGSkX&;JFI<8;^BaawfEEBaxaBR@4GgB}0XC^tjv zdyyQF8u!8jTIV&$V_);n<?#eGOca??Bk>D#*H`;cAIG+>=Z-FQxEm>|SBV@WJL6 zK7|uqBTibl8Q6daeDPgbQnAx%_PBiB6=1#m@FhuG|FEc{uPzk%0{mC|e4w za&T7QD$83|F2lP~V`pwwrqJpABU(u7J@+YuBB`y-mn&=JiBdtH6oI5frBLDa#9J;% zpe^R}hNiCnYL@g*T(w1YBS>kduK-)djK4u33){~LK}56%Oo+3fLW{YWRtiOH5#YD3 zPKKdJN10aSbIP*BXer9q>i&H^2tM)@poQaOun5W<9N5jb*X7mQXBdFVR`ZsZVnxPP zYzR(~-KOZJLBWUWUup>|hW5O?yEtfJPo!X#LH$wQ`<9@O;)fKCR)_#K9R_9;=CfhT zjB8IdPn(TzjaBgH;BoIsWE52LWEmUYB$5ZDX)Zwdg)lUWqUTZiH9*x6KO3_&bqY{t z%}=|^8GQ`QX{?@rrkEA@Ryx-2R{dFXrb!=+IXi#P>dnA)k;|}O4V^jD!u)g8u5dh^ zxpaD*8>g5{FuQDRIjUEUoX1<|a*qEYz+cL$zX-hyo0#8alA@5|(0ri#(bHCwdZpxG zr|7eH!ESkARUf-=TCscJSCR{$u@U$D84Y&we5f)ke zqmCU12Rd&`!R3*6P;4B7KT^#hAHb^79=i}X=6z=1I{r0lAYb>|x`zy|WOyYa_C`{qz^~?3IL7un8G^TUx(Z2nYOSNY0R_<{pq;Ai@?Q5Xcj7IFJcoE>Iu16v!nIyer%f$8f}S z+32;vL<3#=iCOLCU@_SN!c;wodQDxj`jpYTclW4N9Y z;c4L|XU9XKPw21v6XD9wYT!KZ8deteSzfH~1B=UjKl*@4$d;Dx>VgnoGGWde|!MT@QaJHVBzyjrz0Hns3iBig$ZD`EG=+Fg|B-d83#pFHgDX`v*AB9_Q&- z_?hihG0Gara{twCTkG}dnY+yEGs@jGxyY;#l(Wfo_uo8Q%2{et==(N5k3C@}wdi?J z@7LkAn9F%Q-oAenK7*W_(?ROkzYN<5fYl@&`}m@-*TE=#9aOAy2w}>v4WA&7#|_!X z;><-#+xzV>%X#J$%?l%+0)jGmkCNENLb;H89tlRm;*TzA>SC0dE>Rq88bYR!UB{q? z@oxw6Y{1Q7U}Zx-jYq?74hSS)UfrCopp?E~S65a}PG<~-Zj(eetk;QE)?_gE4(3Wq z5BlP<&AzTDA!C~*$l;d_(3QXUjad7$F;h{;E@#XHJxOIxCOqDxh^^FOjR~uAJ6O>R zUgiau2TX21C%f}L-=Xu?m;pW&jMj81nW$?pM_NwQCnSdl0@?vI^e6-Z8dARY*+liB zNA(OWc+5Z{eZ`bpnp1bMq~fN*^iX<1k6o!XBb zx2_qEm=#k}cu#%nhI#;ylQqm#aWgf5BLjz6;W4eyxhCx70(+8hd^oc@5D^)2kiIT_xQ|Yv+CH zfBQmliK{iNLww!Xpf>A}GU#TAQ${iKKl$Bvf|Hu&08Y*j{RQXa>q#Rfk zp()Hz_q!jwOXy+pM$BP-0|GeTz=mzAXy|==JeGmNhZ2sOySAMdvU z?1)q|aC#x%WK)!z}qRKk3;Za5luW<>gmB&rxTi5CT>`7_5;wBD4m zBsxn;%JKnC7fKHm#rCwchf|givlSWKthu6)L4xShyP@oU59cbb<6FGKAB3l!m%r1; zZn9Roh9o6iExDPpJ<>I2#i;Mg+FbECpc+JZ$Nh#_KzT>&P&sNw%0SAs>f_un!W_mF zE6LsjvfwHn)q=>*(!Qv=5N*0`yFP6#e>6vwfHL;|XvT%WafY>(n2BKFId~x-paFF& z@~7tyK^GozFh2zvDPMQH^qwOkvIfiUo&*emXb?%MN zp$-o^!8aUg_v)P<>8?>sFN*HE92Z~jR=Vy^WRNT}?7YB<5~kBL$ZW$^N){_3h^j6; zptVv}9dEH2SuEbB$9gS#Va?G#i;$vSyvF~8e~q5)yJrv3(=MgL2Ppt?eT)-u?dh9T z0=JN5%RGlHl~ZvPagmFq^MDbw?ZrH5o1ce@bxeY2-4N;PiB5=<$m4b0s1m-?t=Y+sAUX*7ko^hHO;s*OqydlQ+t_^m}@{ElfR)1wQZ^<5f$ zZgOo-gV?XA0NPL=Srf%yFQAs9LOMqUpip}Zq2y)*(Kd0=D@8R66M-DtmXBO!f3R@C zjm`xkvcapw6v`!?QZ2?k8q!Fc0;gs5ku`ReVPwOav#I~+qsWLFKV*v;Xa0#FF!3y< zK=2Kz&!zr*&q3rSnkJk9%`zcs;T={rthLZR(qSic-kGXpD_(TNki#JJ4`PmU^Dy`_ z>lX37(0Sm+m;}z^iX8}dfWynB=#yLy9JjexT!_}Jyvsd zeFKZWD{GcZl0!ZvUv25DWwT(WaCz6K!&PjQj< z{g^*ZOH}U7v5;no*g8E6B0Wxm1*fOnZ0j{MvbZ$!vK_Uft|39rhZE=uQ~*y~{FyD%%}Oi)370fp{$*NxwLLmh zRIWr_CFY=MnGY-A9sg{CpXv7s zjH4=w)-`QKyrrOjyrYZ#Rpc7<3r7Ut)C2e#1hx5FH46#; zREgXDRLL(o%yGiul+KG^it3G#Ghft!fwspt9PcgT=YIwhH4BisoW$MFKQoZL0w10J z;B=3-!s>8|)v}$X5x3g(YOSAesr{%v%Whd$&Q{UXK4B;?NcNy&;mD3izs%P17N{Tr z5@eVDHEuAeli8`M%}|$^&KNe=*eHWWzCo^CH3?&DYKH%y@{{xV9SN=L`%mBLj2i~( zUf}voA}rD3NC1P-Eh4ItjB~P92Ax=-0Qu)wc_H%N2>Ha#w-p)7RUDwMp{nYFwVj#b zY|m%V{(PdqsoU*u@>3RO3yK0MUD&HwxKQL^uHz`)fN55SotZCReLwFE8z}-HJknMa zEOnBd!D?EjVwfix*MXYuIPI>Ksb2Ie`NaR64r>)QL9Q5!!)}#zF_-_9?ZY~JaA~2 zH|mS=(=l2pei!ZP07*W+zUurKrL9;@vu&)J*@#TFgaW8dv1&v{My{38q1WuXE(%RJ zs`JoOTgkr#8D#5sz4z!x;Yz#3ng1e=?eyHmpI!(1eAAq!(QuImwFJ^nBJE8nI}7Gw ziTh2G2Xa~+cS0!^On+>%_pr;od$V(9wAHF=D<`X>5Pq$bDt&8^Ya5(41r-)cuzh zOb>+n7?h0X>Ao9n!`ikRn>=F;I?xctmwR__hvV6 zQU0q95)FD16;DdOm$bqNb*fl+TN3*Cemk}qeu{&^E7e@nM`_Wt)2To!0~AKc%gM<| zNvfl%x*Y^t`GT4uyAuZz0psgtYoeZsW*>UZI403%?4AETys8;mH(42&`FM@3E(gcmW8&v!{ zzm<7GFyzX~`QC<-HU(#%b)$Rbn-hNR-mIWAX_bp6%O=O589^bCX^!ghB@c$jKCRw{ zanUDs{1sat9lQ7@D%Hs3s-e5}%6qVcS+zmG*S2lnrLw5J`oUqxROblIGl~61#7wUI ztKT5EnZ(TN`1!X&p*|LBoZ%?;8z*5_MrIBLaFppkR$bYm zmbRaLqfHcW0Z4RlXIx2c;fF6joPgJh=@B>Qm?864Qp^@oozxn+nORdZr=7fJ*n=Ga z?kXT-$g=x1oZm<>Rf68Fp#~>rva*2MU}qO{kl^5X#Q@g!eu@`!f)_FInZ1%~m^|>8 zi(vBaKHUy%uh?N3TU{d`P6-D{_xN3l2B>S0EXR3f>o=i`4$|dF4w!qO2B|>DPy)u) ztjXzji-W-!#|5$W{?lptrj$Rod1iJ|Dx*4F)Og2Pp+2 z{AMt!2*;>oSza8nLGDAMU4rh|=o@v7B}&|}Bce%++Z-H2^KZ{NeYsZB2OUVB_WF)Y zy_#Z*44@>6@N%VwDTop;x+D}0cGj3-BdZYJ$nff4eV2`K>>@YN4qc5?b%|wSw}-cY zNQ9b6H?ro-?NwU%Cl@2U{g!%i7@VWZmBDt}3hwICLH{`(i@l7>o7rjG^V^9`HWHgf zJu1h9h7Ag}l@}N&4b9I@mRFWfQdKRJFytpI^&-!U|KXL*7BIFFqe*m(S|yEs&;E2k zS;p*rT3=-Yg^x{mEwXdd0nfgGaSQ-iu>c_kUdnLf%(ZAe9{gU~#T9?X4*6F0>%)f0 z6Y%h#!-kY_Sk4NJYObW)uF0GV!yHu#9MhZ%+JxDgDNOJ$XFg z0F5J>z-4FbnJr))^psWfb7+{S#&bMwWjsLuQRw}CU1g$_E_1@1)8pd35o3xkHXr<6 zX88@IF_7elql+NX11F(u^4f%L*Wt9V`kOd*O4@W@gzEe6?zAxopb$p@-Ur_ub5#?W zqv7C={2lxP9l(!RuFrRD$ZtMYGD1}d{G8$hVOlnni&HjR;6}m@CL5H)XW5P^A4)b^ z$OM&CV2%|Ud*^n8ej)O8z>?FK%Q=J}BSD#|&2>j2dQh--He6B2xHmPiQh9w9YpX>e zNs2x5rJPjNIe&1Wsevrub6@&iVVf!$fi+cUTR#A=Koa41%IbL~zaK9sAJ6=DYPuwk zkTUQ-+-A;o#}wEDJY1oFs|c+AOIy@$B9x(!g|aT%5Bmqqn8w;Q)>}rJgwp$Q@oFXT zY8Xb<6J6K$D;jrlyJHYIS7hRLs3YA55uCL*RT!?;=mM%h5eL+hsMWBF(yqu76|tH) zF9a23&LCo_(qU*tb87uSeTuD2PWSLter7x=Y*B$Brr<8e31 z?OoW~P16T$N}H4?J@FJ=O=2PMBxEP-gdz;1%dRnsX~|}t3WF@A3#<&wN>6sgsGKEf zi1yDQ-s$A6W<;k_g)5e@ggPa{iu~pF(e?QF0)XJd=#RFodNNIdK1_DO?Q}S}kZNIO z>v|3BnU^7hH!b+N>-4(WrcPF)IevD=wX}ha5GQj)!aI|o?RRF2es%KDu?;NI8-}WP zw_gLu6CPMD%UF{A<25CHW-u$x@dOXkOv$Z2SlckcBFh&QO&=G3Di&lG7(i82GJQcs zHSu%t1dIhnBsHkDoa~5>E>lyGE745#6gHZKY0;Fx_{nq%LFD3;Fkt7TBRQsTzFK>( zN9#QPj}g^MpT#Pd(MpHF-fl;m4RHg)RgW(pZju!(EA~NhS99~w{EIHBgo+vy^nrd}kMsa2cS4h!yX*EPc(M|B|J?qC0 zjON`Li<_GdYeIV;QAHG>TjX!KPs)CAFIRP6FU>^BU>1e2D0X2lNGE)!e-!4U z=V-vLM;Gj456Q~%Ge52qFtjwyln8)h-xUbU77GOTV(n8P#_n1P6Wc5#hIHr5YpbeZ zNix!_{7ot)gIX5+4Di?7D1Wf5Tp@4^TP3HKv5KM5yKO)3Zh`fMHWakJ$~~had!}=X zgNBWa&i;~qA^#e>6Vqq#!|Y zBc>u4^I?q~CNqH_TZ_>|5R)B!CTjuo<7O)+>CUtsrc#ulwo55Ei}`BQB75G(c|hU` zfY)Wih4#3!#*Bz*l`N*E>^vXR(T02&XGgiciV1lIde$w4~BRRip>fVEkb% zDFwNjUI#sHXR!_paaq}*O9H3Ywvf5cP&iuD7}6fdHf&2|sWqfHueOiHw|$TAv{d&N zFfDv34zExMm0O^Bl&x2Z6>?jRXRB2A%R?MBA$TE=&CGzUII^Ni8|7ep{A+pVMq7_H zuTHmX>m|d~wm9Z)#Ed=(#+W{>YqcGQZB&| z7L%1zVqg;O+iC?n;yQC_P-bj19%ffuwRH&|0tDDmhh)J-yChQx?)d&F*&1A9 zI!Ce%;an&MP61MpqX~$?{tb@(dd;1!7lKcy$SgmtP2YBu0t9){#y{X7zun9YA;P@O zyK0!~^6RO_P&-{r^vcET=RiC0;GfN-=c3zf)NSCVi@ zv*2TG9-taD&y050`Nf4Oh)qOlbC>cPs0@^H`L?kr`CAt%k zQ8#G-WuOahkO4bpCD&XlNorJM$)*az5>e_FbZP!$&dQkijXG`lzVnw4xH0HM!sJAJ z(Aj>?3++%r8#6ztFI>jkcFH6njD!uRp-Y7y=PN4q=ko zk&iDhLu1Q$L1in5fY}i1UbGYVo5emc<5>KCYgt(RAiOY640rRKwkbA>TzfH+H?Uz_ zcO6EnIwx@f1TsGloL9?>9(_w;=|d~-i8UxdAhueYgU(o?OCeH3OBTPB`yFP#OL9Ti?=^c+|)9)~(IL zU|+9z(Q13DZ=ajtW@Qb_8URfNL5{QN5x8!_@9}IsQW$2GDYo6g44d9K5=!03R;aI; zp=45xvSyCq0J=3k6LnCP9=#{3vj3;s~LJ=BRk{mm7$Wac!}qZ3ECDCK8?HAQ+?8=7~-+;$aFnLlBuY= z9rY=jd8`%lXVAMGZ7&C)9@Oq52)g!zJ?P$7lH|;6WT775^7MWxSg~xcCjZOt6u*5H z$B)EU=@Y9ZFUNRO-SD~5!RKXRZw*bcTNu<=30Akr84z49VEIf1tSnjHDeL)uZ3~l) z7h=Mo&9!Qm>4Fa=B;j5uAedy8ol@}_bnHyodZbXor0$#@drOrT54mXzlzwWe+M`!4 zStlMmg}Bh814$=lc3--%oF90lFJ$9RgX#O#yy_UqJxdCWoQzOKup@hE=(g5H9pzi8 z+E@!#d#h%i86@k!fL;oAbh5%ZRGMFvmoogd4{Q!}e>-seb~^wmZ6Aft(2S6~Pks?V z=k|&Ydn&TKjAR_*RvL{a?ROAy{C{y3Y!2n@@b-8`b2mw56N{(Fui~Jqj%_T--4Rud z<}Y9Z*>~OrPdYkD;=U^#>NpK=4Ggbl`se&srVCBVbmD;@x;mAXxEd`mXSA4nRu`21C?>4GG>zwlseDhj=lzqE@Z^?GenrC1;K1s=lah4jb z@N;{H+{3n|G+mQm>q8h{ld#W5(qVAcNKNn9sW{;?B>+6a7QX(iL2dhTnNVbnXT=we z{7XKjo%3{WYVikZoxhWqhvz(qdSF054@l}`KNTB_di$k+?RUL1UoW{lFQ%VbO?n~_ zw4AZ=;mj6y-rjxuT^OlJqeY46?0VXCt~~fm@m+=LD;=;*PK`1Mv?2*s5t&W5?5G@+ zujjm9cDRns@l&(R-(85jf;NjeKE785*@yFOk!7JHLl#ygdTDSevLm=X_;M+;XQbstmPL$|GGMDbQqb#Z-H+Qs5E{jNjmbIL zGp5v$%=Nvr(Fgh7SGRRhoFTt?(|<@*7%tkf+;W@`Begc<_9U-#@OK zayw5OITRtSM`lHp!zfG4$sDov$S})1w@XUc&(2opOU)BX)09xD(jub6X?r`||A=F6 zU9l1I$qd>%^b=LyMx*Pbmoz-1%{C+NUsv8ToSk7^5kdM^O%!+sZ503VREc|#N2F1N z-M~bbWS6$i};a~tmKY8TC% z4m}L2d&I<+k%2Ainx~3@`5tDUMK%AEZX_8MOjqa&>te!c*%2T|c_gjOZ;QHXyQTrc zxe*EiX&yF>YOs3iu)9>>RvBd0uvL>EIbAdyidB5M*iTWvTKt@tv(T&S=zw~?Yp|W6 zBXZjZ*NZtCI~S=i$zpynf;YP5wQR`C&XtE#o&KlsfxQ@;k?B18iK>k4R74)Zv?$Np z+kh;hT8cqL@hFzi`uz5%zjz8GP~zKvXYdyK^&{W4o$%&TvT#dwALfl)MP_)sustg3 ze!H0oNgN>z81`m6=nZ@QGh$6=&-z~kCv?lcz1RK? zh}E;wI4+-XTIQk12Tg&_-s$@R5v&w*Ut*#Tgz(UfyEOj$fc^p6c`fyviX_Ewdu*o~HlB zRkzWXv+eFSBlk8sGBQ^1@yQLhXpMv?P(?_`=$4VI z9(u0r_oIgl;+XB3V-^RR0}yS0cj#MOF^0lAZhn(atasCPN8pa z7-*YP^6yy25~AHqQ+cA3VV%cS##nt7r^vWcUq}ca#wJEJmYTE6-gmUbTaF&)gM(4o z_W0J6+TCe~?|jatDrS|ERbce4D*8tQ%gIn1-+~00>RNY{DhC1~Z9cApBA7kaR78?1 z`#sbnK{jhIl!pPw+Yf)!^@ zl4m-{5{Y;mfFYwzx=+6j7Dir<*0N^bN(f8llG zq_-6pNGqaOEr8^Nk3JLQ5~^ zU>X9&L}MM`TNq44YS9)7Q3+EsW2m*HnlaJfz`TO!n!wnYFtcR79|ZIF^>Sv9t`y7f zk3|~r-UFj3O43JnxC>De>^_!{>!WC2BQ$G{ahw|g?9my(yD*4}Vvrw;1qvDp+7IBZ^4UvK^_5h>UyAv(=crDCX zwd0ww6d}xgv1+PnhYYUC%LAVW(~;gYUh}}a6Yu-ir4PbwJuvP42XHD7iKN=4n!Nt+ zGMDXpd_KhVx5F%uEhuaJU@ zH_@aTKYXyY&m-R;A$?wG5C(tgyMIoVuza6^;lzb7Y-GNnbm|67^m8J@Im8I@5TNz|k@-;1ugKe(zPfSfZyW+}i)rpZw%< zqZpci5y=W}H@Nw#Jy-MLPV-zpyd(9`-_v9}3xTV42*Uu!)Qi*N$gyiixROz&CM23{ z*o|e?cB;#P(}p$9b@&)PV0um*g`u+TyJp@yeVnVI;{zG>I-OR_t?v(})?6xTIyy#I zPo3@S+!sCxeJ`V^veu6L6CwO}bOQYilD*Ld6m20G0~*PA>3G>#j50ImMb;DYLgNO~ ztM#-Zr!sbo)V|`TqJrW=cL_P%bp--hZb6Id(es&!t7=j9opP2w%L11M{i-z(`b7H} z1qDQCtxjxNN(p~L@9Q0t+8m#CaWs`L;9?!j)WZ~<13I<$Y2KL)Ief0RYA*rz#vQ7@X71hFL^r>078XWYPflv0L- z30JR*YLvx&Z1aKHI3G@xaFY$u*Fq;-VQX97OL;9A_rCmfl~9>dHjS+XF<-_Zhr0LUUz__NAf2^$yBl&kt#Q> zPaq%08G>S`^_?fSRS!chFG(pqMlfTqRXw4F?^xdehk$2DeGbC_S%Wx!k z!yG;a{f4sr<%nIRkBjj-tgv|+jqn`9`dgrBj!Tw=LFJiM+;XrF+@yNvQh^jqsw3U z+k^O8_UD^Xp!>Uz{Nv97egJzv{WT5bvLtUNK?P)ciK|e)_#~=n>8SO2bf69n&Q7n9 z6A z!pZ#)t03C_F_$30-6`|$Z0NBril65F&dWBrI;K@dvAMpb_sloKE8nmpqpKrv1OoIdbB!OZ74A7@%}NG5HZVf*CRm)v(E zZ!`xCQ}M@&`XuTUS!-^cm$ql?u-^?6k#3nd+ydG8qf)`;10Ih1%e;2+4}Trj2!QqX zQ#GJ|wSXYAj2$nUwpC3ljc}k?lxi^IN1>#lHmR6Mw9T&GUTPA_L!%dpuJOg{7bGj1O;2f9z=K2vgV`a(5d*0o^Ad z_hv))0^#H0Ct)<^0)exS3U9ef3?cYb?dEd=kxJ{PUAf9di_XYv^#yf7Bo2`nzRuNp z8!}?8Yo_O%o)nvw$ylDXPcuqAtEQ$cCIZ}F$j4w(hC;>^cglN7XX8JV`Wvj^c!)@U zL^zOT{N5}-H%VXmDG8d(CEbi$z!GtEfrl}|n0HdlCD)Swyg62m;Z*WMawzTmR{hfZ z5+2(ioWiq~!{(KtNGF1u@~DK%&^u4weN6h5z;(al5@a?W^}ZIM&y;~U_LH*FQU{7j zh&Ub=*ny?~j`;j#VE3X@_Zj{r8)#kzwY201W{Wd121 zyzkHJANwv2)TQ}-Kh$@rih6*{$zs@|3^sSM@SIRK(DBGf* z0j~=1cXJZ)HKVD4k2AnE@0qY=Y2j#DZR5a!qX>{w%cwH8aYC0&A6IJ-enPQo{+Gtd z#545FqiEx(<6JD^a6~L{2Se0Fe;R)R{)p54RJ`WV{;G9`+9~itpz*nM+ek zluMo8*Lr`J{c-)>8E^#U{^?wziKJD6qqQJCt4e1d(S*yJoM zi!W`;zK0~NGXu^f<;d9h8M3)ope-2q^0yG@yj;#=8Gj-ywqJNzOPvj8Z`*ENb#UJk ze*Gze#3LY+dMo4w|0$w_vG`tqh6`6<4OyQS6(?fgE2@ z-9w_0ES1O|4w@t;iklOe^isQ>ehhm9B{cAHr{#V*RmKc`aoAETm_fWYC&gOEDIv`sBZGJ3 z+XCX1e06<>eVZ_nL@+Xn3c2{+`RYa+xadB?!+VbF_b1{6DokBB?kzP}s^9mI89&h> z1SDDgR6t}$byEa9=l`z8_|EUw%zlpZA&<^RsSNt-8cuFOriNyKvv_Ig%7`)V&p-C; z;3eNaXZQ9T{+%t?uX)W7{VNFu))lw0O!SB(({%~H!DRUc{DqA1a=bvIfcbd}iCD(|{iaX~OjOkbaQ;-to0ehv zybXGokN4^DN_eM?8^1=Zt#Ji8RUGu2{QUY1oWL0{WqZp`+qP* z{-@A`iG_>lzjH+XAEAf0g{SJu<_m(<%Pw0`tO!IQt12+Ha!Mf^@EgE^gE!dnl$J$5O>TI$b= zSC_rHO@z=TqQQXyqr27>+HMQRO#Un{mWdmV@$bwdb=OPwve7-f>;WbTrs;~)7B78< zm-`~aeW-J$Us%Wc{W%>h96Su%)MQ56rR)2Np}(DIjf;ZE{1Shmp`hZSd=cuCK@I-= z#kxgr5THGBuZ{j8GIBf2EN+>&e7gh(##IL4EQPR>hIL2W!0R--U1dtRvqRn@^$_(E z_whUXSec>%SWz^(4AdK0~c^b{3_K?WU>a^Wl>s>O)y5E1+W8G-lZvSU~ z46voZKt_$CAMP3^Gp4;KV9vNB+9hg5R*AeGRyBrXj=Lk7O)n zSl$?&Ip&UB;NQ$g9knTneHhP}o;h7pVw!Xn)d3YZ(jzi%*w+}fA<&T$10_A`WLWah z@1GToa5*%_JOPVQl*AzfbDAS*Hgxn`Vt!PUDA$4z9K|TINaVjxp)nS+ z92_MBGV`$VAv#GK)&bM7QTY&)G2$%vcVzTiS%~MnPRJ&{1^2Go=8@}=SIeWAGqje} zcMOk2*-{=HfuJT&yA_m$z;#d<~&Wx}ju&e7)a2P|dOt?T zZQZ3#Q)4^0gaQXO(BPf#{Oa5 zH2)GHw^ioi3#yj*J zF>Z+jZ%jv-Pg^E>23KO*(8*S8)A%S$_%{E-kpa$To~ooeqy@&0*?%pQUsf(`D-YHj z8w=h@R$<7z*^OrBg=vRD=~Nsnsm6u`QlJiJSsu8x;mBkgETWkFNc)3hsym8xw|(IZ zR*UsB<;evWK1f_!?IUrOwIN`JcB$=FLX#OWt5oe{CHVzia&fLeNi2y=;(8jzYs?xynCNHycR;xh-JbXXbM5 z*ZS-F5nT-mPK$#1Nwfp6heQJI7;xM~PmB^L?BMct|K~&0t(!s!715H=vtRsR|J+#C z5&hU@UpnRFD;Lo>wuoern7NzEGBC!-RY|F8;9Z{a;(?fvDn=E;tXPm_{?d|U`{t`7 zj@ihL_o#gZ=FwI=2-X)hJ_vIk*ODsxVA8uij~Nza&Vh9BpX&K9*#2F1kz;)aTNhqR z2rq2{W3}W`*~Jnds6bM~GGD80HdUYJZA0@PkiDqN#y!o`+y^-GnTc5``JChCHOba8 zjJKRdv8Dp;nM~a~5Z}U99k>6h?wUi{roWJp5>g(@V@M{l0 z#<*2>>5Q}Rf*GN?o8_qcDS%T#(_wEypG-UV;?rL^(h+kVoNtn^I!L0&zF2q2iU7Rh zY=027f>M~*i+7fRMlZOI1thgzu(%(1XW_K;VsbSb!x0+~o|SQxJV-QAbWD*5aE*(Z ziiJu72TV+qPgFtNptMzBTV{Y(`L}guKmaY?rP}b&C4EwDXs7&n-Dj4{C>t@9L$qHJ zCPi~P?Mo#uuifq8v6eNvkfy1Dq3fT7KoxjXyhfS3$j#pr{2aoYxu!Hww^or_xQ9Eb zi^M}>-ymL>B$eh~hU_6CXSVI**9o>%r0t7AhOAzwi=LLoRqq@}(?XS2Bz^-NJi@c2 z7W`_U$HiIRhUXh)jsLeUA$B+9ukxAl^6!g4tK=a*ZxbHmi}V zsntdw@UI~}O+!?YUMQX>^)`3thq>%Ww*3T^Iy}$9y&m9;A5nj*78^~iQOIWoE_HYD z^okgPi5nIk2*VuPJO~Ncb?y3JSb-O+PXq~*o*hmF*itr+U2)nW7yR&en-zJo`1~7blGo>|OuwK|dy2{pl&4Tr zQB5hGgslzW?h6EK$Ak@fQiO|B%}y?tbj%9Q3SF`8gGE<_?+4BsB%-UQAFp#eBv%U;*pbqo)4JkY-8xirOf4h2bjd1m40+UX7v zYoZ`Xj58~d3{eHI*ty}oUFSSDQe z2c~pP^WN{;GF!78@$^2Bd?!BcFoHTBz6!Y6C6NF)5E6Xc<-J4zyMVC=@&MIQ#??^A zwA{AY7JiFYR5R>|~qq>T`Rq1AS)VmJC_GN6lC_o+?HtE56+aU$M zm)9&*7RTeMj{dZYTBNSBFZ_Az^AA;B9uK4^G+|y@zSd2ft^<3Czf5lS8+wRn6gSYQT#LYIXnaEqeES4Mx2ztMYEgoc!aUR`h zc~VYwlHu|MDdVnr(~~(xnPUTM22p~@AZ%PJHHjf8MiF6?vjr(8*&fwhpiO&mE5U%Z z3Qv{G>PbvDecIpv<2+(E;M(+#jabk`KKp^COZ&dic-Ow&DgpL3nY{;8nEP9Q0dH-! z{QjdE@y3%Iu_q9FxZzHB3=K*6vS`Za3TW;W55zhAe~pBMwHsW4kO{Wi^^}#JnUE8E zt`RO{;iUsMDDWb8DrM#xwr%GhRws1vqZ0!tKeJE~>h35lJ)R+1x%K(Gcii#dTOnj$ zv6rK=K`1}&6aNhEbeuI04qqWAM`m%#jQq&kS0D7Am_;CcIL4mh5s@X;)zN@K5l)Hw|)vwW_T zq1g@pW_?oi<0((K$Tx8=ueCS$c-Q@v>b}oQS8S{@)zCxIZYd_}C7cB5%E3n)vln&G z=Yhd-gWxKuW-%0vi(-4N<*Wu+$tq(SGZ!Z{!;k*9hkjYpq|1e-UFXf{{C3)3u{7rnhMh)fq;+>R*=`PVq|}0Tr-{OIczmN3t#DF49*Rn~X{+ zcaqhBxddkqzE>xAA9p8pjS4yx^uo0z@(gAfOj+m;Q0^#UDRzaMCP6tzo3+}o^yylY z)$z`px9z(29%SWGxZZ${YUy3qM&f8}R*Wf=9a|F7`#2YP4VpMKlWC-dSgOzd129YE zB^h7PSsW@g8}P7j;a@%ZWKbtP=Fr+lu9RTH`Ju;#LSr(0qS)5Y>8oE>xugr>aOSZE zqzMqB3GhK0vSjBzP+W!~vro#aJH{+%`9{?XDb){--dNU}%YVW6{^_yu z>J2-huDx{UC?wrKDg5xV=u+-X^xLM4tdgOT#&L3AN{C@|>scfMgGymh3_6Is1Lkis z0LrVF0O=LbB`Mc(^V&hb>XWIJtwdcOAi|ioVCTx?l(HQ)*q%iwp_1)*{*Z7E1+If> z2BVYMHi@%_i;)q7SM=1MzYGAyI03`N2{N^RmTJNoxA+f>g_vMr@xb`7(9lp(;b1wP zZ{$6WtYLHymk{BZv5@{Wq`d83ls)95-#G*Sgb^YJDI#EznDM@n-9FkL+CJJZj8md( zIGR0V&Uo=+gPWbWc&K!`mc5q#uap|pMCR~eUx0_dQOY8*%J^L9Qkxg1`jOv<7q@)g;n<_i2o6(zTFspLoE4;S}nLKKSdFKIoy@o&9`)e+r((t^A_*a&-Vkr%5@!|n1 zYtnB**THK275)5vEv}T8xW@tp7a)0b^eJ7586ptJqHk6)=iH}!4eOEO$M?d=b zv%U*P+6!%MGu9Rz6N{ksz|toCmGTqlX)~i<;R^dgCbxMC=Srd$AeDc~u#O|BTnErz zH&7YgZ-f!mEH!Jgd7j^df#=N^KbSZdCZMsy;jzv52yTaae!>7WM?mFJcOEoVU4wFJIacxtLq6qkaKy-1_c@v z$R2ik*Ur1`=A|7gUU9ONsDN-ex|L(6n>r&)oPLEPf|uO)@NwQPRSuS^@y*c5(@QZ9 zY)^+tD=uz9(tnu=IU_&ty!Ys(pUbzyA{Bv^mHOdZa#ZCL+<6gA&{#iwp3n|>bN0Ft9vdb?wLst4_#Y3@)VTT>Xe&@FgOv_AgRBLeyo7 zM0D6OI+Ss+_pwi42l{rtjy~Q+rNSL?zKjhS%{OL?rtG=Hnp=N+rZwvXBi&MLrGz-W zf$jDjhX;ORF&)j-6r}vUfYFpK8k5ogmACJ&(aRTCuTKYoV8>536=LMcS2l~~PI&Vq zdHt5~l6B+rjige&f+#muZ0vpehE@OeA@)KPt=&Ut*iX?01@ZU`-d4Nk_BxW#F(}qy zS{Ka$v)j=+y}vz1JgC=_jVOne$`#7d14;5F5zEBCydZfR^w%}9`~>E!`2b!QY~=J> zd>rqxLfg`qg-A;>5O6)kaQcgFVttUaugvv^B_>IMj(d^vnHbt^XFp!n(uBg!%P>UQ z1m^q%ZgSr1TK^r_bO(X85dyijpp3JyJbLhD$M~=UBe&ZG#^U>|FRW@WLdOOj0a@*D z$ewEI6Kat7Ic;6)@6AAF1#qvcAUh`Sy2v?*x&|E>9bo?0{~~9HQ1aruPPv~u=UL>1 z*<;ksn(*ZniUX_s{(8SdzzzI$M2Pu{;xTN04CI4j0kcypPd2k%8Q3ceDQo}pxx#!4 zy_qt*N546d!f?s!FZP^NJ>}Z8b)R1?o3>vh8*iU9hZ+yR6%nCM~^)Yo?k z_^!T>!Kc?YXl6anpOTaqaM8&3JDU7Ngqtt}mk=^K25b%l0;EN>sW2;~r=|n?B=l~> z;1XoR(hSq+hBOvXNl7s)ZWjC3t0;aY0X?3gsiL}wrP_ie{UjcL?H@~fQ4^ust>l!^ zp;I>cjWd2etR(2-;-P6HZ2d1emYRG>Kts(sQ>jFS${vW63={39-rg>paQ%i>1guCf z0o02+$=yAW>Wi#gtRTKBIm25MC>*>XB%#Hj*?|q)*vgq8i_e|J3?H2PR2#Zoxod zC1?+J-As@%hBFXtK4;fNIP?bzHfPS1aVj~8d;HV$=1A;n5vlk9{Qi=xtR3pdWskk^ zOd|2)&WmM?RLS5A{!)sH$P+a3bh_sCE?Spx=?FZY4!y~kl;H`ZKU%`~b-Pe>P<8I; zYT)MWzx0lPy%4SzeYTf|)R%~sp_C~!%8_N9BLYejV$?USAzOwd_b|S4H0mj?x ztp9L62>uIQsTk7NS+$NUQ3s!i~xar?b{?N$S{i%^SM^97@sn@GF zH5X`Xz!4{SW~|Gy9tiq?Tb7-?g_d6I_c8!n32gF3611w>fYE(_6$b3o?6PFTaBFCMIlYC2vvgANbl-PNLIfL12>OzpiaINm^Z`Uf3;Ue@WogPxX4*! zpCl;Ld1y7!8ezI6xIqJ<+5{?ut0w!n$CX1GyvI_Jjl#mnQ6JHdZQo>(-clD%As~7?g1j2*uq+P^-q^jZGR`!GI2sbd(PRvjETXqbOyfu;D^B`(Cg!0*~}UDhK`;U zPoXG}(~dLzq6|J)dhAS#zjqD~)NO?|IablD2H#Sz@#EeaaefECQN(tQjL1Jwvqdh2 zbM;?}XKs!RD4lcl)o-E8M|u3=A(uu;3gzPR9H0Dtu@Lhlm~Y2sjL(6J`GQq`oT@JXbSiSYy9$$mz9du@p>+n{)sL%SmkIk}dUqRph@MZf} zARK~OGh3bA^faNdl5zL*Ua)dq5Ia$q64dWc5XN?Yb<|gv zQAY9bf9sY1Z~+RLt@eSC^oZjC3!m6~FBrc(9Ta~75NPv%m(rF|*2^~2)=^dtEOtR# z17p4s^fFWk6AYpob<<|4-<-7}cOj#+1)FM_z{yxz7cY{(1Vk*;!*$9c1Na7#ciEzx z;ATeo%@(Yo_42mEm6Mbe<~^_;;f5s52<~7GaRrm)2y;1iQI?>v0(u9Pd{J#X;n?^4 zoF%nQf04<=u-6G#CdUWlZXEI7z8Eu~14 zp;89VNfV@t(!vji8#EovV|LP>{^f~zMobLo5Bxzhg}p0`+On`7vmWcvoT+3^MG5fk zB+p~b%A=YVJC)6-Q=Le~7{b7H7ys9qVp?MJV+yeKIR_9gL~_lPniS|H;IE4Px>jwh z-d0gfNX4G}+CiN%AXTu?oli&kwYvy(41lg$tMw!^Frz4Oczib28<6XzcN75=As}`X z0xJLQG7bgxyLB1P14^V#`gPFT=`$ zcL`&^J8ifv1_}>2b7;2iZw)>!y6nOW-Zk)GgfS+D1 zAXMMG+jXJ55xxi5wrS%l=Pm04qFc}akUZ$2Zd7r z6zA}@`2(YlG1`x>V9)cl6-4&AZn8E{{6VIY|2KRlCBY{lBavz?yN<2WwA=zV!*UD!U88kq@Y>Q#|>Gb9ZBTaU(+mXKK2%R{=j5yeq#GIcOY*1xYUO=s@InF#f8{zTrpN z)#&B78%`k49A>Yb$$G;(HyGuf(VdF2hCweStRzU%3cwk31EzZNQDR!HvZxu5(5NKY zYc!El@gTI@ymp*`EfPjH1QoUYx2tfg+f6j=R*>a5)OU`StZ(UWaq+RKfT0TlB@iDy zJ1Z#0d<@k56Y6P|=}@I?{=_8N~B6IKq{K810* z;*nBwz2@Gawfib&PRTFC4DS{0xUI1#xbRZ;5(Au^tH}@04qp9Mbc<@gl$!`E$QNnY z?Ffk3`z*LUiT4ds_G;E5c6Z#>T2FyHalYo>9T;C@yarZdxut4MBNp;DeV4p0V`dE6 zHfFV+BT9Q-GRf5!ifSW=CLY;N)i!juf7?izMs1SGREr_OSY^hjE2N;7OGpPzUHq1Y z7r_|Rxm`%$-=LII>y%t_eo1O!NLe;G?ArB>LbYVb|~yqo?eVGuPCcOHXZy-cfo z>s#YMXMc-t0aJ)bfdLy!p^k&C#Zzx3lI>N&qVhh=c?^Hwh>%@OghVdy}2Ot$m}`P!O}` zG)uYV=UiJ(*i39pWI`x# zM1oF|lF+^Ktukxudd4ujx@YuFlOa z$44451u=C3$dSucFfa_DoyKt%gq)h%Z(@blxz{e7kI^Ht>h=qVy%ZPt+EBel*5yg} zIN5z(EZBfy*J{A*gLNbO#gP3a?#=d$R2bz_Orx!d%Ufdf(O;wO07zxT9^{@*i@!}H zZ@Md82CbpllKzO(+n^ZF96oLMd*KpwbyJ7S#Vh8^#VsYeQ6PNJ+wABTYPFS}sYA6* z@dizHmJNY4-q*{y4Onu%v$?$-3_PSU)rtq;c}pcQCP$Ic_rT!5_rPLrId4JZghq>g z=r0OB??{KH`$S?L(7dJ^Cs2!%s9aKu!5VdGoRsn6~g zmS+pldJS4L*f$aSeh81=iT(+eUu;q?v%8)W_yd<#k4yXcgF6&ygloA zQ2B2?Ij7N6n)8(vNIyd@101Yi@LN}o-ISMcK1jT=LogYYq6Tx7ZVlRnaK<^C&LzH$w# zL^|x(uDIC7yL`iyQwTo!gq#TRhcy~#?fKne9;Rv#t1k7qmS3j` zew@(LgsKgk!e7l}f=6ITfZc>2Fg{%^Vsw^M=2Qq=2s9L=4D!9K!dFAs z6sZ)&4Bah%%YppTfVo7M)KHMRwBdWm!6B*#)!uFWSm?IrJh!KGGO$9`gxN`sJb^md zLCsN3ZqU$x*lXm#7z5M(_i!;;`m}w`|H-5MkLVyff;TU+;qkPmwyXOLqSh=cN$Myn zk(&q3BM)LbY48vUL)h1Iv`x94rhv_O`BsKY%4S5i^V#{w?;Vmd_*FiUghl ziYL^BdE5$t-I#R}OzLk$qGyR;_O2hpG_w}c@eMYa za@bKOUk@uBTzbx585sCs&cXd;&f)z}xO!d(herGM;E_1Qg>&TtgHY=*YcP?SiFvVI z66kw>jLb2a3RH~!lhoPheGJ+pewZSN{O56B46+E0&Q+Id*_KHI#+}QuBJDXqZ&I!P z#O`x{3(yCE@#h`iC)wrh4*&NGmwsM*|Gv+)3Hq}*Y=(wB3wBHId3XwvDMq-X@-rrI8#{v(EpVLF7rA|Ik3MPVDk&w zV}R3G7Ajc@=zNE%QLLz;3Wj?HX?URU&;b;{Su9|5=s&Qq!2(mKUd$+usBH=rgK<^H zcBsi_ z%JNyC7MB~BiZv4EldQG#xDAc_v!TImwcg)xTnVMu^%yF7jjbRN0nB6eb^95~Z z643c?{bm?g@gWn(0u1H%Y8(^5HU)JxrF8hQ0l~BbvHszqr+@8OGoe8xU=@NuDgwOu z(R#K$3GBdpf6`As_kHW$dw*^-Z&On<>IV4Jw&52o1*2gFo;6kY8cyxQZ!ar^Mby_FQSvMF@mPO87-r%tCZ&SrAq5U0TqwjtfW zKASE;Nc1_#7B(bTMhjT)5S~dU1hi3MbZ>TR6sx|2e5f@g8t9OWnJ=&}OF@-as4-Fo zi0pD?(Vz}2zm@%=@2|C&r>k8}S$yf?r=x_44Fz*QCiDPan*93Wsr9Vb!fIbPDHEhn z(QBF=NYQJ!sw3N}!lY&Q-n^J@9yN9{3h{nU?(AtUB_PR|P>*18dJA;a*?Y6w*c*Uj z96}4{Gb&%49esN|CtSO^`O?RA0G*kRwt+V05&nwGQvU$YI_ZQ@+5uzD)V`detJ=dfFBI;U@qU!A={0s1Dce zHG6k?2h;r0)R7-w`F9XX&us|X=s%dir!tS+h)xuo$(+e31&Ztw#GiHe*f3c7`%ug# z6+UH&h5Mt$r>&ivtxnjN-QPbZ6@oe;Ql;_0oa56e7e>%T&x+GApyxLjENI}iXmHy; zr(eF`ou8hX9a6fZvZ#?6pe{21xPG_%n%~AAQ!QXI-osw7_uC)*ZsN11{!T|;PNLBl z{CN5lf9<$W=O2G#J37AiTwx%pB5l+bO&#@&Rsy!3wAF2OX{GOvCsdR&P$UG!3f9w; z)68YYOGRb=(a4J-dsNS*4$+%7DIW4Zmep0&sj5S|F=LR)Hh{E>2f=(bOVD2#nEBvbO40 zqE)B*yq;wuI~m1Ba*p0n>D&1c+ave;Ej>i$>5N8n7FZ?>0AWBrJ|HWQQNng!g%4 zP%SPfc;TdM*?KQB*!B*5x~$SYQirENcGN?9U?=4FQAUgrI+LODb9$Uf*?oR{yg^ibSp(BGrZwX$ z%`4a|j#@12!0s@ak#Ga)f0Ss9R+RrJ(RkXSvLm5JOZABB&ex#VF4v;gU~EWP5ix_N z21N}U>loH>Y-m|gKag4BywSXYs75%|@U2ishH-{I4M^4atN2$DDTZ+daYiU&Mm_a_ z>pttq8g|!$*M!z|P!wYNVNj*7=W9hZ_W?p)bA-HGk)LQE_e|fwhKvu)6{LQbE}y*( zZymSz?tVUZJrrt(zwf@25D!fZS)%zP!bG+}{Reu{rZ>-k?-*#4OP>eV!Gq9;BPOiQ?BdmZy zD-aS1I^-xJ&6uC`9ARgUnmQ#F5(JV?u4F2N0Sm@}WK=I^c{LEPLq4a+5XcM#RiBV5 zRFE?V5yY7&BO4cvn*U1t2AB!`sA%6JqQRU|cMvL~mc?aiM$we7VBGUXoTeWiu>F${ zvi#Kys z!Et9_jg?KWV&82AZmkpNN-!tL0_)}>?Q;2m=7R6CdqMk@z8?BbrcAQf|D$9kBAY<( z57XVI*{`-uR}%r#DOrdXhSLcOBiV^)+noVo)(n{lG>Io|dz&t0TsoPm=ODLBU>mq_ zWwV*NTS;2G8LE4y(VuJ~0vqGVg9|;FCMXLslhCQ1iv))AiF(~@%KTN-`z2t@i~LL^ zZ_}4z0;xk>S)%fZHV4=a7WJ1pTC~mHrZ?H-&RAE0zZe4>AXJEIIm{6424;%Bcs5Ca zN-)wbQkkq^It7P{x-(tMmZk6!`jEGG(|9v$f6KC?xTfjdPJR4jZS&p2E}#kJFGr3L z%7%t|c4M+=V*8vie>Lv=+#76!DSwlYyEVSdw@&@^`CV))L?RvuId>A2$}jfqAUF`h zJ3`Lhp~KN;wCHMdh)StXRSA$1RITFYA|9tfUyg7PbrSdNJ_y@oFJ+ZG} zomt%bhBjH}6n@fT0h!{hv6OtbIQV<}tgIjdS^>OY2nTUe? zsGQNj(gc<^{yfOz&iUmh=sq3MOQwHEJnbJaq|WpMetF&%fw2iXX9Df|FZAJda2r^> zLuQ0XC_In`wn9;JZa7Qmg8=$Mv?IyAX_)fJDg{VNSjAQwQyWmCbgd~UU6+OhdY?E9 zo}|sLECU0_6*~&wtj^o@c06#y))J8M%Pu||MsSw2)v_&ik@_$W&$R_ox2aTPQj1%jL!pE^ zU>Ax5&L~49A4zhvm@=o zQR5>1WeFcSdslr*WpCz?Sd3^iBg!Ygdo2;y?{-GXy(G%RXl12sm^2xoE87sEar>5!`^~bF;z9`gZTr)7B3q zG|whqj)nITY`3wSKc7DS-C)pKz9Df;(AQ0YEORYnAerO9>b!P#+$I(G3fVS7X`+oi z%{vJ)^XL)5bX00pF_~PIRNVs&knu(tc!7k|)c2F@{)FcpbF0whTp~QwMukHpWtEC% zv1%!5nW9RcoI;05Wk>@O+YVuS`|~v0Ikf8@9Sm-NbhYDk*3n}-vMAZdgo%uHWW-PT zwu|8Cy_5LhJ!u&^363d-Pn~6_M&KfB$T0Zgw+xU zP92oniX3`1sS-k&C_k{k_H$KZ>fT`Ew{viQ2@E2w)sgC|Qi_0UmF|_7iyoJ}Ugi}3 zOX1-=*kA^YqSx`t)6#keHI8v2)$?oH+G<@JrJ<9TTI|&Vaa&%raqxJBh`fb}4J_4C zMS+f8ioE~EI1SUr4(Vh?e;_ivk}CcED+O_@0}meT2`|UbfhPXk1?85b%Mb~dvTqsKcO4=FA&HE_)fA2F8<0}m4|9>`sPPalK?=PAGIB*~zyiqf#b7C-#;iYN z2r|a|G`s0fMe50^x&@f@+-1 zFADh|J z=Pry*XBE?OsApm2Z{KXARNI4q1KOh8^D3Y>)O+tE7;@)AkzkYBDG%&u2S|ah+(+?2 z(i^=7=0^2$=Kj3UmW`RDfPdBom&uB5$!y7F{bZFs=uJ$LumU@Rw27b7rB22AC>Jfg ztH*pE1V>D9aV-j#4@CZQeC?0m49u!0vxXGF>n6#PVCG;J@8|g*Y@6cW*gCPYDbZ%T zvlcMq3J_m>ttlr$ZH7A2`A=VcnM036U=ABd@b3LN@jw{y4~oQJ=ey8DN>*xf9naFT z;NWVS;kIgcd#%I-waIWS)*E;4sB|;u=Jj>U@jewtAVPBBV|vK!(-Et$vjl>praV2w z+WF#6!EQp`m7@5Js; z+19N{*HynVq8}CU%eZ+dz!`^ZIcsVpiWZ4j;$l9REa(H~P}xrhMR*y|fri#;%C;8ZU^)favn7vZ->sU|Rs8;CV2B41usYO})|i-z7)*r>L2{|i zaGzQubNCLC@u%3~BUlUAxP*m5N}(gk)6t*Q-Y!bIpR+J4)yt3H@xk*dZ4GG(h^mC; z&0D=lUKf7ch8-m>JJ8^mNDETDz@olm@HxdUg^`Dt zZj<-fXHfl^?fd##4GPHty=&eIKFCj^JCVhn+*;YO9OmT~h9O(hf+ZE|)JE+>gu<#s z%OP_HX}5tWY`gDP)4bktG>;`ghJHr8tkt4ZqHLHR9JZ!ic4d}}>f zb=c=eg1?79&OK$&kl+S<=4^`sT-^?Mx3SV`s$zD|Nrb~(a%{%#&&R@O(VcCH_K{7j zO?6L<2(4CUuKlA?Bnv@S;>74}+0B#%Y$Jz6tme2E_e=;ggG1}R5Z`|{<>KY=r_Fuv z*0d5O>+qCzR&|pV4i;D0zWc0*3|ZzP=1&ih-c7Z7;P|~ebKnQi z1JDuO5F2QMvwfi$s1F32&)N{aO;!7H?A8Ld4}?jkWla>%4vm6qc7w%t$|LfqkrvDP zkYo@POc1s1o!rFf2)h8`pB*Xs|be`Z4i3IHS%t_jE+#di)F z^E#sR1R!`G(p=XTeM!Xcx@f3~fofwW=E4HMkf#dU=>{Su%A&@wfx|xZFow5g_jS3! zA>gm%tu{q1ApGSP#l6$iLCJ&U1!>3L9A{1KWYJjX@?^x8`ArTP01vVVEhwgtmDKyq z9R5k&j+aOBl;pmH-o}&^z*JI&!Jj~!{^kQ$AV4Is7vdE>MD(HSC0d{)SF|f*^pVo-g|=k1=A48#j>h;w%K9Hz zV;_!tmcp3v`!NEgXZD~wj-HotuuOG+r0RjHNy-rPdu?=^yDfCuK70i5`r!lRyvAC2 zy{`8Hsz_)^`VFQKhXjSoF880$P31|~qm_8?#pvlC18w&K%LXXVD7~6`h;wDwZ(3ad zn8MqFa#Euy2wl`Rt#YgYKI4%g>gSn6H7Ix6x4ZGCfluI8&PH)tb6&W|{my0ag<&Lt zIEu1Vftllt{m?c9Vl*(8$j|lCXB*nUQLFP4>xDob9by@f8bqYGl4e)ip;pk$vJyg8 z&DSA2+=(z^S)sEXGd=Q={mk5!wouIED-B+17f&@RGuZHH-s(h8%?2Y0-KwryXjo3W zLVCOITba=)oP{KPIam?qwn=mEPNO>FEOY`rDXmnL+UqE&8n#dm=UDeuEet z3m%vYL0N*xZum66r0Q{|)BC(*FIG!2Cp_LCH;V2C8v0~-^|n20Q&To?&ORM~=YJd1 zy3|@t5H&*E(crY)eyrX##_D`0r3K)Y=-DDYQR{$TTDb(U9A@pb0xBslT0Cy>Iz=~Z z&4G`_{~QkCH>K0FUjCEhoJV^iy87E78;+|GrEVx@tQ!}_v;#m~iFgK-7WWqP*21L= z2F*l!ib0-<*XSu+Rnp?VJU$=rxWPhI_RLxcfmE(dPG^GgQc3Xcdx#PuTIjdyH0ws; zu3!JiXIt|~_yp(uj4*u2Pg}vhHV!-}Yhmr&0(=Kuk;dL`xxLTdx=iUL=ThjkSWYZP zw!vrG!dG}k7~pd-znQo051U|!t35KJw=tjcUvV9CMG)ltVaTuI&b!ggS%=p&7k1kW zq6bXCuAFp}o?7UGMQj7P4|T39-&VUk4i?6mr1eV*^agDO3|WE$?xS@y$DH)4dMW7j z2kGYJiLIb4hr|G?HZ^Ll$kg`rvlgRL+Gyc7iikj%v}2WHO-ua;Zs2O$HMoZcJ{OjH zpv?z@NX*K-qVsnmFp&+!=m-O#7+*!mAhHgt-uYst^lPF8mZ0y2y>2Zg#pOA>(x*ev z9+r@lA?_~@y|`N26XfR^S-j#kI?J;as>^;)V{ZS817yzr(xc4UvZ6#D$~imG)?A&D z6I=8qWJaSspYO>{Wo@=y)mvC|Ko>ZU^bgs#*nL^@MvRR@wBk}+Y5Du%}=jr7pO`s{8xP5 z$Aqe*;gD?(5|bM3n^tH~ zM7}72!h2E!7`ATvnIuzz_KG0p+1yyIjfzJo5;OJ9vC$tY0Sc6MJJT}+0YcKCyVOOS zX(^gZO>46;a~v*7SLTg~%b*EATW zAMbeMtref!&3vhv;-uCB;TFU=iy5(q(7}n~m=K%QfDwNq%G1*0>QjN!9WOf93`6V^ zg5hO8yZSy%ih}?kefPM!QMKtTdcqmIypt_{9b#On38LcQH<%z_iiHT?-ivBTqSG;h z8;~;_e%t^h9ve8KqwvD$#Bo3VzsG#2A-PrvBcKG4iK?>8@CZNFyZr6RD4#J1^fFxOk$Fjc6%1Xmr~nzh_9gh;n=x-Vs;T|4Q|4e zFBFPXXpOz>*s35RPr$~PI2myzLF1^*!;%bv{D|L3a}`jL<(L7Ex%~YlBS>A5JPEO# znm=<}f(gq23~yV!MW8Oi)G=$!fsG-)^V6#k#_Q;p)1b-Dhd4UnY>!Tf*e6x%I5a4v zPm)VSh;0ue;%V%xQ+vAU`a(hu>K6&Y*f+b0T zP6sz%6{$woa>c{c#c7RfpDscnqKw^GX5sVz(R(ML_NR=wav|?$0b=f#?4V1FR4KL;=f$LgX-9L%c9~?`@Z7DQwhDi#<@+fOOK>bBNHb2Xql?c|8E3_H#r-4r7%pw=bcpyq7-B0|wKXu`( zUXd6rYBf((>j6O?RH#vADf&25ycy2>8>O&xL)}qSi8si>&s1$U$Veyg&d1RcVrOJ- zM@%^nPYub>k7xVmyAhSx-?^7iNEv@!Y88$=F{j+Bf9<0388Y% zsvz%A&DDbAmbB~{D{^i{r>^RL(G@G~d2zc}@o)y~?SF$o+p_iYsMpH) zjKq>|SD_d75D}xsN_g=K*>Xvclu>35y;{VAr+g`W2R9tf14$c(#Mi8MFy8S$2CR;8 z+XS{%{bVJpM@kuj;Sv;vrJQ*!j5+T!Yndcx^QlCUn&yor&K z7_j?SHt&q_Uzx8JK`RF=^);=kS8L{Lmg~x^7;OMoKCein9k&ZpyLufrW$`U(QjWoE za)6qzv|X5wr3PFO)h$=Y{j*WTQ)p3${#GSOFqejh*)l(HOaMFvw_@#U-|rMtsW@{w zBK-0uQqpCWf()^qmQhR@K52DHbx1WkOjRNjD^h_$DWR#3l40ewcPQnb&Dx*m>c7F; zhx1-bfGynS@j($2lcK4H9Gf>n!K8<1>6TqjXzsL0M!-7{TwO5$ZY1KFgHWNGCsdlXyy`&BX+DuP1f!XQnVj7oX->dgqYLx%B6Iw>Firg)1`q7L08wg*m zN5*n$*~T8qBVV_k$H}G1*Vn7(&QUX?gxU5rKvD+7La6D zv(u&vVr& z1*R#x^+cq652UQXrthS)xR@-#64K1KU*@V~?sqmxPA$$iqK|w=U&6Mg_Q1AAr>R`1 zBJCm|Vk?<(gz}9hPlAi<44%L?Y)Qg&bTh#d$u^F5hS%N*bn_(rZde<#mAif`re*O5 zx)9iiIguDw_-~Nb2;1|ZrAAI~J&7Ykn(3tpivin^7CinW6F|dovjcTz~ z&|+{PSGj;E7sztY=6=P~W|*Wr5Qa#^ft;{2;jp$Ws|Uswk5fZa7~T>J*Y2ta#{-E} zSzar-QMd`U3(DQf=EkvO3xA)G9loWv zM(K~AOBWGFqIT*}p`qXF4RZP-hE+YKeSFNOm=BoMvIuD7n@xS%<56JTS<>9A(LfJCM=OXSQB$ALLk(_(sQ&C zkUbrM$t)&MlUHvbTdyjy#w`<`7BDsDc_1d4IEO6V$^+Z*SoiDS!ymle_jBPDs?j7! zt>^c4qGE*>JAsDN3wERqWx)6yaP@UT)Q$fN!uX#S-dxtR3GZCOS801aIl9gF!O|1Y zWH1A~KORsU?3``64*W{8)E%7}S}qLKuLI8O=k@VTCLw6?C~6?9u?1TK-cdlE{<>q{ z$8a=x4Xu~YU@`M;wvvK7Kqm~DgQgZNZN!Co!`I0 zLj<%Wg)GCLj(GhpjKuIZwd=oNckt2r;dkn()1cl9&-1a)9<6&cXCSW;D;fmmb?aE~ zDna%G80}vuZ=LMkpBD4X$A$IvH+R}+P#$=$?}FUr!-32~VZc&j)$F+WY#nn(!}yGg z9`F5ch?AWx3?V!F$BGGa>73_RBbH=8!uH4u!ba82n4EFdsTTz>qm|Ifp(yjFJmzh! z1x!}jp|HCC_hxWhJlhOxFlMdAvCe4#d2}&cni_{N9{T?=c1}IQaAC4-+qP}nwr$(C zZQHi}wr$(C?LObRpJe7I?3*V$m0GKk#9gW>M$aTiSm}M~c!0(KPzNqNHKv2@8JaT( z1&cGj!2B3-&{mc8zA-qEwwi;N2M=`QY1Nh(g7Jz7@iLG18fYqtFW2N|PoB$I)>H2ed|< zO!$WlRap?D%qYaOi)U8MtgpbK|6UreJgAT)8t!7xA3Sj2D7iw-g9pjVJ1kaIs>ZetLm9lSyf()q5fVN-(i$TM9CZ6h2!h0DY|-cw@q3SOB>_~UCzoJ zs$+Z**G25Cs=TlhowENif^v<=ZMM0lL~{P_;q)9$IO{^xp1b`gU!j5CaN}0*Q6#HK zsJ=XLmEf5Sj>-t__Bi-l5!Q~Q)f?MuvJ{q{flHsVa$Wr19`^SFIg@@R#QnMR94$Z0 z0jz!J^vL?x=j#9Tn|3uOg%oJQfC%G_td~GiTon;hHN*PlqQ37n8qS(DA~|J0>f zZ-WDaFak5q^kX>3XecWb_bo+(5=CNymm2T)7|ire5daVHJ`Wi&+ktWa-XpO;k^f@?sVyo5XFbe0gQZK30 z=B48NVsHc$DZ%KiIwHgnrh+JNs5&-nf@JSXcIXlVaCfurU*hZjX~4`O`aAF}A5%EZ z4)kZkq!}K`h11cKu{8+WcEZ_8zFFHOSPn#00CNs37SZFDbpR9woN*XBEeuRv?IK0Y zBo~!Hg%u6*O$fzzuZpGwzz4aHTt761d5C4C(x-%%l;*@|!LWyDfBJN1X4rJKV>7VZ zye9V_I0EIYA~%xyF+zt(kgqku!}1$S4D&iv8ho}Fy_n}_jlC{5NW_>k9b=H^ALN;c zImz=v{6x-i`y9D^`;#(u^jt)?OVU8yR^tmiEruIP3D*aA2iy=5vL1f#xNiQYc2R{g z!ZVW6ymxOXel*tZ=kusz_Wa2HPCt`L|7j!tQ_Jv86=tfb5#~fUhbF}N&76)IJVi@| zngFP&f!mtkz`_jGJ{c^v{3bwQ5?j>QzaX-V1=U!qfSIB#X@-Imt6fxu&;?PEiuQ41 zm6=RySy)0EC+9&kVkG&vMw*z*p0-6oY>z!*X65riA=;gFdAjcjFwMLh#mC zjGN2Pz5z`B;V5=Z;@5e1|Fi5H`U~0PZMgmi450_+@;~)@*#8f`9!3^+X14!bF1F(X z?t-(@`3H5dFc0L>>=8O16H5uX;?gi+5aUhn)V|`adt>#lEq$ZsUvm$7ZHexSnRb%x z#Lg%zwvDk&BqJ{)BuG;Cm@aeovA;MIWOsV{)8vFFx&NsI-(iiLdGD8R-{a05Rj~7q zYfel|4EE}@{y&RRXJ)eoei`{Qo%G)S<}wQJy7W9>Pw9qoR~)IpxG zPNC5w(AemZ=+jVW64B^VQE8LW>FG5UyP9lNKQ5zKQ?09=s@GO({=OG5<98n3)8)T! zC9jm*`G2OH{eA3C+24HM2Y>K)+#b*Neh-uH_;$Ycb>V+(u7=b-;*SxEab|aNP;>s< zPvWHJWafNxP;)eMygjKo-5%2%>dbFWdvcbeJ44TL%>n1Q;)Hi%bI?2G8QvJrobt?Q zj(mbTL!336<(Vzc`N+Y{0nHiXfaZ*MMsv*Dr#a=B&>ZxPYEF4#F~>Z^oOPOQ&SB^H z?_Fmyry3kOv)MWMPIL|)IM`=!aUsm1LuV$=y9e7769?U66NlZSgd7}aAmu^L!Op;E zOlM-Vn6u$I$~kA8?9Nsk?G9F)&7AxPIcLj`b~7D}M`W|*IYccO=ji;?1@&Lbryp15 zvRZr>?(6EBgg!U0@!K7jqwe;f$EQKsYI{yEtDWt?A8da&Z##Z&PWbJQ<-9x_xZm8b za)>e?OkXG0{+&+_v*qq9J1-}5n~_$Jevf-)xA|>3me^c^(HF@4k)cl^NKRgfJz)u+;!cTv;14GcV z(fFE7UvrqQ@s56v|2evOM;qNOM-(=4cPTP+;PSQKKBlFWJA^Or)`yR7HL!JhUgZ?j zt&p@GVatH}jVuN`m0H6e5w>^W!erh4KVg~6s3c#CLTV|6(kJR6NqFaw2cX3 z^y&kfP3}Eeb43DM?x0Wz@rs{%rff&#!@{?*?mDHFSj=0h{mVk&H!4J0Ze0{n5k>Lu9 z7D(8lCPHPZMa3@y=(fnz`H}#&*m#BdS-43j=a(rnyuMk-4M6G%>wA4 z@rFB4ZW_3peDKU8T0UFZaQkO6Kp+CBR+o*=S4%|1vEteA5b@5G>5|bq9KF+aj-0u5 z|BZ!>y2ZlA$KCSYKyGT!5CCqcNZSv|Vz3jcD@En7p3h(hAV64d2UxnAwK~XSq$C%( zX7jg(}&`oezyh*wi@6@QHO!+dWgZ(agIgyw2183;tO!cDaDGBchL= zoI2nl1<-ZthV0=tZByj*f-NQKgmj(1B-L4`1c*A)=0+jbD+WK&Oo;>JiGvNgF-RE9#fTCwu&oAqPAuR|ga{WO zegF4%v)N}(A9v{Mhdee~_Jc0TvChl+^Ce%!ZdxS3d%Mh4WJsPW~_FxO-2v-kFky=OfMIhd`RC1 zs62;e`_Cz8dDE4x7T5e(mZt-Ro43OZb^_O^&VS#LthsUmEGuG+v3B)T8iXsi_So$m zT75Fxsm1(XGLLxnzm=6u=<<7C;j{z9fFbY(#Uley4Zyt2x5Fo;N=cxS0N=9g+%+aV zGrfRB=uC!Oz1jbiH|5N4fTwKVp}O(Ha0g@ixq=?N-O&C2?xR8b%$OqenMmdsq=$9d z4N@g^XlsGqS0@YC#s$)j-tE>Q?m#$Vajm9y%z((MBXcCwwBJ@@?5L%dgDp9ErkL9- zx1MX8f)|6&uNd+lvwGC%Cz%oM8pf_YP@dLv@%9ILz`_l;*r05LGaYC_RWRNPGMh4; zpZg68TRIq}Z^R|{`?CbW?HU3qNTw`40UDrpRwFlx()P8oaa?l*xd zp2c8+;N}8<;2Z*Mz$u-fJOZ}lz8txrU3wJkWe#U)_Dg@<=IlBG?rg08V-LvpcpMMj zv$x)^`P>%qi$DD8iO*R%V!3KTEhJC5WX#uqROGmNBx$i?y09gL(GnoBmnT8#Y4cmp zMT2)Mmax~?_2~{H{?+qs=K|t?b39oIf6nn7JaFDn1I|Wd*v5eQp7{?S_nRN+I$%Zp z=Ju5An7RX|kJuK#QiRvCf@v<$-}(m(^lLTr);hWGxb0A2fa1D@GpB}dk`qAM^M~JA zY@lw$Dj6)k{>x9wOc4;X;x%K_;QKdY{)Q98>ch1Uvk64;N>{v8L8GOqE-!No)?9{6 z`hAEpnJ{tTieBd6K4kmP)3){Db7vpd;WKvjATYR+d7s6v?slNsp zC}F3bwkeC20??s$?9#4LtoSJ6159YTLz;%(ifS<}L7w5yDoNu~rOGB(50phsC%zC# zH7SjzPUWn$JA#1+?;kn%Zk*Kc;R~=n9P|~5JG(=+=6a%mEJQ1fpxcO!Tl>xmto;!; z(QE{7kT7=4gUpo_2Cj23X1l>hkg#tAbimv}v-j75N^f6&+#5_wi<|p1pvsT1iUDy2 z0GL5b9KAwujPaH9UnTdE)#!>P4_S8Rz%fHSqkECRy=mDKtXB_Qg=6j7CE+6Te+IR# z?Z~vHNahD-Hf{m*>@KfJ)jB5RMbtsi&eAKXbj|dlmmb(){@}TRi+f5ARK6$B;ded# z`EZT7^5%=kJG_zOej}Dj*A<@=uM4jgE`vs`0MIpWnY@l>-a`VTFRrmsrQ%BAwqP4f zb0d2Ki!LwRt|i%Y=E951XrAB^8vrnzgXnyV5iMjjNK4UWk&D~kTh3lGeeP<-s@2s7 ziU6Wr6U8bZHF!vI1QfV+uy2Q1zYJJ)pbnh==~M(`yT}Pk%tM15&R9i&ofuU4bY`Hu zEYB0q(IE{5T`60n0Q-KJ2x)|`u8r0J_p-0|JT~Pux&Z&U8vx={y@TDT(-i|$Xy^(W z|Hrc!cP9=RyLLEJ%)!EClUsuC83y@IwRWvE>k8XNR3M$03tBSZ`J+>s>;=irnht1e zfZ%6<{$z~yyp!<)P?D@ zDivfj8CDeyr8p?TLXNQxvy-&^;PFAK3oG?%p|Fh3ATWV9wYj6}sC=l~bM9g`4=KRm@doC4OZ+qpP=dggvrX1JDd@w2r6YJww@$ZiH%`wF zfU%E}qn+kIUg2T;eilrcGn{*l8a$S&5v)MU#1kgI z>IIa~%RJysrW6sU_@In8qd@^~T)=^Depl5K#kJzNdhApT)#7*a#V;&smnTKBgDcwA zY$>$uH3HJ{hE}OFSA{)p@Sadz7V#taU>o@cyuqhQVzn3&lDTWgkQ=~!S?lo7H+GxG z&gNT9K3?T=R>_4-2=o=nd_lX~HptDM*^C|`N9g7HB0<)=pMH6)(q;-2fe`6oSxCAI zamR_CyIHc<)zPFOj5cuAKv5dO@(_A}Md7WUYYn1jEZ)cj>(?Yc+(h=*BI}1XDiJeC z$ssv&C8H}DY}$y6A&}gwrtYM?xZ(vN_!ne?qmC{FCEo4OadGsk z1mC3Hlg=$oJYC>)N8Ume zb%@#lZ~*|fE8L2LXGYs{J+oy~itLI|N!%4LgeKBPw2sj(-Bh$(W{>V{y=CjFe?ZbQ zR=JY0PFI?CL><8sD7n)ItN8qaxN{F+wg(o#2ni@^^bI51EnBBV@j{$8g@cL-Ka8s@ z@~D#YTH_Mds~of3jR>;toJHLGFnqX9VK;;zsd&xGw5?KE6hdt^g%5;jsvDQ0)QhXX zJp}KE8e};xY3VW~o5qs~HBxGt>~BT;E5Cs1CE441j4miR;Na2WZ6MJ;(zN>!e$Nhb zd385cQ1FC}6ea>e>>t^OxCgVx4=0tdPfk&w)>VozuiSA zq}jp3vJp;9&3$4i5UM*D-U!ru&x&TB@Z_Fs)1-0e;eqh;g4tvI)9}}Cc+v4(#*E@1 zk@jk29SJf59k)!JF>O!^VyZ#51o^vQ4z&XXq6p`N-VL*OZ0k3$yn3Gr5Qc-D< zYeB9C!pL`2KW}`Te%Zz9%G8*!O3)Rt~fH+=8?{0yz-MWQGY~^Y$b9nr?JA6{_Jbo%xHXUn1Z2))|Vb~;?4#J;!d4*c) zrWL!RH2c|!1|U*C&hopChqhY-Q#0Y$a^cV%j`6dG2@0{YVnR9Ut#&^?v|TkXCe)9-r=>#AGM!_xu=CV`TpBC6B4Xsi5o^GJsZgmY(q6Ji2Nq@5_KXq6#!1ybp0%Du=50wrGogPYdw2#mjN>Lj1z4P z*6%61Wr~0%%Qh%MH5{jM99lV%h8)p%TL-UwaNiMh=LENtKz7MP23sI(3fTg*6@}Id z*jANfrZB^j!%D~A3Ribl1375{=Z-G6}eDk1O zm2@s+RIu;3I1(Ua>zHIhvQOCaJ>z`hTwXeqI%uDXqMbhEE%~H$cTaHY|`SW2^8o;v0cL6j| zg*59aBdil%IG`@rObkTS3Lzo_cxqlMfRxmUk*2EI8AO*<-mA(IbIt$h+k3{kkasiA z2gNVnP5W}*9lTf#W|S7s41h?4;fOd?TyZqXI(Ah;PPi|j8R{wLb;Z=z&Z`GG`H1dK z?jh*G{0_SV(8&(wb8O=L_lc9Ya#LyU`j{2QDKT5jf~*I2o_$E50c7dtR<4{4;aSuu zps34mF8W@9-{XFD!@r1|XB2;~erD;l$GFA?mD@K;=0|V{*%m|ha$pP7qNW>3Q43X# ztfEj<5=uX&NPuLDed6CTR6HOf!An2^^*~>R^rCSCwf2e`f5tFp_#3tNQ5ujjCzEu< zRRge6%F3m=as7#l-R`+L7dtREFLRhH^LI=5Z)u(a7cjm5{%T@PD_&F&I?sg!2NI)pTs}Q#YiE;bS^9f?)6W|wF{0U_ z!VcU6n#KPF8>^lf{B~s{HkK`PTc#p96*>x_Zh54b()}Mdt6U{t_}=MUo4IJJU#JLS zvSNrwL0c!vG+?#5IvUF9>Krh1|KoZRbE}~8JQ9d^e>HEU;5P!ypD4rbh{<1RVyk*Si^!9|{=Mu0ehso& z$ZXLLU;31z)%E&-a`6_oxYrmDp!YV6RzS8B5w)qf(1AZJB^V{Dibt4%y% zCm;wZY-|&H+Ke8Tx231Armtx%^@&`Ul8rsJ%laKIs0Z_@Xcl;+-D)PL4fcSU6>tS$ ztINzDsM-peJ|zHfoM5^AEI;+iAD3g8rA|3kVg7mx%;_BKf)$a&7eyX z@I9Zpy7TqgSu=2N0ROqE;DN5%7)w9)!NjqN-TxmO(N=3F&H{kkwwGC6kpg+ihNA)t zcXN9YPhf&M#<#c@j(eT|SY2`>kCl1lfS9VKYR%jyW21aRuN$vSbbTGgk#J4;ZQR!| zpy}ZEIl+M-M#P+#l7XU3;!)~jl5|-=l1vzUI~ALRg0_-I4e=?cqehMm=IGDi=z0afT)&cheNaUnb$=K=(^ zHxshu3PF2&@ZdJ#lqd{C*LAamD-5vesM9Ty8faGkf^ zyV$zjLgWf6d3sn12jIJ$s$~EQUKyVR_MLkhHh6bmZ3dNYAvBs-!4xHNUUCOo0i#rn z%7am+PSkC%{#K*!U$SmwGq`&oD>7bKIL~0l5s1Q*G!NIaL8OQknb?swXYvW)kFk7V z)1OBz3nuBNt%b*9A@_rB3eKQ6h)+^8xq&etK75J=pdS!v1N(OlPV*411DBj1LQl{w zZQ9RI?kP)OV@k#k10ry3e`(oHnkAS{TKC_2h&E+?oi49d&)<}EI{!7faByDk?hX2Z zAc(0OeRYKkp=EAdXj7<+P*0`Bj#h&0K$~Uw*SJF3oyWVu6QaI&1+L41_tmEw0`&99 zTowB-#)w~tVU?9?YgwF)j@~~2nt`+H8E)ozTpLNGy?XBJ;5SI6WW!TXw8HdK`vcYt zQm2-jsDkcJ{_QPEkLH*!wic9mh;792y?&nYZO)_fX#Jk2*Zu{@#S0gC76)3p9&z{< z0b15e=yuvF4Y>R6!@732_ptBa@s#dlj{9)^GnW86*@CzmN{~Cz=qhA=(!_n~{^u2J zgb1$THZgBw=w^yx8cS2Ojpr7&@j-Qrj-OkAZaBK|)Fg3%}HQh^x<;4Dgl!k+|J=qg$Xcy2+Ay>mv zqPg+5Mp!V|i5YR#DqsX)Zv7x*rBVX zP9JfPZhY4ny`DzRoYS6xuIcaR*!y-K!(fGZf}EE_ZyJoH0Dzby~VMcMa%eXM`Qa^ImA2EM;unb3}F8+;&t;7W+j=Ogf2`J0Ns{Zz+d4R?4fD zr7$CSX4D#zk}we%#cFa{ymiLFv-;C}6>XO66>T)8TAvrS8@3lUS;177YDjfM zuxLa^B!-TJ>9O2(0*sRO;?nUy_^H{~7~h)G^Bi5c*+oS1Etg3N`eSaPI0U9`wrd9F zBao-@>gi%(8L+*|vJu3#XjJfmP&fh#=7*s%pMP+nJZ#SdY&-!?n+JI8t!m{HhD|4d zwkmiCz>KtJ&VtpWWTgEc-JmeAd~_E6ay zB4=bVV%3yqfH7Ar*tVlMv704}?=yhuQ(!sA9=<+%@k=EJsz`MP!XOV%^GqzEU!Wn^ zqXKeEE@0r(J@3uj_(7TI{@gxB+=ATKQ9 zllKlH{?*Zmr@Zx>TP}!aUihl7J!Fhxnei#ZWr*c;UcfNJ_JW8y;Ia&PdHi@~jF`M3 zK!s1!iIW-#i3SX5!tAp$zSSG%8QFQ&)Yvs62*Nvia}N81mgfdp8_bp;zqTcw(;YuE zi+Pe#5{Wt#oth38(124;8r)bf$b_w>*IbFm+O&OXo@Sb$B z=pFKA0o@Ayl&E0sWa&#gYXS2Sq@gUKLtKr1e1~%D<2?3NPa09DAKDkA=Lp}`{tp}% zQazeC&_B}Cb~ci<{eAT9C~s89(Yv{r(_s|aor&H~+iV&%rrc)v~RKlmPSB{l+X+)KbOVniMEH z!?xW@1&z~^nE=&n2i9%jAK-3ytJW4xMUjtQ`d)qM*o%5GEz(O1hw(9mPUL^+33$^% z4@NgE-#ccGQ;Cy}F}3D;j`>#DAa8`G236Z(pGn<))?Q3ErZ=-o%3aJR(GA?Uej_n@ z%3M9-G)*1ZRP|FSb)7ME)K>$ zyw~UWedBuM&VQh)b%`yIRT8!UCSPK^%x&u1el+aFv!IGFywAHQQ|`wA%I!(Zd_S{# zrnLZSZ+E%ZHGA=W7_x*v*Nu%n_yxLga;Kg#Ny!MQDTR%-Y`M3DdD;wei7t!JZI6GT zG=L9WI6H0V8^JLu;}s^f^4PXhQYZH8&8?#+SLE&w&kh0yFZh0^HhawyVbE$~^Vdjv z;zJ-b0XWuv_W=TKTR7nD$ge8vH-iTb!`!SmD(j12H`{Qs8D#opC{Sp9`2OvlAvVT# zVl%Ubu9zNikjm#SB$3Cp44F!)z3JMc(6`Y0dkDZw1CWY6i$*|ZMWGcfD_X1SbWs-) zAX(BzBZoo+t59=>37J3mv{Wz-MH4EYs&hJZ<|9J5f$I~!{@2Mrf~~L{hxBE->TGg7 zNvrCtOF_!KzGAbp(8PH85J-dejfDJ%^Bk|nRy{)^m65?#PZYaBlrn<~>g9_o0%Y^g z*we$!a_XUY_CeFQ=r*fUUzC_+RYwY$u8P6kWtlF7=<2A zJNz5*E3nP{N>JW@>EbJlJPEPLOfGC_q(c{ zEjnC0t`2q?pFKFHE*7S0iBP;2F^w3B-muL)4je8ZxTIT}rQXTF+8K^!(&?IIu^IH50JVKsz?7Ew~@K;%rI#j?kJPAMqjMOTk zfyr6``^7IJO8v<#F)?vtmg};{|4WOC$MTj49KdmP&R|fvh~sKk>G~p*YzLoeI>pdb6MW423F44$<10KO2X|v zrZzmoe%fggvgl4u_Rd!g;15I0X)C~K+z!&K%5uCMM(5q$Ic*-t1cGy#>M?^gU`_xn zseh@Oqvb7Ttn)?aSUYCs9lkBD-Rf?!ymci8BOLK!df9_LNQQccqCkf9M4@J`3IfZ@ zOoSHwPp|ARey*R7*DrH5sEBCDTG+^SPg**)akCZS5qCHo zAvng-(nK@rC>GEY>KTG7qnF^P0XDGni&)@r@;baz)+Jz zAwfPTUj+JeSbHBGUS}PX?z+Blb2@W}9UV?qoBtoTOQ3OCk zQs!xztY)wZ23+xq+ea3WOCj$B=LGKvx2<0*BJaZb%CqinmR8Ob-`~1<|7e`#9e`6F z!)C}s4R~jN5XNJns)J@Oj?0|Sl>E0b32a(o!Ja#8P&ZCw&4iRoGU7pkHSpu#!ng$K zA$R?kGtqe$mflt)aSiR8`Qu`u=B^haWspFt^h4!$H&?cI8C6!4CUnxLw$*vF9a=m$ zb86EMz`w0gXWSSjr8QTojmFZ41sn=dF%?4v?bL`eHYVE77l=u#do9nHiv77{U+vui+ zXFl&1(F5F$5sdksEFI8u&w=em>tlq?J@p7D?vUd|AOZ<;T{Hf3Vagn#BA#lHQjk;! z)Tw8!j%iu2B^1`EE}e=%X%(SolRRS*{to+U=Y6?f1#``DXx#oa630DGhuwnrT=f%Z39Yz#$)PJ8&Ml;QAkg}giCN>vpJ z9<@;&$1?O49Ii^S6g8v#V0r2+6Hu<89&fX#m@zaKrL0oR zhm(A#5@1$Fw5whocr857@Q(;#FS91kBHK8+q5ooa2N+g&)+}}o6(@{0eW{gCf2=(M zHob;hIv+zd2(X34e0-uvkSC}Wga)WvRe(yW>zRjVTLki>PR2HN+RA8UOBF?F-M2~h zJ#LQU?F)m~3xqGYtL5hYdi^MNBR`2f@XHm})|3pC5&|M=G3iz(Be%TJSXUf%bHEP$ z4=SHnOfnsj&a{jPvR^)LC4;4AMx*(fYT2ap_?L1Aw4!Tduhd8jYNL~bkC>uL{NJu zHkbFmhqEYqx7AYTwE&Q|I*V41VfMt#YcDpt-=1}@ABEWD-TkjIH#T$2NeqUFX*%gM zU^Pr2WlX`c=}5pM6ZS^-8a(ulcd%ftx1rV4i7{~Xg4y0g{o;b|w9xRj81^*(@!Rc; zy_+-0>+kBcIqQtBlzqXI<2^gGeQjX($iRNoMU$$FOWJ{+707%JrxW6?8^U@1)oA~Y z3!YTphZnbwYr(s2$rcQ|X5d;d8^o0LuLlsq2G-3@X3%1u4ae>kJ-OegxgZBs8Wd|n z2~2<5r0{OVdc9S!2(jI098av4F~?gFToO3PDoDAOk9rVtzIA8CYG^29>=f_L8NLyF zcw_{Hf~=~(Rmv5tc2gcy^479^=yW;~^hIe=4`iCH#c4yJF69#CDPs$baocmc7L4^% zAax|3VO$>f%sCt;!92vh;I7}+wEaUzb9Mo~T43*;LRXOxG%5lt##<0{S-^HI6~jG( zX`=#%wn0gZ^^oBf6qeidXvoWFS<405kBoh|#+9j)!cV}~8cJcjjSX0R9{U4_?+wxC z*ZKV#Uf0-blfR#}DGl*MG+=vjk1@u9fj73i^%Beeao&+rA{#uNos=JX~2{Pn9uy=W|MBW6B(j(P)Nx z2Jp0yS$2hG8}1GFray{bu3z?EYrC`l+O7ce-fh7G5z*5P zTPG~YTfvG^tKJudfvu=;#KSY~;eHmV$dh_x- zwh#G%xOekk#W{@i=sx8)YO-XH=+5YJ|B6?_1DFk>kL%hi(V)dJCIx`1&lzJ6251;h zH!pWKeMz5%ugv<$UlII%_uC4ckMv#IF69532il0II|;iZOKIm(t)z3z zXYaDsIxlbW@sv#=b)&GFHN-B#aPxBR06rktpkEb?Zpo{77dd*uF9{KSLv`0>6Y{xTe`(PzCGgQG#`-a(m!Ht1vDST_@@+SiBizIo|zK~9a z>`-4G?JB5kU=ZM=DI4L)by-uo}! zNL_rP3npwDlG11bG^XqeUffO!?`gF1b_5-=OwJraUd`M1KS+E~_2BOVxgXFO*ywye z=T1(}E2E>|-?Z9M>Xd03l+e(P*8<}K-?k1Iwfga$y4!SPonI!oJCnOP^8LN^T6*Hg zM}TzO)vo8HU9quFLK?cXR3Omtc=wY9Z`9Wg5S(SC5uyPqfbd849l^j+x(391^& zyEQ^-vt=)h+NHkhzj>E0@OGLR`a-MHQ>v&2!0U_Fem~-gdn6O@{4t-_3C9Z%rZ}S_ zuzR%+7CTuD8FvEw{{U~FzPd|RBsnR=t3vDln^ULeO0o56UbhzVrGy;WxV#o&iv3&~l0i>8gU zFK7b;GvP*gkAJNsYcn->3;Z&*-gn zFrveCoGNoh%~_6k)xffWjX>4Ofmt9OnB7pTR*{)+My}0hMpvM)FoB*_odIm6D>STY z(PpZnYxIS#hrFqne$~s?T4y`+l$`Lv?Ttg<`E!5wNx5UM|KEWv&!&}mIjT}^W!rxR zht}jP*R3cC-!$8S^Tqa=8vhI^@}Ucd&5y%)|He`QEq$rm&}`*RH)Zf~KJZz$JA79z zoogAinYMaMWbVYGq7<;yN-1s28W7TgX=Sw0&c5PUuynvk_~=EWjG=4M8Y@HP6y8Pz znD!4K5q&FyIGz35;ni*@-_W|GWwc5tQB>BCY$>NiwQ+pEXj>Z=n*EY|gdv^ri2!QD zRdR-C-tGtPyx$X4=`~v>Wt7nG<^@}qFz+v8M5DFcfv0`eZsjd~6)kmDxr`>Pq$~}N z%g0a14EJ!q*(lS$nY}Kw#W*r(`wA!D&N?Bc>V!-e{C;-GHyi<(lg|oR zy|>ec)zdfu^nP(=QEiPu7JXu00=c`OB|XkLs$7k(^f!*De&p#E9kV*cGJ6mv9VEpO z_fEHKD1`9+vgo61FXx?KhmakQ#U-&w&C*q5*V2*BY=Fr@pu2naW<(FPWmMHp;X<_j zR}CI7t-T7`v~vi_!V(273k8Z1G*Y{nMK4;s!AMZsTJS7o37ngCZ;y9Y^vO%cqwXP2 zLj=dLoHc~X#;42Li5PnL0bO!n<4gw-y#rpo&!>%3Z%5nJ82L8c0RF&sUts&2mFGz1 zkP=k{OB=rY?`PY4$luXP-?_aRowxDj)SZ{&P!OL8&EX}!kKsdefZhg&Kg7lKlTMe< z8TQYdmZ=M)z#(8QEa5C%RAZqAs6a$0v|qJPwLHUtO;~H{nVLgjiOs7(n!49C*N0F> zQRs8Ayw$JM5n*#cznVUWV$6(Rq<#GP*Wm*wHTXZ_ICY(XzCZiN$5KeW-@->!_x?u} zFEXLuO!UIHm+5CNkCe4{1HNGXUL}h9wG|-1if0yA5?fkF5+Go*zRl8t4^;ue0xAb( zwcEn5V*l^=$5P?pOnUHr6oBUYsTX~QHU8h}!SttyFrW{t21ItVg=VA3;fhwcKgJJMm0^y`_`sL#KnBmyH^*_E~7K=gPA=Li+ur>fH(MMx<5XDO*=cp zq3FGaS9)kOwV`RsZRGh}zf5)EmDM0Cop~DR*HoOHB2}AY*0nze3kW^_-`z}fx}mFCT@Oqn<&M6uF`8Y-ciQ^2s#;*y z^DrJc*x2JI_*|D<)!03ZLevNo&+q-jcTpP4LF>O2G< z==(9hM&Mnarl;0nzb4`BqFd1SUEnSFH;Jt%of1T9;M6KCO1?H+wSMe35^45%5#gR9 zsoRBC3`uS_6bvj5JujEtpwcV$wB!|SQfpf4fWf+gR0*#u1b78aHD=R7iq<6kmP#MI zjCf}E!3*hw*nHnSo)0j`)8jhoef9l_bBPUz*=wD_=Crl+RG{ucsui4_zMeTjt_BYY z+K>2`2$N_2Et&P6d*HknrSSk5iL6#aTcJD!l4D8AJweWCxe|~Ce?d)=JMvERud69} z2&aV0veD{nSvU`MBsu}RC$)EyIg>&JIl!K5C(Zk3dj|J_*AEvr-jAI#4Aup1UXUkG zu&a)FR*o4zq95z&qt-7Tuky6-+vEpz#l0=$G%t9r^tsR9(1$ktXX6s{iSyG9X#X?(Zm&#&*Ez?AkWP%npD)IPc}wU4Vi z_0HPdq4D88wn9qY^RVm#o_6kiAiZrk7?f__?t0sT%z!syh>yF}g?$ms+#hVf-)93d8&N*WX+N7scjPFN6l`LqPcSxF9__@c5)*@{_cZGQjQM*lD5Cz z7i<8|{VT6=&lB8cZ@V7CH;jfHIrwJV9%3$xIVPOdXKG`X7mC9>|8K*srC0= zW?)Xn{ZtQCKuwiQfcFD;x{vijmyI^jf(RP$GG^@r%pM`dBdX)ggD}AH#<;T`M~@v) zwal9z^2y#DmgAL-tIGtTV?8U2G$PcEmGqQhY>3-k|MJm}bSv$YBQG1Fn42E5(chj3 zn_jdq`riws!}f98AA#(q!IXVHqgKjCxW7ua9u~wn`nphKA8H+{-!_PecpI_d;7xmU zd9LtITRFGs&GNRZ2yaHhhUi!)v6u3*54ETF1%=qXL*m~6PhYk^ed{&dZ15@MW}>$S z65<||OAoHV%3TTyL%ic`Q!PZMS{r@N>+PLOApt#sFN^${bXai)b_Z@7JRXX6O9D#jw|CRH&L2r^5k)n~`VZSUAmhMvMj zxo6>YeQ4xv0(wBZD`3sIj_S>|nBiLt$@B9DD0W7x4>M5BgH+;i)>8rlfP3L#}GohYiQ%Lx%}9h*9|$C~T5&W-+>gj?0cupmc+ImErZ zzlrk`tHcbz&%uOM!W3*%C42E1tnKdIuQCR730EuD*^!U$#(> z%g}5z>navZ*s@2HJ7x|_atMHxw*nT|`?Ktbmadju(oxux(vVsi3he;|5L`cYXgutF z@A{FYDM<=S8^+(GmXDT$mR&?EQ~u_&x3vTCaxBP0_pcZvT7+vGmsYLqhYC@c)-UH& zCT5SZ=YsRjc2{$I{*2$%jeT)j1S&|tg;7K;rH`ScG*qNRgj_P<5_4{^7`W3?XHPe9 z+0H(}8^z*@Gy^65Izo~v=4nfv*9XGJl0Q{OC(>CXW{rwAh)4JA4v1AhwbMYMrbxAO?91jjBz@?Jx1{h zLdzL3K?OpsdCB+zsTuI^=iT^kPKuum#yZxSH9?d^O|V|pgw_7muzl>y%O8$YzP+JK zEYb0&V0E4O7AyWc)*RfC-joF^;HaG2^mUo|Xfb2$X`6_Y*1cgnYuE5~PZ_`>pOi6Q zNkgWtP0EybN!+Xz4XlW=80y4{8l{}m>NN_Wr+2k-(HKtaz!(h7$h|Pkzkbw(ilb(( zz}=j*H(~LE+1N?f0~39zo0mLGYCCOWWXh8!P#IS)pjDor{BpK-KmIsGu(9uMp6;zf zmOgV4?(4jV5!~4~jyP3}!lj71q|;=*UY%Zn?m#*ciKPC9V?BH$#TiX8V-(+B%GYLl#O1QKD{?O4f-<=pU!yZ+4 zDhQr{>Qu{n;ZVG0&sHQUIv^zHM=5@u*6Q8FoL!xq@$*K_Eid#_=;u(bprb-LipJZs z@tdXhn&VTF0V$kSNmz-EsKtB0Q2NR-gWlR&|KRB0PHZS)N>?UeY-v@I7DC{i~% z!7qdHLDGKZi?wE^5g6jr8RIXZRG*;h1?{St7h_C-qR301C!A$8Y8xd%uFn{rgh9Km{(R9f)+sVSCMsIpYOr-16#u&H*Pb`u-nk64+$eUC&wD`7BW=$aiz(EI$K}#Ek(f2fZ$yLK0t4jaYWiDIkCYp`)N5Xys z?1*H6009m+@=6RMZ)|od##)iOZy#HwR% z%hx`yV<@dbn^LyLZOWy~a#rZA;a-7WQJdmD61C-Q3fPtA6|&1xQnI{4loPO&pfBM? zYi2x}JbfH&&KuL!&0rTjQP|_*zrL?u&Z~2xPxIaGr}6X|Jq9c6qG|LMpnE9%ZP7?$ zG+H*3C+-($C-O!kS@G!%$sbeeH85)L)Ig}eWT+lKJRSndBUhj%<+c6Zx@gBr?=^;m zFciY%!yF_4u?sQ3&z5tRt}Hk)j?KXqSanrDfeBeoZ2NQ9RjG*>HCaM9u9(@LI#?xPs~&D_7wtTk`RP4dy5cEHtk0 zw~GTL=nDUjv2*Os1X$K+Y}>YN+jjECw(U%kiEU48+qP}n6Z_8nc0Zi8?)eX0y{fvO z+WXbVBtq)*olq^vkq8SMnv59e_bm0jfUh9sZeH_V7WRiMwa-fNbnQ69gv~BsAo@`C zRWRkIh|=hsxhSiHtq#LgF$)o@YZcnwG|E#{)Rza_l(aOFQ&E)h*W(}Lb@Y6X%WZtd z$uFAPK@-Su%d`4nV#t15CKqHv_2QeaFNXkFOA-y;-gYCv2gb~ftrY~kFO2-O7=epO5hw0`I6{)H0rp(IaMOL95Y3Rw2jz)Lc@}y`h&SfPhjA3XyqGypFF4XFh$@Lz z+5pN~3&8|UD!?rJxcH?<)A^+weBkp+c4xmEbueMfUIwB5I9XU{W8mi_`Kc236)5Lj zFfRY6xR)k^V8(Rg`k&AF&EeNGuVX`(Kj$DsQ(pwN!!8cX0neH~G#9b>hgGP{qTzes zmbCR(q}^oUL^fhj^}sx!JJD#A*IjNnN;Fl=)&L&;(XYEp!Zx?ucZE%r@J3Wy5a&7f z##K}J&cb$7;4vp{t{Ft&NU@?Dr69%Z*c&i-&(3%QVBv%%Dt|R}DJf{?AV;tlsyET) z22Fh^NZ)Bq!!1mBv0*8MO11raSc4rh+^+lHJFLa9m+JSIi!Xvu1x(wJ(wep1jtPycxqhoL@zH_8`Nn#Y9L)%$fFVrb?h}TPk(yO5Et~%>ChyIS zVh-`af*3R1m_>_qhEP|(GxHN2CTtlI82Og$&F6z4^XBm~~go?Sda*zI{Zku(DeN*C|E0CsE$ESMgVKkIk z6hPbD5vOs(V=z{zaNfC%gx71nXz?rT@keCfrolx!BRnZmYqwECw2TFV{CYhX#X~G%y7>W&#s|j>-y0pybS=lr>iOjdk9Ih-i zR@q~_T@t5gM~8RMJK6m>GO&IlqG|aljP<~%{~8#sWKdqH6VI9#&=NRUr8h46OyDby zw)AELOs{J~aK>~Uw1!t3hCXYw*5|utZCf3uqxg*B8-=iuzCl77N%0G_GQPY#d|hZw z}O1rdUyL^=$rqdj{hmaty+-dHwsM4A|*jTS>ov2G?s&=W{`NH z0B1c__iR$A*4?C{Rg-siYSi1Jnq{(&TNZ>Lv-HT~Hpk^6_ZZtm#10V5uZ3bCTkWP2 z^9U_wh)pn7C{KC~X~8()?{DkP!hK!c1A;3`l5NajCAA1sjAJ0j%izyTE7u)FlKWGKBh`dlmsb;4B25xte)N51;KW`9NC-|8;QC|1uLz#Gna)mtm!Q2^SI4IQC6T#tr;Xuft9Z6{u%_#2eBNyg-nQpjGJ0^@zuM_*8wAk@} zpNzXPtJOaV?w2YE_#^Urjl{h!uw}Qm8$t zgOvO)-zsCzEfqE%!g=K1*`Eq5-oGx%NYCLzcl3Cw5d}hE161Z|wxbn#pUf*HuCGFf zZmp8i2a7t;o?iQ_!I7Mu^1;~mFea_!+<5p{2o0xqu;;s7LFcuG=cZR$Un?bU*PGl~ zC5_NH^EKrqVM!}Rl;<|qyQmoC zBQAvEHGf&Ah;R0}Z^&_kx7XER&=tFC`|?JWxpS`cWADs|fSoFXP z7-SCETCMY0ZOE~35AgnKrp4^k&Lfm6_^sQp^qLp=WQ#AzgN7Ml-*)%Wm_9q)CWvf6 z(WMb!r8!#~;@Pj>Cw}MskkPPD3JN&@`6cC_bJNtxTa7F*QGD*&Ysj*p>jmnerSBr8 zFQ3~rU&ziS&w!5=aX3#A-kEfWE|4|^8Om1lWrf0Y-srgF65D3Qg@y}QhSEmpf#=b5 z%#w-rwb3%Q4gwaL2txI2r(t2}5&%E*dp|jyLGhZ23t*TfA*I`BOBc(76eQLNNIx1& zyHaPw^6=i)#~kNaNk=+=L7%Y#;ot?&lY7vK>J+hE4*5(GBT$JT+K>|B_^|Qa^ETyCS#* z-3FwrGZ`q;Ip2?AJ<7&a%#4QD%4(;2j?Vx*?Qd_(&OH(%d}j2P-@H>S=LJRq8ZmkK z8Ik=)vqa3;G#BT6`^K351!z0 z?~0Iz#v(w@w)a0?XCMFr|0u*(;ZH7l%~v#oDK8fW-cqjT&#rZ}Cj+h^e9r$goN@mb z!x=LZD;x9wKb=kgQ*5a%Z#+X2pW;RSi53Mp!;VC=`Xkj$q~3u9g%2j;ow1Z*nt1N_ z{Fs>Wn9(Gi5sxH-Li9*NM3fH|X)KCN8`%NLx!e6CAbscE?!WU<9q_gLz6z%e-FTYa zo%hu<`@wfBKrD8vFD4}=mA4E_nx>e`?8Wji`g?Cqzn@a{F#f#N}5Va*5AhM#o)4Zlvz`Gw0ta7$kY4-c%)gR}3avnUw|zru7=> zZ;ZU1!?IwQ_oOyc`P3UmIDRZ-<4-ylrbW5e{PyB^H`}Us9s9K$VUzeLakZqX^Xuy9 zG&k?>z><}q0-=D(&~`?n{-^0wsfmT4OTgkQB_&_8*+EeOTpW=EQY|WLM8c4aF~&Vx zeWYgsmbe&7VZ`zfmN7hY>Ti-<;uVy`-)=*m#{A4_=|Z+dC&;J~F+;w_DD@FqLVU7x zWE7~i$g{{%zfU8EhtQ0n??LL5(`D#LOOPic&WB)(A?`WuvF^E;_S7mI8eT!sR26H@SET7u%u73D$G8P+H{E&a~5|sP8 zt_1eP@&8SKug8r`gTERr;L^_CLtWn5->MKs4v`u35I;gVffHagJ&Df?H;4B-m3~l- z6MTXYWxD;#fS3gTVe7eZ2TE!l*T(LAyx5(~fS1B~@qKp1%oxip$l-sro2&gamElz5 zWDvK@YBcsMF~X0chyQ-2NR#60?&@>gC`oePQp($Y`(d+`>W84=5c*)gHyojwEE>2? z07;J#&dU0TLPc`QXgAT}d;mnvx~X*B?n!H_3t*s=-KP!)AF6EsUUkST6voVy1`FATxCitoyjB=++iVPGRGtJAob&}s>}S=_p6*Paz1Ts? zm6oL?vHaew>sgk#kzl{p=h%Qkt#E1OvNlUbd0E8}jS2--b|m-k52PajIPdzNj`#y# z%Ssg&skm6oBGTaVxqfqBdjGl-7%wBw%6V@H0uB;3()Aww?3O!i2%esk-aU0EFIL5H z1+3Kn1M{nC+b1RKtgx%kf=XROMkXn$0aO=67pFDbk9eSn^)e8n%v7b>`D*~mA5Jip zFe!=n-}9C{%21gxP>xg%Q+PkpUzRAxROWIhwM7mXG)+w(nJ@b0*JB218fvOS`F}7M z)eSXvX5?xc_gedyhBAD*1qd0eN+iRoFRV+jcvA%jiSj$kUprN-ps-9?(-KLYRK~0Q z+mKm2|6Gi;w6Zw9K6stL`1u-_14@j}LELl+6>LjU!Z_8%fnqf36ASJoBX=f8#_d`Z z32Bj0#pDk?aq-C|A3<5q*Y2JfK%pMT;f%}>D8Qc^h`68<`$AmLsITX(4*^i8P1hF6 z1zUY4uAttU)#jA9{_yBIWzz8z(L!ORJ~0QTAHEQ~b0K^2-Gd$tH*?UAz(OFQcsWPy zM_)x1M>d&_la}Bdmq=K#`nFy{`&ZL59Hxlk-E*sYT$ywr!z0Ff5a1B4`~2(oZ-xk- zUiW{wBYph@#vs{n5_#E57xIAUe@ zK_logwlpuPvHN5Kf0=wVCD=UD8n^siQ2F!l=RR8lKh-av5sbi(`0JnPuXhWM-dOmr ziU5ZzntI~G+CSUUzxw+*QHx_5sH(S&vBnWs(M^J*#WEwYg^focVwZB9E81}6l2oSf z2->`%DW@rCD1Rr2ji9Qet*))8s9p!m4cOm$1J=NtsXIH5|6$$n5`-8|>@r)O`cK-e zLuCpQ7QIyt)p(a?ia?Q}!w(S2@gljg3+gNJ`lPd{8(YC&*Cq{pzXgJjAZ}uye`R;z zx-N(BkcRvW`+d!ooQwmB1GB-(tK>H~rHkpjxxA?E!e_5Xzm!|t9TSDDC6BNO&6 zhZ7POtZk@qIpU%_^SdU*W-t3e}ocbg2?#o+3hD4dzUwad0?K9y|opSLi1I7-4&Hw#Yq zqeiy5h(%KN#f2f(0Zm_K3-AwSAAXR*o>0ZUxl`hZwglw?d|uB|YAhk^8uyhBW)t9U zBYg`<#2z8wR^V_JzfYNuEJ0VI3vzfRT(+y^$3iF*)w@<9skY6#4ruraej|NU*2*@xSjakGR-CdevD%@qSy?tLlz{shdJShLR_RiT_RG zbl80yQTbxQ>syrO!m=MQwErc#FOjtf_}w^5+9-9OBRaZG5e=`>-)#gBx__3{Pwg?l zjgSAsaQ>{EjhF3;F1l~ag&6J%ksT|=1xn9w7!lnr^oz2xSMYab)Z(x#D$7ZY4yz;* zV;rbcV6GfYqkN2eA^w_Yzb_0IJOds%{cwxUh@?iKj0t!faL+jJy&Ta(GZI|^D^=k7 zc>WOcu1&*B-m~4Ryw;DvKh)^iq8_OZ3PxIW)Wv$OJq*MO&YL{I75@YAFe^SR-D1r6vM4)<#q&6)gUzY1x9o^qFwLH-$5*mZ{)!} z{0e;UJN@s59ZV5tA(lL*IdcVPkTrt@$_S_mu)dD2UwD<*sALScwfR-zH_Tm=m zk(*n`^iP?a6;{&;rxP%-v;qzTarOof`5+!KxGTzTr0kmP(sAm)*5Syx8BEI3KeX&m zMO9aYT_-1p^YMloOW30At5-;BbzODdzGpB#upBEu$@b$#sm}u>8S4o9FHm=^!}e&% zMQu62Fs_EKS;>+&cP4|?V`r@)b#!GtX+lMAdVEL#*%>Xu_S z88Qr6O8ePAuMX;gAKLJKl(e5+YtQM7tD-w(7x=xGrL&f0QOZ`z*&?1&LnE$O*3bTa zq&~Z|2&sCqiCwI$aaNCp#+1#=5i-qj1UpesPwR}H;o0KNvPe zgIY3o4ap#*X8=APrej zGh2?VZS%$r$V0j8M!Y0av}JHf_xr;zM$y|R-a+<++^Lc^5h30w1jtV0KJz+I-fmu? z=YxxOv^0Gy7;PAtnDm>C%;im|1|W3CFc4;$@8?IYe)nN_CImho%fr)%(xbFKuRFV^ zRZYg0TW(D+oW9Zi=~CJ`A=I5^l&MsT4^i8<_kQ|L%7ePC-}(}Zb0yN#NweJ}oop}R zIf8*vd*OfIS=n&SkZqSQ2?Hmc&Jo&E|6NfPu&2JTdGI@@Gm1iA60E+F4-fKcLcwIJ zD$u>oqy1fef!}LB4}7)M86d`mSgdQf4cxSaWR!?3bFg{P`4)T56{R!meVm|m1qHO( zHCqGwOVbKBHI5aFVh=ZmP3pMnLh_>&)+ppCr16Jk8-)2qLQ^A-A{BOsh!_L3W$_yc z`P<`n>x$CY@veTErgklEsEJVe6DJMBETU0YMTR*K2_)zHgUb>Aqk<3X`wIFyKn({++hc3)GAP_IDkhMPO_ zxIpN@N3@Y0vuWw#MT`H^U-KzIilJpv!5-?#x~j&wz032lWqCVo(yAV|av@q_@$XnF z$s8J$GV>6l1kOW$cf)4Rr84Nj+(+{V+8bN=_C{`G6e*yc;s&9CI+F-VGZD4jjTRdU z1*a_KN^CO3jRL`RxM5LfwGwgl%+TSD-MJsZSX$r<3Z!0_(s|k z;Y5BaR0-|gX#K!2Mh%EN`?GobzUMT#jLxdz*PdXHUw5MDAYO0#{o+pPN|0F{`nUd( z<%B%2bTm?MS`y&--Uq8(_@F=aT(r!d>{U5ge2lFDG4cm`%YtgA0Nkb!!+u+@*I-hM zD$v8t@(G@n*2*H*IONR%V0QRI_TxZ4=FgE%zD=_dsea`OldR?i1AAF#i}|zG*JTMph_?ob=+6Ec!jT>kd{HLsU%E?Di4yd$mXJbJ@mVgWkP!8b-1& z2s!+P&^{m>oKBz6qA!y?cZ$tCgGD<@{d|1(P}1V)A&J$}GYAAd zbpc>rTaM@HGsG}cQgguDMVnUzcsm-_2Ea_f4_By!N*>J!^78(nDrJT?%#~6Yp9B!L z+r6=6LvA|qKeuEJ3XpJb#_l^uKZuwrqwn24rKWnr@Qd4j&fd84*fNwZFnCbix!KB; zqBO>c6~$wD*94kfa=yYqcGeITItR{CIUjM6nH(jW{-vOpAgPQ zdG~AGPMQu#S6kzQeFJ@d_?oTq%`t>|nRS3V_&J9uYwr2Nl66O3uj*^GD?r>?MjHs0 zT2;2Y!w)E<`)9pA+se}~l|%|*ifawyYBjqw6jix|XtB6ERkd*O5}#zG5ry z$)Qno1!$XOEyg|xu(*#jM$#G*HB7V@P`%e7&mOaH8JVmfw=F*lZ;N}0{5znh1WX4A zfVHiFIcvKDp>)z|7tXxLc-!CLgoVVBt>V6 zqvSEa_>#6yhA=&hAtsn%or~p0aCP40RP%QjuR z$iQk6;7SY3&=+E1i^M5d$D*CL&fu5ZYaIcVASX&BM2nVU4UfNvEk{7u{Gc<-XH<`@ z9)97sf4Dr0_Y+s{BYmpsqZd5?+g}O1f!NR@gsX(*)&j466^T2PO8A3WU)V(WEUP%#@Tg4$JE!X8{y2KO zS=YO4Z!z60X{^G8Sd8s19-}F;FG)88t|r^;PmXwrC9(X8 zm9sXh@p;LNx{#3t?L5jG2L+HdT~`h)$qs?Z8RjCRg>wpqQjb>wrv9!1_yurSINWZ# zEk3DTu{Glw_oM{b{>a(dHY?oGG7H!=8gd$eg(+WolK(eThkBjKh{?;z+seaDOQ&vL zCk_;tFW50R6vW1Cza_PBoV-{Iz+pZQb7|iBZ7Nh?KHT)mY?p?FCJrzuNXJrF`FI=w zw-rJtJcRf4*v8$Fo!75#Lp~^uus#5+2?VoXc^MBrUyM0@=3PIRfZ{| zpoU$l8n;C}%J9}Ww7Cs~$na6#a%!Nik20^(p2~wLnYR@1MWGM%eZ|R!&I{AUlS2mA zESPr`kTH#a_BXI+uF3KPaYVv%tQfg&UClfLL7C|aGB0qlyu;P1P>k3=zWn>N<~1hG zHpC1okO&HS(FIYvGYN6(>!WA}(>GqnNciuHT*GM@5a5!rpz!}ylXOr0PA0LnxOBxF zrk&RLOqsEfU}kmP?_u}@gyR^+WAOauqmZV5YWg=t26?QIyE?;c;9?^&a$=^r2S4eA z4rlmL>j(m?EW`q54Pp08_inp$jm-5^(ygabH&oph7Exv&;@`*m0wo|mKL_nplnW}x z3-!s2YZvAh#Qxm;83*Qy3VX2z;{)^?H{5dmJ(X{^Ay2LDTfUu}AlRO`-nXB$(HU+F> zxoroHbj}Fk*=1|2uR3k*WcA)Toi%OULTG6he6&unS);ZZ83Xl`rwr)l-{iKlt!!I~ z(p76VAX#>Lr8RYoOqbQ92@f*(oTsY}Xz81xNVr^Wn`PrT%ltqV#j^2L=_eH;VCGQn zWog$1sefqj#(q-*=PT1;ffJlJf>$}f9Zq$L&|`^d4~a5fX$+)n|C#l%m+yHYk`rth zYVqkVuWqmV{KOKw?7|{!S9^4so@PPi1+)a3+lw(i*;hXcUZTg|l+No)&RjlVnm+4V zq(Bhly%UIx(XjPu34MFw%n!Z^-KNMz_uIIAIy;)2h^f~ zijtmmp5~Fn8{t|EmN+D}Z%C-`0OpP+*3kuV)U~F5PR7~9_3Fe|wg)gE8WLw4A_{~j zH8`JUNT&CBI0`#*W7MVmEQJKEWXW>D?oFISckuVP{4W~a!~!j-{L{`mX)$gq#{GAB zbDR~@M?EbQ{73x^zinWO3jFd0eb)z;8v1jDjJHn%ai(%oz{2*&CB;v#dzv$qe}xw& zVu@1zjuPXu#kGNM6LJ;Ug9@t#&fY3A){*2veS$K3pMoIY+_G7Sg+I*c(a+|bLjcty zeSAr5py;O<0)yf#41LDkqA|lI)-opnHE3`1I5comn(=n`+H zb=!byDbDJ(n7(U?x%PdFNZf)cgd~kX=hsgTL1Y9AJ0(w$G2SgTX`(Z2`%Y9080_~E z6j2xz^>hr892bMQZA|mIhJk>Xa3IauG^3tG{jEyn8<9%c$f?SDu}CvFHH}iy0E`JZ z?%R$Ii$JXEU6Gy{@-@RnexrBTY|r^U0r6Y`VuFZBtLBm{K5Z%HLsJNE^fXKzG!=OLvfh2e|A{H{epViYjuCBfr9&1l9+Qgc@Gg&xS}E>|GC6Rd1o6L zXoWGKLE3-uXEVLiaJ^?Bp}xH{P0bL_at>Gy7zHNvc11oq@tcE2cAh z`=pYHWeoVkG4*#-2O$}}I29nOVe_HH4P~0lcE)M| z#1n1{rF?4GvzF6VV0^~e+a6hbzd+mC+L}9xCztJCzqm$WyDkp+vPcH1m@9%Hj%Xj2 zjnRUdQ^~|L>Ex9oj!5hnbgLrO#4iTgW=ReNV?M~f7rQk2_3!n&hYkCmX5_Rhb)wk= zTj~+0*?cBMDXYq_c(WIuOT()`x;9$uztQVHJ1;%H0WyH~pM4L?NX}|QrSIF%P4=Fw z0B~Ew=C8QXsIp|WXC;mV5cn`5e}MT%2g$S2^K$akv1f5))B~+X@Ufx2?{g%iMN{`_ z%X6UqOt5~eQ{_@8xqSrpYN`jYc&>OQW4?L|pK5B_^>=a>o%+h3uXqn6mJgy?ZL?Zu z^$r%i^O;9bOHZ+{qmomb+~yNqmWjuv5IW|p?v(-lIz`%|``Ly<^3jR7RQ%?=-r<{N z3c#ADGjXRTaGinmj`X4DxaNt~kj;D zUtG5=b8NgxU&3*0{jo%P69xU;#Sl#bkD!(7)!8d^1DSXt8F#DtNq)uADpo5W`j8HAJ&Kd$P+=J||X~`>ya=?`kVY(UXhT zIKkhYtReA)b8bsyArq^i&CoxFN7qr5g3wIrr}DRfvU{@BaDaPv+slgzcF&4evwY>B z50iEXNJ9gQKI}`YNvF~K!)Y`7Jopch*I)QCxYMEqXhPj;WiQcM^*jcY5Wj&knB-u6 zKrwxsR~q$A_Vfd~7#A)Jb|HF46AysaEF2h1Z!wj$W+YhdBkn$c#tG_v19da>cg11y ziE=kHlq=;jVzOqbYS69(LZ-cQtsz5kJ7by$;J?e)E>N;!@{tMVtH59|zP7G#cgfW%KJQRP$uN`xiU=eS&E@t#Icg_$ z$n!kKd(92gTYA~Ln@2r~TVkQ1SX6M6R217fV=}h;zWBpx-b9`HM zxoJ2#n&11_>R2%;&IuIQ?lIX92;EZ`lOWp zfs0Bs$SfElumn%nO!1Akuc(4zGxbz87N;(!f(vu_L3<8f@~K0dm9`Shdb)w*oaXje z57=MR6tyQd-fhzUy(R#yxR9)o8&d3V7g}4xP{^0^0_2WG0A+5dBj{oByzDrWqr;U1 zn`NHJi~cu4tLJ)H!%CGR?4!^CWH3aRs&yuVQWG1EhJsc*K0SvG|H%iMp9lU)o!Zr! zXxI%%5@H)|&v@SiF)VBwvDhEd)R5lypYGIXtQNSe7u_3nL65Z#Ybx;h!}q%)dqT}W zmC5R;IxXPWm*JUa{J$NqI!?gA3H9px%X5*x_;>Dhr^wG!&|FriW_4C{ z)DaF-F%!gnFK@ab?G+X!DUu>D|YvoONFW&zR((>jKAxK@JN&=qniShFdnTbcB@ z#)T({n}HC(XbO~RrYp=84J%q#pzyB5q2iVq_1QJ~ebtEuasG2o-)rZwE(mGc2$7bY z3x*zvHe=wreLPV;sW+)S0{z{bzd(E6+_!2Ua0a>a`ix#$oKTkz;Nl_3{;jl#xG4C{ z8B*e035!NMi}pyy6_$;eQZZx9Wh-7W6#<2{8A46BTiUm9WM!TMuzO4;+c7D|cC`=ncQ1mjM;8(MM?$P92e>)+d0xp^g{qMm^%EL}QwpbnMZAv`VTqntuRI74 zxN;nQ{YaAl$u}5ii`h076Ek|AK2Wdb1K$}OI&unltsqZ2T@DZXrn!JiC0^G@v8LF` z_6YVkGpIJltEGCLsB!c9U=XD=Z5qKuUmMDFMa7Kg5!yxrs}GT2cJF_Azg1~K7c$r1 z=k>L90*JF{oqW(rPP?tt@@1Rmf2my-S#j_zYs`1`QtwwzDcC&+&yQS(P-IAveKpTC zU0Iy36`jUi_YwaU{}C-i1~7Z2Dtffeq86y^;&k>iFeizUB|7E_&sgo6j96`nc4^=v zHeWj>dT49Am{gw4tY=O;hXaR3e3!f`3XN)xGxlgT`cxYyD=fGoPY=`Q<7cy$WAh&N z<+0MAJHCIYi=SPA1y^8#Pr_9%BmC%cc!9S-vtZGE@R|fL(xWr#C7&Frl*&nPC!(1Q z4RKAM?&TD2^iMir{jk{5eDkfbw#WVBL5F5mNQTs%zM}i$tD8s(s0o>A^V~Hi<9BFoZnxCdpDz9{K;?lKh?LuU#iqNP1!F*pusL$ z_fD{3VJ-{?FY`t zEvd7~T=AUPG2R|EC(iiM)K@{rT1H$VaI{?uCj0ph zM<62*mkVMx{Kw$8d*`OTyOG72lD3*RDAU=x(06fiH3I7=ajAWt!|T2I+4i2Fz5K^` zVo0EXlgQ`Q^XF6%+0H?GaE2ikeIv~lL*?}F)t|moyEFzuYC0-+Pre=)56~Jp52SS_ zI}VZq3bGW~ujS(&>i~Bv98khehxbaB`|wba>9@Jodl7c;91L6oE3+EW2=v9m&iN=H zdGNDgOEye+Ay75zaun3^i3+E^0WDJ+dJjw?h9brOP@tXx0VCKnhTsHCzM}fX0Ua4c zI3b7ppF2k=2(qQSI7l&p&t4y)qSO2MUT-e2+!9x&sH}?VX`pkqQ(b>?o$Z=PoPd-% zfuMcgvc-eXu=x=0z>D{;g!%;dSYyW6fY&}Ek!qvcng?4P5;*vCNi!CLL5A3+4eJ>ay4r-nDCMQkkz%`Dw)#73rD6qoGuw| zJdIM+A526*=&BT-RKXloN&uSGAn075K%+aj{j2>T4WIYt%Hf}0O9Q^C4`V&jY`Zd zT8LUL(3A&XR&_*VWj#`ciA__#N}JoocVP=lBGqgRBIEJ)YHOM<&W&|4U}l<*HAJl= zE0ghA=^^h={5PTT+#4j^!o1r>FX2nvKM(}bybN$PF}jeEYPuLk0<>!RZ@kbpT;%Bx zNF^vtvoEn=Mh5&g>tg&OM71+to^myENRoKhEYl8Al0<7fuA3XK~O%YnhxjVt1 z%};7yuIpCcmg2fYXb=Wo%!4qB8G& z6VdWBrIedna0fC{_W1?a7q<%HNIj)~@Fak~yqE8O!bGuaW5Ksw{V5k_x1hPlE`ryL z8#`K~^jzSqzYa~hO8dsurXwLWk#xR~{*0M`Z)s#%8t8+KoP?~St+icT*9nhdhtS{m7;v9PAy!n%L{Jq>3o^-Zf zFaJXk0jJhNK}lFNpZ)y!w<>ew&qakS<4Uz&2M;GlB=S|nKO}4}q~)E$8xs!XIx5$g zKN+`t>n^`1zY@c&jUkJvmvbMaab=s3(X!5y#FMBLmNopq4?NMg_g0>%3dNd6Hk>}> zq)QMT?ae9zBkM$ zz06^U_YV2wN~T-!Q1yhGOa+g&T|pT-7QNt~E)MFid}s<;B*+q?AL6sgdl<&l&2Bi+ zH)19lfaO0V7+IPAiv%MJ8yCy}A7HFj_i{yh!q1c6v`_>Yjg-;ZS1{0eQ6gPY9kXNLn%I@pPJ3%w^^zB_eDa%l7zCD z7}Lc!@8)*y$wrJVkG0B7f4%A2WeC&bvVfQrL6$)E(4S2s-P}}DgiJG#SB1lBXryC1#4s zJ;XZHI3!|>T%Z2RU;k^2JQX=9Le7|kIbsaSoLp)^&X>|!K}Ob_^zFv&F$s4PsXI z_pm<8AK0Zi7ogI)x`up)DW%)1!$Vr&=5Y1$3o~H5=zi^d{xy^+s%~d3wmY6>U3`KEFo(R9e`Q|xN^3ZcDKyID#9u8sjQ#Eb#Yhw(f591F06pFST zy0Wy8<)S`VWoDmS@+>8~>$Cl}$qvp(++&-~m^_1dOqA#Hx_R-X3H}7|QA)Tsg0Anc ziw-~Z9u zr$il<_uC5vq7dJ(j@V%h?yi$QV&fjC^-aa(%}M=2g4|g6oXwPd#re7+Px)z;_C}gt zD?BOJ%b!-~KJkTW=EhC^?I1w7a_g8}E zjTU%9EQe)W9{|+-bbt@X;&gvvSznJWodV09(VPENm)*+A#x2KeVIW=zhdXO%>faUFt$`T$C4xS2#En=g#KYmk`3(d=@A*JAl zEtN2^!$`4^%j1ns5hwIlNqSQ~l@H+0NR7FvQkEWgqRgUx^gLe> zG$EfCL)r6aiiQuwI7|rqOIL}}uI8Yt0?Q?k5kzfQl~G2Y3UK>-u`MSWM>OT}FnGgN zrC+-si!OWM7O{j-Acot>fZLxZfdXNjWQ8esH2fmh_!au|kIZ*$_>YFK`(E75^LF$< z9NY#{69}ZW`fE(Bl}CX+7%yKCk_qXwos0*6Fba`HwNdzR>=m58$e*b7d&ulMxUOU~ zEj{c?hg&#F9^2q>Tez8Hizq_U@*Z;7rOK#WSVF#cvom?j?&)=UkdB0UBUt{l$ef6)sVK!F=2yEvgVcInGba->qp&hsVvXmP;NhaBATc2f$7 zTPWQ+e~5U>s_#9vK{29FwW=}gCJDvi(8oa8Ym{qQNf8?R?;Xa$M!naI$292gYd~aS ze`5Uh{IWw%sjX3VknpB}9lZw-UA=^NTwKA!2_8eOt>UhuE?+JwbckbH!V#^ZN4pc~ z9hG^TE8J&aCg1%P&h|UHBhqI*GhOVAt1raZcH;|q7Y-4WokGeI2G{~7DZm;D0&WUu z(#YyE>N&WhF-X~9A*pwN7nLp0zN`ve;L0=W);llrzwG<0=;L8IN<1(4%?IT?i`In> zslIlSyev)z&M%lU>*OOWRdoCzs2n{p;!y`|+?elUNrvsVuNB{_wX#lYBq<4vbx)!< ztRMS}qlGH3IYbFVw)+6ex?JT!Cl|WzDXX#8r!-7M!sFate7en#3SL6E-Y3pkH^5+Q zFFkOdcZ)3L>RXrIR!u0>Ysktce|ir_CV!H$91@`4A9Xn7=RVrbK%I0!2eD1F{Q6w_a^}|0(}OZDxOOS(TbpGD?U{A(~e83+<*Q?QiXDWe?pS zOrPG3k7H=QC$ijkj@*k!pHC{Fa!aN44Ix#?tgPwj)$!lsDC;7!y}^{^hZw{Z92_!c zWgw|;>MskMzqrIU_h&NcYZpH2omxBHuH)_+=C6v`w0=z+7ndtXsKq~&o3Q!g5B&nO zv4dmN!tQ}rdUz5+FvE$Tw`Cf{Ro$Z&R18{l;2jpL=1jOAdFT$b$T{*T|5!L%GQGW} zC_)$*tpr4{)^^MSKeilYoX^Ocay;NR|QGtD}=KP zU#?uDQ)2Vi&5c?^v$$;2v=|9{@=?`;!wXmV7ovj)=x3tU?GzWgs#yYQ6KW9D)uaaZ9 zrQ&u^rx@aDAa=#AzQ|pA0{`&*;YOrLWWGy*0pZ)5$S81Jvu>{vG~Ed$`j4COr>R@M z>NP6mChug(1LXBof5lyb#MAW^0L0`jBE{;UOOiLhl(7B%)V0J&>Xu%s5BrsGx?+-K zQ_dnbq?(U669#c{Id{5rpr6rX4SpNdHfuHQuKHCECUtdHx`{RIi`y|tYc@-rThRY& zf7taTCAgK_U7y||>MzgEUgUJ8b7OVmpX|h(HnYte;vMiquC6SDQ+5I2N$QrG%1a(y zyG-oI>VKS1v=^uHq&@Df=ToSgQ@2a*()DKw1G7zIf#tT7fqGkM(!mRA*Y)w}(uh@o zrXMB6=0JCeI+{nNY(RL{mgwnnWRJ3Tox1h9Q?t%(3zNTk>~vbGK)nK6BXFLwenQB- z;@!_9ct8o)_A*nCp@X~Z-9Oa{%1W?-u!o*tc$o2ZrrFOC2<3l!^m$X}xd057CW%>r%dOlg=In9qiu zt9!Ij)kTv44u7!4L=R ze@htMt#2Q#dPhq@0=wC`=Z0i)8 zSvQ!qBsq@sZ zGCMkLcn~H8*{$+#!IWfMsQ5;LOQ>*gUD(0cdD3&Sy%R9Jc$` zhw5S*9w9@_9|^#A<1x%3mVYCU|L1utrv{8)kbg$cG~G2Z*v>)idUW5adDg7_gDm*N z;Wz@nTrN5O*cCFZvH@9cIXGHzRbCd-thal>bKVB)6p%eAwIDs9n9bIEZg(Y~m3n=dA(VRQI}r_O!Pi#`dXw+BX!4wsgOyPD{Vv#lac z9hGm9BR79v`d69m(cVK`X^VaIvYqMgE5HVzfu}T`_?7FCQvi|<+=Bn=KZg)bz(tc( zcKLNOaAEsH(uxRBCFMR_400L+pTT`-cWj^vF{d!jvv$E+fK-)f+}BwJuqESLY{#a{ zPl7EqS7*a@pG;e}QT?pQa7SEOopMUf$*#G<+ozs7x9mDloxPzqy6#RI=aL9T!VomJ zO3#ogbJ^su)ih&}DGa#UKA}caF5^HY&1^x8w)B#6=OYhl?gwVa#XLIYn4y#?2)(h)pclXfCT%stD@Sfa_BVNE7x+LD;$#Qx zQYXZP5QWWVZAIn&?V7&C`PS+VNG@K0+0N|+4nz=!(8#+x!KqJMB^|^g0~@lVNFCCx zON6_EB1e7Mo~3w4foNIbE8ADH+1C>NxE-%70Byhi*|^}G*$A~MHibIjvBWGg%uliu z-V8Rw%oWMfAoPYQSZ`xJ71l7SYL$MM9TV0+sEfHz_ZEU_#~m2KGp?s0f+ zc-NT){G{E-^+{!TgtS1vpT%hryh}y2#%T#_yv&1o@vRBmU-(y*H-zN$Y^&E{|KzgL1WD8Q56F5LU&i->Y;59* zRLXXS1cst)oq~{26u$25QFYJZIcESC!ZB;mxs5pCLB^J(Spf*fK|dWeKqG*y`s{2g zT+?n)?@qy#u$G=KozrSWw`9DDW&X70j1h_hTjTO%U6}_#jX~;Gh)~i9Zw}EB7fj-! zv-nVHP>vg={4AkT01E^gVRT_1M=|0f@zd;{YM~VuY&1*0IFo}6T(T0CvN;`rhLHULxKU=?xc;?s)Wj7z{B=#Ej@LYM&8YvA?6)Sf7>ruSrK3u z(jNz65k87mFP=Eh_03;hb|?GSYO{mYbWVDkP-xS|>rSu1?-ibPfN$7FT{n0T8KFES zoj*D+7g*qbgljgzht@ZN_8_Q#!{@2$u#$NXEwjtHvW_Aj$KpA%)CL!O{Kqa(w;{Arl;?Ukq6M=>LWLU#IB+8p!vaqy=#;UQFIIU*p z?3+x*;&D!B?}}&w*QXN@h6A8>;lh45k6kv3FlU>ivwLvknqrXeH@Y&hbb^DZ82ks| z1&?*MK{&;hYA#haskV4>_I`9yqHV;FsH`ahAwgkIB91moTIQzccW(0Cwrus{i**em zdtG;mciMtX0E>}Ce4$XIyP32FodDTsK!a)qy=;KXa@0@s>j*u4D;lWwiF5pD{swmJ zyH0=RW-$hiS5<3N1HIgA>!pG486r(7U(%L1iKhQzgoR3>SVtg2R81dp(Gw~TK_G2a z@JoLHxkJxDAl3Q}#-_{@P`rbM<1~{9NbB#E~Mf8DPmvr0iRV{P|%Q(~*v#Roe@0#h6#>Ly4e1s<88RX`O_iiQ~xXorM}g z;tSh;Z3%DzJyh6WIbfU#1*?)~8Mr$e1ye@#7@2b$1#X1~JZ-a$;oh`)l#Nhhg5?G! z@g-R`XPl1cBL62=p*!=mnRH#iSV#Et;<^!r1b`FPx%DcxtI8U(JN$BC2f5HG(imzo zxN*kwJZhhsle1c;IQF~@Ysu;RX4wael=zOdt($ldmxY$#a&x)G7INxdOaY3a`DBrk zJXNAQSw6b+K-tw<3h6cpw*7e1{nY*NDNH4B{3a;!5nJW}HYt7ww;%5RRw4}K>x}t| zjkWDi3O|^cE90p?@H0#)2307ZH3%&R4TS)QDswzPVvz1|qJ2c`ytp|EUD6EAl@N?e2mlEalcOpM=2bomvGJ!M3T7u1jDXSK($>{)Aa}=ny1S;EkAu) zdm)f9AzYd%YNe`D+1Yw^krnkqyK*cco(xzqDu>ttU-`b(R!vO>Hq|yGT$$z&?(qSK zBn?xlCI&b#_bDq2Sw(xBX2Or)z_>Xf5f4F37r-KcjlFz@xEB9$23bfD!zQp&qfaZG zRt%Ae;&@XXTuvT)ro_V2u!;%t7GO-(>lqFPL-mtGEEw)}-rhe0ee8nT-iSQPnGznCDRG}v>%fGhu! z9V%F)ml!2)BIFf=pkwt3VnM%Avz%P@T8dy&{8*!HqGwnBl6L^KXiTJ&L!>50!qVdK zII0|ik8Gc#w#jM1D(gLkW@1rcs$1`yE60hVoM3_w(MgrN)Gl2w2%B`Pqs=LFyY8VY zP<4!wJSBkO`cu8vz3m1>jA@Sc_8q!Uay7n|Eo~13lIwGfSF*u9L+Y#ZuDSGS#S2i# z8hKxBaBuxEFXA|bvuNC7_LFbAKLpDPeXBevjKq8a&yjt?{1OtU39$eS&1FuOn+h0T zLmUXm_!a(#V->28i`ybjN4Kga_Y0F<&e~2|BRCEc;PAL)E2s38QZa49Gi(Te7(a1S zC<$iSWdZ>)mQleekAYEu6;5X#6L9hFJ8<*lWQ|MKAIv+p3*k(uKXygZ9vRRI?7yoz z+k2aU&l_gan5g;tugFiI%P0<&OC9L2ZpPn|m@wxWgkQ}b2G!yV7qz{}%>F%mNH4xC zun48x;jmeScMXEt_va1s-DKvtTn%VOfh^wgj9L~|^^s&+&(An`4-%Hp6|1rthsA9B zV@Fr4M5y53#W z$ICVNwDpWoWtOW2K4|V$PEI|U2M5Pc$iB)B;s9Gt2gKeniV-K9b)3x zwo;?C;gqG^FR+Hbt9$-|eZES)%`k=JK7%&~Yxz;Thr5*hn#>yqnvqpKQ^cN!PH7J_ zgfQ9scN%Z2lMfHxJuWJ$V~xXn+_-p938gz=+w@0|g~3`L5YK$T_087zKF$j@fr-RuUxR*xlW^k;d@Bh$FILhWrkLFg`ehzV!`KdIpVHcCkJmf&!?a(gJqbAU zaO};u)%pA*emjHBs%m1B1Vm*L7{K@A6;w70MnCA?`>*QuvA}yDg>$&Aol+s#_Xz-Y zK{*cj;WNNID|@2icubY0PF#-&KVv)n0wS}q9yv25VPf9d==xEonU->tK*%PIB($v4 zXd6(B<~9_6#Ais2Glr3K5`JjmEGW33TLWS7q^ofrE+ z^3i$QRK6H^w+YJEWVW9>8ctv9vOoq-fT_QON?(kqA&Ndos0bj!rEq1y7U!gqY;L2-YI+1@Fd31J((XOv!bEE~&~p9qmDa$QD*lkBxm{Q=YOjfM(*Kd>@&7}j z);JTfSkZo$w`W3uA8-NEO-n`9647z9l)6$3+QJCuer$?W1$DK{ab$XgkS&vy@#CAf z{6;R0fpWgyWmLF?-Bu>I*PBZ_=$GvoW{-eE`UD;&OW=UN*RC=Zb)d&RmdtY9y1a;m z7{qSd46eJoWwu7RuLQ{x0wQHrVTJqnb4uWsJ%iw1O>Wan%|CHt=}U7I7mgB^&8FQ( zX*XEXNJKF>PyhpN#h1tb+qc0JF25iZgR3*mukLD3ALeFed9_a|z6R2iq5AXoo0&YG zfXI7)7m|<+65gWxpS zI#wo@|36HpE3a>%WQg;_SO2E>cUCIV4r96o@w%|)ys@oyeF7_@Dcr&qiYv+@jP(v? zX1%O5-#F1U9qSh}&^w_q{oImXwxXR0Ek(ox*64}Q>vLy%yrG+SED$&x_zYf53w8+O z$`_b#PcnVFl^CZoPzT4()%d7((^a%>lpbhCcv$t4%ZG7c~=Llmw$!SE?SgOJQo@lbD(wfRz} z3l*}HLh=wrE1cp_MJp%;wT#>>wB#Z+jAzuBK9*CfjbqjCM8XQJFzFJcNCi4#NbjeI8i6Cb|)$%#58* z+eXc5Ct$cIdIcSmOjQ>-EK(N^TSe>=S6mU@(q2QUMPKklq}cBFM+sxP>~P!@w~BSM`K4i#kxRtt7>jT`Wc(=WRF`X{o%gc*kfe}Ec z;?6#~ge-hmF+@vib^!~`xJt1V1Y6T)1y?CJG({DmFXwzFAh4lM^HIk0?E4neyM6p5 zFSAWr?yhEs^QP%dp40RL{x2l}fSCyp0Hg~LnotTcXomx!45-)?@CgGBK{h5+nP$X6 zX2=t-eX7uFT$S1$9M@uz-IhD3lebY*$2GB{oQBL@O0ynEtv`_=2__V z&FhaR(X>6mH-`sv3w1T}6v=6SNPS7F09~jmX{hJHFDLhcX>?)43j1bIFh>Ol-!-H2px(ozS!aGeTE1NEF?F!QF|8dg)XMV7-{1*eHR$J2jNGz%Eiite83!ZCMdJb7}0yeTH+sd()YO-j% zZTyxUo=B012B`p0!Rn1;=^cqBBlPkkwl)z2mN>d zKE5b05#+S4|AG^MQurYf7N`5nBTGfuK_GQXg7OAMnE~>o{^JH@b zIwP8AcWEegQ>0ZUJY3ut7@1n=*4M}P@h&k8OakZZXA^5Uj)WW8ro1P=B*xapEA8}a z;?Cyrm_wFBWy{RG=lEg`Vmyvqa7guzJO4_4mn`+mm7#?_??$$J>i)WOW^$!2JY7Nf zmmglcX|K5M`XXp{7Qidkq`WKe$nBTQmxj;MN8gKLaV)R|E-yj4Ox%+F85E`xMFUlo z;MFQiVtQ=60&&kvMdsKTzZji0$=YY&L+w9)YU{{fIFKYa5UdQVG@_MN zBVT2JoopEcJOpACcxPxxV0d}bqhXTm=Q|d`bG;n*MtM*j^fAQu7gn6bv8IOGfr86P zD_ALHY`MCo2by|_SQ6wmO$~gBt9-Py@>R(vOXA%w2MwQW`S$FY7@{uI99u;|qbQlk zo=fw(A_h=b(uAGXB2dLcMz1IMqh@?rLO{f}o2H4ZN6=UEPrK??-XkQnMtbWGWOoX< zcjjs})F%e)D-dy>`1{td6Vd}LHO#r<#_V#{>@#y<_N>=hGH$L3T783xh1M!N|3@RjcUD6hH=m6q)XM?kV>^-PTbTJ#<^W zWGrkJ>422z#6PxV6vpVoYKlja9Y|}FUGdY-9h)jWeEkE2NJ~1P%*LsB?n8?u9NaFqHIo0#wal?a=q*dw_@yOK*hAc6Li$Z{? zr#Ee-p>t_86~ zFL^aqX+QrQo79!Q&Hr%Hf^sYc^|I<}TmDn(n{A1*XCdN8e=}>q1M%jGxpp5XUTDtv zbx&rY!QpdrWc|`7uk5ooBy;J%R5J|zX+Wh3$_W>~zs>=drC23ABNVKL=Z-(95Ch`b zvbm(nM#dQ*izba?c1;U-a#lSGCfxW{GzYeVxx)F%Iy1l|BQ5t(Gpimivk>Y|VV6?& zSX!;2!+MnK0Kq`?J5uW+_2=pOg$D?f$A2~8`tu5zA*Xt*#4L6(f*E9Cj>wRb(b{jD z77?!*xw|D5w1KXwNmNBz(hR6O$j#@39P+2<1`IH;mIy)L3fA>D(T!1Pvpi723szzc zcxD3x1EdiAeT{!U8U|$v6YKzw^g$tY=wDuC>-4^?4ic%{ZPdFi1*kb(;GzGX}yh^XGeLRdnWU`m`K z@Ru}FuXhZnBcHeyXF-*6n%DekgRw~m8(m`09A6~ZBMg0Nslo`Bw-P^Yh=g|l4t7FZ z?2jancbqbu5Sr=(j+`VAxd&QL8^~#IdC+dDGKh(gY%D*r5F72!c`u7xn7sw}q<`h& zR6i0UuE`^mP6$wobV($M7*oS~gc^mRsa)nh6M>&We&JE#e1_>24B|S;al?GteME;f zO=R%F#y#2Y&dNrUMmT!GmwjF}ph9Je7_225Akb(#aCah)R%~Zl{??l!Q zApFU7`m4(@&YH)upvBtZ@($zMZNglpjuT(!ewJUjM|C6pCqe4<;jAOz*t26kTXrievhc2qmm zH+PIx};FfV|#HqZ( ze?ou3Q4VY(M|6?#?+<<7gQR`5Z6gW-1}z!utLJU^7MirK4C1zdVA)<%22L-|hAkCY_r0P=V6*!ygj?~W7)vX<5=2{iNqRrf*ft0ZN3}lb}S=2 z2O?2&DZl<5M+YE$pVCJ(ZUCS08iFjjve#tS1M<8b1ObQOuR92xqAcOJ*0h#6BaRM=(gd%-E1V`!22)j)>> zKTl&F$~L*gP9hY5L)f4tNXhJ(RjbmoLJ$w2zy+sWl$gDigdyxo#`;$iuyPQhmXs-z zDWdBVJ4n`FZP`xlftd>sy6#EnFqs91-|SI3c7`@wGZX^ZV^Z=ueGp*H9{Xs=)7&ER zukU_26T>!gNRh-c;y+;Ml^+_b?VT=%{RDz|5ua{Q;d}WSsX%28We0Faf}uj(D~5^? z$>0)5m(2{j!#d{I1dcAktic2zw44&kgfFI>K3u&<2B7;81uUo=02zrCgU|Y@vOHyF z!V?KVcQ2=S8o1)Lcp>ry7<0}Ca1pi5BT4JH@`qRKsGjm|B zIghzt$?D2>iH!T=lO7+(a*8s3xpGSdo^I?A`#GCErPGelIFqSm=!fEyz88M!#0O$_- z2ma$CNG&3vk9s0Ysu5+jgO(FSs{SfG3u`?bfgm>HGrQ`~HUQ&@ne8j3ZzS-JsVJly z=p{N)$eX!vD!9S*Q{V=bR%y0&l@CGn1~7lmG!!zhZ?*#4-&1ZK{Nw^YDY?nrVEoya zC?7dFxklz=4EA^%CLCqtlR~Z?PeeIL{AM!z7x&wXeZk5{3PL7NvWvTr&GPdz(rx0F z9k3nS~dV^(bx5sp_*9L6kz7UT#s zxaP|sS4<*Ov&p4>ynHDSPFWQUxU)8tio%1#QLP7{-vC>aez0q{dgfM<*J#LlpxqN4 z@KfzrJEt%~;S+(LC~?XA_|mK(AY9`#{%53EX;s1E6TL*f6dV_{KM>J6Nup1C;Y{B2 z@j9Kty2_K~;t`*Ly8{oX!=DH-N)8{W{Gp|nj`8qQVl8HW4H@HJZEyZ&YoV z7V#V@Y;hkj&QzzgcNu?d!V(-?lL0k1kX?QG3?ssHJ5ZHb`+F4U+#joG&-)C**X%>L zY;;y#A%!Gp!xXIeVCg9}9UZ;Rr`whZL3t zBD^a;Q=mX4Mz94{9Y%JN_S72 zFVF1or3^nit8-a^HOzdj_jymI7SR2Fd1uZUWR-#ZHP1+aI?Hvg@ehTeGJ9F{J0laq zfaY>23D#$ks)q*6rS63Oq-=LG;C2mv%|cL#O<~%x5MWA(UdEA0OoABxMq-7b-m;8FJ# z;v})}vlRRQn^vQqW)j5koIluL$LMXl@e-^Iq^QL! zukw;Udw}qc{L#5B@IvJayvaj9LB2(WHmvaYy9TXDrTD2q+9iw-b0+KvB@ix`vTG%? z!Y1nv3xymFk82fLfuCX_BFq1?ZoZo0(XY!3hg<9^{jLQ$3hHN2hJ?Y)}i z6%!z+>7itNg<~l|`ZnLF9&DRSMBzyC%f%gs)An`I)#u4N3{`MRqfZc%W&gVF2Ryel zQRk9JJG97RrVn0gN)fhk_Ro%MOK)7@^@xs2F(riVNVfmV$7LjS50%qcZ1YOwake=E zzmZ#L9;xoa;AN8Jxza_uc25FA1GIy+M6na`sAMCKHJ`6WyKbzLkH_?cg)_mpIc26T z;D9jLW&|n7znB}1j;76fLcqX2+QHzr;`SYJdOWBR7_9Ipe-~AlqXSj9V3ug}f+RnQ z@Zq%6f`=6O%RW8yD{?4>1{2))K&I50%esJ2gTtH>aK?8g9p2bwhXWoIiRv!c5_XtY zXDE6-J)aT@>DT9hh`Xu7K6-%@*4?@DrvCWg9V5uX?yh$vBP))Drin88$1LCa@fkyP zCpW#@ZDSlvsu{ls*5@_lnRlvtym*l$%QvqPAVe3z68?)o-%&HkM6-FW3j zjA%j#BJ-1?xo8G}G_4G7gIxd;&I>!%U3PcTwdwFYv zfJlKb@4&txUJV1eMyS{uNW64t|G)$-Xu*QGC5{)u#L@=o88Qb0AY@5s8MMD5hAS61 zyfpXj97E7jfr|U?k#ESVA#q+|8(imr=AI9uAl)ROf9Z&Bt&8RZK|H9ka`biP?v)y! zO7Pizyvx0>hapIIzR>N^5Ws7>4*_=}|H=f>b1vuduIWOb3@sFCp_{55#^V3l1z{c{ z7b^2g&7$XY-Dufm|02d4;5~{O^uR|O;Xt^@8gNDzS$=(RkJ^A*wyU|BuananY-JiJ7YqhGnFUh)pP)31It&! zr8K5|YRZ%(@XP06yI*h5otRku^7E-*)PL{5j}b(&UOT7Tc3bB0d%Hm|?e2;cFwaB5 zeH+9)UEnyu#3NvN)_x&3D-KWJ%V+&o+cEb{pOG1Gz)dX9 z5$l9x%Dn=}z#OK+C`H>-?j`G*bOX5xBl2OSl3dm-9V*T~m91pGSR6aQRcW`y*H)cN zm+9xHpb2fGiRKCTuXxb=f2t+h|ET5v>t7iY(?8uKy_ltqi>cGUw~e8Tsfekuy~)3U z$_M4_;$&)Q3+1uds_N~c{DRwWx#*z1ZH7L zXt~KEn=kp)H~F*MvrIBuY$>gE4%~o(L0=sdBnMA5ad3{a1+o$BdHU?(2ic$3x%KC! zBR6-Jzm5O)%6;x9Z}0921>wRnM1TO{#C}ty!R-5{@YqK1@wep|$B@-+mp7fsOrQ*b zY5D6eA1yZ&=&c|Ov%-ihkq|*;4b%44$5b}C%F4b}tIOa`=wgZ0T8%c1ZhBgq*jGxi zu4Y*jnN?C-?Gf9Th?AdC!mMZcHxB`| zu=H>sa}J&}e>j=L`+1_U?_h$FKtDX`bciXZG=16DX$&Ct_-P!q{1!);6R>^OdZAej zSGZDbYU&Nh(k^IF#Jj#eJ1#!X2Nr!o8&sFHvlKK7ZR+u}uF#eMI95k{S*MZ~ER4m= zb~j)c*IhXm-@VrdW5REJuZnS!wzZ}6&ppy|8+QF1+gc#5P08He(hlM)cyG$`5w^?t zWR1$dj9;tC0E6eZAr+37jCD*9a1Mz$+cj+)w&YAIop}w+Fi=Qn8uN#?F?YQgKqU49 zu$rM*57ewUQHxRaey^Vi#?M#s#q50|T{rveT{}aKDwO&G&Z|@4kNcizi^WO?tyj=2 zLH(}J4{Y}Uk)dd6N3GOZ7&0w<5OWBeTstai%n?#xiwBZ2>v~A}Bkz$@tL_ItJ%^xST;k>=Z1UqsDmjR`o?x1_kckv60yo!t zmTrmuQQH+ht0r~@FtvFh=1ORRW0;Isw1#Z!3Zx$OZqc`(J{cpmOw@*a9mm(oJwNm@ zbiJAyb>&T?hrz@}=iPRACGNv{2E+=ib+GL8umpL{OQt)@7@h;uk8*G=+#WsH$h&ym z))o&LLIjXN)e~@?t}LZoIN~~Usm4A)9_GwoX9b|X=NHG^CK3UIweBVjachOXM5t+ietbR zh*ZL@KB`l(s4={Sdd?r<7q``mEe2spit7FR&W>3L`2U*V=Jgh$@wY&kFIb=PH=SYb zpUnpYx;ztng4VaI`lA%JFPbubHbT^jY)HJnj9>AMye+|gII=#0iO%NzsHHlrIxnZrnvdXz+(If6+LQ*^jqI3C$xNjMz^p^ zkJMKpu|dc7yOqpafuAJY$Z2DW^eKOM+4x+KKZN%uWq5S{kc_D!wv~;!C*lMXt$p?C zv?C7TIvF4|Lz`M2I_>%7S={s-JpMT+N-@$fZutAjA;3{2TAZ-QD#q77h+Buj3=DAEn6?raRLo)Iw4$?rdb<;QD;{ zW!Myyw6uPb;@0%#C6(4hO}8^yHqFa`zjYJaLR;$6rRi7WIth67vA%cdncC~$VqkQ8 z-EylU81_8aNdj~!Xf2k%Pkb_M#^!4~w2p(=lQ@P=B&t-eE;1}e6Z>n{n8~@N_iO$; z2AJmabc?Onz&CvA+tbvC*fe)S*vONkj(|)BDeJgT_@@_9=R&eyva4WU99MU84R(nv z;9FeazCVWKu4t%lR$?*21e{Rj!1aBoG@Q#G7;u6_E51Z*D-W;jAx}znCKaD2FHn5S zs%gMGxwjpEA#&fF&eF=+J+O3=5*U^G_jHO`l(SaDc2+Z3)}%DbSln3?tPg@U;+fGn zWX(zM&byR49R-S_wXSG_?iX+u_wFhp*ot_ShlChVW=pNAG{>Uc+ALJ_)>~+tV8x8 zWH9kYBfgR$?RE8!PM-|72v3L-Z*Oz}&;se%jqx53b;2Q%;}=O2G2Bg-fb(6<={bMt zxO=M&uV+VD52#h(vp{dS7^w(13vRuGiu!tmJ9|HGS1CW33K8`!%3AbYWF>{(cF5G& zKgXV5%$rHv8C!rRRIN>mQ5%^1p?sb=ws{GV<_59I=kr-@{W!;R!BJ$i^0_)tSexUx zuAmqZZS~ZTj?x)5(P+O1Aqp(w&qB|n$Hyk5Giof1($pWQDTwpV<(+DK-VN z+L~3?;oR`2`+-|GFqEFib%5C)mMFr7=?H<1qV8QS`%tG;NqEtev~#zgsoYSzj+1j& zkQWiE46CLviP7uazZvbh$ar$n3k~!-mN#_)LjpRGTJ_iZM@T~;;OSteE?s?)?{LFw zuW{skBT60TrJ>4nK%PH9>YC=3;Hum!ot<1ON5jK9ztv6UGk9z8+YCBEl-_H9Geb|y zy`v$Ugem~2sq^mj;g#S2WP0MWM(lMHR@g~|Sogh588EJ@DriaKN=jm?w)Y`ZoQA5e z(d>O0ej2d(KHLYuN^XRQ?%+S4vZcTD!&e@^lDssnjatx0RhHz~D)_@p6e}o-M0-aV zE8N;uRMzqfI_jwFTd#>c+P6sy5PhEQ;m2FkYj2*nAP!49_OX}K)mFxAZT0;H(jo2d z_52w`Cno)LUQq|nT4@DkMN6NBH6dqlS%LxtU3$Jp=izwVO($4nm)}r^1*0N#1T_Jw zB<*oP9^)hXVGSgcPFD_ZX`*g*H)L_R9ZJ14pbC#TSy=jOvLb(u9wPhn{sti&>n{-^ zzhHNQm{~lMWkA`kIw~kxP`tzR{H+g+FS~bsqxkZigB`d{h(TqGiuSjtzjz4e^-`*WY#Mt&_JeME8+KuYljBQrGoS9eLd5uYxzI0P#_$$t z#dcy{V~-S>RThNb{8G~@&Lvm;S4KZqhi{cGKicpw0M|U5(|?+B|1&53|4^D382^iF zUaj_Y(w11-&YB*#y9jMEB2mbUNFNI0qTr-yCrN6lTIKk%@e7cskZ6ZQ0GPyN>3jx# zM#Kjby*DQ#VsqB$HaHT1CQWNMskI_VLZ-xa&s)ydBYn9gk-YNv+D~r1{L=DOvc4S2 z4)4wH%v!npZXS2*pYP<@xk zjy=Kg0kWP@E?*uA4{-X&ZAzK^$87>jgBgr`0%;5D??rA_RYW0P>?R>br7OguV` zEk~43m))?Nt{TurE2W{kQdL!F0V|_QREPe)**6aMXq-_WE%FA=nP@)QLo-b#-x-wP z-E2uBoa~7j=cp#b0zr}?YfH9F1M|}E>#ECk>ehj8-(7>%k$|KRNxUa_%|x7@WC<4Hw~$s&9r_5YC}-fy7|`yle3Z?e_pEX~1RF8u z#r-L(tXCW5@|%B0=#eaWesC6;Ary_&bkrO46C%r7LEbJ}Be=a1m2+q96ZBE9o09#| z=ZOK4lHd28cbEOAEeO9;Ipm*O#Swx3lujQ7LT&WLXg?a`I?yHRjrroNZxdBM(f)m{ z4D2mciY#}1>9Zgc8`^4;L0R=NQuslYTwFls7{n&uP|MxU0ZNW@a?aC#EUB5U+DYj# zxJN=`)S_`9{lE%MuQhE0@Of}OHi^Z#OGke3b!T?z8PjPg^-jrlgs{MRLLT4qb-%@IZp>A zmIqvzXiML^Df9h4UsToIYngP_)vqU0tPpily$|M3i}siN`(w%-{5NmTY+hXZ6DA2? z6a&`7FbMk&NCmgRxeB0Txj=x}Anh@i@+YMd`2ImD(G2-LAqT<-KzhIn!Uw1hWGr@H zn5)RdN)%jJ&q`?j zY9~$$niwG#AycEP#e2@oUGWqs;X-xw% zYal$ZZdD49%kVIpTI6d6m~;S1!8Rg^DIe)#p87Z;@g&pHp$BGLK-DIrSZ;o9?t!Wp z5N`l-5A3lciX4g`M(tth!2IIefC040b&vot{?T+*>_84F6pW#K2)^`o;-D}nQHCr^ z0LE0|M%vzt^bl(#_QB60slAD+9=!M9M?~J;h(yYnydyURmekB#cRENR*hH8!CmhNE zek6+-3`!;mrSG|9tKQ8Y&|w6ej06^mL<}w##!3=Sb249m@B|n40VRz`9#oQ^-D`o@ zhej}I0P&Ct4?VOk#KL5pqtn;p_b$p!G)f-f)#;Jxu2JYvu}(0l1M?1i+Pt}QAXWsx zS<;=Z{YgnwE@ybItZux_FU{4I{_7|X-1PpbkJ=6XR2^5u^L{0A+O1J}PQHyHkA)K9lE|R4H)jdVM&! zVOPfOKD)=+$Bx9pYLjs`GayAiTYS0et)|Yw>KKt9H%&CycLEyGSgo|mpJmF0zlvFk zwy6Q{%sJnlUF}FYlxxN7S7DwwvT>6LAe?*UCq?}_DXXkVT4Dy&mv8Udh={<$kH8FJ z=Ax=LfjVE|*HtDIs*bOKE;;uAyhnCm*4817F*&VY0*8c=aW?xK8_VVO8bS}J1p}_m$hBZ~e(%%6f;ZsJA5eY)V08}+^JQd=&C;^!wj4L*!DdUdG}do0}xkJGfKxafSg*9punLVzdGE2m$j-I)q2m*b!}jAT$pX z06;vnMgsh5O}CW7U2in~R#GEhcF#{G-XP;|PT=+NPXqqfF%+C;B2x!t-YlIuXRzx< zk5P#Y+i9ZE6-gHpp0_fSVxq?%ss8?T1B-m%bP7Ch6D006J@mCLE)}gM4BsLFtVQzC1?#bd^flu*W$6@MNTi+7&n$>jP{S&;Am`w^WeQP6T4Fs}`{x8bT zsXG&34YNtUIA3h5W81bnwr$(CZ95%zY?~e1?%0|>Gqcvr#krWN-%wR+y-(G%caAR; zQO;4wRw*4-Cu$vxYl$++On+Zn{`FAy{B^qLu(%b~-aHeLw=2V&V?o?TQBB(SdyfV_ zBKPiP&Z5X%Vx8Sa(x#w1D-l#SE7EwEsCfhyrVnr2HXqfHX(Gq{wxOmQ%kHAVKcQ}( zJtKT5IWP_9)5;-%rPE#}N3q`p!DoK@!HFH_2Q}&T$a{p{6~@nYPOC#93M6Dd@uh(9 zM%s$bj<6?sKqLk}>leQGajuj^+2j+S2fV~1{677`2!UYOJb6W!9|h5re~#uo(S48i zKa7{LLN5--AO*PtczH&(1Gcj1q1nrT#8ehi{W{Ted;tW{lUtM`Xr2wZfwqTO6ilH+ zn%gW`U)nnub2=#ec)n54xTk=T^D6O+!dslg^$3O*NK(XCXo$XqVnqcCX!DTlz>9kI^qthP^SQ^Pr z`)f5U0uaJ($3*sFj8Mup$Iy7av#mR5)eu{Tj$f}3wOAbib*nbfkLZd%V}k#fg^*|{ zq83(FZ>0ee2x?!_IdQM>x$TL$w{QcA5i2}SD}9kpE}>&?hc*GGH5ckW~GRQtRhLQyE?kU;aXnQYx(lGeUmlb|y zL8ybbYQvZTBUq0E>47i?h0sor&te z>+z-X0%Jf<*~?F(zL7PQ#ars(Ad>5Zkb0^e2l}}CZ7Tdba{D{iK*QHjLcL<>fRE zCz-5{)wFy{L*H;QFoN4bviU66g+_0>Rmx@NbBya7hs#wEKXqN%2{Nc1SC%H>ld{`> z3X`!8Y_&l;ULPY9)7fQV$8ejGnKl4*E~tgBb6?Kj;0;~2suI}+bg1S<&{GT{fnmZk zUrqlw1zrw6w2vXWs5CH+(?_t%N0rr-01EQ9=(U+vfR6a4j02n~8u4nnq#E39SOnVX zDLTB2^w-I2SH9aK9qi0htE+1+$28-@P*z+y;{%&NypezU$>Frl4;KCxVl=GEsiP{H zVCfrJE0-&_Qs+-|fFBJp>>^)Ju#=@cuuOVu+!2aia0|1?d0ZBXUQDrwCnHh!XoFu{ zfZ8|KDoEE)3BsZ4e($MD8(I{VN~R%R1MkFaHDjAa1`T{@V`^GyZESj-N=wb_V>-nv zCs#;*LQ7zV{%(7TcG#UuyGcFPbA1SgemQ zzl=cB*vehspA}g^6ZTzI{ohnZG#g6f0sRx(^cQSin_a{}+Hd=JOI18m8QP7DRc@vY zTkB(qIBo;mgC_+;c9oUbgPPe^F>SqrQO=nRB;9$4h}ZXx>N1_7@XQy4m^em)oW1xf zC(7leY)%+T6M4MwrB&00)n5_%`>;9-9pOui>IMQLRvDPEmV_#KSK?ez{W63q%;h4| zFiv?Zon_S_;aCn$9YOjh2$A)4M)AHIgh$ z`UJLk0S5Eo#-7$k6uv5j0m^}F-Ry`yW&{Y25UEvlnElj1L{PqF`DEJE_ZaA>o70n$ zmv+d(s27MdsKe0Lxl0+PDWptnveCLL1>T&XLe8#N0(qz)A@|duZc00Wpe&q0sc5nH zHdK}x&Z%je{nSeqyL5c6KyQpClmo2mNY~5r74umvY z2A-@+3J{O2evYmW>;|j1eVeU>pLI>K<8JJ*pI?gY05gWgB)5C}S6Bt@t$29Dt85c+ zQ`Eg&-2N~RN}mZd__@`(4IO%`ey!?~-el+DU|gbWjQDAM&0KV>jdEhipWyj;FXK_L z)x38E934_>HEXe)wUzGq#h^AB&cy}d>pr?)j(#6Fl(s=%Ud*b7Xt&GR@MW%?bQM2| z$AYpu_{j+qINM-nw{C}kw-rBG^mf2>o-G|(;T8ZLKCV=tE(|8kp;} zteYAGxzmt;`y;D8yWt7*#nxh#sOk$UXdnvtckbot83Y`|*oK#`?u$4?@VjVMh_^Bu z!qOE(Mu{T^`E~|wLEw%nSK>1m%5nA9dHAn{yP$=4XZo%Mgl9r~A2NPoPAQd10T_wrcb$fmlAcTkJej?7w#o6k#p=)C_U#+m!4p!C+!kCZD`b~tFO7ib#Q%b8qjXy1a$}3AxOb!SLJqr{}SXKSBtv6J&QkOirQDKHo*i8wC3Uar_C-(SR! z<9auTt%Ip;RFq4NNNEQ9@zlizDMnaU53mR^338~(}V*@`$|B=Ynw z*z_sWZaTWf<@_xGtlBj3-c+dU(^j|PytYxM)W<7$cRoH@S}a%pm&DJw zOPNz5PrG|awq8gQ!^gy_N}UnFuRv+rg)PwGY&oN!F_{F#|G;}kd634Lf)(X~<(rn4 zjLQx@fL_tYNsdg&iOV1*UH;cXZR`s#KkB_HWxn5ig38}Z)-L2-as0OVWx@~@}44O2_ukUQ^{^H0!LOT?vArnt6NQrpaz{gy1#^Prm=^K3HyLOZ2 ziFAD1bH)Aw3k+7PKgW~*Hio+>>4c$OAGrPgyx+PfX-&7SI=L4? zK`q);wIb&oenR^@A(~UOGgF<_{+JC(In}o3hBF zri%*U@>|OGSMohY2Tjwcnc<6Fny(IEtR`(up}Aw&43EjtLK=$f{us+Rj;X zTSUDfCdV?wG0ld^qg28^0Tf5+WSK#HzAwvJx^U9<$F>^db}uSoQCTg)u-`bxi9gko zRJNE`i)P}G=~7(%VW!R+j+SiI5>>r}-02oM)ztO|u8;W<$*I5FN^XgBTiaiDiVPLt znU{L?=Fk6bc2a{$#L*$!RdxQz+7hXs{E~@2yY5thhaZpD3a8hatYoG5n-j+LV_pSy zVQOiyZM}=(LWM%f^&jN|A15h9wFnzW$x}6gX zc1$B!RT2gmlmue44{2?gYiAFxt`QarH5%B#fvDxEX5xL%4CY0Z?ruz$7#G|@!NHE1 zoJ`cY0qh(>okZx&P{=k$Xc*Q6BP^O*1(9`Wip9!fZ>+U|(oA##%*`0h7&B9~y~CVe zwqYToU9Vt_%CDAA25mq1-`ENV2)S)Gl6EB+2OU~Fp^#+q*~6|&Z|u>@5y^OSxVn8^ z%D9YYRK&BCJsXOvQEEa?#_Lc~{Q6xG}RvYJ52v@S>N5o-7% zoPB`+#ifaf+3?xLuI8Wt&%ml!wrp1=_d%~IFG)T6cb{fd+2;pq*8te|$ARS#i{96B zFeN-HXk}9Y0W9o%I3*Q}bIdT;$}f}?A|VpxqaKI?T1g4Dhm!}8&@VfX=j8IFw_pl0 z80d>QXnj@p)*HvLOLi8c#y4WT(xh$$V)leBk0fPoe6V;Pw}>eD7z-+V?6@W?iQF}U z1-PlwRMq}#56V<#D6I{UQh5o=-+Yw5;d{`~aFl1wHWVhoS>tdHt_Yh!XDYUL1#Hyl z91d(#FucuR=U#1z*tqj~nR6S$I$o%$hZpX?4S#?bD@a5HK#`%d8kiaFfp2+n%7lzUf!1_kwx| z#SYRpga>6hYQZM-(){@)hH0m})Q*Gv(qrhqZjy~Vk zT~PWFNu*pB`OIT%s)iW3q)@z?ZXjOtAH?=a@L?QEws1d1@CyK&f7+0X_3GON&{-djt>RByI)) z8KGJ-uQ5Vl5`!J(jRcL$cKL$2C3vpKW@5M~=4xNH9Aes^gEx0_ys>9_soYM55Y($M z)bv{N>h+tRy9`2^__sr#3#mp9T{+gcLx~AttX{yLR13VG^fd)O9(tnh7kqUF(-OsB znBenHKW=TSLOu`;5?kgZyh4-}+aHh9n}`9i-Bc^t%K+H8xR3gA=3^`S9lahH!s4UW z`{0{E2e>_hS(y8gLLU9F+rxq?XCsuKu*`HZSy-cl~0LxgVd_{Hk2(X zfF&OQqGT%Kh>(ACvf!GWI2UdvGvuwM^uTGf6)K>g4&4W^3`zgIMbn^o zFyl;vaEP=o(5Z2DlUOJ0XbI=Po$*1a;=Fo4M8DR+rLAcD`$ z!sSx>FFW)5$nz6C{!c|6%?!Q*ZcTdg$5sE#K!>>ks8Z0BNu`6gu&XAhMBQXe(l_hJ zWF3e)@n%qVS2o&=8iV{qVe>NQg}crM!`z@mD?Ubtk5>X*NPL-iXFd`m3RAkj3yl^8pi*{^Fj&reCp8ZKV!@m%)r;ib3F0^E@LZr;-UP;o*6!%jr`+Pm{8Vkm z&hi)ds(MuPblav|7bT|tMLE9#N$rhA5C#&%{XT;ky|`u-pN&%yiE#{kWK@ilv8;;6 zieMyqclyDNtnPcwx1MB~^Y*soB3auT6_EY2Yw#iAuc>UjjA;!^jXzU-p)X%3@UH`S z*yT9BJRp1j@yH} zjaOcOm4NoMhS)CH5cb==Q3h$6Yh-jHnQ;{$&`xqsQaZ((y@urwPjM5Deo$qF0_a=B z2J;J=#i+T0`~$Jgd#Z-)ht{05rmNG&V0%HUqXN*IAvUI;UPQu8bcJ2JVIxr}aguFM zu5Mx`6+tR=qegY9IM<3j!ZTWzc8OZ@$JPc>vR7$UyKg+q%a9Ns&K3(b1xcplF%loQ zVaWubKMh|1n|-Lorn16FJi4{XbcO0vnAUBKq^`S1)T0%{dh%}{cXx6h@^G?KdR!vF z0%WHR)-X`2P!w)_E_m5@ze*8FRA69ZO@`NOSq(S;`_V(Wf2wx&8|G@#WN##j26$7W zfUyJ&i0ru@_sqm76(}}2QUwd;^#HS+<=dWJ<_=%8l!Y(}cq2TRo&oIqW}4g5BmDNnNnXUaNWo{Q2w* zGNRGE=7!l!umJ^O<%?`j*pT?vYQ`b9bI~%tf%Jtc)uL;qb|Z`7Q4a+kz9{``2RePa zUq}lrNNH=cm1BR<$gi1e((Ifnt|7Tm7$cB;3gAQ9;JAJsA1I8s1-iN<3_8D~`@;Pt zQfutsuoT75BvZQZw$pk1S>9w;p=KxixNdXPG&*>g$Zffa_x0P#^ za8lrOe~8_=`;0hhv^C0XOz=@4d!ETl%_meMJ???h88K_AmDs=qG)-ueXW!A*06aM8 z(~mN0!-wVoU+Qn09tv<~(_YP0z07eCOHh*KG@p z&Ro}+M$`QUe>gh1|AQheyx1?sZ<~vEo$m@4LhAr3%M?OVXM1q zf?OpOsxc^!Zx30)AN+d52emqY&-MZb<$FKSpm8@=P@tl36l5E-s*|3=6V|8if%cngvF-!tX^i;U-_^PvFuJXNOA1%3vq7=~9 z-Bc0Sz25oJMtH3ZW>+3&2{#eZ>0KiAynA7fR-u{ygLt!wLdCmrT^~;SU7UpXh;~oM zcZ66vVg+%9{kP$fH=|y-=x8psvzedu5|hR3nh&R$d0+DW@WqvVMP+O*?Cp;)R$=$Y zx$&o?FJ0=|-_*Ak3F0((qm9*6D}v4^x*zp2uCxQ%L%$h#8Q*D}+d z&kAEiyp5KK_&@HtUG+tTb|4r<)3MY;4|_D3NW zAetiDLve*0jX1I`qbAV<#zbdZhu$Q*rd^SMLXpaJ@0N~dlTh9JVLKjAf{vsmMt7nZ z-jx{|R}KwxfscWV#!@kDkJcnfK}s@@CbRY?%&2<2T3}JOV`A@Vb`s>R-%p zEk>&Jb_nmDw(ne7C%87rDhp^+&#hPPHW-?DO9Qe5>@P(&R8c!Y=WDz(%lf6bZBF7F(8sLi{*_WM8Ypbov5gvfRmth)Y{jk&h{%`Md)X@;|_+`Dbb+=pfIu)U8Cw;B% zxbz=CKCbhe{e1llJJe)2N??2NCL=F4i;*_y`Izt?mDVn5+vYoY zrPJ3aZ>3bM5NQuCMwv4Zw@?ym^T`yKaTlVU{ti9>M|<{^$T>>m)0sa{0=}-em>hAa z5UQiE&{3kkz11a~`vr+J&sJQnzqGoP2V{yx6#Xpy1Ag$VP>;uB@F%)5Z_s*I1~gWQ zWXHHoAoA>VQLp5GZn~hZ`pc!GBY6{yq}S}8kEh*>;*!o7(^h{~p735my*Q;5zRx8r z9c@`nhx(sujQu=>Pzi?^Fw;*6!wjJdG6gD@_BLXC06pf0T>A0p!c~YwkI>v;wS9z`YkBK~0jCSi zO7ot=C;=J0DSRq#g}t_k!-CYA8p*Few=qfLE?BnEn za{*f3n*3tklar%E!0SyhJY@bK=}6cAl8(s!Pw9y4Kk0}LLH)h^3Byc$d3t_fGX7&C zuk(z?se3!80Yzb;N-+lc;`%Xj^Gjb zkpFRxcpi62aQPb$#=QP-=SU}|nix#AGw&n>j~+%fEat^tqW4ornUN~n(!D3Q@T2D6oa=ms@a_Z9!*fEE+Jb{lYTVdoGFvdcVM$8BG zyjO#L0(BTPy+-5Tv+e3bS_nohu zs4ZU3Vb!IdGv$zwl5#8gky#z%oS)*tyM%{Xu)5EV1z~HqVDjL4j4=Xo@8U8<^yv-$j_c_ z<%+L;yo>FgU8_^HmTbP|-c+R;W|p0ii4 z#=_9d^)Qf83K!%92y*qI#Ftmgp&M#AWuj?wBNCKz z610O!Ytv=x?X}7L!brCTgg9U+0Ez_lDUVZq$t!vKq;QXAHZ*IGT1c$|DhS=D&;d0K zRBsv(5u#S_KFCiT1+@EPyU$_Iw?X1pg4}JEyT!vbbgbdJ=ugLm5T zFC976F%5XNOF44G8TyaP!HYs)cscXgrd`-VEe8}XCo6ubzhpX3Z-ax7z}E(|fZl;V zQ4kWeQ`90$2sKTD%6MVYbC~Y@hR%a$XY5_;^n;l>)Y17?oI?c37@w{UmHtk2ycs*` zV|^g52me>!O&09`CHkcPnh4UgJD?ELf(-?af~�Gayr^$wOMng`ME*5FAnO#4z%ha&lv1TPrT zr$H=~dtX({4a=g(1zf7fQNvV@bQ`~gheRgkyFmg1+cf?>MVrBqXXcRw3t}6gzXvgp zg%+_QFqVyu1bWEJ1yH*ku_XcDz!VL)*jEHY(jqpzff9mK3ls71gfE!0XkxG~HVnze zor%WzNzXxh+688lRl~obb>QaaAI7D2MWY{JJfux*9;O?dee#v@&Vp_KwK;kA(Ti7B z-W7!d8#MXB6&<_~2fY>|Og8HD6{$v)y8PJop{CiKI4zMbA(*SFChBbivYPk!p4zPT z0VFF-&4q0UT{H;iiQPIJwo6;I9P#6tmlLm`eUf%}lFKAL1q)-T|9DCfVm#!yA(o6N z7v6!qJ+gbAga#li$YvC8)IwH!!8NZEVLyqd3RQgJ1S2=iA`?K_H6Dq;UlL2Zh6L$= zi8q9&?i!m+kWF&!u=s25-+dyU2YF1t^;S?fKaH7x7EZ?TTRi@|_7!|pYU$t-W7!!)@jJG6yf@F2!5$b7Y z0FSQ?x-tRSZj?c~AM~q2R=Znpl3@e8K^5$)c38Mrglji?sA>lPF7`$kIuR1L?Ts85 z;=GTka3n&1sBSKr=Luc^?YOq#4Gd^MG%}KwZp4Mqc6iCj@sVFge@gK)0{6Vg zUmd%}Wo#qN8Sm)$!o5f)#DQN9x)5+w;my)pYcZb|e33u$T6 zED(wD-52w+8gmZTprIiojh7kkkkz%P+y3S&?E+cmsXxsVv%fbFs>DNU{n7DjpC{TD zLre~U)hN)&m<6)@)Kk@lf$(h9xWTHsN+To=G7z(OxGf z$R$e$doVH~s~JZXE8h!$f#nRz7PMGgr9_2Qv`!MAmRK~f`!QWpYb>rHZsnpQ>NUN{ z19vsQ4}Ot4RqEn-d)nnrwz8^8k+|M*N2;-MOq0LlwSLo6)LPVxg5187hc8W$}H9+5m(&$&t z!Dyfyw2ud7^A0z))N~k$TnQoA)_>H0A$#t`;Ta*D1JpxkE%e+B8xwc-ToUlQYKJR$fl!O23H}?@2TzK%temKBE{5HCv6jfsVT-AVNvq%H*9yw&#iNHK=hfPBRzw*kdl_%*I zB89?9KsExsly^YhEN}7 z5Qk5Q^~6kJ_>LsKz38QwnA&J-i}8%kLluE$l!XL!@?8&hq_;GKs$t!3M8>luW4xjE zu%|o2GiF&@)g?xubyrYD@6RQ=xF2_e^ZY<%~nj7fw`0U90Pgxrzm!xY1$Z z9tblq>+#QCVHTT>$`H^GZvxzBEPTK{5bt^%7MF+)^Hw=uT*0VwmP%G> zG8G<@q>WllHW_N}&@?bo?xLvDh#AsPctiYk$^Fj3ex1vQ#9aK|ju%e8C|gQ1XEs(& z0UjP1F}_1Ix0Mq5ziAh{n~Yvfl`WGgnocbO=1cU+JY>=Q*dd-xL-as7ykNv1(+1Rt z(NG7t;0%okQ734g0T--~J2$}08B-Cy{-26xKk~SEOjq%Mt*6}FHZxqMOeze0=k%#% z`?0}Am96)lOB^y}M5r^V3h(W7Wdts=@D5!0&urt_Rh7{aN_@0!Ppce5g^i^Xbod6= zl58_}r=O^Vx?jYf2%h5&uENsUmCSX5%_LZ zej$k6bQYRYO8Q-4k_vlCa4i^l1m3c`wu5-rOuk%!e&UxxeFoX+7qdE5DE@7rg`sv> zk7t$`ML#+`iL4yO;MsqLvYLgEey)P;RV+K5ofwWHe08Q&7x1*YX8HkCr*TUjaDWgx z(a<1PzcwCWmu5s1hDE>DNthsbw$@4db7<`bJ=!O;FPN8$cxIGBSAb4glHX>dUcv`{ z!@6hL)Iq&LY=Sy#Vww}1Bm)vlG~c}AXo((F7tLh(Y#;#ey~Lw?$?=fv)Q!xGL|k& zR*6tJr6|zA>8Nt$?H>`EIi;jbxi2E-g6tav=1r{jN%Aw;h$z6tL9@SfX1rwwNz8zR z9w1ZK=uZ3C1$`T+h4OVbR)RM`z_A+J-YFn@)ddSZq`j%O8_I%vaU@P~j2qJj`4_EH z){0;vqZCjAMb{pYbhoft3Sl+U+x#njhZ5D--{D(6r0kAqY`Kc zP`;IxO#ew?3k4wbnwnp>q(?R!1c*6Yt|`=dYB%`N`vwCumGTOr(_}aG<4&bFRQUjKYgdrWsf3Wu_mDcLP;c z;`@wsP&@;09m8RG_7_HR2?C*f(f#Q^6{(&5bPOY-5LS^%(w{Or{QrU(I8df|Rh@G{ zqH?DW{QB8G)AZz}Qu%5OZ1C3<4%OuUNf4N>oYI@FaP*YfN`E91fRKbj&B`{(TB_L| zij@|xr+q*ck6FN(k^QRQ-+N|-TEGSE-Q4$OmpoBV?VKOk!S_-rMrnfVH(=cLhGfLU zAE22BmVTybCV@u8{!XUyv(zxM&S`d^OoKL(wkO07Od;ht-^LLnZ+_`V&?K89LlFPu z_K(vL#s?7Y!qj-EJL9RcR+kkQQK-t&4-p%*O%t6CigTp*3*~=e6*n$C<=nasq8+4! zi}NfU+O4v6?IG68A)Y^oO$gZ1`Xhwg-OlO#->w{a^&wTm;6piFgM{nvkU!{OPNj)L zI!vYwP6Gr5M9HWkV7c@9O~?5-`f{9}m(g>>gW z(e?ms83DGFg|(9Ia|bA(k{HD~RaTRpoYPzB{8YJA_W3vRylE4*3N5oOq(=1w@}Xz4 zoTfTiDIGL(>C-KxF{HW(g{X3Oc58&ffV@qC5CmN?=#TJcVq$3O)@!e1#z9LK#Ed?~ zcCK|U5&kI$S>xNt0r04Zc2izd_}@kbnNSnDD>3 zi8(m_AE*V>+LKP3{|dg&bxRLLZprAsz!DWqtR$pbWXF?C6WtWTm*zn+3Q20H1H=m- zZp^0%A33JnFq{b^F=3RD>FCJu3^4+HPTEiMyjdPcN#qVmhQIfB9{*D6+QfZvCOx>(B=0fUdW>}2&^qms&OV1Cy=odbVM?3E*TKhY>~Z6dWVPXeeZ1_Z$fb$N zVd?+WdP^C0T=fmWil2KUzqM+cm-poyyLA*L^!~gz{>q-cu~CA$Ad&x9jg>h{Qt={Q z3?B@ZPAzowb@zL?PYi>^AT)%?$M26o)kCvJ#l+|&&0bIN*y>#)$lYq6K~X~|Kb!K% z4OS?CZy$5XdO>o?^+)Y}eIT4n4J8*D8@e}XXGWYGywa z?W6v*lj1S23Mcf@@c`-fLKIar?*8gkSs%x;PhA&PRif3VWOX_n&GAZJr^XO<)$+2* zG3N?`9~Z@6`;1xN_Ck}e4RM{{8_EApRpmRBJ z&~rN#kLv0u*$ODqs;bu*XN6(E(OU$=+WaDtPeb{#%8k_BK74On*Fl#8&-hb3JnLc< zl#Gw(50+yQx|m#4F_R9uLJqh86MYi>0o}kk*n)sgJ(JB5*TCBVs||LzyrtslXcz-_ zBc&?Jy?$(Xl#De?nnrA^>$*qtw$|>ZACU&4?}Z8iN@2U!I-xRlcErolPyn*=WV zOc(#ud${6KRBwQky=JoyNl}XS$=qK*BZ6o}HfUS*dG^+}F{XiKEWD{x3OWWpl?=aC z*+lz6EMF1(G0UT|D#{d#ZVqemVBP)g*j$i3W*6tll)UbU1%03{A9n2tglu@JqWO2E z^zSC1M#_3=H}NCKh|$6f8D3lRzkyVKcs z^YC%N54!#W{Kn}@z+I$uL@^1h$61DU_KQAD=_BtDWQ>Hj59-j*+c1Lm4@q^IRt`w}rISqp++ajM+IOSj!Ga!-qXpJM1BfQi_a71ccTKxHK{loM zHmZa`R1V~K;=SBsRTF{L#1+0R3IKqgCSrueLp!lrFwZk$BLn3dE&X(YaArZ5BWxU{ z+it>2SI>UF#kLA^qouNNS)awe^u;f>dQnk?Nu|J5k_}K<`*hXk>Z$qr19_`IqY`qW zvR%%>3%9nSV8^EwU^QHs$dSGeGovFFFnCeO0a0t~)@>V>#$JRcQ%%Oth%5;8_*5Uv zl_9u6V^gN;Gk{I^?%0#9Eo&g;pk8SGJ(_0~omZ2|ZJ8|6t6{)eciwWscN4f!-rNmF=0UDe+Ad=S!$J zc?px0$_WIoVHXo|<<%TS6!X-yDNag+{Saz8;|K8YYCpLtvG&1&&$1>Ln5^cMoZ>3i zJ04byy@I^;qTl173NnM;3?=%Z8L9*ckMVHyXEkDOCFUen*r#Hbi>Xx9>jg+7*~xTOIRB%o#}Z48Bf&EUt|+ z6VV3J&4l;s&`OPK1)t+gW!L+60CJ_Q(A+RzK!b|+nYq#D26k7=CjDSP($p`ET*D?Z zhM+TcIzA?|AOyVn6Azdc4AU=VxS^PhtCKG7*nVeGq+I zm)Vh8cWHk)>*0wne9Q_Ya+2BoTvH7IZIwyV#R56n01r)`v`}9Qy)^fQ_9*_b$al4s zHZkUXV<#G1K~^knHUkh0;?Jt3;zqi{v*qD3#RJkEj*nYDr%X#5JBjM_B~X{S4QOn% z)N_zo?bc#ZUsFc%*=)CLKFCe~jR~i8qt(|`^)i|-MSK4n`;$EbYNFH7)$rTs$K)=)H}!IWN}nuikzB7u8RgZ|kZiYUwgd~?VOj7? z`xLSGkwhpD$_ZCjY8m2|x^DJjv<@3oZl zY%r#>xtqsN3_7MJ$5tWj_uTm1PsgI`iKu5Tx5@kBvhUVWiQcgBz?fJwNE9V4Bc z^e+x{zY%YVXxSrxXlm>@W2&$$J`=5AZW=2E=iX$DAsUvJu>&*)ihFR1>imeG_=Vqd z^|!oX-SF;!qs7mfT13qFp47Ghh{pv6+9Br|Zv^YQ^(U|I)0dZn(RiuXcW#E{tdH20 zdcy33N*vL}soW9;8=y;Ny^`cN!qzIwFdK#_j?!ql&~73|;Om&t|9IMt;FC@)^x zbeUX#Tc#FK|H_k9?gy$z=5B_lH^>s^YDxt)A1rBnL=(4&KtGO!Qt=Cq7Dm<^r}*~A zoU_6~n5ZbB)+~hw;C8ssZ{j;Ni4#Dhr}@vcOCM(2bY<<3fEp_$eA^i|CU~r-!zeLK zJnh%}=>v)Nvj2cAO_P6{_wawgMx0Lx85ovtCAk(g3m*(2!O3EkDyF-;ja4TdM4IWx zEqkGb6xrY6iTG1Kn@02e>0UiP?TR!p6 zK;j9yixvsa>JLX4vsJF|4BEqS95y5sQi)Xe^cM4kzlJf3&%RUkTzw{s`KJI{I%gAX zrRq>i^-v?^SHzB?BRMq=pN;;GnD)Fij?_1R$?nh=b>YaMp4%lc|0Rf<{1VtZdwvBa zOI4?HOoiG!oh^sFRMiO5hfDc=kf6_=L*t%OKnvFDsiMauW z4vmR;*mKG;Wu_b7;vR zaxh3400uSFUN1kQ46BK&m9M$hyO)jv<|n`9LX1{AYL_E}CKq`ImG+V8!JEw%wo224 zx;hKv4YM|Hsm*Nn89vT-@6UU!mm(bgWQ*U_9vFa$xn* zpajwiVP%bQ?mAlTYSJLeO89w_$!3?^=*>TxdO6|Uq!*`nRo#gKOrLeRA^B9+Vbll5 zhG4E)D%aiXpN0> z2Aizprfd4O9UnI{=Oc_9DzVMQtq_o*ktWw_6JH#Hf4~cUaaOm(Y#=^G8t--40z5ui z*zjA};mlEsJCNm#R>kfdO1`+Tb?hIV@>r1J26Q0cJboouM)+bEi+pE?^?ZxR2uFA!L44R|%w1&Wc1CfD#!7t&an(B`i z(M!hT09$bjdWPL(>8p2nS5LlDC3URXpl|GVRYeN0Cep)Sx17|SKgQgk7oay>QuP}2 zoO_Ml3!S<(hox_xO?D@gWXMP%=*vun7cgpp!|2s-?;6DJRSbfHz%TzN zQp!M5{{Tnms%k3M>yoz^UMg*i{~buma+Opih?Sx;JP3$69+&0t)E-j*nc^^vxNm>t zroI0N?(?_ShckE$j>QgjjkAH>wg|zdZ$^e3R$~zmzF`B;j zc~!=4PM))=%;^(^ToO&0oy z$B>UnEc|GYKZs?1u(O!7z#wCwI8BKFqAZ)fb#T7`mbav}g&U?xNC1++KkXufo7#pFnj~IthS+;SkYWVJ!MEh+ zY*s8L?wGK84NAXIPMep{IQzVO%GLm)N^tYx7Kh;T<1m*snj4zU#`Q{P1%SqR8ghPY zSP_OorlE0}3L;+yCws)`NFd0|>*qRbqs;u8?g$d&p`?OSs6nL#E7uu{WDj>!X8Bmf zrn#k8eG42n6YTpV)V9Sh>%+=U@Gm_B{~h9(WHqms?T4)gTX9{BCv&*;5WYzXuG*8G z!*eN?B?!9g694;xP0vuE+0-jb?BcHz%srk6_Vh$7z!syA9P^Pyj&qIDAU}h6jzlfP>JH|*))g4%C0u)qnhHGmnDq_`RM#yl` z{}lO?Mu&;hnMicaCk`tDW-Nm2MjqiUpFzoNczcfq;w(sfM@Q8ShpP&{STN0%KN5?e znlHjj%lR1hkD2c(Dk_0p2`vP`8S#V0IhcLX4asE4xafvsCqJlRUwOk|qx>dN+y83j z<{$-Xs*6bOj4q|08ID-)01;GRpJkf{?3mLR6zMcjm@ih7G6{p0T+%b7DKGdzVD_Ak zLwb5$=C=9`#|(VtE+qAq=I#4}*Q3;J5p-rTCv&LjSQ-n);@l#{d`uj4>#suh&Yfv& z*o{^E0k1grOoQNHDTo#iK{#e!B54|K+a24SER^i9LHq=7!&9~wtTCcwzt~(lxy@Ad zTtkZ_?b@yFpJV#o<@(LhWpMitQ$Lhgwhz~yTySx`cDJMn+j~Q)7;s+fNJu`?@ZhAlgp>hHWQwwu_)eJN{E?6A-9!!nXr{~ z2^-2J%?9jyl&vpxd;7Le3JstKL$(7hhqi8rN3>7^Q!wyZZJxC8V*z_fDyB$f+1=hw z)kxS_852;i>uIFZGcuL{Kg$F>gYX|#sp2GD1&|zEhaE2lP_I7qiZk*-UwtySTe$US zHHVI1P31;ZcvM4%zXqUOOv&x(1Kv^w9+}R(8gb7vPEGHAn5S~d$l+JsL{Db=>QyIC zsx9J%Ry2BUlVt8(H299no6=W7hMy9e#y&Vf#}z%Nc8X04`C>p&!3+7R28h=WSXgZ1 zCD(X$4d+oZaY^7hDkxJqGSSi)Tn3ZNE)-6%u}hhKx5(2phS+41ET-lho-U4g`h~8 z{X>q*UDTO{S|c`-=;Y~BF?|6{<3Sm*0Qp9++~g5|6<`4juh+NJA0P@Yuz~-CyZ&G2 z@BfXwvamD%FK}0lmbCMR#6Lyn^#-2DI;~tf!g(Z>KAXwf7VX;{mgGS$R3|(FRCK*y z5Ny!>f%(G$k3G{TPO%ix=^3h&+unNb*4vX1W9|`C4`H%gg)AwY9Ow zcMz9xa9bqA^hqmj*VNtf8)rKTt+kWO>;*_HO_=i$Sokz%e8v6!uJ zWXm<9r2ZR@lT^ND;q}DJ>-`0D-DB#XGhLi~*|;iW&XQpI^vCAn_?J%AVu0@ISScm7 zRa)^TQ!d{%y$BUb&Bfc(TwWXr48t2J+UFajTDC`#MGgkLChnO^=ivCYoU4m#T@I%P zs@{BgK~9dW0Sw=mNv6ih?6m!wc9~z^Gz-j`&T#IL@}7a=dPAoZV-6+k%^^%ze<_vS z!=Hc>YHOr4&1Y>*RnzW%F*ZpsZux%s+Eq)~x;Xw8lc$l@F{{)%^a53?wK1=`PhPkE z%SJsk$yE|M2zr8KS(uwYk1iQ$7r?`F26xHW&yCsB9G^(BHRkkDX1d#S53WC-zFWIs zwE_)k3z|_k#e299mM=~1C5-bjmTI#^X23FpO*d!6Nm=EC@`*bi|ncm>%;0>8`KqvaDQp;UXR!EL*+sw zbE1YhyuTJyW{*uTS0>ceA%%Nen=ZSySueW!J(bf}>b-FB`@gH%I?+}*JHT^_mxnM7 zUP1(gnQC_!zwO=!qr~NAZG&6EE=}iyF)oQ%+Cgg`i`#cF9e9D2QUHvLBnS0!qO}H! zY4Qf}vnb3C@R2}jud-v+`h@6g+;sel45r-3zMii)HDW@ruE|4{G^kAcn(bqZiSU-ztc6)x8gSQ-$wAoJ$j zVTFf4|ZhD_A zjp5neg5E)5!HC~^03!)|uoH3yB*vp^c&nb;%b>oJy{ea$TeRJqBkVPeZB|Z(em8WP zPQ3J5B)?T&u`cN^?qg%<=V9C+H+l5hVT>lwcSbp(zM`(49WO~f98Yseyu0|yel<4< zy<^zVVWzV;0F?U2ASoXZigAFIpW}(rN;x@zvWZ)o=ZJUv{pu$`>0GQTjODpq`q*yz zq)Rb9Z35DSPWFk0>&xPeF&=a$fz(-|@7sv*T@Z-NA=)m?4-0BiVwtV75G8W$*^fHD zGe=T|>eVHGi1=5UvE_i6kNqreXJVIB0lKqYyGkU35x#NPAV_im<1PxSj(*h!{U|H% ztefQ`!ZASu_FCZ&%GKu2G|oM^XJ{V`v8ngJKffPKJ(4MLURPwbYY}oDVcbsM&xzce zFpO|A0GsYBM_y}-og1cIeu< z+n^?l7x3e^oe?^eObUoR_tvaDbZDYPkjx8#iZSJYCRy6OC_1XDQ#j2wISvkC>lG9n^77tEL6A_Gl9J90 zRek#dtKR!rOBoP%fj|(Lu4ekdjobutgc}5eLo}qc>VP&j%h~J8v%e4T72e z@(@OZ$*~s|1V5RrTu%l^HXXjH{?noe~r1J2NzwI&whe45+QmnD#aQLJI81P zNcLdJh}|B9h(@!uMi?Iwc%zNz7TQqMQ6T?y_=LCIc`Q%q87YwvwzEESSl@7{$S3p* zO$wegzTei&oFiqCFWNGCsxg8+s8iS?hi-zCa5$){h=PuJ{XF{GuDj@RCH{b09z6K? z;%(8p%JlBZakAsn>AAZ7q|mpJY?UaARa8j8QaA{t_w37PF5t0n6In2IXcHL}d{D_C zqlY6-m(yESIIexh5fv)PD6swW`7V7I6(L`sNTd_?sRBMktEoNCL*j!n?C-S`;2j-Z zKu$eDiGY<_Bf4L+fDFQvz_s%-_fxI5uWBe9cn*tjj}{%R*__ki_9hx3mT@&`$kMU* zCWxfDc=w^2@}`S~9kb@t=5OuI^$?7S4%AIFjo>=)jo0?dl3gezP&~-kg!lb(4f`8y z2S`VdsEIH3H$_teF2RxO9i@^&Q;#cAvmM`XIlR{0(+=K?F)#J8&6wsBv>Q>;(tm*E=i34$+Stfd5Z%p#v(2x!{FBJPiyYi6%e=c6iF{%ZUF z*~?-5IO%%iUNK`hm`3rlw(B7{U=O_+TT;m_QbkfZik@87uZd1EhMEpS2XhLR^YAF3 z`=$u`r4c-XNU1-$47x087}Lpki_~wC%k`xnzcHkn{#}K?)EM1!RGH9~;5QzMbtr&T z?vfYEEiCy#v)wl9-fRW@%X*F(k&uw;>@hxmr<(`zxNpL~H84C(gramB(vZ7;)Tf>9 zXBAa3tTTR8xiQOY3tA4d@Vr3Fe}pbqw>Ki0TyXl;e!4p=b{G_aUMe%8`StPq{!KEk z>v^sAge)9P?yom=p|FIs>Q-=ZYMCVvCf!?_k1CzfViG6|CRG@~;w}vKCJLP(jQ+wF z@k7v|qE))k4f@(ZJcH=SiHz(1kWbvM`jgM46f;yb$oxFX2!3YqWI1rr03Fx9(q2G1 z_|j0V0%X#Q5JHMDc<2TYoebCuED9b#fSZoKpKq)S#mGbn{Cz(_q4oq8+?a;k%0kT+ zh!Ka#D0F)Hco;L>&puz&@b8K|S3daRceCh?$f;deTUuE8ON<|qkdY+h$jc0_hI&^NBS#`XwZO0kS(yGe=JDa@S|$o zeMlF)ePgoIlD0gvw5#T?3gT{J`vC-L^TgKFj^Pm*vAkD1wV?A9ldwSkTtBnHys9Jn z50FF0}nz`+Rs)y(?6T5W$|4rps!w?pi&qEg@k+45?gD!vd4;i!l=(yRZ@oEXy>UgaG*y}P zaX0xC1esMEtA><<3@~dlW~AAE?+^{RN#Cidk}TXMsD_v+9vG+mMrwpqB_aVpI|TO6 z;HXoId%_BIQDrVChE@Q0@n>3zHkq`CLKL zDyV+B^2W0h<{{DAJ7?&$?|dS&sZGbivnRzXg3Y6Fq}%=MgTBvmjEIeYk&PkZR=KK` z$(mgv%PDN8nK!54nwI;EDpiSRtXw1N(4jy>iiiouc}-HfG8Gzakj%1Ut`*o1i<>40 z>#RwqV2PV7Aih8YqG}p0(M$pxpEAqMg;fKEi9f!MG6I1&;ns<03hs4$>Mp$aRlyxK zf~;UonP>Zh&#wO)s1$rb%#NtI<(2Jo0>JKHNg%VoTsb*0mSm=@yu|-70`sDtQ4S$+ zDBAFx0M|(nYE+C(is()>_Kvg=?JE1Upkno0L50>j1S6ljR;x3c@W%~0Lp~lkzAy4E*x73?ZT1r%=5%FwmOIhK0$`yUK+#yumhSKD+-(B z!0`B`12P8BS#?@9=|L4f8Hys6l<~9lBcd{#1fKglhlN6=(cx>xSzNI3crt)n$^+|G zg+Jlntm~GrS=-;PBXJy45G7GEGNl8_Fl=JT(^fcJLf2f${D^t){oK^F;^yw|{O=j2 zaO^Y-YqIhl!q$t(B*CtcIIC{LSS;e$_*e^sNZ}Q>2=0RUT>RIN!NcHBAWld-I=cn? ziR{>J-%y7X|3@3ffL+_RxLnrn#mlK9(Jq6Hj`cUB3Veb3jb37sU*E}i=F$Eyrh?V% z$~wpyamfU!&~?wg&!Ev~bmLrVcq?~pYUD!r+3^Y*4pDHWAfaX(zw@vp^(P4lA7pQ- zGviL-5r(PWgJ*N1|5?R)h|fYVJ^Ew8NWK+bK^Xq+QGTy8L19vgW@Hm0nrlGB8e!%? zd1vOAkR8atA3yRwr3hoqA5p}o&NsF-4Dp6DMH3))B-?E*?0tJ{KQ5}bFL-1ST}5ETy6TRKjG-b;Bf_UK)hv8T?BPH<=B z(l)sQgDvE&{M?#zlXBSy9G9nQ*dM<><9X@)+rSqj)jIene}~;uXZy3I%NsZuZI?b7 zvB8F7ZR3rNlk>wjxf4>3-3^W23*k$G&S)&tnGtc9D2CBn2ehYV!PqOVxGD)=@07a@ z5)tt0C|Hvi9alakX-7M;fjo)0#mZ!CW_Wg~mBNom}EAwh^Mu+8WG^G2j8Hf1^MG;(Qw} zZ924%whGS6P8%WYx;n^zk8CQq3cZ55<6VJZ^=R0Rhsjk0kBlNw?}vlEwiCOhgOobW z;&t~{OQ;u&-}Rt^{?5wj5n@L|LfNFMe`4xl?mxJt$%n`-<3*JXGbW|Jh8+kGts^xwFtEP$d{L<`Xj0mvNch3!T`9G zZwhSRssL}?t9Oek!*!CQ{Q6_P>$GR$_X+EE>Nj%5+(7vEufW@sg9p#DxpUXQ9pcDN zX=IJ=xOYjy(NOV^L%}kU;u&fQGD7n!5jzQxP;b2u-#A2Bktw{NUgjT%udirqPVOM{ zR<1hl_835myFBAQVR+Ch)&M&85!t(LATTEQtaty8R~r-3^nP=M(Py+mj-HpmRlQtN(?M_KJj@qPT3xZgOqoiTF$JiR~dqppccI=RI`H z6s1x=UtOb_AI8zQiyeHcdKft&Z<=U5JtXsOj+tNNdhFD}ixlq;lp~|HF^P$8vA@$w zW`My3-MNq+uS=pMIxb?LGXmzFV-VZ9%Jx zy0V!@E-AE|sI{}Z%^pN^LL}cP)p72mNUey!`??bxqT2iKr017IF=OwYxWw%lq8g58 z0S5BgrH8YLQIco{-gb2y-F0$Ph)RQGmdsK18O0EATu$sAF)?QkTAgY3w<#Mt26Jo; zhbFrHgVow%lV@VdzT!5^P$IsCvsgBZtfmX6n?dssHo4xRZa1B?T^wPu$7k#3!zb!H z1y9v#r0GNJh3;L(2ge=F3wff{T>(wzEPSDdPW1-_EnQUDMi#{clOt16LvbTU6-}id z9XTGVs;Z0Y$%q&-G2OUow_yTRgt8fjB&K-A*k_>bv{H1f*rECnHy0#`V9DxBioOF) z4n<~dDSg4|m6B+R8rabn~2daY*|hP??{KEqy9^L`{cgGq{dbffN~@r*sRg>sU3A3{kkRYl*W6T?Q6 z{Dy0+uTIS`CbR)5$9b0{e1h?(*YmDSOVN+iLh9v<`;KxnU-;7z`u(z`2Yoa654Z}B znmL1wP}-Rf%NLj83H~M3zm&h4Gsz>5A8J}T2r?&33+*~|RoyHCZny|*?u2$BCL7ZM zxn}G`gh>V1e(m#-kfs3 zbxk!*e#1ndc?Klg&G(BAlUdG7zjt{lvt#2}!luk?j%WEYIa69TB6p_p;iN;tx}a4t z%=ubaZZ|PdEk| zrkup#_OwyX2lx`jXfGcLccm%JK-d>GdIc5uWBm1oQV%GaDD z1oYcspc#+WKi3$Sbm6`KtKAM}mm-tSZp=V1R2M~Pftm-Jpo7#h)?VL3T zGYIf_v{dyX0ok3flN;8F(rfZjstwlTe?Y+B)RBZB*c$KV4BNHW8}I%kL~M0XpRg*d zCBpY+?x022e(s*kE1`s};1&F2z(Plq4vx4D31oVW%+a;^R>Pt-ZUQjt&Wc`iHcG>E zZP$a%P&)I&-7^v*lLrW&=RMVKGNT58uY2?LdJ1+xx71v!G&Sl+MsHe~i@HX40kCT2 zKv*=K{Gi;JcUESjIugUSeN#rEX0ScrFtln1DCn%PlIe{B!QGtAbp$0XN=!tx)^Veh z(>8+PBKW5?8^G6F)U>Z-RF45oWUMbW$s{eMtzMNYP?R_2E5&`7C`24&2(^R7ALPPO za#{wib*GWC8q>w6R)q(VOE#kV3yu{F76#)Gt~|e!e>e51W?x|~Z*YYE+Vt`b&P1z@ zEB4}4WG>OT3>KXIbv$H$+In0>3Sz+h3@*Ui7zuD-9s?I}O6dbax$v94Uou1;k!rp| zm$}w1m3b+4?SX`?ahu2mBF&}(li+v%rY7i4kydS@zvh+8GQ31G97I6 z#YinWEQZIy2}n(_38X(m`8d;9;SB32tiToN9d| zVwm@H;KEv`KD&rhz4IHvZ43cvR~MC=FA(QCnQ&6Z0gu!WANPSJu`x*!fg^C#vj4;V zZ$4~LOO0X~(H2+wrc3L3z4neK7dxWq9LxKKA)Z7agnLn!Xpqr+`HGTcXweJ}?$9tZ zeqlc@a|{kL_rez~SrGfF#IZcmG268moa|4`X0EL@!4k{s$D}fG>)=QkHog}I2Gz+4!L~)#4$0=KU8WCynM%;-{{tkC!NDv{2Y-B2y-S&SJ@vf zu(vaD39$zb8{I>_a=x$aj|23Z(JcwJnq3kVnp$h2W_ke0{?|18C6ME`u5(xWqur)P z7SZ^HS&{sUS%Od&y5KH2&OZU=6pj7HeaHIaU_x_-Tr|m2)4$>~^xoz)$+*B5NB!V; za3{(B7J@vgtZKF7Lt;!0!dtfAF6`yphrCi_+|{bTa@^dq)4YC^Zc8x=7~WrlSKfGs z6ni#2be-u$sg3^jp{2~A>eQaY687M)MMPiUY@N8y z#ialXQl_B>9cjDO3GilPE}on2f=5(Z&8YqkM0;Eec)bXrq#@+ZTpi?L%u6x^!v>RW zW4~qa10+1QhCcjCrBb`i}fyacqxPO3Gx(cAq-HY zGjrTMUOqm4FaV=OirprM&ND1=AfU+r4*M{IfO`}LT&{8IImXQs{5WGic)CP&evlmb zO>ta?%U=s$PkZLz{BlcNOw^(}Jki&I(DXiideXvjB#12##Q&fOfZ29>JO-&zZcy{0iP7LzQ-f9}-O&CRLQMVa?XcyX_6r$v_XE?34 zArAR;c6z;f+tMenF)hb9N2;#aZiK_*(X1$TuZp-jfAxxgV?w&^Kf{CRWuh`_b*+7R zw|}xn7%vBtO|yT%Dk(7Btr~^Wfq*3&!|kJ`40-ilCTV*-SW{fVlhd19jvL~18jlMS z(Q!ZUB_*b18!JR|K;@0~@tKnx)92IS!{5U}_)b6aHHGr4vwh5aUPL~8XAtp2t0AUm zH@{WdtX`SnX?I`s0CUao-y>0=b{UdSYmRSU-6P0lhr)#v&rApe+f7#3q5yt6$oRMhqscm^%y>GdFv{o!!Z~pFN=?eY zg#A=&C$5hPMUGjb&xjDVUCv3Z`PIg=zwh3hIiT-hR|n5Te*lg%X3^>0pTI2X z!pZgf9pnNC7Do1jM+c3%A%{Ta!s2gDll6UkGlEDI=6@{)q)9)hMEXnp+xLA0c@KZ3 zi^4>Px$>vt?WcVI7QI#XViNfG3>aV{0ViFEgh4g6W*(1oYSh=nmlyKL5EG&QXxqvnOc8yrOV;GM zarRPBL2&S9C}R&?H34p*AAJYnBd?*SDKT)XGE!%mBb*A*Fk$nGvA>sXzdo-dfvOkI zzcoi{{;r6kumN(Xa=dp|-;5TU&8k&e|JkP2Sz@uye#`MvjTESs;loq!vtKCe_ z^KN47);Jg|qmSg(M9=eSAw{|A?RU4o7j4B3QW_;xvXfUI8j{_Z%j~M#BL2CwDUFvj z6_xlpI(PI34NKihwZ(7pQYoU!{8NY_XsarFsPZQXz93VG;?>wg#+$x&UqeRKHnf$xm0nfLbQaU z$(`_RwKv(Hzvkfh63Xt}Y1G1q*K`I=-kNi&to#7rdaS22jJ$NoT}8?~{IcrMVr{nP z_SU37V6yG5SYFo_G^n*g&M~Nk7yqf7KbUaijRRVl|<&=O2SY6So?~6#aQ+ldD>!^ zBu{VDam6*_^T8@}B9ZT2wtbd4zcni`!RV7ktoQ&e$B}ZH53NNIJHSEQ@B2+Xo)5}z zdj&EQyv528>VU!djNDX^#Pk^$9^SVtV7>dB2Dw;PTP|jALiVrKTfSwabCemNHPH)F`Y^8_wRsCr z9cze8bx-eRsx(;xXvHutR*PF%7i4WJ$e z!ebqgM!um=#5uq8^SlVnvkkbDjzC#g0vwCr77t2SuJsgsb|YY3+_zX*-gPAvnTF6@ zDnMzDHNU4HNXkupTdal2JDBF-f!L1f3Q8QupEQQxR-^cLsy#`D{qLHuZ>Xgjd=^k# zbQ6G;Pi`V-?K5xXg&lJ0DoPjXNU!OAIsG}$9h3Cs$+!8;?%=clPT)gg~g9 z=%a2T{xII@ZOH}GZ$%+rRM2NCQ^k7_XEC&|rwi@Dg9uVBKP05m3aN!)jHaVHers*J zwXP{4&~wb)_*53-2HBanA8N?gT0K$b2p?bOr&+SWf+qKDJL};>XB9{5EblCP{0`3l zrAb+slWnXzW8KhP2x8RpP|Uu_3Cy2_e?!E#cCe?AHk1pbKk{YyR z946IOj=}%D7p{YXc-a3He|YexA%Hc^pn~H=godg*^74Gdo?IVb_7@oXwXw&2zYZxP%<+2Z}0@uxw+eG27ILG{m@4m%hBoB! zncWBkKi1?l`kAr=3h}YLna`O5qQ60sx#3OdIRt$j3F6{g)%BcJ#W3XcXNVOoBfVi) zj*thE@;{OZ_Dz&4D;eaY2_I>z z!MUCiCnJd`5+hBvp+fAeFm262)IDYOpLKM6miUrjJoT@ytZ!iSPE|^{9jZo`$O$>u zs~`6$$dK!Adw3+Cg-i$47J5?M0Tg%jxx#H83Qj8rf$&nD)VFNAo+N6iAv@ove9CPU zSksqv|CqmRAk>m=tBIF&het7xhpu}+b-)FE#Vk()b3ed^-_xG|i5{~2|MZZBosIE- zriWMB(n%WL3FXHi$!zVRo71*?#&P^P^^Rrpb!Mw)L$F0 zKfvL2-VS_~K=s(No;AE2IB=p8pyz>=)syr+lQds!lE|Erck-|9-F%Yd>f}8+5?o8s>ku7vfn+NKOvMJI|dDM=aUkgJI*bD80l2t7z7+pWApBpZvEoe96S1w}s_l0&m8ehBN@t-m zH!tv1s$t8nZjSez$gs#Ve8Hf3xxG=yXi+Vasq)L@Y%QMqt}kY;PJU(Cq!OrX85srH z&N&_$yMt#aeuoq%`a^X}CH|SyG?N*n86I8P(@M!{O;RYzuNl&mEf#9QFGni<`@hFU zxN|&hEe_B86t4};+gTY#qeh;Nt)2_1*nQK;hmI>R_Cy;rL)f5fpE}l6dak5>LW!n_ zvW^>9N%j_t@h&Bf@+swpnE7;dmtA z$PD?}PB6D^+?P7uwJs~Mma7C14Sk;=cotAn%LeTEi(`LVMno2KQPtvs1UTae4>aj@G7177>7j> zNjMOzu^vA$uLTYqt?qz#ib7J*>Y}}rgI`moETkJ~UyxXK4uCXB@@mEnQM00^c_4fz zL_4dm#R2zengP=fz_|xD?V;CNJanbw11{T97nDAdv_MLl+rc~Dk8`?oJf=qAw4UDNkHZmn&OYaR$fspSz z4TL~E3YJCFXEO?m@OwkRC%~kkDf)16sG(UYIYppyXK{t0E(`7{DuuDxl@{b?M}Mn; zV&cX&f6}F-%BDZz*_Ds?=Z=a9On}9EyK1Lv!era~2zkfXz z%)jS09xJHJ%%Xy^70>gWO4V}fDuU>GO0NRjp$U#-h=|2m-J-(HS%_#2g-_bHzphxd z7_#1Db*r}&=wA=6EuHW1|0yb?h4m42=i6fyIGrrT^v=BcQxVPS3EI4Fx)VY+(L)Qu z!nfN;?*YX&!J}d9+7c;S%*5u}yv#ZS#Rd1Qy!WmJHZPMy?>;S{h2L#J-uMnIeEvxc%uYE10`$wu!5 zWU|V*?rAL2-o8#NzC!U^zL-0N04EvB7JqzkGpXX@QS?^^cR0Dco_?T~s>KDk>RP1x z$Df+0sSnFpiZcWB_RimY?*c{XR)Eh!ZHSdUOZT#mM$9GHGGqu6Mw-_b+KtFPFaYpZ z8wLAAqW)e5U*$`dD0auIj!&=f(XdSRmuF`sHCrtvFAG4~LAiVi{d62)3Al;|dkw%B z8!A`MXuG7~$!`dUo4N390&oP_0r}E3e1nH#;yqLubyu^Qa%d?0NxK}}Re6Jmh*j9= zMA7{Mj}t*M{r8`T@gi<~#d*XZ4oqR2yP8TXiYhQkq~tBdfOoyV#w_6~SKN2yRXz1& zCaz5)>?)?B0!eRXC9WOCQ3QYPNlXUxl%7%yURdeXo$o#a0`Sso8NER(*~?TZ0zbpN zh>_}5LFK>Qr0o0dqrajkd*xNR%Wi}S2vglHa0Sfwy&XS6bAMp~I0$(#%ppyoITq+? zg|gi;8G@(h?zY`d1(nUKHQ?qbT);P=Z;fI5V>vGWNaIU@EP zkYFn@M2Hj0$clJa*@5^qH}{4TV=8jp@Zf?)X?X?VarwW&F~PYa2q(<|lPA_|nk1EF zE;1}gga5CP>or4wOOnos5;L_IP)NKn*N^Myrfk@8!@?=BuJ6xJaIU} z*$%OtR`$iEycoj-ctTJTM}kC->ui%bt}zZULh{hXg&Cjfm=HL<0@+--K1KGb3bClw zMe!nXx)dsX>e#G1W&aVr0rikU-y9%_is{Qr{Wp&T1pTVYk*Tg*MNyZb6^I$IUN2ng zM{w^3LD$jsj>LU|kTDWjHPuBK?H160SltV86v@KwV{71A0n%XiQkBl;AVDk|?Z#Dz zC~zRCeQ!7@2NVqWz_3~!)ZE(JkKc8~F}Rw&FO6y5a|B%2JBlU)LuJ4)PYPLtC}4LS z#_rdb=GEG^Bi~FQm}Kih^r0nhIlU81_M)3;-V7Zyjj%-V-T8T;zX8~Ba3+0AS&awL ziq_ezT9D>V-+0e-Zhi#-0+{LCz?z`02F?57VnsEFsw! zJ^>QUij3636N$^2IGg%EU&NRzb29mN>6S|a_HJ@~c|enMFSDkK-$tYF9TmJU(U}G|2oipsx4~5si7`>rGr?(kt0f zCIqS>@n=0fWEmu?f}mBnl(sQ^QPN2$vmE4`rf5@;xxSA4*S_>!>Dn?3L*iKUD&I*C zg2z&s+d_H@@SA9`7$Na8Myjl|=+)$3WVIo!&MvZK8X&XxnhG!uU#}C^Am#aEV24JS zj4(K{?R#^?dO`ZQnJs-b80&JAp)ntJ0v&*O)9LL%BP!W!_8l;zFsBR{qf0ZMU<}ss z0OLmjZkz=e!-8Lt`p$aq2t#ckM4I9)@XX}-5p-i+?FrW~1n9ygoq=$h;#`@Q?Zmne ztH(gQU1#-Uk3iwg5=gy@qfP<#ZQQnDpPakcMrzv!*o@)a{v1U&-(39fM^!XTRIhk?oTVvXN$u%vHYCz)Now3Y>gen^`O?UIH7@n_` zd~igl^%+Qruy&n~cW0UI@IJ6B_>^NHPly7%lh&8DQB2Gq#VnS{<;p?K%#{zA4k@o1 zz`O+XLvaXiL~{%*x<7(}+3XHgqNi?Qf>_9^9 zO1m$qiV)x`llU}}^oxQLgD_72#*!ROV`4y91!ZN&G+Zl!CZ_i}J!f3~)n&EwEBTPZ zFN6<<00kBT2ny;aUSc;Am-K7TJ+5YtdsS?9oH@uCkLc{@VCG*TG4{xNG5_=TwP?Zk zwJ6@L*NNzxB3f|<$gezD3HQ?>9Fl+5D|uhxhDB#ZA#=!!eQ0lKw8$dZQtjx9fOs6; zEJ8Ls8a)K1fD9cwmG=sR*@fE#nwh5L>@o~jElU(kr^60U&Rj1nh8r!6uzEjqeNV_a z!-VVR-){DhppkcTzDCMlpK4F4rE7&&?y|1!CyVEkNOoZ`n0)y%?fVk6D6NWP+L5NZ zQxLE!4vVpcE5V4k9HDOY1Z_257(2hwumi2leF@+HlZjO{q2-_| zF%AR+zkX}n+XvG{6w4lmkGg6stKHiR(}`k={`1y6rGH-m^}h}&LK?4fk6PVvxU;(m z!jrYdsE`!Z4+O?wwA+bu8=UH4apK`WqjHH+Ph)ySz%Tq;)3WnR4N%BI!xK`@^4qcu zQek%g7H)G&KxV=i$n>mIOg^58g~!5+@9H{=K+Ta3<3N!vUv0%pA~T2p z*u~VzA$(BjX&=%_lM1%Y{+8r)ftC)*MEEShgIC)PR03S>R-;)fOv($4*mSwyFpED&@ zG~{8&UkvM>^840KLv4HRPt}7TpVdv^N;N_`zE<;9bkv9<yz6JA%mV>%lKqy|3&xOH)6tW&wVIWP7^E%`KEiO8(lIR zTqkA+O_lq0$=0VJe#MR^F;`qox&^70-V})2Duk7&xD1H()UUOWiXzhcHt*tyX83V^ zzPzXZQS>Vj%b7>?9^&YS zth1qt@qkJ@A`8-x_0HciRXwSo97+FMN><-VaiTiG$|hgyq3e_Ryu=(vq+%KKUVQ?T zY?xkpL5HtEtTuJN#QOT|FDNpQ*@b??c=&(<8I)A%LF6_ppXJpvG?L8r3BQ0j99F^f zh03?C$8(WJj{fsm<$+Y^_?Dc?+}619ecLOw1ho|ZvmdsV3=Fnt% zLRi~CZ%~dCIKRQEzW!*B50>kTUaX*ec3U$L>SHpn?R7fHtqm1(o5Vqes5tCr7>w$2E<2y zIwcj@58O!&!ZAgtgH^G|Z9g>djF=O)xbEi$R5pFfJD}QQExO*Fxy%5{#Ie z2|E0Svr{vY+5ZwTA^{NY%uZpIEI`z6xF4~d-6K&YM9TII3{}FYSBC4O`~+`N_^$kf zC)8ti$(h66n6+>GpWdMkpzVr2$@8zDg5wx8O-}`yD&xQhjD;=U>fuo^c4U>k-NNtc zL5|41u|26|~D|L_Yhr^{{M9%SD0 zE`3v_msc-Np=h$)p%ac~FT$)wN9|iNIQ07QKz{>s!bDyZ^?QgD)||O2C9NoO`d%kB zb%#gR{uU5?Vn4`!5V@??Rf~QAC4VT|YPGy|5B@?%{||LFYp~I_PB@)~PV#STm+N5a ztt>nQ*TJ!XWB3Mu_Hgt@H4&g#Hx=Sc!P(eNOkR9TdONt)Ws??1#nDVEOqWdapm8{H zjfhG*`4MkR{QHXKX@(~sgm9)tl6a`4uUrX_pFE6EeCyD}v#PuzL$QsAlFXJv*-bSK zkc&0;^-^Am%7$FBYqxdX{deiWVWsaHJy&|p*R7hOj6Ep7D>CcOIa^BvQ#_P@6s8Wm z)s4(~YUI(3nkxI5V0dXPY(vYbt*jb#D*6_tdWK6Emm;kE+Ur!^=Txe$Wc;u7Rn#B6 zGDJ%fya*JmsHDpyLdtD0S@xpK$jR%5k>zK2!YFye?lAn{>|D2KPk)n$qHOtie|6NZ zkDa9_6Vf-Ar!H0Ntw>wt6Hw5`Zdh?GD}ceb_hjL?Z>kAp@((FWk?YYePgS@77h~rX zBU;p@+tMl9wr$(CZQIr<+qP}nb;`DF+g+XXP5R+aZoll;mF$PTX2zc1H>#Iq(6Mxq zkw>4nooT8ORvIneYSQ)YPHr2r@v`~&*ndXZ_PHFqKhM)RHJM4XlZM?6Mi0UlY0+{Q zQj?1NbK=(x^NZ8ww>>mKd~!=`WG4rLAP4MLriwbzqlZzUWj6vyr>1da{K>@(T;M}29jvk zu0$>!3=Y1Tx;o66s7<|lQjdu@BVFDAlkdqn9HTnFRxVN|3 zg)=zN%~V;HHWK|~fn*7-TUxYb>yXzhn1)8)N=&Qm$)=(v9Q2fO#Y#j0pKTbM%&>Oy_I zIRHQGg*@_%9b3%JHj*tlYP8ZsTGT@Pyme@_KEW8sNR$NyWGO*IJ&^>wSA+tyty0az z6(@{xRHqn8sp|U^yyEtN9yYVsLdu~Y4pXe(CL>mTK8Bdy?^jvvB<6>dD0Usq;Vbix zQS%17^j0lr(uWMOF4prEF@&e`r!Fisrdmi+^}J=&0AGtk9bKb6Si?@HGkGfua3?;> zaNTp99oqrQri|btf%e>ri4%#RE~A26n=en&4y+>|SHG*NX1Pv09amT`5S{L8bgQp4 zM(P>|`J(;<2zx*(f1i#>+VVq9MxS-Z1W81LH&734w&!i)T<`L9FCHc_?!c6e)X`iJ z&hZ&+^jR%5@Y?QZs$Jx8t9rWmi#KT5>B>}j_QxO}QBhKzcxOOcRIK&66Rb}6L{T6o476a3DhD`Er z>vD;lF&B)uZ%CdAZXbVU9y8g>Qj{G_=E7R2sgD+&)~dQNu{qV10uUj7uNx^4!YnI6 z{1+DD7fkrKU3{1j(iRayc}cAJP%R>S{6kPVJ+1;XP%x)a@MA>~EWiTY`t1*rYP!v$GIZVfAK<8@*TX4Goo8Cl_?w5-_4KwcLFHm2TP8<}0YX5Ag}`=zjW&4CCe&L?Hz!YXIY zFD*I$rJ%Sb5c1suQ9HW}URgpsNM^3_5JtS=RE zgXH!wSVyT^k~3$aKq!J{H9cT);LcYNI1RU&G30=ep@(%2#@1-8pgXV%L~y^gV5-F< zV>1{Y1HA)46_PBbrcOmmAVx+tQ6hzXsq6Q=cm*Qphhkhak4bguWQ&*Obn0!nQL8Mh zCp-|rZ~lb9!vnxdYYp?P1h1ChmfYV!y{y!RjfDu9cSFuOE3|x*dH7~nS6?n4AOxRi zF-$(ty;4?484DG^Gu?_ zsDnxjYZfL`CT^|L(sdv=K;XqzMXMA-6d`>x;>*SXpf}^t?$_V1YnG2@OdRnY0(#Z6 zxef`#x_Qp*q@V#)x4y0l8rv+#g2TP?f&bU{r~wMHhP|FjAQ_p@;E~JN#~%d19r=t` zc&=CHXToF>8DP}3-r!R|k=7Mqodg`_GbG9&Y3ZM~?1tDl`_^UW)d;?PoV{;8>SgL! z=wf)a)MSzAQg=K6DF^|J8p+keGV7Axd&Pso18$q>)nA>safk(knKx&@6(!*CCrW@f zZuZ3hzIT}i^hyMhJehitewvY>-j6lbFoMQ)2-yiQ#@r$mCa^zGuYWzAkQN#Yhiu7m ztI5V&kPCUAa2bcJXA}Nq!fiI}B-3=*ZGXb_`Jt;v2BzCxm&QtVGzqGQf>bQ`3d#f} zi-5U8Eq1uT#&A8)`=(3<<1sD>qXz2kkw?&US|V+^v~)EE9yUm4GgOejNAIB3NL%(} z#a0j!3I(~5MsW$55H+g|ULU=Kq`QaAIMf;&n$zcGu)f1EH(Ai7gD6z6zQa)a$MdOp(dk4Qpjtlp|Qw8L6E&M_# zFEJC^o;`^HMdR>B1cxXEIGq{blK`7Mu?<%MW>>WIzQKvk=tQq1ZB3`Ua56P?&$pW_%q932zU%}}^sg&i+ItVb#Qi%t!$7|$sGMh&=H_JFoyP=pyV&mN@yiyn^N7@tZD?==d#c~4e>n5vCkKEPmpy#ANY7{FW ztg_yd-0*+dO(%A3%p9z#eKAu8#u)!hCs__T8}PK3sXnD}bVGNF>#m)16%iv=Jew{N zlR;$bG#Y}G`fO?I$jzg9?cX5`O|n<+lM4IlPo+r)x!k~Q?G+|_;!D+I7@||Sb+MDp zPwKNz7tM6EPTIfjGOh*`ub*T?NV!A3&jz>^T=qj%<5m{97kSSVtbUWMY+c@@@CK3l z76tJ1=$l7;6NA9W%ebWeBq^(Y5>%>gF&{J^cGRkvwkP)vC(8Gz;g=K4#PIAbW0 zhl*Lk%hCn8`gr119ej3V&PjlDl(v{%^w!vIc=}ZUr#)4;v5IMNGjPJ^jk=6m#G!1_ z&}9iL8}Qe^#p&ct;#h2l#_{!ybRKb8B8NFLx=(P!&vibGHD$j13f7}QB)z2AzMHOY zk*hl~6g*Nm2uO06e=TV_junU43gP+H8>T{C++831N3XL8=5H)?%O}b!yBLTc><)GlmoJhw^>5zpiv%A()#0LubiD4FNXLGQ>w-FbBx zzwMLXS$Ry9za35|k<#gYj7}d`c#aV*b2|Yjv~J_QihN0@wfEo(kUB{l`hk?&`32R9 zyl}TrG5-?_ZuOR4dm~+BWGSAYc#g3C=9#hQ&r$VJ96F)Ipiqf&P_ratd9ioU&l@!_ z6$Oy^Q(_|8MrrAUXR~Cv7`Cq%D4l_1e4GjdIJ`(-g<^97GF|T#2c;9A#fBKQC~uf+ zkhr+M8xd-X5&a|XeO~a#o-CQMq5z(!u_28wd>XQJbzC5m=j|XUA$V>iSxj7fDY#T@48?CZbR^)VA$hL7|5wB zfEHWB-|AYP^F<;214w~gUYEcfmm}bM$#@h|JEfgQxUeH7o=JylCR7uj0Q9c*_hIdI z6pU>CE-CBvzFg208Q)O9gVFo==W+nlX<6cG!j^_x3fSceH7b@W7nOi0u@Z+y5^_(G zx4s!8`n5$frt)(H#sp4hB zg{dOa<3?p1i)`1E@^Y^DrzJlpcTvu@DOedZ>q#J#m5ww`NSzBX3zIDV!zrhP1f<@*kO z{Q>{@^#)%{w^JiWakFkbiPYoUDVP0LDPJldV9*${^juGzalVb(Qx(zRNes1=|NVJ{ z91%Fp|lA^qqkiBkT#P4iIynZ>RLduW1S2=~KR zIbDC@`I4tgw*sI0ywm4zn~RNoc?PzXdNojXe_N_(&zK>E6ZrhcxhL`_KbBJwZn76L z$4YdU$BdvQ{mwMKcq^6m^qvt3Nvw3aut3d#rs=)Ccy&~93<-jSRZmd zxr+AU78B-#ZsEQdIss?zqUCv-xT~^o<}yB3Ha{qKqT~{$+o$&k=;?`dz&>cW)8D4= z;-KJWSSjFd8~s%bum#j_Xm5K~#s^m4_u4HIPuO@g9Y5|aPfQrK)Nd`bhl{~>y&#@B z>~`BBWkh!9niXzulZ$T$4qbw+lWeVf9%lxJ67E&knwjN72}Nw zny{*Mx&T>R=C>%8Gw#f@u2aUczqHD;#qkD~?q$8mvn7$Cbi2{#j&JQoW>lI^0kSbQ zCLv5;RKTbj2raTewM9L8rtZ>201iT#xDaebBHxHukBQ+f0&Cx@iIqw_q)J9IQ)T@* zSluq$GNB;-Ny0pp zm-N4d16ZR8TTVAi?4)&>13|Nwo^BRmJk%4*6NEmsI6Z8(%mJu#h`P1_ha`Mv-_|LO zX~RRGeOZ$gw#?%4DI~wnfri?r4y z6Grv#`=+lzn8)oOQeTXiU*~{LGiHo?fgttDlj)lF!P+70Z)&x&B*&qXIyE@%?eEs# z@#H-h9v-guGjSvTeiY)@OK_^`0{N;bZS=`lYg-}J)9b0r(??wnDRGoc=MuV;yhb1> zZ1?N~F?^2SGZIHWeb^h&nP#%xIj|+Dr8Q~B^C8_!yW5WS5GQnCRRlI(_9J#EEM{$s zqDMFIWkZ4smlsLpty03_1f(l8L^fDICAz9d7b%e$!z|XMPa}=avL{U+EXjur*~c5< z>(rcU(9l#K@TZ8X-8E=m)fsMEGKQzsBi-+MouQd)Rth>Wc=e#&iei{}24qg_e6 zyKC68E_AHVV@0y*c7=BcdfA<4UKj|}LQymnaVWkNRdkzXo3XI>;|CZY4xO2DwI9Qx zvl%BSOBhnI4wX!}nO7oLD`ef1LgiNCj~qca)=bEBYs*8Oh6aj;1iD%av*0F-AIe6a z6@2?>RWH!@hq+CJ;6iIgq;9O0;L^;55{WU6b-U|Dvptc#q_DS9!oZIGJU$ez+(Bfb zmQJjZfNWhi*9t%;)QNPiLHmg76d}%_5!;zd>xQi`O%90N6qjMX$C(vT&SV^!N|x5< z1+XiIm3f9D8DkXZ7lwHwn7Jn}4h~xLIiD*}of20Mp(l*Irbd96SRg_`amR|`A#2j2 znkdqf4}e{lsNWTU%I9uw>oDM`lQYBU(sk&{DkCwTJ^fW>?V-c6U^0=R%37%;a50o% z6Ot+g@7z-=MtTtK!kI_a|J63AZ4!|nx$Jfc6|AlEnooAJ?|-UKF( zN(MKF0;~p5d5V;OX{^B)G)uyK#r|pt(#Mj9Ilu78RiIWw!lNF-ZX$TSV1!#$_e0;) zIu!f5`ggU~J13p{QSA@cNP8@C(h#C(ej+8Jct6!D-81C$Y<-pCgR@ z5c;6-N?ezyGtBhu;T#J!r`c}t!OA6c0?}?sd(RGnyT$4hJ*>S|aRS*c4RD zz5sc0>mIwak~&tt5|)Q9Q~=-fC)A&}6zQ6k)ozVM5-PHO9~Z`>JT4_yJ?n8$dy#If zK(|^HEJEP>I25BP^Kx4)hfDZtc}^oDfu5BFnfS)Y%{lfQf>SRG?JSxMUKl)pO0ubf zP1#p=@nVxUlYF}KbDeb)djVlJGKY|TLrt6(A8a`OBBjqIOII0HkVeNS)|r(nqGebr zQr^PowB<--aXF^F#T03GpSVE-gXs_4aHc;l2C${+W#kT;tR@9K>_ZEl zt{hE6DbXewxtb)1$*+CSy({6mAP`qAA3)Nf0**Ul*Dp54FGDwktC%8(0o1H!m>&!y zf0$BBpl+`fiHX7q$J|ec5>Nfk>W?H=MTqwqCA~C7IAypzRcR{BN&F*a%zEbp{sEb> zY_r5PdExJyij8ui)2-O|$y@RFP_ zF!-Z!*M}ZW#|9JUxV_SL0Y`Qjmfgnz)OOD*E+Tk^alRzb8mxF7;aToa7k{^z)lbw0 z&52A@`88S7pnG4|(8Z_gTY^(G;Wi4VG<8BSGkoBRZ0t1|wZ6GeQ^L5qlz^qj;I{?X z)9BK}vyaFA5Fs`9-rzQ$V8Fn{=e=kS7F8EngSLG7b8t zbt7F>u{o5Yh3NmhUsiVwQ$H6KYUiAFt)@;x>_KxR3>#wDT-CWpOeXRYzj5>VJf82# zBF}Cq2xwl4%?^1V!d1hQoB6CB@DiO#mq0qDihO)%+l2-%bPI#w>xQODwU^rwhK@LD z_GCOT>d_)B{q_r;epLBb^g%9m15W_1rlg zg9U`F+tPXqSW#c<>al2ND7J%iFIyXKC}&&X=yuHV43l0(m&NdwswyPAid;<|f~xRQ zh+f{^RKbKJe+D|Bb2XCjaOubVo!nv!`NN}G53W2snX6tmaYuQ|m!F3B^XN#O?H(vz z>P`d~S{CBDb+r9)9uc8vXu$gato(jJX#AZP4^B~49J}UFirO-jSkc3@^X}uX#6oIz z$?a)(VypzC)A||mz_Ip?CwH?D8g)LmoFdRJ2g|}qMIH4w`m}cSb{5FR#ORuU>(zGf zU(dG*fVS7bmv5K214Vz8Z}pTj+6tSHGsRrKxhj#7=DAfQl`0x$5vp;G=YD`Lt@5T( zi@6`TpP_&Ia_NTs?zYM(MwbG_A#hQz-Qflz;!touZ)G3lAMr1ZWj)y&=9v~qqcF4M z=1xE#u_+tG9#S^SaN9#fx$_hZtv}j+CoV=b3pLt+X$;&rEZUV!lD;RcxT#KZ{sZ*Do?}#40z!DEO`5(azHl+)_$&PdT*4VCP|HC7V zxlNOKUXVK@C#UY7HTXaj$`C3(R+Chgo{CC_NAk4QC!@U};wiV`uRD;}NxWgCKL)@l zeG_9iRG_2CpYlU#fnWL`EVdz?Z(nTzB=fZAup2i(5QXeZ8?i+HST-YwqF^aP!f^DJ z7)<3Q7SJr{r&7u_{W}N8l}!V$N5Lu0TDFJ5Q&fqX8+XL7xWvQi1NXG$B2eJ7hSUX| zEKSjt$K;=pm_X?-kn;KkWP&sfRmUWZ{H=D-*>CePiT*ih7%U10URi{ z2UQ;eDOL;byS`)p-clfRW87%Mu*)Em?=n*jVC9ZZ;F^Cml>0l=3fJkeOc+R|(1y}T z3YfqlfUN#i_8L01a`$cnm5&S{cv@i1*>G;usK&#^)6>y0NhE|Arzv&Syp^84sG?l2Idq+pK8YK(x36SJ3KFipQ8yyV4 zbyfc$G0Fz=EG1VZBxPLQnnTF+i;1!4M^g#Pk*1@sN6?ftJu2{F+z+Mx0!{z$^-`}S z4=;7oY_@1|?iAi!*f_P9onFDJO9S`(WTc;L*^x`VVb!7Rx*4c!X-D$I05udp1jn@B z2+LCtOhIs`n+7Wy(4klAnaF}4kfKuIixxEkBOD9umI~3sIf;g!c8eaWc3L{vdLq_6 z2jmOiGnciuq|R|V+8!NthXG3!x#Tz+Fw1v3c)Lt8_-(~EV0R( z@QQwpmeD?HM9&>~D`-IDo;lClIhF?omJLd10$!RM<93%rz!`n|PLUSM{Sqa{^%DIo z66mZ6oV4zSxp*>e(~2f$_VYca0T4EOk{@f5wV(G}6c@FumO7++syPtpS-%~pxUdDU zQ0VE-T)5e3BL#K*_QaDy7QMZA={;KVpm5`${kN!Auwh3~@jCJo_2?*6&q3 zB=RBjdbndS&-O%fDdvbO`c~x!|2i!cgNkwiIJmJCii(XfR!~t~08u>-GV0-~&j&$X zZwV0TxP1P)NdEbAT${Yl4#>S4sW_k1y32gNtoIMrH)Rs5FiOxQg+xsWdK z-OaV$qhTRJDcTgMud#jT!Z*vmb{;GWhOh_)6&|NHzEVr?Os!{5gV#JJ>-{N83=k&0 zBV5pYX~$~v#|3aBS;kYlWmk!JMEs9<~9{gIk zH-=@^M(tYqCKnQ!PIpeEM&)&k>dCUb9QC(N=ig>#?mKCn|I&P=x>|_%bi0%_g$5yow7=Oy`LT858cL+5yoH^R`|@P{EagN zX7QuJ!Z>bgh2n(3^<~0@Y7So_KbO7RCVHsROoBCj}(HZB93E z$jXTB3S*F(wY|ZxGv_d|mkV2AV92#>FTa04NuuYyOo$kgenUO$(&p8)mT;Vf8jf7`0aIz?km zRi-SIG#Bu&s{{cK-q&b3BLjWVnHu?~UDwh~jK;Kd8^YIUzu|*ayN=bF9kkrqfES;Q z(01!l{&?D?iKZ^TFZcISk1ENEHL?{$Wwyy9rRuOkQNIALc!r(C>h*}jKy?_Vr0~`4 zcb6?}g@l3*Q>G{zzvG1J_S_QEGSAjU>VTtei1OSP@*}C=4Z6B^^DXT;LiZ&*d_`e8 z1@70(GsFf4-d=&5?t)>aW#8;sXda~a1;X3P}(Y-zdUFp(qu*& zjo&#pBd82pIzqEoXFS~%Kac7S7(2Nut>pGpy8=}+UEA&zT%|UaCKbXGqNMJVm$KZp z&gLxC#d@Y1A%qn>8EPubncLT2-zu!}cbYARSN@Y@Q)y5x6R+j=(6zB`)55JKx)He+ z9BsOp=uo@Y9KqlQTj842wPj_f)IsmamuHYu!=;r7b!pR~VAMGTE%^JxStD?=9_bI$ zOG4TqNX?XKu&o>pb5dfLd=q81!7}N_T-_qmeGKhtJj1$O&DoPne+Sg;qujZXTylBb zeQ8Tyfw7}|%l`)jXR6{Ui%HDFJdSZl{9({V8up*QN$j+ch*8ok$AIN#vcC-4Q<_1j zr94#Xcd=*%(oJz^2267}E=yW`lgICJ`Xb#Sq-sau_(M88Jf+OD84|z zA6TLgD+gV=bs$n(n&F+Hh1T9r8+ZDBAkp;OHx2FA{g0pIIFN2qlAXe4v?tRJp4hJB z8(i+FqH;19jsg&j@Pu)#>DWw2+0uf`(Ncp1zB@$VI4%S0%4! zzp8`tBQu8o1lrjC7id#*F?9B@H^HY<{(UkC;nV$(`+$Xmjs5>0`>a-na92|4=`DWT zjBGHcix<4p6;n{&1Q0L~?qdWGHmtV}q!wK3VDp+MmdPlxaB5)r=wY#dOac-TK`tW? zge;2iM+RAR_`Sx@+p^hH*>19vM6zC{UQ%v+^fGRwZrWe6ex9{#VHw=v;}a8^ima~_ zNaVP`hLG0qUQ8juOh%5+IV4PX9Okz}y%ws3~>va$w3BP}a0 zga@KOy(BFs*zT>aH3%y=$oK4qf8I)|Km?>~%Bw3+W>COI9)=UtiZ0N`S^rJ2*-~zG znOJEICyZB^EKRKNeh&>vB<5x@K9!f2%H%RU%#DKY|0zgPm+lfzLO2K~8B#RFxTCMn zQWxMB>k>vrh()9d=NO{gXB*Pq_b?=6h)t72Mv_B>3ZobjEC_OsP#5NwnIyc%zb3wi z$Uv944eJb35C&w(Ko>$rgm!>q$U+z34p-uj3rc+@9pyA%#TU9h-Q4;s24_2gm2v`5 zybE`@Q^t9|3eSGnxzV!u{IbEe>FatbHCPdHBi|b*$oxrX|62S9~4+vC?z^NvdK%zcx<@$?+0C&Y3MeN1j&@|8?8Tv*3eZP%wJa&3>l_wU2KX+I3fJ}# zVlBB0nLP%N*LgttX~_6Ny~Y%3%x)I^Pe2JsBx#`g?(x7A05033%uU`YZl$7uDryh- z1j&U}nH8AjwGi1nhh45v{I|C86!oq{()kLUpe?=24M~hJ(Olic&iCqwyST8-djM9y zeIYvqoMtqudGRzh$BN&({TMabg7CC9kHvlzOX|DxWb=vRlLa$D^tP9s!p@Wa6#HuL zuKwt~YvdGR3G`{+bWw1UdIiGm;5xaK-`n~Ue$4AYZ1fBRgHTc+~gwhU|!QykrO#5HTW+-ldgvo%23Y&5H%U-59R)xzcXW?fSZO ziy$%xK_VA8uyRp9)$Dnc{EW1y$L>6!Fn&gPqxO1e*1W#D)KC?Y6xTR@3_& zC9Ec;4H6r~`mXYjcn7!qC-EOb4Fr~;jqV;0r}!c|u_t!qcdF{;GOELG(9;ReZV9!2 zYLvv(UnqP9k-B{G1Uz^=p&|KTC#4TFMe&t($T-3WoF2VLy(L?4<&Eaz*hko2( zJdx912#&;Bj((MC7J!wz8F8-wB5K2nCI5~Gs%Z!zKENp+$&CDAE-tE&l|EOXN%6wrvkK-7ID{J`}6~@Zjah}X*(Es~#A`Km7iLYB$#@a7k z?p=EX-2{8-a2uvkY$b&)mlL2q6eli%W#~lJ0SFhhV|jp%3(v4J(G^MJ8JYr;w^0;g zVd)gEbGB-KBmX;Ktc_Q0SOOH^C4mplI=rgoX|LCU5NDhROaiP!AG|1F<4S#jm5dm! zwlqzMVcV-ei%#A2S3K0itBpZ^W4um?J`}o#T%sbHNaTzb*ix4{8 z_pslb6T@xz@P(agI`E1lMq+#LYxz29VaqKbemo1ru&zp$4_4LTGv;<&Pw_gJ=0^VQ z(qOc_R(`dDo%!Z$(;i7v>#hy_i)W!jy3#<*AcCbs0n{5LwgHp`9;yaVhkO3k`uq@R ziATv^Z}4vt-1}Vi74z{UAc0 ze<6W`nk|hA$jC>Xy&aheh;X&!!qEp zT-FEJ3S^cOC-5zD@GUJ>BWsLq#cv8^<~MYg}D$Vv@NI&-gCKt!bHBOms4SqCH&Y1n#Y` zr|a5QKO>ioO;fgJ7^iY7+|*8xOZUx}_kgjK%7#Q5Xv_UJ$cJ_m7rDb`Aek4Tf37g{ zqN76B?9&#~*9azN>akKjcyU4&R!%faf|8#V=pi6nO;|LZa@m#3Dl05DACK7!I0-W--G|?z=8LTIdCMnxrXLk#J zQ9)YK@wx5K@_N(xzKh|-DA5kE1uH zMbmxZVYm<_%5{cnCfEUd80Jq$Pd!9(aTeh&LKv{E?J zRfreUnB|3ddb`Z0qtvEz+wwtR$maefo<2LvmfQ-`Qu0T7l%ZL>fg19q_|9xiZ9Tnq zV-H7*43^|=Y+pp;UYp|3iUD|1Fs)%GZVcA}We}f(I!uusp|V4vb{DftrOVz2l6mg% z0)Om&qqLmZ6~LL!$12d)P*aWsyc8NvuEGU+niPtM9YKA-tb73MQY{g-YhhTL%=yAU zcn~oJXU@ckb(^DukFF;`v?tCu3hjXKY~j>0KISHjvw%S&)%VG^9Q8TgaEkMMT{OdO z4djdGimp5J=G>EZo%9EApeZK`dL+gbqY(B|_RoUhXk32~<)OH4t_5EOSR3I*v9*Cn zv!L=(zp|aYJ={{ghMxG194gjzS4w8o zicLw#7L>-~afXk}9=YMxa)s)Cx{ux|FuzRkZd2`_AK9Z^8$4FR_q+(=zzMhOC)#(JkXBAvRrPYcGhk`W zMYb@LGpy%i?w?_k=1AY)k7pcaCDhrE^>Y-em1?cE@fQE&I`7)==v=NnaZI>=8rF;P z=7B2Wvzwn@=;Zl`E#$82l3Bq9tz}X}T;O}v=5tW%o%yue%h&{RRIQ`1f0VzA>#S7I zp6v$4I!1*>{Y!CI|KJ2VS0YuCt^)FfEgHQxhSx^lWWK}PWm`f5Ud&6oI!=9ox zYwnp>EjLwcRN_2e@Akotv8lQG%{rBPZLV!C?PM_x&voXXW$LmhDRf%PoHY**L`#;8 zp3k;4YY|H$*R2mDHEfB;M&Be>iRM_3RV~}&i%+AeyTcoulf`J~q6>z?k;@l2|4) zxgUu`;toBS3-J7IrTrqcz_uJzV|`P0yL zci%5n^ZZt}72?k0!Z!7up1@7KfzdVUK5$}Iqq1CR_k{@YEnO1B8kirKzm18VDE zg=uF`*;t7c-sKk6W+t&tA!2dsQ)-;v~f6E{v($hvVcHcoL=sw9=67 zMIY^j{@eCIi^`4JDE$tW8RrG>Oi(D%{1ukDiso7}W20V!rNuxA{!zG)k`DBR8$Su| zbdLQ3Y-w`~tBrwWa|<+F7l=fe${)dJ2)y}h6)|Scx4`#-q1jS#9ol5BK06T^1$H}Ntl=S-p_;}LAt7Lnr@G%HjRm} zPQk3HIs_H6H-rmQrQQDeq;PRNd6@1&I-vH6)|K{2S)HAWa?=tDd+TdG);cgv_Oh(h zcuGTwAozvXdZc;e3OwvU-qP7{2lxdS1Jj@1_;FxBjf}sGwjVGXYz~Td6+4gSipgd( zpZ2xZ{ig4HT!|DZ2TMCiMdFWtU`Cgu^k7YU#S+285E;`8Z{m4b#) z6F!iS@}W<6Pe+kMe197G|Kglxq#glZs6uK)wij7Vvb+-f%DYbeuTO9CfGqWanJ<%T{ik2gydngdqE3s8&{2k1St= z6`V~l`{^eg_Rht2Fl1bc+3>;aJG@Y!I?f2%B5PmLM$>KRF0y?nU4r7i0ni8*##HQ} z;F_L&$5NhT`*g%@?^x5bqB(M31I>sH!13T+aS6_Sm7hMJP74Z?w(nH6`9Zxbwwbn| z*mK0inwraP?T`9m2JI=wvTa8{rekg|LBz(dp{97Ktc^VYbxfNwwV;gfeA6%2&56qK zJeZ3D*TA_d9wDu;MViP&IXf+-JYl+66>4l%&0{Ici(rT={guP;%TC&)k?6 z^GFvt``8a<;f{fe>3!<&e~IaGdYyH6#RwL4%!?($JprXT2pdmhPlhC0@3y3K zW}HQW@Y1R05A6n4G^4Fb=lt$jgGy;keo}Esexh1IV`i4b%8#p%XbM7qBnca#?p(7W z)s;|J`%r zcC?^@Q%?0h)zNMbti4~cy=-aAf<>dD?!vJ-Ya|UZ=q;{7e*XHB$Dl>$7vQKT^4G8m z!;A1V1h0y>p5BGY#81j>MAC6~?V$bCi21LZA49s&0`U{25w#|aAg>=em!9?A4lT^2 zn|g5bKj%jnb2edChGc!B#c*SnhijT1vuk#nLec2E)_|%ftAxoKW;m4sx`#HpF{}ZA z;P=@f>N9UwnPy!9r8&RbSs-PZb+Eb}I3Lu|m3?Y@gLpWb5%^`YRgP7Lv9kOp_0gnY!YA@RdPq|Wmn|jl5&<<_gOC_$Mxemz7FW|C~DQzvAO+(`FRI_{ShJOY2vgRJ5x z8%pc%Jk8or%L+~cyU`XJ7gODrhfQYM(99v8C->bRrQylzh!rf@{ebbeYevivP`>*= zsY=fY&KFcwe+xZk4V&~K?)J<&J>~GU)S2zhvdkiuv)GsafruA&qinos#WVPinvJwq zDPX501ETi9oLu|HWELDqMrjtJT9$j}lJ~0%bPnSTJWon-&Ow#VGBcw_KY4c9q8GG4 zP&GD)6HXTq3tz8QU#`c}0+%eXhJqy~ktunFP-Q zJyOV(t6?g?A2nC%g+LLcJ`6hTV3Cu(0xGccpq+HWOjVkD!*^FKHSO+_H^uPxEBX5z) zQAHbx)lpW&uBbY>a?@48i3gAg{%+?7wXnCUN$F; zo2{UsT){9^ly5veG8~too;W(=-Q;7T5yWyfe za$!kq58UjbHJrnft+NH)?WN|jA{z2-&%fya2KuuQc&XwI(V`s-Q$3l{E;j%6*pZkn z!R;ZHh)7A^E|z27Xi#D{%U8z=2Lf!_J2|OK0|2@Tq!oqCW^hX*g~1u2){62SIlST; z5zb^+^sv+00EGbO#~bmDlwk6m%_hgC3yoJAw7sq~W?nQn#BbJ!wO$@0#d>p5s-+vw zX!@7pLs3Y+G{G`QM@bR%rZ9;(i4!>=l@i1dK?`j|W_>zq^+^^jC)~y!E!eUBt|wjk zS0#h{HRXrW(;Md7-=~GcGA=}l-N}jtuc`|HdNmaPDhMdUZ)?gSjn)Xdt+}s>ZkJ%8 z95wP?fNg&1rxdd%Q#0uV%e?W2y*~Q$y+x>%sr9JC3>HTa{s+Po%Y*PgQG53PMeY9| zq^D<_Xecvgt`0W}(6}v&cdS)BFChpo|2=T&J=J4E6w_^&v%4!jl5C>Scaa~{h z3dJWl`3c@|?EJc43>oT(v=UhlrrGOYB&Xh&oC@nfxRii4e$Rq8rz&8N1>G^qCG9R~ zSK5GW*315-UZkT+pI!RA_)O$5OVfCr6KRC!;MJ}c{F~)6>*D@~BjkO=jAS?)iFD9b z70QF5$O&2#3Z8?-?Q$4#PSN6D9ZEB#T5GTCnR}asq&G`$KokFD%K93rB(BeAMd`?r zemTgLJmKc;q>1}hZzWYKOsr{GWEz)32x`jJJ|kapHYa$MKH2oPv5Dap+KkZy3Hez{ z5Qmn32J?I}9{A=Id;Ba zlMyQ)j_WJT@{om^7u+Kz(#%;I7sO{w?;w`B^OO=-d%!r_8R|x2PhYnLSAfkas$|pc z@p5}wCF<2?DsSkdNqOUB(b57pWMzkC>FNGs2Z>EUkMVdozkP^yF>nNH0#Hi*BM&eYVqJmUkjqGps+xS*3)fz)~Yf`$;ckO;}4&lc}1s^6fG6DQDE$_ zsONtTx?rLj%wsY>h%UIy@?71O5A@Ugs*oTVWA=_pv=K-Rm`&m45H1dC`_Abu_=hEF zUqO-TGqyea1?zoWERKMOVC>N~`AgUIK>D}kL&tTl5TVtsE8jAAcfj~iIGy!|%iJ-N zm4|t?3gHZ*Eh!Au52m{Q{gy}I*cpx-%izBXl&e7JpIY-+azvutKZeNrD6au9u0nv3 z8?b)N7;+<~(4dF_eGq=;$b;bZTgnrbO51nhvXOyef(u=H3fTG6Rp#Rxzkn#rk!6E5 zvqPBcNL2Ok;;A6(np(x;?h+=~sdEoULnd@KI6I0gk?#kE8_Nc-t|!e_oN)x!p`=S^KNYg#4JZl6$IKo>Vv8*n2WSBM6*4b?EER_ZV@*5;<5@tCZft>V0!lxf z=y?}11peTOM;O)m+hd_dL=L(W9o7&JH2f>f+m&-RN zm&k4m0o?be2$%_UlsjI7Fxs>|(9Z?<$CJhE5pFwh<1E@KeSvpXDf`>)#&XAP+BWRi zzkJ)IdOT-GC6eZvLu@@in`~+ltyKj^OON%d)A7F;JI5wbpe4(eZQHhO+qTVHwr$(S zE!(zj+cx^nmx+$(H_`Ja@|;+?bM4_?V4Ptyz39JasIQLYO-JQTlM;(7te*gs59eUs z4Wg~5$rOG2m~l0dDAz-dU~2^B2Wto7cd7GSeS=Z8{4Q;4>@1yqCJ8 z0ZI;@bqZ?9+FB^(N@RZqupIUUY0^j|SkGj*>pOVu+-Ukz(AwVGdKi)?2*Je|79i2` zHxcwa?>|a?n(#RJ_g8z<=Sv1!fM% z1f(>HM>O(c5-$-J_IbpT7BUlj?HQ2wlV!{f)W~%sqnpp>c46Za2j{;aYoE&WWtcq#^lRVvn^! zCzA#S1D^7EAw4tiEsWx_!tz;K3L^QR`4yJ0n^RL!q1S;csSKfEqOuY{p{tviMWU2w zf4+RXrjCjgG`r$C54XS>Dj+nrSd;m-ffdt5!8;0(>ZRIXAblY+Bpe8He#eL<7q{8q zPYVJo?yb2_5gX})5Pfo<2qyc7^hckS`eY%#3<&3c6I1LZkq>ur@bkw7x`735^YV9_ zOHQ$Qn@gfFibxNn7@eVd0B&4VUK_sVh}k%*(XMXFUJ4xUS+Jg(`@!98zvVLxPBi0R z{2B2LYGD#of*}txB9V)?1hyW~_uq<+rEI#(-^j0iC%a8P0!{+u_(a))@Xhi3G#6K_?M(V* zcE)p|AJ9%=GXEAB?$QwsbLxwHFFIjTU03^;o8S37`PLgTQ+O5G@=aa2Il%Ul1+o90Q?ma4hhz!)fwAQ`_g*Ld%+9s_Pw^<`?hS zfuPsZoUBnPD11!8Fo@V(s+s0ANiA@q`xgEhSEE;aa0t@r#=J^!T6u-wm|p$M67!fE zuWVV~L!)NzEQIOncF_zD>5g9N?K>}c2w(h8e`+gC?wF2$4ZkKV#a;CeI-96JqI}f# zN$VN{?YoCwbCTBmcDhBp*nlutx+%+-rk;8fPfHNRBPF;lfzH(Nvaa;xh)@m*hV;B+ zy{Jx=3Ha>g%+O>pu1e(NaYMPbHnZH+JB4QgoZP>yF>wFjW1uNebbo;Yj%5gndKWf9 zAc6w1^*}c9aQTIxba2q%mYsP7r_%yJk<&+%ak^-IUptECdOVBpS?3Ua_I_tInwvq6 z?IM9)heRb7i&u($OZWJwZhnc35jMjmyFH!$v$F5+HK?CZffP7M7I8$|N0@tHeHbk! z8D4xfIt)Do%W<_vv%Y_tE}~(508>*flz+PCRze*pj+}4|>izztQ8#7kY7R^R?JFp1 zOVZpQ1`NnY|9eDP19}8R;qrpMd>@l1Zhs7{lcyH$apMzJimpz%X@Hzm(5M>Q9%8PF zWaxSH5Sw^M+{$Rz2w1u%WYKZV&~wThnjVGXJlnVL$_*6R=oZ5`oS4QGW}d?0@_MAS z#Y8!Zk%eKwxG3qS7lAg<_8LJE(d`eFf4yxUJv_0JV0Cws@W&^ug(HoZKuhUqG4nti zko5fbomT$V`sz*$OdCpjcfTxo^LPf3*)k9)p?UJ0G4KF)zBZ1p=Wr#pjDW#Bh$`3M z8moon@MP~L?2y5JxQ*{^e4{oXu@Dkr2ok-_)VWQLzMoKUZla!N6z zYz#yXdOG2t-QAZDkHsv=Wabd|Wgg;&n41v?=#6Y`qsce9dfmYUvex&G4_FepMb3>V z{OHOupNgLP$^bDXgcBTJdJVef5veBu+5BRCbZcGx!-UVz;<-UeYwt3Wne=G zxL!jvi+68!Uu$dPJI%CR;HHU8??s8*Bf&$Zh9(BQa?1CW(hosOxeS|2Opcwo!r6Xa zFCS%JWhg(;KEwO`)51jufa-YX05pV`k8{X<;<(Pz@}?{65qN9kkCm2C4X(oN7IV2C zdi2_hjlp2*$5e5WlhHtoa_N|C7A<5G7J;+V>{%X~HlH}BQQq38)B6J$6=LrF=2s{`JIdGoFwjb0JeAW^n|$6- z>eCo0LT?LX)vN9x`e`4q2g|41s@+W?Ik*JK!mM*)GB=sGNZa>&OBC4*lY`#{3PSF;Hz0s2Y?sO~eCv-&!2*r2OKQ&t za0Ic1MqGUHdYN(*d2QKJA+K{IJk(lBX`VjnKc18E3GuKwv}OkWYoCaOw@A;v8yJCb z-jfW7X3OHDSJzND5ad0>TzMzqSm44}8`+0I8!F#irzfPc2jG9#A(#PN++jqeK}MuD zkJ_qV2S6d>QK1HhHwJ5dShcRxSJ)7(tW;vou9d~&%s5-}Eqoe`qN1;IX9m&f8>Q1udTrpG<`gMjyRGvy-&Do%QEUCsN}_K{P8oFsCO^6~Iwcyn;ij#f6m7?Bt-dF26(#5UJ6US;b7?RJiIrG_ z`!Q^Q01o=Ajw2UJ!lcZdIS@cH*P-uwn{ZO;RguHL!oP+;xDWGX(3k1f7^-hN``LC`Bi)Y_N2bh$7xDTw(CLwu`S{Ru-E z5eL|h9p4%Vl;1+zZX*PvLu+Yh-sgym>`#$wnqdY!aGYJX{s+>G>l{H||GmQVl+!)m z?cu8-NYL>OuH?@N5Pz#Uo3nk(bcjWObm6DGbgn9T)@*4jC?v& zXZGqg31k$X;T_u)ue|uuO1pVim@$&6)b?vw7dD7bc;{!QCem>uN1M1~h zOL&OMwSOs9XBu%spM3vgYX)*_mZM_wqDQPIO9v<>v%KzlRnK*%xg~0N*L$O4llfQ>!y{Nbc`yBUy_u>S#XiM2*|ZKSVE1hm3HHjx$T zy1}Kwed=AF+BeyLGCR1j6zrerP*HgMb13Hha>nSQdz8YYLp*M7zL2xq)^STkG-wEY zX_=Fc=+*m%xbiU0wqINWy7D;a%p`BfAGRYx`2g2bM-rXk^{$aNM0gj94lahe7JiiaQjoB z>)d|{r|~395;a(1xiv1*2jS~bh((^zf(uwkDy4=L;LHzI?`0zGK&n}v#u7gGc^E0X zL)-q4(?_Jm-A+pXPz26?WHoc`2!OR~Gq>ej{Y@$D&Vj)b`bOzF^8JH#hxR7H1!|r8 zs$T}0!-^38#b3LqwOJ1lD%`;$g`No#a1Cf%PdQ}ddHT@j=gJ|)eSO?{QJTwV+KT^iJ$1T>a$r#tU@A)AaLUK z&J?gK<}o79`F~WDQAOU)Zcm$tWu5<2lpPp3y7)QX@kU3?_Q9CCyLch&Z)5K=?u241 zZ!%WA9eOzRQSj*WbaJDsy#vy&cdpDfsiq_e#d^CcoLKv;E%0%3g9`_ zYO7l8@bI5GCdCUAV+bxz~1h}my$4`G*Xp1~Eq7_{#E?%ISkKmsAiLk# z(D79Dq&yM8$YTb;4b!0+I7#;MR%Nwl$W1;am%TPK)`j|{ z{{A4+=W!x*RSO`Z{|4|nO%N1Ap4OS?!a zKW7I47Cd^iONebe2s!Xuy9aBb#ZO-6gQk;$gUgzEw_A>2tx|*?>&CE(J3-VD{R|MI z8|xU7^O|#Dg&aO!*>+o|g6>Zr(rEZ#w6fBxM+jwwEa5*7<^8oS|iTeNA2yx;nLgM64bGTY645`c;lE#Y7gvIQ=m z2T@1kc2@J&5p@e__5G@FiNFS@z-va00irGy>Ca3%mV(`$+`{-FDbJvR9bZ{|Y{+1!uSTNj1oc$Ud~X8`d8A z_Z3K&L;eoNoAHK=YYgZ60xr5CPJ-Ssw(n>X4$HRVa?JyGggS@CMsJjDpe}4}gwlcM ztnovv`#@K?$CIl!NU`NL%B5?RICK%zsLi@NW(#hEkI)Oikco2@jsjX7815D+ojB8H zFC8$myOcXkUI2rvsLcy$K#Z~m39hCGZ5k+IAby@M)fSRYLV_uBx~lUJC#k^6fT-ur zJ4svYP(C+6>Vb&;1M(L>Ke`YeH0c?j=n3r|KrRUkx&P=S2ws*SIFHcg-q{g98bxQ6 z4;m)rzOet}V5X>haBGWkUR|P!U?6xi;KuDfp$!A*d#k{gM}m>Q)wjx>@aG1J;fpuA z4NF4%WFKuy_m|H)+Us#Cp}o-7214A}Z?x#Y;$3n0JtJ}l1LSDVJr|6cX}K+gfTE<{ zkd;4s_9K4=mTsf)5`)-ONKrkQ)#6OJ^ z(7oqpj6qJfBLFoSuT+V)QL|Xu|9vX=&~b6{5rCc0G2gQ#9t1QZs_fU)36 zU2n2uVieQhsK6M23f9TJ>_^C*`v7O9s1r6jUQrL?H0(fIy7tX)^v16WRtOvW^aD}Y zqyJz6WYHP(SLkpMI00z(=y8U7NPD^YO7SY-KO3nkhov zL#WDy;rfI93d{?>a@skMj_f~^yF{`YUmPHom$(Am#*gxwJ;Bl{Oc7-OHptHi_y$TC z>yC(n_N~FW$=Y8x^o$Rt2O3rA#J{k&GXWW>7`DBY3sM_svhLn$`LNPbh#15XPk|JXt6gu*r1`{jhvZ!WaZ?viqcrA zqePS;J-ty39iu? z5JoO!ECtyrRHWoH^*IDXFsbQt#s*Hb4|DxK(U7QOaL(SE8ZVN$L(SM*_DeHe74 zJ&=R}($8-_QG2gd#~y~6E4*Q$G78<#g`_jMQ8oo^%i?j(0Gt16;OeJ2N(PvIn$;4+ zPJ!8xjXwEDw#v}bL^t0ZjTZ)}8kJj!d`}Flz;CSnP}!lL(3B5fUI!1UU)d&rIKERS zORvj{y)jI91#QGk%`F|cm}?>d^$4^+7_gftvmPK2|5N)b=72f*o`r;^5xWkYj4R$S zWg53SfJnpNJw~gLXu>w8PaPrHU78g#cG_qTMJMXIgbobM<)Yk_{ljZ5bzvQi!l)6B-VE z8CeicP$vXWp9NQwzcc}+6p+Mj%1V98UZ8bh_|_ks+yx2J{CP-H(+89iVsQ?FCfABe zCofZNI&b${A~X_}uGFAlVl9 z{jMa?-D?lJ?X=)Hz62%0Yr4(Lj?<$Q)P5@N?|5E`3BFc2?F{&!CbD8gGk@YJ_b_22 zJr+bL6{Z{kl2aCvkhtkdp7hocB4V-na-m)-LL+>ZXi@fA+mCKD4YbrnJo36rb3RH8 z8vn)oI09FF!4?wSKv;*h3_LFF)vO~UdHLs?9F+rHl4e?jslc#eNp|{hJ^b;%hCEB{ znFXi*np(QmJp!&)k_q|G?V#;0Xmvorw{kRqLG6mX}Q+4BjW3_mdUT=55 z5TP+dO4y5UL-hy?m5^PXaYe5yfz#!iVV4ohfmI zE6J7ubR62IeOfF_v1N#106jEBv(@tM1Fm0->QTE@=&I199JQwJ0D~qz`2)ddf3m?wF*4Z4rD(C>2?mx9eJ?liLj>=rHEklra zp|jbkN@<{&yVbXNjNRjinfndKrzZX+s zYBJb+E$?fs?$DH*ETui01Hzq-xB0HKn;aG#Q6tg5kh=aL5AhVk40gF>CQ7h~{Y-NT<*VXdxJ| z57BY96Rdg{w}8!;1+S*F$;B*iV=|_eSDQ3~MJ(R@L~IXAKkyT12 zT9o~?Z|Mof_kiS=lnctM7(#^H2E(!2aH%8Y{V9I2QO_QLCdO;aJRV+IMV=`RYgs zGL(rO&balHVs}u!Ap_xVNxIGj)1xwsV;V-c=j^x8t<} z;!c`5?0l(0_;Rg^v)J{h+8G?@f?mB0T$GsvZlpp`HwC(o{-ww6#-a+B9NPCyxxQU@ zPcck@>AqXiO(5E)wLL@q^0Pq`{%9Gs0xGf>4Ftw;o@wi;d}WafR|lw!haWdpqH8yH zB+hip&N?cs;}mN*kP+rD9dwSc96Hd}UxI}zC zRO>tet7j_NH3+vL`FQ&&i+dp6M%@sa@CHBBHYiK*YTFsUPIY2lIj&v+O%G*$R5f0# z_F(R5Wlq4@s1z;i1s_8QXJz2}UyYQ6zAg%(jpnaBYgQT|nEf+k1wH>9A@}GmnfgSj zvs`rDgZ*HHZmuJEIpk=g0^Wf-nm`2Mggs(!C=C9?zi9D;TyzwM3wUnMBO zzwd7W4hpXs^HgI=pDnt7^YXH{08`o6F(uZLs8qfUru`|ai~=2TS6Az&zfOuN`%Mfg zDE51(wXXsaDCWrE5;y-Q=)^iQpVY{dwUHhMF>m?teW1s;BQ^a8io?qA-zd)i>%5wg ziGlfl$2uWOUNXvVC_T34-?~frhovZ??ds%-qOBG>rP_stfc@YN6AD3NU{I#Y6ch>@ zLKPV2kI8{4Ej%C<)NzdcmSHpK&`iqa&Ww(HY-0}ELeZKBCk@#Rc@dqPKb!pAE=YVl!hV4QpN1PhZQ{>a{;!8d$Qa;^@?UAE!$?El6k-27)rb2kGnXGMAON?7wR;-0fGWGvAx6Ijz*7 zzUt*7xy#(xG)7RC7L0y47uR8W{g+&Jk6dCj=GW%_S)$L9Dzd3QdI(6nZ32xk z4$yS`7PuDo?LK=!jBS>$xBG8^%2k6b={P3xIMB}Fi@~qX8yzrYPQ>Y;(COW1?4zIB z)>d1rUH1}D!*InDqZK5|LD~<})F)N6dn{x#b^A<%K z-piF1#lb#skc^={p8sabqKY!iQLOpwf%Av;N@1KADC;f^7ec5Rnf|UUFN$g5b-+ zJDvXMYxmkrR*=%S8t2&y0fUGA0Mp{!23!-1Cb#hM4Lsrlj4xuMUer!C4PqI@b{-19 zVY73R6gpJ*veo0DS+#8+V&=e6Vr3=8!k#&8CcTn6^f6K-hU*|(jTlWw?11%WxM--U zc1;U@A`ccnTd%fyF`v01poFuHa_$ED%}4TdR=3v;je~y^y8A|NkH{oWSaNa4I4E{i z?XP>`Ra9MK4KTh$L7#b21LSz;cY0xXOZ` zSvmB4V6Sct&xa0P-!uV_rvAiPq%eDR} z0!gt|RGs~`jm;3GJG^MKl)^&?VM1i=?%usX$W)_du__23j)l*`9UPXA+kYrPIkH(W z8P5*h#$u6RelX`@?;*-Fb&d)RGovC`8J7g%j$wt+x48J!?I1^HX%|g8%0ASPB=+&5 z%;wXpSpzGCTPgCe`aOnTHeM)7v)L8BAo0Lu1@)NgHAtA0P8~WV0y!gc+6eOi79uOs zMv%WGS?fr42gH}WI9Dn|sB-M((=|A5fI+Ba$h81<8aG0B9rgLy2y>kvW>MrTxwhjV z4fIlJgy{7)7|ZBETC!T?^mam``l6vGbpQ2Bpy|~%&3^QpZwXdcpkWE>qiu|RxN=JK ztR%XSfOURf$w-pTWq|FuhPpCq7zz6H+&=50qd0CMh6UA4w6y?wAqG*me)_W>ZMM>7 zS||id-L^Knn^uOzGGAS;PD5&kkyKCR`^}vR&#rJef3kIYfNDdJ^dh7rq8ha-q{N;K z$*M8)=ON9yx>%Ry5J5_gNYO8U7g2?*L-yvjtiG*dvg*e7{j~J@%uGg)P!C5}FUR+k zrKA{+rT9-~Br^GAROMtAl&i8jCMReGvYCsLI{rhbse}c2RSL2SX;T%~z-bBM8qbfc zJq|6&jb`=vOAAf*Tb)(?%iti2lOcU;x;1O<5=vF1Y%S@76tTy2>tl;oQY%a7l~nOP zF=plHO=ioz=&~MrC+eVVWXu#x@|74*Khp-`+x4sG7@a@p?07qXA`0vy(T8fa>KD<;<1JiDubtxoI z;mN$*O=HJ5x1S2nWajei`a{>)H^8_Qf92V*LI|T?bz~-3BY$LLpEU8Xhwt_hGNzu{ zFaZFa{$bW#e3$2xwOSi>QdnvKT%o^`Wwv`gN`XOk_fbA3=@tDJ?qn68SP%9`>ixPf0>Bb$EQp+3h^G3lT( zGk9I4I20C{-syJelt9#TwJ~Kid8%2UnphI9$kTm^S{EEe$;39Q;lcb3nLP(XV(Dk1WZ-PR8O!1liAo?S#V&uy`%~BOZ_;dKPCbIycbnY636v#<^P!Se{j;00W3EeQXo2>CDkZ^YJm3}L zFN*!HUT$Z+rd(P~yXYDP$AltEf*VVI)T{|l)Rbd1 ztZe^Ok7iUj22n}S8F)a5Mv;^Fa2QW9o&vvmC$DaOc5usmM=qaBx2=guZuN*`G9TAF zL?8@LgvJgkoUx<>g8kH>S0#m+TV%yenQo@n!YhBGBt-KhWTg_UMuw1%*d6n>?>zgk zVbwCZS^^BX>tLA;vCB;R!TQmd`g>CqhM1fKeU3zr*L;I9+DIJW5aBM3q>C)Dn}K<% zaCGvnS3Jbh5J&GJ-+-=857~{Bc!KTm0g2~LP)QO*7S|48=c>*dE(@L6axX-b!w>ty zU=V+O zcC<<_#B)s#EhaXK!VQ9hXdLt#8fGs}M)V0wJoSDpvsAy*)e>JAgWf4DyC2mf(9%32 z>OpYmM!&7H^-SpH71-s#h-OGAs^4eRhGA#(Pv+4Sm*wVq!KKmx3$BGC6E?&xA2>o~ zpZ3;krLpC{3%J&V;s375^c6qo=!Hre2p@M-OM>4v_)T>Z`IDS{b~uR=n)A&t6~-sj zyU}@y0iMNipjVi%geUD_D~*?)m>hvrVM&-KQ@!Z2v+5e;t%IBeIvD1-tUnCSI9jo9 z!JOaWsFM_*`FQtjRqWHT3hAQo^=-{wua5ay^bR>na+~yWCFtxLoWBDW^W!5cd%`^9 zGENqJ%)i~+Ipdt81C#WHqdZ{ zDwc?ML2bzo{4{zhH@gfQ(>@-Vb2@Oo1|9#LJ$iQrbTaH7+$R?pP#J~s;DtUC$C#6V z7!eSc=}bnBPfyy+cQzBZB&M6M>vc*QWh<%luDjn`{ylOL{>YvHhud#3WST`(YVnZ7 z+GiPvJPr7ieMduwWa4}b~noQ@sZT&9?v{XFBa|Q!^Sq{SN#h)DJ$8gx&7tIz?m}8mo zHm=ClcMVxRuZZB++(tLh{*P5Ti~ zhfoCOMc3@#7j@eFXm?*WpQg9%`4;$;bO8FU@3@_%OHkZv)1D9y{Tdko^4@8!!?MQ@ z#+dhfrpv+6Oku?N4u66EM$~`o!|`8C{F><(2LA~cRY}V4dU;ySy&qv1Hkw**5hlx& z=PASbs<8fU-G8e=nHDmF&pqX?q!|!d%l}ZlK0m=v@shz1Vfr|Pb5Xm^v@-Zp0H1#) zZiZsvO2vd*WG<2IO|JoNvVvrnG+^nO7vM5ze$`8t6aq@BQCE;31M<@-^YZqo5U>N6 zRY?3z+fJE{-oES1g-=isVN(6R^_^7lxvJeDlC!}_<=X30d}&cGaeo`Xm1H}pixq2d z(^?3@AX3#C35;(_m~^1RkmgKLP9>9MFU_@<7cI@z)!0@o%$3x>5nn1Ske2=if>Z(s zb(5Bs6!`-&n@zVw*KbtkIhX4|5LJyFan>X6t-m5$KjTT8zAh7zCus%i&@D<-kC5H) zc3ws~xe!db7ziF9;t1nd(z!eWd$?7LM79KOi$}fVm$rs6 z9RNF@y>9XS> zlApB%B)HII%}vL*hbQV_CUE1H=Q+C75-t+OV=pw{bx!6VdnPFR<(ky|WOVe9&+*46 z0y;P5_7Qw^Mq-MWOE?E{419uIRpy(WJ$d2)LLF0K6V_E{(TD4W7S*a_0D4#9HnpfI z=X|@vgDBNx4+kM4_0Kc@Eg6z%gyESZQh%t?C4x<_YM0GB^A&VeS7116_hj2)uu+nf z=Xje8t7{79vg!C;>>~WT@@a2Ch&<^6Of7Q60ip5@`rKu|s^_@Xhh@E@7u5X1@=V1v z-&M3BMjb)1Zapm@ryXya-pHB@svjwVQ$qqw(__;|2}UQFOY?Nq%zoPK&N8r3jn|jZ z3fR{lkWh-Ew^m;srmPQPx79WnkWP&GeXk&)w% zzK{LJEI04%-sH&DaZt+5V!Dc4*OwJAkvyr<@y77cLSE{qM?#v0hvKwW+0f9Q9q`=m zg*B#0nC$$acQH;_n?WI@;z@=1Q1R%q0PzjT(7{uhC&=ZW(XNb{p66L!?3fr-VJT_R zq9wyw)U?*S6>&QXsbffe#7 zXE-u@3R9r>yXm(u-FXJA&y%KN86+=O4{y2E6kw4B? zv25;>`!Vx%H?y)9wXy(-CW(Y0LG`aBkFl8?p{gs(H|*~tl1YQwUm5W6RpvSnG=ds6 zlUumF1aX~rwvC!V#y}NiKFrp-9$~X_3XJTISK08|gjrk7d0#;)QEBTJE$`01^}WoA z3~d!K5{IAn0B3qid5Kvz<1D!~S)Zu5e;aZ=eVY5h8uQsg(mGmMDj_Mtwa^WZ1=)hOK-ZUuX-|G=D?S#XVcw3G-DyzOYRPtwf14T8N@A3@Lx-r!3f z2fIW2V$cZzruX7QE!CHDA*59^|y7J6Q3P%+A7Rc_#wsuZk z$6*hW|1*vBu~kIRUZV^&CXeT4EmiYh~Wg zUse|&gFzfVzSCSO#gfN`zI-)#-bxjkSeuY0k)Vo%ae!A11;TGF$yns9xYu) zaO`)uk{`Hx8=Xbn4CV5Jwr1QK4}6X$L#3w_*XZ*2386FY`X9(p8#`s6R;4gXZyTNe zvBiEM6j4XniwfpiEVv&;Q4Ouk3cck2b_5gUTmx1cm7#I8C+Ltj;@X8>M#~F4>K^ek3!qLn2d+et`UPzJdg{y z54aiP0)^FAJNSzJUK@^^{Az*zGR3;%17B)7<|9$=OS9|);KLGge zaDDNnAZ{7ks_qB6B%C*#X|&Kp5+{N2CH5*EAPWJTtV0lZyP;$?gFT&n13Z0jMDMgN zLsJzX6NEB!wE!4fpuo3*srPo0<}3|!b}0I=_sy-HZ?CZ`NFNSlJ7<`G^7-oQYwxZ* zkvY$U`ughYP|#g?>Q9~|{qdA$9;LC_=_Ro1!_QGx(&^)4-aaY2r+5U-?BAf*8G86N zMT`8W@{s9hN1YHfd=1t#y2tjkE2E#q_bTn}_9;kWV`JCY?TjgkjHq=X(`;z_vvEgH zAZU#3-#*`T`edam9}i~zIRQ@V4y(`o8MAL2AA`1g&6*GH=l=C>S2P1vjpAivZm*PMhW-rYfAQ9) zNAMRx_wPk-!+iM}T5kPmbe!wsXp?CiUr)kdqszi9`szSb&*@kw?|5!HUh7tUG;vea zmF<>lwl%V_ka{C0%nhf4^{k73M+Q%|IMj!>HEWjIHyw0Qn?y?%6G}&#bRru==+mWD zbdxP>YIoK>4pPFS4E=F$*niE|hHoA!w+@<%CDB~A^K;e{cd+z6)|yz3fYjhz;~PW# zmoe}|FDCR|kIm$AFs55n0kzGWOSU^GH;i=cCeiEz`hyLor(3kt^$#%I(P>6}P0Y(~ zeY49-uCGI9nHlV~wAl0^`dPb%?2!($Su{`zSXr|FHwnlCR$*ZwHDp0MLjk7~DzTRApv2*vG*+Xc#09 z6#{P!rQ=Am>7+UY4W82(=v0zwp}d&YT`nPy^cK+k=P*+BS#DZ(ZNAAJF$^)>oz;!a zwbfkIbE`Tj^}BtWemr$Vs)J1=v#p-S6cN#QCaQ6oRh84$0cq-&V1$6ys{6M+_b&?q zo%lg&IjYS11L2Udbuu1o;wHJ$Mp+}JQ;ya|JHay#g6j?o_tW#`{p}=pcamb5S9$rw z0Isi<%6jFwn*~_B&o;CH_L2!zh3YYqI_*o6d5r_kiLVz8Py_ z=yY+U3#Gkb^N;~WgeWX!308nJIBW)h{ki$mEKr1;2=;9t{y zDuO~{5cwg%r9pznz@J@KTW1+{xV@pj4t}Wpj;OH){OC~?lSdC|mvEI*coIJzY+Fvt zh1;ldx}l>&8VU!}KxQNX#f?}ao~us?F3N##0Em?t5$%sC!@KOds(O}%0b(5he!o#c zr;+N-M&+j3Cs+)A3Wzf$tOaASEKRRoBxExoLWyjKE2f8>V9G?HHk(Rw2G#y{?k75Z|3zZm z0B-89n1U_-vOQ=xhzT7{yYc=bNPFX7IAaqHPPLrem zcK7Rno)X@d>#Ey2Anfp8QQ7GX4Esd{?h$(y@t15#ZJ=*-PPi$~Xu)beib{fYIjs1t?j@8kLTA&9O=1+VgdB{B8twI3(M4KPNTwq^$YT76EAz4o z6j#f3!n^}m^9%zCz``o+{4|wKFB$v3W+BtbB-^|oDRp7}Nra8{deLst-`S|TjZ+p( zI49*iY~MbNAQ522^vsj8r6Y(W?zx(*1)y8T#1w8IGCm;lY6X)AYW4dKBa?RA5<-zo zU5R)&m@gpJfQn$+xrVg?<-OEg3T^s)`emp2s}w|z8Uyb>wt}sIS$K13P{Hx3-7Oqh zN?#l(V|F1{P#6#{zxgN=31fXG1btk0N&8+}h6mX-K*(Ql=I0N$=Cykq%U zpW)KfDhB3`MZYRtuj3 z9!p&>76S&{a$P(2&Nd8R935pm$YoF8j_YYY=14qu7MB+Kgm=EJ5@2Av(e_Phhs}X^ zZfG;!O@nJ3WO)=y1M5!rWzc-3T)^&9Kf$EvT0J2sol|tdh{6L{GhoP4vGM1tQ*!9m zp*SFv&b^O*jS)&;DR#pkeZZ0O&qI$csM|bi00h<)yW#4=qzMAo;X{OG@0*U2F#%6e z<2~0WIY7IBNYUJd+n~-+@90Ay8aWmx!uK+v0(=ILvz>6zx^qZ7prf;Z{K*7C1F-ym z72)yf<-~VurHh}=B zDgJg-(f6TYkpb**!oZw8D&CZ~;v8Z!@kHR_@^<}(2^ACX?km`;o8#~YhelN#@Lo3GE9eOL7IlIKBguAE{$@U}6x-79 z5zg(-cB8V7BV`ZmQY2GTPcVO_*G*YZxIsKaixjFIN0 zP&>c!atsv_mK6Q3?Y|=S$}MwMH8o+jFJ-c`VKe^30uqIH04@a~$#jt0prr?h6}dts zZ=osOUwuSVhf3PDJ@Nr-cV@#1`H9;eME`JWT_!mQm??}fsDp{t*9X#uXMXI$mWYn? zpIU=PG-0{o>w4sc0Xs%EN1%xSaR*BI52G}IvRE|@_lYXoIzU5M@n8&xS!arA zKTBm|W`39fl-WQN-;77%U2>C4ip0tkNF;(RWMx@yJr-l%_UD9i{sb3^5{|OHZ;R+< ze?koTHlK0g$dDw`CD6qLYyd)R7xLV>CN8khG0)KRFxAz3I=`{N|pZN0u z!sW2@;7Db;CV$~4flZO}lSExvMels~I=!)+Tz5d$PpT;)zvxJj!&5{sAV~v-#mxZU zqD^Ek1Ql|CkuKEDLTQ6-Ba`zIr6~YHUu7QjOa%zOfi9HmmlemhpiUBI1r#`D{09HLQ+4Y5GVUNEGlfio`JAZ$fM1h;*7Xl_yi zSOixJCS{0@B^S?!%iF(@ZrM$gr-?y0`EPC7whhHyoSnbK2mtOM3Am4sg9(1H!XiA_ zoM@@ya>yIx*WB0e448-ZZ0yOaK6`yp_>@`_wU~c!$dR(ha|Zc9o=-JsL_ir}g6u$% zIxna(@tPsRx20|E)TKp*DEo3MI$O%VC1?hzhMdN3cxuS;Fx+oKP@2W&TXQC#fL#zEKenW^Dp})Zp191 z!T_M$aqXZnr>Q9oL9&lR;N<&x`*Ij5ij^1x9I;H@pIKq!8}MaiWv`)Aw4(+@4BEjO zAT$d6Tr)h-$7K%Lsas@=O^>_dk(V_-OZS7k&W4Z9xtz(xav~p{$Uat8^nMT3>f=vgm9rm3hzC|8h|%;W**%pJ@5;? z=UAN)GXjL)i^D;gR|o2CK9V^;ihE1PDf66a7kCf~J7PDI9=Raq3u;GY2~G5!Ys*GYc3RmNmQO9!UgRqV!Z%NUf3dL^5Cv zVR^fak+lihZykXTKYpx&w<+DXTGr#$(APl~KhWO1O1@`oIg1ld)zD<3oA11C@Usf` zO^^a|X9hnaGVBK8X*##GJRVh*?GB6^sBoIAc#1Iyd5S(Angh&4tDgN_KD(IVGRmK0hldPiS9r z!+!-iOo{qfZbYylNCS2S8s`qWv^d;dMeHNT0qMg3EX3oH%{^VX*8)so%f_*Hp~J^(tQQ`tkk@*loUV{ z<5v*yb9-N%5|^^PwQ`4&_ceX6HyGo>Jr7d|)o%>73zB8^Mzy`Ujmr-7pr8*lzeXO1NWg9FV|jQ+}oh(GuLfg~vOqn5b|okcZ& z*b-^BUJg|W^Ig4bZm)#8<>K@LxZhkt54CwiM4Fjm3|LYN9Fdt6<^t?jyO{yOglxzY z#*TpGg1I{h`~f!9DU|)M{E6v5@+Wr2|3m$@N^8skM=ZYUV0uCOZxoghDbq8=7C)3> zK?VvX5RDZqCAJhwJb$yLe|^2U0{;5XEGs_4!zQrWCa&bav<5IXl`MrK{K7H^o!z2q793mGh+zbz{>A zdIhol^YP_HL&Dk+-`tqoySYZtF88~RHg0%HnLld+T6OJm3TKnspY*hQ5g3K=iM1GX z-!7(1x?iI>8yY%uCQYTP0N#wVLNcS~x5K(`9~jxqPK58|@h>*9xI1ggJ)eMeqxd&> zy0w-BZwpJh)q>B_*U-v>CRa z2YlwB_|(4*vc#T7psp@%5c`EXY8@4uA2ZQW&|Wr?NBZIsER(?8xC zHXZh&kKaCMnWY;Z8phbM9z^(61TT!wXb%ybU`oRHgYH{X44N_n;PJ7jbgYFuNXclC zmSe!vDnT+s2W>&4*3t)=D=HnS7NrcD#02~W<^_tKJB^Qi#p+8FTlH3~DGeNUIE-Ir zU8m{rOwI6TB3T)tblEnlZZx#m;?k9WFf;<+EY$DLo2^u*YEVNoU$ipQ{pzu74L)el z7raZrC+Ogvc3Vx;Lx-}Qo8ijP7h?cTla(vdw9nqvn%&f~Kb`b0O z*LyTUf;}ohW2y&M4(@mqSc*$m)fJ=TQj@ZpeQb(l!!1uA>1?21*a|gun53hbGJmZN zXXJ8tJd+C7pp6VQdSS|tZ?8JsDFT@5Q?LW03{AG~1Bm|uM|2#crbrsdic1biND$9X z5+4Y~?Euj2^&GAv+%%*xt4e@DIxlX}#K<76>~Fjq@rk4G#pc+o=YaV!?Qh_}<{hdS zITehVRSI_hx&sm6RAsOEDxyt0jcdPXA8oRY+im%?^`f+Y;j!|S-ffCOF@^$Qa!VQF zHQ1CD3&?N2z%il&`64G=6Q`!~?QHKopFfhY)|j5B!u}1usxC z2!bHC^ZodNDUa7(IaqosJbFgXP1JD^JAo`J%+e(OIWNB!Plf z?DdLw88XQL0LMWn;t$BiBPT$!(u00fDQ$+yg0}~EIxx{C`^=wsA2o7okfqZi_@Gyp zkz+vz=|0tGhen(yWG5jOR3$nrggdT-zd-svkZHcYE-vz`%P4SnmVbffCacy`H}Gc=&$jOz(;ZtXLKN2e$8!S??6Sgeq6tlosV{4hgt??c zHz>TxQ!HW#@l8$R=Ww;4Dmx%Z7ZrrDtoCJ` zH}}mp!(1BZt>|o3r#nBRc}$&6esqURGnd=_z_czQ0L#kk#iBQe>qpBzowL>%j{@s! zId?SQK$5aRb{;E+3RZB{I&N2j&39_4!t`-!M%$5JIP1_DBs`ASQqM4I4jA}dQbUd1 z`lLcgjm}A>Aa@fjl$QB&5>&?(MABkC*TCq5e?ik;HKrrc%|ps-roF-ypImjUsm&+e zA-<-id!^K_5iOVo{ndZ@`il>5X2#(_C10800lDWvz9yKhssn!AeZdK}N2uf!vFDtm z)z^TYiRL+|AbkG(RY;nFoi308bJV*$CEl-EVcBs;uC#x1RnK8=N&hA2u`+R(p+SPo z5Z1Yd#@KVwP_c_!)4frp(e3SFeY*{ON$fwGi;R*~Oy?>+~g>dB4!K`>Q4Y2e!S%%6DItAUKnM)_PE)d~tv* zFjb(8MnuF1IImlqdxIKrtToiQ)R%%#s9YX|4PUWqCk3RnA-z>$RZ7 zHFd_mSA?HlUm7te?XQ(2>~Gyjlx{pIWF&L)j=ql`i6aR39z-;snYXv|(zc%ZLGi^{ z;@A5f7F|JAM;L`SM6Zv44!=QiWW*?qIF?-IF$eBdqh6@W7eH<% zn80;KeqF;+DK1xT0(V72C=@?_s1jp&OQiS$0Cku8w>S}|TZ57=n)dsl>Q(FOj5sqU z4>s*_>A7MPTw-Pw+|9;D@8-jFTQ%~X{~CJTP?AOD$(^9){#%zM`N$4bS?1dnU9QQ@ zfpwRX?H(LZgpfVgLo_O$GP|KYSOO9}vV!O8w!zOVm&a_wE2Q)nuS~sMGE6dOtYu0xx8+E_u z+UUNo^j{Iz?rJLed@s^M)!u{haaaU7QRsrgy`@-6w4JfvE>$v+cGo)%`3 z6|9%=fJEdDA0gBs-iB(eUC4PTunXfE9U;2W*AS@Glb|z}g*ov?$A-;A(cQ15^X9f~ zr>$f9x*#88oNyDJBCO(bD<{Va;?RvNbrca!=LO#|*MOSj0}=ZRUy0 z0K_%b?`V&E{!=C63mnUo`#PQkq4br?Bky zn=$e}g|u?V6P>T@h|W)#g(JbYV_yS*-g;j?aTt=s!+vB21u0bj<<&Rd?o zM=yRd6L^iLPnZj~;@S4iik9?5fFDxaEOTLiRnQy_`8neiT{-QKw^c9;wQ*p7dF6kG z`;g{YFpvZko-9TOzItrXXFj2&p zB=i<0YEr3?IrW{@P=^mb13ks&Y~EGJ8sEAKsV?%@in)VAp9WPzm@D5Y{gFHbV89Dx z!ns+w5g6j2DU9lkht|%{EXrMsz5yt9kOhxl=nD?ST2_6f$0X*ad^vBVV zKR)@eVxFvg-fxz}VgguX-Ut9Dgx~<&op^;E@`^opBXbbCv=s&--CD z#!CY++_SB%?g8phhps)*O6rVXPHM~GJ9GEvu?%)WqS4LEyM_qjTGlRf2S|bS^RF`F!FiZna5mwy)NSu%Fd5}q zFvABUD>}-WXn5=qfy>#QJb;dkyXnyy>?*o7>WNctu`jgO+Svb9yRVdk?Zj2GaOdoN z+LgTKtdF5Cg!sGhOZ)K|p6r4#*)r^!IBP-M>bx$~c0_q~x%@ojLJ(PsB)TS?%18(w zJHUBri6Zsx>Z*Rva+@Trr$zKO$6=hFnS4~TG`4{pX-0?aQI@^@h!0BW8*j{OD#vyE zXbteRa^A!Dq*bb&+$7AvBf2Qji#*?*tw4C_h1>vZ;}!Ts%jvo@Gpm}2xC?0Z4DgIx zh2A=O2A>N z5j>YZuT7U{_(d!!C9uQuY@8xRZ5q9CM9p?)!bF)<6zgu&o-jk$WY`IoM0vgEbBa!| z{-KY)sP5T6Mey3?u0R%Xfsl$#VvF%_EhprItby;WKk&-Qn*;4dz%W#JFzqUt?q)$s z3|Yjo&J#EJ?_cdnr>Dc7)|FcAow2olR_wX6)(*@6sX0}1>Ytjn$Y8Q-sB*D@9SQTH zsd%hHyvMbCi1N>--SlhIUOcJjH41-2NX@zidwZYD-vJEYqde(Y)d(;#K4K5?0vBol_Yk0wG)*^=+D2GLP~b@PsXuctCxI1rpq?Nc;v{=!@1V# z*j{wDef{m$yeOI5xFtr5Icd>}DcSrH`9Npg;2@7?F6w4;q(QAM1vdLeAPpYlT2^+N#zKF* zR*Emc8}vGzC!{O+LITU`2K%_Q-NA`@VP1||C$`)cJXqSwC0J8iqw9Vn#!386Ll7u^ ze>9=aw*Wr-Xt{s`Ejxy{1jDyNtGdVxa^&dacSV^Mw^v11>>W(d+r^7Lp(&y`X zes$0Jzoaut8ntbfRF{*tSA%}ir}q+CJvCaX;;KLBEobnRL2g?7!yPS1oCQc;cxkkl zR)pm7pVz-aa2ETmx&KPYnExXoV`ch3a0jone!q;z;=4cU$z?s~wk5Gu7oIvIuz`J| z7{K@{UtEmz4GU=4RN)m{#4P(BKCf>6UoUHmMgs`Gwyq8l8>g~*Cn@ReC`DuFrhxmP z0Ox0@r{PF}x1HntD@^H(;yaB-j87&8uP5n$$%8U{IKg;chhAiz2*g%iWKDQ7DPot8 zS7#3kB?o7Avtw}Y;@KQ#)Q=gu_;opgz-3@G(#1_vCgyk1Dnjl==+quuZ*gIj{YXTvSNFcFoQxb##EkQ#G3Wb_248 zIiraHV78fYicrJxw@z~&lV6qo4q2JQHKiM4fZY}hmy*}OI$Vb?M_!Wa3glAC!6nm= zO%;(yU{Mu+yNl5pq>{5-%eV;1Uix6j0xc-uCi0~LIc0#X)TY)QN0Vz!W(_aveF&Fz#qv`)?>PNv5}v& z&RU^Al(CR6*U2b-4suzWW}I}HoilvG^e#M_k-6y-n0bhh-3sjY|MiM2s?U)V!+aWV z$fIU$Xz1wT7F2TZyIXCLh%N0xM4F61+?&_gW1E__FuV-_x6kHC5F$M$9cq}T)FNO5bq{w9ufroR>(w4c(M#W#~A! zGbYYDwAJ}8d-G~+%CO|bB%TV@$s+#1)zm;NCI+aAZtZI4Bf*%;GnaG$KUayI!i)`Kzp0SsFlm zK%gl7BxtP0{k=2-5U8CrTrz~7w|tgb`mUo;^E6O6-tF|5Xv;*61hNJI=||0baJn2* z2E;@}B~4+z7>bZK!L>)QUKDSuQVg=7eJM;M50j&Z>%})`2vGFTrdPIdkrjU-BgzW4 z`B4>>7m6`ZfGndaeaZl=ziJ|32Lw3+RJZtxMeu%x0RoW$Gz1~&s9MhkR9`FSm za?EaZ)4>B#ClId5Ptl9T4t1f>+`jTb5X%0Pj3axVQBFI73K*VzYT6GOvJ>3}HY-#m zfcu4$2zm0+Dl5ZI(Wg>~sGkMPF0qf8w_N{X?cwq4t(p45bZ&EVNHud{T(+3+U|gPC zf}$q6_~$$N>wZQB)hhsaABMMzPI=DT6^_(4WwWUTS@3j)6QHUuM zfAxvXdKlpOEZO6pwQ3$l8C0n>R}E6D`#`bX4-Fu;+fp+=4NL!hc{WQ0YN`l$M4`y< zqN|)4y*1UK<>@#XAPym4yK3UTz0g+NtDt`CX`u5$f?RRHtFisf(O401ZeBGeYB6m# zTD-%Vv6qUO=ms#lK*$Bqav!>>dn?72SB{G|%PMjTIEq7F>QM@)j)N6*FC z62GEIRes5U8FL~R9&x2HJrqHWmvUlGI0$x zh{J}poT;Rf{pQ)&7^HN@0?b042tiVS$T!emjl~PYkZHp$7OV4HWb4zzg;{(BZ}5vy z9cj2e5d3%+BL_~nL0$}$RvRZ)+*k4H?? zUvNh(o_SppC^VV6O=l$@f0kZ`*AW6v_HEwiTd+ni^GDx#Zhy=?4_I=T5>|h8_Jg(# zx!g~oDkUiqeymC*E{Vj$&eA`*oaSaoC{=LN$gg@_BedzvpPgyw&MRZ&F%@A)I@1*~ z!u_QzI2-Ow*)Lq#ueY- z=g?EzzYU73V|Oazz@Q51z+mXGJmk@8>BHb%XG5{?Qcpy&Y#eN=7(-Tb5+X29o?ZX| z$NO*Nz|;flJPB3_t!wd8AW2=3K)^XNfgIU|_eWDv+&Zm>mb;nN^TU;Qht1u-*!vq4m zlx2CZMH*UV%Mj3+2+`<~NAOV7OLjXX^91<}MB^FAys(qmA ziiP8`7o?WR6T6@&OKLzLaX^-^P1nJKn&dhrSDg#zPEX8EVePFKV;8VcOxNIL)@@zl z^$5pcB%M^nq>OUVkDg>CiWs2G4QT-u4OFXtsRDPgZaFoYMpj~6`)#rePsG(u$(GIa zs(Cda3sp54oT&UaJy{0NGn(bz;RXWQAs>C;Lhk(ReEoe{`$FLTeO`EC4Z_(&! zxju^O7bwB#UU(cF9u8zQv zPlyhK-c*;bfeQFHrdwDSNOcL-NgGhs!%s_GKf1wW&_u4lq z_aXio&Q$m+(=;`BJ5k5Pye6xb~jF35#e!z&nKb;YE=Y|m{vGOmtkwK$VEyrhF-FtP9Y3o#J9FL>Zf z1IJ8@fG?ms(E&<=bW>@Y`W<8oU;^5{Qr{s5`l19?!489@JK9%zZz8<{&pF0mDj_7alqkcf3E!|IBJLN_9gitZVP^d55XhCyyWt2YtzbZCw z#vzpX*<$ep_@UI^__SMk+`pKmoi3=bw*aWbhxmdQ$-@~cvt+VxT0%-%y$GA#=QW#C z{`|>j;KaYN>kkA#XJ4S}voaE{&l6SH^0+JoUtwCQBRo{aveN2!_XJ7x@&_RV}8_Y!N@3zpU{A$Qehdp=?ApvNpS41(5n!unlBXA+=lP;JkGsfSASqdWvAOG|$wgVc)c zpu_d4)q%Rp#;ld1yBE_PKln}tpX?Rcc|Mp#%;%$zjtly$0_i-8_dsg_B2WepvpeLP zMl88uJ=LDGCsoE&z4*fD)f~-<*ERCB2pdb5w<&ub@-YeJy`xY8p5Rsx##gi`Bda5; zx+GSaAS^YjW60Gx>^)aFVzisvPm>Ky>tBuhp(IbS+F%I!>yq0&xJrVIsmLyBVHICS zi!8PiVL{mp!NQ_H=??w8MxEV%Q92^Y=`(P_(HWw!eT7#k}oZp_oR0hS{_epV-U3hBdacX9_Pk!fyq zQ^FF3q4*AAF(2Y8*IUY$xHmmcNVdO_Hn zM~Ome+OI@`ubVemA|F8ugUTWNKJ|1)tGc%is%Y9zjU~f0&Sz>Ev9KdE?0Q>OaorH^R%}gMQIsH@Oc`49pZnSpbv@+0QV=<9`^LU^cTYoEU@w zi+P@WED(mc4=*I{s)4k+i@lt0bEZqIZ1Ie5lwGm=+)l4e?l#~D8nu5^Oq*aD@SPrB zn6f;2?9k1%;;p|hDuj$D57ZF~F4)43(Jt%d5n(swH&fSlkn<(SE5PF3oiw;h$iFJs zNxGF`&It>7HGp1R38$Y9T(H{T_==DZethj+^6;q30wL*o!)27qHhLm99EnNQgdOIy zupt-e6b0|711Z*w=Mo|&@ij1gU3_rd>bswa=b~93o7yyI87VSAIN}uv9uoNjzl-;e z>BrUDKC@bnqT%?4PxI}rx%7Tlf;JE9>C&i6w)`L#VFCJoK6*9XAYC zLiMgyz+6V`?~|brQdlTjV7*Y7o|TO^PEr1yG;HtG{CC&$6kO6L<(D0T2cukE9V!kU zbSQfKEr(eOU&)1G?qwR&^7HMZ{jDQcpEW7}`SW0kNqI@H^DZ+8z&_#s!wwP}E~gX~ zt#H|CW49uQ$Jbr9{&q~2!7?$T;)VrorK7TS{#~hQv6A*O(NnyDlW^jDE?qtt2x%4FbWE&eY1q$SR zcJlzDAz|F;h-%^9&S+dGhI0cZ zBkgLVm>dxvC0b}+3VW81kE0FnfwoDB0!1@>iGj39^P{7^xaoUsumQX2ced5u(l4_+ z{08Exrq^HfsTg*o~5yAUBtVwr$(AXDAHT{R75^afv(jtL0Bi5FLu1bfj z>5N8v{2FfH1IG$x-)l+bs{HcYRE75Sf|$m+xR8fCK6NSRb;+Wky=**o%f`^X=ZoV15%&;3s5>tgG?+ZS^Xerjcd?4LSHe)s_g4Ac6fAznt&GyTCP*^Z|Kp37p3hyo|R$(!@+-Y6BCR&zvdSF>Sw z)tO*lOH|!sLSxi5o6>s|@4|YiZqJPyHlT;Ua9xs3qP_fI02mWo4&cHa6>JZZK-GNR zq8$k8k|~F-V3opYN$p9iW!HISmL^XLDdxbqK0jz=nSG!=Luaifx&_`{xDwZ+czlV2 z)4Qr`pQ?4x%up=}uF8J%X)#Ikv~l3Vus0{#1C-mx+s+a#Ay!(k?CW1tY!6+_)vQ4H z$`u;z>|b!PZHj4hXUMzNfXoE~AbENx`hUe6i#95@A$`)1?&lI@KZL_(@H4tB_8@mY zE+iJ&{DyVG)5Q z3Ql{3Pgh#&JH|)ozV#A&>CujxtR}40P+Ym;QF3@3OH$_jD@HedIEJ(f2x}B!U2|80 z0zpBF|DC^*3`rd_9By{~Z9CEJIn$Zr?WBEHVK_4j6QQ5yTbl8@X9_2=g~tAp2HpBd z8yE8t4Oa+8z>OS88K`bW)N^OEq;8q$3wlx?TxFAHvDw4vr|3B6tPb11%&=stQDu^} z*#@1PBBAM_-$xxX9DBB(cO9WBryWno22__t8~x+#+9J&*CfH@Mt_u_Fwt}FWH$}i5 zuQiB{#F1-#!X#6WF1R@dJwBq~ZNXS-2j=IbqSR}C!D}XHHvd*@2f|m~;i2&}ejSP^ z*eCcgt5`^spbY4#sXQ^&m~SvGk}CiB;Q%`87sPIN;k88qhLlxW*MfIJ#0-RND2LOOAq|CqVCcA8i?DJnK-2mdpp8Rrj1dNhoIZ zm_4b2a`k3OrAi>AQ3Eb8Yiszo7pTBsE}c}@X(#;|8Z?L~q{TQkn}*01Yg($k@TP7m z;Q@ZJuPF7-sZ-X7sMT8>5To1*WB+iyu$BukriNjsaF0@-;%16s&=S>PvAy#sw;1JO z{1m;b4Rv#Z;hE@<^f9(zCzhQ-=hCG+AQ1J0zEoRSq`O*OBV<)?}@;3q0G_9E_Grq zzwDmG;h=XXNdbE@|J~S#_MkE~tdJg{r6^U#o`sHPD|gJ-I9l>0@&dbXToF3p+6G$(*01 z=H1%$&qjI#2c=hXiDkHY`nzqhw%D|33$`QIty1}wqi7hPDRBp6Z;Fc~%=ghWTtS(@(#JysOK;@!23o17uv6xW`OoCL4L0@mL}9 zLcUo%j}u9vZQ?*fGpaYG6t*b>l~D{hlmA^Fcs`3rU`9@g1wsnVz{TR%t$?a#*I|45 z1A!f&Hd1kKg{Bdp%s=CqP1HEkC*lYob&jO=AU$SPk3wwYpJ{K7z*_kawT7Lr4ypEK zB(@&I%r6SS&#a8_(^Z~-t)}$Vb956~Nxt>FSPTPA$&hJ^ zvW4h&N`g6uNDFyx#_t|=zb!i6k8K6o(wrJom_>Qm=Il?kI1M9Fdo}X{H{G|!aQM_1 zq#4~NXuUWkM|h^40JO(LZ8JQ*Z(eJjBrEg(4D{+AWf2R;U%3>EZ7&;M?}f{+ii1lGAYRxM14AJ7?&UAhF(!UA6afVLl!!fO24wP$Chz=lH<> zl3G{7;#w$CfzMqkhNjv5*TL}GMBJ1oqYu`jox zZiv0_5kaIvPMonpO*b7~Hw2N$e`d^^2!f9?tc<%Jj;lTz8LHZUN%S=ET&YjP{>Ma= z1x>D~TE?+swVOnYweog?`W+o}06S<;WQy(zPUsW?)gv3MBw|g(y-8=PB_SCcemapd z#80ME%^@>9FX23OLqx3xy^tDCB>tRf=dKWjSS)tWyA-$h}{mY zwyOKX8z(_f;~`9jXk5O?S#Pgbz(%RK-rrf7wN`sAA;2m&h|yn+;+g4unx@jW1$7*# z4%swDAcqhBAQgWngZEf&ffX!gD6B)0Qy+2<@)}Au)O^!*B?KN+x(|hRE0|PM}Rog2M|yKhbEn4MEhX+}A-4O0l~wLw3nCHv!4=fW$UhtOHyv~h*@Mh% z8FVbpf7m@l*!y?X@9MOHUKSemf+IQ*$FX?xK})O z77Oz#XtWxEwlPr8$`$M*`Sy|`d$yGd$^9uFEl%R*g&*1t(E=#)@*#)lXbs_cpMfOK zNAzOS(TyPz+ySF%{~mPS?oyK(>a`A&)XAHIpYf4c8>`46K~I+_{!eR=+T{u{os`qShdt)WJ&TW zr(v=aWqB(!f%u!PFxUiqr6vc(Y;V?#la5YLbZcX{cH{59x&`J#ui%9%jw~{k6GfF$ z#feLypk%l;+p<7v1C~#=w%VZjPVDEFWHL>)Ym?K;g=#7{5==*}0?H|gDYhB1(p%^2 zX66o#A^980W)B^(M?~d%-~uV6hwU%Re9AELrd=Ilh`(5pnC??OOxfy?lnc8ah(3e3 z1LX){9NC2!|0BKWr=GpDLqQrg+pw!xD(~Kd5)1py6(U4NAY1kAI|Y(y)V@4PX|MO~ znQ-I0r8ICNxVx+_WOA~GJ_r>eBT2;-cI42!1=T8jXoa}oz&9-8^a1j>I)46*`}6cN z`0#M{yx;tk`D8_Lr|)oDCmhKy* z`MmB1&+<`2I4@SkI;H1;M}AYho-UsvR{(Ic)ajfx1<1amr~I@)iEOOE`R;Fn7B1q6 zFuC7<{k&TH*ve%>|Pj0kRc{&-PBMBB; znvD+G%Q)<=s8b@NeIf>4gTh!)q--9P1!e+qv<(_a(JT9A#MmyN)Kn8T_0vd<*v)xG zv-ORh!eGqoE$3b$ZL;j?G_+=ir4UFjXtWV<=`kA6wov#fZxdpMmY*7dmlDkBh(_|0 z;rd48B>D_{$4so8x35gaizOJPzUD9=yR<1n+=EI!2$927iE8)0uW@?jXObsQ%5hZo5UsQI0C-(_hB|xTF*WFY2sq=dE`Y zN3Vf2(g9GO!KOa-ID;E`-dWY_cd+Q#YpB1F(r?-ctyOZW4(&9)={N4%k3^;{5$^^vz(F(b%7O3EYVY2RD!>%+^lj#STGgIR(9w55u$5Pt#Ys5!bHroiqQuMqW8S$_2x1P#0^jKuXA3Rv^ zeBub|krTKU06FPH;WE*q&RejS4F)^RTxH3%A)AM*uZzN$t{e?Si?Y|*l1mG05V^`)LBw)h_ z#Wmrx<@n@m&q$)3vB4(@n@dhynsh~=0fCN_PgFX+5o-Cf;d%=CIy`KnBMlaOEnP!( z80qprr(Ul*tW@iR&|DBfdp%s_L$eH&?q-8_Dph{H?!a3yXYb%tfKy&%@+mFd1Oxa~ zYdduw=T3m_JW1I2w~yUMj~%fTMr{_sNyZVQ!^ob)Jzj+k9ce7a>@PETinFdQV^+2l ze2}8~oPW`G&Ape_Z*SM(bZuMWXk~aC7)*_2b8c=|JzPx~>R00}uz3+getJQvEcbJ0 zCq9gjmD!kfci0%)sShVd2VL+ceoTytUZ<9qd$4!L#}C@~Nin#P#+C1Ml}U=vuU%sl z{++4zhe-?i^iZu}Q#F@Xis~me)#+GXE|TU&IGf@&82#H}w+vuwbdY;F9o?;GMD6Hg z*GwXnM2D{;Oe{AV(GE2>vVV!`|6!JqIxG8*10irJCsFXMhpxaDS<L08`zl!78>h9BI0tUZ8|M(26e$*oshNA|nJfMH zCz(Cd5B%+_F3)~48 ze&_Ggzz<$LtkCP>%hUM=#eJS)(>4}5;B|Ls;DjSz=TH3&o2YtfJI3pY^eXg0i_9N9 zI;=&vzlJ?wrmv3783KhfXRcHwgfHWoaK@zR;*_;F04{pgAIc+Fu5RY2wIfr`{EM(v zq38aE%agSkpfh+UtIxhjKVeE};@k8`p^d4jW+5twAJ+n7_ZC{>_6V?w!?-{rtIgpb zPhA_VW*NXv34vlPa;UO^l%-`b=3gH!DIsSEu-TSbv?oE|xTOin~3~E&?vbwtSbs?@b z2#+y?h$3rr^Ei4sI_?urLo)>oMlIhX@7BtucdNyCwvy#M6V1|LK0eM1iw{n^?zAww zPC?oTTV=1F{RB1O>oKa{Ns_&K%!mqWs!Z=`Xb1<PIr2@n2odH)7=V()&kICc^3mOqJ9X(oQPujK_B`ZTAS)gU+U&fW-RZ?i zDrc-?Eg7_A{^L1SRsek^5<7*rV6fB3r+Rs0r za#<;OFnTxIFVUAWz4B2$gYafLN-$qj>9S|Ko-RbY z(a59VfZk&M>1IUg6J?a$oA>6mpbVW}@ONIS*<+=;5|s~&X&7rs(2{HNWb59nvBnGS zU%%-@U1dR|r|z!=Y$g4x9lX$Occ7(XDYQvrGg%$gG98>(qr|PC)29$B$5EKBPuQQ` zy7n~vFFeGdsw#a2m<t!oD=S4-i-ovUXYZ3%6BKBT^&P2^Kca3= z_pmw|d=P4K58#g~o#OM^mL(1l2$9@)d%3~YE0_xBMr9n}E+HVYR}^bDixiC9J``M* zh+;TW7HM2i(uV-jXD_eQBKbSbiW&`@^Za@_os$nZ*KCcqkF3Ws>!aA@wc17GXF4Qk z`JCojza?FzTh1l&(i}?1c>`-ki_Nv#6V%M$UhGhmm8$*~E4oju?W!UH-!S+w&^jTC z#N4c)$6aZMUE7m+@IYUdD)OUoF|J}!(`zdS9fV)B#O`(&=1|AE^dcieoPIB4|CukI z9{AmLnqXH-{4F_^CctCkusA8SmjV;F5|e*m^@Z-Es$>+6fU}@@@)RV!T*ZYGWD>>6 zSWfPe^SpS3Z*>(`F~FsC?2kgTKD9#)uPga)>7b~77sl-392sNmc4$Vs)^|zm%NGSC zhq7t=eA7y|<-y8JvxW+%{q|IJ0>rhRw2T+VL5Fq8$01f3^a6hV(hBO#fT3x6hyAD! zyg0`hnL@fv$Q01lcrhrIwF{a3kNcX6v`<}e5A3A}D?PCW8WPX9Rc+GvAF}}rf*4kl zl-doCL(WwgLkWL<)b}BeMKv6AN$}nRvyJB@J5ak*i;FP~z2~ARw%wknra*XJ>=NRK z!829D-UTYBlh5#DI<#FZPidAYs22$kB_!*FPmB6)FyeYJT!4S0EkqfmX_L{nCGgq+ z)Cp>VVI+Bd;R+2=b&8~YqCOQwE%`o(@bf?k9-waV^zwQ9etBl>EGXX8yELZUSd9MO zslGdeme2rIQ=`bM{!9%Zq>>Qz3$X~rfC>TpqFfXjeT<&)AA9|m;O`Hi4fbyQ9 z9kOlDF$>)b%lGy`IbiglMu_?MO%hbTo}1s5W_qqwJz*KM7nmSjpYJYttqMrl7>DvE zvCW>G3P~4+WtC&=p!!+pmxSS?s-RTgjQt|WD*p(6aCrM^vOi_QLMq(CQ6#0nY>n7{ z!uZ^H$L{p{RDQRBr8UYE@7hEw8;8$l;tX84(dWNbF*w!TtX(w5*=t74_7w@?h*Cju zcm|9L+(5xZU5vg72%?v7Aq`w=BY~H0QACV8v-a8vFIh{J|Z+89rKZs9Nrqp zv>qYA&08XO5s3G9m1+w!(}1rf0|J*3Z5!L$VwSG#YG4%@Lm-;BcPv~2RVSgW7u8pqD0rl{4cC+v0Kmbx zj=r^$@Mflfe5fqH47mYA@{k;i_2HuB`%low(QKQbNrY8@k8PwuVeA|rdPr#V(&a(C zLF>gT{({ff0c#qohQx>@wiQ@;Pmkux#JhE(*r$`sU3_!R zK3iXLZZR%4B7mGCn3Vj~i8rGMB(2-QA`Kc>7=Bgod^CLBK| zp}6vaRKsITj~~GfVc!cSvB)f490QbE97OuM#u(I)mfM5?8D%=j(kN@VL)(zLPR+6K z9Tf^?QAJNeoFf--zb8Mw@pXwY&@tUAd=Yir8AiDKSn=47=C};vOK$v?hrvCHuLAvx zys6=CFSGf`t|3Q%-c~fwfo-m+^ZdQGC)yO`VzPO`5Ouw*Mlm0Q8IN2pynxBD8~EJI zn2AH@P2txGTM#?C9ZA&AhU&kN&+Ce&%^Hw881T|n0Po5XH^%q7N+STz?ZM} zn?6E#SlWTFb!|51ORgzQg9|Ln8itPBBd&{w#_{wgtYlI&SW5rm3)dXk`I%f_6HcTyNffB?QVF^b|w4qyV;8)-Yiznv-7a11E z&Rv{c#Vh?5NNWLvQc{Rv6FILXm0Cu%#3tDhFpYJDDvXDy<9(cip9qo@IdthJn)LB( zM_Tk!cvVGjjRj8NQMRQydJu2;@qMZ9&lDdxoYr#np{M?JRxOt$1n|CB`tZ6$k33V@q z%S)H`bjhArrI|B<>{e8Cswh=R)(6>VXT-nV8@R;%L-_sqUiM`Pp0hXO7%V@rUP6?- zGK?f5_I;$0c+O#chTQVqy&9*~`llW*OnFFx=1la(h`6sx*7c<`|3ecy)pU zh4R68k18L$;I3ShPAYYgD}w}>gs@#8#u6VxJm#+Dz{zEttxzc;+f-f;{q) zh%`}aPJXj*LUCiuB7q|^0kDHAB`e%goiB`Udd3&X`ds((EDK`!qB0r-nK3m`;HIHd z4ArLL^2Ce~7p&|%F0`Pr(Q!f;kN$gSn_7BZKtwvA3Ukx3^9oCzh*|y`Es5wF@^OgN z1nk4aM5RqulmoJazBja67!9K+I}0CxIAS>x+w^zZ6edd^s3s03NtOmC7$>BEfIM0) zDFze<;i}F!Vw#;pMnn0D69~(mIR>8ngMn0JgBXB9yCh^Z@ozux(UsHPyXnM|dd638r@suiQh%^WFa4?6@D`Wapqi zl};Ii&J{yGE4BnqzP-8H6=Qf%*JXO}xC0V)XgpkBsQRr!*?82G2uN*HN^s;KE{XHd>Pn8v< zXLq`2?u=j}09i{;5@<2!9i>%Oud^D_F^mQ(eps-wAA?e18&5PSi7WjDP?w^9Fe}eeT2D#+r~sM8kOQ1xCKSBDcJKn@NbM9Y`bi zC`KiYIFz1DM)g{%P>&O&Fj|M64R+~>1wE*_&9I`Gw`RX<7UY4rHg&1?J_gzjpH1FS z0%Cnx^!Gc;tomzj=?@F1-v*(ViFAvLgpMZ2v6e)i2IkM;zFC4{m^DUmSFubh3v;n4 zRd)a0IK~6+x`L17VD0^mpln7pzD7Nza!^CCPCv)MGCD{`*8TJBFiqnjmHIId$>XvF zvjy9n+ywja(ooInktb-C3rgcN+Gf(P(@q)If-rGMZW%AY8sHDz{mA6N3G$$w=q((*}_q4j3+g?%Ygq-)f$|W02ayh3<6z)?lR0F>lAvQx35?sUefy* z`528!YU2!+rfbtoe=xIn$SVM1-rvo#(7X|D(kI)uGOLNOvw0 zzU$!RhyK|f-TMx^uMWUW1{87R6%M{`I6PbN{Tt~cA6_BfsSrl*5Twc>fMl;|aWk-a z)>C~n)_$l#wUX|sSP_PNrAt5_n}|yLeLhuT$G@`E#$-60R_bf?y-%ybbe(;ZWzY^f z8DAmE#=JvCaR^Qv7N_JCnKN}l)kBzn}GdS;GbU)f;4KpSDH zwPojY`@vpRAKxk8V<@{jJA&Q=v?WNvkrIYJ1V{wFr5Ku|1Mf0oG=~y@Frvt6sv--P zo{;}SS*zSKK^I!(eg@Y8^y7qS-&(}^sS!96Lw_d)CNR%+o*jd_?var*X2D~tS14U3 zD@d33Jl7N7sPLZf;|(;^uQ|~AUb}LUwbWQ9UrVsNrDOJB_ttcLi}Jj&64jnj&MU*wWUjB{s7@|BmGQO_)drGvUJUUheCzd(R|z%7y=_mkcP{ciZb;YarW zi62@1kEuV||KP{2|BoNH_Cp~H_!Mtt6FaufQ`fO1vFo{WBVfY=Nt+}EU<035dwliH zQ28Y0tmOXY{* z7R;7=NY;%iVQ4oZtjd~+A36~mYy9@U+n99HOqQxPGW}Q@AIcrSN)q+>wo4 z&QO|6=by_8)u$4HnfE1|YsdaJbaPkr7Un3}s?Napv?|dYX347G|k(l7@9d!fmnOJ)|fWI*FJ|H*n(nBu{e?ZS<*E9$5m8QAG{683LBH~4LtB!?nDsG zFC>#w6iAsYQ*A@FvSF~d#AZmUxPH>)a#>3Up(K*_gHMH27or?E#tehGveLlinKliU zm6E4gJZLX0F=K@4H^GvcRpm-jtLR|kQaenJu&Ol!iVFdojX)BV%@o>rQ}qyjRFxf0 zp3unoDf6N09~_vppR#K;k(sc_p%i>Gov=a}XdoRDf0(eqN)=NLhkvz$zPJX(f8_2M z&8|@?zquV(RB9b;m}DD^cpjW(Z0jNyp=5jN5lKT>DS0)PMQ}xqIXk=k7tsAHk;;$!7!4QCfBz|v(4Q1#kmkip2_9Ly?i)gjpTC> zIfD14pN>dz^{8svqDWqmdYYDqaQs{L3#!T^X4cHszJ)q;J~v+!_a6%^*g(DN$;|c( zq`Yt!ln43&*JYl$++@Tvg?vvL7K$(Z!)ID;Ma60?#NyuAZJV#_i0q`HUb`(1faASc z7fdjg!~F&8R4F7wb9+)i@R-4rClQi;YYgnBD=bno;Hlpf(BrJD@iV%gFu)<-+vIRfK7qF85QPiQgAYf#oMQQ#-=>esOhT12- zVuyK-_gLy4N-dM-^a50y{?KN_yocvyJ$GM*HS9N3;O~!`h`RrQw;ytCd{gQqiD76O zK)gBlE+r_**P4Ix^-GHu*T~J>r34T|v>F;o#Jj0@_r(RAqjv0aIA~rkCKvUDB{rM-N7+%YWvc3h z$$^9nKxT(L#zhEo#^vI;pq-X&i(}!WIzK-J8BGAGQx~7&O;k-Esl3d3?=`Jwy{$kTY;2G+4vwn0GimixS3N* zX)*-?i@)Rw2-?Q!TyVe+T)C83@L20^OZw)jZV$wZ%*J#gHR*DnpUD~PQpU$@%xA`}T^h_ZX)(>3zGAz@6J(q>EHSWbXRq(s6XLKRi)Yh}cU5!x!&7)L&;KH&=B zqINHt#lgoMm}Ba}CZd0Q;K%sz3D)EZ-aItFE5n!Ox&bxtn30%prSC=s*JBUf9 zSSbIYp1_98HF?d_FfB#*L!Z98MsWT%h#>Hf;uwxe5WWwCRQHu4sA{%W0ql4#jN(W&SqtqP*HVj$H_n^Qq9ty- zi1GuJ>znxSWR@w^cXhH;Wmu6#X!32f8`RT9;hsk%I`PQC*@4pv# z>dbsUK*pOYON}O{LE7$i-mW1cB8q9=`nQ`9S71M6lLaBMy62^E5S!E;tn)u7S`6EgQm%uBhSfqey+KR3R3^!>YPG^TZFPrcPC0t~|d~ zz)LFogn+u5s=nkrn3v!2jPHGHrQQ2*U3Qh%;~qXF92kn< zTOUgo7^C3J*kBvSP`dGu9c&fO*bR5hI}s(q6r6y5+u9XGxe_NjgF-kL z>8DMmrK~Z%=*t5ye9Vo2qkjR>{oz59FTF1YO^W4d*LL|PSg)OZFN`2&gUW4qGZ7pp&n?@r*f&-}l-OxXFZS^B}v_^iq&!PReJyUWlhnad$Aut;XM} zpL`*TXigwObN)dE$1d5kpkR_<@-i^0^x5?)U2k*Vm0H`Uv}y4+f-x6-IVH(>`e`1? zH{q(f$fzF-zgAO=G*oe^B?not+l{{=@k)|CRoJt8%zQ6;$WmzR%?y>$*U)%5g$* zWE!FIF>QYrE*F1uzU=6n70#Mfm8*l%*l$r2XiP?$cpPKcD&mD6p+PY}wuL5dBI=e{hnDNL3@I%<{)WxXy(=HSwP^9HEnTWDUh^1>P!J*#` zetH~oIu~uiNi-b&97*M1)TCFK;99p$XP=>~_!he+1A(4RZ0ErWncs;oK}wx605}2O z2*QwH_F@XkQ88{y*KuVsZeyw48e8p_!5@iy>PvSgCVU7}%RfrK^dHP7Y!FJ8DGol} z3z>giKT*k!q;+<`#ss|1JY+=ig%N9XLxDzAq6o_?J-jIH2BKj*-JT@dm-(RX63}!> zv6gOdD{L#=;C`VN4A_&x<3OU{&xi1wgD~eXOz1oGpqb-Tex*P!EkD*}3!8xn<_ zRJ5x4~Z?!yW~4HZJ`X|Hgz zx4iYwud)ToB$YU)+p19-w=I##eET)0k*;fw4Ew@nPk%IB91QnVCHi9J&P_2gSqj4q znWJ`Nx3QO1LalL;w)f9ppFYjli$@AANb6e24S6r2d#m#HQXG!O%v;dU$A&DUcXrL? zDoQK%91&TOcDB{2m?M&&r>A;TzLO_q}WAzb-7@i>Dy=104 zSWKAdD*RI1AL5vyvQhSZZxi?3IIrU7Zu;ER7GqwZktkEv==5VFUdtFD-gm3s*x?@3 z4UliG`mhcIuXRxrD7a?fKc2~T267Re$kUoHNd2~P+JXJql%UMD*Qf-t7Iavp2S;8*fBH#ipU zs8N&UEtciKk);o^%y1b`igEk{s&fHg!)UfE4)M-PaFv@!t6n@&^An1!N#=u2HT!&m zVj z9h0+m*IcENxxLF%%a0oV{g&|8NS}Fs*SiYy$&+)veg81SA$k8m(O;0;hC4cyKWje= z#`W`J$`vO{*|vNK`wk$k>IFGU&2txf?_D|bI z(bK_IA81UOJjq31CN4HNyiB1pyuw9@e-rO|GgzkGIj0aWAyxLM86@T=pKbH}ZPi72 zvDS1OE4q{2@2F*5X-+h%IR$(ch-BVyXlZp0U#)aE|FZ0sL^N>@)=Zx8?muSN!|mCE z@LT!mes@ETVyt62s?t&1u->=SY|$8X)5T(UUF$&u1+Op4)y-1DH2`C9Vb>H!H!RY{ z8z-9X5EN??w+G*~FxzvRv_(jQI@A>sm-f?et1Rb+%q;d+Z4T0r*So`Xi~)y2i_9sm z7wpz)zmm=8Vovi7Spy5Qw&n7azs56+{ge`>*+CrrGk$-(hMUTFxub*DytnT-Ek{bZ+Ax|lGcXo1r>8-e=RmY+0ie_K z-1JLup476P_=dlWy=MN>bmw9Wm{=+ZWb}TUa_2J5g)|R^D$ndpT`xO(^g4MV2}Dej z?#!&RQ;Jca=95)eg`EGfN6-gj@9>3SFCHw-uOiv00~~bjUnxtXmivU|7S->6X#cSp z^z>{eWd04ETA&6Ht$BmTOBo$p$>Or0q6vh&@OX6LPE4h9*gPQ5Ph269G!yNf=s>+V z%oGEBB=V>tEUHH##xXvf{iT3Vf%=VIG4ZK%&dANwQY3_t4}pSB;5&mQTLYOMy)x}w zRKOe|(BPP`n!)T#6nE$#4+`w})DF;3mxf2(m7ah+pI-Ms?+LKYcqT`0jnx7DfX^3_ zB9m=S61=+sBpBTl{Fu2R_bB^sb{BJQ&uoJBr;~gI4O&i`?_czow@UFpeFhGj*GOtn; zWHnG++SoqfVdpa)BYSW^`2!`sTK*e$GRGil8CWiXo)%}yIOsO<* zd=od$uage+)==j*WU--vN-LD(aO+4lX?;Rg%-eKqJXU0l!_U^kvAs#fsQ)zl~{H`5mk;U z;8q@jrm8$(oaW;TS{S`oXa-S?xfgmgnKb6xN}!(afXaFQA!^57aIq1x#Vp?f zJmWq|np${p=C-JNr1wJW5dQs==~pAA`R*mrPi4&DH9{ zMNo~myBK^LmGw`e4*=>U}ez4TGX8p};s;0)E z43MnLRxJ<197w0Zrga1c5$w|EL{jk)}f(^xfKZyUbMmB>u~TDRc~dyZYLio3`%KMZ>WmPn9+L z5dA0uB9r(&`E0jN%7sa*@m)o{HFiI+0vOIL(?7df^Q@@~DomH%ZOr1bs;K8M)LP3J zZ7=Xm$@!xj1;|UU#+~{#JW-rW>9uWW3^53DTQX6PJWwGFeU!K+bTUJkUX%s{9U25A zrd}s>e4Uy-ojH-fbJ@W-nM2U!fI71=@ch4jnoGn7Z~)EvL(iJ0+$!(Q&Sg>V9i0W} zVRb}1WlZ(!`@SleHs{&ax>8cqQcR=To6d#No>cxnDb_6Q)2s~%d&5C!NxsJ7|Ov~gD>N@!khSWQg( zB{QsTTF=qai}xrvKzM9N0c|tJIl}XSvr97d&C}gqw}xJVx%c34Ds`0@VC3)vWDf~` zj*XK#evgQMiYFwD`dZp=K-e-Qhn}Gfjfl=H0QgvmY^26RjU@$0iNqN}uZ$I)3XgVI zu3yzyFTbH*<=4Exz;8|t!>{7h-9KIKzpY(dC(Y2=EoW_)CG=gcY^e}dFn)cvtkbOr zAYH$U{$*(wM$pj>P`a#H5qAaU7;A1J##z>4masWiH@yUM1u1s?p4|%a0Kt~}7jx=4 zC-O}opoON|W_*Q4fXR;QQ;BhnfeK<{*tT;`w05@3D~8}JP$(v+V-Z#ots;MCu_hRW zGeQW4GgR<(<=UJ{Ke2S&o*5@Ju|XN;i?HALT|NHo>$vQi0;Pc(Nmt{KX29;gxq;TY zrLB|+rq}~yUI-dIx)X$eaOm6Q<>YG@{q@zM2u?%snEJ8B;&2U?kx);W!PF+W_6dg+ z$yBx@;m|{mT`Wq=2rcR{AR&kEkA2>mtEbpkP|twU?Cy-1m#`Kthm0yg=IXMFM&*LwIi# z$Ak%71l?w2lNFqiDNKFp2Wh8?=;(wtoQYF4Joc-wEqBFLlEMhnqrv>bb*+X--#i_} zL?qK{t=e#J*Y~0ndr`nRSBQ&QSBaYTTX~$2r&tMhW{z@**==<0;tlm;nZgP9b_tiJ z+;*2K>Y7yJvLetqWelHywQRFM3jv%ju^x0NGJN=27jU8~jWk0KAlILJ9*qk)$hwpc zg6zXfv8BUP^##;lcgD%S-3%`4w~rfsvTF{R=qxcf0!n}(!e$6Tc}<%052n9NCUaj zRfso1k4GuwOkc@V7$Cq*C@`i^eqznc@9oe!!Sc$qCMMegBhNXu?dr*{)icpS<3w$% z!YMoV_ngD|W7=i)@N41Qnq)WC53y`5%gbEy)rE5P%D7Yku619&);_|%4p1;OPI_$cA9kHzBcElIZGFk4rCf@ATO3nK_J#q#R zQfUj-y=t!t-{VK|f>;0)e$oj)h&3KKR#lX0Y~4GvpDFc@U3n+Hn$OYlt?M3~_OO-vfc;Nfm{!_*V z?gVJk8zquI=xfaf6=+4lZkah&!IxxeuZ{L6uA(L?9U55J@jl(U>a49Um$(VZMVP-b z>`nk7hdZKfajf}=+nsA(O6LK|-yLIQw2#%8Fr_BOO@_Ky7rWjCu4Zm2aqmyEgkO z3XV{@z-7a=Sr<`<8UI7Yb^)i{ME*mP_36S&ozCr}g-35(!h(^Hn>qIueyACyw@u`bvD`xSb zTWk|cCEAk!1$xpfupXCi?7uto36RWCDIWXiI4-WO(x+wr?^?qp4KcKlr+aGlDpj7D zWahac=ce`OK;xREGU`;O5m;gRywZskKR2vhf4fWMynPpx>vhZ(=EWrc$VycgH_R?> zx5zv3gC%w1uYfo6B5@09BP_n6pM7Om`jxM3qL$~Q0Xu}%GxNad%;sH%NR1-F(R)(ekkjEKgt z*Y=!PCE)X@r6+L64=z=xgO3V4-#qp-96f&?fcg(P*uTwb$9+}>jdk$>FK*!KSN2Qf zlLR0ux@?GrI7s)h+Dm8o$vzXoL77A?4{_%U$cAHk&1^#n??8V$*`#nNZ*Xu#*K6H7 zMYoR+MIfmrCxFp7%3_?BxAN?#=v&K*V0|DjfZ%qQEVjLC_z=M7o@;gf_#Z}}J*1ih zNAP!YcAoU6feAGIMdZVU1!%Iaguj8(fG^((q4DN7ku>wgieE|#gzWHPscrz1Q;fmP zCfHAYq{EXZiBFKBlTCBx0?T++;sas&QlR59V3+IcZVC*+)QcKN zMoFkkpvfb*73|uEAG=tNAq+xt8?nb28?0d}5(g5#$TkNqa(^htRXDWaH|KQ~xEz$ihWtcLjMVHoEXsQf2WZa-JSJu8w%Rjv$A zm^mJ<7jv2jn|&tIUJT28nM+r40!(s{)=MMAG@d9T)S%Je?N3{-L-}RCnuwSXzcbD# z%43SZfGwQu+F$78SKcrMBxKZ@QuI3D1g6S9G0QNL%y2>!3TFD6WvJb z=Fa4BtCa(oss21TUytIIO%yPdaqQkdk`n~QI}>#6eeJD=qjyKUc7cX~g4<|dj89j*A`87ZRM7{{MKRN?!a!} zb6A2RMd%zJO@vp~IksS5vyDq?!H&2@aaiK~ps^Ok`sr`dl-*iclQ8K!zF;q@>Zt|r z!UJbaq%gmNob=r=coI$}XEJrUd!5(k1}X7fRCCG}s~8|WXsSBH=0dAysyh^Cf8ZqQ zmWmtLoDHQi- zGn^Bb8$))AWeZ#C3~LLkz3oP?_)**<{@x{knAt^Jr9zx*i|ew+&g10?174^w8M%){ z7hfP1Bs^P6fxir_pSU~zSmS6UNmTUP`y>8W9J)p9f3kcENZ8^ZPW`(5&7{wAu|SoH z^SG`yVZ~Q859nDi^ISQx&;k-rdjI%5W4nN>8Bv-}^0@b#?`Pg`IqNy37|bz@#Fzhu zZ>lnE`)?49^S?keHV%gWy><3VOWJu;!m0bACJVQ3L&d(1aIqsFThftaIvlHdRM(q! z3eYCCov;8KheZqP9RJ;%qaPjL$Kkad+}unR&;oHWnum|8I37!Y_u=l>j_~w5weD^J z=V$N-{UYeo!Si+ku&g!gu4m}|_V9f7&YBV~s*0f>N3V+;g7>w6nn@AD(Ly#svKqd2 z8T|6|>22ZRU~S3!)I|Ib$xJpg+Pm%+PF>QhK!qZ7lBKOPrdFrnlFVwALx{mjV_vQA zs}V<*W{s%LWQ_wyjwK}yfA$2;jAy~wgyuUCHb$4PnOQ zZPkAS2_vnxs>&TSC$V>*79I#~3QaYQl|lUY62Gmp$P%|}zpb;-LUUZqLwoN6C+(xz zmphB#Y4GD2K`MqxVlhWLqmz9=@YR!h;sH*+1KSYFpFCOC7|b)U-rZSK4^N%f~pUN>MB(m=_D5l1)c_e=L6qBCz{gHe)zXc!BRdz4ac_;m-z5I zXR7@BsOzAN3rwF+7L;2K$&mDm)PozFDY1V_ zh8mqDgljd&*}j|Bw{#p^p~Kds!}q8|P6+zH;5zqr7`nHLr;RJccc_^qH$^wncaPZr z$JjYFX##X>wrtzBZQFL2ZL7<+ZQC~9vaK$=%eJS_%v_v^`6A}rIgiS2D*S4s%*{x9MV^BW%%dkMv)R{ z;NJ>9o2k*@M(89ifm*9>0vrR8{fhR1CJk3XfC=g$jlK&gFj#RD@Bn8jvZ**u&RWJp zF`{HlHe;dIY-^OH5VXv3x5YtbTTQ8rYhDBU>Wiv69uV+bLbVPhQ>*swCp#|?p_lHq z4fj2fee{|Oy`@n|KvU`ND~2r-7ma75qXE!H*Za!uG%bKqjS;Hm0q75tNFlQ3y4|Mk znqO}&TS?Ej+y;mteCQp&1zViCZ7>$oY>u^ciVey!=sGAC13tfz-uK6~W}3CD=k zYt8dYJ<-838r*l21I@aNft|q1uCWrV@0+E}j31nNm@+XI!nf{Yqn)@36-#S`?>gGE zG(e5|ohh6Ptpq+ww++m{XOC_$bYL%?R)qUAHu3ot&mXdQbOiz$$NHiMO6f2sq*Q#)O{GyDb2FI7Aavv; zP6;oa-fbf>;P=!MP6BUh`T(?w1z`LFW)I5}W{|nJZKQAFiGS>l1k3Lag;=hn)mRjF z1O!zUSt2H+roWd9;!7^c4gR6Pp5CDEeC6Lj!Mr{aXlHP?2R!m>ydqRc*1tUt7V|6@ z*2($-n|!49Zk#yhMLk2;ROVA5a-!I}zMi$24a_HbHJh}<1|5Z=rsf@dYf+rT?s1<-j2><#qDuYg9qbc5yWAx z1SJ2}07A5`9t@)YeyzIX7w}ISFvFklnk?JE-s*)ZN zT1oW!Wj}+dKLCHiK)Nn@`v|$FRQKAMhh4}(FLkd?*T1jR+k|>+$99B|6 zM}~)NETP1Gc@c03!5A3Ue@uQ33?9hWjd2iy9yrjO#l00yzM_qcwdDH~=*Fd>8}}uI zWLyj4K~RpDy}a9yzJB`=MKKPX(x%>CE3mAN1y_o#7nT*IoEHSzX6B`FnJ+R3Hs6~~ z5g0nNWhhXe-DU5N^rB1VH}-^5q|rXebQDnt9)w8-$q_aZ%C9HXFr?;!cWYWmmGOx& zcEAF;-Lg6LC6cA#WtWZ)nQKEyi)5azVh0K_ zPVZen6nE4Lgp^ebFdNTUKbwazN8P+f>_*vt$?atc4hL%|oryFl`BVhB6hWQVwVs9! zQe9hVs#lO__DQs8Y9M{lSXB<`r?1o~aaCXzlKPe!SL>GEeufQAObg&qs26D=t3x>q z#79pNk>cM;p*5&g9WEfF<}_j1H?CuJ**Of!1wfayLctXmz5L->^>g&vA~7E!0yNd) zE|Qg3xGq=_X?s~UDp*Uk@0Iu|tK2B*_5s9-40#R0S0hvzzj^U_3nqN>@(q4A{om$% zDaZ6Lzg^f4s~s^w+LT3twv7lXbpzt1cBv071XD_0l#+gI#&sH&s-AG^9Qc${WkAI6 zxsj87qkOY~GN;jXlAY&stbb@Vs;I;ZRMai8V{-h0V}iFs@wSyZ9rzN==e(gv=%R^d zpODZt0&_*S#Ee<3(mI2%j9x1036?nfGsv0iIhYW1df-!c93N3>-Q_pQ+i)eeEVNDx z&EK8uE2ZL^%>CDelqA4V8tisK z-o-Jzo$zV3r4Ig7Z?ikjp&NU?;zZG~QJ79NBH>!#wr!_n+jR=1@0P`pQp{}&{r zfqE!5;;bLSo1qblUdDkyvLh-Kaif~#n6Y8VsRbH0J+D=SfJ{+X_sdK;y)RrQ12=K1 z;$20JPRKmiXg7`T+^8LwW}!+!L6R|52aY>ORLX@gp3UxPq0q%ri|8K^_sz|p%T~_K#vr$RO+Yb zvr0tLK;Hv2O9!>FWdpTH zbLEB_z~=MK3Hfye591_Z>{@)l-1Q%Y7vf;{20{&`pMq-cCS9 zN_aY(U<;uL)cv->6ABO!yY7QQw)bf!%iGeC#T`k8!+>%FmVqe{f*t*X&9b;Kg{D@? z3wG)nK#Q~%`@IUfH9^V|p(!J0x~SV{#yZPZI**}hNf{T$!s%&v=l}P|6Lk=6F9M*H zg8uQ182Noz-Q1sNPHaK>6W$4)fNA$c=;+VpX#Zj^KL(vJBXsRw|9v=G4wwskp3Z+O z(<=i(Cjib%5%hi8&~)>~go!g;pmM;<43Xy}qYT(!?fYia`sG&d<@v)iu+}tP ztacj=o|BvR%|^PthmXnx&>CDw4uhvEzjw;j*+<9Zlf6-dQD7wK)4WBwNeIbNYVHq| z2x#@k%027(++1*k0kc`COaObO8tDb>zk;R}Pqex<;1%&&a;0z_jX#CKOv_JFL@@!X z`<|7qjmv~j6r=mB8)e!u7{p&a`trhTv>3FM=WIsTLf!9xmO|+Enrs3c>1q4H`yg+H z$1cWR^-#`)(A)iPDE^q^+C*gl#IMH!%VfSs@KYY-6&JhOA}`^G9h-mgff;U?P5Ow+ zznQIHkz^0DCVElvAy6L;vr&0kMMrwE_>&Nx5n#je-A;?MKpgd>P6VJLY6Zhwcg2&c z{ZL7DNZ&D!Cm*CDloR%szOu~vOt4pX|eA`c)e%+_%ZDNZ1+n+0*#@Fk>9&rYtCPd(D53Y(I=8htF?L!#X}%)^qw}?1B(gmD9>tao3Tly z*Eh{^)&pziolqGRFmTtrZU&9qsI5jv%D+^ZvKfB;cwYS}kA&2x!Vt=O;U6f(m)oz8 z(v8XPFV*!n{rI81s0NJHdQ_>#NJnLyWd}&I?^;B@FC*k>hnY~qw#`;up7!`W!hiW) zRT!ETUIx3jwilnE8;}mJ(Nn=n5YnydkpCtKFUV`4U+3E|c|d9V-v|_gO*|5JK%N&3 zk*6id6js@dm!Af^G3O5ksxu#{CTjyTNu3R21^30Gi}HNm#(MA>aO!|-^qhBXUL}vk zryG8DU%;Uhjj4|Z09|jMm(iSL+!7OUw^i2@wg-IC?B;>EB*nq>ACMF)JM;elN&Rbc+Zs>&hoq8yPf5K*;a5Kj zry!^&Oyby+Brf%MMfNj*vXP(_J%S~Q*?Hf2zs&@$7aG^NZOt{%3BC#}lac%6PKki@ z)TvC5(|7+FJNp&1W+~x*Mt(4>Q8_xf-!7ob%adPsjoG>L^K<+Xj1OEKLNccXb3+Jx z4tyATkWg)W7+UkDH7VCe!lN}YEtcL43X<_36Zt%5wN5$v1tw-i|F^NTuGNgCYU?6xucOdr3 z0F} zQx46^vQmM~w)ys2jHOJJn4K^AP6k|0wfagyInlp59TPFlTQ-F6EBOV%%=5zLyLRbz z?A{*ixdSKOkoiyP&I=m(Nxf@ zA=f_CIPXT^YI^BSY!Y)d+52F2=^szkx56|tFA*;e>{aOoO{SU&lc~dO=(y){e@PSG!ccgrLQa(c2btCU6Lrydgbkj()oA@~u-w4S@%$aD z(DyszQfC_PnH%; z!%tBTbO@pfpLc|JPg$GK-!yzUYo<@m-89Lua^|V~3we8Yg9!NWUonF#<662$qaBZN zoQ6=ZmY_EAOW|)8(e9fe!ZluTC&dj57=wt5^h1*vLY}@^(fNC9 zu)!z!^?9r)jF*@HQgs4@P868Apc(`UyvVQDsAw^heEdFlZII!Fm!3+0VMf!ZU}H|w z@LI$>A*n+gmICNqcVuHh(#8rJ#q`uwV>ZO)FngcZX776qvF#m_>5~$9nx_D=k-$or zvPOPE03HP4iave8els}^8FH&^SgIHr*8%OOZ+3T^b9DW3sUK~U7cILW{OO;J7WY%n8#T0bW9Yz# zJcRd|Jp*?eVI;MsV5Bgy0Kk;TGgC-GF`W);fzskEOT+d4bhJDyje8zucF4EqbioyTeSs8}q z{072Z4!(h7JG$)4m5a6&_!%2 z*#JWarm%u+>`eM1Mc{9~kp<*$2H6C=5H3XPV+{FP=$QPn?aOna^PQjsYZ2HZVv!uO zx+kD)?cwxx!);%miSaH_A7r*n@y1p;cy}`Dhz>#n01Zn{Oe|5+nO5~)Yqgb#hbN4m zA>L`nqcE}wjnwi76GiJ3yryCdvo2QY$S@P=Fdz~M%2*5r%OZA5JCXSnq?8D2u=>{FJ`xw2=tZ{13{$9oMH;V2~3OCg8@^EdM6K|#ck@2D4{gi}SfCb06L=_ZHL`?N^R(8$+z)3yG~ z(%&|S-8`oOC4D}i`V{KRgB2P1JSReZ@SeQjbOCyOWeh7K0y(NLWeMG3wUWc32ym(D zm?aZTtU`4PbVgxo=Vz))==Uj+%Xt{Eq=OGjvtC&@OYro$+~p6oGlb&QMQi+d8SG{R z%h)y0mZZ52WLXtjqN_ux5rtLplB6sZe0iLIrs=WpMY4`yBzLLgG!;eWTVGsM!-5H| z^aP|ccc$*CT-GTr@f`lD5<(yu;cyjNY&R-Yw4T1nC?lxWa-PFO!5YYorgHJ|D5-!^ zjhodtJgWUA$3r4bJ%jX%i!R_h zBi(V{!6Y7(sr9MNvW*t?Muq<1MIC$1Q{b>+R1F`9{er9Ik473KNN1!jRPme22w++W z62TOcmRStF)Z&~Xz6<+$V$OgUP2|�GLYgit%9p zU1w3*${r4muKGn$Ep^W57@xPtnnv}xY<|fVgf?PicuMeljLT2xu_Suwiwy;}L; zz1k}}mcVKWlNgR$2?h8>+9hSUwZ37ZXh~Lzaa@?kJ2n(#PdSW=`?_X9ic$$xoLYgU zP{yn;dd2)`GOU0$xD0g(r{DnGAXJiPF^HNH3sPF_i3#i2qFRj*ezlzTwcaOuG4#&> zf7Jema?KZ5AIJN~ArpGhP}5VHCY_4sE5i;U>x8?>rvY1Plhr zn*`waQ)1YXV9!)=QpxmRh0ExmwQ0aAaH(^M3~Nk z_Pvm>>ZeY=%E+O5^EpQxIKPN~wv}~!Rnf1~tGoq0JuBjrWix$$ne@nbIFKkM5lbN4 zk5!67hT>Ie4T7xVLm0ulT;c0T`K4?r4XrSkIVzyRhXoc{dSY|LA?&;ci#^5%Kz*>ySPhoI>_^c z-tqwuFh-bI+pM=dLc~GILKpGK)Ds##9K`F_IC^BDQEvit;w$*_X!)C7_OB!fU|>s# z@dN3tc??|_T?WJDYu9toGt@h($V= z;CiP;!t?UQ+tnAs));IAIG`QumV8?QB4A{(;i&EaeDP`kEa}F5b9=+6@=8m` zfP@2clGZ|=dr(e%?u5W2zl{~O!*8q#MD{=4YJ!drSdo%)CS%D%UwQ}}!1%xo7dl)-qD?Jbd@cxLO(eyM* zm{L*jwTYDoQ3Q7-cm@c?xRU}iPoD9A4ItiRj3?lul0qrAC~i_DA%j1^iz;LE^y5d; zODn}bN%p^aIiZ(8jAZt|s2mcC!tb=l*AL1k2$67A4?s}NWcZ6afR2Xjg>x_!*U1!V zFJR*VOtW8gfS5>M3=<|@jYT;@AX)2Fg8Cl(WGa$)g(ZJJ7rJx zeU$=_4vX5fpt|p%T@doNQRj!J&oij_y67nl?eX*EcwX>x}OyQ3PI#sLp1g>=I z2GT>CY?p&Mg7;poe)}%0;Y0}>5g)_eKa=w`7O*7XsApGPLlDd~O?kaxql=vobA68s z<%B8fPVx_G{IG`W2u232!w|H5&glIOYFNCw9A}q83v?p-L(Ds$aA)p4Z;&*`WGQNoA ziQUm2A>L2b8dM+13*Bf^(+mGF)Anm~8~&j}KM`vP0o#t-@ah4X_XKG)1iUwNQ~DgB zfv2b~stFVXL{7hpYIx5~KE2S1P9s53Wm;A_9A$2R@y@Bi_B8~X)d;bK$##|@MP<3@ zfLLQ;tShK9TO#}hI+>}@{}zcX)RF=Ov{i+lgSGw$6%hEQaPJ4&yrq8}?S;;oESd>CXJGxw1>6ORCL86x_*0it+_H0O~TuAX-{y_4D40d<4$ z`xkItl(Oe38m#BTz*B+eTyEhC;iZs?rtaTjGsW{^<5tR;GhR%@-hP$+78=ev-ckO1 zuVrm5#^C|_$)d}Zix%Gz=kCKhtTyg7xMLK3>i8wA3(we-Q-dGM?fJsi8whaid*=hoOadrk(0@PP1UcC`ehIs0XWgU!)N3{dRvULCx$c`ue31zHu$JdbYCz zA>RjJtWi)2sokN)GYWSRBX*(cW9>JPqFAWzxb#hd-#7&2!=2$I0ngTIU|c@W1jgDt zw70I3lBXR_S&k82Jj8p|W7Og0sWWQ|v=_Zglb_TIv?lC8ux>8|#0x)Bedzajf)fW+ zGf#WEoM9bCM8zrVAn&STq+&wGVaT=Af`y8VlQTnfhKSQ*z(V+%hnYS3FzLcFiQjQD z<|5Dauir;MzzYap8vlu;F#o?Kg@uXz{|{3168I4YnczZSa!5EUWM7e0j=?b?kU$s` zIuIs>a6(xey;0Ixn0z$!ZKwR~-NV0#!<0Es_~uh3uS73d1;aHdR^x7ntK17egbp>O zOTy)=*75DSdQmS_0Dg;K znEXPkY&59?oa!doVSp5xLO`%|Fsvq0;{~utxF#z1iOUX zi{qu*44kprStYa(h3xH)Ufv4+ZjYOC9=ev|Inx#uJ&CgPwtw}o zSOszGa=vVb#clYfW>_Cf-R{5n@^U4jUCL*S{MC@DaDqY%JbrUrslmqXmTYoas2Bu)iF0!T zLOvT*Y-KBs5jB4BaOH+s3LF4xG>Vp~%C1c`f(GiY>YuZEsVEvcuJEt~C}JbD4KZ^k zaadt4X(Po#1#Xt4i^a82-Cr9o>4JHAwTX4l#3p|>Tyz}Rmi8=TSoh`fr=rZpQB?&d zuCQ)1x^5)w$K!YUTv#o*ND0?M?HJ-WabcFyRE08XT5pzb%HN8i+>ik(vYw#ke>^{8aC(c?TGZ zj93dab8*PdevZtQ;QC%HoeawYP?rb<=BK!Kuvt1@;0xe|Rb!f$Vl@5JTD^9`53at? z94tb({i2pv)Aea+X~{p^-zn6;N=VL{KDiK4h{44ffVcXQ={p1j{Ai=zfGYv+3tg`53aK<9Y)hVQVd}tw1Jz>{=L!3 zKv7AJJ0f%^Yug+bKGqL~z;;3Xg<{GNO{sNa)b+I`!Z#S~#M70GsIcGy3~nZx-y$Pp zxa{X?z5<(Se8Q(>Vd|0%loUlE#E7${Wn%Z#A#>OZ7+35I_G=H#o3Fq^BImW#kR+2q z96Gu_;nf0O#lO(9U|BDzu@+d5_bd$u!dt5p!&`duZUnD>zvsm4>gZ=k4O4<9YVqL^`-)Q9IsiGf zc<;Zksf~!MTrkk)8dH#NPFcIE=(_k_ZM%iF&UnjaK~_=6LJWOFx{;A#{--g<`rpRb z|D9TFOiXP5|M7nHkCWq$A^+p#s?GqX1eP@38*Dulw2|=^a#@T-F6xP6kOA~gh8RoH+dV!K|3v)ssDjJOHoZu+POA~YVvht6Qcjiy06bqk}Lxay;M#TGtV$IUjB>HhwN2>XC7a9VphUKJI_1&G9!mPhyGHS2=6`EC$3sXH!dK<&uYIf`!Rf zPlt9TJ&y~5_trr_l z8@+S4U-~Mqv1%oxG69`juG@;Egs)~BBX{%6<$VXC1W9w#YTx-edwtjcpxsoG*6B2? zpOg5dG@q%r=G3ZMjLc>kWWPj{MDvcDIk>58$Z4lPd-FUKki&ienW*|dzk)HM&w%Fr zW#XOfG37vw!DozLO~mkr@+e^V1(aLqap@H865bI;p9@B$g{(c=dPS}sv@F) zo-^J_;WVKxsBEv1+J^Ji{VuGCM(?}d`Iq}>7RQlnL_H0L0amHT$S$kVFp}+&-0q?l zE{%#_-fHG_YmI?daz0yZI<$s19(CX<)ozs`T#jN-zL|6rjxH$;94}KuEc5po(O1G{ zqSaWw__IxmkuKs@rJZQHyGBRPdd8>N1bMaBS%l9w0#2orq}xn74|QrX0FiB=K|ZOn zPtZWw{Pa|lhv%T7Ss}9Fl5~4lqpx=H2L|JboCY@r)~mT_+hqf5jA86bLLVc#D2@=! zDh%;2NqfzDO%!_^Jken9^UBelqUhY$^`I4W`42_1W> ze`>ttRdVZ2armU^V7wxHVOU_xbySS%9k3n(si0^$cUIq{GcDMS3gepJ4Jp5yD~D9g z#*2LZ)H+l#)3s@hQ4-xcqt^**kE3wF52Pd)OuN0D(@`bf8j?ITY0Zq`q=O%Oh21}j_IW$Dc^oINNY-_pJS5~^S*l#_EPo)GgN0?hJ&QEM2YvqqIn%?vP(#T&We zWkP6c7!N7u7xU4cSct2TFSq}k5v980Byq4`NP!7Zuqiibc&Y6wjOyMHL(kg%?ZCyh zVadA?LcvsjLwZg;>Es7%%`M0Tuo*g`)Wai(9uYV=?qoS#l}5F5Q)PSAAoImN8c-jx$yV-BxaU&Z-f z;9dgcA|Q2&5x5@hTEe~(y`da)4nvkB*F+q?jr2ix2ks+~N@iKr2y|LgC_uUJNfr)` z3FZY!INphYz>_7L*`31!zJyJSp##nSLM=&VLzpk8Jgghv{gb+I#|OjOlAK?UBX9~!N@hN!TD>9$1ca~=PQ+oPqXYIdNSYEPCx5t$nvWNn7!!!PAWoxZ1xf{2 zJwpv`#CZ43uv5)%PK;_LA-{%$J&n4{1=9khA;^h>iN@0NM%M%#sGqxQ?CU&hg9Cv! zglUy-hPvtS0@ZLLqK&gN;QZOIIYv$D1@dgsKnN8i!@^=lictgU%^qeXs03%m??~D? z*qZrLi3WmApNx@X03;p|%F^TEo$+>v|4}5uj7*`i^Vdn#C8C*eJ|pS$Tu+MLmSqso zr0?8g_p)Bvam&|56WD-Gt03Q#G(=aU=$)Yn7z?2gNC%aH*87)k0Ex<-FBtlkqAnoV zu;*>3N3>i!1UC6aaGP_g+%w@ z9=I;xuoUQ#ye04M z1^aOEzu{VXJVx?30wGG7Nt9Gzi^?X{DQP%eDFOOtSUM#bkV)bDaoSSLl8LE{QQ^3{ zJNxv}rbyn+v)BXIQpm#5Qn~|>3K_hKe`;WH26ni-JstKSh^`(Tt2E_Zdi^{-J)vQ` zx##|(ODhBu_VdwQ+!y75qk*Ogo%0Lk6iRELO#-WeG?zmdvLT&sO?%%);Ql~Yl;aIs zSQBkVcrDTfSan{gboW+RbW)H?E;s8U+sS#a;h%SIA=}s$6T}j=>7A|WQq-an7&3)v zu^9p|LB!q(z~Z2htkevn5kW{SqyE~3c!P;XC474FR5M0TV84!tPr|;*dGm`!B9y0Jqpk?$PgRdK}hg5vA(+qv*O#15I)>2x6@P z_+RNW{U%CSlstd7%a4ESiM55zVv57bbQs5<&&(kmjbiB}6n|7`>853yc@BP~U_j+_ z&vh|-nNtV_S8CF(hjDGkwDgU zn-T>n>ZG|HP4?7ZR%a2aG!o5j(i$uK(uY{0L&-G?uH0AK4|pCCe}yR;Pc~LxHqCAp ze-E*UpK@dUH52XrXK4kz2K)iL5qS&@jlOn@IuJJF#Liv*IA?2Mm9CN?_#F6RE0(R` zjVQWt7xiVmp`nm!$d0;LEEBKcH&pE|gfHSmP)fpCO+7?B!hy|X%4rV&8~afNQAt4v zu)C0O|CX*=^hGt};JoFl!CV$?#8vUz@H$-E~WQwQu&7u`*oMpP((ndo*apxJNaCWwe0bx4Yb z2ASa;r-~xmE=C{qx(ffD)UOm|G%Akf42eVjp#Ym18bqrMrHeI=kfE#}m#?Qb&l3Ixcd zYw}7*=&U3)t}aJr>Cn8;u`KWyNw$3Vx2>G(T{{y75zjGf4SAk$!V^O zk4Wx&QMIa2Dw!T0inrhr#sJ+r?f{WVt$kDlZ$F%3POAk94SJDzZ?M9dw<28wVa9iy zHfD@=O!e$U%g1Kt59b7(?4e2aG$ z7}0+~b+soZiv!OUR5JC$Ix7-f$joQ*xuM2AKYqN5G(m2?38^7JH5d{TdcWFQo$T zE>%1hD{H}m4rfRvEh|cqpYE%k9;fHWCUX&iDKAY<7|mR@0Gb(jcT}$1AtlyUCrMxGUn`i*q@|?G*VQ1h)SQHUpMtF`2oh0Z=D{X_yn{8_dhDq=UNGeF= z2jkQnl3P1svgxU($IcuV)Z^*BwJ$Gr4$DpzfLEY@^B-;`dN)seK4>9%^uC`mXo zSvOO8NfDy6e?JXfhHEqvCmiLgu#fGn1(hNQL%L7}%>yI^f6s^zw(7>pk1>dj)j zXapn!&EW(U4urT*fi!D?=&2JOFc!2pO$-CwTFJg>!U2$9hogYGs58M=hKj`mZq{}I zIxLh=3=macv&GN!5HRv}295-WZgam|fQ#I7B=H;W+)oxzq#$;cyMbpzXYUbq7u?)m ztxx7N5pj?ekRGF3f%p=!)hRAl*F#tvCXH2kGubppjv_>$cydU8#lXDKRgoWID>VKp{)mzKvmP_|!{O0s8vm)^NRNrc%Tl zra&f?@fT3(sy5ar(+urNhShnLmPB6bk=b?i#4sfL+!DJPoDLm7F(l^;IiONEk$(v; z+gX@!UcC;3zNk7sEOycZV7u8KIkF_)K-8NaN*8(lzA_={1v9V-IjVG?2{!eZe3}|! zkwT6S7woawul*(Ou=Rf|0|*#yrtoTL57E=kY>Hcu}A^7nleY&H3XERdF%_7G2X*wUMq)+)cf8POKkp5xcvyvRN zPT5rE_`Ti zNAgJEM2u`L&JG)-1tj(c7DoSq`QkRCrx684_w19_tq58ry1PmRUF&S(o*KmsR$wtK)?X>}Epiqf zGW~cWJBrM0UBM!h0^M!mA;oVgLEKs3{j26($4w8~tZf;E#s3#W`6Bs66?Cm_Vo&C* zREX!-3pKP7R3QorMmU24*~;4b#K`V<`TOJ{6oz+a58RRk2Y0_c<}uAb(_Ztp|3|%i zDufM&H5qW}1KKF=^qc2)v5mppRN33$ho(vykIN?Ik6vq4ecIE#HO~}lPWMtq@nyC4 zhtsNoX7gcPIb(~Jt<(17lkHMcx(p&2?}i{?Z)*Ai#^H(IMu`GuZ;p;rpHxu&p|3BO z1!7NDTy)+++h znETi67TZdZ&OMlRA|*efym?r%6cM!+I~wn}n`|?f;hK178SaV?+NV_L(jfH_xa$Dz zdH7M+je~|_d|X|_)RQ=m119e49pFe9B0--lprSZQp|f`KtijhB;8~Dl{Y`C{l#m4A z@CGBY+H(l=wXc0)Dhic8k{ECb&=BLEx z@|K0pTgd}wYK2bn`%ZD*1Lu#mKDB7gO3gq+L%<}44wz<-S$4Uq4lJL*=*J)TXlL&5 z#_=2s1F|d0cX`QPlP2;^y7na8#)@50CjH!GAl) zYEnW;=k6P|XAZjQnR6Ndr)ta3%)dB}9jA}3AD^jFBONKY^%-ahfHOSR-R@xXdO@@h zYlA^<8)&SBEc2ubp3JVCBzWwS`tSj@zwk%#F^chnqy|ijvU1#2hoqp|%MmpX0WojsuyOf=lBM55-Qo)qAbcdxuehBcJS9)~?jr%?F^?UH*KLMqaRPT|?I=NE$Tp;7u*UH( z%bNm*WSYD*pGbrTMqhyeOBzzp6Wh#_BH%2(k68N?bTh~fN14*oe~M z+0YP91~8|aS#&CaQxG5eC3YU{_*Ab8J@;49IKyG2cjQmU{bIfE+depr zAVIT{PPe};g)IIyb`e#ho--6OJioi2u>6fV+Qc7q8PuMwmLk1B{+Wi z=g#Z$kFV-f0AcuT+_ofg$%6?d;iecSq2$Tc*zX^I20z1T29&?vKKyJk&-}0NBj?m4 zm@437o-FtAf%Z!hu-FA74MX#Yi*}Jpl*zkU_SIXZ5D1Hal`J2z_WU>^^Ftw$KTM-w zpuvQbS9O$~pn_t}YTsI~iGV#2LrXwiAPlcZZA zFQ5EQlg`gX+FP2XM(f1O7LJz;b$7tY7{wv_xKUF)!zyrBA4J2sn)%{^H0~rMuiIFP z_y7%dz%o_p!-T{?M*w}6FOh{3o(?9~H+GX~^3%Pha=g)-lQ~lgws|Vc`M28k>}LK{ ztjQ-f%FK$%R-c=ri@UePft*aD7wTxN3cOr?8S&7e723Fyp>kA>tdBPa`a>EH!mM9( z3ve-eIs7O#=)<3oz(s7*>xNkXZpke6;&nmFX1q#PasokV%L!79N&@2u9BIm=Z-FCZ zDI4J*P^#3A+o2(}8AwcaK*}uFt2b8Jearkgg4~-AT_(6lm7s1duY)simZ|kI!Z!Ju zLoH%CKaO}id=4RL3`z*o>R-I&{{FKIad_?;5s#4A9&8(UQ6DTDFbg)ind0Z;!-EbF z4PVGyq0?NP5s#vTDbMJKqn5eprN@^9pr%w!!4&1N&H6G@3$o5T)?ZY!VGJAUn zpi8%q0_x3FRYhkIu5&CVDx_VM$M$_>3LI{re^Qx*s8LHFs`|R}auGt}lYw!c*`sUQ z_0)T{IO)FEo1N9)67aHQB&PtJn7UDP5C#DCWowHM35v z%WE>$Q$?fi=Fo8)&!H6!SN*#8QG>kQt;Az4>8fN1+%clL|Cy_w8Zp|VGUDId{M3D! z-+>4p!W+vcIgH;^Gf(tk%?CLNWl~jOD{N)i6adF@7CQqXl<1l!qd8IqPheGDhrkb| zcg81Rg_8QIAp}n=D4ow+sDkKlwPLu$??78e&~4L`hQ=W>mOme91nya2udKeaNbiyM zMF>xR`^=7!!)WOe>ZST zxAj9A6>faQnnKA%I_Wub!i12Et2kMpC-gJFUH&3d)AqOK7cnE9%%&YUZ(rxx#-gN4 zraN@ZPKc}RZrH9k)FPbWNsA5RB&^{~lyXi#s#`A4*o1L}P9oaAfy0*!y>#dlRTxt` zp2WIZHF#k94TGl|Q&_JsViwvjsq>cO*VKe%Q-SU`FFgi}`Zf74PCu*Uc6Cj5yJ?_z z;x-O_VV_CUY1v5%@{9X{Cs9CXhtMB6xX7u+6A*kdMT)upLTY$QBiQ4M#Nx}p;zQ=u z)E9jP6aq%Rs~qig#OkLB_zeOV&(}9#5+A}`j6aLL-Rj4q$}G$Hj1-R z*E$kA$P0Th@n0Z(Yg3RxU!!$M1?7eNGf*Qt%Dc*no~jI3x{5s1WwIJ#1K~q`wWD=P z{vXo|&G2U@ST-wY%TbAx0d3>krI6q?NUMLlf)OwO7%9e~+MU$P*s%g->#|v^c=Wrw zwP1w!yO}r&`&tE#@9`ep2WtQF$M=*8aed!oF;k|7L) zhp_TPTwdqBjvV&olGztp5Cx`(SK8|zBrAi5d~8=; zPYtyh0BmpNr$6T}M=VH3i%HkV9`+6~KjUOC!Va;r0%Rfa_HNtN{dZ7`21oKM>x!#4 zLZaCZ!%t!AvdK-^{?kN(Bz!QrDemaC3H8E(zES1zDt&Uq6G+k~xlZm-Ta*W5Vm1fH zKv#+!D8DyybZbUgnnj@y?{%9n7lZyTaS|AE`7B>%f?kzt;q zJ8mED_ij-Uq#w}&qBIDI0zsl*89?YQl(lcr%VbtfFB)Q5ivTHvI8H0#s`O4|ZMOS0JldpJ9#_$l zYY7!0WWr_w$ca#cHhg=5JEB055j$cS)OSSF@cziL02#YWCsZsqzEF(HUw2Nk;%&c# z5{?tD*^7$?t#2mQs*1j?Oq_*69f$9|Yoy9(tQ(GiHocJgqY?&dlaWeqWqSHaR&KE% zOl+^Ohq603({y$y{9jQGuHZGnzai3X86%yF2EJ5Uh#eA2>UDy z=}v3q7avrz=0*RG-X}97R>6bFeTsaw?QarO^?@7bc3r`%u^CQ;`sZM*Z_y>? zSB{{eSj}*-%2xbuQZ`5r+haO3)x!O)Sp&;eUpOx9TeR0q*_oP2*7g*d(==Sl59ulA zW?^?f+2P*7x7i>1J18*Q&2$f1Lh8|oeb--J@=smHbS`rQ^}skpn$ml*!i`<-0LbT6 zJXjAw;iKQ&!%Zz3{OdlAvfg9$gq_MC7nOf4a}QIc$M=x&oU6NFOrTYXROYMW*WP07 zyi*cuw*V)+ersdTOU7(E156hYg2uz@ktb0WVgtP=q%jfxI>>D^)xBC5%0>Rk_6i6` zW!l+|OHvQ}hly!#0V5~nB*&YOW;vq-t~8wy>O>TRzDE$Az|ndOfWpGS5k38|l+j1A$16OW-__X7708FtH{9kOc)~)PQ9t2n zdVGQOh9C8vo0`^;3^QaMCi$Yt2_d;4k1|!i{Ul z+JUc|lJa&8NFUA>^B@=J{04ky)5-sx{~DL4NZg5C;9opG`#d>%l&VymDp0m{c5)ja z`Nv(85iJ(m8J5xJ?$FE$$55xw>EgiI?Ftg~We}UB0XsF^n!Xara)Pvh;MFm!o@IS; zr>>~~s|lwVv3ZZku1_Fy%1mD^l|2kEXU<%y8ebSO^H)bqnigOKYi)4%UHv+HFY+JD z=99Y+0ZhS;%4WwKRIluf0NsK0whHV8;Ft+2o&UnGOb1JA^?3SU;D$|(=>r7K#~tv? zoCfU=GluviYUXAd_;$BumJUCqIkZy9Kg(va(~_n?cx>0qK2br|Ze*+02Ehm_tkX$5 zs(Ix->8#O`AUES5-w<@Xe6q-gima4Mg3*=10a>(RPxy~IqR3OMmw{=1JV^0vH+W|k zh9va@jba4kl5)Q{hgEA!N*YEB2WI+AN988nrnah7Q|7a|>2Y)6Q$xTK)Qr zr=+gm8CI?57_Q882yqbGMBM&Ok3 zK?IUB1l-m#w>pHgdZ|T7iVJSvkwdCGWb(kA-Bf~&Th04Od0DA)(RNz$?L>p@wqwao z$el*Su3xXVeYXV`0ES#OhUVyk&_QG4b!Ad&=2IPNNRB6o!EE;c$SQ%LBGg%dt7ES8 zps+>y`e}t;ifbo@x!SaqN$lC#QxjP8sK0nwhRUr;5m83kG?6t?5M*1{nFZM*Jmu*8 zYSb5d3P@sIf$)|7LMcXibY+!tw1h6!3~GH|Xc(AEXp;Nq?f7m=-b0Sk+W2J?+77xk zzQypT2VEA}8-dIW&4ILA10(kz3sDX03u+4sZZDz_v%g$coGq$qnG#WazWzM)PBvWA z>7iGCxax6SIfkY+cr4{a5@ba(F*0|6cCk;d6{rP-U)G0!PB;HrWV8m8lOC|ti20G` zH($QMuNKKq!CYXi)LdeBl(j72NZH5#t7Erj-ub$!uuBBu$2NI zC&XA?!Xt;h^vn1@8Awe1HJaRDcwK|DiIgJ4<_Tb_O7reZueRwJ;5Y=2SzmTzDa4v} z>E+F*&*wEGNQ*l$`W*^XzboZb)nzVWMFyK0Wv-v_WEvNGVzpgvWWyl&FRl?cwdIB^L#Qk&wbmd?mL!j4gMf^85NH!K8v0sBHHEYl&KW_GZrfPKaOFC$_^R z0<1qX0n|g{F#J6pHiH8|Ww-CeAQ9pmzE`z9r^zn8k}e<)g5N>a_>38bPm?ACu_`#^ zhkO;3nxc)y#pOH~hYi=$zZ={1V;VyP-di}_q)`-Wudr%5FEeUFvJ&pWEsV5!bz9|R z!FbyaDx#R;l#&E8f z3~ZOC&3u=Q@P%T4^Dc*%-k+7fJrq8`Nsqz0AEzSMC9VQe&a>}&=QYIcfaG~OOw-&u zvC%&C;Mhue-b`d~_)*%fyNh`BaQS1oqn>G`_<)P$Q8U|?DcPCAD_OWIsDC(iNviWk zTU*#2LMCFyfgh)SNadkr(r3JoLrF@2>MIbaPv5F)B*B3XMG1qqu2O`Gyqf1-=ETEs zj{5>$UsLj63j9g#GY?9&E|Y2XMF?@RWzvC2iPSyzZRCjLE-)@=p+P+H=p87XLFpT` z$#p|Ue!cbaX{3dmN8vr{ftO&*z*`3)r0X1(11>(R)wRBjQc@Uipfd1SbNS{_4fcY& zWdBDjQC%BRLvg;5c*Hv>NB-t{DpNv%#CZxXK+z593=-+Jsw&efSs}J&0|bID4QSx;;V&2KX-LYbGaNFOg<$HJAI=^M<1$y?`Z= zhKS{wh!mE6U~JjSu%{k75B$Zw?gyyh0gwwc>*lTQUDsx0FAh+S>yvHLkVW1ia1eM( z?TsRszsKsOkruw&cioNPR#Ls_aU!9#p*PHK&-h(BhPU7>7qEMsi5{pNJed0?{*1VF zY4_gwX~tBSHRq7^Kk2Bbc!EUW{gAH@;I|a9;gWkZljGXhPTFBF8u~L};PI%Heo4)Bq*d~5ZwN09Su=$+=?qJ1UKJGntf9lEva!9m z?GgYSjQdiu0{Tb(WbCSE>4-d#c+l}U8YI)oX@4jwD3=t*T&8{o|9VzWuD7<@1+u{TC!fpA(wyYrJhW=fM8sgE?h-_8A=30{EwlR^T%*DYX~;~?vSc;KuArFKX#8QXh9vLV zQKI4<(LbAAFm7K|D$g?2qdnTjYe#vNcUikUWkiCH@))7>i`=s%Fvyt0SH{x@hWFzu z`9SNij({+03~m?S93SMZq(Q z27qtBOFmzLEkOE#&fcsaI3l#h`!!^KpT(YJ$3TO^VM1YVH-H5R7P^V3gNRxA_qrLR zxX%$@m#qWzzN70=T*21?i#D8tH#uw54{?|Y1{E3s#+H55BP&FqXaqeP1!RgPJ^WJ2AKJ13DBC;d?o}!s}%o)DIotnm3GSTe}<$s}g#n5?A z=p8z&;oHw__oWj*Vg?bb1D(nOQ>>_nBV!D3+D-8hHFGBYKv3d;!=Qb^*Ck<*neF@H za>I=Pc}00$Oj^j}CYLfETsU|_BzOKU!EIkG>iZ`R$}&*;r@s}ufzz?p0%4vEStUeVIshgZ zgbZaAev8N7{qyqO|nd>*@%=K``B8D&@mjp$1% z=D^ftGufwieebPn{Y!@q{kS9()2O{6Y3873!$D6w;<$~Pz39rN^CLgcKEJBK2hSu| zv+OfUb;*iG3U_|a$5WcmQePOJ@w@x8|GGfONLnsZOeR4AI6#>V>w}40O!QYITPU znh0t;;yoch$Y?j;-ox@oAsFrfE5;Lpa&Rm(>Al%E7*=b(X#k^5+U9W<<7G3vBmz>< zPtHmd)7PV5(;XZ_m!9_)d2!rI!O9Oy8NoHBO)_`GU@!b-Ky+i?M$M|;d)h4vhmRMV zJHA!jIH#tAUcmufQbq|Lfkipz3HT-*y}a$WRL&HtsCKng=;`4NVjklf0(Pw%4eT5x z%izm01RhpU4|?m7w9y{JwgHiYR;ignKLsmC@p~D@SiL~kEtmjC@;6IVe z5B^IqhS-bjzCV!=oZM6G3Rn`Fbo~mc7XKc2H%Nz?xD?>-`anXh~RUsz8xj3|)v)_kA8pRbs8U=fPYzTL8u zwhb+XD6^MzE`RfVU}32?yy1KxtUI&8U6z#~`adU)Yh5d+dkJI-1EHE)@opSAaju|k z;sbTJJ@AF8V3zeM!#Tb1XGn|EpNGI|%3cJ`K*ZON?;-LGUVHi-FhpXyaS9%@%t+)8 zpSO@^x7QimK;VrEWuFrB?at5Rl=N@3K>&xolkI0Ta3(UbAm}?>k@ja~|8x_Z6gK>W zt{r4d0w2!hdt|241q3Cvs2ffM8k!(4J-MEk+W~iGYVCra!&tI26_Be#nKpYG>SA&H z&f59B84?hiOe+MWiLISj$5D9Xw2m_foG3kV>-lu4gR2}rQG84~cxu99Rb`K9TRVp5 z_ejXJ%k~g8GDf(b+vedsp;LJgQx>Edip=&LQC2SRT~%hL+G?XL+Ur}bB)liz53i@i z5rnpN9=Jqr3xe8Rg#t0%f;@86h`-L6y&_m=TJ}`b(tr_NZnj{9+rwEPAhLyvKDP%@k zQuEkt95tibg)#L_`<%=@qpuEM%RS^v%4(vaf!VM5l~a0o3*9oMgmpvvaiD9PZ0!B2 z=8Z#SyR?2?I=rMu{HQbf*CrIJ_FnZH;hJUJtb9Shenvy~pORPkKr$r{2*ei_Jw)fb)R_1Y%P+0}#{mM0%&8h~zznL?E;&DS0!kvp5a7U;sY zs#|je%7sBswKK;6n+AOU-ZM#L5asQ*`e9>Rw`X0OXUhUX~b!t24bR2u8M?$7fs_) zqn%w%0yQ!j5QCcd@IPM%mRbGm4PG-kIkuCud8(uQ}sQDP@!qm1jZ*j4m(wEp*HPWg`(n+OD_*x>;Z`W6?T^|u}^x@y2WG!6#^ORBT(gl9?F@P)5f1WQ*Y z>^+F?<^E(VbY#{MwOMWmNf#8x=NW#0#z8z z&@6)q;-AQOw_j>#-_3$X{_+&b=KMlv_r#NY@^FM~1hGeY*MP%1z|j#ZdeUL=m~>Q$ z5gSss>Jb7@>BTO~FoX*7ZK796(&=!420OU^%`9>R`-3YGx&A9iN>N+=beKm^=r?C1 z6q*!gFW`3gH&Z{&E^C;qUKgJaS3B1wOb-U293X<^h>963VcGV+zKQndkoSszEj$~b367LZ>A~~tFaU`x;A3XmDX#9-dn@V7E#+78F#pil{ne@!BZLm48a+ezVuZ-~mq=N#Jb_xM zq#7@K`eo>T^Ds9Sam3tDoPzzV(DcM-fA03mNS-GSz6q;|lIf3bjd7G)0X{VVoFJKl z;OF{km3y##O0rVP*gM>7q zsk*pxuzm!7M2uj&;W={MpyNsOP?`yuLOpoN#kKh{m@@)YC?CHU#aqP@B~;HwJz&ve z?u5fs`@!DWy$E^y??7tZz#z;iea=MxHFnPPKop;~8^La@Z$Ks&0Z09ekV!oh)lvai z8!}xss{TDUQuSK9CiFSs*A8&6*ZcZ+bZ_y@)cVTO3NaX_H|@-Dz=^xhro_OogyPAH zra&<+C+u)=0u5MsaI8NYbt&i|tf$K@QBJn46F#;sA*^yfp;)fys#&3ZR*(lEwL~-X z=I5eQ{~pS8wLT~8Mwwpw?eHttG`>LVSvR>1gJ-uW(~$UqW)10-)B1$H@q$?^Y0EyN zv~^OIwe+SdhEQ*}>cZozSZ8^KxNo-z~5ydxyZ2rUQW)u9SxWGHXinz!~W2$}0n z-;+LLQa6#(!lV5b!rLckTG45Y{zk!2L{Co3^lnbi3M~+B%}aEFQrq{iI8E<%k#3dR zS$-4lz*$Q@%F;SeRrOr@X80Zio_?~w)igPGNJVU&l>c>DgXAPxfS6ft%JZKOA`)nuIH|^13Fi4S)vTl`zn=olW-A(W_dXYKFd6KzWPzz7A6ZHd;YT7}*M0hY zLkUVM8S#o~4L%4oDfKj))(pEU=F+JWQ$e~4@DZz#xn2Cz^fp#iMexAW`-d3Xf~E}* zv)fbdP}wNvmg<45lA}35&}<;|b>I?^?Cgyk8gfh~x{q}KK5 zNb}4fM4On*r8l~2%^ELYQVG0}XGAsg%a6iLh?h`Te%d11rpEA-^y%j&z3{;zbMYOe zJ*(fq@cKIZk2zb?J9K0|)aKb}<`S+-bm}$+ULt>W!y9c(?P0F%87#J*Tija*;^9|Y zgvK868EEB>8w`+CKy4~B-Oj$fo}GmwG#R|*#NcVBsu5e6<`rtuj3*WrGAMp9u={F@ zEf>bw%RF|IQqGZk_MOq-KfC$k6t{e4Gsz$qdDFvKWG-GrF#q<6K%;V74{yrUkJ&hDe}DBm+h|Lt>@X;Z|AgorM@X# zG^@G>@$g$Yp%MWmm@F+|TD6@As{^hgrS!|GiLjlj&(_Qo+O$;o=?cP3f7Rw=6x^sQ z==UIu*-1_4&u@m7tLbC=tz#E3No_PeXJa7!*_uobNuXh$CJx=N-HXX8DcUg-5$1mv z>J}j5Szf~f ze)3*ofidhiS(VnruVz-TGri)EE)zdhE(PLY*wy85)G2)`?T$jTAI}+m8ho@fjRD{{7xlCJi zS6^s3K2m&p>MjZ&h41-?XaJ(VKjzRa(2PuGM~`CJp9z6aj{jzSxR>--t!<}e<5{3k zkO$)piJ~SEMy=s~*}9D8k-KFJ_~mrbj>lNwtR0WE*UcLvT{)_{&=96g0dcktz=$B+tA#He`rB_kUVXONurCDN*=lwpuH_&2b?vR^KwoBE`M}KPC#enA%8wPv{bhkkQZWM zTqd_s1|+`>t9a;BhsM6)gY&CbCdsl5$23{uL{skeW`6N{aldB;QW;aG5x)kpBZEHV zs7I-C%JpAGvmvLW9n4>b;#GzLdJWW0SY@KwD~RugMu;5Yl1`%yA&ukB7!b(}T3aW% zTnLo}uWv)rEYXs<0#U=)e-T=^OkcVXwmoFCC%=>F&WJ%od z6UA{I@*xX}SQ*YFq-2+dp8}@vY`mf4Xu@ij_dsR9Px=cFtPeAZlT$|(@zxM*og(KmL2d!ZsWk%4 zMuF+IdnOFNF?h|EQMR$8Rt~wj+NfAAq?e^jVHh_TVFz+s{qoTAX_Wg+mzOv zLIvggH2_ouF z6JrrAhb92T0>){WJm;iq8JwjNl$x>!Mz;l+(fx=WXeJOBHQ2+^I}N|@t51S*fEL8i zX=pDHSp^#LE*de}vOTGW@Ip>(AfgsL%KZ%Zl{t2=PE$vXs=TEBMO5o+rV1nwm>Vu3 z(m=z{kLuJ|4py-2^zyml11uf!9!;s^T6D*a9aJFO*~!{+hXCVA+%^4osr7z?-n}*5 zoICUcD?%Ro+$0@Ion|_fyXiIy5FFN#&mmyf`ZL+DI0hj`KLPQU#Ov4yrhx=AL6y?1 z=h?P!vB9cxn2B|4e8~griip1BoNZoq_m;6zUB!T@vI4xR_Hy&nKscy5KVEr%kIBQQM=R>O4P%S`JZ^nqn7$ z%^`Jocl=LK>l9rMxd)g-9|w{KGA3%|n4sQN#rp#>I&2%+L1_&Y!ZilN0KGgn)dm6` zh`G@W1iYUyylYx{#4*}vKL%CtxXz9wG>X<) z>}ktPVu(7((Q}ub;-$1M2Mz5xO`?fNerV>|kywk;=snVs)hO&Km-XgjZ;SWp^EowW zbi&34y(+fdn`Fwft?QM*(}%Qk8R@8f>T2TCHZ(>{$$fKzJ2Fn|^G!A{6%}uLUmW`n z+=I8O4Ox89roPib+=7BvhCYarN*7ofRW8S8_322FeCPS%xx3B3rGpT@;K=6s@LH-k znys0x^{)!1+c^BBSq^%l*pQc;rz~6$)c)q`+vc_SA(Lw($T}0j#z3WBSaQ|h;2?vF zujqdD^EMRjj%qvo^tnuVK@d1se--&r29Dd=V_Dxt1K5`ry+NL$qOh*#X&he#=ByKTE^K^%?9ys(RNsQq{0ss8W0# zWDW4R2S7w@je7I=qFPXOdBSX04q%2-_ZSbGSo7>Rb zUz5Cl)a^>b;lQ|Q*sW#_XAe-Hf`R33>*)BZN2;0<6INCV6C;QPR*=PnM)$YIj>Z0F@DbYY^sod z+;pWZ9To8I5DDk!7kIe$3{*g`pYiIde8Y65$?B1y>M11hDJ^V{nXB~Pm~~6%OVp8q zV^Li2XNJ?+zUzc1m4b}IhS*8NwTY@(3f!~g>T;+wJFLOx2)}I8F^`a-y_K!Rjce`@{mm#!5-At zLJ?0@^qb)v@6nq+77?*6m?8owi|jBUpaR|t6s?PZKSoXX&Lpwf;oB9n#}%K62e+lf zSByvHm^uAh%VrPK?8U;p_A|T&M$@##TV>t1hTlxqt64wiMYZ#|C%D0E1Cd_e$Lo1U zIdBlJ)Q8Xq(m{%H*_G}uka|n0d(y!7U7%9%u2!fiQ^(sn(S{!H>RZt~g$JP6W9a+EX%7XX^Z@ZZQ z#Fe+x_rpe;!Zim~_yKGP4?exoxBPU^L&3EwY(-49HVssBxkH#+cm+e-hQ#)qZ_Yx85>RpJjl}7jm`0fFP5EudUDeT=2sj4 zCT=j^f_zv4woPSl=*Nh9zUSPkJ4#{?=v$sZ7#j58x!YCfkH^o6f#CbbY1nZI(a&55 z1;!lGd{PbLL6pJEJXtGCbkPSH@}pmXl?EF7sBX$!9vopSuBS6vqF-EDO+n{=$k5@= zI@k9^Qapir5Z(iIx<1kxw6=$52hG|uXpl=*D-pnLWR#D`EFF1e;wYfSVqxiae zvUub&J7JOE-WJzG07E#GRQN?>c}q~mxs_9mUU4SMD+a`Md=zNib+>>2N%ZZxv{kxb z?#fJlf6F28@~J2u98(+157E;Pxl8ZOX_|wrJJjd z2FCz`vN!=HRT|HE^^-8H46JKJQ@)vg%I^t3SA>GUVYv4yq~n*VY3zJ*WHaA{OW1= z5>Oj=s?IM+!F)mDVPs>Lx4V21o*fLK6r*Jlu|l4bKMbih6(X7C5~w@t*X8+Oz>$Rc z|8$!cR5ZD?CW$3mmhX-$J^|p-JAMZIQy5B8uKTSI;Ibx~-P9W$?P2CAW_cK{T{e-4 z8)Q_i(WSGYkO?rNj@i0+UvClvK#j9`{kO3K2c(pum!wXO(+P6T(p6Wx$>IS0WoF3g z0b*v>mByq9aA0$>lDonD9R4~)r3`{oSa{yq`_GvIKhsv|38 zP&Ms^^rzrH63VaL&t$H^=XmC~e=i_5BUmKC&F%hf)`4m@ymLioIS$HQY|re*V)s0_ zq_DHFhY$SnYSeE=gcOpT+84l}gpx7HGQMoLpG#}*ITvw(X!#{B;{ov%>AmXm>*{)^ zwo-8rH#OFa$bf7s0Q!!_VIw4)?@z}s=`=_-g+BQ~y3MZ%_77c@>Zr_wlSUe3t-mwq z?4}POX0ve9zB%-(qQ{1m@CQMUwXsI z`{)i#X1TENFtGsR2BvkC`Qzrz3ymk2I}=a;d~r^sBkAda#pMPfmo3{p?%&?)N2&`n z0{I5Kaxg767rCDevKJ>OZ*2NIC+y$+Y{ywaDuJvd5jphr~9r1O` zni<2@%xq@|AkO)Hx&Oi)I3^T?gY6j?2zup7fx5gr@at|?2B4|m%J>jK^!0JwB9t1? zZm4?1efP+)JCK{LS;w8GdMM;`P|RCc7^~S{&?V8D_32+>78eE z{nqP_ZxXg8i@_`J9)a{=<2T%}qb%taG(&ICwL^}D1g4smC-U17Q39F@)gFpdWUOcs zu#Z|!p@2z_%|iy=*@rHgjqPh6d)h%yXi{#vFO0jQ5yT#mcU-%%R;SDaOEy2)olDXI ziA~SH?zj?l>MkCKIQE#qLa?*Y-zRPMuMntOpHx8h^{D>X1tScOdfTo(zr335NGON; zci*`mDUV4#9=ZK?SC$Wo$|b+3fY!|eA;!-?f=<*!#Gkb0kg)&eZBOhb0qIz-NDo76 z7SLdMDWCG+8nUR!hcVp7)_i|~e#oD}+PERlBWYcZsPndrB~SyYQ34`ow&Un058g$? zwyrQ~a>9_D4l0jvz~N&Z0~Up4Rm$lWNOFf|Fy;=4L>g+URBs%3C9ZJCNn#z zHlE14gv5IuizynV;j28?7YMmftI7udfRpRV3Z^Rjp696{a=!xwBeNrKf~m}cRjC05 z-aODx1@B0f0S6CURgH3hVqE$?JaGVlQ3L*a*}e+0DE9Vag= zuFoG^DBPowKB_A-U6<}7_3|L@SG#a(KGE7B125sIOY;6WhqN&!b7cISH}B(>&`~3_ zK)BX-&D!m!D~|WZz~BD7Hy2Z4^Z498+A)1HV&b*}m5m&J6}{ z!&6CXVA6t`PuqoTjx(sXHN&zNs5`WBNXUZt>hTG%J~52q^O!=)Y8QsFb4pqp;&zXL zBSHFt#wkR?`=0W-DQI5*qNUuxb4l=(-`n~4UUK>F{+xSgYqY{rg51+ydI7R?8FXh+ z;JWX4D2Pj>B155iLGP)TmM#6L_Akhr3jRG?-oZbMaSs@KOinx92>6Xa;^aV&QP=o& z7ca!5c~GWHG(nt2_K?v>6loOeF(l($NfuSwRv(VgUAD?HOP}4n?Uc7E1WVr=4LpNKJ8t z6_tg1zS}JR(imadpXV!K;Kh*7eFfY;_H!25z;g%=?vZ7_ZU5_-^Ux`V?K6QaVHSn3 zPA7H~plO5_j0ngAroKNQ*ClDAP&}_-?nP*l1gMC?P;kKAPJzntP9@K)SxFrzS>7oe zMagC{Og_hBeV!cutap$Uf|&3mYTsQOhktIh(Xvw-QxoM5yVL3 zGx9x`@*AdF%2s64oN z`yif;1C*=1+d0PFd_!`DB@M2YwrQ}$J}INk%=KH&$}RBy^_Oef3MHY9$x&dyV&{O; zIElx2ma3|h>zDagkX={eIoG4%A` z!^ltuu3~Kuc^GJ6Yd54tnO%8GKH3wJLtZGuB;Wvj{maeLt68__9#K!_0p#7;%uILW14}a*bjPF&Px*UgOT7GZ^n#4G3f8 zv2HO@7p%@5SDt=G!1VP75)8KkLW}L@k-w?5eD=L?)vZwI`B=>bB@4MrPvouON;m+5 zd8`49_kG(L-yFsHq*Mwc{`9^F|6{e_(VY?r22rINGi?_Jw{|U6R8+Du@4e}giFxR` zbHHyW|Dw$Mm@u@(0~gNd*#lXrqsCaLtRQl2v*J72r51@`_N^x!E5Nsn3`sa~w#R%{#N8v$A8 zq>GZ>c2g0I(8Ojc1puvjSi|1&Z9E_k)QHtaWxg4p7rrFbk;6-{Wvyg=T{Zw2#3zbo zB?u^RVxX2MqXVDPJ;?WXO~hBGxS)>X_6#9Or#H+6AJHR|vF~+q3skJlCV<#^EvSV} zpjr>;vNwdpP8zKuN<+fS;&~8pu;uN{+;(`teZWz2@q86`R1+hzXdIUe8}hVeaC5~e zXmH;euhMBXpAZFcyuAr9o%cbHoXEaGZ4!-ibNO4#zO zV0bhn#fBqdGl^00^_ykObePdqnEzn#&J|22!C|r;(dh1*$ z<)=1GGja`ykGDA-;JgmzEB}^MIec1QUvHcN1va}DY!Lq=3-SIU&!`mW@W&jAQ)pCZ zt%6R^KonVEftR0;S5{;L9D;DCUjigPjF}vyD%xdvA^req zgA_tsT6Dk*u)t<>9Aco4 zn>jeRN(RogpxvDu`~g< z_5`Icy=xcvMm^kCrI+RUa8D$Zz<%TrJCS%El=z^AYYQ|-!4fc|b~QtJF#_TEF>n0x zG^w0~vodWip`s?peOPA11Az=RD2=c{@(6E{ADUFP0!@}Iv_V84${7pmbFXC3=Bhqr#Og$Q6?LbEuV)nq7&Cf+Q`7A*8Zd zpv=3y>fZp%2_QMM<1rqLjRj7H;{>Nod-0=^@zy^>VHR~z3|oAZ@4q^I^q!juWW&Z{ zc~lpnc`DH>je^K>GiHKVK(I*_Z$Cg3cHF_B-uKo@Ax84$_xf$vrK-e3_QApK1W3^{ zKiP1NLl(aU{;iM=ZQloLeV`W%5#*G$LtWIW1YDlVh{lJQMmr0GhU~uW5M`w{gMTtU z%;WDL-+gqElVdDfOaKy#AdG!ZyXVgEsit=~iy}Ejp*+Jv1%YAe3&qbk=fT(?4qQz^ z@X?euCIwF-J4`*w;OLKTCe=(uD|U%@nYMx{`Mv_hZiB?vQp`}0m1{6>)7XvKffh{e zerh`nFrvuePMZDMHL!f?eVTR|t3T2pC5|iHSDCS(C%@EiC!V)UTw3 ze&hH0OYb78xge~aq>U3IPwpN~F)U~aQ;WHT@fyEmdg9IKF=G z40CqDZcg1O0OV|6a0MTRq}CfaeVKJT)Srwfq|*akK7jDs=QLr)Fs}m<&{%{m4b0ga zd3VG_x`zt4#TQCpwCFDgC7(#AZi2_NUz(p6fLPZ^q5$Kq#Rn`CKNXamJ<&s3p}LoB zGKJj*>u%2z9LD& zWMf89OmZUt%P=zOVgV?imYk8x%#ZOYDJeqSY7|f{Xv(#1u(0u?8WH=DwHcIpD6b^X zLcti6x@?Z@U}DbKf?*26cB+QGG0Z$aKzE5@R*?0Sf0t^#=^yN^RqTAxwt(hxNOO6> ze;_dRb!jel6QYxAv-0I~87kacS?QufWjQ+&NH0l+Z5BE^P~m6jon*T6PB;LSIKI$q z#S1}2^jmh+MJyRf$mnhBvtvG^rQr4}cxTpr?RRySMQ4TscG0pWEK2HA zmt0DJyNVLU+Ul4DG%Kk9d(U)~F<8r#V%e-Ff&ewVzl$Qq5n1-kuQrNHB-AB6Cko<#fOxRn zs&Q54z1QxU(EIcIpg@@WHj1s&XKU$_7PbSBa#fpX((R!a?8LDqQrAnQpC@P-T_2~S z!*BORzUx2n#qUl#C?in-BTu6qme7?8-Lbv>9R)o;GJcyI`vY$?3o83xz>niUfFBbZ z=l@T6b=GNA-Qd5qpQM+791PLHuWL4j9+kg*$B%UYwJoeYEg(n=|4vr6BqQ~p( zH=PzfR~L_mbM<0Qj7{}?|M)rJk`Uero3e7rFl9pm+sVD12Z8*Qd{#b&EVVaZneC#` zBu9BW2|-5q%#LjiKXL@$1vLq8aV<_}X}aenPadrlX{L&;_ZqE{dbWrZQtww0%iX2_@c z-|^#gcjS;=F`mjvmw@Er`u=8=(Wn+JH`V$AvL=lPvP{xeiF-O~Irgsx4(aH$=R&dr zkK@iEtuCq+W6h|abTpIA=pFl}j8>B~v%eg67gRhWzPZndi3bH53k;mRW1^b1gDeb{ zMy&A4O@=ZXh#}i?XaaE>9rd=o^rq3JFwlA^$H67Jf4)+#|EMNlI>@?CCJ$HKr3{fp zM>@)tX^7(mGhy2(ln^HaYvMK3-qvhDIY}{1VsKJ-8Y$YYWie;8M=;*0g6}NVX}gQ3&uP3Y93QBW^Ko$OUqz%VelTtcf}6-+#kj z2eb327o=l!lm?dYmO#YnjCwRa&jU!2p z?q}~p$_Tr{BmsOz>=K7bC<=x{#vrMW2DvvJpu-)E$jlI5J~$?Eds3?DEU>fkY}sQE z>g3U$n+$hO19d0M3vOhfW*7)@vWcs~rKh-)lxrID4vg0qx@vOHYECm>;)@%WWn!jcB}zzC6k!! zuO{VmBx0JRf#FvX((1^qx4^T%;|b(m$9K64HWIzvbq0NE@FR~L_dPn?T6yA=Ta?3l za^Qdq9v&Md6>YITIl6LD^+sOP`FzWxwM&T*Ucge z)fC9m;qf|E_eTJzNx!hjD5LbJnO>@+_eWE%PsOXcIT!o};O%2Z{r8(Got39#xvU&G z!Q?uENLrReY?}v?d_0A4HF`!%*YV(KPW}ViM3C_dNqfw$tBOfMQtk%a&@l7z>a9G= z8OA6*gQ;c1UE26hSKXChQ+fNSR*e)&?NOQ11bFoalAkEr;=V&e&qGIODkAeo@q+y~ z=GGkSD04b8d$3=5f|8q$UsQj%C&7vubO-T0)HU`s!;g$AQSgBWyS!6epwhUQ?&1ROn+V8oo(WidK=J43Lh;$dl_yk2)Hyadh4D3duZkF7rGtL zd%q8^trLj2t8oLEXba;BX_0`zOmebLY;+m_y+=7EWSS_gdjC<+cC`;Tu;)V^m z$a0?xWH0S;Pk%tNn;z{o%gllZaRFyzoPkv0lds4saVoBOU$`^#G%4xs)>2mz3vrWD zxVafD>)88q>5Q=@wXLmK5N((TdS$XSSdxKFjs@I8scb8S7M_w0T}1=kd%mHVk$NdP zeHKOQ+q;F4Is1L3IXAz0-q~u!XJq@!OwFhR=8s7Oo(#SmUCi9P7SjDe{4b>co-Nua z4K3f2sM~gv^zE{QfpX*FGEaLnad{{)E4xj-j?wM6tjt}L#g7~xS zqT04m`)Y&>)N{fSu?#S`F|lW(6XUMYB+RI&HoFxN`}jh1a`FS^yfhFqqc(HjAJ)Lp zjx0t2Bv=GAnMRf|iRa|2LY@{!!$B;#UwZN#Ed@~#l^;t@4{jwON9L~NQMt822x&gq zt$w6lu*}-%UT!>KrTzfIN23+gZ`qtmPn>HC8# z-q!U<`1&2``vevvt`hf#?CxAQ%22{4vkFw*_HR%BV_6V;U-cBlH|lSueaM`Z?E{^i zVkNZ*;99TVWW|(yn)tuqThggE&A_?nL89}yNVah z<=B>|q1os?@CDlrhe`}jh^NP^cOZTIVQkdY9dho!u{Em#pJh!BJ38-i7=uit@A??e zT~)r?$+j6Ytsg=msJ(q`JIM`Hp5ag*b4hBD_@AiLB+s+6Ez1zdbUXj%q zXq=WGja0wQa1-3lU5>e$^0KQhn%aDzILLl?> zXew(me^lth_H~ypYkM4lEHwZ8bpR4q7xqQqoonG@hWxgAmB`z&xwO zw0kWtE(!A|RmaZgCOW`iBFJG@8JO??rBbtK>M6C)8Un9qZ_Us1;}19xA`K(^2Ah>O zG)a#4iq9?X#VIc400;BHnVb`cMwYe5WsXlaUJGF{AAUK_V_XA2m6-r# zRAJd&NGI8`A>Nk9;JlF53`@GhwqmAK>ItK%hrf6R-Xj@+v~tuN3$oesOA%`$kMD5SKTk&ZNi4aGp~%HBQlh!4vDZ}IzjJw2c zfKk4*l75w1qJ=Gm98PJ@qye0jJ5)ked_Q*cM*tE0vU7C4kL`}th2owq&{Uv$4~I)Z z?UH764frkCZ62SMo7HUvi9|G0Kwi3Z=HQrtBU4p7$;9PxYGULPCZphgEN<2^9gzd$ z1g~njVvGb+_(N2HF|&krt2E5t6+1)2P|#tcXX1v}_-G*#*Wmz{f(QdcxXVR)(Ljs6 z7_^1F07Hd>i&yVl#8{_JD(F(-QOF=g_t%IVN71zF;gAL7cF^Bq$RdfTR)VfS2dfS| zYm%e?NYOZ8UWB-~JZ?vmlp+T$#R#Rus}!S1K;}$f#Phzsn)9nxJq<-T5YVu%m?5Lz zhN~%yS;FW-(M71T;OO^-PTU#BNlqDl-Z93pc*T&{(MohY3K#ieh<6@FcqkkC#)CNL zexY&Fgu&htWtbn8@MjFO&*Z-3M8(-l)L|B)ziJ!9M;IGPUY9Nyn7EcIytZ_G0{KiU zq9)k&_+GFJj4SHz(a5UQV#dg2-H;VLUTlrKgYgt~q%p2-_lx88&7iXHu*Z4RZ5S1d zhd5kA8A=5KUw9^&n%C1^ILZ$6_Yxzv!j)k+>@2_ZfpOUnt~itfG0 z_0OQJw3*;l1(l)=2Gvxk0&Y(jgkl)-S++Ej+eSsnMtPT?O`NON)1CIvp4rH6o|0yF z&DXY6sWBW^>7?}B6tj{_jG5E}5MT5`ek; z#dp-w4QZz$=I*vbfxi@=VMjKuJ5@kVe};Nam7de(vd3t0#^Zp_G+0J;!$fI{6NLIV zpH4UIy-iIZFY(gbb`Ch04IC&&(4(r>2t~Q9+Dhocn@~?d<8?RILu^~67CM?V2_9)G zZdkPCCHO8LjCBxU$695_B}{#gQ|l+S(zJXz04#7%oqB$s&+O(U#Yw-cIIKjav~k`y zx2g}DcGu`6DbpqXs?CtZADjlc?bLb&sGs|XBF73ewmFm1k4*+YS{pOI{cZ9_tdP?n zMGbVVsB(}3?@Ivwa%k+J(`~0(O|PV-tYLBjK9UW_zK$8!qp_pbGNW*-CsyQ^W|cHo zlhZaBdgpSU&@qhiBW9=KuqYr1uT*@^4_;9x9RTT32Mx~8{#hcl@pk(NKyRO2MA)P} z#X6Qt5I|2sBS7D{&WWg_Yqj;^OT>b@ov4mYJ;sLKdTtF1YY1e@Tac-r-6>EM9xh(Ozs=Q`+ye#R-w*=!a=`i?r zfzuGLp;&K>%Ox0^jfR6s(O7^xGSh3pcHO1H>&hx#fJV_*k=0Z7yHSS)xCW?Wxp@*Yw|Kww<1&Sk|9G<{kgUeD)-&t33z10sD zU|%XETB?)OKGt^V2~8v?S1ei^N+s8X9F1(|+J=M5{fgGjglXdX7HW9AKBDlnptJYFN^ht4_bk^o^VKU1dzaS3 zu(JH;^=}6HZNI{!kYzp|pnr}OA?mA2R}A_vlRb6X;XzcTKwEuBt?+Y`wbclRg+6BK zrVBkLgysN7`uHuf5NOpbZH+P_=d!k$#2N+$OR5eC!eb)P$ynZNt@;Ntik$c|9SMJ z?=$(cFZaSiXJf#(!9zijQ1v&Q>z6Bt!Zwzr*Bg;CQ^-Szl%IBw0(#EC7Wco! zRX{nx%F%k~%vKh#Z*=YQi|?%#8_2$S=XG{2@A%7}5JZP4m}NdZGoYSAr`&(!UN3&} z{yyee#;U>bc0ECHZCkjrvkPPW2zFnwc28Gf37d5D7qe>@h2=CnCav{HmuC2 zJ4(K-bHL=yBY<%b8%HiryfKdk{NaRY3wZu!K4kcizr17WGN%}h(Z|I0_)Y%7CTT5; z*X)|p=1#FEZDVX%NV{Nh^@hS-EJgK(fB$JbjZUxfV6ne@H{DsPIk}w?K;NP^OzWZT5Y-NGBWCESjxyPFj+!64y_|>;c1t=w zO|;8?`s)7F@{ED7+LH5$dMPoVoEeY3^Jr|LWBbubX67rFGCC#B2>}=!biKFlWm{^@o$eohGpN8BO71xIS3!EBz}vlj?Lj} zC?+05zYJ_?h!e2-#QszWvOwW{SM$d7qU8#}5<3 z)9y$|ncZ~lnzHryQO=&B5wG|CJ{cFBJP-X_Wvu)s@jy@7K@vo9r;c;p{DnZt4IjpZ zqmgcp@(kJjXTfmAN0%Xq{Xq3re()ut(U=$gBD4Qv>-ZSRqMp3R^-S4`rZHW~TdTc{ zL(mJOXa_tdn##w%R2oM|cQxYrNa9PFdp)&!@WTZ^SyYnmSudVoj}+-UGTn-3R86NY zzt6)F*!%PeRR-%MIVygbi9u*fhWc<4)pU~uXG2vbxeX+&bo1?V?0B*Dtc>}r)7X1; zxXIG!QN5In7#y7k69Nd3w^Os>1t5Up$tW9fQP=4Dpp>vKDG0mC`5ljaF<76_L%fU6p`7f!L_H9Pj$L$AZMAuocjy;C< zC&7h>f1>cR{aJCap(+AQ3n&;q$mU-?J7()_C@>BMrGBf?9h1rSzGlo>+XhsbPKAy9 zH9NQ8H6w0GD=zrzJ)f6HI=jajTpH3!HC^smJrmi)REIzo)hl{-QSB2qr?5xNnX>hPM52!jiPeIsvfFGm#+iX z&az2*QJ25;9M>16b1MD^68}Ky;TqRjS0?w_RtLd==E_YTx)Bl@Y#IU>E76+a7o?h- zudf4Ij>5Fxw;}#jm27Zhmp0~^@SYyPRZm`wV4*?xvF%@G3E0_$ez}q?Jk4*5+!h)Q z4)8ybe|oYHnYzJr0{g5b0}rt^=pf}u8=x_7R$22v8C=6Jn3 zG|)Vni}teu<-KU7J49epLZpjm%`&etN{T?S$wGLN?zn4`yj%hGDmuihptu1|Z-a9S ze&5Kn3m=uvfN?w9Xg6v-l!0K*X|1Wz_ZHfu!E(X2LrNg>lXF<%N5N`d-)IN{5VmHc z!@dE&RDX`L4-qiJ9^QVdj^g{_h&7Jc=%s@eOa4J!$SuI;t{ z@*t~N^GBa>#}Tm&zTIG3A20#@?#A;_iX^;_q7P~i z$u5R7&!qxa<{U&vfvV4K0XXAoOs0DKi0TDdiIWNHcZCBFherCGrKW*b0`qrQRI2hm zaS(M*rLZqeBR<78HvCIYs>aVt|MRZ?c;7qb1`rZtV_{<);HeUTt->pj&dY$9m6DT_ zzYPh8W1Y((LC^_wwH@aTv$YXTuIpZb{G4}W zXBcgwVO})=?h~R2>;xBRDTs5%R9xa&B|Z^?GIl0GBZcy4)9Z!80J?s>efRqKxh?q#X`BIVmis7MU5nbb!w&PijmRd<2`fe3CwmyMUhzUfRE zd=S#5G~J*S^vm{aB)1)OW86nX8j`(*c%4`w%C%eY2!BM;mhfyO%3FS6H_%r|k zh}-3?0TSJX8F*oAwsN{ig#6JA=Zov1wR;JJmVj)=E=+d?`a#I*CvXW#2O8iveB>_L)6IF?dzS1nD~#R`4Gd2D^tq zV<4<6V6RQgdT== ziY*2PH!fWlx}8$W_$xFnUSiM+>lP-qRs&mx@bcjX^8}!II3(khS!R(ImUZoMuiPy| zLA?<7N9YZ>mw@J1mJo+1F<~&+2Lyf3aH+K}S$=SZV!}iG#-(R3+NfF)L~vq_w1M`` zQ;8L{=&8Oi@yf&lRVa=;6ru5{8xsIwR3wrw)hAosLz^bQwvk08ZHzPH`3SnG`HX?- zsCW&8SVutY6hYUZcGTsW(k@gWMv#{)v5QzNj;!nnkI~b|5$8(o&sZpq0?~F~Fc9Tb zI9~8RirFNeh_VEps5QW4%jI9KeiEQM2tX(r3`PPTtH%=f7wQSALE}i>ouh!WqOH2d zu5L(7ON@EtG@nN36Pbn#P{s!aZ3LfO(#a*V&`LzJW~hDxMZV~rxPAOy;TO~f7pQfP zns|~GOG%Vlxs@fv zJ5dST6y-ML5A`IT=Xs0GS#x}Uzgy_eEC3dqD}tveSAl|le`W5tjd?h5ZU`G|Gl*>| zd=?OcP%_rTxR$dW*(8X-Ly3(L@(&KmrX z&L&EMa3z@L+7c#2W#`DwU;e|4TA>32=p|A4xZD#PE#}Zo_NU{kJW%#iz*Xm%*kr8} zMC>M|sCgwsdrC5`R1d+Ud7X0>3a6WkuV&v0k$eO-B%g*cIHs+fa;R8LK`NS%GiCD12-i$W|NPn}-v3g_Hhq*1jK)3k zOE%^q3@e(k%L@IFk&vi3A2!=*d+~a{n%iLZHl%(VWuZ*6_!K>}QL1`+qDYhhfGn%o z0Reh?r_)$-aB_#XH_m$!i8Ao7cdru%RT4T>W)McD-@nB=f08gJSamvBaFeZ$>?EF0 z>;(Z!*b~Z?8CoNnyo`!G7F^aexFBi)wM>Ic$OxC7Yv|NR|4>Lx0+N(kEcY*0 zox?D7=nt(c2oLD0K@%T3a*9?X)N|$7xU3Hd+*sn%I^IV}y+T=5q7q>-ZOB5vja42A zPjP-D5HeH?Rr%Cgpgt}`z=pBQjou+$ab@E3;;48Fm%WB31ab2~Bh`{$b=m@nl z=X8#gLyC$}C?ZGB$H-ppL@01|t~dfx+x@X~tB7=XzdU%l==~x8?Lf~3l85wibcMIV z3QH5B#c7h2iv4_IS3r_*$|B7OGfVnp;z@_-e;;z8mPf?&ta9CKi9jtSGQM~3N`z(2 zX6gWza0>1a7nc*#LtPW*M%e8jQV(I}VmZAPEL7+1jR^+M^7(!TrBW5q+~C1GG!)W6 zTlEMj1!q&qz!&BjqGz5DS4mND{0fK>r3HyQyt(FBgf~M7l*odTL16iz!Gka$h0Fpm ztq-_zy;?zh==ZQESoWtN_v@7@=7tfM9t9?wP_f-`eLQ%7+}zwWUZ-KF@bg1TrMY~5 zlvS>EoqX}A@2<#^2l=E;rdVU7BP(!jrpUtzgxkMK&+=P1;>Zy!!fl0=CrZQNo15qO z$^&`U1>38dETD<{9s$)*>DYbaw~MH9gquGp1B~T_1Ibd(i2}9$8jrc7p=22;_g|)Y zltfdDkoA z@S3$5aheXa#1P$`rAzBgMT0rSBUF8KagX>^18>pg(-$z}C;o;SQm@5qs`-(VO`4MT zHY|TZ{k{lt@NEyGXN5Y{IiokvlxyD#es~of|4kiM4>lg*FPaWae~F$SGf@(mJRxi? z{nmVZ;`?Ky;WCXF5 zmnK>qUP1MAB-D%nqumq_0-}*vOfvUf%FdxjzXW86=H@PKf4?&@fFTd!5q1D03GrOq zbYnj%$b<(4I2aHl{6kd#G*}?|;>W`XLYU};JQ97k5%3M06xsY`GFZ`Nx+*Kc1^gI?7 z#Btdd?JRcD9$8NWXYC!6pvZ3owgtn8{HdB#T~t#QIgUX3`5@5RAO;M9Q2(>o)>k#+ zg;J`umvblEzpRD^NQjsk>@oH5J$0*?9Dlo}&Y-<|uWWux#KJ6-$ZhW__z1QcD`LYJ zk90o$?<7UFJ_T&Uk>2|x^7>U3O&Flj{tgxZy1)DsH=aYbCm5cG-IHJ8z(0#%K3pW~ z^4JFWR18+bcl~mxWKoi)5{}+XL=QFFP%uNG(WN^W&D6CWzMgL1si3QUo4s@pj?H{7 zxU~IS2vzs|@=SK|0T4&eywTtyuA;z&3coKgZWYQW0KVwF=WxPztNTCznyux~-!RjY zBA=0gAM|kJwZ_6DVQ)e(3;p<$F;iX;oadK!a)!j(k--&#qS#^w6-#(GYChk-Qh-Hs z^5cW_9d-G)e6?^&)B(x#;0$zE*#HNp8{%B`^lZKwh;IZP4!ERW4Up%R3g;^VDhySA zQ3@F(yeT2411c=v*@?l$Gv1Tw)qNR$%Ezx7_z1mgw1QvvH)5JijQ^mor-VJ-vpN$k6=DQ4(p7G2>zH`NfE#$xm8f6Yk(WAjs4`rsi*;7 zMXZlg>RG~Mh*asLsrFSCXJwCxjrSC{y(Xei(YbS~f&DuKA|Il@2>fg8`OmAFP^$g! zy}Djr01I{Up=T~qfEjeID+<7U$<1!wKK?0v5$94#FIPlr|9dTAn-HI4WzJxlSY^PT zZNR=B`t|fpB!!KjwncyIK5310fGRW$am@8Z<*6*(JYbh(&$kK4>P(IdW7yAPe=mSxwL$8CfFX8<{{TbmY;6BK82X>4>%Wb)p358FYr#@ThJirE z+d$ifFAT5hmeGZtftxjd2@NB7{C^8;!auD(nVeIGe`>wf=o>2l;(|%Mu5W?>)$;K) z)AfDRH9x-V#)b_Ae@FdBAA=6=F7{r5`+M<5{6-!JOVPpQcc?uyPKuLX?hY0O--K{J zm{NELOQ+txukIz=4{%2=x3zGQH=~ZATbGza~)F5#(dI$wc`(L*G8S#N9EV zS{8UdM&jm!y{`e3@`+w^?-bn3ZBnDZO1)&|rz4%}O?{tmDHKl<$tnCg`0-M0A#o66 z)PQN}YGGmE(Q))t#MV(eAZ??_J{$DmNTCyjiIg-ZI-%gBiq*ytV%gYW) zS}w2LQ9!7%@(ejs4<3m9kEkamH}~k#m($HjQ>j5}h-T+?&F-3-OW0cTu8&f;NqUz@ zi$SI5G|~Mlopf4pjc3P}$X$D(MPLsXm8XO{WNRcS<}TzPsW;9aq$fez0s)nR`X#b z_9#et15vF!$#fl82i?RG1@wL67Y?}12@Pny-|;KUb(z)GN7Wop<5Pv2VbyJ(n-1RU z%ys8^IQoWE7v@n$sjH?Zra=)bc{Sr}ai~ZNYmPKxNLG{`!$~^!xwJYnib!7t#VY7d zm`EYJCXbz9mzZIvhBOwJWGs^xU(U=Hto#<1z$aCbziTW0^YM?M^v01E*@e{+SI(UH z{aEy+QBY)Kr-wLlqLI~&{BWNrb)^gA*M&Zx9wH?Xyi3!k7T=&qRFYcvBvQlNT)GGD zb}QH`0G6*QWWD8eU54988et6D9nF6L*;=(y=&oOlL2PJ>lGe{L$?D_KShC30s~isQ zOBak!Gf$j+I&z?y`@3e|F}p8RhqqY;IOaO*lf>Sw0Q=BU@ zv7d@TwZt5((Zsa><|~DUsR2&BgJS#wBt!vicimF*NdO*hnFR_;Ao*f}D-{Mb!PN?x zdouETl@jiA0Ln5j!Fu-W$do)st`1QQSy5XzDe7VG2gQemyjlXz3MkKI&IHC1gsH;c z0FYZib+!gwh5oGqfdn9nM@$tG;8O2G>W_u}&z?gu z>DT6kBoM&XTC5E$A3t{l(+zxW`^)PFDSQSgwAiuFQWV=Uu5akGI+=N9*fiOV#yx-! z4qqtasw>ad2yg;|X-{;UaeQ2r17F}VzVw0zZ{$c{*WYxyIw?DdRfeGkhzQ^cX`T{LdgfX4u4OeFMi9j7iWBRB;}{-xM;o?+*m|&(1x{ zLDVMwd~|+)+1Qwv1p*9KkiY#V?-8X-)mJ)sUAm)n6X62J$#Ff7wVlg{Br5{cF8tJZ zYP3DrQb_vPRcf6GM-_vNjES9a5_CyCTmVjBY$yV752=j-Z`}C%b=Y9JP&x4fJB1>} z__>Lj0w#XRb>P^NUR^5Qm=Ufb*6{}tjmKg#d@E#hVRgIq(A-V1*S1{HNu)mHzTT^e z#lP33c9&owAN)MH!Z{~)tG|+d$#sRPZ zqZm#{0IFpVos$<_&yOi{bx|@Ii!g?jSNbHe5r?bh-TTIE0>}zY>*^j)_b!-w$lYK+ z=fINQi)?f;P!7b7Vy`cd9fmIG>yD0oC4`D1fvIcD5L~jK^JCOv2ZU$-&aN-%-g7Y+ zpFvahs(H*;vu50CA2Fx_x>y`X`7RGfWTWq4c2xN#W!?j>Q^vdjxnkB85i~H_m>soq zwCUjYs3IY z_a;wmKL=uea^MZE`6<=_1?W|2^2ZplkU?6{y!m$bdoZ5x5iHZZt}uSjDExIpM372^ zt3!WT}%D3FcHmCZr2D;7vVWe_7bui zLT?c(AV9evU;Q6wg=Yk|eqp+r(zt9~x48n9g001{5l5p-)qA!f@2oOa_0QjH_+$b{ zr;{w;l|jrnBppCeWvVPOFzk|vFPX^c@7AMLZZuO-G=ZoBszAxfzt;QL7VuBI4amFX)K|U{5t9UfLZi)H zAuHni@i~$J%^FXc)>q%tX>y2h@OFKe9QMEwUV2dWHzft_@9ZH(085sBBP4;_az~3eGz%gzTmt7$6R9(}3=3F96$MA| z?U&3C5Ra7uT-DC6JApn>)z$LJQsa$4RQzmxO{pvpD#)h6+IURG_Ai}p8VxdK54T6c zK~&&#`ESR%D2KF@QjSVv)a)ZL#-9UH<~(+|=x2f$#0pXq_Nb|~Q&-k3=Nh%pfK78d zWTl`n=(?!fI`#pXh_&!bX)qA=GOxWG7Xk!X^c)J*zXQSnZP^nXD$CKbm-ICi#IZ}a1vrKWYf$z1MBIfIhdryioDBWfX+nf zP(Z(MXOG5|^&nqCJ1mmSCLGdJeDzZUGhtk@a?VT6JRwtI1Cw@~BeKkE2| z8p!$|SaoNVEZGoZHNporoH-QDlGX?+hva6Ne49bbpN zm$id9D-Mcv;)ar+q7ac>Z_Lqt8)03jzJLWdW4tha7Wm=I75wTqoX%usbv=N`GwO z_he`@;Pn(9na=$#&a$;eUw=*J-W>g*RJ8)tnR#iJJOgMflC)hY6eR<}{` zKWM_`s^9gfihOK~`BY3GPtb&wmy0vIJ`i(EOyDLj;B&sQBG5P@6+gx3TB5rk)n)=@ zwWnQ4Gc8#hhj(O1iPibqD3}~?&V8(oDqtv6Rgxur{-}!8t1~guPPB|!G~b=MA+MhI zacZ_o?FJQy=<{iit22;sDu7T6)tIyjA-5+^t7-#F#MGhRlOEaEn-Xgh=rlykAwshl zCd9DX2Y`XJQO^b6-0VaBMQx}57{}7(!tu^LrGX5m>9?cU7&_(lHipD6 zf3alKaDr&`p7_`Nvkv6w$dGjop!6)>euRWWvvscN*@x?gUpCN6J;k^8c)%`zTTZ#+ zB%^psu8dd$sYP7uUkDwtPKA^EcGzJIo#RoaveoJ-Kp=`ln5rY)j<2G0vh(RRD1;h* zkD*et!6lDTMJz%^9xd`pK<-$|hIn>8&;YFIOe27ZazyUfa`3L|SIfmx%HhEg6Wpxs_k%~c^rdLUGRk?HtNoN1%X3QFFI z{9U4;3dm8e)pNnj`h)1|?S!v-P&2(WX zvOx8QCZg{~Nu1T~q>Zi#6t9FfK_Q75@)(qZt*7ISN*N7JL#dH!% zQorJ${nRG6Lw!^}VGqopH23dOoV0U#k4cEQ3m~&zIh3q4`qBbTJ<01fsIu>=sOY5` z=SX<0AV@>AaAW_4@FRxS>*oR^%e^{6w8tT59}%Zha;ctQA`t1vUH0$YIa1*TxZ6kA zY~VFO8nMDf)_?9G)AKb%S10khePC0Ue+k6p*6oNhxrJbFhIUdAa2wmiK=g$-4GNHx zx7CMT+09bI4!>YEnpFHZzZ+)#v2^mJHBxI(F;|TkG0iUaw)jT$2stL6z26ICjj_u~ zn)2olIs8*aoz;ht>K!U=BI9`Ze3Ly*GFkU;gg!yw8Z4W?ty%3%Y-fr71^qct;#l)m zc`RS$s0+;>;fc$V4B1{K#qqVc>dPD31-pD(9r93>3#B+r%9AS^+%S@~fPitoL-g%@Or zFJCGls4+ZRsVHN{Fy3ZFLtTQ|ffTab<2Oj6Z@_T?P~XR%UsslL6g(ux2ZR*-4BBVE zZCf7oFE;7tnAkJ~en>)t<#(j#Khc10ZP8>ip+Jm!RVXQ>8;Xg=orPDedQSs;_B+%q z(i%nL>c~)VI;y03^FPx|U@AW0>Tq=7@9sRubMI^(yKyr2FFV{ATRG6z)!X#ZsH`NBULn1pmmdXS`OI+u-xK zwj$xz+XQFPM?KF5yrHabNB(B7wWbX z+BW5#rGvcS;lYy8k_xbK4`sjCc&GYTlT}mzNfe!Cf7o#h0{||lFOMfvHQx?7WRD!A zeg}5fKPY(haK0T#-ftYe9^>BH=^OozSb!r|C|INAegz(Vm6l5ulvbjaDoxEB4 zkj(I9b#rj1>wE+My(;^gBB^+Uc3pEKez4S=*Eo;uVH7Rf>~~mLMm|ASV=tv6|Kt07 zqUXj|zfgA9&{l9GW7I@-bgx2gL6yBvPJf6@W8nN-__p~!_0xTpiv^Ge~N~M32_iW~GTh$UH4v!y&}nekk}ZhQEC2SWoFC?Hm?li!2Uu=v?TJ zEmY)-TY1##;rK{8e*`TmhR?}p*pVC8MNl=OCWjg>sH&zd|6NGXZ+bE%OIDD0-jtD# zR{N!)aMEQ_v|CVBzcDGgmkd>H*tSOnvdLzpcDf#7FU;!CDI`^@wpSlYudO96m&z^< zui77)AemJURaR8nUXJwMy8ks{hNq5ickkAutzes(j#!1X3mZ_A07sF~806^vSJYcR zTPfI;O$FhRBUE zc5pnoO&qsXGg-r=EJl-E4izg`DzG&oCT^MZ1@V_Q6?QX&ogG0{vaW0ifMu6$hZ5$d zrehFMW{1{WVWAnS-9_`|on^P$%UU&tQnazUAkHv^;l$7fEtTqumE!iyBKJrcr^$#w zt8kLuu62{;&BTa6tU}@$y_o|5Eh(;!;*AqGv9`=7?@m6tCHnPwpqZRXq*?ufKktY1w_5uOhTyYN}OhYX2wvWuFlQL;>}H% zXOy%~Z7QF4WqDH{d(2Zds}rGnym#L@QidgOWcWm>bmB_hQM!9II=a20@)&V-A6(su z?0sN$P&vuY_9Vw}+s7rZ|GmfJZk8GwWO@P_RCg-q>lYsJqOT4O_^9q}*PIVs>?tF% z<#5ZC8oudUy_-sH&0LNVls?Kh?F*m8OlaGrS?k}2SBi{Uanr~PiU2I1^Hb%Q5amBL zUkI^J0OO`Mwbc>)n zN34Ekz(*wTs}eb>smt0A+6c3aqmdPie?B#20YDv!gY>OSw8BcDqDSvSp{M@brKQUp zTvr@Sc!i3s-3~MVg*P5p*PjRheqbw9V$bz>LyH+;d`rS|%ZbITOi&47&kpn!o2HXe z8l8;pjB7MR5lCo>D?%kUYp~fi;WV<4byi$Qd7@rDR>o!~1=zg^-#ct(tQeR%#XRt2 zgR0+jux#n{H^!XNy#?xT9qp}M%X7z1qj$8jNe;Otp|}{m693EaLx?`fJ06gLx?7#b ziBitgtn73g@IANVp64Wh%0}NnkLCm%f;Dh1l{)Ln;*!Z0g$5O?t6w458AG+$k%~ht zjrmK7b?art{vI8nTT6WK{-zx}uG3x5i3ZG*ZgmA6QYWyA`~dtJBq8H$5CeryaDenE zLm!DmWyLj;E+!H6V443_R~Q7=;Ko`A>S^b@x0`|tg%vXCCoVrBIjIHd&b)I&QuKv=CQyTmksgp> zE3#UOnkNv&079h=Z}*lFd*C#HocYZ+Y265CS)_~FC3XHWUJR9NA=q^Z<9U$(QOj5o6JUk8DVh-5uZk#Znff`7!vST=@ z=E#GJw?VDb5lHqfI)nv#By=&kyoO^-k!ecdZup;8?MS~c9p#Im6XKfLXfx{cUpf>L zi8&PzBpI*S3#MfGIX)w75ddp5Uc4>!CUo0z%zw(567ZlpSds!W)WoLe46t`t$eT-NEe2$tX!dV3eG+ zb`QEdywmNfu6wQiW0b@@g+}OPg$%c|le~MHuZFq9>ri=Zng7GsIW-3su4y#3osMnW zww-ir+qP}n=-9Sx+qN;8xt&ur^AC2_zFXh(uC>AQXwGqe)==zf9^SPE(|J zXtQt+jvD@a0s0KN_@M(0AAZaHjuo@|x7fPHogi$4Fxn;Rp(1b{nDA3ITPM@iMt9Ma z`ap`zMJ()g%!B2>O4$-ymsnCh3)tu_PY~>?bH%801dlwjmVn%0GHf#zD}LzBM5+H0 zA^qcA=Q|`wxl}b;G$Up*VbW0DY|Xx4G&{<<m9YelQdsDek1VnN(Pa`5`mH z&q+HCus)XG5x5ASa=L6Rz2ke0io`#6Z~kE@f}0-r6TV=HNr*pqo%6KZwg{UjAo={v zL-Q~bb!t}#1vV+*D`7CxiV3c@G*eO7vXajE==bZGTt^n?YW>dRYr>D_aw zC99i~oAZz&>g$L^WXi(V6XKB!xkGn?ncXPaI^rT>SX{JKFG`BBo*8H|g4{7Gd;m1HK%2}2wQWLO@psDm&ETBGE7;v%h2t*^;5^fO`? zKLB}HAL@DF{=p~!503tgXI*Iz(4xpz%&pQ@e@LZ2F=VA!!e$Umg0hkXen+Xzm!aiq zs~|Snf8kcUHw#E4wrTuc#_B@OB)}6$cgoOR?(a;$1Xs5xG2LwsLZwN(7>ZlL4k4e zDkx-0T|seVFNN8Jt>(L2?xNC7?GV|4k?ljGRl4jqZBhOM@FY$PEALq0&`DvMX=#KG zuAMn(#fh?4&CYEXH$_qJtM>GE&CNSo7Y2G#P&G&+b9L^HdfS}4OBOr{fB6*NLKl4I9_=jWn;qw>^sn)J zSCYpQmloC|9Wi2oC_R65PI+8<5(RkuX+bo>wn+j0Z4-AI!U7%;6ka_=SYf<_0ANA2 zOJ*bhxE$o`SREzP1X%Gwzk&nbD3!a3Eyt%ep>hirwOU2o%7)1>)?J6Y_ z)>@0hK4o|!!@hHu`eO7=KT&w!aLxY6MZNjBF}7ADytfiQ8l-PaYI+PqsRWG~OTtV3 zG^CfO(Ul4Cj3ofr`-KjJcSpDu;KXmoJwm&*AEEX!nG*1xG#JAJfVpQ`d3aYE_pwWt zGKO%wTzDSjIyboWPD6aQzNovSI_5~erwMk0I3+>FPNzCs@YF6^qn?{{r}4-@#nu22 z?c=TK73K(5ch-Q&$@4Kkk83z?r3<#plq!AhG4PO z8y@03fykpuiYSZVr~+N)QyYei5EKW?R3G^)2r@_Jqq)k(8ultA>nlGplhEIJ(f`Wh zB$xx;vxUo-w~CQ?m1NV1h1m3~Z`$sn-r}M{DunV@$LiL{H-cZDO)gXke7Rsem-96) z|8vaDPc$FNIuba-#w()s#QyJFGWSi*5j(Uw`qEy5J|_}TvPdSk@7Fz&q=F{jU=poF z+-x5QJ6j5P95r1((d_PpjONdn1jCM>#Rq&uwX(++oVT{>9eCPVJ>`amF4f2wsuD{@NDfv{00L8RT&>-hxw!;Bowo~EI5^kgOr1O6&kvk%UYSo zK*&eYJCyW@=H@Re(9*W79m^L=vZwc5?R^)3jiGQ?wfmRy#rP-t^Ta8ZIf$92LILR+ zWa-qR2_TD!!|#ALN9k>`HPkYAL+Ih3%AQ{K0lnCbdR@uKyd6jCsqa%TXlOjA;27=&h@MooC*Vg5F}SNLd^ zd#){+Q)crF)GnPs$7}~AhNn(i<6hWo%G(=(-bIy%xM*ly$G z_6d%=Ouv$R=zrKuV7|LJ$WaeS6nx1~X4pN~^g~_Fx#0fi^pJMjhQDnaN0m98Aj#pt=H$)coy2b}=8%73%7b8ux26`Ry46cAwQS z7)IVfs!Jg{O!QAV)lXEI5fu>$N314}XT#~ebmZ5ve-yrAJ}>@Qs9Eq4sg-B8+r*Em z-XEiEP!UG@zTJzxWy)SMBwRc4U3(Zuh{2^6>_Crvli!s>$wA#&ZT~(knTKj}V(qps zhEKZrnO2_q=Y_QC%xHry8fX47?VM^R^Y#p(7aI1YEH%?aVEb%770;=EDR~9m)xMt~ zH_Cru-8hg_;4igtd$XaOER-uCPT_t9WQjmojN73MQ@`d0Yj9ng;pX70W^DCeyp4MjG69Q2mHt5l_jt-yLNI zL_A#)Mew_OKBco!7}iV0bZ&^%W{DWH1=$(xP2BQBiq|TXAOdTWV}nnfeAw-mYIE3z zO9z;nAHT>x`r4Ra$t!9CT|Kyg7!8?IPjl&#uX7+GuVU1F7jD<5~;&fmsm! zi~$@3hv4tiP}6}g20?HZxp7z(E%7=TTL>U$mudW_!q08FYl&)mHCCC{e`Ad??FWx% zllgk1@1|D}?^3`})9P)$tKvg%T%)dkgN(AP|BG*DlWN&6;>k}m=DfJF> zqr*+9iB^;TuoKOzn=D@1?OD_d3N&`+W~|31<;*XCg89@(B6?QNrE!?bz}QxP~@hun^7T+eeEQanazzIbSxXx=%xRyeC<#eO6||llri!sVu(} zx4)OvJ`SaE&`owu?gu0?$xs8z_0Zpa(I?;MGXfAKhp~Ni#bLEeZ+CKIlN`xx#X2XX z37*&N;MkRK*B#vX&_C%*g@B$rxW3Kw#o_+>xQm|NG#%!wyg|6%KKbHS@$kBf%~+hZ zMV(kB!J$*3Hx9JE9DHsbFnHw8@H9U>hE!k4vCG8?>YB&@x;AZW+fC(iX?OWLE_#e<38^iyn)A@_mHpLyf9;y>@u)H0Vxfb!yp+U*}!6Wv6lffj_ zqd4x|4SN6`0<0mw2ZsDo18)N{U%E~vl*-0kUT=)hJ(K(5 z(*~apKXhzjcbC_vsYDX4rr$kBCeQcpr>}vk4b=rl*Ib-i=Z7%=WW49?rx%4#3u=Gzp`B(JgK63dv8#MKm%f?3sQp zp56-gv<{0@y6akp_pFSXB#*4}f8}ZE#bl;?kkon(KVo+`55dcSy~vv%K)c9MWz?Q! zJ^)roW&d0GzmbC46lstSg~^9Ur(6Xyb#eKAIwST6EX^AXx|=ruQ_Kp*vXD5_I>ySX zfqZEo`~390(&hIAOWnwz#AVtksUO5tcI1{wi0g;khq>=l1Y_Za2knr zY`#yfS{nj!0u$nx?uLH%1@pJd>iBhFh!S~=v?ljMOoKE6b)Dc|HPwcbh=?eAwN|Vq zvi)>|#7PDQhVbgRD!n1DV|}>ia;Nf^@#1F0;x{BQ_1-$5WtfRON_-uuq#FAf~1| z?W9PtPZt2h2R#2LFoF_4F*1MF$qP*QVE8duNy?x4{)v){h>ySyXpkGDKPd~=f7TPE z`#NoOC(&gWl3OV+Ix74SnVPAtIJE~=wJs~VwK{v=+`9p!hNsmPdTk<%b$y_`_mw+$ zo{vgCr-1v6CC;v!g<<-I`WZF+l*FW>9a*X{BZi5 zB$(^j^n^7cf|pk{KAyGTdzN^pka7G&Or9Di5OJ4Qx6Un z;K^ zHxW3m9x^j?D}ZV4o;??zc>s&CyS8C0y*Jw0jXuPWjkd26r0c(KUYu@h#@TG&&k}NT zgWhs_2F6&RY<#5_IjI%y_TV{J9ys>af*dOfaR}fDGQYRWg{&ZmT1O>spdVn{B%&2oqp9 zz)yi-qHl&@myZ#0;1u~&gvU=I@uZ5$FFS|w4KfT?0ajm+OD4H~0Gb^)U+UGHjvV~f z{ycN%mjTyN+Gax8nKVRsovQ!?Iou|u{H=F~#NloZ*0#90i(psq!n;F;nzi0iraDhW zSm*e3nSp0`?60w#`J0Nsi3-FY9s*n%Z!oqQ{Ve!S6EcCC2fo#-es8>~_h@;W?xX$s zu^To2Y}v_JgodAk$c-}Ca8TJ&Gt(^=3pUQW?`@WRIb?tKeBg$Gxvd_Gi^=49tcV|? z-8G)R@n&xFwOL!7E)MG(SdzrSuj;(gXxy`n9TWscD?Q0fX;yr?rK!htT@MlcJZZ6D z%72%`sLx6k1r96Uod#kKL7Q$mFt>wOr_ZTBz&2$&L<2CIR{?0IEBwt?dp}V^7jY~8 z)+<>%I*3U)-@-gwKW*zp891GE12+#bNvb%{EkLA776zR+)7R$N8Cl~G`%kZ>cM0z{ zg~;t%FvrN16$1WjJ`Zk;PoG`w*(PLJ7f%}GNWQr(+&sh!MK(0 ziM)~OP>s%;f2=l-leawvUfcxfxa^*Q2lPB&MnqEHf{u71Pa>-)Zv_jz{<$m8Jd5R{ zKSr92%U8f*oG|Oyi2nvjZD=%VINs`K_igGe*t!?=rn6$|Z-SMKVQeX(5Qg92Ylkhz z3x+?BGYVj~)|}ODztTi$83SC8|HQ0!qhts=$R80)6p%l{2|L6Rh$s$i9Cc;~sSAs% zSlHCsu|cS7m%QVIa>o z%`oJpq*bZ0YqC6)-w!4$%SFrxQqcsgD9bl8I;UI3L>@xo#j+u zS;x|Wy|NQe`gDMF^x%4>qE75ziSm3?x)?ZoOr*H8rO)i(39JBxVt7`KV!2UdryYd4 z20a7bB*&tz1#Bp*+$M_2U}Vt)B6^BDGk0w<;slFjNI>5s>uK45 zJ+D~}8)6+Lt?-%AjZ;%P(614tcl#72l#al(z{AGF-@ zv4m4-&`{iWmy+muD;QJ8?o_I>p@-n8k`k`T5C;ai0f9Ro53~VJ z#EGBlu!j%$w1K%HbqPG5|uY!1Bk2687?l3`6916?4W~h6YS;NefVJS5#j5 zZvqk$r}<5n+iM~Hj?*egi8EQVP=L0ILG-Ly^uKFQ=9KI#?ztmL5b*h*J7?=O-`B%{ zZVXF3zdgD32Z_ZySh}Ip`D6^BAE+fD$PWURUE)|{el>Y}1^8f~k_BS4+4A!Qe1nst zSH=L!2Mo_r$pvsJwGZ0CZ8D;jA12phq);#7Jj$>siYg`IW7*YhObDyh!!E?}1Z(TT zvgK}aXB=5pL~tm<KqKZ#6cdkX_#Mg1=HomR+)t>|HqUn2hT6RN|0A=1nyd@?B zu6{56rnSfmC!wW`LnZ0-OvI;#Z|Mw0uSR0OO9^BUuxO#L=qQd?$KA5KqbI0uG<0h!Zb(#cYP_s#d+*MS9~oWEL4_%*^df>mP0@{5f!{|SIr(zH8#$f580oVV z^=C5Z0j$S>Y;cnhZWT_4q7YLxCdhJn6Ws5mgj zJ|F1Z(k6Msbv3NRwjkScWwpN&Fjef`-2^!jv;ze4@2kjeY{lGYv7wg4P0-UzPVN7m zBUg#;zV}yjbG3i3;EAjNzq0C3Q$bjL3D7m)7UYflx)$Syb=8sgI=~{Yiq+E*seviA?0VNZIleu~Nc-*567Yu;T7o|O54jG3^p^e*1I$I5m4LkIB$P8KO z_Om9$hT73Z@h#xf-@gh&5hPtiy9M{MTh*6R#GTU&o`cdmnXDXKNJO}3md`Oi8Cj0J4l8jCV`|t; zCO3V;iKR_a!~<&qc5QmPZWC zQ1n|KggP&g4+yAi6K>MA*I|mFpQkz;b`R}$czd|3OvhcJKv=)=EAh7QJL1I$5K3-a z?w7d+l7fSAIP#KhSv^Py3eEEE7s`7=q&T*;L}qbHr8-wkf_n9Cke5BwKIEX_H8|*t z=ggCy`i?rhs2J#w5;kuVxSytV31wtGw2YVoYYDMLC&juP47^=%PdXTCJk1Be{4Ev? zR}Nd%d}LwvCQT?^=FT%pn+uFMi$E=9n}sL44&z(4u)Z)`e6NT#qnQ>N`QpvCfY z`ilQMzD}@p6tFrI+ptGvL6MsT1*E++#;hoabU~4DIXuq)6$HL_EElou5VJQ4G5YVJ z#YR9Kg@=9YqVMnFehImn$kEIUBY@Z3wnAmlC53*IZ_ka0G+4M0s}gCn)@2z6!=j{t z>tt@UpD(EpFESKjZ$&=7LGW-cch6wi9q%E%gj>j|+K>6qC>imtmSmq*Q)JfI@cfvd zV9?^b)EoWRF+_vj)$8O{l#iytoRs3myaHi?cK^PC-i(d2$Fc{3@&aY-Vncxws2+a)Hp$$mTYJQE3z>bK zq|uso)LTh1BDdD01V;uPlI+cEK-SGCP#fF< zd4N*u6J)eIg8LagI-RDslo$-gq;6fne2_isksGQ~iUAjt5f-1LKs;vyj7yg*>uzi& z1XgO_#6}9=E7yY+f^WStuUI;l`;X#3PlRd*d-*T1qk?X?K>?9@jK z@DR|Z2np|g!fvttC zXSf721E7IBA-no6|3dX#A2=?A;%6fme`a^W|MSGJqI#UFYm)NK8cD{CDPS%p>0&6(#zD| ze_?IZB6)mGP|Tj1Ns4A9AC&diik0{p`zB~f!;9|v0P0PS>a6MF`~c7+ljWy1xTmdD zU9O8Vp*;Hfb}KYAKHDBc2j{~+OW${iOZ@r_po(25UlpBT)yz>_H&vkgV!I4}qMO_R zQr=IB%#`9-Qiy`a?#LMu3UlmO1Hg1jO4jbn| zjAJc8Eeq=Xrk|toA$h_er;8hfWO-7^Y*kVr=kej$`&2yf{VCLruLf5OOTt@4i%RBv=oTGDXLO zqy2tOY(2T+qKLRRVHqs|V7P!0rbJbjOwq?9N@*mom8jbnX;b$vqIJ(o+1Hndy?s-; zvt>h%1qVK_tUID(C80 z>2@Z<6e<*hx{4(M0+NS}ws_K(T&7-o*qo%1(r^yrW7e*Y*bGSoh#_c$}U4_ z_1-YcH=|3+>XWh>^_MwZdI5vXAEpt6BV)yPqeI6`)X%9)w z7+o~?$BBOxs1pwEtHf0mqSzbxV+`ku#<*Su!`RjEDI@u)c z-q8tY3I$t5S>uh4P_kzQzXC5-CX_}Uh^ucm?`Y3x^iC#mixcqj1wvKFKtf!r@x4@i zTxYE4gfzQdcJ&jpg@gEpxTap`pXSFOCOG7@pU+%bb=y^scJTg5JUB314L7vr@Lk<} zasj%w?&xC`S`DFso@rgDv9#=d9#Hx;Q0tY;=*6g`we7v^V+YQWv~!+b(>_Z|W}Aoc z?u8q@v!u{L?<8lxHD2o%Loa=!!DjoT-M@PN5}#+9vj2GcbjIAACSHuZ>?IbL;y*AY z%sV3XJ3?dwB@F93K;TrFaT*V)7I#N=>mSYK`U~y0&QqyyH)_Efh}o<2ub%iKF)SDG z-)vYl+v67VQth5aBlau)EcSi7tyx(9UhahyzpJunBEYpX%~RrKi2R`)h8I?2BFT%*Vi z6!?4T!@!9Tj-=OL7dyhPSo40MXn9CC{XbyfH`xF5Bso}W(}aNa0)Jd)x z;Wv@@73wz_qcZeb^ertS4S@F}Cl!Oh_aiX&2kx~lvxx$Yb~=5N`rpEuw#WSIf!Pxy zF4vh+EuA=>rXM5QtW-2yNi1Fj+O`%p{g^+B^{d%+0Y$>1vNfMh`ano?I#!0 zqeucUddFUQqCn2Xr~x=}il|>RRF7+~-eU?wEPRFaE{-U1!S1ReZqTRBy2@Jv`!epZ zNz1VfAESBms~^guVAzyfIS~SMB00(Kj-90)8`Y%uHsPC`g9Nw$I@kpXP-xeo{x7?| zL1@4Zy49T>K(IaQ)njr`mt4??ABU&6m)z>n0`j#pfdZI=88ktBL&tBudUER}=vu#7 ztaMbr)edVFn4Zl5#OV#vY3?iNL~%GjUdFP40ggAs(xJWZlO{hyQ<=Huu8)QNq$F_z z1o?lKwqCS27Z-X$Q`WYc`=@qp+YC%tlhQ+OP_dome>B$AyY#OlL`kV*-c9S99kzg` z1PRZ)4msh_xPs1zrA9NhMfNnpTQ;s!y_Oya^coh=N29Y}b^zG20xv&e^fZg0nc-hc z@V>GT;39fZTzQ@S(v+%vCt5_9aLP#Om;6TgoAGOrQjI~Wa}N~b8l@MjkGjC*mQl3p z$h%?ewOPm#Lq&wlvi-6PwWjAjM+A=7Lt`O_I|Oo(;rT*U@Mc}Qq1Tq%hi@@^7vKZg zVT9&bCyNd3!}jd12l|B#cSpTh5`VRrS;lp=__!d6{a5+Uvsz{ow5;0|_=9?t18S|& zOmO1y<<%3$;oo7NWZWH*>B-%+4^pHTtsw}5X^HHp$qlE z^nvDI)Wks*tez?7+5D)4jNhESSxxruqx+Ae`*4Va`0S>9s@l$p;}^e0zq2o9hIeOq z31R)mpKyXijPMd~NLez0iw&McIub>CYyiLAaO?h*=Qj0vp9Kx)fySChx~_4wi+39( ziW}tnN1vG;t(Lj4llfZ zwrlxJDK{lMMiW_kX35~L!n{JnLV@N|mr#db{u=T2-V!uEkNnnxmm9!l@ zW%8x_Op9w^?jdo+pfftNQC0Tn_OOTT9^%${wW*8k|e$ zr{4;=lcuL^4LQVa3zYKlb#*KqLbo|P*yvO9xdU-Qqxc)_4PLk4U9$l%3tHucyEDUx~EJS&l~kK#B+IDu~5g*WpGxbG2lwt{o^@CE9NoJu8t zeN;f+y*AF@ew*6Mn9SlEPcO&Dd!kGO!`8_q^d0m>gv#S#%wI*Iqy=nstXmd6*&_!| zO+bONgf)=ATY!J26{@{Et#e5z?;C-*r_Hyi5xIc`e6=^X>%H|4*B@Jt16k!1zq#cQ zQS`W)O7XD?ygnboE~M0Zvknk!m*)ZC&T-*o;KNKn!fw86+VaD691Nfp?wa{d``OrM zAgO1uoWyP(o(8xFN!&p@v-USocqpC#X?(S{WCJRLR<&WFT=un@0N{ zyixP|*T~qF=E3*msr}6VzAgl*NEEy?%WBGspuM^_?d8n9s3kq8g_uICzP;_7tQPRIsaQTmN zVsv(2Mw`S@Q)z7S_+9LhEIZ!AXra6*C6j%XIexzfgRqvao&&dyKiFvmbGM5*5xVpp z^~#cs!_yfx=!gOfc<3!rV}8EIOoi5me)Rf?cNMg{0iVQ|_n;>-_?@!#Ja&P*M*hoE z{fn<@@|?XHZf>?w$@4AxRzyEaV{`Mln@AJQM^#31oLnl!P^%Q=Qt)g2fUQa0 zc*e@|0DX)-->Va><#X(D#6bo7ke$7OJSfhP5g#$X2krF$TqlX{#xYgZvDeC09D#Qx z4tE3CC6{NvjOK{mf(do(`2XdMSAa6|ew^c--}+}(ptN0S)=h_dJ$FSA z?B-3Bg}*;&pf%xhEtTe{|J0M3v@Pfm-E`%LlRu(^=rum*Pg=yfRqjY>H}N-9>^4)n zx2A3q8Ly9y&F!@qFAoZ$Cnsea+$QzqYHa}n25#jXH<0f{s>Dh7$eRSuXHGv5K2)rE zl}n(FJjI(LZ9SfjImKvkt;;+#Eo2JhqQPrWLydYb%f+|Y52sK(l9FyWS4|Bd9)Zr2fNEYM@{TwX;awn)Q3xHjQI>vkFROt^Hqw_{ zs#=;<5f-nLZtt^sNbsZqv!1O>xQHsdl5~d@g!z4^_`8+gH-fqWu*)?b| zP5XLVI*9j-MNKW?{8ZGN2kVXdE>Rn}>p&oTlJ1+2dYOEscr&Ho8J;rXP1zlx1myt8 zKaVb&I(0kD6iP6XdLZ#VA}Tne7&J$T`W8>ts5N7#y;lde=-^J%IVGIL*yc z*{epwrFwUNBbRnsqe{VJ-9~Z2&1RQUJO5OxKDL0?Z*2gLEJf3a8vZS$d|z~&vvWI( z%rGl@`EJJj*r3QMMcyBeZHV*$(0ZCTT4TcM=uW33@DA7AyV@D~n=E6dvOloKqZ78o z?+}2QO2i-t7PTs-dw%R#$*eS-G1SoRtK7g$LNKr$`{Tz85AEb=u$&S~2lVpk)`}DDj-8TUKdC(VZltt&k52PaeM$(T)rn!O}fN02ml{L&8Y3 z$Yry-R&|3W(Ba&d^q$OAUrdP_^pbKB}@9X#~ zIB2wua3`6(`;X`y(-(mBiZ|SCgwDGq7?ONxMYA$WQI1P!BVft(c8(H*UFp|FA&QPt z%69brV-QbSX^xk-9-S;%*(yaJ4}5Yp0gPPQWTCrsyVfqcV)PZwCeNcj`#bt(3*#<^ zf0_Zp;ERsz#W88=EakD6WcN@e0|CW3*>i#OKZ$sWJCVi1Sz!#YjZ_h$0L4qj=b+^l*{aC4yF2-wO z43~F{JE#r2#)ec&H$BKN|H^_{kS9sMOv_5=m zH&(IHNj?+R_Lp>0K7a8 z(q6%)l*uO^`Kj9C;+UseCPRRB-)Tdc3OBWA9x|PmTdPbNb9-fNY3xRp87<8V9@57r z0EbL~e3>*ow@S{+!h7Gz*VNg;y(WvKC_>1@gxs_SITTjbA$vcRjKi0zo8|AfK4*hE zlMK_{{>X93rnISS`k?1s5e2Jl>G)o{YeANz)q0uqF37f#M8%+SU3>OA^<7)y?~$Wa zhBP>0F56M(alE>wp!ie!X^Kl*>!+a%)dzO_Tlf3P)&^zFlTM^T#(qswS_s7>R_U_= zBKnJBX9U%ej+at$rHqP??i3%f!k8jFNFGx#@FVG+7t<0#b&F`DvC`zmd2^4#y;m8y zn2QNxqFd9GzD~G)6${n*Qun`e83BuBipw#~I=D!TI^DgsC#MvR^=#ywEtIPE zsqiyPI?czoVssEln3!R$7l_pu*L{&TC5{6T>E9=hniIDYd6t_KQgxE8o5Jq{K*{lj zk~`>|eFF3Us0J7|P9)Geoc4vQJyopJ90*}-oh7L{wIMx}2VbKfy?eWzXaAu-7(D8v zWP70B|=M1iT#QELP#`{{grQe{)dMZ-q)AYYTEc#;ai_#RZkCv+H+rTP>Yp z*{FA}3HIUfu9A_J%_okzi-GT3(fKC9W3!lNLa!s66TsEH`WN(h1@l4JG7;6=d`fu=dOv@cF86OX@XH5iSZq(k zqqoH=adN=bue?#DncQlGR5Za|FU7&$kRO*NfE#M1Xqs#GA?`LAgo_{BmTNFyqI`02 zq20U|Vur91dI5!N14EGl>#`!S;x?ixu%b5Az6ff<06bF!`_x||W|;{MJT~!4t}qce zYhoI_XI&*TGjA%Yu{@^f%}GP)7UAZZq?4D4LZh<`sGuzg;i`myYlhi+HQWF_w2!bw z=B*^i>)6-}E>C;G1Mw5>C<1)J0|US|0S(Yew9J-?7%+d@`1#ROew`P%A~Eby^@O%d?678%K;ZPO8J2Sa z=EwlK69Jgwxj+DVD?UTq#irrX`4o;VY0ihI75np%o10UVKlCdihsY)a4*qMtj8#Odf z528meoy*LOoZ~A0hc7VJ&HI$>qt>Y$t183P2#DUuH$)H!ojrsovFyKlEUY&&w-RmC z?{PkwOH-Ex)CeG&r#>bs!{QF>jZ3;SR2W25UB7IpXlC49q|eJmrxW&Cps$jEOpQ}` zJ;R3zu2D_t26IMoAHpd$T?rnzfo!N+%jtk62#7DMjtHquDD5+?!^zUueNel-X=Z|z za4|s6{;6drVujvyTg+xx!4`6-KOHAsrfVles=lBH0GK#()bZ4E`yt!7iK0W`Egq(J zRku4QX4od3MAiftVU(v!3Hc||B=4HReB!_((L_EJvSp&={CGZErOxKu<&FNFs{2-4 zm(BtK$Ay%=&!^W28e)tw*fZqc8TFBxl?5&gk!Elm#I;Lrhiec#fCNH}UiJFC-R08= z!V|GPv{ogg90VBcbEHR;_Py|cjTSArz#7kGQ3{@f4C$ecf`zU%8htfdjlnj0dkm7Qds9zrTgE{6MAEDQha6P44woLSAy0Yr zHn2WxPB@w(nI`NO07%HgZTMBJyu|FE=1>uYHoE@OGf|qxR6dvWOQYM zEdi~-t^4b&R{Z;LR0On^E_wu2(K8dUP`s*TTXy7)eJ~hQdx|WL5_sY&nr#5o9`2yH zql)dTf}(h#b?%-(eEV2sG_iwuehlGw67sE`nD)%Waw+-PB=`fqCPOAFGM)8a<>~}$j3kt zAF#?|w~|Jj?)sTT{t<#N<$1Sf6QUc4kCvm4n+ub1#6MMb{$3=fI24BB z!*XD{^;ECi9f=F~hOu4r*nrnuAz=B#Y5Wgauy52i{n!ij4AP>9E9-CzGA3%W8qAppkQ^`LgEyrpZw_wKb?dIi5LSwwu4FaeR_7FUyTl(dc?2kb zj%jk?*H+@{S2A$~N=@GAg(v*TwGly@nM99S0`4@TK5j3GHsT^Mb!OHY^1|a8!DAms zc6n8U#hH4`zKu``6G^WE-yRwonEKU+>fpoq_TVpTAkCnE$XU5Kk#YDW6$#Nc?_{5& zs0}~ZJ9sz^Q%1ZTy0%(gY_ueWF*vH=gQ)u<&$!T)0}RIvPjE+-pM=0#*%R&jrDJD3 zKYrfse!f552AENYpL{8(aG&a_$Flg`*4A*pg>MA$%UZmnujRE52(8({J&_m?=(oL* zDdR99$&0!^!yH0t@sTF#A&ogEcfUd~wJONw2OohHHj>?!!gk$Q7 zOqs3L#!GOXLo2c!kNxG0`34i%OnThvrv0)ximn0;vIjk{X_mup67DA2r1X8$B@UF`YBD*$QYW!Hoc{a! zc;GFqX@tM8`@cv8OW_7zp%n2@|-)gU?UYB6k@Xx{sRcASqukAEBj zD`^PB0`{EeS^Xkk#!jQk%QgnLa|N|d6?Oy86j@mIj25E=o8b+}D%g%Plxt^?jT z_1bCFjc4)NV`YiD38%vD*0mn^At3U)O{c=V%hEsvU3j0Kf^`B4ZOyxsz!oV;3Uq5e zjUfnWCVtUrcUGacR}&2@n`;W)H)5#|=|`s>!rhi5LI+Ou;RdAU#sD)$2p<#SB^0C z&65C6oO3wyyjbmA*czY;Jjl2Sz;W)&@m~y_F$MQD&$T%ZJJm$Z+MuSghH%ngP7anZ zb$!aXOQxyd2Ifhd*&OZ)1d$V(CG@6$(M;j`p!YNkaH~4zHYP#VcD47lG^AW-Nq%?q zwzgzDJ#CVk%gZMN0g}GsGt|Tn_eZkI*DN*EsR!)bzt~ww-@Avl;yNRt%6Bil#!hn4TL)e~b)VN8gUi zpP>zJ#6U9fLH|s#!f)3!lJbupDb2)}RZ~TCM*>t}I`r3x+Fl|!knbxTpB|J5LAoOg z;L_BDk0_j{zYe^uyU*KUBzazIB9UG8(Kr`Ov&S(oDK6F;0P>f&DB3*j0w$(zdzyZz zsf(V`R`vs-Itq@Qu^5_jkWRTwBPND&>ab{_AV#4Bi$knHn}-&1Z6>aWIx_1n6M8&{ zR8itm2Oiw-&`~WE*jsh$^1m2+r{GMZZf!TVZQHhO+qP|Y+_7!jb~>HBv2CMcYv)^g z)vjIjud08Y%&O;n96a-y_qZ=W6?GR*V|&z{XCEz=?5fD;&8$YD3=puWnW7%O0OPuc zwlWB+fG`jUeRg)2J>-F)`h_^hz%n7CXFvs~^9etuVs_u8Sg!9#w&hWw`pzwF?ihtY z`bnc@L*+9sM^D0e2q_3Cz#%mQWn|~UScs{$Bl|!{QY*FTwCoBj?(u2L{xv|<Qs) z-hqrfAM!2Ql|IH^_o-B#_7b=e;=pu@eUbSW7(f2=_1)YKJJ{^KkZ7Fj59t!ZG z%^TE5GbMQ@t%y1NoE+iQ|F%tF{;4A1Qh>u=VRUR7I%dCR;gLF6`D01Z@JtU?9=Iz{ z|MOcvpojeyvXq#Uz;39!T)^&PZi;AgGIRI_e{kN{#D6->u^(eJoDfB$;el|KYU@S- zC*O#^zf9oWOtuWo)isyoqXQs_*N>4Ydpps=$C=F6W#=7B;*Hvt3>_EW7>65{eGaGe z45{+kwO9W2uOw4c+~wf6?YolZD6}X8k@Zz}@}Cd|ZFjCRd)lI`L`%iV+Y>P@?r@-) z?g4UVOyUA%@$L+4XM7K4Hd&Jzo8gduWLU>iKLaqu`@Orl$4NZp_p=8JjLPRZ~%vj@+B0*vj=7m zDQFknpXvGRdYh!*NsEh0KC=*8p6@|ReG3l7&A=vV8wO1(g8f@eza9aWrQ}BID~L93 zR2DI{^xq@3IwL0#${dF=^i<=K_4p`1lpeIluSO8oJA{i(qFBqg8o`zj?&{N&l|UdY z3%qf79y*_-Ji>)a8%|3SW33Iv^rfCvr`_efgE6EIcO(x-ePcelOQ-D;+%jL?gk^Rk*m+X#3E8kj1f>n57nys_ z2>reyqX5sBX-s#3?`@tdmqTSaTrGdWXt}}I^}`D-nsrAK&OwuOCYW@xU_zQtpy$+x-^KnNN}Br{B7ACKe_ zG;d|#C=MPjkHYoqcLqkSUs)`M%_kjME-X~snAq|nN1`TtdaiT~q3&Fc2)XoTjwd|j zQZN|-^ucqUqO(0igN4nX8Y9xPCQ+9_8=CU9A-3#q6fVFdO|?q?{lIinvD&dj9q$<+ z;H{jGS0IGsC$?96O`L<3Mh7DDh!)MrS9Y50RAfuUjw@>pJ@{NvfK2H-A=HxMOf*8z z{O!JEzTw~*Psra@xqkG;leuyssXpL>6jY|82Cg*v(`EJi{ieKwxAA#$Lr%CB73_!q zaxtf1Fc;wq^!U$EJctf~=+cw&8#spvOyGawUhMw^_u^*#Ulg<*{g?CZ}ph zmv+c1%77N`Y^^Kzw8-fexru9-q#w!zvm!-R?*;vtR7(50(mxj56rV1}9Y_#e?^01$ z={%yb4GJiD=j*%AU6?X_28jGie=&UX^k~)mXL$M9MR|4=@Mh`J`h974J&s%ZWQv^Y z@@{$k>=m|&*r=pj>3?lfSi2?Iy^cEO829bw$He^(3wKw{tb;+!g?KQ1?7NTcIcu>v zjK2J);u_yxlu2bcf|SaY|Kaj9TzlrXSE^pw%r$&wXWRh1v*z(BVT}weYBtfX_wh;b|R)HV}j17mGjeb`1>#l2oZS`wk=5EtkLaY3e=lKcbEs6($|MS%!ijQ;&pxwqP z4$PFJpVUrM=&9HkUp$qH@m*e;Qg01A!ifMp?niXl*4tbHFT7LrWrIN;6rE=I2%PIc zID2Wk$M~dLhBk`7)sCEX=`Sy0a$uKKq}?w`1f+_KI|BvjQ{aYr?c^FqXQk)0t51xo5|m(<;;I66XIYnrn;U zdb^Dc$5v1>J%A9PUqWE04wrEK_CQ#!8pX+SI+-)p!~?|*{VWHMLkH2=Q*O5o2CpZV zU_{t?&=m};gpVGJUcvXfWEMn{*C(AEal$cnhR8?`wA zusEdK?UkSm&!-k4eWfX=S0a%$`!ATA5EYxj4V|9EOdCtb=qWXN9`J9*vwz@QCwH)P7Ts8o3DJo z;F~Zn062tjkt)$Ak4*-lhEKJ<3sg*j#m#d5yWn_Rv1W(@w$^FGPv<_()BnohsUI5j z`|8&KAQc4$(_W)6dT_)cI4xcQZeB0Z4~F2Vzx?U zjzp9pvP&Pmv?kr%g%iM^%!~cZOsP(xaZ(Ky(2Lb0gN_&=7QfBCi3|M#5gj)F!zdv`t{k>$F84f z8Z{G<(R>+l{5q6g3_w^qwS=Hr4xm7=9Zd;Hv5C+@jco03C%l)?;gECa6t9s5P#ZB; z0v)6+1mmkGk@6$4{wNx(PlA%9&}S8tVKWg42sJ=nT3E!nbo@R)Cp#4zd4N^g0%3KU zq0!}ApUq&BepCQ8^lBI0_@Jy{L7FVg56ijJ2cJt(8%1Qp@F4WE%H^Y}HJ4?IxH3_$ zAGe>C%HguMpcnHlQBRyTF_XG7l4|fdceF>-eBl|8ND+cnu0TEYf87~pWnK;Dy1%;^ zEjO2^2b#AnVBGv||2qxYuZdkpbkWR~AeVSehe%=c&vN~iK;(HB7-MwrD=?6;6a?w{ zvXE64jCN@=*&=~+J_Sp1xC698+J$VGnuY$DEYl?UAoH|XeeoqQHt!kO-{{7SkTmSR zU!fUeck@bFJ!}t>NGC;v5HH?tH(Y_6+woIEF1hTcc5Ra{QrJKj!1meF6{K>HWn@sG zl8M8*k>v+~W)F7hR z7qho#^45a{IBhZ!m&_5|u@~KkK+HS9dxYbIh3vulrC8?t;6lJcUr{WgOxPEKmS&N(A-yiVEF^;DC>K9tsI)IiTzoOUvMVb{ zH&{hFaAXY$#S>2mGX-r9Zyg;tpdII};8SwwAUh8tJ>3qE_^4R@+6~$U3%hL5-ef}; zkS)@Wu-BCLfy<7+sWD(){!=hwE~=`9hEFOXyV2Y^gvI zN%KmT3--AAw-14`Vztz^bBpvLR>_!djQ#2Ss}(o;0tr;e z30rRicqxwhJ0mcEDJ6lq=3+F5Nb2mEDVtgnXqK4k(e1VVesonhR;wI?oaL2^xAUA| z;#^5AMo5Kzd}hN)?OO_3p-v2d4$?>d0QW2wJ*Gdz2TO@y7p9sVAN8n0m4b@MwW$SHwdUhskmU&e_e?6axO0yQ;cvyyNf zqM(CaU0l@VGF>(8w;#iRFi_OZfrq$GG#AB;5Cs|4Emy)fja>`s7eg0L0V$t4P%*|W zmq7gui7aq-jUc9fO>Ve$U&(PCKuoNfh`d9itS{d|1@bmo4kt-3Uqzx`xu^n{5|onu z^I+FZ&_$tP(^616Pc^&h6~3a+9%|=osW*&ANHUe7XS4W|Js43NOf%HDcqFF|_F^a) znltyDt6SlolWDed#97=@8HD~Mi%##<1G5$~tguP1_R(<8mHALhWRSHSfhXT*Lt7np zZHPU_i-k-+Lc8D~ZTX}}uSbgTA(`Bf=gW}QWk_$M%A*L$fHf(9oEm|sqKLM+*EO_= zL39;y+C)kMYDcw?OWd4Hf{qrycWKt*T+m(tB0Y;Sv@)*uQbl{Khj#LV_8mP8ABSOh zV09KVajCg1PS1$emCOR(iS1^i2WD}*F{L}T1% zHgO|jq5M>kNk{VEb^pU$FB(Vj<1HXpwrVn3M+J)t?mf~latvFA9`Ib}P-+xe*UJf* zd;_|4D&1&X6x*u+0YY($4`M$g8NdjXWd|v7EUBQ&HA@;fO#)^uCLJgVkvAkX7LnYX zvk-wqy7QtkDunQ*ih0jyNkQQZ+YJmdouf(zq}g&Rs%nx;>JwAFGuJCnH2h(ms>q&H~C0>E1E9;e; zs41)2cH(xx-GX-O=YD?#>O%QmTf{6l;G@E9tyn`XcRc?92Z{|4!5TLMhdrd$u)vNL;{wCt zSg(Bn(Xjm~&`7a6j8k#(xx5(nC^#)GY#PcE0u1qWHpDv4P@t)hY3xOof}Rn0u`O#? zQc@j0yu8i@BHjPqHr?E(X1iP^4UAON)YorY5k)P8`;>qP+Tu&7@o3J`v}5T}Y1p`x zxkl@TjO*#JA`Uk&IHD0sT1s>83sp1^`8t0c>>g(rUWD*&zyTkF`_pa`O7}=gt)8eD zh*w8}UX5UQjD&0gax3;9^N_!Xh$)qDn(E(b`wSdwqS(ujH*15rr@g=5gg0X@y`FGi z6ak<&gz_&-yM)(4{m}xsgHOAP^+FVWL1_4Mx4gSE@WzlYCnzWUyNIt|Yxz-lbdaId z-{_r)#jZk<3LKEM<6jB31(gI-)_A*ub5tVVc4da)UqWMYPCH!WWnnp(|(RC6q>E8fB#`8<%B05h4?cc4|Of>rL)Un3KjTqKQ0_!l>E z-UEMmPSvHB_we(Bwi*#^63&5bm?m8Qaic{2TIAmm5M@IQrj2K${p(1XNntlY7ZRT$ z`mI0#%j#suSG!`3WOsB@W-y7mCF!2xAo(z$ z3L*qHSi;IJwZ5E!izVv5sSIsDXS5XO70T@6mO5B+ff9JZaL$Lezt>!+6L6}C{0aeP z=;L)`hY5EXT&#~yP?6@G5t0k!Wd}eYCCDK0&$U4+OG)8)-eYXPOOvtSY*+BGj_r4E z(w#?-ia{<=<0>sLXVRx5 z%$w;Bt=4g3GyTY`{O}Yk+Yubm>4|c;N!otvy=vH9#s4w%la?^Q72By2zP$>Ij+%%c zKFf*!o48cRSnbZk6tSeI;`kDrmkwEUvgb&s9~f=YS*MY>g4C=Y-W%R*Dorhz9a(A; zTycE1It_|#4|Uiw#0v^vUjtV+lK>C4W31|xhxuN0El9korPb}q3swO8*C~91WWj>l z(E_~#TAk!ZCb-Zjp&zP1Wq!?+-|3Nov;h?}=T{K)SC{bsX`-K&-@YpKq)4Uw!Ff49dBLQZomH)yCI^=^E;Te=Z8vkZo3+$()f+ndtD zl)Pe-M)bl6LZm!&I*ftbiQ3F{ct`luu5bydJR7#m04r#PTWs7}C*B*`hp#ZxoMKFP zmhzcWy*7E?ok6&(mi4nLKrm-ksdbu5f%jdc zWp6lR*h4B{MbFn__@%J+)D6KlSMZ5b^>pNS)m6dfE%wey*}v=akJ4|0GCVyp|1VAm zv0pS7ciPa`KC*R=9QQw~dQp3@hJtd!d}gdJTs75QUqBT&W^R86Tm=^{Pm&);J~G_}sp%hMu}(V; zNcL_e?p1V;D$cM^I^Qn1sz+UIom{B3$K)>}^>@5r4{nF!ruYR>K%`V=jyG3MBRR$v z7rYV#`uOznj*la;`$Oyr&QVz9ms1P%L2kve>I^Icjx9RSA*#~vFx2_6Y?Op@tgrW9 zsJ}SORwcjnW~h|PV~nhqy!3$ZJj;`Tx|RR@o-qTJeaoO>cJr?3U#og;I(o6FK8__D zWd=yiaO!rp%W>*1=)Ia$x1k{u_N~AawcZhuy(b!`eg>)Y-)Vean+xvToL6!DLUbXN zB_Yn*($uu?m$|+xj~~U z%kKWOUQe4MAyj`G_KYr*y3!6)DrBJFAzS?@C{=)!eBg}}xkC57Emg?Bv&C!C6(0;7 z8eS%&O9X~nYc*9#aFcoD&kV{@`mvnfU?tB!aNtKO69IC zTO#LU)J@!?{X#9ij@i>#0Eoqse$|SMnJGC>d<(%PBF86RYK|l@in5hnzmsx8pB9mA zIg*QY>($q5D(3Tb0@2$~$7uULmh6E+yKS-Gmu2mDQj>@xWskll*9FWZ>3uR6-1BOf z@a^Kg9eI&!EutYQ+!>Zx7jE4|30{99UT0E5|4%TE`Y9o|Mw8uf3%>R5=lSH zvAvUcUNw2jp%tE~CX8gGjnukLe-6F)h`Bk0P)Lj)Lqi=@(!Qs^Mtc7l_fOLBb*-he zwu$_w&CT+^H)DGZFtht!*yhQxP6(}av6GAt2@w5TXX4Z)rM^5pBK@Yw|pFxyodaU!yyYQ{D;H&`}1br zi3!tbPWh|_j5Jf@|8O|R4U0cl|Ha|T|HI+-+;elTKm8R=gFX=xb~SyqoD;)L?DzJ+ z{ASWPFod5SEqpEGCw`QtLI2@!%_)6QESJ9e7{n4u_qdIIoS9iF#)`@{jy66C;<3tp z>Wt^^uH1gL7@7}CCchPOHW@8FFDyDMe$?pkm}7v~tF5l0=g$0CuTC$`cGQ?puq*N2 zjuA&?!BUJw$oKQZmzQn_68KE1)ESwHIRQ1dU-xCa+DRm8l>1hW)fYbE@#Ge6Lt5V> zT&jnhrRWi;=(pos&L2q;c^;wiqx&CPJgR7w*)C>?1gJ}bGL@tXEJ4}qK)^NZARFB&Zy2kMr}%Jjc<3h^(V8I z$l4kzW5K`6NQBQsIR&5@!{7egDx(8gQI&eGZy!pqlyvgpFvZGh_%T8;Q;?rq~ZxDi=Z`#|h zTtuuiY-j_lkGVcZAI)&|cU_I;qT=ATVjM4QTBmn{QVavZY;1Kpw&mrWbT}=)y+B!Qx_XR@DQq0Ny9(~^2e@LO&_|p#E{58Vry3JFIFX`yq zmXFn2yYvGYCOJ(TBR&yJWQ-GILE>)P!|0*0?4Q0Bdw1xb8c!0;1l=?bzY$$5cqSTq zr(CreVs^lLs35S1NiG3}u$SFyhrW(}!!)??G`JHd5z{V)Bb{$XGa#Hez2#?V)zu$` zg3zSBue>G%M`ur_9Xdfm+bF@Cq!hpuNgkkmnvB%XWB{M=?mxLS0?8e^*4!l8}5QB*pd{8dQrB01W2 z>J>KAEIy0G6tazxtR876DLhnG=lsMZR8DV<mygs0fLXMM*%|ho#>kYGtS-o%IfDs6)l0!fD$u z+4pzWIiAN8LP3vjI&o-LZ+(a$zh9dLMkRy5$H$_D<=L)B&#Od3h=IyJgjIpQTb-$S z;#YwVhVX7L0&2yF(Xh!SV!r%qPTBF!TXeTrG6~`gdoKPW<@}5u;>19a1wJB54Ef8V zIjo?p0oRmyDb7MNin7q@<9#jB?{ue7bEzqKlZ3*~vQ;~#j_*~6UT!X|cuLA#-S1Q# zQZ6Ol2nt<_kQ7|O!aoEK>~H#_pl)n66^74dzT~(h?z`juM$h4m(eMV|y9AjX&nO*B z4ag+hF+%bvEhcSDJ^HaYhMp=Orbrji=a77g9#^Oc^hhuW?`i|~@y1`QpNm|U7<3Jy zKCPy_69bgt#+>e+Z?lKupXE4E>1a453X7u*=n{V`VM3LY31J(bO&u=By!w%y5^ou0 zs9T9nDCF7DP9LV>8_4~M#>5eRzX7%j3z;tt4h~r!HL`zKjlLINSHC&1$=?-NBP;jR ztUr5J@4e>tO4im1Uwg1>#?lmTvdrOsxOZk4?67ODT_E2jUL%bE%G-3cwaUYa7)U#g z@bvQL)a6__0{b-&v`apT#9R;Zd5-J9(1*5|xGUXj#~f~W=AIFG}>wKKyx8KGu| zjicR!$RxrKA}TFFN2UyE$BnXR3t3%b^%RZgcl8&KiYe#qVOSxP8_27bHi2UDFi)WrjPJ^bv#nvf!WkqK1v)a!q1yHjo?(coGfgNf63RcWa7q-vqQ6lPHF+mo zTzq_dQ&%#gnK8l0@h({V=Div8-Lb#W>aod51k{SIs_vDbMKAey(=5m;Iny=f7-ornN33BtLRh@ryggW3CwxE!t1_zgKk+lZk7 zthUSjeU=Z@x!A&o5JF5PPYR)+;?qIY0KG&j7m6OhQkGM3yaD3{y02b~`+i^LX@s#v zdsjEs?TP!ez?sX&$9AMTe>J1(zsWx-OJ0x@$plRP6qmd&JZHZEe!1(E5h)^ZVWp2( zRhhsmS)=N%+JlpxDw>gV)$l#UAn}#hjzRxUI04*awDXNq1nO6BD`(1Ei_J@24yRtV z^1pK=aXMUMdj&D$BOH)zIAJy>1Sm+Qy$VR-!J4MH~t#4^GgqU%+>XTxeb-`I0Ig ztl;(b;lT}ZK*}(Kh!co{qM+v*JU2B=kP10R^TB9UCTNOR3><#|*dsK6_son{=!4ub z5>#IjmnA84sM;Jv9i6p|M%#sf++Q&}1gf_Wut{P3 z<8V^46U#^PDAsbASR{Rk;KC6=l(bewlZWQ5|~jN#(CNKLX95XU&}n{_zR`$Mw(C@xP0 z%AjS38(w>?cght@Dh1I#bn)6^jxM8#7Rd&_(E{YVAMMbY0w7sc;qHPnGw*1?GuXcg zs)Wu&B@F%{w?_+|nXl4;ICC!Dvt>y+EC!*K>zitV7IBX|c+4CBt{4Z*K~v$s%yLff zs(G6F`wT>}ILa2={iVnBQCYGqo*l-Wawm9tG~^I3hN^9lP4;Lio;U5iXLvYlQA#UX z6-+y46SE$3^{7;^YllNrrU{^sacFG{(sLAMN=w;2D3EKWM83e|S`YF=Y$F9nrq zD8a)W1Ta{Mb!S)oD+dmy3*R$1-WlOjIKF>lJ;yr4l2d##(1`qe^E<}hY~33XC|7ld zB%~;_A`(>}A}`7~cV>_jm(gJ`O+_G1w_reJ<7RY7qt|Q|F_PAg)v~YwEq{acTVqfv zQgJXu*wNQ&U~r>e;^c-XDlvVB4VkA#|E!>L;YGv#OJ8@N9>laacKnqI2WD!{bla%z z#jfCOQ=t(@Hk2-FDYG8CQQpKHITBu*VPx+eG>U+;w&^*I`7oJ1MfYGOny8SfS_=uTKZPB_vIjd#X#hQ?6oTn` z{pl_Zj{{gyU`+KFASMWOO*+=qlmeL){m_VwTD=?fWZ(KM_H6groHAw>@GUsR4B&aA;PIfE-VpTI;>> z`jI1f$$+W78Y!Y3K<$%H{}Qgr$m@veug`#j5TmO?sc+30Gjj4_oiPZF13G9%r4kB! zx2)Q)WLFA({BcDwp1O_T_Zp2Y%u#D%^dua+#*9sq7}nyZsRwdkt4tqR&8sy(R?|JlCP z3jyq9^{A;cUM_VW4l^&PFZ)kqeR{KPw5RIxZ3~WDi8o!0ynyi7U?J!-A7;Ga@`X*` zg zB$wkOYhz1UE+jC^FS@Vx*7}7hxTtR#cfaBXpeO6_&dnVBMKzGT+d&((>Hbkw9U?Nt zH^<*^pm{0%30OjERTa<|l17%10|yI`Y{gd)aYD`6dSU?d46AGINuZT2g*F+sr`bF? zFV%QRvi@Sl)mkSHMVyIKd|Mrk!SsyjdWU*vkY`Z0WR+~O@6Hx6eXf!437F^TjDc!`3=UxEx4;^f{pbZN1{BB&y_)Fv828-1%$g1EvRGQ=yKs_`zK1%OQ%fo zE=tq=3rzw#s!Ln5T2h9si^+7WO<+wTU%Vh&Ag8lk0Q@6|p>!w3qAtlWCMA*BvRh`( z>+3^^Fc_t=VPnv!N{K4`G}+L>H?$~r)L&3IR|e-{&#QAkD-1`Upi8$z7gF)wolYi5 zDVXWE+{ek2hs0VqMJg3aR|2t5wu2-7of=ZW-P`3Kkt`1Om$7IxcK6hM^J$5NUvbk9 z-V7>dlX6}90o|To>74tvLU7Gxc&3|1+-L@8XrumyoUjQ=vIAR1kiXdj%VYGeOe&Zh z$%)kqyU7lxGs2|G2xMkAn4x!=<%y+b@;r#4^;wTzm`2GC&ynD-5u}+86KHmGoVs$~ zd-}*!57>|l*DqO8jxc*X_K+Pr{5=MOuh*ipAKx(01ogWk(exe9H-Sws?fSIZRv?!~ z&xIUb#skv(?~lRZmb`iu>xScVOTb37`oOxeZ+Q7VlH~7_^s=)M+chHjt65Tkv81rr zT~_gXyaJm)790rT@7-x)Yx&BYZcqPp`5HmU3Xa2M9;{`6K-M*QfG?vAG!+kIq28_S z+A?KYzjD-f{VTm>uLcAA!3lNZa@--j5%p;f;wFnVD`~uQdh1$(k!Mu< zI0r&3*bVb!u27WjuBiskUnsf%t=Me^hg@SEgnStOsKjy1eWW&90r{rC?GiHw>OJoi zH8G$@C8ot~y2vTHjn7^vP{Hl}03*PjUzWn-*kiZkQVOjivrPPbx|?eHU%c@h#9;?j zx%AZM1!$46`xidW{=el_$wX5Hp}6ITe%-L+Veb`QS zs5`~;28vrxVlT0GXT~>!sPs`8H_8bClaxmWW$jpWe0rAL}ycsSv89)xrK z`T9j`+(P7hcgGuzOiD-rG-lBa4DkK^*RG*xVXG6pOOtK7wxsLF?0_&1)V1l+hZKax z19$W-X~VM%Sg+b)I6ut7`cc{^R8*#=R62Yf8cb*+0Q8KRx;{ymPu12lyYOJ@(mneeA<_tet=F9 zfBU_`xA=7XeMMTRFnt{B=@JhKxl$nIi8*@Y5-{81rvTwr)@f5keJiEHLjBJI6Vm>( z&f(1A;{3kt%(Et}fI_Od6GcW8`c(hf(5+=QAY=d9_Zt$@Rh#vHqIy5qcOpj<8yG%5 zB1Tnr6E|-sb0S8ypOaaXh*8em!P3o&h>iPyLHWev`0Y@P@gjCElx6lt;=_bx_T*!s zOW>8#qnRnfM-s4jW`)`24n-J$jU3#(sfggr{ycUEhLmVgd+eH?6)3nDj{FNKvBkus zEG`Co-^j-@b$;CO0h05D+^jo!cCbSyLH*o>6Fb7< z8TNbcvATuCgq8C7U~Myd@^s``&gEZDMW|Z16zjo5iG6g{AJjw?7Ahu_qG_5F`N{w~ zB&7{}kK1xAX`RSWhc4XxHFdA-{*P&ndB-Yr(xwwxMfJ3_4{>Q0Zqgd)JblN6idw3y z+GVF~Rkc(;Sx6*4{;pwisI-ZZ(nxZ*#r+a%4s4}iK7u%MxX;?3s?{rW&w0DEai0%#XSO!SEB}xVkndhn#uwO5-}OfwDL{qiyIVjRvlRKIUXR*87@P!eNF5I9UWQE?Rf3aB)i$$^=u(JxrUCBt`+4tOK1z#us z+gM^rJLOsb+~wii?Jw$Vphq!SMJSx1{rZAC*wdH>I;l7@ws_MR z8r4LK;aya+*!2F~GloW>z0dv8`H-TutzN1-@ZoIANHONkIxYX5)13adZXxsk@?pqq zNcEqGFP?sZw~NGZUV^WMw?rqh6FrZ&b)KB$Dg3I1~!gayl0jz(u*Oi1D4ao}#RF z&nz%7IT*qOiU6&DLDTR4vOK7JHIii!-H=K3RgNR&RV}7lBAV3I4{QrlZZ9w|`RKJ`saDIiwX-Pn-Tizn9J$?qgu$W-0)B7Eni&oup%4}4)V>9i8xd?} z>Psy_Z5ASzA$2L*A_BO>2z*PCJT)Z-=Wv~F*aA|2>1_d^%Al@+x~g!Yh`=(+9=|i< zwFQ7e3n5ypwa4I;`_kRAug*?+5vg2rVeNy68~7TiolP%y^M12`od2!+9j8;R6eN)| zb>i;8FumOFo-G(&0w%MsF!E@|T%9IuXDv_1LFi?zSshp#kiWdTv#UKAKI@l;c-Q9V zsOuD9P?gdC*oCi!)}(5}+|WTuD$Gqkjjz9RQT3m@r53^zg9 zsQp2u#Zzv{*3Pf6{0(kSwH4{X)AD+H{9SUOw8;Kg6NQaO7VcC#pb)UthJ zIKK1S*X|`-q2VeaN@iTngP7z2AoZ*UY>2{NDmcDXz)N_Ms6qLbyt~U!L$x&ig+e>N zs14*gd~}TGds}bEd(Z1}EJTvHZI;~<=+;gMZNsu1*yWg%r;tvF>~E( zh*K|ucLAIiuf@0=5YEgF(k(`a@Q%O=e$Pa}?&nSP$-UU-jlCPpXQqLUqWilGFi01<*Ev+RYjYA<3sD;bzQ{t|6Bf zA{(;NIfSZj_xo|cW_)|fIDbLs&gHiTIsdV48fiX5BSfR#k3Y2@jFf|Y89gpnUXcc>#;=h6 zkt-hGilq3InB-Qx2vUB_0LYJ4Ke~+ypg@E7tm^Z$x%EtRtFzh$SbA-{9>Z8={q8%V zuf2D}Ii&Y5`6M^wSS&=Vr<`i(1ndNT*BaXcj!dQ=(T~1*Ef-$lED38%)*7HKCI$%VfZ`8sd)cr*TJ+ue=}$$>tSlVo-!H2E zTncm*F}Kj1G}K`o%T@=FT)x??HB_|}2P7%Q9boMW)_+Q(#@7B+vR;0jSWLr)s=k0* z6TWWPBx;Boc@9)tV)ycOeR(>as{s8?^$I-}=;H>L7HH{~v{1ENX} zG-J*nNx~y#D-^Gj7$J59WgX|JfuGT;*%2406^Il(CX0>JSL*8-?rSvkK6*6cW%zJ6 zT$?4BZC`{Wf9Z6sN}+lXkNRkoTL`Jb?=pG>ecA-XoWshQ$1e&s%zN^@C$(Yt=ikK6 z*BN#F?T&7#2K?astf!*YLn$$iTPz~_n$c}cmVa1ROBKWvp|-3Rf{`IlJdr6rZo^bf z#JjlJq~yoOfAzq{G{!J+oK<6S>WFNzC@rk91oM}sI`X#3;J`7|+S?K5v4pzwm#^Tk z=|d3Cn7z2)CQZ}|$ly{$3uu?qh=F|;@e8`^LERK-p0EoJ0K2FKzVpO?T;mhK5l_P2 zh}=Njz$hEULELPR7mbDDk%H55wXTOZjF}QSHbXT7O3!LSkD9kZlqxSzfU>N*>FC08 zb!BEH7b8N6rLxSNe)Hit#oHaKvqCl=as3|dHOhrG3IF|<0O2if6GqP4zqv;9BNpeQ z%UpM-+^0dWAN_S?pbw>SvnuK4T6{?%qCnw(XMvY#EWc~EB`Z!i@pui&YoPYCJfQ&X zo7*0D$QX@FJeBH{$^ab?a@S0=8azVPiI2a1<}GQ^&{vWt;740Ru;%?hc+HWqq>@o! z#Kwc#QWMDpR2L=BUevpUM^3HA;{#s^uu9sH0a0~9LQ_J%Z#P7!AYllZv~{F{A;sB( z6yK!zL%?r2f6t7UP=Hk&RPeY;1&69t*A@Ze$0dxueu+0-+?C65zaXF^)>0dbF;Z*P z`PDs&0;Bf#=V=UoHCo}27ZR0ejv9ikz#2-%$QviH7;b=BM|B^S8(~jK+|Gw?Fl=Z? z{{=#%ZZI@d@to0CkXUl3HgpGi>Kl$!lCKeZKZzgp5u!`AoC_JQ^eVwRekf?^rNR@6 zD@?@!5%}3KYDEP#?nmj1#57d>7)Uo}lRAFkMr_Ia5 zsBU4VJQGPefiwHa!JsZ(t{vpanP;XcJ+qv_o)M%#a~JP?3G6-&2hBmSefF1U6XZ=F z9rrJ$9mp)H+K?_m^AaU=CdLK+<(G_sDI{;D6nd6$7l#yj5^EdlXd@53A2&_Q??k${ zTa$nd2Ozi3$$UO@IC!3xj%@V_`nD#S7ZmOh-P@Os+cJ2CXO9L=rGc|a4BXa~R9Nl{ z)4w~$6cin%>JswuQat3;9j}(Od`S#;oL%swf@7H zU}=Je@6IE|n;Bz|Y!xsHrt_YlGU$7cy2ut*f4l-A;Om`WxbgQ@e|JSqo8>UYrAdm} zDQ{~_K3lmc@I4Z1mDbAOyTVg+VrX?fg03W2&GXQxA5rGrBA5s;R;%VpMO~2mRjcF( ziCEe;Rsi(@&P*KeoKpv!DbYSWT%ub37%>NOB|s@`>~IQ_*wVIR^gOW-o$=MOmNI}3 zm=93UGucVOX(NY;UQ3ZFofQUVhDEDxq(%mD9Lr2GXsF=7PnspMnY zvj@&dQ=U3&*E2^?hyrz2aAh8?9r;VtGTn*TmdRHSAJ?#voW?Ej0TXLuS0s7;L1vx} zVHQ(EBQKASuH9)u5^VoMm`Gb4RJGYte+-ya-<;=~?V7kBXCV892}dwf<)POuc=xo~mkguLnerp+P3L9pfI9&IremABZyetDMb7*kDIm2=jWhI{` zUwe2|nM`cCAzb-2mOzNR>kd0mUfmcl+G{Vh$H0BDh;2^7_mFC1t0y5kA4J(gUA!h0 zOP@M7?-Y!^O3-LKOM9f3j|9FAgi3PkRymhe($hS){oJf-X(!I(!awO$v$Zr#Rh@ca z91a>|$VCkfkyd3Q7HU!ztsD|NJg)}(9WZANX7qe37K)k3aamB*SgzuYc@+-#3uN2t z<$->b$lR#0PSAQttd>}dOa!a6wJC@l;Htn2Xvjx z;Tc}6MvApisvRTZE}D!6o|WUHF6!;@BT1dP)2jnXK1~kRydQM^T5WR$CZ)f7rI?J4 zq+F_Il+d?hPGFUQiliFi-BSezgv-pF2*O9f3rQAkUr8`61r%)AY!1laze-FH;l@4+ zu~iB1nW}PE7Q>Trq~jY>$9+aj{U?s-^U@TM&zguK5Ra*i@e5j=L8Qggr+=xj!sk|9+3b9Xt*k<22f~1; zX3;Bf@$nxawT}kUMXbqNp2xJLj|jk^vWq}xDg|dKiU3a3*;03ZFRsQxo%ztRyFqvU z*bEYgKvCkrL2JF_Pc!&{AG*_{78k1z254*HRHg`799F|wBs=s zy2k@{pPHcR06(8I-6_}wItoO@rfc$s`M~O*EhFIT^B&`rx;#l6>a5+_%V*lkzj!R6 zfc2~Wy6Kp?2uS`_mOgtqkeF)qfGG>|;5B=YIux=pLYf7(Nn`vIwTnd6J8Nr#f89!h zgBoD(bi3`aDUUG+;nVuoPE1q@`J{va%InWlz{LK&pv(9DkoJz^zbqz>>gy4mu|C)Y z3#K)|*NtY9T2%#gRY~{zmnMU(s`RG4x{8YS`nrL1R@UfN5?f}&3|8jbik~@@rgzV4 z=J}j;S3V;Hft|zJo=JPAz%BfAU+Bc;48{<(ogwmpoMFEQf2j@oiD%9Y#l)uEj%uk# zArwWJL1q@pudHI&Iwje@ht(xWkvgUO)~3JGn0*b)=oQ_(mWHbFX#?VNH=k41(sdk$ zL8fY$DrcsZQK#h_J(~O)E7ek_o=AgV(QLlb@RFt;B#Vq`6@%UZ?PgGNyF_9d3>XTd zg8Ljf%DC-?>Wh6R9EUrlUuk627MQ{2k7=n}5MuIKG*i3*qW-x%4dL1V;{IS_QkJNI zrZ5NfGzhtlvE!&omyP_c?8&98oK9gmPC!PV&0^@Y-&_v*nhF=DZ+gLoP4AG~-!D4| zE?A)=0f6Vkt-E1tZSMZZj(&GX@x0Vl9nPYzgZQ7Z-ys*zH~YUQUcC@yTGXq6nsH3~ zoaV~ZjJ{ihh+A7fXSX6;}E`B%(7)vG)4TOjvg}v;5GowJlz=7$pVk(_;9* z?9Q|k*9RF4=v5jqOiIONk^ABe!z{rI^V}gBz;)lzwc({9Czufp|0-z1Q+)Q;Km?BR zH<~f$mrl0jUN5>bNye-BSLc~_f_Vh`42Lna4w5nnU5Yzm1Mj|ZI>a6Tv8x(X>5CiZ z@N)zt*GX2ehU?F2W4@Xw3_IoGjc`v!W%3v*=(zBWWJpURBv*0uyK(;Y!OUga8f=mQ zY$mBQejA6aKb$u>mHRdM|GXtlA&js3^=Xy*z4moUQ~5RUlS@|}^Fwx>8teii^y-jX z7Q^W%p0H<|_bFG6O7@Auoa<^{m*<~?f_1sUkH|*;v+cP{l*ArvRKyMYeDGJl5 z{3VImqw9J5c5$Pa-uGON^M0!+D_I?s4lq zDzfu?#dv0H3j7RgRuJGNqLFLNap3ZWP!;I#V6(`1b-$VhB3E)wSN8UWpjq(u4~Cc| zX)=hRoM>X1;MqIVku)87mGe%6l>j5b!3bs^c~?bJH$gwp$q~8k#Ji)4;-(4|MNDf9 z6;mB#YV%RTyA-W7P|2X$qCnvBHkZE2ioCby)HIbFP|#G0%mIHKYHO?nlC1-(+Y=7; zuKcDqM5U#i%Qq5mTX+=XxFt)_wLLkY`18td9thf1r8c!(p;15cnG1 z`GyzDp0|;GnSaKCPDeJ_rkBXT5?Knhz*|Mc^lvkrl4&h`isY@cHF#82)KtR!lO==4 zq;6X8j1G;Ld@S<^&I(tRs?CkJR&$2E%j}m_D)AyI0XZvf6ol1&&8E##zj|^5kC5xu zy1Ew=FAVSl^^*P(b4U$)`fmxN!2g>t za`UtQFJSP?|MXA)BWF=Hy?}Bt>@I}Wa7lB zF#j$7gFqDUXnwEZ%MD{cvP8^hL%9Mt%SwMe%*-n7%`b@&R=Dkb@o5lHRC7-J^rt?( zRQPCG$rbJp6%49gRFDtTx;pr$-+E4>GmzI-KO5>yvUO0a<&)dnyBwa6Yg&C&n;A=d zv(2)~`cKB_N1g8qI#I+53^OnbOYFv-EI*SfADUke<$#u$cSGi2lwvjAL zjD$=7mofT#2R)2Zp+MyZLL!KYh2hXKV>@8ssq)J4wor_D{c-Er?R{tDhYVck9|HIzv2SIQJ}+C;6u!4qz00ZxZ922Uv8J59wHd}-OeWY|G5h>K z-V=`1Rk7VLvIcJ5vgTrFx7%{Ms#1I_G!uDiY804>UA05o5M|*><*=+93w`YXohfRlCu%{mKi7W zA(8%y)@f_LHRvmA1VoJ@X4ZcyirMdCe*Mn9xmOTWvmBtWo?fhI7q#xX*R|#N?38d_ zZj;QQAF8p5PwgTzUTIUmL0-Bb*&FruYhnx1sNjkmvo$*FNl7#FvOq4!KiEhK-^ki6 z2gLQ{AP997)!kFeMT$c_L^Q1pqjAu14pIPdy!RfX@bKL@4$beCcM00!@f5j?dZ3XN zvAQQZO5)ks;-B*{8UqVV;{5ZjE3ty__GC;hW2;PC%)2jp@^x?=zCrS&OC+RA=3ctQ zl?k0RWQ1h{MIWOcBGP{sX0Jh&&SN_D;b$!4_Cp=k%FG5<5Y>ml1j%B}Qf^v;5|*Tz~)! zXV=m5>^NrFbj##VVP9b#tbCw&3&4Sfs4&)TY-svbAbuFM)I<;}qj)kYE`H@!+WM&c zc+m%ACyx^0(UGuCl*crDQ8kcYM6eS-HR9}JL#|^M;BvT7xUVNl^@cRTS%c3lzXwV6 zvv;wc-*GUnTP>B9w;z%2iqaMjqba?_-seJlx&El12se0zJ{nhW{U*5PJkxy%u3N2@ zc!#Kdt-Acej@2C2R?%q13G4dL=kn>AcKk}{STfeoG9m)!h7gH_xs7-3OvqsJ=Q^Vd zfdJq5_Won2`;#YqdF!NfVswl=XX_!9lc_UcH4Zz4m(wrWM-J0hWsaM~p1Kd{WA?KRpB%x+c_p2kikTOf#`%^>I_^h#mH7hqIM2DmKw={K6 zRXpI#$y4Y0ly7Ipr{?fpVJ8Udg ziZP6dK*~%xK1t?lf;&-Nd>#@)w}m~0D*O!o==jmEstLnYoWZE19z3SKUmm8;7Va@9 z#SDg3{a_dD9Ch`DXxk$Ym4ZDoG&fXx;_%O^6KgxPh}u?%vn5z3h{62%-9~D4Gv~6~ zH+%bP?T`KDa~C5o4Le$}>@!o7_5|DlO%-hQL~+#OAr%=PJcxAVm;dyc1gTwkE}Kq- z`f5NZ)*P2$``cjJMyRb&mRE-01ym5+)g_CIY0+16J0hyHvQ3psX!(riE1;-?Y_U_Y z;RUz_I(gM{2lZtVZ%+sIR1-33bt%_YTEV;_rKjG zn_}OSkzws&xh9p7h0WvM`3RLRz>|kUHxirXJeXViHwqp>v~$(CJ^63-whzXYf2XTi z?7>|;R?WQJCCZYeRUzY-^tp?h8#%tyhwVtdWdv4qM<8RvE3qDI{;4|mHeP`k1aoR9 z6ukZw=_r<-`vRUXgmnC7DAez=p@>=0w)3KJ6ee60hzZM;o;24Nvnc--b4tfE-w8G=%e9+5PI zzFEoxG`1Dm-8*de!&Nmie%1s8?@S|%N)<2%t(E$|36!!lTig-V3@?-ZBP%FT3sA&f zgg4Fzc|QEU*!#7WFS{f|{$04@fm-S_Fh^jr52*>BPX>)Y0(oxwy zCVFJCWl>K>Uv3a-e9>1RPtfyLhF=evd>f5L$$9?BG2|aCUdts;=HTp%z1aYo%cd$( z7?ok+bPgxFt`CML{=D6@hh}{(9^v0C$p%0xp31`8I+Ba~8;fI`Hy8STOn>K9g-qLe zj3-ZdyDwgWFY?c&>X%dM05%!O;ZL(!(v#2;aW(OwNOL5QJ`G*w``cJPDpPJqT-@Xa zmzR)N6!4oOLZ9DbvS+RUn_Ux9E1~#(QHsuZgX7mgP@_^jBifeEOB;K@6v}cK9&PIV z?5`BM&lEkd2^ph>KB9;v;(^t- zR_Af=7IV|cH1w_(L2cO>UZrxIenasMq$(sf{)@1@%W;~n!B9GpB5iue=X7{9WZAu4 z>3(CQ{Ql&x4CX5n{@Xcv;j&O~ax*eH<=f=ZO_$!rxQ8 zDkeF(beGKNe|rp(kOue=Yx+;+V!V(tmRP#UyF&Kn`9M?u0!HgsVzQxa7AuE~vhWF9 z9Hbp2hvs~$rR=2l{*(&$@ji6xLppAJgg)jEt+cX zp>{)__7IDoGOOl;G69%TZvLHWr(>p=Uo|~!UqttdEN$)XM>_cU;hzd4XGj_gBxZp$ zrr^L9ER+O6!q*ppA}s<0qXCuCIxZ`fxS4f=kj4S8w-iKbuYhsig>Np$JJIkbc$(4? z3j2C7T2J|qB+M|xg_BpW{Y?Zk;$Z-Qy`_;z3+{@OO@PizA!FosSOLe9b$KGa4~alP z{Dp~BTC~C%c2+h_=N5{DWas$$zhwlA^@ZX{o@J^%=FLy@%IUfk28c-!K(Kvp1elpb zVw0#ggrcuGjl|7>TJ{~kM$l~;?!YML_SZt+FZ8koIS@|w3ZJ#ta?rH{AFElyZ_L3M z&u?X}#mYiAhecVeivYT0snR_QD9+u*ep9R?Br~6!+ca`+CICMkqRX4)i%H ztJb$gK%8r^GJM-GHc6`L3}Wd+_nV4JR+KV^Q>bl^;aHV^LZe=Q>br&zf_4jR*X9Kn z&*UQ^IP8RnLKE0tdKvlc#hegrkUR?ROp?|^s$<8l!_f7HD!#u-X7Ltl2hIf{HW=Xp zyP}YQEBL3t{WWuah2>E1PGYI%W;u4H7Tysfh6QaCn|jAYZ$ieI>|vBQ}Di9lUQmj-Iu z!I}FZc*KUUz+A5?U$FD)wAwnaJyR)-MswN}WH@hhtxN|Y zAK7CPK1Pu%R(z0Tk@{MCfL@-567ZTf&Nn-;4`;`awI-sWxr3`i0=?`7n_kjwdF-m+ zd&~uIw}*G+Q$c)ImHEP@!vH^bbz)*;SupP|w1a*p|-eW{8;Wg5OY|j1eY2m%KUfG0F6|5RO z+Z#I@<%HuUnyliq2a)<&)3Eq5$%I%lDK+%fdOoe5C_^2!F*u>Z6_2%6AuHZj;yuNIy`#$-N0{PE)d3qh>g27T5qcymWIpu3 zWZpaTs05LqzUf)Hm<@YT>C1z=$OJfe`OD!)W-I7)@wes8h3pz<P;-mdc?)k`i@86IolSi|(NbCJ#`mC(C(QTFjXuGp&0NwD7sl8QusV~o zxjShUj?OrAt?iBp$m3f~IRUZby(Yy&o_iF^pZ5ZuQ$yd}Po%L?BsX8tTy9oVmaR%6 z=-j@Ca2NiVT+>+|iuw9Jd$3MX8XukQ1IEsLuTDKDCm4PHH*y-9Nn%oT)-cRiERMsK?|efIV+@Z}CPguiQ;{WF?VjZ&Awh`p4W5^lqETN%O1DNQQ52^?Dw zh{8|f7%e($h*`LN3*N&C@_e`*?Md~`Vml)Ow@AR63Du1V=#Z-1m&Qd0E;Fh38~^G1 zqxwMU?wNy}kX>Dx>D6SMzUVrQB1JXbkg$M5zxrPjGrc-?XXnU`M zsdm%-i-o5jm2Ita=pYMWvwR?e(#H@L3%@_Pg2Sm+D>JN1%I5k@UeH_WHs{dI&N7W_}9R^4<9U4k1n{zZqBqXRR52r*&3cTI_)g;|EhSnNB1r> zTBhfhE~d@^kZK9a^@x;$>Cw{tU@>{IdQHpwvPb1cU=NR|8<;z5%^oF_8uVaKdo#r` zo;ZIkzC4HBsRDe-y^i-1T!@D&^Zb0^^hT6*jP;f|qRJ2oj>FB-qawu3Lfd>vey$+f zW=fyz&loN`P{NSK?1;&v1mpUmHFN1$_G$af7gE5ld`Kcl*qGBJKj}b`&`2Vn(ltG}~iU;e1ApC5M6sa`=|xU0ItX=YxVo~K}~(IVO4btjf2 zcsGf7=DK0wg-NuN((_B|tDOoJlKaLphCcA|g$%A_XE!8?0;ipW$fdD%&AT%o(BB^^ zbnuhbU?FF|aTk-I&$xs6UVstS9?2xPL%<2q%vOJH?T;YZ5uL6D@V5<+29C*(OKPH* z<;7HkJY+72g%R1~&+W-y0Nclw;eUz3_}TwU6voT-zwv>5nebj8ch7UE)!^KrugirpEdMIMQ_ra$}a4+?|csmh#Gp|JKe(~V+aCYB=hK>SBcWvzOxOhcMIq;WN zlKY6~i8^#Xu;Zq$~YIRQDsNbJS*S^1+a=S2uwH)KRfiR{<%eI+Cm zZ1HNN^Y|l==si2Z;JBAP9<+%i7oNT$_iGnR@6%@Nf>fhV`SjN8ip@awB!O5K!jB>Z zH{~4gTB~)uOFHD6U~5aS-5EV-gy(l=G6m!y+__YSz{ls{*miT&@Srwh}17i&iK@n|C3L(6>yAQuJe? zpbtw_r2zUWA^he69?ov%-7jt>+lIoTi$xgfZjAcH$jIb)9|VKxgK=8g)aUL19N#v4xBDWasuRYLAwJI`)D(7J+C|s7@|eS-&DgSGzio zammQ}_y2ZA8gr7-kZ67l8^n4_m;so%bzdunOJZ_!q(o-JQ&BL5ml&YCa*KllA z79Z3XrCb({Dk7*>Hn-=)Q(YU%wRVVTBl8=wiSSQE=wTy;00k&V{2(;1Lj;hFs8+%? zItND`!V(=s9mODLlgv)EJO4HgdV72fe;nrFkezvC^PC#9P;0aHPr^+jg[mpuU; z2f+>gp19t2z>hu*3%qy~IY3gjD7n12FEV0_R`?mqB-2_xyr;_%fg_R}mZy7oZCK(9 zreJ=DJsL7y$Pi)TOGXiBUk6g<+}V#xh<`=8kPMtuUBW=R0g;p{N4MViCUT-JQQkfh z$kMIkO3-HE{jrgd-*e

A+q@59p}ZFK&vtj6$;(vRW@(0&sjji1xxr>;x7_~$bH zBJTK?TQU?Z@buw;s3pG2&E_ETJ0*WGWeAX>MIP6{8t*slUH6b9d=iW+uL z2RkZyadv4^NBjs*5f z7aE^j(q60#cTrF5AhR&W%rayn?+D};$6a`*n&M&Xq9=)CK-pdW3|$x4E5bKOy_VM?#4rgFC%fSV{JK8^r;GUoDTji&{mutl1XBDKb^C|bR07w zz%Uq5T0x2}h1hx(pZ;Z5XFjI}#V4SQyW$nwoC2PtnenXZP1!V$By>6)Gp8Txr?~q^ zo1M54a5dIipf0vH%j%ftg;|r8C%upOQ72tA@K;E^_(&YwK%SHv)n;(JRFr|#$MDS= zOMm77lL=JV*0ywf^zW^tI;Bu|3|osP2QN|r3Toq(Z<8l7=%!Z~Q4uAD{Mr><;nt!T zU9T~iyZD@qa$_`r_oOD2pYfO+(gVF{IiX}`TA?X&&TDYx?)YR_{3NK|Q00&Y*p$2t zOe5{6IV%Ws3NIzz=xcB6;<1?>`}TS^U_T!GfT)pd^RmU%VIxYJG!YMG@yE0Y|HWc&zH3vrYg> z#Pc-%i-9-ZtHT*O zv!4!0dBupMlIjNmS78gNmad)OiBcGURh9_o<`dG-h0aRYviDu0-j4{8Qo@qnWmb+N zfy%&d8wcVRu_IBH(jFuvVUD@dk(?aZqEhPvSUW$xW46mnetECa6D`X2TLsHag~$^z ze5D`3@DNBN;N)Z$0gxL7Lv?NZHc2bch}6W=Z%9o|ZF)g!n6b-Ywbz3F`bVh`H;>c( z=20kOyUA`oEtP}7(6_rMJw0nH?+ zlb%gYScxvE>X}~aUP+kkOm4$I4+FiRlTdh~d;Q6hRvi)qlI0r`j@}}`tA(N_FFDy? zA5tM7)SvJ!WEP7Ut4H?NEIiDU&cse4q$u1+#y0pU@(l7%NHS2E*cshSx*9j5MDN*D z>Nw%CJdgy*v^c{~QAp(q)yNuP49KdEeSwN{UQy3=y0517nEG7U4}1ds6WVen1Q? zn2-=Tck;e>bnJ;@32&4t^Ym@{q4xLm7uC^y7oG+?K(1z17RDA^4&`gg9pX)}WjAlS z=*-=6sE@=lLCe+i>iMwzedF_dsi$bT|QfB)2OeU`4{DlyIC? z_?rjRBRb=x92YYoIyda)UADD|qE(KE{WXT~zK5O&qjmssv^mUb^~V>Ydu(vdx68#F*;J=G16YIDRrNm>x}qTwZkN-8m7t&8>Q5WD)fti+a!+-K2(ka*|O z_ZAIKvD!H3BrNuFY>^)%1eaAtP-J`@u3C?MQ1EN+I>Cu#aU@9_Wa32;ydy7gP3+eT zwdjiOI8kH8Mx~nJ-vy%}MO4W|!CFLS!Sg-YX^DrM;_#~*2J5cAJ|Ik?7K zNBtb5637hrd>e@%eLyS?NPoo@`JstUqFdGCNUi!vDgMsAD75k$x0@K#;(evMS0K8Ib!$@&i79>MDKd@M@i-}5tj+8y?$?4I7LjA~I#uuCliLnf1pUH3?$TGXwP zg?Y5crGSX&Yg4zL(8S9g~@aWyuiW1~H7W)|D_RV{J$ zdjezP#j+^4pM{Ym+;i<>d};7R=G8OBsM)-&_a#p??8Sz@tCyuNyNh5diK5^8ZhB6I zt@@4etvz9ki)SEyfMPd3!%>;G>$9*7k}Loc;#J?H1#bfUQ+Vc%asGNPe`@L>lo;ae5 z!rgXPILZyiIQ5qZpokTkD+-^CLdL$1`|-N-@w6@tRcgcjc;YP_c4@%-?$8?W#H<6f z0whR@!>hOkxK2YOTk5cFoA$rgz2$bu{?rM{(iA}ds&=C5MuY|VM*c&tz3_fq>2>H~h^+EG&d`=H3BF^SEL@Y20jw3TL@Woh7&jO_H*KV+Txxuf8URPRJ`z=b^;ordkW!aRLYT-sYVUX`7vYfX0pdy3<13D$>-WOmo% zihez)Xokx4RQ&J_zje>{2FL2lzz!JKYMhnPd~+q5MxrrXZYRUVTW_3?^WlhDa;WZ_i-VWbD;gSZ!lCHaTTLi}JDHtlyEPgo+;y5?aTeE8 zG>l@wpH;Qf_KXm{l@(aqTP9}=Rrf=Ukmzy9hdV3nV)AgJ)+wdQ^( z<{hYk-zV?{-3g-Kx1~J1Wc>p?U+c(LWi)X9EPWEAu%6y;ntV4SyCY(J=i4`Ar)--0 z$fZqUN5u1X^ZTHO%=3G^qf4Y5k~q7fmwa$i7{#E}`_=CTZO{BZ53g{A_nbG6G=hv> z#7moD?0oZJwLwBNjb^pcLv79Faz&3k_^Rgb8{KJpuum))j@_}Q|T!Q zszg73aUn09U}8RSxC-7`fM{F^uEHCxHQ}t;oN88Ypw(%f06)je7%v8OAAy<4O$!Jp z4}n{Y52VV7UJZF|o%6gyP+x2XQ2_Z^GnX30W%HydIg=CN{1S93C(MwXLF;6f_{2-Z z+7%VL8Nb&MRZnl4ir5&ZCY-~6%zrKqKONcNMVx4>`iB({f<%zzJ+dZ?1asqgl9p^WKZ4&;99~h|3bdowR{^Z|$0W}S6ya%3knV6>T*B83znB&1Ua!v(J-HAX-OxJ)ck0lDBi9Y_kV35E3br% z&^VOEnVV!sDOVevTzJ(;#d=uR51$xwma9jtH-5rFI*iL?P2YSFjrh4PoSnI8i30(+ z3e|rH?ZC;&ws80u{vp&q)EjfFbkpV;%iuCT9#29bDt^fW{#(Mt@m~|B{{_zZpRyH^ z+<(UY)=6F0=pFH=YJJivd3kNKsI+)#vaBvHB1Cp=Az6x8Kt#@f1$i;g>w7HPBXHD| zoL8|#81~?t+W>}5G1w&U|JJRx<@{*n1U@rARWs#YPmWHWsT6(|KYRrf2L*U}1{N;T z+srcXf&G1ks1E3dlx0d~w}zDsg}`4(W18Zcu79hJ==JuKi{Ter((Ui{Rj|O%PUnod)>Saa;Aa@Iif6_#=NEsEweCo5<)83`ZBN!JrWA1~MGdaFN*Z3_K zSK!3@S1@1y2iJ5Se_lKt=R|tM@u7Y{wFXSX&q4#Z6Uvqsiu3&N7w`3+auZT$qdgxP zB&iyHHr^b+SBL;YTL6Of>8jlApHYHm9$(HqYs1MnYk!iom%^H_!xj=E#8)!dT|E@9 zJ@!Xld@joOVLM%TA*B1Q?x)g&<-3JYq64tu-z}cZH$-qRnU&tiAH{XoLRY$#5ZxL# zxhE{NRBm5(CoM7z#j<}1$MH5o2AVpr+s7Nzt_|`QVY5wu9nH11e4M@nR`7a!Sj&&I zDWx=OYb;n3BDs{{xvU?Rv1CkC_4uKdQ)*1g=58e@nBI8ca&z;U#I3s}Y&(HE_7Hw* z74;}ggwe-sPS(Yq>NBPMeOI`5hQ!QwT(i)v)$&2JaL@`Nj+rZDdo=fa+@D7zQ;(stwW&3>OD&3Z`3i+XWP5XU({Kns{aRiD;yUEI(4Ef;DqgQK@4li~wt~Z)o ze;G%44FAd!E3EB)V zVfrQjH9H8a?CewUBS2GX#`Yt$v%z!R({s~MQ3?jupf_Cx~k1gzZ` z8PvAslrW9#+8Nr`Y^OIL<2U13qXbx=b8<|pcbn$-_};?8fl$b!4?dG_Xu9-TGoe&> zSSt6c+WS5DXy=wnTW+(Yoe&ubnV2uAyTAR(1ugc4Yd&!EBAduDQ?1TFvaFFNdQD?~R}iRCB@x&AY0+kVB)|X*9wU zvnzh{d;)fz#r%pBCm`^p`jUSZ$_LkCu_(?0&9x2TV%1IygqBtXP5gO4^=vQ{g*Hxw zgPw&3Cu1;$vQA}*XL{eIi{m{m`1t!n2GnP7$bYr6l2|*)S~A}bg9Rc4G=`$TyeIA? z@T{V(_y`e%HKWulRO=-6JB6=AvmcfIC>QO(xQ0iwQ|(UdZwo(+$PSE})-ZBjvu`on zW7%c#{SrE_g=yTq@Xr%3 ze%xmSKS;a#D{07q7xh$s?a|G!#ZlT>;Y=&U?QIi8AP;U|&K`PA>;3H2eLLu+HZ<}p zH-H?97x$y+5@Ad_JAZOpLEGj)cnEHRurz!OkWv)OZ#b1&A zDxi6iQ3i3j80r!FpckY8k3b@>Pm?8hBbe}S9R}%Yt$0|%L6^cx@p0(`^Cy;oHpJvG zxQmOr6u_ZB-A*FRBBUq|#h((C_t_xv`@6+#*w=BK-Ax(=@)zQ?a^|)C{mF(n>fqP* zbOj^XVYwP+o4M?*xJEAu@z6@37luA#|EC}%6HzwaF8)R2M&}{d3Ll##ibfX&7Qd}d zeSVCT9%=*3muk;mcr+WBu7j-P&zy*K#jluQ=|YM>8}oQWvch=~bB*y0ZqQ)8!PjO& z_mqRUm>1KBu=%}x9gwkhZEYsUM*Y~x5mL{&)qX`Di`gm&LD{X)AJ2n?Qzb6P#ypKy z!};iPk57ZfhrjzR)EJ^PP#)moRZw+3s2^ij^k$-m5A3ys*nq11fjD{&5P3q9TcdnC zsr-E!S&uf`chL`E`mt5IOxPh287zpBVbLiW8RnoJ2RH1ir%rd{&4o&&6~q_ERvXVU zEl{#uXwm>Ro&v5=tO)*!(7O~WR+>r9C~?^W=&}!~V>F+p5kPGBIAb3fIfjLrDV=#1 zg=|RF>YhFy8kRtR8<%7TQWZl={x`CT&pTRoaET3!BQpRJ1teYxdKqo05^boZVQvi8 z4c7@&bi6_aETB{HV?{DOK_*gZPZ?oKK`ciUuH-BJQ4jtAEW|ou6~(W<`3b7fz)49Y zdCMW@fOA1N=&T&rmQ+t5PRl$kt<)_KYzKj>t;zMq+{r2$>2%*2q^r9gR$btnBdcue z^Ts4@j8wmHU3Iph-U^5N%E~RD%8>OU?{b{G-%Q<33=ocxabKEB_7y4=7(}ScFvt%R zg0`RLo^bKdwa20)p5)hoKhI$XhE}#;${21v7Xm>qe1nKOy2elunDC2_u>J7H=Pao- z!CiLJVU4X-LU&0yFnj1?@D@=!Cy;KZ=V4!^FW%5g`um5Q#a>wDfD#?{rbAv%ve)(m z{sxcuMr}Qu9}a#`qfa5g8v^2l2pal6A5%R;2Gd<)qmZE}W6ynki?XNxH$~UKmcmUM zD(EXvTv9gJ9QclxbJz5zi{I|wvjameIy4GG(yv+Wxa>L6X>u~ii!a=qkByuc%}G$5 zo{)xPcOeuqLyA52Z7Ldyj+p)7A}$N{RaRY&Yu>=mO(IkZhz~N#%oT@7W#G>jhhxw4 z$&)}up<^xiq1(xf=K|lPMTX|V4wt4z0@V2WWTp*~pa{QFpirkARcg-`He-sS#j!P+ zM@W(Qz^8_D3{-M~S>sn-SoIqb||UH#K`tGu44vHA73YXkd! zIwDxm!wp}1wv3!`KQWoOh;>30QqnIS8t};mQ}gW3S>N^OLqBcM%>YpsVEdSX|Hm3( zL;}HgzVwEpV-641KN@zb9clJo!*`Q{`{5dc5IPN=a3KY-=8=1U-D!R-W4w_j9?zTm zG-~#WY1Vmk;L9$X7KMXC`M^GTYx{*OdPtBQF&~eK5yy67AO(6kcm8T_dhH03{5tDa zRH-2(+fU6qvl>@lZ`IqG0@EkKl+Az`I|9Yqio5EB%9bx~X(SD5msf<3upaFMT#>CDF*%)okxw>d?Gm z*W(-&c2E5l`8tzq3~``t?IWV_cWx7pDH_1g?jAP(N zEYRa51|Hyes6Z9OXs*2`XRTPDb7-$H-|2VHJaZM)9KX<1 z{BC6WC^7(~Q+%oAfB8~OSfj<5R(^j%{XVJYrA6geQEp8Ljr=m$Rnxg8-jZ`zW!w2N z3eCPrU8M~aBeNmo@TA?64r{`!32dk)+~#&_$s$9w-aXjylV|3RQSu!L@^K-G}>lnZ9y{UNR6<`Ca?>Gw}|KG4ud&QNeis`N+|dZXNouG48#B6lRIX zC1V(T9*wd5*GZIx{*YDxHsHIWGB~kaO9X6482RhSqtslcgHcJYn*RKdfZzrWydvr{ z>;e@6hvwmP*SkP?8BBw}P=I~5tH_Q3L9AW{U7c1iXz+s_w6+P*$ zGS0j;?=Q+$zR~$u9^RZD!%sb)_vwdSQ%ndZW$la0LZnaxL*{q5c!q%kvb{brbx@{K zSUTP?PehURI*3t0749>pBU%nkNFaL(0qFy55SI!PSwC4I*aa4FRItth|f%8tzDCg4;%aM@)zY zj5#*7+g?;hj5$IDA-Q`T{Y?P1k(Ind~^x4 zNv`VqOJ!va4Q!Zj4W7Z9n+YJ|-ncFwkL9KXGm~sNCn}}OjI*qhb>umRIOmAfR>22P znO3D;iLctAVaggpR0Go>lNDciR1fFgFTL zyP`j8b5TVOYgEl};`(-*^>%-(-nz`=v!DuOdB7HPRkk}=0)v8-y)pZIF0Cc&Hcekt z$Sjy~3~A~*)V&c3<5~IgQYHjtf9f@-X?fZwVyEOP|1_B3jz5jykkliQ0I%%JQU^fj zhqL!P$el_b@}=*2*2CV+m#Qkn6c*&XV`@%n>{4{ zp=LUe@0iP)^#Sp7mL0;1Jm|vOTX&E&z~YPns^IFIZMw39iRFx}8DOtd!2m&R<`?(( zjk=fG3rl$Zer_tVUBBDUeiA51+jqZMiUJ|00)h2Wcux`K$5;+#7O(vm`#rKshM#M9 zBwKvB8Ruz(;%Ua35(xFdO68{_X8i2d!un>ruJt%irAfF6tuD1%XH@Xtl}~MP_VP=M zOCemFC)~NNQ{n+F*n0X2w%h&5nPhJ%*yu%rMcM`@qJ`6Nfq;t=H+iOI=B#5#AL5ZZs8HJSHd5f@$LYtZ+|v#j@L zT6Ztd1JVGh`*oR?UUU03L_DVA>xx$lT8CH{fNTM4vw)VIH}8EP5!nOYAGWODq68mD zsG6?ynVK-4BrM>?BVVwYI1+kD)+e!{#(Yx z`Tu52Jna0u|35;Bb!TW7Ly>-14K|P>1`!c8jnacUS}O5`r5~yp)>-5tin?vH)!DZ5 zkkJu)RD8$PKc)``G=^(k=R~$3-rdC6`@xCzQC~poE;jsku(xT^1T41wUjHAt>&w^H zHL;!zA>n2FZ8}k{)z8CD5mJihJhe1Ed$H-jqMw_+Mwc=3_w#h+a;Xyq=IWw^#nyX9 z8H4Q;=OS*CB+Uemzc3x{wf`e`&8`r>{mtaY^SXZ6=2JG2YV-w1Fit7=H?1xc-E+C1 zg4QWZi*P)PD{%I`Z7*s}+kxfxM9x?mu6vPzp7H$FK$8MjQXD_nuA1giN%J&98SQ_7 zzIT8Bj_1w%NQ5V`Q0Rio{sgRl8I!`%vt9sii>Mnoq-mq`rRYCEA8J-sLvFV!v7@Wl zGQ-CW{c+pg&-6E!)XHxH^W0c?sdm}EXiC1b>8aRiCEeK9vEVG_JWqwHAq9@Q9Pq#~ zvY>;8_pJ|YtTp8EyPsJf9xW|M-fOzc)m zWHli(i$Z?>=EACK(jDloc#Z5hwcDvAn3gJ=rsN2586~B4Ju#05=LX^LAzIkcj9r(| zA%p$M^A~vJlCc;#drM$@+_By(yx#eN<@gD%s^eve;eNL05zk8nfm}23m<7|n%z7&Q zNLt&$pNs4LARwDVt`j(85$zAQB0UCl?47=wcle?(F=&$~iEhWWSjbinYP?ELd}2KD zn##rFJ5;3*a(6Dn>x&C?0dC!jU5%H@P+PB3Hom6#K(?Fp0Fe%*_ud$xnWZH{EKhIm2La!hKnlQEPv1P^f-P|N5mG!&wZBdqR$-G`3KFd&D+up&mTUl z7QGlTGZBj~L9bl8mSaMkRGr<*CdxQ9NxyG#b>-U*`)V~0d$k0=Ax^HMJ0I@Re>U&s z#L%Hx$Qag)t8pVLBcemRTQzU`s1U*SJ%MEi(jdXv8;(rFejNsQdP@A&{c%u({+hMb ze(}?uQ}Ct@b`FUX)*Z;*Mu0H(1BI5WwKK$phLCcJ2;&BNK9**krp+VR0a#bGEjIJ7 zU~9o0GL?)s`bL)=6+^uC80@L(JClBLbf|o>cH9ls+J|3_Jkcs;%=CHHL)Po=Zwe;` zw23WKWc@92Q~(h40ms-5r+Gx9w0KR!`(zm=2a?d%^pX-%^>q##it0;WB z{}>nHenK|Uzg>M3qI6~|k2`i2m>ze_uSMCT9 zf&U3YMrGblQruf6zvWWs$IWJB6glCacKWkdXD;aaj3pTz{xC21`g-gG!%s+{A|k9D zt*K6P6M|+b_q>HhP{g)x*GBPrgGCR7bQkIfEPHkHwr&y5Dw^$@4FWY_hf&HXs{60e zZ^3NT<#cM!^!bsA8(d{aWs-J+KI^S!oi_c30@+<-ob$WvP0_bX#$#Tm`iyH9G#0X< z!*h;#8Xwa4tBD$FlKoavwLmlo@m;D8LHjg?UcD;ajB__Rd-2=?S3dJMXXK#Z$CD2~ zo(i;M)E^obcQ{dbG)3z+03|=mX-wpR4K^?gLLK1p2(KS-0cBq#UaRmhWX8_ULzFNT zFgO+*zYCCsDW(ubN`U!JijM#9u5L?~NJEpSR!RV0AlejRrPwW?FQBA1JTC2-qS z(Ls~0D7>w%q-y`s^#u>*r;4~g*aVOexS<-(5+0J#&5}AXDI-wRZA~DQ zrW&p{K6e4J{$qsI(1I`rUNdncMx3l^+nzY#6&)Y02&(?7ZY_G%oQlfgYX~N$qx6{E z+JgxD@3d~4mmuvw+H(;5C4u7Qkc?{Zp-SLPa+5RTdm;|u zZa2-;G@%uQ@lI7g`$@XXdN}3mwY;aW>rsmYsq2TFJ^p#PT@!a|tKPub6mOSlh!@!N zu*FKbkP7lAb65P+AIoOIrF@JyS8<1DYvR5U+h{mpOp*P>xJHxuKazo9hCSy0!`L|m zi4wF4wr$(?ZQHhO+qP|Y-?nYrwr$%s=I+BrOvL^hGjH`=k&#(-;>&Yl5ea#{!@g8b zk3?u$hr(ney@0FIWQe9q)8|-hB01sW55lGewkoaz!c4*Jx|(5}PU)%jWvsD{VkYND ze`vql>fvj|fIB2K(b0HBN+^e1z02MAo1|Y%t@`KRH6ZY8XE& zDueyEGHlewxoztn%0Vs}(>v^dY>k2VU+*Bw`yXKt7WB7bHV%ft7dVP_>cC8J5r0qb zpPkO=OTpysL`)a*XsH@tgQn{6&DAkcsfgDg(_~;v7~$>sVRcdTAhrn-6-|dQCA0`S zuG!K3^PCo_v6rwwFg?K?wkqpo`?O><{iZr3vXAVA?9DSbuCX~UvUgn^7XRds=Fw{= z4*vKaC}=}L5K}i*apD1YJ-D7Z>pHUwdqnO~*e_6;pz6|1v}mxn?7~!>Xt>%a(707N zULgNH2PbCp-ywwc)I@HKrB(N8C?U)%L?;m1N_|ODgvbl6l!=LGiO1L%)*4HvjQH>t z)1;fcK?WT;y4FUkBR>@XIB?zm`^Q&QC;S}|Mb%Weoyq>+)lNmf;qz}2oVTv6O0m34 ztu6!*)PKKD!x}!h=pP9JSvG3~d-?~tXMHZBw*v2*jy`>K^iV#kHY>-|GW00yT9{m3 z_cDwy+}BXfmRC5Af#r=dd`#}K=rXwt=Mxn+qE=Zb+on8~V_xck&#PHet{7n`f5LrM zE~%!XHx__-kY3Sc)Iv_a&!sHEmcL- zaD@F1#WeW)YGoMcEOMo>Ihq5LC^+fy7Uu&{>5V;8FfwaYeV{O%YD&q)%xW39P7NcM zZWPU-`?9!t{~jrWguOOqnNuK)SY0Z58_K^4rZ7om;W0*-b^D*mviDQ24<1bIB-Mn1 zkn;?{8U!-k^vc!P#7d#M&VUjYVW}QQ?OQ(qBP6?L+7W5uZiI9m1 z<3o?CY*zgx!F$uI?cCgvD<(F=uPX1d?LXC!d@=Gc($Gew>9NlW=J&}1Uwb7z6?6vM z(+*qZz_LOHXLk@7pN@`}=Cd}g6FB~eZdJx@GrnYYXy!`MmjJ_%SiPNaZ3F8-Qdo%0 zu|B?fo00$v|t)_$qppxGK2-d5= zpDQ-dl66c%D~a>zKeJyaVIG|0N(p8x!s^GM7oB4ONm=e|#4E!t^0ijnFfD4p_cFwn zmOPvhAd|%!1q1_)?39@R#>Xkhv`ISHP_k3gZVeoCRHQ(Idx*}&ueN$Qj=z5nIP*;6)-|QUHCs4+9-v@36c@;xPzAHFKAz*~Q|K_2qJQ}-; z&3rix3<+uixT4WUUNDjTuu^O=j&Kax%ivZ1 zB&{O7y|c~{iZq(~e$r&TWb0H%z4n4pcRMz=x2|Xzc;A*t z8_wX?KRP`>U-tWz`!~V}kHXpaUcLun@7x60B5-bt8YPp6_t9Bm(cE_S)B4#(E#xNY zs!voY=i>rf?4f|V(f9Y(kasHUCHRH6P*B|fJ1<SCmmi#THkRE>!zZ5~VjlmXGR+ ziP|(##hj*Appoa`{Bt(Ml6{NEAE4sYv~{dc-0Zk-I4VHSels~aS=|r~3IeTd80&Jl zBnN%48A6h-uN!rqpgR|wLqnVpk?72A7BMpWg?MX{yGk*nN~kZmq~?EF*xsIKMQ1lad{KrO0DkMt;I}KASL+v<=OM*I8)SU*_#O9QirzH><~(OaP={KgvoWCwRYEb`%G`L2Jv$XZOiVCr-+<}mA*|qVY#V{Ty%;(P^VN}%#ZKk^oO3O%B z_oCXN-5-K?31`FWahewO|nBJsXk++*+jkbNFCd-|8O<|)1nKlrpn}se68IkpRqoJ zhR&gTx!t3L-(00xf5vx<=o*rCO820kA-?V=%KJ$`S!p17Oo0(An7nxv9#+tF)R^nhcH zTC`NrO9}>i>OuPqAAluAqKg0_$&53&Dep6W0{wt#BJ}ENbY1!pmCJ4HedCHYlE05; zG?Da8#OK+7U3i=+XiE5CQMp(o;;S+ou_>|o$uVNYA&$#p>Pu@vTE_q`Dm)Q`X!J{d zM^17?5|yX*2<|})|HI?dw*yD&z1bfj>jL~qZN5a^=jmjeM=mF_k=G4t-yqgMOOy4C ze-wwnT{KFwFe5I;g}&H_WPp{3)%>UP$OWE_>{P$EJ7A(m+I2Z+Hy&Ke6oPieyg_o7 zsdp2Mxixm#RJ49RZRE|464q1iCM?pEH(H`ILIJn@q;DclFzkN@zIJ z!v$;w7K@DY2hDxMI5%lR-zEfpP9vp@4xT#>d}&B4yrgU`*f4 z7@Xf&Ba(n&HBzi9vMGKxGA+PLGZcTkzcIM=uM!j;e|gMSY5pcz=uxJ`0PbG~{fenW zI4q+7ID!}hw`3cy*dt6h?Cd>Hh_V%pB}&|1&7O z){=4B7IocrN)2h4DB7V~rb85PF60Barij3sG+M_>U~(Yx0PF!zsM}2i~gE;Mw0Uhl8)0{8^TC6?HR^anEfji*(HDA`13ui*4p6|kt7;2auC?T z6fv$$Yc{HpSkgfMbwthE4S%{+&bvW5X!Mc#?Bf>fr{2d9d9+Z_a>p20&q>G9R1F%*_QN15~9a8ac$u0PnR za=4f*YDP>j_JZ5ctdokqviiccWp9mn$i8XF zU|MtHy<*gLCnaz@^|+4ikZ*5J5(89fMC7T!i;lAwU%8-$z_(V6#PLCsJ zrfw!lPc90hn>(Kzv%bHcpHcEV-cPWJjMAT+eiL(o5GF}xm(As~&o#|q-5(dwTZh|zCWv^o$OV*;g{F<_?OES+TSKYgzPeeRJ1j%mkq^gP3*Tn8$RkP-$Iu{kzSJ= zC1{&>99{3;n@eh&BpW=~>pYE&^&0pjwKKguKR1w`#|9b>7iL)|!5YJB8oPIbNoI#p zU~p~atWF&iRmj`-iw)KE1Q>{JL+FqHDnb1y$J$6>{cT6JPuOUFd}hXi_I8>0=QJ6i zh_@}UI`Qb3fZ7n~AoP2n{N$K1t;SW4kw zcD3#aP?V$`K5w>QWqauWH6RB6>X))#mxWP&zQjG-F3%l5eH8S3D~j&|)iwI&y*K3p z##|Xz7fp!$*wM}H>W!e$DtG7Q;s6Ea=mm--fbeh!tP<6UK2=@pTwH^kN6g_s&IbUH z7GVEsU;hTh90iPQ!d#i}s85>#=Kt-09$fHI4I4~jh@47;Kqrs&*QGoR&VNfHT87!f zX)tvuK2?qFPUV~{@GLBZa7>}}iGq}lsRZ+mIG9-W0~n>BL^Cy<_=csJM0hjCY!N#Zkk{e~I%1z8FKWWIc}e6bd_acZ0z0$_3N|AB$3PAfB>L z0ZgH7{l$BCmY@?!Q$}>7NelrOc35@Xfqc8%eGZOqAVanolA+o~E7sA!lSY>^45Ypc z4#}rQ%Yv@vge&HvCz^(g%9w_Q5P$;n-m=q?%8Yfrup{yC(G+N)%B?+T+% za0Z$zIqId(&n)5J$_JO$G?1N56VXmwJ zO3#0_%|X8Ht6pU*N(xHVty}X1@X%M2)4koaKJ8cxjFV2vWU@+$}bwV)G5L%ZT>6030P`0 z8D@`lB_@ER`mpW`*oDHk@Y|A%Ip>c1%^dEKvF;JlbBf2#j1qObh+?@!`H;<>&-~KY z`V%W%u)jAloceS|s*c+JuUKmb$`Y0zV4VNn--c`+(51%aCl^DCFe&?Tjda9b9#j2$ zJR?4$bC}9tA*U19#=tK#3Hg~$;?(`TGZQ!K`{EsRWqVnSSapVTN^z>A8!6r_cAaSW9{C;%D$=fh=b-Ba+ zDYNatC&=l})HORbh3@~Gx=H4mov6599=cQX%gf}8Wp{e8dy@5Rbbuj>t<*RLE?&aa zGQhSuNx*<$tMwDn=3wZ7Y$lnbTYdNm(QC^^c;8F??alAobVo$8+Jq$TdkIS55LiQ) z>uhiwUgIG^U&Fs@JkP({VrHd+Aa(}Xazb8o0p5Z)6n<(W#mM!HRdjm&wu8?qyuuL> z7io-(XPS^nSmLfZY3Xh!JaZ-NpbIX`T*rcq|5Xk4@Riji!wj9z$s|d-kZ!R9ruABW z+}*@u?{s^WPy{V$uRkM4Vv%i@m*UO6!cT`ehu4OCCUVTveS}|(etc9-?SG2`-w+iz zh>Cz0!1#r&u)+88%<{mg7wr^|Ty`xkQvdue{>s2l5uxzo;*BoN@^-|GDsWNJMeeTz zG>J&)8Qf2id^gXt=0!B6Fp9sDdYN$s`!(pQ>#4WKVEm zay-95kKnu+fY@?KZ9p|HN`EBDCy!qjkuvgf5)0a%w+W&D8{tPpxKC6(l~vZJ*|)BI z{s**Y>F;U0QrL>`b!8TS=v_dqqh%LrWG?;8fv4qGvH@l}9F=!ZQwz3-5(mMU${sjU zwm5i}YaFow#ssu7#pplR{tTDxc*{Ydg9$q()+Fm-)U&>$%>(mDvK>z(<)QdICBD)Z zO5Oki3)RP7z-dL^<4uCySc$S-O{1PG@%%~yljeBvq=YR++1u2fetS*E#fV=bj#Dyo z?asV^b!owKODFG%r*_65K`-U1%C9mJ+~nV0+O+ZC{BmEX!}T>uL_GQAWF2czFlbSN z;Cq9s z?!g$(#hZ@?qiyBM{rO|=K!9~ARPGchI~S;<$ShOTO{MRKKW?j=ri^xB05T6>`BZqy z4sh3NYWHxjs7;H!ENMy9BrzO1GmEs!d*4OjUx7DDDHZaq4g8Vr#{mgMyoGK~3|FS= zlUPr7Fb+rBX^RGkdiNy&Z!NWvBYodEkUT47L3yj{NW{<8P`SzEOFUVs& zlh%pEl{h|h@tp+`hG4DZJW^r=&S1JMl$|U9BZJUYKA2G>XDhLY@!}c8aGp8;?g(|W z!+Yl@$(C|PBY9CYhhjG(_{+cS3i$h|rGE9D;^$nPBd+XYx}1B3tsQT($Y@E1dixKw zR+A@42v}DP46FQSQdoaF{`f*9|iI4&-B~OgtM^mjM+@5n;=Zq^5WZRaV7?O0pO;lt78t zIJ$0xg%LO2Vo3)K@ptZCtFko4PqkCa)ZFG`)enA}AX7$TQOqO@1=NdA1z^)%wpeRS z%e*m^5N3ab6UR+pnR&?pC9(#);%y5ofV}s-=zq67t8MxqLk!#l)rp%m%qdYdwB;%h z)OgA&cpRY&jWZXEBcUZBGD@;UVGz9}G@a0klaPB#4s~l~lWA^0)3Q9&(0NzHp4Yj( z72;(40hmc$9v>8iSEb%XE$!>U*#C;%TZcLz=^zpy%q6k!!Fd2x12;!I(EHT%ZC_WO z=BdkEi2po$w^oGF{gG`wA*L4RVkK5m+ALJx6_*7AQq%LEe1!1uc4fc{8(ltk2 zCLK`SgXH>6z3?xQ&14HKZZEIbsF=yc8{Xk{*sypJ!~AbO)X(rNzt z-9djMEHpw-hLeIbVyq=wQS_FydsqMju3i8`2Y91PUJuN)5oC2S-Jt!TfIt432?(J$ zS#!6LOTF#{N#OkStT!-faai+V|8?*%`>W!U7W=K;fBRv4r=oS8s&zL~p$l@bH^uur zm56zZS34;povF{De(@r^EVYN2Dmj!G2^=yz?w#Ip=P^9oMZm;fbKwXY{3kecho%J!v1 z!6-DI{Rh?lMJXC8tsj==^`>=l{%XBqtF_*VMmE#YNk^6rJCW(!v)EoBbAARi?(OC3 z$wO#0W80)@`^rlB?fGy^5i)M=sX{Q2I)Kq3NGE{R?so7i1Cv4h`EGW8b;fQHCjOU6F`-s(v zoUciKG(YyiNwcg-mXEY?<*c8`T@*hW67XwD!Uz3wbo5$xx^Rf9wP=Jomu4B~{du|N z@-4^Y={`W(gLL4=xzW0Uu`7O|96l5h$QuQrC)o! zxH~oRXMZpsXjjYQ^scoGNQdTPB^N6Hwd_jeXH6}>@V`zxtyvwI3tdgQPW30y3x+bBc=7jN6+{+dl7WN^i^ zp#Y&MgI~6#zpOo*o&A)ZqOgujCkMfxh{&cBW6#=NHkuM)NdkvLfQl} z_{iOv0o*@`((1-?@IHT{_T@)g;DoK~wQGKiTk^uMF(wQE=dOPcI9`W3^S5h54L@|d zq#`QYoc^7Jm_SK<$zI%#?F!pk&1Kzn`UP?s0+Mo=QYRE_FFI4EoeUPX;<2osi1+MD zjcuF2WF-BgDLP>#m#3SiqxtPMvIHB~zWus5h{TUkVoMN6V&aN+Gj^6goAd`BA3{1O zA=i;U=<4V54Rw?Ub#8+2uS1v@h-PL_w8KjrO89o(&!7PusBFxJ1;ZD|0Ph~=wW4bM z46&{V1E@Kj5M@g6a^xb7rhY1M2`g$+jh-JTf*CSx9j`k;QpI*5u!FJuxH*Y0xCOS# z=%;+dDW4Za3P-k>5#0y9KxEqqgLrO0=%WibxnxjTwMXg2e|LVaUYIXA)%|kA#xdlRZqKkE5d_DD({-GR19D$3BO7A!v3Bdx?>C z6FXAm{zFlDpJd-o#kWh{9+_x?z&i)?e!sYQhe83$?+hifXE1j!-v_?`pP~H!MDdva z8^!zom1Y^3m>K_PtjCk+C%Mc36pE$`J(Js$)teW_}WvQbcB%kq=rz5lR$| zKd`noasqaR{t37{yVt2VFa$W59~fayWxh&Cjs%&Iq46SSX_#%Uwt_nA#(ToDba~^E zcT<7Sr#0Z%NRcvMnUVuvcTW!gxWWbKW?6{PgQPQ$B>Z5`=v4ksMuHY8^4 z1E?Y+GC^5(=0OxlB85_-DHo|+l7N1KP7&hdcmlPH5RpOxg+7%_ht%^2aT408K`OX> zeFCVSL<{qu4Qa8MBgC*+l%2etzdX5{$1X>+T6Q9ZBC+V<%ftyexxI)@i#e$>EXXflc^gRoAudS3~cBs;l02J4c7#$YcjCw#7-4wzBxY$^th%{!{@Cpn9m`{zVALCA~>wr&0 zVJ;0IwXKIg@B58@+AhsP;4YH&dbH!zRdp-#Fw?JdfY?X=NL1*h*i)CTRa5ma8?u?; zu&qnvtU%8n8LyJma2xv^mpmuv&VuCb4#vo6!%#{}uGKR`OA8s>@{-q3$*s+P^Xq42 z^;*~7Njm*~5)geopF&|iZs4LFX$K#xY1*?|u=u*5mRApKxUx~(rV|UO$&jo;Bei1M zRH5Gwl(MK#oJWVVg*e@)BQH~3)r1-54=s>Y7WWD7I=f(O()UjA$wQO@I`)pZd zPVk*^jM>d%0UuGFtj!=|sC$@RI|&AcUK!)Ui;Q_Ij?4NW9w8%T)d0vQ^^4WsPi*KS zyYKI9lQ66@B@Igy^p#dsnEd-TTUNw2t+EC0)xFMnWjT+^(%x_4BkJ~^(C_qI$^q`K z{;xDW-uCnKWLq7anb0`;t!(dLbX)BaoO3R)aE-?24Eec`9`bmMGj2y4GSUic^4qJk z%&P%|MQYeIU;$MrwiR&7ZUYQ;z@Fvs+jlRWaQrqDEx27Os+Gf*$3|Sd-~CimgN-YG zIBM=9n@M4&lB3IREb!)TF@qS4Z^9;js->w1AtBXkzsk%tc+d1prCi&@jn-jIp(^Pl>z}>*-3C`>)iaR zEjk_HNa30vW(!5a=86C!q3?6J*J99QsUHOGLCM=9bcP?B`78BcHTw%4YE)q>82PTy z>ki|P$WZ()nToYRO)AjtLQt9x%NxL%e}64P$pobDT1PE@!ZJ0-ZUfEOuPT3kCI)XC z9Zp0ux1k(V%gW0ergSl_8rG}{`L%TZj-ETeX};r?y4iJOoA79&UaoNV0mt;w-+Wu| z3?A3J$d16uPFOu*NSkJk48e5ER+eoDw;n=OVby#{ zma&)*Hat@QRo|jlTn*~CX62%;JU5WZ?ctywYJeo3A^Y1@V3v&0SxpMdN>~YHWruAW zR#Z;8DDR9mZy1@AB`ICWNH{dlM}e>)BK-03i+ey4H_m(AtlGUos$2G*Q?-z*hx-;p z_mS7s(V~6A`!sY1mj{BeHQMQpIExk)!a~tL`DeH|l^Ch?^$Q++F*U#~sHs0>cIz?Z z)w{PD_QtVX9nSq4z8}li`k<7mxl1O@2j8hD8*Ax;i5`1TYz={`&dQX4>GTwK=Pb)A zza9E7K#O!*NV;cb*WsvW>8dZB%;hRz^LH2=PHQ?mItWZZTla)r+bamH%xbPtzcwI~ zKlT^h?BmCQATNCFhpx52upKV6SKHn~Fm*3${sV*kBQYS`{r3qOO5Pf{U0zeu{!507 zlxNg<^>asGuC*JUyKKid-l@6Dm1iq$BsuCm3gIsRiT|$3e_Cj)|81fD-?ltE2gm<> z;&*GxIBl^fc3<8s%1SR;05|U!gwp}dnFysUw{WNY%Y=V3Y&+#i#z*p~7_*EgZvKQl zO0mAw`RUVjrbFHpP9&EW^3?A`REh@@#Jj&gm;m)oEdJC0(-<1L#%!MpOBuXd z6DBbq%PuB8-atz0JO7Hke0viX8{dAr_T<59sIrJ?E#vQN7Gti3WZPaE{)_k|YmF?H z4u*(_6gXt-;eWr23xYSo>-VFJ(I1dth+L95HBKkUHcM4q-B!ib;59RWR}PRcv9vTM zMT7v(L`(07332_7`7SN|xb)1;nPr$>ai`NY!1(=fO-Xgy*H7E1fAK@RJsD{DLgpQH zz&g?Tszy3xOsPdq`>TMJ$>bMBW}ay7U^`McE=&}1vvR7#vkJsVqt3Ewu2sG6Si|9L z#qnql{~MLsD&n!brv@66*8yE9!C`hiF$zA@fZp?DQ;(B#a%}UOoXw|vsiUQA5?cvE zv!>C!ezSpdJ0Fj(y#bV_6mI;plv_TM{^*cO*7!YW98aMSwxH-5st zbb7(iO65EgGV5FodwPHMt7n#s9S)|JyN2!M{g#QexJPjsFQw+5^AZycg#$bKSe1EC z{q#3xBeluEs-Vi`02eZ|vOd}cXCuI^@}+sS#e)p0`Pc9L8i$ckR701NtQc6Zu7J3iXCL+E9a9JkG0rmR?$7I#i@g#5~L85V5hZDcU|l7(7wk>}JF&^D^^YU}rT z{<-%Xg7=`p!;_=?AC}J!Wg4 ze?&l9t_3juj5t0CMm+jhB5Py6M@UlkV}|XE0u&) zaRN?g-#02=p|c+o7GJU;V5QV@13Dq07elpYU#d~xNU49H1^!Cm>_^&J0r$Uq4AjMM zn67H4*uMs{1AX?Y%YIQZfC+9YxSRdx_<698m4@vv+cNg}dTIzueOHA38W`EhpHaex zsws;k8=Sv~NOTD-%pI$iv;lRnAfGp^1jQM2@y%l*QkHv7fvJ@)tEj;{6`$AoIIW$< zajUYpzF-J0x0$?icvWrkNbJ>0ZW8itUR7pEQ*BkRrBRQMPG>m!{#LF$Y;X}!Rr6fF zx~%U=y{jVwIHUOC#z5z+Z2j*&gSDTT?+6b$|qmlYO~6539EEOK7Buh&}&T|sxM=-S@$aSpf7OunXjHk^^q4DqV>Pu?KRaww-RUoev(}SAcC$`zR5~-a(oI z;hewC)!$%+;SRzP?DTWTV1m_Q2#vysz%lEPFbLz=zJ}}sz=4VYy$+vIz8zMT=q^1v zsJnEu`ugkub1vux(KtJ!xw7-8gPw8|xNt|y#iU7dGl|lWtI!3qDlCXHd6YlRz$pfH zt%NWJ@zsmSGEHM-{1ymFu}*0nJPL_? z3JWiBkkQ$BKmtq96%0_r5Fxu$Hu|$)C4I|eh#^lf)(U8AwL!)8f!Qf0v9m1GADcy= z)tmvKPL+~xVitHq&iZ08WI&`zN_?1*zy-XX%k`)4S$D8kck)|sJUJ>Z=cODt-8!&U zz(nMBk(q`%O~K+zoO$~wC4BTTeT|Zd0~Q;eVFg$W<;ev_dnIF;?tc?g;8F$^aOGcY z?&*4765Io1BnZmQX@3#CAyKu*%*i6a42yn>(-c(?>`~$`f4s0ScMk+jGYfaRPTv(MqZs64}&;1m$-B8Atwa`m;Sjj!;I-p7Kr*W8qMUVeo#v0i8j#u zgzmmgZK*|N2P<^8JtJ`ldI&15)Yc{xqcIuV8Ykaqr^=;35^zkUK^*k!|1qjEXow9f zzM?7{X?$EHt`vL0R5;)RvXPF#EmuWJLra1YV_|i(&958)j7PhgBjlXo`Y-+&lmg&4 zB$;r=KT?%4Q6w-PC)n{9JfinPUS!_}a5_POw?swASJUu5fPq7qU!@Fqt!e1(>5hyH zD>qDdY9H$Ut^Be}jdp@;(jTZCm?`NX?k8uNKK2ewDB9|h?Z669`{1hd8KaTh@0?hh61fPE#M|NESRc{d$X7RAOuwJA3_zaJe`PeXtak0i~R-#YB7pG-FD^%W~WZ#f=NnW zI;Uui)paUP#h0sgFMz+f+)}*^!0)cVwfx+2uR%YStGDVJ(s=Pe_M=eh>&=Ee{YHu& zGR71y%EI!X=vcPte%546tS__DP0&80IBN5XbV+dKzdS?CDof?d^8=acym=z|fiZ>| znE6*YV-{awdTD(Olfl2|f{v)8lJc+!)VTLKw+~&2wSgl-EX_`TBxwKhkT$=F4Y?%P86trnJyD+{G00x-g2j=Rffp2n1tTZlnwRC*-DI+$ zjde=jV2U&Daw}Q_2Ml?I^;E3DgFuF05)Osu**c5oY;cZpZV2TuCL>yzcbPjHfAst; z?6C@Zi#6H|p=?2Pvg1d(qDDY9A?Y9RSD-Zs918ls%kOE|TR~8bVd>BW6d)v^Ay^6t zi7qOz@b<2YxB^=&z_1|f?VQ8t?&Hb5Y+k~kqn{?o%gbPsxXU)7w1$d;HkE&{N|I8N z6Of6jP*bG=QqwSK9I>UNZmVnYa40cQw8XMy{Y)vA$JCMfRBRg z3+vkniC_%?8{OC4yoS=x*1A#4n$H{6OW>M}Rs|oUMgkOEAcT1+=@yA0SrfLzti?%_FIo6@eKw;6aKXT0e_J|^0AR>6-oPlvF*VXNv$q+&#< z`B;RIkyoQz=_!&4t;=^*slqSgw}|G4Xvuj(2|>mCSL;n~LVT8i@U{hjd#^$-U zzv-JpAROIeRZi9?*kbf=Vz`O8eq*051N@+~blS0aRGv>6T=V!g?Cj*gWPiVef_FX` zo&r&alrxV$l+gXv(%$l7b5(eJk4nkPQR+0 zv4Y^7Mqz|X2%!m(;{N4_C#dVYD6ATw0#=0(zf%;Nq(-|g!<~9rXnnUt5fj9g;ws=D zJ9G{nrJAzig<=UY1JxkHt9j3=(n~bskA#=>LuPg0e-UKv!Y3uB3uV36hZkiCazWo% zn75SkD8ae(&?))EH@WAh3ty7|f;YhietBOL3D#e@v0s!D*A+A&AC#HcQ8SRAZ{~l2 z60!URF!!#FMwSI3qAf#_Ga@Z(235@;z^x=&34bNj+Cf~f7I)LVr3VRCyH6fC>6QE~ zx3VT6TlztZ5*$eQ^t&HIxT%TqMU-vbbg93R8vfoPoDH8JBXCv?!Ouzy>deheL#8lT z4Ej2SWb|Ptpirj33Uk8vWQUH4_NlUzZg${oU5!akQgao8>prNb?Ju{XSnqkdY#R

$-&F3-u#zFJhAt! zqwC47evna{1w!u(?D|$>e`N6FTm7#Rv?ZFj?NqzW4X4JW;iQhh7$6O>GtW#j z$x5SP+`1`t0jdsKudf2g3?;($9Q26WedU)nycIqsoKF#-V)kJ#kQWarJaRY^Gxqh& z9{c1V?f3I*=!Am5i}USS^uB8JYgf+`wiqLFN2^`-A3QrnmyBV7}BDuFWEQTKJk{qIJ-Xm9U7d-ik^iGp3l=6V)i za;2-*tBqUSC$sBZ6leGK1xs9%VfG)~+$%e2YsHkyuZy$MLJ&IkA6P7J2apJ2g~;ZF zKe23LtSutRRrxB^3&ln}=n~07eTUQ$4hr!(yA`b~WPaJxG&`81vN}^I zSb-LqW#Ovv`(z4cCcCYdp7RrOA_BySx6EUuBQ@bs#T)Wm!Di>;`j$F+jLf@r_qRx7{JIX zeB@UA(;tQr@$g35FDg2* zS~BlovmY0D)I|b)zzIQExENFQ)=VbxpfJ%~xcJENNoD$>GhBrf;w}H^HqmC5kQoK* z(W0|(x!B)b7x7!b^xsHvqq62VjrL8oJ< z>7Z|PYA3+Znu*c6xxPGjBXL<{t!rm8xFZ-J!QdIo75brbFR+`;65Rnf4lafGi(zSr zqC#MQinV#I_qiUKi3&A@=F_r?SgJ_CrJqXVqXvz0N>6}q<%LT%0J04L{F$M4$|MMV zK>%Tp8xm~7>8jWzj-P6rIku!hLKpYI5pJM`4oK~^*LI#ZuOu*~=3R;thzYPdd7&mihUo#~3Sh6Gf_4eLsn%n&;70SQ5sDZ=^%P$v~{e}D$D=1z=F9m zp{Cb}gTvcNuQ#PrW4BRrqn7z?wbPpISK}j_+8^=J>MCkXoItRUJeZ=HlyDWZU`TsbgdM`##s7gtMoMBIi%T*c|oDuigc%R2QJ8J-t>8+55kAS8#WhrBL$ z=xERQ;^J%83{e>HmWj5-LY7!0kK8g`!d7ZxOvcmtp3ej_a6HneXp4G<*>gsC7A>si zN5|zCyDQ28Ft%N|0L=v$=hZ2|9WDWB$>+ZqJEtzug2fDuZR?C}+d5<0wr$%!W81cE z+qUh=B(J$^C6D_b?5_U0ir3{MDy_VgN2F;P$Ik9sm^|!p+fbm;D_hTh0pqT}Foq3~Nwg$@;XNUp*Hrol?OF2U~b10RbE5R-Xb~ zgTM9`oZb>5#XUgpg@7=JnsGemQ5-8W?jdKBv6kBX*9f(Q8%Oz#T6ey~GDZqBO9J~{ z$xT%YM|;F8a({_T+!Ai+lr88W`N)O`Tp#5E?q}tAQQ&+AjxPloH}D6rUJD!lek5Ul zN>(A~CpaV)sh|0)5Q`8S=%u6FGF6vzD!IHhzF5wsk6^Wg@eU%3NmPgXutZQIuXgn~ zwAlEvQANddt@>uQIVA+hM2AW2*bc9rs%QT9hEq=^Et>kH^^5sjGAoMmA2(5fDL(F~ zkb6H~2X)8%j(P2*hc{b;OlS3GEYfy=;G+3L)cuIrtR!U&|5%=c@PYLj&AEe5yUyp= z0k2|;Hu;!vE5o~zWch-C9B=S5NPy;p8%4SgSccO+!z5`NT5MY7RfrL{wS_nfgdUh&*t!;_aKYlk9SKdb%kXl4oT9?plC(YkZ1ph`< zeVw;Sc;81klW(WQQ&ZVuK{q=p8NdB`NV^iozZ%IV(}?AV_5m#YAh0(G*l8{m63}9* zLH)+k^PUQ44t1X1JZcQr+85Lf3j5AnJl7+Ia>&3A1*sIVc&GNQWwxne3y&Xchws+M zT)Dh_I@(r(K7n5%eL^`v`g3{TTJJ67USv7UuchSFqF_BZ;yTxP|LjQf2>9n3EU1u% z$lFz(PRGppG5UoFfjzQV)!=91h6j%Vnlb=kRgIUra)R8V75dV-)l7RW`a>lYsFKI; z5MI52hp~D&qM1VAb^S~x;Zk0tDV#kgk??Umlg@-ZP6xOVqK}G0Ts@B2SHehMa$k>= zW+pND+hpI$i0~iMJn;Tvd-T7{bBX7bF;PFj8H=p^a~5g??t{wB5dUK!pDRrQ#cqMN zLH2<`ifb$e2RZyuUGw8C?7mrHeN(8rOusKMp|4itD6TNeTrVnR! zqROJJ854$n^qj(BCguHDZt1Qf!#_EfjS9IiXR;`Q5SQ12c8XmQ#k&6u z6fy{(wzMSVa_x>0Cd+k#H#y3Ul4#sQcpy!xr{EZa+hYXo?Q&XsI;yZw?&5i`EqJW$xvr~l0fT@= zLlosjfYX7wELzdA0C%@}8!&gzQ=eZ{0k7Llqts*0y-ZT#x^d4zR+&uBt;|lzyPBVBW=8dBDxbYH_$S5%8 z9 z-TPu$Z?;?%y^#P_Z&>2XijY+a{tN6Rc+i^y)WxZ$U@J+jeniQ9(`{|DidO-mAE3^3 zo`0>CObns^%<|5(b;zxxd_{jcx7uq?Oi8#auZKnyuInZqlk%gtIO>SH$>E<8s{nx+ z5lr7o)#mI473^$#N!#9wElLUdi&mpPs4-8XZ^pHM?0?L1nm#4rQ@)H_G~+m~ytmYY<;$=vfP4CZd1U0T2TvZ!$X|l4 z9M{t~{s|V?m6!mfp1G`|j4M1)OE!?BL`+wNH^6tOHEr?sxVfXUhD??=;7Y*Sdd@#*H6DPXL{QTaa9a{Lv(Q-I_LlqGry_fe)A5K1R$mVB3L#CB*Q*hA7i_ih zhHZYJ+iUBxVOQgmel*DjD8wecB)#%25_QSrlsJ78;>2Tf-lp|V(rD-R_|+NgkveIj z@ufh+&ga%GouSPEek-8)r%k}JAL*tum1deN;y~B6sRWw87v9l`*De3XsUqXgBqL@+ zqa$((y*T>lZ?rwwr|sgp+;blJBq6A4@;pG2XuqCVfKV%?vuf2*eY_CF&~{KAdd8SX zCwa>*NdIUjH59Ik_Ha~T4M40Ankj^w20jtR@Th=eno4C3a z?`HK=)hnxxwFhz~-M_9aEh;^L+7cIMEZ!T6%qv5Z+bcADH6sE!dqMCp)=KSBopj3n+<|r4OTcAhR#wMKZzDO`>M<*UmExbW$?;O)#Q)^t8MIWSdS!QGQ z9{ze??V1n#@Iu;=P@51@^&UQG0Xrd4ut??p8^yH&yk?mGvap2`uEAl^v214rk8zTu zZu5+kRGzFrL1MwW(jD3UWvAbF;5Kcxi!fORoR_Mj>ioQ#BP(s36*s34sT74RcWQLX z{F{*vyRG55yckE?TTQUO3nviQjITI67RH&EsEpS{KV^V=NQ%=?7Ou8^dg5-#rVbLq zF}ZQK3KQqG%JFw07Ldf;YfP{_Au+F=*gudT-;Akj`y}tLT)IE zWc6C(ywUu}Em!67q*PT`-1o4m2t{-i4jt9k94}@@Z}jG?6bEc*=cat2NJAp8NcSU0 z<(K0g%DfI3wBcH;#P_!4+GHb35Yk6+g_n4w`#+KOz*RIW&U}k>FL}{Q!fD=md!YQd z)$kYbxxw^ekaBh~>#t|OOrZS<{tNrk6VtaE80(nf$gHyu%s%js z1PO4>`?t%-(8CFv-`bXQqZ!+bB}ab{#?`s8_Vk$JSeZkUrY_$NWs=S=?)Tu~4e61U zU;^`keD#8mxZHRzjcvJ|3%7jPmhb*Z!4K%HtJL#PD;WnV`9w>q{eb`$Dy5L-T!ZZs@Q8;V5xGKrzoGJfD@8@^)+h60yXL!v{f- z*_rcl#gg*%o``?IFo5RPu{jtCyN zLMKdn!R(F0rV1AW6HkC#Uv}^3k7h1!wx>pay@_Wsnb6*~b?|Ec<@qZWpph=DA2T+- zCMz@PGu?FRsW$j(XK!Ioo1|DGVmCn(cyoyB>H!zL^wj5=I8u;& z$k~`UrjOPYS2Rm|g3nO0;iJ1WnpgZruN|&2W^&-+3DF~#ogLh72mFHo;d%lP3f4xp}89wL-BBrKB+#Gu|I0ov<6b$ zZB{GRxc}O#)2J~8=43n4s!P}KOkS!pbcSf#yTEoJg6XI~8?t`#>Y!>a{n<8-Xf z##FNC2S%2wS>N0`%0(R%cl#f&4%<(RdmsuE;WXwX^vtFyv5uiwtJ!nbPc_FA2-+O>_B*WFg@%062ke3ENX>KH{0-6UU^Q4DKh%9E;H@AXL@pb&^l za1P*AsDuA6AG;gnuH;|qiw^39$WrtyzklrzkM$=blO7M6Ou1RbJ`}9ZNs2Z0C-&`K zv6@na1+-b>w)t;(#DaI;R}E_|UO7FOpp;4)QmhUTKM4y^uZKtKB^FS1xsMQ?0GlZmj?S=>Th)Njj3 zY6FZR8j=ethElMK@c*F|SI`+X)lpj{p#5y!+qkzxF~G5u4da`^@pl5jhrc|#n&H`p zl9}56H>NzS)YvQB1EJiYJ$sC{A98?$>lc?kO?Fvm#+e}^KO2r+&2**$1&Jb@u zbyEkyUv_;=dm6k_nCUsZjf_zPgw%^0mDlytp=DJ;*Oxz+Bb^8#$67+j0N_aS-f5c0 z&zN-kl|55`%8-TS>FdFz4^0ZN2rMg0w0e3rS3p4@W#tg8Wn`nB3LNrK6nbwWqSe$35yX4K^R?*jJ1uV6%ja$<{<1`BPIelu^0B^#zfkQH zOIdmrW{m<+yTT35mgZ&X23*@r%M0FtZ&A8dY>jZ759R0Hm}n&#hxJN^f~v~umQ8f4 zwzSJ?C(e&Ro1Q`Qo#3R@=VQ-;j0`kE-{wrzRe%hZ-v1yg<=0GM>jCRt%t5XLAV zHj)6;3=ZFV=xN8UAjGRM@X}?H+y@Ih!ajXllLv}JH&IWuBJzbJtArHm5pD_rD?Yq~ z?(Dwc-7T~zV;$!)JEn)7dT97-=g2*{4LH4XXMOC+B*+uwp%gQBv6V2}s zpwu-A!uFxSQx#8~I7fIgH$SUpDy>{#EryMMU1kQM!NDoVXobZe)hWpH@@8;jdW4nzMSgb~ksn`5@aM4&rAB5`}?X2y1~N9)p%5mYwv8nNWHQwl(9v zu#M*I0HUADljfy+Y15m1eyHUadsVHX4Bo!^#5_{KH<#FHVl--CG>R7i>h6c*_FUhd z9ghw~3Ix_^;WFePT5%?=7`H^x=o_jR#f(#6u?{Ns;fUeKt*qP#aJmQxmIIemIazqR z_H;Dqc@I0ZR4qH;WIRV3zyv_oBz5D@f#{f0x%HJa1LU9jnj;KvGPO#P?jgPaN zIZhwWpPxEOz>vt`NyY?I22KVD5}@ujNkCn~1dNLiFDr*eYqY=J7YbpW4Vm%MnI2dJ zV1kiFTPee%H?6|kpDBr@(Z{j+*AcAlJU4)<3MWRkiPK5_V)F)$0oJqd#?`a(^X-;!Lw|7 z_VHH(QG%X$9}CyIbp{LKC9kObxlXQYW4rabMMscy*>67>VDQ_%LlTFhGg-K84-gex zN_svUzt=3%*{)>~juoIQzTQ^^kQ+6aezA*J=I|qVkxB1#MVQLCQo)~EwBmq$ zIfxy>ro%KLw|f(#CpbAht?b_3_Xkq=u}vvm*zK_c?s8Y#Pr@eH%bhb;mO&Z<(+8Rn z+SFndYfCGIl79f%XyO5I2Jf*1duedwF40uL`oi_aX-W?0R(VbgOGu@5$L7i)ryv`P zJd5e~$PjLEj%W>3K3_lGJ))NaF>dAF05I!{4Wd-uYrQ*R2A+i70ho>zHiUu$6bX~S zt255A<4`Kqd*arq{@qfq==HQ6^~W6h=HD}^$IsdsNR-|s^=bF2ze_=X&F+*VVm)JN z3G$mb$VQ;(&#WR~esL4wMRAx5V^{^(=%Pzo{fcKN>dlUOlvBrxlI}GnDyz%%EE2TLOkGGcbMV_!MCzSY5T^ z6yZR@y#%pa$=9o22k*^oh5$t$b!hyFXdb@f8?%I~oIl*-wtsz>vGlL(XOh+sk#~SZ zaUrL)PlyEXTA;0hmq2G$;Mg|r?%Y`-H+0rMu`nh{X{(=K;{(n2K_S@-WS%{^^EA(3@(w+#& z)u=B_P$P+JBQ_`xlbqI{LL^l^kn{B2mTGRGG$9#SNVlh1#HsDCIXpv~{3oo7#A#M+}WMdzoGfHDO@A@&5}?qdd8kOfPE*?sz;<-o778g)R{cQ_2;~Egzu_HtS}+FDljR;Y(A(BDm5x+Y!&$maxM;tG+GKBqM@*51C6qu`eP#J|*PLMC9x#r0l2TO-#Z`@?ox2 z{z(hnfdBc^=$VZ>$Uek3{izGD?ThRB2llOW;0R4EoXN9Ea6C!K%v;+^l%>;JSGA&4 zYM7;VxogpabA!-Cj|PXMn+2tT4j|@2ecW&xb_%^h`!5j`-2Gb2gsV`Njo7VWd#N+5;GlWR{Fi}9faeL;az{nu;CR)@@HlKL01Mt+A>`g5g|0%6pHd4J0!9iuzcN+4~+aYnH?3Vu`HEdKlw zC5j6_9JSR@7v65PM-y}*GpPb_P1QS~{)uw99l2sYQ23BLBY?#KSO{zNdQZ$)nv$S1ziR$7S#Iy68Ec1w{xuL8~sx+D%FrkUVoy7uD^w z8%nJ9K5K~%hvmuszuo#1cb9T0@I%o2he4?Onwy`Sa=(^lv!+P-Y!sA z6GhMRo1JE(&EKnX{!OW6oOMl?dt)V5=mr64Lz0R_o7fWFbh?MwyG~ zf2<=RzA<`^x#k#?Q1n!Rn2k^TL!Sicc-yJ%9H~VAT2)!eZ@L2zkw^oud%*57WH7m} zXXW>TBY4l%(Fgkg8~+?{Fg%$MNC2$+myiS65q&_`%;m{ZOx3EM%p3$R8QrFO?*{fM zRx4lRMH-Nu+z2v7KBbNI@_`2oj=-Sc<(bvisdR*R?1^7{*%*6Nf&OJdW_Aj9Yv+ z!~DY+Os=;r8|9j9O3>!utB4?F4Y=xo`xHc;Gwh+RcysOWVfB)KYygbQzB->pReh4& zwJF`3AdJZ;Cwg?0bE{llG3Pl|7nL@6#4`yuCj`DaL6?366PhzE1Yxj`#I@pVkVb`K zIL@_eTqIIH1DqUD_ON`gbF^V$yy9)k>JIsYg9W9x+q$e)Ir2&l0YQp@q4pZqG(EI) z`0OI~E7|zVj7DxCHZbV_LI=)${3oOX&^*xaVe8pR`x& z3|C{}6=*TE{@Xj8Q-(R{WsdEh6r4U#7buS~nq@G}APANL$>x@}t`%SjhBB%mw>At3XG%r*er-7O7bJ%sA-&9jYvP)mR}@s*UU$bBBSqY!g`KKvD? zoY;XM*?QGSS@p0M6$|7buTcQ~Qce`_c{#B6tKmux|F_m^iA}OxYkf^fx+c;O3F(;w zcQOKtTgUG=GFw3qK76KitZERLSnyx!{rFVba-Q^D!DL>Olw+hxoEyDMzy-2ld2yB? zpp@A1)xNGDZFL41jV_*`SSs!j(ppV^#V1{k67xa1qF~~-i1M`wB)?sUv z4seb=JK(PB4E$&Fkt;^7{8G*Pjpe-`*jyl(2gOv99-GuAvQhht2SLtuJ978hmqizAngvu^jheKZ&!*JPUXr^s^HWV2iRC1>TyyBH`BwA-SJ3UqI$g|E7Y z7Gx^<+#9&gfo%zT^S=I{EQmX3MU&M1(7+#HGkl%W{{#s+{x>9K=3rz0zu%cw{R(4$ zmv3Fa^>P^y*^VVkmY{bK3AXSA>bW_g&PdEd$K)p%h<{zUUW$kGJ=`Ex=>0@CP|V@3(JjeTl9b zf8GbVj^Ji^af8SX^{n5&gkr5(n>O#D_>Gio3Ur8RXMn`uZX8h*tQ)R z?*xEKUH72(Ode}k-c@&|PrZq>m2Q+*WN2z-EcVgvuPG~5wdy8}YmGkMx#cMsyXnx) z!1-{_V(i|6p?Nv`VUqQcFOyE=Rl{3Zcqd%!O`q>R3jDVvHqyz&!n7t42bp`)DFQG1 z@>8d?#r=I{ZPQ?q*?E71a;Vbl-%PjT8Ov zgu_5jf>OAFBcg9f+N(@~iTp z)t64Q3F!95{2?YR?`4u1ds_YM&7)99NFp`7)#MU8r7;|ARA)~kr> za<)y`&!TL5yplCL%%{GNlqY4Jzu^_dyrz<-Y5rL%N9|WtjSWQ*sD@IH8^l5WlcYTN zgRPUXE}iWf%14uw0vQ7vC4=H~`r)a&`LJx>HbXtnd`;}zdf^v7fVWs4r%v+88AFrv zq8LQuUk?-d#%CWbe8mhDoQ0!&5O4a=%&E#;ps^`)mFl=26^ncd^9o5A!CCgLMAEk@ zZT%sVuOnLf)||>7>h+QTvVd~3|Vksd{^`^u{Vu|zq4sh>oF@J7kF4T431tGXL(lZ z%ShKnJVT~pM|D)}=MtQ}l_pu!TC__MP5G41h~4~YQqGy1Rfqkb+c$S7`z)%MY3H4K zZRe#-ZumgA*c92L{EF%|ZIVj4c_x6s#jyACisdD|IVz{=Qhb|o8_bV5WK|TbJsKIW z?ZtslTa}|BW4`oxf)F+TyOH#yqbMQ(O5yfY3Vzfmq)kuUCZ~ATy_K%kr1Nj<5 z1c_IYJo+iSSjT>G1g&%Ez5^ui-1;w_>p_1POmJ>b@m9Ni6s8^utQaU;mW{~t+78o( zs6M1u)O19U$fIH~qj**E)}^DXZH#(zgC*t=avbu3QOpfN37SH8C@T-*241bs6%<;Z zTY2BG-U)I!UP2F}QA#fM=cr%9?5>F-c&i5%o{GbL?HJ&sOX|$TPUO>41X_eKK=rEW z8=K30{^fYFP5HKo6W5IJP>*yi-Y4TO0!~Bw_19}PV%*>RO_>*{$T-O3K8V~-Tm*h; zgk3&hAONoMb2N@Av5DeH3_XO_ql-{5+!>_o(e!f=?mbV%xF z_hl&aev-qSj+oq{=cF+igLpi+meP-)zgd0e!HvjL$Y(lbk?>!4wkh%j8$r(HHWvf= z=R3>MqS}v=jfdt-yQ~tNjFO{>eb)9btpK6XFZGGyePbBg$|@b}v$6m+=ybQrL_f&H z7i4Y>2^bG}Zj1>nfzQ!Q|F{Rr%2CdL(EP)-#ZUmE{#Qu=5inb1EgV~e=G>av{ju#w z-`u6_A8`KZy@1?Dv_V!o-dHOTkzV`#wurSb@@^#HhUE@*xy0&l{|HR`xhO;=bbviM zzexX?5VJv0?c$DVVKtyo=XkbRz7=>luEAi)H8bazAR4PNd<|dsbZ)Q3tR6s|vL-;v zrl8^t&P^25LF>$jst|tBZ&=d28k%f@$XCMnX=|l0XGAe3$Hm@vL+(B{TcC@8kN*vl zK{_fZRW|5!l&l-JJjWZD!qZVaH3Fm>qZjAn?Y}&l_%|nGFN(~iud0`hn_O}#g8$>~ z?g#AAjUGP-81ILP@sH zYmhm9{_r45>EJu`ny_0;>b>mh_v<8JHY6V2*ES4imY|; zKRl!c4dQ+uNn!O~`27O#Rg)jrw_KPMd~@m!9RdJ2wt6+NJ?s!*;}9qXcoQZ@ErU+r z#2BLIfM|MI_)pjAXOC!Hbv6`uTqBX(6s%s!3-=|1fHIDrxMqF>6Y0`pT3$JSk|d)T zJx2kCu3l)XVc8;sr90pFp@Hc35Qk(!@804gBA=UmK@ z^}DIeANF0U%EKkP4yO#|N4=u50i@sf2@v$|A5YzenVBDe!Z=9v(g}|6M1~V=I1VGvIRPtuB2J0puCfezdVn<;VY*Cn;x$6v09g*G{g7ooK0Zvz zKERjyI}C&kz`FJC8f^k4**+$?uv7-yaQj#&*d?+N-ycQt2g>xc-;6 z=t-o+?L1bjXT5(tfhaZsD1Rpa0(zl9Y3Y5Zhv1O~vu6XwNl5*KN)u9dE!^V5VU;^s zA6oVp#%k^F^}0aDr!T}`xdcBq!<}mq5ys`o+IVpcNr;K!QSkh3b(b>aX#7$-VY}kv zHDm@4TBAT|Nei2Gv_tPuL?L_ukNdE{BgP`0NuWJozR@ArIRGeeNr;>I|-TKIHUuGuLpu%3~>sCBojJAQeb?s(C0d5lk<3M>$cH@ z<$5`k+xBfDU{ogg9PJdx@^r&q_j(3@ND=t%@uJY_WWvoH?@K@N`~L|yEgY>G zmtI4uIERN~5+_*@!3+@Br@aJn`nM0c=K4D&6xm58g6b&^n?pyg!aVCCS=kZ#A$nO1 zf9`$(Lp~04R>??6hLK9ZB&B@fQ-B1p%~+=!1lo>I?E5g9u!VGwGxPDvIxL9wrItWw zbQvG2g0m2vfbS=q;~>K0AZsIu^ReUnDQRl}Om6CR1#U^n9X-?wn$Rcp*@olUHg#9=EKjoHM!&`uA@{4NdQdL z&SqvzUW@b5`Iq5h^>uwSF;&PwOk^Qc+u8XwNpw66T85cA@9L{tHEyD<0)Ho;NH!1W zb@}_3tARE$jo{t5Z#A5`uQ-pmio)uP&VJpXLWjU5A0 zN;!1DFJqrOrCxEIW)puKHDd-okA{%hk;|DLEtNXae=nOvSkCwkEA^WdBl}!-qMZU= zYfc{jDaV#CID)IJ5M9i0kxs~Hv${D4vhX4E>uYM~T;J`k^Sa$*qROWcCKia-T_JdH zK`O#nKyn|AV*62nlvaO5N_nNKz+)K(RGyJ{n+%gFx3l?t?>;-;fB|jSM62|M&M-O_ zjs^lamHVV}QYbo>t~BMSUmQG#=}|dIJ{^`jy<;OQ07!qwMnK>MaNt7QW@hGCNdZ!@ z3E=3M2+7i0mz-p|o#G&EZ=Z+>Y7& z@Vg#(hNFPHw|)84&q;|`%7;7af%fuZ)xg8oM%5&_bq^W@ImaIO8PHyP|j*; zL82Qv{2 zNF9JM2IIZMeI|NM3x*l8ZvLR7k)sVP3P}>E8F~3-wM{4xKVlC5Wac6j=VhMahX*;f z?v04PnoEMFuHY%MF{}}OQrlUufFRIkR=c9nkGUik0b_LayW7N!$+9EIuBo2m0;+Nn zv>5RpbYsSU^Kd=Ks0-bJep1w-rskNsU@7|6*gImDU=ts`JxyMHAEwF_B;=D|#H4)YuGH_O> zyX?$BW!tRIUPWXvfBy`SyS(U?Yc;{68Z00}A{~22EJPkTD8T(#lv{s*Z$V((Jpg4t zr9o}6%AP60$q>cF^jyq&+a5rulg1?MswPR~fSQWH-NZeudpxw(Yz2i&NrK-5qMI-# zxyVL{DEs7FPK29HI*N>p5z9Jm=h2ZM50d4n_Ut~zLgcuo9BawMpB0!2#lP;ihm}q% zYu(2%d-2O#zyCe_1IY4ICm`hkMLmj+gC<|a;{96mlA_- z@y<4@_)9@bn_5lPDAl`-ut|BV-kL_VaUosux&Z1 zDAx|*K1_mUuB@#6_G4Q$eb+Lv3c)cKV+E$N5n7tRD@#yJA>j_n{9Z6`93Mqnqm7nn z05^a11SC&G#$l2%dp}D(G&>vlSVc%1VV%bLc~UMSX-!f{tz-&kdP}nX9p8K*b<4i^ zAS~Fh{d*uHWe(w6;nYxp&te!`BwbUB+HU+8tfQsPQ-uFCsaa~40d~s$ew<+^U{shd z3%o2ZmcT&zcl{26DVdvzS5kUl?GC=ACVMJ&WPipUggR*C?i&eWzH}$v+c!k>l-a(G zLn*}a&g#r~MVrU_{;-%^mUYVn+mU6L$<@`Mv~!hoYbx>Q8V#EDGO^-H&PbS<+U@Ns zaXV!o)@ZG&r})sXFjf_Z^6+vJ6z*YmukV%uRo=o|xAg6-M2SqKp;Txvlqj&=>~*5p z;-(B1LNS4Y#|KmyTmuS2mpbybOWVYA7>n;;8L<3H%4r)KpdgOI6+kT4lBtDJwR=Bm zc($f)qqij5qLLfppn-Fn*4Fci*XW8%-gG z*>i^`?l|bxtM_coJ-hEIHZSEavNtAEd0pP}v&&mPS;uFY8us)gVc_@k+h}YkU1G7< z`{jwh3=v69o)}&Gw|g$5(B&x{U_W9n*T`$1J@uXa1~l`oRQ3*?j=S@YHt8x-7s>X3 zX8<*`dmPY)iG$5#SAYC#1t+7zWNRn&%)m*LovOUe%Y`dspaw_lw)hhl-eKXMSp-K6 z4CIl)2~1{VBzz~8nO#3U9^a+dn^1@8s6|VX7E$kzEI-EEb5i6n(tZY|hGH*Ajg664%>Mb*v+w)ar4n+)PP4Hp+@H|(`u$9M1#et6 z;G#Tglr8-8ULB&b$mi5`+g*#7NY8ar(Tcl)#tL(Nc=b{$lY~kAH|+k(P}qCBt$`4{ zNcxoXyFmFdSSwt(p;Rx10z9064C0T#qkxc|r%?Jv)%aHVLGQOn(oy$B*`7K3%jUPy zX`n8-0~LyoFf)Og2*!88XBqXT6B}TuxEM>7CDKg{H`QT#gggKWH@V}bgC~A?#*P5|HojG5q)alEdJha>G z(4Ue(CH!PoTw#P3GXYxKJ>-oCBl(&fC4U+~>WuSz%HzU9b9je3?JPd?@mOO;!LjC_ z7Go1n69a_CHs%hdJP0&U-|UG_N$AArn%A>kU%*21KEVF~?>OlH1H5BrVflZ+I}K^q z4bha}h2PR+K*iZU4zj3sp4>J?!ljDAT1=T!yX^e$$d6=@IM*+C{=R3*LZ#u{(S+xv zi_$Mwn4)BaS0fDI{K0$Eb*JLpIf2w6VP5yMN9|8xVnv`XJaJFfpl{^TVa7usI$mS? z>!2xRMy4+}1kWMyOZdHnwAL}Q_0^%5GrkO8CTa0xPh0kX?B>c=)YInwL+&`z&WNRpT0V~oe*$2m zwttAfS+mTmhJL3koCo0!bCng<=D!igc^C)8P&VtPLl<#MM}mcu4e(^jfeTI!zSnPl zL%%WGxZMljN_I%WI@YmIBWFFui$pDW+Z6DJzem(xAd&WEI1}m;{|*^**t1C6UWiX} z1_{Q%KMR11%KEY@9%$PK0Nv;_wz;DNftjp@mfi>hk)GXL#2xh+S0r`N`P` zRH^kfvjRx<{mJ6+J1|koxv0nffp4R(?-3^xaUgTj>0+jY@J1}C-b)<`vodYwMdAAK znCgmrn)8YBYP}@&^4dV@30MWzW83oP)AoGW91^PYj^~H1cgGbkVWxvOm;Zr0K} z%1ApD(I_K$3fV1Kui6VuSvl-=s>b0&wU<;qpUfTx-I-K~HCpAN%aBA?k=Z_IhHLc2 z5YX23NT?kg@F8rv{RlKPk{I?K9##u^rIS+B`G_lmtXPo0P4#`VyQ2grjPfhP1s&bZ ztQEc{%|zUcQJPc3T_a;Mryfs}O~%l=XO-3Az?N6VKin|lTsqGnWTB3W8gq=`u!B-D zvS~BUNm7Ubap{&BpIjtDESX3SVd_W4gkQmvTAguHx1=)no==>FKhG!GDnCm zu@XkF0MDoL31d<7o}gNflcfEjnL#+D#=xx4#AIN@WR{#s4r*>IYZ&I^(nhGQ$EBtR9vyVMT8*~mUrpLa0-^|tNAU; zCzVRGFOYtQ3y;#49UT|_YKZjDqyx=PL)QlFPpDuCwe3Nr7EG^$^qe7Tcf2GMP;f|C z>|jDuw-!;tZXo8%!tU< z@z0Pb#Ml&}t340x>V9(iu*f5AD~h3va@^=U={o9j&<$^lv4BM8UF3?cQRoQsLLVUa zI+>z{K9Fe%c9kZ-1@@70jsV5JfM-K$@^E_>kh*k#ai~Am1BiDC9nE(Qas0aGITTF0 zRGYdIyA}Gk_4`h|zLGn{*iDm6Ttba(;)Q6n&aIb1NP2&(-z?I>;g%CRmr*{OoTF$j z!<~NwMTdgG`C{%H|J0}5tOsbP|2lHZwQgm2d@1R|AFIJ%uqIUEITx%Z{*2|7AHb{iUew26Md}Wi+%wAsZw3G9g6YENA*?Hhm9SqaUO#FnD#lR?o>Wb&S1yuBZup8mlN zU+l}^zdO2B9*CjHO>Evd@GTl@K1efY7LxJ3Ye1ZO+-$t(7wix_RfpGTkHLUu-pWw z?(z;j3K$55G&PsWU_!_xWo3cOY4Y4?s1Zkh^SR7Ww~XjnTm!Jc@HGtt9~l$gUG6#w zGDx=Xgt-RY6y9Z0`U7O-|vk$0$Ib-Sb=CQ$N@ ze+`*7nFLDxW4WkO%6%x*|GV^5L4CZMvsFndJy6y-frS3POuoPQshWPN=R`X;6!N8! z#l&36zE-rJ!qjkigYd^x=A`W&=d_a6lrN?IcHbCFhy)Gw# z0z6d3L3j*@>=h}mt)n=#Xm4lY_?*^@z|nX;p_u!&BI5XnZWTx*0w22Eu+Ty^h)fUG z_|NH|QTPp4QAl3m_hNwWhT}}-+}p_gkmHzef+lDKE?VHLYd(3zFhg*7kY(xX@#r+E z+fpVA(qC+U5m_0~y{3QJqV~YbQL_b8h z$N=`1x&E%D*>rmSBnJ>$`ytC?mL&t_Y(BdPCAR@>;3_>WYNQWkhy7it_h9RJ$WH(f?m`HOP z5XKYMD1S{P3IV~qQR4_5WX%gNs&87jI6WEoey4t_od`FtL?agq9wpQRiGce)ht9`( zu(`t*7snxgU!qEDTDP+`^gb>I=#d0CPlJ8BjdgA^Y`oQlQ?+<{{@EI`U6q&K6#M%4 zebdI5r4X?qDTuEidqKOnFEfG72us}LGsq(rJ$f@w&pR3UJRtNkM&5BgLyQC33YIcs z(Ns&8{~*5+s!5-+&Bz{Ornf1*b!)*S2OLc%opg$InzCL%X0c7s!`Ro_e(392IU3mK z9eQc|9Crg=#76c-*p}^AWJ2F^EpjD%8+SVw)LU1y>jI*5D-N$;KQYpRX~+JSsT~%r zwtN-1F1UrrOVNz^n5FK_1Hp$9xyl%w$Jv!*5Tp*3tq|J^N8(7pRsBqoUE#I+3JHeo z*R4{S-1BHKJ%7$dB4E;UhB`H@n+c-QJ`y2sB{^XsP=^ni+oOvxr`R1eh%yMM9ET$` zRf2$AArz(^VTHV+49L|o#kW!oUXy6vSySO^!6uEBLnd<+aNCr}XE%lOrH24hm%0v} zMX4a9IN?Xcb{%=VW1v^#$&$?zv#(OK&n@vvc~>kAp1Dc)zWe;8Q^w;}|0G9RsLSuo zsDoV}P=3QXEA~5yMQR&uC*yJab(*$&c2Lq2!5)-`F74W0 zzZ?Y=jr==7q-tqFzHY`ttU~ef`d6k{T9l}RZP`S4gz{CsbY6IN4M|p2n4(Ua`@n5r zfn*UUh}JG7&I{JOcT)C#$&9zHvAj2y1c5PMFp;(myX`?R`8a9?AqKTFpW%Noc22>a zMemx9ZQHhO+qTuQZ6_V8W7|o`PJXd%+cu`pOij(j`B$CU7wc-@uC@1i>wBLk^lxE1 zP5M$eE=%KtKor8*#6@EP&2vYP2km*xRFZ+$l#!Q! zxzdV^?YiO-egHHi^Ic?~WXMKZ z;tLeh$}2ObpN;-eQcoW2XeMlZDX6J@>eV5F&djHS>+-KDwqw(-Yw=d-pkNkV=i+SW z-Hp;BI4V9bp56T&Bzq=?Xo0;V^apVOGDWV_wO}I)lTevtXoOI4X`^8(usv4s zw~9vk(LD$<_p;^q^SKunpyV7PxylaToD0xFTHBYoJB%&L;ktpV`ToR$ZaZG;AIy}i zWj-`jf{@GhtY7mz~00nYH4Ed>1MS{HSZ{ia~N^7 z)%O&z&TJMHmOCZLiw z+Cd`R6e3^K)6AP*L-eXs%WM{P;m_jusIS5EZ zU5`3aC7f*sZ3GGyiSB}m0@^!{t+iOuMvYHR_=8OSRGrS;x`jhlt@e{7xAw3xtD0Y< zqw^~<8vSjR9y=NPJJOpy=J~K|P__28I8pUZMb3g??eOtH`xQq3#{0>c?7<^za#qf# ztbaEyniyZxq(VwVRPy&WbtDJlF?Gwv0lMp>=sVHWbn(ZHloGfe?MsOV znHXRNa{!DN)_M1=Gaa+f=wY)m0;!mgy5!SCwdTN>4*B@PqMfeOKv&9h-+{t%qMJzg zWho^Fo2zAIpq#K4Jq`Ms&){OEO~@3E1SDT;oouCGz-`No^CF$mz_&1rOuZ@Re5VJN zwB1aQz;9%IDXkjB>b*HaZ`KkKz23#A_>@Nm$h~oZxT3{ZM0y0t?(els4~oek@=d3? za5g*QQV@4~-EK>6G-6)3`xkfG32kKL_k{XfxYL^jRNE$d`C`!sl8r8$pWqW)dhvsb z;xTlO|M%c0Ilc!M1){4*C@g-#_>6W;sv0L2l+jwNhF+@8NQ5PYB9fGcKADJZn}CSh zvAwF~A)@)Arm=doja?PXX1@@ZxCQP<5J9*KtDDj)I3%T_mmx5yA%dX-ufg=yZ2qOC zdy3<7Or?V?^ce@P)=La%^HX`D`8zF{v=-_H7m_fIu)(hcDgK=i4|2I3t`|@B z;qNbD|e=xsF3AssZ!YXdU|pR|6A zch2t1l(f5~W7l1LF~#?Lzftx9GzvW zhbXcNNCE!ny^4VTM(M?8ED)E@M_C1u#uUUm3D)lbk)-)}{j2oV-yp0FcM6t-$+dl6 zoS0*U+TGS$XV2B4qwWk-*D~ePbVdqa ze!ldRryZv+kUx-lInK2IDM4fYPYD{w{|ESO>;I0wHBxSIaJ;vi!m~%dL-;0J$4Mrc z{}S_Zw?ehM#j1=Htd6#$O{d{y$bf76Q6Wx5fK4Z|!n?zufhFQVToC|JP zJlwu^@uYnJxU^80EAXfH8#!m7I9}O0_ao51jxjEYu_UM=8AHI0!AwG@OZvPW!8#F> zhdb}*_I)j9PXs4rp~2}sGu)+ul}EEFOu9)9J+3z!!LUAX?1OVGU=njU<51m=-|$w< zYu^66=gR)hmKCv^OtH4}pDBG_?kH-U;D*4plzzhP{r!|t5(0EmKq$nJPCpxU*I<~2j64=3`V@+NtQf5T3o2> zxpt+oF?CIc_m1jJ`0nDP_F`LxV)0VBdpAw>uUo97-f0&__)3ZvM!kFUC4B>RYqcI0 z)qa?~;|{Erb6a(W6Nrgv=0%@VDpKa8vMGT3+KQd11I+R|Hz#ejy#0uUrJp8e`XiRs ztL9QiPE#85`qZYWpUzsbo?S6))KBe|$&a4sD+oWRhLZl$ggO9)QLRs|9wN0Qj8<$` z^Qz^fz3c8Yv+Hu*rP(lFI@h`;`N=}-+Lh2bFI*3?O8x|km6PoK!b@RYGsCNvD1#yC zNG~~Zn(UwYoldh^)XYg4*d7+{Z)8Z|^QP{m@BW38C2VSrq=rLz z#rwXpub-RUd~wl+Lm(-fJkep3h918!)pN2F!>|t`hDn_6=BsMwAf|=*hWqb_)f#zh z#mOQktT?ahr%^|Ld4s*z*iz<^?;mA$6(y3frEzxlx5<)e1BEIJ{^gS5<7+g zk$y7F8|S0?ydo71qO_`Cf`Qqa=2F_1WxEdmf|Ku2b7>Y8Q6KtOSOLf$OmL#j-+g^! zY@>lJXLVtq#`w)uGk!wY~81 z{R+J6`QaVAIlOJOjaDHP6Iz%QlozOk&4Cg3Cc4i{w_AfPo|8aT9X4xmV|g3stwuG@ zvNfw$oVRBz{yr=#&6@pWm=_&E2AxSldv}2_W72EvOXm9Y210IjYUA~j{6Akk6%G;; z)jYfG{K8c5e8lufjbb4sHJ;ipU&p?%c(?{2h6at9K!{_g{xNnF@`cg6h--BWlUBWZ zJzqMMFj=dN5P#b(-2|m0Fud+TBALTvg(~%M+De)gg3* zb1?js_-(Xlx`UB*;T*N)vdRjiJFi}Kf1*;q%3N;6!VF_s43kP%6=Nr9ao%h>K^%f< zN0rJ`^1>O+-v_3}!eKHdg(jJvZa_z3*7fM}hw{O(*UqIOs%9704_DWCxOS@%6_bKf z0ND(;N8E!ea$3g{X;z#jK_VUc-~MOk@Fdjr)qNsdzy3N-a8TB2BArk6vTCeMpnBq% zb2u@ix*(u-c#E-IqOo=DFa=4_~sk@UvRr{sO8_qPK@t&C~;2C1a9 z#w60F9$0i`?2}>f+f?(jVXN0j{H3Ied<-n%kzk$gQu#a2YxOt80THm7jgYU6T%sL6Xnmw2vok3|dBJxnw>z%C6~&pZ-#o`C9rU77J<; zqKFdl*)={%cL^0TDgPSCz+Dsb04AMN2ke#Nzb7I-vI)v1OAxM}-H~`^V3J$L@OQBV zawG6CfLSfdv3D?fvW$)(0b#8+Z}d*LVSOM79pu66JaQ=#tU}{BSG4L96R*RF`6iIP zdnLOciUbqo$*^xPLmFeOMF%kLmFWA>%RDs(+`TjcSa|q(&pSUbJ`M&$;QlZYWE{`i zi^ByD(}NnwbEu)P_Y1XfMt@d0`FTN02AXB(=2NfCj zrfcY;ksj1+?A`di3B(9)+|CH+f@thzLxi&XuD>KtHK|g(0wi0S2L2FOr>LmV6SGOe z;H%ic_N?a|UQfQ6I}k<9C?HCxv*Q}PQ}JyWQ6Br0$#z^5xK{P6SOl#P!fq6JSK zF;r1;cCEwI)DF3su+?=G2zH2~6Eh3Nv|E|@=WA1^oeth$#uk_j*}i+^eY$03B|lfq zhnh>{*+eRFsD>I}Q<`%u(dJI^8y-w!3^PDaYa+kRPx&8K#rpwp$NZSyvUSv8Qp5d9 za6H@2^R2~xC7z{2h?EpP6J=W}_4WN|YM?n;4kIj1zOP%4PD3*xaKy7O?-m2*Fztk( zd|>h*&gSd5sc$(8d@>w7?`_XgYlZBihvetl2%X43p2p>*XvsyrW!qWXok9bRc_i0J zaa+t1Io!1rZO!X9?|6S9-<{MbY3s|r0y2YkaB5lmuebh=4^_Z5j4$F{4<&e5!F>9w zm{8{EC8Woj>l71v#y0{WSV}@E9D@=6uhW0!%#Gn%d)$B!I@!J37GY)@5$eRktc0z+ z5IJ+iK9VXODNKYH-VnK`zOw?qLf%fef7*0qp*ljrjWYD&5+{;NLoGTOxy&Lbm7A^~ zpx#@CDWW@lek^hdPyeog0W?G>sH}@EBaFK07d!`qgjO{TfvUx4Jm-n9x>;Js5Uw#+ zb*|4&t>~^+^qWvvgJS8zTv1(Oh$0Tp(BT3WJ8WDXqhHb)YuuPp&EM+VOp5(1^?)AV zhAZTD2e13Pm2)9dtyTkzn2MX7>r1;0qr#eoiF}oX9YAA4$Q6wY3MSsA@B3oZ0vIGs zeFNCiqE)m!2l2!;R^W_tIgr%6sD&%5QHma(4QKb-avWud4O56(D}!oW=J0PR#>ZNA zuETCal4&@KwodF)o<%M@po$-n=<|sP5j1(oKgvE=@hEz5fA#ZpfFx#{7|Y=UHFXuh{!Dmq8q`{36EH>fGH6uI(&ER?dCzac zu4G$x3dp;}0!eT;QE9ycZKiHsW4>w|n5+tV0Ie1|e;O65iEZ!t^|2gjr)K*x2xoGa z*sVDq-Qvd2^3X8a=kkwDhP}P$R>(Mt6<0Lu{}pv2nSl+PJsjuA(|lX0B0j%oCx)}z_ZuUe6c(dEp~@&pAQ>1KI@ zdcm^%_jyrNq6%?h>=~eYYjW9dwn=6sXpT$WzP_p)M{d9E3PQM#E9!IW+8WYiMqBjs z|4xwyaKv_59%%jBI*^i({dA_HydAaS8w$K#hX(}Gj~h>(P_9kjp|<1KsSrdeMcusMEu$c(!w{jpqWA^J*!IVQ9)-M8$E^O9fAx*N}KI1a8rNu=r3g#yR0L4o<;1qQsfE{m}J87 zPiYN{z4vUWUQM0N$Q02Bfr*`fs>h|@>813WCW(*4JFfzAqIvh+PaD%|Wm`=qQ~#Y& zwbg$jbZKNOh~-NW`qH&jI&~{TZ>Qlm_njW4UTuzFe$!HP2b)WIgM!}zArZht`g6^M z#;DA9YHHwJTu3Bs6L%>l8DO_36l)qOH?2Y<5>j#*Q*r?d?P95dm5bDRvrXzGC3j#? z`}=Tt$R>jgT76gJDPlR;R8nI1d@9RPNw28mztpR2$CU8fL*?`7t;gV-Y^I~$FO*gV ztZNJuTeT}{oJ3%^FuT6*4y)#EJk<8AhDp+|uSKA_*{9wSqFS*CK@EsoY>e=KkNQ1JneDVhSdu2oQfaNs9KTI2J z)YN`GB`@ZkDmnA)8V|98lfmAn%{|H6U-Vk9dxt_ zk}@y~j?zxH$i&$oSN7d>S$u4I=qcp8j5S=+^drDH*JHEq0d=QI4N#q{tT41oDanxktd>yu$l{3mqMHrh^jD%=UgNmD$fl z&^FJtX^LUok`4rU6#l{!Gt$ZeT^9_l8381^7%oh)1&LjU2KuTiJhJ!-Jv%`JzG=D9woM`> z6&y_7EDe~txzG{&V_b*!u`TGV}o)qt!wdXzv+=yUucoVKEwh;2xu>OR7@SIhD8{9*SpbM zVx4j}eL+q{wGbGZGc(iruH+K%z3N^qGbm$a1mQ6M-mW|~&WY!nqn#kSasOds!+L6ceD$Dji+(>I zI1)0^hF)5jYkZ?6{dPRWGMEOzXb~tJ-Vc|?5RZLK#|NWJn(%saE?ekrZv@>rJV><@ zEC2>3N!>M6VS-A#*d^~@|{pm%THdEwSAqU@H%!@t zK&|Gp1^EI@S(_?>T$@UsV$pbz;1@l*ecxEB9yge0hL}eJ?rv;chTE}@4052X|L{b7zy5pOpf}*_{ON4}+(Bt<>YOUk z%K?9}vxD1&Q53kulHArJ_JAhkHB24$ta0-sNuTlC`^ln?JDuCujBc{Fo;w+vyqn9Q zGS(##_uV&4~Jf2`5jK2>U>GVJ$R`S}Xr;AzJ7;D`y=hNH~<%)J+gZEn-T=_M?)Ms5}S zPrdWs;l0B)l(?Md_$+yD*QID12J{Hx!*fq+BOZ!)b`0*?u(|3%Y#(DsJy>$5AihpR zAdc~lfjO!8BB#s7@(F)cwp)r5rP%I^^^1FVyUY*L>rz!J24D8Nae`Oz7dL*Y6wzn7r}S8UT3^EdSIZK)P< z$Gx8HGi?t9sUvr%EeMuTPknE^`U9=pva_h$covPE;Pm8F0U|9;j;v}3_dh;q3SAA; zqcJ!AyO)bD`%!Rd2|H~kcx?&9q`GlM_dFaY$}tdV>_JO*&X0d<{vIvl-%9y4IDiuJ zEIs8aNPQ)OE<|8`R=tt68Ly-F7M%Ht(p4dXQfRKf4p(!?W@F38U&B_9*~b2*gntys znE=>JMa_G$zfx!+@buw~Gpp;0(x;_H$bvVU~`dot>ITtJI_aycT_I`t$g1aL# zi_2aH#SudxT=6FQgol9yepXpo^ahVk!OE9r3~nYNFT*FNDdRr_^Tps^KCVxBz4ad8 z)wS6UBf?OklACf^61oYCkgG&bFs3#aSEXX4PEg_$Val%)$-;jL5%)&CcYSvcS1&i1 z9I?RQaHc?>UD$0qZ$TTN;cY}c-wW`^2?A`^pDUwgcB7`(r?xQMdC(yO>na&naqvGiCOBTonVLg1Q@1uJdeERp8>pES)okqgrx>I4%AJsUe^T; z#(YI!HQ&OtK>UGLGYOYZ^Re3I=TJk_8q3WC$ZU9~;hCQ*iq>~-apzf8nOky#SvFkc zqBREa4s%jteqW1jI9*_Y#VJ3l7$5lWyK08@UBnW;f`h_S2$4t< z*$;u)5+o7uLHKxmO%C8GA*B_7joIOo@C}6#MqCzH#zTi5y#7X@s4mStJb?zb2fkmo z1#NwtMG!&aiLDS{K>$cx8~miuut#tHKGPvf3rNgXss}ztDA{S?UQ5+^2Sb@bMlS|IbC*~;|4%Ta(r~)G zWA-CJ9hIF2gzZBOAbFFc?G1E9%Lsnq?FJtRx_@JK{Cfa0Au0!q2kPe0vWIqtwBbJ0V_{rZ zj)NP$vr)Q57o7QWsy+MmOVn}59e9KDC@*t%M&W=BirE1-3Xr+zzAIA=+AM@G9}A{$ z;#Qy*dfsE_nP`nm+9kBlg=0L0it@sW5|CFg3?Nt&$2er@(8+&90F=2F_bC4iJ4AzE zVWVGsv`8i@l8=ds%PY{Y=h!$5Te8NO5OAIvA=Y;c<;+MvcSym-&9j8zwC5Yx)8r?e zd&fjV?Li>a1YAa-pz|>+P7`26wZYDl6|@-j+QDwZkl3vPMlciq8tb;Rc0yf_Z`kRF z)@!<;iE>wQc)0x);R>H}o|SS@>Vm#zu{+jnW*lT;v=i`d7P{6JK$YN!6(>W!)0(OplW4?8X!^3nkWSzaU@;li#mPOq{q{9Wsq1z}x zwS<+l(|bo~sD58ZFYO7T@FmIqdZz9PtM`a4HO$M)g9I7s29Jte2-m8{9o#EC6RsGD ze6PS07app}%-IpWmA8g}L{bAeg*({IuvkUNYxY8GRXOF3O1jsWlz;&zee})~24^b2 zo5}XhWuy6kXb5fZzUPr;B$wzkROWCF89!Qu5^q?kwLN@;PT)z&Amb*Y`5`%iilq@I zo#v5qr}rw&6{}SFgte$)u7ffx)c!?W#zyKddJ<$If0VD!c5#2qOvkypWxT+Sz-Q>+ zSAk@O=Hu~3`Xw}u5s{EsB_{&J@c^+bBLmk3dye+(dl<-M^}Cj`B(>-w@pNTI#0s7` z)KXM)aYpnzBY6cFn^Gy`-YK#fS^&#;tMp62OM8}$=6n~ zhin1Z3Je|oU>;3#k_A7XJ9a_wmelgq3tfuYL#W(J8HHW!$-%X!U)rskRG4=2^{dk^ z03LU%j70?0IjwAu7I{tltqsei8^5Gv09rJ`SPSO)BFRQ;e<-h#2ahf^$x8;>GNLfR zTD`P8t1$l`?ZjsCS{cJg=6NfS;Zl?;c>a$oB^uVUxe@uLuarFz$2mY_Z=IHZVBNg*h^ayLJ?CL0f(yYH*;h`F!VU^r!NY?oqoz zyBKZD4HLgx1FLpBiDFOSSdv;37>ZwB(1;f-(QDMZkI&+q4^uH#a%v&8JmmHkxCOr2 zLbX~426N6pEc8|w4AwDHW{pvV2|Pp~U-&raFErp0xKO1a-0i*i8Ij_YUF%v~i~@LL zD&da~w38)NDx*@z=F$;bRIVF3n=ZQQA5B$zTlJ@fAh9F}SuK2c+s4iyOh6s6K~oT} zKz&S;CT76&Rs?`}lL-98-G948blr2(b|eyZ_id-ouusIz>BDziT^Q?lUjjA`)H_Kg z!MkOy?yg=@bsOW28Q+|Gx1vByuqb5mvj->bM#LZU(Im!=PS|!Frg_S)G9fOXEvJFY zT^ZoZ1TJch-x-?f3#|<3sBtK_;7WMo6UGTl&lrqp3)>YRxcG@kU*^fF>|-6~1lE5f zkf;R_=bC@tE7=Cst|wiEaGdpd#!;5Ej*jHvNm)rrkwEGr9aH=z4Fj1oh;GD|9m0+9 z*;Uwz(!xX#GPm)@to0|82{PV4n&B9_e)2RQDl{#J-(Us-H4CN71Y5aaR-}eO=>nq?1{ept}g{ z6q6ceFHO!n_gQ#F%9CE=wPsm0#Q39}l7Hf1sO2_#kdUin&<))suXy$lNXTK`Em=ii zHgFZ-s>3N~E_(Bm$brKJU7DP41H;vk+}l-xJALtNl9P)yFYAj)!AEvGaoGEy2|I=g`$sj z{NBoQ9Qx+=TP^-qQB)20&fgF_ir)|Mp9(<}lWun|r)HN-<_{DH} zquKs8+7F@^>DqkA3A` z7s=E3$MD^Z9|c^#o4{NXPiVEjy-bnVH?Dbd-RH#7hHk0x3$q+A-*2D{lHX4Nw6;Irv# zxkg$EWc|3bCm$5{e(z<4k%e8Z#@{e9430q&qzm##MF}VjOxA34bE3S0qpO}e2~gbH zE6BDN1&h=l2{J*W!M4y8dr_pUM|B%*h(_dhRHL6m51&|i@0j1choSLylvS_L{~f26 z{S-hSjNu-9rXV9eJ(&IrFHRZq*yRGutgkl&ChWJMjwCJ+Pj7O4=Q6JwQ~>iKKCipB zi!4pEN6ax)LM#|0NVJTWJUY%i40{6)WQ8HME#C+&+EEw<|FM;isICO_KG5S5vAWht z!IZqO8oL9(l&66n6hYQm0tED>Y}-z;Wqiv4kb~cjLxCfm4zU)FuznJ~+qxiRBwcdJ zy9ZI$#y0w(PQc^?cu(i)w1iIw?l~ktuoBS!iv1YC^Y7%ZrJ5u@RvU(IEy5BDG@4$sYl2?WpEiC~4E#G6XGS3~oq z3q5NpLXfeTo4c}wUH_H9>w25s)SFsW5KSN5OzhfR@JawTu(!Q~YJR{k(_t0Et+j6L z-hO2vkIy%a2~y}$a*`!GZ4hwY7(a3==BwQiuQ3OaeD51;flkl8m448mrtlo}5_qfB zo;RKV|0=}fQ%0OCX_Fl%lm?O^bc;U*5`A>}c{!6EpoK@IWQ5tb2K(! zLD5Z8JgW~&sPO4g!FQ;+h_*RRcUvzmoeIsWdhlVZHpy zJ8sHk8!Nf4{k2c5O*cHI6>nw~;3`Kcjbc43Hs9IM_}X#uXE&1N^bqUD=6Yum{rM23 zWnm}=#^_*L_94g#`Jq-;hWtLoUa|DIa9509-r&w;>4tF0bQtV(tex0`}MEe@ZJ^|9e`=#=*w)KbL2(bY)#|#}a!U z1ULyD>1?5DYCTd(I5^nQGzS}`=12ygs}EPLK{ol+j&eElG@1)~?_IKXi>GJ4i{J9Sv-BErfSCVgTrgM9fP4-zG zLjP>~@xhY54!y~F5>qU{%GvO@>*h5~c%;`c7%JTOhYPoGO59I!n;LI!?Z^5N@<=UR z7_-#aF^I;$RA<6ud7XfF=6xSskA$4rTC`&{uyzifxt5kV`m!Wz=epEY9FhF(gYdrD zX%bYlRFQeE_066>6^*lNS+Za74Yy3iO@Q@zTUDbyPe0mZ&K0~M5Mw~Ax!AwHM6^Vq z$oPao@%4kCoeoxNo{r6{o^h@9J2v~UaC7%j`9GqJO2+>!%E;FGGl&@P>_4-OJNCnp zW1Yb|1JBEtxn#kcLk&Ko*O-p|t?~Hb3+i*B2|j-oAA$>x))TF~;F$d!MH7o0)XjWtp|)uW5G=D%x!-nEs%TKH{jQm$}sN55#Dr??>&vfG3a%ODv>mfyYUV?$2spz$k63Um6>8 zo&Xgj#s^;S@qX6dRWB}a;|4cRRC63hMw1rY4Vxg zD;8wBGnSDrzr<)}OW%{T6bZ!8hW-fQ4vWDBqemgtA#>qVlKr>p_|Bq3ns89qD4!n1 z>kR*m?AfXP2^Ek6FFekyDU${$FSh4N`}p=Qvt_xf6vRA%6Rx<1NjcKhK8GeoqRk;Xu8Cu1C#{yHZxe zoa2PFV@Qxm;OKG?4Qw?t{S?F>Wb4BC>iA# z9-AO=s^2Z7TJ1uU_UN^Y`UqIp&gmQ#9*A6YMo=;h6}o#gEY>MrUkz{&*JQr1Ms?(7Fy2>e6{)xPNZOLuF!S!3F(P`N9e6yBgQxXh87N zK+_kgK|Ezd8WmNOxIB57X4p?t?>$4a`xVSQMcyQRxRJV-usw8&t#?Ar155f?$gACV znq?)8W}unsI6IwV6yrFwmhLz;=P-87ObkctPtbD^#`UNR!K%XZ8k=ba(&`fH$&IW3j zieVTvW&4)52EAi$*r0Jr;6#;qZpKa6KNc2w(I70dC<>3^dJ_7B!;}2qUfV#jDPCZd zis5vSt-c`Cl@n-{n2x%t$CSAo8#`wMJ05B`2Xb5Pz#$2h7;l9W;qz{?AV&5YBKzAO zB250$T$s1xQzGLjEj&I-2PT4wfMmylAl)kDwhlizXA1e;RS=m)&2NR56gU-h&ELc* z@>4_pb^4dkB9M#gs)8Jmu2!K`o9Qp@o>Y2yU#+UQ`>ZzCw+I1Y9Coda$R~GDV|CZ0 zv5TMu0|kz4eaV&TLwBoH#12Ez3%Vg2Hwh-DRrnM^Krp3li9Z7`lOm_?l{ zk*TQ#=LKdspM`5j`qKzmYmC$vNQ-S$ei&Myr+ms)U&-7Hen{of6~U_4cMfD`0dWOp zTlEBg8=L8*&eqwx)Fc?}0|L(>fXLJ)9Dv3P?XJ)9AwJh4$&B61!PQb_(=KUYO_I_z ziNXq9g?;%PuUh}dS$lFjjP!}iWbPBb^z}X#4Cl|<^BNV)fL==s_%*?JTE7aAJ#v7a zUVbVtqC@5^bWf<0CeZ9A6bk**os{H&1M!-hFdygw^1ZF8C+_t&Qh0FJh9XNu;TV=i#nIFH zMP}j!8n)X_t}FsYW|k4wBlsT;iBh^TL1h-5X4n)tuO!#_Bmg~mTb#l*82JTYulwo_ zY(~fXUhAGWaWo0+Kz+aOg1jpZ3L&qzT3RT78PtX_Q`qPErOM_5C@kz)_8ht)E`_nqkm^bZ281xUo#G3?rXQooR7O5urvdBl1YP zY9%0Jj<>%RFz(w-x^wrw|M^Kz5yS_s2;xce*QEZJ@afa%ja%J^`kUZVSF(gnfBmoC z_syvTAl(fwGtL%km?Rs1?oRti_d?+VDdy6M<^v1=x>R(HHT4S zDRqFD*B|YZNMZMl0ctS;Kj}w6DcJjP{K^VO z>ushM)0PaJq?40zC0eueoq>XTt$6);vm|jkRDdMXzaKK;;ExR2;Mj=d`ONN6Yjh=W zHRsO{?+p_PwpCv|EkSqNB1lU13r(eDm4$ry4$NkmB2oU8t+F+X-L~p;-3DROi8Cc# z+He#q+p&$jb`TKD{<{|TGmO+FX+RN4omiq6PyE8wKmy7f%+zo@)ZRHjxJ>!s>{qSb zL={@|Swn z9@=b1QBG(hXniM2lp@@AmaXX7;>+@&z8(aQC4E-(CHM3xq*q?+P zeNNki&GYvrvY-_*m3Wr zQldeefybD8p?%^K2vtO-(ZSJ%_{aX4TvebfFMmF1H9|R>euj`r)|8-LLdd zm{!HN)OwY`Wg^B`uKW$y7WI>B#Fs@xOay_T_Qr+uUW8wqWB$x|Yj z3U4CGsO%Q3c2wp5#R0qBt%+3J%!48ibb%QPGNvj}5JO3^biT}lGc2MLkrEE7h9(yO z-0C-Ps`AIAjdA)!_J!=jG_Y>59YmO1ix0VW5UPC;g!_OtAoN)l2N zWJC*JH9*0di@x>xbBa@*oC?`+uemVLZUa1_C`8!6B3fXSaWJ*E)!S{^%P9B*6B>Gp z@!YlA6KNVac3WDbgf z+L1W$1=N4>vmA3T`w6uF>}<#mkDlr}N*>B)ZQGLpi3a-cb)iQ35NIgg6ypN(Y>2c} zD-|Z1BTIpXOYW95ZXq23k%%7zhhDiUb^ps9VH&&~Z>f)3CM84-^Slee(dR|UO`U_s zWGkuJ|5Bl;oO0YbDVV39p>}&%o6lfs)NKQTpu9XL?*(9=i~%I}O2sE8oc{+HU}e4a zNqnaS_>6UN{>5+z)`xFInsYYR)=6lG$^%hMs47>$TJT{_h(5|P;Bg;-??!qAqt3kt z%?m7l|2U+mek5kmF?60Kq-__F7;y^UsmPsQzmhC(0c4-Q0Ba=N#_30>vpwLA|WGZK_$#G|qD|TTJOl3%%eMQ|@2* zQ=is&_4Hz>?|-*iz7#c9>~2M{u7#I{cUShwiG#|pTjUt#StJ3y5pHCweDDi*4Y&nr zMKaSc82!Xzod$spdU4|rPm&API1)8ag4w}2o%?(pxtGOyxrBhJ)_;JzK{~<6K>Kws z$=-MekuVIOMsPTDO#dR=81-z*rlQTKY9iqe}^c_w%g zB6(Q?sASP8g1r)OPb^`Kd**xU>Z00%Ujj;^s190Vj+LC8tt(CfMmhmKUA}3#bO$Cc zNKI}Kqp`fR--EjP27s8DveP1rJ(i`d#eK+PBLGC5BHAJbi+y@$f>N??970-06nNfhH6s)e*>!YlA^Hc2{`N|Y@rI5uvUu>^ehW!$Dr{A9RR!rnxw`|mKV z$2T;f)uIVJAXm}ta=ARtV0+4aEbRG{76#U~i_zqCqzaB}I|wU$`mKrO`H^cr8M)X_ zt1}rlPRoerYBi4y_hND#p6z>yRj!`iW)O%^!^zgQ8;)`GiTvihLFxDU^ z|JYK%Sf6SuC@A6Xx4t9!IXb|a}pr6)nemRg9+1asGc=6kG%HLbR7C4_h_kN1| zO%#-|Zze3h5x9Yr|>Vfv1cCjdDM8;w)R8rq2;g0YIw(D(wF$ zYGV5jQ4=>8=l>ivmHb!BwB^4zZQ4^JRn(6uo$^SAV(yLfBhGY^PPBK$*{%5<*6fs9 z#ROS)>7VgCy`C?iL~|@h#TGQ`(lKB{yZf6iGu=YpUYYcV*5N;)Keg_ZYiCzuf{JAP z{O~{8R|!;Q&kx178Jn79#LLo_wJZUZaA(~|OJ`WGi|Ieb!ewn$>tEZ1alx_6 zvsba)qs(I$U!VKQh(KW3UVkt@o&YTRRf-XsYW}U9^r=^of>82}q@yC&RDBio;-!k5 zZcS8%ex)*|*S+}}f?=_$Cw@~W5F^tJG=92W_O6!@RQh$9C7QscVo6bF(>-Oz)z_2G zp6!%+Vd{dimAR&j0G0RJk}%{@lq>RgO1`PCu#&;niJtRUmsB)EkG9T@?a^b&uBvw| zY|17{?*hFAiFsGQ>x1Y8)?-E5J~{`{dgu%Zm?R_&8SMnL$@-2k>h4^$HgAjJm3t-D z!LX5a!OIUvge()>%WAbk&4U@fm>swAmy>o@n5)M#Tp4|D!}NEnwH5vN_CVTcCh`eW zyzhAP!_pg5@Rdu)fM-SDdW^3}#jY;937zi98)!^-)AC*veIbyMabW4k9LLEZFS{Wx zYcLRZg)8b;+KJfQzC^Vwhu2B;otmnvW(&Ms(RXuAQ2JGpr1n>yNeAk^$*br3s#_6J zdKT8k+l{Ct8hXg{tov9t?SL$^3?vcr!yFX1Bb#A@ywTYPBAtG{=@{=sA*0hmhdtFU zt4?2Lzb)HVYq^*UWVNs?o9>;z(>C78K@o0lCP9n0d`bKG>^Rj$7LV=UG;bPM3uPkYY%spZhSy+^ z4`I@+R#tx6D?OAii~Hr?p}$Gp)oELvY?i7aH8pg@EupcH9e~maa%7Xhn%lbC;K6}F zwOy9*x5R-p=V0`b4~pH4nGwGI7MCtIA~X9jvN6)kxPBDtdr# zgVfnhp{5_jy6-XHjEu!9Bp=|q0l=}xwrdtJIf7l-HCYPcdVUqE(`U2O$n#ne47X#{&0(M%3 z->^Vq=o40|Jo%%IdM~h{w;(XT6jp!=p~7)IAV`#Mp|j$24Jv@$hou+dsCrrvsdp(i zP)XxhzTN2yTZh!4o<&>a8tOK~ea!prBJ*WqOaLyP8p7q=tPw8zC zqFtf4TI_zy0TlBEx=V22?4E%%XTw<_lEts{Q&Y_R&0!&CO{Y&9j7PxG#Cg-!@ttxC ze`gMuTNFU?xp70*wLkf&5l?3FKO=#Wy{})0{4F)7iAQ8(62Rwy51;p$T;p+}zNw1Q zp>}5}tj##@^s+{7+g(l}gP(p8d8H>ZUmC&=OOQ&IV2J+kb71oqq0ydWN3125JM6zp zEj_+Gs39`E-=f04+6riSBhQO2LrZfD_+r8lh4m&la=3CsMhjhak&1q zpS6DU*o6MtJ^bK(eKXK|`tmg|8?HMv0iSC;^w0FN88RMAK%diUV>#@b;BV@?EwY`V ziIHwfpyEOu-=a}}y}ShcIFkB~C*(FvK>I+XgnT(Xt2EgRUPy~ML@FZtq!*x=rp=KA zw}Z_O810RbZ@+b1{$sM~=4f2*g$bb)iE}0#5S_tx`vkG2_olz;{kJP=xfYEhDRaVb zo$5Cmx2Kh78o%m8rjQ_BGJBktSyb9X0({e_c?)5B+NvjPSs!=3{WrW+heoI!siQwf z9y|^4Cr1KA&>J~o@m8|CxMJdQTYTNZOvOwCUsqDT3bK+X(Gy8Xf|yWYYIShu6Lr*c zJ{;yT%*1n-k#o8&5b(rkF3dYn<7u!sNHd`j8hfY-hf$J!WCKvFOGdIe$3s5}0vTe7 z^u8EBfYnc+#zZX(!ZQSb-oy{vMqAv&QIp8T4a5v=Bmj-Iz>cK2W_-X5*S_W;wz`iG2*7jPQ00~jkz2>mX0%|TxE@@O1MFBZ83L7>PI)7sEw&pck{L_ezORFkQEv|=2+Zd0zX^W0#MXCNKGP^-~ zBXxc8@qpV3ohHt!lwyDxV~L#}q5^Yb98Ne~1>|Wi%SNi)*tx6_oEAL_d@6745Y7|l z2yDx{TAf2Ig0OmPT~XyK0f*m=DvV{ByAu|qJ@s_`bC};fp}hp`RXy8spJ7M_RADel zkn{%;p|g7^17GjqKSU7z9Mi6u^9rASP>NXrf2RK-rKM~=Gei)7^!~P^#?SvuQhkh>FrSkUt_=9I z@Rl9TJ)41Rus0qg+8&=p>XjM9{Ae7s*NKHeqfu2>k&z)I3Gy9uI@b@y*;q3U2R9i; zG9t*44UIBHE8i3wE7%bpnotytrry3S^Q;7j<2&Mu2l&FqWy7IRR4AxAyl0ln7t0e^ z$~u;Kk7q$}04x@l$sTwXrk$|7m8f1=HvBHs9Wu!55VZNdd|xz#-S~9p96{Z}7W|EI z7$Y=sC752oiL;C{=sBG%J${RCtWn{^0EtY(t5|BOlZd80ID9~!$i$z3>z~g6QXOM! zq(_|VM7@wPxzOtoj^LAy=3?w7^()<{LDdmU-6gN>7Vg;R${_ptst-Iq>z_9qG#aH! zo2&-hC}b~~l;w*v!KA8Kd}TXYIhb`_P~ia=IYpsMAtK-Y6FfCITFL&5rsCSKwxTCa z1(<&MFDU743wE2+5`Fs-L`R6@K zFmTx}>(2KUycxSk_iYe?*P2j!&U+g~hpu!E(EuSfm!~HyZa6ncfpvJX;&`wp9UrHB zDlnB2104KAqvRCN+wt_{tvDIRwymf$kmINI~8q~#a`a&kQ@)<^h)Tc+Iyb-H34#k1YZ?*h`jz~W_`d5SieP<5QOY~`bErEdBi4=Z)1Wx(USxrmbt&NG!4vV(do?2bR+^3B_Zcb7nT1h56zHG#&nEPeVCkN&n; zy{^9)LcyLj0;Ajc`u_aegQ)Y6q&yJpVbpIvSbAwG`7j&#@Cf7i_9=vEf*|asAP@=+Tqc7ownaxEcvEmVFVARLAZ!61GdZw@E5=E+QIu8EvX1A z<|!Wu1O<5%zatdn#9s12wBI?Z*Y!1?iJ}*)@NO=O0|!GGV zJ!Yb0)RLNs(CD^l2kffx;mkB>^XP^;#fk0?il$8TeDEWaRX$OdbtsO2$%#^qIc=-M z(6K9ULC;)ucmZywxfh_<@smoSuJcAOP`I0H+wD^e32!k_)?GA zb*;R=b?zJ{AUv}R6wWu`cE~}p?j?n?3aBp`VB2>_)f!GGt#shXru(A?i_s9Ryqanf zNgZESPU@BTzQ-Cahdb2}#fD~Y=+hXRS8woP>+^U8S#~cFn>Um57u^8@u}1gg%O#vJ zmoihz)y0wXgCyEUy?K@C@_~MGCAt`jDc~3B&L0TNk4kF@z-PW=%#(B5okfY3Dq%@` z#BxcvvZL<8Xs2QDn1uWIbM$aIIPqHdFyRbUmbg`QrFO6Sg|=ZGM_@X;MX-b9nkGK# zAxxogb)mvM2OWT$VM}Zch>_;q2?r` zFlA3I!NFg)N@+qt!c)0qYCpa@xx-UrWm(vK+RSg?CU{A@%O2w|W2EXoPSdPdJ=T(B zxG2|JtXHl$mLRvb?-Zy9WlfYCR@1mp5iTq6h5O7==*gqb8DVgIxN)JjZ88;z{S$D1 zLLi63I5QRcMkQXsmH*P4sV%VtxO$QRfnHDc3oDt~R&|1fNR;h2;Z>YF(FJ4rZzDn}kaJ>*@@vilxh&py5i& zJkkhX+`iUolJnRBv$Jh2L%jdo zG1d}ZrwuXeYPzX55Eqryq1^xbC?GN`J)-mh)rSWqxe!3Dwf7?_KR{~6)|>r_Q?4-( z@Si9i>;I;Btjx^J|2M^}(VBMJWJ~S6s9l`;T?XeVRTyD6;M8bB-f&RY(0sT2OG!yN zN|pKA)>6C$?WlKaaXI&;Bg}^YXgrm|%QC5$$}$iLvMX-qqZ{IoK+NM;oZ~0^w~j6R zeNk1AAFjA1Z=_#pR#DjrMBKKO3u@zTGB?+cdmqW60umEiGS@ww@%nnv!*qcA-S==4 zh8+4yZyzB)R*{J+>dau>Kc%`QbI?tp_*0DtD_uqW9J3=T$bDQVX&6zsZrK7qQ)(i49C~+==8`j^10U9QJv_ zTrlUogLy*(kGXVKHasbar}+A9K8>G*3`6aSv~TZmOoOZGem7n$^x@74KRV#9W82Yi(Aw6m znKX@7z0Pw3bLxU z!vK?*D}n+9A`AF?MZ+-7aRLwv#Mklt&~6(ME9lj4Qdh`YmToyIheu0eIw7XC*c2id z8tt-J^ZU>u&}duU(H;dZz)wNN1c&E~)Fd?Ynb}<+cK)I>j@6s)lkAJFsX5VMNFOIB zE4^|g=ua&BRK(tsl8ka=5;cmXiN&75nJ^l-obm~9xe^N| zO8nr{;ue=h#A2v3^5TtiqNwd)8#&(JFpqAjsgw+`4BuxMNiW!a1_WE|{_;s56A>T2 z-1vx?hw^=XMbEb1hD5A^T1E?+EK&*wJ(7Ug)#Iho+weR2WcSU0C?Cfi$ zmuhRV$>o79MQ-YVNv~R%DkL^8Gzb}W183KFnt1qwD zL{mozLb!MzI<^8*dUP}7&v0l=Z{L1;gJ`;QwQj3+pei^t;!hgbkl;ylm`%?!p+r5; zl>ED%f?uquNY-B=Xl+_kBiDkp-_yzou;^yft)t2-5&tTH_|sCu&rA{@fciUt6M~;i z_%Yk0E2w^fC>pJth*uWXq6X0X*nfHPF+F)uTtSEmP>)Aufd9)}hV`0-TGU&JS;X6b zAi2mQa_IMqV^Z660ybj_@=7389y^=zD6|L0j%WuNqpzB9u!ghgcM$9Tm5ZL;`A&Gy z>@1F%HUG(Bh{k+<^_jBS@cFBD<>nFmyPr6u<$ZLU16rm|9OS(4t!Gl z)4+_0Jf5Y*4uc9$8po6&1Yf@>c!Pv39Smbi2Sdqa+uv9AS#YPZNi_Gv477WNhJyn4 z&n}{38=C|x(8})egENFbr#t``Z+HEvJQKdT7Z(|=|AJNgODz2vIR6?27g8gR5SSBq zRJ3=dX6`jhdJ={fgwG)qoFWC|juGMZ)9uSkTz-rhf%IQ2N2&q-Y;#38pf9i$}Dn;fU;it);x}Kjz_>u+QALN6wU957q=mT zIg*c|K#YoYuia<>fPjmkHsAs>1UkOe8e;tEVK-F%_z(lyh7g%(kouP@Tila0 zA(`ROed0aRg*K7AFYeZ_7%5SlDXe#4o{NT>H3D0w*geK<#r2}E99ReG*JU-t_>U!1 z*WSLL>^15U#weO#9G@ez0zX}z&_gK#u?a}n`y<O5b?DKsoO)ZlopvsZ+rE4Wo!B6!R-6Lu}D;9Em) zM$H*kbSg!UvlY)l*fY?7I>4j+KL=%SyF=`F-M7j<)$b~WG@_A{2XQJc9txr~fWg<~ zcfoYKG>4Je^HIuL&a;2l>ti*1VgLEUQ%%jrMdX6(x}q?@TQA zwbDzH3U~imno2wGTWVuA;8$46ms-%#Kl3D}qTFv-mG;3hdT_l|fwjE=wZt*`RvC1H3ItYt9h0|sPgry_m<2O#bzC+e;4WEoFH1>xAocrm11_A~hC@Qv>nXz<-GSo^!+7Z;qSdp>lc}JEkMoA7~X4!*z zkkNx?PHud~0PYjBCU9ADxbq8*R2BXxKs8VKK7+Yi_Jkm?HBVYoND$QbdqhI-RjLV2 z(?Or>7r*veTs!fyu@l!pXJyvk`AD;)(#vBQNV+ml3#uz?2FgMJY3A$4x#hrQ-|0H$ zk0OM)pX!|jftCi^z6vw4I+tWg?KytoOhzON;^kAC44AksQ>ig!F?f`C!Htvg2oEpd zr8*@rL>%K7G7VHkaE1Z_?R4>CP#?VINFj^P5=EvcA8a$A+E6rW0qM4$EzU@vY6xaZ z_@MpSep~m?#?Pj%YVRI3&A%FofWbUSIsWCN7HO2eP6O2ZJ>)>^qz-gK{}wP4xSXP| zT`hqlHH}Py{2GfY8)hpWj%G{*7{s;g!u0#n4vF@T09R!e^oE^MeCSfVq*y7)IS?tj z>+`E&R|@5qB{WS!w5otO-k&aU2$5+|vE1ziB*G*&c&~ZGnAtni{>F}LZXP)DkwT*s z+-=^Pj&Eb>+02%r#rf>QyCez8fg1I?uB-RA%(hUlQ^qeK1FqVQ;-ZSAzeaXNb^YY= z7@T}7-r$G#k-3~kVz%Ko@oF{9Jd%{^(|UCF8PYP8iW3}keHjpaG-*CO0|B3RnNGxW zYCM=tsx)EeGQ%mUkC<|!s46~2_cJGP1Qw!qhtqEwnK(C-`e0BS2zPNr&hM^!4SZR0 zRfE&NWHs9)-dnB~%Lr3&n4(O#tS|Nuq2s2rSk!M+me2X=8_vhIX`zn;f~}=g4YLo; z79}todE4z!75MI#E|X+POHsj58OVsO3pt zj;7>+gBZD_10?TQW87)V$X2+LxuiL%>n*KM&VhCoxnkk;Un<;Xj~uCTf85UD{+9Qc zS1IG4(BDMF)sNY~9Z;5(^5rWoq5)I8_NoeuR$RwQP*S{D{Q2i~%92>{1gFa^f&cwF z7=Pk%;WX4&H@p&B3!3NyT8_K8yhIRZh+<%=|iq3gyO!;EN~dH{Rn znHZ-!9)Kbg_9~Xd6-iT9*#Y8bv(OdpbAyyXcM!J6U8m^h9p)>W!3Nqhj;Y#k$v@4UIq4fu%BL%gmV)$`yO)1Um$+Z560vM{!#&F?xXxfS1r^-GR39+kfM7Vil%uR|B0K z=E!qMY$hKVPzgPn)#6l3bOw7{d2rpX(?6z3u!k`NGk zd7^8Qu1UflbMW#;c@FaU(8S^6`#wE%^UzQa>&`00PBb|qMQl_3b%K&+y3x)f@Wuo`BUMh0M*T&rTsmJ>?@UHn4PvcN9V`fh*mk z1V6_L4l(qlCcy?Yzh1pno!iC{Vu)G;9t~L{1Rcc|mZA?RQ1<3*zfzv9_s+u+r}dq8 zh8xAFc)gItRgWh-Pa0CttXA~q2VWhT4#n^yuDqxa-;PMpAqg4(a;WJfAsutSY!~}s z7e@!S|C6+R?P8-2`zd#T-JfpJaQIAngtoMjS)hdSRPknOo*>UrH7Wz|VF96{1CwWt zht_(A+cOqun&-E~unIwMP5U;GhE|_sW$l~X;_o8S8BvITF|a6ti6>f|Ai*4)%zn6A z;b95l``OAUBRVR5LSqIu?H;csAnX<+ZC<^lr4DSq{ND6Glr$q*L=C3)5@$8B3&f=S z8iL~0>zKkBld=|D_Egg-*Ea8O7p?B+;w!I^H^(!%*2jbD1cJ!YWT?#D-1XZA!44cT zi_m`K&T5XPPWS=k>D!2ywn}||*W2*+?q?W*3SD@+r-qPQg2y|*Wo#jarZ>Bz@8)8N zslTR149|r}8*a}K}xOPKa`%j8+PeUlaYkg-^fFH5R zr=3}U@-BiT`5${p1C}Jon9Zi$ed@hFPv_C>$F@XTtl8o$Yb}Fjf!?YnQH0Twy@10# z?#gTURB164VGH!OO7k?JmJ6XbkDSRMtpZ_cVEYGTN#9n2en5_OG8q07A7%UBe3XTQ ziRu5tM`hgpil=UQ&Dx^tip~(347-Kr!0@o3%A-Cz%Mf&65G$=xEl{>H*$}*THsu{v zpY&iLB`(Oq^4gzg6Qf1WzBugo&>#BYr-ty*ob^85?BD0}@bJH$1RO38KF;pn{|lZ6 z6}f5RsfrvWbMt(-BZ`d;%A-{(p3ikw&Mo8T>fncC-01RnIq-0QpagvyN2iKF&KPa3 z-3xpo{|_65w(_YkihC){q100@P~0E9Brv1Dkv@T;uaV3d#*#6EsZxm#bJ9sJsix`W z9JjOgm-z6y(`#y9*6gyXHFeUZ2)A9d?fhoWl{q8CaPVGgrM^~6W`fkbm`O_59C@YRvYMngWYAD^XvvSlFpc&tBmcGWzuD`QaQZ@nO?jt#kjq+*N2PKJyj7`mxgd#z$kG+p=!G-M2@GgI8{tepr5>j^|Ec+f!wk=o0 zY|&u{2Jp*^HMQehmiL`J#};lLF5NP!C}uTtZQqa*>bisBrrqHiDjcMEtIE|KC**Q@ z-{@q?v+NDXZ8gbAUaq{fsK2o@L;~KlRhTHWmrmej&XNz$e9j0wrqVPgBypRXB{>);vv{4ld4dkZ-fJdW56+W&MWN`Jtp&lDIpzi+do>W!H$_4aA>dQr0p;_^1nN zZuaUE@`A8`4y%P0ff_QtRd>BY$rh9}ibIuWoH932fx?02sam)ge@7S!>2^1f()q`l zqy}&ZpW5LC`WCLZuFsy9(7d5-O942GsWrcGhKYNtTV-~ZtH}AKh@lNtWeGZfX3Nw_ z6QZu3&&TH#Iw@z;#HMAsGfjYi)Ia@cnwvx5eLK_>!E6SBmQEOYBtuAH`#RqE)K?Pc zxc0b2en~7S?YnHzb9pL9XUqb%oalJUm5#t;LO|RRFo>`n^k^J2?!Q%%sPxUIv_(h~ zQH--KJQ$;qXbq$p67*2`H!p$o)WetujpYwTnG9}IH#l~WPirE3RCFM#ArhmC-fC z?sydyG$?PR+W=Y#w-kJ5vtA`Pc!uM_a&CAY#jZ;>424~Eh4NAimJp0t==Yje)ZSVR z-r}AOm8S8*$!g6^ZuPqk@m>Q^db4O#?DN?gQUmwpZ^F7}dQu9R#49Y44$?EpJJ%6$ zcg%HCSXMJ3#=rP{0F#~z50f83WhSI=}!zrHzrah&diixul4H8;r#RVz>&Xn zuDuXa5>vo~-k@P|xa?=ZqkcZ!1F-qXH1Le@O*(QNl%NTK;@EyYSG_T(_$nl#m0h#3f<&Qv?zeS;+ zN@HU85+;2LZF8>%jm691~24_Y@R|e25G7($gH_NQe1a zibOG`H4N~oijuUEe>FLGcTa`%goNRwRE`9$cXG4wXZ1++nN_>o6F;X{1;IM2*4(B! z^+WUH5N9n=V#*=}DEh0IpG(pLu44TT#>k=i^9LxkTr(SIZh*U2o8Et&T9CqRze1f7 zjc~=TBm>qGOMDkv^X`=heTIU{8n_|+WEsQ_Fn2eHc{UDEr7+3nxB60({q5+l3&Ng7 zWt5U(wG+OyKBiHJ72Kh-3pk+JOeW;p@RmzF6!GJ^wG5w}?MsF{ec+hVJv)iu3XmAQ zUH!;@EjFNYmdCWq>84#V(G^g<7&INIh;}!!JhZzNb4o;w1%P;oNxI*PhYDTOycEWR zLQ}E>Yff8y_;b-6vs0alRZ{)aN9ekJaV~)`9MnBMiJ9BwtU}6vzYkV!X|sHprYl7) zauBaiiZdp}w`Pf?K}8|U#i8I#ZfwXbNsdduk<0r+HN)(#qyKa)T$BD5U?csV@&|gz z2^?u*R{jg2hoY z`v&on{3%EO5Vq8NchdUk>;B|`#)nE#sD?V*zSyJG1I>gDeVb0UG$aHC8Ml7PuWQ{C zD1n8Z{Sspj;RDhux;(OXUp~p~et93clWP-!d>z`wS|GrasW!B>b=J!yAe?@+S1w^Z zwrD?@6dGMXKFq8Re&+kD`?W+bNr^Bn{3woIhLlE@h7OVyj`~{X5NimvZtM)*LA+Lx zstzl@BvXo+v~E+3QALtGvg)$p**tWC(qA^;d5T~i^G_gK;I0dSTPpI5HMx3Gc8A<) z@}3NIMZwr-dC$n{ZkagNujtq+3ogTNfn$zQ{R9(YjFQp2WU|Arv4cVt@Ce~sc%g56 z!>|L6n$okfKiWD|JhbyH<>6k5?5#nvnC4OHu=_b!(ZY{o3j-VJY)-TGWRZ)WC$m)Gi2>uwHSk&7_=!>?OJ5I zO9^xC5m>Bf5v?A{p|t$+$sz+MGJyK<0XJzYew@5r>G#3d?EER5)ydx+mPhwS@0WX`pm{aFB%ZLOZS|5d_)a_X>>p3ajFo6PBFpq|L$pA>_W zN(1sd_Ph6@$bnAtS_gYmf1$d)d!xDr;&sp(0tQt9Y9R0YQ+lNFH9lH2ByU&&Rga;n z=ErRp04d#4^QKIF-`(=J4(95PIxodj4pnr`AxYVQzZUAM@zdh@5^~4g3_f;<^30VY zD^v`D4XH|*lfRNvM>f&oEr9R09Bd>YJT+HeR`r#jaxZj12sVipj#XO;)Ts^@a{>ZS zVMF~S*OPeOq62@NPPC9*7*~hpnE|xdT}z}=9)#XgcAO-z)q@VY1Szw;O0aRWIw3|o znmO;=F0Lf&E51%DaEqt|PibB4TX~$op6WM(nQ^o*#c46$`)`xMXa3JfD{eUH^n4g9n| z9h#`m5uVoM`jlVytYjz`vA@R|%caxzO0jCM&xz{5qBfmIpzm*e^)T;oLrL`oI&Q^P zZOLIP?oYO4QnWvL|MpPXH*&ivDdH;_DvJN27{jG12GRWXCsA6O)8y5m*OMlqbX(qp zsrW6+^Okq>4l4h^b=VQ;)*$;|JR<+USnnOPuXqj7f=_wC6w+}yUx|CNg zQH;-W7$wkK7L0^}lFp6}QEl%HRygzSkDcxMu`4}95j|s<_grs)vXROAN8@ZCSwfo5 zvN80`W-i_M8#~JQ>AG^k#>u)vrJIrJU@mXXb6i{FPA{3`uY)DoaWRJ@38Z=B7hjgi zc3g*P<*~(EIKP)=?Vn4<=;eYr`F>mPwCmH^!Z#FM{!7>>>ipjd*e@ikSj6Sq(rMGI z8PG6ZUOP`qy~pnh#r`>dAN@~fj1RhknTiz>JgL`Ou~(7fHsGUv7*b&LyN>pqCJVIp z#kQaErbtGhdP5P5q1Tn6o0i#zjuj&^RN#8wa9LkQsz^%$?zLgb8pK>3+|{>hf46`E z3f6XsZi}wh?ya@y1_HYCovh5pUsqOomtx0le^y+GXpM|z-|9oOXxi2umLL|^^PN{w z09)>I{`%rj8N!0#!8x07gLFEwV~F838z{Vwy~cve8~wriH~qoYU}>Jb3fM z+Acpn8ldt6W29wI)IV**F{ut49?Ql{)EGol#;@X438fL`4vkM=t66@G>DxnABVcN) zxRXFIzZ)G_YzmqTTr#?mvxK-Fp=QX06I1}FfkeY)^Rgm$xZ#=QwuGpjfV_Z#o@{J^L zUeKS}mHF;EcX&iU?kWgJ!CGyjg&B{2s&yO3-{HTW_C9xAx+~Z*Y`n$h+&z7^M+75+ zPnOd-=T;o#jonD=?&w@R5mx7IXl7>qO7PTqsh_8YA^&*g!_o(ET2^D>A%GftC^^}1 zr_w?S7YDnwTs$63$5XIZ2%nz`$@os9U0_&!D>HE+vX}1PUi0FRVoIH6sqcpmQaikR zG)Agtwez*`k?Pm$~XdUdVzN*QM-~~pK<5*?A zMWd1L9A~*joaJ}@?pTn7kn!3p3?o|_Pn0>*V3z+gzD&;_4O&Ye^`lbz>7|Z-kvADN zn0SUSoTIDy1wy?tO{NaI z!0MiOdt2_44#{Q#n1n{rqWuAFTg7Lo4`umE#kF3hK^@oUv_npebAMmNrOZ;ViJb+9 z>LXt9HK>ztX!;~NZ%vbP)iyd$6>@eob&FP-Gv`_ zm_C!!;@QJ(gbu9|w+>$}><-n?uM9`Y$u!;3I{a$8)KrC^?;MHGRiTLTpN)+T!+D!b zPZc1dhZnls9@XEk6Ri;)3@=u@1}8=-#JWNcD>qp-Ih6(`tf?cr9it+25YIl?7p13_?{&KDlZ|pYY4L|`lS`ek%rR`5zPk7?4iS!ezuJ?aa9{#{M4w89T@S z;{##0rj*mBxMR=5OwwR1-`Wy9Yp`dLGs)AgILVB?#LSGSI1YS0j^UsnNVhqQ$@$Lp zzj7Fe>Jw40qS{tUeza_7b(zC>~jQ<}$4 zUAzXR9f3*(m?U!>r86yW!z1n0Arsm9thc;6ejd!3U#V6|*iBZrvgDeP(X;4G5=z%B zJ{~x|ec+-u{0;om7>bgv8rCN;nG{CTwU?RC_`X$t{k2!eTWF~*(~5bFczhT?5i8`I zYg<>u{5TRB`Yv(AZ!aNLv)v2Kb1>OAaL<%GNt3>^&QH%JIGy6DtQX1~vU8-F)OV*Q zQ1Njm-cUY_nLbY51P9YCELZCuUi#A4y(&=QKt4*FHBdC zoZvU+MHyDJ6bSXn*F3iAaCT_W0=wy4J#c&-M`toEU6NW`SAUn?Mjky;dCQV{X{GEu z;G`p-_{WK~#`kAw<0) ziE!hz9i>Gs3T~P|+WuZ~tu;`tMWrYYT8oy*cDX$FT|vvj>#2Xu@S#?ydc+cTZ`@Kn zLhfknqR#SU|APtNIolkeR6tG%-^e80F&k}`cT(4O$ip4_6H+i67)-86*L95nu#u#m zWBW{@d=UyDF(wgn&uLXbNrm16sqO-y4$Is2lz4ia#&Z3AJoZ~p!n-1WP@AK9Iob!L zN`N~}wJDHCgrfy}4GJ{+wPWiEoeQ0DmIHEa*hjCKbPqMUE5j;UT$33QPwaPKQy=#y z{ik;SV;fr9!g414H&*M@bFIhM1K4LMVYCdGyNWr~JLia)?8t?<(%#pG&xP5v{x|J3 z2yjSgNE$K7T`;9M;!|4SGK?i4s$61t`!Bk-5M*lq-JG<8+bM-)ynubc68<#NXndux zWWkY3)7mvR6N-1lh^hLT+n@wHyGf1^A>I7;nM-CvUMpTl=naQ~EsyqcmzP<) zI{)tAMamT6HGXJNe-G&Qm+2u}U1}tiIEX(DQ|kN!a6GD~W6%X#j+_66R#f%TYVP*q zd=^rujf5s)()`1;yWTqA+cz>BiOv*^&>g#FgdIRjg`TNzi0brji>fa}IJO%-LpmYA zL~MpC`#f{{D_en5xhQs~92E|-oruRQ@kk|5Gblb#SFf4@gMUv^62<)O3%;b!oe`!$ zXs^ks(2e39?id)}L`W)rfsxM;6c`D^O(aHEgnr2Gfs$Wp`9t|Hoh26w^*d5f1U&~P z>@2=q(5+p`S{b^({q{@-N!K^_Xle%Z<`wm#nGeX?{67a{hA zS+2%-wuHjj7;@dY_!zYy%IzZ~Fny-xS~={y|6=T%+AD#!ZM$OIwrwXXwq3Dp+qP|^ zVw;szY}>Yzs@yvFe)nnbhw~TaHy`HgWAxVED@dt#VFvo4Q*0$@Jq-e^fCD=OXB2?= zn52GV3B1ld`-p+r5Cldvw>=*L_y?f1EqF2}H;6r<8POa)n<;r_K0TkqhK@|v#OV;x z;~%VsNZyS=L#)r7>aCwK?pO$5BhJU4o;BX5t>)TrL$n@~3cAd7I2 z+zYp8KPl3_8ujjz^z^2D{sRAk)P-$%C8i!Yv(K-!K5!P8gGRb@5ZPK4$_202W4c36UuO5|Q{Z@={Z}<5hUg-k?*ROo)v^ zC$c0_x4WQ;0s|%h7CgnCyFWr42g_|>QPKIB5iMSb)kVjB5~~j*ht&=otLHZ!%_N`3 zhf_&iBRB3Q0W?1&Y$eQ~ZNpeh`A2{27Agjr-E>2O=G+dIg&{oG&XZV~t^(n@zrt%D zb_uhyMV@Y&LL!e;rQ~ds3YP@ zy;LUsWE3v3F%k;4_Bp!?awILzeBfd(QzZB!2-(9|*RQcFS@7aSXSD2TbAa5bjfxlx zT}5;^Zm7@GbvapxB=OJUnQ9t?AXpyo#|&qX_`m^#p>z$JzYo8$04a`OrNu~Bir<6p zmR7T=50tWUB4k`Fhs#eGU|e8&TBVx?fE1Atw3DFCd_!-byWVOXR6%{PJ~)w^hPE-= z_UWgGS@$q$NILic7sP&H9l3-ZG#BH6$@RU&Y;aWq3?Wn|jGQT&fe8RBJ{dLK-**a} z+p+u1B!Jb76E4u1a~Z8lDOP2FIr1+RHiO-GVu^}~FHf*O)SUu1QoKUd&xBqg6QMcg zl^*IZSjo`6+HXigl+Qp}{hN?5m2R!5$7a44&L@#b$Ax+sqpHM8mZM5NZFvjT!o?ns zLIY?FhEnnspjj~lg<$PrS|t~)-QbX4oQ$``Yx!DQ`}1A*rM9GKND#rugb;mvm)6$32@QFTSl|m)xRlC5p!8Do`*9j`+f024koWCUl-LfPn(m>f@JH7VZuW%O%2{ zZpi2(j&k8k!`AYWE%Ec2L5rDgBth&+A5~p1Cc9TJ1$&z(h4m{&+Oz9bKf{j-WAhK@ z0jiSC3FB$RZ|KQvq-s+9OyTg1Dugl3jlol8kamw_JH?VBUWW*a+Pas-kAJA|cH?Pr zd6HBRJMhokAVH#Fr)D&DH=1-_&>i*SWG0#TNk)mi5zm8eYgb63&Vyfa_SvhVFUpNIUSVo9Avx!WJ!Z5ErK$4$BL5Gv4jJw<0L zTSr`G5bTr^{?+#%(4ISsKnK0k@i^99AWYs{n~pUu9TtEZH%l; zlU3=N7-_@sY_{q|jIVFGkn|@(P?96@{o>P^lh^-#X5GyQiD0#q2Gi_k)_6{hBR*# zB_!4Ah5bkEwCQJJs!0bhwa%FSWF5X;$l*9c5wPFR35j5Q?3gH4{X+g*eATr*-Pfy^ zaA8}N4*J1^(|BM|ry#K^(!usXWVTh6Vo7FtSMtY$zg8?A`u%n|f1yzcAj%bR3y0x$ z1`u>!S=#;nfl;J-2lJ8M$mLbweLJx{2X+6$w=eURtGN0miY=C2Hgp^^ldbC!Fe9av zkequQul9J-H^XIkMdB8Gq54`Fj><|Q*4*7>XDyRA#I}q$MCO?(R0KOKRDe@06DbV6GzZ-%ZG^uw+5FAO$ZdoMWx7ohT*vAbAJjf{UJ-6d~?7dzEefJMa0k^$nyTEVYzt! ze%0N-lS+-lo2WStO)-i$tZ?DT7sXhvckMu&ee8BVYh4v*SaIEFZ;a;#jsHhB#4i2s z03z$^Dwf#y!6;S0;?-6G&ux^=fwa)n-#)8?J%5^XDG73sqW8fcl|3BJenia}uz<5y zX*W1B_e=RiMM$exlJCb@(GNsVS=kS^I=VR`{o2O@@H-6$o6!w}juLlI?lqLR;9Y<5 zisMzmBg}n{x2a{`rJ{qF$pgyzyN0FyDQZtB(fk%^E#LC|W@$NDZ|?4xQ_OM6P}6VT zicO(2H>B~HI}gx5$8aJe!x+5Pk3T*K%9#G_Lt%<}?2kw}k#fD^77wv^Cl#6r2y6cI z8U2jz=1Xt-mcU5$`~#?m#eS`H22Hf~{X>jSq9(&v`h!ESA$ZKsjjfAf4s^(~_yn3> z|LN>XV0>;BGVNaOxRxP$x%WFREVh}mf(9XvE@-dj*X4#sWDr$LIamnyQqDgi@@MrW zlljh=(VFC0i_Xa=$tl)ik<1*<3HCOD$}g>A_%mo%On%iyx9iR*;thzvfeh0-+8Ote z-8;f^2huFG7<9H)V)Gab_nr_LC@K&~Ws?LDm|m#>YyoibkJx3;fJBWs%VheLX!pK$ z)BdG0sCd%W60vxw;a4YQw39vbMa2VxW;Npk#spD5-e&s?LM&u^0sfyObakp=du~yy zUU~Ho!=zMEGQyr{{ii_0rb)E+buZ3g9ZdGbX}U^I_id4ya}Yi9Zl(9z7c!9o+BKC`k;JGb@QvgxV5~_mxspvOa1`w6u}LiUD(S z_jh-8XYYH1;qBBXVnDlPgZjgG0(AhvHCMh$DqHJTb%IGG#-qUqBS`JU^LKL8;~xbt zDn(=T*8ylmbJX(0UfvQ^hsb8u0cH)P2Rhm<_{B%(ABlnpdN4Mvj>a)GXyR zknzBBITfQ-R^>cQu76#iQVr+F>iOv4D-8|COp*8mj$Y)r;spa~5{TTT-# zmr1I0ww2SW`}oKMS}PV)byI$OeOxA{7aN`pvT}ojt`L!8tmfQ$bV)W9GzhqX9z%|^ zOE0xWqfNuX?!cC44B|j*g{L}EVqTUkJ|kw?J>~GYX!xwpcB=N#{)xYp|yW9`M?9MJRwJb+1_jh$`(&SlcU%qM!|~E``!qd2GT-VZo+5xja?gl6m0`u>~ zlmpF~D7~S~no4Ju_SeR^rvK3so>p)14lA5=E#xnkhF&n}yIHQ0Yp0}I3cCiryl$R_ zCEthR(Y5`}Y_S=(Y8mRciBUh$bm{>H1!V)h@o%w|Skj6^Mk_a{LC@OAkxJpAQRIAs z*nX0yZ}u#-T>4XISXzJ8Rb)&05oTU`%LB8`7Qm9Q4FWHwg4n5VAWL@`w^` zf!DJu_Y{?o=TWfX3Wo={;&katM0%DWabEQ7AJqi1l0?ev*LZWkw6&MmZZdIL=*1u& zrFPMZorBEvk2O0))2xQD%&ypuFe*qir(G3gWSXy59eRqETd$_?3_R~ zI*Jx$@}T-}8GN4dp?r4f$DqA-tXeqMl;-`z#PbX0LrTdY6q6~OjN=}=#Y&=cPuv?9;7`A!fsiu(XZ?=bm;TTi3CyIwHTr74hj-bQN)y@3=I z`wPKPecMQfbKJp&lN~(YeyV{A#Zf;-xT3_xlD8$Y&2FuLHc`>+j9xTGC+-;lS z8pZF&2@dP2i|HZd*w;i%$nqYVKe^VSz;pzxjZ7ij8cOG&dXAaGa2J@-w-iLLm-k#l(i*oXMy)#!)(dAbPIiqk z2D!-$NP6!No`Ug+s_nl6s5Ek%m3M@2nrTkae*VEC70uwb;l21GFbt4KD~0zCZ5u%j z@eIky#usgX6XZ$2Oz_4gB))+WE%Y$7KF-#{8qIXH1-d3?gU(wm`x`D>Qm@Aowa#F5 zqrm_BuKT7w+qM#scdq#!kIP=Mzdf@_hYVTfq@3GgwGvru_?DvRjBU zeO=bjz4P0~3l!9VY7V*;b6AH1PMF4S>4DcE30%sqi>_oG#7-gzMYYhXbynJ0m~d-w2r7oc@M zj4(cicV^Q)NK9z!?0_a(4vVIp)AsADSQpfn<5*sBj|Z^ptp>*scEuA(|A(4g3wdO6 zmbL5vt0~wxow-KxD|zsImtNGwuy!eyKC}{Y262&sMOc&uF3w-!+9*;9`4FG>C&L*g z=ZBdW?zvp?-kdzp>j{+u&9Ism`#rPklThK`-24z5OyNOXp^vSB=rSPZ1S6CtKpe5& zkNC&E-s^gd$8>qe#CZwTlK-;b(t#cX;~=RQaA*yAzEIfIGj; z`=aqd%%N3j%rv91yFa%vWE*+|;vM%o?K2}1#LiqWi|p|+Tb1f^xfY*`c{vmc;LKE# zD2_WvVYdw&C9`MQ8Z_s(d+vRS7KfugwrBaj4{3Zi^jgbxs}))o>IN1@172itcPmQ_ z49**_@{C}7Nl5tKjtR?A(tcR}zC70`gJL{C%rj--)d<{{sMI=f7CHpf6zwj{A93ZP zV(Ez`Y6XQGV|>#nX3VS#81!u6H(3}>fwb-D)GZ`CZ9Y>@sfw}t6TcTN2-pZS>ZkCoa%UKNLrXGGZ%)PwY{ zf82!2u{(L2JDM!NME7!Vlsjk@ze`U|xfQmyIn8P3ZtH$+_`|5_;NGoxHfI-TWix#b zrKYH#ER$+SHpJAI+m5Mrlh`M3HdDWO70~3`GcJu6-S=#@xwg*THPRkQEq+Y8Lj-&J z5trBaYQ*;Y<6~2$`z~cz6Psi4c&KH_vuju}$A%pD;qxs)D}<*;D&U3)0~tvh3J9J6#HLJ-r}am#NSA?ioi*1>qj|IfEtg7IF>UjT}|R zb^@DN_iGXAUR`;*D5gygVRq8kFCig2mn5Wgp(urqBXl$mtA_GKPZgTL?#ew@ov~fS$(o>C6v8k_Y9zqYC95HvU2b2Rl;{wnD6D_9(`W#+C zG(Lw}044nu7fOs;>?#oDhD^!|sK5`dcSw<<*|#B-E>&hJ2H}*+$w{h6Z2Z;>VXm<& zmY6c`#Q-RL%RwuamUIuT%IL7u;;20k1CIA*=#A)(J!H+4y<68)KrwDW|MccZJ2-Ja zUM+bJc(7owR}gir z0FC~U{4kF<)_BaP8d{xi<5%dS^`oo(y2>>5i5bh_mh9p$a5l}k%z$4XT0)w-(gmT!bXpIiDE@bWRhPO_s_TrEPb|K>JR8V-!&y4=}$Cq?T) zK%ByLMwE|1Ok^f*_DjP^7sd^ToKWpQJ6oTE6HiB^GyPe2zFr z;dq#zE$!FiuMCmJ2RFzx2uw8TH!U!Km@CcNrAhoFwHSck9s~`D+#uT&p?#$*76T>~ zZcBxjpPmzuxYMdB>ony(FCfOci+Z@@b8`nw2hY&wUiW zDjM>%*=ONfW9yXH$r+tw;o5o$O%{QN9DCP_wwv8siRsa7yIFSg5qpK1gOyHqZ#k+o z;mb0vm1O(PypPnjwH*{yfDxt)Z$#+%tBLXAhIG0duQUTMV~{e-0cp9WHxok_W~4R- zG(;GBg8U=>92JD$B|G!ksn5QE8GhULREwLS3AS*4;^|&qMNadw07C8(_W|KF))(0|Gf3N_xiEvWv|%d zii1&DnQ>0_B&d49N(d~A>#^s@!5(dj!H7%`4+M%gUC8QTo0xb*CKrbn|WeZ2=vpOkpC?ZmuY2Bkg2qsgPxU>`wI~h;rJkX*ziugzkw66(d zskOrD3lO-t+TP5RFRFs<#Vu~W$tY+on?!EcgU7Ua%Zwf>Nn-_oj!M3Z;q2lN<5u~K zBsR;3I@T)f&7HAQtN#O;R{W@{sD-Y}uWhHRd(>RUSnU>BOq2<^&QTNG(W8g9&T(oT z`?SDS#@_{^hjFn+Q*u_26@IXq&=Wd+0GAeLk0H73A!z3*ae>0bp!+rIdwC-cC#29w zEx1M$3po*tq;3(&x$?-bN}vD3^QcBwDCHfcs#c3yvOYXCfnCu2F1o7kc!LO`fb~^$ z&?dMBzYwGR3iPOE0_m5wBhS-pJXye;dlLq0;s6NS2}AmhBr_r=E?`bGyPHHCm=R3) z`K*Q<6lc}`U>=2qQcajn|GA??qc&Eo@hdcP;rnTizdYBNE3)}-CjrU8=irx+K57HO z3cu*ADT%54@*W4{HS4;**C!?AJ<-vJUt4$VqX+ye$dCmwveU7C4=-*k)!x;B`fELV zqRI1h8YHuNRwiR*Jj`D%Y^h>Z_pCKiptY+5aq4P-ox@e|^^W2U zPZXTgvW2gC&J8j%Y-3*G`$hv@Cam`^g7x5-&)sQ1^%j$Nrsl&b0LD{)G=a^eG|5D= zR=l2xUL+fTeRQ@x!hMXPw<)(mzhd}nU7_=h=Cz*ib@Hg+D)1fHNAZBLJ%&-QkdUz3 zN@fZ#83JRnm#u%jopUF)vFU@BIsrfQc{c?`Do-0Y*iriPS>>5kf0W06*a5^`I9F}_ z5f_9v{b|K_G&muvj$#fw5C}xRlVV#xO`Dd7>v78RY4ZNGpyfxlpMz9p!B{1pEoYkh zp2rXfDx?-hulnU^py1yv+XlPn3|wZpWz6$;oLpHG{h8)kh1Z@ebv9ZLrUQ*;YW;F* z+l~m!tZm0s@NLdPi{Od!;SzN@ZQknd@5x=C?AsREN_&oaNtip^)XAHhR@?UbYVp8e z(u*o0zzX7V^0+W|uwN3J#&M^iuc97L6g?Gq^c17w-@EyXNNoMe%yeMp-T@)pIztOz z{gW$^V&p)-4F?;ISFy7H2qw(}a;y%y$|U*~b@DBT@ZrPzt~_x`Yx>gmlYI9Gd(7D{ z>SpzBTkv1e8jw=iM`HMeuS?%)rsyjE(+sMTtdQ6+R zD|1G}gHI3Ls#9qy8i>Wei#r)sbNvr(cPWmgLMntZ8?OIJV>c{YrYU#sPE$zTA?h4+ z9Z3Gtcl3bX;XJlN`oZgsvLY5bXw7i4daRwp4sWO3`MSa{jFX+8k>Zf^ANk zw&~)h`x1d|`K!)M`UW54xH9V7)=UnU0~|uR#8v}xYhOMmy{ zF*!Ho@Ba*bgM>NQQ~tNql;=NFQ+6J<|AjWB^?$S>+{wNFfP*u_l1t$+l)ylMZ;Dx~ zB;jv&osIRp1}e%*+O83Ij4DcP_swtjc)zVrLQWw-GDX^wz~#>FQ{+MZD?-t&-ci_Ti$(=UrJZZV|T>fke>1-K?%47Dn0Z_`@-I(v}>1KMg+lkHfT)`u; zMRmVL|1gWZlfr+k+D#3ci;U6*m@GjYBDHGv-)`P+cZb4|X!w302z~s3C^U^|7U`P= zPAEzaL8tXQb3e!T!h8x#49u*{s(in6ZsfahbDFQa%G1|_-n5U~=FT9VMRu%u&CPnY z3|Y$iN(?Yai7}gF0WwD z=s%-0_K>cCkVlQ4pBHsiRUBmycZ9|@lcPv%TQqA^O+TwZEF7~i5p2j{M>iVc%|7X3 zQAcsnqkQc1Q2e6Y9h|dUbc>a?Q#xe$w9Rgkw}MK-)bsOtyga;NM1PW@sW+Isg_e__ zd=cX7_SjiK`k@^;m_t5i;WP$%6gq(69THd*5@x9PUL0`8b@NV5YY3}aklI+~J{#lc z^TI{nC?i?^80CcrhYODT)M|Uv)*`@GMj+BKΞ4`1`c%SPu3y{_G$p<=Z( z5FgY_W`%FrbS0x-9d1NlD%+|Rh-J>xjcTWsuEQFrnu-Igq!Wd5temXZGqCUK)qr%b z{+aMGrneYfc-^LzR>>1*fDTBBh&0*CpU~B1tggudBTwphYqwi>tP4;TR zdMkA{atjnaVZonhq58vyxQ|vB8yK1GJ3`_QIy$BDMQ287$o&^`>*(iF{B*u(2;KmU z0Uc3kBr_*bGL4Be`;n$t? zG43pTw+LK?opFH9F&VbZ6>Yvj5v#dkexo9JloQdfe-2mi=J3{&ofMmd%#~+|kU4&Z zj*^j>rs^CH(ohBAxo2vQ{^7g`|?=l1rOlS|zr@kQHOSpA)zx+V+v&J)PGz z^30=D_im6+rTs7(7=2wOiQR9KMVq&KLDgZf9wHk%$=e7gZ z1qfPsLeb0dc|wM*WPaeazbLD)Yk%{i4}N&9|D;>!Hn#oS%FsW0)5-mQS{!OA0_HgO zRkC-Dtq#*8oqFYDjdLWWj+q-l35?2-pC6S5=KT47+Fd#JuO(vVSZ@NJuJMH=sUd?+ z+(8Hbk;M9eaK7)n**0B$);Tr~MWZ`HlV6gC;Q8;mDf%38FIZM!6b9OpGO{~9q}1{u$Hv~D{L&-CCXIQE*{qMDAXhv02qE3+;M zfm|?WHm8Qoyi0k{h5j={p=glti{T@M_2Yo4R7KKO! z-*w#aoZ13|p8hTVzpnmbI$mW;bVDfzxzu)_9YUn#!Kq;(|MdXY%^F}R zNt|=WF2<5>h)5cAeoNBp4q94~W34COLA%i=j-?D;#(|vXRmTqzvv`aZ&N@FQA5R(o+KK2#_Jd}i3-1T8-ilsTNBQ82bk%#U<*|xoh7;bpzaR z24Uo+7_hg29;t*PA1S9?G;}R2VKPEn8a;l;iMMnqC=Jxf%1VZ-MJkm_S%M-Fn#$Mq zI$LDn$aYn}%R+u(V@4}$`dqo7T=DMFvkMZ6tfSH}Qax6Flxm!p8Zf3Ap_MA}jxFvy zh0nH4@D`rDZ?oGV6|9jStbAX&;$S-;n4B)0;Nq~r?vO@D&;@?kX@+kNy96IE#7=!G z`XNyBxQh`kG0gSML(_K>JVQXl8|ZB{?(t>_scfJ%BI>0-9D4>$3d-8Ix7w7&W{CRDtVoYmf&H%5|{y zDTBGiKd1cpg|L^_nb!r*rm>c5(aGClUa4+4DqI$z%b$X*(!8t)5^juX(_cp|2Y!~sloQg8k$N=BBbZEFXUs_T}j?dcxnu1z6xJ5*d{H;VT~k) z!r(yy&I`heISzhVSpK(4I3hpDt+^QHlXd>@C_07qpbTMPW%?-V6h!)6-I|)%^I@pT zh!-?Ss$0}dIM4;24w9)Gc*;kVi)3)Uf>cRpeMC^ui|?mUk^2LGH@F@h&be64r&dbI z39=!0OSz&$>O}$##S&4 zaER4xIfb?o+HE|j<={{WV8-|xLP;Em16!Cp07YE{SN%vVoW4D2uW>!TilW*&rbSw{ zCEag?<-YEI)i~V7u>zg@My#Nq_>1w*i&lhYUvTQh8umf0J4F^R){J(+!me?J=BOw} z-jiGzWbGBABET?ks$`%#g3F*WRN5wbc<=a9Q_)dGwiE&*W#m1Ln@>O>jsh*+G8+kn z#-hE993^r{uU-9o#LggHzFPzq6Ok&P%5qxH$UXR?-5TS^oo^U60YXsj4P@kzou%4= zP!(m&Vb1f|!(gTLb>L5?35YAjLS?PKn`-0?Q{1ceYOLjJTriah3S30>rJGSN^$@!7 zQ*;{nX6u+g&*l(XR;_>j>{Fc~T%JW+>HpTKAIOeSC8`eH+*Rk?B}+{KyuJ64cHptZ z(_nxHihAZAsV|&b)w+W@%(3=s3M)HDFLii{UN+e z)kZ6Z-=g<{nS6Kc81GmF`=0hY$4{DNpUUrY=~&PFIwZ269u;rJzS!b|z(7}>;hasG zgG-)#tiG+HA4qxI52ad%($0jcSmD=&0hD@!cK}9cYMGA<2hbEw5`|@5#uTs@ot+`l zUb(&k17Wqe{CixeNMrj^WkEp{jF@+(Q?5r$F+UqDK)tg}++14BEtN8hzIs z-JP8TqXDj4370Y31D#BjmJ#BKjh#1C9I&ZX*qbYgL#05wfUWe^t$ClDjy)NLrPcSr z<$kaxEVoAb{E_$ZVLGlWF5hrXBWCW|MLMjtodND=hh3@1f*|t2GeS4Y*~Tj$Adc zKb;=6$w`OdAO0x$Ajt+9f9T!6SWThj=?RlVX?sNzK+(0LU5Yl$e=qGi6Dmvq&Y(^W zLN#J@bqh^=c)$#3Kawe(@d?~b%ZZWjC2J4gzIL+az>LVIsn~y8&r~;bRk?N5sI0M| zD%rT_Pnvf#iV6|Ik}g3uyjzRS6_fdkASitCM4<)Zarz#D)w2PZ$AYQh-1-SOt{Eah z;c=h!Ae0G)@Cis%+%Y0)cq$+`QD}>blCZKLQGO9vV}@U;u;HE%@$N~e9G!#ntpJSx z4`i+Ve`$g?E&@9L!rZxKf~QM!ij8gVHrJS zK#rs*{d#oO%^Mzv?Y0t9lYL2hKHJ<<5W=RVV4XzE*k$UeIse(-?1-9u zqE*It(Udz^;YkR(aMs?BCp4#o8xL&KV31WxDU~eogJ&yHfp$NP#1&Dr_Lzo=43o89DIX^-3a2} zSkG}0V$r_sx}Qn(6&DQ?U%v-$e|obWc=<+-r8&9! zz|Z98`dG|RWk^(qVKymx85$tpH}^=y9svr|i9v2ry1 zv^P|SV;|%i!AfV-T5+)3KwB!RqVg+9P+|M0&(@e*a)MdfDt5=pM|PTpeQi#^}b8qr)Li2fA3^Au}kpWE{Dq4u9SlTBM`2EzS;&dxPU6V;S#5ymnnM z*gOMx7Y7D*=>z_~L@|+WGA=Q?+vW?|u9?)j*ZgeX3nPDJKT=t(ul^YI8xV`{P0Fkn3W8@ z>a7myQHa6`Un~SEAQ8P_o5@0+30zOXmoil|DX=cO4!W$tvX^J{7o}MPALg+ zn-((&nco|1qk!6O6UXRYJVyTu^W@b{dr_qOZ@mBPzf&zoI)+!rR2?w<%L#BbZM;GXH zRetNxZJ}6DN4&eHPZeYS*m5vgZzy8dS{uzn*%>*5n0WbNrQ?5OE%N#)lDkb7_3LM- z-$oL+P46D1+1Dva7PI@;NCMP19L*k^_kU(`T+Gb>C6nV~YI8tcIv){=x-d<&?NK@qRr2>+njyHLuC(>k|+JBpN4EWTBWlySW_s z+(k{ICc&^YQp%RDq|{8K@O?VIT-XZsen9Z9D--=jJBQ^vUK%3*v8ncf zNude+3&2 z#i!Dif(3h$iM*zL&j0W5e*M1=?-onGAd!Eh8uh;w#^w3lcyf5=cTiMcKX5vQ9;Frq zInMQe9bSv5|J&hxsBu0aR$nVsHsP^gb!r%hN~7U7r9Mr#D#3UmTFeS(SzKzEJg1j$ zf-fIm%cvSpR&yMEQ16@OYOs#0NxxFx*(YoXE|1qcc|LU`>>tAJzI2oAhmgt7yREL) zl?w|QRGAc1)y#~yNnf2@OK66xrybK)U1jIe48Lx(@h&j4T|F_g%!hqZMJ?cv>zok9 zLcVa%Ta}*U%pjANB2jvt=M+SbDbAPCEnBeV%L^LX({A$h;^M$1x-v|Wj`D%=uII3dgSVp&{=ZyVSTiDxFqcNU`n1I}wNlT!B49E8RjD)=ml+HD; zKCKM5{eY6j{^9Mw4oRDtO&?yt0<7dv> zZs`~A9WqJ_A%kBAS!}U6H0Pk9+-Uf_bVG&heEpmccG(w+ov^2!6-9il- zgcABSih_Q;)Md%A%OTHTiKy8RPX4eBdAA7tQwW8gPr!>M((&Ytfsci37A46KNQ{`eAspR*t zN^+>64Pcn2CC{DoKMLd|SD=8ac1bvfuBtd%V%x~k<}Ex84YR6oK9HM}HD504qlpkQ zLNyckcqQ?MWc3qY&5{swgWMCdf{B4B9BHSbxzwW2mAPp8xfz!|w8P9DrQcb+#q!uD z*4Rl0`~z4D{+TVMw2f^EI}W%Gq`1?WmY@=O7}*jHH8A;JlrZv>_(>P2a45o0WVQ z09j8#a8=3>(niGiz%i7{C#|zm)SHK1P&omM03U$D>k-teT(esp`Jba?o$ECyl{bpn zYqi80I+D_5L~Mo=*@s^=4V%lP*;q8@uD*RKf18~_@NHSw^CZsn-y}y;-AGOWXSgl5 zcT5Yw*FN`%rDcB}#mqps=^f~u^?JDF69@%@sTcK7S_N*nhS~D8si|EfMlC&!&2DkO zoi8_IC}-j8>fJ0_;9@ufEqCh-ZZU^#sR4)~%XLiu3ActrEdaN`JG(jh`lzR)W_E?( z_(4){&b>BI`I6_CY z++ns8e{8uyv-I&vJwFhRf0V;9)tVGoRyW5?24Weln|!4NvVy$7Fw0PG-I(!0vWWmT zHhsfBS^I7Vh)M4(fa>bDpmgE2LSa{FU{$80SMDBv4(g4QGQ+f_o(mm{1!BIW*qeNu zwo=59e;bNOBcM_{belHbh2CRpdibhpj5;O4inRF+2u#a2B3F-N-Gp}=Td{d9x+v-$ zy{j*hMU-1qe(G?F`uY zVuF{>wVOy5Qr5Fnm5^K>$$-`53=}T~M-c%8-RM0Oyw-7xwfz~DUGAsAAgV*ZhTneR z4)5Z;*^Yq+vc|Yj8b!xbPpCifcloWrx^Vf85RBtyGx9hRNW zkk*U*Z{k+z@1{^yYW4)%V%99JQ+_r2oCk=h!LsAAe(=AT+q|6+E@IJPF&}L4a3C@k z_WVxpT~$&5G9t576+vfRY5E^GJdBhw#ARH-+Iex50EiW`N|p_*AJt-LP(s{_2gJX+ zX=u?!yvT-Aa`*32I28ZWe zNobQJ@aScLKcsi6I8HJLH(OucnMJ?_!L$UPZiWS6N5pB|1|;LyO>>S)BzRHJK!`Ys zYTszM%N?xO%GXZV$Zdj#w#h`}_~kzq-pkutoe)*k^lV8#;dvQdXa7_WW#lULaqgu< z-ALsr-9)yWeGr`4wFUKD(rH5j3z{B&t8F;bFP9oP`5T8Q@&uH>%9I@?TF-t)?ef52 z7h`a+kWclaLz9aU(F_nJInmtv1QDMgGSlpu$PtpD!(JtTfUjDC!rK)?BOXH&6Ki@% zm?n^QdjpW0Q#js09+-uBAed)K1!Z4QS$DqZUJ4FyWa;C-d=^vfQ0}l&{|{s5*d&M& zW!tiCb=kJrW!tuG+qP}nwr$(CtES)kd?RN5LuN$gI{U0W2S^vKCsnApFwXL+8Vngx z8$L&sa%&(7u@#AyxE-isg73qJX^ybicl`NY`XmVn1f1J zM%eiET5^S5CO=-%4!#X$3^kTI94HuF5J}M$oSmdZ6O()xC`_JA;@9xN|1%?63J~*( zHAXt7S=R;c&+-BnDF^r+@ktlQlMsO#baNwoM*UXCBvt6l4Q2kA;q}SjUOJ_7r95up z_cDRA7v*83{ut4>A*TEE*Qj*J=u1ojB7NMiO}PD;g^)VU*pt90vh18o5Cj8b-aAZp3kyQL(fa##=g047NqDwvlt+ql$3K*z;%jVFsjO$j(`L# zuoUw{fa2Jd1D0BaHwrul7RaI-z??Tx>)Du=b1|$~&Kmh`1aGqjbNxn^*hnTGRndRR zLvh&jZUnp}BAyWzO?W&U5Wb7*at=_*xyRD>dz?Ba0gi=I3hRt|!sn12rhtmg_kvzZ zUVIrZgo5&XBJS@V_+4)$op5(|Zgxr~OHM@pWoc|rUNE09^a#TtrmWHnzT^kT#R8Nb z_huHiqQs;`oGrJ{I-L6rPn4wSF{XNr?@A$Xt&65lYhELi@aaLw%yrywu(HZ=1RAAT zbpABi%j4ITrx6sn!zGn4nY$W1p>XixK78b{H6Ifz=Vcv+c()hqwGnRUaiQ_GN~aCB zhd*mO&?nTStC9&S`1M**I2QU#GQn_+%%n;FbdJ(#j|td2ixpJ{2b!y&p=b>QFD&ln z`y1F`tnjt#^nF(gez&Brq+ZUO=>)c1!a^yiZKg*u4y|GPP}z={Xy{>)B1$#3dO069 z8OZ7gKPR+wJYeM#x|bo|IKJt7O&A$*nz|*t?ECfg65bsgpwC_C7nLc+3@PiSaE{&| zjwSiongLn@aeVOt-n+iRf^w@5#t~;tTGipluq^~(f7zSzQ!m^>w~z{{a2QIl+9`qK zC-JK(ew;6Prx$=Y6qKo@3zWn($Zi4zuCl@Y-v zUNo#fgu*X2^LyYfZTjR7&!<`O;mMOt;k}IJS+!3iuYvI2Vid@Y;Is77_J;W+!NTON zCWC$b0So!YIK1r25YBT>@<)sY`}h5jS;9)7$&Sp~d5We56Rz*U`^oXrrr(C*YnH?E zR_#W7ATSLF$>Q$kYU&8-W6A^m<`&{WD^UMPrMV%30+Ma^h7I_W@OmQ(o9z1GM(yY6 z2jCu0r{sllk885&7m?Kd%F&{^cUl=7GRRw1Bxn#Cg?iu^W1)eM-_I=4|4BYVQAhcl zdje117eS{o!ygJGGi1gMiiwHvO*7b@0ty^Td~gG)aw1>N&$rBd=z(iNw~>IXqyEmz zU^r`In>^jGVlhM$CS#;S?il=Q2i)b}Hi^Xd>un*=kEX(3O}^$$$^ro=alSGwH`48m zD`TQad>U|(n$_*gQ+kiEI#f2ioD>m)gdwt-5N*v5>Ad)C%{(CGcb_FA`UMBz6R~Ms z-~J*5NYy|ivk#v#EH#`G)29T(aQN=um^OMOc5fSBC}^G@a^x;XjEwhaUwtvFh+fBU3U+TA2e=ag-leL#6w2??I0t?S!It%~tx~21&=ZDboYP<-9ZRw)FIM%Dx z+9a~%$EMlRPzB{Csf?Pa;LTG;kY(ozXbBImA^=?Fn{0-*H>$__%Ts8xdr>|s3%~() z`vHHnF^=S4Vv?P)$@-Yh*v}n<8)wFluJd#ExBS#n>9aj*Ko0Xe`2a~j)f=?PAD#A2;W{MoZJa+g#3E?j- zP{Z*}Ix1&K!JMV8hH)#u10KgTx-Xxfo3WSc?QMz}iwvHx$0ie7M}tY?n{-3t;^N|) zeXQj#>M3*x8ErE%oEOvcYDZ`Tp=_gJoIc^lrDq#`Cgz>`Pw5PrQNgyG%&hT`?8S7|q>|eb?{)$Pj5D5`sc0KRaQEU6aRVK82h`VX+4PYvH zG-BwTrdA|wE+ajq`;sD~4Oqpm>$6_~sR_1_|3X;||ADeN7}@`Kly&_tUiJ@Vbzk0W z;<#-d1VQHW${%O;h@^}*ET@#Xl5+p~g8(URmIwe&Y%KY^Qjc9L5Nrb9nxEBT)2rzAc8<#b+aa zuJ4SfYVRH{iv2#az}aBStiDk38i{I??J@h2Y&CtbVt_dMqP8(HJQy&BF?|3&^ztjG zdlu!_n0tQCBm>KZ8C%vZ0Sc~{&7G-oKT!<%6lZgG^ls3k+|Ek7W`)BiprPDI4sx=u z4~J@xB^0$KBU?ceXJS=t*?xw^^1%Sr0tzmgHMLFwE`kDBGIemn%y17eq@kix4(XzM zC+XFmKIB{9Dqrp*uifGU&z)`Jv7jPiO^i}aXSyS{bl5`qq~5b;-Dg|tb6*c|d}i|f zS`GWHnt3u&k3s82MX`_69{i<0Kf+}P=AZJj1otqfWTJkvp#xVIg~ohjg-zIN@5+iA zR3ltp*io4c8pHcjx$y(am+DNXPi^&Gy53x`c=G&W*v&u*%y4E+2ld}*kmign_ai6T zh;6$x(1QWY*4`{3)p!{@0n}03zk5MUFUPr=(yWBJa4LpY=T__MhlgvIS(O8BR!yK1 zSg@mf-etWiXLZ%}X}GqIu3)XxOM+*g$%w!Mw$BjiT8qJ>mY$U=ZF?^<{ze?wdWv;KQ@Am8@ARFhvb_;f< z4bbgic2JZa0Z=&s4kV8xA5x)tq430<{@9~$9CV5^aea&(8jiF(arWW?R4T!-+O@t_ zo7d($S!D6@8~ThVjFNIWOr!f}Bif0Ge~9{6&5N)~x(#vD=kecKvEg>qICVdgsov?1 z6v*vzurQbuyvjqUc%aC;n`0k?GUZ1LGoVR5rU6H&!9a)%;DZqSgU<2d2a+qWbm+9% zs-8r;q3vCFT^LFN>f>oWf@!ER1F&RQyVa`qiBQ!LR>hROBRz~NGf)|=$yHSb z_A$MXt9l;*sFG@OR%#G~Hv?t>zq~+%`^fCX0BZN;T8N6pb<|coF})J9x+;_>;4)?t z4-n>c-!v8-5E+phz?F*;)2xpM2(LxbL1rWpj5&Uzh4!MQkT&a#30l+~WmTrtsaJ2a9i+|kM7<18&J3lO=e09~p2 z6wx~t`Ys1P1z!HjK+xdUp$_2bU%eUp#OW5m}p{s1;$yL?6}gQXA; z2m`0w*csi{Rx1H%I-~7cM)1Dci6weMQRb;cE)EVj*sYfvSQGO1#ZFSHI6Q}(RVrwQ zB|bNjOmOfchW*<~CjK!FEMCxT0f zrW|Yj&%tG7zG%1bHk8@6TyrcV*--7&q&AeO zG+{ko{e*IJo`QRXzx`MFYsq%xktSNX*-J7u{Zz|h#*x-w693ph6Cg{dJ@y=_-StmX zNCAGKZ0!?L>rJLwvS2IbDeyzr@xlS)y{&w%(;C}8rzC_H-kA`~5l{pn_APLN#>NYv z?MF&V$+i3960>9=(F7rT;Wjw*@G##K)RYAO$--v>SjRS8t)STUTPnUejNtCyw#OV{ zjVs8}2W)JdNyll!Q3vMonR+8Kk!161C_MF@efMSTW9k{DCmiX8zh2w(Aw*ZLbdj>$n$l(oX{;6fK0rZrsc}5!Z?`uqvY39D&$lyol;4Yn}Wg{B^O-2evOo%3EHJwJ-i94iV@BMb8iWj z5d&13(MtyHtECw$H;+cvPaJe=Y@nQQX~yFqjhhR)xj5ZW@nAf6bFtw*peH+)O-q;c zf-v#)PbT}`>;n*<_E35|3ZFfXm=4%tG{F z{|;*A6)bdyLq>x=6Pp+pzq63qDwC2MIO-<|^1WWW?8=&j&RqsSFnBlV8Tk^eAO`?5 z{HD>xp9!&PeXyzSWK$AWM9obYW4AVWQmn|H2sQ8A7Xu0e+dTL&*G6?si@zTSy9JUs z)Z1wqFWZ=n@#>b8-+lss=DrjjLuJR1Y_T3Zo%hW5k)P_qhGG_bDJA()YMK*m-6`*1 z|6r{J?Y=5~+;l}}lkRTsfwa~m5B6J?19VYHwhXH&EQI4AS!&P@!CMjo%jfueD9(Py zKgoVerjXI~s((2|I;59BJ{nFe0Lj#%s_gCt(8bxG2Eb%Y|5CWjXj@{;yPhe_*P3v^ zOExF{N4P+DBHdr-Ul;CbcNS(b4uxrmf7MbNQ=t@L=@8}*9e2o3I`QL6I8!(SlaMg4 z7clG@<7ABIT?Awk!|f?$X%AY7pV!bO(;B>dRDk_G#4xEn7(9?#uJj>qqr^HhC%Cdd zKkU4PSUKXnztV7FNNHZYBoj>ZVT`}(Hekf@M>?ouw3xY^Ver&fzqxB05|cqs4jg7Z z!ku%~Et3JRzoCef`v{78YDwfNBeGI=`9Du9)n(yo*IQ6HiI$p67lr@LPxOf%Q$r>& zx2*PZei$^?h{i_6<7`s$qdadI5GO5&JUyA}gS6R8h0eFB>Mus>;0MT%J5uM$RaF(Y z&VbVjOY>z}2cO{cGKXj9kC*03z5Tck?ddyZi@~~2c57hk)isV=T9P1`izaKkZHu%Q z%tXkcgYq6jT0#bEUf{h~I}Wc7=(a%zQH!f6IGL#^oxw%sV9lYd%hnOJ+RnI759P)Z z+sHfIrmEXSmtVGM-=tY<=TUp;>QZG`_jE_eVviSZFZ8=Z-K}3h+CbT9>$Sf=#%7CY zUq!Es^%F-;|z0vZ6t!b@E&bo_@Qz+Xzx23G;_rKZ4yY#Dd~ z`4!6|`orcEbGQRENj1ixK(A)nlg9)oa#{aO_*~WeAkZ;t-YAxmstL;2X|>?Q7)Vb1 zb-h;~FHI(&^wUc9ckJRMiu9CW;%>G=rd=2-6(5dK!H76%PB6XiS}%v6_6!dugp_TR z*KU&@h$^%}==&%ApdZe8?N-zCI_6B=7&;!91f1woT=(Ta zTd%UXQ5Lee9Wfe3u!2`!b324hB0D%B%TA!&v%wTT<;{JjA8EJ$us4X$eqo(x) zoF=$2XZl+TFS2y>!cl%MTrm=0X@>p>*#F8=B3>C5F#pLVhy5DBUS=YKI;kUhx=Nbu z)qO)tE~87)!Fg5R1PbT7%2$nk*um+y8!k=`(<6gI?l+xyY0R7LJyA%7LQJ6bEi#6w zz-PNzw~@Y8-t@uJYk~?#f*EJUT*V zv_@yg_CR&M{f>=YA)b>~qjCWY?JDIU_8}3^5+A?W@|og-zL8P5;_<8gQ1{w<82G5| z`H31~fuNXR6QdI!8g2ocwOp#gDiXmX;*qimIO0Dfzb(e;?~!l5vTGstIy~}_ zVe8e%Tp&91EH(Hx5$H`H@z3ddHf_gS&np?pq?;yUuICe#^1ha$W7b_t>r~xRlB|7&M|E4=9(woG31E>mg8DEUk2y4|xyV-9DMKpVNi%FtP5PAVai%=F)0stiKg~mk#bj~+;o=9;eMUNV>vLsu zqlRv5Zkesw*xRoo4M)s{&+!>BWzVLratYX7YP*H5cHQ{RqCA z%3j^S!r&ZAM<6XAP~g1IwAXc-4dAP4n8r4V-d!p6`Thl5piw_&C~ij_MryqP8NrO* z7Jul0>kcK6&wq~Jgtx2j-wV2($5qSr>i&gN!}|sCL5^x(qp;HgddEZbK020Nrp{10@5%gLoyP0`4T!x|@ zB?_HOfq zSH>=$>+RCbjdo|}Gao)HZ01AuOd~H#Zhh}LObrtROnYd|Yp4-*5Dak`ncyGXf3Xcx z6E2_aZ3rv`_DTxsT6@+Vl?^VXoc8c3P(;$HCLZIyeKgI6X{g5t;W)O4C>=!P;|=+; z!v~y$cjZM&uGsy^R4oeT*|*vzZxX*!s!uE|ER=)c4m{Y#8~sE|e6^p(2twCX9dCqgn&(2EOf{uT^PTb$Mly8^%LZPibi0& zBMTt`WJ?l6hDD5gY@H>vl9Rs&UGkF-eV-=S&OvMD;whbMYx0Ac{BYUO1|=~WnHT$g_-dG)-TDm1~qbKv9P@c!QGy>#~VT}~$wO(~hQv{DkpxdnS$VIZL=NcL;k?CYL_v>Tsf@?O5q=oQ z>|hO3epf*H1|Z8s#f`)Fhht&=7CFA~EMUZrHOP@a@JXOh78E$!;nNS-uBXUr(Fa9* z1HI*3vp{T0F7V|_nvLkZl291K8WI7fjI?V_bT&`SBfp?l-y`3^bI4UXmt}kcgVH1j zO>^JiDlzU=bkqlD(NDow@(P`f!b|Zk)p$;j?rqX%c^CQ9KL-U!Bond0CFq_{6Oy3! z?lz%~pH=DSZA$lebGsXieMgHFAJA3O7YDd=GpUg=H;;f&L%k(k_l6aeoe4v!Zmd>@ z2^#|vmsP!{MKoYs32X)Jmm%Kh`_rHA`U`mLPx9b@p*Y6>KymaO4F3ldry*&JBbw0j zSd+Q_lVLT% zI5tRIhXA6YYHVUG;&VTY9{;?VM}3r6WUl~MBjdvr@8ufzD~0~Adz$RfRRh)l zG20<9QddR+o&%y@PQKjVwSH~|juwsVV7XIuRyP}t&PSMR_p;C!G2rpcm4ORURuS0R zgGNign$FqcUS~=0IP%|9Hg&L9>+5~QuIFx{VoObH&aS0S1C#^fynx!c>FY@N5d`{d z+n4UgkbOZT(skK`{Qw=beq*I``2^V~bD^L1z^S|ZbgPEk7@@&y>szjvk-HWhos2Kb z42J#<0@}wj026(jd(f6QR2!RcVyj+lYLyg3bKl2C-A7%kJBMOR-SBRBjMRXg>-tNE9EU$UVtjoS1r zUvE1v!(WLT$CM98-Vik!b#tjXNu{gOwUh8RcImsFOEqYLwe$U9@wlX|1AnYtxdbZ?447hxD9P2m zm9SV}F@u2$i}uh(YKbHQoi6Y9epUTcf6a(m8`x<9z8_-<{U76XqU!p<>Fpl!Ys4rs z;TbdV91{a1p&HlD=a+FeFXh{s*VJJr7Au%e3|mN`NJwBKFK_Sn%?`oG)UZ8re(Q4R zLaE#11BPnLB47kv){z(g)G9q(9acFdA@lm_u(v;UQ|Jd-ta-mHUMYtDx-RaEyYPav%5JLLEi^sF&M&S=+MGJ7{#dXdHa7{voGvi@Xoyh_ zgRB{3!5o4OaA)-Z4_K2W|H*GM+aJ2L;Hs_M>Qp+oy6sxYD@4h>aAc}- zducUqE^iU#5#WNs!4hTvbHV&-%GdoN`v9(n+7Kdhcg*vVjDU zXD8Kz?EL`wAwX2GXy&#g`|QHEuM$7d&8{tTlG}n}kx_C4X&zOA&e1LDGl)oWk(X_z z9fl3EdAs(Q%v^(lci-rKBh;NPRr#`B_=UelnlRUqx!l9s{FHE)h`nzPYc6{Gg~yd=D_?1Rm5^0GarOOGLA2yT*%8A!N#|&`WC^-& zKgm2rI}tportMLTe^iTL*dczRu~y+!z&5jq=3<6 zwPR8NSU)kmUoH3__P@EetC#MxNy`NqaENa%)GJjq>E#HM71FTayaB$}u%tM+oFJge z7?ohGAB5uQK);E^y7&Zt5|V(CqemN+wX0ZF4Y!)d*hg~JRY9$Vnzg1?g6|jB@^1D~ zp+L-w@YeRA42@=~j%dnMEXnVlh_NOlH|&n2^^M!ENnR?ugRtL^NLK1tEe7JLfCmeQ zd7Gy(0r_OAGV-h%u`b^M;jJ6+5ZFF;l~wojYmF;wuo#Y=3RZv_mUoQh9%6@H!}$?% zfRM%TK6z+RiAw?4>hStzDa9!)5NyOLZcg9^fc^Sdy^?6hA=gnEf6S3J$o6jN7ty`& zp)bH=02c;v^!<8*yO})w-Yo2V5P(|LX2M*On7C3bDM~f1AFMT;RE%E3P(qTbxU!7+ z6rAp8lC21X?SrYmt?}%BMM=_;O}nf9X_B)le{JWC+sERZFjLVQ`VPQL5>4LO#2aY? ze}z(BclrPwwEaNIsMv1ep3t=oz&(I3v2`YDV@~qk{dEq@&qeBr>l1ajPV9_;edAv9 zf^VDaKuI@9ylqznE5{IYlrr;4fb}ssq(~bbUfc((ZbMQqK$8@#scuxToHmO#@h2!q zFRk?CJyon*EPVo;6)YLHm+>kJ!?BO86`HFcgWR)0oZJW*)a<5diV`cHv%vLGjrfxW z;KX#@4-*FoT@IS>#Lq{12_sy7T4>*{&1GKqAOpWy#J1sdA|WWG$2GU>YYi+vhI-5` ziOc5)yBgw&WpN{{Xa+DGbbuejJvev1eEmCT?Rd8+dUq_gwXaLi2bQ6*H6LjMZXlhgOAIx>J^?Rdv~qUdd-W~oFz z(5iQq{EGJHxlwS2MaN=SzPT!sf^vv%Q&APVHRC56j?BiS3gSs1xS;nYpaqfJ}S@AXaI20ggMSQ=Z!yD zl(lDS8&Om?MzK>gxXUdD*vCZ?5)s00yC3c?cUnn!{fsOWx(DY1G?B_Yeq%IfT#VS4tpU% z+6!cV*Q)6NF;f@Nq^4!_K2sIleQ__(i|U$X=_{evVT{Y0)#`%MZUzia>4@EXD_>$j z7cb|K{F&M~bwR+|`|GVfO^Q^D!A9^D0}}WJx}>VbP~$Q!Lr>)L@-u5tvA!Vb+q`L> z7fugZmd!W+5qWIeP4L-U)%ycNJ}ZGHa_CHaM>goBhK{hYxo3SA(o+4c9yQm6g+>;0 zs0`FQz2AQe>%~f9OD&FLh^|OIay%h+lp5sNFV#C0zGjtLD0&KNyqsOmd z+AV)lI0{OApFy&k1QuZXb@Vi7_v9f+RO`nY?ftY*Po=1kdW;tc;@4_K1x>g*MJiTI zfHWknSK&bxeOw733YoRfYY-Gdmd1PpT|BrPRmG}+`I2wOg60^38+m#gRXCPp5Vb9( z(+3FY`QNnopn|Cc|8{sUU2E4Ts!hIb2TxU?(hRNlGES#pri#X~Ncewod-3oF^lR>3 zZtX57m?{z;O!+!N|bB#m>rf?{6i@hC|2}uhj3O@>oozJf zj(NrDp`rxK$loWW=R4n~7APr<_u`L{dHApnPVhg09xhEvbNhdT8MT(`NiIJ<*0l(w z2!3nY%NQnk#C3{3C{i?KTeUa6KrT6Yy47^%S535d7^wSWE6G45-H!zRj8)6ji-Ns2 z${7OZ5uNcaa5`xuW@4s*d^1Hgc-6fbalfp>h~>UBK9RUl0s;6#{IE^1D3NiL%JEb1 z%F051i4vk6&I*R5AUr0PB}2T5f9bQ{eej#>x*>o5Gu0@*Z;rCb_*Ngz0U|&y_Z&VU zf^{U-Nb(|Y_I7`C58$V%9$%gC9OiljX#MYb~$8DruLwaYLL<%f! zDK+`NefaP)?pe9o5XxEC_6iO=mU<&%Seb*A+kFug#B^Cq%Q=ry^`aC%k7O+mxi@GL zdkopV<-leepYEknAR^o-6?DHl24Cn(I#+7%`Kt;xp}Ec5+dZFfkwMCCr`qlHc{L^CVf2W>Z&}q!oniMP5CRC@~X!P$#YVK zHAbAlS@LP9BKJ{KdmkSpmKBYV{r|H}?=eI@kgzr=e5Jv_luE*iLDxcLaEu0W9WA9 zhH)R(LZv{9W1!I)t*`HgujAeBUxr=jWK7X}^A#D+3yn{@spf7EF2uN_@ywXs2aH$H z+GBLT2rtqp8}ByXRC#1`0xyg7Y6l8OFV2xdiI~Ey;t2m4wkW9& z-zz>_fqjVrr-Qa{R(Eo49B`Y7sLAimF1sG14dOl+TnI+OCmKC7HS!jh(w@=w9v;G~nNC7!Yc zdiemRebFKny%4VIyzKm3q4Asx66Ltu3Vz?xPNRCx@Otp2YY&x-eqi{$dmWnb8Ly!s zuN|+^66-0Cf-KKpZSPQnK3qwDv|Ev$+(b=;whf~RH`fIanIXBLZ9U|j43hXE+K<8u z_h{k(AUj8~+T<i z+dRzAc8Eo#QgkPDUo#uNx*@k0`&K|#Yk}fG$^RbD8q-7 zRo?R9zWbHnwLF@&+IKlI*|-1sI@l*jg;hW<&L{I_BN1{UL$p9go&#^%X`C(e6GAA^@{W+E6fcPHYqxR z*dy$qF~JVoliZO3VA$<8I6xxV-AIIVWB3bSD#0vs*3L?Jt8%-}&@zU7V7`^p{PUv}@-V)@4Ic z{s~46g`-ULO7QI8cQ<(mhCCdGD5&WsF zugst=4fsk?Bu#euprIzmTI)lgKRP^#;RF?!Y>Tmv-*qERU^gQm=j_Ms15@C06%{zO z%0XQ*X4Wu;JlnL*t@HE=5Et}8Nk_&nvzkUTW$Jd{E3Z7d9 z%4Dv;Bafv@5!AQ~Bo3zqbCl|?p{`_G{*NP5o7fn=A{Xw1dvgURwUf=-zsu_6zFl=0 zK6xQF{{S@E6K)%+g*Z<9+$eK>3L|hZrj&LkY>DzIb`aSiC$%o4s(N|B=8*zpzRQwY z`7sEi;!o!veIu39??0WxZlt0YVt~*8c!+CzE z808pdm2}ynZm-IFTZYorvlY>rS!wd_qjj5`;Q!LOnQ(Oe;Gj)INW>Tvhj_+?#hp5p z+9O285Dev^L?rx!a-i$_>`Q%-&_WAbd*PgHCXX+ThsBN6ks`xmsvXM}FHh(+6(4bM z)v!=u#qYJo{!I7CxYr!W+bzrRvztkJXRZ_t}0+2Ji zRStlQU?ac9kJV^lQy5_oDCFL!AqH8z<^(P{Q81<#yD3HoJ7A;uBxO_N@RKhF|1<73 zBW4RLpc(;u?yrYQL*hItfbjByI2D9a^AWq;fA)tI_<}Km3gBCDGVc;46pav(jk-Qn zz-j<$1LO{3$fiOw#<)#bB*``65U;*w;`qy^X%0z;LjCSKhJlG&6)~sUQKt&|V6dDh zWJksS688boOGJSy<2}dmf&Vbual@ZhoFZW!)M2XQfG1G~fMVRg+luw_9uoygp1=qD zrCPD=&~Z+ulY)~30t8milwa}AFsd!}4i;xA9_F9!^Y_eO$Y-bP_M`X=1#Q9pI*?9G z7oNWzN*JX?-}3uTu@R@>2QQZPC(*xW?`g#yVUG<3?V0thd%O8&1`vWzVkfwIVG;dEr?q8}EH(SxW=!Qk zV01iDUR+dsA!+JNa3N`2GFH6}wqPmF{aDE@@v!7wCU6V*5ytlIZernjhJFC_&CQtpW@IoAmx(9=2tR5Y>aXcd`;NlOCLcpZ*b(pMB9F&9v#1no^P zA$b~t_|)cwcXLREN;nLXvt_}$2khrKw~8e=mJRH_M^%6 zFooBY01Qe3FfNG=E*aIG3Ne^OI>XvJ{YYe)!?#L-K@QubC$5@&8*EhQj5aL=Mj_il z?V72@pgNy5VF5u!e(KO}T2TWa6T5#+OlTrZCdOFMz>P2-kcQzF`rN3=1bhbc69*Sp zANOehOTab*Zk?LK2|jjPwDGAx-Ta5OsPRSQ#TYS=cUuYRv`c zBg5F)GqaF{V=J3>G9lHvkGO-}C5TU+VAz!!xK0*HL^(!9fH6+D-~->z+E zqF^t#C){tfh#lBb1qu%~tC(8W%Z5VHX|%7wUoW#sV|RXZ@&)Yi>qqHjhbJ7_yC^~C zxM;GV*ob*+ITYc)Q9`NHa$RPd(!#Pyo+b$4u(+hFg5N>O({z{%$SjRGB-Z79mH7sO z`XD&VO5RPxxkX})dz5fy&}@TRKm6L{E^))?@~EQwgK!lN&0>fZ4ozMN%8Ra16*CTc z-GLRf9z{VFx&eyHZQy<7?qt+^OYL`KsP7;$yVM<46|-m0(*uIW0F57NgUJI zDxy(XS8%ksnPv98i%!#OeUpn>sn5L@ou}6xBgg3)9L>MSR1J!ezUso4`7MVDPrAB} zpz_N?Myawehv%*m>>nEaWqM<&h*)&;V&gu+Ftx1Ks+U!i9 ziNK}t{yd+{#d^F@9*^Q9y%o23bAOz#I(gMczHLSgkjJF2;DuyqAk_Y{PD#aRzoi;~n7s*=7({bksxSHiHv(V`?3R)vJLx65kZVY}dsbd|W`p6Nq<=;G zcXlpIE+03GR%R!j8_tot*qs9c*K}hbKttqcCVJ3Sz8=-A+R5DTB;tHpvwm?TsFD+$%MI8gcnL`(X;l)~x!k`mVX z$}~5FL^NMOt3*GU^?TG}AxFkPYgUUbyjp*#6-rtWSJhTkzP((iSnk;Y-8D{K?Czo5 zfH#tcsX+8|2{9t^&ASqw*!3&eJZU}jFe1xBeTEx%eaC^h<>P0!Jtqyj_WGwXMk1c8 ze?q{)gXgCoZNb)lL_{~FeDjHlfnIga;t*1jhjXHfBS2aPV;#7O()kkP$Sn-PncKGz zi?!CKIt(-?oHh$DYF)%SG#h(v+RHY2*87J!iNMmFy{NrGF`%dhbNn{A)28lWQwifa zGHqywS~a?+7Hr!(AEz%WwD8KT4Qy@h_-sexL_lky z?SZyP0c`+K-*N01M&RTR_4n$CN_k^aBLhQ3J zF%W1}$NfkJp>irRr^NDe_{UAu=>SAkga?$I&)M$#qS!IC-`|li8TGJby#H0CQ0~Yb z5Ob0FAZ?$3Wqt7bf2b_PmU*rau{w3OhgV02tT3%v%i|jwf;!`YDopvnbP%5)i|S?X zPI{pPS$&~Iz&MhuYgRGLhjLltwW|pl5!R4D>mPWxF7|VV0j8S;@kv(_FNZZvMZi3& zp{vJ>JC8upVD0-uDE!@QTd>~XW+V;diD^8#$rR3&i-&RYE0n8sDUdakz2o~6B4JB@ zR_$-$JN*Z7D2sG8kpf-|k;2I2TPFNyvy^ssrgU6VcMKSFlD>Ur)s%X?Na-Ha)cfvP zr<&(ecSN-snlvQddJfoZnnS?*Vw>d}_7xHRd365)5NHb+lM+V-!$!EfT^o>`2u9`d z-|RX>+W#iQZ5B$Bg$L^Mg!3r023&E(4*_$JjenIGEm(?~O!J`p{m5xy7+C~gpllMc zT_JH3=?UK${`!PqECzU=$cpq?+R84KBmRW6gPP);&E1<^AEfqb8XZ|PIKgKR$aIL` zOnXS`me2x~vVd}R&M0U!->clQXGNf|;t}J#r`<5)Nt`#aq&pxa|vLfld8X6}P z7X>{?K3T8vvh6g$-^-{w-akyMc_rA*A5Pt3syIr$q%jy`u(p*q zCE{`B{1{>kK#s$)T8oFbNjI~$ z_bW7(tflhq@J$#At)ePMvd$HGf}R3uMsP!Ka~0~Yo>fvlcNg7?;lH7A&SOdNr_`Bt z9;|Y+va|$)G9^3C-4#UX6Z0^im9=LmR7JB!f$`{MLmv1)>AA>Xs)qWeJg->!AC+uOauWA_-FU^`iIkA3$XzxBv4fwjwv$%u%dtNyKlX3 zpC9Q=g=uG?zL0;jt7Re@KNEFjK(=`zf2UMT>1<_kx>X;jkNPD%T^wBpY{QuF959Qx z9a)bi(i$_D#a-@y4!#bCk4{Y8#-n%S314?O-EK{M61tDTBw{tzwlGx5ZSDY?7W1tSTM~5c=m5SI;4(^qhW2uXGd36(~J*tw|L> z;$}0f|J@zaww2&u=W!jY4G;0_hre%9D7vud4RIMG&mSj8xnxkB?4PJAchitj?5DyE zq^1`+F24+43LkHF6U+U1BL_@w zFQ1Jx_PCW0)H%{M{2LqA-B~il&Y0ZNY3|TtdMY*2BA-(-lpM3%a8uRW?h%T{16j=# zYc@Q1D%mD2BJupLAqc2e9oypWtk)_lT@389_YIG$Dov4z?A;hEL(nt0tmyRcIX#!F zTL1JCL7i3VxYjeFk7ToK#WSw+>?YAiIy6_TzA_^U^YnH|=+@T>2n)nAJ7uri2rOns z0DS<$^v80i^(}U&4F*(A*4V6&*tGevpaO!ss`j6PR3+G{H|N~D70prx0oA}xy<>sN z+t~RLm}RnjfkowFo3~|at7g_{Y%e6@>yK~lGY!c8%SkMhm0SLfLG6+2@sv~1-K9>P z(1%!W^f81O0SrHqN;ir8<9+br?m9W;%DK-1c!(?Er$V0HjP}?{W9O)|?|KPAse;OgAZUc!lFda8lfB^YQaTmpn44B+J~-$bLHtU93~I=e~tV5l~4Ev z-A|fMndsFbLgd0dipBPUt~<7n!3&eNU)oH)oCQuR0*e4?H`1_s0QqjT7=*AIxBn2e zAZ%nk*2ZrLOroO~QUI}yDi}IJKO?W*jMQ4Gn`rE};lj%UJy@6W7$^buw}eLm_J$1n z^|v5g?aX~xprLqRx7e}>)6msTqFlbFYPn-~g#(fzUIVW{xx*NMG7Th2v%)pfjqSHliX#iw8RU-g)w={!Dj)plc1nSPisdOjH?0TLiRnr0_x}Y8+VoR@x}!W+`D-BEmZ5%_D0AQi5)l~3Ou zta}rV_B3{LQ}z3tmR=2oG&5AEjkOY^j}|H9NfTO|Ki6d{6f3sue)Gh+7LFXgvk)d< z`Fa8ltoY)p$t7&>f_0~-+Rtcsz8r9$S(&RyM5?px+UC0|OCIFo{dI57nxjcZ@Z}L(h=|nn2rNqS& zW!|(~k~IJjA`$luPKsbg6NlIyC=A)464_dU5v<8Q^j8van9>f2XOy6;?04RPgEVS0`nnkrse`0${??mm=Ba%k@jTQuYZ?XUrx}1 z8c`3r1KK&tv$uk=LZ<-H|IXFu?%(An<%50QbOHU%!Z&g4dSMtGRt)ZmX8Huiu17Dl z{ap1?`zdKOrVO^2KCFbG~b8w`tnN zC*OyRXzFERA5c$j=zk60^^+Pu3Y5ODZXtvUY1D3Q7VWF35cLYNPNq{x=+Gu{l2?QP zl5^a=KK9r~`MVw=$o6SRa7$a{bezPVatgzO7n$#!bX*CkJ+ZodFEob>xV%#8EYW9;KoE*%7IRyAx4 zuCC$4!C#q}W0z@f!q<6i+gLt*3sYqj4j8EhT8< z(*a&L5(b&oKzi6#h?>(I26bbta@%nXxVjpNx|u^Apx?KMP3zFsTQk9|VCk}5T{Q$Q zuzPM|eJ~8{Spc&_9dM%p32Be6A)Z*;&wz+|(AzM*By11ya^V%q#xjBEG6%_E+%gMHF>Ee+gX zMBDUX_Kq5K9ZZoiQKdB%8AVeUo^K9Dib=4>Hs@-J*Ff{@LSqecs9w*((wal?H>OFY?b>mTQS_qLA0)Sav zcZ^y9En^NyQ-fl|b)_X!^v6naUXVyaa_~XBw+I_)yM440Eq@4 zLnfi6+#l_yqaKIpKa+wrjvW$6k(jBm1+QM9(fdj|kNW32gkjTCW!0t;IzuGBr|-t3 z5qg_W!aat3AeJx_oMm|GT?7F@-3v5Tk4NT(VS;maB{|YvXDg%GqSbvy^>Pb#N>JIpn;@F`GsCcb;gL6tp_+9hmoQ2~}XV7&G>^0Vmulg6}oVH&vk z;p(Bizn@~L%OB{^;=b&Tk?h<2m56^Ry+d}y159$6ha{b;EFGZCQckEcA4(YV@T7uT@+tk(=piCd>Mi9Qlu_i0cL?`M4}M=@siu{jf~7{TvE3ciz^wy zm|uO2aYFns`>KI01>XHynF@tPJ>Zl|0D1i(;Qn;Z21GRx8kb_c^ zPvtMTr*qSEz3Ecr)dlna~L>fJwJ{~SlS0hs=?#F*3m>KI#)`{pZwpxcD z!`Uy{Z$i~<(xb<>wQU0?K6Q>@EU)A2!9|dV^IIXYBHe{2^9i6t9Vg?U!3NHRr5m+^ zB-qvi8=t9)yq4LJ2lMI`14>!S$8JWYjMs%q0~R{;ByzvT%Xm+{^rM{&LrK%7gQKw+ zMi%82JB~?OrY!T4Y{obyCYL z|Ee>6q#V(!7bs^HFlCp5k5xvrwBmP`b|Pf29pfgat+4dT;)~G3MO^mrV9*C3_7tN! z^b(gAyw4boYY+<$aKqVzhXS82!~$N%L1{ejiugAA<6uk7OYMMHpL*z{of_+)dg zVeInWh)OteZu6o)rtS&T(>}<{OK|H&Cg5FvV?c=rs(d^azo63Nf2e-4VuFVvstQw- zNBgn>?rp1j?bQ8g?0w7E&yF9E1esP1xUG*Y+~~*Zw)v{4@?_Gex#Q*a@Lo^V!&(7` z_m;%&?TqF#Fg{yCoVtc8-~yt%XsHfuv$v zZBY7^=Z}DBfj#jMUv+ES955lxQ+lFUfD*mPL4(Oo2qnvpSJbKWK#|W~R|2x7jVX7p z4h^YXVI$A3ewttIaQt8E(p+5h`3{r`VVHfUpYUR*ojkAyL_FbZEBk=t><(Rc)6@9p zZ?u}u6-0!>k+@VS_TIwJIy}$fPunKi43PGKXX3L(-r)Z7=Pq4Xw_QT4A@vJId#O5W zR)r9;3UI~=sDcVENmo$Bv4#u==C?DZk{1Pis2l1$ z$eQiO#NKfRQ%9ePP@3z)W7kxLd4`*|%XwheL67O@R^6p4=a3{3Qe3hyr~t90*W8}9 zt9Ib3&il$CaQ}E|-d@Eds73s7cP$@*+oqNQ?YMw9`PZWH%;TIiEcRGgb*JP-& z(WoI{2tw`2(O;{YX_O|O+8a%dN~YQ^{gBEkT>qp)aNYz1qIH?`bZLP6Yo9%|X zYHC->Pr-xQeI`Nkm+<}b;Xt2Im13^`Tn3ftThDpt_^ruo2Ojph)2278_;c)@{l-k(H(wH5{W0pVqZSUGXZ-YT9|d3 zxx!9RN2;Q0@vzI>rDFx-fsl}TGfFf6uX@dnV6JhtwU)hR2e~CQJeDUA+El~6jFHam zf<4(AMgM?4rma*zGyu=J0J2d-|8=|r(2{4lP7ie$x3+E^YR!^X`l{|Zd#&bW88juh z9L93*MVDHhy)LtX2O1xn+b;|P8j4Kan&pCX6}K4!cuof}n_2S{2bnwgl5ePWoAr7D%kTg7EmJ%9Qc)yVv?bYz~^e|~^WQIWf0 z#=CNB+632LT-w%mjtRGNSF9&Nrm#+2xJIsq7u*qb+U7RTYb}|4+P{2AuVSM`gQ6aV zZ{onB@|%#Yf?fq|%BL&$QGR&e1V-sjEB+IbWBdP+92*n!f3YZ6qa)+AE&jiYa!c0% zZU2%=ManVoJ0pnYR0?Th#v9C5b3xaCCB!cX;t0y5{xl$bLwc<Bo}x@jd@kvyV}%NzDDVa>%OK#;Hv z<(<8k1W4nRN#R>q#no|092|U&oQ*_{JW0q(n>jT{H_}Qm-FW(t%5MYNn35!pZ?dl1 z4#yR0m$Obb=|*IY2-NED=+MU~Tci-s)>?4kXw@)#bT2F6PHo?(HQjh2)+GG67B%-?SvcK& z!6He9u1!mjKpbGe>Z4q zS7VJuxHoA^Ba+vkj}XV+>q=LQFmB&{XkMtQ5}b*cCXu4>nPTJe%#D~F`HnFoT{PYi zG>T>?e<`waHqEfmCPZCyxvB(bWT0sb{lNFFkj3Su zl>|ylC*d-Xm2Pa;c&P;zhJ<=MT7hvEu>}2!c30XBY~QzC5`gSsHuKbHHS|X}>qF17 z(OZPwgeGXyF^6bS0gOgks==vkz1}cmh~SV%?Z=)_9iOSKY=^qs^Mc2jf~n2CiV2rJ zWV8PAlUV!tBB=o}l&rpP1__%IZF7}Q z5R_bR3X^OB7#`@Y{apr-Ls`_uXd#oGndd5EqXZSQVqiGiZgM0Yz}>jOK^z~bvG#li z_{pTMNnOy%lNDaRTPSh`@8-A0j;|XJMyl6NFn^QpR2fOkqVu2>qT@YttX z+(sLV;8@txvlr&{dgr%cZ4GRcwf>I(WwIHLCr34oHE*UIkalGn)i!gbQfD|R%ANXw z{aiaVJFFEVM_Un@0J1FE2%n93bIS13K59ugg*j73%Eke~jZ!8$jS!U#7E`lzeezwq z6D#86o~MbQn7~S{Y_T_@k_iM~n(^AKXq`M?>e&{MiDe=bJRr^eq%iLhZ|^@+5m`N< zrum#W3kAp_Fk!poei2lD3O0JX*(|smYA~RP`T14l3P;y09~iSniZE`_q*3RIza($HB@qdtZ)zJyQhP$tLz%h9Ah_irL)>DzhKx) z#Ed9KvEy`o+bP|U*p>nT=Xg+lZpv1L6*l^-(f{mEwL>0R7XJk8eJAMgI4!sjXZj>> z!0oB$@q&JeSW~yP8`^8yXmJA-?;n+GUKaziu8aKGDOSiB1AS4@k_QD*S8Ku|uAM7U zx5M8gq=3M*aIL8mtL!!(0`)XYeD&j`wOH=PDvgEvn2*7t!do(=8(qe*Ley|7lN|y$ zQ0+}hwW6pF^0AB&Xr7F!#_xLXt$$z7k+muogV#NzgpZM0CJAgr)d zNFzWwR-b9CCRI*F-BqoOBN#h$#0BIk?9y=4)5)LRQH|)&JM9J5H4K>x5eQ+)>LcXL z9zX&KpaLJi;ikwp%2(?MH$uDl!AXn=RVD(UiR#Z}%t(#h-0O_k^H9iDBJOKnRa@{**kM@IeI?Ojl{laxpfmftTTBO>~ z_D5Q$isHmD*JU@&Vi7EDV%F1Lli%5h^XjJ6#A#aA88_5cA{o z7BnkKVL>f=HI7uACUO=E_j#058CRxa5|6y8%5@+R02Sc6E{SFs?%K23UX$%UL4=07 ziSGx4LO_UdXYrSa?aankS{^B%-oHL3geR@GJ(`XP29iX=Y+VoI5YD)a2uzVBN*l)< zs(CbXQXR`C=Z%hVHMP9aiL94j)ciF%8UPOWwW!gR7h$C}8R??MX3cb&>hRZ>a0>hU-dgx3L*(afDRBVbO^5vw&ItpcUFWLs3>-9ct2fW z)`uiC?RkREo-Ep5FQAc3ge|`dr)XqEMM*5{rWPP-@YZoI(}$eBHEQGEbQ5aiR{%d@ zq*3z1916{eNORL$33B%ZupWWguORdrBxW53>e~t2bLc%y z#h%KMePq0c2HXA_ntW)XC6)k-KZ9~0IB-ut&O3OG7Hkf>yL_qvMg7E=;G$rF(h%@9 zp>u9xEhA7_MLOn;Lvf#pbz}Vtl!V0`wI~6wPvo1;&g=1hV^0^TmOmE}KDZB7_Da2~ zzi>r2Gv$m2YHKn%EuR6PRHDj`eh+5AxIf*;=)+2QLaLp4*v!yBblp3p7mn=WInLg5 zsbBfx=R_m#e)&A-mkOoJaaFXylj!Amj}3uI>PUCTJt-eY{5Ji&OE^oK{xZKy%HsF1 z)>*VZ9YU$W6kls(ul?_)y!plYQ0Pq1#2B?GozkI|ignK?TxM=J^`3tipRFD<0KvJ6 z!M>NwJsG{uCd1#}v~{ccns@89t2l_D1F;tQi&MrFY)TA9>JO^ZCun~`o}fQq3F`M$ z=T2hL9js2V2v6&Hib6FhaNOf>bB8HFC@g`7%~wPtFO0fBWgz+#Taq}VS@hmMoA4>I zS+z28rISUlhV=dg*G?%$tXD7l0)hBC(@^Lbk}EQqc-b@XplQ}Elnw>72EjGnhc6Hq z*YR(lWa}a7Zz)OtlVyG(PFQYymFyr(_vLA+SdVhIGhF4?5r!@SN+42s;Cws1hC(!y zw0AwUr?L>TQKWNCo`}Zpa3{eMfygrc*rA$SVhiev`gKhjRWEUM4Hg``Pv^R|X2wQ^ zu+Q-gjG4;wiuN%XVa?q@euL&Acf|?H3GrFbFDxlvA!~=?0U@q-+EBzHc=}H4D);$l z)((vk5`}c-Gb*=hHnJCZ*`|vj&_uPw(hs`KR)_1UfMO8W225BtKy^4riH9DW%pVP3 zvg|Gjg>A3Ion9lsy^A5l{u5v9-|oVO1zC>mhumby>76}agF)GV*QV;dvncvt4}SYn zVXPZNC^x0Sbm*qAL>Yu4vx^Y@Q+Ab^a>rZ3LnA{yjkeh$Op#Tx$ObUa`EJQz;3Ztv zI@E~QYEQqXjjNgb*`I+D)(z0@Xs(pA1G|jQa#%@P5Ec`dik8?T!nl6RJ%T!!Qi{?Hc8kR%_bujPV1Hyz?4T$&O0TCBqz|}A(kqv z#IEl$^ENqhZRNOL2n64PZ}Pm@K2nG>Gv9Ip1;lA~vRDx1<@0NcFWeG&4bEH-HDJK| zK&E>=DwxDH!w?& zswiCv4F4?yli}u9#m}dqocK(rj8`0xsWs6JR)OoOG#_EE9{$=kNdR#sWgQz6MW>s5 zJAz;a5-$MnS~r@Zg(GmtBKi&YM~VmcSid@9&v^NeCDSs-v@*fK+gG+jh1P<1d?8`w zCI0Q=e#-2XyOrs6&E;qNL2Y6~SF<|DF*vF1rgl8G#+7P&<83P_Urj=L)Up;)hU$TG z20q#5KBHvLGSPmbzLpcP=cf%1iYA)4TpxNV){|KD8`%u*(NxAds?&0tH4cI)T-f1G z2aVkPPGCe%=>y#HJU%3P#3nc5=@o~{`c6|YM1o^;8lxxh3YrH*&9vpmn@5I%=zMTvcAL`V1*_@L0JTvTr>#1Z*i*|~ zt-oquKQ`04ykhr(mldCM+{txAq!d+Q?NHQZsh|lLZH;%rCfd}6N#Lid%eIImRA1G- zdhjT9-lGZR*$&ov5z+T$W1WS_)?&QrZDJi_*^zE;WzN_m0a&6>Y66`sENZG|XOEUA zUX!&JZ#*%_v$LJf0u>^FOk+SZdgF4b;}>5EU5hFW$I413Lsgb7yU*1{HFL=G4gvwi z&b5Z1NSnK4!{Nv>G-DCtF~1+P@L(mwI2CeSR;dfocYV{`UZD*?$mmY$mbN`@yyQ#8 zdBxa`eEG;PV0vM(<@k^+&a1++SG!1C`x>VU`xg_3Lr>#9D|%l6T}XD-i}g6{adzkL zdKwCVY>fVDzXKn}NdM(h7sz!TkuTP$N3#c)Dk$2{ejyiAfvS@pD=)bVvl(wS8^s=B zA9b49(~ZrNYe`-xD#Q&(E%LO?NrA7+Jie}wORzJS#=rFfLVZ@%5L2n(;Yg$YAl`2) zvHddmnj~S}-`ypPFlkuD<8LyfkUiDe4GnA6V_eJp`GbBiUXOA{oh8VrZLW46dj~mI z-6zGrYCn!>h3QK0+DOVKQvlm!!%dsqRB=_iCfj+&+GqD-T4{fhJlT@}5O`g=O^tLJ zU6d2AE#ZmgV?$A96ba;zb$^zLnPY$8cgf=w0x69??9Al2O~+T@n?sH!WE6Xxg|qis z+1IVig2QpOZtDFC;<@@nVHaqj^9E*Q-Qfpse&PC1Zifw#hOhz+jw^9A+n-yzKZrKDT zei1gg4GB5{i^5FEtEfnUKekUR2^ZJ8&XjyrMLn#Ql@)*v&|GnBJo8?;za2A;N%Mz_ zo6yS0hs40<)MY)h9F55#?4jRO+#34mbShPktMy^?4O}WOl`W!$3Ga_^a@cFbh3v{l za*V6DL))}(P@L)>KF6^^H5YbU`a;sCq^|igMFy{%a6Xe4EZ^9{ws&~JWvzUxm4A`= zS^A9R#j-QJvLvqhDbFT6$grNMEwo!YxAz%#OSlW+$KbT=6;Fm>VI<8chK05(VB_vY0&nhM8aZ`idDk!!08I(PchI(rjL z!~w+$szzAc5N)GG(+JbCs+jPkB|AxM6=1q`;x_F!v`@ogSb;&F=CF`v?@Hp?qE3 zZVA(KWM?*&ma50g{2#rK3!fgfOBI}r##k9jrWgvkhI_+rMz{Ydgw+7e6&Fs4j-9)S zElc++5LeAkzIHSVFsGLHt-SH5R*=vr$$sZ`v!Rz;HJR-?eZ9cnp?+PS>mIkY{$dz!yuR?O6kH57*dY#u& zrYg_c)FeU+9Lui$bZqV>`G4#gaq}rcoeMvN>-P&G;X9JVtT%I@xnyqdfuV+Mp^ppc zXfPE7RS)_qv;O5JJn$RjHE&S$t&y}KAxyTO_MLTY3Vm>Wd?wrdWAmB_lP@aWh6rj( zm)WbyMS|}|cDIbgq0%%wAUFYwm>ssU+}jCk#@H3yxDjt=uvJjH*!&E}eAN8dXNhIwD=p&T0RMb3-Byt<7zseKva)kxo*b%LA zeW-Tq!OQNcT2*T(j)Eoi;v9NE*3$6Jzd)h!^Ltd7A~^$=ru6UcexdS^AcO18F8=xz zYW&;>@$LBN=Veyc`JY%K`+s1G%pCs(OYG5^bK8=1{1@e@o7i2ZvN)ZN`0h5uA+;LK z_Gd#oq08>!(&vK^8NGoVnjT2H=*{jWedrUCtGKfBe@FaSCN7WP4E#@RSTo3`FeTYFObUg^nUB2~+_=B3sxZS3+N z6Z_5j7?OtG>lYZ&$ zQmm3gChyuu8%I7jnGvMAPZEPY$(0ll^b21nV^A#A@q3t`tM-Q#+e(u~?iUY(DZo+d zxL)LyY>L!N=b$B^EHH;*a#BEw%S;SMoNMadBv8L(zec{^D3;hfkRWDO%vi{b}DQ(yr;XmG;y1>-!+? z@ejnq>6jgZKzAu#bccNeYraY@dB$v~hTp%eVx|D8&4V?b&$AuMkvZ4J`nVzoH@*_3 zNN0#5dIgHLM_5bQLk0q0KeSNe$NLK;n_*5@Pn45V^t^3ddEfAxBM`p~S%$g;%M8Oe z&)%j-DV>hm8Rtxd>3~R|~bPTGsM<6ldLTJnSeA=bx z2d1=rrdbi9u?4KIS_$kJTvV-nPYY@`P;e>rXx1>K{4fx^6To|I1oz*DcDU~O^KxVD zOfV+_pae9H!{fFZ(CgG5 zvEVk#g!zTyRjTc3&$W~%Blqa}z&0XS>dS7As_}+C8p~DQa{k~!1QPIT3y~p2Q1pS* zNlv$L^BHdSOp?EjC=D*y9ei>2R!#R-Z`z8OBtTX3rR;6NNmOblI_|Jz4!XKGf^)Ru zE%f$DM0ALvRB=@y&0BNV#56(Vte8b|0_BXY6EL3AP9JA03&c3H?E7^QnlU^cUVAeE zQ+B;~P_qEVIe!)E3sRWBcFgN3{cb~Ez?kzFG>An^q-}kuO|H5t`ISh6mX@Seu-grX zV5<7(-fFz_EVf)#=XFAd$i?|HwUd3SY}3JpKiZ#~f9YYnkw~jO7LwhpJi`NlNJ3<` z;dF?pk$trgZ&i$_;B`%WHDRAoyhh8c+>mU;y&C70l%e{k=DfPT~Sk72;m$z zf#R#--!p;A(5S_SNynW|Ip~!J?QsBH6}`;}9&(Zi1*L`>#p z6T2oGm3sr-*Wcm|ckR;R2J{5KLF%J;m6d2|MzgN2viScFN(-ZSEl3lr%{`#kA`Ghl zJE*o_uJ_LxSNibkHD4XiOiT<7i7+!nb*bIdO}OgWo=9!vvU>dMSKrGiaZuPzO~5_3 zXZT@RUw}inFgDVJ&uwSy%X7<&#|jm&133S^-XMhAXod+Q#~>L!17R3DJRZo)rEUjw z`pz{F7$zd?xvAx87oz#Dn(>@KyP0>Q{N->Oe(?{A zU?JMSW{0k(1=iPt_N#)|9NgrOjUJ}Sfv<}U3uh74zEz~=*VAu}7`h)Cuq(q2RXw&3 z)2~T(+&!RX(eDLi*&`HNE`GaxEH%2vaFZnV*WS}v&M1tpYiR|c9J@$~C~u-vC=*tI zCG=ILB6be(f&F!x*u%zvVXCtv9)rWdlVrHxTZ0tDIx=w2bkU~JCmIJpC7-gEo6BTn z_#=m33L<8>6sGxd-0+a@{;?{CJOgK?63^8MaBw|sAgXb*Vv+OR9QYte2L^u#;De3A|;!`KmR1k=Bsd2a_V^8@lRih8T@s< z#kzIm_DD>I*@oT$0r}!R596QS|VIo8!MQS7k z0p}3c-Igk9%1F`!^6yKjD5@lB{D3XnqMODVo?a8?c#sTA ziNKOu)WM@jPbkJzxWqHq$|_n$e&zZd-nN=G)7b~X34b`hgrT+*3t4Q(EqpzN=UAZ@ zkM7_RP;`0UyqqcPR*v|J!kTF zsJLrxz~`M$eKLp+KoS`}vW2?iAKe?T%r-}+7FUst{9F4{6!ms2<7NcsAdR(PH9 z;klah-rh{e1CQbjw4Y(*qVqjf4M%E&E8rl(lXlTccz>XwpX!D`t9i$*o`u`RDbO>_ zgB@&d_&e+X#d~>1Qc(=b{+|Ek68Y?>Tk{!izYFpF_T$(i8!q{_$m)5v8@#BHZ9((Qx2c}b*~nCs}w2x=Vpj6Qe~Xj~alnwdgh zs5AN{70lty0ok6by58!{Wq}pRl?%plIb?Gg4vGEH3o5I}zn*yru%lmQMfAY*(O z5HO$%485*rZ{eAbpvljHFIbZBDso|WrteqD{3p;lq5|s~cb=}*ss7{o*ty)RX)XF} z=JrKH+LFRy<}=hB?KECB^>EvdG5czj$6ACJkms?f16F}E&a<7{%ls&A zL7PHe8_BNkdVcj$t5-)N%)7av@-;yfl566gSFWg)ceZk;4!@Va$^J2XvxT)fZ2GsM1 zrxt$f08m$|8xtn!YO@RcIKiPN5C0SnErL=DlB;j$#9McNp=pAg5?Go1bAJ1%0hrc<{1C~YX(%2h?{KK%vG~DT z=vmmDE`sXPI1@D-Y>R%2IFwV;tS1XMR(LO$e5EHX`BqKvIt%|4GTTQIa02!-)Q%P@ zHZTSA!)uxjf?2wK^7LE8Hxjmf4`oyl$w|4vC92Zxh()1@wEm0dr^tUVtxlaOX&u8R zlU&$jNE>A!5=-<(@FnH$3u-K6kVkCbA!D$4cnWyyhS~~A>zS(WCz)2+*TjdeMK{c< zxE$|9BiVIQU*bJwl=%4@(v=9J5d(%wy-NfA8Sl+E)4t|sKR7Q8T#Ru0b#Jn zDYorJhBU4ygUyUx&2WR=me;RUS}{Y++~KBLs0P zjjRjM2PBanp8?3jE%6?j3Trm%n5%yrAC%^WM?d8E*}LhIMfu0jTDMns&6I~V7Nt9C zrNRk8A6H+gL&0_&O25CC;EO%Dh5hOiq*1Ah`YgWLGR7Xi$N2!BACRgi^!5OFHVdfz zTE+RKmhuNWvvHNcNkI+_(cK5>?Of%vOPaYgc=CdRg!R1ZuL zq#GRq^#Q*u_S#T}^K=|aK&f`>W4K%K^uyaNI8Ufo$}m1gaN!WHT?@1r3u~qcip=xb z6m;THV(}#+kL4rQyR+aPyfBWz9X#j((!CoEZ#{u&0NEJP+!N7Rdf;cv3SOTYS^+;Y zZ?NawIfSIbAEu>OHA#e<`G8F(YNlAS2rC5+q_+?|&->{f;#BGq6A-U(QR48D$MCvS z*jN06sP%^8u~ojX*U7arkONS^g5w*&(WQtx7Gq>J$1g!kE^*b+FDJfLzc#{vOEJ*h z_}Fs1f;&cW_X6N5=8eDm{)l|u(cuH2=A-hL{a2Zont~7qbDwR)@W6wU#r&z7q)N|k z>%h2)PBh;?f5LWX7ND8J@pe5=TdN`9=&)(qq~v{p&QO2DdV1=ivyH$>2?;kyp;b>Q zaS?^)-8m~A(`1hMYDN<~TU(TxDex;D_1?X-&fY*l+~IqTNO_Nv7fAo%>S8dPQ@vgs zPKPI#C&?`poiwt0$@Qen*oVpn=v`OY;Jed3tj(rrUVtFC5PQcNG^Ha^Cuz25R8^jH zc?hHRmBA{!XU3-iQPaW*Q|-^KEAAx<*;mz$b?X`Fbl4FhbWScaQB2L0oBaO! zaX?6(=%Cl$eS-s-s`(jIB%1wbVsh`5f?>$FTuQCFr(lQ?afwq6x=P33UcTj>vDU!q z=x%Qp1Xxry$VmXDxj;6)bq?TyB0z^^LaF}{A8#0BsX^8`n~wWQ%6GBnM!7HhxGpVY z?HvdnPkZR=zfHfMEa2GcQ|+i34A(c#>{F1uP^ICv&j1>S-$SKcPPvcuB^qj#Ylw14 zOH+*6>If)AGNa2Z`$K9c#d>KERSejEpn>s2s}1So@?ax2)*892-|NHiaS_&M{efo~ zcbQWG={Rl`4{-o$yC_9e$ir;4Bb<7T*j`|y+p+KcC(71e51r-}udzCYO*THe|ELj* z8TKjfDm)NgxXpM*E;h<*H>;1?mBoz2oqlbF@`cvNVB9lSa$QAeV}GYEhRx|(KpCnP zb%SygPRF9BOqrnfR978+u6_(fTA9W^bh^t!`=+y8O(j)Zl`VRT`WIlUF(agqZul~g zRd58LZ<#9{0;wA+ULDOq1VnYp@=n}|43u&3Awj!LLlO2^3v)zJ4`L3J0|MmbR?Xu| z1j@b;0XM(k4g2OJWkTzTAHWk|y)?6zy<@6{{vi3Q!3I>hOx?BVLchTR@cWz)`e}KN z*Qj#Oj40yHH5kf&D8dt&1&Kr42yTFVSJjj!xMjW}o3Fk>3E54i7WmlF0b`lyn`Wg%TguTXYg;Sc=m5YVnty%DSG;-IgAOkE)8 zcThs>hUZ#P`ki;B%SzZ5wYk`~An!_x$nH4)Aqrm;(h9=T<34ktZQibp)Gz#Ax|>!p ztrG`=;&IOVkAYDEXB>=6@3#l)JT=G|ZBu%oz6WeRJ+JasnB{H_+L@c{^Ffu;B##X^ zTci}_-O#8M&HV*;Bz)jvMW6zwEI|~zt?tWGXbAK+n@uQNtk9AT$F0|(;mKF;!vfN9 zcoF))`q>J0^zSMySooDDOIcG3eF9ew%v@3_Y#hGK%N99_rN?J zmdpBuBQ(zrOGEq{p?i#l0$puegg~ObXL0@PVrw|-T=x$13{5+ghGQJ4ljp=`4c%nJNbUH-Z1}ejEc+Fk4G2+{XH$c=>cG14cID8VB z87lM&e9Be`R!7kU-!;$xgEpX4@_!d)8(Sw_?I!_tU-IV%WkbKg-{*NI!e_nF{1gXa z+>8Cu9eO4F6iDcx29J*fl&+I3u_SwsoK?9WRUce%Yd1#XY#MSu*dq!1&=j*t7$~ZTE3*=e5ALF&b^@Pw1Ho<%dVZ`R{VIm^ zdDfQ+lH3~x;{1RM-}8Us*J%3SwRNq7!nhTzgT`X_)&Tiq2-~IPB!*uYuTF+a{{DbCP?*fg=Ci^cist8Nv5X z)?M9v(BQFJlK8$2MX%=e+vhd4OJ~lBOI0p`;;O{4i1ZxgSEV# zs{_;MPGgK%A0K3F6i>a6_Fsj$uoIu3xd$&XQEGDH1j-`ks8-D;^`Z<3qDXSSFp51e0~dVi0P zOEW$7V(9DI@6(y;8h@v0Xlqog&87ZW5e6G$vg6Rm$g5tYp)_XEl6XysHwLQ4%NbKl z5N8N;3ZrlcEs)(|I|R8aBd3~5w#11M@QXAOLK*7%I5z?dPH{*$rM@IPFt9T~qW5ZRHN@TT?=P zT1C9FXR4uP`!B}`YYRvBbtLgHE8$Z@TVyjANoY`x)i5JiMr`n&Gmc}*THF{tDI|Px zVg{$hVpAXuz;Ul@u&iG#Xs&q>cnw|c*zmTSIp3_>Sd3M;%1v8q#AfD73TD=@O)`L1 zU?T=NgL{bF!g~zfmswz@zb2{>mXS7{npk^C-n5S-_r|HjzeDJ|Nxm0UG;DKxU*0u- zjo4KPo6g;}1wboL2%6EJ{t^<5QRbc$*TJTh7^rm={?o=bJg$;6b{^;={boY}HAZtO z>;2$%AoGw1o+MxPW+88#>?y!q=ty2(3-ag`3hzwR`TRF9G( zX!Lz~7X;Q3MEZ+g>C~K?(^@KctVhYS)at61R`BoIw{7E+)(o_4HSFgpWEKXO2)y}% ze)Arn%`Q9L`ARG2-?d z{k@sL*r#hw8-Tq2SHnk3U^_K*Djr5xGWX9B2YIM3W_QUa{x@};c_1CXgwg2JjG!6u zdVcT%&{Pjqe7UC0kM5LHiv}9d^nMwdMGM+Mrpr z3ePjJw|{!HOYB2!k<*YC!B>j%Yt+tRwqDH&$wRB{$8!WxZ^# zj4CUHEepk3HfD(>Wp{DuP+?e=Ge*UkcK_zA;nl{{7pANNGm9}xD@rb92OFT915a?A z=S?BbH;(HioI-V`otGjsE&%xSCg=Ytn*XmxvN5x<{WliXDc64^<~vSxdJOHE zvzbMK@d&)iSzOJMQ^{nJa*mhJlKf8n#FE0|z^s+hmS;=HcAi{Ozo*X#(s2?b38TO# zv8SAx6@kaijE7@zkS6qB4kfS09_xPm8n&Ze?4Eshl0IGge(V~ye~yiACNk7N*c&UP zUl?j%bP+o8wj{Mmo|~4%xuu>NQoRmdFJEt84P5Wwa~kvGQk8Pcu-^4_uOjSpL2Uxj zEs{yKs`a8c3NtHc8#~kv9G@RXqP&Jol*mz5IM6Xzb&<)m0gisPMVzVZ|Mp_PVc8M8 z{YbvKv#e@Hv^+Sn9z;9LHC8&W_zyOQsg8)@uRAM`U6p8^hz`dNgCbM*UGVfSzf2VV z_+EUb2>%TZvaaI|k%Gxx%nA`lIwg3i90Mu>uEvUq$b51e6!6)57ac@PlYv#v zQ|h#q{pED@$o2AQX}rq%es+ETAslw~nJ?rgqf#qM;p*rMNKzO7W6Wl`d&gElIC zeyM53-JV!y#>@r*+^X?&jJ=!`Vs_+ii)sqQWOyEL2LP3Se>$;85ln5wf=>4-r}8S; zH3rOxT;?7UH6+B-?Q~t&9J4`X_ng+@rZ`+(o!e;*X1NY32Rn!a=ZDfko4PM53Vhxq z!o`FJs2AVM2c7(kFvc%Eyjy?Ph8{p|-ke0`D&rydm*;;~VZT(xJ)O`?m z+|u`8$mWbD{bNsH!k#S8+cV(>$pU{9MC;unn`m}Gg~gJu0Br=0v#h$SO|i$LZ$7C9 zt%Cwt{t$ojSZ8f>#DIPOq3^pX`^(IbNRUgz0+BAQwwl9A$jJwK1DUSv4+95+4$y{Q zl=HV#g4|C9=HsY86mCO6Yfta>&H-=qD1x&6GV<>^r=hUQ$-Z3$93Db@bxgTHI^OKb z>U=7?K-zUn@;nJ}SB=Mg%{B`HT6tj5KU?BesJFe=l!*nj?oP)}8!40P z>Da%b-=hI2Xr)!-&+%@LX^L7btW^V`q1b zTqq}(#HyqqUPI}bJcaJk#4wJL(XJNdB`Dv($RksUC*WxXx+2TsVZEc@^bVQzONAEJ zzZ3Cm(w4vKoUt;$wh(*k`FsF%nUVUtc5Crnl2sVbEFK7UDHSlqn#N zP9+u79U3avR&r7CEI>do-5_{DzA#>ol+nc}_uSZXyDJdNueL$6_w+#QPo+)6w`{gj zI5yuPmQN`qAC1}!13fSa%lvJmY90lGY?o1Bx-=sqT2L5<^% z)j#91=t3SoKE`vg({NyJNYOCErak!vUPKMb(0oI(h=EWsOAUy%ulNVE1kEbPcn(^J z;2Nrb=ep~GW5TTb1(lO3H)?M}jY?8bfEB=DNS1Cg@26nqmm2qIvWRw`(#Gl95gTto zaL+kUIo0UL2+nwQH1i7&w4!BqMaZrDs+#V4k45AzY<7#Ldi5JfU8JaFkt*2>p}aUG zcd|8Ji=0ci2vD?S@3*iGbkPxTl+e}LU-1PV;CbOKLlOftMYG)#W=ryndxx^Ujp3|{ zX#HLlYAVvo$s9ypY->F*UiJ;~!9$hWI%uluGG7NjEI7o-%ofqg>j*#vj@*20f*)}N zNk^6Kws{FG8!vDK8lCJ3wg4d1^)^y8u0%kdFEYaJ_jSj-4IZMb_SAs@C(~do- zi^gM>hxO4lrp%2j$FX@yt}dE(BHw zkvt(#_#=&;@BDd1Xuan1O@EdqL$TCSO5e6D+)|o#AK$8?b`D@F1KdyxV1G89pY@*> zrYlwCxL?FX%3LB*6D(4DZ*mhE1d7^eKhlWXV{_pT!D`X~*EKdg?B7}DU#L{9l+3TM zB9JIqI{nLz-_T!AU6%qDEp?b5R7Spbdp?mOUjn=O7q3S)05FkkaFNepYo-{n7`m6I zix$gaww|FZra4)xL(Q4HVeIxojZ(u5HVa8f_^InWdsCSqyDn18J^1)2p(J4MHW;nn zFC9;LWsLdk&nOYooef=YlCvcQ%Jv^47*9Wu^8=tlI{_&{OqM@@n8!I2MrV2L+pIKC zAf4&A5Ax#(45;;`O~0%1NayM0rlHK(&!V)U*}4TXJkI7*Q?|XLY+wB-pbN}J50a(N z78`&c?BjP`hc%VJAN+nkJ-FF7Yxm}IK+O_2KO~})W}vZLNTi?7fS6tw1!q-fX9te5 zuM?5Q6xdnUU)_1-H8T^c7w?{QperlOL$I$C&K|Sl>wlx#DU;cxfD zO;Z2=&l@)cQcxRx9_YHNG9 zV=D8WePS_?VW_mgG^WKnGY7+BmRcES)rOb}MNL{fhTvKVF$&fRPEU8yb&Aa2wmy`j zkkP3yVKXlvJnVp9Xgwc<$}O6|(A{ghP-Z=8LchZqw!`0~@K)U0-zar8q_yOjzk0Kh z=z2)ej!LTZ4X!J0<+_55w66v>;j_m0fsETpidZU2GS{91quS5M?)E?-sieGi@W0QzC(7HvM&6`ZNA5)z73<7kR;8CZ zTN>A#V)4jiqjO`tW2Xv>EQIVVh0(j%?>j^Zm$DG0^F=TfY}{ZprP^iie~(6OZ6AS)Q;GvnV_1%*Q#V zEzgZgeg2R9d2Ul%(Zz`o3{!pUvgwocz?66)IFX(Ful2EjiZfv0+v>v!%6W{Y!@bet zVp?FG+c)yW3hHKBpzg|iOte=IkK+$F{#0rS1UchN*_S1FF}m%T>8^5xiQdFbzl}xs z!hz1IIOMSulgl#PtCnlD=v;RE`P`<~h}5&Y6%J;WQav_dmQI$k=NOMZq^k(roYl56^~2bi zW>?&#+_rN;q_L}dTf_X=LFT-MNR5iJtdPjW*4>;aE9Skl1dBn+9;^l=I4mG1?;{&0UDh%>yZX-7j!2+Y7M z_chX+CdlYt12swhFJaP~?l}Q-amJ3RvX%_EJEf7q8?;V(=lzGrdqvTY}7pvuq|<=T>ny5tfwK8s7sHsUbI z-H-jiy$Bvnm!p5eTAHxOf*;G!FnovUlw(hqmid%(F zs8}v|Bg38IjqIrc6BC$IBPvbPyK9acEJdpQ*AmX{920Y5Ifiu3v#@S&8v_#)k0BlZ zY)_S$TD!5Y z9tNRSW-x&|t+@=$aS_JOTYR6}D>xV^A3MMQv#RLp{tZ$zFCTz}ZIT1I&`UasQSp#x z7!7+4C1wLMB@Oy*KLh+ZsoboPAm#Y!>mNhg!u*nSwgDFVL~|>APO78ghgr1C;};L% zcgr`L1cMY2UHR>P0`tPjo#nL5=ti*ypz%O|#r42eNk^VFUY|WwRmTvKvTL;74tL>L zl{e$sB~u9>b{zZaq7hPAyYk3VVhU-4ko|NIADH7fcOA5zxw5jtv^wOeTUnc2*SmH2 z=ei+IJvBCVI|TXi21VMrnc)ILSF+jad#GDGlo}CZ~7nQ;(Xp<+U7tc$Mqpn!Ur8*6r{1mB)YeQEj5>Ztx3{ zk_`&lQ7(XAxd-esRYKydD)*g}5D8QwH%-YLJtzw%RZ7|nGCuoL=5bIQhSr-uq4xe> z|J)OTe&>I+Fo(PII~m!20@64@&3AH>`S1ioijkkQk(}PNJ+(@yeTrePG;;Mnu7ZK$ zZTu)l?kY!$#BW%aRVDr%F%7Y|hC2aTe#!g$7Y%eY_H9soh5n>!BS4YFi=idq`sbTK#S>J13m$0NC{Z1IUL z{NTVmh_|DEx#MaLtwE>xM(aFTb5_ymztm$0m9xT(;UllcDN+ zpHqauaAuo4T*K@9Y{wO9S?KTr8+}T12%*9FGH+(mNWk|l>3&D~*O%`*(q@H|Ct)Ap zcziprd-K_Ue9eD{wQ`K9!+yP=eG}|O2yi2uE=rx9M_S)2$WVCUsM%j@= zpC~am(i33NyrxZ8l#JyI-?!3hc@L;Oq#^h_d3$Y9sj60A_!18wO(ppg!jasmgWWFl zj7#sxzC;PC(CGuIL2EPoN4USoo*x_%Rp#DK@UISd+FzW_$2-`3m}ApsoF!Vmd1t9G zqU#;q{C?ARuM@2ym4XYWwgqh`vQGo3GG)S3F!JlI-#=%h*L##@4fnoDHv)5uiVEo{ z!3)7@OaIZb%JS(NNm*(8vW<(npjR_?A}XU(i?y(Lp-Z$U$vTS_%#-)#Qr2cuO_5s+ z)}PzB+7N!y_GG;r0D(v(1jOjk;gyl+{OaRGf36R&Rg|7x|Br;R(Q2$~h_)d`f#rQ< zt(u1XhOaxYmgY0P+a^&OWM<-Nn|*_5YE4J{hE#zXbIE8M;H--7=l}#3?iBfa7f!>9 z;QRL>la?=2BYf02Nv82H%0Pb{tDb>FFhqRGKAP|9yWRj9Z^3^NO|Ji!XtHv#asKZ_ zvszoqab5DK$M$OeLG&aoi;?&sK(!srI)$XwH2KvP|i{(4=dvj z);~ZCaGmGufQmo+B7ZZbTUHMad$FY73740(h$%0&wz1a->kOQh(5vy%N0?9>e>k-% z)-!k2HjWnx5L(FEeSo8QIS1iO2vJI*X!>^XOYaRxlwVy}#n2NZH3q5N#xLO|HO3S! zcqZI4>|v&_JzOVR1V|a34k*kTTUPaRlOMSo zxhYq%5ZFlFkZjl9F4SGYoRu-r=B@wC$tX$8*Rj(TCuQNp%-j8?p!%DP)f;=oV22bP`}SuW;k0KTun0F zompaeWIZen^2QOjHfH7eZuD8lXEa;mD(yC_CfB&a*~hY%I1YSY0(tcBfF6*&n#I%vp*b>5DtDGs9*n9s>|E)_GtcKM*{RS`m9HQL^cORuu!z+ z>Dt-$o1S3xn~FE9s6g3zuxrXdO1?DekJa-F24Yz{=B0NXc&1m|ja#?pXyX?BC`~Vj z+QKT%(qPDSliwmbB)qQFn!7CqA$dkpaQZz?OAEPCw(NXR7AvxEwi}8K3{oX5=W2zy z8*hFO9sDn2*t)^m0lS+$f&15ou8UzdE~JuXkC))k#^@@75OunuSXIbuqkSvVKg(5C`ow` zHfMWnSVkwyYPwc0X+XrrSVE%>RvNZaH(bqycbhzx3?D95cmQPxMQA;CJra0IHW`C= zjW%`|A6;DCIBX8yS&9U}r|GXdkaY8v6{2xF zxLn-3UW5jaq`uLIcz*Bti3N7d^%ms1fHVq|O>-Z5fT=p@wNtw&B(>0WBy@{}Uu8oqMTcY8Oz==VqNG!`X6 zxQ-m8s3ixilC4{vq#Pm0k8^dwR%Cj4^-eJBlp-v-8J67Jiy5sKyPXG_vXbz^t;Bgc zpyLVXU?%!(YSmtM{orRk4g|dradGyPE{(5oF3#}gxLlmWf~X^{)z|+86}(Gi%q5=&0w_4`achftvDfiYV7f_= zWmS-obJY;a&lx)qx$8%d?Sm2fkf@H}AF1BLX#@LpChjd;V%XEiX^@3okzANJz5UYM z?X?0Y3i2Fv8;rTj*ExRx9-iM8oL)h#-FoH)L(e`2pSeRA0H!F~ka*n*>UB6=f$sOL zM!mcEse8-aRm6UkT>)*B+%l2Y5ZR;d0%GuhRnFl<;_y2+S?fGK9Pr#a>Hpe!`64D1 z^qO1sjdX`0X;Isrv?UTAFofD@!RPHyLnvg~lO`(g7_eYu)98prn09f7a@yTO?+lM5 zPLOLWm!!|?MFxV89gmYu(DJ?F9dkqI1<|)Fs~LP}L~Iyk^aRs;!_gkRe-X1+D4e5*izz?hKsXbNcuFqB4?az zPQzZm{X=-$Q)u@|`hO@3a$YKf+L23=>+1+-br_ z?M^%1)EI*R*PBq-0&U+dFQt;l>+fQ^f zJfd6z!D&L!>mM|5{ElQ*R*+;TC;eBIawv|6<)I&&(Eiw(8(o=U^*D;04mH70j{J!oy1Np;FjYKTUm-Q4yNlLgZ*N0-2yJysV7kL$@8X?a5u5F!Kj@`4zI{9#H1GrU{Tr~4n zD}11wowh!0r}K5oZiw%IwSh@e1K&_C8mz{arN%&T$Npi%)cxr=Q~O@LoBULt5fThh zECnu@YKv~59tS{>BROek+-X}X*GX9936#39s~dF>*2Qg2ZA>?}i455x+-3rQhs$b< zJ3%eR`&MGWH_s#+&kar&iYX$fU@K~{;rfzk zZxQq@bc|-vyz8VlP?+YgZtO?u27Qv~;a3G;Q&#@8)L!rj5qL*^3Jk$UOmrr5qv4TW z+oAq43R5#Gf)APEFK|lL^Cn_mqgba_U2#RnRS~Ny!q%Y4Oh^r#aw{z^S_@kIUs0Kr(aN_o6~VrLBO3jA__P6Qld`&iXl_ zh(sHg`d7}WHAyd<&$;?mNCMrz%{P583A zOc}-+b$dw)r0|8`<>c&Pw-62?k>u}~pbe=WJk&wsNoM{al^z$9P%YFab1u4h$-f;j zip^`)(6IT4!s3d2vYK(Gu zu1m6YCJD{@aaih>s*rEoXsT2hrO3Plgzj!r+Liq?jwf{T4w8ZlgtnTHcLuN8~Fyby~mWxTYGsI;RCPq{<-lb6X zpgn|v-T=SoETv~|$dI_o>d&STbZ#P>jnu?7*Ns1fJ@gkFD=V=*e_lZNb}(AwLJ+O% zE61I=YX5Xn0C$n&Qyqc|I8%`swEmsP`!oYW*#6VZJrllR9O;kbu1DS!FKlyIkYDw0 z8THu8W&;h;dlTUtul0byJj#_Wg%IiXEQmYuDXm9kou^&nLLLah`s4(d)&uQKtFXfl z4a35)HgWdxIY}K3b8CJ!XN@(QciF=eXlpo>$7iYqjg(Y{G89e^TX z)^ct*&N+-C*T(eKHsZbo>o^dcfI71ncR6EWq0EL6&%y=AD}cI{WbtZRm=~~N2-!b4uGWhpKLpjhm@yK z(0BL5Z{z$l;FsT&H%l~%z9EW#i%FDmZ4lETe10{uGD0n7GGRaRtf1@RH+fRKlryCXn4th*_#Ub{hs3E#I z#QZW^n!k3B+rS4jZ-VP&9=J9&F~Az&*|)m`iC5!{$_ryK{8*${3Gfpx>c||9bS=d! z4IPiH^_07xqxJ*>wiR~SGT|RwX^!>E`dc2_`6F3Vte}P#Guke8r~FkiqGmoASy_sR30raHQBLMEmzPPTBN(UXfuymjbAM6pFPk1 zdjyi+C*8K8L-W{#5Q*(mij<9;Hevd0pJn7g&$EFI8k5az&bS?|Q-OwW-@1 zbLH|ndFJ!|{B9__#*SdyG2y_Kw@cCb*yaAJ63(~0Q3IOr{OYK$)mvlYbF>8ma6l)i zeQZ)~BseqOd3X*1C#A9f>3}%1qNRxIo*F|P7v264J$P#7Ow78 zLgGJJCYMu4J6&UEXWX!I+by>%In7K__$m`uzsIk%pU2Z|EAO}~xkd>TdEMi7c5cL z9M1d$LYf~mjZ9s4Iia#M)?7qugI(^zU8EMagycW>J~me)m1Dr+aMw>rEBjoMJLl)P zyWPHO94+ZScAUcgDu0+6ASsjWri6aDiNchB{4L50eCE9xelPONn|(Z9`L`HZ(h!i~ z+BKin^gLB_b90R| z*mT6*e-Eg)bVHHef;Rcnw-$rUcqP7P&N6LW}nsL(w+K2{r6z z8(eq`pCU3`D}kfuNiF%l&d*<&NQl4hPc32{zd-Tehd(3=sy?x0gfbf ztc23I4bVcHbZ|u{F6W^Vu7>xhjdYKfgN_V;+G(hy89Bh=A_aaHcuv3Skw3X66^wZQ zX&NcMBsQ^`wjdrZ`1rrVsmS+}{{VGNjQ;`ZSlHSBcTjh!^%FfDiT{ZnZql|*;JEsa zxRU=ZAgC_sSWF-z5QsSe*8p!IQUf*tDqg;RS?;<;bYnY?`bB~oOC5mMUVMm0rvmoV zt75YeHn1HtA4FLC!tm9-f>%7uA>qrj_wc~^)w+Tw@Q;(*{nucQkJH!Hi#;%nQq>g9UOK3vp?_V>Bjb?gukP$ zja+gE8jEPws=guJs+K!CE2}N^OAGW;?MA$x1}`E-!m}MY0KCs#$v&WA!dq<8Nc{ho>#aBfE+^sluV-3zI#4|6#R~oBn;SZlmk02jA>Twy5-y zkk72eIu$yVy}t$m6xk z?(J3gR?E%kS$}hSDYr_q4y|?PbuQeqG2!vn9?ez8zgQh1LOE!)UFA+^bV)aD(+rmX zO{;lxOl|FG=|c?jX`4JcSaL*sCN2KNd$`@b`uMQ2O2e1=SJ-Is3&9q;CeFY=3+nF; z6O+kgOMYX`OQkT(uE?R(4_>}(Z~@7!B7~O!gal$-S}#ntOe?+A`&%N#0;hX7NI zT(nFt;gRkO^8a0DNe?S#P2M5}HDs7L5o9@zg7EWM^J<{^Zigyh!MyQq zADZj{YlwEjXgEBa4X4iswE+I<9y?!z_|59S-*rr-<}+P=y90WATR%q(QLNht(3g+6 zi-ZtS6&EMr3g6aVF-q^3rqUFU zh`#sWnkkGS6oQf5!6xas6+N-mJ8*Qi8b|xxnHKxxf@Ay>FZ#{R2S?=6uv-c4pTVj2I;!PmnT<2W$@8<(MxU zI(6Pgs7Z9{$}V!~-Btqyjj((=BCH22 z`39R@w+n%v%r8If0Ro`y7bxxv_ROv?m>F^Ht{kZY9zlJzV>-MH_BJzM)L;&^x)x8% z)dDT`8xVI{swZqcqGEG3>vcnAuhBJ-tm8eW`NCd<{5&d*dy5Lgh8{&z&w>MEn)cPH2VHHM|Z<>^;KG9p8N`g!_VVjGER`>b?a_RH z>3`0`jT!YTftkAWy589P&VJ|Ney-K*th|49U$u%v;+OJI$h%h#XCFS_4&=0Htb-$@z7o4pw7|S!tQbX0Pb>+VYwEG|rt15h96N?E(La z;m)u|p@$|3*{9K~*AV{C4(s+{LJ7h!dW=B(sL(vqy!GDXbxl7u|jY%8N!_`lywRI~c% zfw`W*RKD=+0$$bEX$z>sZqNeQ3Ik6O-?XPri zp@3E$ZzVlcYu6p1&ajAp9TLEMk0mi7b90Z?!h&MLm%?MJ<1S;wR zgJgCUvFDQ@HTa%OYJiMXIE&*0=^dtaWyKUC3C!h(h~AlZ(qmwUlT=_^poxV{t<{+T zbXs3Z$+XVxIV@bDORj!`jK&k`m%iw}p! zB!<@iQbJ`guE18(0UdNcH6%Vc;NgEVLiNq)8kP+H2w{S&S>My888M8ny4M)|OaCS{ zwdnVCorXAGq;_>_9)EI@NdL^Lg?~be(qTm5 zi##aKym(cX#cwARTq?K4jG(qg z_w8p-bywa<0!*=C#(BXKrSsLLyh_MM(9)n=?@jj4Ost6|_86i$1Yp;7hT#;oYt5# zb}iSDLY1Bp=@qt@u;{-h`PPf*T7f9ry}JeHe#wE7IERfkHd{89!~O& zAu=P?54$>o)BC-Ekll5zNW{MT8N9?n^9wtdC@Y>C8hT|*R(9@u!~mx3+-{yUpzC+? zMefj`WQo!c28^y{gs6wk_{1Qi4~Nd(TfDme7|!OB1P`s5J$R1{2%9ulDlTdAgnI$W z6=Z+Cz3o4vA={9GGNc31<@1?Ck9yJx;mpKV_@OMhj685>i|46=`d%~rl{eZz6jgwq zCkFJME$a6Bu}0ee716r5BU%94$%)JvD_`*U*QOC$ePgd|l1S6go!HtD zkJ7lKoCzrl1zVpnYO4~Zi|C3I(m@eifv^GqAbwl&4m*go>>^7`t3p*F)p#g5!zAo1 zgZ9Osgh^NtK+7LI*iE&Sx_RL}Dm&K-<&Df*^KvWDmx@>p2F!*}<@eNlj_mw(MxnU& z4Q?@TYGqSQVU;Zg6Uy7a;*IDP6?d6s(>!{%-m{pro>60mOpl=3-$Q(iNiY}YY$er7 zcL7311TjIaE94GE45)7Jq$6)pxxo7{9bvpJA(!t%l&=2d_Q4r@OKf)OiSwjf6b)Dq z;YRBN4UJRv4KVQ=2FJ2K&+8#Vu{kfW;<@~J2RB20SA99&?k4d#fx?z$wQ;o|^)+Wj zP}X8t&t0zQU)XKe8bXq*&nN(SG$(u7)MPE4LRa&iCX*cK5}h>JeP~-c#0lo>2a&2jH;D2VuJ>T*nrPn*N4la{j=p}^ zJ<#MTjZM%m@4?}rzJW~$SD+a&p*{ey7WRI%%HC$*ani)shW^M4=l=ka-J0qPCF!yE zEtOB{v;Vjd3A!#SU=7=rvahJ^9<{w=_QC>xJ2ujQ9|>*wd%P!C0#bOG^Tk?D6j5=1 zm4c1H2szj?;OaJ$fNqkw?5&9p5-L-~g%quUds=}gqeB*7nRwx`)M#Es=fUZq7K%Zc zP=^3p1qzg&iIih7?KVlkZ#XB;SY(TD%!RaIEwPyah{o7Vn2_^75_k&Wpa3SsZS}4< zHdloI#nD+wUKMAfWzf|&AdjPoT!hT;8Wb1^P`KBKi$XNbL$`0bm%C?(PNfFJqhBMPn==k#+jq1bpnTkf9W&|X1G znp14HDx#TFA)0Yz)ONBXX;7uv?cZ)5H#)y4%GQF{FFIqs$)xozI%w5X=JJ=9C-iui3CACg=ra190^VDmF{ z#j0hWbjN*0Ok9~P)!fM~#oV{gn)ICCY)YLeuFjho?;LRK=J%XUn>tPh^M(G!=5a>c zxhC~PY9NHBe@5w1c^FzlbjJg$PS_^6YXSIQ@3h@MO%Bf6_z)5hDKPYGIiXi5g&C?N zLC&BvBwv>^9*H1;=F?7_<5r2ZCXcLPJ zh5f5qTAVyKV5qDL_s1EJC=AT=WaS2)#&8yjvA7K!FjVsD9U;Aq3hs>+^Z|ABKf`68 zQHEB;eAw5BWg=l*uCQ*MJROB*)BvIP4xk-M=10~oCFZe%FqW+N7u=7AN9?o2nZdPk zDzJ{;w*IL%Y7{?f-rQY@s~@nb+EB($aU0f~d}gR=>=MY8zsu;W`S@cKjv%rPHe*7qjOw(HD zym=2bbK+~*2gzP!*o80NaXRnRZTE|7J^xs2ZS0pl>FK5YN23N|cNZ`BYtf?pNPzOy z%iTD#FL7*STzTF?y({tj5`kV#?7(ViRjc0lGTZ!#}* zyy~(2h@d;nqr?j%B5z>vqBd245bl_Tp*kXakPN|$rBanR_LzIz5u3KTONy?}zxO9A zmTTr%-Tb1mId#OUM4VBx#eNCOo4MitlYg?-t}W9`85bV;c>NMO9GmIb(joXEEB!#j z&qTesG5Ukpu!Cz8!TXQ#lhz|Jo86lD05A0^lp>f+CsWGfyhacVw+m+k0g$m*%v%|Q zNj)=1SfitMR6sYg5NsV~l{GjqUum-NqC}tvcZ3w+zyuX*ePb&VYjIz3od} z;lSYfrlq!Z1EouOs7ye!xrQ)r(Y*&~;q+1P5W}FHw>QhRjgmBDMQF2%j8)L6BU#E} z%Z7mSdJFY3Aj7!H77`#BWS?Ngp4!4#XKvpTE~EC_i@Mg_`2mdP`j{=!|Dfzl%wKhVnfq40`L_HtX~?m(tMbOsrX2DdzC89${}`guxX zH-6N&8pXXT?bINFjO~5oz$g$r^;H)#?_i4ZodM*w&3jTAfjBZCfkyuDyE}+yBRE$&{k@ncHF4hZ{0y9JK4T!R2+k^9S?6XIux!fOrzb_IuiE>#76L z}mR9zk-k$(#(4B?FlSQt>^nMdP zD+!zJt2VZJSy{B|n(Z29a-QDI=c*P;w$L!Ed*g52XNSI)`}c8(DfdC4=}lf3E|y8y zH9!-HJ}cfxUXk0ItNn4l>pT?{mb^f4ld5(UETbzPo z{?mP5EZMr3|JIb!_4fAE$eTY|nHzbqZhmX&!epcZ@6X508MSVl5Qktba8zVw`6fgj zZaZNHb0L&ULtb%UtZJVR#@qb`_=?`W>=;pDujof$yA%;UmV)C6u;np9c{Xm@W-f60 z6y+SfA_r?PLDjpq2OTt!t*eO#Vy=cb7dr?Pe*m)R0#*Ewe{omGD?KqVV^9uT4v7#E zh4BD(TZuy?+`yj#=soZLxG5n>OiDck8``kQ3M))Fe(M2vb|gro(#rD^@TQviIw3EM zHBJ&MOIUc zDd#eVGLT9Hw!r7ktbBtcaaQdOgl+`hJj1drTcycYPkpRfeGb*gd4&o{tPjy%ibc>a z6i)5*fzJxf?JIrWqGYAaZ@zlS^mI^rUf`!I$a-Br^ehNZ;Ct{|ikX^cuxW3IF3^?T zOtTvWtU$hr6(@?SV|Es1gUw)q3m>Q_kA9tmq3$h(M?rEf#FLWakQ2f(&_}ue$otug zcN++&pP7@YDI~$@ml2gAa|G2b_RiWoVNN@y-;`&FE%9h$BZ7#>hXpnNu6TUS(>Qp3 zU9H;pPXoN`&aseo7(q!B-+{udkbEPw!9)l+Mf_LgMp00hL7@3rQx*rgksP~h*M zp~70Sm>ZS{S%+C<9^nMHOxs;(h|*({oUO=|c#S`1mM=J%bvK=*R<0ljp2%&1AsYXC zR6@51i*ktznTo2KK-~PYAY2&Y{RT(Se$!alK*1l`l~CyG%O#RR0z{BZC>Pk{D<=>p z4fqnyDmPGbACEZ8c|;TCG}Ix}Yd1y?c`Q5z&oe0C7iFA3Ds=UHFP)o7aUW6J;}#O; zi`%^iBD24!ZwMhmK%TZDtKd{fGUoe3gWpa^Oayp+BP%5GO{Qw|X3J!WZqn$U z{TR_u;v`H~A4mo+KYhf~xl@#hU`3xqG`u8hV?32(V=~D#`-A19BBxLTm42Dk!F6pK-<%M}2*D44 zis@l?_d{|%@8drtH5l}~0nplHsIG1$5PXW+vb8LjS5GCeShUjd}hn#?ae7VPIbn z;}6X&6xMNeZrNC->n9x;PBt9835OlWfOWDDdV=x0ZA&*`^3)KvY!6gq&j0t-lP!np zAes>7TDXZ7&~gkpP;t;}d2Qy2DZNRV+BH!3m<@@aQ_IA zemOWQ5lL$r|948xR|K|IwQIHfL$Uq*CDoMA54a&L(`ApwYY($?C$^i{KTl5DQUl(f zzP^Tm6UYK2Zyu*7FC90Xygt}(-&pIp>&?7DPUG8stJ`3`#K>8ypkh4kr+%w#7}E=uL)21~cwwzD}};%&lU z(CB)0>4wgN56b#W2PeaqnAxDuD$+NomumW+)g&PIaR3RW|@Hut|`-7RULqooLe>k`Dag`wK#O#F@k?4>y=@^SMtVz^1=k8o$MSx3HJ6EOT zD>1{KE^W1xXj4S`Rgx42n{A%dxKH-{Lhp{TK;s3YH$moi|c`_qD)zKPKlZf$CMW)S}n(b-Qdi{P7-j5gmp~Hf8N!!mzn5AmpRAnk{+Xq^-Bw z_1sT8*wN)UY3=j5^D<`dclD~)VEVRnUu_7X24$_P8w#ZwUrG$t8@ByLAWtB?eBfPP zKg7>lEtM$SJJ+Smf;>&l4+LR?G73FkoHEpNDAb7Gd!QY%Uo{{Gt*n<^KgJ>b5}8N! z2rU5#;N4dRPR%eQSOFQC>NHAlyN_0Kk(fj&Yoh#En$Hotf7c@V!1vhe(<>;Cnc{sl ze!7{JA}?lIOeUO3j@CW(iZdd&YI7R(#~AR>8$do4vkef`So04hQKnNt)JWtcR%D}^ zAwjhFXIy;DiXKCr97UAAdICa`pmY`y(vvuQY z!Y1hd3p0ZMow8Yxvvx&epOGjlVr<^hduh^59eXnJK}}w7Oh-IC{MG!+Gfy#1@r-fMzcZb!+)cB5;zV!$2)mqhiNG=S)ZaAFRWAx4ShebhtK)@Vs*9#^2W`S|R%IvKPMaen}Wi*A`k0M!N ztV@0%V4yhK{*_5G?1oBHUbN4o7fjgxED!8UcRoUT%a9%CLkJH0R=~)()4D#gn`YWp zBq}~{^9!Kc#HOqqNj3R7e5BG+N3Q)i!p^S?{h?}wQ^Ibi$svsTHmdfc)TRm z{kU}cp+Iycml(WEPBUE`^{=Cpswj52>dZ-9p6#)d<9U(F=*q7u>Ywo7@Y}$+Jli;< zW2>OFP}Jzi2Js#{X1*R$&GRG*xfL&*@6;mw!?pBMncAW7TdhMOuGS*5#-Qc-q0=kR zMWGSP^Dtv@O7386Z>`yLnks6{cv)477NW7HZ0kDz;yTSqPOK34f-XE@(u3yW6+7?3 z6Z-0Tv+Y|1NGYbU0nb)i8MB*UY!%#f=M<cLDw_fs>NDAR_>16yM zsWv??X8I1y(FQkDj7KC9dmnU|;0Rarm-oSO;6>d{m01W#<$%!kH1+`x47W)njT{qw zBGZ;EZQ_0s1Q|1m!qz+|G>;y$$`OL`M2tIMUFL+*6UNcu9v^R5xtcD{?Y>p6xoxYf z^{fPAb6hRV0MdQ{{iBP|Bdvbu)cvH!R15e6u0k-1ZNvI%7yn;Zux>>?=N;zNX$`5m z3=raQX!eq-G>ypAL4;mqg2`n?1H*E%?#13#Mu?BrvQ!v_1EwZeE_(s z#Z3FVBRrUmyI5&lRyDGr=s?ZtF|Ko31$F3JZJqqah0Ka^X6;8~#ganJaDOBwl6IVW zP$#`~uluyJ(}-V`0PYt<^=&fhe}OLM{{UV8x_R9w_3gn-{r+^QDIuKfYsw!}1M+x(WMB?5sSp1|Ai8?!v>Z}l$J!Mu`YeTq_sb!{ zYkUX^7Cn{wi5_EPV`!-w)bZ)>S$44nVChGkLg1p<3_0lW%-ze6Fs~e9BSyP@P}|^Q z`5@QQd=isMlx)`ApZm+rg44jmO6@nTycMU8W^E8}&Sep;S@Y-A)D0%B;nP>{>!EF3 z3x@sW<$4YUjB$@u>*Zh69l7%Y%=@cOyR@x3N;9M;AKvZ?W$d)e=x`0l*5*d2&Yvb;&JLA16zvh>Ty)8XaMzgjmpeaI0x>Br5bM9nR6XPx0byK&d5hD1j2>fTwws9{k8|&Liw>Inj z@w6}zdF6r^v9KWV7<(C&srYZv@3t-U=rQ^pO@`I$s>~THryABM;@j-yVm2zv)|hQ_ z0i7AvW%OEPofs(A2wT2!U0c&K~!|p3J$ESy5nMH0Zmcb%lLT9z2TrHF#6RT>u z7D9B}Ft-xKLs|y1N;?s6Oc4J8w-XLdK}YVpvyJALu}GRCttRn zfh)EQ_3Tfvt;$h|m+Tyh7!URsc0IL#U-JkcW@9{@&t<0Rhcm&!5Z?LworBmUatvp( zjT-%`Z8LpzS=O1gpw!;mJ{2$65p1FGN;i#2Jd|%eii}8=UnJ*htEAG*i5YUDI1k2| zvZ7+!*%4kDk!Q?J=r80&rACyTJ-P;JzpVGx_2gTBy1u$=x-b?f>=((jvcaT^5sWODJFA?oR;>TLae3S`06gJ%+={gz z9ZflYIx%Z)DZOa_b(@^=kh;AX$0OeJ!BiGZ9lC>+D-!`FzijZTmv7NV51aNGLK;cILQQ2YRcTXuU(nM6QA;aNh*i(nhmsPrKfi zBsftor!^gWJ+nTjxh|qT2UNMG*a}NK$eM@_zTn9j1z@br9dR3R4n5(S)JB=I|HYor~yU@YiCnKYM|p^(au1k6%gZ`CE4;h~R6SZUzS{)CWWSm~JUj=hYNe^|pm?A0kB-xIsW=I*j8Y4`1PVepptma4IlK>Ps#+zK|?6reF+!=Rf08zvWXSrw=Wg zjwUBjj19w4`;|2V*^H_vzmMxc_bG#t#XVpdHEmkaa$cU(ii>}E30)gzN$b{S<=g+& z1}hA?O=%*iVr>UKJ4LIK9{WVMN-v>GIFL1?;~TnMfFV|XNR<>t!O#_XB5k8(4KRjN z9$OK6?$BECMuKr93h|wYyh4@^mEd+`1<3~Yxky4OK?45ey194_O2-#xTxA4tOUYTb zi*`%8G)9Sy6Tdo4(jfKgW@;#Z_064gJX>r86d6y{vTAn2T+Ma6#RFjLx8IRpCB9INVGTm2|eOk^nCA(7e2jz`=%USxkS-+nC&=?UyJd57m!|6A%>q?&JD0q&2fpXGcW zY(vytwS0T}Y|`*!I$!4rnZN*nZPFv-w$nj4V%qB#kcx3X#CH@mfVN=?*+gmgVZ@ka zxoy)r!gV_iG`$})%&Kph0v{DXinEdGF9^8$saHMc+jxvLa#+#fc3`xzj1_j1Sx3*b zp7fO7h8h8gS_#BVHRC425yL^=`tevl0F4(jOcG{q)) zlTdtC4NPxm-w|*do_8G2@aqBdW(HjUs%-z@O#ASZ#K!`gRi!b(pLzWyYL)g@Kfz&q%E zhqxXRYmlWsPt(u1-f zVp`{E7l;1T`8AR9!gq4P8~!3nWTKYysjL_!=R>k`4AbbhjEV4={>Yz_CRhr*j;o-A z$*r*Je93fO>fdl?IbV^~0=vm+A%l{r(+EDl7MtDqd@Qt=2RMwfVcgpRG(m`Zc+ZZ^ z-B$a{B^|v6bIM9n5e89(t?;F<^(O_3L|Sa&WhtS{H?97^`^!_O<1~ctGrFBcXrPQ1 zf9(;VUCUgMlv3l0;z})|BS=&8c5UiosB{E`yuDKlkO)qj5iO-rqTOc0fu$Yk!C12)q-UItNJB@y z;fCb6U@4{tu<>0g=G$DJbudGDA{@x9&#m!|m(@Z*baAaK4=O+Si7UTXEkt;KJFm^q zwkuUcz08VNqsC^rX+Nfh+L_n#@JgB3MHS!rQ<^J7zsp%NG`}Mk@$+8$7<%6jf z#}okc+V$ayei}l^G{3PtD(F>qe?`PsLdD;mank!}*fCtD!Dzo8Sw*RB(!Nj5lD)~x zu4>-MX~ZxDwRCABFMn|wnN`V4k8mQJ!i@6#`_^&p%&LiV3bk>xFLII6YE|UMgepM2 z$B?aZ^WOO!^~lK1r204oD@i*!&Dil4d92jDmyE^-j#N8J7vn-{=Ocz1efN`Iu*Ycf zgeF3ovzsw-bNdULlcXv!m$P5YZjldCg;$fRph;Ez&_Tz2Sjal4^55l}h=tuGUk^%o zoV8HVTgTPMI}mU^*cUkcmNa?(-gdk@4DpVfylf7PoIF)E6h6k5EqZn9d81)F#Y3yF za=UM{{`OU|Plk$%lpOLdOa7xT96beS)qg5*&Wy_&aU&>%qPC}?0GSs(cs4n>HSLUf zBzc8~dnh-}NwUc}Lt>6oT26k#sQ`D;uo6US`c=&2_CC!7&tciE*z8 zzxLs#nuwncJ`$XB;}^zosM9WmITjyR4DG%Uyf(N*<-YNUl$em*K1hi{NSfcjeGHS_ zP-q0|DM-!~Or%oK#I>JU39584GhhmK5YpC?$l?f5Ct{KP5Is+_m|d8e&`(jaEvFDz zND=`)rJjCC4!2i} zjP^1HLhsk*aGu%yth1@Tm9r_kG<@CdB(=)_{t0W6>{6BWp1I;br064Nzek6(7niV) zN&mbIzEcLdTy{W>xLf?t@tCNNyQ{f~GBB%(h@bv40etD6s)V^|$;I)rq^roIdwI1g z{L+FmyJVQ+Z%jfOEL?=hd$g$%DNPf>k;ZPkbW-J1{0!n%4J@*xkRU#0RfRCXa{o|e z;JvFNQq4RB<6{wy^qSh`TBuF~$4ucv72UzC7+aun>7#pTU{CmM@_7yGZjo)bna`jW zZC6E-+^G8^A@Z_fNO3bdVlMr3_9-n!UrH{zkGY1tT{bz>5GAo~>*NA&+iu%58|u)R zBn}chE!ub?TC$nyA~+sc^O2-{=mN<}r%e1UuP{uu?WNR3anK{Kz6qZmjxwJggTzjz z`jxeM&Rlc2Wt%?H>2miOo_V@86XnWtIXJ6;gcahw4f@yhVt#v6CU;x3tS)rA3aX)O zpZIhK)fH$P>IBY3;`x<0zuuVihq4T_Ga=`use4)A4@ z?z|w1YS#O^pR5zMgB_1OJ4Kq`-u_#cAra}h{G^J96YJDrV3)fP2C+NA>MdgV@2n|H zb0UkDPFWBPPjWHfalE{}%TZNI*6Ae@eti8b>)JFsBY&9leZzj~a9?=v+>8axgq-_d z2xTk=WxWxV<=x7;qxU8ZBvKk-6!eY_G__F+rF@y^WkGz5JrD56$lZF+tS7$dZhI(8 zeIy-fO4BlvTpG9sVn8!c2y1fa-J?$>KBNPjWmYsi#knnobyJlj2)*%L^*OpQv!_%U z!T__W6a@N}?LP^2+NPJP@}Bq8PjX&WIFo)%-?n)Nmi2=#X}(ZU(3m&z2y!i^g8t&hf7^0jeBtK>@&__L zz#RO)$PUYYkR4{O|3TextuvMMOTumYB~J&^jh-{Y_?Rad6+t4pgq1Fy625B;KQl{= zO)-`pR!|5$0Pey46Y`tbd!lqmhrx#+^ZATkIYm@~Qy7N6Q}M0W^Jc=?9>YHay(0tt zUqwvs_gvKa`&BrLkiEtIoDtUiQS>OAe|@J87tcE(^VP>bU0=^~ zo^&1~)0!MTBX@ERX)h+7if&3#M%4$})P4|%TLW|dJcbnI^=NAWgLy%I8-(XZtKEKk zJ#S8kH9#Pb>df#j4(;m*_B5wOF;?;><7QswV8~YW z&CJQwLysGVp_*pNiU<#6s&~fTbR`Hr$i|!5-$n1|^tVC5d@&qs=Uf#@PdElW4F?y!?B1ZTdhPtbm@AH6`+kT^Vm9FtM(!Nh z6JhDJ+gff=sCOct75mJMTvN(y?+(GG-B&B`pBehzRILu3d*?gqE+4(tO-(T8o;5rg z#x7*V3fI=k)s^2|9L&`YS&ZO0&HOuMM0hQh?ySNcg^>&)4IQr5tw@G&s2+nqyzcl+ zW`MGn*tL?S$>Dhb(y}>`>K-)BHJHiyOaBZX!TU2NC+CUpJh}Xe$`ZJ~B=R+z0 zE;|~^hB$8ZL8;GByrF&$m9Q2JAQaJyzlaUmXUml}S;g4%HB*O`$EzEIHfwyWA>(-* zV;2eJ_)T8nb9vJbCF?f|hMB5WfGIP%)|j%|vNTH7@mL2ddakVX1SR?b0=LVmQnl)< z7d7~7sFUZAK8Hkr*IlRWJkB1F%o^Yp~M_4?(6+`9~0cQ9A-L%3f#SZ`y@kH5nRI z^a+D6y8Az)diuX}0DDq^E?`;3xp_!e85kjze-`^Q*gt5daBJWK&?&+Ll{XN*d-LE0 z;IAMDaxT4wMyz+S(~|pp=@5LVJEyStRY4VJdcb)72{aXF;IHCIn0L5x4D|0N7waO4>$E-o=QH%<2scnOc&qR_Vf8c{4UQV^X>$YzILES%mev;kV1E-r(pgn zdtqmYLj*AvE%FeN4J!>`wwZ>%;4lVf^CvKN$2mX( zDC;0H(D5ws@U>79Z8Q@`xZ5+E&aT0on2oXC0ioUTMgqeQ6CSSM+TTBz`a?g^`p*W4 z^*h=xzx4g#>b+3vK|}GvJZjg^@JUd1U>UFN{+&&k6K|}dC&m+JIa_|{gCxRsBDH%z$vU~SlL=y7iedssb*02G)aXr-d==J`#t_G)C6hrj*&CC z>7Rk~_38l_V7B2C+1pz@Q9;YnE2M-Zm*1@ns}9Bu4y_K{yEC$pko=CE?H_t%SKB>T z21uZme(PIlJhGoiGjRj!y*4LOcwt^(K#LLbJYWD7lLjGeS-7g-{XwoH}>Z||f zEN0$O!z-P-vTB)DQHiudNqt#WWK3gqskCU$Y`N{5MKwRjlDZ%cUzu?!6q3g3HSgrZ ze+eIp#}Cw;6)M>cvf^&4Lx6SR=l$9{;_a!@o)3PLi%x9>^Tq$85s;uHTs^=m;IPE$ z30k&K$vO*P{>M=x1Gt#0BqW<<5kChZ z7m`6Kpx^{*{Z?`Cm(#RaFcEKO7s-UUMLyGdB^Ta11POOoTaf(B8a(rl*OG2>)I9Dy zEAEU=0fTiGTGh->nvXn4aQKyEgJ^vWo{X|laNH%G1D!H^dS>r*P@MMA;Ko2r{|m&i zQmofJzQicYFALKZGADF89qHVjNNV^rM$W6V)EHmysB7wmv({hOBxqBMRFNV%*mRp* zz{Ypn7kobarRGbZk&HMIL8NCpaIu|4!PJKD6+G>yCVUEPIBCE|g7vZxqS#HixL%He zD~Oi206*=W+wrgoH-FBq_pbEyjM&sA({d;sSR zX`p5hG%7#h$#xQ7zxZ)}NcgglD=H$sxDOw4#&FuIrn zhqr3MaPN)YxsE>Z=%P~K!1Ci13dB3jz?Uz}FdS{2_g$Vz z8a8GWUHJ9-yFQUJb1^)Xh7sMoCOW}s|0(7ib)K{lXlbM?3ZYO{avupDzN8~0cOYHN>RZ!O73cZOTkI(bL)mG7gxF_D+y^-awTLQ7sB1g!yt75RB;8alA z>;TZn9Hp7)Kk+*)6gJMVYT#SG@KVy8R6H~%c)m}lJv#KX{l-PWpd2jw^y7u=6~f+| zUHZM%#_iIvP0rLf>s;UUQR=K|zEJqH53?1SU=B@0gzCQYHM&f-nuIlz4*=x0okRG&dTWYXaXm1(mYijFaNi3PCfrTLf+bC$a-{lfW@ONUAZ{bfE>}+m%C8^AFa8KTE6oa_{md=lIq7b8r;A@l-8(1nL;+{^(k?U;`ud!d#zcwe zM)%zjTm)Yp9q7gY9ts;^y6FW;p-IibiA9ef2GmH5m7ZFgwW?8_ET6|3QNZhXrBq4|@r4+@U| zcGd+aNtG$7!#4o?9BpjF5yC+HM+W{i;d%Q#GubDHbl zxsETDXb>^*t_$dZ^_w<=DpmmJ!9-rh0EJ8*;Pb*fV}%v?e)HSsY>cTG(l5sWMGk{= zWo?44pp*=j3VrF)Z1;2Ao{QKBVOXymDW3nH`XyML%l74R$%+6L6vxpQD-U6ksR2Kzt zKtx}?AZ<4p4^sGVr^JO-AJIIQbpO4P)Ew5dTDVYFufEB=@!I;mX0dfc!R0pwq^H>Z z=BC}cUp^_8B%(J(>Rsn3k9v3|DdRtH(B~vlc65CsdgSinXco?+KEb4xNTRx8e2Z||#(P!!r-G9V!P)9%p0O+5O&S3n!)t|dQ zOb^KT>)X}d{yq0owX=xO)Wp4eT?Iyr?_!$!sB$F7d$%8R5FgH~b9m9Zo+e3>7vDK` z_=CIqyGnIC91(xwDK;9TkJ4{|hhe%E(I@fupwDDo89^bUSw(C#2hZK%EFI_y=1V;` zhE?wZ|9f$-jW8Y5V{{{&dAZ|bnK3ps&7HAkh*eX+y>NJb4~~duJK7Xh)Fg9;u?HdJ z?7oK5x*>K2Hk^^HQ1faiBZPn~h`QOfOrU=IufgE~I0;(uJ9mX*nX3eZAk z?>H$cg=XMlhGF4sVWEQRuR*^~-X_)#5Pf79slka2qnhHB02AK5{kunyUu3LgRgj_R8JXUNBPnTSKZCm55XOH@gR{m6X>h7h3 zNrvDTWnD^aYe0p^N1JR9%QgoGtDW7;-p!I7>Xsac*3i5YeS7@vldLP6{Vy8A`hRH% z8zVEv|3Bwat25=aC7$+gDQ9Imh{wXULC};OC2~2g$W#l3YJ#+D8==w!q(-`_pblhG zFxAqUX?JnKpCy046a$cvu`SAQ!jBS9E!51ju@BV+Mf%|QxZ^zAllgtl7u26>%$I#5}k|?9*<>BhwZ!kVIEjCHzYN>*0Wr9$z4tX## z5?#(9Pp(0*N5bAfa|HjpV zgeDdyb_Ctcg1A_X9I>1`%H$lb$Ma8bAUIrJAF6+zY|Ed6!Fmc697|5SY}M2E%Si0IWJC(lSK39g}J+t?wK+_%$Yc6EO9FTLZwG$T$yFq(<4^* z&EL8ME_Tx5XcTPS(fj36OR`6(ujeHx+P3G7!IfN0r*kGAKI*U8QFNf>m@C(VD)tCF z+Kz8;-1h9Dr}xA85;<~fsFn`jdC@|JqhRA!T+j@Uf zSl4=V%uJn6T{|&u9;8cgh~Y2=N?|$11rOjjqN2eO;a&BFq>bKV)!upFS4duN+RWO0mT{t7y=i{febR zU1R;O9EO>efMt%%uuu1lwRgxz_erY%+zu)GU0QL`w@6xc-C#?YeSgNG7F!&M&P?+u z@eS8nrTv>jOzi{wc2ONKKt}qm7eL$ ztX&W?ch*;BE6tZ+Mea;^kob$-;MGnqpm+wSK{~7KNx1U2@(ZKM{$ppMSgW_co5ORs z`*ONlSqtNps?ENePFSJ2vO_`5b}E#3KLz73Y(}1Z^Q&9Zz+EG%tNv`1x%z^MW(`_- zv~gPcQ-HYfPk}%VTK$E>%S<`E9Hc-1dUutxOS0yO`WCvM8=C$}5FDU?&-nE)v1Hp|+YiahOoGCbd- z&m4r`iKWn17tYIzvLJBJ3Bi-kx$%hngXU7M=6Pt|Y zB@`^o&tYX>KMY>jDlbAv;!D-o%-o(N_F(qYIS>~uhr*Lqtx za#diFAcCl!9Q)Z6q=#B$*%rlF(?2(*6q^VZJgiADxKfS;`xkk%ry6hAx^!lS3A}n+ zRs0O=VgruZ(f}O3B1kb@sY<~M=`RErAyHh0Z6acC+D1M_8co+Zm-^2yJRoFb>;q6R zp4*zFV1b7v{m6aQ7F(c0j%r?Zp^6>XHbr8eCp8czKvQngOFN~ zVI=aks94Jtgl^+?DhZT{N?3*S@@4}mgfLfvXu1eKeuopMEbvUO$*5-sJV-EVfI>r; zB#&onU>U1HZ$ePQ_R;iZpjCWB!c@~a%wxM1x70`ia%=>+~qq)-*6NveCGKFCc9kBi7$gjsF1SeiX8prMQ28m`@H+EKoBL7DHQ%L?9$!+|;{2 zEW@&f*k>ZWO)d7<(jFkA84b~d3_9e}jNe2WxclSsiCd8=Mr*;OT>1X7S=&D9pBI$C zQ@{cD?+HZ6ArNbVElPTlLhnHjN#Y+Z0In@%|1(JTFQ>L-%1?-7J~48)w2-Z-s%}2v zu7)22G=P@=91T9?9oB0iTJ(qqm9UIY+@>tKO?jsY^BMat?RacXagT>M>bNAPuIjLT z;{;=t;@Q9fuj?#4ei0Y6PD;xTzqrdN;er%9cN5+>%_sKS>~Ne*{+YiRG!q2VHI+f&NWu4kTSm^_5p%&U#!Ne|@@u<6ofei2gxpq{jnr~f zaGqE+RtK&E^Lu=8rS{|oU;DiL?)(@oMRF2wAx^?hb>EstoulS|PA|U_S`?gQCF7Lm_`n|BXw7I;_I>=R+uW!`~c1Td_GQ6Q6K@-i5!Q>RGrm5rq+ zAGacCh*%WZF;Gm3E1Df0RwWwdstIx*qM@k45D=XB@-mqSqZd#J;n2cHv-#(tS1c92 z2)t!lC4|zP6TNvTpoB8qKoYHs&QX)5t|J%=8cMgMkY96F!}SMpeB2WpYF{LU>P#T< zye^s*pM`SfQmeo{q5J0FvV9`TXbAZ3;M^{lVj)Aucs)sKERby25dgEn3Jl@T{_9E+ zWc$%n&TqXyVN{OgtJjZsw%3Nl>5Oh@EbLE52USIIolrhvKSCzI1lvi9tMhRMq z!X+KJUJ@fMOXQ*qRUU?<^%G3Mk2Uam8N=9;1hnPm(kP+#7H!#F+XyatM4}OTE(%A9 zeX^h_scmWdRtPIfU%o3J9U~yWrKdd?ZpFR|=@w(%m0_2^+G{mG@Dn&TDN(u4lp`(2 zmSGSU>LnTfZL}4SJnMo;f?QNADw%mG>#GO^QKB_*Z?$)4?^jfkQ4i4zN6LUrF#L#> zK0`Hma}1M+5iheKb9T4=OB0Ch^rjoMgVUx-OO2%}@$6gqp#NDf^U#BIhmO9{z`u-% zf68z|(!)uo8MOZG?>N84T#cz`kS>ZFuLK^UVaxT<$<1zaW3Z{Co_C$1_5)qp^U7Yb zNHfRF{(_a~#riWU&kM7b&+))2uG>!NO5DAPg7r%uz%f55XKcB2W-)lLulE}c-VAco z*V_o^IN^JyQ4DGh4w6%fpO66IsRlm~(^;(Qd5we~P&%@+VP5qTLe{M1!L0Rcj0LO7 z#O(DDnvqt0b*w!|T7W?Ld7iFmjhvJVa{}_WNl>i5hGLM6F|z0!rTZlmhjC69h2D0w zR!#@|#575YJ726poy{7?-wQ?S*I7^?TKB>IzZLZ=yf18bylc*F4I^1oE zZ0r2FU7X+Bg3L3@gAQO5tN_z#4AP|-kzm0j{h$<+gc1!TLBybvB=)4L$11{MB;zwM&#!=%F+;>!O3^VUR3F_DZ#aYZ1Gj-%A{d=( z!@#t*=c3$JvGb`WQgT?4-=v>&mu6~hKng-dYsO&@8jUv#ScZ-|2JUt&4=v`P1|OS) z@j>;2DYLx&@AOSnq=mJK27@6oAih&vY%346rzpG6L1whL{iw>-VQksciSbiAflV)i zKr7)5tnAVNzW2sS{w6E00U@aF zeqGaWgttpGr${+?MO7mx_c|1x%;1nQYM#g%V%13PuR!LU@KLRR8l@6tLJnJ*fx?FZ zA@%YuznGBE(s0R|HPZLXNL^xownQ1*ODRg+f6V|sxSX%`G5WS}i>Q6Dq=Tph%+Utk6UH(W!(jg&XuTk1*W3jVPEdcQtq5=)2DC!-={ zTNHX0diLz+!^6X_uC59ycK!_ry;R^O+lMaRtmCmCu54#Uy!y7)8K$h`y1>ZVQ;r{D zU}j(;m!OnHX8l5*=$FJyh{dlS|kYXML@ z-+x-yYfgTw8az@v+#F0Ya^0Fr)6p#d1FZWhc;kw+wC)b?oK z$oBKaZ(U$|-KV8R{)bQ`8|wAD@_j>bV+T9!k>}}RwRieD>TL6 z2k;VQM^0%=v-nWW29hQTMBU?~&e?EYCZrF-aQHmSf|T_WHgx7i&a%N!`|o<_dL#Ap zz5wwf)cMdKR$mp3RBz%CO>bXoFXXQdD$gM-0*f(r?i3_o5?IH%SH}CTgnI4jd$B{_*3H zpo_)Ga24u{vqls&UCA3IG)>&KxmlG&;kP8 zbxoHiz(6%#;lFVi-3NfnC!(j8q@YKAW9LRQ!%!itZxqnUE5h&|uH`!h3g>RCLSGHP z7JZujjfGX=pCGRGo%(SFmZ|dI^z7?T{2$WwNYW{RR1amtfyi58towmEcCmI&;K!H3 zZANh19p4I#ZII3bT=>f2HeIbUg&R$y2UAk@riPQ4xaPaZYm`_{>OVv@7cUTlQe_BY zq8gC-t`>aJO@<`c5__)uJcX?%U0h7e?mf`wDra6}x*C{cKAV@CLYG!GN#By$2Ds1V zaQ?Kg{F$LOI28m(`=@_Pku65p@LH62Abj>dZu6Em^S_+VvwLqT{5*ARHAG@JsW$5= zyv%7K3I+tuf1;{Rf<rELqlXZ6N;AL7lCr_bkC+yn9r_=t4SB8afmg$@~;&>)g?WoL*L)AP_&!`uX-UrQf` z;~*1|T~o^Ec}>^8%wMtmTuJV}+dNB3w=OO&ou`GqeVVvPXZ2rpY2kbtWZUTF7C-E3 ztBOCQGHV9a>5eSVWP7bNKSf)+qYrTmbjs2*XQI^iLCX`Cn%csP4Va;;vX#OCUNrc7hcYIa7e9`zE@i|BgZH z(@z=KILv-drW2tT{7GuX-KO)=f?RWhpirCRL2bM#Z)ic$R#%dhi23J^d$e<3UBMka*3cZZX#jN zqL_`<7JfwTi<3d<0{Db*J9O3xh!|lkuBycy_8^D0-6SmEbk$2(lA(Qe(PCwh7e`&g zwxH;=N^zkS-W1&Jam1=R-Q`6(aHAb&D|oI)ng6W>|76MD&bFjaOnJ)zo*S)&d=+k^L}HeVVg1xy}0v$COovzkT#i1K@P?7^3(w_;&W+g=c77(zkr&q7!J_ z;8_AYft6kalwt=?K1YWOqk8-%AHZ7*tCVq>mElw|VElOOD*~=Pwl?S4U=sPK?a)Tc zB}WP#USIt4Ws@$QxD>WD{GOzeb&;or7n0>0qhe}*~+ zI@v7c9p1#zv(HLIC(x80GSdEjsZ|DXh}TV)<>qL=)u_BI1N=0%N&0U_G?Trgo<}9< zczw&*-39(mCEoxD>GnKz^^D=q-}(!)#F=;N>EkpF_((kIzId^rmm`nUy0-&8^4Bzd z_ru^_Q0y<0$4JFc<;sR7pWE3WD8ef+27s}QMdm)v-uh1%lO71sht>PRd-~6P{^$sJ zgJi37$4$P>uT;N1|JVfSAP4*uiR`hlv9|ue%}Y!5+Bn!CJ{l&`6EXe~mfy>&nQQd* z@ix4zqZ>Ytn@wUo%*n%(l*tn}vJGWKgaG>=e+asotMJ$Fy&vFF9x#jlqJ3=tm-ewT zvHdT**gA5~xMPVuho%G`4`C+qaX8USXYX~3@CEJ6SS4VnYZ0-8=y3xD(9OIA4s(Lq z7rXBj!HB=Sp9dHKRo-6wK*C@?+M&e5=QmXsRW;M;3^&HD4S&+PTRbQ(3LnJalp>?|-B0UV!yZG^)sIIHa>4+i zE7aoyl9*;x#m5v3n-rRYx8FSXyF-b8w7YA6hmrUxAw`?@XtNSH+DSupMHFY3^Pb+nrU*~A06+WYU)FYl&E$h_bb<+YC!`=lp{rYaeDFO(I-{5 zgvA)}(mbh3R*d*@f7!_}Mo3!|EDVR=pN-h`$}TZ_IJCc=pO;#w9>OGcvTKsc8cMC( zZ+$R`M;#8+!r@js2u`2Bpo%$T&=@M=pR(OZ4d~9xlRPASS*x?E8Zops)oe#hEG(G~`k>^`{ofyuMn|w*Bvy$o zFAUZ-)z?&eDR_C*KVGB;$b==^V`tg+h2p%sxL7$i5!J zN1oDRn9CRzXCJlnY0TK$t8G}WapZ4%JnkPvw?^+ucHv()6tplv>Brd$5m=(!kIZ~n zuXL)5B(buZkRCM0XDsf^pPiWWV#6aqhx^d^;ZR=yuy=RfIbhg4VNA+nI59@#mLe9fA=oPZz@JS{4%wx=qhkS3CVo*6@b~G2bXQW z()+5(sz1Qth|VW~#dk@&zRr(CmG#R?FPTs_D{0R-v}p|}?XLGnuIn%FokSVQq|Gh7 z>MKz_dzOrt@*aTWnwzb*WjZ92vG5g?^Or+M;}?X;uu@UF8chAxGE1t8Q+{(P>wBa zY1#=aynMJNg(2#%cyd^dtR0cQ!w@Eg&#^@BCWt>ym09W2j%*x=x8@)4bZkBysR?yeN?Os72#r z@R0<7YOq&7c%?K5hPgQ=&b!aG1W~(ezSQ9>q&-3T$^K(0PR* z;0s)&n5eGVCMNH{G!w)6&eOuqFco+`2RTA8f-#XMF`>Dqtl2{uPP&+gRrFb?Hy7gp z{y#4as-EE}iN#8ax+XeA5$BLQ^MQ?LeWwf{Bdvo=Rf4h4&UjP^u6-6&u3J-9-)Tpu zFaA^tPu`0Yj;uXZoO6`|&4uU*H1+lOWCILVlq+D7k44~Iqc@D%IZ_XS1bBm-c?r=9McmBNR+PwMk6yP z;o|#Yw2b1WAI>v>3LD?yRyMY;@DBT9q4mk#;o z(l)_KV7eIh-6t}jMiiOI3LPSFp{9H(YFf@f_f!j#xwn>&VF3tF;$yXPui%E5n5c0x z5_ck}$I)YQV`jhXu1ullwzd6=8sY)8Rt6q7u4iB;cc{`A1?ExJy(g__7n>O?#Ch}l zGNDE7p_J_go+e_x!auVq{0vkD#Alv}Iy6;mT8cfvNAVTC9+4;DJaly9!zJb*P~s6G zqLj;Z-bYj1XTupC0^g0(<aJ++FQH~ z5Mk?D+zqE2mTX3obzBEO`@yf&&OrgRjJB8#=i#d0@}xA|ct(zdc*DvvnjAdJNjFV= zvR)qjB8=5KxWKL-_-<6jl)kZ;vI&t@u#O?FBe4Nev_(|uCX~*~^$k`IpPn^)!;cpb z(|o8mht)UF#-1Y}qf4tB!W-tNAR^f_#v>{v@+8&`33yQxjk$U>Tah+)rKGnsIKGzw zLmEm#%qcas)nt^vN{+=l1hZvjHm7%UTDmYc;PP;oO?*p1BkqZV#H2TY_!$~$gy>D0 zwp-i|UGEZVG>B&19%6T?A;Oa&S9n|YCXZ3JTJ}7OEZM}NKtuOpQreDqm6%``pz(Xl z27h{k1UiXoqHl>`_@6p(_H0nWR*^1S4%vs~I!aOo{^Ti-+BeTgr0rE5Hx*oQawBLq z_FM&c<*8oeUIj#ur29*_6yNYCQllcItWOp&ez{3c8OH?_$ z{^w=$Q(lmHhJzL9a+&T*I8Irq`u76!B===i(p{}dVkB9b6(a_P zUFq9<6i-DePT8|6YdO{+$?X;O`P>>(pQf%d_7I%3Fw5h;m$jy5UK=dY`8$49Sc+o? zWPvDBIMT}Fiw0nv_ro>2tvC7fF3injQA}C6H^wm;$~#dit?5+?UNMdQbgJG`s&O;@ zYHy$V6QI6`%n#AZSO&=!=?Zx^hKfP;524eI&A;VZQG9b;;3r`b{JIdN`*}1USEZNc zlOtP$xYqM@;?2yOlk(*Op1UaqR?P*H?@ugq;eYqhn>8(GF58YKNGO{!l>S91xJF8~6n z;+Jz#MA}}0X>rFHJKDY-8HwIk0&fb*%Lix*h?=7;kU7J36LZPoC${H$rI_@HZA7(O(c4`dpuB)S@Wk(wKsTN@OruNA z_{qBK3FHXVSIlg#T_VVy*aaR?(LbL(djc^upuh$2wE57F3*g02spcUE?+fjrhb?!I z-t#ol{x3=Fy^pTMWpFn&rteyhK4dswlfp;qdGA5L4}!cVSWCSzB%_UhA`ze+G$%*t zz34LXFD3M3rb~T9T&O_2r^!D**#NU6M5u|Yw=OIseAddugp?%%IAGDmx%Wg1n)(DBK*{Fa1$?Id=T!FnZ7rj)WbN#fI< zkSU>*HyGyog14IpB=t9gF@^s8v-#$g%vvQ42ejd*4AN!-FSG)W;?I1 z6pT{l@qSCOx$4(g>Cq6Pn6#2{=hcbYhW4B@*=px|A;PjBEI9T18Hn-wjF=3vYPLES zP&*4#_vM*3>)d~2dTv053%m4bMsBo^z4|IQnSE|>GcTDvpW;TCU`bKM9tXz;(!P$N zLi;K!lv=ME$<|_hc*1jaJ)$L$6e1m?9ssNd!*F2W*tYRJ8Uas^h{>x_D86W$+K)s1}rY$D_hY1Qs$8)cqSd;pCC9#A8kVwwi z9b1+k2nT($;}g9F4!38;Y_7|PP!}41 z33QKEu7kg6OwKW@fkoLL52KcCav9Ec%d6Gbm&@kLcgjzE1Ii|OU@sV7h28z{!o}s6 zI#4Cj7x4G7q;lav{?rmB2)+cv?nw|a{&aF+`=6n$l3Cy-#i)`cfMC0K8h`-ij$MBf zt&}!@14OAbFBcDbAmgAI+zl_XR5#sX@-Xm2Yqrfm65q8AT91UAyHBGRPK*Zm(921;H5N*IzQyVw5s!5cocIIV@o zhXZ396Pd!2bNO@UxV6;YGx)xJivaLuy&I&<9w3%&J%c^-aMzCpShO?IxUer&eSx$9 zR=Dg7;T42v9X^{(zd4zhbcXJYoQd;NW$jnR?kN|W7nK+!-2B<<@P3<&Y-z@c)X@Z$=;$HsCy_av2e5$w)$I**B@^^YI zx%^Cg48JtvZ5rF~U&uLL2%gW84_PlFl9ex6OMZ4u{OWmZjlY*KGw&BH_o^WNIMaDl zsPAzE(Hjd(g4M#1sn(WGI9k3Y3o6_5474lb=lpACz=5q^pco zl&B*wAOHLQU!E)3dVYV3IDrFNYJrQ{uZeH5<0UhBCwpaKBiHM+;NSL2QJB{|lM}(g z$P|vB=i(jt^SI-8$)(J2nuq;rW-qI+w%Up7r~|-~=hM+WQEgOH*Vf!V+Ox-2S_Akj zFRPoHP5EG>PUm@N_s2I&I9~^1K!Fuc>@_D<-@Kmw=cDgrP4E>++iSMNB>No3e+Y5bt*F)|D&ZiZP1} zRFTixoc5J;N?Y3YnwVmZIX<1%mCg?>wykyk^s6)5R?V_VoZ$OU`?GTXoL=5y(xgAD zva@@5xxb&}FGd%JJ$O$$D3=djRHA9nH<~+Nc`QuZ-ZRX`GEQMOT*@6mtWGajtJIv~ zdqF+)n?|ZDZ$Vz1l)h1`5SJ2TNN87B78Giy+^ImvsC zPJuTKzBwh$z~><8nK!=AEtyt3mBLIvBw6IUt|v7)4f`Sy(#9yNez_f`sK=Sz?+k5>)mWy z(BC4^o;pZ64DTDx?y$6>Qm{)zMKQ$Fo*v$t&G8S3x%keEvB~}fLb@tJzNjCQsgE>m zkCK-+^!bDQHEdcQY)r0nu*eGI^T?uRSe;6 z?fwl7LvVc9?K0f*1_*8^@y@d>C7;;|47fxLINa}5%8GPko4Q8^S}v-~3n0j$_N4*v zWW51dGaz0yc^jnUM%d5WXNHObDvAd}#+in+3km?lfFG?CfSrfV0_F92Ai^J#W`K3d z-SUtGBjjMgBhVdG0C3fi9ai{zLL5g$4PmnTgS&uC3npV9k++OE9rJ~xEj$O3y3E=B zDokp_v{vWL3xedJ9&l7<0|?au_4bNp2&lUdFh2st5Aw!M9c?Dp)2(*Jlm#b3O9;<% zVFD_l)S7OMFtbVUVm^#!javI=!kqedS*bBQjYb;#h`?oTQLVm9mcB8Y5He4#*(;)V ziOwwrMLg#Y>aO_u0}kv9iG*AQQs^FJaiC*NM;X*_F_=;v3Y>(2)dC==r;@EXKZ=W| zxV~Eo-m)3S7R}CcNiDmgo#QEQcULu9v_o@w##{&y2X29-2KGg+se#3QLl*#gwzlOI z9#|v%Aqt>D4(S+Jg74EIE041%v4H%A!nGkL`UNRp!N-2QqnV`;rL!SfQpuRX4CV`th54&KaBp|q=Kg8ia1V&NyfW3u-qAgT}HpnO1 zG{t5dFg7W>n_Q54jS>E@?9l;QH@4r)A3B~6)(oO1VmK3ov_`sx9a_o(sL~jfw^M5fh$#s>d zO@!Hm{$t*6^78&SG$V{QQDMAOz-9hdMn6N2<p+sAkR) zyC_qr2%;dpaVQEp!ss9hA8zV`xQ?x%7;+2@7*8=2Em+gMTs^d{rhMjq!NjIe?E&r< zruJa>1f_Voul}MWc<{d2n-^;sKQ4&=ym~Ww`sjM#qF_$KAo4bWbl!Zf79WFyYG@hc zmgkng31l%~jC2n&9dy$Fp2sezKtC}FYI&)*jhJ>=SJ5_nU9T${w1*=41@suIxng@; z66R4>CXKrT)RP8W1iaSl&{Bo4*FxAqsxRGU@ETy81HoU3wXh{>dP-+$Q0@RIx9d6n`5cf!Oped zn8583BM+9iaXgT_ZM3P&$Hmvz*EiE%E6Yb_+vwHDPt7!KrjYt^<*fk=e*)StZs4!0 zHk3Vp2PEJ(>{NZ_SM_kpb?B#YBdUg|g0oaccmDV`z{`!>;S(GM+Dc5WmVEMDBl3KP zmzW8eDQ!fq7Km7fRbqP7D&4qQk%=iYdj#H}=P9KLB|#ycsrC?ks+-sc^-LPM$L~=f zA(gFd)hQ4D`;mO&#IQK{cR(5WmwCgy2P66&X}Uhl9qYbT$AN3}2HVm~uoA-YqUB zg}>`z2!>;1hCx?HH}mbys0%zKe3t{x&dG@rBZ;sw%JkKcqW43HL`A^9JUM{fW4Q&3 zh5b>l?Ui9ga)X2f?{zmBGK7wN(vlEoKSu^RU(@^ol6dg(2DPq>+L*BVS*r(SQYB?w zS-PQhJ^`yp_fSA9Bk@ErHDoD|_bLIu{9sc5z*-gAS@u(<){sTHxS56su%<^8eqfRr zW{7)nrIr0F6-vpQMFxPsxC-!x zVrWV;&)4v;h6zdSidVb~c$;3_o!{ijJ^M+#Q{MCqxKnQDu3oK_FYrof%MM7@MbRc( z>{C3P-!*+#R}@=X93j9iP_}>{fp&qY6oU@{13y5s$dHQZ@4$5z-*Air*@)3}415Oe zRMvV4QpN-Ub%+pdY3($BsOmxaW++GkZ+HR=z)_%7!Es#pBMOT75%({4r89u!k`W+` zAt}SmmV8BA`haI70J#Cu?~f?GmNP>?)`J#5{xXS1w_l|2pibtO%a=GB+-LQ1{3<=P zH&>1l0%!XqVi~a)WMMRi)WP{fOYL8sZE&xl0cgK@Z2VT*(&KLQC<};6KqhifnF=<3(&W)Y6jN445 z(u86nqVaDsdrZh8jhka3Jt{prK4!~alH^A}bOaDFnY}eF70X23H0-?~1SwUd7t#VB z9~@9PyMX(M@GD;)nS?qe`<*%)m$DCXMHpf@1cJWBL8%b+nTeVaeL4MN;||K=u4`8| zaj(s_qX<@RX1&iVt6GyMLKC1eoB@mnD8TI8sGut7{I^NMowByf!x!~N073M`~ zMXo^DqWD}Lbxm}qMnP40=+GTnOHePamPJIotoFfbrv~bl5Y6pWqOE>$&C`w2q- zkUl&FKUX?t(q)B`F6lxDF*YGo5ushsv%s!%Mt5c&l+{~D_9Kqxj%)icWm?O*J+BJ9 zyzaH0n;`BmL1O7L$A_PW$PlLw>~L9tUOR+lb2hH~%h^<#jt&$0VCs=0lA7(33*O_> z&ojxwIN?+#CrSy8$eHVv!y#>axM$mf&vKXDV+i$NYfG9rU}~OA9cWHTZ;hX^bXG3( zR`-R0K^RKw=E9o2$M|Elw=VLvW$|Y8>dsG(9@s8{Dsl9!SGE0)frX!IH0VH=^(*4{ z15UwHncbx_m2>e$(Qy5!Z|WZv5fYIal`vrJ0pdi7*hgQyKv2F(&8P(O0JYnFB&5S8 zhM{8HNm3l=FG%+H;@&La2_Rp%7rd_%Z`7OsUo1j0j|+fDe3YCBsp&*Xmm7fQY#Lmm z`CLHp^1&Yd;3zx+s(F;sT!{uC3Fsgm-Hib_BXr%O1xA;JLdc8kP>cQJcLJwqep!Hiz5Oroi;Z;K!ap7435GYHy|5jXD!nGTs z`OzgU!^~AA@wfuOv`uk)N2eZVqZfjzog1Od0uRgupVTm0ZogPDApr^2t`2WL-mZ%p zOda-ku0}hv>pj#oftTZaO@H+>2agA=B=CH?PlE)-0t)u-`{6_;20EPfS2{X$b$16- zfD3gRG;1<~XEqB#oioehg_^-Ln?bpQY;ho1^TY^vg0Arq6l3-zh}p9{W5vadBS8#m zmv%D89_a`EEX+0_h$vVa0I7|Qg9*~vHK&POzQkrMM)Bx1527K^OhlanYlw1{oRmpR zhQOBLaA`1fu$_WsytJKd1)lY7+ja&a=2}xL z1p%#z+rxWs-CW_pw{l~bMI}89bPWM;(MyZIPED~cTF{<$Su<>+<$q44oPi9`? zj(Fu#`Iv*)eH3C8`k3GXj;4v}JIP=k^fQUL#ULr|&?mQj(L`UJ43J-GsCwkiwp`rD z8XsxCPp@p`c(#GAyJ*}^neiM005I1$@83<8g)Ul)v|RsedkA?4r&#N~V3K>>n7xBY zmPMfh3x^kO8Tw{Nm3M_H38*c-JK4BFADQE%*DZEdh#-(N03xca25Ibh> zN(wP+0OJ{yPNULmur>C83h9;yl5$tC=8E2^5t=ioJPro=ZsdpK1ce7)KfHp;Dm3+) zW@p{lpJ%Wur2h1?b!`EuUF$Mj*s#j9lEdKjQS%v9`59CVD0Ljf|5>SoH2}u2#*0N| zOrAn~dwj*Lb}d9|Ft3Tf|D&G7IznwZPY3Z_M?tUdLKt3G3>@g8tdCZ3ZEF*ZxzwM) z?x9<5eM71@-$8?*KNzu`L8UbRr*c(2f9ZHZIVb?f0g3(vWU3fvA@c&CvigV;Sl)8Hn?Y?n7t59z6`e6lt~8zZ9cbw?5b?h-v*ng44(kl zwr?6M259-iQF49^EYV;1n4SZv`3W~|T6R-Orv|HRIdBl~? zN&g6N@|n)vVd|+*)Y{iRl6IzjDgOL#u5a_Vl7!bUbyPT+?Z1LfEumKykfSYK1E$kj2W(@?Th@|K-q}x5birA7^{6LdPy(Swp-Et6Yjn<7hNf{%PM> z^t_FLZG=c@SOki@NT>ST>Alze84#P_D}d%n#_!|EaST}%;8jkJrl#wo`0-CNc_#n$ z&~B}K6YzEMzkCSJuNA+{Tlqdt5BJYA)x-&*L#g8E`0(x^JdcITf|SJTATrxt{u=ob z=;ii!-CuY(-Xp^}B{OQ7?V`ht>gwP|*#8M;OU4s5tDRzfeG^>%=aktGMoN$F)9dbK z&Yj9>p>jh*=g6I@QPabtbxk{C(y-$E-yYh;C;qTLnsjX|N3|Wg^hW~vPT8*8x4H+L zC!iDJSiA3zLLW`K-uTzVTY;0Qg)TWdxIfQ4!@vT1!u}MnhU2JEBOCSL!bkrpn~5I0 zeOB-z_z8_*Rl~%f%n6RmsbQeF+}Ry*L5@EHKS=JcYhw$8c^>+qr>d>N31Rt4&pK@y zXU&9#Mb*}N0=HtDdSyD|+17nqKq_QknY%Y+lR#o1w*k!C>H@Mq_F z+Z%+g387r|wQP9&;l+lLWzdWe!kF>rdmDW_G-SR?(V=UVLCv3^C|?0Dy{LHI{8Rej z_LaJN_U$FKP;{QnfebIyGu0?pNd4Nj=eCJmYuIZ`O47fmHFl7{VAvk?W9{v- zJyb^-1`fPN(cBPOFvil^uz$Q#GXxsL=e3#2!apX#&XR3>l&J@IANjfc^iWmCp@FeTu zO%has)~Z8d3PoTU*wPO;Ga=APfayTT#*zUOCnFPcSM&Q4{X!5pIHn!Q%Xj&~mjUzR z%%baW_Q26=_zq~2x;>wT4pR1mRi}M#TMf{YX4sGMo%h>y{y=rkYs@zeURMql9{&pD zPX||IeWbq;YOJuU{#q9DVEF?!`)g@5d_2>}QuiaZdT7^omTi}`dfs9m@>=Y)cgNdv9xZxJB>O!Md^GUTua z^-35l1ikv#fxpzOXCYC}bouSJX5ZH_C}=>Ub@V{kOM!NxLO50b2>nsDwR-ZQRV2~t z2A)f-Ku~xX?)Y2U5{h&w`l0%p#X~wPzY8iKhx7V<*VweTT3vQX7Lxu*pWOCL7R$J> zailam6|b6ugXSzFtb?F_a*Z|}HorFTv0=6J^I8I-RNxl@xz%fp<)VJXGdMp&R1nKa z>}I>8H`h_do(c5pG0f^~7OYGs`MrXV#?&5CF$S$_VnQ*Oemk`lI%29pUFG*!=L_tW z`CTyC;%s-=4;Dy`GI+`l6$wYPZ381l0A)(qqHP@i5gy#x z!aOi~OqUGs6gtNYikKEOqpHp?{1?R8=R;Y|pGFC^g2 zA_xBg@J<3q2scSwYl6!c9A8^pU-}Wyq1fHV61ItfVKW(AlkxpnI-vAi2g(>F5b}yR zLbY`R2APP?P#Azm`e!n3yU$7+>aUx?Ti=X}i@xvZl7>TJa5p} z;J@^CovdZC^I$7Odu@GeE8WxqXMugO9oty3k4<*TS-7dTYA*Oo_V-~IG&-4R)K92q zh-|LI>n5S>ECK1Fy#3Siew6B48mNJ|8Akf8!Ob(NI2LAB8kB#$3@g;o>*uYVSPs>^yULdyNwx{Qk9em!=kFC0!W+LEyf?8ItXl=i|c z7=$w!jjDdSfRg^0NhYyxC%2D@7^2%vt`8tVO~I$M*G?2y9E7>O^S{4`A_(T0OJYzY z{pUi{ps zM;i(K1~LUhkl|BX*kkP%=@J5|SQKIK_!}c_<7S~sx`3$SW%t^i&mQ2=sV3 zVPl`9A{dj_x&?t%MZ#V}AVDj>O2^in=B)sVF@;*G`+9!2O18?Sh^9ftQBp$~NhkJM znZMFQjr9_6%8}Tt*&mV>CgVM+$)F&S=f_sj1mR&rLX29 zA>t3xygu-(m3R5Y_>OW92*Rjn|7`8VeIx{UVKh;D@v}lO#Bb@t9x_2c4nbyq9hoQu z^a|p}0gF(^J4E4Q!b1b0kOhlDO@KDlf~2%?M1k1d@YrBTh?Amc^3{nvb_0?OWitG5 zJ{!>29S5L`$5{jCR2Hnk^;DE78A1qrOMbNlnSS}8y^k+23${}`6QloNHZS}sNO+yZX^0nZc z%4iv*I9#ReQL(F>bw%M+h@|2L@Xo$+#P=gav&^(!^86*FB0{7rg3*~i-_Pf!`uYof zO3)Q9@s}2yEk)qKMHn^=C-pJs8fg`fGEaz68PA#UMr%JxL+lf;ZvoI!XUK^;ym$&{ zrgMpDKq-OsejW9*CrHA_H|#VP_&x)I*$W2Bngwe@qc+n8f zs5MFkL~-!dC)eiMQ(7{N1h2-q2r4N`03~!`QsMbLhy@*dqLCTh$kN_xMDb4+E&-8> z-V!(3UUTpM{!M=@i8X>m07lxT5ufJFaIhK6C_$mXVR6!3r9Xm@x5x)ca9 z!hzs$n^WwxKd8R)H)<$87KJf9IE&P#D+*yiMIqd( z7KR*L$d!1-%|j2I=L51&cV1(yJtZ>X9Ee}EjB>B8OAP0-31PDDf*$;YhwO6=B9AB@afY=h1vwtNOh>i}RRUQ5n%@FVY+E>-Pdpu|@e9x`>oI#eN|e0A^=VU-N|LfU;LYP{Tl0sRl@Y&FW|8CWQWm z9Mfi{lK&YmG}09E|D~UcYz@y*&g3NP?f%ed4Nljw7XZVoO`jw7lZS%A2Sv~E<00gUJ*oPmgF0RVZtu5A4P!#&=-1>rtAN>1)q%`tNn{Iz;#q9#* zZ9W&AqeVJb4EQ!f^P{%#Y=LsZ@6o&FlD&+MLxVlWdus%HoB(0d*|_%T0(4eMm{$A7 zk;C(1(DrO58VqCjnystNhYA1oh1-^CFgz*XE5F(M<2g2}vd?K|^N>-apVqt%h`&m> z2Y)~1zK{(laO3^BXe0}+l);&KL?%n_JWLH@_Ciaz-~gfcj2%!OFp}Uhyzn3Z&pJv= z{y8nPDzRNYW*aqUIipO08N7m%AV6d|#sk)3%G zxT`dmUU2r&RO!|?Y+vBJcKy)R6rD?O{@!-AsH3L{l%Q}Db&@vR@8ryv3a=m{m{551 zPn7jIh_quSDyiR`VvQQVN{a9^Ln6Oz9vpMVFkRLf)LC;udC#`*?=D|TVZMBL{Dj-~ zXbExJ{YcKj<5l7%ylD1Vym1_gv492^G4Rl^=2|n95^aN9NVUoCZ;A%Yz&14%X(KhLKL_c5K|rs z9-G<3nn-8dCWc++gl}wZ05w0JF|k-4vox)moB0I|=&7*M$E;2?YsD9CN)IiW)p)y7o5S95~K*}xu`UFqL(Yqcdl5p5&)gw$) zkp*s`UqH(6hl@+3e~dlu3f8AvP=*}cMW-pZ{w*lD?XXO1tj>ZW85`Ty2Ri$mRxzk7 zZTzbh6-{%f9gp0cWf+!uSGT_yU~0_W2c`f1qyMQ7zJphjU%_uNlbvAQ&HwtcNXda!1TssHsyL zF((@(CSxYw31+fxh_M#B#)+{*&hXvo{vrQOQDc;Z2Yf8*YIa9o^tc3sv88wfj4ufC zs7w4gp-!5IscC#bSpw3=J5JlbZMh}c14ddZ-YlOP8)(|;lZU~NSyH+@U?v>(5nCdU zp}^@3Wp9VZd}#w>c=)ANoq~J4{I{_I*d@i2@p8hHYKDTa!jZQxh2TQGk}y0`%PBV| zVxcgQc?4r}v=?#~L>0ERvwwXEt9LqYU)zuGU+&W)0N>*G=^`@PH#ZOt4q7?WJGz9ccM8;bBp~Ta#`ipqnYU6?SoSB&b0N? zl|N(2{iaESc^J1#V`4u;$_m#Vg9Lu?fEN3&4S+ag55gLftuBsXvjv4|coXYxPKX}d z0nRIr=j@FUPpJT#t`-NgWm_zH@B;h+;G?&~0%3Y42PVMx*A8M*8~QxNrP|mH46`F6 z;qdrI?h19&<}>1fv{Sqnld+z-x=g5{%i$v#a{Un^`uFQy?g7^{NaBPO7fJ(Lzhq#(v0&y%~u{y@%zKdf;;+sGdWiwJJ1w ze$$`lji0{%uf~63r5yjCm9nsM{ukTBT5VnDEwLuwwXVH3yYutahRG1dJ&eA0;Yd^Q z$hAeaHmZ$IP(_AN8b}6$isARMPlw+j-ko52rIzE%5*<_Ie}$_h=g!ZZy!-lcOCovY zkEQP}jr@}Gsbqb5l0Dw*AI%zheZ5?s&w>?8i6G^hr;8&=AK^SRLPfz!Rv-&yx={mmwX}^cKG|4p1_}4l+x#6jGfow_Wbd5`8TPx!=X#hk+QZD zOuXBgxjKHUvq)+qB)OTJoPSA%0OWh*?uip}{gL=Vi+@h)49!SsEfGehts3{I5+oH>Slm*h(3z_bi}topR)Xr3EMq9z zzHrc`Qc%A~c}h{xM>;L0L|+vSd{6HFWn2a$7&o5V@YnJTb0t$0alP`8#7?<}4ohxa z=nN6C4ji#=7U`gxbx*ZOie%42;1O7!;256)X{d#t< zHk-8k>|S=?{0KmR0w_S}=>XrXG+9ENhonjM&}OP?@_HyIvyEI~_r@(IQYf-cmiz{h zb`v%u065@%kPULBTU0d7B??sK=RXRW39+;rchHcJW>1rT7L(HSqSYb4$<}6uHGh5k zJ1y)m@4sYT)DF^dfV2I=r=#3ORZ-dV%`km2=fDasDkXXPZFKFE&^>UcMCZSH-(s(_ z9{~K;TTVK^iY@-CJ+=7TOiVU*%ng5Lsv#a?=yqtJeotxX{IGep#l2cC)Y(Btafzcz zQKn+pmqDi@skLy_#}V4p^~lk5t`FUPhV4+e0+4OP^~L0T;B^==>g*bDfw1K_1p`iE zSQt>Ww2Qt9P&#V<$Q6d<5Ae=bw^>wm(_T{P$F|e+*RlsM5%=D#elrWSVEOB(M3cqI zCRQI%AMfeu97k)9f(e-mNM$Tt%U~-$3j?AYE`DO*?q02Wvg+y4O%jbJCU6cX8k?1c zGI(kDgt5K{qh$MN4g)aVVv8<)(6p@_L@xBSgHV(FTzms?3Yzsk1{Z+yGa&PGc)Je7 zCA_PE-s6@MO@(>?-aKCx6S6vw`%pavlLDvy!=qt>_$+Jt<MH^q;93|ysxVWo%`_isV%>D@3sD0GCpS=1$zPH(S-|LIk6iYTmxLx~+HiKd*`8mL zy1Cb60o%iJq~#!_9_Nb|LF%s0hs~HB+NafpkOVq5n?}C8GI_cD(H#qOrB~TTW>vR2vJS)d zH9PLzmC#K`u(^g7jNR}3+s9vDUlcBgiY1%Db*^7t(>fkx7b|heC?g&+h>RPq2{(Fq z83|vxnHzYgciSO^j=tz)ww)xqH-}iDK2?8_kjZ|1cj6{&>Dzk_w^58M%3U@-E&Arg zozXZkC{CBfQMJWNyZ=N;N(`m}^$`NYb<{YEpX&zQ?#N5K9md^0plODJNjr`j9XJi^ z2!BU6pEOmp@oI~h34tBUSmvC3eap%)>aln}+k3@+2=G{KHDdA{$o6P09=~&H^+}TT zU1sd_Osm+xla4tuFfrXI5*P0G_fCU z_$MBi8w}WjLqM-muk_0Il(ZHOqvoKuUv|Hda`*Y z2j|_Js6Mq51~TOrr%7$6daeN|tjCtwXL)6|k0=6_8(5i^ql{jpIJ(bDOs%`>*F89u zNYP(bS#*Uv2^D9&Z8u&c=vC2qx#O=hk8r1v;FT1`G2DrmRwrd9$Z=Y^y1bn?uE&nB zELQt1>!)Yb8jg-cX8IY=Gcsx)9vzNEqo27Hat1y#rqI-WBM5ham3UTRU`A*#d4`d5 z0$x0{i%vxHyhxqns`o_KO`goJG_C-GrWjsSXWiL$`9MX005+ND0oa^Nd~?Z@ik&S2 zFS#!I?40WyDZgYKEAm{=6iTxwfiSw4E2QePvEkOl|q@V1V(IW$xJvAcNS z3?tFS$~*DYBXhS}T}o6KWZK-Cc*RFp5*N)Q3up4)3lWvbqzGz*d=*ajrg724#UL=g znuLHbzG_OiEBQ-zcVUOEPq66Ys$K8g$(AVQtFFfWZs!r2)$r@!%>XsbH--__NWXg% zymzl50t%Q5+pExq|F{B`?-Erhag|RGkYhcFFoP_-_p$)B3^HtpPo7PtAK=bB*@Ees z+Qu07($vrw3Zka&uYH46R+q0V-Ig9Tk$7bhh~gp2pWOaj7m1^hgedf+g30k$#8;}) z>d=U)>K@#u#@bjy@YTf<8&dgf8v>_&Ogz4Z_$^Ui$mABj(vZz{-taRI(+94_0}kfjOwdieIl)n{f|EHh_Y#%&CC8i3Ni#SSlC(UdhvSVc zI4x2a^d4|%HXHP0s<=#zO}YdQo`6(ZpkowjjabA(Vgh5HZ(KaWwUV}~g#y-pT?8sG zDcxC1(Vz<16s;OISxY9Iq)ah0jF3x6{OX3IB7s_#EIxtpNQ~*Fd7ghRkD5NeLgvMG zo|{E!GnolyZd+Yq14M}*ZV>k)sUk#K=Lo^P{j!(CKQ+7L(s4T8;`nGVuxa@|+Ig`klf5G-ZR zdHrXpjc3;M4^ACasUH5!^(~2cv>V0I(IkgU<+))4*v3{AHb;~5wjKp!Rs!XDTf9B8DN*qX;qI_T)V}&IzSfkcga|AI6#)0zaS&FJm z`Jf|!S(_}rpiy6z&%vs}@u+2t??qsm>mp$*R9AWF(zV`WS>_XSD5bdsZ;v9y_-a=> zhY?>Aba!H>jvbfTYU9wW1Y1jf{7^1E`(WzqZx$6QVIq@&h8-^F*~x`7$grxo=NoK1 zP9hmos_}qeW{8>PgTi&NG5t*kGE<5PkoBp*x<$#xG%CJJ>n9^)?l$@(+eyR*H#o1y zbjLB@$57o;W8N>-fJl6fZ`?AX^rBzLOi}=ojVbkOFTq;~)Z{ob4+Xl+DH@I7&;^Ye zWP$@XuG%5^R#A3?l!`BVU)bw!lR~7XbU+SRZ(*=UQ*Kq|TjJyk>AJ zO}5XUQX?sos6|mnD#f{eL~airv~sa6+N?Pt<~bqk8(`Wkz9AU=MGUBk zjSu2CS?r&t1cQJGartKHk1k}Do5QTx&U1v4frZ~nG6anl;RlN*+A(`8%SQ&3QBdo} z1ow12abS7Rd18!FS7@eD(NYJ?w6hea+9dVE(hBS}LUI29On-SZOgsEXZMf)Dc4RZ));{whrIDJlqh_l0a4g@4P1 z)-V0VLUa;Vj8Ik7(U`|;?(|tyI+yZwOrroE`tsz9j_I@VAV5=cqd^d)Dh{oQ0Fz%~ z&^3u_Ir1a_Z2n3xfPg!i1y^|*yvLcsW4)oXaPlkT zrc1TQ?&A5`;|Tz&dfpeIx2y`DrMBf#W^Pgv-+)bhxJv%ZU~rm`^MSkJa{$f6Z+EC( zrkNoOUxX7J{@|ek@r62iF7!!bH5!#6 zDRJMX@3lROD7@%i!7yfk@?ZWqumry-=^l+}Twk#gLN?7?1({JkBu?~se*!cHa!7pR z-_N`km+B;msr|Z#85&uFRyj18lOXOpx$`{TQiDEWENUvo5CS-!_Iz`N#D%Aws#FJm zvv>*O6mIr^1tjs1PvUqRd+m>w&j`tlf-_6jpaYF}Dtk58-Jmg<1z=o#%;&&8{$Zlq zog$(G-ni+y51WCKc2^p+ZrZB1uI`)g!r|<0JYi<`B}RWHZRQ3V55Z$k9Y~aiQhQ7UJGxNCOP&k_UA(+r#xOGh$C^wUD+TcU@`2`*?w3){HG?`2{Sc?|!<_jU3H{?jO^3n5b`NlfP>mcq);5tvE_^Gqyv| z+$F(B-~h$4tvP%sfvEBuD@r9C;ktOOYiKDW_nqQu!tP2X?VZq$R03C-vcm|ID+=#y{J-KH~f@_=ugE z;s3)K)slAKY){(pzFpo5RdX)nM3?OjyuumHZA*^h-F7Z>#%LDNjv-Y_Pq~uwyZwyO zlu-rt87s`qTlDR&m*9a~I;^ULd0i!ueI*(0ef96*lOSh@{NdVrbVL4`UQIqthaAl> z$#@<5CyS=9LK=eQbL2zO4-H}GOVp63UxGfrFJ5BQ$(N?<8zkVyB=R`UXkfZ`a2U)A zLsmfW$`rN8nx5VNqAndE+wttLE&b{Cc`)G0)~FS|nxb>$%C(@x$)7SwD&4ZS;YHhMR4wR@<{;FQ2*vsh7>)1yGb=4fONAjUg= z$9|o(e4YN2MQ2!8uZ~KsT2-N=Ul;MNKNS2VVSl#;Y|(wjkZL zC$tQ;XNWVsIw*}}ut-aleGzLEDQnN4;ViBzuF(B6I7s*SdiNw~?LJ4fwu z+Ae$lgQKwJikY~gQw6GDZdUFGL4WWj$Tu$zTB4L~zURU04rkhNInA~)yBm5Soqz?^ z3((_QSm$=2!0nA1LvkGEOOxf>2Pf9DPpw9AIs$pRaY7;URPz9d5z+4*{1!wLu5~yajP5j7A6%_zEtnx$N=u~k1qw@ zBh|=j^X->wbyZ8*Q@V~pmig$cG=?l(b6aIQTm=9I-|8wQF#-Ot?AX;Inu#v?njWk+T} zbwfjgtOtXk4&6Tx3kiFNb!k*z4^~UuKLFqvH7FeS*88@XtQ{>x*;CkV^VtAiuuD61 z1R)TkTHTmCW-1GyR+U16>ZkpLnCi?{qNoqh08J3Ol~o>ZT$TkdH0`j;wj>1j6(YC6S$4`=KGk<+Z=AMg(lnkP)*pEJw&li%wtmx`b@P~ zFO7z2&M1ENg5E3(Jk9a98>j=Sd`t2yoQvHeH%fOZx*M)IN73;f(J>(SNtzpI6QCDZ zD$}Jv@o7F{e(>;^$@7Ry<1_;NE13d$62(@P1yA61#p5&MUg8lyqQ`MYFfQ^ZIyw(~T z;d1`U^_s$Has#0JacinF?&`0{M2GbSj#!h#0?B8CNQh8_=b33jQ-oGpUA14V6`$`7 z!wObO{0^MI-5=3pv4=Yset@Wjq~3jT^X=6Tx`V{1dksAA$Z-hvxv~ z1uMaP1Y}GtpnW#&3vlGb)TAe2Xa%<2`!n-ZcqHyigD&v)-MCjpGo~O5mr)3`64TER zrk8>mKq(>56X1xT=JZKwdGpXkRk&P|%TA>F!91@iz$4(zx<#V&%EvX^&JxLd!(GpI z-rhUjCowL`Q8*aUi*#y@OP6rKzU~!81K5RvVIAb0oA+vO_0B6r&UPdGaqXyqUJ2+Y za2D;Chz;ZmF{b3Db%-Vr2s?Z0>sFBxbJs%h1e0ZKZX)WO5C+CxO88Z8lWGNnfbq*^ zt()&c5PU5qhJg&YuIAta+t2s@vwNC=B-v{t^S2Zllw=FRs|FaWj|?UDbBsVV$a1qk zR%*~kn;>OjKg4f~6u3P)3UJm1_yK4I9)T_WWy z@4dTtDv%^+YRITPO0Qb;e^VroSF{tlf#&Ow0;}>k!AX#xb1RSwT~;S5@1X;g?+#{gbV1 zAG{aAXO1`EB92$zDs}0}reSFeS&B=Eapi{yY|IkV4fd`MF?`a?AHlCNb zrVQ)UI7KiH=ZR>#%GKOU><&R8J8HRSlxq&H1t=JBgEaQn?Gh@DASJ=}7#@cSzU~i) zJwJntdvTtUEvBA@>Em}2R!PBWNz@GmG56{o0I96iY&0XurHN{x^kL8hCUlAtR*HIVNx|CoNx@GGexL-HjJqgnuBo*w`{o#LyYR3JaJJAA5ZerP=iI|A!AAK z4lD>NYn9z_|Ft=?*pi^l#Odb~W1SUukf0V`pDeR#-&(P~P&iAH0~u&rlz6VB;O+zJ zFABsFeU;zDOo7OH*;OBwjpUTav03}rqh#=AjK^%6QhW1d{9%(CF}(=bDlZ#rcO$aG zRx%z#zl0GSk=*cC6dMLzb`W^}hLTj{1G~p>FiCbW%{Le?CpVPAkf|UjNg{rY|U1wc=Lsnx4cd-hq)u3J^j?cn9u+c3Ms`q6R zPXhm>dDVHKCvzU9#V=s8PS8@m#djxI^&Jql@mLKOj5@eSf5D|X=&L?$s~p-#WURY+ zq6()Q_sZeIKu@8(L5A@rjHcWAZTcxRwB=T)Ax?TH5YG1c{^7kKHY8}_K(@g^Z=mgI zwhlKeljNrzk(rRAG+VIgy0<}&o}F+P`bohcJqmxwf#I|;hmdJagj$*RUU+<>s;u3KIvtm&BfMPBmwZgYuy^`$0DA#7Dkwpn|f^1mLalPJPnt<>cPEzGSqWZ+1q_%%yA04rkjK-&WZGO zXZb`Rs7boUCy0)s>&<4e=qMps^K96ttMdw^z&5Zt_9--nT_Y2Xpw#jR5v?}ZW32rm zB*pYC@r;WMKx}=73hD1}3H9RNQN`jskclPZi~;l&HYbn@IPY4|JsFVc0jaH+TksJn z$vwdG^2&05qgreCQK5-T@wW#cqPOJvpL2%CM6~X~HXy};qm{MzWWb7q1t-vC^ev^4 zFwO;_R&Nx?;&uIj-bP5b;530wwxH1KCL`1X2KY%8BMhc*D+Bo|p;$dCk@|+_ZvOxe za}C;6kk?!jE(nD3g&fF*?v4L}F@9sLcGd8UdHk{H4nVr+TyhQ9GYvAy z1J$G(`!>e4lIRL}6Waj?W1o|unxD%#dSwt&dnw@5{ z-XP*^#{MFBDQ1@7xtB!>i&*l_^#uxCNCFb*IqFL)B9@*wujlT3A>R=EcbC9;>$Vus zjDrp*xNX`LuLOWNV2<4+hsa?rkM=bEtvy>ztPDwz8y%A(SZjLGU@(m4Dd~1O>xLoc zCn8DziQ@;A{Zn4s_QGh4us&m<$|6e;i1oM4Ltt3>i$;ieOpo^nKy(})l-(am-tCy1 z2p50Hu0yHwMR1KnHE&8rnT<6h`gr zHrLAIe$TfBjxzY();9;a_=!t@NC#PBVBN|MdA&qEk9Gt%fH=-ga8FHBX|-Jf00wKOWq=Z!bSul3=5MWJlY7@r@~Lhh zOY}Bjx^H^wE=)ag;o!O@2K|Uto5w}p+o3yxpN>ifG#6;Y(H)T&10fVI2iSGeOmpW- zJ$De2TtjJ9(TgBt79~p53uGAde6{YA@QFjNxVYGGS|Wr~zzwD#gQTCQo7X#GUWFhD zC`Oe2Uhxc!#lUZ@vPfUvn1L*qQ1`rd;VCL_cHZ+D0r}mwCvdzQWSo{m7H)DW0b!Dl z(Ubv}JhzoLao^a7PkiyO8qCD!ucW4Wj!+Dxv`(KJiAXogtF1R9GQs!6+%un@#|IQ5d7V=>&gzz->3y8jVANEaz2AQZKmH&)!4fcXgAYlYDSz&;g`LCP*7(laddC2Z6nCnvh zt?dIL#65v>OKk3C^={F2ufDTl%`peqPW0oUJ3@{vRrcKoh7A3+kb_lFv~8;*>cxHv z%aca1C;d`M(ZnKxPhe?0%=qTm+I#LCPylQsyqS|a9C!rB-gFy;&TuI`(wVdx;`QrI z`Q;wKLs&PU%6}8oOG(!T@5NwXS9fQ5Xu7UP%W|aW>Yj~)El7T(EjWVT zxOoS3zo*>LUPv9WKQQ5`eqG>cuzTA`jeS9Z8z&69UD_b-Tffj+rCI6ma@g(c@tluA zPFqnNVDi%_Y$r1u(})r$X^U;2LEE+|`*8L8`u)vUk!71Gv#A`PeR1W`$GDL&ATg(bNxPbHPIc zoohp!{1tPgXcSBu%G{LUvUMD!6u-b+rAxpH}${&~(*#O!YbW*(-uTrsT^E_u3@aCQUX{TpV0em+Si> zptUyxX@VX(QH$NpTEfT7Y-4$!K;vx*$Lj`yyxzC|9kyEYprJL#A*$Kxr_PjXzeD>JKHjf6+kuA>s(CoAG%6mArj7mw=@F13X{;- zhDQiUll|J*KExqE=?tG1N5ht7v#hzE3?r6dQJAfsqVJue`7%y2cS<_+zrXcyO^~NU z{C4~&jFNn3){)8or@U^-3awc`AveR1>ww64PpB+d5!pnzI3hEn4ta32J^qkfeEuds zHy-IjXU;wDfD7xvDWk6soc};PuwTslS6;8GEo06ai!QDDDE$+-D{3c5cW~BOlfFbZ z`5S}Dzx%7uNp-ICUl;z{tH3Z|e3qae1G4@gCe%pAY|zBjzzKz^E}VTbxXJr}>!Om{ zG5<#wMJaRkfTB1>JP>yHd%Kt!$oQ}o^@3GBEu|68vB8H|S82+dPZj^m<)HFCC?EBz z0C4P;xPkW3c`6mv<^!b{RY@x8xxoXs;ckjPD|&IC^=CFl^&EA#x~xeE8&f#)yK(>@ zAuQoQbDi$O6&h5Q6>19HF&Li~ufu|G{=qK3sqz%PL!|-SXRL)o>shsn?&BR69vm52 zHT@l2vCcA=*)nGH8<^&l44}qW@m-LZhH^Gz|(&7 z#7MKIv-vzXI0zI$RAnT_Rl8FWG=E*EL$J+qrJ~T80!+S6G2Fmy?E&VQX%zPiB+cZt z5zCTi}{ru`{p=9w@qP_I#Ts-) zFTGtLKAQpzujRR2fK*Lcpc-IcolOe$Zo~!p1U=(nRePAN%fql#j@)Ar?PjU zpC?iZQt46bcI-L$)ec)?J!q|3)4@TWF=|YX_hgR{aVp4w4pilYXEVQOYNR2}9N7n_er` zwPi_Q5SJSIH`z0RSV2?b8|AQ~0o2>5B?r#GQ(h@pC`#{75~qc&P)PP@%5f)^z@yEE zqRNBtP!)gXqaw1IU#N);RvXV3F#Wh8ERNlk?G+gXGCQq$@QG6gVqZy+P>H1Jt%%N^ z;R9#`H4TBhu-1Yo@4D3$4*^Q~tYB)s+WI!F4h*2DDVq3hg$ND~W-MKu>eCDdv#%@2 zTs(Dxz}!MSJc}SU4c<(-o)v4~o!QH-yFhHIlgIFgVP+2y%Xlq8VvwA8wDkRr?^UWc z#Kqs(*OGee`=(FQA$~)b*{?_>U$qx~2hUV=Qllqs)$L7>cIH*&5$kDZ!5-=R&IGQ$ zlCxKXdGNE~caTZkYsPP$Ru0413NMvcwN_xWfHCGz1u@m|8YvB?eY6;rT6K79>^FwX z@nXI@x`MiK=Z)oiWQTnsMR_k=d|Sgc*JT9ZkTdY|Xb$Rh{nC-5 zj$k{GI8{gHf_eH4s~xyWh$l^gnsYqz$T)aj=)>~iEa0`DVZ=t!-g6lZ|NiU7lIF^8}^&2og7?c5R&&D_nSFp z9wPa6aAbup`dUETxtOFgPT0eycpaAW0fK(cpRi%qtuT<7}nZe&C-SPTjk%#C>d;@&jgwxK9BSs|FTmrF)$rU zif;h0WMurjXms-7q46^t5N$zq7YIi_7l}%MY78~7aaCt(3nMKfyZV7AP z8(2lH8dO3ym6pP zcM&qq!^7_r4>gy8dT)tLkzDJRM~6B9iC5=?yNSFd=j7=Eo63h)AaTzEt1!GmhVY?!`TYbriESNSVi#-m7 zNnXpO@LpIqFY2}73K!WySq*6u%TKsB#YmP}=-fOa5VD+Zq9g^%T=P~x-}#)K)(VCz zsYKv!$-uamsC-k1-LzyVE@aQyK3jtfKP>A>2%ZeOF_qFG@!0|7Q|cqiC}j(f8To`1 zCqEiArjF@eNvDln{M{qPwpiOVgTq*Fqf#^ z*&MNNpfQ>O_sx8YMl6`I;sPq=B%rIQ0Kg)l??cL#67 z_Jf-PLF|1v>Wm;qEyHj!Pz^^>48qehC=(B0Q7~OGuq)mg+UF(}&+QN9=p~)DU;m7N z)t)CHnrfzuc)s|(W%SPhfkfuk z=MB7-)xhw)esA9yY;?rAYS4te2Uw=|vC=L5lu01aY?~{;yT5v%@}}$}}^vyA5V5b>?_Poq6PgD!mdgmo9we-7|4sIczbp9&hhb<4o#NYky^J{!uZ>BuK z^V(#eqNRKuXx=f5aLMciPBBZBu>xwdlYF7o!14_U+x*K+QGFi2kZP6jD_+I7SxE-H zVUV(~?pdTIA#P3RUokJLc;1q#mp?9TZf&d~xm6;Afu3n-^vW3~w?q5tcwM!FXtiJF z^Xe2^rdqf0ykJ%U9}Z!dbRw#BZ(Enek(ed2G*E47fqLlvy>V+#9*sB%92SLg!g5_N zjg}I|H;!^>?P0VtOFjQZ+^qm}I6+WhRxn0}o?t0+HKQ}TW@0u)iUzK$S6=p)_PF-F zp3FIc5CF}wJkJ8EFH`WZu5Tt4`JTUQPjOtAA#PXEJrY)aAPOR)hz6aqmrI1XZ&zj+ zi*ww+l*S2VQry6U&P}1u5{QSeFMLV{^KPE=zk21;cVER=H`R!kfu^I&y~ycbB(UMv zBqf2dP^XvCVHrRKa9w@@_ZTgP^yoTr9DN|?=4Sh5zm|;D?}FHXk>4g!!T8*kY1zAi z%qCEo5PpG?$j~`%za%9cnnK~+m2drekvPUxQ!Ky`#>?Wx1o#KmpmInK$%!R+@tHsX zNo>JrIYU{{0fgnwnSap}Erm4-B;=gwB3V@2Oxlm+Q$$0|B^|Lz1sRaQ3x%bEN)*UR zS9R6Sv^jwo_+hvyDhH~?)kp7o9m1{>3g*y`(A9)f&BwYhAA!DaQsy9?OE3`v{q<{Z zC>tr}sO+DDNhOMH7$mLAjt!tC6|rY6PB4d7$M# z89{Nje{kO^uY%9sx(}$J`0{Hi^SscG%8ZW(6xKZbma8Z1mRcc$TC1O{EP?fWFC7CL zHO7Y%qP^0bZ1H5C?F8qIhH)Yvo@iGRxg*xZRZLh(qQ&TYOS!US?mBQ#L;;M zz&q(rT8ZTA%a?dsTKUQDeN|wrs}5hvmqd;pzU9#+#=qq;xpfHhAs)L4(}s*VdMh~(p+{1Y#T#1DB z9P)^dItMr46Hb&5%tmJN@_0!T`tXIll$i=G;~J zmyLKZ=$n&c{Cw3#5!q!Uj_DF4apxywhaJg?b(u4z)*aAz z?hGiKX8YZ_mxRuIg^LU>QXHSEmsfaFho`<`!|iDLIvwn8oKo!cPRCl-cDARdmfd=D z*1Bk4_i}jfqT;aU%_n946&6!HgN32k$8~yy)-FbB$ z7q$-qDo5!82_YUOt>Rc)3J1cYa>7k-7-1}>hyGN8x4zIru@$yQE2*ca3(F}}#K?k! zG|#+YG|ks$!&Y8wYU#E4>Z7Xjs`X76$BZ=DT(Q2%ESWQD6%^p#l+@tJyt?Vaj+XF|veJL|>c6>_O7Jq9s=WT&z z-CYEdS5nAWPhu=zc{$%7A$GBV1r#$Zh76*{ zOsM<%e;wW3b#~Wo&}pw~^Sf@FK5+VT#27KP*J#g-bMwmux*}1&N+B@_J9va3Qih&v z$-Oh_@&>I205+#K8Z({oLrx8bL_LKa62cU_*!`L=hs$>Nxna6T3Z6u`KGhLUzKQTx z&{J4?|BSVn!iZnfKc#C}fRco@;{*eVVHuCGwl{uhe;u=122zVXAnstfOlor{rcE$u z^PUMHN$IGrG^d9ji@H+;i%c0*hyQ`xf4uF$4|AwCye_WLocU?<6^!xz;g3j^GFw6I z+UyK*@ODm|&R#a0?!f70RH2tR22A{UGHUAPz%do?vCOucm4X>n?{iZNLeNOjq4{?T zGqz=~m;FB=pXIWO{|)Fd{=Yzvm4$)%e+GIj{{ebqi9HXsi*h!1&9!j~!sQqN3pJvS zBvQh$Gz-j^%N#fc>>@@>+XcWZrWdDI{14N7ug&cFZ1}1Itk9B8#p`o|@Jpbonwk23 znVO%IhFN2Ss6TH%>Dj@LcYpWjA(M9S-ha}v2gy$%<994v&ydqAuI@g!f#VaxL#0!c zFBoQZ{<)1jyEqZa?Bw$D=uHp#h6;GG^GuXcG~te?RiO7F+!iQLW60JID$en3MR^rG zj-kp2M$csr81AK$$vRrc+|kZCGh}L2LU=Q-Nu>*#R-6d`DsX3C+V}ZQ9sjUiMXkB> zy8>ZW?M?PGdY)`npw7@iXU*yoy`(9j$-9G(LIqQE?cqc4F1|&^-W{mK-8yhJhhd>c zw#kiyvw`wuk`SW(3Z!{{N)uSc02wpN-wtr_3H!mGCrqWodXezVaKOR00ZrL3V;BiXoqen7VLdvXGNPIDh}l=P!rn0 zKZi!Gt!ZEn=iejAP#PCn=(PU6`L68i(M{X#f;8_+CmZf@0(ZtVi{tXurnmRzQueFi z%6jcWXK_iZ+gi&f;Hf05E5)N7d`v!)_ze1Du5FsjWkTd(CY~t%9gAfx0d6Hk>)&L2L`JS7}xseDNclaPG2B(bf8%(4QU zX|J|^aBR{Tw-6n%M4^A6MinCr%5n+R5~qc^F*nWjAX^uKBVDR>yaSB7xfAFdH4^x& zG#(W095YPdi4-R`WH(y9&g_nRIu>`Di|O7suK$%igr!oa( zT5zZ_=63nh%l+QNC1Ormvj}5lZSZ+gtNn)A0E(^>X|{04Rk3%c78M1?utmlVuxpo;fRmEF(fBfKs#Hdy5V8fBOo|0j$0pd3$wRh(QQ{p4I|M z{Pya8aU`3*aEEEbm+y-SWVqOXgB7x!$aXhRITM8epjM`G1B56)1T4?5u}q&!bb84n zK{wR0nVJLMyDh!u>Qy_SQ((&9-c-@JP@!zqRl6Vv-GnhHNdyYQ(dF9+UY(F#LS)6N z`b`?gOnRubqcoag(n5!_{Xlix)R(=49dSg4hdkbZ$i5OnUU?>s*h40%#^8~z8}wa- zCo~5vAB885L+p6%bU5QWm`qHUKqiINw(USaPWl-zh751}_|7$e(jzUsZ}ek>=K+=? zdJ*RRe+SstF(e^Xa%CL|3U|==Eu5hekG`>fNaW)3X#KMGGf}fppIx_KwnJYO>f0x& ztwP4E!xAC-W{u8tBlvUflZp7>jM{pR0|1DtGo*@FL}fr)adN0;nU5s(U1N9*c~ZEL zhqB(1GhBxT7p5KFkfS&|0`aR|(XP{R81C?1bB!;HU;uMN7Q1_G2w3tBr>^QQ(?eX# z*YCWDIP_rgUnYZ0j4@1g=9se55wYll+Od_kRAtPdl0C0 zaEnS7ztsc8d^w^02tH`k6QxYsVF4VRSeYm_UN8gU8T{Q02cy(8$bf^DFYGT^IuHt6 zkn$OxYMFsfFuo+oUp;TM>#_)Z-RrjMgc;$S|0Tto;x@;wbqn977VvksMc7$#|ahJ`hamM~EtwNfdqv5Er_lCIw-sX>cwK z()Qd2q*)$>H>(lJP2_bPKIsOSqG=$tmmQx;_7r43!HL33VSd4wMyCmqVX{_olq|ed zo4Ose!qykMe6uD+nKbKIAj&t_G%|s$Vk+NFtq>rJW(1b<7mdQ}l|00W1-t5;%{vR@t@I8mIlW_6)A!)p13v7*0#{i*uf(+OrvW9is ze)takFU}A>|M_X;3stv@L8BQ^16SaRPW6WkLrY!A6U3Ke#d0&y6ns03!t<`=ISJ;Yc9+8S}X4woZ2D(63DP`qF6II5z)s z3oLnCAb*ei^D^CP=K+#3PbnWZATUS=6KBConNJcbu@#jVGKK@KBgfc?lLRe-?8^i^ zewR_T9X4Y(SD29a#C0?tHnIPw8>Iz*mxb_6I?OoisIE2UGKZ0PK(41fmQS}fB3V0a z0;>*@C?siuXagP0gz%~@r5hmi!GeVleNy%~k?k1G?)q;fI_gxNhEv2rwU)Ox7~-F+ zFZ-$c_iFq{Ph8JMr$^0Il6bdHIUu;cum!A3-u|pX7y6rR6X8zvGLxh|dZAWPPl|`+ zp%VzqezjIaA;32r<|FQHKf?yihc34bgOq=-l~-dgoMW~@wjYs_a#Ah~*_t|hhiATQ z{)lNa?XN>7Z_xw>oVDkqI~L5nhe|zF>lm#+6qV0TP)iIv7Kse+3b7s0O=k4$%wj2& zWURi8+|n*?yMvu9ja$jGqTZxQ{#&lc?1-Zz3;5<5H1831ffekfD<2!?xeBEw9a$MR zH`ZcP*6*4U1mA~TL{aL;tq0=_=1iq-VdVJo4>_Vuig3REIdCiwJyj_R#y5qXslVuC_)VBP z#Z;?-oD8+=HLo+(McqfQv)or4$e^Q2O&9MtV|T4SgH^t2>)cRVbTVDC?e&gH>Up!a z!7>QPHmFK?kKbSKbkFR16)5qb41rCSm`2=O_YJS;uiDg^ff1D>Qojtc?dn8NuEqFEkLnBVA2686(muB!$>^?- zzb?Ki3D9+4Dv+>lMxXx%Y^63U>Alp6?P4;DoYh;}c-s@n} zmptoo$nzl5V@M7O??5?qQW|}-$BQ^Ab%MI2hjZXf^g%yC+t-VgRh$x_aBF`Vcs#{= zf>1unj@@2rO?HLQ&T@RD`-i*n$sWoq)5eTuNs(d4UYI=BdBfh+0%rIG^m#_iz4=7% zpF+&Kh(ux6JNUa}ocs%JsNQOro~hajbpHZ4CG=jERx2y;2a5uUPU8u5!W8nY!U_&4 zzsZE)<#tXcVj-kx>-nSkDd35>@(TymOCI`fWNHsu;2BO3<{gyB86cvkZ%XOVEjBUX zY}??|*YQ7!RhY1C1XhbmuI}}{>o=k>lMz^LX`Y26z)xmZY32$7?{z;oHWsN(L`+X1 zxoCGrv`2Y;Pi&skSaM_yk-^I#m}FsoTl_GwQbz4XBj@HQmN6xg?PlN0qZGzT8K5X# zFGH$a^r?@1#E|qrgsgj{mKpcub2@pAw&5gl2^7B!k_KA&4m%n#k}X;^A5p=^uN>HH zihN#tx%3YQBpnf(*+6d04JJyA*Pq!BPPtw{g}o^uRKn9V9_0g}?MET92iE52|MBCI zLy_l`A&Z;X#?##!7oZ4{P-0@gt1}GGqc<2S|9tuVqAf)_oX3w&KQTQ7Ejfi8*T4k> z>zt1&W{g)*F!Vz?S2n^=WC}Xtf|ZxJ`&bAc(kn+Hy~n3bp6K8*>|t(E-DT&^$Mv{z zxJh_YN)7QrRzVOXIkdvb zFmy2a#)`iv(K#?sD}0HAxgyXy$5|3q9>U$TKa^gdf@mII9;Ws5MJ=9e2wzaL>8zD>s~GwTek~kTOLCNnwrkS!s!e0G!)bK43^lXnEOTUL`y!J5;3l~=;EJT?bwEfVW z)_p%aFQ+Y51I{Al!Uuh1147mV8*ARlQjnbsblnGvNY7?E>Dr`ATt}4f4s#5W_X|zk1Gx zsJPEfhsDLWmOy%xX~~WxG&>tV-j+h#AAMJ;Wau$91~1}ef!A6I1Va;`y~_t^J5R@i zX-$|wUx=fqXI1b=F6eY=~3kR%ODmG3B7S_Z!Fw@T)z7LTJl;n=rR}*wV zn_^*Z+0~u?``Y6Eq|GuEOSRmw-CGI6s~>H4r6t&>{pupxojtQ7`fPsE)6kdcvl<$j*kNS*_n^2XQIvj2uqOC&vrfh6CReI0#r>2a(S-fC6~`sByK#!aRRO|svY$2nWDRa5e7px{NqSuCqI0$N>)3I{gLKT8PnUZ7wg*a51`zX=NU&Q5aVGh>8Wii}X3Jl4Z%9OG z$GmaX2TPQ!bnK}3o?uzso00hol@#{-udkuA>;2`?LTNr4+3=lS?Mk%VdwN;V$r$!+ zoUKkiZ6?EZC)rMI#9L7_t5MZ!qCEc;!2~4xKBGSB`^qYq$VcusE+;e(l#;$Dhb<%V z^R^C|z#HMxC4gL4PGji^51S}qKk+&!=UFwtH=ooG zBZOajfrRm|+~@bMFL1jzde?uVtjzzLva)ip|5w#@#)OiKKQ z1NP%^EJGUT<{8ER7Od(RrY<4Ms{_UkF zVtkqkJu?(^ko$V$@Vre zrjFfwhM0B?K5d-oxfT>Hkp>K z)%=5v$AHn`?9HX6h;froKZ|@!MhKDo$9TUlbEx#@@zl zcj-!3|9r7afsPnNDyzmdAC4FUrm}6}Dm+fxg;_fdCoc7GOvwbQpcB*%7C5QL8tr}F zta62`L09nUTGk{9ae6CCq=Izq#vA2=z!v?-N~Fv;+4|&NgT4C)1?!vaOo#{r=o>vF zI{<5)#;pc8L~Zfsg3Z3fzMY(_lGnjLI2E^(q!8-$zr&{SA(LxsBUhHal`WYdGydB! zQ?jiWonJM5FAl#EZ`g07G`s(_n;t=^(g$MgC+KEn=X}Q)Gf8#kKut@(*2|vT|ikMi4DoJu~s3-7On0v5;GIWqT5YXW1JqhfTd~$XM_JR5`uH zs-#0GT~>5a+cuIpE}^yu8L9mxkJppyE?u%IZK9Dcl2d9>f~C|tFvfVAXar=rE2vGm zvUE~sB=qdE)6-%MZ`HhuZ@I;2%Izm8`k#&GA6xI~qf4&&?Jt32;WFTypjMr+`l?Jz zn20(5rRD@rUq&=T!~%Zca`lHX!?WZRcU7A!EN$*Z@C!3-ui%r0uL7{SDwHfL(>&kr zJlUjo1OP|ti`Mjpp$~MzCH2OlQ{joJ6KgQF-O6%(l4}LoyTLIN1t~eR0~2SFoeBi4 z_V+_#3txN+3^n1Oagxjf)68ly`2o|2Us(L*qodjg1wRu6FAA$e61zjxYzC`3>akl& zkPq5I|F>+T#l@7lLAiVmx5KG;nn+}xo%lAd#hGB-!^|;i+3)3m)Cb9`hf0)U%_&aK zpih)yzL)?c>-f)NLNS)$J;?A0CW8b79n19m^>+H+YUf}A7G&MH2GP$OnO$`15`eA+ z*}d*W={-^?U`4FswJ-wOkG%O@>Z;Fn4qHjWWo>DAj657ehM>F2U#NTY`U1X&@t3_= zJqcaUd$SwVu>Kh@niU(Wx{PziTg=#>1+)1#%7_gDeu-&RN@#g0D=~zvM}#l(v>)5> zhU97Dk-A2@?~u9ExJ<`JEPj)6maz?`??DQXZlE-JEBlYr*=%;W?wCBFsx|c9@NWWZ zBQ=Z%0fg~7a{L~l24Ge?)k=E4diSRC^yiG8uAODUT>a)U*?QcVn}XDv zBxX>F$5kq|oJrg@aVI9y+BhHiML)lz8j*nmoa8SuE8m`TAHE&-I)NguJ&Q(vv%eM) z6c3U-0xW{NdX5s!a1o(Fg~b#U6a)UDDc0+ZRBwLufJ<$7`u$mO(cj>jf{ZCvm8*T| zPcP7+OaZ9872?f;qNI1Tr0vG8i(TM&g{+{31EAtv!g79f0iGWsP&JK<1LPwL$8!#F zcHAQhsAQrezj*##B-#T%iSE|R!-BUd%hN1{Q$;?UIcTPI)BXqYG`U!}%UKx2(;C*5 zW_3u!O#lcmU~`GPAn+G3T_A;Lj9YOyZX7LOCX>AVFMyW(6$>>};%F3i-1y3`^R=DG zCwz~Np}1uTerVr7#@qAr;jv2+q?#oCz|)T19Tq+7FWS#q-= zwzaf?Dg8nilJO7o%-8}OA_!q~TtoH;(I(OKCZ7Yu;TkB6h(Xp)4}LLQt9kAK2`?n$ z0o>au%ng@j=u4?uK{}asD%?i?1i5FN)~IKfe)ltN!|V+1vNEsOU_E zr$-a}(Po@>3g6ynEELTNsv->uu{6PORlGJDMa#Xyf|1^`p3*XX6=Y$;Zc@kq#fYC; zhRa3_k5Ib@jDk9uvz-S>%1bN^nwY@{g9J^Sss-|UYA#mVpyr|yE!e?jDY)BLlRGT# zdc#fc4k|s0;b}aRa`H-DU_;$sFhgy|r^IhN1&QGvfO`22CgKr6}9}|#R-hP(nCD}OHafBj~0`NM1faMO< zAsu1Ms+_imK{bOHMH#bJ0PP*XzRe9}Kqzu~*lDX3NTS->yuRq`d1hzTnc?Z|+EYNJ zTa#WKX{tbGyTP(iOuLm*0sqttv{t-+%zmdU2F$j^FCF~-8@7dca*dr|Q;j#x_>=JJ zyqnXGt(~B4#I4wd=8Rjnk1?^Bl~rlAjW*t%Jhk$3CDg6WTgDGB7eD0~KiOb+^p{|J z&~*vGIWK*Rh>fVP)S47(e>?2*OTR*t)dMb^QE!FyZB2SLn(_(3VA5Xt?eH?_=XQpe zkm60zCaOR%mBtOK%y#?wXnVy(T$94!O6c0iNtoS72U zb=tSCU;hO_i8jlBW@!2Gjclb!MEj@Xp@?MRbC;qGt&Vm z@mu&Pm7u|*-wk+8I@gxEV50C?-U~ZHUnBqyc7K@XnI+KhP2911u?}*uA}yXbA$l*) zYaq~zhMJ}TAN5rjM>N0$cwQPi`;@4x&!QSeqwqDaC*=#~S{cPNe5D6|_5R=TCGinX znXSrIKrkdG6X^{|kJFETjEv!_jM{SL75Qc$qu@*(@xtKvjzghEwQ&OIv;78g0oV>T zCO6&a86$RSU;96mN3C%bdc^DG}qekh#9`>w*{M9_uG4bRq#>ujN`O5RG-LeLxJW1{w9fv9vKX3jo# zpE#eo=cXY@sAnR^%j+Ml+SR00!QPX|zmZZ69)4-$$LHe#vKqc3*0S+|6IF@XJfQ29 ze(C+wRJ+P}#Xf#I_P%ez@0$q~$0CjFrx~%=*vnUcB#{j~O*1vtc=>&vN($$098j5wz4_>De#VwVn4LAaJNIWM`ART=~(`K;@r`y>uR$yXNL$YI#P zN?>SLer%@@ve9>8N*VOFoZZ*5Ro_YuSMrD`1N9qgu3I*6*U{sNxU_&11^<32v+(yl z>u}aj?IQJz{#!M-Dp(xC11kgZPGEQI$fV+Aj{XAgHx9~rBt>7URWXRWpZCB=sXz|f z4@Nwr1$!5QRL+7d37+{uu`jbDV*Yz2Ba9<96Usx~yL5_6tP;x8Xhvz7E@k7vp{JV$q&JcLuo*K-QN}TlPh&6mM8i!;-WT|2Yqx7nEZ({ z>=;x@ILK|ri#;W!ZBP{aKo;qD+X&9V24t4sk-wWt{cL>@(Q15R}XbXOAX z_$a`qHr#P}bX@Cao^sF+=WcN2EgojbYVyJ}`Nu`Uh6*$l@}^PSe<$Pqm;n@= zv#p`}06hGem`FJaWw|Lg?f|wIJu=nWZ3H4D;twW-5i3Vqr$FZrLtC>O|NX5>l6^wJ zYpC5Yk;enTj(0V(v61EQJx;ze$M8OccI~ zf_TKo!9^o~j%kCKA`>78PPN=lpVnD|&5f>GBe5NL9V&HFW_)^ksBb&N_irC6+T&CY zj0@H%+{w~aPt`d|ayF26^l(VNJ0|WqQVC&FnE+gORRl#B@LVqa#m1Fw>bQPi451M# z`RO%}1%ar4(l_E0Q$lt~SRIrdl8P@CHNw4q^h3p2+hl&%z2_gK(?+Q=qWhms<*9`h z@pBl~dTDfp2DQ3k#eUV%{CT8pf4B=BwH=q{TK<$=u>}Ya)Z{K-lQQZ^+Ybd5Dn6l; z_)O=8Od}K%FwvuKrDXANX1ofr1eGu0!G(Hz$-oX@596V##4dGk2 zTqR*-Fs8Thf_8c9sznJh2pzzts1#{J6mwkb@5r@8Z|gzRP1E_DO7Ea3ZhMVN8DXOS z1-clN;acM_fS9I`@l1MD1xC z;v@~lh!^Y8K35`4I zi~()2Fybr0<*GrUP)bmA#|3g20l6e4SFY5})818**M{O-Dnug?oc;N*oT#VdmJdk2 zE1l=P$0}OU;0zPPrfQ0J>S;zQTYRn-(8`8&>Mlh@m4j&ZjD}-S`!{1_W8iGC{`+@F z!!8lW&b5wY%Gi1-xE0P3=-RwFc*+F1Ibys0bUh4 zAgSFGdwauPy>D{EpEci4EV2X*9dCshq&x&+UTv>iUK&o0G@G!u!M*W^5}uC4qqTjc zphZ&aLRSDTwk1={PR=*^HxHZ;@>oPE!=^F9$upPn0z##fmML!6yZgsS!w{3&o07sj z2@_~Tk#K|y86#>gitF03f#BeFaf==Zieit$-ZUuc20RkkfH8Y+PcJX6tCWvS>mD7# zg(uq9-Py&Z0vc0sG%8KVo|!2Joq)QDmg)USEX#lSV1Jsd*}>#I=(_zD=^sXsUn>() zKi{YP;4T<@{|R)l{0Gp*$@=U6zfSo7e6lqQoa}BJ|Aew(9%^R|f>KE$8uRudAy@F8 z2>Pg&hKuk`bLMmN20xcM$8{@!D}x6GfkF%UXE(eUq&WfJh6d)|Pv*+c1&5rm1H2z8 zU(AcbPdE3+h4`Y@k+T^?@6V^DxiRkQ7$b~q#}T!g*tSp}gkMSuK417&>{#U|*$PV#q_rs7{SaLrjP_n57`Yubkq3o@v4hmn%oa=xC)QtQtWBF??%iq{c~a-*Qyafw>GU7JA@9DdYnm}`#;2@hQNib|D{CzH`|T3B zXCpbcn*N7%wZjP=$G-3@*T9XrxVk*1lEQ=H?f66J;rS0?(+;ms4j(+lPuWZc5Nub) zFc2;_f>jTYWMf|d&guO8Ms!bZkrZb5!w3zXeY25+Wq5gn3?QxzgHhhb7ovP(^pIm!ZUQj{6K)bSuEn!dif)~Y)!8)@=ukN0w? zjZtx&Et0F5YI9^@EHEtZx9)HBI$N?v8>M?Qy@ zTIP}6XqPD0NYTXJn(--%=gy59Dv{|b1*iILBTXUgtnMjROt8=>Yoh~tO|x)MB_e~Q zxYQVAa&U^f)&+Pz+kQ+NYn$Q{6}JEEgHae!rvl91K=ayLs`RzcKAVp>!wV%ode{6( zPloD~vuhEl!wpq(J#S(<^sRqXwvy?sBNeP|cBZ@p%3_gnqBC>oGdeWXB5i(Xu7#$F+>;lO*H(K@oX7yh!43HVoJvi-Gx}@%g=AFG}kPC@=m{afE24RhYG=frL$uFwt__ zij2Z(?=C8sIxJofyJy_1y%qWX)bk1pr^5>ht$&VMg4{+H@~l{t?Ul%L$lHPNhbzCa zZ}o4?Du;hqg`qMZInbR;UWjiEnKnam1UU8)a`nZqV3KE`Z97}aS}!ao3}oFSU|rVW zdpd5Gn+Q>=)E{mCV71wf+cZ}}j{v(`D1rik_mUTq@(Fno4mZ$GWbUDYnaD%xpPk&7 z#A5o1a157(!mXAhKyCvylsEvLt)uXyA2{%%5eaXg!RGAXIB=gewod~Y^z1hBDQ!#a zR6cxOUczP3f}#9aLj2e-Mj=HD`VNB(avgxKl3Xq3yUMq(0|qi9j}iX(`9LgQ;r$gBN z28}DO1{ojVyNb_>ns>ET%7_=4jabt@O5m2}S600y6|te1H4H3OAY;x|jEw4cZRfFf z{Al|*_LWM=7BC*9t#Unx4y!(U`soV^Q_Hz|2;q0~&cVA=2znF z96tWuwk&>M@9*b|A3Q-Ub$6KwA)R}uP_LOZ;hvWpeJ)GO;c@?cZSn1Uw*WVC6I(w@ zLv`{B>)`qK_<4V3R4G1h;JH++-%L2x#k8j4(`ZDSTW|oxd&B}Oq$U1*mw4D0Pysr3 zt?fNT3Ajs@HCriL44~29Q@U5@4XJ_JAch9-XP$bptX~b*{Hl>X&C2^08ijC7jp%uU z{B&VB&>BUJBN#jrSQ7P>hy_R|a_9Tr&1;UVO27P;UO@*0D&zcC+Y?t$Uqt24WJ31l zJTlYso{`bkM~rqS`BH~8hG{nEO&S2Kt*_~(nWPUHdp8~T}dWAQ(fxZPzZC9PdZw(O8`mD_6s0g|Ux)rZ}}>UaDq{YKvm z>kT6a)%oySylrND2933bs%>|e2Bh15P8k;0>$d~H zM#{xq8cX;Ken~;eG#JgY$$i+RxuQd`IQJPI(c=n0JW*3f%>8Pi_P7OVY@bU3O2wR( zqbZZXmNC+}>r5#~`ztA5;mG99ynj@hIW*+A$F2e|?T2!6wn{HklAJMG4nn4= z-y%?^jtwLtIqF?-bV+)45U1}RWVAxLroT=Ri)u?PX8UHtHqDHgbi_s0!ls6bKTAaC zDhkrd;*2e~%yz%$+FE2hOf6>oa{Er8OjS)lmPz? zEtWYn%NU7{ufbPvK+vxdq^0@-ZrOR6I(B|y4VDTw2%90^d|$%sx;Sxcwj9=@1Bhoe zOSU~!%R|((R~#Cev-osGwA+KIDoFcc^NE;yTRcqvuDWKn8Q@xf7YCAB zVpr?ZM7VT#z~7`=NiYq3Uu3kZV@;qro+;{AhP5Q>D$7~FISzB-lWW&+lzRMyVu}JUxIOWQMhWxNq>$0+~IE&AHuwEwfw}5 zJ1vThhb#0{x2SN~{AKz|J(nbAQ5e(Hte=yH2`5VEcZA`W&&;=%IZVwQnG+mj-SJP` zEqYTtwpB+;|Di%#Epv=(4BeJ=fMHVFLCuDIKy;dsPCn7Lg*!*gERT@Zn=UgWD%{MZ zp0DKFk>7R#zsgBgWF+_POqP_k`)b1wb@}93Drhps3(f+A>dv0ntlu(l_-tbOUdxCb{o32W(>9dAFvcfEE5z@HPbIwpq3)Sb>I7NIj{ zE+sHClv7*pwfB8%Nj@8#n@R~g=A_5u63U^!q2&yW8JL(oHB?B^JSuEUj3`Na0-Hp4 zsj+un;DJWt$Jg7nq&j%?+!KdDXNuf$L>nn9h~ZkV=;1jzhS3}3iFe^`=01>WWG0c%lk_|5Ds z3w#dkrv6x)rKCP)CNDA!Jss%g^v&u`v=N6BUV4-iD^y(^NXoR7

GlLb;u!;w!Cupw&`d>~>;nd% z_kJ0IbKou0zxG8!I>${3*k^$S2*F^B3eREM?YzDom3w zI8D;zV`j%&lsl5=f1(NGWHliCp5>#tr|AIokH|Na%c8-<&louvS``epyTvNJiD)uv zX3(+}u-<}a7pBkSU!T17(r59e+|@L*c>}O1?PnHjp>6ocS`0_-cgVbyl8u!9`Puc< z+vF$TE&r%&=&`6oeBA)atz^%HoECqd@)tT}Uac>^m1(<^d=@am{INzFvd zZ$dvRc(?S=N@XmJqL0}jncJ+dS0e3bA3I`?o^<}MGWdY-aBr;@kB(to&m4=ytUF^E zQcxDM7A@l~t$^Lx6MuJ(I905jRpvMbMu2t>N4KU+yKCvGSoW=Tfb#H%r6Z2C${+J% z6u&H7)8jo#$-CQna+$-J`kvw#j*|16&lZv&ezK|GM$>kPlBVrR!|xM6s08#EiAMXqC=kv<8^o*EdTtIVwLbtJNLOAP~6D8m`cdcMQOU za$vZM4s1HgA`=^WEa`k?gPEDvYCxNh&G=P%G50dewBS+3P$V(6mEE`~OtY8vPldZx z;Pa4@=&0wh%vlICKBb6@$`EDj-gzaj0#k2dy3eC>FP_49u8+IZz&fn$av{D4#i07f ze5=U5!%pfEAV3Q6bSX5z?ySyV8d`&D3p?4my3+J;&bvB#k7i##PGGWUjGubYr}5Jn zy#5;?fm$H+t2$<$y}m zc~JP>Kerk+-4+h!_u_r=0}V9l{wHR}`X87XJL`YZsaLHj<+LH5&~s6<@K@xtNnw-8 zV6jt@qgnH+Wu!KqHRj4*&RT#>+CUlv2dQb}`_G5c`)Vk_`jT*UvWf<%h{(OfOc)Zm z*2CC4eJ54psWp-3fD-UM{1Z8);M2hMzJXlYNqM(p=ybpJG`Ao{E$ek0%|y9E7&CzR z`ENhWjaatoE`9O+(8FmcN00C4`)%#=z4kU9-<#C^III44>*+qs8&fyV@-bwrsihsx zuFGsmdUcYUadqmFafJ~fqcQzRb0~FU2I@Z>r@Y^8sF?V%x`Xd0A7v{HUAaG~)Jc+=kRH&sr0CQOJ)kVZM9FqJF!E z9K{Lydyb3c+t{hd>57H&gZm$1fxMdRuV9iWq`0Dcm~@YI^_kgz;Cb;)pJHRpaGI!L z$%68_B;zjJ;iMh(NdveMw^mv2%gjW0VCz9V^w7B4G?$&08_g)PY}Y^ZV6!f^RHd3= ztOQKlab}NI)ma?nG@G=8u}Jpk8_h7}cbY1B)2F&ImiA{&6^%N^D~t%Y+vYq09H`?5IIj~TkIVsfgp1T z`vfht#f0tU*XJ`p@ZhQ?*90EM_g2w8*)j(;9n3S(2lZVW0pAd-9vH+Z8FOH$)&yEy zGa(O(Kj6F)7(KiJ|KHAmyI@HIM0?FM)B6Jn) z9D22|7qLZiG=q`@SE#5SOU;@JChSz_b*wjeL#G2rgV@1>5_(N(hS{C1k+j~bTN#$l zqr@KOn_`6^JFPL_2;rL?yC!ZEZW;&pmFg%Mk4Q27{2_;Eu@FvU10U$3Cu zsPKE5U;|vKEV!z458%!g==*LXr~!rs9~T<`u*4dJEmJRZ-wy4XBJ~&ZLbdKfFZCwd zXfP&V0&e#l@kyw+4SXSPpmarb)L~S2LyjOn8AO1!lNLd^R$rD{)T`GW*_wodUQ>A1 ziQ!v@8#HZ7^ywY^6#1`pk+eM89qUQO{ELyR+Ht$Wyr4YQh|kGw!};`C_BML}e~lBf z*i&D!*yRy>Oy{f|Y-J&9hx_t5+ZjsJIIcPf%4cM6kzb&AAs=bdV55v39jqPY7~^9= zfRjFk#Dz?>$Dx0tkcPJ72zG1wR#+A#;;Hn7C9S`&aB*V%5qJ*-&9|b82_gkvs?_N& zn)#*M@S_{Z5*GscZaHyFlc5C+HLh|FnRWiri&;7s#1UTF&E$A z$O=QX>~kygu|o1hzb2|9{f=4_tLq^1DHcD2lB^rq9ngeHW4e_zE}Zm-Mv; z6rLimZ!2r_ zrb~lET(^z(7RgRpV~?OfCwFh{6~GwYOTZ5lrO2R|m&3 z9PGa3te^^~H$w1<`7?^WhE8VwRAGIV8Q1u!t~%l99q1g!@~iH5p)(RrY!tcdqpbeUaN@h~UU2U;6_!SWsPU)&ROU9`jaYiYtm zK|F)k-ZDsX)sm-oEL9AocrLK6E*b%4>xt{sdk7IFm7;H=8*kik_*o~1Mgs5qni0`@ z-PDN^{C9E>D5tr2^V#XyT8s|_;jToHAp(i8N$k;6(&s`}2Pj!_QlMNa^D4M~c>HSf zO8N-wHZh@oNtrtj{)p1fUlD~Hpw+(NGQexu)oMr{of}i^TL;8;5|~>yr?1G{5yHKt zI#tm9pB1#f4mE)gEcd#zEZVie6n--fuEv$MENHOM6|Rz^2lIODqYHc2V|E(0xwTie zxDit)^_H>sE*-77fg0XP0nMO1HHZ>}Ds_WkWAQ^s>#87|5imy<_{t0~#YAT4yeCKJJtt_YAFwYT6e!bS^2-lYa_J~kg~-DY#@wcl}{kaHk(Ew&HeJs3a%qy zOs$hlTO3YGMJrZD%d{}44-3A#($GlFyALg8m3W8KY<~qS62=P?t$&n(dXZ3?5LiBu z`O3HcXb>lS@HZT%oa1y5>gU9IDN}WJ@cE`#mKWL(yw!BBoehRgCmk~6g^tsXHnC03UjRU{WRoUhyZT_5>=<7Zd^Ueu>Ws6*+v8>t;p>eDx&-vG6)x z44(!RW!6`uhWXKcp*6tr8%GGBrvsTWuYbYrX}kNet6MEnXG*paYk_REgT8#@hyPi^ zKpMQJ+{Tk~@?pFQ5kTx1kwJ=hSH;YpTMDDdlb*P+H@&v{{iJ+K=YowLwz4zQa z#*^426;-dB&b43l#do*55u)6$G0he%vjMhMa8bGdK#IDBcV~|qSAdO!?G@mIn28dQ@<8n3|BrTTSo|CSjXZRM84$}gR?`B zef9kvbNP=W2u6o6A`YW*PK?4i>L~O^O4#U8gREK7ow3ulbOC(u6eajDNC(Bd&p_xM zB=3<$1#s~&DcJIOB8t#0VemI=Bn?q{m*=HOx>W>9Qv?fCIIcc-dloP^gPBtr&$<>u?`QZ2Z2C~X0lT?>GTfp7MD>G4% zm11Jtld4N|dCOuFIM-v0q3QxUwZ>n?2G|5+#5;T#KWaE(an|d*DavYxT5piSq1X^6 z=_6V(tP(@ZKKuC}3#;)U`1px@`DT0?^&J{Vr!(gCq0>Zpt@QN=m(L_z&&ld5Q0C<} zV89N{fGbFcWGKkV;H(PQB(H63$Tx`d!V>K|5xUXpG$-gGLdDt2S$2_yzOW_#dYv1> zgo5*tV^6`=dyp}G**e_7pE9IbUaK{;wwUEn8`I=1jy;*t8zhD$U!zJo!A z3MDWBKNB_o-(ML43@X9ij=;&ggae~G)?1@rOxw{tf6ebgOBmIQ!tgiX`2>qIr$))4 zE^q4!@zF6)TWQx#w!8#F%#H(+s1J^cQ*{S==~gcJq#L~bb-Lh1HAREC?m^da8O!Mo?feBsc{--}~J z)(+s#$H8^XFm%%%ae|H$=`u&TDpukPXqg!pkB|p=qcTb2g8%V;PFfv1AfOxCw_Gj9 zgC}RV6wkn`TW&>EwihQ%T6@rE>YjZzqwvkiUe}O&Y3FST;D^?_=JZF_*-OCY?^pF1 zh*Yw;@vHv5QK#qW02w3m@(`#v9^ooWZerRtAtQ!SW&rf-y#x-j7LJXk`0oNgRDf7|HKI#nPen{2{INzyWpbz&hho9uJCV;`p%N^|DPn<^yPS2bjW3L{YjE-i28}(c*Q9^pf*H&Csnv8Cd zl*taWg>u$J47h*fz>r0*Sur6PN{>*LszNlbUOWtOj>woWy?I}O*F33-x8leBoh2#b zrxJ}b%ZhXOON)2+#o#xYh@0n5>d%ww?{s&`?)}09+$eL)JP&-PG;USxYn$xMtscsk%4C>Yk~2G zZ7Ux5E8~JY*ZXQ)%c+p9T(c;0g#V7()rcE7fvdOHofg9lDEhH~BEj#zPHsgFr0rqlC+M`;;Fp zJgBm|C0qm&u55g0`Cg-OO3;XpRet0-49xbU1WT<>E$U^k+-w0FC$ZHW8tc<%fBc9@ zaa{c#`RhT^rnmog;JFGJee>c9)cGLh=8C}^>kl-6I&ZlA*;pu*Q$u=n<-p#3wLsql zrdU*+s7Nbho9m0YTIP^DAZ`=gn9Ky^`fZ&<6G)W;{BlX{GJ%X^(ntxToyw=9jIFvObGX-4A7D;C9D43zX ze7@c%P8La$gA1G;$eMNjYahA8a-kDs5Q;&Iu#qK(c(}>^Bal4?Aa&fy8pjDiX|AL7PZ-(Cb zhKXVt%su6Z)7Qz3LCW+d}nlJ4BYQ zNF@bV9Ss+7Hms`Iu!*}r6w?*&2l@31Ih2Y)FYOqjPS{-J`*dC^$NL{O8we`MZwGN> zUgiJ(yK+JXJwXJ|u=K=-CmN_b1BS}TF-(RLZ>F`a@J{ws+OH~n%2a!FBpsGqKZ!yz z=Y1h0?S0Zt$X=!O`JqL!aOwF7gsL6ar8U&1nt!U4L0D!l|E^1;xvbSjhNyKXX^J_( zkjI;Iv};{itmrNtz1x?gZc{+{#FOaisxnNXkW*??Q>c-Mwg*|^Mv-I{e_#Z{bB zj(Eh(_XT)W?H|S`%y%D7m0@w`qG5ihgc$UhH=bTY0$Ds$l~6bwEGj`yEL#7h+B{*g zjH<5}_TEmXhAYNPIx#-58s%%NdUs-z=*mX1grh&XSa|@2KCeSsBV+W48?l;1q#_&V zFxsP7(3O8MF9^W9oZRjtwdUqI7>m}21V6@hC(SwFRHElAA>F^NoxR19DF-Ig((5M3 z-T-j(!p0u8;Zsm>to#F|(nT$7i$9348#AyU65j(Z5u_wXBSKSh4B0pE5BQP_3SEX0 zs-N8FlufQaj>FyBtCvO9Sb1L*w)f8)s7v^6B3%@Al&9clh=(OQuf~u0*Y6rF1i;Qx zmbE*nv)K5Kf{OJJ#t-yIHr8Q+?^dFaxo@ZGbv zo!~*ywd@U99BpsaXTxOR73NOFF`{rk1Yu0glmL(j)#@GZanF>+(m_6S3y<-W zm#bA!Le__SXe6m`d^WPC>z^IBJ$PX{$wAk%mZ$Y5!DPIq!Aq|Aye~ zblUY%uc$xduUU;6mT>VtIBuerL*}mZ$PF5=)lQ30>ZG}42j zgV(4f@s$Zn(C3g1uSVIH>`sE~`itMVo2@<}y(j9V!*+Bi&5iP$Qyh>4tn&^f@X0)f)h*65^)RQ}>$D zj=Oy3C3KUnY$+%i)oQ}oIhXPz+){L~i} z&+f4q-uZVlnX~fF0(>gS7?-I2Zos&~ofVNs&>#+G%>!Slh_KfJ$Ggf%kblMe+(iL%CrjM&y@G)Opo zq{n+O0;r_e|m-AP9Hxa{J zO7=mcOZPh5YZ~Iz{D@1aFV*x7<22qTk9QK+&p17$4bkBQz=)I|s}Cz#ME9tNVP5Du z9HKJ}%0wijtNeoUQh}|yh%hAV+DwB8^w&GDSP9A$;LcnTm}0Gy8f*cHs*n%|sVqX;757YGN0@9*T^I$9ZF#aM(H-nk@TvYBf?=!aKk57wx+yU%eWny<`GE z1~em)Cw_nQI_ep>k4}-HslH3?EdPGQxKQ6|kF-)TNT19)-I@5RfDNn?k8+OV*DwcR zGl9kPmxhW_KCP}}vQr_w{jM4z>@-ecEOw)W)%-tnJ1T<;soOpnz=T5P$TX2v2*y5T zd@*aF5{*oTG#8=Q$JM@5EQ3kO8Jaf&9jK*u2gfp(YQ)q%7GcuQK7O%%+xZ6!)|Rf< zC3GE^4(qj0VoNg220F>Z00A=eD2W+8P8bl$HIvKU{nQzhDlpnXIyB>$YXEA;1NK%Y z!^fvaEYmabEhN%3glNzCj|C#ZpVt3|(GB8EsO~@RheJutOrrw3-AJN+@SO$U%EwqQ z$yF^)CWG9q*SG9EiqjR&myp&mX>l}y(fju_pn89rGG>Eb=`aAb4G&4NB3YyEd=Ed1 zNG}r?uEbx;n#_rKj%iz^7m^JMGJEC;l`1j0VDHo(ez}Xc(C>Z#>&+$FJE50BMllK7cpkw-Olk%=!Vm>5?AuD(rB$v-PHf|Fjw zH+5Sh;_L6-WF*u(PI3)OMhCI9ykLyin^Y_?gD2Xczq=@_#b#-cY58idZX865;#ecc z8bn*b12Z;lnmv}6SkSOv2nw|Rp!%`+mbk@ztbygrxE>Nemq1Pb2yd6qx0HARZ@r3UvV+QD{ z=^{{_0Yi!pfS-5H*HYp0G5Z+gAF{n-3n+HRcw@$Aa2Csnm+MLjs>9*AVt5_JzxM*P z#<^Zvsxn&0sHCg_yuh+DtS}wn0!Wh@EnH#oixRyLh%Q>%g}M-Ea;CSnt=^+u*{@b^ z%SOlVU2lbYG(K^6(;EpWf`|n)mD8b~;Qn z(vOk9gP_pWMbmYZ50Vm0jrS-~)Ra2i=P*49`)x!6Iu7jra`STmR}qS{J;qh>P=A=t zQ@%T$H-{h(MrLdJM80!UeOQh(GJsiskL$n@%tdXl%#AuVTi#W&nICBxw$YT)9yLYI%R49UT?0Z zyphpIF!|0vl`8-#d@C_SEF*^!8(7CuLg9~+7MIb+kHb3i%Ru67`Zg~|jKvQmCOovN zyzL_6jXYtE{~fU z;Xv-D(hf^jg-1DuesfWy>@Rw9dCWr*Y&IC7>m|LUlDV@O_s=G` zxLv4G9YeSAGN{-uuKB|3qrrMZNPcWJ!I-k)B8sbp;_OH+3dz!Sj1r5QS)B|S!x7Y4!ka<_B!t9y-J!?t^Ak+=moRe4Qw+FvGdzumo~{s0cF zCybl&4bFEv8Y*4+IFJ9i5KhQIgG*#H2|4ztNw?MYUs7*_z2K9 z3!v7jEXGa}-@>hMwBNpnsbPbM0w)z*At%%e`6dMUpSl}6XE!gUkQ- zNalTY(q=AMxg&3KN92P0Jeq^|VI7tWf4K$AsM|5WwQ3+b5P>vP=qOer6_v z!VtR?XF&ahL`sZGwX@M(e?US{+?w70pn0~1oDs;b@br{wEcy45l?4g9j!TndBC6R7 zZd~v$5;1VXrI*lySIE)tN#r;av^?l-6U$Jkox-k~sQ2QHEHMJPq$*Qkir0IU0X*?I zHJk8rd>EhRAF_?;!(4{11$3Et*bRYaf{-$=69<;U`8H$N4fFJHYMgvqM2Y5JMJ>^H z9F9p$1xbr=yNJci(!d>=d@fS2G8Wg!+tRJy%M|H=8JQ7ky z)qOc^EV#OIZrqObtc=0Q$j2$)V)5?5I_Fu_X6vJS%V^RC=-V$Q4nzQl0 zY@*(u>!~POc%LYaDuWLNXOq(O###K!c{}5#56w`1p9!j|w!X#r#-mJ=4l_I6t@$E4 z5`dh?$81uDQIKL?6B=f_a_W>n3t(gpfd&owxAORg^kFIZZQvnl5*ACTF_4IM(SdT+ zi@+VC1-nOPrBcy+T~F-K<*GGcp?ry@5Evc#AE~G&df8b?KTMaOegA+@v=)PVPbTFd z*~yk)L`TG;qOm9O%dRj(o1zuO7LUB^Ya)DMCMkiF<8(BQGE_N^=SM8at(PgvSTrz3 z@=*{>f50wf{5LrW*$R;hVOfS#c&klH&T9E#ptT}f3!zkI{F2nIUu7TTER}~1aV#CS zXpNSM4IC?`SO9N8JlA;hjY|6LE#FN;gms1-fSX#ndWzZU@XpT}AW>B0UZeB^+f{=L zxUQM_(XAxD*Z%D(rEY@=(+(Owgx8aMnxLkJNFUq*olZ}#p-{T2TVMBE0~_6`=8tOX zVww1L44}vTPq~EX8D+;s6;yC}xeE)r3KH*gR=pj?Rc0eDDx6r)JRa(xAUTxxSHHbaT1F42zfW(msbnfG+T zNb7Pz%$Ry6M(yO5x`xUY1G=>k9MH27|Eit1Xn=VhN#LOYZ_csoqT!ap19G#`6-Y8? zB-^{j(RSq=jUp$zAhc3lQ5B3?t6(}Vv>1?CrCG=8o@OxZAWBDMQ9Y4D*C}UG9ULI^ zmIT>aeF8QwuVcjVLU?Iio=fuX^E$qNiv9#BWTYnb zxfR!Z&G&+z;49%6+vczrOJ51`xs+X2rM|r#ERlno`F%F4$NGj13otyP92-aY=L}*Wd8pzu=upd z7+XyG8SzcsOnh?ket1P5ca=ZnpS(U#tV|TIZRUCopH*@3czz-2B9wQ`rKnvhpY-}C zu;|{%2|KOL^?m;GvVVr7Zy)E@B6H8->~d}9LX;l>uqGpJ7}83;U^s{_Lz^JwdzI3m z|8aTVu0Gc}C{nerYb!XnGHRYUK2MihQ1O_Q&v}-w^B@0aKiEY#)B1bgu|+%cCfsiR zqPp6IZ*RQ-e%Jxe=WFCPOFMUZ-UsDrWS9GO#Bhe*#^S_2+nibuvnWB|V&b|;L!+`+|FfLKwl^L%@N-Ti z+k9ptRYlrED%$V3PBS@d(^{U=B_*MSqPx8GIoj3F49Qxtl$EnjYwNwXUYXJZ21s2s zmJp6fnKHUb6FRA?T3b(+iH${((YoI{3rkDM=hu|Gs=L1O^=u^EYG`m5R4V)30>d;# zn9-mYB443%H)#y1Ac?!%S96Qzg+Z%#qmfE;5$@EwT>JHGUlJ+U&=lJSsUd@4wH6;8 zKx$tfr{m-Cd~kFDZ@VWgSVuy=EWK4c0gcA!WC&aPL-P|P%cCiA|8*>?fhv}>NUR0W zMbB!S6=Pf)oMfg13$IVtdBO!TPN~>(518U{*^GV*x0>lHs(Z(a6pEIbc_OT-E^51| z?5lD_x9}n+YimMjn@J`o#?iNvWMbDmlGEZkNJaL!v>n}Fw3FSzLhk!Vqx-yL`{%w; zd{y=l4-Q;`dr^vX;{A`K4*Rr+GTEvL#(Syd^QdU;;!`dC%jw(W)7Mw$xYn)UwC!pMjR1MtqP?bP+0`!k`C4aRa1O_9~#cQw?*y<)o1YZo1#3rmQ}F!N53y z=yy(zff&9Zjn2_Q1NAP_!96R@(=Hlk)UPi%$HL<9pLqwB?N-^>=K;@!HZ<#cT73E$ z2iy|Q(&Z8>nBA1V@o$t@gMmTiHaXUNWNZY8#|NI2({J~DLys9?^4-~tB(`w`v|WK1 zv-W2kgPc`BXO@7<`t|yZ&uVIV%&o(fb1=zOFhN4|I~9FYXwgJ1dOnrPw-l8FQ6&Ab z-pGsy*nqWwE`VV>nnM1N`k!^H# zv3@cOyjY_MXfz>oMQt4xEugc8grQOXZRVZ&vpGqs?dz^4s}kdh;z;X{0)@qZzmOd*flojqjUii3z0-Au2b{ButgfS8?Vg3kwiCBy%F+ zX|m}WGaB2vhQ_q;{b|O_oDeu8Q*V7ZRyCcA_ zw8V4D2H8#iemRTQ>y|EHy#kF}OJt*oz2=}`j;kjvBp1EURm5IL2(D}k ziv&t(rR6#yiVtB3FM2MEn!Z<-3+1gDl)aKf*|e@RLDe~JFxD7ohAsMwmM;S)#3Ncp zph@b52U!x*Wo#O|FL4e&0!z`aBJ7Js`h~9LZ;$!7&q)nIBl*m@>HyH=w}tVApk^~6 ziy4j_WCvK$(KQIyCD~B5`%XOof{JlP9|nV~@17h#t1S_(;Hi&B+a#|o=GGYwpI6c< zD~%qr&&-?nja*RGn^f6YanhOvtWq`UL!;ISB4c+02I|vCulR%H4Vgprj987{SM7|w z{6(o|OrWQAF>4U)0=)fJ!`UQrgh@p4*8>|7Xox=!2II)eps>s;D$5@t`G=$-7cExz z0-uBSvy2Fm!We(wxCZIZ6|I!%BE2Z2T4y_|1d&u*oDE2l9idX4w}R5S=O={n(d&k| z+zErk_eFo`(TeU<=ijS0?Q>JsD_J)ncr4HoBWQ6!8H(Uysy#^SWdE)8UEK|@u3b2=kEM4^&SD9QY;r^-X z{1=*QNqxL~US)nr$Y^N`Jc(itqxg^T5#P`+U``t#PR*(7EV$(mRb@f~Ql;Y?#m2vr z^9pfD(sqA2hHS?uAtaeypQZ(cf0EYBsVu(Ih2)a7yM(Bh5{5f^BFP5&yed5Sg-3=||7jcQYsqwK%z|XV+@p{?2VL0p`ajIvQDHRNJ@3~w zsX%Aq0(SmFR_jjeJp_9Z3>lLYj^GX7jR6V+Jz20A+Rcdr)N+7LR?$h%{#`Nb;0<1mT<@fiZEdFM-NkPAb7hD+{6V!o#-&7qM;z!ha*@TH$BO|WO0 zhQZ^5g($RqN4|-bA-BhcJX`kvQqhl+f4fCQA1IlPseA1HDaTGm8w7;KD1a(6aa>>b@wYZ-RO)Sl>i(7K+ zGv~<>GPsen>g+=ak8xSq;=UoQjh12tF6%>Ks5Ri3RSR#%4f%RcH%?`Jdt(5bULKhfT1)Dns=U5e(6rH1rEfFkvXmRS zOiEJf1AO`LpiqGk))V5YIcd$DO?q^cY%Gy%Ym z9pM0^TjUm0?V@TQ*wpJldv2uSrW)|lKU!FcQ52ymZ?`V4`MD0jket`rKpnlXCTq#e z)Q(gIm#3rAM?e~$8xaS#va&I?t!P9!BAMxaK|bjL5`>8 zlNhctP?rk*7{I$3M z!(lPz*uNCV9&WQ94rXhiq&e;JASG&HEd=h;5TQAglbm143QMCgBL(T4aL7sCU|Gr` ztYsFns&LkHwWlFEYI>0Tdrt$myFC}DFxRtuw$Gl-XZd7sP$&jJTf{l+{qLFtF9`LG zSJE^%o7|3+YRT{XqR^U>Shj5ir)^1I93#wpH#CR(%17atjZC$s?_P(VIvF84A8E3KCD>P#xo%hm@+c?3 zoU{s05q`*h;2?f79b{7Vqh9IYIBh~#hBs-m3QxSGn7_DNvE-%PvRM#<(a+JwUNUXp zLwD^*wpdh>OUyFSL(l&M!%x~IgD9_7|0yx{?`Q4VU+Y=B$`*|j<_WO8w5KvQmy}b0 zPr$-b7K07xpGQT0H)Yh#u*ihHj!$P@-9(Kx#eETm8y=8{UIp?40R?NTv!q;d-lfxn z&%s*f_3KR0UA4Eq+7Pml%08++qT|ok$i6ilKVvePgV2VHjsVN~=a=4Hb>NLnylc|X zqSVX_VRI9pk;$IarB*=1M}1XmcohEHt=JpdmLLT;{qBp(vR+_D6tr$n&_LA&I6h|vN+Zi5unp+~=M zS)HdcJJvzupfV^|@Xa<=^3LBG#T8^&A=rqvBkShX&6*&4B_SX41h1V^pJ~h!NB|*) z0~g7aI<<-5tj#|-np&_n$EE^>wlkoG>wk7q&xw!qo(q=wZQSuH3mu$~P1{K(v^Z0gCrfWDVCxkh3OY`>8I?~57;hP1b>iVtW zY4sR#s1{NcHs=4nQ8B1+EzYGd+}@=?F;++%NF(T(FTE}$=)W4d0XjerNSg^`JX^@OIbk%!vqi|%`AM?|0$Tq^S9(AHihN+=& z#^;!3JsZ9sCvbhL0F-DvS)N@a<0|+A`0E2bc%?=q3d3*i^{fOj{LqweL-YWqJS1s2 zqwo5k|DvX~LDmOWQptJo_sHXCT(o#cOf0LDz|Nr>mOv>sWjd^A*tfiXd~+f|*R zjm{^T0a@heb`>%1H^Da3m@0xAZbw^q>Co3l`B9-bdaBgtW9Bnef3Rqi*0m5n2K3Jw z->9I^KVIFiQA5&p`6mYrT|mf-DGmHtMt4BsX?7tS1|r9t#k(i}1B;(@>K`mVj8!`o z14vGy{|H0slKLzlVN`>sDM|K^Jym*5BH6!`jyaXVmL0nlWLLO`mfW8K9S6KqEawy9Y9{9JL7x%MK2(wuV4~sNX`?Wj=q%!L z!P+nyr%UMkV0BuV00n)I1<-}9U24)IlP}4CUSHx)a4yYpB zo~nuch|cl5NT!96ZL(>VVTiy_z}t40cgo||9c*^rH`n`)ra9xgT=}s2qYI;nU(rH~ z9}{|M9`72swaqZKX9qH9r55IwJIw1%e#2?X$lP|y>Dh+=Ds+A=H#Pa4YKWIyCQxL| z<=;h^m+@;GM?MW~Wyb^WNWTZW(|ZH?Q$+O>+-JrQ?X|dwQk(b#90E`isRLWWe{6Zv z)(Fme^ykpC3=n-X-$4fVrH47H)(YI`(wd_WYsqejDQ94+>W=l{OJ=riD<;d(kK zYN^m5?kgGJvs&Z6wp&QRJyW42JT%pZ!uO^Sx!7BAXu8PY)rUs(_bxTu=ZFLB%A1Wc zMCq(6=&M0)_xg$>GQP(P`^nJUF4;UjF9Pqr1T3RhKW5`OwVJyV8vxU;A6dOXWm}Et61}8go z|20%}L{x&SMHO|`OO6;W(7)NqXgUo;ll>Cx&ZTjCb}+fn@`yLtGPUiz{<~5^s1)w$@q{sLpbaJQ%*_#W+LX;(~zP$@}isZ}KYimjKiWQSzeeI^P=eKZt`5$=LSaGcXl%%o$ zrzDO2|3Yrl`tM4hjMZ6FU2g{JCD8T_Z~ zEB=-Hzd9hAbio!cGj2?s1_^}USSr0f+1=!Saoe(fhI<>A6@9umK8__Gwh!OTTDW}d zou94()kaO!&?-23KDRLYm?sV8K#J-wQ_kj(IBdbqjMb-sgAxAt&pnQx&( zUu#ap4@AxLrt?JOH7^}-wY*TKMcS*Lijh_mdv&|pSF&)R(fStJL)H9trZl*= zdMDU<+_G*}8DpyN(G_94<+9wO$6_v@TP^it>C=dkyBTA~k(>h7FsdAG(qNLd{t^Nf zPg))JW}-elugnbH%Al!e%{Fb+Jy2ojMdLeODF)9-$9LhOS!)FD*0RZRDe6^IRCZ#O zpya6=cER@Vtjf1iRTrp$jC6YlNE4M+clLmyXj>8;`$&WHbuQmXy`HR!#{8#4pYW%t z08#ce3zvcXKs^17!<+q$WBTr|x5mF3z2s3#UQ(p^3~~j$j#~-AhqNycC`3ayt+%a5`sj-% znwE&)7G4t5BoqNOW-7^Rj4RkO{A=#?K~g|!(T5@GmH;$P4Zn+*f#_=2%%a=MEd`+e z_UX7K6kj5JK~nYo`+nnW0Q`4B`Qe!Tx}>LDvbwqLYL{lnSC&WDFZ#l!Cv!SOv5ZP1 z^g#XV>Nf0Tg*k1TqkxH0PX*g(>JFK)TI+X;q8NK9^|X&*v53`3kA8F&ZSvOZ9v0$l zifC?=C7Rv32)6J(KcOoqRU=T8q+m;!i*=*E{I$+J9aRUewfj(Xw%LWyEx2=wH@+P) z74U(86uO2KSW_e`LIPi{AgxtKSi^q(1>^6<^|oVy+Hf!ZWewpkOGgmYJ~{+(?%0L* zFDp*f?>y5IrNFfI5LZC`PywBr%7Z=D$D(mkwQBr%0S*A#AQLUducUdN>gEHa7SFak z?vRGDr-{D>#U$4>mgg34%^X5M9u@c>0T>f|oirxu0Ea$xmRUpWFyQwqayFJezyaqU zKT19*lj2Mz4FCrRYADUID6`gXGT3-CUwoNI>+o{AY5@qa_tcaT7$Pco$ z?L*`q7J}C&2t|fxs5@_h^+B$#xdPI3k!nB0A`3`1*L&5w3r;?F zlWdp)Rvi_1j2r=d$ps2VD80}-8-Fbn~tuYlKu@x@^14a zA+y~`k&2*rRXix|ekhUb3LPNDt4yYJ zWu9tq&v-}9K(4M{IxL7&k_Yn~%wYwQN1ZI*e>jy=*=BX}W2^z|>q?zKcBd5H-1pdB z&e)FtI}G5kI3ioRUF0Hs@N9{bJrYU7H11b#Ry{OXW8Yv4(vo67WktjJg>cHh=nS)T zNDOc$;2Z?JtR+0q7w%~`pHZqbN?%!qo*y2YuF;k|-dU+W?v;w+PTwX$#6Y^jX<)s{(^;vBr|?@LMi zVPe2M{wy}GbC53EhN7?CRPEd>YBBf`rNYr5ahyHm;=tg14efK&t-WMm7*-q#`H*(2 zuk##fW$TMA6}U!G^*gX8UHSYNjoC-}cyOz`-roD$=78|9JzEb+z;{cbWhW!bb2HlT z>TRs@jU~ajS^D=*MII6oXabXqNe_Hw_JEa(SC`G#BO_%#JTlyjt1e&@gKh7%bDA}( zJKu#0o>GmCSF#Filt*Le@q-Sfgr(s)W^D}xkzs-Xhm%q{rIMYWZBn-LL{0E=$*}_3 zsP8Qrd>QTO((6_geVeZ)h{MP~dC#n9`2N6J#pmot`l+jYwz2tx>c*Sp*PCkJolp#W5KV zp?(R_ga$R$V-*$2x3-N;F5MFaj3WvAi@+1-C^*uI4U9UHe5GnxHWzJ8r)f{wXQ)_KE>HAW)FKB&n}W|DNqNhMm$M6g{sndJ zlC+rtE; zfc`r=&jkLpY@IbH7gkY4n?2AhP;?B#MedHRTFaYs#{FVdmU6(9vS5%Zt2#kdudYU+NK0C-jcv0E12x)h`c@(R z7d66S_`45p2a}J9k*Tpus!ii6_F=lTCSZhL2%SED+v;Pxw(pfFId@AEmkLXPDvS9( zFA;-q9QJYYvwmG*(znKO*uug0z1EG8wu>`;n-?CdqnmI%JH4pTfx&Q{$mExtA|g0w z0B)Nh*%amgA|dv=K(VSFv1LzUKyOh*euYXpa^p(N(;rS(A$TOHys^~=Y05jA<$DMy zd(ubXn9N{*1qn_th$GSbu ze({hyZ=~T>NKvk5k7hZgp-ijYy!C7=u_0_~aCwzB#H3TlJ}m@T2-J>Vkxoz=+t7+Y zj24iAn>*;$zv{L`;S&*@L2Na#^yVb6$@fc0r=~7ZI~kwFoU^u{8`ig$woU8kA8W3A zLa5ptRKPE!iBkf#f0;0e3MqBAR^7KSWzW+`63K&IC-x-2 z!gwfIWl+&=j)ysyU;gZ0fw}*2^Lt)5gP2C59BCuQGTt&pn5=@sI{4&P0@!B{E=~2V zHQju}fZQ5)n0FZ?RClU0-YN7Ry94%-Z=kOth*Oqw+zU`oMOWvK;lTfR8A)N&p z{ftnUadAFyvg-~V3uy68{JR@GjXH;AyD43AxwpGsmNdFH((WB@pn(aIKYOblKs^Y$ z`E_*l0Q%i9)y?0q@T}q|eB-saQ<4%!lgj?wyFAIcYd;n1`%Q?NEf{y+B?#m1VF^|! zFuG1b%DX^Jy-dic;9n(wX12+7Rf?KLe>8P_P|$Gj=`44@~Zv(I$ibtZ<0}pHK`8?Y`XT@D6v}92Nm(_SIi)Csun-N`WGC*Y8 zdYoktKB@Y-^AmV zO18-`>9nx5{UDT)aD3+kl3?xJhtl$J3rR|$LYYsI{#hE(io5qw_nObtj?vauo~PyC zvcrrnF{9V-k%^zVLGDbG0C$WQY4Y!0YP}LAfot_eWF{F&ziLP{|@sd_D*;nz{Nl)ZX)m4{UmdFyY|H425NK zE)%*d&V`lpRkX?Go5PAjf5ro8l`SJaFx2NXZKQ7pVO`bMasJ``J+iF|-z4mp;3^ z2$!3oitFwpE)1YSAt6A5=|a!+OLhr0KkOY6d7oJ!ZQ=Y}lZqy2J;p2z$*ahTlgN@T zW4h`97=lw(bA8ZYh9Vx^^=5356A-)W`(}EPHy{Au;*|gyQ!?gNMyy;UMr8)|g)lGV zGp1A(yhhX`^;G7#Np@;=jP+GO%l1MORcYv-@p(j`tv*z~`LP|JAFt^o7|1dWu z%!fiRF*UZcGrsrwX5{vjO@@=hb&xNXyEbFCiFgFI-pec$_1x)`1+BnjP0tiwb_QST zKEBP_V-V(rmqCK!lS{zdkATNu%@@h6v{wG7aOp5c#z?Egx^_2S1HJuEA%-RoV+7Kr zoebK~5VmJMDA6GN@(nYJL5DCMw@faynS_8j8&@d5KhM6dxvWIOosRyIvTBMls#d-H z(&Fa!^CSNi2*Oi-{e^~-CgT-iWajnDi@OG7U(ejhPJgc1?1g*x%!za2wdwlx%ldEK zTHPHxqqesx+`odGijzPXJdk#U8`-grWyh>w$^>uR+NHG%_wN@y4#vXDBB7;b+0@A{ z2Ypdg6zi40{qLD2#jP#T?@{9zPS{}lg`Gwj!s)GkE{tWtWslp%&~Ord%!;2xDbCUT zF9yKBh8CeWBTK`K6ZMzpxBY1J~*@B)IAYOfHQ~n0r=^~ufx-C>R$uN|CFe4{Le&8{_0F-m4+4a@U5a`Uzd0Aawwb>`LC_Jm^jsAyUAFn9!GH@4;q&s`)_hfRs6&*Zw z*PL}cVTzKy-Oaa;_lU$ed?G2sEtPa*V|@5QAghP-`{v2q;~v`SA&^gtjF1EVo2~|K zsC7#qQxG~;<5GE=b|=E2!Yd(ICVcX|moDcP_7s`EIwE=)6wVB}QZ+8|zY1xlDLq+D z`Ub#(ueQVdrOegMZpu|9%o!A69=bLcCvZ8k7Xow#=F3UYmZ?Qd5b1mu_$n1J)t5&H zd$-_NU};_ekbK;JsN^&#f00vVQ^47pHIbZNP3`S|^0G)JP}$C>%Cej@f@p*}dr@so zF0Ok^n4tqp79tP3G?)ImrF8YBLjHs_8y}praIATfU0CW)J+?$6yE_NNBdO63V?yKj zpk@~A&k)excVUl(l|~wAs;x@$AeMtysxrhZH}LirLUr^{82)%v)OR`@xIlTtnZTWx z+e{t^TOoCo?GE133=<=PvJ8THP+knXD6go-j0WV|=f59x*i1}MJ8ZkO7F3x_GGpE) zT^p0-S5Lc9O6Tl`T?2u2X3b>DGUbBaOEc~!*)zdD4&0l9CCe>U%C^(+pK!Q$u4fl2 z5=uIwJ<}UHr})ebkj`LOXxV3N!L=4-hf+p!)TCBkpX_TYF+de=Fe-dU5$P`924zXz z+kA-NNr~gzAq$zNU(1na_W6IP94O>1F)3CgN5S7G-Ob%6RaTrh^MrRN<2^^yjZ!0xd~s0|RR+uzA5Vhp*P&(WZ%&^oN*1IRlzfRlFzpWXf#OJf5ts zsbkh#UP#Z?DJuFMxE+Q9<(9KZ_q7;E&&eH1)fO1GAr2~lMfb%UQv9)4eWeNH^TQcu zYD?}g!`vE+H=+2WkjiI)$&?65;Yi)RYJf>a(XcQri-@n31zt_uXD<5q&zI)TdfVe+ z-=v3r`P@*(b-1=CE}4-%m_|9FE=5~&@R=vs|4^c{TJ6mN>d{^cYVI|1STAYcQ@$@C zF`1Pnzp2mu?x8$WpKeU)(jo%1)l3@afhaB$=r-M=!g4tu7>qo|wIn~sVK1$hHIM?6 zV^=Z!It~;2Z;|0#j>z9%AKuZmb4Trm#*zNx*dx(L6vadmZ~$fK{au#5b!(ms*1Cjw zuDK&)?*gD1FOO(lfoaPm*K0vQS$t3=9KgRID8Wm!_>+ZEaeI;vvH%~N$uUaJ|@%vu}H_|0SoX(1^;FeW+{OsCNMRW6GNUgVkK07b_R2R7!yyjd&B@#R$aGjaFPq%)9L zb)^rcyEV56k>I-O*|U_Ke8}8iz>h#6-37#+@wtq5PpyEkKwsn3Q3tL@-$KYuD|4af z=G664wq6IhUbTpx{xKqlB%2l&+8IO>Z&$a#WT7XrgsP8luFeW6^bMI4VBoCe)|sWn z9(2?u5Fev{VlBP`Ub(INr~93oKoeMCxg<2YbpP#4Tl z$h`W#S}=xvlt4>pBu)TniRLM=#9!4R1t_QS@qI5urnDy$;<)%|QDIE}UgN+-fxQI0pw|PU z07KcS6Fr)sLkhTHh+OV2b8f(7{&eM9n6K+v<;y#-zOehFa_*_e-Q$zPfN+ewS@`#Z zn6L_!vbaesV6n7gIyoTAH(h(lXilL+yJ$Yi`+}XVBiUasdu6o90&Cd8p)dl>!!}-B zY>a8WyByF63&<@d> z)aOa`Z#O^9+enP`yi3u5s2){N#kn^EMR+AVqW#M?Tmy0cd_`t?iD#z^;0=Ft#tr|+ z8cr_D*RO-UgI)rxip=R3F8#cttbAG==QJGUKXR_hJmHA=Z#k%ikT$Up!WFbO=|I8% z(8{Z*m{xNHDPce6WXsQPO=YqLE=WXJqM`P%6H+er(JMI`;SO5&$Q=#$ZD3h-Qy+@I zBsKFf;Yv||B|s7mvA%*P)<;>FFlPBiSg#RdMp>Ee({Y0az;xLN{>h^4oc>r9DB(M! z?8R@~iqWAZLRb$0as845LgVc*82InqJM4Lgr$GUS(%e-Wz@L2L@*m+5s#{#Q(N{Mqw4xVwwMY^XFWr5}D;E1B@12A@Tbmp*jL&u3w%&p&+-0u=xy013B+l`JpJfP}aEo zTPBlxo4O?QOIlUF3rB%4Om*^c|HBliOdW8v46B*JnD7&}Np0+=L4`$J_FthzcG!yQ zBM8#^=39vo5ZQwOEGsCBZ#?nlo#HR6-ZRc%4$VI|k%t0vDX_^Rf(yMT^34YU#JzSW zyBt+Z;BQ#pP>9_2L>WPRsjNUQ&hnP8JJW>#H6mnBcDd4p7*V^=wwt33k58`F!&1;| zG~yO3_xbg;gt`-I*<_*%xoW=xX`vNpSEqmcDl`uq(hxmXvB_m7V%n^}oJSq4eUJ{=g_CP$I zoNM%iDXMd7SweS0ND4NuT^5;d0&oW`S~|+SRE{M{{Nm`cBf-F^S$!*-6w=i>tXiG8*WH@>g1!VYdZP|$r zA+Gpmil7~85H4r~oY|o5N?;6+#i%b28hoB5Z~!0-5%>-fd8iF)$PfKj6Vb`vJ7B~2 zO2}{hoD>sHoYMCR?DZB<2zDb@dNZFOYQ~t6nm| z<6pm6O{X`wXhV9MAIdx|iE!sA= zDz+Ih6 z76!rJ8vZ$(BU=6#r!J$N%%7X#j_4RwV|!pfJD%E%Q1XdNRn`^euuVz!yRN_iRuG^( z8zznJb`g3H`IRSeQ6L5>P^9UXT^CUvU1cOew}8?%qer0E$gJo8jXEW= zc`Fdb$TER?`Z#U15ItY*$cj78Tnq`ha07iNS9Q>>fS(v50lkY*;X%SA!WZZa{yG|} zR=ygBeEx;h?po(|8x!t!u^k%nT^&WGY&PgJQRw+Vq*>u#c-Bn zt&d(7d~eL8>@EI%te1^|EUQH-R^%M%0*ov1PF8>7Yt)4&w=tjv1uf&;OO3PyRm}yG z5MwZ%a5>Fitq2nm4G>>eRZjM_q5d@=FD_l*O&Xh`d1s>y1WAo^q)do;fX=`rOo&u8 z10?TF2(k7w+8Z6WM0F)y!fO?wFmdrjmmDPI8x&o%5%NqvdJ%<}#1Sd?5@liHoGrN7 z6wMwTwsKZ41+M2hw-J@ZPt~lG0@tImVGU%8z6Jf1bjBE9HXB9~rIf0WbVQ63*1+E% zh8ANn08~iG4j@7yYUxkv+Re{`NwsL3`g!4fIiO6CO3wEr; z({zL&Q$n7}SD0T=KnN{#BXFjiXcEq$4N%p)*ED`zLhE5H?4y@3$Ti9xbsYF1_?oDU zlK2I`Tw-j2zGwV)kZRE^o(2=IH$)VV&xVw&Fe<2;FW+FcRl4ex&2C%^nc|qo?hZ1L z+X=YwCBA|hCr)6`rc!c{I{f_Q6NQz^K&gUA@}x@aDaU!C>l5V5FPNcfBr=zm5zZKf zfjiE`knw=@h!fC3F?tooU)KY3ut78$R?hYw(?coeykK@`mduk}Y1}ZLkBt9Jv;QJp zyNf!??NR3Fwu(c@7fXssmiB>Mc2=3qFrXu5*1ik<@ze?8RBL|+Xlcj&pTW1I;y4-;@9+w+MyBAvVV z@iHNN6@KWmGaSId_Hv^{mwC$ss_~{lyJ@e%pWlB=t|@O+XTJu~h1%-d=$<`j6|C2? z+zcCHQ}mYG#}&Y{j>yB-QCRreh@rc20fdJFq*K#;V8z40kfB7~KpCdH0@#^KfY}}5 zGC8&TO-1NrCos;!iXUdpo~xJzjLq6c5PF+)$A=Z-jR)dc@R0RcI-YhHipp?5ot%_= zU>~>PV;e!p>8B6q01GJ25%(H0>i_|z=uYD!=Y#o9T0eXalO2&^HEs3PQdF{D1f{$M zZSv14tiyY<@$^$(hBaGO3Z&ke)|)32h{_CbMqM@_^Q>mpJlp7aAFeXL!{SYmT27W{ z_Kn!?FhobR51PYccs#TnIi)_&7m>|f1SBWdz1{&(#re;h&nd4;HE1U;3eUyXhUoY1 zjIOE>Avxh>sGwqDL{j3$EImDWl&n2l_uo5G)``F_US^x_^I#7f_teKE)Dy zOa>F#k7PSkS`DpbJLF5AbW*f?? zg0lL%0|SK)CZ!tr6+Z&kwfI)50Z@I~`+)Q_)~KPI4>a^CfIxs9NO=mI@bK9TbY%8C z)zjCPd=mhCazn_Udax=Y)5G+M7%Y}pi&~+2vlU5(cdIl#$8<}f`u2AGMONi$Cdw{; z4~Xr@-peq4-_d6=hyf;Kwvx{RmWa#4>&{a0Q(vI>v!TawKdV&b=gD={q1!cUsC zo}p&R;rIRJJe02TarbF)`{<*M_LT&77ZWA~mAo!DqRA|(j2QM(EHB*!cE3<^TFGB_ zE7IfJJM{rTOer}?1_yM&g@U8I-<3vk^8Gxk-W)oT4Z;P3{ehzgznV3p-l6iVe`o8m z@o%Q|gS-?^cB7=$*r%E<@1%M7+(d}dDDK?~jV`*wjwYuGFaL_oWAwTb59r45WW(pC z5hShXoLb2;+F7yh8*3;R_XhTCSX?~8D{GROOzdFCCvMLfiX4tQ!XTKW8{I@ zX0KbPzOB3Aqk(i9w#I-Kla}33-dHrSVKT7hp3ncj{TL~W==)q=BcQ;}_~CVp-qd9V z5)00MHA=O>_f4jG3{b8sBAcFUiHOEqLMR#z?GH2%c{}D+qy)Q z^@NZAl>IOVk>Z_fy>QJ??~dlRE@Y7F7Ov6rU;G$IGOz1B)Am!Z20nN<4|0De0S`Ad zzy}}6?Kd~Mw7w7*aD5ov$yd{>{^H95q{~yCCdx#e>xU7v#yw_Z!tTy`D zo!po+P$v;PDeD}_T;Z92n}Bhg^VV7fmpy+a5G(Wyr%gB>GIqJS+DV$eSt>ulXaaBS zoxsnFl{YHEF0L_~wje31aI-4d>{WLY;Ef37yiRrvrO+ z`)l|#v5Mvi4s>k~#Uo`%Z6!{%zC9(r`|r^uVYg`IfK-{f#)bQAVl) z(5(==(#c>zF)>02N~ZT!boE{sdQ}egZ%^x>`t+l5#_@*Ggko5;U8@{UeB}TI9@$(D zb$M=7D6d8D13v>z|M<#Fe^&w2#!gf&&rz;q$kT5-6`*?jjYj)sjf+7(IrQVJwT8N5 zNpsJ9|CjDb6rEd+B&R$dFAeZym{IqjtM8{dm(b?{4tJ%bC8NjCEc`Ss8fWh)6sf_c zs*E^&V}Z}<3t5(g%RnYEZN}>d3Ti(|egWw;5IVd6RIr(|fkly;-=It+u6NI;5ngqr)V@Sdcp zTuGc|9dy~F;K$d2nmR74M!RSx@!VSUBKEe0jj%s@YK5|J3 zvuQ{lF}z=2rMEXijFntoqr46RUghR9@*3YXH8ubqRYGBa)tp;|U`E4)@d6{V6y zOjL`8!vcTr9>h;A|3*SZ6uG>x^EOFiZGOeDnSOyOS%Vu2r z6wRxNl7zfJFSVTXYr1;HvoQ^C3oeB~U-q-uu*IYWx99dd->)AyLcfGjNJ=KIzBTC7 zZQmxM!~(O$EU$>Zo^@7>Uz0fFg#GZoN1T#)x_Qp-F{z}U)W?_vE5tKB~?o( zt8Un)LSPw~hhMM)r&mLi%dimT0PASM=}7{L_hDLu6Port)`jp2BBlG4A8ntL%U&Im{O-`Z&;X^&Pi$> zv=R0kK0CV6T)|x^?R^7v#i6WWZSXhLx+?Pr4@+2HD%UK!m}q6hQn37>Q1eOOHM1r} zFYlWWUZSevArxv{Kd;Vym)8Jm`nUa&-M#5Bm(Q!MqSUu73YZ9~l~*UTHfnya|7u#* zoY8uQ`Gc4pWbsVLMOB4mt9MjA!HuFn@#hA}?w%MX2X9V2N}SoC+`W2u zC4L_${Q$UuC4U5DOD%SZc+l)%lo96ic~vlD5~pw70wGf+Q`9mQtckbV<71j7ov4%( z>Jpw{7!W*UqN@S~cn$k~f)hqb`tvl6AG78UG68gZg9s-hw#`C*JDX*cMWXgA6=@r^ zSQr9PBJ?z(4zGaZfHeIPzE=fF^=Olh-Wb8iAeB_MjC4t>aSh`mukNJ=qYzUAIz%NT zK{S$C*kP_UDFF%CQV|$~RaCSgmxyj%6lI4MhZo#?{j`nT}L6(a~y!4GeH64zUR=# z6oMxA&VB`vZH-q7ys-d>BRRy;jQPxh;rZdI?nBx&p6RDjtU zbQQKhN-^!jvnOaivq{+>Ul=_f@hytD9%+!Uw5+-LU=p)jMl#xXdc%nzCBk<(O`3HB z>;*FXzCmL<`g|ZAH{}c2hZx{my*JpF_u{EeuP)m?;R(U0j0+n^v<#iaReWfB2Yi@^ zHkC4h6?cBvzK%o#+Dn>(0X?|rn+aTR5N0wm$+sCJ8r5Ztky{DfkROlN2ak&8mO2(w zK8qt*iG9VCKnX;S8#!bP%C~bDpD&8r*7v;sDY=khPhR3N+7uXsrptsBmhA}!v}(m- z-GhW`E+&g9d;r%Vibf#-{j_+Hk#L4mba@NiBN4+cNKgDd*u`iecvQ^O5&$A;gV@dJ zP+vw+bYQ$F#@e(v=1EG9yZ7_x{%UM`2R7xhv7g>#%o)R)y5>g(TvsfOAbju=@ZjE> zp8_SyG*{~SG6kzSE@=+8$$Um027oe1ny*JSGp5A~96?gIEp20Y(epb9vX~4Sa7Zgcy*`z`vIPOPrtrd1K-{E=St<;U!21KCIUYdw4B? zzzC5`RVL3w&(%*u3$^&f&iH9L<-kmdr{ZZip^cSA7AN{`L@!4xMpauLd^N2+tW(cBbv5Cc3-)x&ag6S=&DA>d!&MZMA zt5SX+rY7vXRk5fFjJlZjq+)DOQh=O?w@FN*LHYHb;M9A11;A z<-o|l(Sc1$6LAEQIQKI8sO+#G3B%~;NNeHoE=S#&5p%BgsI^5t>jmgm5yYT zTI{27;MaY{OJ-U-@34s;%}5xCSE``8Nu!#6fO0C?(-$vvnFnIb7vxg&Kqw0}keU#z zARykbTar)KvAU)@eGvs2)PSZSFrdW_Bpi!?6~2m>(S+hw*)S(rdHd-;dMU2tQ^Z{9*k^^G7)w|Zh%x!xf>e+v0bsl>^iB|Q*fYM2#pmVraG&maa`J{-bnwo(+Aip>tryJq6T538NXl9VJOM6Crm{ZZw@!)0uTL`nO}9P`}Tpi(*)7&0cZ00>4Frg!XQR==i?pOG^3@s<%8?cO$Y0eR*L=f**z$G@ zShc+d!r$?&_y&9G!#!k$8;-Hq#vCX4Dzl9S(__wco@)&6wv1+w?Ci1~zwiBj0crh{8YX z#5kACZJnd1V7I{XnBjz(QI36ubR|e-e|7>j!Os5`+U8<31oE+L&LMe0FM~Y7h@<-1 z+c#pKeb(WIifG?EHLro6O%J%P4aK9I1fV02L_kJ8h*SP7qH5uWBW;0&MW0dBn8nQU zGx9%!@!v(xqNFix>iu`Rx_l(A|d`v=d2&Ml~yIowD;`gQeGY$xVRI;>E{Vavnh<)BZI4qE2a+Zw<5w@ z`oU!EtYIw((a*-`9RG6JZ4F|?tw3^z(PDtOu;^hfvEhdD_yPweFGm6j?Sa<`y&p8J zpTH9R6d$~-#%kPCo_TD@99=Y&k@0x}T(=i+dPm*(GFG~qd9 z%5}*zU+SgmLK9?|HXo^@FwH=s%>`21n7<+^2gvpck4VvWuG@LB9VEoPPf@KW$p`7w-8`a!P)LgW{S(sQ zg#yR?AOq!JOf@|d>FX8B84FT%}(IyJq$Gp#e^A#=~TDL6YV zk$kU!o+F~rLx)*Y`C=6me*Vk>0!4cnA+dTKM$!&Zr}VR!#WK7$Q8IC_#*iyA!9F;C zjYeHYy5Ziw3~fU{afl@eOr?u6Cm$(Wkd6Llf*7cg~*JBKe= z2D$9N7LV|TU~E8$ZwNM3W@Chu3{`~;AF6G(tz4d6=T5m8e9t9S*d8V3^p#o|`V(=d zQnmwL8fa7R?Pz56LmBDX>=NC`_n4L~#Oe{73Hf_t-o9%JBD|@3~4` zeW^{C4d}t^ac9CmPoe5T{T*JKibC3;nv(nP9BJOz;+mnvsrzb-N{1Yxy&xQyeA`rg zdK321<1dMpsNKqXL9EcyIRg#_0=1xk{$Sr6LBy2QpSK!S%U+=xQ7M!)P7}Dh4r2CI zW9FlRlq}ui1yLn3#|Wt;nY7aTinU-7iB9!d8Kgok*S;t zyutbj$~C}ekx()NL4pq>Km5UsCfVzyJduv^s&@2PEsVKycf9FLJwuCIoY23EsCbTq zRlvI22Mu^D$?i2v;F<&BD5li6b2rtl9f3tv9UJYiB z=&0esRpR%FJf} zhFp&9uqr7?Yz)&sv^fpy9|wVo%xR;356d@DeipT!<_@?$S}FyOvd6Bgm$9lmYb=^w~ef|0Hn{d)Ykb;3^l@OnyEX!iYdfsaV9k-|0pS5A*>gML`kysi6(5s#uouBl@#iLvhA17y;R74{ildD&O zI=VSNf31%{o~y%6h#4beX=a&Wy0Zx&mGy`eit{aurPIp!#I+V?SJFDIP&x8_y)4JQ zluj1PV54zI(`3$wk}Imo#U@h5mauHyciHzth~4Z)`OC&7lK#RN-hPY_Fad0*>TdfC zHN`5A0kXH)juSrs4b8$M;aQMqG}9Meey+ERLP5~v1cdz<;XOYbuQ{p}Q4u1U)e=HJ;U6skR+YpHbI0bRI9u801I zuxXF&q6t0O>9TL?r9;H_olfXp!%AJndc|R^zlz#X95s*N!z8`tjOODU4w^NNcy&*P zJ{t8n446h+J& z!?4uc4YwgaDxxOph*AoRek(<@5Z9f21$#t>yQS{hvMW6vY$@rU6tcSY^OM+LGvhol zCdfj<`q4aIM3;42q;MAGp=T`|l)3J>0Gd7}>Lj|wJ~isfnvIjF=T-D#EfSc^hZ=>t zda|$^RoTNU6d#qMx@aycw;wywvCo#(n!ynxQqVjH>xdrLjaHvk;C=$6->&c$o@0?= zFJiqb%u}=W$fMk6?SddoFOf1)OU%ZG<=2#x&Ff3eiOe6+V@tEa&E{CMd! z-L7ZqD=-48dRQrv^nUf#b9pP&;2`sKxY$j)a*Zz4+ocy4r;%I1Qxnmf$qgq1FgDSV zctl_AtjMm@scF#oN9VkBZo2jcg&aZQwqR61qOrZktX^!mR@_1lbVZUnSd~z*#F(g+hk@X)rROKEGhL zuL#Vh$LqJ&oC3~N26!t}XyAGkJU-a(&toR!V$$>Zyj48cbew=~_X|063-NUb2Rr}cWfX|X{ZioA82~7*>i0=?a?MGJ!$(L0uzVvE#rg50LP+0aMFMS zgMMM)|21`$CNCgm2$~in>sqs8D!KL`Zcu%u`~39n)=2_Cg6gEL z^wZTYBu>|J?{1)$O$_+jLd7qk^(ibj1NHbda4{5#F`LKdjR!Fl3HbpGQN^e^pjMne zn!`zPVC}+VEOPk+(vZnwuN|zvZWSegw+Dg^EGSf{zRiuek3)!otoDuSLDUK7e z9B_ORDo;T$Gknc46gTbJVksORcL^QlTD!d2dNfp93d(CDDgn)X3M#r##qeGqz3F%_ zRW$h?^imc~P-ODem8m1F?NmF1JS(FVDPUrmK}>QJ_br)vQBwr5Az;+?8YUT zM$gM&a!#13LAZn{Cn9MSvZGS0Mp6`S8ZHzVT}X@djGX?+5+;(? zkpZyeAx%`~{$nqIT594@bG&kFB=X{jW~FBpZtc13;B;+`33uZ}$2SiHUyv6qya9mf zj2!&p-}B5eUO2Dxf=ShL0$6;OnH>f%2wVew-tSqd5s^&Qi*0DDQUyC(5TK$6lR6AKaMMHr zoj&S?>|D%-U=Yo?vyE0A)UG=QP^6Nfi(2BgH zXtArr6D6HF&{#j$(f0_GpMKTdebNX}1kytk_Pzm!{7-V=hz$K>illoT6F{0U6$_go zE^&-v5G7P&4@9T}=1ARp+R<4yI$F;2*A%}dZ8c%msiHm$usZA*4UZRI4sj@V#YZV{ zn9yvyCa(`UMhZA*=p|c^8#lWe>NAIDtMLlI(~QCH2aRDpRVG(VE#{eGe_OPX1OkI4 zQ8T2F+O)`z@v)hv#kovLgJ;zaIZap0*$a$D+pmDNWeaZ!Ja#NkJTEOu4N(rMiZ-6r zz-#*VGIf{EAL2Z@U;46Jb>IDipQ^`-wC5j$IiS?a5`CX3+<4eH9O?O^9l2=s<=voz zp8PYO?yTgfMQcC#G|@|>L52jk@(6`rrl3Le7vxAB%V=qpE8D%i>IC^#p?P%fr*qFg zd2qOPcCi9vhjUMr#Z5qzL4vI#Yu6_lR_VHm0-}MC#ZJeLBrt8`+JWUO{@>{Gy=VIs zj>?Uh%vcQXblk>`C1v2UoOqJvK#@CF=F}Qo3oc#Dx9*L#rb+fUC=+ptPWiPvJE~s8 zhKNSVFI)k+QA-|Eys4x&txtxKNSnPnkrFb#mm#KIyEY10rb*g58T+~H5+}s{P zo^dcjb~#Oni*mt&tFnc03TnBJ*4Sw77mM#ANmp!t6BCwkbj)%JdB6Sz1vlFZh?PYA zAMELKawGk)Ux?q}m>=l=f=A>dx@E7~TL*dhz9H{;2!_u*|Gb6;q6aMDdk`2Cd_^oX zbY9{4=9Lc5E8dj6N_=~-%S)a+#m^=*W*>c6FssMYxJ{)K4G=ul1>FV25nB|pyx=MJ zYoxosEL7EZf6w(&q;BkSQjLH60kc@2r}hlZfT5SOP)@zmbK z(SaOuOc183576HtqN?jQh!Pe?B$nt*erqJ!fAF&&HP+gNn{is~EDgdHq2^oo5KlYC zF4}-<@RcAtoXTtIF53#=6u%UfX9-5U;H)JAt!$=P;|sZhZrwK95IJom9Q?hRJnn3q z5Q*b|5v+R6IDgu_#hFCC2=Xqmf%nd!kOU|)S4KjCHZt`lbD+^|k;3`^%}Ncj@brV{ z0RquAc70VkyLbjljcVh;1lW_?&gQ)$+K=z$EyfH*IpvM@1AN7o67H%6N}wa_fd2i% zPA;cSJ%}Q$6qBhe!f2Kb|IW(2C#;t;8It`3_X?Tn>@=y2M0kDWvW<7&wwrsV%sv>( z(eyCzVsf*Y9jL8lSv8ztK!K@zNu=GrxZh_(n{FEeU8!$1Y07Y66Xg~2xnhOIgp4eD z*S|5hCiA;6oAS#$+&K;2RLV5ndgbqnSr`d4po$l<(DIA!4|$~4Vd2rjEk&*5rDL)$ zXZ0aLeg>k23{ub@mD&l5Yb~O%<*}O(f)B6dndYG6p<2WO z8m7LQUA!Fh=pQOiYN0nh_Fi~Vvrpg&eus^&`giLvq>Y8a8^cpw&9MXz@n1dpC;PFd zW?2NJFII0}gWlhVGDr=NYf{=Y>{0Y185*&7juzyp>h7b!*9>2&74}C(vO#7BFrAqr zIj_;g$xTXl=7l{#U5We_u#(UaEGlZ8<_!>%L48tkZYFXfPM;lLpIpV*x&_3kR755I zW<-EK_6<1Z29?4eqfDVXi7b)&v&Fm88A%~o7UxDWR@uk(RIw^1m+J!Khp3YHB24aF z1$HO*#k2rznk<{`6XDXatx-#ngX|`5k|(d=CDwCa3VySflS_B6>Ol&w=SSHo4`_Nc zu@pdAA`0im!#0q6_3KII6OEH@iTZlvyxS+Og%_aD$n}XJm`%nSOn+wVKD>Q9#pe2! z@OYTpj77YRRD=p2+#qQCQTWz_(_Q@Dg5T4q9=Hq%179e2RgPJ6{_8f^M}C3>(m8>R z!s5%C%=TTLJ408|H{N_`^CMI`;lEE*8f8871Fs0cKtrnNO({P|6djdGNz|4}@d7KQ z4qW=f;+;{M02z}}>4?^H(x@LRoj@{Q@-iywFD(ymE4F$y^Di(wZA8x*Cxa8b|Vv`4PlI6+vtG z<2{Y9$wC#!)F??e5|d)UhB$fmmhX)9;nIu{LD@dzsp zv{73U4!q&&9E51Ae$sXP6WZvornOG!tyKSQa0@c6kqn)cm7SuB(9M^b%9^)($ED72UNyRN*yr&E{)1{L$(mGJ=2;ji#j zc^ZcLpf^9uQBC;R(o3B+Dczhk6R zGu=dZh(|i{OuI3$Dvl;x@3!emk(&7CchyW=mfgOU?b00zGQ>9j*LNoDBYgvCc(kZM zET=MCQ7(MaRR|o+Q?EOF>MW!{6mj&o16+FOI)~r6GncCl6K^U_%Fnsx@aF;OFhJ-Y zzP?+i8~_mz``(%HCOjHLIsNmilsRy0``_Oq?!=d;1E=zkKZuQgcmV&68ZiAAYQV|# ze==tNqXrU*J^zfEZvSRYhv88~yQ{h=Bv(9+!mAsWm{_c5fr$;F(V^>smF(Vj<)-ww zpL@#qY|PZqpeB;N2A>_Tf`#&P-(K%0tM3^;JV#!fKXHC?=%4Sus=D=`IlH&xeq=B{ z!ZEy)+?6grrf>|O8ThY5@qKpRWITwTYwj|Zync)tM&)*Lcm=pH|D{aj+~&mZr?`v@ zc86`^@WLRJ;I-0atg|NP4*06X2F!Lm`{~8~dp$huIkPlt0X9?g4xBlbRJgkYleDw# ztB=Q=zJc&?JH6)qc)v@MFLm2-e_IldX4Y!4UEzJIwF9@;2b*e7SLmiXk9vH&Jct|# z&v(Q!Jb!ml7zTXJ{^JJT^#`<2fQzZuui@O8KeO4FyHhVS1j#Ir6$9kIH>k`r%xAp+ zC4QcB2lIO4MzA5K+f5-Kq&3>Z`J^T%+?M+&)&RNI&!Y(}w|kn}OZt`AZia0WUFw&v z>(f=te^f3nTU$PZjDw16%*B5&9ogwN8;kfE6|XfEXN_QI>1(q9^CjQ8Id$LWPo{ci zipDw!ZX{h3U>cH3H+Q|si)b;G4rHlp1cT1>VC9QEYe|l6d)LN`^iitIOL)Jz(Z$Tw zUcZ=TS?YW-;XAEUiu&spp9*cNEP*Q*7m}=I#SHv+u%@G0G-B%UXvJC9ynect6ZR1agqi~# zgm3$-akq7a0YF`BsGRpsz$Of8_f9~*w&SzX`isaqdN_{efxF-4iinuw`4D-rK?iTz zn}E1rjg5dNgn19$=7i1Vf}@IXcj(oZ3*V7PaHQq@CiUW2uTej{b>^+yr@+h*4%=Ag zlvVhMV-0mgQR|tT+PBeXAy4>YaviH)QSAN<&6R*jLRYVUF*1PpkCEum>HV0wDg_V^tHsqpsGYrG)Dn3}=qt+|jL-I#{(8pK1q%(Dk^{ zzqL6lHoI7<ou1WtA<0Av#!?>{GXXtx3hP~*^WiWODA{s0-wzjA9R%9>oyL_%B zz+l{4=9Q6_rj(Q%u&+5b)8i*o6%3dY?zBaLl@TOEP-^oS)H^cp~8+E;I0hbqCx3X9H&`GJSapvrBX?aVs1)a|XdMj~|cv;*o zp-%|J#@5(T*|Tvr{BN25Oi1B`D}7fI?*&}jqg&3q`Lf(1S1e8B>fJlyG0lJgdCweS z2QKdAm0U|7o=SZgzRv>%p~bh@U{hYX$s|Op8bu%2=IBZ}tqG6#(1x$udMu zlx5nqq>D9SoF-qMkKM*PGL$ICTbC%NxC%(9EY6%pcV>0kLutBek^yM0`PNRf8Y}d* zK-s*9WBD$i?l)2<@uyZny0v>$^5t1me27zL;2Jic3 z`o{pa!kYz&{PG7?>DcZ;C@v)@YsfKkLc4=8^VQiydOmJ{^emEciZ&?0?C5(q^La0TsQgy|hshgUFVD4AFSVM#m;X_99Q ze?gUtTkfF<*{Y+tS~0Xl?#T_8Pig%?HMj`2y>EOvHKvMm;sDFp-r8ce>#;d zPy%(Fzra-5NIb~=WRazIiMo*~7;dk5i};OpNYGS~@t4G~vI-sDiK`qJnhDtL<=R%@ zeBX!v&Vs6YXCq-lB5lK4Su_gJ4Bf#Ve)UBLoXep=#Sy}s(e(U>jD{?!MGht*Q-C0c zsFMT(6a6m~6?W8`SMrJn@}4kuiqX;02FNj3Xl-qjrBJO@DAsqTOiXUjG75t~i9uI) zwk-OKjPWuFuM6a@L+(qosTZmf4baKpqyrSuNh0GoA(zl*km>EhD{Bm9s)@b`(ZsJdukW0 zXYpIQ`otb}TJ-rnVK~S`uv`4V?e;5D#6}Dhus1AcOah*DZXRPs7hdQ@JQ=L$I}nbC z`pgQ`yeN9$-YtX^NRSe{G=pz|NQ$)9E{-7as}F~q^N|5p=LwYWE_^s(AM z<+e!to|s*F6ZP7TWH&6#NW)dC0U^J%2~aHZ5lZtO|Gq#h0U`62yWW9ksr$~5WaQ97 ztVQ2x>k=l^Dc(^hEO9*b9z;aApJb(tXeDgM;Wn8ekU0I8((MNHE(oDsML{cZQa`u& zR0^tUFda*#0MiB%@S};(QB|Phh)_#$qobZAIs^L~!q`Q4iPqNZlvQgS?R1AXI7{xw z_*)jq7&bc@PCCxo@Jl?Fgn0uuXL|caICze#Y z)0#M~;gblV*}-41cXK@>V9!HXPDI=sea)U|-ar9x-qnW*5#ehNOg2O!U}TI#&0% zf&6{V2@-QRn(F!)c9jv51NFw#Pw;F_uRV@{o#hPQ;kW zf~SZwJhlIB5uP-;9@Y?7UK$JxO|lUvg!Kd`ILJ15s*z*&>D}6Waz8dIx?48A3Clu`mknU(PM+$J^!selE$lg5E29(S_m3hKm@aC9ZiE z^4=ex?(v1GM@qv4RuaWzIsHzoN|9)&286bt_0JBK5%*3`7F2r^Fund)i++2H_LT}r z!<+*8Y{3%@p(O?S>c?3V5E#~^@4BMV?}qQmvb-y>x8oTG#j))ahKBfSU9V^$V57v5 zzvyZ046aX_ZHL`Ti`yHDp&8Is)8UWXRsy+DhZW~45u+uMnhfKWhJc~O>xT)=mk4U&`=TBKLO(ucxQB*eG>`j@>K*b9Ty~Agy)6a&0V@2I?nLuo0OHzd=U6UgWY|6aY zFydYMz&iJJAv;1THK8$s$ttkJ$HpRngfJ^WuE6nrS$e1jTkir!fXC9%KG`a|7Yf}QU5HrqP)*gz-uouxubwW~8|ejgZCNXUnIidLykaJFduB?7#@rUQU&g0fS|Sgjm6M*O|_6Bm2z5 zsh&5ngT9Vfui1_~m)iVhyL*Qsn+Tye_7#K?T1tDRO+ubbnfy{;nSz4x{g7^QoFEWC zl@!OwlH?95%>0t7nt^a%X4}IjglWBxtLBgxm9^)Pu^!;@6sF7h^|Y>SWWUMfDHN2P zH)ajnu?MFP-HBKinBed$g3s$HlTTFM;GYqV9wxS7AO1>Ph^0vizU(UvYH-WfzdkVY z|4q+fDQ)#D^##(%#yQRT6C}>rZjrq9rSS#?GP{ z^w{fN%O>aj>KDaurc_ND6*sFAE%@*IorZ^#lQUAEGQ*4Wwp1pe$9?p)QcJgvp~YNS;TW>vpE;cvUflIA}rd{LVbG>^wO@tN4m@MLU3n4xtz;V8mdE;%BqEaf1Cq zMb#Y?6Qd2U}$}U{>o8v_19h_jMJ#Z)4Vu-oVs3vhPw_LixI{W9B)NQ_3)U z25V#R_>J*Q(A1us*!|?*#3idzZ>H@k`+=3KY9~m4d0JbuvP3`W+3)e~e4gK)4PK5WyI7>`zH5oqL<_##ar5kSCRll?~<3{K% zOtq5C{+fO`uGT3w<_$<})!(-~+oH>7PbRREO39;33g~?}kqa%!(LA3$e0rT2#9CdY zExFV0D+mdPJLj6qb9K!n{J=jwr$(#*mgQjI<{?gY}>YN+qP}v^vs7d zU*4*j|FCQCU01DpEx(arLdPeE_Z2J3OMt;k)5{04&7tU9 zucH%XP5virbgBB%$z}IB)CNTP2$n(lb+GEGsqCo zP+{@~#Y)2Fe95Oc~1>xcDj#i19UJ{2! zhz?&Vc;}Njj~iDZ0|Y5PB^YNmKP*patbq;K^l0hBVk>7o$UNcY*PKXyQT~@INeFf} zUu^J^<{w1}>dA$(A3UE6S%3hHP-^X zPKqJ5+>S85qy3N&q)**691R5W&w+@2b!;lIV+O&Df#`yExHJcm_B?^gcbu);!g`|! zUm2jG&h4W(0@PZgm3EOvZ_hp7wF=oA7{>8V;uufYJ9bo6u5lMv_ zg>yg~TAqA|OdGe2lo~R34^Lgxf%4~IxN}Nh3^w`RyiO`&(4q~Ypdh-479I2l;=MD? z108^WT4*^y_u@C7ZUFh?y}~B1m;0!o(0-FWJx>WBQ<@6>dFCW zBFg}a^m51N@nhC&_US0PgLN^m{q#PKcO3Eo4>XbeV48wBeS3IdUBlE3v+5@gB0sNA z;nI^V5dEmrkYNWK76%u#0GQ#HEL&_` z-315DHKhHx9(A*H%xzTvQqlkbyE;(2oQDd$o!P@ERlomK+g@5MFE=Lz1cYg}gB>Xn zAwZs+)9=?ixL1>A$|_auvHI>_DIVvqop{gV%u7Z>=qS3?it;Z^ZhuZ+k=`k?U`BE4 zt;&B&|H15RDih^i%1Y4hIhuJ67U^ze%=v2F`7|`GE)o+@v|mjFZ-E=-f$iEEf*{egecA0gtg9>TEBrl zRon6R2P2}$BTu48__#kJVvrvN8B;1G~=OyTiJ6p$sj=Bcgx!Z-L z6%P&LS40Rb@>fv9IDwT>?eO*9z3fWs6mRtI;x@n3{}zW$wuV09p2&v34J<38%Em~# zvWD(35d$x6ZViF!SZP!&0$jJo`6RREs?Sk2HnQTuLgOxS(%Zh?{>C4@g zs0Zz z9L^n_7Tq>#Fv4BN#0N2!xM5gtVz{_^fV_bhezed^a$Xbn5pc_tMHto3KR_HEBiL)s zmH#Kz_w;*}v0yR?A8MqEPBmX~pM*-w6mSLoYI#gYu?hhMuM6;Qy`AEcxwUXN1E=Nu zDT_1SVXyGmSWIAty&s3($oXKKS1b@xI#1eBYhmLL-GrFJ+rR z^n-}EOw0$JvAuR!eyc~}Nd+pWt-#eW=G|_A7ZWI~vP4V8qY})Uy93L|a^U*@f zJUF|%yDPLlWM|QivKp{hggHLsA>W?r24-^u=f~$q9c-i1cZ&9B<^UhSflLe(E>`}H z_&1-3({#zDqkxSw9#x9*Q|&T`egpj0F)&IP(S`XG8lQbVkR4o+f;-4XCUm|%@Dv&; z*|%?{j5@9!u%tUg$SE~2mWJ9+LU`|@UY(N_A58TtZxYJK1)L#_s2#GKc&IznV;_;8KNK_+#V14RU|JDk$AH~6H|?nt6f_t$jJ7LE`veA#Qud=xe35v zQ1u3t1X|^4l-1f#6Qm6?dO!O{l!oAWCuFt$^}*jE*}Z)h3xa?>ph{4R(oWJEQ3#2| z`gzwx+k(A-#rWkS2e}>OeY&3M#uDlqF`HQxyIX(5Fe5yjgs)|yM+b@dx^q1Qp7@_z7cgftK}?wvR& zE7{#!Lzu{P6osap&p=U-?tc=f8$jjSWU^Ink>rZ>S@u=(K;%TvaVIEu=gyk(L0)&|qsxo_wkY}M<&si$AblWIgeRcTJAF6-jOF`Oci(gXkjS@T z%fz^tXim6QSiVL=whTAwBXOt(+bMR8OacKQG^DaS;$9kH3?snb!xH00*j& z+OyK@8v-ZH15i+H)(aEHh8e{*GLhh+K&Ktfod|x#xNh(OLn;EiPnXZxU<=IE4{ywE zAv-p`=m2sVFbc|9bU|f!1;h~A)sk%Am&_4xsVxviGinq}Szs_V@@&JE+Bun{nfFL_iVL?_0=w2-M4POwLdkYga%+uc8Hjr*PHcpqjyj%O zW;S-(cZDd?Iw9DtOZN!lH`RUoO~dz~VLw#jQ~)a~YTs?dRP_k>G#mc4LBv$u7zg!& zpWJrn8ZWv@toMzz#e8*FNwR1{hdxWF2Mb^2XWEh}(1nXYlW^it&8GHP_Ka980s-=- zog&50$W>B7eoVHWcZO6DFJtF+O9uxGM4aP0)EyE9)$ z1mzjvV}nJPC!wW@*pb>-Z?1j59Q^n@Yn;D7e1o*nf6p$3otfM{y;=-kA5_!ZPM9l$ z*<%P_y5Qc#=GBQ8&>2R3R$|RltAr4u1l(R*gt_TIx9)WN3&oRSB$(58xq?Fb^*V}r~;eJi(U^aM!>mywi5~d6#^6N zeRuTzHtsSP0;4$WFUuX%eal$HcWNHk$-FljpsH?|6I3|QfOQ%rq!-`GX@$$XwK(3k z4s+V0N$vau39&(=g4S$La`Nk&zXQTu(m%Vi1ZAvWrV@7j_bs3j9{Ei_$!UBR5DY3i zno24x9u5QB_LF~0tsP3xztpg>U`~0m`w#n3MiS4RFh){6IAWFoHn71ns~V(`?7o_t zFn;?-BWWUtfL%nT=(H*d&&vK395hm1dh3N+pG0R=!Hp$J;xgxsvEm+d{Mp zaBNa??#iIurx7Q5?yId?>d}31Z&r|fy;&CQM}-{evl z-}OB5|B$Et%F;bi4F+l~98U0d)OR6}*s z@e~=-IvK&j-I@A%dgw9u;bO!8mc6(ZRye1x)-|+3ii9=r!!11)O+y^W>B;m_M_*js z3Wl&VcXUmTXx}T)@xN@9&n(in+<^cDZUuN1I=}mc^)5z~W(G*7?3jw8&1D3a5@OZZ z!?&Ur#`#tF^&gE1uSa2HuCIkMO_-GVKy3%UI>i^*nRCWNgfe@~Tuw*lMYmGmK$%+j z^B8N#@&4-g1=nRbH|e~&Y?LICVat90FyRvAyv8u$MsAy-!I~~PyKdCzcD4quTDm6C zrB6j*nW}2hRXD(}OVdK!L~VAO@wGa}xO)DiVy*pUqnx19s%ru7w4aeehEbFogXWOA z`|A4v`YaUl_)ile%YT><895mLFB9Vbr|Yf-Tj*_48PA~-ZR>wqPgt+nNFwp^Jr9D5 z2oVnx=7LLl9`1Cc@Lm2&9XKw41{h0pEL`38ftD-OOn%-^OwQkgniC@^e`0*=RU*s} zWCZRYkaqFG{?ILlGSDMrXIVO*qTD3%bawLZG8`IQqZ}o1PiD$DHif2KfY^KZydV6V z5btkp71`g#CP-&=z;|@$5e8wdbEir}r7NeDrug{7EQ;#~PjvrQtnJY2?cmT$Gn=cP z($d>=W@uD7xO1*Zrc2qD9g#No{=3Ti71lF({M*uAcYFGnRgr#Ow#mAK$Bn%#&=E z*A%J*I>X82{+})XaQlGE7m=WF@omS%@t-Z98RcJLYmW5@JUO*5WASg!_ZnMHd;S}? zZeMwX2^K*iXWPf1X=NtmS(Dq-Ognvtvt|_b`ssBrTWSmfmr-@%vBxNheWYp4t7N~f@$M)w+r{il-i^GG%AK` zlQ#b%GZH&a68k$s3n~L45QwWTv=R)4ZNma&{)rQh>&SFzlBBE-Qki%PwKZIFf4EZK z$pGMo1mFg6iYZC`hPFYrKnVe-22^=S8G!DAyaFLZ_dWzRjz;D(C#M|^y*gomHP^t6 z{bxw}pd~KpOjAle9XhEUBXM2%GR|tjQ)jA!B+KD)o3>x0IVxvtXyLVrVIzSwUKNdc zzVVwS4SJV`V6PDsgP<-BI-yiW*YM738^(qz7WOWuO0y1}ELx5Xe7$*duI86J>rk^r zMn1eL!t>b?9}t4HInKNk_Iy;+ys_KmY;=#H@LuckkfADD9n(_&F`pk4xU9F7_5S1Q zrR-^ug4R40TymN`X|4iieVk7hUC%DaemGjs)UND?cfO10h1-YLPMVm!kh0CtT%-za zer0oc1-m}Oxlu>OhR!GZLyTc`a;_nHb|ER-++r~u2VG1-6qZ2qNFZ{V*mJFb5wkoo?5`{)qp+x&Gv6L+NIVqx{lnn4dS`DzW zL0Fm1p2zj6Cp#I(q0tDee0~+`r0i(7xDA$sTEk4f?#CINjJS8vs;+SZ?r*Avr~1RM zP91%W)To594C-+o>mRRKK6h4YEmvBTtZC@Lqoh|chu`$d8^O4-+`)z$bSH*MhW1VT zDvpwac)%OoW?EtUx(Hh^Y}jG+_PSdbkkxJ7QdnhH4$$aYqK-X#cL@a`!QC4w#_1Qt zQbeItO|0`BsE9zTNWDUl=YnRkwgZV*KX&5)u%@Dq9J(#5pJRSNU4@Vz$U6y1w zR4FSmA9pDP079KQEe77TRM!U_&$7Ekt~$=*UqdrPCQB-p!+g9gq!B8n`;MQ8q)w>X z2GG%NZ58g+ppAN-1@vl5%$=bNd^9U`miZ+tUl*#*3o4Mx>_hgu|D^mL4`t^Rcl$az z-qjjJEcM?&Oyj4r%R}j3x4T5*@xLnCMWHa`BvF&lZ6lnuF!1t)vMu6=C3R@qgnOkN zGIbj--pB|tT-+dA`Nl$HL7qEH&(Yu_k&DKjCz~zs?ppZHOC|8^|k5hd|zFvq7>%H89gd{6+`_oCqAulwf zIcY*o=nTDks9W*r_IY;5S|UUiA=Ba+XpUTdsn$HOBy!1m)aRj5$mBNV?ezVekg0&L zFHWp~J$6nkiy49WP64P#QKu!7jqPfMNlxPP*%_$sY4ekk%D2`5J%jXe!stAfMof%N z|A{k!nO(qHt^rgbQ_R=cP9) z903)m&5Rb#yLVDSWJ4K(w|_K9D+k$_$(LyI=i+{<448c7GA6i2O|} zIEyJ3C;JkI(y5xfuV8s;@L~r_eI@ee4=9v1SC61A)X6+PA!}P0j65;#E(oW4{kdzY)AVb45OGL2wVrBpCkRZ z^_u)M5c^-CF^v&$O#|=gaDU1WQNO_+)UurL?ZAkLKTBOj(qj*c`EI@*Ca^sGEOL&w z>&KF;u(x&9{JY?q)5r|##uQtTOZZ%?QhIhm)+%wo{pwWNe(`a*>BwYp<)9h7GQjME zUM}_lbV+^wN?+Q{*1_=|6 zefF*`MRfaqc^UNi2!%KSAcxPEgX|ZPc4PrX&}ycYX(%Awy!gbwj6W+f!Otw+|1l>2 zpt(dl!W&#Hh&D)~)xyciY^M>V))_i(-^K3;sX?V*toxi2t)b-N#X!WE??F@msF@Ku z2yr>-P$0}pK131R1A;TtM(gV%|KF{?YVyJ=BooneO+^MDx-1q-A}FY&c!|L)s=2B~ zNv&ua?U>`Q;pD`xNo)JSNO^Z?c9H6wL;gO~d({2c{m(;6R?i8dS>-6{m(T6Ub}wRY14z9j67k?- zGAc?zQz$?`=(i#)O|LY6Trb6K6loQK=>ehPh0$+W*zuRLX2ho*gGd1)hsl}XvJp8y zM|{7Z9e+lNLVb51Gv>6zv_qBiR3?2!4o)9DDxgce2QRbjKP6@L+z2sLyHP(gm-2R? zM+>AAk8Y~|B)!`NVMJ4yme9jVxG2pIku+S2!+em~C&($AAXYEV$lL=a!&G>;*_Y2s zgjXGy68RIS%0ax@(?pLRfG_}~6aJF#OW5I8$erd>L zD7{u!c5D;7q{WcG2fNh5#Ki4VY+3ONN)&C_I0o5L4$kBt$$T7x4Rp7cowfyy^3Pjq zO?jII-z%$4V~eReB{upi`1y}m+SifsL(!#+@4<|8Zu;1d?|mw#P%~ItkPYi z1>Sf^1;^%609k3zUC+8V=kN5zmQT6unK^LVA^$vd4$6+eYj(Tr^Oj!0)J;?u4oHgF zuYzOECxSETvS8hj4vHbE#~AltJv?bPfGKGGZ?JEZaFER^mHZ}D!YXPHoSu<(0@ojXGCIqCkLGe1@hq#?54 zPX1nQWeBL5Yvx>vHAZ&Pi_-(WdQNpk@J1SSBg(z7l*!-K6#P?HIE#+MECWH`j{M0; zG2piwC`_Igj9@8}+v-uxt+F*6d9gy}0170>dc7GcBWv=Ssm(xZ0z# z^kp@F1teetDS!Y&XQJIuN%YO3_x8AFFE(x7`8R4v-CvdaA^=+P3>g{OU>q*%)NCe9F8@J zbgo?cnzNa_vwBT})$w3Mk*m_O=Go9Bb|`m(OIz4dER(W*bqf+#xL4nrmE5l3+NZUN{#&t1^M!@N#Vk=|Uq{6`TtH)q zW0i7@gE^fNJ{wrh0cVzd61eK0RcUWiU8|C>TznymIzbpc#lkfVHFd7X`0L(6TUW@YRn$`|w+_({M^-{uk#{N)3ZBdDgU-4<-p$H`<54c2~ z4-7Y0+~5m|*{IL$IIM|kq{>S^)>$^tvS|UOC>7Ku(**xCFJ?#-t_;mx%CEjYR^0si zH8(US52DQ$l#ui&*r%^>L>0;2vhr#hBUKH_?&SD{7@!6)Vqxd5$commVS2J2nJmAb z&g%5H8PN%EU9_vlHX&1<*JgEBb-ZH8D0Xf4@5pWt85Kjp1Ql=wFs!KxyafWU&~YJM zDwo!?@C{W=ctU$1dku9qCRLU+MfKaVg?l~t-}u4qw`=NVnT#KpV?C0c+bQK>RO?p2 zxle#u0ITg~EUk|VYEHKb3eFFPbsb^Sl7%5Ny` zRsb1?)i1O8+zIkvomMIx{GEP76EUYdT156|u8Z&DVbwVxttuERIjPtb2PZweru)I^_CdegyH0lR1w1iCQjGkfhFC(oO|(%Qwf*^o z9wOs@=~9L-k|Xjc0P3s*Ndg3A(yuOP_3oIoy*e}9S+zBYVi#I9o=qo5PK!~HBl77v zQn{dAb7EieYp{`8pBmYX6a}N>liX)bWYi2N?Sl-*N$p{G=Tlz?GO>ed2Jp~*5HY;f z?YlPw4nqVZW!uZ?EnUCD$mnO$j~ zANK5jV>B8TJzeFk!lsgkx-y1%Z16|Txzb0+0HYeB!YY{k%mn!NK;Wg*>vQ(Q zcT*Q(@{CFiul5B;rP}6x;PT1XWq6@d3lv?(4)&N`%;5fLRKM$OR;?T*+i=-E(hSJ3 z_x2udz^7o`aMXpio=?5moYtx1!)l|k(7km5du#=i$Ph9<`$6 ziR~El?tHBgC3u|A;+Dr6nF#9GrGk60=D$i7y#zs@!YKxQIt7P~qP?>72LkQE&SM9t z=FGe*ML85ie6wv;JY3d&u%2)8#fHL!Bl4YTe;bfWa>c~(_7!TM2{-C9#)T=p>|!sj zGNtv&=459^Cc6i~X}|TT8cB5v>D*z=VuMZ7-QvTjb)g~E7EjZxQ|Aam_= zB{~w~qdfxXpkGgMSKoJ;hXVHzFYgpi5g~V%%whDJi=7d**GQC%(r$->LVf0zF-XZ= zz{gJp#*tE!uM8!kv%nRCxrt?iW=VMf!{g|qHpnX&1=}5Gwu;yu%ZvfUb6VUO+=l=r?~u_ zE}s-&5Oo@n5jM7#dfkcNszE$yF2i#{c2>H4(P z0RQs}NU+cN(W8x(t@IDK*Z=ag?B#9%bcXpt<_F7%Qu%Pb>g7@r{KsLC^*;@R%>M^j zSC!U;%SLN*mS?s(!xZ#6+QMK2fZ36(tdoG_xvIQ-3~v_Mh>7fPd*Wf@m*ke-e9qhe z{$t~aH$Fg;tMS4`wW=~HmRJG+adO!;19TrFnLQ>M?YrO8u%yV}#```JSExsEm$PPZ z|MfOE22n*$1QSe6ev%XO3deOvq%2tWmqw_#Dl4-ldN2ZkA-j_^E8EvUWpfvgj^zdh z;wEb*X7J^|I+_v04a-OAT3)D6Ds8l{xr0NGHzy3X-qXf%ztM0a96vLsD%Eu5qMegZ zm^D1^S^4~-V10fleX{W_tA;`?Son&gJI$J(s?BTmwhzHpMKf*IJtVK{wLOR)LcgNM zlk)8O_GW3RV3dKB~qBOQ4}M?t&sg_-Ke>$gle{8P0-4~J)D za7VyUKIuD*b89CzZraL~V8ddl5aH+G{k-mdy~)SuJ79Mx?V~c>9WWMsIjGwvqkRLD z%tUeOJX9=);XOH3s`ydZIM5nkvfClVWEg|cHMI`MS$GL%>e#w2h2QkK>^m;(M-c~p zae^&?p1>&5ont+Z$``)QUb;TLGy9B0yjVNdRjO^g=52=8JlVwY(+s(H<+s2^y_9(H z*|rn;ZjoQWrszt;RnKceL~5CNY5m2pd{v@12g{a;F$fU+R~_IglC-3#w$M(|Hg2_w z$#Bu(`_oVuQ?HpBlgzN82@-8>ZzK9yj4(|-C(|+0vYEs@9sVBgv!aWU-Z_^bAWb58By(b=xwSlAxUp0?U-5ILaC;*18J2XP5^!0v-FZ3vD8n0N9>lzC#~ zoTl>x_vyh0$(Rm6I&v$BBH8iz{%Pqu4ym zL7MwDHK$^Vx#NgJ=sVA8s=)y*c_>E*7h$mH^nYquImJi)Rxp)OuwfHT~uwyRNHh{7R%z@iB6R}n-7W5r*aG3lmr{BJ0U#3 zSf9W=7#|(Ff|3hr{rM|SnXMJ^a9LX}Ck>`xxx$cnJkqbH;k75Ta&c zYomHgRiKSQ`BMAVs;w#S&+sGv<+UnQ%$R6W?$`wU8ld;famMqp;T+HYm~@7-CNt6C zyYbS(w2Eh5EqHIVxh-sX4*as9Ap|xHcQ{>z#AGi4xUKbKaG&TcxWUz)_nq*MJOfm3 zb{9?={HW1f>42`KV3Qtgs{KUokN~2HOXz_LiHn#cOj?mMTE^XaqNHdJRCy}jlBeQ}bGp)2?P9PKcFfk!pUD@~2eq}eA}EH4YYw3Z}z`kUuZuoRvO0Ri#V`psD! zPb8RuLECcOf0j~0rXVBX6-Bp5wk>{P4)R%8Bpx*d6g4=kGI$Qw0W(nRKjR()c6CxO z1XF)xG(aj${DAbEUV%VwS;C62U(J7LM2SPkHfX!fhBD4gblE?+hk0dSfRpw)TzhvP zx{uCws|Yx?1I>OupaE6wfk(zB5z!-ZOvE5v-82l(xJ{UG8~nfj6}IZI1a^v}fWY&J zO_scN)Xwry!bHpVnl*LzAoN9cD)F(ZXRqmlDj@YN;_Fe8CKd@n<{6G>eG3EUC`k zLE@p>(Pf_O-4z4kwDz^m%!fYEC|4Rhq-o&=zYYG?IVuEZLsKV~!?N%-0zV*8D@;2q zW>hL^iGRFhE@#d7A7P~L-~n7c0ey`B-VDgU(gVE0Q#o~CsO4~qZ(K3-sKGOUVc}uo zvd1KFHW3p|2Bvo7mwOSVyJAK)axI9U469mk^_MMzS(%#A89j}sVt0}R!(@PmdVv{U zN_PoGg6o13dts$CzRTWtO^D)CtF;bRBI5+=f?^x+jJOI+gm^H=s-E08E=NQ7W2Ix^ zx=Qx=qD-cn%$BFhvxi*FWSccc#|&-9xAK4cyH2k%k2f6MGCgKv+QOEA+$M0>MlGNs zOMakIeq|~M^3S^Si{=>^z}3G(VxiPiDrrw+3A{cTi&oA0kF8`9$GGV(5%MZnh-9c= z9qryVl!Ix`m6ctTjj&M{h6zzeIqCDNNYN9YW(kdLhURi=W^Mrm06Qc1Hq&L7&ihLV= zOt>)@!Fd(yl{CO{lMj=;E=T5$n;p_H=vuXSPKq{?LgA1x6oF$qtb**|fSNO=o^%>D zg1)*KnMp?VZk6k_l|EZ{ZRN>Fx1n*i!#(4#DgOy(s?R_?M_0maw~K7e`G@@8W`4X2 zCE)Nq>KBALg!8u%jm@1|-yZYMBn-O5cAWSOI$nBb8!g6q`b&#G;u}GG-10_55dwAq zS(oFlhze{b%V9#_GTOUYy&60}jhLJz%Br0X3tGbSG3T8}ANX<;$1Esl*}-peK@xm|IE| zak8$nv6&c!BP-_l-{kzxgsO8eFxY*wPzG4nh-=%%x*I@izvBnu$It`ifO?uILh8k& z?LWweu3}9yo|@B>sM`Hgw;kigflF^N89C?mOtwOO{oDhV=B1$xG91WrkM)W(xdOK> zjLwMRo_mvhD_l9GO-xW}a&>f*CN9jS#>bz^iOQ7_=T=rF=7E2e(c?R$?xtcIp2h_? z8n;l|_lnAOI&_!PfRmmJ>`Z(~txsYu8I!3C#3Ul?-^qwalH77;`-}Ul$I;%0EMIO3 z&=%2h`rPWz%kOj!%x981Lm-unETbD@sf6JMwFo_dp`Ta4#C4*Phuszl(R3Q%IY2mT zy_Um(oDWxyBRgM6Y{eH*`?NwX8WmdL*OQbNTmC&ev1p(`kVN6mSVB7|4e&)l>$M&U zM%vJhlA*6>m;y#$KZS^89i$3)AU*<~Y2LyS_PAnxoA~3cqChWlA`k(NEKvQl+&;GhwGrlE?mCcc%i!dk#+TSXV7wg zgjN?=KF9Z^TzAIxp!BcNEZ$URby}SL8~g?2W5pwcsK^x-`&wR$Yhwww9bYCLUW$io}uu*#orGfah zT+UD8bNldR4bVX?iPD_jRfe% zxhr@0{IA!HN%?wPgrQWuYeKE6^>4qMN6F^v2Q_b~W1+QYSPeDxlnOJ{Y5D^hk7FD8JX0ib%Pn!@&mpKKDLL-qTNBh{_iL26WjB~AJ{ z*s9N<1jKb<-q+8Ra_Su#c0wWeW$p6-JU~=P;4ls)<{6`k zLD9RPZ-Z{ji;HX%PY#j z^kTT@5Se<)mviUQNzsSaewxuFVx8ZE7Q0jim5yRS6tN6yEYpS9a2|+gH7T>sV_QIZ z$rZ3+BPs9ceC~?n4?jZU-O6=G69<(82@wn&U`q*0Woidb?ZAT9IOh7;q!}uk>(t7w z4IOgaFP46d9rCJd41rNm&}-lFr(DF_v|h$v{|RU%Lcd~?Xh%EV$WmHrZ2ZZA-gVR;#!kFC^WRpbM zEL_nzb0H55i-yq@o#K@X4uK>s>+OhsHUyOLq(|@8CFa2(8%O*;I`h+rJi!6oQKIm8 z(%yoT*T=L;EEsMqeT@vY0hlCKj1zmD_&2~t+aesN^T|c+n)qB;!|TY&vi8hFfVM2O zg@NGyr4BwKd0fdg*zNA|Ou6&m{`vzw@~OnyfRKNkSBAAWr`}0FccU-Xozp-f6qB|e zXsW=T1)Ji}uQq6WobT=}SS;4ZBXVqH6u#97zuQq>YqxFkRXc`DOksEFK|J9!_fz{& zmwduty@K5F{Q{9;X%d8v7TVgzCY!$ zFNT_*GbE}MF0ddraRXPNq#vhK}PZA-S<-EZJ^7^D)ceUWMvHd_# zrooY+pi||4#n6b++%FAVHhA|nsASa499tnND9EN+EY~Lz?hQSG7YG4&0N{bQ&u)Mc zcmP_x1Ke709Rv|*z{m$u)bGPUL=V+vjtFp)(0L8mOn4geo~787FPzf^6@6Vh$~bZ9 zI}Uz!zn$NSD!l=&gf&&lD6vFGotpF$L&9q05}%j$>2vQOwn+&I1u4-dySevHJqf5N zy&7hZhixCvz~+fD)M$n)$v?~oT290gx9AHR#H&waaC)a&WKmawT)J{Oe-VM%U2=iEkoAkYI;(Pe6NL8p9Dyk*nm@D;V z7-Y$^m79S}R7z(+v^&qYoPDQhM3axoOXtz$?MaN|^tv1w#Pc|bT3{LnB(A^m9&pQAF?k{l3u85<}?XN`1Ge_&u!_lAF%02f3%r7C{htif;OL!i5sxOe)78?wxm{AGOn0 zB#uk4D&EN|;>@oSUKp7(+tM)+BX}kG1InnurbI&W6Bu(LuF(S zUPxIG^@nXI>8vg?Q2bKK!02_iXZhjWB=eJqjin72SCw)4VEL!GO5rX)oJB6;iT#RG zVGZOrSN{ZE9ODGnzW&JXz{83IVT6r)M-PzhEwv|Xhf___4$?a-B`IcT0Y+ECzeC7s zc6|7r*kP+QzP6Xbhc6?m?|5t{WspLYheDO$)96p_;Rtt$3kKV`p?zoyZ#>aI`q&kl zeR16L3-zz0hb3nu=8c|&Q0#6is|#yd0hCJf=m`evwvGAR+m{UZUu)}{<_SZ}oP{8A za^025)-EYd-;@gAR^zfuHuBY_azs#1wAuWWa`b(U7fS2(J>%>TGrp&th=eIED7=UJ zNKoJ>!4A}&J~)2{jQmZH?!%vRaM5Q@R3hkKd0|4=gty@W_kK~WIE!}FWXrR&BR+gN zzWP^Kek?+6yKikKT5?zIhc~_)j;6g^-+r%3^~t?dzgm z|KUT8Y2vPH28Z*@TrL9ltAevFmA68<-~)5FqbbDFkK$hxX5~~*6l8M`=1axt1@lZ5 zvcjpzea07WABDTlD(icho$5ag&2fPb0y!c?c)b{(U%8CO26_|V8Q<=0;r$0FZE3q_ zbBLF8bO{{^Rg+n2BbXPb^ty*_F|N7N!q<-7fuc?QpGdI4Q1HAQxKK6YKeowIJ5N2* z0&-Gd`Y@7QQ^zFhCWn5{|e{}Tkr_Mf&s4wnBN0_^&q zu;wtRe8&=w1>_UBR{^P%a9ZLY_K|Ezb3>5J5DASf09-Cv!q4&u*J~|POBA0KKXi;u zfcDL5T^t~0O0BGtTx*4{I#KTT4Hd)ngmw93+SBRVj+|1OCG zUoWhxo6t|*rmW;($nq`1o6BwY)&}7ZiKA`K4Q0RbHOPgAborx-Cgmf?4fB)iXhlDa|!jU214^gkNuR%UYVyW8qBVSiK&cdqa1RW0S`Hl-P5AlWq$zSA<^=`?Oj9DooRFblRVC(fL$zk*2b5pPHyB z*OU6g3C2e5liD#oZahcZVQhfnJAJ`j(?EIZw@C74#vo{FRl+3B0+;_%rs18&6;ZEH zJIP#9HN34+m#=xe-iufMXh$z7!+tbpth)Wjxp{1(U8A2J4PPpP-F6)u-&ktoHE6$p zoopxsT_y=-!&z&y($Pw245zM=ol$E}#h8OLE}wE>bv68%`!V`Jq|z-0nZA^&dVDrF zv6SDLByrTYTkt8+#))HyoWXR=y9=9Jyr249rf}FPo6c4@ zO%|XoO;@>mL+rxn+T3j&Uc8ZV7`X5#7GyLc7ZL87QM^x+vn8!jun<*ubqZ=l=K@4X zJdATD&UdVng2%>5wHMAoN}-YLv=cs54=|Xl9Q^Vd=1`^~C%?~y&0;53!xZf$hH#YG zkMScRFx@F#O{JiADZYR(Rbq0H+1*ETGEB-DD3zdr=?E%PL2mxLY%8|>LtL?<8mJ)V zflBM?le~0QW9;*STJ^i&aHzv1ZAnL*MpxzOoPewe#|+`6(o2St!yuKF@fyyNe5lYG z7{juFMLY7HWT`ii`ZmH|#F*^hV{wN~@m=<+0bg-Zjxb8-0 zKUZGt4lbV`3(Pdm+_c%G_V+nhe<|P#7aEoEaU{3puUwHvbC5kcVv)Y`P>!A(ajhTk zB>sY~q<9xHU5}{5Z?y~RR2s_JtJ$u0^9Wgc4AkiMg1V}j@^-Tt4%w9j#>ZQC(t%LL zXI+n*Wmj5^o%N9{F!vEBb7()2F_D)G#!2a4qd}h=>Lhbkm|d zzO5F{T$VyxeUapA4@P3#Am|rzXcRwj-kQ~?00Fh98ublE@qfq%jEEz6eMpxLQe{;m zXmZJtj+)v6J+et7^Vw8wH|;vNaJ-SSX!4!87JN7FDc3hG^>c<)q983*Td$6^(UWz+ zJ9ohtkr&t!UUqMh$;umgr-2qNjuQ$KwGH)lehx>-^ljPn!}VCi6Z&p5$k#<4MgzDP zfVA|%itlH98GO2{d?gPl@36Rv(~|AYSwB`?MBjV8@d@2Vr_FDJkt$H93$f5thu;ayN9n@685@4n zyasKudl%2{D+ZsqMQVP}O@Q4)*DzbC6l;+{=Xw3Y^(a;8_RHbbsxARMVsZJ`aoxN` zJ11M-NCaNJ^lwmCqNW|Ru@1Z)m~U!G-rDYhj0`s16>1lS1O?(f2QLUz9(4ZtTvAQ~ zik#6wmt(vH_l)Rcf6IkC)Tsw$)|os?9RN!@SZls>&W9NePHe6?IkUtq zu7}~Axh~-TPYS^z91&4j)EaWuOKKM7VmwdznmDOPPev&;vlxPbYZ3m)x9D9%?&&c* zgo}q{!BFGE0hlIYg)7t*ni%4w0JaZ&o>?e@qN)!Zv)$gWy-2fIRY+LTyBHlB7i{Vf5S(Xd<0YF)-L)-dAB)y1C6>yC+yCfEp)8v>TC z6!Ab2#d}HN!yjR*=z@B$W_cHfv&+mi(xnHtSJ4smrnn?698D1`sO%wsjMn1jeavRVYnWGyzYfc3X;f{<*a;R-9syIX z&(uKj-UOC{(LW3!MzD+aLPQ$3|0Eb$#*SSIAq`2|QiSJ#s1yU( z#=BbsP6`HgkaQ)ibCj}nTR%NwubSZbg8C*`J}9|;Tb#VVXEv@18%7cZRVUd+j0j|b zyqO0k6rYn}Qc@*Q!1J$M(Rnz2vQ0n@h%KBx^lJfl_YvMEg|Eh22z^QJ1MQ=f=9BT@ z6K?^0uP~mo9+u5OI~3$|eMxG}HfbNZKVq&!Kv9ZF8|QSK zI9_E87785+aY^S0+JD0~YjK>sc3w;wqkt`E-_ySceQ&PDS&~u@4bqj=Z&Zz3AwR`% z)17#>dSn}{G{B~u>!(<2LfE%?U??J26&4ayqQm)i))MQMqM)R$EdY@UaF7}qS+qQ9 z7!g%1%zLG#u7B1X0FeXx&!5b&4FR=CbI)p1G7kk=av=>+!s#I2qGCJW7ZH?JMKOx; zRKRmAhZOMIhkc8uE5iQ1e$REZ^CEbd--GOVWrn=`8lpCx!M51Qx;3_R$oWOL<9cgm z=<^CxZCB;m2Ufq_3sTjLGx|&2ph*j5gIvg1Kyc51jHr~Q4pk*YsazUUapG31i)W5e zCRTHLJTe{0N%>82Gla||&GxbmJ1FcF9I9&O!g%SpUi=sQMi0(1XEm_G)cj>N1H|6f zKVTG^h1*EV3{tfhCQBvo805!b*U9J6L;6oi#A@Mt*}>+Oowr_06OIQMp6vI{`(s)U zo_iZ!jY$!E8P|s|upvosKeDJ*B^j%SGGc1$&(5N9b|I8X(Ffx|73Mh+kjv1EiEpeVm{FspNDo~Y=6o_qR=%w=O9UG?q5 zOsDC85mH@qXrtPnAU@W4ykQ*G2@7TAY;k2yeGz1WJ=cZFRvKmnKMM<(My^!>*1eNb|*YgG6tgcfQJOoc%WWLxbQ2 zHSbwW+ACD=OV4(L7v}Clj2}6dUDwsj*sL}qu^$8oy{<_y!e45KoHSnoS_fUieGrlQ zVv6E*U=~%&uK7^Z-=3iO+QI!o2Nute+H@3vlxpX>jIA(Nn;Dej3fl@0%`24O^NeJB zYI8NUdVB@PYwI657$3_riQ)-_h~>xhL!%K6V|@Dx=cXL2?+wBi-lD4*UrLP!bGf)( z82_v*E7187g2zQ!Q|{LgM~5JYotuhaFXgdV0%i~RjQc~SW;TR=LFS=r6sT2@s7rNr z$yLq4YyJ)PRzH&jw=FVXJMz>-wKGm^UmXRdqX17m7xxr?p;WJUIVAs#0k(pYu@fy` z$bm{sTKZV_`r9b(thXIfcB6_!0Q%8eGPdU_%!$&P?yCvrS8`o!MB9b*(T`Rdd!-s; zfqh2qSJ7zG^?7Vh&(4{%E2hqU`YIJ>T3W?@@%w*Q8wP_bjchua|fn1$ zgSFHbNX5+9ji5OeB~lEMwGIxpjv*~PJ$)Tc(hm}f6>IcVU=FG*taU`(3js@4bcd^S z3)vW1#6&73-b2a$ix^saeQ}aXn7gj=emRVgr&o8t}P}3 zRbgrb0p!;e1*Xw5Qa1cVJgxpldQuqc|FI@5-T7pmnLK#Yxe(hD5wWzk1(7m4`C3(T zpcHsQT`_k9y`?Z9Jy)fs)CFD?j=%4wrhx!c8~eppyx5~@v9Bzg(x88ODQ%iIE5>n6 zeQ}>3!CjQnN=_mJRfaG8WQCbB=$WQwL>zPq{6>8Ox4d-8g!Aj zv;xPRO-zL=m1A-_am(t3CW zbVV&}s#*Ns*5_slyXRbyIC`hHvWO*i`_Z}z9KCdQO%IPTAb5)U$XQe`aCBvz8#^tC zC6Uz9D>shu6n*M0a|5;WNBfgwQ8B@?A)BAtl1 z&0YQmEqlXMkQh5TW4sV~R; zQ1>LDZ-I(JZx+r=X6JN!(ja8cu$aXtz~ z3#u?FH7jyo6&Oz1PO(u=RRJ3~GBfM99V!s-n+1^0x#V|y6j;%@4HbxP`?dVzOtXc? z9(Q~CeA5OWE!RM(4tSB8sk5B@fgMVto`FEM^BhY}4QD_6=+}rsOu912N@p`_i!t!Q z&3AMpMkP+#>&>1&5H-iIzW+@d2DV?@i!DH%XXY)oJ9d0c8_GcXtaH&o4;%yZaC6o;ENx!w_!UsI(0?Qih`yLMW+M zOnJRMm|^^Vbj+JP%=s?y$E+^?e0@JVC7pDSUlfd9zJGiyro3(CLC|uT!p}FrA0>EL z3zq{gjf2HxvAw>3yO8Sf`*}aUv}k@LoUhFIJJ0gCgMQ}j=Sh|~4zVGp-7&pEvf4Nb zsr*MR5u7lkBjfA(lhJz^U5mI@CU%q?-h#C{HGSMU_qY=-EkEb4Vne{>7ve#`xpQr+ zhr#yj$tMY8mi#vRwS{i>)*!>dx~&dlr(T*Yqv^Ltk5VTKUw!gpaX*0-oWmzHhL39q zw#7Wvia7&beVmP*=*g3Vg{O;KeJ*)TlsJ#`+$_%@A1KpP#Dn0J(2q?UJ7!A;*!v!u z?zV51Y+PtoPx?Hw*IFk!o(K)IO^%G-`9O=G88>!S;I!OoA^4Jtjk@y>jDwnWo2fPZ zluN)XQMyczy$Ye#mOh~$Av*x}o{-hPDPSA*I6c38Ii1e7a4Gqi+Q3ln%P-$2r_o9j zY39MCw{OgnfHP$T&O88d2U#rr=0t)n%d9!hx8d{8U!i*eI|2JH>rwk0+RJ8;`fLCn zEXRn39y`6ZPI#U+Tl+sRT>l8BHl-cl#@xrNn{M^;8(Y6xtha0zqndW4suQ(8C_^^m=4$KIz6s%2vcbWSV;Ovdrw2H-q*CC1KNT9-2m2Z7Kxcza)U&aOX%+ zjvkMrON_yJoHed4AaW$$+*h^n`YT4a7LyfFXCdY~L9%195(*3%KS>BoP~U*QiVklx zDVWPN#>YZ2O{Q;>o7LvC=)fDls(EjtRkt^pz2Aw+n8$xJ&g`(=Gw}`}($CMyDWSDF zk2ZJhzGnL2%5k{^*T(IhDXQzzx@FEM9m_>qmjYu-YLcs7Sv!386G?pzD9#26V;L8} zu~MHdPnZ8fE2G584mM#0C&&oUlFAIHWK`J67XWzZF_f1%G9cc)3FiTq8`a0_-GL$b z`T8`b@TkIpxC{x5=aN9sP{94Q;V0##e)WHecZF}}mM{W-U*Af*{v37l6|E8Blfb<% zvPo}oY|KR7RlB^rw3Z%o2g3&Pw?3PgbV4jy< zEy!xep0TGPL@^XJMw1(}YZ5=Z+ zmgYZX7eD&{5b4`qs;_@D;x2=Y2Oj7xx@}%MO}4`rgE?xKX-L2F*kN*ahr;~Td9at; z?HI6EU{2N!^ba`-qE%(Wdv4 zCDJ1x)w*N8Bj+)yIZW}*bxDflW~(kP@W+`Eq+yPeR?OMe@@Kzjp{ zZQd63hd6_rV0jHL)Uq8kyIo!qtUOm#CywruwE`O6N@FfBpZ;P!r{=m9n;V^xccPLg zv_@%^Z0axZKUa8jU!8UmOISHndo^SAv4U>@$#|6Sq55(iQlK=M5w{^f&IC?k7E*7O zj|4$(a1PitbG@c@KJ}b5S!{vi!kvG@*bkxwwK*L_*9B1>(eFR86)emI+Cl5hpR6Nsu z;$jC4G-w)Y96u6$(e(t@tmaOd#`0*Gg)Y^khtDn@DX{}6bzi81uDxFsrF@>oCE>Z9 zjhN%*rjvJQXicXwoZx9oMh-y`P*)T#ZrIX#5T>r-DChCB@dBiPian>5rr8Mhr)h>e zD}#~x7+(WW=~7S?^dqMDq2m#Rpjm;J;uRI$VV1x<`*BK z6V?qSfSm{KV)m9%n63LV-EZ?rsCW28bA%%6eTc(-fI_FWgO*pwv!8f^zBK`H2N*YudVG3PRSx(J-x5p=g>CPYI zPQzDnSTw8QWvv2k65CRI9`K3&Ub!7S@fVNT0-e=v1G?R*F`|q_8m$f2=1$9a8mo_5 zNgXamDvJr$7tICo9WJ^xgNmW1h)1NoB)2tU*|u_?z(H~~2})32+wcbO`l?om{>?P2 zjQ7JL!1N7{(t6M3)!Q1^RF0G@SzYhI4*=%?_v|IE)dN-etnfT#jfyP!)il={wE3Z1 z>i&T5vVA!k4Ed{1O#;ZOem2g67wcq@x}d#q&YOgCMl}^W*=-k_t|q{dmpSzX&M>^Rw1MSCQeNb z>QhI46eV0KLan0o>{XhMP@gwvQnA-yetKx{O9t|U+hU8fgupjgGt%iu_)d>TDl~`s z?P>k7eI6!d8c5q2UVlnb8t5%chqJw3XB!%s_i=$Rn?&ehO4Cpwl@gklZ!ivhAzO

4s;=ljQaxNdcWk%hnDgWaYOLtJBLo8Y6-kOnZ%cr{F zLbt#2#OSu66Jn#q4IlpI(pgte-%{(blJM-LO84~#lO72jz1zs{>qjeBcOT9LS}}9} zdfTfLvB%tyZ+DHCtoY6eFH~A;9ta!Nu20i6B$m^)L$=VQ!OBZOqrCim@C!v>`E7?Z zhD-!;{k)p&{0LFxhspcG`C~$MdE6Zf1X)*sl&c){V?qrw;hgF%xlNM70E85g7h%O2*z#NHt+JVCf zsb_9fBGf2fp{=qCrWefpWB`4Mj4VymQ3wQnGwpIp7dHSwB1c{vjI^&ichm{TS<)cS z3UZ$bt<)sEND;Y0tr7b-mKofDc!9{A@O8`{7ySZ4E70Zp!SbR=sAOp3ry6U*KD78P zbi!2|6l6vOjGwTmnZd02er~(O7ofc*6r9mlyU}9%|1-kXk>9 z_?5>C%3a_}Jv3AS$7LNm(MWN0p4|sHBPBZym2VOi0wMi)cYTNKhiF%S857edDZSah zrzSF)p(AZ#nSQdc@sGkI@Z-Ec9GOGorNAEmF^NwYtP$jI?LL6B6UfB$kvxY7f)yAV zF35P{Ds37!BzC-z1%VJFM#V7J&w+Bzat$nT1*`(=5}NC!eCM8P+_i_XZpiY1-a4`} zoEN)57M{Ivxq!X(O>TJ*Sd6K<;oOS|H_nu>*zIVmK{^&rL-#!IUR-6y2e7}zKx@tc zPr8WlTF7rHC9ZJ3fjjf5yT+V!)wC7cHuzd{yI0yOQ7!a8Mhi5DNxKdha^uf?jJkMX zG+Wku<>qOZ$Kc0z?#JK*L(EMw?y%HlPuEcRYu;quH^9e(TU&yZ{h%Q2SJsZG^%M*p zXJ1%7Okx|IDGb4SZvXZ|)0iwIM#`VRvzt<#*{?YDMx;lAdFVN)xFPY@JGwL6DoU{E zMC_-!3~fD;p=&mq@rfIa-@_ZhWvYB)yGJ?1gfhyYq>_5&>+>=3*!J;7>L)O7#9x@q zs3F+)mdcGpirSvO&_<1=yixQ>(lORQB?eVa{6BfJ;}aSbZgY?5H`+F#*dC z1{~WlELV!f13|&03w3MfT8)#9LWu{;t_F@2F&;UVbTmrSC2N1(LlWa{CX%9uD|pNK zn72K3d@ugEw7mMg{sXW0XDJPD-iEgeRUk-*+xF?c@jX&^xNHgI@ttl?t z8XQpaB=hFWYVkk}vK_LN;KY-9UUR%bXHsmr<3OXsyfi~D%`WM^76fmV)~>5G{O%>V z^>^o;3>rTSCtIy}dey@bHHhrBmyMuAp88(O?|8+Zdo-z2^y#H3uMk7}+`he{NZ|0K zn--UPcIpV6l*KQf!Q3Edhlx&xy_VDP4DvoCh(6 zGA9gh)KGP*Ba)*^QZm8M<)$2!mycel`chJ_f4qFx;q2dCUVCa9GSe^i;l{jJk*>F~ zh=vOy6CCa~`wno!1yFuTNi3@kuc`I1-dKq=!)D@L&QY{2@%Wdt;`fO zdN_d7mPFdNbN#Z%pxH_Yj7T3)iJR9PnwgAo^woK00}sUji5fU8xEd(c>oYP#7U|6353K=Rf?JpW{||lZ#kCZ$BEF$*(TH5Amqd?uuXj)z{~b z*~$OWeq0k})ttN@??b#N7N^Chs9m!t7#3FYc5DV1tJgZuxiVN*xvD*vou zxuujA`_MiEN8j-W{g|@<5r3{QWWAIrvg)p~p9$*Za{}~6n)P&kEipuyQJ8-I6Mq_6 z*y@|JllyZlvG)GLYTkW{L9raY)VgRsCvc{$&+cXDm z(bB}Dd=$6$yK@8TTO1+yV_y9dg;%-IZT0K3nuZ(U{;nSQ7_|6msp@#7%Z*%;)TjRh zmBJyTtspYij(F}?HOvo*3J#)Oln+$mL(DsQLk(qHaGaKxkJwVFF^l1z>91zazbtjx4t!8Zv)cZAG*!+^bl4qusW%OqZA295 z)6<2e2QBAi$1f^dhqdp6lw;TCtPqC0HF#Xq+750=AJoDx7N77W$F~&ku}1Ys@tk@@ zrtg{lT8HZz2;fQE0uLGsgx8M@yOnN$@B2gk>V2_>!JkckbwDNmcfQCb%2&tD zx3g%^?dpR(&DL25T?a?Xhx!A)-(90F zjO)8)>$dA&o`BDxF}8!Y>fw=Ot=h_T{TbMlAc`=T zEX8(5wVSSa4->pC8kzlJ6zY#CSsh0qE|+!bTMPuG*gHo13hxSg^>mFWN~$B&NopeM3<( zAf!~peuXtr(!H{Zr%I|Xv2gu@s$I5@hP+rRgh zaWRx-mE%pQkH`3HckWy#AGJZ{4%$XJKp z$dR^*9NNbLMyPs-wvZZ?$6f(g1+2N#x3uDObll^*wK+Kok&k04C``QZHkr7I^7y3L zZOrs%=*$S2C{0$h)OCMgVTeE{P5Gks6nmp za470Lo`az5M2wGp#T5*v0*sSEsDhifWFG5DeH-(G{jD}9QZxwgQUWT+8t@6+cW8Fd zWnlM+uN(l3aYAf$oShnT07!~a*dy1)Sn#&sR2hD?P+)gVmN95V+~4W-4ULY1NZCv% z;-NykTBZgK#>TA2e*!H)n?{Z#m7C$fPq<>6bL07d5fPyI-Xmav)Ld6iIu~-m7>Xv;csW7TRo$H1=Acehi z8lACteE<*6#hsrs4>E}{8O~I*;||P&^xXlgZv_>gvjb##x}m6!7`$8)s4DN*md}rF z%gGt(1&nziYRt4Tbgk=V7pbt8@qI#>A{!<~d!;209<|_5uMc$K%w7g)(at7m3LEaw zg~=+s8ajzC!|_y3v8k(kmF+<`7_T1x;co_u>`XQ5 zBrK6A#f!u|^$!G{o)yPxR0z3`OAZeA%rGmi*f3KmG}nMaeL9Dh^-SqkE|Vp>qt>wmGbzW_Ku}^Mjz;qi$-00pX>>{bk9IRP{i0KwTGW#4ohe{ zCaQlOR%+DTs8Z)(y~aUXC`#G`mURF%nuwY2DltFI3yZzg1^lE#w&HP#a!ZI(e*JVo zKFqo?httAD&@SAvES1NZ$&4sfl!SblLOwNGUns{cY`ioEa4Y{phd_iFc64IGg(S_) z9c`pPPYmcsxjH+{LvM)Fw$Q8_m61Xn144qDM{cp&@S(`~~i`HFv=7m2I#E;F^5G7e3BCDj+Sr1V6p!8P(< z!bb@BrhGup^`}WdAR&~6suY#{u6@=|L0He~Ua2*f*KQgg{YVsNgwl4ox1V0os3}6= zql!)h#8J<%Jw>mY1k-ux3l<_=SZXw3)E~1ihj}-xtrz)TpB)ZS;14tXH;C|{QY@B@ zX6x-1C0yg#l@y1B0pIqWc40sDxCS^!dDO8|IfQc%mWAl5^<7|jj*N#?vDeYW1ULVF zhm0Df=gIDeHZrIen)hBp9@z?rtfjJ-=8O~(qEc043Cb`p1EgHJ#^;Bl$YuOM&&%>abU;f%so8v-sfVV8=CgVvE(rD|^r`YIJUGGEN=wsiG(v9T z`+;hS?Tg;G-f1>3k-=bUmTTrqKb^!Xp+0o5lU_2jMyRQ7j7EB*q(#O<%M|b~fq888 zNL5`uYrIMbL=;&mzsh=t%AF0b{bk^5=*do$xyk_(wmPWmi&j{Y=`CS!bkF9~aoeXE z%K4j=?A$d0D@j2oYYm5ot~=c%3;uT+$?r-(+0o9gAZ29WEe`*N!CC9rAMC7u>>qn8zNkjfs}-;y_x zl$rc-l$plFG0qah7Iey}ON%StC&5X8v<~8=N>;_SzH=suw7VkdE^N#{S*M3j{I3Zo zbD0_3Avy`a)7RPDWq@RIr;lb`l6dQ8rfuEYl9cXh$E? z58{#;%9JeG6rIT(jFTtp7q+IQ`iFD@h{;JbjMhHTN)T3woKUo0B*2VhLDsS`NSp#4dY-C5l1h2V%cv=YT(=$OUzQjHlagmU-dLn~ zs2tdN8Cz+BvzUAt)r|3ZP|7GKu+Qwc(Td45%MXa_Nh;;>(G+DPf2p`pEH&9L&X4PZO9!oY;g$%}iGjWtt}7h1{T2 zOsMX*yKhY(CBXQ}^+}RqxUadRkIUcP^*R_g3dY-I;|xESBp`L%Nf76WAZO5_vg{OE zK(e_C&D6FOvB&`BaOoUUSiR2M)I|uKX%t|L{>IA5e0_Lsxu%~+k*$hU4!PhrFiTAf zd5UZ5IgE^#0@tuIXT_JdG|={1^zRq59ohW6Des}T&_I501dFjU{LbSpC>Bzoup(D4 zLi7=dQdb540+ZV6wMR&^$>@KB0fs_qUTRD!&>1GieDIr52jzD>Xw&6rr>bSu`+;^H z`y*<*_ZCm)?7I^j!=JWBwUuH&__H%Qd3B#%jSJWj{i9Dse+X`{hO#pfz1N*o`{zsYD)x4)~}2( zxJ=>aqW@m&4qhN-pMXI#u??2v5gvWigH)8@XYjr&bu$a_S+>E=ksK0usiE=9n)^$= zvfVm^b;kH0Q2#K8Jap{2{?tQ18FAlKBaq78E`*3l7WUB`pItOH98<7DLtAmwQQE<9 zVetD0uHnaXr)~JNAjk{DgK|Ep&7PnxE&q1a^W)FEy7Biz^4Li7Cm72gGq&$-XF<= zF;CA&bNXg?Oc|qmF^lX_l=P5#{aoVGPeo>7%B%-e0Qo%ixY>4spOEi&lBt`LM4!Q2 zqAXC`y`h%siV)5wF1HkIRc?U3LNbu4m(rtfj1}|YG>C$s8{VHlsT9GTL5k{#R zn+Zj#g(R_~2ib@*(IJq=zfp7vk-;nyTNb7M8C%OZP*OYsK4bZ~5B|l$<~xwkH7n@Y zYq8}rx}x4(u3GD|7B@@r2$l70>lt#+eax_XgdR!4VU~Hp-YVO3bLgoypZ4~BTA2Fm zTwGiVX`fjkz>C#EGyVpK5i0K}Z5L*@PDR1-n@dcw+OS9N z(Y9>721&<3v*G!{VO0M;ifYjwvIy9MbC2X4p~ZUwy5KUC!lSGDTWSFl*fMTGZ&}5J zlDA>Eg?M7-T*X<$Sa1H2Jc40Jw(MxBVD))nG4meihl7KeO;!-r6;YzlT9`e}4A|PJ z$>Vx=?w`~EQ%5xO3K6#VKtx4qeL;WR*+}a|`Pn1}R=T@12EA_`b}FUa@Sh-Y8Zlqz z4#q}^^7UPb0o=Tl9-XjpZ$LMTL6PIVGfQK=U49Gr8cI7|*RF<787nI@k{PyAYjCEa zTC|T|h#>F$ko@=A(Z9(A;|VnwZO$&1a%EFbI z0Fk2FTs_;kk` zo{Dxe9Zq)#+Xc8Il3BZKC;7u9H80}DT9Ov6vs7-BjEu~-(&*PnnSxMFRHb>hq@;n{*BW4R<&j+razP+LHRnrGx2d>GO*&zADt*{4g zmRHd|+B-A863KJC*xUZ9_4ZK|amfck`h$|0Tz>#X=i%DGtS*Nv_oBU39+r37e3$yk z5}%siy>&^JU@GxoCF)Wt%h}V9K6?6`o;trc@7f?s1D>|f)aTP+y}s1St&D@Rb7c|| z;6Jtd`YN-V^laORQHaU);>gX?{B*T&Z zHFq}- z{hl(81!Z|35%<%pTYaj-MHRvz)Hv?y5kDG`9>|a4GO`{Sf8epdXv;^I_=CP78+Ec6 zK<@VO8pPOmpMQE?Md$Vx#V%hjc=exMmnnCmJ=$lx_heK&R&ilT8YxOL|4Bv83kI&4 zM)jrTW09f$1k92stL^gz9QWMH0Sa4XDgWMGrq7=pF>qxA&GVnZ`7iV^s5eE^?cZQ_SC@aKSgYb^&QH`r;~2;ftH?qhAk>P+!6)t zrrr7MMN;GE2|A_24zTGOng2OO>k4DA2os|(?Fj8eR|dy2A(nj>LnH&E>iunjYSDAc ze0u2pmsS(~>J;T9+_gZ}w}Qd~(q&SD8KIGBFDL|$Navemn(?vxK{M;dQ>GD$cwSb! z4@8_y^fk~uwbalqHkUnonCo?X_^!wZlwimj;)J60_^zo2P~L!9nd|fLyZ1Lu6g~}* zHbuLudVx_;dTFsZp9mqvaF)EQaSFD7I=P>9Ng!{OuNFA}YZLa{13J`*r7Y10PL5YH z;{-k$l%+b`rnF=NH5FHd0;mO(Uex>IfpeejJq>vPZM8V)a}egnrQBZ(`x?u6$aD7B zg=Rp!M<7D6*d`jd0Wsd`OpyF9oko&J^esH)vLMR`z3$0s znCW>g*}jkWeDq12a|mz<6Rr)ttF)(gH*rzcjSb#`HV6#D@Cf%yA5frm_`!978etwR zfGUt~7W`_!QONwl;O}bxwjZi@bRc66PEXYqNA3C%PnMbkMzcn$z8pECVSi9FWKa1y z`21&MVZyYXNc7u@XneU8Y`8(eloqQyVO@&Arb)WNJ?FKhHhSbbONWa9hSY5MT4v3x z3QdyYtvtC=$5-Q|^&KPwzuHwad1uoCh~hfu7?W~XLrQVcHEBd4DZ&V~&tRk{uxoIW zD0Id@f+Ln#7KBzf!o%W1o6eH?9#%(C$u1X>FFwa-Ap1}&5`0LaKt<@&C8r~0#u*uF zL|`m;cYZrB(}H;K0jYRs7}s_Cn>Joq86B^UcG0#yFd*`hiy|c+`u2A2a$;zxqX2O& zHQ20qT+%43q`(2L#=*@`2*Hreq0SWb@d0_>64@@w%#IMb0%PrHccQjH*~CGR+OQ|J zl7Cb!bCry%HE+7d!y7L>I}N58X-TJnum|~`p#sOZS}~bM_&Xx$G6oK+2}K_`gR7q;aBO)d z-6}6;E`J@lO(z8&_W-RcWTYdogbh?WQ%5(q)p)*}nCobbR@r3_9fGuwlMGe#p1uL^ z95`TRxyoiE=tLFD4ZW0vOJ%TXOn-jcS6?H4kjTqIe0+2n2=3*vN`N>JOptlw7S?lu z28XPMe724>dMw`9G;XmTfcp-D+Mkh{Ff$L%qYFz5hGc*@A+J_6R)NflDi+&`wz(6V z3O$}M>Y4{f3!E?i(L*MW*H9fUkR0oCf)1$zdF6%uX#0x71!n20=_Nh6%@X- zV$nOh8_+x2D_q0c;#q{3fo1aRp5p;}j?3~sC4zh+L>b|649HD&m+j3Q71df)_RDiq zFAAUsO6ug3Ye9}~Kk0meGb0VECCn#}U$6+gk_24rm*PNs9fo!fj$0LeOh09tu9|Pn zRN}C!c>@`K4P3A|7{V_uE8_d3wsECtu2t^?<%^V7Y|e^Rs=`^Q<*wPBvB8Sw^y_Sy zYMNlBhE1z~NaImxi#!5#Z(C!16AQgrWb4sdhv1h zih^JNXyyUYyaZO^?}gx;n}#caNj8cETKdnYi$&Q6awW3j3Y#}&Nf^l#Q!*K)^q(7q zgnrIg``XB#fp8~aLUSw!g2>D%kaK+(oS*GMC}GK+8UzFZ-#Pl8#Y^{q;`Kp+NwSBpA>4{wa0+Xd2?- zWB&jT!<-D$+HZg?#Ta@A$eEmnE_0%ge3~14vlS8wp5{-eQlAR#y00{_7i{nsjkGIOVHX}+V67myJAB6Ej`-R<_PV5eCe|`#V1+>;Bc&EEegz$j$@WAL0Mbuz4qkUnFsQCks;~u*e#y5`BC#kTQ!x(c5W58{Pa5VOLi$;&IM_J=cFi3)NazV1P`O{x zCuE{bOwPH26#jymHkl%9@u(_4IBCs=NV& z;!hU}6WO@-=C?PG4uo9xt~R^(0Ex*euUIJTmM8)s;)TJ_t{lLBhQAAPrXS!kUOJjyeMy&ON{t-Y5BS*>E3CZIm zKl&>spi3n1_u;pJBu$kE?yt?4%M3$q1|0#o0ReJ|8Ao0PX|q$aPU?U7_Zn)}Y%FKR z!`rG)w`}KpQEh8$9GxLm+cviQO57KkjwKu}ecW;1=t~NN)NBtrptPT` zIB+E)(0ZYoOTueHiV4pDVeA}&M2QwHTefZ6wr$(CZQru3Tefk_wr$(CtA4$X=!lN! zH~6ER<|HFduCw>rDyyuqA5AWDf?TBSz%ady+k$dZx(g^}->)D8#`TKDn;DT8eDv!+ zOXylTK`#_?X|oH2`GW@}Fe@1EHxYKpwK-cxjfqh%Rx=aVW3W0~nzZn$#bq~8*#>2@ z8H>0F1)=MU9Z#_la0qes52Fm3`P)9rS?5IhXxJb`nb1e0tTtvyzXq^uww=}+Fgrk| zs|f?G?A*UF^ud8{XY@v#B}VY``sMgzTo%P)P~;rFNyLp=`xXFUrzNTKuJKv>40UNY z4hu=6bck4MYKdn_Sn-?V-CDO`Xl`2zVVNngjVt=IcN$_4y_s;Buink9ke>iPX9m%d zJ2=FsBgDWv?LqGZ#A|057^0=dVlE6bGUa2#qNOes=_192VganKQY(Y`1@okY5kXUA zxJlQs1e@UI&KhT_3xxxYnPMSjF4p1JT_wN@cs|omYWy!J5??ysVoxK(MpNLuwrJIRsK z=f$p*#=P$D$@K4sGGno>VoVL;M9q(QER$$AX=maqBc+Ph=NXqzW6xFgoo$RFm|_Zz|E9H zwuQO$Gji=@O@CmVdL&^L`|G{n4uBP61)1ED@|Uydm=r+Cg@Dx^x|taUjZNxj zlIK!$k1EM_T=urZ8(rHbqm15*Ql*i3T`I$f72a&!L8O!lv0K5AcOQ&|N7Q17-e(#4 zO{!LWg8rOKbr=r%qG2#gWq2&PE6CVbA<+*COA!1qC;er|QjYvr_oO*3I<*X)ECPZXa@E)FtRMAyz}*~u@zIzKkTtJXe{rpi#T{)W63!5+6#e?Y z;ac2j)Uo0Efqf{YX`PAX79n9@S@Gxcuu;mRi&-WkqCTCcgHa!XJ1u|A*Fy-9%r$eF z9#vam;H3>WuKkb%;DvbNC(vJ_9Jb=SpbmC`x51&%4r_SDar4lOX&R7eR*-4s zw!q+^_==B=wT1ra-Y4MZ2g^FuoBOf%91>vs!C(d4ci!z`k>CfW2X8h=&#YR1cKdi7 zOZw#iS2anZU7ku#%Gj5sfg+szL+jK~(@|UR?~=+sbXb>ecK-N|j>y!Ea{FynUn#qi z27wWJRn%69m$wnQvES%U0Tqq^4McJN7ZAnC`M+8=kaph?OWn4<(M8jmqI_mF{N+$7 ze`z=)vq#-TT<0^2&=DY|B?JTjB|B1bdH4;~x=ugJfCNFEm@=}nkr4%WZw@+kx>Uc9 zIA#tVV}FKvCY}`ie7qkg5{gr={xfP2Tf2kr2=93*ToR--QUj6E z{QTM&wj{HY-roQxLBZurcXM1kD z!r$9C1|1*8(ROnYx^CL`D0DpY0#BtLy8LwhewsxL3yk9r0-1x~7s-tFpG||<6>+u} z!Ur#p1svL($e z+9S8XH;(5WeE&UZt2I+AAk|Q`r8-Z;K3#dZ*bG0~TEt^rO-!M>0)X8y2ork6|MTtm z+%~_L(U8d(ZyYGzJLsJjAW9((U>w7Hog0c~NiKt@!?3W8(}C0NIYicTem-Le2YGyJ zMC^QIZqPT6CiTk6Ar#2u2tw0 zs~B*&7d1^f0fWmDjq~_9wPWsV46!)aftNi}z#ItvtwP&EAaU?$EMHRx?v;)db^Ya5 z0>o8^f*DR_1(l?2i9kB|Tq6bFY4mM7 zWZBYPi(SX-cR<={iKW3Q!+b_MZ8APqpY{prW{5-3Lm3+t9~{_#bltr1F?CV|2`hn} z%zq1bw?A~{l(E&^QFUSea`D=8uFo|%=b*!(AjI$+4*ZGCQ=3L0FHY71xzV_Hj#JoM zWkTUFZ-7aOqd9P56BkZZi*n+J$lKaH4RDTph(+K6xoxbdwf7&E2f2uz0NVsb^`Bqf zU~&b{(nNao#vI)G+3rPvuq#1OWwyj9V=@kf@KL4h0IcR`c=XEtG3>rN-ui0ROOx&T zD`)|qAHNSMFr6Xb2ToxA5R~_?n(rQ1Le$0*!=5IcK~#xclzqo$V#8{uu(L<}1whh6 z8-))Fw7?ueIuo`|mBD7J7;q)^8sVvBb^~O+4S-dx}V7XwpGL1=X_oiqgzm0lfmRa0M>7zxTn!%$_fsrUNA6;!dhoN!dG zH5%xEP0Np{j$k_PN)$F<{+b|QcNwNaC+a6C7cF(4=MM}@{c(K0#yU75@NGYcD@A7y zkCDbxn-AF)0u@Xs6f$nnY=-NYpwB9II1cA1^>V~aic-OM%2gjCYc{z=l|r%B9*7lu zzUq#;tNuQ^5!|k12mFt6Nvc46yP2!9yep@(;ZjM<&{~lTcQJ>{0~&xPIJpLM;&oKn#1O%Z04@adr%y!^hb z&?5+$0=(VSOFfkOIdGt9&i+~NsN*=rbI&FA{WUP9r9tH$L<|gwZSPCW#0>&hG`CX} zfuL#;8REeZnj$`!v11ePz@}x0yyykGKRjw+#a-f3rwjWMA(16Qpc*QP=(5sCxVIuR%aRSi{n`huB%#-BHdGJQq+#bC z9U+KH!HnvZxRzz7Z96}Ib+Cn~S6Yijzkn7XUtjI1!pe>VIHVka$sSpy@I!@{s}`#FZ`aYGwTRa& zT_cK&0pA6Ao=!~OXLJG&2gAR_H=4mxcvdoVgU)n?1I*D1!?LbQ8Q(% zq;a!(cq;DN^~eVX zJy(;YP5EJ(i~%};_j|beX%23a-7vI}1Sm|O4-{O1Q|6&`xJlO2ywND(>YTb5XO=bb zR4?F+C--b*gYGU%s^xs}9*nbF)0Q7bo{!OUPbGTQISL&jeTRF4IJo(`B+D3~>dY5-8!g()uxpFzg_q@&z&$V)Xg25STO(p1;v(2D8; z*{_k&m}9WhNM8rBp!Hkdjc&jnvG=8+uH{8_&zQ=uLf*={_@?53ARtwvbIOfW=o9;n zf|Ujs?Qo>Y>sDE=`;2-j@c4j8MYF~gljHB4^-0Bj4gIWjwdN-15&U~~nmF`jgsR4& zo>~*#>r#OZBD_>oxck;wj=^m=>OD%(5140?516fw0SEXq2K=x} zO>js?Y(pMie31OrYp{tCu0@XW;^kPcp=|l*mF7&?jJt%fUUAq-8Y7dz_YHs(h=OSTFX5P$ zB2^}>M0eEY7W^%VD67q)^b{#X^~lGA%eVg7eE~hV5@dp)d_*XWDikQHMW(%17v^;W z4s-|=1)O^@GZWr2Oef(^BS^ufX{jWN2K(&J$v{@M!r&0V_geAoL|3^P<(?>>l1&UM z&HyjX6uJ{rh?`4sZ!Ce5d|zhNxQoP8NSSHasNoK|%za8nT9~UYR-CH0q#{sO`tD_D zEZn?nyU7M5+j&o0v|KNz_#Bd9dFZSX!?t*2Ai3N@TSgt3Pz|vb=%e z5J}TsYIxZbSyr4zg*(xtxc-ax=W8>Ga!yJ|P7~33L{)8L{p9^tz&GgBDF-VH-hOc` ze3xmFpwO@kb=?Knj8EEBRaip)UWcz7B)=)kyFQGlcl`rkH3o` z!Gn{EZQug+Kc3DjTm0tOVY8!lg9k|Xut|NmZQp<$O`+C~A4#)___l5`&;P)|^^`d? zoj|dT-ndRnCqTrGg%MBL6Xm(J0?<{mNI*p2$n1a`lN!AmFJGS4ZDBY@gPLG+cWuF# zj}UaHHDc#6h!bqZMF!3-x=bzxiG#_SNp_`4A+aFt*ZEZvA=nn2P|UI;?mS?I`_`sar*&%uPe!q6O<(soEWMh7I4|jU-fo49@uW zg^FIgSeAJ+hULs?S%s_wn=p6^`U>4ts)kI7u-b##Bhpkql-o+qfihj{>b^><{Y%Ia z^)KzNB}j}^_fkUoJ;5Gowx5p}QqfO*U9tItll)n;Z%9B=45Rbn89sGNEs2yywxxVP zi>k1f^FoW<=yMY^X5t$-rHB7%TxaH|0t}iZHU)lhl!M=U%ca;R<1qZ$0Na{iK=IN3 z(NgCP$X@Z`a!ABs%^UIgm7I7P8mVH0pf)F5AFH)IF(gR zR4o_}4-SS=PuCbYZ>63|8tR(6AVp;!djQF3yopmT6Y16vcpOZidcXs8{&*We8-d@` zL!pS!;2z14f_3XIV>j~!6_Yhlh|%6w$tWIu^uqXIhQfV7J*$ktRP~hsb|r7CCs+!` zju%hv&Bn>m-&GeUyEr<2FRfwnEr`cs3P^IB)QC>E2oGOYFcfkzKLyVf+8tZchGpn) z_!jMo7`8>+fzIEc;$m5iry}r&S=qeg_qzJA~yKR2{xgGtvPd9cXeHkaTjlb|&WAKReDl9kaV^x9qmF2OkhJD5MrS;KKh0 z#V4vY!Uj%MzH<&HphFw})m_*Xb3pXR3cF+1&?$*m9#{ZJC&r2I7Lp2ScliTnfDMt>AXdf|rxniTve9(P_aVK&wsN$8Ryy-N z6^GQ6hXpcMs^pv)&rkRs?Cb+zo92@3iUD2}tKaRF=n{*y z1X*sWG}~TM2EGIXxJCyowDDj2wc6CH7_lnIP*?58_F*45qkojHFG!&hw{CCo7)z zew;)hMw9&T?t%KP??HOpT}vXPurN}~aQDLd^KXnq`|bRB=k{a@lUTh{AmL}?nE}jr zm?DfMi;GoP3!0Jq)2-+>WRqodg2kO<4uA=t>KJMDotgATP8s2Su&xqrjfXvUK@{Y3 zdPdcvgW1hqGq%DkXcntHHT&*TaN%}YRX}Mn2Ti`Bx#)Pm8kKYKg5Dw1^3peD&@4(M zmEA?~KRWGngd-V9__Sy1rmZSx zeNipTpU;Qs7t2=3qav^G#GcXO9`FYu!wR8iaPm@?zJ*EiE-_R&$xf`~f0MbGUe6Zh zdA$yrxz(Q37cQkg@*K-TO-LgzsEda@*w&V&J7ilf}lG0r>Lh zBB^rV_Y4+TSL_KA5bWN*wF0q9<_8ZtuwuL~jTYP${I02&A4-S^r5`I4h2^W`kF%Mb zSd_Vt4#K5@yX@ka;4iTvD!s@DX+%D2zlwd9i?(F^*aLZ1I0Bnt=BNg!|T7=UO%Kw3pS9?eP zF%zW9>u1~EKg2Gg*V_!jzqD&GZShk?{^t1IW>QUarjaYm?pkGKr+0T z&(eUzqw}~O3r^|G%GYA7-Sf8qU@2Xdznhx#2ber4o&7H~kA>mC&^!)S&i|#Kj3cq@ z;(9^O=CrvC0$mhZ9?m9=gL1LmodyU13;6e z?eH3x+lbw$1$WLYfdGK)zRM>gz<)~~xh;1-_w~i5E$w>e`m3G%?b`3tx_Q&zvCaKf zuF^Nxqj2*3XiCpR_?kPHqIR*Hm2b|o+rMoS+5Bd@o0~gRr+Y-&?J+hr!g+YVYvO|U zj)-M;t?ZBF{bmW*zgJBaWK-muP9*iD9^G!A#lNyuld3i}bmbj9V@*y@{uN0qDSRj7 zy{F$Z`}9fW%UYe(H)P6qqY-ARuCgBiJDOGo=#DNcYS4CRMe~pud^?{C<>MZD zkFp_KX_<7TLaNa2FX0~(L4IrN0!&frD5DrT~Xv$G={1^MPqL3nxzpBIhUV} zubNb}KWrJtrz|OIpB#>yzIlkIyo#~iqmf}1!^yZM=3N}N=i2W;@{Lo2{EVe$x_i+ zYEKpgsPf0{trrz>$jxS{5l&1~Oc6QJ8m_i;DJ}3W;)Xu-YDWdxuByrl)yZL%=dT*a z`|a;74MXDAIeuhZlewi#DUX2L*Yc#>?x5;<%%MGj zH&=$Zr_}kyK0T(Jc*@$fd2P?k$5vB6h?rI69&flMCXGcSO_7lM&p0z?9hzulqFm4Y z?9W_Obn6;sPm&z6P06-CZ%5kc&Hw%&AN{MEsL8a=nGdV;RfWR=itp=Q{-c@g6YWL| zA)_hRT{h_0*s1LVA?;ENvCm$QP^T>-KAXn0nDkYhl z=V=e(wZ`#UisC+G+eRclcPXxUo@?rqIPc?=3Y`Q(PGpROXr@#i&A}IEPXQX7y0{74 z<}iTjZo4L}pZ0QRN^o3@oHnORnfW~MV3Ea9NxS3yr5=O|pRwodL4_e6sGSI#wOccu zkyQ`!akWrI^MGqZ7$fzM?PE9z(ZwL@qJ(P@L&@wd(1mM?EU^w)9MJC!jL(53kX|I{ zFHEt3B4(>M3mF6O`QM6UF;Hsh&Oc^-a3XqEB6|FQ`mK2rL9nsv|H`gc>gr)mBQ8_U zqTdc!v(xZ^Jj#J7QSU5I$c>!wBBq5&S1SbE2eAtmt{ zRZ`t%O7~e`L=IP~!HAi{3ZI&Nogoj=WYy#EG#$a#?^P9fKW=4MOn}a~pxOr-T(9kx zgF;o<;oSt!byz97PR3F8_Y3cYjW9gXJk~=8rb};wACL?_)8+t4pDku=C)b{XArn+2 zYE+5lk%+K4`fAJRqf{nZz{H0YTYyDsp6nYgrA+)i2+l$V;F8X1h=MN%OAn3@*QWadoc*|kkXa7J2|aL zr|<8?aGmY2+R+d^CTvv;Bvb+_MLxLzH4|Qe5(aMsnKb2eyw^EB2Q;fD-rZD7_ zn*;#yfg^ika71AbI2irt!ff{JDK8b}hQ8CKW~@y;a+TZ{I3by)arh2vKOa|ZirzoE zlafjyK;Y|NU0#2ZZ5HDrk_3W|Xtxu3-0Jdx3-lo-p_4#G=GBA<#XO^g(0`wn$*xwj zZU}k)OnRl~nIBZ8fe{?lZPQR8aj@`JqHLcX*9N}_hBU@oFueWB}vLhVLBuD$D*u)RQ>>RA<-6&YY>C(e(PI-3*Of z02-9s?N=5>ib%f$AUi0>(-0n~y0JL%+JqE6e^o?zgKS-eRlBEB0Dt5f9oLGnUx+s} zlGF@^m`L;-HX0F+utVx1YlU`KFK?138=_1w$g;(u&3&{*IRaUH@n^fv6AMfM0XR$q7&tkjchVO_o z#_c;Ofch~uGFcA6ele;^ur@UxQaK8K^wBE1E5%QdMZ)+N#8O0)UouljfYFfW!ZzBZ z4;05Gr~#cG>IF8Bq{Bdb5+44OAnbZt zNyec35G_R~E*1^;+p6Uw7Q8cSOO#V7852oFGB(THATsP~^lUF2)gD@9$(P?;h|i>k zMGWq8khtj?8cMgQ2`SjWkG8SIPWA4GnhtGTJexDc_)`x5(ZK zK&;ZE=lI~^Ql`L5^zGp#12(6@rBn;i% z*9XzI>v6XeASqG$RYH9IEmT{_jjGPRRVI@lLPgW3OCzdLFU_Y5+g)E|+1W5e{?I?i zMecne)764(=tBiMw49)^?%T~%O9eXDi}AP5ljDd!Rk9g))8`fd0iRCWG*JRII+1sA zoF79w!MIAdL*QBvBaDKdE~;D0sHC>*p_>n#k!Zpu>Bc25OP9(%fr)>5@fOi&-qRGs z^WPL2z>!&fmCOB!h0|t%k1sTq!ejz9Qcz{FG+DzlbR?=QexxIq)fn0s<+7ul5|WZu zk*gfle)p7*+q%`}`hF=eIH3OJML?}$2 z33F#`e~0ZX=TL)hHp9FEkh8Smg;V*9UJZ#hlx|Gn{STR)k|vEJynr0?pq4G;AbX=B zc1`6+GO&cRJk%zlS!qrUp7IO!icaKxR9&3q1e#z-d(@+h*4VauEEXzhGMs%63aITl zm>kJ8M!3>>?}BTDa$ z_qIi&ClD7R&&pqfN<&F7wry7jG+K9P#~s1~K98-KSV;&1)?ZX8qCxqbG^|W9iySzx z-YSAN*lDHLZz<62hu1#o!h*pyi@W1@u%NWv9LCb7+N(83HJAz1wI|kImo4__`dL-c zF|@`Z>4YADIV+k1@nxSiwsy8?E5aqi*Y?L?cnja7=ZGgkTn$4e7xY8EO`XES?`a;I zSCDB4LwBEO8}(bsloTC&_j_M9*6gj|)sQG(6mlAr1%mpKTu@Sz{f;^rH?1(hG`y;RX^N zb{l}h6yjYhRH8F&M7b4+*xqf-Te0b%olbL$L}Jmlu*DJ)_6Z%Zrklg*{l%*Ucz3+@ z28tN6r_lt6JmH8e!p|S=^;cwsPO!!{k+S3HK`~`a_UH*cfA@2~-F&|C1Un!0>#`?f z!Fm*3#y|IkW3w=`%IP`JYiAOvt(>*Ah&6AmXSfn}URL61F1<^=@IfXESMO>VkRo44 z-4TJuw%txKMcpKq=V0(B8(;c(x#dFh*3nv|hjS6t_dw+-g9RMRF&d+(zaV7~?Zd3G zEX&J1xsN1ef*o5T?+Rg-zWs@n5^DNKMn#;k=5MoUggHo+igYT(;8Xe-yh>~N}czQ%_x6&*;3ZK=oUX_x&JocAucZKgB> zo30H>X)^p|fcZQqi1{m-d0lHgTCV_lQX^Aa0iw+=m&-Id-b`V`$eR2^+ zLS-gCuw^7$F*VQ-wSolQF5!iU!X2BN%XkPb`pJ2ZD9w0Zm(2}~aSz(GmXwGSjAu)o z|A(W5BA-N2s0eVKZK8i0htl?ZseYbps`~fd3&R&g!`pf@A#__}9iKckI+mu&lE3PB zO818i4Qespy|F+xUM&WKa>dY^8U>TE$5Pr!bNCaiQH*BOjmETy#BfxVNyb@xJl``? znCuC7MIIo2YGu0JwXEWG(t%TyNnz01C+Bn&@GTbRqe|o$WL}_I_T8HMu7K3l6^u9O5+D+uS7h-nT$oh0<4iI%^ z?IfS`+5w>kFb9v{1KTGYIpKV?ALq1m?hyFvb98rEkADN+-ksg*?OOPD!8NT+@#>Ep z*7z8l!3xYrsn{;)Y}L*0^mw0Jfli*$T|Bl=2U*KCm={1 zM=f|`ntT8!>0d+lMPA7K08{P?Q@%hkI<3U%a0?ieZvpHU9IKIa+a#rd5MG+T!hglj zm^6)#KMXWeavGUGVMpi&XJb}9z^QFtxdEnKYoM8+wXgbM5)WAcIJuhKwB_%3l_X4g z4Q>1D;`;tg*)}?%y6`bi?NCn~b#n?Nj)epEg$Kk@OFksht$JB8@+w!`poQr@oV9Mx z_phfmndc{jz7Fq=Y0+bjQwahOh&}(HSfZkcQI2I9yP=v#F(A6+}Ju@ z(HTe)<%JCC>crkC?iWKg1B|}b5UmgQ{FPhsY}i2;8&4HKBrX@FOgkzqx%O58@=|s# zheh}+0IxX4i!PU5|0ephT@udU-J53wd<<)25mSGm%h}fZ`EAzHcwUn1G`LtaTh>?A zKydK+S5=wC!})FvE{P|z@VYddJHzUM{^l}K&SZ<&wKX_HD|hs_{8DmgI5T*P%mgL~$w?7h?H$pmuzzawE7M~%vX<~dQ< z2VG9?!4}GGt-^EAoV&Bo8rq8X&TX4+7)}@eSf)kX*IqPycZ{x(jm*CYbA`~X0DM2B zMq}|(K3T$$j@~uIov(GSv%jWm)uo|3{z~@=66bH|Bkr`r@@FUIf+Fmg)3;;#O> zxe^g>k&?`z=Zgns0Wb`Rncl|+?Wj;hW*Hnr# zHA`4O~VWAC?2%MpPbdEa@l6PZ&*b`d!o0&=#(!#GdbWw^Ur%SwMOLf$%F{-rs zBt1E!^!!Xse{ST`Aum>zqIspVthu%_O#54k>Eu95FQf^x6~bL6?lfpS6uFsCiBcHh z!=RJ)$HpYKMV6}Gxh-7<&VP=h;&qu^E`u9$vNmUr5=4ItebS{rWkxegp$0-wkGehd z6wfBtXP%vtU4IWBvaMzi({E*_Wl^^KrwulhiYQed=-I8Ok%9FcR;)~#D<|l8|KOWe zgM@I%(ZwrOnTt#zR^%lOAoZBDoG(cNWv3C9OWqlVxql<4yz@wTYCkZKna=@tE(V;g zJJ;zA4zFl zN~8whJ)9&1OEmEw6YHYuIHwfAYE{hPSPKg++~pG_iBc6m_mr3rYo$MHa6w3mqRqZC z$7lk&`lO9!o(g|r3`1_4+iM~zMd69$$d{ukgZGk;&tIxTw-}MP9g^0bWwzx;SL`Rg zch)WTuFS&uufCA6)r|uF8%1OM|0xcQv;YN7BU$_fyq8 zvZI5`+cA625i>7BI-<@lzA23N!J<^IG(smy{OkB%dX~u6&Fow}Uw?DyM~t6-A9REv80IAD6}z zX`G?;?zHY72rIMwL-F0IVO2dO+O17vNko6%e0qKi&6C9e*cP6otDO96c`HU|^5fzw z)4A~BN7fyg5(Fj)5jSWtYYmOwy zD<%Nv)Y(lP(QyfS3R;=XE9lV;RKjB#=SN4ebtQqRca7YUSYF;eY| zrYLQFC6Z1TEcL9I;|ZQWmHwdAa!L_P1n;RWl9zWk)q7rsgAKg!~Q z?n?5{#Xd-+OjoSFnl>}j^tpvLu+d;nC%N@9w;Mvi0?O8d-IMJFHK{9_S+$m%#SMIs z%D$b+b#Pht9;N*5o=ZX#L-QE3GUp<`13ku+?Xq}&)Rfchonu@WG1^@4&?94p!4YVc zRfF&hUF`czGnq?usjz+wB(*J`X}0&i%Kvoh3jzmaw?YsF@6&MMDEC^VL8AkDs*tj( z?B#_c(OEB5qAm)?d!eAyvD_s-oW5VrKP5UhT!)%prGzv-uWRZ9hP<#%sED5F&q`nm z1~DwX%6q_IRa*uBa%|F#NsvR@H?M+OzPxA29rfuW5q~7iy1`si0iwSV9EJ@&kofN7 zbgBQYmb9BFYe>aW38wS~7&!BQX*K@Uc(}kCtQ!;fBd4i_q~X${MR9s)8hJkgpC`|x zUW#--K0){MPk}PC71`4RpPT7r_Is#gw9>ONnaas+bs3Tp+ogG0;5HZS@O5~%PEY%; zyS#SnvrB3Y{r;+_C(>iM`6A1fyj2 z@WNLT*eYv2gh3iQ*2q|VI5IW}4tnuS2&WbMc+R|`L7Jm_=_TQf`K@*SplwX z@T5p-2pl=|pJ$-q1(zZJ8YsK8MD1)$?EU%dDhKwEH5Gec7V{+UkcIXgd%3`z`%uE0$<@}MA2;YK*LrQ9aEDon_`8i>U z(}gHf-eJQz5&PJ8vJ-Lj9`TKUbih6l71gbT+~2!eRj2te4~!i+0ryN+(}Wh>^Rf6D z+ffbDvU})Z!bBVmjYb@Rqn=}uIv2Wmq(w2gqTVbpaw)UF5~MJtFWKRIk9tK&Q+>b! z^)NF8aCb~$33lebVqTb%;l74p^xIv&KiBtOj7F{^=3W-FqY3W+^bfOL)e zG>mY8S>bjNv_r2*jyfZUK0$dX9q#~#3@ndzMmv48TLt3R^nC+Wkw~R*0|T*N5Q8H1 zv%-hx{@QavSP%!QME5>jn>YRJq*8i`vb`aVB|<&f%yI%kmEbRcn}Fj$!lko>N(F*4&YG|h+yxco#0{p z^@c}JyQY2UA0s&UD|O{)W(qX6F0PdtiX^t5@m{bDJGMdqa*iGDK)3FPh$uXNNf3L= zDzmb;)`Ja^kPvW*IauWorIZKVAj+d@>Xod5Xn4|b<3XngFcVByN3+!e962t?;|)6r zt1lZiaGGs;O(_&CKR{0PQMP;agJngwi9Iu)^#1>H)&X`HW)x7 zuWj{GT;bn3V}XCY3b?_N=EMrD_#d~BH~QHy@ItY+2Dg2br5aDv2voTOC>QT7ffTv1 z2uEZHLqN&%RxYTrQjZYI5K4^AyB>6ettUsykq-M|W#S6N(;qufP0CeHcEG$W!C1Z_ zVd$F5!VJX(5oEB@L><-NFMwnKl)?Df9&998tH=wwflElN4git>q~Unml=ZbQwici< z@t2PD;_*@;EOHY4nr%DJxP-s;;jZdly6g4vZd1PahH}$A@>j*(F^QxOBlUegWnudG zQj*Cyzj1Zol!U?fGQ_$?8y(M;IQR=Ofz-0l%(mE-T@ur&7(|kuXyHX=KPh|1h>aY) z;1bZlpXjDC7ErfLo@qVWNnVU|zevkgrYsV@OQ14CG~`tqt_c^;?_i*7ft3ziSZYEj zB0a@a0Hf=>=(jroole|{s03ceJA>4RNl<{kx5s!6>TzCDMRfV_(hT8&-)^2%IiewH z-6;nL59!}oK~1T>>iUtXfw7K%%#x&$I@aDczn>((p;B8-)CmFHtjGN&630L7pvnTb zY6Yowc%smhK;kkMAAr6HkA@;~DTu~$S|hW9q%$gwFA^);Nf^IZZ%ukm9uctqnBH&l zRzc6Y;&=+V8sKojLyeQ8hA}qn2Gk&LfFiFX+XvUvz-1(bkhw=yy&v%RkVFYDazP$) zc|}T25H>KLceKv-6D{2@f%(1pQAt*TDZrmyA7o;(<19z1qMkAJCtKCJG)&26dOmdB z9g?A#qYgenFu&s@Vd5~vXV{{X%fSfYIP*X+#lXsf;)ljR=4PtBQA6gJ0Q7{A7JSW( zj%W}Aog%&1oBh~=#*j}CP)y`U5VOao81%OiU`i*>uxWG zg4ELCCqj{xBW9rVW6Iqg60)iDB#0z2VO*2}ivUu%08{xi>4JL-2>fJiC_$MdnLMQk zbr7|eF#7~|ni41$1ZcAZe1Sj3ilZle@zZkh^6(2DTcvmF;#i4CAJ-Xjn-4@>=Ap>| zOF~M;yrNg!v61x3E-E&A6kFz9igafIgr3p*2w7ytpLMqs)&^s)p23DYBqH>{jCO^Z zW|usNo09Z#l0Z_S%Qs-D!a+#2RH7x(~VaxE7AjlE|2;*IkHeigm;ulf_ z@yA_qVt4MPO%T`IJ`|h)#b60N_VuO*L9Ff7NtB4tp>S9aTfi)2@b^0R^>)xq*T7K$ z%5Uyxstib&{Q(6Fo#DCE_d9xEo^V0e9OtnCD5j_i$lYOjhmnC>vy1kS-L(C*G>G^C zYV)XNXpw}OJB2)gF_B3v1MKgq^iOcf{MZcN0zDeZKHgKZ_R@2vK(WH3GIz7Xe#KS% zRzU{f+;2Lc?jQB(qK4!2NrH9efLM$ZAv)=t%4MH@TFBC-_zC`4>G(zUhwZYXJSo6) z*6ybV;3zQ~8AB^@Vk2;iqnK&LRhg_|~jIpih5sYn_-v+SeTD!ro);J`Qdtl5X2{8&vmvw}Ps+tKJE zIG8wG)WlAQ^0(1DJ@|Ttl%7t(Z-D8%T*D6*(iuX-;}5Z%&?aPX5FvtDmO>$8kw7|H z{pZvBw%HEiZMoTsIApSt_8%7pk%XXS{98JF(UUF0ku1VV=;ic57+^m#&`$Tu{LB`-9{d?>B&BPu8;|=a;mY2Ocnkv!RmE zPh-W2sm^2G4I8QEWI##=GWzZnw*+v3JWZO#ns!_!C9X2$HC>VUiRL;faFo%l0J!NT zKUC$l!>aj|kU!x}1iH2tDyK8(m8w7$pB8&X9FY%VtaQ)7D54#Q$ivymP3NL0XOTkE za#G|#bDXSmehx2G^*c;-FDPUx{$_xm4=f9fqAe=6mA~4~`JMk3XsgrOVlN6ryJ2m)SKJZl-VN%^y-+ah47O?BA!;27A10E!p6Fh;k^=i*hLTVoa*#JUPTPJ++ z&^eL@B6sQ^j#JBS&{+$fRLO!WQ*gS&$yNdV@UkrX#YlXSN7Z1QOb}+kzoB4vb7?2651H{|w2%(k{ z*$zPfH6KF<@(Bk=fcTf&6B14|>?`2XBdpa1ScF#H|4N>7?<$qiHWJ+Ye{gX@EO! zhjawnGsI!IIW{!_wm1=Om}!Dg!VT!YNe{X$BZdPnfE{OCJ8 zc=c1Qe=Q;BOa-WO2P;b@BrSh^M>N&2Kt{6 z8WNv;i*_q04{ejiJcNzWj*$Yg%dzwN%SVSrn2GJufVwH^8eB=&-5QR&qlb%Q*5_A(hj(1lkDIEY(hoziWk~`cofYb*;i|{15adNp4 z-ZRzH#p4$00;2G-Lydu+;foK^#oU6=()I|jjnXwiZi9uM{-&m(y(O3R-3_SPz)2O; zsyEV_zZ?}our*g@1aSEs8!V?=ka)|1mOU99UKzINU5@OoE%|mh7T0&67`d%LF$X-( zpu?d8XxT@PfxQ8xz(sH7s1%?iHdC+PM*Ny1?R4&0ep42~aSnm?sf>*I{Hhn7mfYzX zn3c_kh3OM1Z!@AVpgWvcS#t5Lx#-(_fb&m0dHB0qMOK~ITTo+w13b$wgrd?W% z60}f@3&O^_MLm25XonyHT7ZQz*bM&tfUhY92m-Y1ZG376eUTCH(V=i5v{uo~PbdH& zJ-|R!d2RueR zv*Ue6ywHnakKZC{o%ZyR#e%m(Z9@sAmQ=Z`5IB~{YfKL?-AUi94wH)gEtVOFYsH%r zH5Lz^rDkUWz}aDL6wC{Y!@G%d*E^&`0`y?L1kHDszf(Dn2Y5fCLFzT$g6wG!Op{kNGeQF^uZ(Hux%aPVle=+ zuq9QWC=oYyy?;sZda`zAN(2tcKOPpTJBPA*)pgy&^aG4I@I`|kqW)LRzYFhp^e*$f zOo)+$S6?03SU1WOmd!=T-G3ro)rZ0e_Y2Z>j>fU7WIdPnu{*ae-ZE#|JtI*4&Wqlo z^m1xNyN3{k5?jYk{Twj%O!5rJG-p1oMD4(}SpnWh+6kR>m9oM;k(#g)D>c)|77?q@wG`2y?WUfzN z^7LugO~{(RF0=eC3Y%FBsmfimKp)znGNdR!HHn$VBEq4Qn@5GDuM$lOJo;Bf79{51 zL&yU_@;Fh|=a8DL_dcn9$)bqlPUl6YZG7x0vvOCQ)A==M7C04Hb-zAU@na1aUH^1` zUc2sX__498&~K!G8>t?Bj%`%k`2xzK*M%cVelEQ_`kq^0t^|huSOmxwb81c2avhgoF`wVtwq#jwb($1}kFVHj}TgP8Jo)!u3j*SMkPi|!Kp6>ewNVxZ38n60*Lxqjlf z%;@9UYk`ufg>I?3I6uB+`tIxhV`g)FySpAOVUuU#{cPZE|18NSl|W@Xo2<%q%{1|h zb@~?lBuaVm*QBA9$&{ksX34@zwd1{^@e{|=aKa<~ zxo)m?TUsfo#mZ(38=j0hM&|1vQI+Zp0YlK+VbgwB?9{jTGd?alht|hVsgzU5|B8FD3y1~Cxr@@u zYqEeLPG80R?*7`#5bfJcbZws8+?~`pmy)5l&XGDbDdhJD<>BbCbHyS`1+!HlRH%+% zypD7QSTz;a6}+E%TGuMJ_om;5|L@#-n&iqMO#e$V&NRcuUmQ!$h?yyt-B zd`jyl;T^u(+h5>eVcQAVG|O8LguY}H#zIvA%u27VbH9WZ)svCmeg6&GIm(%wNgY39 zU)gF19fT&(ug*UYU6&=h?O1BV9=*;&GXu%n1sroEZtVqbd}V?c<PKJJX zQcl~|S51+J9;KQt(l_??uTC*f;crgH?r*9Ney?owBW$8SI9-0{oYHHeDW;28NkvPl^*Wg`d$3CsGaX zPyvz}Au4j8FVST3M5I~gPNEgA&o9mKqF54&A~9Ef&|elDZ-{w2IywFUrh)ec(fhM7 z@%%7GR_n7{rR~Gy67a)-Gfgz5@rXa`@zTeUfV02QbI;5T!kdlcG&$?rzBD-?YqEje zvcvrzagN@sLQoFqg89$MtXFcNoMb5Z@4KsYSN-)5k;c77zmM_Rm1zHhsjM!gL%Use zi&6PYhK;V+Hq*=%HOMB9jQkjmmUgni@X55|PO zg{@LHdM!nx7;Ho~ms46<>Nq!(3IKNkdPNm;n=^Rr55=mdZ`B4JCUeBes5tP|hN!8l z6pZedbEQV1X%ykO+VEW=(S5ZQ+Jtgj1Q4pbjxPjIT8JkXDK@XJSc94*6L&~CcJ^+o zZ;^inXrkaa2bzT~@AvVp{~b@l+iEdgx!Q0_QvvMATp8yK0J$nU5sDIErtSuVE*A&@ zOmCH>9xE~ZMH(crQ}!DopwM8V&@`FR@g2Wx{U#p!VuHgui96y0k8WF4TM;OxO4>YE&mowdrw$i{fgNC*^ z8P7{EuwQ-TM~a3KOhWtKrT-c=Th-B^jA7vauXXCAV-H?!aQskUwp-=^3Jy zc4<$k>OqFG#b$;Y&!|R19sTtVF|!0o4mt##ecSOrU`3+cVz*a@$wA0lBjTERRh(M*l%(;% z(sM;Lza$suCT4Nuy`1}s)#gaLW=BI*K~C_Dz7f>A$tN9_kWgrTxoa2yX1Ss!g!>Jv!I*!~e}@3|Lv`4@miGWpBdy($K9bDo`VXg{!`}qS{EP+?Nz_1jgaFeU zkmVQ7g`DtUGz6`G&osT1sV|A!#9?RaLlo=Jg}&c?^!*ZJAZUj%{Tey))h{ePMdUsA zcG^t3a*ml9ssET6IJ{s6NkzC=2#SyQC{eQ76Y|3fXm@1$$oY|xQn#s2IN>(6hJj-bPUA35sL~Ej$2U15d|4ir2~>wK z#_gfal)gj4nUhE`-C|37=zTg;u2D^;5qMy~1NlUPKof5SkD!SWkKn`HoV4s@@}TaA z)bvM?$GG32mK{q&(k75Ag!9BUl#^~#EbztY*(d)&nU!eSW7afvZ#?m%S~mWI@iHF< zGrP6Sc*Sd-s=)Z?YH^)CNRckVzl1XiBh4cn_G~{-BpnRNsEo?b!k6OeDKtcE zForRh@7%DAC5oUgfK!+)4R^%}%1*LWbs3gJCZR02vxPA%u`yNz%lB=ka@#`-UOTc^ z`)Rgm-St6ICuACdg9j_ON7CTyPkyaYK_NMxF(A@%k3_>;4eHn!7Dl=+t;-Jos%m3FbSKzW5~!iYq&-LO|5_JKa07yT&o5y%hX0>QzYpIh3fyq@SB zdr5iq)Nd+SyxQ9Q>}wn?x)oZ)&cHQsPJlR*JnB-2$zq~lVJgEg-THJ#;18ruHWAmq zn{FV-vxMdPe3ofO*eci4xjMPH)}!#Et%x~3W59612l%*m-bDAiU86W8 z#Ijk{2!d1(#89yzJ5c3wUO^0Vsp=e}X$H#aKG!K8T#7pTSN15rr5GcO%{P<Zcz1IQc8vz<_i=}P zqCJwU{Gu>1$)7878mPUH+A2PolGCoPHArX-a2s=OYuxV}doVHL;C%Mm07-YGB#0hm zM^il8J>rb*ad&B0zu~L`ZD`y<=S-7bG9#6r9*%Wgda`w$P5|n~mbT`w-DL$R(aw(!u-q4lE7?!4m0z2fOQ&vhwaKmY=GeY(}|DWck=Y773xxx7@i#A zOHJ?16$=1tO4H4W{Mk zr*)8+_my=5_d=G8`YF6jtRTxg{We`cXbg_Yh&TU>Qd!dgI&h3>Jsv1f-NqdNdinM1ZVt+;R~JK9Iyi1qje!jFtP^jHe(Bjb-J{RenbZJD&i^-6 z7Z+In`r0%?m%yq=^9*)JES{E|yKX5i!!ra?9Ff-x7#0XQF|yZ)^qqMzue>ro0JNPR zkJ9;i_jR{;r)m1z45c(w7qt+Mv6pV;WA~*#_4zb{?GD{C%O?cJBBu`wA-Fb0*FV0U z1@-d-1mqzyujIUd-}r5aI|Yx3)Hy1WxPpeB!9l7Vj~A}Z@}~t^7FQQ#=P;&Ck4ghQTXHNVmgRrFnlZ`)RtREG zAeWzudoiul4SH1V9G8E+xsGwLr}XSfK`X^2wM$T`>9=o!u_@fqnU}*bt^k$03;d*i z)1H#U?c4(^)gn@%J#c`#>&C4Wlf^PwnUhWXaCgF$9eng7p2`%MgmQF;Uf`Gk6Q?C32GF(~PNDDXx^`VuuK+0g?LO$ymzi-_c>*57j z9~;EX51hZ7`tR@K2aV+NO*s-y>=x|A^S?Nw-!Rt;j1Wu=rGyn!ZTmCX#yj- zPTMj8XE~V3DPD+>k%#-z-^yq*l(dKcIQ&%&!_wKmMPYw;ffe$K1shAgMM8FlX~sh! z+@kroS|Dx!2htp^ddw~IR8|#I4EnMAmvOa0`@Nf1L)qT^^|I82-(7hrLB)~g9fs_D z0_|Dm1y({8a{as{!x-h8YKA&#QB9R>2Uos_8963dxgQJe2OB+Hx+a^Z@~KCRX$0?(uEH$Kp7kU)Kx_^hZ#loH}hs^cYA= zwCy5M-aUNdd&4Mj3o-H?U{P9KvFMqL*YQuf>Bsk{y9U>pbT{Cq!%Z-pvyY(1$Q6yH zD`-(J-BBdk@wgbV;qpZaj6nj2$#3As7(VkXS-K}f6Xyb|6&SYn4t@Fnmd~{1GJ+Dh zZ%~I=%I$f$Wc@XqwZp+vTSmfEgm|){A#4jOyu&$2s7`rOLPk#ArtsU2bCAD)wzUMk zI<_d;|DHuW*6+Lsb<35+W_fdOKYVXD6wSwf5+FnEt*$vh`7XS)Zv9_&NZ7rqPf=y| zQf-9TU-hxnL7i6T*5uI_^Y#3W^^FjYenaH8qoMuCLcJNCIpn?3a1d-*x5=3V9|LJ(pd1EuedlVl){$n# z%rgE5Rdj&Z2w723x#zw&*A)E$!HnsP(<1?#Od&~HO%`ynR`zjG+ZpedJtitg`i=3E z`zP-3@cFcrM9iMzE`QDHe*gXSF?cSScD(C3`6B+`<)GX=?Of(^@^0f0Cr7&;UL*&> zc9)kkd-pdQcc>6SaT4nj(7Wzz{1A+t!DT_HLJrKlLv+3<*AmhAyliy#nLust4h}u3 z5EmX!v^DMw+03b`oV+aY?+ILAF8kg-aPS*{+Bq39y5w57A1XeT{ddxK&|3^)2LZL8l0%O%q9-BPP@Byx_+a#@4iMVn*mv0+Ok zUhY>99X>UJ1Il44&dhv{=*xyp4(j{K;ijO5wu)=0X0s{E;zJ0?@N^$y?fvigN*hX! zzc5GJ_M$`w-pd=H%7qi4vj>&R%9WE*G_8)Am$M0?!e?DsLX%VGxta6@;@2yd3K%4j zQ@E^7-qze5Uvrm2hOB=ZWzT^XAcHSj{1TM(1T(k5EEp@QJeTXCy0G$rvP4$abY(4@ zECrVfr<1AREB9gwr%32jm8Z2<43%BX7u0`meJlO?Bc>;(q@i3jG+3&G_6p{@4ZTACzE3LU!-^nTwU!D$uQI=v-k+-7+cXsXVA@K&9-(;74bcecyv) z$GE4fvO^oKws{{X**ttK5}AjYfM$RifhZe6c9%Ra^o>$FzyiwW4aHu}cb4hV{nbsy zK8uLVo4M5Kq|V^@w!TLn+*V$twml_@3)E+M$J0E564Wq+o%w-;fw;6Ir@&XNV0R)QkX zgr+owq_8quWEI}eZ7^NIn(GuK9c*d+M-0qZ8>RtbuBSTPbR0(Cp992hZRB zdzdpn8+abVy;k2^{*!6yXT`QoIUvbvhuY;RQ%`Nx#tM+b;DttfAPJ~WttOC?#o`K! zD4#X?qM`VO05<@I7DxVa^n2Vp&#%6<&I_6}vCi&7R)E?SCTt1ds)grYe<+5X?)snO zkF$@7uGPe;qNE!Ojg^1TO!TXs@1)QNL@@h`_3R3jBaaKj4fH45wgs*ih%~_Q59_{e zw~#%sfDmd?^}Yh)^Q4z+E11?0$9!a!*7$tzTX!guII3JA%KUPSP1TcjMQJjVH2z>1 z@-`x82wpE3H<$-m7(5K^MB|`Qvj|S29Y%>h#xyh$4M?%%c&p}IyO4rJ;?OnVZk>f6 zln{tdv>Zq@r_f4@6bf9hh-w1&_ud#VeHqo)UkkZcW*N9$$vMHsQt_w1m+Jb*aDYF^ zq-F|`esgZ))vs|QfnWcE!i=LuBrs&9P>JXLt26^Ka0mXMM!R%FL#!zyfQ|n@X$c4< zh%m=;bD2|j-@;2^!r6@q?lL$D{$Hity#Z7)X92b-GR!4J9>TG%isK_%2pE5SF(^Q8 zrOgYerZ+bWbEb$w2q2C-0ZSQbosh5Zem1Q&fWsmxYQ``0LV3m-aWkVRFe;<&Mm)S_ z7scKcy=NpVJ>W064FQ0>Ci#n?C&*RLHv41W&_}rf#Q;gBTAHc7x~)_W!{}Eh`z*ApEIqb6Z+Ke1+fDpW#zjY z8}wbFdp2P{*KjQQ`}0=Y@i>^1TJ+6mq14%NYvInDi>I5f1&B-2u0l0`jcq%_0UsTR zy`1246Zh|0pRkNt<$*M@lZT%9f!Cg;fmOrqZmnPuBREjPZH+$HdQs}YxB(d%J+DTK z^Q$Z3p1+Xpjis7sb2~v@?B!LDXqS%5Rk!8L7MW!9as+dNzmX|`Ob-8&M}gA<_K}I% z{0H;n827&L3|MPWG!=pgVUWR!kIxEI$((-#0q~D%s24%plDSbij(CrrEu0|}$_PqW z#PrC;%=}sPckDpi^)zB+TTFlN=p}FW)tIb};R4F_s(ykJghXEV3VVZJnU_9@g73?E z$s0SQSS#G2`$Ccjp$!B5iKzu!5Q1UNV{w(WALq#5AqkUnE3yJBdTOc|xR@;ofPvmG zY}<9d^>VvvR(`hL=^u}W)0H|E9h_>a8{q-6*i&F;Z|v)kZX#b=XxU)jH0J{iAV z#qtG`Xkyg`)9nR7Kf8Jhn~q@I7SCC(<_xjZM)Yu>LJ49!9@ zb^H%-q)6(LiN)LsXJ*4ul2d{_A*c3CNEe*|`f6C@`lVqDtr13S{Zof#Z6{JUO7<$+ z(r!_FHPXNBLu55uApj26Hhe@pC!)l*BI}g3Jw&sLMDD zbNV&RBkgNw0rp5&>*dB5k5@2_a75iQxuTMjd{YeOzstMp0+@?e}ljw z$HmY~sHsg7Db_u0pN*>uNH^qqA>v1!{M3gHwV|FM<6rGAQ7kb#Bp*_1KR=y!%e0S5 zYcU<8tRmJIFRHQ#iv}8wsf*tQh1iUP+W(Y%!;QbJu;>()Rk3epLE1LcZF*dcyg>Gm z^jnZ0rs?F(>0%<3sJbi`)dRV}0WyiYVBP>}kn7q2F)2}NV-yr9Q7eJ8l&Ff>BaSuw z1txrEk*Y@b2RMMcuS#&z=D4wTKTtJ;KQQx)HU+$iy+C(Vy!5OfGtGCkzy(z7H)(w{ zU$z5CwueA|@fSDpwWo^T00j<7pcVW)m6yuQkiR^T&p8>T%r<+7X0y%+xo({l3H4pn zy(c3AVthmPSR_I6b1S-eGzhCez>o_?(hFKqMA7bn2v8nYI~Z9ufMNm%n6;r_HK*G% z4gvIY-z#emnvq9=IyUBo<-^}(c-jZ=4{$_zs@cmFf^5he_TpcJjPO3LR1hZAjHN*; zKsCmxq>9Mnk8g}ff*z`<1g#dFxe%p)iWRU~+ydVb&QR5TON>oOqlgBJnIm`*E@zNg zH|5PKjk&fDXgh~3KeX`smE2^|2p8;-zuvn9?Qw~10||Q0;47ei+x8o*eQa!1QoP&e zS1}xA)5Bp~cWFrwlP#d?f1j1)xOJ1f1aOMlze$-ZHplw0k4 zHU*}SuZXgPYZNLZtc}RTN?~y)txCdjl3zo?6RT`(6Yj)HT0Z}LWaV6z%{29LRZwi-y$gMn{hW;aeRDq5mnw#A_z#?qJ0HPSiuzb zkWsUOa@Fe;diM0z``g*ssMi}65JiS+wpF(36V*mB$4rD5{H>ZFx9t^9m&j_PJ+d4{ zLb!g%7ebg8h5JMNj@vtr#JnZ0tWCh`FNaP`>_}T#HW}dc9DgHl&+hVUxliCUKn*Gk zJ#5HY({=O+zI$jwh{I~j-I>^|6Sv%)1=f6qu`4YX|E+Dyj|T0WP!*v-9} zp24O(5zQ%bJ?+dGc2p2S1i#X$%voO#De7q7kW)9PljI!H>9Pn<+jecTBIPM#nwgj3 zyd(p7+4v>>&?0+X^Jf?vQaMS^g%^krouhur*3;6FhnuP}BL=t_xy-PLf*5&m`?{0Z zFvVzU-e*f?Rl`pN8Tq$u`;h3k^5ed)l4Z$x;qIf_EFxF*Uy9LRH7BgDtnY##i~<16=MivAk5 zq0tj5b%e`EE+d0kO-eTX`G=0=Sf=~nM)qCW(UfSAvM3JfNB6BJVcK)Y+P+E%d-@u(xXVfat0kqw!wMlbi*O}A%cXF7 zJp{vvs=}wtB_iVOnnI&^p3*KlWLI`P?5J#1U=j@&{Ui_i#0?!a4WuP7N`_9rl%kML~>=tlTGGi<@3cmOfouA zbeT(R*AQu6zlS*rLmMJ^M8qQsk@L*Gufo_1MHNR2(z?Vt7FYnuK%$hiy z=)~pxfp(m(K7DlVVN!Bw*ZSe}OlBK}vV$3P8zS%$t?id9r*oq=i=rb{&?dx^j>KMy z*sR{ze7EFLP$D9Gj8n-dvm-)7Aj0B+JvS=|%w4jm@HUx`uO*Sx+I_KZl-;UGH7fyY zzL;Iv-4FZpcvSXO(A()L#@B&;VRU1Lo-#~^SF8@^#jTrzTBhfH9)6~0=BN(732DYN zyZJjD->~~DuiKVlPfp$e{IKNcq+sZfu;+oYjm}7nMxkA_>AdQZn&Wm4jO3c*T+WUQ z4@6$T22l>yF>kJM>o?knRR}_sqhfvIU2?K{aCvvJ;c$o&?kkjfbo=U>_-8kxN&>e;`FGs7!69%io>m+%{Rc%5S9@c^vPgSlMh9h< zQL;)Z;eA4TN>l5c3jA60Ov=cibw(#~EWHmN$#co&o!K#(^rC>X=-*(S?4ur!zVoxU zc$c|Sp?6K!4_Mylh|^_MY|H@JX>vD+$&Xs62h;I)N!^%NB7J_@jST);b++_Sp)X>U zqwed7J2l6L>TqiMa&9sBYuj$EL%APVNj2K&7PjN6JVEg#bIa#>xeV#S^1MY0@C;2> zq>C;1cY%W@-1qB;Xd%2$K8*^Nk>)wtL~xR4w{TR>o^kJ;@p5A!uXK>%&nPM3RS}@8 zSaCG5V1WSNL-wY18PYe2d&w}nr%X&eLt0F%qIpAz8<0qdwq6v=d1yjo#mh#R))dKsPPN(B7XI_al{IqV z@n`4gq`P>hFjnfCF?eOUpI+nu_V8~%ncpa!#X~%u+S%=3f7iBm|Ku(tJtOxq?8SzD zf))bVqW#z%95+$%qr;^Muk57FA&}7YCsE`qT}va!n0Ld2FW^?oB1tHRG(7;NKT0OMaQLU`oM%15ypQUF~zD~ZkgOG)# z##{SlK5su?r#q!6mh?7R6Y~{)`A?M3sSZ(2q8Zs^htby6k^NPbIlrnrP^m z6y|t#-80H$*hhe z^*0#oAkmy@>AHHM**5I$+xx|gc$0Y6U70b?m+{WL^5d|>*tW}L{bHN4W%doAU1*Y6WtM0$2N{ZA;In;h%XO2X0aTDgtI_Uvm1Y{X`u&6+^k+3I zuKG=`+YHJ&Im_^jjYDeM6$U+2{3)hsOVMQb3m$JC?a=lLbl!7|`cZC$cYK*fV%4?E?Kvy)+jMJ$&&Pv>*Mwqmmg@S;r#w%$H>y zm(hky7o423sNk9$8xIa*9z4OO%q_iJ?k42+WBk?ddV^caW3P}?(gD83tz`zO0H@^h z(`!)NT&V{7F0hqi=11~O*6j;9t7`wl(UwzgpJZVb>gS%eQ4}H0Z#2d6tmhYJf2iuQ z?H_51h+ofU`g}wB%vr_Zr285$wtH%fYMD+y)UfzWP$Ey$u69>L{}OoxasaSk*?A@3Sm45s&0e&p zIu{6lgd6H2DFnFedcxD1C=IKa85D`n5~-^$?e?a>FWF+fUK2u$)p+6!o$1XI{V3Q$sOZ+ zlq>`C0fu~YxNp)>>MBDyz9MmrHqr&efiq`gS8sag6!=!bfegF@DzC}zu_Vjm{`V}w zjG(;WFBYu1?<(IHU8O(fd=^4Z4mfxoT3hUNqA%zLwW(wU#D>;Klm~iGbc`&@`7L2J zIt&Ra0U#Si8G2=pyWxzNQJez`am-m9Jo4?g8lftJP>px9hGChxbl}2@+p|LolNBnT4hT3^;^J z#357l8{zF_ebEpB3(!Wmy<@E~NpuO=_UrTJhO#6H(SG;OPV*BnsqP|X)V~n@A5(zD zv6LfoNXbfApjdZ_-P!;JQhdSqd`E$#q>rdX3+b+$XiO_O(pu?KsHjn2ZM7Fg{4bVc z?e-P4yDWTTw90BpIYErfuSSu$Lh~R^hJG5V0qk8=Pp6YaaJCQJdi`uDo@Ii;{GxQQ zV|JBB7vUo=De(e}D0P2E)6;!9;NKnwJnBQ1;eA4C+wmMpg|bebR3lFC?&g|w&O zhhST&+yKrq!w_g6OmU#O#3-&g$AB1{sF;?v-(DgxLRP`Mfpc+6fOFU+pXd}K&O!hw zVt;>qQKWmnEH4EhTu*&?Cz>+4)B|n4EYqC)Rd|tH}G{$>~Siss**`=Tu51c2hCfMO6+$S zVP%n0AK`EhU1yJu>z&qU538h82`lFXhnPojy#dxlFSw%r(?K=yal&XSS5dtUap+js z>M(MDIKBk@jntQA*7}hhM}^MBzz&!+V?q^Z3?@r6u6O#NODh-MXs;JH7cD|_hLVtv zZ1Sd$*{71o#o)04+k3Zzu`3!@((BcYW+}5 z0c8xl{rUxhpnLpae%>n|UXKZ!qsSy5ifd0v;^ ztHDbeBfZe zeZOKA6;rN2P}vnca-)aZlr$1~3J5*!3{9j^wek5w_S~2bOaI~ty6!L-o!XJVM?PXD zQBWf4#|{hWzjab>T|KbZR>XkN^}QEqhJW-0lE|0uG4GScz@TXz#NXK!p(uQ3YCIJL zj^znH4&YjmmHSund469QqzRH%97_(WuU`#DYbFUX>mLo36?DZfm~Abnew_+%%X9p#(4Hou&_}Tr%HU1> zC>U?(eq(*wZv2;JfTIr4>syl`JWo-|>?y>778R{g3;QC zDop^0%@*Z)QPpk1rVqYWgkQV|^lTp86N8z;dLy=u zRz0?RNV9=2dI_fCoum>kvvY0bqOjAD)fBPz=1$ar!Bd0I$VB!%1yMhV?@t7md!&*Z zHp#mC(lvMmbXVQ5i4dOtAgBS@tTHv1aP~11UK{D;=I*c*<$;TON|71l}UinPR6bJQ|UBld=uFn$sAm@M*1j*&2QCZ;i8*yGK+}#WDB3s!T6c z5xuKApHg>-&>t3d7^U1D_!8noZWHD+hEMNUMEH@kPuNGiPjMcCi&7u(g?+?c3=Tz% z4jY?}PknDJO&T1{6B7@kr{9u|qL$yy$4+N+qLXN*Y|>3Gv^|Lyj-J8BHEDT0w1^AQ z2&;n}VOi_xuBYLE139+kd#=-3C4yQC_bWHsarINutxC(s@vI`mIfo?SYx3oXZ*HK0{Vz1;IeO+U20=gRC?d z>(VI%^LqKsxF;{F;)Ua7)koJB_?ZHk?khaKUP>8G{ySgL@IgT z)rv@6{DMCH2=2o5!BT$q@iQe6rLU1T;$Z&v)s5a)O1 z&F^ym@7X$vTYvJCA)bU2O_@1bNxnd%^k-Gd^r1L$c83$u<0ind-U{(I zB1r4VLi*j?$XbTB5*(`xz7$WiE1$?BV{A4h`;;C2s(d*(hRTiCudv;34Jul*#gCuW z&lcb^g&twquOtQXu_R*igGSQt>uPY_ei3p8 zyqM4DazYbXSCkS%t`)v`feO>tyOqa{%s?IC9kphz`^uT-4&!kMBb%I9o=CkhM{z4m zcr9_LeYpfjRPI<$nv(m5u3c^LzA~mmb5E!Yw5X0ayyWm<*K@JmHLo+rOS6xC z!sHH_C{L&0UGv}_{dC?kEM*Qq8m{WhE57_3Rxmo@6_gN@4er9%RbqRm0jeW8JW67q zS+}&Rrk$0q-fIXr0|*BOB;}uIy`xg2D0ak1EIPVu$+NhXX43Nb9Q29gC_herUP#qT zrts}yhUxS)Px8f2hxiOZUpMpW81Mw(U9qpAVAkQEA6wU_lw*hNyxBn*xCUhW&ERjz zU{A^hAX%g=0_rmw!{fnkt;sk**j!ajL>dSKZNuo%b!ByI=&;wX#FtJxS4LX9YLXlX zINyzXYNI_ZjHZc3ryQhsBJdf*a~aVbkUhZ}-jZV1Av}VJ62nCRHe;c8S~KhW%hAn( zgT(dkNGG5nR4w+MMr(h9~WTEDjJQlax+v#0*;-h|7TlZ5E z*La`@j6of+AA|bW%luI61PeYn%McbE0;iJ*_|dqc4qNe+OU=I8@peph|=`{)9bu ztg1N%=jL_g{Y(W7P7bv}+&k(cY;`8=m>#1%XB_+{CojT{AD!V%in^n=8upSrMM#z? z%Zb+Tty&w+oI!qQ%O`jlFKDCe@LOY>WZ-(@Mjc;mf!r#A-N)JJpg{!Ln}%-XG3%7( zt{>ePUM+Wu7o{8lqc!8R`%1!pa&(L)&SU{Tco3VN1{L_dcwUG zr!^=WJ$Nj!C$nXN!gZyR06NWVp2z+xq5P@!%_>14yTNZCg_AXPL88@M=qpBh@8%v2 zo}K#Z4rYo_^%iDY3yc=Ut~3`~O#d6Ql)2WsP?Aec81^(Ja`)KL3DGg`8uksY*K}xp zc;|j_{K?$&V0$xfzoKhea}>yp;X<^DeUs&YS-}>ch;!d*6ciQy8{zHz0+Wot-j@#IkxNp{aX4 z5y60Cai|WbV9sKEv3Cpd2WV99rB2@r4PP#3ah;_f1V>PFEg?icQ}zZ>feh% z-v)lac&T3>o^*!~`UgF`_yhw&VYZ3eDN#n%$>r{?Z#F?9Gde}(oKjAw=@>WxAj@CAI zN8t8|gX~hBhc?Qz%!Iq-J>lVr+a7ract7sHn!Beoq_6h?p17`mpgUyuCa$`ASEbSq zo{Qjzf1o=eHC$$>+-RGA;%Y;9WXN!70Hnt>e-xe*LvmAbfs?i&_R-1l$s#OiwBSP7=_mvZ!C-PmfVLSd=rfDuI5kB2B z7=Dx7Xc}Ap_Suc^4cUal2!6uUQD$t(t8Gq39N7zDMAj57{|GAi4;GwT1Rp!{UD2Ac zs~u|WJXtbREi*Xd1d~^2*@e+8cd0MMjC{5C&(suL)aGm0_f}F3;e4KscN-j@VjNPu z>RnUWl@{80#;>HqNmI5sU>9j;nh#j=#a{ZHd0J9QYDQ9%=9yu{y@yK!I~?Mu+8L#) zt1;-NV!daJ>RmK$(^E-@a8jxGk{yV(RkN;*;~Nb%19F$&An~yWqC!_`pJJxK3`&~* z+Uixe@T)6tbb6Fh&m&PSt2S?_4|Y3xQI^N8wFl;np1|qEo&>Y1K}^z-mFshHvckwH z6s}M0x8BnQ4@LJkaI-{0>*Z3XxyRr2p9MBQn!MPRz5cY;q3L*9O785vZ6%F&k?$Nd zw8CjaBc;R91HIVccix*i&m0axb#W+W*B)s^OQB+t(z&7CTy3wd`NKH92*D4DKFKL3 z{0KLFLq0;!vj`4p$p;QX@9%?ej^=#En+_iS$0p`{to@Ye@lk7jnxD*(LGvZi$d-ky zw1o`@A0Z5DkotOPM^CPizf+7rJyIKgU&xApz;(8fPY>WIM3~XgesR zyXe+Sdx&5uKx9R`>03Wh$(t&(8@ktJBURRE2MQ3ai3vKMH(hJ$Pe|pT=MFsQRG;8S zq(&%IsFsfw@3~`TtI1PR{gAE5hs)Mjem@p6S9*hsY9uHs>Khd{v%;M_dnxAQRNOp9 z$Cs%Oifd<(^~IojrJihSM0Ch=Y=xRldF=}rq-2tU4 z+T85BWtBom6GQKNMPlrzIksL!35U1=+QvJfCbBDfP7R++H%amP!)@De+tH^ z0GRiZ;7mp_?IZD%64{HS0v2qyn;dqVFM@?sG3z7|B1|d*j=RMUl12qPk?oq?@X=k~ zf}``>IozlW`#n!LBE?|k2Os>Vh(7J`WY#(sKD>Rn)QZe4eaZ)@NA%5;70dT*?d6Ny zlu#lWrQ{}AIA=&#M9!H+)(6lp=bv$i&8s$Ph_#Ig%0Y(=JI(RAev9(L9p9PK8Vd{) zE8Ez)(T#q{l3;t+os^v+N>CP!6UFVGE<1{%FeoUM&hU9d2BBy{c%&%^1chG3%6JUqnyIHYgiZ<77q0;fsP&y#eRc2pM z7$Vnf`(6?Wzje#-s|f}Nj$Q`0#~mPH=fq1Q+p2g6_&l|W&7#zLgKJg~UW^;>xD7HL zhC6lMK0>lPsMK$GZJU!B%(@G}a1q+OF3SERURV$i%D4*1uPIdm z-W!x+V=QfVivnvj6m{tP*y^A=IV=FpgbzGVH<}ZJmTD1BHo2vVCn?aeZG&OdS+vgh z#qrsKLw}D0j1P{UP1>e+YomFf#xVkon2cy|7|z{k%tG+_!3|x%VtdQ$sIl&q83BJPNx+K$tr-Y5e+`Zy-~Yg&e;T z1nOKCEjbkVZ?@V-1QNeED8_jvgDi?G5!xnaz)~>;xgz2W`7kVcwJAIN*?ip;uhSHp zjlzY9CfApl31I{wRq!mcXo4`&CjhCx`K!U(2-7g^oROs2cxp8mb};*!UVRKn{ZrFE zdKUFJN=v)Ot_GZLY1dW4+h44q{E4RyQEbc+YDe@tLC2O2yTn9!?wXJWTl*yQPBs`x zZ|n)gWOk_x2XlciQV25 z#jBy^1<{HMhHVoufUCgAoqYO`()3)2UtNYgG4bby*O6B<)1%gt#Q>Hdi> z3kpAnDg8sW_gTOM(t8N$q9(o^Jt=dlaA>Rnmg!L{CBs3I)oC2vbC$d}`N$vgI#}n? z3(Uoh5<=s~K55bm|CkXia3rm(xJ}KT&h`-4My5*NJ=oWEq)w=$(daz@d{hzZRpeYW zqpp%mLU7~&qqpyQ8K`pB#tRxk1!rm%3g0e=700ds8_1C;X8ZP$$R6#R0II6N_>8~c z&r~tBSg1!He#pME%3Wk66d%HzZqp}6TZmyi40i1MD? z7(_d|*`1fF`Pfu)uR!VQ!IAusum|<&P(tMa_^cIyLc2o0l7# zM9#B=iz1B9X0l}wINEKq2o^1+Pg6;^tc9^`&J2C~_R;Iy9{lH<8o zF}oIkPJhaUFsq5ct!vAIiA}d&6rH8xXUgEywGXPQjf0*85;Tim57JGzRUeWRPF+zEO}-3){iT%yLU$7k(wYC zv4p@SIARsz%?dEJio^<37!$sQ1mkhko>US@n>Yu=`f3tU)L_y5QP@LPk!FjfErF9o zsZ7ZJ0D$PN9ZKu8Hi+i=H!+GDM~8m2uH}P7UPf->444r<1CaX8l7&LB5dS^Mka9If zAZ%pB@`@S4b7ZD2j>1ncKfBp&KNh0)XUu!XLArtq-A`4^n` z-i88VmSN9f=J4Mg1dVp1d)rEU`y*lP{GQIC&_UULTZFso@fY40?Jdiv>*<@mmWM>Q zgvqSCuro9vcG8so!A>%`8LW0i=96lsBs}DUw}O0;W^gO*hp=SQX+D`233Ns8gn7U* z3#dCUClyaH$;M*4o!37A!zFpUhS}McwwY=|q3w@gainTe8>s3c(zssZ@rW1(%y;r6 zkqYx(q3wH@sJhgl1O}C3MB3$c$MY)%^cCRMMFK_;u|~!dBpSA;00%Lh8D#?97nCnv zEuYOJXpcR2G&KFGm<~|j6(kvpQ?AQ7>Zoq-1HtV(p}WqWaa$S;C=+9$MSC_w4kytw zv%%TxBE3jWt`mZ|5*?>V{M8_Owtgpz%C&!#5A;ql5&rxfL>Jhx_GG6I7@`8#^V7lC zwIoTnA4$)aMStF;pK57}sMKxGOMwhDjZmk4o+T>#rjxGvBPh!*Tk`eiO3Yh9I1{%-;y$9pN|sP(YVx2jyMfOkID!7~X9M>e6W zZ#~y}VZn6yt;kih)Z-E}hyIgI0uO&kU`bubn@}J8X!@Qq=!J({Q(RSLp7POUIFnUu;5nP;Uw+M^@eFI3w<3XD=*bLrE|F%_&gd zGUw`RfyWO%^q9h0pHmMCoX+Co8${Vs9@+9&ucqe1nu`Y7rhbNA}sSjcvY6f=pVS>PjgeUGopJ5;bYS?tQ< z0BIW%2S;EFnOD@E*S09P03wmU;23zO;mTfkR;U+A^>a;(V@%Tlwbkb6Pco!d)XP1& z6m#~)r|`0Vp2s=l1yn!fctWng`aCzFN!;gPJ) zJPo)P*i`!;VfQG>1n>pjM^Yg%!l;AWme$c2VEc4|D%*shbs_Q7!_j)6L!D(T@fdUh z-$m0mk7#m>H$cNDsb1(7+wPax5Iayh?nJ08I~6&=gtD(b8?LCJkOVTyocoK1Kd+NO z>+*ujG`k?(dlP6~xGLSHZplV=brmriN)JN5h6f*~-*Y0}`f2-%6jCRfTr&So)Xmdw zvaS=ZnSXKGb*$QBS4LSpGp06*vVg5&k$jOJrMhCCW`*BwY#B!hZ zwL!b zWO|=LdHHl4j~33Y+E(fjRw=Z8F2&{j)j{RX5pDw<_QtQndH473%jSl=e>>r~#wrd9 zvnYaJ`}y;)McU<3lWm^6S1InpOnRVuK+N4zu)+m%3b$)IgL&Wl@MB|-7&_yQViEPT z$2GPtpG{a%3{8j=7y6vw&FNMKi}x3OpZi`7><8T}wpNqAf#hijw%_*ec zY}0OZi?Woxv3v0W%LLN+QwRC=2%o8v)Tq==>2Q3J3}e)fIpyR}p79-$%h+B72)j1oKZwP*4BI5g#}Ya}8V@Lb-Q6&1EHa1wr`N$q67E zJ>tQ&U;NJ~OUV;)p0o5H>IO4?Wom%08z!x#tTIw{gLC8ZQk~TI>_$hS1Q?m42TNrD zbpY~x=!{z{nw+%=z`cHy7k432&4jwhnW~of2CSX?#erWyTesiI|HP8m{y$4%Wn^Xl zuS$>h*pi98mp7YNL2a@?HA=_ai~UCX;Tx%&@G+!UQ~BNu6bYmaHi;#aiELZHZh^g5 zjt%{}C-wJZ)l^c=qs(krc1@KJn3(s&=!oZiauY&wQeX7nyxO=aNm)gGc#>Y+lK#_b zNuz1v*sT^6qACIHB|Gp1OS1B1gM*LkLv_d^P(m*$* zEGnTdd;hnb_Vi*{R0zM6KU+I{7?4DqNv#k&7gn9#~GaDCb}pD z;)!?j`=+JFug9mg!Tge^1I=wW_UEUy0h5w{>qioO($u-?lcd$#)@v3p2i+BIx>@|< zj`QrU2h)x3461;B-jo-%(%XBA5mvmjMcDVv=R1eUG-z^XM(p~{leP(PQWx`N$*-Y7 z!)3(LiGh zvazO`BG~8kdCix*+gF`Xa*ltYTFK?gpbrbDsQ+o~&q-;-vTFom0iT5;|ya8Kq&wG-e~jH zxNquL5Y7afFP%uny!~a-z}W~~vj3QG_E3^FEQg3h5nZLL>b9ZH6|xsKKNJYrhNmg% zHcd^0?|xLDj!wg?)=+sy@-9T;RKh~VwS21z+70%RQ?Kfj&M96t3Jg=g(n6{LZ3#B^ z(Q8sM>Nh=VFxawmXqO@Hv})5;y=09skQ8~wxwd;0UcuvE!RwD%c-?T;b5)KKFBM>- zh(vHn268jFSK0fM+KyW@M?jlBg0d5=MLn-adx`4DNTKdHPsh)4nYDxx)_zh71!kxI z-hy05JvZVP&jkx$bx%-XsI@BrYnXD1UROYx&5U(`|NBUDYf~LFf~F)K*UX$8{!CJB z;!lv_)}g(hy)xj|9bey>j^AQY2QN95H8_MPs^8d7qlNTkFV-aqG?c(AVD)t1+Ev5u zIbNt)mIozc<=VBmg4;_m&PQh^+mky_Fo1z^)W1VgUd8Ct&Hm49ZC8bz8SX}9;scm2GC~=YOPfNXbjn-B0(hF_OD(98xO15 z$J&mZq#H&G9NRNlA9j6E_)jU<0fYUH90&|EjUdVw>=oK~$ec|pGj{Q4p-anei#3N0 zo{MEBI9;|2AiD|r!Sj!p!?&H~&8rA*cU`su=V#2H#>ardpk-<`)rqU#Y8xCUu<>Aj zB6(ngM2quV7-#((K09ifOGK;C#~do$ah*2XoN7nkm>UJ-f&Gf#(`)OIivHN5*{rhS zoa(s%-?1G)Y(2{%x^5lDJz$4UO+$0<|7nH}U2!}*?^ly_+B=tk=@`z%$W_^HBwa;c zci$%zmkfUe`Gy|s1K!i1Q$#fyw|AV^Thw+@4S``S_z+YO_n`2NQ#4Knix(O(0wtD# zNeryI!9NW^%qlj#`%`H!?`^s-1$~&WbksI`;R<`o*lkc!{3Dj_2M{lcsXq4AxWjbS za3!hdIsW-zwYtiBXeodeP%lA(3g&_h<@MI{6%E!`b7SMs zu)Sl4mzzD6Cd#Uf8o+-pVkc(?Yi*vO0Gk~Z>&+rERjd9xKlPhKQZ%_w@T15nVUXV2 zj0_H5I|>J8Gb~Y)6bv{s&yC^QKY+1@c?QmjSw!QkCnT&vNN|@ZfCN z$!k9vT5Fo!hvWX0{JtRV1d)&n+4gv)z<_@+KZ4|&*vd>YYk@#}@L7WaWID>|M)MEI zf)s9;rFogYP6gyzr>5`jegY)JOcG|RR8tBiO=NKW_K8nOf*X%sbH8o3xm#`RLQme@ z+REXv86XOpy~(aji#gfb5L}hUPEoyHmvCT43cXd#6edekyaaS$PS{!vgNZo| zm+1$;n0(78-UQAY<3>?JJohM(o>BmO2|Hk7w9;8eqglY)AqU&RQMPQ^ScPbi{PP*g zIUdRQKcgRU;QNH6X5<5c(V3+Pyw)~}oL2#GYBm%@EM@Y|cuS|m(}AW~`)=5V2f9Jd zNWaFN9Lg!lLgoHMZfAM3p^ilFwHF5qRs@0}q+ldg?fvP} zTbj&F#fd>~vfUC}N`;n9-i#n=LRXFtBr!YM?GL zfiL9{@NB`i*}pyFK;gtpamMLpmX(A<0{G)T=5$o2pK?4`?_>X(kLa69y`94{Owk}1 z!4n$M=-JaPu#XV~!FD5^-eMjl0^ zO8ksoCi($Ea)eq=Q%|o`1mv)z9W~%1lTndcdZ+)D&RqmiwT+rCu&Z(^vTGRJ*l;Hh zd9%N1w^_T1N`pPpknP=7j=9;5LWvKd<$+PB@ZOCJI;iJ7V(miK%tJYR*s0BSX=181 zF<8zGxVzJ!fMYs>@TNL?AU68YYNbZYg(HdOC8w!$j!K)XHwgIM`eFO7-Bi2!Rh!1i z`lo2=A{Kdd2}0u#8cQT1ZP{odp&UsY=ahK~eJuma=z$MkpRgGeGs0>1a>!co<=HDX zY9bkynRcnxFW2XWr<3&r=#Y@bUXk1+h4|s7Myx z9DUp0I|HSo{f+|Ryxdf1HlARM^dv=#fF`RPJOIS3=e{c|Db#!OK7P=ldF~Bz#_r=6 zf*C`ottZq9VLBYjB5LnE$6aCR{H1kaWd%TS-%(-Z7@{q8Z>iEv2l8?baGBoe8-hUN z$9-lL5Q&`azvt0x{nM}=!r*ZGfmk}+9f?8MM3-RS?z=Ib)Ch6`@hbq@)H%FE-J3V9 z@bA!>-NjFnZ6M!-LH6re&kWaUcpkgkJO}CvcDyrwytU}4oI;>_d%}xuC*Xkgc@a7T1B@xCWt*TS)RsyDI�v zf;ugK>kV`*^onxrcTOPc&{0Yk+BjC5BBfygmTkf+q1mvTtZ)v-5qHcC#*%{R{hR`$GXBU( zW+Yvg-9P$~Yy;m0$Ylw|*7YSX`9X-y7g3Cw6wOHnpcsdNA$oqu_M6v{K;)jo`@=E7 z?qoRYR1)Mv+A>=F6Zy9tU~so43zH^zQ>)UUtcH-Jt;;r7Ws1M%pRcc|=a{yK2PvyK z|JRNFTW(n4`PgFSa?ZH_oQ^XfQRyq>O)n;KN9Xd1lmVVaP8{_f{kql|lZ)!4^ljF> z?z7`9u50c%vd)#>y=Gjrd5l*-Z~;v!R6it)Vvh2=-*Qn|ixThJE^;e&5axSZCzP?F zeov#JdMqQcApHEX|8|e_^?Vnu`hLJS&FML+t&|t!~=d&~e z(@c=aRL~Fz5MmO}p8d}0Yi>6{AKStAeP!)R_+Re5UaJ7~sW|Y6PSG?f5Ae)|*5<%~ zldWs8R9!TVIXtEAp%K#ME`qQCKu}%1h-?j0!{k7OD@Qx57vP^?;DMtl2XF6hVJ%2B zvtEaFfdkhFhKU9f;cR$ySU#1*dW3VgaZR}7Gqjljr@p}Gt={M5U#Vf8!yTd z52p837(V{FH@ywuOL59*gfPcH5EUrZ0i(Qlwt%h0U=p+#D1CaR5wb=$`|boJu}2w) zspTxM>XU;Z#YKM8ezOiB5oEtF@7I_8zvW}vmSv7MvGqd~521^$*;8I)%CM;Ktct^%bmD<2NqEHukP0XF2 z^W2-|EgIOjy`c}1pGm#W9t&o1HL?;ar55^>IYx=hF@ph`)0L=!Ng+Kt-?j9BWJDAc5SkHa3E38 zF2@Nt<};J!QJJpe|xkPTRSd%ld? zZ2IFUM_a#f0HEqhxOyk2++}b#U!&Zb44h7b*Hq7e4Gxs6wK;kZzL0~7!sl#fLHU7P z`DBuKPp^~IdaS_ZnaZ%EAcHA~28oEz#GOFTiK}heRWh~et@nEMx+g+~y@btTdIRBz zUuFUuFV1lA;MSDGz7Kh|x9KQ7j#UWp2+nXlIlKa7gPX>%0g$_{tzum|YGe!iFFfSb zp@BlYQJep6qOBl#=q+s)pnk$k*4o(DmOWRmGZY$*0^l9<&bW2%L(sd_6Y1Px)fEGZ7m$=Z_(}m$gxy& zsRIaj{85&*Ay5zw7aBrN=JZ#UL2 zz_RM%doSw$PQ~&T#GsQ$U~5&kv&^fbWanOiawsRUeWNv{KKBTe{cqsCJ=~qp|`4Vk@OM%gNE}A zCJ!0?7TD|o4d;yzij9zax>Ql@%mlz&2&{J`mQU_^^c-7NfpHvi>*@+?4;O|&wh-i( zeix@h&{~Q!w^4OAb++N~7WJEYN!kFCRybOP^g9%r@-xT}J(yk9WX2!e1t_tli`wMo zx^QQUuh3ST!M@CxON&R6|M759Too8+$mQ|qy7>gIW)$E7Zu_F{#5W9st|xiuKUowA zeQp@Li~r4^MPp_7pLiqt|L2V?>};(64R4%u+LUPAw!XcFy(#4|>@RmuaZjwiAG2M@ zl4NFNexVV7LYpT8WRuuP{e2PJ!f$u<-!uD{TPr4P>Dr!$h88Kn((i;d0;+xeN>HEuf4-&vOVIm;1R@0Q>SCyl~!9%|=w%hJ8^hQ5>K zYb5g-&|XvXD%74(PH#B-EZUxbwZpaJB|~?u`es|>b=8eFf7xJ%eNDy~e;>LsNih17 zq%hmg1dy(=D$4F%)XBFVy?QnX$wg=^y*yN zDAaE#XT_^dJwK`{?yUhBEQ^=GOs`Q*UW*s9-5Vo(xEMHk3tt*Y+wNf9u5LeM{eYNX z&!4zAPclsPP0` z%YFyLcO~X@^nNlIP&cYO#1qH12$Y6XKBC78)$3ftRjK$Qf2aT;@1@D@%6{}ASeViw zV`c%R?}BSr2+$+0wENc0RYN92t;FfJd6K=dc1;t+`nxPIZJUi@M>Vy3`}zUP9<5sg z?VyHGQ0pi{y4{9LPf6_syV)}%U#oz_gX(_64i&5q^Hl+(W!wY4hOPQOv}4;&YX z*8{h$d1_e+FHIKeC#~#E^Vc!3dXC`@GQCAU*bn|bKENx?P`$m3HoAl}#9Ze!?>p5z zit7*wpN&P*ZuHEcI5n>Y9vruJyaM5Vqe^CeM!Zx2Mu5LFwz_=aNsjpu=4XBVi#YA6 zx7bML;Xh$8qa;6;`QS4SsoS?4+b_5aS(sCw2@sqUTU)@7&diK(1_5;jl|%H9;hvTS z27YucJ9i9%B(97E&sm^~RgY>#*e-6m#{gu6s51zVy;KWmC{7Hp^oA~*1qZbGav0*} zn{-e<0>7^|dg{uN*8L56rAi{VRyP1{qF-Pb;5bBngX>f*zyQ~&0^|S`MF|i%pB?G; zlkx@p8bY7E44(HD*q}IZ5`PA1U2ipeyQ{LCx-??En|TQH zpiGt!sIx(wEV*{l=|@tb$D~sZiLXG#Y>98Z$#&)QtZDNsigI73Kc`MCTKPu5+@Y}rfBI4~`!3EO>VyS@2Aov{V)q0+80J^|AfG*c(O&s{9;koRbz zk6^Bm&qXTXA@<4+ew<-T!OKmgnl`5&ExZT7^MX(XBU?*nu@e#dV+ly4VAd@xYc%im zhbCX&LX<=sU6pOvM1>7~ZR7(vYA}f~x_3+E#u2gNccp=$EC7h_h*_aj!nH2**=IN6RybyKkmAl@w`C=g&`XklqgvDx>z7F) zrK?ctOtmp7KniaIO_}&OeE?!w1QKWCPY4w2ajTR)QfvW+hIRh|NfM@AYb3Q2Dp3fD zCn+v*Sd|{MbF2XwLNLIK4J*aJR^;1W%6ac!wgp)+Lt5^6GahRVWi@8U* zJIm&}o6r*rqaY@@wn{a9u2xU}qcY|kR7fHVy8+ck>Lw$ZhL~`bd^;}ZOTUguNTKeP ze&u1Nlx}DpYK&4FnZDO1friv2Pzf)ArmzsA4|PU(54+C3Y7j8yRDiBxflqMqv|GI@ zi(3jb_#xc^q+*OE5iRqpa+>BVR z{|=&IQywcK#zccGR6o_a2Eu8Wg8@0@6)+uDKR%XgN<-HlOXoE8f(r3>(n)jc*wZ;G zu9-#TqO^RZ04kT>6MQ+V0$z|N@7h5E*gNt-m?$qG5Zw_k01*g*+d~STE-E-kVP(t22;(Zj_!Xx7l3;zTvRw?ZG-~t_ESpjq*(b&sC!o4lK5vcM?{qPMy{G!<*tABJ8Tb9t6 z<1bp@wZ3q)dMDWR`BR5aoqk-Ca}lGJGW7=JU?DlGRj!QYz*++TKrRg?6!@(lOV=YQ zH3p+R@_j|aAcUF(#XEpUoTpQiQzu)Iyc+~y{g&cTz=+Ws_^RrXwmm}?Xl}LIJ_$Po z0P|)5V#u0FU<0C2Si@@wcvgdJ+!<#5HIy7u-nKc!z6*MciaPoS4+dc6Xh>1RN;itw z7(ihA+8zkqz&3desE@3O&`1vpfF+#K?s{#a+Q40hK)rqSx2GYRG0*ZpDUmYLtj1<{xukE2z@K3nF(Hz(R>j%<#dN7@G}|jL zY+(+KOJMHk%07dmz({n7825pZSFlB*mm(3INsS1DFBaPj;P*P`9f0i;BzyHo0hE{p zoOhM_u-hNA9uAwQ`7TX+g`G-aJ{41v_XD*?Z`eWnid{6m&W9QKBnI!=w0dkQ*3C8p z{I|CGM<0^OJ|_RBy`ADb)N~Vum%{IrmA-*2F-)8xG0<%HN3aCpIymXYh>_R-+5i43 zBe1!g?HTJSA<4x6lJ5}^RdX~L8zKN5RN ze&F6x93%X)B5${Mc-cN8($A0usoKw*x`_aa8#ef5ry~ebyB|3<=eo z%8!1g(wsAxxryDMju8p=7kJ!IF_MX|ep6rQtF?N( zs_u!otFjHceh=JePvOj8#7tN&p6XGUN_)7t^@e&A|86y&H(WqeTso+SrV@yP`v)MY zWUShQ! zk0u1>0Mz@jPVfrB7h9v>0!4Cz^vglhN39U!E*WH{;RSHFHtC&8K60KI2S%)B)NRy& z&!PYrm7-}Ba!@1Fy6oi+6Hz>!X$67PP;yBF)eg~y09PB;bUxae*8q96Ymyuy*0g~( zRpYWD)~k!Qe^$ZLcevcrmZNkBWfES}FRQMrlesqla1yOSt#}Z3iFs2Y^mv#}g)@%L zhFDe81)WyR7E)-!EH1^hzYqcP-Q)WopWiP;4#BZIGuzG94lua;3svb%==T09V~`^H zN&fFumCzy>@VgJ*tC1cZUViwZN-aAnDvfSFrz z7cQUEuRm~c?N+4s=?k|aR=_Z%XAvKNjbC0TQxZSwk8MuszZG*QrigET&Y4>_2Q3Pz zoX|!~nEjQBUtn}Tn;A<n%YvAJI4sa{MLcO&21heJVF#Cl^ zpV0wK`vjEu&!Km1)Ec7Rd1eRdnTEp|pNXX8AhlXBg+qeG85mkac#6GcP6fSIC8DS* z4t_6*+yB>+7uFy0Uv2<;0@=hwF8oBvjZ+pK%NZk_^}ZA3{wZw@mjv0n)oTvDWQaxN zW*J#X5UnII-xx35EN1y3pYRae`^))(|B;Syz=G(E1Kg>Idedx{y6%MS{K*M?6j-{IQ#_QAQ1YwKFCWnC7jv`L#3wL4 zjt(_~$rpGnZv07_6~~1Kz#aj|+kV6J>plODFk&$BqQi-e$sV23ni!Gis}r;<@baQ7 zJ&prEVcR4_5ziILAid8an7Z9r7EIhbJgMv^HpenpmBPNpW}^p5>?-vmETp&jUqa^8 z#`=slurJg)5fOL!11NYNN&)D`7<$^YQv5L#+a-{qk{N1#z;jY?tDav&Ybk=#~ErlXs%w49|?2(8i0nZ#SlJ zLy3gpbBd=Ps2{!Q9hb!4}q26m6DgXa8jcLrOmk2B@DqaMeNOKzif z_BNB6f<*YdEr6(bdGfU2(e98(<@a_2FYd}JdGva&hS@+~ zYp1lIEnOHGcQY$=kd-BR*_+!wf8`o+Z)G|aG z5)VEL_0m{r^@igE(bBRWvKmXln(suq%P%-xXPi?aUY8V3XcF5%<1ovVw(2sy28<)y zrZYG5g)<`tdq${|k;q+jKOCKj8SPCmpgS*FI((km^~QB zm65WLHofE)U3Qz_YQI`*x47v83!GTdtW3!RAr|9d4q;ZZ=iy~ExN1x)(Ar#C>Blgf zgWsL(+x1LkyBw7ug1IP*GtzE0E2dT8ax||?IlRALu0UqVWE((p=}>JIXD1^_k|JvM zR!904 z^mk{l&riE9?j#E$)@({-YdZ2|vHqLk7Q|HBCxvmF_o+^kZP54yJ5QQ4w=x&i`=C+{ zxsT`bs+fw(3t;fXftW@BiD}uI5k%!{AZAv%SASc@xWhz!1RvqEOCiUNf1q$Em<>oK z50*u?tJ>k^b5I?Zpq3iQx?F;5*{AR&y7X?jLr8l$UjGu}-s} ze0J*?cxh15O0cN?wioD4eaa#slS^fn)wZ=$xBi8~?y$>JJ);;tX@!CRAfx~z^Mw-5 zIs0rUpSSV3OhYt$4yZGG$*e0mp096zVHzq*q-Zq!hL4j-69FG zv@EO;)gu9z)%`~q@oJdJM)1omB7pH4v5A2yQ0!^Nc40-X5gaA5b96a+S!xW*6C*m) z0lVjyHi8;?@!5VcoZ)NXoXR)~P*NArXF>PxT0a&9MF2q5vvQpjDY!2o&D?wbct%t_ z3XxPSLA^KipRi#OY}s%C67%|KOhcGa6_CtV<=q@StdUt3)8>NGX#`6#8>vLA7;#yA zaRGTm79X5(SA6lQLq`(Rm9_9h8EH1iNVMkO?kF@$7CPN|#!RqQLT5(1#48$>_YS3$)?!(aGIHgA# zrU{cSDSqMR4Eq({dJYI8b3_0QFejV4zac=zP=m2+LoSI}kYNUPHbb4oF?VbRKy5O0 zu+Cq<-L#hhjsbCxN#~D~KkI@_mm`r;lkDGVmxI$qm_3!pPF&`=9BgAS-8$`|ARJrA z>aNQ2tE7XLjM_>sgM>)wi))YErhXNd`; zgCN&QCLz-6qrL?D#RT_qduIeeg>gYPS~eD2u}nhp)kwmWC8+kP!>;(!vy*N;R2N=z z-xK#{u;mmV1Y3REKQ(8glWxM#?#q?;!NknG*M?2DJV1mz943kf-?op_z%tOnWn_Qe z*R#y+4N8+y;}PV2DnsYuNxTP+>@;oq)6|=@l27`9zum5kcO2PXp#+gqH%uF|_9Q49 z$HPozg9&sWF~BuLb7xAbrJ*D5R=D<#1oL^pLa z=%`|d-#Ty{0Z?Dz2qMrzV7`sGwbd@Q(^4$S+J-m8Uue*8!yJ>r?0$z0DwU8ii$gTf zc=(PDpy($op8M!{h>1iy^g9etApfea5Yi1Eq0?O(P*`}4_@n4x64Jx2NI>$# zvW@K)0TfLAJOC8v=5K}1T|rMs=x{Aj6FZ(q*5#6p@OmEo8$>utFv4k@5*Ca-v;Yfn zZJ-C`#x7=GH6-B{(Y*nw!$rx7tRF;NDnWdhI&CUQ5!Pkrc=+2&PQtqsOTCv>XHI|@ zYCDDVFGI)=9&pJg5#El*EBOXI1dEF=u45%y&mf|+y4ZjrO-k5rzkeFgGUN8R56~eZ z+XeEPbPGxE3FGAJdTuy|OD}Luwz;9Cu^oWsG#D02zY?hL(Zl$mZH9*rowjzIZ6Et! zj%!M_9jG(jAhwwo3bGh4-+UFH4t3b4?^VpY1PE~iqFuTDL@{$<9MRmtPM&3^IRqWA znN2AD1}muq)L0UYBX_+R4wAt${^Q&?+oupl1eFadrD%fOvx{xKe%i4nDcmeRu-6@R z51=Bx?#jpyeV@Tl{t$m%y7hSTbKKuCnh~#(FvESEx1_uMg-52Y6v5%H*vkC8kTM(7 zl3}sV+BIra_Mr@h66#yx&3Q8J71Qji3a5jb4(3A0K**mwWYQ46ZIt}V>vW3S8N8A$ ze9Z;bHyk-S3oJH?Twi-LXF0{+A-3lD({wtUM~rT)mnFP)_p`HZ04&PE2qF9L12f}@ z;czsHb?W;{uEC5&zQKWqRV@9A!MOftlmJh@ICh%A_^mcuyq5<_`3FGht?S)Qh}`dm zp+K0@`b7%g(a-1kOMcdCH`c?chJ^FQ{@!{`yL#GTLtzX2Rw^l>Q2+`4>!XlrGxY(R z59MNk!0&2diR6&vodv2d#zWP{P%BE2JxQT7u_p@_s!@z9cf@d-_boFoA&hzC@j$ud zB6g+~_*Gkx*zEWK39a#`HC=Vmin*XtG&Az^PBRiZ?IIS7SJcPyXE>i_%$VQ&9&%hW z2GexmlH0R#;VUB6^%Zl7Z`Ri)qjXH}3gT7jD81Hc_-C1t8lH5A$4taecLkLHfii|e z9}W5-F3`@m{Ks6!o>N`suhRv2zb@}AQNh>hbAZ1-$X~_pHzomo2 zFEjZF)(0FF8Hj>BB)yVcEuDbv$?|rhRMF2~A@!S70IE|7ZV8G?NIE*4E8i+~aoCl# zZ$Wv%hsT}IW!2hLx=tZ@)YavR=r&`V`gCi}JIG=ad?mp>JI4gJK4s~J)1)X;X#OCy zt8Td_)ee`Gu%g-asuE8yVCRgzIAi?6yGSepM!@#05b(!dk9<5b_@1&03%ay_F2)(3 zfK~}HTA3J4Y!tV^%B?3%u{jMCozKV+Af|W8_}h|T6BlrWnIq`{$@U63+4*HOd?raK zoFyLPxQL~CvIxNb@(sPM)Xi2c`^tBZ z?F7gS+%=|EUT0i6i7ho{LgdGPbC0j~@k_t;TW;ae_`Z5#LZ*WAo)+X6oHDkm)4~OF z0OBunY9WPEU2tbGNj}@nemV*pP8~cKO(%~@Hw>XlXBxhgp>j`ulnGf+{|@K!DN78D${d&^=s*+i1L0% z$a65I|AyNk#2WF8+@*J~`;@<}@wt@y`(PPWBtIP^T$UmKu6%RRTjk5U zUz$l#SRi+)D|BV6`?yS*n~i(405&#h#e`gR2?AEMpOb3V0#8U79w$wS0a`@27e_jm z7kEZr$a}k%`yzRs%X>HF9O@nM)Gi$N#@`F7Z3nnHV8P_Ykt{dzgE#FzC#$RSej7Fm zj6(92)z$0#^MW5Cd%E8;Y*{dj;8NzQLkDaZTGnwsyOXQ$h!;BI{|{s5*qlkYaO>E% zZQHhOPi)(sXky!TGO=w;?Bt1U>&>ZizP-EZ?7z^}eOLGDYpqLoELk6HZJOqfqL@c5 z5SD@n5!&X>O9U~sN#{mjL}&nY?9%wm~vf?o}_;0%laVj zHfhc@?@{3bcihQrTt{*JuVQ&L{mES2_8d8)B3h=*#D85VhUsFrmVD6( zq@sFWnb64fnx#Fmvz`!nUwf)C4xN|NKJ~#o0Kj#08|ZKmps;xF+g`mXdsssKqNv?j z%n2%=^OGTJ!@e~hHkWWQ-C2YKuO;RP6vI^kE%3Clf|dQ!lUMvD2r%boL8`to-nPrs zw}~=f0!I#MUx|gJtdx)Wrb=nUiIay{CRzlk&f;KUc!lY9TUmY-!i?DbZ5joyFiPFq z&EiXG@S=f%IzcJIi_nvAjoSWLj95U8jZCc$B$-`_0%DFy%8SMWCjoIz<{7 zNM?rVwqYM-`ea$(tCTjqk4L~^&Ym8iSh=8~t#41xm?dR$&@8pM;vu{TZl+JU6y7 znp5MsZLGwUh1PAZz?{?+2enCIxiyGRD(p{@MKny|`sz|*lDuM3wzoe=>7_#HJJ;Xg zWM>DDe9^A)=E|>ysY}}s)SkWPJTZ=C6Zx7yvNBeFgk+V`L$k~O24#P9;hMRQXkP> zkq(&(!XNj*Mh2WQWU*Fcp%*?|Jz%NJGsqiyUwiG=hV7aDwZcL=gZQ9#=^;ijK1pE8 z1Wo@g>1*uCf(ux>Cy-)DTdW}0W-NNpT4k^xmcb&FWN$mt%TCLx*PmNZH@t(P8=thq zT8U~MhDY207u>lI+9YWnD<{1Z62F6Fg}_ZTqRh{tUx(u7C1%YED_e{@GQo4N^oA8& zu*G_MB+5|c*8F6Ek_(9pd-T37J)K0zgByqxsw>(-QE%&BdnC%6TMnMrES6HMTxOb> zLMTa%OyFSv=aT^)@yv^K>C1b|`~>BMSItgds@&8I{)TG_m~KHNFhf0FTzuJy#x&~$ zr&T>+;I5@!AQPpRe#)=m`w0?wXhenkEnQXx@DPeKf~cWUwdxRRBz-2aN;Tvk(J;ss z7WrR~HYQz}wKOWL$3IQ)>O%9hwbZU4{#V`dkem9;k^HZ(Nx;)D+J$_qU84O)RT3y850Iq3$OWP&I>M?8Cx0A2Gn!V$%qIYh^H`daPvxeszNYE;jSgk z-6FrrMHL7EAIv2Y_b=oF#cAZ1xeumgJ7E@|ffeyg$<{{?u!T2pvV9^bJtdk%QU-zt zJfpFv7|_Vl(@$rIx&?mrm^gx*rsP%GgRBAl!Um>`-6I?@vMaNL3I)C44;-2ReRnjE z8Y))CPQa3gk&TMd(RCgq0v=?2x0kp@X5Eu zr$J)M_%JvM3^QAdm>?87UZhlLr{;ll$SY*+ugJ@HiH9s;_0=E?gc^yh#sk8JWC4!P z2(ECA`7w)c*G~bR$hi~F+(Bo5hH-`ot}Kl<_KZ-OB5RO*H${;wcj^zZFeN<(N3J61 zqAsA$Mn&n!vItR_zCZ(AhC)U6_W(>a7gslW$Lh$L4;*G?s0Ht*T zj8OK}3?iuFEf1-gK$Ga=y0)?^cy7-Sz(5@lJ$4+Mz-d0)t>T6TJ0s>9FRNhTBssEj zmCy8HEAgOt-cMWA85Yo7VcArhugMWxR{R#dGUWe`8eTX0J_C!{ZMAUHhfZ{0Sxc~J znTTQ_QmKi!##DG&%>l5vEQBKY+Xf_AO-mj*|DQ(w^R z^b20+E<~xEsI@0cG}$NBMQXCU+^KiL>QEY2OeU zO8PxOS_P=fMN=&_kUTg)aQuFWw10O6j?Klx9;IrXJgeh}_e)JEh+h$IGLxCw%vd?! zjWWG7fxPgGq^3V}%!NB!&u%rB$~ic4*1P&cuk)?zaoR1_wKnK8^r)fvRRr~37av%# z6xONmIK9Q91ZWZMgFFIvpH|5)F;c8QDgxc|B`(yWDb!3;Zp|d0j3l%)WlP;YG)jqiSbmJI2MCn z5>9`K9K1RXD}c&#>5w1|vE@t~u8d&D!`nCOFFz~TE4zJkN$?vB=i1!o{=sz0U(*p> zcC+i5Q7asgbmGSei=vUK~w_VG-ye3Hpf>D6oft*#D8)XMay@E|GJn zE^IOLLY2>C#f!+`E*%p{LpP8qaT{rV*^8}`YDAIj`>$uc25J2Q$$x6vQMkER)fXp#tdBM2M|USq{m zX*qAQ#ivKN#H1e%-+n)tZ7HXwzQ$|<>^Ag7dRTbA+?Q|q?97S}jceO1P^%ubG`!X> zbb{d~K-w74{((pA9M;x!{4R&nQmbsT+D5&U`D6MS!bFInNUZ945Q9snY5b4l1{Kn= z?m8m?2{!Z-C`&1iJC>Z%fa;2E|-)Pxuf-R%*1oC|livSO1eYJ*>i48$viRL(^&KU!9f= z%|iWae5^l+6=zl7WgL%jw&qWZgv`f?iS0Yim?HK55`)HJ5jwQQV0Q(`kE#)#Sg9{N-N?p7me@Ox#Oj7TV%-}c*Nw$ViFTd zlBQM5iA^XFGSX@>FKkc4TrXt>(c#qv?F?sLu%p(C@!xrA8`fT1sfRGIDD8y_%c7oM z9as;(#9>1B?B>gWA_B|O6jJ7b5pYg) z;^ti9%tWvWEx=_!{<~MS28w9klzi3rpyukf?Rc?)w|tSkk;8uG?vyE14ug<{*7}(W zwPZsr1at4QzI#J9IKdr}nCe^lcl`~mS`B{rDB-TGgf>}AEye^=>?YVGXY$I1TM!fo zbB<}CGyR~x>{dORMTLXVGX7Tmo_K%l8yN5iXGxDoW}Xz5y~4#Bb&DSEeN`^=@-5>1?^X~|2eiV0$Qsd(Zduqa$eM!6N#$?s5{*L+Pxj!rqAYHzS(PX+Sx*UiMnL~ zI@Xnck#U%Q!;<6=3(zZzcD4xf^f8Rs1(F@o#jq!XLw)ZY!MUa*)`(Ct22U zq#xp&=;{8y#2Q(dYBX6Jg2vmuHnn>tGPB2&=?kF14*oq)dD)5%ei2E62 z=pkzAwadm%mchq#uTf6(Q!VtEPp!9sBA9P?4ToT(kXEp!AH>E!4C58Shs(J#FT#g5 zu8RY$zC0cL^9VcipH-qF)es!K{<>O{T7t_)@~rQ7mA2T%=Re@Tmp7>7&xH;E2ihA1 z_%VT02Xe9kk6F(5(y}43rqjn{*;IW@L$8@Aw%_A2bEx&=`{3;K#dH zqBRs_Q#a!vryx!6+xl#G*r^jR(NV%Wv){7G7hiM>*soZrJ7E*k#jimsyB{7JQ!p~Xbx{U0$o9x*I0davQ!BDG2 zkl!5q3&LmgaGSbctqgoWljB20qFe3uDlOI$E*k_@_JVu)W*OMGWYMFE69q3i@poWk zYI8Qq`?;a-zCugCNEhdI>)-Y8!2HwwY1q^cT7mUe5*%acJ{kWlWWYX3)-)fBuXzqH z=CFS+J-!zMzzEb*fLt*>TMM=L0w&fME-l$ZloDY7`XV1kSl$a?2~f3!Letsoc2j!7mAt(h7q5Fw=Kj zpSG~^S$l|6pRYdwlIN2?PNA?WU4 zNZUcv=k6Y=!`J_ZdEBaRXy5s|-c>;I5W&Q=(yTsK)Ay%Fou;Q4Rz^l8D7L9#;r^cU zR)ndomGkspNqwPUwH~4zEQVvmoWoUqbhDzrjKGpS=eBppC5o02U#zB9F+RgogB{E^ za@6}y2xjDpK@4}Y#J`LYAjV$1R+{&TUZF%H6C1W`kF96Utp!TOT3}H3H0aP3N^}rw zm=~gTYjt$lgQuwkkE}^bJ7_V2AF;_N51Jb*ZU2U3-^H`$9fqC^+KSnSg(x#yYI1rfBD&^Xx{bZ_bT*+mw zwvTZL%{)y@Wz-T~YWLHx@2e4lC&C|?PA)Jmk*b|AHr5hTnh({`|5fO0C4?&xlKB(! zD~J9W5OUsa_)ONz5A{8j3aQBO6uWKhwo77>!r%5%f9x|BG9^Ai?G~?tX=Q@Ikm>_C zxEz0`98HQgoTJ8nm}i+_$tv6#Udr_Z;GUSy4M>a>H$2gAM>-e&8#7agoE#TqG}y#p z_@iQ_W=Y3j$%Tqt!{WuXq>LwP@iN);1k1tX`K|Eb$Figo-8$^eau~^?Y&TI_DY(xa zNq;~RyW-Zq`+rfo!T*QSb*{c|YJ)~3?pC>F>Di>)ZugAQNX znB#nUW-P!NJ!ndzO}&{?oDO2{t9*6!>@3)Dy}5X*hhzr?rEdMO*jsea*d3XG4fIQ6 z@`vXm%_XwwR@~Uk?VO({qa{i_=sevwu!O#M~i z_jm7-Nvgj+t08M))+zP1#-q$2~5{P$G-i>ZkB>AE;s$G4h-%0ogj`?gu9wpD&+09&^sdN&hr=>l579PrMMknwL1p+jxWzRXAoT~e{sbjh(Gh0=oIh-KnM3gGoJZw#_u_87PLn$8`OSvK>0APZ?DDT{03VO&iWRr;bt^ z=13Wiu!zNc|0Y`;yOy zd4CqN;){g~`9sx|42IIm)q=&9Xx~fw`m;ge8NiBHsC6<(;oeoQl{J^#pbB9J9uh&=R1vDhR!y7yD}Eu77SPPkN6 z-*E*c11t$5ykD^i_A{BUM(syB57FxI@$U;eat)iKK0e{PPz`R!OHT>=^%9azNBAk! zhr%3R;X9U&!f8ryZ-myUl~V|(FKd+50U$|AVJ;Wu0d7`5NU*% z6Se*bGYub?!tAT*NP_evLJI6lNE6I@0!|^n%y{bCMSvPbCZkqRNGt-gLHHeVTxP@mGS79py6tOBYu@d)-z4xNpe=w`n9@lrNf_#0y3+O7RKUvcYC!h(yn=%}Pp50AQMCTqaq z9)yG~K>q+`jrJ{ANkdC8r>@rk!=3&-dJXsj%xYJI>MUcgQ_Wnhego;<*5b7JMLh3- z^gd(?fTXQc{bk%?=?w%Zog%gI!ao5FZ5y7qkxiX6lfq z;3=Nd&RIwKT3E^7fw-x(kD=HMu5UFC!zdgw$EA*T(+9^0e`$(nvR9YfRF;(m6_)&x z`=El63l0~$h1TJ;EC{J+#L@%7nYp!7rDokoKwtPOAVPy)()FJ1mrnB-f|!tATKmz6ya!f+Mv8*;zan z&!j+d>!IS1h1szLI>Q8#X7qI*%=$Cbc$J+}<$Ssa$J?Y2!G(iK9NaaGBvi#o$rH$3 z4c{ZIk5*I+O_9X#4ugfdhW~o*X5k%e^e@MzvlCcn(Vh^FJp;1{cmb&vTSYD{|VZ&2$=2Gd?Lm%RWCc4@)B@zKl8L3q9v-^Vb+5K;}zXdi5a$=JP_s$ogO z2Ah_4z>1x}xc_mC!7X;gGl@X*HRXbvq(fvX-@O^+cjmv#66xOf`?QKT1ygQnd#ktoqLT-;47=J8nValND0W z#nVPr>lKJe!y%xnU^ifg!xtnE_}E>an)T!j2M&_!`|W(z5}+5I(oWfuHiZ;xH=M>} zAzMs-CZJlSCvz^4z?u(Zmg#nlJq?rs;ORdu$GlocXVc&q} z-85bhS#z3`gYigthQ|8^NETb(RYvjA;I(Jl!boBy6WbE1P23@)G?E5inNZBKjk=zm z_C}4>7gvw;y%uB=mkl;UAqImIrEYV?S2vn?C`+0z?s0=eGu}rxaUmOaQ;9w)wI5Hm1bZ)*v&WY1=2|z}o7} zE2z(N%9&%jFs(_3d(2r+o=TSRB^u+YV7NuIE2bq*Cbs$`##Iz}_!)VE3qC4V^&q2R zcIWQ-ca;;AD>*ecyu*{44$}Hr_~6?uBI^&Mc5ri}7pF@FTt+3Q35}a%g%*y(W)Enn zLXE&$x6+fO?%cDJY6o%pt~196XRg2g`vfXFl4g zWH0p{X#QLSBKOLBC3wib80Ng57|{U#_JUaGHiU<$hO2elIc?LnRar484Qhjhn{4w{ z5X${pF%)g*uoSjKOCyna`>gkd;2$;j4*Ef_h(@*^yM0;>v^CNpp~rrAc_UT0d_add z4PY{&4fNq_Dcf(%rDSkzzuoYUdnzj4dn?=ds_Jn6W?uZ+*n!+j!fe`37;Ep`^S!Ss zzqA>qe)y~P>9OP=+vJ#%(7#WmvE)8@>bYO1HFJl(W5TO6#MfIn050BHw3WQ^6fo{F zmHeYK|K!x(Gi()C#e0jE%AD;?!4_8V&0BG6&}yEoc=0mehz77;Mooi17wO5?Cu@Rs z-t_51%tgmnbxuJwFdSxZwMz6h?1c4l-=5%y&kIP%0(MtDiD)3j;0M6(7CtdkU)T}J z1zvuQL05G!4^?%oJ?&%e@z0p-Fb3WQIc#V5o~7MC=Z)MOa6K^;FuCVpE0~ZWmhNqD zPK0nfL!Zde7nr5y_qTAulCbHmAh;Sh+E2$Vx9`P3!pTI;*_mOZd zj|qE%SWWKHNcnOLTr5|qdjcPnneLoxmT^3ZDr1ula@&+bASL;aBAx|-;fJB0J`$=9 z!u1fU*fgiWZny^8rV;`}T?KY<&Ao3*fEA&tlZIX9hqnR4sKXdNQ2zVep5TU!b%!b4 z##m$}3l0U;r2}H@-gm%ua@I|FrLDl6!>p5HB|XMLLD1EK>D9@uOZAEmg#r%6hk&H& z6vgG?qpx}}3FqR33y&4lBhlpDNSbrLg8?VM0wnOBt+LX6=wSA}V#w1Z6P(oUW9Ah{ z&*pg$&4jZDaesPJjETZ&k6qai>9nsM^w#C3>ef65^t}SKlB19Z`QydD0{5TC@y0r@ z>0J@lEB3PP;$0!ta)(WUC$2FCg1Yl^d4HRKCZE#`YNr|*_QmBeeL&%zn3WB@)=_ay z<$o>Hb;O~t@VXiW3(AwR)U7T4lz=!vgKk{RTXaExLKf- z#q8X3;|UmsCp-EJ-o55yzE*eGW;uJ?w#FZNB3o;%0>5zIJ<>5&zQW{;pf8swm;TG4 zv(^3WJkRTSA1u4wrZ)^9t*f%?GGL^uA>JD5fyb?3e(C~roa_Z}u4n$VJ}I(k|B5zD=$4ZVB&&Y6Paxs%wl@O(TS;-&pMy9}S){8| zduJ|TC2zT!sWEw39y)qYy zmBpc$i-{(sSnh64@?rW1S24vUbQdc#ZUJXSf0LFh!Ls}HrTrBMb zp&sP@0tz2=?78jW122vDb$IGLzjLI>wuKehu7{wRU>DA=4#G-DQ7sPEbK0pO61brs zV6iah?>7ym5g}xZ4sQ#`Sq%#2CxBB>-n!ZyLW7C57UVK3;sFonDw?>g^x)u2uh47I zhJAJ#*%@7Gt_JK6{hWiZ0Ost%te*OtL3j3hoZGf7)c&+8z4@`UR##Y^WB8>%p)6u2 zBen@9PuJ#rm;1_-Qf3xghRrc)THzGDyu7*f#W`9vG=y1yRi2Sh zv4ViDKqU|x^t^#=8N8&HWqz)9#*c6{<>Qb|{=h2gJ1k&wOWM0b;Pp`K-^Rj?Vmq&& zub1;vw?>)3ZN`KG<4&$4qSu*kVV6po*Ac4S<>}#-7~l8TE9dMiX4-YWs=#TH89*;p zw0CkT(sK&FIiqgvB(F9oE)=grzhEFEx>7e68%i_&YXwXtBrtwIt zp3?hy;q>-FP+SL~2-2nM*xdTLb!+_-Q()8I;=F|a;IM*lMm}D#>n?GZrpbvv9lZT5 z^P|b&@8)~G5*q|6BNPq}5DY@6q(iewn##LUu(kFBJlvT3INxe>%f{2#vamKeFIxe+ zB)PxI{1bUvzjf&7;W8)m_^8+I7U)wmRO=NfbJZ`uT0bk_LhH!b<_lNr0Eo2Pw&#ej zhxv94;!A+vY(G?FG>V}QW%ugGc0gq!k$(lHtg2Nm7UArfZ^KkhPyrj>G75RJxegm- zx_?-n5Z`uWAw4DzCS;R?kuy>NdJhk2ZB#5SuBGkJMk#zqFv#L0#K<%qTeD7A)kJQ` zpn%andW{3Mi(U@5zRYU)V^{roOZ&P8FH-c3qawC^lU6X4N|H@O6k&|z)60c4M^(pp zw4e5$Eh$$XMJMhh=cS7?KiggLouTl;T{CP_fn9@N8!Bv%QNJa~X@90|^bsy80hJt^ z*m;7B^`Asf6%v-@s5N&~&!!D*Fw4o_w|HrF#h&wYM&AtTGSHr3k}%fCzka^I_`rh7Qj~}t+{9!NYUBeS&2j#)1=e2t`S^)LaO_KN!ni_`75*Ih-+zp2}3L6jzOzG4s4j~&dEN#Dj!DR%`WZkf7!(IGeQy3}xGpwHrxX;LF^p_)Ds7%jsB{l7 znZH5`LJoBnn&|;yL{tJ1au{kcK!m8!yJ2W81$+57MkSH7AdxWk5K$%w`*;sXPM7S8}bJYR<>*y&*GC$ zPf20u)PFyh+gyBGu?6Bkd<#W52h@E-+Cs|bAOWQuhJwTuZD^7iQqIg@LK@<42`Ho$ z{EBXk27bZSa7Qt~eH&cQC9W-P2mZq?r{h{x`WZ(gu`6zHGz#P44?mNKU)cg!@s^BI za7<57&WAKQF@{7Uj!!1A#JEX;;pc5l18jszew!OP!RuyBqU&IDLM4awp){YNXHMP8 z=hJhiFXOMX0G%%&hIA-)!!ke#KlygruXpTRm@gUf5~ZB_o%P1Z#~$2Q2I^mrSr^?? zw@k?zkq6!EW|YqNCIGd|E8s#PJDks}1qU9>p+N!F?_m2V7j+*9!ooxR^yGUAW#`)! z8ufC_QKmG(aqxYkmIe4XQifgmDVy)CP57jt^LXvVmd3UeMz^N z$Iy-=r#QK2+Mhes50c2R0y6wNQxRKq41q0 z9rejlr4QO0$PU5#z}zxI521-tnA|9pt3;oLqjv;u;CXQ|PQ*s%PvHd+s#h%p!m1kz z3eizkItnx-Ne-dmu~PXtRB4#t3~U1oAs!7}#A2*U1U+ucFf$)H(dq$1+Fsq0f|SgD}s=;!{c?8EZT%{ zFksRB;E2J~i@w0Ov(jfc-oK|kFxNB%tspvyH85@MkbAOt{5nP#NSZLdPtkmj#$~|Z z94Z!whfoAG9_6Q*+&?^OZhMz|@!s-P1X5)Gp8W7#=(PdLcrZ`pn7w3^bTA|L$is<7 zhBLk`D^L&RDl^bIevJ9sZ`eI_tTOPY;V3k2on*TH^(rKNC^0C~LnPE_6-;4-{SrYoKbD+bmlSA3<5FpA4iXmkWpMxAQl(ZSL76FPQr=St- z-MOV6xX=K5qb1UuJ*|~(_;tsS9?o} z7-Yfb4y@8x^B%;xGxR0VMHimSDfN*hN{Cru;(S%b#a#vhMD5Bu_b9OYD9@sT7p4WnP)<+cQlI*ltM~8`xo6&?Cp!(*A-&~p9 zLsL0fvYFD+C*~#3hvn`KPVTJ*D)W-nRaxsZ$%ni2YAc27WX(Fx;cceivh1h9vTzB3 zGwJ*b(d*h#!67F-9HNb&zTaZV%il_B>4?XKUyc^bQcp}w{)p7ux4dFcf}Iixw~b-( zGG_Io$zI{+1!|(h0jW1q2kk8H$8W`v(b(p9k7FRV* z7Q!wayJF;R$+w6UamSthPkuC7a;eI~i@X#}t~pg(t3I-85^ufdOILfr728B}>UHkG zWCk*bq&dNI7A-WGKLE-YyrxoM<~U!eiUikpY=VDupv1eFSfjn(V}cDb?PHFp5m z#p|5WfZ)u0L!1aE5P{7gDe3L$+_~$Z=WD1N>v4>Y%sgVsbZpIxlvCKA&Q#@&I|j97 z7{4<{AYJ~CoFk}QPzZxZ8b1Mi;WiZ?q#tHLAr!+FdRI6*YT6jksV(vLW0RqAV4*`btrxr6L;~o*zVpvWjqzMd+7Bskhb4R#- zH9$$u8<*3`1C{w@r}e^%!PC#fHxwKcVqn=#bkGbak5 zu#HhbB3t)+xR)!UhjWac-kKyiN}TljpkuJ_=*d((mxc;s)^td=m>|FN5__R!m1Bru!`1e;4pt|B{L|VRDzXU|t`tq6l>u z78ZQB3In}mNAx?y;^DZE$(9ZOzMWm?KjuCWtcv!%B%E;CW=v#W;u)c0t! zO|Tg!t4SF#ThY3*U3^1)T>~T>e&LG__v85q2YV;8yx}ZEg8%{T27s3AJbXSDH8s^% ztEb^D+o`w>9zJNRUvQYYJdtxmJsTQjEl^jmL1Z4ivW-H583`1W)qrLKdie@GZXvG6 zA{7J0okfDAH*#71`l^P_HJh)G6~K@-V!P-~Z;f!5GqCfS7w8B=U$1XA<#8tL5Q;_8 z5pIG&FMoLA`%Yk>tOGQ(erSaCp+1)pD5@@t1v(u0{lTwf0y zRHyUZuk)}F?;(3Pk)g7RWRb*?^%kc~FHDq4-|i&xDrTY};=m6Nc=thE`Cm*sna5+? z)AOTgIF^jD4EU#;LQUXEQBD4HU?>O?L@>B%C29>O~?t+OweFkR047EdVYkE*0}mm~ZS z0a*l=?Trr?YQz$okZ=bUD&W2#0{ME+mI*(EtTPxG6hI$b=vSbv;33eexXDx zZodi1Z>t4h#gE)LE&U0)GK1a6P#i!pc+m7hDd3EDSKYDZ#IGzN&K|_)w6|4N2>JHF z+%q-|>)VW%K~OO4zWiK{mK`<&LAlA9O&3IP2r*xDa0Wp(%gZnWs!61=lnTRy)r_ny zKcY=KwIpM%t%d!Yxf0JFs~w$cQA|-V$vAA4nRGN+7;6=7e7&t)i5g}kz!%eb=7 z6flSSTSDuG9|fHrxlfdzO7|K|9acK2q^z-V2~6efnREnqn(5NxfcWk_*1bT5^TJo? z*kug$oNQ!)0OkRASMG|f&O?n**{JlShOxq4f_wtrmf*%(DmtonTp3O8jAi3?^7drD zeNKD8`b>_C7TV-KhDK*AAdsWR5vvOn9P{HZ3z`;K zBpYhbVwv@oL6N@)|E7%^^pyRnfoQ2=;E5K&WlgpME)u@YX=4G3k;>}vW_Op|`-l4< zA(ZXlD}K-uAiklY(1Y~@%|}MJF?p&Wj8Ix&B42I9Kgllz44Km;##05cpg5N9P?t=u ze)C3Fe2l`s&fq(#>U($ZWJY< z$K9ch;Hj;~u+!hE@jtyn`o{zhD2Y*oXroO)`G6BLXQ4cy%u703-Ipd@yNC{C;h`6l zxkDexEa-lx8^|lj5{o2h#n&?3y5gciVqig(8|y2%4}q|vY7qs<$5n<4VORMNGOPfn zR%hC=>J8T(GoxJey$MpG-sw@01#yh@Mc7O)^nwt|?A&|EpbrSts#?`y=Z&~d1w)g4&B?W7&YHOmpjUC@@g_7I z6hIFYCQy2LUk#b+%Xj`$7|8O!!@wVl#eWe7*67OrpD?idzP8&Ff$vg7G6LxgRGn&z zYCO@B-t5nq_G~D718}1-hsM?)jt*Ux&%Ea!6w%OJ9oP~k`LbaUnFObt2JLSDaB}KS zany?<&gG5qtJ?toaI{rY5MR2-2k%?2E;dh}fZwZbsbRKD{pscIh-xZ+v}86}^K4Z+ z@3fGsM-M-;>DhvDM!_QOJ4D)^O>(lDu4b@n@-@K|1HzIO=T}pqV><$=Kiafh`&61i zWYwA=V{d0`_GC7l8Xg^kRXYi7GtJ>`8c(#3q`S^aLUaH7Z+;2h}&G6^P%forYI5rHfFF5EPo`KjVqU6G{ zYC!{hbg#W8m1mDsNle5EZIQ|sDI1Ez>ig=4+=Gvz7v-~m-!ASdM)P_+x0Uw2r`GG; zCAeWP97Q%EIo!7$+OudF3?-PI?)Zc+B5LPe*$HWYPyf#9|6%MLn?wP&WZSlF+qP}n zwr$(CZQHi{v`^c%_0D}?=SIx@hN`F?J1cYLTj7c3t!Jo^j>`1Kp-~x+m?rT{H)k4m z;@hzd;pUwH_dR8FR+85fYSc)z?$YI_S6{!~YCD3dTsHNHo#Hb5W-#KTnQDEzYE3}?sHVvM@~FxzZOQ?@321XJi(^3UXrj$9 zWslOYk>boBPT#L>y`3}Q<;TaIiYAn}hX(O|uDL1zrnO@b(=m0?c!<|)V}QBc?B%Z2 z_(3{rCPXIM>!^OB-cIkAkFE(ZE4^3H!?WK7>D02PHy)AnjcH~cIdpBfGzWnIx_89A5H{4k?VYo~4N*04P%%nU-MH;-eBtZl`;0V=(7x_{Z8EIMEf1S7SL zLL-j%4DSX(A`Nk9tx-?zQuoXg1hw=}h-xGuAl~_QNGvHG?fzbA+Mt_Q(1h2@stM7@ z(b#)ryL`S1eG2~ZS;805swj&2pay@QO#JZ6NHw`naBbaYllF+dKf3-9P=02=+uD-E z4{d7x`!x1iXI)=a9mboW16V(h6-q7AT<5=;Y|X|>96(cpr0q9G)K4gh9ycIr^IH8A zX*hmt4e=4>kTPbyc(eBC#m93IAaPAq2RQ4S1(?&|>*2AcvM^PmDz7w(Gb`EuU>;TnrZWHGH|o zPAMlZFRzpl^vm8?S=^%$WZGt_lr3l=%gn>eABVokOAZ06N$5jZ8;ho7i`74ntI%gy z%ywkMOa0K?Yrd(o_7K`hH44j!+G*&nh5yio022m+G)EUeV42}m&;w^&M|46gX|UI= z*V+>zP4(dO84hlFjPod9zS@Q0s-mSe#m@{aWd~`_{;u?>zK|C1k=HFe@wm)u(CLW+dNTb=Lw0H4P zmKXQ3*T!`_`p~9pzmALRqWoN@TYG_9Xm{trvx#R=o35z4Sew1K!kE%rU1>QNT0h^@ zVqA;RND&wm?JB?yA^f#KbJ_i-yFS-HPR2{z?|J}i=J9LzV(n6hLHX_*4KPzP z8<5KwlkN#1vZP13kyiMCYn*?I4(hthp`HYvU}8Y}=#>cuYBUiB8TXm#S!m<}ykSD^ z6#-ecrQ1;}oiQU^IBeqPQ0!0XlA-fpH&F@`NX#-*ID90$Ug7;|cxgx)63G|H!bb<1 zFmgFw7RF%6YB<$MC`|W;e??a($KeZ;f>nd2R(|;`IQ?@)Rk@VW0FSUvCL9`ej&ytk6hTzm zh*^0sMXZnrhCcxvcr*F2d3wZxW!q>+7=jas+flDR-Bw73a)&X-i>IQfJ@j*4-hWp8 zYC?U^qdKm0{?g5N8h<~diSbV=Abx(S_1jgp72-JPQ$!we&`fhC$~VtErCZ190gsbw zeT0?Fql*^PZE}3`N||+LQSt+yW7mwE{Fqc{7W#Bmcda}^Dx1r6LwcStfv;8UI}Qc! zNjPGLD=yxlm7-zm(-1<(fm%(e;v7_g` z-Bc0;e@$Y4lEW}jvPs|2RiKT;7pIpDn3Ys%Oy1ORVCw=NN83ugR3?l(;&_-ONgOTn z`CpJi$3r7o;mDcD!aLw%40FqCaR71!Po>`kU`dvnYtiOjJ5BJKUcFD;bakkXUrR5p z@{W7Tj3{y`?V~i~r_s7C{ovOT?FFcFY-i2g`Lq>RdrJIszd^E}7=3Ag$47_9mC*r+I+aD!l=HkD}$kT6E95i-tPV zpaz>NRRo-k`{-}aZg?75!$r~FBlSD1tGr&5v3+`_$WO+45QWskg!CvB6dR+Fb7A1` zm_RsADf>?sS%J}?ad?9;?vXB0AYs2rV9MD?=c=TpWxbqUug3$>ZyJnkED@@X%{XZ{ z!xVT$05BIzg zTJDz8rZiSv8YDGtD|Z$vHP~S0qNZ9rW9?t!c4Worq2!;82o%^HYh_)rrV+_3X99Mi z)Oe@{QjzZ)Bk3*Xl3+JwxBoU1vS_YE(Ft1-v}xeB3W~HWjn|nlE0(q*sJd(PiKg!G z>W`+33N~tkOvj(~0j*S+(G$-to>6ou%P4Xr?b%N&Fc(qw5~bo-|I$PT%{?}UeIS8< z%J;a3zf2BxO+Yk<07dzY=6UwV4YTQ>jcjO!eZH8jQK=zaK&eaJ+DT8P-5TvUK*X{a zo3tZxU^L7I6*5(feSdtZvjZ8WG3%)#4nXMNKdJ-s48;iA3PyTi>Skc#1y#;<;{)1k zy0{MuEC4iD%FU%+oAltXjb5n<*#`2LBect>I1)ypSK$K72(;$YLRf+rT(?!TMV%)B z8WwsUsq0`#Y*@UQZh{j8!H(;^fo05~j~_v-srqB=+Fo$P{k(mI*e$|Hzg2o`xRRj= z%bsSQ#<-mnBuj|m{HBtUT1_SL!vITb65Ec1OKdAtx+8!`i_yuR@YKTSL=)JWhLzFY zf@HR#1!~NU&Q!uvV0*$c+P;S=3-ev(CAI&gE0}3x)oW zIxqPPhQg4$(H-%c1~i>sTQ%=KpV@N2$E2{v`m;){3N^xTRv83KtxUJ@WP0g?`xf}X z30nFp2sy=EEQv{!;`F;aq0?RgpRHY0PJa`ZuRy>F@1M&D@tqs??cgalt_O_Z8s*am z`l_X|72u~rI={M?>Yo2XyjhC&@%WiPvDD1uhSBrP;-^@5D(nL_`(}q)GmW0P%YKD!x+33Hx%F;0G(f{DnB=G1k;)4 zgrI>HtvhN7G^R~+`yzp`qMHWP;G5;`vYhJ-%2;) zGCA$|SymjH9bUS5O1th2_#5D9cCm#A`h5lje$1p6pU8V}saB(Il__(EUsfBKHE1<> zkxTo{zh>(;TeV$XF=!eD6n`9;GH6n7<^jRLepO}LA2u=ki9&nKF{q2Qdh~55cK5|C zzNqqa?v~)g>^_mDU}n1Q)HVHdZO%8Jy_xiDwyA!9nKaX^AKbXOPF}O^ z<|TwD1m!|vt#8QO{Q_h!HN#18=@~yO3-7&B5|ux={tJ65nDB~}N)xy`LNz+>$k9aQ z#i%-tCiaKmZ;=GJ_51X7pEPe5 z&mRflZx(piZeV@g7492}CjGHxYZ8%63tg4g2O{bZjaBCVfHj0K#5sZqjDC(5~J*Hh6P9RFOH9%sdE61j?@RXCCf> zE&+i6bk4~RYTZdanQeie#FtrB#Q?omD~?V>0cMxTer5vSe)^yBC4mjXLz@qEOL+#) zU%E5bg)JwV`oZqcg~8qf`nzPbj>(6eDAsnNr#HiegJ{E#_3`+6g+R4+{d!&M zrwhx&Q(QoQdvCDcTw3H=hBso|I8X;`^0!`(1w;K^+{Q)q6GG@uletTZn9c|a-gAVC@2c?>GqUv-s7lIG5x`_a-@A^oKC&M1c zNe2!sglebWXBh0R#ND+MGS8Pw3PDuyhYmUp;sDWfUzz=_JA@?|_1seZn^(XG*LQ4f z2!=??y!kWt_*W4>BM4_djJY^& zZck5H*uh~T`2APQEvv#*cAM%CbwSya`VTx%rHL6d^W7b7#2YP+gMX@Sdo?*Mv!+Q< zk@bnJtB|U2lP*K=@k(-C*heK;>$k1ODVdv4xCf#RoV$=XgotP@Ij*(#gkHG&*Ynvg zAbYv@*MF0`nE#8^#lg<5wRxTYN+dm>=af3z!DxA}pj#n4M+Wgk9mKH|*P~)K zwliu0UxlqFbYbhTWMQ4}=6RlscAEd4t{Dwv<6F+2xJ%N6S7ePZuY@ZQ+_r{!8iZ^S)28e>J9IYSXij1k`jWcPTnScM+ z`P^j_NwV7L*g=xsKUPj^`nYp$N#zVJKj&?p0Z{R0f5dOj?7tQZT~Frh$MBvCcU5h6 zeFs}5@a~A^>@qFqZjw~Jf6zaNLM2lVozrx2f6gV3!FzL!yK&+e4oI;^Hpln-lZi4m z`JS8Gns~aK*5(jZMoDwB>zd_}Q_hWd^!sR*%I{AbyZa>-vjflGKDz9UzZg9O()T>) zeus!BA^MmxnzvnBYA9Q#?uH}BRXZqO4B@#k9Tva!-(7Uc7CxD=;A6y7-KAX?mzk-O zdDP@+)tMt6Ds)l0+O&6k=Km@$U}$U5EUNE7ZAXp5rj52+9S@5B0jQFTEY8SfO&YzJ zvbn=qr()k8X^JD?Wd3|!H(6ABNgKNjf*A@uG0ogOotWt^lHV~Ni`rnDGqW`-flV(P zQWB2Lw2mv4-gwY0leBwgoW6Uh|0Bo=GcbSvJ9cff9@CT}dXkF<&mRt%Q_kk!oGMAt zvI)met8V=I={D8YC>QC@p1jqXT!?IFHfW-2`x(~tx*2sIt|bJr+@)tVu(honp5CGWMp$X3ewnC}>byqk9+y!1xAQvt; zA540@TK$cwtG{*x?-owwtXgt$4G$JkQrUTKejxG6Fm8Xd?~v5aDR7Caix+jFh<3rl8Ea=6xV zgd}DLQ}H+}yP^WEXVeF(HlW(t8;2V)%r_EF3E|V9+rCofFHFbxN47^n?W=Q>u97}m zR4W)Gc^veS3ho%}``nvrdXk#{;+bzq6nP_{2SC*eOH7JgIW!~ru%1ODEwjwDS?q6^ zeuF8dm`pBRlxyeKUDN}3jK|g=^-Cw_Nak#^u`b5B8;?Y3n1HSB#rHl)AE=T)?1O*? zMNSTA9}Hx25FeF0?Am!%tUWU(lClm17rmW^>d9Qlrpz3unLxTgbu?%Ms^Jm-Wyea@L3!amwePI!vK#zf#)``P)z?mA39 z!v*_G&;Rb>k5((QtUhQ!-$=NnM*YIVZXg>Bk+;REd*U^(_Dj(T1+YGg5kMzA;IEtb zPk#B8NC1WL#&AqK;8dfg4`r-;VoAZiBYkPmWoQ^@LZ^oMbkcpecF5>-bo+fL5`9}Y zT8Nt5QmvXo%Eqvrf$v2(8c)Rqbug;~^Iq!K)dC)46^3FP?YIUIWQO|^B6HCKunU!5 z?N@I>U*)(DjsR8aB%Y+H?Cuzn0ykCIcnMo>(SZ+CWv9ttT1fux9?EV*l!Qp6x~THC zdn4}NLNKejE2$n3-e}%v6lq|#``#R}=X_6bnA{&&5tw_&JSIvV8VxV=fSamyec2)1 za*cqR*PUk9#_o58dg_=}u5puMf-RouHklkZ1I?W|jY~|BEx&uh0Y+hG@+7q*A)JxI z=G8S<<<@H0Z732i#ky(GdUiX{7q~Ed_rNqBQXBV&i^mvK?zft&tOb$jw(f?Ix6871 z8?N+^ungOxR8N3J7B|Y~ITDg50!phSg&pgRd?jdVVB|+7jY`i7d*V>zg3+RDmTGde zcwCb^pO4tmA2D3P9lc`h2GW`OCtU5gK805lUtCSRfNuXC3!7sEJFz$dSkE8^XK4Bc zI$x7s-lxF0nn=0UYy;~|x%YsT7&SMbpYK{^kg&FrSe#_omc{bAPPF+~+vy2*K3O3I z9XI*fH%y|uJvl+dxTmJCrPZx>WrjSXlca-3H>4*j=E!q~yJ(K>c8AWkl2jE!qns7N z6VHZ1Bz~RSY$72Rwt~PAQzkT{b$q&Ih~+FYw6t8mBvE=fNv8yE-8l^(rzw^CVRTF* zVn8NY>x2hMv(IB0uG-sy@#L)B9LZ&b7PNlfqn$3L(pJ$FtFF=q`qs)OOjP#3&Zkn;HJhIjG!;_z(@D^l9&c8hOOZeKyNT)Uu6X~wcAkmentK|4mRRMVU zu7hX|*uI~ET+=2$^$adm^)Y^)QZp$e+E1E{#Ox^s|LkC#PdxHMY@#i9#=T%7>$;1*Hg)B?kUm!* zru?HCt{A&$JYz$_MeyBd`~pRPcImQ`Ks-GFu&CQf$>Y_^uW&~xG)Cr7FoU)T#goQx zy0*zYVi7-CbKryB(}Ggb@h7qEg|8Np&NxbP0p2wnWcINh-JPz?z?;UNx~Qptrc_b#m9O;a!#@k?p`aFf_m%Du zm7;BZ_%Sk2F8X*SU`p;KPo4`k^=-D2Oz19QA?Bc{M=jY%6(j6!8R3qDFgUoq2if0& zS=L{hExQ1KPwiyTbE2#A=xWTBUM~{oyT(NjByytLAYkB=-gu$`aP53E8B2*`v3p;h zReV)PE_0GKn0(A=RP&h*H)?M~eq*H4HI~&XL^{Z~`=4!tkixa^f?#Zo8o6om_58Q0*ENbLB$UF+G{=pm7-MI-@uv3` zjPZ(7$lZd_KJelxrL2`%4;W{Y@*dcUJnQI#*(M;Okhp86y6FcyQ6_Oa;-@lgL^?}8 zJ`O2AvfXzf-6p?)o1VQ{EHH6A6=uke@CiZ7p&ef1eFyf+Jy8zBrnyerjkm(+*p&3} z=eQNo2czEoDU5rAXS#8#=15ERI}JI$J^b8W%Dlhq|M(Y`bZPcifY>DLMU#G@dAsO-xwa`4^Qk6{*n) zpc$>x8M^@k@;Z|rT0HfI`9?oL&n7~5LdIWKc;LE5{(49CkMiiw6BG}yN_lf-0wL^G z+7Kkv5DG23`s1#efrq#=f6Ug9kpiRrVY)}G^1(jkl=j+F_6?$grB^}F=Ky#g6@mw{ zFiAhRb`Izi_2@MCdQP0L_T1N74gfs@Q}|e;moBluJQ+O;P`_HJf%b42oCI|frdyPL zY@0r%Yy%(57@}{#>gvq@FAc8OeXBknsUkGdrn2{j!~{9?Gwn4WYDU8ItL8EphPhTu z>yv&3Ko{b>s^8YgKs(y(3)#$;53Z#}&pD>u=CoJ3O~OcNnM~u_FP&~#uc;fRGe@SI zv(Q*~3jw%^?N+XwByxueY`zh`nR|>N?sNbmUl}7|I;5_+Fb3#X+>>QnzM!GL=t0*P zl3>lE&>VTJXKr!m+#h-PL}ZfRG6oXeC(=M846S8nv@qdbW-fe_Jia8B;p`SsG}$DT zhvBK9=TJ!H^>T=DrJs8IpYGW0pN23wYvWdz6vZP2!Qr=%& zYw~T)?`!can<`l1!0WFrsJU2(TbFK}6#T5eu>FKaSLK$Qyr^;CzUn~8;e)EWA7p$? zM-9NrgEYSDO?_SP6|oG#9TpH5}R9#TbOS}^hY*izghsYNzO6j;<0A-$!Oq3Z0?#vDuPXNQj?mi@2={7R^kJA%K z$lVAx^FHK@p!`B7wl@%}3dwj;v}%N+R`mCau_$e@4(`SMXxkf0a|mmdj?;{_9*^)e7(Qm zZZe&Y6VnGj6l7bs@k@>=2>gRI%Ui`t!$rehide_YVt@_5BxHbaUk0}*%_de%^xe!r z5Kj(?Pw~}0L`(*rag=M#Z%-ldbQffq&*Hc!4Y?tV-+2n4r;Z~7Fd3+W$3V-bPf2 ztlB8hQp8O?awIWa5D~SpJtu)mi8L)cx)QqA$v|sBOV9;(R0cw`XzzjEdTc0ePS$_g4n=NO91{B$z1MA zQk)}BOs?l^H`~D@p}DI55ncWFng3WlhI|?&pE!DB`6!Yel>XjQ-YC2k^;FmdBQ^P) z47m!z7>;~pVXsPvO_^QQAAvNq+~h^B^Gt54>P#WX0|ja=geNmphuie}AzN-PWK1|R z-1@C?q`0HYSV%E}%>CuW_6x(FV=@NcI!%>k+etnW{kVuXC~Ec$H@(m#pP5P@*pd_% z@W|Pp%tmO0xzZ$L7%O@PF$8~EaFQiY{&@b;)atiVYLoZbJvFvC=}_cvltPvSrN7SY zI}bD^!KNp`AEquCCL%PUTL94N8?n-S!{BW2Hj7tT{{fVGH_N{bRoYX)x>Jel24t}2 z7+nS1ysQUCilM5+6d9JPG`3k5>~(HZQ}u!{=nULvbZYeb-F%C}dX@(_7MirvbP5VX z8f$ivK&-l)Zlu8cS`1}iFEoP4vcnIX-+J@b4Y-#+tOsf7}5 zo4f)Tr};GGqw}Lw5qCPxji13oGxU>R^}_9o*kF=<)Hjjq3rS`bD}9=eVWoxxys_9| zM+((Wv*yZgi8yh?G-$`L`9Vd^OqL1bJ(yT3H0WlX$chVCMLZ{&oC2QWuc5$1gqR+> z_NVn1w%e=Pd_qV{E7L1y435e-YOL_;VKT=CGR0v!&nL0y5A>9E#_2TCjRUKS3Vzbv8VOgu-#fZFW6OUXZhu;cwheC9-Nbvj!TV1aqC}J|!hfX%pz|BC&2K{m z1+I^nA8x?+&*$Pgt+ukalarG#t8}mH2@2TB zRU5b>dMXRLPe644&BFo_`7ETtimQ2f-!=T^$V7m)$h8v(5_@wJpSo$gWl7>KjMR*O zFX``DY%lzO6V+J$i>Su=f9uhZb>13F+qS-4>B(BMD&r&_In8yq7_d}k<>eA z51`Nxl0kxz*pz=CZ_XIKGG}}(#s;vQaFbPX90gg}^_~A$qG9*zmtp?EF#db6+dL}h z=iz<#`JWPv`}&Q8yYrR7L+DyKVYD2laP?dGjxZh?p|T*Q@SRW@ZEufO4uo>_@_G4q za`k>eQvQl!Q#F)1;ojjbQq{(W zj-eA9vsS93eHC+0y5LfJ@^fezgS#K~`$vPYHGHw%JXclz{$=GkUE5vtJPRL%h*|i&mzNY zVAdnNeIv&z2bR9-bOu>-mRY}=zRt|`Y=4kk5~)X-%f(VWlznT`RrSAfJ-l8{wn_RR zP(DY#cUh}FhP??CGZ zk8A4Enj=#!pu|)z7wcThf*~!R7hhtD3MwGAB2tu~8+z<6hY0s3PLZhPm9KMK{dKqf ztu`UGZ0u?}H2EW4Q*r);>IFb5^*5L;G_RNpwHS3QwHSz>(#DlI+-dIJNCE|b-wdoZ zxCt)E;k%YOgNU6v-uD`8Ev^oPQM2R~<8N8Bc>0+pyE^VF(Nx3kPh}JE$vU7gVJx=b zkD20B(p80C}uP-)G~GeYISWV{{6sgHVW zw?gpWw)@WFU#t`TJK`F`hy!+wp!Q^yPO%h~Hwsk*l|adWM!4W<=dEU_75Ne-I6G42 z`Yu~OF%1hA>U(ZE7{~qKbsO~aVC3qi?LW<)sEfERWdlZTG&bLnR%q+?-2!4xBRsQE zZv8*Ofq?eBaJOZ({7#MNHf4P!;Ft}MnDp`$&0y_aL?*uqPhvJ31-n?Up1 zHm5xZe?s2LDu!ufAN&s$z{`(xv>Mx2&hNn(MKgwz*z=6Ml z_Ap&XuxJD@*>RNH6}hj<(HuM?7A(R_^hrKP=a z1t35jxB7otN(GTi2|X>LC6q_-CA@Z^-tOkRjuep%01`YrN z3j-2+u@n#>_?ynBZ9fwiXc94R4#7)4Un>*P#)ck?xXrmB7mT&(s$^GQcI(Ih zRTHR3+%^DJz##bsudt>xTRVmO77U07{22TkF7AkrM)>bKdw}oO8d{qdW0N;cq^ju^ zAsu7^vsmT6pN@_|Jkcx!urMm=7a%DSA;I4Ib?T-un}P2cmq?^3L_ypi8Z+(hy*D~z zUieT`6Vo&N#%W+{q`=t?(3dnC-llqxFSd^)kC@l(DPJ8x<@OSj=#-G)U`9ND(h(DB z9$-p739H)vqI8RpJl?%s9P#Iy9%0rhb6_Q1E`=S;E5PnaXXfe=1EGddlbiztc7)>s ztpEhhg9Dhf+om8dOQ!+jy`92&f_$x}8HfU*4I~qs^$@@QWFbm3_##XwvQdMTKIY+R zKufc2qN^_NkTJpx9hJve2a2PVUzdAfev0tJ1K|QSx`J;M?o$CMCbTjCB{i^-bPJ_h zkW@P{8uM3@b(4aLGF3ZDY1`9Y=5u5BFk2Le0C)RP``2wbyb@}0zX1JvalDr zSqxzy3+3gWs;|asQDhC#L-__(WMz+>g{lDu!a|@69AY$by4moRKr)9SQm+~H%7lga(^D_sn^TcoTsE{%scpaebDgDF2`wgV{&|xAbX~O0IY_y}3eDj#pcJqlF zD{UVv=LY-&+b4GeI-H(eA+5+T&vovjURSIwjQo6X-V=i#$j&E2ctgGmG4?4>Lo#~& z?WIh+`Xmtj-bi-{jG^%S*{0pbx|L>7(<_w_w3~$XJsPlZB3o}!pU5%gnY290Cw^7O6>licP7R^- zfiot+ZoT{})N((9@jlUUG{Jy!5&K!M@J9|%6D1tY5bR>lS(NsX$R#dDgn5$+13=CGU!`Ik+F9_WSlk-oSN zx~k;4hsqwCD@7iie)5u3>CD{yG}s?kN><@tVIAv0d=&N&EL)yS92g~wO?c`#fJeTJ zDLHfua=UD}LJD3{R(Sn^l{kmw3oLRxpLwnX&GWX)t5>I!QEbUO?eg5eZ4qzo1#z#* zhWQt;D6J!Ejo22@LVu8IxTp-Y6D|qRk|p<*d~i{3GWY?$OTQQJJ)qcNIBWvC3}@eZ zvh;QqslaTD;TR<@`*!6HEfPmB-%^L|mN3tel_@3px6WaVHTa9>_I8_=F~?AkHj3sr zu$*g_SweH4H`B2*AX0n(CQzmV+#rQU6qP!MX{H=C?Yo@kHAsl^p`)Eo2M-BZre_lN zdq~P5xq?=`Eu2G@|A!?7g$0Q*+fgSlTm513Radu{>3RS(oNHG?771;|-%deC?S~_4gun%186+LXYimD`7kuXZdu&$TqlCxEf*!1sV zM3iMbZ=Xhtq?`LhM5MhU_p|f zZqroL#ip!{#oxay*pqjE?!jXBN09C+mvOm{B1!6~;9FfG_i+nh{;OQhLG2Qjwysk= zBEa2p&~CNbsrN)$sf9t3KqY{3OV1%MkHWmZ_t@5%myFd;n9#mIX`lx6OT{tDN_~3F zDik1foKkRbfC`v4)%i#3mk4ec0uDDpBkMifix~(U`>f)fP(ic9upZu|oA;$HJ(XL`Fx;$LhplJi@cf48 zF%*_pm5q81m_igyY!RZP)zvkbX*CgCkmC0VH+LBDfx&N>Jx8Q~7!d`DQrWMaVhdg) zD86TPE32g`2;(6z1gwJUX?yrjdlHwJ|L4zNc!>+jd;J1GJv&RDAQIrx>i2}}ps72; zyXBIpZ2 zU;KJC+hIpumpTSBiA8QR&<-L5+l`&4#rHZp(@bC`Afaho16PzM&ME->xr@8ehV4wj z^&i?Sg~DuQeZH!fm&nxk8cBjs13@5F7bDnt)~g6`s<2MtzoA6sGoW0c*PMt?PTH** zZRrH2mRU4Pc>HN$Fjnsoa9s3akIhx{p)(a=uz17HPC>k z`0y2U`R6|wc`rK6_MD)&3U&w_V#|T2(QQ?BH_CwPTP{%vDr$+SR`PTW0?Aq88Gclb zVG&Cmlvp@OTridsXY|!uqB5^2dqM2yGpbF-*WZyQkf}xrvVo79%(B)aG0SuMN4*xDB)MrOBw4P%Y2d*67I(Yu5qW*-w)OK}ZbkM6Uh!x5vFg>rFyLjzpUx#{a`-B!GEk^;nYZmEm*H4wRm z8o~7E`MvpMG%B%@1W24onE4GVBkC^z)zxDoR{3)4&m1@db>F=RNyw%38$eX}z#<%w zEN?2k5#mt3h4Ar;ZN~k8g3@K7rG!PxMeqdjy#m*+Q322B?+y>oEhUwPvS~>R>Ocm8 zd$^ySH4Xuk32g2pvNW0z?)$5wjs8=3Q-px$*}A#LBXfoG`a|N(RC+v0b%tOrDE7I_ z=bFDMSAz!eMn$lD#S`}LD5TsQ?KGoX>cMPz>diE~@43^W2yUaQ2ZT2#Mqq=X$9qOs zI0h5PT==Y(qAzBXCFif`v{(^j7#rG8$FHa9%gS@=#hH10gBCN1B7AZDfFtcINb4Sa zy0)Bk;JVS~4=Hpb95F)N@^t8r&|1--mP0}}#}F6{N5bNB?%Jz|nH-_&s49#wx~$}8 zcrd!p1UKCHOD5^>o(m6|(1y|r{G>xq~5v`}NC-(ImyEU7uq4 zQUxpH-Tt}v&aneyyXd2E_yd3rU2xn~K|n;?fQ5S@#c*RIJ9y^?q^UF%xR0PbBkZpp z!_h$B#VHxbcJ67^Kq6#qcbW%wYnXjk#TWIs0hXLrZ(7(!sp^_vx*7CZet;@B+=gy1 zv1U>1n|Tm6y`jF@YqDC9_g%DW(&hA>HO`@ zNcbe0Guck>;AJ(uWt@YuzLbwNg4M3-+2aTn{dqi2dGOSf%K~1ij8{4-$@BrSXA$(i z$C(~bJud=du0!VT9A%azpMBOW=yB%s;=N~{jt?;NOe|#wsM}c)5lO{$*b*p0P4&k< z_TJmWani}hp5*wRXBGVfPBZLDhUa;?q!g2aK~clO?HMhS8&y9{iv*Hea7haIU5cRh zFSog?O5Ny3CP}&mCUHe&zWA&?qmE9$f%2JeUwHKpOR_h=x$}f&HCA#u(xOM6jlGs9 zjJgnA@2=g)4oXf*Va3n^^X@rQutl`#_(sJ8O!V3QfaA?FtVilMh5?!Ax!DO=tUO+5 zDcGNtk}e!h;|wty{3L~483{6Lj0T`8hA;8@A6v7^wH&a|oB{&8@BmGEkGAZrFijqx zi9-RKggP60-1F+fJLN-aUm0|K0TZJT^?G(U|47js2eZqM&G^Qm2}K6Oo?j8w)8L@5 z!R+;IBicFJ2vr$fN}f z3;f(HOBr5JavTWvZz77@!s&`mT{+LT?ur>RzFc%mpyza5UE+EJL=xw2_!OnG-|d*X z*b=P^BX>e_Jt<7}_U_k~F~T$ib1=zoD1k#pu<(I*GKhl=UJxj_fwMiE%`^3RP?NuF zp)3Sz;lCTPApoPA!y5zLvu3S+4{!IhBFkHsHmfvvYd$zIH)XZ<7fw-%NO5HdTkp?@ zZ9pt9-*ek9~2q@VlEYpHu@&2###^Fb}QfBlEc}WyduJcp@^hqq>aQ= zx`=x`z4B)p$9ZX3zr=a@t{}41b~8)3vRwp^rfJf6zVP1t+^PP@A6gQLQ~E;xnPV3} zxA1&QCzf;<-=Ftk=IZ3k(z9fJh@4kp>w0$>v7bnMjC_!aeV8HN++3VJk;u`{&CkP+ zxAzN3e;DM}BAYYi>~U}AKqCJNU<)R18rkGo`5JLA!m8%EpQC2P{yjh04!(IHNmmOE zABA=Mj~6bvI8ts&#Y>vjyb+X)q2o9A?e$*7nzn9g>pq0wpNeml$D5)0XIfQKe)uhyLtDO21C<~35=uY8fD78p4b3(q(ulWOXudM3 zXkDpYN#~b`dQ(@`?z7&9+YOALKy)L;xZ|`mMXM&cp-KtftR_~fW}XO)ZX65zu$Lms zNU_yiUzV=V>z_ZQg+($eo(!q0B`@Ef`DGzpU617%EG%6bYbe{QopuY&o&uvE`POm( z30(<$aSj?FaLnA9e(Ekm<-kL`=2nXVU~a@L6PeaMifn!5JJ|e}Wt`I(9ijxcZIAoF zH#Q>B-b`yI+7$z7=Cq3@i*t5Pn^>)pnmTT6I*V_@%&KaeBhHx>W1=kf1RDr!kkZz? z&xB{v223A32fZD_1s=UY#x+qs2;b;?2liRI`sq0RyW%ZuO~NqXAQ0-7=cGNv$}mt@u;M2RF?M?L z*7ud_t=U~hlZe*py5b{?uB7)sEKAM5a;6i?7S2(-md$HBecAGAqg<7(GAbk{c7ikE zV0Q8|gQI`l1(~c;a`f`MG8;xxRt9M9_`wD+e&gn9M z11c9+6u=4P1HKaNkPllUrII(SF~s1vi9B#e;E%swcR2hSnad-)3t8Byf(wLd4Gg*u z-7Mj$TTQ#ils)1sMy6tDB3O_|h(0|>a+8UfJev;WtRonWtW05IO?y9A#%+7-s12a; z&|nG)kuA%a20(!3Q;Ap|=HehgHH}_TgiRve*av1qh8icqm zV4D6AFRW-pUVia(BU?_pDh&gVY)?F2ZIXlEDpwfBm5C<)=l$p#uZQY$ljwR{R)%Jt z?Yl%qu|yb^C^$FmM!=vU)-5=utE6gLX2GJmy+Q918gG`=%R1CY%*jU!Kh{1T9Ge7Z%nH&uMH_iSBE zsOv7=RBb*l-CebJ>L$+cydih2ays-_k9A4DnqYaUN&9eI+!qC#WBx!|Bpw`M6vu*y ztad*XFjFUpQk()2$W6Ct*(I_lU1WqI(tFt#aqsyA2*1h3QR^*ZOlV%tUi+!D1r`L} zANxXad@^=X!*o=#9ffer>0C9qS5rQGA@V@fqs)GT6r+Z_#|*EyVdlv^GRRDOd^g#+ zoA$`blHe4JGJ@NBE|8+ej`1EQ!QF%N2N&9nagI4cY8 zpb#4>)<^$x{u=jCAdz%4{UFA0(zhPyFc* z%fs*}P$_sGctu?L-gb1brJa3{e_67}rAksz1rWg`e zw%{8GB)H{Hi$GQzqm+M~PktQ())a?V%(Ia5IknGys1#H4eL#Gr$w+P!4hCLm%X{3#8f+YiI^eb4}q-J8< zqP?Wq7wkvae_xmy+|KUsTKZ}z^GY%i^C>?5k=sI(=u?KIk7{60gKo791y^y--Zrr3 zcm+5?tCfL`OrVoY+s7F|B?i64a`f|kCgllOd~V2`j0UC2E0rnV91uh@7=)6TAO`w% z`-}i)*_w&71L+t<^fPK(66$C+%Vf=?qY=+t ztCUjZlsS2O`?}(|zzW=g%g#6K-pw1xO=y2(q~Gg@qsIAX#;~5#Qi-7(C%8BbjE@=a zp8{i*SC$?53Jxy;or|7nM+NTF{DFVjE-@F#g&_Y3Zd}13k*`m}{kb;PzCm|Y-l?nR z>}2+dPd0Hwk*=&?H*AxJ4cMws7Q#l<@;J588_rUx+4iJ3kX5bg&hx3~S8(J(BM2wy zZjmI*PXtGXwNy_tgAbX@fdD892@qr~{r98pA_K!|X33C3b84Y(zfUN|*Q3GqMdxAa z*v-X3HL>#~%vx_T)D|B9u1B$ewiI|Njp@qUfbe9Kgdj2SsY`}GZsg8^My^A>57V|m z%b3p6S(fqt7(1seL85Kjrfp}XZQHhOv(mP0+qP}nwry5gH}`!#yR~y(;tN)b8EeGs zqYod6q95omy35S|85kF--^GP1_7f}YAb(iCuGyG-|2;*&R4S|mMqWL2_trWNfK^*tFa2SO|6b4d*<`0I@HzrIm^Y|G=O(hjrx@(yd`^B^bQZW|Ov${+MGF0ELND!L?bn3tInM3steqEs{Z%3Bn>u?G|^MnNK`KAq0 zD{T-+q|jj(I|5?w#%kVu`8|=No0g3)>9{-=-89Kb;l4*J@XXDYq0WLfeYB zxcRZM3`B8;^5eCQURt-6+*I*Ty>tD#s{QT=-2`fDLm1>?-{c?paUb&^hWKK$@j+xk zXt2nqk-z94^B|(rsgSl1@+lUV0P@X!2O)+}^Y0zh?P|>~9UKuV{0I-X04he@CDrM7 z=g@)~gVIlWBixShA)IN*>b^mfjoJu*JVcTik;+yOj^tKgATkwjPQHA0oJnNsXDm26 z-wVuz81w|MC^9(wF^SP7%mYk1Xjut>pY~3VJ78z` zrPbkzrO!s)=GzHUt#PO;k?w{1(pJ0F;tov(6j<>(VlU$UHiIkC6z{u=LD=5%4<~pX zu+dHE)L%qKg$7Gxv4p(L2r*y> z_2Z$)%ULr+O_OHYI?2z$S3*GpAmguDQ|{iLwzYpdsWVfv!i)y~wN#;zlwZJEOr|m< z&kmuRH1=1BuHm%}Uho~qZL5FVT8JZ@?|c8W8AfXS0b?DiMOChvF0;$RTvoGDIxLR- z6%~ek&VK(&m-|lZ8dx!usTq$o43CTqERS}jj@k@@?~VA6!sMkniQ+Afl-z%<-r?H zUB_UN;QR*HuuP!hY)0-`%aB6Fd&J)U^&v;e$yJ`Rgey|0fag+;ow$>7?MBW-%9^E> z`o!{SQ1}cdW6ZpW<4I9B6jItADp{yU{#EdgBuw(JYw=wBSsZ#Ln8y_A6c{*b15I%k zuf`n$K5jstm`qj3k)I#>On8+Cp6c1R?!kVs;~TMpIevt*vSYtofh69>(bC0p)WDWv z%+i!#z_Kc^#?%LRTrM%^Z4n?CM1P?nQwcdhMJ$^qhL8AyRPvoTb(^nwUd<3ZIyzcgx;n}hzyF{;)fZ&dVcfh~!X6x#C&+IT6bCV+VG~0tGN-OTq~R)oe|%qqG+@S^!Uo zil*&D5w@UF+5M2Iq?d^(?&iy?ZQeuRjKfG!FvgMz4(i`=iu=vI${phuKA+ky89X4l z-lGG_2(61U7I<`TefqdJIDfacDaUqB&`b~0Y#eap;wZRI=(h50I7iYYC(P7n_#>ty z^J85Fz+K00QBc}b3KCwxLFH>o<`QGiL8rY-!2=@ALjDP!A%oF4ty_;kbaCEhaQ5|)gDdz z?I<4JuRj`xK7ch##oYb44Xt#6*8m{*efr|q-&Dm0>H4XW`!cscV<2dF;`EAOR#y*T z-bWEKCv!n5uIJ+>OuYU(+03(PbZOczYbks_#63lFv&;2&iBk;qn{%OdIIAS_qzicH zYz7pjkuNA1_@<3Rst7*KW#=icWj4rdD@T#hFQX`gR5=&u@xYyBM&0pVs8{LttEA$m zNX%XB*1B8wAtu3KK^m;rvSJOP&tfBiNxL4hQzgU#tSeLK>oiC4vM%xpiRg^%crQvv>5J9|YR-`#u2FV<=gCv$+hAlG5VE;|Bor&3mu9&6MZZG_^akI$g0J@$&Vg^R>y`Yb%apdpGxrzpm z>4B^+##$e@_sx&GD)0yY#k#}>8+vW5+)vy|Eq+-shjj~#5m{ z&^5>rmnXT;B#%xSv;b~&w@p}BFXJyiHXdHCSP~E*`3rt()RIW({$M`R15FjRp(Y_Y z?cE>mZ%xB+AmFoUaVCrg3n~exssBV zy5d`D!)P-Z<_TUT42&t(T;1%w`&tlblXf0y39&8Sn5nA(w>GQJR}`^kes)@oZ)>oi z01MXNLGy5hDFs*0c(mZ;UrWB3@ixTC(u|hxj%RKSMvyVBxxdXeOx5bTlFEdU6>0kG zXjC1tSF4SGl?Dn@bYL&hP0W1siAe*gk(W-oB`R5?sKA{O?3dX~;t>?SV^yyCOz=8a#_vzZsi_lmZ zmAd(L2~(7v+^9~}&gGg8OhiWHAx9B>AQrBrP1bD;NBVpInpED#C#uk*NLMX;^5N0j6G*MGB{2q$m zw=X}my)lqJL9eLFkbM=D-rZeeNvCddE1sQLG2vwe?L1pfvF*ovYyVp`)k5iRv*Yjj z52fWvo9NPtgAGYlogZvK>uc|;n}%&oOPjnR7-gDIz~1%~w&!3YG4xx@hO$mKHfLS* zzLxq%Wv8igWCqQ=?_yf&^!rImWaii*H?s|IUypi&PofGIf~*N@JL_6Q(|T*IZjaaG zd$=W*1{8Y_M3io44-_kfOfN9eZaYP71xFM(hf0ggb^61D{2L*se;h5BC^!J1)AIw0 z7D@WQ&Xx@u&)tUg%nPGG1&w&?N`14L&$+V)_--DLMs$9H;ipI)m2tsdebfjN8W8|1 zYZ@Qy&r)3e5*vQr>xL&l2f?P&9aa~od6X2;%@_;`d*OaekU=Av z+5vNgu>sYCrqs-5`EzHB^pdFa<%mR)qYMl4A)SCk-OO|P7)0Rn{gK&-VT3^`+|ndX zMcyP${}LG32y=k5pbWsNp+7x9r0DF0fUy{-Ya1djO?7M*<+HP0e=dA}+Qmkl=M=L8!syi>8FOKJ|zulZS zNrj_~ewg9O%jPi;w#AV86B%UllNb?xl>Z^_wDC{_Vu~1en*a2)TeAUd0vPC@_@Rd( zgvd`7bn4?nl3I4Y$4++Kf^XV4oKw0<%l1^%j?E&bDFfCCW+>>R>b~ksXCbnpBEVEy z4->)o$xKco;H0c30>HSK6mQu3{w_$}v^#vuL#JvZJwIJHU~%vYPN+FbgEDsJnZg}} z1$pnL#A?BbdV2|foUE2-P79> z2zxY=GebHGn5*nPAU&UJP904eEe>1(4$^|nQvoOS1?$BJVh*X9;MQhC$@67yvZbVi z^vFv5H`)x}^W%a<%M?=>e*3Tpb<=Tj>qzid#Hi|pwRdxd0rd&d9b>mtD-rcHbZswk zZ7Z#@0nt*Upk6>>3^tpmJn&(DJ+x9du!uxn`(YRaHR@8o*nZx6(>Qg+1TV977AB3= zSI6f4WRAqvy`gK~Y*1lpi1ujOa|5OAMvgNU*3+S+&&? zKX61NV?rVi@lNa7s*DSHt9Xj1hv^;g&S{p8?3j>+C$U-t$r@QBGM?{#G zKu}BrCRM>L(eRO11%4xk-b&G*Ec9*11rEU5WXMDFd~QUM&!~eiDri>-4bl-v5z_2z zA>NMm11njEXD_woI3NC+Jn>Vwy}bzC&asS5=XL@yVP71i$|bV7o}z6uJsK#$iBmNB zML|PT{c8*~MFgN$4#Neou)X&yB1*1TL;rjKv6_+G7_{4{KN}=LIEcIPd%kM#&;e;c zlxR)a8e>TSPa*hVUxP68^U&d6==y1Kk`+4PKib>AAZE=+r>vb@ys@5e%4QhQ^|*?i zO~<8Vap2&lA8qI>b2-4_LUa#2;h7D%&_T4IMHrR>;O&qX1YgYu;rdvHBKRU&W-P%q zyc=LqC>ez`&Moi~wZQWqUQuHIo`AfDUZ1rJs%M?g`u7ZR(+XOTNnZItQ-Fj<@#2Vy z=>0#3njwp*i=t6^^*PHM)GjkSs#75ulZQL{Z+5ynGKos?E$P%+DT2^5@?eBQsHBB^ zSyIsKcqUA#rSgHUTtjz=xIZgSp%Z}capSRv`(SK_O8%P89k4C8^lV6j*^uBEC&<|w zXInD)fiLQVPHqFJjEn`HY1I8=Mes8%B_R(H|7K-*Z!(HPLAJ_B1_Pt0-mXL-_Q4UL zy|-k6brt`J&JmB^?h_i;V;T$n;q<+^GC)-9fp!Tt2$V7LU&lKf#w~7ofTT)I%8z0% zRuW2B9f4D)m=x>9vy&VZX2xuRLD3?EHn749h68C_vB#a=AT_x9X0HDXRw0!-rkiv! z;=KI@K*f_ot_pP3BGIR>Pmmx;{3)l(i?>-GAEDHxh|^b$aY&?Q%GKiwrx5qd$}~Af zzng?8N(eDie3CKerNzG)Y7mx%Nn@OcQ-2Lr!Lf?Si<#&M^G2b%92gD$Nm(ph_f?i5 zK6Dc`m0ZN7aF$Ntgt9?DA_-Pv@bO?6HA2bPAr$D#6N&?SH0vkIP~lE9OsiN_V8aCN zLZ)%@26$0bMuN=|Liw{~VM$6>fwgiwJ(wqqi5g^SQ7@)P8nj^F;7j8HoB`k$AVuVg z+3ar89e_-v0AfH^Xb!QGA~?^)lzpvtTmjoO4y1bkD1u`fphj3PSGuSjT_QFzsh z7Bg9dW`rXUV83PD`57+;g!=B$iWF z+S5Wb2>i01dsxEr^y`v1vTw4%y7CqQVshg)Cbp3t0)?j$?;Jr~3c-emQ7B{hotsgL%91MM*tg9W@b1-@6?) zF{yUsYP0HrqJ>wKIju#iZ2rJ}xgFR0#aK90KodRQuA=&v%xv2c#kcO)jW3{9iUPML zb?Q49k~=TRVp@X>BzNq(U`3N;11rr1bk&Xra9jQSvpNj(w@Ih%cUKE%CkjK zk4(${69qsheUhnjLG>cA3pH&raKj0<|7x-Vjg*=+6f*#ThCa{6hX`hu-)4dh7R6+2 zYV!wIq$ct{y{;c0Z2NXBVopxc1Q`auRXJSg6l~yJJm0wU5ZDC{ zc+H9&MjjLufGh-1vymPJEaa(_9)Knn6U?dy8PqME0RqNZr8hNM(-WdEIp%kJhb{Ea z#ByGwTcF4z<`of(9H$JHMGiqSS2Jerr~tcseGb5hRWc1}+$gt-+#L*Q>?$&M;+ogL z4+9>9Qp!EoNNA_Ehs+97!zMtw30(7=565DQ$#b^kl_y-_W+%AhFTHIpn48vhsT@7k zOzXu37IcHM6>n8*+q5?{+$g8U>I1GYsNEzTj=$7(oD?~fcK*d-h_#qd+ShbXI2}PG zOl<~upzk{Q+xXb(yk9h%co&_gNVjkNUJr?i07iDhGx>ps7UXN8-ySA*`;Gt?=^-5b znQ64zuG`a(g@s6|iEP))3LM%jZoNE`Etn3~uGpklsvwdd-qHdG!>stJ7$;R2c&h`nd9ECmucqPNlpp8b|_vC`g}goX?Ol$9rQB5ANC#%#UAjT6Vd8l zmHQloq>Xpa8BBpy9ZQM8uK^TE9f{)d4wYlooFgIuB+X(qiveu`!_iBEmS1oe1t>qJ zXi$n5-p42kU=n_ObUDq5^VJWe!7WJkVK*^_@<%yXVB%Ri`0_yAGia3{%&vr<+pifR z<;J|}!7q|{0mu;~!L_qJI4bc&vNk3H106qlBmrm8O#(JQw*vv-g|lP>1Jwqp8~1R7 zW9v99rx3ocMB*@p9p)MVl9%DVX1wD!E?{Uhko-ZH@2Rp9Aq|C{N+ zA?NB{`*EWVcLj`ib9MvEJGS3~&C0D~?@%~#_7#?_a&4ZQXX(-)YYV)yba@h1QP{YA zj+2O>nvA_m55{UJKZD$HG3FYk+w2w{m+NBd!PS9^(tyYQtys=l&dnQT17__p;wdN8 zcw6~W2i3(w@k%?iXvU|owmtW?2Nd%bl{xM|n|5yDuy?_n&9Dfa{I~u36_CM|u3*lu z&huFP<9otB3uJONb7`G_p}VVw?Vcp6`{-e4tXy{9CJa z>Vn*7bDZ*6>eP*k4&ZH%cP?G#lfv{1MW5Todjl=Hx(;+X9E7^^j(F|Co*2P1V@L`ztliq;BjQv<^PuHu>F@zhn0!#f6sKLe?_31 zV=g}EEonna@{$UL1>YeNFdpIGjOYRhw3QkI@%APMy#86a$crkYRNM`XqF3ChSU z$Bz3v*mWinHOAv$h`ij>C+}0Ik5&ondx@x@f!*V0@|-Q)FPEY7YvNB08!|cDn3LM# zw6-x4!t|{!o(G=Sp$>6&0#Z`@_$3dAMjk`-TDqQI_UxUmKv~l{v4^pZA^zUXWkk;l zEO8d|cwx;<^>?)11}{qKqbg)8hOdw5N4XHD_E^_y3=d) zwZ6Tt+N;!aX~}V2v(ED<(uS(4$C9d>eT`wrvEb~R3NNASf;6p>!^3ZZA;I~M8QYj| z4`KtqaUAixb;K(62f=wx$t`2gBBN_Us!q1d!b|YuDS?W*apC9*jgfd~xZ_W%k7O$A zZ{F;`7|bUjU-Zp)M66YkkuPE1&*tRa}hTv}@GIHDn9Y4rXS*ixkf1?I4wnNaA%0W^&z`4syU zE%YYIx0m#ZOjrN8mOrmMxTK?3I^C-MHF`E`Z=a^j2>Zea?j@;$HAPem%!!eBxT^|o zNyLmLD$QQKlv@w&>a$1b%$i}ZQf5z)xerBx?6jRQd~TxrN`N(U1)aG-p^j>7Qa4bO zcsNtVJo4E7lzX&K6zTudj%@)YoJ795adWO5fQT_V!}@k$%qbOwzJBghl0@vxGaW(F z?4Viy#9F;`$mWHHA1jhvKtLG`lz3&W(X6+Ae$rC)VA@^mJhhSDjSxjr$ww7=YxAPA zab&X3bKO^igo$}9+^LCTpD;It4jQq(uu&^7Ix|e3((rt11P6w)Z@a{V*HW%}4#vWS zJX5e6u=1kPGd0=s)L4dhT=j)hh`gdvKOv83Z)Wl0{JyQfk~`mpqbcQ{Svo!1l?jz1 z=~1YWcl+2dv*Dw!rK{KIG7CQPP`eXkGOz&a21%orFhw(|iUAu??hi1{?KuRJ zP*EQV!~{GmQAtQ=qfk_*OBEnP0(K(7-vir$-O?&qFXNMmDucjxb+}wJ-ufoCuvzbm z<2Y^A2q)ARVi#5+p7}Sb#K7RXz15KlmFm8_8Of?$oUK}y<_rPpC0p0xv)HTSP{)AQXZ1;RjF=JHVI*)=a}iiw?ll9d%_3 z#?8E#K;Vl~q&@Bal-84U-$aq4kPKj*J!(EKrU1qs2p8jpB`6PK3 z6uQ{R*nS1ZHO5gtQ7}q?7J%j@Y6QO}9+;*)z`7N96J)9|8SQ-28P(Z$Dv!4UPTQMJ z1VGKlwE!HUKYrXG97rCD(5F02-(P)l3vca$p>aU}>!1^znVQ*d6BD*#*2G}w2s<)d91M<{Qwx_OCZQO@Q3}TX7l;+s z=U6_8Y394$9kN@3{FTUetQD##NE3h@;RqD(igAvUp%Cp_4W(_RQ)h4;_nD73r$`DS zuxk${6P^H|91IP50K-)B*=%6M2asoL%!-&mF;Q)D?3*BFbV@i}h&NY{bhejxy!gS^ zDPr;U;(~FgmsGS^ywPnN1hGE`SHM&b23N$49^r^0ki@CZTzkPQ5C_a$k&~yrI}fQa zy+>gi4%EHYYZ7Ywf_{H}42p4}dH-d}dbS&hfVLz)4$i`v-ny5Mt1CLqrFpNq=%iDR z1bQ~x@}tN~(PL$WS1L`WhYzX?vu{wUM1G&}tFX}+%fumxy`n%IVF`{N07*bb@=8)A z02f${(lUm8els!KYn+gv)JhF^@mi5HGH+n;M@ddQ@2>65ATyj04%|&3-)kpnh@E+t z#An?-p}t@Gzu#t7>;T79FeovM-Sv0?VZ0JQVIYPQqCjxb?uv(|JX2x6+>*?$yEiOm z%Xi$>bJivx+82}t#nBr!4V_l0Fi*99RSbnWl`^qZ10YAI9TjusIMEI0Zj87ZmDRnTQPI|4Z`~B+N~-k>38gU{;4x=OVN;+_HDyOZ&r*{denifjb7meUmY7>>2htr zwp0bMzmQ0?E9a((fV) z(b*D1tBuh4N07dRPIsupyWNPs(1e!$nxDN`?XK?9rG@TJsHetl*TLj*#dkLT|aVJ>_A+U-J{4^=7DXf168 zL8y#!qZb0YqSAA(S@%6Q4KyMM*AP1MyZmGXs3fzMV(xUVzsAO!OGe2*8*5KBv+&(w zh2EXErn`(S@yw6lK1xDJ<#nQZ-LJ9s(AA0G;EEPGAH!Y9^edojSK zuQuSI?gE{%o8@Q2J)=-Pbu zie!Phpu$6*k(TX^anowSg+HfmQDE?P5{P+snly0uK+SdT?r-w;n?SQEd|D;QA$sQJCE@DTOfrMJikcoT}Wej?i; z174J-`e3i2zC&0Th=^#N$zq^R1nDW&4xcnEM+D&55R`e}q!aTp;cF^U%ZW!L=lo4L z;x*e#v?UHBD?T8DLY+MA7T`aAPMlv!2LKYuJ0sXs_4aRi)Yq2)PliorA};QOni@-I zk|kSim5AaPSQ8PXtwGWvRky_Z#Pe7mme!_6rN9CX7JM+2JmuSoVh@3W6_2}Fq z2t|zJ-KoxB^UjMGdpV|FeU-|)XfHT#^%{c%`9gtkQ}b#qePv<}A()MX1-z zKyk#M?CnEbZ^wI)L!ZgA@csR|i#vOCyayoa$9YDr45j)848=`Bd66Vp=Ijgf+u}2K zx(ruN+1Av51^cD$Ck_`1M(2AC3N?=>?PLS_V0ain5^6s;^)IxPD4K6rj$pJxE{z_^vS1w0}5huq-sJL3E;wxibTsk8OKlB1(*up50-Tls5R_^p3Y zZUh-|`e(J&A7YaieYJOKgm-PUa0WCm+C}LL6uJW%Qcu*TL(1d|3D=Q(UWH~JcUnI_ zwLiSIsmU*@Rn5#A?6s)=p)OIiT+kzr_Hk52P2x$TDSq@MuW6~zb5;IG4sbq{hB$nH zqb@ifhv3VCV|W2!MyL`j2aV%>W^_3{7|nutj@&ET8ML%|Jb}bmi)hpeIyyN&g0X4! ztw%V7t7|9xO!X_fZ9~5xZq(9m3?{)HpSLxgYX3OQs+sa@7oEO*um*PG$vj19vi))n zcNeG38Z21JCrJtPGVa;3?nAJrpVk@k+CanTfN4h=FspE7ex1qLo0XP^}YSgf*`zLR1 z|CRybh32hw~W^fT#gXeajQy zmdS*DA8MTRH7va7^9bTH7pEOfXF&cdIhNtPXkT4qXDUd|Mx?oyYDTFQ5pQcjQ+ZIl-l3ixCfDO zB(Oria?~!dR7YcDwVfQfc(pF=BUr8hyJ3vblRQrfkauL8Uts`4G8Qzf8g>>kNL_L< z^i=*oJ)ShcLe$e29V%v{q`V5JcXsz2;-*~oizQyNOJ%|DQ{+i1&fR>_20x!xyonhYg$bRgV8A(v%42Ya9PsAd` zR=s~8G}iIFsqZ>~a!e(jZQ^wEP=ogOPwC|kzd#nu$NbE&$?*^dgAc7V=%1FuumJ^gN8p%L(0d3pKg%E*%m}phgM!)wYN?vF` zqZ#;$2E~x7hqa-S`PnmZ4v|@JyTU$HVSMssZ z_Df0DnP@y3MRDoZ)!COCKfO|!3wU`F9p;wruzFu~o7R4(YZ}=qZr)o(H8G&D_rp5_ z<>Xv370%Hw;rdUxa`nF0*B|-%KuvAz9}1LBUA?>ksBj{Szi~H9Q{`Uwbs6BLl%k&+ zPg6g+|E7LWJVXIMalyZh+K7+Pzipw>K#+) z!|jyS^&!4w4tGuL5F)CFZ2YJJjzv|0C_3VAT-cGgMGV{#6J~(}3qJN|ul3@7(GeE` zT>rP6iT!`(Obl!c4F89mDZ|=!lNDif_e1G*n;c)!&Ff@`#HOwotE1&U1n{pOdrXF# z_|=k&=D*F`8{&6@mPTqx%Y1%h$))(N-P}?x(cI1kznUW7r`dgLF_&7-T&_N~&ZpRY z>zB4&4zF4jHne$))ob4iWjTJfAwiMij-6patg@jEX&Y(-|(fys5n|=E|Q8ZyD{4b>>nlVLVC5&N&^H#LhkG^ z2SzQQX}d@vQWrZ92gkSrdVbh^sb**2KZFtO~6U&Q#hZE3p78CvNOapMs_uwj}8 zBNBg-Tl9BjW@_R6GR!E%S#8VgJa+ugcDOUw^A3(N z!VPsgZ)PmGlR#+^YzfbM7Nw>n^Pz{0R%#5*3Pj@{p8T&X5b5LutFdHFrMnkXi z9tdrSLjhL#pTQ2HUHpJZCMdXo)h>qp*f~b^V;GO?LKO>EPbYD;sMK!F9xnu-*aI`? z^)p-eGYH3mcN1ago$AsWg0;j|$w^5xRO3S`YT{BAl+nn{rz`w5m6uPvRp2y~IIl^+ zh{FLxNTUgyl?+`Xhp4qYg$*a9cBKu)(^3YJpqW234g8(>36wLI8a4-nFUBVHfu9Ak z#CcXkgbxok62gF{w-)WF`AMc~<_seD_TuLdm!w8w8Xs(W044?5L==N`>_|iVkcGlM zA!mfy=*=E7JrHMH=(4UIIYw_I*I9?IHY3)rK^rYWPgY;pc9Fy{pdmGug%<*cSV@?_q{V*paM#qISH6i@oW+>Mds@GoUF#JZyJ^>4{+v4V*9XY$(x9lr{wQ8u(8w{f-qXwdby6vy!m2vfx2jsOou z#Vpui|1mz!>EYoB5BcClAPViEoq1xVTOD<`IprpZKx@Bu@4pK^IY$ShznKw>jfZSo zDtV!feJrBsM$c7_YL>(s0~l*9FPk29;68MoyR9a-A0Z4pP05yP(5rJXy}fhqKYZ#m z29T^Rs8(ChYi79_uXh9bSh$%(Ex4f+Z$kYQf2yn`h*cXj(9rU$wGHUB^J%pADYV~| zsIYZ3zU`*}Y-Znl8_S%F@oE5L&ub`oOFHx9?oRvga(DfiuDzv;*gtx<8-jaC^OF@z zEfk13$MK#H2*3H79^P;IRF<8*ugj{Qx{R+q{Y**6m{Jd9y<~Ea{6-F}Wb`&G0oC7b z#yHeD)dPY0GydNuAIJZke9Y{u|A(#FU%RsF=9t58JEjiWMz`q-wC>35Q6Dk2AqR67DYU*P-~s=HZ=LJ+F_wqtk?@S5sUMwo!wUHZ%&I{ODl;-~f_C3SD?U?1<_Q6UAgIBj~wYadU zWbI9q%xG-PSu$0s!Q45Q#4^Q=pU2!j{a~WEzNkM8nU>YVofkH&r{PfZ))&^Rz9Wqi zOuuI3ohRqP>jY&q!IIr;Jmm_Q+RM+Avs$q~z}S5J0esj7!{O)=%@QXut75Fpv=smD zN;o=v6lahWhe?^3D9o}+gB^)a^@eK~*q>h3w+!c$e+fI5%u?dIEz0@TOlzH%*9hA3 zgoA??iv)bJaxup?7(N-qVxM-)mZz={`b;xS&XV@mYE7%l4@_IEMdjCo18gWF+OMEs z7e%MIr~og=`YQYe>4j8jZ8ccu+l^Prh*+IlJoEY0&Q7ib3J(ja4j(Mu;`CSd6FX`S zuzm<3Jdtf`fwNXUA+*iH!3@ zj_O&?-J7bbOdF<5EGe?g$VJTQ0%uvGJE_uPOx9}bzraHa%sAc(UDu;}d2A)(f5#7+ zc5mKi;1fzjKh>YEf0`yQUf5|fwr(zZYTFWx8IrFZe%`zol-BZ;f`1#X*?S0Sge3|N z(b|!JG;DosD|q1m3{<5QucuuVgx|!4HO=U!^8gP%dn{d?3`wW~TbN>VyS*L^tH>gI zc2s%rYW@EtAa)xR+jY#Pw`I=9pO!lqYSn%zNNmjBl1%eoP7-)g|9n2c!U|+6&{r?X zW*1LA2&21mOI_q(MwJcoND#}|SK!E;7Y$9CvT08I@vvl$5r(ujjsO?>12O!?PLhZn znmBGyvC71$o@!avoBRsOqq~RwWtkYSf|0p+rncR(A#881&}yI;(`4w~8rb!$SBvhv zOzSo_fDy&mY-xo2m&nfNbhZc}u4p}fW+%ConW<$|v1x<5kyPp|I<(5NPs4N4V;_sd zTo2as3!y2Qbz%3xY?#NV?43nE=x(0l-aX{+`jsFur!8ldF?^7EY{QjmRsHDKBAmzu zWZvQiDQJ%|+$f}zS)P{5ljD^VkhZKJxTD`6n6+Y6^LN;^1X}c!qAF|FNhrjfLH*zq zk@y9s{36AH&(s;Q$*+jX%pgN06v(bYw_raI7t;kCWc}iVjag?nL>N=#n6o6|@Z*S< z>lQ2E*}-@!jti>@V!Amp^IeWU2we%I%{lw~y=gqtrKjApW>aQTqN82eDzl*XaundS zK|nyT28a`KAmimtfUBUX{Q?3=TGK9Ly+}&jF_8N6dcfh%bo0zKgO*`!rwA=gpG+Suv@wd8 zV7t^~ou%#9X!30IIARuW5RX{$Vov=eO-2X~xyGF>1~GZ0B$LS`DmK-te$xq_0h6PV zX#eB*Zc_M!o!i6Ju zLq-lj$LCG@RChLmN zSZ7IV%_uodO49Tj4zDSulT=l~e{~8!sN|!SGAwv#Gd-`181al@UwO=EVJ1=kqRgAO zcX;?X+}+&7eYDuT7oJ*yF><$(UD+p)7ap9CWi85NHIwdb9a=wM!OvocabX9_=N$r= z*!U&NmxD6Qh*)~voL!t*&Q<(jHOu=4;JVe=yN2@*H$E6#4EzH;W%0yY*J^-R$j&;< z`;9bN9Ta(XhXl|sU9dcFc^JweLA&N51<&280jQh#XxdCq`#Ik(uwvKsI4&>!Hdto*zrqbKrx)VMj8E0l^56NhJ403h# zt95p%ji{tNyRS7;GG)B9Uc1Kw>P<#Gy_Vf63%ev76BH(zF0SPU#7sh7mEuYac})`jJi_4F1TG+*WvZLO2@#}sAgdB{ zU_n$8D>7Z}=SJBmG8%H^FgrM>x#n*Nxt|g}euoQ*4AB-4X#$H1rNbF{`SuMVkb?Ds?}>l3W(Z1Sd(@@*8{VCL5 zo{~?o^z|_Pc7Eb+C%?xz!hIk@=RE@h7K#v5CFjokx6`HPYK6rwVuZ<W zTO-0*1g4*Svg4E@E7qs70UnPZqDANGz-v#H^@nx5J%5O@fmkxe`@Fwc`hoiWl_E?u z?blg&awp&+;P6{WX1Mh*4^HN5Y+eH9236Lu4r2cv_b8u*EnS~HDG`&(6VcgLF1Y65 z;`5xUEE`(Qs1!a0FTGpJL%z{RA7z>%GZ}Gg%K4H<4vXh{^;D;$ z%fN0N26p1)g2(k* z7W{f4-p?39MIxKISyS6unU<&RO8zDgoCJ? z29|gv(z-5-0|2)}spa+m7(1un%%VkG$F_}5(#fBWZQHhO+v>Pu+qP}nwr%rf-?v+5 zSKafps@C(WRkOzY#y3a^Ks4Xu!b}OwR+v`#hiXQq#~}>a^Tnn8qxUMhepqik$>U4v zTbFSlF&i9w!4-iuPY-Vw9%ybPhuZ1c=nAPaZ?rERtRQuXa?`Pbz%a2}lUnH~k&t}o z16(gH0W?&C@qZ0h%Bv0~`i%!)n9_kGu>I>?$K07sR#i@@b(>3H0)JZ$G_Q4m29!BV zkTn{V{R|HEjP!}_RC!e`#oo@OOzx%AA7Yi!NvnV4d9(*a|3E5j^eLasIzSU zIZkmfEs+TwGuL8Isc6`@uG_d!0|lm;!T}Sjf_qWgo@-BUkl&MzIuhCx4@Nt0|8?)U zsggLszz_S&suQwataNNZZ>NqTShZ)zz-=j zOSHY`Zy$Me5Fiq>m@evyhD^Miz$JkVr|Y&BB^ebmO{dty{HR zA<&ii*n>luB0d9}?jSosT@-KQ3LmYE_GN%F65ncToST5FbWb+oe=IB4VWD|xy*J6$&Xc>h0>mdHons6U3PuhrVPTkk zDn)<-jhV=q-{5%*P6vWM6zj3>!J33@pCRnsw!O{faV>mGoR>c)Rfefj-iW(%6bqy6 z3h%B2bMo5-hSwmtXB_r%8mdB%LER$C*@n;Uo|b`9I^?f)o44NLhN^en$&xQo)!upd z1i_C>{1$x}syL_mV1Nx|WbhoN3qNiVZ)tlu9mfPKU6IDcwf&wNE<3Sdmn7DUSA~ur zl4${zoSO!Z&6?+Fk--RpOgqx{BT)~ivaCCkOCvl2Idw3<^L8~PvU7dC$f`VrnqmMn z{+CA{kp&(baZzd7OtdY3hfhrT$s5@)-#niz6n@yM;_HwW=9`{7Pp;X(^rOjfd^Fhn zd&DZhaa>g5WRki7WeXNuaeRZ3+z=+jf$cAjAORHgksi*+!X>cWL(i3$8xX^-&>#fY zzeLY?0Z0)YvCU#52HilrOLHjG$hK(BZ799_9ROAFSJ;cu7f%Sybra`Q>0YM4qa%Sx z{vgU>CKog;Gn-V;H#TdpIf4vS1V5JL!j>3o%^ZBwz26{t;BYwrSD3{^p)sFxck?Ld z>{s8VPN!Mfq<9aX3m_VVlz>bJ-LTu9;2S2mSopM%3HJ5#9w`=28^mKj_4A<-7J<#(6o+)~y*5SaM`?H2XqFex65#h2U5)>248 zbAgHqGS*%T&)mx(=m}U>Eb;cm0iex;%U6UaT@iVuDV=WIV>)n=&uM%s3~1Uag0sVQ zI&v{|=+=dHwzoxJmV?-5ofLi;iAp9)KDWgfvOox^@Gw<9#MWOWZUV=52dv=_5b1#Q z0~>AXu)tof3Bp?&As6~xvpuBOKcKprBhb|N+uqvsFcL#4ur0rSjdDnPuQ0Y@?3)0{ ziD}Be6cqlC{5Y!-xn;h2ak|XwP0(yQ;)39mPgZ z;MF6Z%^!MJdD;#lqfrP7`4#EXr)Raa6}YC&U-J#g$Xe9=mS{MluL+!%_!?Kaq;hWEm+m8;Q)c20*Xjx=jk^OttN!rRvFTUTMu5Wx38hsxPs;HX3+QWCkncSEsCQLG72^JN zAkrqJEA%oY(n<2MRNM3B?kHA>>^fwYmnMq*vUGN^gJu)t^p3VlkY({!$1L^&A5#l? zS?jY?>A~#Ug*@gi_LCkJBAoHo`2!V7>k$w2_F5JlbdXOJ1j&Jd5^jjNfj78;TB`kG zS1fvH6Bq68*wQugPltlWDMIxqcaHieILyyD`+2oq_HCGn+oi#oX}L+#Kx;oA3BtE! z%AuP7FwGy9{sVTDc!0%`ze}((!0qT(IOPk~BZMuDDp7XyV79)0i_$Dc`5Di^W-Iq` z@ZD!Y{WMe$^>NUNlc+DieDStxbZpC`jZ6OIDYl(xW;6d{RO3d^S7+oKGp9(9a=g8^ z{-pH+&&3yc@$xxTWX~;OC?NN0Ki9*yC&kBWQIwElUKNb2ZS|Wp^F1Dp?I09T zc%%hCQ)+AbwTA)2Lf_L)LrN;jSYPLucW}$^9a}$>QSX*fml8r1^nJFY+jLmAW=}k1 zzyYMc{H@%j@7E@{Iy$)c%-9t->Uev!dC0t*KHN(TJWh*e%AdfvOT%%bn@!s{diBnU;H! zh+6~i!-wefQvP|{%K72$ZRem;MU)s?rZxW@5pqM2z?fK`zm()!Ag_|UO$R?Tq8W2r zJ42?*H&DiBMs%EXJR`(s`5yE(4%dv)Y!}7SyR3DPQ(h)PRsbWRCH2GO)54Z5+s#~g zSxsx-j-^(^_|B{>o++_+F8TfioQ2NqTjpm8W>F)ed2rK60@XOrX54vmqrYAp|37gW z^I_Z?S?NTORCF^UrE>6+vyb;pGblVLo;L)vk1x<48MPl_B@JFBf|Z3}-09ck-Oiaf zyJQTd)m*AP`#IC703Z90@v%VqvpT-?FTdjH;Op@EDAQl(0ttY;%Ak=&L69uJ7E^kW z&x_g~t{HO0Ncrp2E?L`{^p!5A_Y{){MON!R!>(zHqa4s`=AuzFwiF7E) z&BeF+H|cM+CF!n-a&`ZC_XGvM)1>jX@2}QD+Mk#$Y#isH+)i3#B7)4D-p)+Js|1&T zNILW7)JD5%ETm3jMxlry(A$qj%AS|y24(8IrN2J_fgep-(rI<{DU?@sC5-bGW!A$pe`b`~Td5jPmX$BN); zJ;eFb{}9>K(ETNn{IhnH9A?e4revC|R6=e@GX~Nk%%6Kwu3J&OSZQJwYi+*;pLYd; zN}{!a_cA*@OxcY=VgWVc7pua=Zwv^S)Sz2=@3&$04BU*0q!wRty;-}XIG5Hg>Kovu z43ArHdGvFKrcVjYU`YF~QXLz!*;Ez-c+cKIkW0VQQ!1vBg0b1SQd+Tx3Z2bIauG#S zqJ_3MIi`l139Oko>}nn3JU;Om=w5-T=_wzB(V>5cOz{4E3~85ZHc3WvBNQZ9)8Hxz z?OyVJKpAmxq+a_v{}R8LMIjDh0)FtoSSE{sHndDiAw7smtXdyW@-1GG6Dc{V&l z=6pc5KI~EPZW=~)fB!UKO77VHgv2o%3wM>)l}U&(+u|WpT9DQ9K%Nu8__0dHd7g=%}Fl;ynj2m|&P%P28Zk;>4MW zj{#LiGpjB!78fk6jWZo1EW}y}NsmpogcYwci)06MyTe;;%QxJ)0Ad^n$kaM!X6qmJ z6Icb%=?%m8Lo%EYX15NgD1Zf4FXid49Wg-@icWC(c$#vi2)|KpPPd^dAg{=2Gn>>$ zD^Hd9IaI!RC=MXLZ9iUt3L#@tx_hWalF&bukgDtaO2w2g9xqb422 zB_!kje9$n=8i?)O)LFX^SNbCAyK|L|Zp&=GUEiJ32DUXistAGh zhBwQB?X5P?mAl$B^WM|bsK`ck0A74`>FO zEOJ!3mpV_(AJz*T$UyVaz*)?8=flvKi!tf&kfkGE6;{uj!#5wRb|@eb#E*Y0E2F?q-I#Zu2YKVf{0MC1>Yx ze&d98%-{INw0{9dy%r=gWIp^<+yEMcX^KcS2tFX-^1oIV2BXJdzR~&2dx8?4`-O!O z1=I{h^@zC;VJYd@@q0k5(wXuD!7Yd{4b5f>WU$Lon*<6C&E)vw(Mg!>zAx0xfZAvH zAGNIbRGy_j!Od)-3WkbNS@af!{+*-z%(;jpSaFIBN9i*YlIuz50CEB%^3K8bk!Y9K zP%&KYb15MMFj#k&;I&0eT&3cpdsl;aRSG|L%@=p*$D*G9W5aC3y7b7l} zJI%}R47vBA<-(G1fu9$3zMB?XBN>3w%Q{YgXTBb;=Yd3n0BzTfQr zi2N?BV*l}M+~q+T4nEz{UkYayz<|spebTE~G#yBOBNQ+M?1s`B!rV={CcWM$hWrup zq`wk@tYZZkwco&-5WE^S)qr~VMjVBplMpY4^_OLE>5%J&LL03T&puKWj}TI=w-_Zg zOff6z1^>73o<&u-30g=aSW?E*Edh)(!(+H%tyisC_+S#j!8wBOc$zvRK-TJC&e%+9 z?w^(U&$6l_YC-~V=BSf=lzXqt{rLQoR&7>co~+0iKJE7tZ#2BZp!$hD6(YR9^tU@{ zynaQpmHNdl!?}zK-|a;eOE1Flprh%&3E4U$K;trugjN?MTf2Snv6(eoP(TJX7>G~-^HR)$954v> zMIKt?!3dNd=GlxrUQ*JO!s6Wk$6Fq@k%*$`t#NjXM$!FFaUP`<%t0Dx8`_rmfUM6$ z^O`t*YU=85j*BVHyv2lrM-gU!-FDN)tS~gRv{G_( z3l^RqL+(~E-oirg5rSA>w4zJV6lL+4um{KUTt2!|6uG}C^Krss*(o9mC7xYOyUJ#~ z>`y$F|4nshc!VU@J_+juRRpgQ-CIY_ETh7Jr3`!E`b7E1{(yoC$j2gTu1H#n4z`P% z+biatOjdtl6CTzMQI77>d6#NAA2MgyzSf zr1!|ZJytAe&e1G?bPNP%(6jIX`2Jeu&Z<(-G=buqC94CRdzpP z%6?ymO(J#*{#N2l;5T#Ah9S@=##&_%k7Nb;BeAq)wIB#_0~6!Luxv;qpY0$~V<|oP zLwmMDm<%F`>SafOnd6tB#w6i-{95bgOQcJ5lj3>YYli#L?&FqnX?SWu7Qr|c>^QazQgQEjC<+r zkZ*N_zqag{mx_OwmKT2=GNTBuMy3 zXTgW6Kq$f&MY^+4DT{Fn5KJ0s5uG9GEr1GQ4k6R6L@K0&!qN+|eU22&vB^)Icl|S< z+nHtZ=fAi$_5+Gsf7`0j!DxOZW$2Y9dn1Nv;b7d^=2Q1~5|N4-i_!~-Pg2tkp@YFY zLFT0jiVFemH^Vt}6E2VixOt-xslXm~sLku_~M_4VKyn{&!%Io3IlkKClAM8ec zb)^g7Mi5FKx)>1VOu230kgT$goE@dkhrNJEWDG$HMa067L3y8)O2ETC!8_tU^r^#> zet7%vor>#nf^q~6TR;Ihs}}F}|0!nP>MfZS>NPtA<*gxaTh|xkHIpga2(B)8Q@{HJ zV#=YV;+>ecrydVafEzvxZ=+$>v!3j{A4oqVU%ZY_U`#E8v4_Gb*8~Bno=<*sd4f^> z4ng_I_axw>S5;Cy)1w5@S7F#M@_O9$D;+lvs_L090RSbKg9fqS0#<65(S0P3;F zpJiOI#D#e6oSd$M+Rd1h8+75EPwt=*3104?Wj87tNU%D%PYMcfVUafsQ|NmNMf4#e zwoinxOxg+=C$PqA<)ayY$D`3OG})m_yu*>~IP}Uh!e;Mr?)}4AXenGmkiEo70l>ikC+d?_FyAf`Ys!5Lls+gT^4cx zi#c5_YP*!>9y|w~th3n{SHBgMr^XnJt0KJl+i3F>BI;DID*NN`c3{nxMtC!!ZSK@q z>6KB;TXuaT)Q4d%Ba33j!;caVS>WY1&Q4CLhs);f8T zZKdHUNkELYU718R2G=}&IM{g~6te}HNCuP3B-x)$xIBhL%u)$L2_^?977WBf9^M?9 zq*G`~zfsE1HvS#bGgZGR6>^=l=T$VPkGleFp7W52oAGMerdg*Mwf|*(#G#KR(5E&Y z?4EVPYKu?{=a)czx*E`$lmNMgI>?t{CY33-XUlp3*h857IssAJCKiGR)$8he8_P3$ z6H^L+NWM34pFhP7LKLnKC4#5#eFne7jG-UOcPm=E(18wvdzxjY*SS(v#e#V5gD`@z z^{EVZR-9Vv#joqduT80J#**g56iluy*T>8dSvpDWKHv7Vzx{OmW-bdyN;b8im0R89 z98vJ>&iog>!-*Id${8E8v1J2cTOy-pe4 zVRBCGpOQqr7*C-rsE^mSQ)_1KS=2<<0C~;!$X^sj9uLAz{M>K?t?euOvW=Ze%Rm~C zg@Bf4fKVqGAHVk4Tp4~su@N3UuS_CaO)Yurso0c@f~t-R&Xe`<9}sLeo)=YNjC*<+ zic#L@b>j_}gGcng$M!qO@cJRcU2B=(ZmgNZRA9_PlJWf5954~H{t2Oh-@om=RbLBv zri=Fa`CweSA^fAU335)ZL$CDE(7+y=m4chpZ=rnFwMd$DIFpOH9i^$M1I7z<)qu7K zV}ED4F#FHTLc+7Ix3EWyIf*JI2_0KpW!Qf%Dn2d8vZ2N9+qnQDJW7uu$=JHB2=6^Y zM@~O6VW4{fsNB7O-Kv@xJmN&(w8qcv_`?KvZS;ijKZKrrSe~4z$bW_GAyH zy_K9%84tVc#*?59i#^(QHL}4IzF%Jx63s7=L+RBh%PJL)(X8%7V8eH{tFK=SFQ)N& zfptaubw#mvGbv82&`F=Eq*!td#osYf0w?9%i5kOV|{ZFJg8lPII8#OOPQ2apt(Y}ILI6R)O8%5g9 zjqpvyGE7?)C9`GjtVJqZRJUIqE!q{rh&410tEQ9Ke_eeXzTU4ck~@uw zdN)XAj5|EExAFYHjNbHr7(IYi?m)<@Zk5h9E{&{%xw08m)z1g0R+;gC89ihD*Kzfa z0QktQZ>ldtmSv3y*TqSzDOC1Zr^hPu>l>>%I0xiu+cl5LA4ZRZc)qwlP$F^HIa3?| z^Hj{xZ=W`1r<|mM?N*?cWi)($GFI9;mTj|Jolu*H!W6pHAU-uUp&3Sbug!NV({5nmBo$9j8P=RwfwE$XZP<(f!d;)R}x^PCydoum>U23Aa9Mvke6(0 z3JpW9Kp|nB%r(|+px^lPn~C&Eo!LQK+`)nM8L~x z9Nwp%%@L6H#Mt>PX~FZh-H>zUVphL(eP*J$7il)ts=zSH;L4S3uCUTK-I!{Q$EQEw z(^-Bz2^nc*F4+2_MD5~?B`qmGi)p458{yDtwdT>h=JXKtWIKq-HUFe&@$d6}$viFW zGig{1*6b+5*I0w;{gPQE!K@O-9)H+A@6B1YaoqvmWIU5%{VyS(CLSy`lqIwP9$~kP zc3wt=fD~OD6B{4dR4q=rVY*C&>SI^AO#sF*8zo0ZMsV|9A)zaQ=v}~ACgF{;Hcgte zpoF!GPMmhDb$xD#bN?Xo5~brj3)g>y;RpwVKI%>1?T15Ai}5Zv z=)04%i{0N;I^RoQ+ra~|99dugRLw0<7Kk4#>z=R76TEriyn!Vfu>d1q)|~Z59H}Ui z6Lof#Y7)taw$ZSsj3aEZT1*m1QD-TQ})TIQj4>l1gJqwwHgq z%f~0Zxqv%asm_x8Gl^7NmL!k^*Ul7${ue3RfoWPiJ-t26)wZp8L^|w|&-c1vb|STb zve%P_x5|p2%pzfC{0fN+2gYVElq&A|D(#NIN`Znu| zkC8~Uc}m^yQ7J?c6vrGxeYaR0s0Se880h}%J3zcKW{?xN2t@xCF?AGjv<@(CK|_eB z?~m2H3?BEFK2(#Jd>!Pk_hg0u{wWSmk7xfz-8rVqrnSWgn}w(1L4=K{evqO(gtrI0N_f)o1^caBw$xK1hYfTkP{7~Vm6aoDY>&nl=~c51GL0jzJxU;i-^! z!*WyCwb-6jtE|0P-#K?LuGeqC+^S%Rblw(=Z!%^pJO13)^JbNsZ3Mv9E8O2J z_1Z6m{2mDC9yru!6QKV0Fji1 zv}WO8HRyOQP@9hy-;+iePl#vblhBw3=g1)9Vx`a5DE0BE?q@J9D>K9E^^WXyLBCp? z&|1$@e4a+urjfH$(e;j#Y5>WYU>f&Aa6YWiCZKXX*1My;(?hYzRJrNGj%q&bI{u{{ zY5TD|kROSHxZ!tTj$0He;FWE+nf#w}aP({0b!Hgmq{FH`W`Dt#gX(6FknxQlUFap< zWjQHATo@rT;)7!82`+s4z+u-PLsl4jcWea8`3HSDxX7|ckGX(GaPTwum&LBGsvTWQ z2AN}2Pa~2njaX5aBe`kz*K&HpaXhOk0k8Y2Gav|UH~AN@f^YmhAw&VTN~sGIIP{$4 zE%7POl3(p|GIdH*K82+T(K+jBb9h{;N$k6=1C zdSi*!=QGoF%R8}!xX7#SZW=Z&CiULB=QCTXW^d6{a|I{pFSS|Zdn2~idp7K*n>KZ6 zPTIs=dkfEF=23Kj=P|0-p@ID7#n@%QwtYa1Akmxjg4Je?3&ZFxDN&<;nln0id7wv? zfK`Qc-u<4=TIX6~xy5#!<<5a=?=Gb$V7$qCFweTp#%I$?A}eD=u537-4`Ut=w#AHa zaRS7M?vICmOeg(5<8Lum){4l`-PM1^5OVeBwVinjyo~^dzRcdxR+t@5MgEcF#|07q zcWFIo^L_`SKQbUuD#BDjQt0{al)}8YKCpwUl?`(bCmHCaxPzZ(ZjDt=MW~76kv)|c z6;UE}Ft+Mv_~xcEx;_^%BM(u1Sy^0a(A9s4thN=I)@PA3mVSQ+krbW zZcNW+@y0(*Qpf3q%3*`pj=LLuBU*(#*n2Ued<(8MVJUhhfcj$04=3pGp|a(saB!g> z!l4U?H2lYNcwbbkD2U^X4QDL?3Y%;+sI)g*^8AaigsJuq+ckOU{Cu(C7-Wq(iIu{W zoDBGqO$HYgx0(bYaW)yKC>i)e)TWV={5wu*Bdi&g*|nqua1}|W@Qrp=Q@&f{F?RYL zFq;(e%Zbl4rt<2g=5LTzSqOWU)cXVxuI;EZTX;)9TgdP}cVY|3WO!iP0NG=ek{IX_h)qWQ05_v0-whyNFO6ZS6b^ z{%b+H@O0cKb9XdMG$l3|5gcr-&ft5D&<5_7e>S#~gH`4?!Da=rhlaqPpM2F=?I3zSv;UJ)Mzh`4>gN+iWt`AG^KETq_ z^+l4h=0!^SC{GA5K0QJ7h33oHQkWt7u<8E+#wiKJ)R_g%oRB~r0lGL%fRuIG_cS(0 z(KNnI6!MO8NehEjB%OOUcN>ao#;H*4POn>47Fq6iAwY_`gPzYgW289;I{L2_ z_mfnT{oH0}?ablJtMethn$KlBA4+KE8OZ(lRMhXVYE^;#FR$Dcv7epRU@lf#laV9i ze%{7>$XP)vcHNABzfko2aSs)3NW@|*;bV;y2J(SV>ILozP#b^~7=Ihh1N)~f*+!)d z3MZxN2`Dx-s+O{S!@1t!!ff~EG`iUfAgeJ`b+%W9yCq(~Lb1LgV=%tmktmU`6*-H^H&nF91&(@T>N6qd5svo;D z4bfjtP=5fxa38~vU|S@W?snFy4BCdV+T6$KH9i}Y+b(-noDc+GLUi^cl z{F$PbI|@0wf+hN!loTMY<2v0}`_q(x)d25+tLn0m6jI$r5oRNZC%Ex%bqoU>fK@N= zYjE4=*|H%i<1Q62XM7=J>iFl}_V_@!95V>{j2ll{?0CIsFn=$oxYkyJyWK;9apXV9 zzPG|=p4e5L$;TNox*eHlBa`f*_so9Fg_MgbA)7xz-Rup}ka~R-q>M~MWAMpayCL3l zY0pm}lA1G%1Q-l>`jn4ZVnc2`V=Xv4uanD`J3H3j3UjC6=OpH=Cu6cFmwIbNR{b{5 zj#_&FKFNO`D)H<2BtZvD^IZwiCpxq3$4(0me?PDN4l~BtmWKEqlmw1qUguFF_Q4oW zemB{M2W4T+DRySvxv_f&EGf6Gc6*~Xl?W9_7X<;VW?zxSqHW1G;lR&#X%u`5MR=0f zShNCn%U!!51j!Y_zULuDh;YC4kb@nlOv>smJqF+l<!E6y2%rxO1;S0ZqMsQMbY5L7_?8CP1wqIZ3Zn` zvq|gIwW)XP%MR$fwcGV_Oq;+$s{Rhs_nmvo7R#%HFMax~cV!JSq zm|h>~2sr=T+*oK^G7ka^YEgYBL&LA1=W-3asw1!S21I>{&#I5ivxCQah2-4VxBJ2N z@4}GbOCx78df?tfG#wS@wQ*=@cpZ+b`~d&dQi|_D@g7Jca~kfFsX`?oP=+x7Ec8A` z%&T`)EjrTftAB0_bQav=u-ZADe{`lI=g<8;UZX8@kyacgqV?$`7G@btc0r#7?2r3+ zr&Hk_5DFJl?CbR?YV9qgNWJcZ_3d~eM8}(}5Gfu|Ch+JfnoJZayzGw@-F$dfX;a*| zTHhJU)EPP84fYupF3BKaJNw18w*`uG=7$zeP)nlvBnD)HSjfb}{eHxZ!pT`0SvZg4 zkRUk8+gieJepIdKvD7FFZL^Dic(2n=Jil`^V;XaR!ypn@Zq3w9u%7JPI~+4N$OPl?7J~oWzY7;>d?6Ir-+7l;yJQh0idp8nm$<;&$<KF{}Z@O(H&E^d#1nfBs#_waVQQwa~7 z?m#wi_}r!N>Ls5t?SrMS9WqroAANia)$ZtYbFt^>{DLA}?PZNKox+CQ)~1bGnZF58 z3S^ZkYMeB7+s7|2>NDAL@3krNZTE36W=~hI681{fHgx2eS775w9i^IXU4A&=;`WDz zyXrRl=E)QXtgbt7X3C33Ew-0crdQilI)J+97q1s-%zCO%;qSj@zr)7D3h#4t^Sm93 z34q4a>h&Ir((O@U37=2AjJ}B-ubIlYxT=V%BCe$dD(lCI{998$O1&qz{D|t@rz`E2 zHpRi2Y&GVUHg!MhlvAm97~9Q43HGjHnNOX=8!BG8r}}zI>ZDpy{$8uP(qXJvp_=p5 z3v^a*+GRM#qq>Akk?+$ir&ZkDDeBqIE?VoeB~hC-vhQrcn-D(mLq455>#Aa{*VfGE zwucIzEJnFFs?ZxZSXOaNNg1gXernk01!4*rbdAeoIe($RyfBcr>2uwxZ+e^WF#D4s z{-QJ|WkNjilR%Z^JIJ95ZLxlmgyO}+&cd1uX&$OlocHWn>&7?@ZW2n~V zQ8p`z7P6Rg=sNRjqyKwwT)44y)8Ai@xpTSV%Wr55+V8S160=o*qrYRH1at|g-ex3n zkkSX<(p|-t$FCLIWrnXcH3S%HxCG#*WrK|a@amN*1~~!*=}B`6H%LN?RrFk(#4FVD zPL_9pp1J&;;eYtkRAy5O+uP{DIQz)kX;_S@x~zC^FT#cv z0b5Dk-Uaf^*8`1E(F0Ev5*LScSfQrWJ#6Xa)mS*%hl8|001q+D#w;g0_G@F`k&G+tGoGwfYMl1WncsFD$35EiP z;}AStP$n_FO~g719TgGjY+C7u8b$ITC!Yh`b%l-1JnPadgAw<3zuDiSlhVX2Cd6Us zD$y4$w-!i4W&}3Wp94DWy36~)fzkTHJ<5ig;L8!7+J#GUrK)t3_1NasLH!XqVQjm9 z-F%GSmfnsI!X(sXr7G4(pBo|cWj=9;WIQ%sK<)2+cc@Ej7FIBDr^xF z=1`I1bfl;ee$2`1!{O5j)JFgmcl$Q4@?o;Y*SlFsPZ$2R+)Kws0veK=O$^*>SyTi3 z=fv#hF7OP9X?vX+Q4|rKUqI3+>@?4Uw1;F3voIXuMyqb)BHU=g7D(Bv+SK~leS!fB zio$kZ$4}`esXiU1q{1e~JmAq0vcm$gLKl$)(mWHC(K6Y-+AD!p-D|%Gz4s%$+$hK~ z6#oQ<>v_h-4)7;{Rxcw4VpXgr`LT;QY7kd!U}Og6cT~@}JMLNxUE_`+-o)yxgWF_X zbFO=ebL$_@XzH{9?fXLtVt=Ryiya#U=wJL>Br|8iQsS!}d`bNKj z?SOUsP7!q8aBmb;GAuw0GE~aV5w-7*9Yd%U1Cv%p zOm=wUz)5vq^INu4g0>-L(51bfz-+R+q!SU?q#*2>S!<=uq)I{!%QP71$iFQ(T zocRTuRQVw>?BwyG>ljNAmPF1kQlG@Qv~K+~hvQ3%>EB>V*0w1-ExwNATc$fR1H&9r z2Z3swX9~*>UtH$d-L9}?NYv9+Gz{SWV|x;aq~$|drX8!%bWs-&)V!9Eeua*^MAG4B ze%U5Q;(yrkR-M*{s!&h6=zGnhOiZ3W=nj85Q`q?|4yGlOvf)~QnY3r92QT$;Y#n{W z|NFHy@)lvPPak^+x+3wkPEtW0$)o32OP(lI3j7`~!0E9x16|@jA6W zR@Ekv&tm7(&nUnwo+yx21rAF1u8>Fy_)ufGdd^amiWaQ2Ao9o`BRTPLZ!>wK(b%1F ze`syfaaH|06TaSim+f>g&PlRS4Mw9rOyfohRW-}r*u5YM zvVkePooUI00HwkQh)*DV@P9M#P@tdD`IdSxIhL&cKym77$%0)|m(T~Lj0xc><7U5a zn{(k*%kgNL^k><#+uc$FEn>dSzkyO={WvkKAp^sWru;+Ek4cQKy3!{K9hHVTi5yP@ z6k|hn!Nt3cofR)M+j?~_T`yZeK%iDSXa#uE8>Q8!Y{Ip}jA7AJYzW5#O7<~@HFhzsjkUKOpb0tq3jKlQtB&a&qTAQ#8cLV?Kzr zCQBrWnabY_am&oeDV{%q!y#lxnOFsljRYD+LZ77q>B6tQI|hp^CjJ! z*RVihb}OZ?Bfi`HLw$)hu5HC1tO=?u@dTzeNfn4TgQXQ3kr`(;PVJ>qF{0}Roueg) zM};}kd*qX$2okgrml0E#jrz1)A@HNBF^|Co(9aS@-6YC!yTOMMRV5%iPS)qCFGA5| zDkk1m*sSMy66{kjXB(mjPzTVWu_pZ5_a=HC_h6-6OXE>wiW!F41?k%BWbgpi-f+ro z8hW#X536a&?VuKy6Ce~k-~DBGjiA4xTI|UgwMa&m=n^cK9Od?=bEqOA=tNYBh-nrC zDbWlQ5PquBY4IEZW0=cyP=dY~iWDa(3#UxTKYJBx@CwQl&&kLTh81Yupp`>T02``h z{y{0Utr_0HP$uIR=6VGqPlNfqseB0P+6hMmB@-!z#=Sc^Zstiwj2|4XfRFg^ z!p(CK8ma*g2zH@WRNC9_8{Ax7mu||!#8Vq>5LV{6;pF4bE~JQ_1RV#}1{B?EU=JKc z449O7-v&2~RJ!;)W|wZlpR#P=)Ljk)m$B74EzFT4G_@Y@_U2w+S^xx9D7^4WrcSIm zde|gZyu!NwBQd*d1+Vl8{xbz|ptH+JMw$d=(x@D5q#dXR1Sfu)GK@P_A~zC8>sieK zc?!odSDI0c^Q+?PBPbvk7O7IRYOWiHD+?DYvi!-HRysy%!j;hQub3q1!*P zc^u>2@LHqhoQ*pB2l%aD%Ig<`;nG^a%rJsOgx&}hjn|Ew;EsmH;#~j4Q%d|C>q&xa!o%%o&ds5;F zDWTX}1lcZt%ekyR5@4(i9)(Z>Xq9vQSoMsQK#O}4T(EifVUtqy+=Bi-U?f72-i#LI zt19$8_{epAYUa=+gZ*ymxsiS3Lw$PM!yUxfQQ27k;`dO}AfRDWIZ1rXh5xAfGxV(j ze6O!?>af@E-ns1rYD+Tx7ZlY0m?{+v-Du6pdVKvT)-8g(^My#Bt^gR03?_VIrQS5O z&n~R@j52@wLKt7B2`Z=Gy+jX~BdK-)3i~T9N(5*J8pni=q9;tOh12RG*Dg&l3o2Q=Vex1B6> zeJO*Ah5B57ZPiQp=)H9wSa#j9yiOpq+qd&tyl%HbgO}1)mM*(^1Y8Ujwu|7c@OQ&8 zBNLSbzKlWqUS9b(Hl$pK_HRWF7e*+=#i|Tmt@4{yeW&(jd+@B^?MUAhYm@0d*`?Hq z0Kq`3X4jnN&8jjV(XbjJXRQTl`G&|mC@e)j6%K(-nTHYR^=qwjLz_op*J2OoVTmjj zk`3<~V8m6#3R6KMAaTQh3^<_|D#>neKfUH{eOY_gFSLYJ9vRq-LuF;3l746@s=TG< z@i?8OgaPhKl|#zL!PWY@utl{Se`Sk>Vg;f?aJA3(Y;7w-6BY{Wk2c8l=w+p#iH7Tf z{?nW^acXtN!(j0uT7C}22=GOvw%Am3nKf30 zs;dPYQ>U2i&ouRBZT+{2+pQlUu4UwsPZ)oKJkWJna2hnpvyCJ3`By^cIdPB{Rr>Gt zLPFr&V9XH-JW7;#>oG*bRpE3AD=!_sqgsy3D(=@zDbh)c7%~0z233#!TLh9fqY5{>Oozp zH)Vc$nQb%x50GQMY2R|TjB6evlb7G^7Nn2;O-3*tA+GmuA4Q$jSWcYSVv;|y$oZ`j zqq1bMc`wta&#rQ_vuV^c8TAhs`;es^ZWZ!W-A{e=Ko!g$;4l0u(0*XV&ID|}@R_0n z{s1)u8gojI?M*{R;B&i`is0r_LV{`;foyUBn)^ix6%bBQ4rm@U{OPGj=@>6JLvNWd z4rRqYg1vRy$lE-9LU|XU;A*wc{to?Tq)I_v@7_)2J#CpqCtS(iYEL9fF$)Y|eIOXC z<9@h^O1u3#L)Xmh$%_ql6*eCPN3}y`2XiW*{J`jlRz0~b19E3( ze0-!>0(ozST1q(F#9&JF5C4Nadi(JE#U1K%O3&7s<H`( zbw4rp1L(^B82{fO8uNdFXzZN-6GZz};fbYgS=VY#33ULTL?E1Z?B)!EJ@v)`w zsx`euCGFB-sM$ zb<$3LD=O4>X=eT8MBH2YqcT>xD@RjvFXB+R**<}w`nZO}mJCsh6;txd;$+O8=c@kZ zt}Nd3(?DuNM1S#`8^@LMPE1bllkh|5u5Xem1j{lU#%!3T=6OaiJQL+gQ;#z$4rNts z>=^CSEO}hV;P=*)r*>|zy>mME$TQJDcP0$QnJC$!uUWG?b7VlZ#)7-Mb3CH)J(Ye5 zNNczd>@s{O_i;--LY!hkO0%j1{n+r9Z7SB!LzM0O{J&y5|7<+lO)?G z6xWQXlK=5usgePiT&@i0veiAqq&qUHWsj{ATi$Oo*yfn@nblb%K(1z@rQ%q-J#cKI zc>Q+f(Cc5nLtY)1UXr{Y7323;$R5ChGkVzH=kJIC+esRP|W- z9h3cri9kduYq1LVN>LEzKnLUgXrNR>{b5>+0QU%AGd+;qSALY(1JW~qhZMq73>k1N zqLY^RX2|5c#56d^Ioe_?AX3+P#x5Q^CmFs7Vbry<(+{j9y%eGCG}Vi*U~iNGeI=

n>x33pkhWo4T&dXmB6jUkp3A$m-A1(ywMa16#^azw;G8|J#Rid;ZNEWHiIYH|r3Gye&}0?&!<8V&;-72$LND-T_vio3e1_cbf@s% zuOH^jM}XLPjY`<155z7)0;a1zbv_QCj+yav&382_a8&qc0Q7Yarx2#Q?k(yzaUb5KIpyr%d}SFkxIaoq}lC zcWE(;)ozvqRHv4-kSP-Z+c$Tb_0z9=n4gs|=yGNPV_j(#Id?h+|s`d&uVMu#Tbyo((*G<_M{PqxZV zNkJ!K{RoSY^t|Wx&nu#F!j@=|2bydNX8AME9l zDMnlgM5ix;cDo%4;Z2Hi0K7$c7|$zTg=*R)`lQf1c(P{@zy^ni9{f^lEf zhV_61#3((zT-Orn*Rh$pab?*`M%+!^GYd-FuP{2enU%-Q39#<*B5Wg55CwzS92l_L zk`cST_pr?^^UQt2kYVl*v2ECp4Qh6S`1M;Ub-w;5v}K6%-BfG;;XDssrolohsZ zqhRTDL+OYyVb@zD*hA9+c9n-V_Bl)aZ00l1@AQcx2L}?Rlq&0rzbG+n=qlSMqzGh+ z)QU1MB8OFRk)FoWk4b?psY>s&{oI~Q&Svkfcoqy(Zk;9g-IPqYM-CkBjPfqK9xbTG zb_d6lhB;p$tMPY5!Mmsi9(@QWtfp2NBeU-?OZz%*CApKDp@7K&;~AAmaa}lnx?APS z-n0?uZ=PF7zUM_o1?nwFDk_dgH-^Xeg{}lN3h1#_Ut547CgI*&mIvWIhWIN`q)vTC zH?4aCV^zvwEg_*O{Sz`~MZluco8DZaNSHNEREM5p5~MsOtO`#XCwVbpQB-0k!O~_J zCp4fD4ZCmV>U~urXLb5x)jCWu=aBg^RXGpQ} z;SysJY0CM-L9cwPLwN#0lcb2fX$*q1-qBbXcz2xs{E3(kD}Hm*pKFEJHi!Sp#8NUf ztpirkVG#88r2rSr;#Ab#l}Xe|1-=guzabx_Yb!CD$QIIIhb+9VdqUg(k$CgrE|`rT zq+CVy(Kt-%BG$o)#T>q+KO4F2k>Ri4CFSII{R z&Ya}eB2_%qC!)rT;npKXY?EpJVpc^0Z^wk_m{16DSavZ*jCN=>Vh>c9O4|CBQ>Z1? zh~hF2Oe%+zI)h{(_6ZU{?sAkkupJ3OvRDut3Jw9=yP$??@E>cRqRjvMZW8L{?xH&` zh$x_VwGYb8aZ$8!hXo zt+W=BA>Q{_-kS~+zKE0iRqzcM3!^~cyZh7dG0&Jk;fSO&_5Z0b0|1BOcPD} zrA9uh|L@pvO6nAryf`c!4e5ooci8pnb`@4B+M%=;R`8a6>G%ZIv5g@+N{^1ymR8&QCXm)BHWuWRoB!{8%0fyAs`P18yGkpYBXYaGjKSseuB3}Y9Qa6_4=3G_Tt1;%dvS= zIF~_{HHUU;mFSO3ul)zR>V4RQ{mCs1;ybbzW*ig}cIuPdV=ltA5Z)#oWSd-L%OCOu zaMpmtxKv?hL8C%>48D<#+^PfF`bs}z*=rWyUDsx|Xycq~G^xUN)-{o*T4Htz0<$V6u%OwJIwSyofc1KD%#tUSya zwQkW)j4hUO#oI2S_56wYJ=E7P;mBlfv*-rnS9mO<3ro2t6RT)YzyCl2oXfyYF#e5d z4ya=yn(W*}mC|M{{k~aIrWY{=Yj?}TSVtJ%8PAazAqk#ltuN>Xp9-n&1H@^w~&sbb(*BUFPqjV5i!mg81) zjeFB~Te9;3j~{*Izn7f$%{vyT3%MyOKkVL4^Na|TC>%i3y0NEj|GTC?0#9WVHg)?} z6b!}dCs#meMuTD;~^;>l0PH&aE2k;Bytbmzj5T>m%0X+LbDvKRlkoM_~aJNM0^ow0{rPyu|2)zLFmIu2C2xTu4{gkea4oZKn4#+=2g86GeFKnL?75p7@V{$>#JGqL(PYFEtMcmdFS|hG zy%8?F-VW^n4$XVkvh6{mCN0=c1non@Emb_%r1cmgsxT!PB7Rb8%YoUwQm@`xYUj(Q zg(?a9pU+o3I1J05>nyzR3*klFu8PM583G^`dlRa?XpiEbyO`fr)%5VHhG&QH={&?G zCh|K?!c=LpPKNbOf_!%%{B4H$BnOWS;7lcx7LM4JS6pV8%x)C4q79cmf}BvO%}9u= zH#gL?eT|1E1Vyn6qFaxXqsoqEx_lrsQo=HdbrXwyJsf+{xpwJ$pj&g11acR_qFn_8lH@ zOD<&ezDrWWMzis3_pyI@i8k5ukvx6zq=_nmh*MdCrYcB@;x8$zuk6cZJw9Wqd>=T6 zUn$UZ=5qI}B_XDi!SIGC@Kp4w^17VE+gTwAhTUvcn zB6YjboV^UbDr1yWjA9$g{#ar4sGqKKGc5P~+KFC-D9MV6R4OA<_UpYZ4ni^#wAIUu za|BvRiRb2^Z-62xdjk-@5%_D?h&x{0qUjUayUAPJ`9J38CfMf}`cV zfBkhRgvx>yz&Apre|cLqt|AxU%WCD~%+Y^B` z$7nrKUZp;0QTc;Mm-y)R_%J8T^nauN41vO#Ggqk+!o#_yo-k>=-*fo}{^zbK)i)RS zqEMR+~fJdSe@`X2A#hMAw7(5zTLKThg5DsQRHA0`Taw>Qz-M?Z?>7!f;l{ z>(0t8y{NNOwJ^H^AZtl)b4k4}uP}g|c>4DjJG9|8SG_^z7cZ9#+k8Ju6@h&5_kpu};B$H0V?6fKxR@6y2Wa)ft!1yy`uFIaDabxo^cxgJ4#w`Gk69YOk!gj;aJZ8I_<{Bg=*Je%9?=F4`FnVUP%<+U-9vcrdwzg`Qxww%@e4=Ml=R^jWy3uSwl{xWZWm9(CM`% zDBD4xtU^89SFfMcBhT`bfoDbb?%gr@^(E7-Aol#1P! zPP}Ekf@D~04atsLLp!@Dz92c|o?yiGB&|Gl_-+NxW6+-J~2u@!Lwx0h?4xhY~ zXOyMWpOiQDf`cu4vAxVSV#ulsm3&5HSh?9Kr|GA6oEbJ4IT`esIbbsrjQeuf5G_p+ z>>D08BG@-%w?vvk%{s*?`&u#+L~2_>`?nm=VS2y)Hds`FON$|7c;7iV#+vwn<;)R= z+~zd2oCc#a$f=cpc+mEM7`i&8DQIx_&pg+lEhFiY@YHT|1t78zS(1KE^0UzDIj4cd z#t_J0HJ3(TB;PUImMu|8zrBxn5Aa@jxL&=+2UP7Ef&7?oJp!Qs9Re+~I&?k2!_C>n zW0ZtC>g;>Jz{q|6Z5pva1G>Fu9B45ZWR4XP))gDD)7Ms_da%c#DZr*pcy*Pxg8)GM zt6jgYtSSud1%bZSkC2T+QUq@Tx>IfrL`)?p;(${4qa&&lP@~2<)S4fJ=OE}yOs8Oi zdcv)TwUZf$fnhsZhsQ2_O*EuAHOCJwuAPkGdJ3J+&=9jN$X)Vaz!X$;RuY) zvW%QB7Cl8q94{tYX;M!<@&3;?be<7!3Bu4260x8y*4|6Zm`3ZORPRUIQhhwn=0zy!i zCLBH4JwP<*{b#^nl@IWbv)Y@Y9 zzK3X16@Uh4R$kf^!gaXq91sL9Lo#vU{v^X)7%t>-R~{;c20)!&aZ79jg{Uw9*e#2O z`I8aG*}z$JAh$rayKgo<(}B$62@%Y9bH*HuXo*fCmnOY2Rs-^?`P@C0%^}b+h$J z0i4iSh}J9|1p0u2M`1k&&S*k}Kk{61s28gCA{(Y{6|?ZgWxKwcEW(L8MEpUkMwG7_ zMpY~j0LRWCY2-=!Pg|3Oa)27i6SRNX{*?EMQH10pkt9Ov-R`?90 zsbSW{U59p|oM}z3Zu^B4LKx_e-~3NE<%~g^gbmp3DthiwUS)}_iZ`BJpqqAtxj}Cf zw-=mDp*L1B-z2PJ;HTpo-__TaBU{d+n+4{?mFQx{#6m)dLNk=(6x9cEi{RiTB{8|RJw^H%k0*>RrciMeuvYt9?E z1S4JOV44SJ|g&0F4kxaypfvVkVQs9Dib3Ue94 z<)HDdpo=_YMqehT15C(XLYBo2BA}Yo7Cv-wZsiTK=&rb6+q{BKuWr@qiOEU9aSgMq zzA^-UaVn=xTk`JcJ<@oD27qBPf}g!jTREf6KM2p5o_i(4JYM1HbzM@yMN3^-INUND zk5z>a*e#EGcBOfi2M8((hM}9dzi|Y3_K?Ugr#U_K8F1E&=c|vwq;92FH+lQ#&9(lX zO~C6G4+5ForfT~Rwf#I{q3W-ntfJ6oAE^AQWQnWFnr$OHQ;LXZWO zw$QWJYEYF0bh|z7@B+Y_(?6wHv``(V)=gBcfoT;6kVe)#0sDL!@k2viKRTjl0z#-o zqgAJMBgIJ7PCC1+!`CybH8Okk`Y-L`E#q-43M6U8wT-B<($x>9^z}^! zGOO+MiB-U|?5;?6j!ia_B^Bo?-K6F02LHq6vnqIXkUK9>HdSWDv6P=FPx5K@#5*`GAbPh7QdH*jgEG2HTS zg*8Qtktqnc&ZxA-yszob#t8?l?~$3jtltPDgIh^0_{oMwoy~a`U5zQ(B)TjvtZrXd z-TrkjxyPx1`DB2L%S)nLn^u|l6b*Jx>G$lxl4mk-WR1jOO8gS1Smlt{t&7mEuG+s0 z2tgAH;mY*wsBV)Jh4ZufTL?A`fW;D3Pl(F!+hK?_g^-f+-sKR-x#+}-G1qz)2i;^c zfALktkI6GJv(CEEgi_Xn(mCy-PmgswVy>e4k5I$8HT5kwyBZP{Auf1tAye>4HpFC* zoQI1bBt6mFoBUkK!-#?nRe%*lD9UQ-vX(iL$iP=-{Lbg7OdfVeTF)K*F?2D6IRFG%bS1Sns zn7_l90vC;GEXh+QhTTBY)WvxBQE~kM8d%mZvej>n0f9N_aRH`?Or0LjJs{F|?^83# zeUVbSk;jky8=3)007-j*FC1bAq62oBZpah-%G4B0`m5atP-kbhJc&wTO?UUeHU_e& zR8o!>1&p=8pwhP8a z-J>5qf#9-pT^yAC31Q09RvGP#qOrs;g zn6WI{u^VVCJvq*yeF|_PD+TFI0IIFk8|{Iegt+cZQ;fkfWm}MHH6^$^U-#J)IXQZR zfBHmjI;?V5E0)t0_#;lp0W^~*0he{%(s>dwpvi}x6gn#3KZFAkMu?y5l47zXnYf0J z?w$gnhedeSPEWatCo0@9QmRggv_LdXff+|1WJ6gc-d6trl2k7u9U8!un=?yo!$jAj#R(O-rU;GthGGMR1}_vEXnxt3kVzqkO1h@&GGqdIne@^W9?yq zpcG67MHVOq+#95npm>ES$XZbNyTX#qA>VLux9>yorYXv900WmQ0*_*9FCLpxYs6HD zQ=#M3d*kM?A#z|Nr!l%kdOcY+^HEQp(x!;kaTzp2>bELqEP-uG3I8bOQ_z4uC$ZJr zFrdxQWGU0P3nRT&0B{*EuzMi($)qEHo3A zjp`7{T}|ZgH`9T1H@0+;E#v%#=r-i@BT*g>db!nDJhe}=?MZK8#|spoOZ93`uA=vb z_atvF0TKlgSJF1Fy%^h_ZgaV$5w;wE)Qq#6W=u={d|q~GA@0EHXOyG9*+|npG7iFe zF(V}xlI(S4MZ=FaMTNV)@NJczD2Y}u#1IV^hO86Q+qOyW;r#SU0(%+8ue?-;LwLar zw;tK!@QCjbybj|r@R)tWpJksozrpso^iv^dHI@if-ckIGwG`dd2NteQ6-uyon@zKc zC$Zy&0U;Yi?{3`~2hikB+CT<=^#x69fjkpjJ85Tv(OIRPx@iM{!?IQ-uWHWc?wbdz zf@(a1jsiI~I`Q&Ouh@WqEDCUyROMKmY!P8d(v~L72s(K=Swsxm=+bIJ3=4z43*O4f zw|-kG@z=>;cQQj&GNFv_G82KHEW;WqEi4`9yTN(=qhI}*2{!^%Mc*mNCMEbE`bZ5N z@w@obnFWJAex5Eaun^5%lgC%*7G$EZj@@+dcLWHq(~(#A_vKXpaMVnh_LW;7k*vF# zk_A)OShHUs0NU^JumNyVaL%8eH~s)W8bcuejajk&UuMP1$nigpw>CN4emkbBUW!#9 zDaML}=zv_&IGTT3j^gS!l=fxT8}lnqsaiCKtQyAc-d4z}I5m$=xT8Da156}MJxjgi z=yiw$5Ij6?x=*1Ir<@)&>mk!UC=Ek`Hl6fYYQ)TJ+ z^6ax84xU7gr)0L$$Tu}7RW1Ghn(5cj+2gsn$Jn3kIi@>4=iBY2PYIqQxb+w>lgMVe zrOg9!McEW-r#xy$-mitFXzqcwLCGp1bH~tuwLz<+qeBgHODbn++LL-vGzPElq}RB! z`em(7YW4-=_jq(!b=$0Gv~M;`FgL`qtd`0O{X{8xlXs6-;ga#$>4%Kc0Dp-a}@M z+cm^V4#OgiER#1AR}<>#_WTF?4Y0 zd?dQ8_EzsF6Y#|1EqOI?*!Ul$sVSACLY;$0!zw5Gp^eJTRNG6t!K!;VRR?g7v0zc( zumh7qO?ks?7FVTBBW4=Z=R@Bey-S09Iyl%`vapqD)xwEWR|88Q{7YoKV%OD>;j4^&@ZDi|Tvkk$JS<+04MA~Yn z>b^414AvYNR*;TZYd}W-cZ3V6;56YoYJnswyt5TMkXj*XZHu zem%$+uGv(nK9#KAvy>Fo+N>Y9)2ZvSyybS3k|}ZxoAeXXZIy>h7e66T8Y)j`E`@c1 z+hku1beEze=ZoAkNRNa5r#(gIN;`k3UF0 z!TkND4TyAq?aLaDV0LOq&xhNV^;d9VZI28s*Ga=i^c@YYfZ5|H08vf*UkxW4#ehQ3A$=$5s!D@C&~h|jGQYs^pQL*g*}z_xI+x=h_iL@_1#;^F{D;JLfR8Nkr7 zIC(Spg0r7Rw7(>{W#Lh63nw44|N7FDNLBOfLa*taLy9qM_m^k(#UO92IN@;}d=ghM~ynPEmnaLYUmzMQX;2S~AOv*pdQ5^bkhYwu@Ne6) z`yK!Ke1CGDv868p5FlQ3u>k=Xx!yBt--(pURUuk@{lzG7xiCv4JiC*-zyc!GqjQGhu_q ztN7eav^z`$pnMJP`;C+`CcQjz`&1d40SP{}t=rWD@{?kg?Bmb$W2bZFrO+xAn<0Ck zHnDZIzeTBY=W$#Wsv}?Kt6#8?V%xY?!G4Q&ka0P+qIB74PP?<-LE1F|Atvihl)P@9 zTnDXT532v$usPo-`1utAc7=V*dOI_iQ`;ivY=8s6E78_1vuu=!_+}$9fqn~BAgR6h zoyV7QQm1#esSqYuwy_tQNNiBN4}5t>T)O{yx7j(xqXL zw+WZ=_40J{bEp7F1K9E=D2HG#enjDCn1Jv1N#_In{u!lM5-t>xP?-bll9Ivul&|Xh z`5gU`Tahtp29N3_P$Db|X$~CA9*z_AlDKu3;yE6QUHvCsNsCx`?xY)16EC?i4;?T}!x`7E@&L5C$Lh$@CFn9GW&`DVEnI2$(Fx!mnFS?QF91 zKd+AJfA`hn6WrpP&VRTtKyEp!s(0M8{f%2}n1!=XJnAxzq>B59pv6~=#4-S| z1H8aYhHsZ*CON=lA|dErfIjWc<%{SAVl`479ccXkzHyufPY&$H+6Jvx4Pk8Xe*29X zM|^p?oA^L*C84Nn*)v? zqE7+;{u$up?EHxEquqE+6T!xwlb(d;e9w6AO>K~m&Sr1T~ zPvPd(6jf~;y67C)m^0)BN)0*FLRv%GJYY`NT^*_G8-jX<5tYg+42EP4L@Toh3LFWL z$KTu4w_~)RdvCQHPP`|rBs~)#AY%2GANS58A-2dLhemox001K+!`jAZaGQ2@DusNqYlgGECvhWS>@)Ogklcg z5QxV|FaXsvXodA(ztA|A!l#7IZz)jBuM7YqmE5v1o&6)7-pBIwG=*P2ci&J&b< zv#{!GdZeeIu@lXap#v$py3`-OCx^*cXw^9B2xEs31F6=0iV5QCVoJ9ma!VH zNQ)Jxd!;U|?!l~vPQ?sCo@nP;#pJN{TD zlde43I@9{>%wNv=^I~ToJo|x=g&v}HRabtPDfB{=7a>T+l6JP-Ak@MK6JkJ=`r65=$ZG@|VRJP^g9NJFcYp%_A=}#VJ^D zO2m1}G+0R!QS*Ue_-qmADG>qQzRAD_N^#TGPp5W5 z$HTcXr|KGfhzWOAg^f%kh(xcg)Yp%5Qk`xcaD4IZfxM?zmyy_K2oz%`2f(;56Mexe zun7=Uc%^gIQH{cK;-F%;l`cAYd3hmz1JOb7Y&r8YoD}s+;=q^|mRg!H$^;}E11!6!!Oliit5h#D%4-Q;@b#gEw7xBJcW?P zmgEXMLUt#?DzV`T#p{u=DSYN@vQ(W+o5tZNV0anffde#}#ckuM>pn1f0 zr`33n*A%btLcCf(Y&PJO}Pk3S*Yel5Kv2yP9@@Pc}%w7J=9*%#O%@2f7` zf?Z=Y>_wS0;@JiSr39Ekg_iI_BaYbvetg~VgbJ#c!lMy5pb&zoF@C`ag`LH`B}fTa z1RnHN`iM~Fux9>cE>UsBc)I{hyMpiABM$W6!z<)h@?Ck!xKg? z32l7C=Jk5wXCURL*X=b1!EYc8z5S+$qP+6138V!Dw9Ac^S*(ay23ovw0kz$KTYf)L zSLjmX!qdPpZyxXR34})1gCm%zZaT39$E`yK6=J(yCo-Aa!Eo@YnQ57JSjr-KslCT% zW396iB(+^{&;6V){unuvyBn&?A@)(BHP52T`&W&&Qa zY3lb+U24gq)gC=p#9Q1y&|fc?YPVlBq@n*#1r{y0)^QUkW;g>cP$<5APe`;QQ409U zJnhMl>Fn_D-hXyZ19dP-Bk-0M?+(H^zA&P7KA z%r6ZWSRLd$9>!eQ6O^|FuAW#!0Ggm9fHMUCMJu5T1qLH~yOOSAY9>;d&z-kwp;N-4 zZ%d`4Rqauew4SIa^X{!i_GJJ7^|;MPtf{ANk3bD;4+nbgeODL)wFl?KE(>{6kmAd`$Zr8HYS zR*a{d=9~cv*!vD4VHp6A%bTL4pbo+T#cPn5E!7YQ?dL^w&Es&mANT-yQRgl~CaSSj zI{Aa(yTC8n>&GY=jD2Ebz>3@;gTJ^>Tf!b~AuHiNfrSbex4bLEHfi+!GE8^FG045z+Kqj*b)s&#iU%fQcN}MlSU`H_0KmV)(}Qxy~tqeXVI@r@Tjr zV-XVJA0IKyo@nK^o{MES0GoK8p6Z21d&o)@N43W37=}O~d{|u}d%^~+;3J0)eNx|= zcp}Ljknd(YNW*E(M^NdV3TPLNS=RH!6)c})5iu&wyCLCC zAm!4-EUzo^p9b^$2glUz9N&!L$%uq_CSc!m$F{*Yyj=PNRXBt5)Z<$mGncKE6j*U~ zrJ}6W@Qc;gteXYcb<`>acwHJUlsirp^y644pHnD->>G&}PpuY!2=bW;E4}3vpL#a% z;N&Z!d3}JgJJVR+1_myfmxB+{G*j-Ir^+uR8B0v5Mb9OgbKKPaN%A55JGSwfzjerY zpj}|C+eE&lQH!&L>@R8pWtFbV#%KyYtDc(6Kq2WSSBJ*iB=ymavP`P zL2yVbM{exq;ss*JR(kDOC2d^r_v;R|91035S%e51R^1c_0G8kR^VPjqCRW{gV*(7I zKl}tt;lc_j07yQ(J4&hplx)FXv8S3ATsxugMlSb^xZ6hfQKVGUOlwb8Qd zD6~y^f=capigR0DeBHInU;=+zCKU;TcZ7U0vlliuu2bzxJOm<4mD!^=`ml-%dIb*d zCP<-YUO$wV92|$isLZF@HTN7}3BMsIhX$Q%d-f`qmv$?Qk|I^t)E7yJMZkf7Oyd#| zGI+91*Ufw+w)8Pt-REhLIVG-%mD1sR{}Q{ha4LwKza$D{6+EzS{Zo4n{80pz-(#{4 z(@x$>N=gDXIK0lhmQ~ZS5at8ytcWt+G>rj;_c?tprX^G%%>3W+4K`zz& z&UZG>M$Xuc`5)vLNmb%kG={3b30O()3m$03#2o$z`1!Egm$gl!{|xAi*k;D`wy6{+ zZ_SgHtm=I8uF`ejYh~M?>u#)qV{qIWljE6D z8S!c4+qmU*PX-AFV~8x{B`zMoieyl38Q+|nxj`YTDzym{aMHhYSlBd&!vmYm<=DO#At}Pt0bVciCwFZ4}r7iBlEV;qQe9Bo=O*gVYy$`~_t%v#;r$$9W^BOv*iD z3ClIhSdaw8s2LiMCobWi0!W6XTLd0;w~0uR{aO)Z8x+PY-kWtc0DSsHE}*t}!SzIR z()BrLU(`~lgSJV%8!Ab+D?6>=wG@lXcp|@4TKuh{&af}rOiu<_E8yG!{`n?jB~%i! zf^cpeUs#hm`Y+ML5UVX)2HrE-}7ui8cTtlCu~ez~)6ytR*Q zliVikAEEJ#|tbBSkUE!nmVXORo*&Q9O9@`v_0rX(LW-EX@ODl6+2 z3A7X7El1PwlPu0AV4}N=-^CF54~~bqS>y-sa5y*Ay(CR+UWI@^Wqox^6|Se+bX}qF z+@wfCdBs-(^3_Q_t0xQQ?o#GvILpd|S4Y1K6RukuFy3!767;hj}((MI|VYbWc`Gt8*3L$?r{_6uwuXVtIQtX~e9FTxj7TGq|bJOHOh zLqnn?W@sZI&R>QuWW4Vu*DxE?#Vz1IWTtp|{SBCCt3?bz%z!{dpTCZJc9;Xs4*;Z@ zh*&4dDM5QO=<87|`*Hfby%`awwtAyB@BhcxIW`FvY+1By+qP|X*|u%lwry9JZQHhO zv&))#pXWu){DT`AnGtzc?tRuuZc}1W51DD~+)D00mHUJ?5^$q^oZee+Xovw$0YG^g zT2523F?dE%T(?B~FhCa{eD*@33v3w@st3jfoU`(>OaJx^8x&wpSL?;EtF_RAx$C|7 zO1w5kSl>O&Omw&f!X-)$f&@5+*uN`;uGvCSFsYoQ@$g0q?{gRc$403PS6li&PZhQm zU%xP!VIH?UnEnxwF=?Iv=f#Ho%G(BDw*H&^@TG)#{sVF_*R zR7B-Qy@E1VI(bv2nbu7Q+Az{kXPX85iy(?lli08nd^72M2@al125o2<3yq{8aYKaG zHnZWisp_NJS9Y2ZAcfB#<2L*mu-pkc@|KInqI?7EemC5Cx`!D%zwK5I^GYxjFs``8 z{dxpMJ^rqxxU&28(MKd9k_VVt-FF==%d$S}KoAqLwizp`wOx^onXTz}adef3yh*5Z z(It=x#1L^7#n%Q(>>_`CDA!4aDnaO)bj30~l&xAA%{t(*=NEc1?b|7v66*zm9K`SZqxk`6nvp8) zwAgEjymEtC+V)|&tu&ONg-gN7GD@miWsxG|h6T$sUY>aifi4;>T0@eaiD<4_nfpz> z&Ida|u$h|=zQz=7^SKNo-3a0QY{Mm=Jm#?{N}XUJK#HuNz#YbFwfGz73qSw|J*zU! zwDI*S*Wa%LJTyZatm?iUId)si|t$DJM}e#IW7?uP08z#Qzkn zHEGlF4#Y|{QlMF283FMRQ31n94iy8T^wf{2V9E)`*5%Gb z5ZXMeLYARetz1QiHY((D(4zi<<(?BGft4lWq<1Xp`ik8LE!E{o$LZLL?tQ{unX)JU z3s)DtDu~Vvtr1K_lb08*s$vo(le z%jwA7$gpCQlp+cVi=38dX2j4235?tUBXaA#2bi8k=9}&94lKn;)pPT3l;cCe<$VI* z?}m4PEWzV4hYF)43iqI<=pgdFrHXrT1WBMwNwRaoHiLWdCaO-OZ3I-SqV}uDn{QK{ zYuiavb}Cu{%q)$HvL-0Lql{HkBK9~-kY@ZdbJBmBd0O&F+rW06130nN><20`_Vr|r zdmbte*sR6-OHoU+mtGa;cLwVBt33!Rs*9K$p$tXx;)EJjJB!10Wj_>3-D3c`Q9gRh z7$${6Bf1L^I4eWPc>T_DGY2dmPrqHPG`$Y6Ul9{jL)J1pX9q}?kQ^h2k?x*Xn~Z}} zb)^rK6&ecTC8O}XJqL~bH$AEp(RopZ!np%mdcZKmpG@>>5H zy#Skxi_>mw#t!-V@XjWz`;$qlABn!@HS^rQ8j$H;Z~=c19~hRVTxm`IeH zjeb-Etr1uraQ)e|GU$qS63Yt8&YvihYaZ_Bx^_N+7rR6AG^L}78GSe?hvK80yLT}} z1lk0a4kh~_T-0-AI_@+}AyQe)tupd1VSdF*8Ci=y(P`jS4$iIkMwVe%GvzgihZL3h zr?ebDj2b9!r7gDQwA{ssXV}afzf70r`k!j869RURls*@&xRpbDmGrIadO16DEh zt3^)(Srtbt#Eu6)sfl^n^(3|78A z@pd|DU##77A^EEOex45&1HDtTBRrSZ;VQ`lfxhkHHA-vIlZ+g zr+IP8B*J6tOA4?~fwI15SqGWLBELM`w031mQ8*CX%9ecl<2}7sQ_3I_(;dH^d6Vdj z&csNTXm2An(uFaz!%ekO#36$T3GXl(1AqwDM*a$>)YKWiBFe%Z4~*-M!Yi+h@q%xj z_T|nvdFsCf%WBOLGA2BSx4GQpIco+>`YpI1z#>E?y|sQ}NyjL5O@JurGxHMR-j6-4 zG+Y)B`;lfKP-b-4-~*G{HQFL@cPvo3Ryyjj#vQFFx3f${&)Jm`L)_D&MDAB$0Ez{H zBFHx#PB@SaYyMYK8aI^1xDov9tw6OICJx-5GE6yjn_ih+#J0Th0Zy@N}}mR@IrfLcBK62O<^!gU3P-m0K4W630QMhCQCKoc3Tbi{?v) zw5C#~`7|}auOaJ^n{Ylfmi(!3;$J+afs`!S zo{l0KMv@l~JvwTjyP({H%s^D+cS~3DX{+%078(HM%Q|?GS-a9crMP0|?!!Zf6m*gINdin}1y$$_x zj8vMPj`U@S!1{3#*4xp1bATGz)awb04wkaa`5_J-6yP0v zG>~n-ip6QvIGaY>+w{kpa(B&J*e?4{$%V|V)_p|cuy-|+W|~#@au^{t`mE)%C#j(y zA{`z1f`*hy*%A?t8r?wNC4611#XI{6ZFni8-`{{Tj#A-_L~iPT*>-;+Kjo|c4La7T zDwMb;^8qxU;J>@6%y{QwB3zlP=uA_N!C2)INb`%N|85==sdQR}3uhM&dAw#I--!Dr zxz8l^I0b>K(e5pIJaOJCcP;trbTdG0-DOG>MTkV!bUTgKG(`vlmq!Q}(PH$T)mD1$ z9+r(*aRh!=eBW6QaRtvA;DRjaEcdglJkETqG}*KI4NuQnMH`eCC!d)tn{Hhsukm>E zBkHOsWm2byA1Bn`>=5|Aa zA2HU7KPemznAj$D|B6MEVxX(bf^svSrX5d^45T)XW3(q_e)-z~{#K7tt?h=`VFEUx@NSXYUhftSYpO z`BS}J80BsYp&2*+t1A^ylWzoC?@5we2baJHXy#J_G(9i7`Zo^|6K65zpk1P{reu%F zZ5{h%4-dBYh683qPZvR+>EAD_SntkrW=*DpiAdl;&M^~Y)_Nyo_dB*sxX?mW~AvdmR_k~ zq5Dn@J`@dS$8d*b-_C_^pFOd0fxZOBZvp+dFU!Iy88tPfV|dOwW$C^CdwF^8f11{MKI?oY;KM<*RNt-!q^hIx`)$X|zhvNrcmm^IBB62Al;5 z)Nf-ErjF5!l&Kp-IYz0|dL=f6@(7fkiu8@&4wx;xRAmz^?{Y{Duv8xiIYEvO%Tti; zE=$h-bbU%j?mi6{JQH0LE^|6@!B9c{WmDDu@{X%2w7wv>p+oG!s}ks^%hZ!XbR6B) z{}%wsVjO2BCDt!mQdz7gg1 zZH*)mj-!3MnGA#wu~6+e0c=d=o3cG{O##o4NFP=iz}9k3Xy#ypTX_u;;h8JNH48WLWuE(^&s8tSuiw?&4( z9_dhAAY(A4D=WdQcH(Bxm$rCe5Qjhe6I;c;3jV*LFZTb9zJ8N<4F5O!YSx-`-4sv! zP2%a$Hf72_Q~ntP*WDR!86ll)*iJ05@|^HF5g?HsAc8~#*dYCx^xpgIt1@L|Cqjas zNUkiccy~8I7;_uxb|7YYdUU`zJB3QU&mj|(jQNBkqOnhO{(kyhJV!65mxDiN%Uk%bvGV8zfy^PulkJJvMFvWp ztuhf@Aq8Z0@~cD&auSF}CZ-Gb+sAYz_%x|tA@Lz}CAK9U6p@vp4^Oio#s$p*z1cRp zwEn+u@~7W1XU(cQsGH--+>$EG$u=sxhpo^?RZ(2qE60(OparriaJ~4x>2n--tG{2U z7fXZ1@g0ro?GcO!-aw&%GPHjGP61U!BX09!>f-!$#${DSMTXfKbZI?EFwLZSnztm! z1vl6$SAvKI>r(iA>B_Y1zL22e!Z&ZpnGq_OTCgacVinE9rVda__rRrhg6$#rIK%#U zBUVV~=CboXkbW}!v&H&GwF}}4--H_%SAs>h9h)OtB1?WotYV|E>dTlT8#HImO5d}Y zoE731P_l{~b&(OVY`dI_8j3F_m}*@9u0FN#uo$d{#Y|l$D=ngOA3+tuwNRmXEpL@n z1dp|$JW5yB^o7+J%Y>=eqp6JXYp0!~8(qqkUL`@KA8DcELvg9!Z-&+}T&_Cnu~e}& z{Y5`AqjK2gP*y>Foy}{Tl_+faN_O!9Ypt@Pq=f>T6i(ZgIe9rb3UW=9j5Hh)4uLM&h2mu&$R% z>Sh?8q0@_*q_20VUwj`}TJDjR$v`-;9Kk^H_W%HunGlgBj|vap9jNB#0g>4WUFGn} z3Hknt(s=8p$jyhW7`Qq*J+Gej2VZeV!gjy`M$Bzc#E22Ug8AVNrMMWx8z0Jk#0+Y! z8J3m5E@51W8$Qci7PkFp8#J!P60zFlzK5`RDx+J0s$QC33}={)bnj!J4@p>$lrT>h z5(vsE=@0*v^&}a(L_E-A=v$EQMYM_Esj1DN^lJ^y3#Lb;gD=HAJb3P}7t(Upz9F}FE>SSbP9O>wF zsEz7EBanbi z*n)WQ7Ge1a$ot4OgeH;mIW7phe<>7Qy)3ka+h4`P#@PIjyo!zOmpeUUs_!beh{Q-? zQwuVP>Zmpr?8x+zdZ4k}Y@n?Vf`o}%duQ5R#xMX7laF+8^{VqFa+qgh95(Y&UL{pi2?bBm{nUEQJyi}o$g6*z#hNr3f!bPu8An3<))KT>D~eXEz(>vOm{fM?+})-k43GUL4X+$F&Gp%mxF zk?kS~%h|$4<7*6hz=E;%5pXOG%j2BwZN+JJJu?gFs`Q!??9Fu2FWFPB8Au5U#{wuP zGm9A~;*GyD;J18%m0t36H>r!`9H9xmH#}dQbFyD<)01Cgo!U^zM&@nUxBic8-=;HE zc8ex5^HTlEJEQh3AtkcoTf0Rs5{t-kr<)WW%=D)yyA=3I*uWJA3PM~kiTE;`-1Sg+ ziyOr~n|Dr3BT;s|;7)*+cLe5VE_CpNp@Hscmqp(0doj3o@DdBww{G4cnpHX*d+SCy ziQIX`i#G|*b~buz=Q!UIKL$AjZ2Saeej%NX33-;1b(VK z{gAi2s?O}vTUQctV})MmU*YbcuyaiK9C+SApHIGiDy=*%+^q6^(&U>|FXi7%-E2?B@7|Pv5%Jc-du}3p zr8*b%%a|3;6O*N#lp*9zIIkJOa^O4#?rwSND+*M&EQn83j zx7Jc?tu%yP24j<5Th4RUn+Gg*VyxzqW{k7YZ3XwKTb+!f{svFnkh`}YMZ(tt^;&D# zD5YpcU?XuNraRQ9Rj94dnP$1=;9IOB!D*}t?c59H5Ut4g18chC1xAThP8~t9Cm99S z-M1RtFAmOYI4P=n1_( zFIa3x9;*BUZVxDid~Qyk?a=3${8xvE8Or=Be0}#QeCX5#PRlWsm-^LKMnP2A0<`e? z-JS@@+uDXlQdl^1YOxGfjFE8KN6x$UhJ6{!ieY$2Vq}glp!|`rXAg~mf(-8)F21I2 zj`L@wi6;t6Hz@opKRhy!B32BEH^LBX{%zzEi-MhDBgp?_qB9f4>nS|0j8ZSD1gSg$ zPJ%Sxsj!Vkjau=D4#opoWD9Sp#Xes!@tMGn90bV_k*a`v2)QvgNxUG2!ZKDO>AmnQ z1cEZTpW%{QrU0hN{Y{E`KvpPF@RY6{2#X-Bd7V+jVHkwlRNP&EFjr5K5)S5ebyF8c zRi<4Fo^?#V^E8WeF|#mWIVjP5xiLqbZe$k>?2LY`gkD&aj&m)qUJOueIN!&Vz7525?6 zs@No!?WnHZO}W2aYH5v&%N6FDX#SzwO<$(U7GqsT?4o$BGh;?+UzcOd5k}L0&JL|f zPe#uUX~@Y8g$A@R3eI;dqD1UCmGRL}+DqXEl71@qpwQs6~t&0r9;|7uU%LAlIh_b9O2qn;y! zU^%Ff>`Rw?4F)lScg7k)j-=Xj`gl4%ICNJ8jRw_M=w|noPelf@zZqxQaoQyUA&YEh zdo!Jn-$rz17ZIl$h1qjKZXX_q0HYav5Zv8ixR)bT3QQ?9v=rE8<}-}aGT z%}N(<2)Y1B^B+|maLWJ}b?;1@>W?Qb!ssU?SI0Ji%U?v{HzA5Wy)c)7NzPfGYAk04 zs1_ksH?%%AK?LR`{Szmhf;r0|7L31suDbz9bjf3CqdMu-@X)H6NEIS~(0nOIHGB?x z>nMx|T?qF$791nehWjO9a&}6>RwPR#?j*z*KI(6-f`wX=OHcUf^|(vK1~?qNsl*XS zWD#g4g|a%3y&(~*FBrD1IL<=|Ftf!>eMo7xfP)1fo$ zC-yPa22w9*67n=rY^6kqzlR1-PU@ykXX*YTzR&+X`#T^0_@fP-^TkmR&m%*Cvk|Hq zKpjFkoS}12gGny_#%k`$GjR&)Q7pJHapP)r_kMOV%U|9*pMk*xJ*4fks+zaHYWxRR zq+PuwC!2_)o3Ox+@&z{OdsDapZ2YyWs^{<^5%;_Ga&b@?NpZzwcHMkDe@;^a4w*@V z`T;Mvx3bXe9IQX!P6F(y>tFLcF7-O)Vt&+kqT>&2K3(*~LBp#hPC-I7PEG~9wUZTV^vEI`FJoT4tItua(g|=g`{Woy2E`ZY9Yye939#_2`gBAkj#VFP-O)_ zEuJ~uX`R%Q4C#f_cz>FdMkcftW10}Ph2r(+Kz>7b`5Bnb3*>U!+JVRcV0nPe(a|T2w~eL=BUfUsdwuGDQUPb{G0pxjpw^8+>A|$ ztlp2rJ;s8&f+`md^6qJd-t zYujZXR7hZ9$X}vIV^TO8V&U@mRlJHC%hk#td|HudJQ<)V{(DSa_b86Uao#>NdH@>0 zY=xBnuz%()7I%ZWZ^7;}y@3aX(Be1Yrwd}-1p}DB8^%cFF zaiNlgsXgpbd~Imrwku8la&1D(bIZXKy_Mm>M@F4>{ErE20`~ixUv8`4fu?|to#UUfn$O|6YdvlW?0ZDqpV%R&cU0rTB*SSw>FP?>(si=4H zbtEn{>MOg6yI@aI9s}vBC6oYmO;wj{bUh&qsQBLP6FPo`W@*{-Bj>KiOSrliwet=MEnK+FQ|||ZSa~6|qV-2fT-#$% zhCf-OUlK&%1xw?mx_ClNM3?|-s2qC6qEq0rE0Bm>v!;QLxA(L~@g^_>|ZGl+xBWr;7;5P%ieL)F`jRY$blSA4@ zj}a~JtkswGkMRBY+wBK7@$@=y1Nz29y?+Plzp<;=c1yp0bh(u#sD1xB+%Y6t73zDY z89d$HAD?EDgg>LDAdU*Frm$bRS9=B*fQg&a0DIr2sm%gy45A+ z1KOcJMGo>dVN8gsTo)~?5Oh6UHTn#N)kk26Jt^8HK*>dxT@N!Dg5VR7-b1nn)dX2H zbSyzp#pW2l^N!51P{qfJQ77Vi!wH@E5#1=&)0wZ1D)V=K?$F$vH9|TPH$ldC%i{yA3Q;EN*i^QfjME@^4R(=Us`3{S9kV`Uw|+^{p1stLj93cD9&MC&kPNS!M_+KMUDtP83s?)K>8bxC%3eA4kEfH zfCw-oydA%g?)l-DLh{Sr%e}hQ^Ge34iT&^$IKOawHm>3EcXD!hkq_5PWh0)vo*X9J z_Qx@&L68sCv(hDK;_q6=tsQqx-qM{mX35Qs$G(f_8fVnc-I{s~<-UO_#%QJsTV+Y^ z9-ynv1j}fjfP64~tWHOH{g!|oY5y3x1`VDIOC{VNDrFQ?x@G2S50T#i7Jc;A_0G!v zO}^5p&&pYdVp452QDb+ruU7n3qCV;<*_P=fK1ARDGj2mDmuIPs$|}N-BZ0C11f_ZX zgfPu^FEY=;aO>h-D0h`0`{2$hx&*(T>aVy<5S7``D3Lh9*?2>LhGx(Hv2v|2m`m1j z<>i(^KHbBOC1sEsHjoXSK4+JvT)Z;r`H8d9CGF4MirqPG2!_rOx!kRSJ~N|%nciVJ zuBg)Nnn_fKP@Pp|Nq5fnfpKdOHf4?>Tgjx2BI_~w9lx*1^S8sA|P+01>l%0>;#`VNxw zC`nxEFnR@~PUhr?ZYsi9zYi;X;Pba5eS{>P`^_WUuX3fTNp-~KP3E$Gh%oYea6jVf5LiItYw2qWUY9G7oq`0Km8G235TwWLn zG6S|M&f%C6>~(irR43|ds&nGfXsPL|q|BfGDsu%!m+DHWe;OzpjH_wd&q$?#>F4(( z>t)(6z{A)=(ehwqgCJcW|E_rtjfY-3{ln=uuuea?LZNH3ue!UT*z4@YwN1~2>g``N zD4VRPu-jj>i{jw557@R)0*ZvzD6*t^lH)0M|40bHO!vZtIm;?U12=Ce^# zWjMG57^(@*Pcn75+7$y&K;YYvD+k&bzlS~`N-nb6!XYGBTda?Uad3a*&DR@TNh2cn zY~bnY;&F(*$2ByD8k@ve8)>QJ^%8meaAVSlfX*l6KYI!x9_tkTa`Gqv!ZR~-O8Acu zs(DTW!iPL;F0E*;_hDv`Fjv1nMCew*I_J<4^8sh3N9vCK9T!~wHB)UVHfX)0mQ;2Z zDuk}@1eFG=rTMJr7U>5$HCrcbQ%6f}Nj34hcM@{#Zn*WBsRsrit7k4S&F6j*&`c-9 zmjg)zl-2dhbDOxospgrokVFmP+5YkQo`>1+G@n9J{~bAd3C3 zuu;aK_%7OZ$H>-L>Qjks!KtP8QnXJMc2^xDO8bJla^3r?J_8yix6cFg3Lgc{XiGOM*>wbcQk)v?yjn#KF1_|%vB*s5E~h8 z0Wlna`H}^KC}IqKYM#wQdl}Kcc^ZT9>$c*(@Ay0;O>cS5f|Af|y5WU@P-rQVnj8Qi zBd$g=BDXB&nFW(mK2mQ1d2?}2B9A1c13eb0-Z0mM8wg_H}Z1I=AiK7=7vv5p-- zzl{k19xorGFlO~V7*ocVDY8%F9b3=X1L+7gE4#EnG$9>yP+CEpLSe!f#pzHC5%%G` z!y%MkeRON1sGb2~{=OE4A!rNL(04Tpm7<5>bEc$v03`vr2La28<})yz3!<)3`Wpi3 zMnS`(U3KwqCfA*PG*kjGhhf+1|G6d_)$fN3Rc#cLw_Psq^T{^{Q1YK{R!%PaVRJIa zPqljnEzzc=9O`@9s2P{S7r+SB!RE$XSsuL2qtf@YIYn;yr!HO~BLrxkv%0xm@``$< z6+PTpz;3cQcd5ebJsJoBEF%iwGM**EpYUW%GWcSe0r5_7?|2x0M>eMKHShrh6!03` zfEl8Np8?>%vIX2l7z5`#BDQzvOeT8Hjqe0UY!48=6-LxIZTL{sCSYLz_h+Wj`!M%m zX@fkwlqeK+l|nI8dMgU`6|O+y-QsUtR+tF>Dzc$)AcSD_(jNJQRDy}lz$NljiCNx5 zgTyNUKt2_~nHasFu(50;KIkryfcEq4VHJa2-q1SSQnK~&3$=+8RW7!O?4N* z(9oL&Jbdzf@GG9bx$2NN6IL0_xVRo+KQpCIi6 z-gr-*NvRaJ+tJ$#)$f#uK13f?+;J`(a#II#5_Neo#O!xViN_4qN)T zrlCOEG7^M^(mMo86978um@Es!ZjJbWa{H=)m(3hK!uJ zgn{MwcT>E)|G$AqBiQ-sb`BUQ|KLWs@j{&JS_ld05+;TBxh)2IXdi~^B0)dBG5b4S zV+Tzm84)-^;OlM1ch8A>Uc+S8b;71@nwuT~i^XCj!HHJ6plMKi7f`c>jo7WFQ`x(Y zj5ulaJy3hWJC;uj1B%)pY;ZVnl$)X5%!_Ibuhl>Ad=czS2qzyUAoC68KE3&xObdVF zjVKF4fc|ZboHDt0+UMQ~+ue4G-R3T^3xqT-#cpSHm?$+3ehQ8$P@3=r;+S+f#*Gm| zn!v>qH~rE6A*Tq0FRKK4%Eyasc;9_ll1=FF)EFwD5f9;sE})Qk6))zre}d8<#bW$0N^sL(y4$L6S=$4-uyp)V zs#1+<_CE}j=Bx%4NW`@Ii07d9VR92FNB5#iw=iNt%O*|=KTBs${s&pIWBa1geik99 zXC{tI+exmh4gwJABh!y4kKMI7=CyRgJEyR;ycK;3a(cRYYep^zA&?Jow;-sm4PG@p zvivubg04kiIhWr5I(tTIDp*&>7(9f~V!hS)l<)%!LC*UR|CT2QK!Uv8^JX9(10HsM z{G?j3qIZOB@pHsovj(bFqkiO=mtZd?H{s{C9m+RV%0EI2DX-08yXjbUzx3IuFtMqrsSP(2v{A46^yiYK6yW63@n}&7LgxF2gf``bxxdDClgzXD+EPn4541Xse+zO?@AWdKTWk&}_O!CK|L&A=M ze8w?hJ%H^ICZIX|odv05q;dGPEhI(_z36}4v3)hAZitZ1i5d7V6uh0+R_ySW5R9Cy zJMzq7=PZG{!L`?A8^#SG3pC! zKuj`tW7-}52T`vcOwPb5pDZHgREq7I3{m=8&!n7`!)xrxzA~DasC1#4<1lPgBUH64Jmg`wCg*b-4%?7kylS6f{x(DU4-q$T!faS zBZv_Ojpy(dFzxZ=3;vU7+p!V8Vu4ICxZm3P-djOy2p0-QU9Au$-le>FODVKM8*UMK zdX_zk2a2X#fD$9$8Gh~sdG?1i#&Z$pkTU?bfW7((WFOkO`_p_$@<(#O?UnbP0+`%@ zU06@ux;`Kp2VS)a#8h;{OQ&PKfGH|oscn~7CnY&k;mepK@D4zRMT!Jf*RxhgrDqKlKXlG2WmLWa#F28j?hkTyZD$ z6$p3 zf_C;VV;>WVn6G%Bt`F{u)*)_H^rbS}b=T5af{C9%06IecsGMz74cnaNmvMF8arLUKgQ6I=uZ1N zBIPuBh3)qk2AGFw+zZqyFv2wN=36cz{E4?DBD={#!4jOw?Yw$g+p$J9(5Y@sMJ+^r74Cd8;*4S0TfJ zny>HNkTp%O3MgwF@A8Qk=TA{hq;~~rIYs3?iU`c1f5SKtY5a#E4`X~HnN{K9GC7tJ zGG6joJL3TaOW>4C4-o_4nV45oN>)qum@OqYGT7Hazsgs4}5sW@9fx1gkb`jKz9_G z=L7K|gjb#vUK+=pYUfJ&r_>h_FbmU zbbri9sB%9Q)CO|_A+}-D)S@0effLYnL&z!dC>tbF1cO1rP1V$=bb&}I+z6f$+U0YD z`p|B4>-J*pm0-y)w9#addGgsp9?1=P4)2eDE{8kLimcj4F9wvV&b%U{Mp+AL`Ma?^ zUzsD@vTj3#tP5+Y7m?MUU#6lEiLTYsRtGujJklfbE9QVdp`^>V`YG$C8vmZQ-3BuE z`=q(!@4Qs4RY^8^0HE*w$RiuVUTDd1JQJ|)3u?Y6PVEV1PB03*zZ`=PA<)7Co;7B- zSTXmC{Ug#t$@{7Y9?CNNN%5oYNsg-ZLFVPNTp3ml4^!a_YYo)$RU5iT<(_@HW~29} zK&-0RRH{Va3G~Faf;teJuuYZN-|iFV#4WDP^T$w4O?LAokz~W>A;znnw;dd^XsXR#Ue-Fmq-seNoAH9*hRImLf4b-!{r z*aS{9>O*Laj* zCt0~9l;A<-yP#Kxy`A5QPU-S`@d%p0HKaF);4*G*-{|#-z36(;J+09#f9G5r2C~}v zT|8PIN-@B@xP+lP`aOw}(PYkn`w)J?b&%6x5uH$xlh0@KF!$Ez7T*_9+3t?pIm*Qv zEL`oS)L^2pGpkr3K0}zTpCTF;=!93v9i%UloXn}HfxR#K?QuvIj2Ursi>PeEiyWYd z_aj(B1g4X`&jiWpCw##XZu#>yM$oc+i8P}bIiHgc!ztyKx0ER!7wgu3pXcKd@y&DH zuF~qnsR{^zdrlLeIq4i}(nmw?B~)ZK?4k;v92t(`;IYm?{1oA`q`^5B^0DM|cH#rTOGQme&sfDdIuY>btA?5I58s2I5l4d=E}Ud_r|l{q zFSl08w&xF_#?oeE;sti_OR@faKOI6?ynZZARX{J z_NPFEiSFNBEPmWKdczu%%6|ijoc|jrvT$(zKY-$|6~pN_*Z2z*FGFo%MOg@!pjGZ< zRIN51gjUoeGj?p`oCQ!w45g8<;+r(TTb;HJN4^8Tm@1(FNXM=*KicLIZGp zD*wfO(V%`{exZN!Yu%W9uzUTiA%D8^f7-Qd`8m|P-RX{h5z#tk-0ErH{toWgMJ=Z( z-fHzopfY-w7`!&RB+s@w}Skim!WB(dr6G}(cq~Fkb$)1Ea zn(RZQ>B#*$-JkS1vNR2h&9k-V?0*fhT=o3_F~knsr5(P4Vq>c6wlV~8as`l$Dd&sgfJIsA_ugA%6B4;=ElD-5+%mtv&Y zqrP{GIfGKMw1YjW`;q*ZhCs!dX|>2vRpTNu@aWU(q0*n?n=JziCKEENOY&aJbj>KY zt46i*0*advh=^KPv8Lt>PzasNqgSW4a=7rvGdXm(Mm(0jh(}nl`LSWUSn%@?Y?|uX z(g_{arF9{R3q6u^nOPJRFd7|aZyH#O(!l)!u9*CT6qbzI8Rp)Y{`l}#gGCA3<=Zq&|5xg z(R_R?i0m;SXTZ3ew8N z8hMy6dfPNGQw}uv>;at3MH~TKTvD~tbg<-!=S(9~&f2*N@$l@|7w%0g%=sKnZYmX~R&&oLom8sUXtFpGp%uG!#zd40L*=>;OL!Pvu~> zh~--qPX)rKfxZYTUbb?k5FRpht#PmXQbWrILJfpa#!;ho;(3RE0VWT_VobWyRi^Hs z6D_q1t9L(HktE9lb#!W<3|I9&olavH*53LYR^!N6by(iXelx42n*Gs4{z-lXE0S&q zpyMtV?GwbJ`)8lN}~Em*q=8OXS+$Zx>NpT4>*>LA8)4HRbHE}0*|cQ2KVFJdp~ zljK-P`|x`hzfTh2`#$M7U#Ml^VG1$kG3?CK*>4;h zT~o-f63SB}4zqzwKEV*cLlWFXwxr21SpBR9oE+@W2=*lsEz=jwPum*&<+~>ch#|}g zx6hLiNJAiK_K2M&`0*0LaS57&(mD}`PJw5s6TA?okmap2UFq;q6NnN@RR>?J)gP1d zLyo7QG~U?5t;Ko*9^5&+$)80#c&@D(i^n+*-<2=<(?>4cxDL3TiXx_&TkmN0ov-ma zbDJZ%c3HHi=^E4+^VNscFatefkK8VsjjTi*tPS#xhmI{>`pIP~5Uc}~h3k#Q^=C(g z+Ef2PS*bN}*aH{iBz~Ff{p;3aOE7HnKSBeEMT2Wm?z&_Yh5}uT4ed+fhob(C7-x5& z>p@6YA@-`n_`7S1q1PF34M~cCpLtFyri3@Mbov3uj*1Hq*oJ_J*?M z5`#r#gL`<$Ed{UvRI@@LUA{A4N>SI=4q62R|GAE)T@6F@&DXZ*eu^ylOUHJ32YI5_ zQjc4F@3|aLxKll{97}8d2HH&AJgE#5F^=h-{zs|1-E8Ae;ol*nQL-*{Mt0V5k3p_h zHaSz5_Hg{jBWoZ2tVUSz6O3Vou<9L{T^&c)jn}>XofXr^=UfD)C^vAXyfq~^8S&X8oNzt z(+l=5SRQra%VtjwoBRwerTh$H^YWp-o88}X{078R$~|HFRhOM?2ti4$j*^m(ZvHJ_K8?}L=Pu&Xy>X6(6o<`O4+xB!_E+B1mHU`Z8{hc*1(&Ntrn zQcNAEQ6lz_;+|K#w(2!}u$F8?NFb9@M?5if^;c?;#A=l;OrfYF&1)wzgqi`$jcvaRR7%NnP= zN*Bw;H`Dgl)Fv2vQ>M0n1`E1ox1fU2=N-rj$%sfs=}SHmyNM?b%uLzJja zEFw5g>dFonJeg$z%ynSYPynz5>U5Z3R==s2WE}OMbrPm#Bx98Dl(0kq9DQCRJSqXJ zQ+!@_P|}LFm{Hzcb;6BRd|toBdcd&Hs95fnV^^Uei=~3T%7k%O>zly%U9R5lm&Hv- zHiHd)kq3vsIau#uM>|8}jL-9FgS0?g-0q^61?ET&$)}I&LJO&y`nPn51Ij62U`>6c z_&?D~_PmmFD9d#d!(Sb-PgwRZ+{OBzF@#-R-8xImGsa-7IWmV{^qh^<@r-upS)LmL z9!*g|_MdJK`EtHq$ycv%({-XD?7GCiX})%X8peYns?lUa2s}Ecsz8Uk^bxqurKep< zz>pIlm{Na#&Ug^~j&pX3>|lRG>X{pe5QZ79=M=)M!yY!y?a zK;b#q9Y<&Hr8t3KJ$VI6v=k<*KYb1++(3cq-+WnOr?)CMX`Vo-w?OT)-%8;9RE*=a zAE8ST>yN_4$XVa`!&}rM#)t?|f;6}RUP0kl$y`?E$u51aQit^mYkFPKU!DpJ!EcLR zke%q}4ZzG^%5*KLnn8_Db2!8xVBXl13n}-KNVY+qP}nwr$(CZQHh!H@1zAtqyv6=61fdW`4rCJo~BI#e4r7 z0|sP%;q~}g?6?S`>BQp*LzQ5Pis@-Em?K_fzAQpz=mryF}iz=dn`MXg<1z#kh7$|_3`&VDo;c6h|%- z3nK!_Gv76FiGaG);JsDZz`jeG?%{$V3xuA-pA|4(Ea=c3QQMh^+cU_hV&7ZO5O>^iF{!_1ihr;WxuaMd=b3e!;5( z^&?9vIRk*)HE{K(?PZN$!(iSyn|i}YlGHUFG+X~|}I^~M`-qI8n^To_zU+!rP5 zYv*5$t_kjul-gsVcG{@B=5^qLbvT1szq~0&%D=ijXWodcbvPeQi%oOL1n8ltoZ*&N zmcfN0(;QAYZ3`ehtbK0Q25FZVtgpD4!NsSXv!cSZb;g9`e<@}>2J}QR&C8zVH9tzO zdIvgW#Y;#S(sk9Txv4y7a@c>pz@ODo3!c6P> z0yL5AK|ZgyWTK~e7GLQpO|CKe)D%wCr5wzZODX`bpm1Uu9UuRzHTpUq%y~GBb++|u z)m~Jy=q{sU5!{KzWUo)fWvptFP+`XtAzZHLQH)MAf;t%}k8MeE%Q%}m0=|1v0 zQebysb#!Gq4;Rx8VuX6JepRCDN`@sQz z3Dl1|9TRg^{Ha#}#l=gG^@fhDS{+L?gAt#}H#^Tw7}PHyax66}$^80i4^!S?^d%6V ztFb;rxCNTj^!7oa)dayjA%UsiBKr_wn(GE#JQKx15vQih5~or83dYP#UB1J}Ds!n~ z_cjocQWh|rA3Oui&YnYqRFf-)IEB)N$0G3|u9F#)P#pZ~ApT6hhl?`Zm^0L3@)hYz zf69$`-tukIB?MCO@~{O1vNHSxKkF2`uw#1zdY*r7u84j!Bi{4_p;vp&>Cc94q4*KcL0f5&Fe1 z#SqLRu9V<56}x1B8S1ATg+j_BY}9caQ;h9Ql_Z09<=%YRfbxFv6g4z-{i&IZp;@^p z>PuL6CIHn{03?6gwE*<)Y#!39A<5lsPp95cGv9uHg8aljXPxGoc<_|K(#DE2L<81uGzb22PSGx9{Qgh4bFBQ480 z1O2ky!GK`UHQ4H%DC%{Vf+REg56;8Kyr~a}kOJJNFO9yBUc>U5M5?&Tc?+qBSmJAk zH+&LS>2E553wkHag3kdf_Qvv`599dO`slhOqyVNv;UCI;Vu)$$cJ+p9ELV`3lN13P z_xhEe$p*}qg#ACBgQtxi&wCUf$c;a!9#}hsk$`c2emCVuuH*`ocgwoXHi|47sirZ$ zJhq1!QJU3_Y2?gL-hO$ zlqyGtizrcONaM%D*{$Fb6OXpFk5$SX0hJ$O$l7I6Eo=^bN9nQQj>)NLu+3G{y*x`V zQ3HkB<=RQ0t(>9rOtmBi?vDd@`$NET>5S(r*zV(^4fzO+`AI-{ofPn=?@m}Dnej`s zOPFKTT`YqX)W{3srQpdS<5?)H^4Lp?E*w_ti@i7Nn;J@z#hiwCgxbTjxuV~y57)rGb3OO&Q zJ!UVvQ8Z6&M3dz+JSKQ~`#2Foq#qOMn{rhY z>90NABp2)Ah~|)bKTvkwJ)oMnM}zhqGKVV~;KNGe;sppcNJsEhl(;A{n9NcQ)aab4 z-KsHr=D`xLR3rj|)E}~Q{lzi9dH^jU)k?hDUa%Gq2?&^L#w`N0=w^S{@+hnN?=7Su zIj!)!ss<*U9UVc`}<+WV_zx&=kiKqn1*I(aQ?uByWnLzI6KTep(Z zEPul|S$U8Za?S0R-UsNEA`|_p5DD1(6gaUA75fjUjf3%jqc%1sCZ_*8yLY1laNeH$ z?+WoQZD-C=C{&TYrzeXLz+R`_^i#Y0Upnu$6F3Pmie0KwXubGo_M1JLge&@@;BP=Xf3!qQ$$mWP0N%%6^Q*;UK*@23cHASS zf+}-&FTXS22_=9ejfPD+g?DK!o_<~8==tU%W%=U$QNvZ*$FDWEnO17?&e@+7zXxLL zS(`Me&8k)}#-kX!hIP6{H>&1=A%nplE&J}fnx#R29>OC0l zd2awkAdW)Y=Ape8OTn{Pc4fP|E@$48-3g>C@xBu66acCR*`{x-EGRJCbp6n0u`LMi>6^xWepPuBGZ z;AzZcNvq<-pD+29s2*d<%e1s)KL#*k!=ZuA!XZl^D<4IVEkDJ6FwZ4hQR{+#$NQJ7 z)e{23a+4R^Y~9(r*Bmp)hhRT4-BjIf-%9;K6a7a4*k8gkkd`xVS4wYhB&$fJOW`bd z#cWcNpP3#sMI5c@<@FWtF}9P>q`eL~OO4mm2_`rjoc9kSoa772(Bzuv1(e=2$D%58 z5Q>!O>#xGsnUXAVEj}!Xk10F6I0WRfbx5?g*!YF^ibdc z)Ew}Qo=GtTRAZV`VA-fvbx!ID`m?5zuYMenA$&Xx*Ly@#OjY_j@JiBVw z!P6^MVzS1>J_l|x(5PbPSg*vnC_ue1&t!dE&IAKt*$>Wnadft%q<@){`I{*gn5)>K zW5&sYlr(W)@P_%asjsCovRUU~k*Kd0@3h{sgmWbdSeeR7XBv*FjT~Djs^ilQIgFLC z;n?fde#I^&<~9JrGt$Y#Q**HVW0veJk^dFbh%71^43fSqr>Yi0j^asjJQ`mT=OSP8 zqY_NgQp7;C)O*-iHOS;$?99u2hBcH75TID4ie<;W6bebj0U)c!YqvY&E@_E!q2T@1 z+sUc46;UoqN#lu-EHSeU4(^CC4zxvDEzx2JX8Q1U@|V2KUDzADa!m(SmKcMeR8PN< z|J->*L0|~Z&lcNuCZPYfr4_<4srB0YkgS1~J%=J9IY6EL$`}-B7?!2dp&m}I0gjFX z_-EX!^_QWmYWi*#oBPVn%_I=N0ch!ho(2qEumjPcW9fe7YzvJaPtrvWh7}>k@PW3 zyh-xz_i*I5AK_H-WJ;y4M|ST}rG8zm{OHhaIT#5Z1nJvkqPZv8Xiu>#6K;S}1)iVc z4Hw=94VWcTfwQOY`*|CF!`S)*1rZG1#g;o_B~mA4mKUxT$u+e zmfr%d#3U(TcDUF{hI#VLm@n;gG=Z(4+#srU_C-JS;0tZ5wVHK64^^qPLkS+M z&^Dx=@oCDzL(u}m!SmJ$w@|SOjAUg!G7}vTe~mVh(Xjun?Q} zPpJAFZ@jIhI%9tH_OKdCpL`?CM>q7FUFvudYLJIm#DL_<)G~h%j21}e!`ENNUb=T% z)@n^d)@hA^fnN_uPi5*J`e$&*)F&Y>zl4)S%QD>#<}2IYz+E6Sc;luvj6gs>IH;t1 zXun0GDP`idvN7GZd9WX0!`ajP!C%NR7%rpze!;5+SqTjGY3o==+$ zYgU+yncIpktN&%M*%`!oA#hcli{qSvXnGc<*^61mpjcj(${ir+8GJjqhdo+Wwcv|o z%QEWnjAeu69iDy;QX_w4Xw1BD^_EruW5$Xv+_CTKAh$z3KQ&j2*rOWe!#Gc-_(3EO z!r{!~hb4i~%p3IGG)Yg%*QcTvm&Vd1LpP_0*m>B#mM{o~f$0(r_ zxk&|H<{jJVbO7-!f!Ka{Yb^50i2X1y-YDG=%yw+Vb~ydOXf1?rq}B|j6THPflPI%| zb^@ry6nTnp&YN@q$081vmi=*jCqgVXHnNE244Bf%7(`Rsi)Q01pz+jZLsC}Yc4eF@ z9y`TG7_2xG)rdXiuRblc5c5z=#L^D?Rqq?TE2MzXce!hC6`ncI~F>ai7FSarcS)TlkJi>ws`b zBC6)9#6o0>fEVa;PgMBO5W$D<>+BD-su=H+7%DmTUZ5#}#z#)Q& zo;1a6-_+rR8*!JM#okk@d}!yywa_i{Eb+43>G4$)UnirG5TPm}-X;0l{|X)N6hm?{x5p zA*F?FeQB;s^Ah()R$qSoG>Xfpk@0s+-dkhwI?#{GULWc8*E4ENJpK;;6m*~aSd zp%RW-_hxP5A7xQ60SnQ6RHsgbfQ}MU zdkFKJUMhsbF_irX{2&kuGp^j5WNPUU{ybuNI&Ep2vRrwlBQ2Hmr}-bPs0(Z09Q%x` zx&UxiYnwlHo@VE(;m(BobEMPJiYJQKxI7h6EGC|o!l^5f-DVoocWRS4L$J|qW49X) z55GiXPJWods0&^}x@*_Ue25`LpxEA^ifbCkx6S;HC(2_%j-_gWl}>#-ocr^W{!>j^+FHTc8cP-?ODtyBC@bPqiA%v2gI`DK(Ix{nb5Ud|y=XiBR zyoHozL!|1b-k5r{;LW*LN@{r_Y>=L~EXP~%?b1?$TdP5`;hH@ z#5{tMeQ~;7(21ek4vFRVs4^R(P3S0=7IWcLjecB%DegP#OzF?~NajTjdjxy((@JNv zKmWn9J0KQjSSl#iUE}|T)vNI{S;`+N<8&A;Sb!;^*F_hQNPMsv+x3(=GSx!hF_m}D zCj6WTb1~pD%b}d@t&2{i0jFYpOfSOAy&h#=Zq*<8fI4IC3nIC>K79K(G~VUlGxW3H zDZ3-1xZ%uJnH);Q+4D&}hsmV&i{*UG_M-WbmfFpL+TE~QRf2u90*}ZU=Ny&-`s{i4 zD;M}`e2@h6&VD(yPiB9%OA_8N_2g0U(J?SG!k0O|xl>bpvZsu=9r7bsAQU7ZH@s{uVwPq$-KEMq1*B|&~ zXg1u$@jM0NZD{@|>>*P9mBwJQH|gLpSRXP<*Y1eC7G~137R}GlDQh)dFWxY*m3y7E zW0mQ$9AGY6F9+_LDvi}x;B7zYDHkqIVIZ0(ox}6|0wi4&y0n>X{k|;P81q|$!M}ZF zxu|t?MVy=f-`$c7Xze#-Q4z!(%7t}*I??g)!XOqC^>bU2bELoM4l!`H-Q;x>x+?xT zOacD7{`#pk431%N0d81r1wM57s66`W`IfB9)C}v1p5G^6~2wfj1 z+9$ivs|TFpW(o_YP+UY>D<`JK+*3id`?>mFE9)*ogtnsmuxHQiLi8$Et>Mt#?bEdv_voE}f5V?cT(Z5wN5K&v$ zwcD>&In)?nUEF5J6ZhoEQb|vk_od19qA18FLLxu*^7&2(OgS?!7)36VyQGo(!kG71 zQB^-B;H-PK(^!P_>L~(mLwQiNOQi;$rUdr7g<4i#_c`D@-{Wf9G3;b6$~Mf#^Y}v7 zY>da>i*1R|POTZ?C7VR@a2E)rFMu$ASt>3)iSYFjqQs(<-_Jo|od$l{^=&pHDCa&M zusf<=doVxB0E}^uY7)>qT+hBi4A^nR!7o#-9%^;j33Mf*xrDb5>UCRWzclOA)tbsN zsCxP-dY-&FTw)Zd-bbiw8SWTqv4d0>Wd70$rD*_mHGlQF_}h2$-v`g=12(&Kj?z7i$?Q1OZ1`o zt94$9+kCJozUJ-hXmuV>K)E;aK)gr+C>^*v<<<7ULPIz_F8#%v$xB(0G=H==x#x%! zn~Oy+_Zc@o%)WP~@CtJW$w(8s23*C00NT$vtysznk&FFThyFtgYc0AKl~YY4G@c!ELpn*5p7(C%JfF4DOf8jVA6 zkjK`$&ham}R|Wu30ZB9%pEG>8k}KN-=qV|658?H_+IJEHlz3oLL4+(my%_IBY7LY7 zebxohOav>UJ)&epylkB(ZEU^2XhJp|!vJ)m%j*MY7GzY3Z6s+V zve9My=FftREes|4 zsOzci#AD;{KItjg8LjYGWFA7|)t`|6z$|4aNA*&UbMct7%6B^rQo195qWf76@bJG4 zISL66$kLpTwFJo;$Pkj1UABDGq_RXDc|G_LT0q-d`l+|1(;|BA(itztfJQ7FDBh0b zayeeD^`WR~CqJX*7Z#}u^s}4Vl?S6prqffaVKEo-E3@5v0bbZiHyN?%s|(84g*$pD*|rh|^{-0c=3mZ$rJU zkqp=XN2x7c*3;f9elQz1Jp|Ez-h!{J;vAY1qUO(Gr0x?UPA=I>3>C)~>Mt@msm06c z4Mq8}N#6!q^0nUm39)6yQ^wx?^}qg2piJ+a?Wp$Ut}#IrQCc8xWz49oG8MaWV*n@V z%u*tv(H6tRo5z+CVnvyLwh~+hA3=d#IIQRouE)ewR0nUTHVpUE_%ztO6x8Ya-J%fG zn#avm$BspYEBv_9ZsLnyfG`5V5?@ZR)3X|tvG#Hyl6}ICR-yi~x?JkU`n>!;e;jQ^ zY?6K&X63U3X^TsG313`~ERO)ARpw(fWwLOG+VL+=UlTNV3G9yuhT86fPc=RWEcaEl zvHQE*+HPpzeXW<~Hw_IkJ)K>Gkwd$B9{R@WTLf@?Zpc|m!9SPYNElvt`-&UQ{(^p> zR9v0&*pgFI&6*K4Xo-Nait2A=cEvZZ3EO^$JzOT+RP#;AGu6uGfU{!jTD`VOaw+n! zPfeR7>qx@VnX1HRg%3Im7Fc7#yt8jKhzYTXEvB};Sy-I?yH4ew&tS#4Y09QwhCf?Q z4|VvMvge&UMozy@9T-D%1kP`*nLa7-4Zf+~M%{+#cUkQyDe85}pD*}T;AxE0mwjIs zpl4)W>I5qdUgc1ecd&^5Lx&B%!6$yE8j%ZeHaoV6zQQg3j#j> z+$h|!a{>Wj7d`0IDT z-3fJ7W^0Uv%S4bZ$=z#ej>|(~E64l4O$rm?YeyFdSS(#Sk*TePGN$h-)UXqCZqMJh z|FwTH3INOhi6AolZ-U6m#q$3kh;{!6;?8*Tu1mw1=u9x+5{fO-8+2LsIsbH&c#=BdzcB7_!Rr7q=&z0;`n*LzzxLj7L9 zhwHJ2>uXT%4Tz-QIDb^_7ZgZrnL(0R!@rRtbZ1TQ|a7!f1b~uIY<77 zrRojM+(k$BCT+7v$NCi3w60SIid_uafBwO|cQf{N?U;^J^M4Z@%t{@r?e_K$+okZ{ zNE12b+RiaK*$tARBPn)0=jq`D`!&WpB#XY zWuF(?A~o8jtfu_3XDRX)k%nbbfXmVoPmKC>BrKt8!+71>$L3D*Z5hVA`BQ4^VI8uXR5_JieiV#9-b|ToHI8PZT5TIg%Z!o&vtchO z;pC_crMcRucwm;XACiDHReyszjXOp5UAs?1Q-RcuTh#desUP4+dy8j9Y2o=RQVGpsDGfB2Ub%K#cn@d!GOV8aXM8BYNnK+QB6^bf- zwsl|FPCRRc*y5M6pD%Hos;vzFIb30R7W?obV5Z`CxKd?Ch4|5Kv7A2NUps5A#?8@} z=2uKH$C(h|s?@}@yO1hbGG+imUc~n~`QXZndyiECE*@HGQK@~s=ffJ6lWJ~KC|V_P-+|Gh(^q`4|9+H+ASLpoCZ%5 z(Z%s|UDREa<0Z3AAam*76t0>86~b8LN=maspIt;{R2}bh+nVhwW?w?wr6Kr9haCH> z+|J+HG!MpdKN%L{(RQ1)URYz!m^T-iu-?BK%n+biEFY?Av)O!kHym}Z^YanEkZ=Kk z_RM3Vyf6|ntrRf)WQFAC6n?xl(=~sbzTo62um3DO?|Keckq|Jp_xb&fi$h!-+^AJ3 z_2|72Lp;g-*(wILxD0}De?^&P>eeg5%C_TlWQ6W9c(h)>*<}R*g@mHJxJ}j0 ze4LD(#9FbQ56`PQU&Vo?!4r=YxlG}GI!+rz=jCq0N3osNJ)YjT4okOL(~NO7HPEZjarzAYV69al+J_Te8XYONIt|&tC+)gF>pf;obfD^0 zxf?ueB&rw_S8iIgMFfQ*L3)VfU#%fVg*Ms2@wPP0)d_gm1@Y}m$YVwzAzCI59;o{GQI|P*pH-Z0t?uJtqxBwr?Fm|`Tt@Gk&6{ippQbu3Z?5D>m7D&nW@mf z%l*hH-c#5?1H2ioclrwWfGY)p58t{-GSOn`ZI^VVW%IJ3SQrE5V4q8&7JMfB(SCE* z!uve)UhMh2gMufjf?{|8yw}s;5S{Zpms!L*w*vHUvgfRG%~<(%QAhDxmth?yHh?PY zD`h0>bOFgD*VcpRe~=Uv z<6yzb7mP}7WQT~j)=|<8|CW+%2*UNZLY!%HIn1fd?O28dTKl^nc|kSoxyOa0(kJe( z?@T*8G{Pf~jDsyOHN8`aO*uC~hLN{+*OS{xStU%~lp{?}rY=xyjnEAoTnq>jOh1#- z?I2{bWE~^Ax-QFU53)LN>Z#%icRMme$$Vr!COlEdFC46(lI+8nxG)2qpOR#h>G6n4 zakYXXfBSi*mtp4OX}|X?d#^NSCx1KUCo8PLb5sW>!DfuLfLM6#e2@R&k;S8dxlx#`t)N=07pufmm3 zA{dV$jNOX(Ex*DACayl3sErEo3aj4*vlX()i|g&LoukDZf1EFu;-E#1)&n!(3zp&2 zx^Q^!ri&Qj)E#{ctm9RPYi~6QXKi-&a2?$i>)T5Lr*G1P?x$yr;zZFZE`7MaVL`#0 zV7}J=OF>_oZQ51_+Gc(t?ZSG+{u1~~aN{8D{U$0Q=ZGY}18JP_g`aQA3EB zu;u3UCblCg^^yK=Zv6UKE)RCJjeo}fe6EGXLLiSJU4uDK=sm<| zRB+)D*Og{g9<23ZIMD?*D@g|4kv{!g`}H+zijkFs84~_E0Owlvi(Avp(z0pt`AraG z!L?Oe7d1qK4B7G%PO^r#mU26%K7>j!iM4u$;V@lZmfG_Nw^53+c& zwpyXq&>)K^71n#YGqf*Bt_c&npFFlX2}M84w$DRtr*%iqmD&DNtFGErG1P#R0JBV+ zufS*c$upiJX|BI%BtcJNg3ZmCb)Y}8H1oep2O6uv*DPs(@aMrJ{|(8jCDxX+Hogs_ zrVY69F7q6dvy{sk9 zCOdZ;>+C2%0LRy0SU84F;2pFM0KJ^Af^i-Yu8c(eiP*Xu4k|z~b}YFUG+<5{n@ztx zpImhPZbu1vmT86+nJ!zpSK%&x@eC>Br(>V2#0FDd*fPEax($^+kymd%@+8PB$kBW*SB>LRu>}k z>#s^AW-u1ZljpGq8^WYD-`{Ibq!n;il=2|`b;j^aD-8a=@1_`zkWabNEF(gO8s@~0 zCcZzD<$htn3l)NV#%92j7cZ)DR0LUvSAx*W(gfKEpaxXxOsPe1Ii$Y`MU(=hfUH2Q z^TPxa5IO+Ps2Sn(@zmfC*dGY1Kh6`)&vQ+{8+5RNAJa zr`u?dBMmQw*|{Yx^MG4N5DUWj;-&1<(ZOgnt=67$+Y7nS4jKRc6l>I&gQT7L2$YE1 zF3pL(#y>|OV71Wl1jk0G#tIBK#fFOWY%OH3W_ddG%nsSi#x^W(TpQjWVG0uYq4a!;$Yx~AwmIA=Sq*j zUJ62%KwYF@0g5&Osy9VZp14-G?C%pte*O+}+~T&LRp@jshW71Z?S^Tj;yGZ|3|~#* zxTzC%fHzK4R?p`*kKDmNsH!u}a>}V4}olH5olPZ<%q@kP$%V zVBe-g0cuVO40Op3;02c+E;y}x{<`+!P~n2;+zNpj1>ORgYU!oR?6RYyeMS`k z@6{I*A6LIUV^+pXVVw+pWVjwkPb<1MMBQS^p8Zj5Uo5G#JQ~}EONOp5#Qd?k5e!=; zAx&wbxAZ_9DUr)CVy3#a z#n}77(unVUP2h+)MoO~lYVoEbnXofqhEO1Hxmp0x%gMz>r0Y?;(2`}*pm@B^f5l$P ze6tWA##H1ECC*hiIY=jBdr)cfO5>w&wQAnc$KgaFTvPG*N>nG`<@;x}JRujy6YpT; zJFa5LHHcjxHGVSl8@Xvt{^tWVay0`sC_O7}I>+HbajDr(lC9=rQ+ z+HNxEqMVM)AfP4nHB;9@6$Pw|THQ^yC`nqEa7pe##!~nXpZ~|-ZGWowiO`rzXfD%@Rt}M(c?4``y;a*kUEdEV%$SyVI(-bFREmmq^gkWA zbfwsMO5d*c0bg@YegBEZG5-e|$HDslm;~AVSDpPI7hnDItx$)MT&Lv`+&g*43I29n z`}WeDd@bLikj5gi&t(yj)@xk3-7T7 z*AK?uJUWHPmhF^)e8+Ad^uJkj^ZEO^ynYm8S@XR}XV+)TQl23NTm?`JqgibHDcS|Q zv|kPO>9yeXFR*swSIGm2|3OhPq>(5cq;E?APi7HeYLvpkIJI)A*UmzR_|W14u#*U4t;dDc4rtwUe$e(Q4+ zNSYn3)a7Y4-J$ZxWuCIF(r$XqrYMZACAN#&4*I~qL#l0(?nkUlMrmYvi6i~U3X6>b zUj>h4GPEt#+vZxPp_6cmjdj`l<%s;$OJq}F5Ou{o?JozVA@#GXORVdxryYU5KZj=I zk5Im&A$;Bt1+U6y`fuotV36W)szE)Lbl&^0)>+a37RL=L%eCMbaIM1AYLj#tIDN-v z-PAS@<8|j{!ujY`-gpq2Zc6guuNq7upW+j1#4`ZSUs1$vxSN6d*~Q= z7D69=L=HCaJ~_{gm^omf)J>Gep!CV(2)+oupx4xK0XiYegAuWYS5e|V>)Ap7PIR2` zFpJ24F&NHQFVWp%oag9t0*3b?>$BS`vAhEnulp1W@tlVSRMiT8im{IRf+h6S-j8`<@tuLSD3Ah)&NnE>QLoZJVEEA)c@B`|E_n z^GM!V91&nN5p43y#Sl#0t%falg2x$SF!M6>S zsy$JLfvePIv{$%pmwZk)dpjjq!e$|kAw_bH1=ddHo`oF+T0I*!c1^QQ1P{eQ%mi0~ zpzg1qOs9n$DVO(l)g10`Dvt&c>j*;As{+R!cmp*NhMt+gR3J>J*oZQjbKScl44((V zfnXRMK^SpMNCFqXcW^-&4{*k_N&h^3etOIK=gZ9gsL=^o)#zwaj~g)xI+MtCLPWOc z&#nWk!82JadalPtd0Y`^tryVYR1SRJ6N`C26Cf&J7|j@G#C@8SSXl>{jqY`#f98I} z3(7@$;sr|c11XYBnM)w8Kr%;z6_s4Nz^k6dGlFJ7mGMmA76<-=lWc-PO$jFlNX>UU z&`b;&Gk%6gBPUp)KPryo-ay@`xS(I5B&tH4-}XWIC49oEjD`i+U+$O}ah}lPFtHO0 z8F{L(wac9L*BoHpMW{>fKO8(v2t!~-Bb|4^G9o%KM3L7x)}eXq(F(c*kYZAH@a+`+}+svJeR-A$q(3_XO5M z-4xDhwoKG)2JGMn(Gi2>R2_-*TjxEk+9w1ci20Lxe!L@imc7H5d6FWSLR5|e&BW(` zr3gM$x=@igR^AF+EGdD`Ta1tthN$1(;l3e!K$t}5Xgt%v9uq;wCkTai0N*^1SAI`M z*BUc}FI)0N;H;A$I$Lx+`$0JXz5b`wDj{ zt>v!P?#3q4MuV{Y#5=+_*7)odrO)xwyCCUSta$jm(= z`3D*a4-#Nx0TlkXl7&-5U4HI>R*H!eR4Y_eErxUENy(y6f|(O^4oDU!FX$*LeC)Y5 z7$4!nrL!W1BT{6~kmb;nA@fof1ug!pfz}DpoJ#5U8hwMxDFetHv>?O?wT*R=17-z1_c)`(F z)jrEoFL`HwDXu64wcuj?+NxH;)F?frZz9nj3>1kf)=m+XIcQeQSPRIJBE=~AjdV9p z(2<3jTe~s+!R|qZl9%#T*idJ#lGX{Bi615+A0SUE>r~EQLeBjFwVS(`=5PfaXXq(? z44^$zMi4vGEXXO)e$?n}jMIr{kn2WGCH+*L#2sE(VgEnN-F-BQeqh+iyE|&9#ief` zL=5Me*mn7`;`X%b|oKjj8`h0Q@@GOxWe8xez*f8OS${89Ci0&eb zV7aRCE-q5NtrN3oTodO9;rB4=9i`X`_q0Z(ks5~ff$1bXV24%~JoHa0gHwUs$+KH> zqVkF$bG5_40s1Kcp=3Jhbw>`j#ER)J?jq|g)GHW^BWOH|%Boqr%GDG^oLs#3Si>lv$>TO@)tbOXH#7$b2~{e`@9_^Nyh@`p-^Y_VNc)M{CN2yk zZIYoHZaArdl*lPOXdF3*vm{&`HlmS=4Kzh?Bl03sL)!CSX+m0GdpjlkVo zmbe=#WR9{Gl+Ua;^eVl$UIXDMcX|U|Ykp#XPzkL~KY_}jr(PB1c%nG0r+o)Tb8zK> zl1e9sh}?{rz$X?48>>-X!9p9Q(0H!8`6ea~$_b9johm+i#C6!9@*dEh*|USa(dGh2 zW&=gXEs`8dk%^guaWSfguCCaWLN&!~tLj9|BoJf_Qd>gG>|sEDZ;2>AZg(N`_LwUd_9ZLK}SWOe5>$X`rD zJHVL}`QSj5<02TAgpuI{B$_ml&8F$DqW?w}3`cdt53CFv(7NJSwCDxDKmckj>i(Mc z9Fbnl^G5`a66H6-gI|oQMvS576&e`r^iT<;F&po6Nua2))U$z!5)Fj71M~f-5We5{ zLhETW7L{nyUYTbt21({~g&>=XnJBn8st`i6Ua=;AGunp7uA0%tQ}e$nZG~U`i4*&Q zbd)=sRMH>+JpIJ(;2kU!Btn7_5XOx=mHP^+cP3=SM!l#3C163m*z8hP!Do9Hm|E}# zq}4&6m4+jo?%_mtX_L>Z7|KFMOPxLuNIHS=%HRWCq+;k`$c_Srm_%W48fb@WA?m)F zgfqaJ`I0-iOImf!aHzVehMwILAI_I1W|lnPXbEyo;%^j|JG{XAM2))+Q!9Fb%35*Z z5h7Newj3%oxfUuMRt9Nv+He3*>;$qURQinGx&ib6TTjFKtRLr8&?XD5Q7(eCV(hJk zL)pq`GGa{_RP{SdS%oV&!QXk>FL8%~nfyW;QS!!G#GnxtF)ycVBSw14Sup3%Zu;!i zKkSlh&p83`AtGgumg3QVM%zUPJ%Qzfo6M>6gk$g$TLJ`Ad6h2sxvxt5p{D?nU1KmprZtCjMiKK6iPp;QaXE-G^g6N(O$8JxUFICm+mZOoV}u395%lt#e06OX}Yr; zCvGTr$OmvFqT}}$0f7ZE3e9YdSY{z&CsWitf{mw35s6&I(|Z%v1)N^$C}M#W2bS4m zq*h{$^AULnY59I*wJ7dvoOVH28dH)(@0{b~FWX&C;0~Jr3!x~yip9K}fkt@*&$j=0 z7L;Pl=e(MWIuIfqE(#zCg+WOA^xgXLN7>Q`vi>25BctKE~DW1ybo z&i34iK(V7(BM~*^^O3vRa4VO-U&q}nDr)TIc+)|(?h*9=n!zQ^O>Q;-LQ5Tjc$C`* zn@$O9Cb$2Zo$z1nrD10GP{RiGUoQ6;+QxvK_dI|hy1?XBST+uSnHSt_a4EiF_D6@{~SfafyU+7|}8uGxeCU_OTYK zl5kABQ-x+izy!K1RPNmO2kvcaQjXL!v6I+}e*LYaK_~1UAmlDbY8@%Op%G1cgcqtW zl*TlF8||IH7Z{qkhaNN|`}@BbJBKJifCR{vtIM`++tp>;wr$(CZQHhO+qSVid);|& z=bpDb<LIx=4B0E7I^5R;r0;q` zB!LnR=(icl;&WCOLopLMe{q(uZCvNVd|CxJ?AUL_S_1(f_5wO^{&L!*0G}7}ik`7W zTIUKLyR!qNaq|Bm^te^|CnW5xT7H0*AdWa z?i)}C(qm`B?^BNEZtV5DcFPx-VYpFT;o~P^L~;vgEEV44;Ow&2j!XX zpcMc{mwKWi-8s%vb>&bW<~jcba&AQ9;90ne$xu#`hR#qhrPsVA6~4FA5LjSdSM4g# z^}SxT-OPqfY%yHx5aY05ed$mdDKlJ${(v!Vhr!Aj!$|lm6E+_Mp#nS2939~RxR{vK zjIgO@uft8Ku?uV~}YazsIWgescU5078xO^S|LNrvHMoI5=4T|Mk>rP3iyOtnSO2O>CzP zVQDqwd*QgA@s^cc>(=VD^y`-)FlFINGgXvI@sy>D-v(bt8OOC$h`YjxBonf$yTAXz zSwoYlbVzr$WcI|e1|MiYGpnNam{Gr}?&D43Cw7c(L%l}b5T7vwRf6GQ9&)+=Z@ql=Mh zD@-^PDNf83>nHE|N##@2@2LxR-SUduy30){2W9O$DKVu+Nfo-vH;i=@iPi`f4;<7{ zD+BOG+VX{Z)Sb4)?HNvk#}AaevemaF@2P4ve|ghSQX9Svm6znzSj$Bf6~r;s6P+Tx zVp`nK zTGdQS%P$?DV1_Io0qSQrzlP*Sq~s$~(b!<9sclsJY&b~+WPMSx#|BeeZh6M~UJQ|8 zK9;T*+JUowxM{L(U!tG*PLGi?q>Cn4GuH2sjO~%F+cmzhIaeUX&z_YgDMT*c6{liV zNp=R-3u)e2>!?*iJV9Rd?Z$_bpD-hI{U@GXU2)ktCm{j-dhh@WlNzy#Xocr&B zzZz+IE)=Vt5t$O|myU$-M1gxiugu-dPyCV#bm}=s-nFy-{3Z}G!9Dgwn9VcBiD&JV zM$iCvv2Wkh;KE%6$Rr!rw2b5|y*4&AAzkI~r$}@r5)(O^&EihBY_rDB&e+)67Iv6L zbmcU5D_RddBx&KHr6WCBG!s&*W|oYBNp}5;c0Kc2aJZR7ZMww*(=m z45vqY?hs)GY8qP{G8HQp&n&H9PU?;Uxi@Kh;ZA?}+cAqLZd`~}JQ0ic3qF}$Ymdb# zGo-=|ORpoVAz6>*Gua=$Cs6CKF%R$2hl6R8r6Km@FVK@6Y#tHxEH`z~8utv_&NlB@ z1Fc(J&@&yV&*YE*4rV5d2na_!>Lr`fLH#P0@^N^lc# zOWW6wG%(?mcWt?C1OT@-pK^S;7}Os+`$SQUFn0)b)mBz2*NY&5&^t7hXl(n@n}y-y zQm41MVOo_WjIBNSxx0VOG}P1W^kf=v8zO8nvKVj|GLqs)AsISarK<)SI-8{SUn~*A zs;E9uFd4fG$z;s4rKmy(7PkJIx(6{-lw8Eve04qD&0qF!|IBHmSH*>{9-5w7khK(I z6Omg9&BBSSx+i%8_VDPKn(X*wRcla-gEq%?UU49@%yv21XnfSf^XWlmlI&Yikb43| z!X&E?W0JI|3qJp{z(i~G=lfVbG7;j+olmI8uaK|IeQ+6tn**(X7{ZhQhplVS6r38i z3!3_xL|$A6N3ioYd(ds>)$kUAKM)>}G|1FNlRj+_88ck#wuI~m#^0nbWlYwBo6jr5 z1LcJ@TxXg?sP1 zgw{@{1boI3&LQVn!mm3Pf$!YAiipG$kEbsCRs9yN`;}zk^)Bvpo=dKw{}WMspUo{- zdXUHs4`#vrwozve+YN7fIP>4Bpo*FI0*cc@(uHSwzB^9CfK>+Qcv~R|yGHQA*Z$aq z9hAR?fmj#B`~F~jPUzah)*fUC_u!MogtX%fD3=E8jX{QDD%WIqqxa~N1NtBU_2+6}xJS_nR zYXH2^Sr513AH@y3^^Q;H_0vHseSBVZG2vE%HK&fkGEzn0=?|E%SZ!Pe;tmL`;wL#hH4*zEl}dcj=O|b=$#2MV&PssIP9Wkb*vRebcEX^ zn3rW_OTZ@XHFpmTxOOphRqxG2EIiTZto400E$ik}~zgK&o~8Gh}Xr0(CApGbNJso;7b` z@M2{zV}$3)-Fw2lT$~3yI|qA`>lq;x*M+KYBytb^CmCbb?^;i3>Z8?lDdT|z$#eMA zOZj;bGY>`QY3C=1>4|yy%!#8xC406=%~e}}tkNMKP55^le!v~|?PswdKSJqsA8y#0 zMSQ3zAE{8}m^jhRyKXbd_n@9=phhqy@ox-~llTva`1GTrZFHAK;9j~f0nYf{end5W(IM$c1vjP<7{7H#(g&|n={;N7N_nY-K(h>l4TuHh<`C}ElOkgoguj?1#*&sn z0=Uh2J%*#UEi!SLs@g#?9htpx`2`kkLGEZEC>Ou4E=L}QDFzdoz~rh8P`aDqR?76* zIo+fKmjg$Nu8)uH4@Qqn{++N6*evloR*2-#!)i}V@+WrIhqSj75fz*WZvT6)>6K=Z zt@+@?QLT?lhOMjz$`XvifR2QY!ZnIHu%TCLx@u;A)pC=wj?LOUqb=zGXQJXrGVA>& zt?m-Xsx%Co8sb>(zKCTm*n+2Bv)|WfqhcwN3e@{u-TE@ow-Xzvjzn%TE1cV?Q z+ydWNk}Ncdo`QE~Fe{^sH=)FsK!Ja)>o^&Z?uk=dkgYm`ujqw2qnL@GBXd%m2jY0? zxKKMyGLenj<)Tl*yO@bttG^bwB(vQh-#=@h(>HNZJgaaV0C;<5>e;$u@ltO_8fOk%u2N`4lBsq{sx*Wcq(!#ByS?74{)WDsE4#=ftu)Mj!C$`aq{&!QCQ* zd`rwd3iM^pw|ZLrT4!kKejb=Hub1|EB5cx)^5 zSKVi{B%jggGj@TbtOL%*H_b9EU6a1;kjxeslWfhpuw|coMBqL3e`=LLt!>CI2-xZn z7P9-lbCajHQn5Kon*YHyht6;G0QhD~s9t5gqCM;}q`A#5ZB$DIyF`D0_38m=QXqfv zP|iWJd_dphr+SI$vJ!Fx+)fLE2UFA1J3SG%H-3l4l57`VoQBZIhV<9<+XXg)?J0Qi zs@;?Cx7g{`o))d3PrsSSE}Qrk*+oi@7RPRnwzqDs06(h=+Rj2yqoWkmrTb7OjaM5- z>!D_8O`XuH*&0ivR1X*pud6G~JriKy`CWVmH)ZUM4R>#;m#LR5E6sli2spL=}C#f}%b zcJ$|TdRCl+_2xLV^-%2kc$Tq(c5jXp>9AR@TcqW70?9fR&RrvN z=k5_aQ2t;qf2&X_r2hEdUya)4EyE)`rpz|YNugvPL1s(TtbXn~85F2VArUTPiYT{6 z$={Gx8m8YTtJcKG3-Wr%vvztlcR;z`Rr9BW1%5s(xuEFrS|qdsi|QcrdU9x=>Z->R z7Lu%=o<>9Mu&5(TyLV&MxJsK{niT}$PwVT%EK|9%@E);5JS8ax1M8+`UOkG|m^=QA z9rl%043_PvPVZp*Q~f+VtPgj}Hm>{nIMNG6JsoXqR)o&2JIKLxc!l+?~Z;t?ayJbZc-l=&aZfk8qXGxZ2EKVVsy!3Si?! zc%t6@NNA<=wGAUKdgj_NygllM^_jbt66a+l_n~};u$6#i8Ea6h5E#i_}5@sT`O zX10PNe#c3h4EXA)S395plOfj{prZ@t4ci(YPL6$l_F*3ASl0dntTN8DudOheUpMr_ z_ZnECXhAVq<=Uv#`ZN)ae!0Hs@F=jXl&>O=kZfT$$5(t*6os}I-^9Jc>8I*xv23$S zy401Tf+nl=T;~jXhhs#996~>=FtFNr@}1DOl~0^vEcDEg;g-VmC<^Pg6cU=Zr<+y{$eC>O!P8~P zDq%bn>iwmDd6X7InZwv*o2MT-fWRJnbaBSii=Q4TQ+hc>;z5aB5c}%&_YIpYu%Gk? zo6J$F;0&A0Rx^H(u3HX;4=$5FI|0Y8|5S`6H}bWAK1SBZB4mf`{_)*&EXlJ%hjo3^ z?fo?Y^IsyZmn4o~-eCHt{P?m#`=+MAy_uIvor^urQk!<-E6G9S0E+y0d0AcCi`F7e ztZ%hf%4usz4EhMZq2xQ>ZybTEHu8oq$c*5*OzP$FUsY5vkx2Z%(LCn=m*%lBuyFh@ zG;h)YOWbAKk)?5O0_!F#3|}I-c0y`hKUAo`w46+{p!C{ITFOr^yb9INpb(&ZIhUcA z-ghtd9?tE$T2<$5D^negUpU5~Xifltk_wOQmgjW~+?Qj5$vtVY_vzQmJq}MB@5=%I zeYNmu*USCs?O?CpvN}cp6;c^nyNfH#c8my*5haP!L3pgGSrly&*#70?YJ9zLe0bJv z9OfU`d+f1uyFIwz=Z*nx!iXYgwA8LS49gW_k*9^?CAa5zUmRZSKG!-aQZ}z^%R9F+ zX?}Pxt7zm-?SDvIe1U+afBPo8|Cb_L(deSJB~#QHi|i<8ef~h|$?Wpe8Y+C!Q6EQ?RJ7AzjQGI;%o_Hj3ae zYDb(Ug`9lbiugzYfxjzJDbSk#BS;syenAN{o9)J0Xd;fZV5Jd( z=!!+0G8Z>qqS3M`J-XSBM&8Jh&!h8iwL;Ps%zh=>b?`n;Bq!6QYjaDlns~i&xkRhh ztsLVVQyj0uO`&!t`7)d&-z+pT{gCN4erIr9^(lU7ChV3^yQZ>on}PwQR};OytY&h# z#CtK4EjL5D-6~;9Y77#aWMb;QZLsy=vyts7WH1P_Bd4=9kwF|iQ+#2{iLn8$@C~4~ z(Q-lH)&WImeGVA033&lEv>{F%J7ZEHD?2Br#a7_nthr2Cu~h$LN6%GF;gqokXd3RF z|52BQa^+h~Pyvs$YKom~Xp}IZ3GAY$hdTy5utIxlX%fS_x=-ZpZu;bxwUzYX$UFtrPdMwRgL1X7B`*s6p&60W_vh?*I5Tt(U0l$UnKj z7(FYUl||FJq&~cMKhHpi&cerUI9m)XPYD*A)ibrda5{}YTw3}!aU{Z|ac72Y)0v|? zi{)7PgGbH2c}Hi|wtc9Tu>k6(p}a;kTHrW{+D8u=HGz+ar`f4OpewEVl*PL^5$ym1 z|G2{j&r;rI!xSJ_c%(IMluQ+5YWlv++PrI<_QR|I@(OWf2J#jIOkXli88M!f9oMBj z5TVwE1f;F_`e8eh-;w|-DF+3hxnz7z2f7Hm$UR%;-rZIT-D(m$5U^JX) zIEeY;`SYZt=XjMQlYv4)gJDCDfz$=YnNFj30@ciT#`VQN30F0w-P5;D&lP?zJYbF1 zzH$WR2sJpKRhCeK306o5avauE^~vsh#+CfnOW?roY7)W=o|*WKMl!vuhj83t0pMZaodmt%2~ zUFd|zfYsInmfPF1bNkIs)4^v$2PInSI0dGak9RiiBI&POO)9Lg3HjWBzjlsPZ_;%3BJ$Ex2p3qd@FG20WQFW)7W)5f?=0pT=}eI2=**B+%ZAnJsV_@c`>AsxR!o&eQ(Y@nM})L((p$=f&1SG z$d`HR#y+j&zSXp7#$J#lKvWqoQm`S4)v&p$tMe4|k^m)S%zq!?9?vMU5O4FB3-M{B zPpYJwIItKlfSmfqVk(KZ7KQt1?V>dLM9#G)VmHa}g2!HY1?vQQP%IYw=Qqa(-Ri9h z8$_wi;fh&D!}%5_q2|%Wo9Lf&LkuU6~h9?`1TikM!x^b%cVkkvW06p73==DjV$I_}##(NXnIFAz!xGA@pccSLy#&NzK ze`LO~-Y_;eESYIv42O@{)J|CAVWKH1gE>XEYs*m4=OV7nyS2%{uGd_`1Q}+5zjC*d zCMiD<(G1*o6MtZ9m@lWrTzcxKvKWE8fXnY2cgd5Usx7&XmN?IFbNIm~W!fNK zod|bRp5kP??J8RNwCW{=7LW(7d{kJ#;xlnS`tUa2bgU*i1DcgB_3f_5hjQ83Mb+op5cP!@< z8jOrBc%v@HCDFmqIcKefeEP0wf*%g_y8qAzZB?7_uBkR~lusSucQ-eQdY2qRGnVzm zjn9=HM7VJL6(*hLFP$|CjmTW>_3I-bF9L8g{il}OF+3l*%|D{{pb$O;BWW36{A_OO zE5Pl;TpGW09T_>Si_CD#lLnDla#HNs1O!8DK+~vr+Tw8cekiM&c)zK@6nm4lxXgyn%u5a&p34L+y1g!9Gy!`Tx<&s?_c`C)GS@}%-S&y_YuQc1&%oD zG3hC|z>EsgG(Wfr7(!^}MgM*_Fm!>}>+S3cBDxeA=IPh}X|)e6u@b?=rgsdR`tmD{1$oHv-)1ZpyY;F? z4~!HfGfzrc#g+8oFa5qV|fh%55FN{a}j~p=X3jQ|dYpD{KP>Kj~m6 z5kFgw`|#pN@8w%Gq9s}6ulmdh%hOI~nm|~1B-aBfDb)ik#2VD}D$15kFqJ*035rCi?EkIr(Eb9n=fnV9u+Ri_7wP$gh@f|IOm?P4*5Q2uEyG9cC-9|WbM z23Alb4KJr|HfNTkzL%tr0I8eRHQCjjHIBK+8;dAj7-$5z+i4Sj%e{odq~4;a zYTR)1oX1TfuINqtV=DVbpONd@;t&u-^w$s~6fi4F4Qta6jO0KlnY@OR=8y1+?6o%O z4Pulzb~vCW(4r{;$E>lR8C2ANNm+|r!spg>>YetF#X6MFTN$hoh$K%M9^^Xgu$i)g z65%B>+s7WIZcp+d*dJ}Dy`zT41HZ8W@&0X_os`I8VLq~`3j(i%U&E$YHiq1FDqJGi z9P|u9!hAw9Sp=8-nWD{|E@L%k=TQn{TP7^XfKr;%T2|)Tc$6BSa5K<)Jbelt9f`1` z2DICsu#brQBGa-a$SBDMuKZg4tx;Vrvv+XQcj~0nybx-ZMW1Um_&cE zy!3pSW1Cfd(&rmgcWm)ve$ak%^gdxV)y5~AvZjy^hZNyaB)gK9QfK8gzhnfss7GL7 zMzBeQB!(99_@Z9I9*7u!^%w~n)tzx~b~E#4k9ZEaRM*iHz6F@liDCQ;`Lv75H5CfE z6g*EfH3~=n+9~WTiH^#|Dd=b1=CtL3k88T%(9_g|43xUjUzncZyR^(WccW zeD(CcUhyN=hMqoWut4AvTcNsYtkEicH9z@uERNO=qi~X`o`Iut?BopNiJ^C`ZA`Y; zpxRkQk-|z`Pg?OU#0Nf}(H$X74vSFVvQHu<-ha}g?T+G~*7PGb8P}hEYfv;b2y6#k zQR70vzc#dbM?8AXGkC0#b5zb&gIzu zV46PLZCB&SH70N}Vi>y9lMUhkXq^EhcBdQx&ciwv(oZ{*-2g(pb_bL7;J}(|SL5ek zhmx4}_QTJjLFd_O0TR6DBFa>POlxt@8%eH}b1Q_5%p+OZ#v^NQI5MYq@k@O=TK)N$ zr+;~2ig`AqTOs^gWDTrP&;Q((DHXZ;OtS9xc3EV+iGBmCx&*E^25DpmPfmu zJyf8Qm zr^%Tie_&owGc~$9*3QI)o9rM&8t|#-xxARFRA`IP#nxUGG4C(u@GdztW<15pOa3e? z-z~eaEU>{Ks%Yd^opBm-S&KfE4SUN0W~D;n#RDtKiNS74ID+wJNAi%T=q2i*cEi=- znC3Y6i6=Joab3GT+|~Tq*WB3}FD6$#IFFXSTw7%*6>HWm# zPj+JO6}ETx8~mndo6#bo*#{737cqawc0YlnF`m}>C$IV;|wP}Ch!m8V!NuY&JJ(x^97gWh%{yoIa(kiTPym`zNKE{ z&WR6k+^YL7uUIwk!1LP4H9^>jjjuu}r~6nFBvFmTb}?tfyd&FnmRIrttlN4;6r}-o zvQkJc;h9%Tv1os|YHH-Vrw^$j-MzBL--uhV!-Sjz6-%LiGrs5tG{&4D?f57Q90ZMR z6+ih}spGeOhBh6L6MjT99w%b2@tsRzEOUt_irRGgwk{fka84o80P0QjjgYLX$pPUTQf+I|BAA1nHKraK;{L}7EpL^pL` zyaJBgE{_*!ZB>hChn}QL&$_+cdT29GZo~1DdQNAnh_eLdMc~;or>!6F`U5T1Mqd~+ zS0}-ho4DK}oSl`YON0>34|S~2QsLSjz*@Cfbx}py;%@hIDcslt9jOrWGe54wAp;amR6% zC;em52Yu`t{Z0s&B-W2WcpX3919GXDlaDy_}_AR7PE&fdTaPm44w{Y*cRy1kkviG~y+ve_WGI33KNvnt++}hYn5&(W-M25y%yqtSRBt|TBG`8Y zYFI+&KOVCotT2e%y%evsIohTWm5_;1Q%9Va^kNIhtM8F-g!=VA^qHDw%t`g^WhRgG zVv5;kVn&Va;^*RCSrldr5HwS@mG=Zn`qVcYo9WSZ{=#634Np1|t-?K1nFjrlw#{wM zG$blt^pb?}Y?ob^4-?D%xG@bSIr^sby-IxndIN~7RgAGfU zSufBPYQ~A*?LmmF#*uEW+bEW1uo{1oQC+EvXp=_ z?vFb2H$sOn5YQ(mjA) zDI#lrB+<)zNyM6L<& zA*TL?!i*WRjXzUq_Y`Y#%T;v;KY^ta*x=6zQ!?_qy}59R)J!L!w!@){xp+3&G{&h( zkZJ-f%lOPa`qLGGQk$AP5>t({yzIO)XNGjA-2{-4z)UC_7e2>ic>l9KKTpAXa6xgA zHX*(?T}6nrC?C}l?M7TGB}m8(4Hg-VB@Ej-Zu5YOXAduM{)Dop)3+;#5e~e;3^D29 zbP(~k)l_MZ!t%5MAX@ zNHt=sP?+rriC5q=2v%Ajn;r2p`fgMNdNjbUTPw2K?$oy%v>O3P2D1;4o5;R~C~h0^ zY=EETnu{MjZdsC^TaTq5fk-2B67F;z8vWl;lW?y96RkyFAo?5hYDW+?&Ykl9H_=7< zpjqP=?mGklA{&d~KRn@{a`IH`T!*hTfl5*SvR*;AATZX3|6c;8!W(6ETwctCT!InudZ7svw%x_Q1(+a=)I}IeM|N7bWA7 zQCs*%SQih|l(iysD|5MY_&D(IAE)ZcrML-E1URq~VqnklIZ{K542Z+YWcIsC%;$n3 z+$%}ElgdL34O6NX)uKpRu0<9K0xYNVD^+2mN3MZpxN!t?9OnTEB-7eL+g|OW=&$g& zQ)#M(4YEcQi#mZGD97bri{XD$T-g)DwnYDHyIOtGn<)fKT+4I+%jky6j6~MtjOL!s z^}A9h*ftFU_XJ=AN(FOASQ9}B3`N(FP$AG(`JAwvTD)?iB9 zCDEIv!V5^>&zN*PN#2a>()6Wqj&F7+?MNB<#;TmzM$%s>t!{V0UMo7s${bmPF>6#) zLJLOv54nM2TfmquovO3q^WVlwDoTjJK@CD^G{D1xd?R(`jlrr0eTp7r40&!B*ioI^CbGLzo&QK7qiMsi<3rk`D9 zqvk;%9j?)&c^Nrdc69jojVi_@iB2TZ;nEf?>OO`cqjUHQzOMu9R+NNj?i?YXF0%}j zBH9I@K#`H}@qSnQHFx#qA`a@cXN&BxQxK0eYX*otpo&J1HAqc1#5R#5Z66O(zNPvlI&A## zbAGQ&s3}$9X)MyF+R&Jpppo|QMA-vMt{*0PnGRlsSEb21^rJi?bVnwYb=cA<_Iozs z9E|dg)Zga_Df#7!w;lA+1_)BZT)xj8QVfuzquWH+Lc%38;6q&MqJT~O$6#f=`$S<( z7MC#gxuaCWQGyLW#^eXoX6h-bnN~h*2(|2QBV50r9)}#7*p1A$_$CmHcWM|8ou_4o z63PK!`vu3(KwkmQ=2c!$-R*N(ERF83zDAWp4IhZ}scjx6n9Nd9DvUC?08Yb*+65B@ zfQ%5nsfj_rN6!eEExS809+DHhBTVK_lb^Az^m&&ryo1=nvgt_g(8DIP}hW! zMXOY4E)QU!`2*;>YzX=&Im|>H(|n2XuMXfOX#vZ%CuHGqs=`DKJ^z+uCCxIAAmYCj z?%(lF1Fi;$NgL;kk9E(E1p@-9s{sls9kRD~AD$T?p}FEi7@|~}Wy1sC$okDD7rWge zV`{zU(vNR}B|?U6j5E)8I%w!xZ^E+D71a@>2zQ#Pp#zAHaS5%)g|tV0DdAFQ|9h87 zQy~biz6>*sUUm=B7lF$D=@3w1oA766lIKsW1bsR_tUkO1dpPrTc}$HBIZmKo3qOjF zeqe{_R%DYo1Rpnq!Zrgo^P+C`4)C5`0ELX{W@i|~h_p(tWiaz}@M`I%jC4t6tBysb ze{Xeew0aQEwXO2X+-CF6Obz3pweXq&^bgbPQ~@z!u%A0kQxUcZ&@LqsvIiZ{dkHuuW$NeJ~EN~cP6bljaM22UQNI%5eL|SClq-y?6@Cq?h(BO zaLQii;IO6r*c(h~OC3|7N6hH4ap&VgnPWlJ?OBY+A+tbFbtx&qoS}>#F?POBfq5-| zWU~e{z7eRI_N3^+SQu$fq=ZKY4-d!d=c|(*kTyXmA?rSBn_WumNQt_f+ND^!KwM_G z388pzsLl!*k*0uO49^d84CFPC7m#684R|SP;eMf82Bkc{uq=&@2T1NiVrroS!6Wq9 zaou=|2?Z3HGPCCG6SUZZ)Sv+XJGAi<89?t%Z_fVa@&vVeqzE)aPHbwfGLTnv9vI9B0R^lMMVN@h-=u{Y z?B(v4BWBHB0^lgg_pYR@kI^%9*XOCGo_elh~>BpQuk3 zMWt6tkxrJO5@hXY!B~)IYW&=qBD(^fYsxbSX%m!*&OOGZ^O9})2qJ~+$IrZ+#xP+@ zqEmn0K+_n6VTf9cRj1IT0Y!BBUe1^1UW|Da(->{_NEpA0zYxzH=yYws8x(ERkcTaf z3XWjK3lt^VzRt&?%qb{_mx$ApXTYTq9Vugze)vuD^pc`E9da1z`nhw(#Sy@#lteh+ ztT7vNhz7-a6Z}w!WPw#5aMQqkkZFNDGaM*59LnMR(ZpQQFITM+Yz8>{3!c+9@S^LJ zEK36yVS1JPTG#(gCkZpwlm1~hZeA)jS%VDtthE50T)h8{NxNcMUGa?>%IA{wxT}3? zMl7BJ}bB`+zVkOX$AfAxzBQf@v-Trca;IB&Yv5< zFAmg(eCixE6J77%yXpbD>lptQJ#qpkB0*CtegMK_?~EZ^f+!ymk*lTLO+h{NXX9f6 zTduug0=u%X%9}DNNkjrzBwvuN8unaT|09vX8`INvd{j;^!tqA;_J|dg;so-uIpg;Y zLD9*j36uVZN)AlTYo64 z9|__!sd7GFj1V?+r%Mg;FcaYCQN4wLc15ay9NxI_XojPv^!n%T8JVp-zl5S}vkcCO@H?>v2npuS^bGFzbFen*qd7@628q%t z4VI*d2vFuht3_)kHt}2;J+PT%R{nmU+;PAP0ENCKroEW-+$L4D5x-cU!}ZxDPzZER zb#si5TeQ00y+G-8d{($s8&xRHS*o27!6ciT7q>|2Z{pXwq;U*&qS6Z#BF53DKl9<> z26?_S29j=$10^hhFI~kd@b7TE@Reu=wl6Oe{hLWr+Ybtc#?+0SlF?=fg%;W%pi@Bc zaA{ZD4l7rzHCM+Ii%M20oSMQ~C_!kuU(ZEm)#PcYdL$d88C0I4tUJU2QqbYP!SK-! zrNY+85ds**&5N3sh&n++(#inrT`)0(`v#I*+{ECZ+xa%QuLF{}AcKJ)cm-->#^@M4 ziI?@waNs*V%v9V6uL#(9#sHB~*jD?QqETG@Y9Kbf-Q_P|A3T|t-o?>OlWZLV`+OY) zNCERnq}}4XGvb+DTiHq=06uV}SMq0eMk)Qc`>ZXX{~v0Pxi%3;d)GWYFo8?eI0mLU zq8r$^JqNW8`j*ZVS%5+9o$*g04kcK;5TP}htg(rEg3rw=OkIKvgk)u_^p1R76GO&R zKA-8^$l+DnI$un`x47+>8eQFcWemSH*J;E>RmwEPvm9h1f2PTD;3n8K4BXck-VO<| z^6Vt*@jrZkRz@|LdcXSdyDve6xjm9WY*I^hR`xl9gcIY`dWaPCB9PVDfRaAa0!!yRaXt;Kvi#ST9) zPqj?x4`7uNi@}0mQ?V;dxl!61fFwke|%giXk)Neaa1B zQdm%h{3`Rwf616?%hC=2PIgPt{UxL7AecrI3#KYIP=tZlI^fk(pW!#*9vhtYamwzr z=UvqKV1v=cNwz=n8~$Lq>*#7DJP%4@*ZG7Rdv{`N zBiM5QlERMpMoB^J({_Aic+oN^@%d(2YbsD9c&0CnJF#Cq>^KAnq|}0CZs>;3$O~x3 z&~d?yUp^B@Lr4tL|>Bx2(r?sI|&4k8l`U zcRHWD4ZK=N^lX=h(Zmmh0ey^U>%|fwGj@*Pfsjd4%xUe0M1B5n?l(n%U(7U6H$uN9 z2NuI0(eAc+xahQxJSX5uOf%O~1xpH{Put=f8z&gHuU@ zCmwD1y9qsuH+BZyfTx_WF_vYQINV25p7VKsZF@`UvmQ&-j(5oS>)N~?`AYZdv!3|m zviqxU{Q@`plH)aJ!57I`y7jZ3%5y{Tx=5{%uJ9Q?Noaljvu+}b?al7t<5b7~iNQ&d z$g3e@et~%A>QYYz+wRYrh)UbQ8qU)CMm!NLQ9wPN#IF1GeD&PDpFke3@;-6^9oiVb zOqu*Ci8EY3MeXg5(9;J-ddnC7t4ZCwav*5mt+pUkZr0IYyP@gES`XYF7I@iGRrYU? zS843+{w>eI*i4&l=H;iURNwC#D{gP{UwUomUoz5R{l{eXKb2>muJRWdyrf2u@&VH9 z^yJ2vLgq=incN|=^s%Gb$l~C?hUsCak;YLE5><%rKt3i45;PVb+5JoDLZbXhn#1c2dP@z4n=Q-bW_ zv)?xjfoLOd(x9 z64j@pCbN1p`gpL}Ad4i{ z;548|H17|HGDIwvuNTCcNODlUy%&U3W{p8XZW7aHA8X^H=6@2MXkL3z?;oRJh4?gK zr=rq4#mT3xewIlp zsqH-j9;PoRsbV^hSB(nF5`+V(j{3p6j<#;m1psH&xm8QYisLITF7yU0jUjJRfW0-@ zYR4+0=S%Y?CO;hUE%ezfS?7yd2SlkRa*T10z>sZOP3)ou2Kct_{2nHNi!v6QD z*v7RBcZ*|j=)4_QOmb6WNlKI18p21i1W0;tj$aV3nhJVUBnx0nxfxO!iz9oH99qUH z^3>)XrN?;G6%ozCOr<^tP_3jLEb8hG0Zqu{_O zC%o6(%CVsvYIh)huIgVEmn?IXUy3TX<`I^sa6ib|cJimz98iLGP-~uje3e1b-y4>n z4g?J{OTK~)d3et`aNw3wRlVAPMi0EdaKsejd|xKZF47g%k1_nuS)VSsSUmVF*_-oMbktM;B(j(ah~Wo4)&P-hAs$Q&x~E_dGc`zmTIGalg$opH4b- zeZd)z|D`h5u*q3qk+y#e2Sl8?H@Jblzd?UOFR^_D2fK2b4D_R&x<73M@$eoF9}yA= zmwpU0OAY7(Zhy*}B_L6z&8Zf}eWT^9M9VBZM0eNCl#GWMW!?z+DPbC>91@`^4j058 z;W0m=z}XOM^8zwGz(8HwH?V2;q&VrsawLKiK*5R`0Uup%=aoWb*cu`%Wvu8bK%smp zdpe72pHI4v1lJH$^aLT6+2>gqqr-47ZRP`oG=YGEoPE;~aO{;U;j1iV$~nn8jl=?k z`f8#&fF*9V+PfXb8A(?4D;J1EYP&eSC(V+N8UHlCt2QBp0fhligX2yfBhQD@OWpfF z0{L(r|7gVWu)v?kR=Qe2;4o4r4px6OxOk#r`<#S4*zV9#@qsH$Q*w-c35Qu^&~)-Q zX|_3HM0#VhKm+W4?g-c5=6>96B;X+nsXtfkGyrtX)|TMx?f~$vTPZlb`}6bO26Bq< zCN;fcO1j?12J;^LWo*ESM>L^snof74wdP!W9O-jovEhi!2y;=j-PIi&vd~zVt4!gbpbmPpO;18) z0BH?boiB{tTL`d?2wCKcM~{s?gJcqKzOA$Y!tgjlx8o?DzbLL5Y6y*z5M1cty<>#M z0XwQ-E`4MAk>oMXsf3jzqk9vA4sl%m-sYf?4*1BCGK`GBzB+qClF;rN%hA&NRFB#h zV8zab6?HR#2O^~Fh>p`1_UPLvhZ_*-Q}u5;t6R%@{Zrv!1%ekSb|BS|kXf#=gC|pH zbGbc3W69#3KukFz2KnZST@?NxUf3TR_pR-aI?ZoWgd*_3+_sxlW4%b+J>OIHAll3|{_X$6u!LhC;CQxa zSGxzv)jd_qF6O+G^#_fa1Q-I7wrpGb!8jq3u~;U$I{Y@|Ps|Mm>g|FJW z*3q2{j0y%>nk0}j4MDwTq@#nnT^2@lbXHV1CAH<)4r*Z8^0nUoc1Ya>*0tg)cu6hv z6r3HlVm=^-(wZHuEkX(VS^M;Uir>+QZ~B<3BuN76oU|eWt9|zM^CjO3xLk|P`Wc+pk@5B5U<7EzD+_x!r)judY5TPHlR)UoAl7e zZ%llR7!n`TwXEM)_We?rE%yYVEMwY;=de4>J1LF4?M!@KKu6+5lRlB82AXce5>s() za%WZ#3;BcWtymapM&6!?l;@o#gib@C%Ow)KzA{5%!WkKmNF9gQk2YoamK=FVlN2)? z92h2%4-6~v;+DPUM#31uPFjEef<99r3+M`T16}}ocHNH(CT(wg8h_|#Ln<#r0RA1w z?S+8Y4`~|2C!%V?7ra4wNSQWfoh0uF68%rITzBhjOSw|`Dl9|cBnfkm5m%L>Vb{iv zPxE`tuav2LO@yy){R>cP=|a^?SC%+0$1166Z&@FO*(b)nK$xcbHyH5j>z4!QN7IX& z8*vRoBoP7@!w>Ktj%_3m4#ThoL(2mDw8!&mi=V?fR-^r86KAoCA11!4RWK#O3c4M@ z=Q(BHsdfA%CL0yofwsI8_%YPn|E%SybcfGePT$#|Pp*R9@Yb2oe@!j+QbfDWLvmys-ez)8>d%RhLEOZ5UHwBBEPlIHB=f9Ov zC}|H+s0JVh$KfA<(UP&=*Dzh!^BqJ$FIE`YA56qR^MD-zxfI(ow^%f=LtqIAn@%L+ zIlB{c5NuuV<}5Izk~>uRfx-k>0+B2iS^4GF!woWY_L03A5TALYhD&mie6=>q7g(0Z z%)gy={274!vx(|v00jSYIUe_C6?WQk#d#kieYMtv-lBB>r(3)M9gw=FeLBN5KBAoa!+gNb|7DhBT=<^Idwj-tmiCS@a z;GtJ{Rx$n>5w5-@`{+tnH(tYU)&;2xgh#~gYwR9hRGZkGEF_G5Sh_=T9-0%c;nKK? z?9H-ixp=rU%%ME%JVOopvFheNuu~H_Bu2gzIy?9*}vIPXSQK73#Hj+N^u?1k^B23BjAZ1Tnf{s$-cTx_=u3{Ihw5L zW4Jv?G2qrp`$g4El1;zBEmYE+OU!@Spr&q9<(f-JsOpy_U?syYklVAElye$`bsJO_ zKgQ7WX@CALHA21hfJJ3gQy^F2A4clD*fdCHSTt025n$yD3lm=1a77wM!a2^y!%^{) zdubg1B`rfrP!~AUK%*gOJ}8{ndawj&EuE|Yh%|>n{k2pq%t)>4G+egs@p=F&`ao88 z0KMAi1kI*--0TC2=I*4f+^|+sDzM#az#D6vepipSZDl#m5*aEQMpKT1U&>XQ0-M8a=Tkqq{cn{Y7c&XqpBet`PA+%1xFWwPe@d zuw=8#77iL)E%PgTEh*$_`!lndLRjH=9XXL)r{hq|R-tpRmiPPk;M#&rJ0$niQNvKOK90Fn4@@$os-QYGK%=onU{{nyDlQ$Y?g z>u-fw$RX9bWNTc6CiVOB(yfVDvR&jB)*WY{VmKz+U!`3%$D0PaX~S##qkvP{`1^-SoNfSJY9N7zTeyIa_g1~%y8CD}DTg?`mRsZI*27h#NzxvW%I(pzq=Ak7k zn3#&A;tO~nk88d{#9IP(7AeL1%6B?Zdvo#q0D)NEGIXXhMc@-n9liOx*hU0>TOTft zq|NCt1MYF`{?l??S+D!}bfSDeh2TJZbYaUIypU?l&Xt1nZ?+n_ks5<}5|G!sr-mWT zU=sU92vxnidSj8NJY7eXFU+|q z#rySANb$1ejbS}^AqKaA^+Pb$%@qiq<0L-cVG1w^aKp&RVBVOI&}>85E|Mh;4P{dG zW^eH_c0&pkX|_WJAN?B~jY~W|*8F<(yE;!Kfzva{K!&$UC}R>k(v^cUc>UbE=^uZk zX!)EGZIN|`ydAto*6#0GvHL)gJ)gB$L>y=tu=?9vm6E+HX5nd2@LJJio=ITB7?P88 zjHpAP!8k^=jijT;o!_2^9m|svcSl%5%uE|FiSWyf4~x?7Ti)%G(ENF5>z#iv1NXJc z_7%4nVY;&7z@*NCxQL~#?&|*pza|ufS=rd+rMdWhx)%pHR_S!TbQgn7I?s zM0-wK;F0+Yv@1s^d!rH4!({<_?n>S7vO1fjL=YS@qjansT@_qU@l{U-NG;Crj@pl` zB?}Hvi<_Yqr9>ls_8X9(H=hk+0ZyC|WnE0*1Z5aaQouNs?{egMYxtwAM)@!N8)9UY zm>bp7ld3+iSf15?e!|?h(DM$cM*d#32YLWwpVLv#lpq(u2H@bMi^|+8FdCGJPvxbangNnj;pq3%BFSM&ZAR!xN^sYxiy#LgN60M!Sef|sfKZTS5lb|DNG;n z(cGEIq+ZP%$n8yX3J}5|=~37nLg+m){HOwttPR zbt8V2Q9L;iqnve3WX$;TqvZrT_j>Eg$hYFMYFF>R`MkV2Pa;dp&$;$!wq%u)l;!OA z-+CyVPIJEr><@UIp39SdNb{=@&7Q6v_J2~uALF7L5W-1NvJr5qC2qkzeOub$1FRV5 zh5j%5kL|zgzyHnV;r0J)9*QRRJkBo729*81fJGGcDRM3%pDZL^WS(en9n4$@=n`O^ z{s+|MS0eh|>A5oa9Wpgl0X#z;e|G0?gB2U+FOn~x^n8CXQN2$KV%`U*_(A*UoZggx z!+G^xIea|hf7x|%yZdyxdDDypNM$3O?7jHT^$ox=ra_Pm2e8m4XyR?tzOKJNzh^xr zOox5gnl-<3CMTIZ}1u>bhDoyhFXwv0%{SxKJekO z2wb}%DWabQDUc?r^Qx;zV-Gr3=jd*f(e9qYSJk+_xB-01meO;Ala(gw<#Jscu#+AQ z%`7lq^-hvB<9qt;bb>17+G?;9h#dGPhZUDeOtCmZW?^1qFlNFV4nQR1PySoQlY4vy zvvmXcW63Sm@xQ#IVV@^jdeR@EsCCju`FyOgjnA2C>(hcJW+Su)ruwPtV1GD05Tgi~ zgsF|vn|@n2YW(B~)Hn6oNOKlUP=3H()gqMwqMP-uI;%$3f7h=5Z>jkxwUtSsB0(GV zW!XJ5Lb$1Ygk^&U>XW|%dhcRr_EimDuh9{$B7z3sWMnBv^lDH_y@ooZL4cGps_I~* zFUgxe$~iqJ0kLivg2fjFLbl5|wcF2n~>OKv5w03>jig4a!zE?n9@r zHl;fMDSX3qw_u=g)18s&hL1{Eng>3jcVcPev5XM;1jNnpkflbo_pSdct7}%zQ8SP} zRj|i^#Q>{zmMvx4JJLtw!N2Bn3@%K)Bg`sNt%fXb%x>N+(GCISgmsoWuVrQ@#t zaEGWvOmR@5%KWN!K&`=-)DDsG@6E~VvT@&c%hmOJ9=R-xnkDKW$m1h7dP}D5wZ|a= zKces#z6&u#qCy)eSE0=VMOov=hMo7;u|7lX1Knd;z>9Ag2?pfud14{4(a<|Cflbc( zZ?{Uk_kJVJA5(*cIw-I60>lLoeoiy%1H~Ks0hR(NgW4PXgUFfS9gK1WfWV)E3{QZ) zo8ocECCE>NSgt4YFfr|Vw2-;_$y6j0rh7}fAS#r0Eb2C%Jb0*QUOR%fQ@;Ab}~$+A)F_Qo>?OLKGd7e z7d0Q(Qw&}RwlMuhK@uejD7#E^UydJ=IQQeE2Ao6vXewBkY`BS%yA_of6y8+GuE^`K zleKP7licv44N2@lqp4EYmB$8^!?zw%Pr7HB@Mu0dbZHI^P6c+~btCv4ZlJOsp$Oem zm6l>>;Qb0RIv{UWL0_!HNt0$RD>an)1c4j$Ma6KML^-a0PP))^-&+ro2VW9U8>we7 zz#5S7GxZ<9P5_u>GI$g|lv9gW_YsX;wE))l-( zG;Bhcj9oU~n`Gvw^zU@AuW7N}9+! z{jI;_TA$_$0d?37yR^&d+m5&UpIsr76^#~%)F5)|wu&1cm#|Pq;%M8G70g?1W4lsj ze`QNnI#VpfvwZ?f47~C0E`B;a@u>kE_{7J80!xeB#WMOlQeqW<%8!nMXmV4ZtechY z&~@ef^6+mzxI^a9=trmhBIPrKWWSPT$Tm9uO;WHUUyzO;5LED(wHp6n!=ieX`0TsCo zE4??)pu<%KUH+44uEgPI0d3v1s(hzXcvB2V0dvXeQs9zeL-R3oerPzIPIUqAiL zVtR*?((ZOs%7+dJgs8L1*SY2?Kt)L^RVz6 zbt-&9UUJT}$NQ7+p|uEGQLcQbBkBuE`yv18!zQx}UlY03{<39B*NR@*L$qJN41RF^ zbk%%|*Z1PVctKhrUX0!uQ6Kqb5Bsi*gbjQ+V~eaR`S0DFD`H`_A^phuszS(rhekOo zf3g1kf>MM~$o$=~lQG4a*ttFCc|ak>j~a*1r=|$QSz3$TiVU zv4OXh1O7l|*+B``ec=`?zt|n={o~z!N7azR#jIa#e5DrT4QqHoK|Ow)y#`jxtq=GH%BDRzMSU zyl_J|_gobJucDy%tZVuyr(Ruyxxgnqz>GluUur`VWkEqu4=HP~P_PuB3KLvR{!IFE z*ff^~7j)sA+t8*g%OH~F+I~;YIlRj+T5uK7I`cpu`T4LSC`_}g$WtO{f2Q&_+$wvl zz{tkzaD<}r#>7a^-Xkla=T2B%Y`$g_8s|t)mb;WQ!NbaU-W?C+6}x!vO!Z10qvIgU z-DQx6C`Z*{(R3}%N+rWq2lUIu1@DMQZ zBtRIYr`S{1nRcHFUVGw!F=9&LQoyo^E1 zbcfGLR~ExQmt{LF0F=rE8(^00;m?I4xNV3xcSqwA82So43$qrIu2`MF}Z{Io~EblFwF^RK6) zhlkxom?>QBH?Y+ATM}j}Z{abNra`-_`N##}ZHyUC(V=ANE;I}o43TOaj2Oz%fCX|j zHBnc^O*UMaPot(OL!?&ZWP@*DP`FJxehl7HcM)Gu1AONGzyb397O7%vO6!t%@V>uY z#k>1lnOGQF6YYGBTFgn1Y2T%i>s*f<650W-NN;~TW?=v>rz)?5dB*v}3AawZRn1Xc z?nno0OHHO%8Xo8?D;P8MZ6zkEKyNaQ5 z+YS?ZgJ{x(`S&wD(~7~)66fBEC7b_NUWFEpwt+(_;)(QAs8ENLn`I+bWvl-2Zrn-( z@w)9S!YhEj30D%CgAO6*mMC3?MbwOz#4$rm=-elj&s@ypk@23b%HUH%Sh{6pMG&Z+k0`(P zOig1`9ibN{o>~{gW_DAY;^{rXLL(eCz4%3(j2Gt$e0EvmLK4Tq4hE2&b95v(fU^-u1 zY1`<5y84{$LZFwqk6;$E~p#leuepV#S_D>DF=`8}nMsEj0W7;)D^EBoA^sz15U*i!#Q; z<6O$fSrvk2FlYO8*43=T*zh{`(bgji|HeyaIm*#xn+xXyrQX$mSXjcN+rQ#X_q@!OtB%0QK;eDjEn5>Ru$ zlW;PMCAF3OhLF^z{O2Q!F4e*3xNAEPnDIr%T;n?#7vIOP9z*}*h)s;0P+)K4X9|p2$=p``54#k6xir7iuAI3=t3nw|xM0aq& zt-P)K0RXx62b>&_e_0mIP#B`6$2u+6sD>Qd)y{jSoc<1v12pt}KsUAU_?{ zDpurS13=%R=#`2sN58?iab6~Ffl=7quw?G!X$L6m(w{Jns^D8H{*=AMo!tl3gpI@F z$_!F}ph(%hT7RA&xb~`0e^^`kubuY3gZ|(xj_%ijHz+~t!t?eJ(*Wz-HMu|IDR?Q!8`FBwThgFI9Z#vXVI`NYTGpyCy2De&JhyBc(R;e z8yQJ~zTj`=Z5{0F-(j%wk?}QixaAw=4=ukdQ2S1k*K5BsgmERuIL|LS%Q8<)jCmKsIkzWausaMOFr@@iZqQGJbx3_> zL1%}vb`xX`K4-uY4B`F4>vJ zzScL4Y+^um1EzkDAX(qPS5>r0Fv^*?A(kxhmwtG)1alvz@fh!~{Gt1!RJU3|sAk#R zCYm*I(|8AfbcbYSz98mi3y{axO{g$xyo#@PC7GF^>{%d-H&w}=LOh&WfUe?G(+ zX~F%xpN4jRfhrarvF7?Q1ky%QgEF!1#H^M52EmWl1y4?wk)e? zWL|~Wo$@7{g}W-t397DMeh%N^y#7H_{Ea1=dv0#(iIj-};e~s>-x&k^@QNe2#n1KL zy40^rz}dWhcapqc@xL^x|Kw`l;PftB~*s(lichrdjSD)Rdb|l8ES%JFtAN9Y=W$ zbqq^X2^u*_3~dZsrc4EO$Sx|Qrsic2lfC_eg|;2!HFf;kQmnW!b@El9&8*#FzDCo| zVjk)YFY0R5b?hQe>4-ls+7~RDyz7>w{g-cCLf_kCQoY;2HO+n#X<(g*3|h&Ru}L`U z>}%TYdb{A1s;1;+WJfED)oPbko+75MUp`DLqfO|DK56f+p*I%rsb+uww%uqh(WB8i zLRZdaQOEZ7OptQWp}BmH@%vYA?wf(?Rxiyk-;7f$$YCkeq%(cmLOFQ3x7NM$A&Nbd zDi?z)+K$N@qX@6vXpN%m#@6dAXiExUqRBiS^Bl@!cJl}n0Vg%iL~gyyd+4<=b23s= zh~XPT&!=)T6d{u67Q*_7-N5E?QbDJ>lK}9afB?{Rj{tnCtEGO}i;;GH>; zR7TGS(6qjg^ELK1d$vX8LM^nZ zZT)MS{^8P6-qoq`*!GCl^|kk~lB=GAH-YQd2S*=$;f!ESjd+{97^7({VXC}O2qn9k zcj0Pk;Zck@C`$K#G^prs!@cM=&@Lm;$HpSP*1~Ev1kJpdFNz+N9`?T0P^%M5_@)%C;`rd z2#+&shNOmxRB^FCB)Q(Xt77d3um$i;TH$U>FGWr57c#BCn9Z`CT~k#@Ee)NEB#rrC zt#g1>WndWM`vrh`PlNfdEbs=<7UUck_fSpj{E)=SxwXRJ5d+KH8cjHf1jxOWq+)~re$U&uCo zM7q6JE1+mQTj59E*16%!t_&5x6PiQu0P`tW_`f5n9qvQwfUuF3Ri4eD5CP(-ZAo}# z4XbOSC>1E@=DoY& z7fB1EKm(d5;2NPMQ;fNUMFzACdCDemKI7l*aOctu3maj$9=7=_ysrj$%Idb<=@lfP zIZchABQMt78_Eq@xZZJEgCw~tQto&%?T)bdGAd89LWT zh8F(*ZPXq9nZ45{hk3ALpS;v)9?F3;q#ZG{JF~qY&-U@QA6PpceHJ*W)=pzX2|O&$G8Sfw8z-peoWEa!>)^F+r$v*W@H8s3Q-sN3r*vjd7P2PP^|5RMU9@{ zj{=O$ST$yfBq|FyQ)yGxn*FkgZ6tvb^ZPV~xulyc6)BWaxNb1aS}qAiw02zD`4R_T zuo>wPs>h}vF@rLgIBT@-WDrn08T+9Tsgb@275;Y_x|4vTIb|xVw1%yXs`I!fIE9Q4 zT*^jWLCFr!Xb^Pu`t6847CHnOv_4Q-Z}}+S7la*D0>;tLZz%Nn_|@#7U1Ov!xXs8m zYd#em$tXiTUb6e^woN=N9e_FmT6sEWpM^wOEJPj`w?tFBe8w+UordTdTcnr+H4$XH!xV8zc?C&6v8TiOc!qP zfC)AI{sA*@UPN*ZSZHZY!Y)&MXooeli6TlxXSW&n)7b!IuUriJ&gRCVI7x2CQA$ro zv2x#~zHQzWi98#H;p$;FYqyQz%XNMQxz13*DNU~Ypz9&@uvi2F4I4rOj*33L;Zr<- zgzbf%N9;t=RxmiYLog8{Q|(_^Gt-O&1Tpfj{$!rsFdC4hE@6n7W_E6NHVoXHR<)x1 zKBHTN1X+4%x-sTVrd&+9Xp=2=IoO(l1AaRTupeCnH}|J~Ub?fzHM+b$-qa81i-!Hf z=(YZrog*}|Nbd9YLuu!IE0wn(D!ITZK}vKoKik<9m?VUFL&WNava2S&_<-oKqc59d zU!X;NG&B!2=%>=FieKWWkJ#d28ngV$zs^CvK2i}#U}YM~7O*jLB}rYm_P@RLk7wd+dD^dG)A_p%g8WNx$O8-K|( z`e>5+Vw9H8pa+hy>;a@I0zuIWjF~84%3*#e1={1gh>)F@@2QI^v-G?#d{8sJe<5u} zDvk@NzD8En#f#Qp@t4C}=`d`Yu?MG2^pu1ft~?pBS+{$9WHSKQrVaQ$t^3ig^b`Ox z@;r*dd`57-HEGzo$;-=gZZ*JL;PqI;>(jTEIA-_nIx_io=53&j_1Nxjsv`nK*BIp~kWpl1m!;+svtI^)Rzw<^q$i>xsKEuxr?!@BwYThDRd%RS%|J zw0(*h%o3&_%~wa6ldXPX4NQ}T@f1JX)|8!sAS#&$j_wDCoqcq(vT;MV%+L(Eb>tiF zq(3 z51{F_l)l95L<`!p=VnDZlMFM%V)LkCTUj=zM4(AGq|8+6^Ttbwf+fo1YrECp7ry(S zcFUt^-sUXL`yW?lQN4+IfILR$h17KAinqZ&Y+f+68*`1(dGnJ?#Y88K{v0wM85w&_ zEQ8y8mx)mN*@F{;-g}Pe6OOulY9=yhw#vOz?W(l8%a`v{jS(_A?z*f?`oo2l1Y7JP z=`ZiWuqd@n53imoAI)JfP6{d4g>qGPhe3qI;M(L(1RQ$=+US8u=$)V^h;ml73`El8 z@pfvU9-w104_<)>8e7)4-rp$D-ZxbKNvZ>aTjjmKdMkT2@Y@~&E}m6izBBj@Q@U%h z{fQoFM)me6>)urlY_+t%4U2ix`J>U>3lq9SlvX>C<(Ff7U*Fu zlj{6AV~Y<%@DVIC09SYbtAmFGBrmr4K3a^Sg|Nj67~zrph1rZ9ED*7sk)x@1LtpvK!UpC{{2xau#3DT^Xlx5o~zEYu*9?PC93w;9iN&+Y!?vBy6UXaKW z-4+Bjl0petuUU?j0j4G1+wJgC0{F^sMm`zl9WVyQ_{Ts2X=vkB#p-P@a?G`v1{rW< zoS?q~OJrAW9A#1=1RJ&(DrqlPT;HEz;2(R&e0C5X1`w)uU&fbvr#?{qYz#{a0*M)Q zD#x|7u2Q*{xOsw)n8Z0hjvFES`s72FALPU}B2WJ$nmirE!&-znk8^LsHy_G8?^?o5 z$}5#mq2!?6IX+7W_I=ladVI5Fi2sufNfBR4624yyPeB8@Y2)>MjR@i;4*_l79o3Hw zS|#71=hQdOagh(oZOe`NY9G;G`e*mmnqic|hI$_Sr*N&4$T?|m`#$~h&jjj$wzOv9 zW;Ng*p~q*FBJ1Nu$XvB)A`bR_+^PR4ReDgk+4_z_fO$4E!v1L@>kCQmy~{5P@{qut z*3p-dFL<~I_@96NL=!t2prfO{14Dhe{xO~${VjFeWGtuzH)BIJmgSI2lSW}7v4Q>L z(!-~^yv)0D%wI=igq$9)9bLk~FZj|n$*vz+A{rou@ z#VC3H`4X$++q}lF*!8m+3VjduoqJ{kp2V)WgOkq zwO6}lH*kYXZ;$5cUm9emIK)ldtIzQq9mrbfm;8y5dxd;4Aoi0g&lLY)W+kMJx_`e+ zkYuRJs8Fge+5@?Likd}{rjV^TT(E(&qTMKR#DbGNRh=zo7nZ<9*$&1TB$cNq*KX>r z+Mhb#rTOBY?Im$o+fB#OGCg}g=SY)|?Sychj_OX84&0mVn8ipAg7uCZLZ~&CZ-zAN zNJio633O?!=EX05OvEPe@{tH9V7_?nr}y!yeeTInHtFwLmo1qv^%Qf9QrWU@Bp)*o z=3;Df@n^-Qg;Ya50STJo6A-xLR|aB@#B2X>SBXe`;SxAb7+Am7n5V5lHH|w7gzxP! zPj&|mp}4(a*nIt7e2((3eCA!V-vS(--{ngkIGG(v2HfF>@Im}s)EH-UBM2w4=DScb z9_tvmf|`sVSJi!Qw7&puvj?rKY2`VQ#7?i36?`;V8rw&o*j5+z_PlW-N9ceSj}YRu z^>Beqsr%o)^82sXwBH7;(#&2Dh)=ovzIO^wib4of0h8pAts8Ne6x7>zckZT5uLe<_fHL!=uhcqLce{} zPYyFH=!i5C?eg99-uIodGvq-vw)hqgoO=X9SJOn^Si60(?PiwGA*#vRN&3{Bk7o3B z$tV?RY#ZS-$7a-W$ zigwTl#>n7ysSQ|sOEbcL?jIDgp%T}xJ&oHIsu~MsxLNt#w;gk+y~D1=O2TQ#N$Mq5$Ix(8~| zfdCxUMv$Or?<~1%?-*Hw-?#($5%Suq9$u;_KoWa%kOxo8e%klmQ)T}>o!8ukJV+?D zKjP1e(E`SNi9-HB@B8}bv{VDRhCaMa@{_*KlIi$Ypz{3g$lhW;NF*L?HSEL~eankF z|3pKkA#~KsxWJEz|AOA1vUya*K@sDJe2D4IM&oAmTSN+T;fR{4xH+7+|fM(Lu#9jx0AY+WUd%o!(|ABv^=}$%u?tlI4NfL za9UH%qo`BzA9O|ZWTi(0C=M~+&K`M-M&6W0uF%lLKf&I-$iMNx zw=-lr{1r4aP039zVEfwo+Pbh{0906BPWnYZ@_W!1H6!5J&iQr{JcB0G!X67eO~%$jY0rg_nC5yM76-Y)&rD?6i-cjO&8P z;~=|X{?^;0KQ|OiF-9|8*eYvs{(y+ebdYq%V}RD|m)+09j6F-eR>XRG)`308oC2FK zdx}`5W9{j<_!9^Vz2|rQhc!!zVk69(D{DzORz+q(^_hmdYS~|V{l1lUb(vP8!-(65 zn>$>YTyt&WeIO5x7{>k$gZS+LKs6hrz&r=tubpGQ;w2*KJL`F1BgpXtUuC^mA;~e- z;-FiM!_OG6gXIgasUuWC$#w%;n=*`sl1jkXX@9Tnn7VFjL)$_<`Sr?W)kdwM()N)! zGIi>{xk6mqvMcDVqTSN0*$N$?xJU%*d&6&o;-at2qw8(|HAM^ zxa20`;Dq5hR|>Gm0)Q|K$5Ji|&EmoW*qv6>99+);Y~LFlSxzZh_E89^skq+fas%t5 zMjqMT|C#Iq=u>l)RJQ4~!DEudYlCggdB=S_pY`fddOlSEY;Eb(7JyOL`IB3eK;vvc zBg1bLJbS7MOk}-zgwQvPU<=A`h{A7UvB7iJfWT#kxdF@E!7E3JXn3|(8-Tj05T+d{ z8EEhhB3U$epdN(CI>&K@cV(m!-`CsVcE`Ax;S2d$4kf$-K{vy%^i%}}&R1>R)5 z?4YwKm|(7&uGeGhD5MG%XV-rxqWb#pwZu+e2PrOGdU2d}VxNCBaZ;(4H&Ik@mA{cvB~gs~lQiP5=nA!M zJ<}9A)b{U{YS;I5d*I;C_k%+ZAUp^5%x$vPO`S*tC!z=BN6!GbmzhY(1L=FHA}}h} zY&f=sn<^CxY)!;^z1%#caqofWx0#>bTGp4@;D-7OovG;`dT$2-nO$y`njc8|s8bA} zScAH0UR6Ckc2lU*w+UMT*QULQ!0Cg@ZTb7KJH0KUvU$AD@Q-DA>)5c>7fQSWwWkKg zA*0GIdQI)Dc}T1#2GM~^g&9%G=K0R9UdjUY@Sgooq|AD1xuc;nC zI0F)Hv35*E^%WH!q!(fRwa{4+$#6L|e___5Yiz_472cC1q`x{fO-p%n@<+MO*)Hn{ zieCeLS>+~?KAGyLNd|!dC?Sf%=dQ|ME2IlevP?x8XpQb#ET*mj!r`lt3Se{+0DTn6j zvf1<&LMqDVNJ2#3csN6nZ|Vl!Qb_HhXF(VM*AdbN^Jj)|*_qIns~n2$Z;U6;mJx`h z58%H6Pv73sa<{78}S3Swh6k@!KxF8%FXsMmjwSbjCq*{`jfRqVUFx zW^XwPo<8;PixjBLn)TD4(V=mqCo{RlP?J#%B#*L5qW-7Ft+Up|h$nx0 z;F9~9;vZig#o)-Dp}vOX)G2uym2ZWr!}in*1r*QD2NVN5Dx3UFs;Z4Tb8N(6JCf5> z9`pW^$p^(U4YAi%)oEzRTyTZiPd-z>f}~s5WdqIhL2f5yqtILBce*%qV6^~jgTxH= zoAv-1t?Js1H*n(h{Zo_!T86U@)G%`5BlJ5Wb;T8`tL`!=> zu;eeMT~P(o$!5Us<|Jg?TUF1r^sMy+dG~0Z0o`r!Illz1@ZYp97~DFTQRa zw3(OPwxY3j79yTU#9y|XXYIrqcWh&EVY3Pb5pMk18w?)tsTqP17zIb|#h}0iKv0T> zM2VQ7Q3?z1(-hq9Fm{$t+K3oj^1n^go{lE{NoG_RB|&UZrl>DHQ)?JDqFjythp}_) z5iRQ0Y}vML+r}x|wr!oVZQHhO^OSAdM%PVuC!O@0-1}w!f}K708f!k!p!PQ~m%0WK zWCNdl($mO7ko5*x^)*TCPwJ~1jWQ}%r_W);_+=u5x!hmu>xyJfe0*l%53AO_q!V=jxV3)FLwCvBx z7J5)HM)?;UTQ?f0g-F_u%vJ3w!yv>G8cXGGG}n>|Nsd0i$RLY@2*;>teD&|nhfdHd zJaEGX2U3}2Ci_*>xJvk^swTjJx}Tx7gQ`xme~cC&Vd_rffMDVrVX(o-1AtHUzLEp{ z3UCBH#tC{OIfC4(Sb*0KA-*=Y2u`-;9fXT|0D~|zEA;F+k26z|*fhmjQtKUo!0@#~jy3dlL+F0$DJEH<;UuH`?T$dcL=;c~Sdd*)yHA;+TJg^5#V!{9Gj zlO#exl%#l}T?vyUL|t(!=G7>yY4hUQSa4?7zHXr0 zWrzr(aAY=uyHFx_%-q?&WFG{qLte;$9t?SmwdBz9Gk}}U7NSnsJ|`?2l0xoVed*V0 zFS=s9L)1sWLw-6eYa2TA3;TXmmVag2Y6iCkIii#p^>B^;8bQT4)O0y8UadW}<#cMn#_C_x$ z{uC<}#+wmyahCVghH#s!%v1=Um8QAQD=~FhZ7F%os$(X5)-{Luc}(HJK=Z?g9TWDQDrym{}D?1%a!`k4RwCsKR$3-epvxl`f_k)39~}9rm?O z!=c6TRKvyD)t=3Xi@>~J$!LIAu}1H9v7umUpzI>KU>r6eD<>=45zcx1oLMr-Enn

>}+0Jd_>BNV#-TTv~2Zt1255$6cmk?ML>a0!CeI4NVp*HJ4Vmu~8FiGW#%T zJ5=L(AYz{gU&-qlVoT`&W>5^F%x z{*yHE8s8X}Mj1(}6d0XpSX+e66%@t!J;Jm%F5_FRFPCb;)4!dUPr&V;0gcdMM*gm_ z6RiZN1PQzgTG30(bkyxq5asx9{fc~x!pVj`f*%qD1{KptyLkU+cR!_;@r25E*T#D5i~~!Je;t+TvGE9*j2)$H2d? z?$a;A4!B*9IUhQK*|Yr_4By)OjoDp3c~B+4xdsm;c-314y}0sa-e%DhN(QGAzgaYP zfZNe}WRS2zT_k5Y*yML#VaH;C`0}hYAtFbvF9Ron7SVJ-a37qO2L6VIaJ;1ib7R=w zlC_mT-($!A(Tc8U%4J4{3tE26tlrOit+J0kXc0;AWO7?306+RuxZZ30EEoZq-07o6 zO-iR>6Ae%qjNX=X^|BxRdaCZ!kgx@LNL0hl_NQq$**yGdLvK3-CZt2Slu|V77K1$m zsl%^LwSscElybgNMHV$EY-_|S>LQjaP*cZ?78vpB7S92(?E;C%(;(9*FFL7v3FoB5 zD<);)CY~!MZA8m!opc96XXx4RLW-MCd@Sd9CztPWdZ&M2dgx1b18<;BLN{ung-+P7 zy2e~#C|O{L3*X?O_%gL3j-2y6`g0w!{Vp~`k1bOT{;reau4iFXEgC0&SeLj4Hex$M z6Bhq=8>+32a}n6qQ?_#zkrQp|CJ^oj4jY865Ium7{#;D4*N9({|10ImkcZrL6HztD zSJ&X970wF{{Pe2VMPDNW$rLBaaXlJa_391FNvEIRihBmo_=xlm2 zRHtndocW)DKh*InLd5{4juuw(!7UiFw=={nLX(hq)e`PP;~A9$Z+@{?_D7Iv0hj=L zRnUFtfc9%tMH>{wABX3EblwF4*cD4H<8G5s7g59(vEP z{metZS|M0)#75&OwxL#b`VfR`c`b&xJNy>i+d))X#yhApQQX(*l34c=mPBB+zbs)l z2NJfFtnAi^k&L!+(+(i+<(Tj}n4wc{n(c%*`koB1m zeH1hz$N!C0q~FucHQAdG3F~Zp#hPiMWZQ6=BFs;?=jOX7#0{rog8MF?YVphDt9=ar zW7Ky|xAJD1vD7g%ZkNAGHtWzK7PwM?Lxr1&n;cy#8OSh|-4uq3aL!nT?DjTa z4@2BigS3ctY!?k@TRZ z!FqV)J5vE}7v{BN2XmuqYBM%kDZs$ZY;?Ol#K{l1y=;nq@ zQ?GkQY-{^Rn|*5r(aT#;a478iJ-9^(4$J;NQ{9M~7O-W~P{t-E+SU?))~k-ThI4UG z;RrtAzm)gR^h8zJ118+kNdM=L?ID%6oXi5E}Rg&&tZf#iV%*5lCk1N;p8-+*VEoo zq#AkSDAeHHaz zDCe5uaU213^@;ouu4VPjlWB#)a*~p5#^K=J%6u76<4|C)GXnyn0WsRsXXx0atjf%P zaH|bDxV~C%UsY9WG`rb=ZZRZu9sPYUg=tOzYqd)jb>%LYs{cm|<&rBBd*sNS5$ahw zvZ=e&MuXlANo*ye`uO(v&7!gJMQhc}Yz=#-8moz;bDJ-BI@y7xI&rmN&br>}#Q)vd`D_8nSC))5ptSkZWG?Xu7y@S0WjvG#d@Rw52FD3PzlF0B_Xg=M3B zth|D7(f19;93c0vtyO4-Oivpy^j-}-|FL~Pa2sd#+vlj_!LYM+C+njH%Gd-TV+UF= zWMSTn&U!&1cXD*^TdG}hOc<$}hHZk97k{3Ky^g9r`7T(1hF`I@f2nb@F=+NG=tGw!!^IrgZLkW4w`aXu~54R;< z)k?GTBM(YrxHI|99~1BgQ$yn}>94o-W2AGEyPpU_JjYw)iD|3~CvPm4>>|YZ&wpz? zQZ!E|Er@|R`ysgI60|eV5@g(Q3PCLcCn6Dm-!s}eAt`uedYA7FC)m}AAW zk(Nbix7YgXM)`dM?}O*haN2L|q%Oe$*8ke&B{Y{fO zvGQGDdqq=cO}M`>{b08okv+e7wmWl)(D_*05V+CNDiH%6gPwC_Q4-q&c*AG~gb;=P z@_XEVe?r@W>jNL5(#M?K>ZmIH+A)rZ#JjKcOPQDfUJJZYJb_N2Y4e1n>lp#}Z$QJV zVy4Jh9js4$k^yeiv1zHNMZZ*9D8a81j*uwKy-B$^Uk45?y^fqwYnGdKih=>XhHI0# z2+W;cD8iTbHl|k6pAu@2prae*1MJ+rC)vCx%X?13?P?9Pozzvv4y!9l5t?MI>i-h7 z<}Du4r!kVY+rpaCN_-WP%Gg9iHM|t2fkabQGav+5-f+ZJ!Da z)(9SeJ5avs49?!=LaxDxPpRh=lpTwm0#FI2jRSbsJ_9#|uW}gkpNLh$Q0_D0ndo{0 zSO#AoJ0hn4iA@(?uOC~&pckl)_f6C~ArN0(i!tvL#P0t+8$bAp2~PE~unx7> zz+3ALphDb~e09MFglY?AWHkvjQ`+&AgGxj7?X<>?>}i>w{IzG^-9}}tUnCcAcx{2x z$BLRto@iRzz-W1*6`6+wfX=3bqK@5LG5$mOEngG=dmsW*^ZxS>0mlhcEmvtyS&(Ju zq7qSB0%Dh;uY3#~MeSc+T8&~TQ8hhF1?BsM`JO&;-7R{<2}+b8i2q?uShwS{7y`q# z*jNDIg3yJqz%j+=K+E)g4H+$&UtSTa-Y~^zu7F((q8NbZtSex7!loNi`JyX)Ie>TA zNI#MX!X~}Ryw$h+mZNa_;M%h09nEz(OEkUrt!MF(SC7N9O47~54# zT&G%IA55D#3>yS8`i#Qvcd4Qn=vX0wA%d{2JxG#6_9gFCVvwDZBDO4q`%#~~fG>fS zl=;udEqb-iXR>J3I<+%azk}3H25nio1>$-m1p+N?q0W#`B4cJO$e?KQ%5zHO z%mz&J^Df=Q0ez}KXYYyd3r^~C;hwyqhwGzn{Z|aMfA4yRZomWB;FkqXOISs+*fBK3 z3PaBd*AnXK;OKs?O8;?5i&rvo;K8H35i#U98d^t9tH^^+=$FMlQl#aC0_jf4;la3I z{l0?3b5j_ns~ysaU@iI%nLz@a6vf`8`pn$vJz;;v^+I-Qq`_i%5I2v;KDO0*Y#0&$ z==qlmR>^N7%+U#uA@2^t3Ik@;fR31<^eV;^{e`-~Q7&|F?eiS?6S-!B$`p`CF>_=1 z=@a8Q9*6C)T`dId+z9RoOZuT9%rrz5cANW0_W~G;0}A?{1Puwg0w1O}f)PuEj~)K; zVzAtEaXk#s$iKki&ERB4ZdWsi1E`bk4AC*jpme(KVujC|p{VreI&~F);ndD_d9Ug+ zXA@>ScM!>&U%!8ww)40GoOD1A5=K%Q0VHcko)g`Xg(t6&zG<9B7FC}up}-9_T){c# z*PXEhAx^p6Q!7Hpy^|XF&L)6MJkdeQ-I@p`P{7e@3ZfTWOb)P)TWpK73CTt}Ub6Vt zP6`KNO8LB=!n9-zwKVPJ&=c`EiWD&Ir2(5U0{`>dX5r1oaakTMT-STq+SX0y)+sVX zcSusW=yE)6!XZRcPPq=Pl~!JXhBf|etrZWc=1GO($AKt0v*-0tK(9$%v5*PRdXLw% zYe<+Ea2M?XrO3q5H&4>!ZKYlAYW?@@Cpm@nwyQArdYD2c`$E;hS%Ju)wsv7)3oq z+Ks-jm{5Ak{2&(n#ibok6?*2%F)n5^OBm+RD0mcNW1q0 zATA;_c7iS45BjD16|sBn90yhg+=R}eh<+`smN}F+pET6d{j%zHTE=(^6b~S>bNA@Z zp4r4jHIXwcY)H&amM7iInwGrfcL9C(0)8n|?6B(#BoO^Yb5b!La;JNd!*)u4a57|3 z`rDov5)}MMvxJ_nS*u%Mo@eJW5My=L?3oJr6qJ}zWDM~iVdh&caT43#)cb&sx$tpe z;D+FKvh$|w)?ymyfK1d4{@U8x8Ustu>4CLkuYhcTOP2o($dZut<4c|<%(d6u?U)LB zD8lMe8PrS9I^ARqj0i;C7m|E1xmV#JKb4O7uxA0o<<&(Gh2orb0eYi=!#LJ^yc)&^Kl~|%4F}j6Z_S(B9#h4eA-f7K)QaJ z3L8FDU1_d<^m~Lbcjo~3J;BfAkXX6y4#Ik$E?ZX{dE|N$u0>UoiE4oYg>?RTeauMw zJoD7bvjXc@EHlzT!+h?P?CTI>kAiu})brYUAJor%r4DLA4FALw8@;9WoE9LU5V<9@ zY^>bqeg8lp!9u4-wJgxsn7#!DCDLB7?bje2DKjWJkv`?xv-Kc3`p;uFpZh?<`a%*$ z_m_ejq5`H7r;p=j*DQ8i_;7b|1$%8VGj%uKrS~MNB@2=1nRybNL#$nB?~Sv95WmHY z$ecd3%KMbOQjxN+IXWr>N)L|<-?IT?nB$xCM~ILgS2aC&e89IHf*K_|9-(@|=-5(G z=Wjd>*k+NaJJAy8m*hM>*8D>P6kD&wowFk1yyp)c4L8(3!`!(ODs9NTQ&{(umO4?g zBl4P0%{4F69h%$l-D0SH_8t5>vrOvlIpPWYS&yvQt@MAa`Jn=V$o`>sD4RNQE1Y8G zs`#2eGg)sUO7yN?X!yD&4|hJ|&aTL_1aNgxd_X#5WiG71n?P?k;_%6fboRu2&F$Kq zc0e@GHJ4Vf!6v}LTA;)U36oFLx$@t9I$xt8_nP}Z@OSsRzwCW-div#?KU{sOXvGH+ zw;Fz@+H*j!uC}ah=D0SCj)JpkG@ym$+VyKxlGN zpaH#8W=U|OX;UFHYf2Kj1WiDnB5IQMn=A&j=PUPm<%yUeL^S;lac9$@;F1Cobe_;3 z*A~+iOCh)v1h}HBl#DE=oKOff?r^OWn#`gPV``&f|f}2b|995!3D2JQGGWa zj@SPvinqBd&6sVuG8P{K@*|@fbJlm4 zpU2K)t}AMw^u_#}obP4P)$6K7!^n6)=8`-v&U`&q=mH^h`Hd35K{~~w5~^Vp?-O9E zj~_T~Pf}5UHdpfBsw{V&BvRBIGYRzhZ158emKt2{hcTUSBMwbOfjX>f_ErmLfw%FE zNzp8XOA4O`jxT@OS!LZz6;lZ^9~TtnR19n=vUb!x@az4@{z)F#PP368_!IgI0=f1;`zIPhMUaLLkF8QQR)ny!iN45%z|=~rDA2yy3cu{0aa^A#?X zb0bzZMV8#u660pyc;}wH@Q4W*n;`=E@5U#rJV3Mcc6dBY3@)T^R!t!b?ONjDc-q^e zR-q28U*3_=&ER})uG|@2)S2_flOkAec2ueqpPARL9HX{Soj0F%eX14{*~R8}>sFVk z9ild^P-R|xU*HB1ESG7q+B9doM3k743Vy~>kij7qi|{E3UXnc{fDZW-#HmlpiH3!d z_P;c;nPwxV7)v!AR5~B>{j-dJkrI6G94gc^72OXq!TWIV+4>vv3e*!HFCKM5m5_Xk z)%3TtbkoFbtuJPZaFhU*a5x+ki@zs{dw(gD%Nmz0`x9a1>a+o?kAgX9<5+gIv{d69 z$QG;9;p};oAQRd`(}MlV@Nh=G4>UuJY90N(Od#*2kEg-|HCL3^c8; z@)l$U&9qE9IS4sj?^9uzEO0;>R%~t=n*N~l4*HdoW&qS}g6}A&UZdZpkne;|GE1_u zL(ODWchB^En1&YSNw3MyZ}h!vc_PMWj%=lhGbrGp`=Mx zc?6X{c3F#S@^|ToggrbMc_8y3x*QPJy_?HQCQ6Fu8`?F#YhbG=G!ycgk+aL#Fjyyq zc?x7_ygMV-MdYzA;c$rZeEI|~WU$!b{xAeCgMw-T*t#(Az>O&vNeq-za{GK;=UuQG zAf*-gNkC5=JH^Q;+{~Nlc22@DF|sf@z;skA?w%UzrwwbZJ??hw%3i7}0Od%nc2X4h zfkm1C<2ZSuFPCUNaSw&a-q*x}${?8P4eM*n2emp!bj@E3-c0GW zC`QRy=Vn1}B=ROmBS#-=*Z1WY!%){tmEKRXqf>{BK9J;vr^v07j-T^Dp(eA=V_2DF zx2~_9%N+DiTr)`KS$gsPY3S^STPqF^H}|!Jn+M=XBC^_aLvdsVy3p;C#OD@_$Yj5$ zn%7-*{V9Gby#ZWr1HDz`6q&QoSnTL!ybQ~1bBr?ZAserRjS1ew7lpH%<5?Io(?j;V zS`yol@wPO_8pU<*Q-KEsH)iK3&Tap-Icsw=AM)s7-{U2&N=tP>qZ6=z%Jk74T?3|2 zH{Y#`7xpmzSsv{8Q|O9zmDxbR4h1-P!gV4qgY00cy(VcH7Zc90A>XpI^qlsrt`zkIcxh2G1F)y? z%bPC_9XRt3t68*(?&G>n{5c3raJwiOnCWQxUQmilGa6dYU=7NZWb?>;2wa|Kd@~t2 z&;n#?B>uP2WVYKqovt24Qqv5KI{y|L`9 zrVK+CSJbk6vfskMbbilZ-D910ZhMb(lR?2#F}`?ibBqf>ESjzE?xp6B^+*5}wP9P5 zH5BfV!uO9MN}6FS$Ah>gc$fUzshYvgIy*e(cbsiUKBM%-K$nC;?Lw}6IQD?34WO*J z4f6vw!b*q?AQ?uN_pyX)Ae#Qi({|@7i+}E5KEUKRusXNz-88T9s|bu9<`Y6x`u2KC zpt)JPndv1?(=Fm8XR-B#j-T$PBL@$!{p1gkS-@zupmE{KZ5P^5;7pn{*;@t|aXbPc zgqO#Kt&!Zc4q7^&lm{`AUQ#c3l>tA2;s?DQ#8s1*>T42ll)dyro<78t{pF1p?V~;V z#dMK@v(%0e>Mn~}L82HuxZ=H`D)Xn)nfyPbc}jpwcL6Cj>)wtff&T=tslF^0m{5s) z5kJSHS5fIV&;gdGZ_L4rNmX_&TF<1P&Ieumw2OuhpoAb2@dajvE3{(}AZTP{}^>ceZYhqmd}= z_-PFKtnRS70YYzAf?hfF2H&KQ@V)?lZDH&iP*DawGh+s~@2e1*ad0!)>OPq#^|7X! zGSyM4HMEO(mlREzSEg?8h9aU;6K-!_wxBaR5jqdA-F(};e48vwQqO*vI9v~mGCr6; zk@nA_WTCP_BS+5`@lA9Vtn_ni$Uz?qCXv7|EHPH^#9EK|1;uQIN1ItB^Cd5bJ&uy zWyRV!XN$HD+c6=7mu0T4lBK-w zn7l>x)|)&Lepy_vu#8Cmve^_w7RPcCGMfR?p|@=wSb;T3N_5p6w`x zjt@70eLr>*K9S5dp`BxH6-&1azJGbO*4UW1HZ0ba1iSr@>2BM^{l?IPASVcH9Y(Wl zXq{#$f0(`?qnu{kyJqOL9q8OdYyQx>coQbDqAurw}OXhD*Vzw`0?--3*@60h; zb_o-5t*KKuJ{9^=ln=&{NOyUPdA@PNiMfJL4Vwf@$zmY<5t+hF$IRo;rgE+s&n!+i z>xSXePNsdfgo9dgW4n|RhEg~-6NC?EwCfRR4NA2J-8q=uZsRtU1G`ir&m@eknnBz$ zWt5}s1I~@+omV{#+9r>iqPE%=f>?_^np5mWHpuIk^{N0FSmVvH{q|HBq3JOh-2u45rjsxeCx%3$BhT^d;V@Y=EPR#l{{mJB@PWWI_WtKXnuoblNw zV{7h(G&vSEUn++Gv@o~&yF90uIui>Fxp9o^9_$fvWz{J5Ys~!}s-_jM`h{5|rwrvR z!*PIVs#N%$?W-$XTK_yLAv!d4WiOST*(rwAu4w%pDL|i7eU%v$H9xLZaA7jOi zmVJ$_MH@j}^*hIv#eQ`EvJK0A9gl;d(sY;C`QPI+X=^k0lv>4g^} zUw$tdz!70kPZMSEj((GzOG>00%1OXcEU!f#9Lu8YJONN&f*V&%vYbo-|H%%|s5|l3 zxx3NBgLkL+vUHTktwjVdaJ?T%>&!!K*qLRFiQT|2s3S23j}%GESbf0#@kYi&-4wT0 z*0ErH%PMtSd@*5aPIvU{*tiwiYlJB$8%99V;f%VE4L4_D6*pz)h$yPVnA8s-A?^$0>VzP8oa^Ukk2-R{4Pkf%8P zeP+S}v#ed3rtx!*-Q+CH1Wm0gQw-^?`!U_&?u+3o60b6X`!@4iDybwcTQwX-1%W!0tU@S{&G35!+_MmsGZt zubw*pf?t+dushC^HXk1x-$i=cZE28abC#iwE}r)twPW2)S>HRwyVVMp-@AFcg6u~y zLdt~Q$FJUMI`F)0xoRHi#6|1YPQ`|c=9Hg9!+9@o>JOhJ<3imM%rkbAVkYIAr3uy^ zn=Q~pog>da;nXj6(joQ1%m_MuPgcQ(ZElr!-2jy}A7}A3KRd%qmSj2+ z#_#gtbcf%_>B$FahiLX^=HR4G zbtiweGu<^+xMvc^Wpn{LAYqu)biyNe<&QS}A;NS{Zz&Lb2%ZedK%egqr>LKRsNtpf z*I!^)CS8YGw%G)P=Gu+tV+Z_*te#T1*x1aA&ZCHarR$hXN3EBKP1d;RQ8fnHdoPy< zc6A9rWTdl-CpyMIU^7ve0L3|$^!xp5jOcv`c=kx__6ey;OA2w4obFvzH|w6vzVje9 zxj@{lnSvbxlDu-9VU5J?Ny!5q@~@|PG#NJCXBmJZT@{;f4tHigLJ9_X-shP}gl4-hQttJ;oSD0M2rZaj+;CFN@l zl+NM~Hn5GL1`3N0q0i=?!Wo61L=gkTMqA=B>Vb^P4Fqn{!Po5)e8%DGBk;}fI3HIE zK2Z9}*6w06n z_#|-|k79xCsSW438VrD{BH*1L>7^*c`Lk9kbW$tpo56mB3JR~z3}vu;>BvS7wi^K( zH`XhIr$awkMBKdew*XxTsEc+(%-4jHiJ8zt`5O84F*0-!1y6h$I1E035aG+?R>Nb2R}?tYqdgM$6tAvKCNz~9yE3+B?!N=K!UXNK?0 ztp@Gt$Tu*^hPH@KBbqog2#FaIQRB_UWwxcz$Wu-@>?%G$u?%gyby(X0Fg8ZRNTC zp?>WUHU+skP&8lp5?Y223@x72&`xyqJZjX3`{|bG8^54P%Y~6pCTBFg+7Av#R9mX> z@|u;t@j?V8^XXsvie9|Px@CtrIWm96Js7h*s^P!u3?@y?=}YftFQry(X%TvR|_NH~oA)ch2i+t6V1A5wmEY`<>=e)CfGO_xuBtMO>C&ph1OM~`}{ zcHvp1C*wxziH&c@BDv-mxqLnJAA)a+l*xu42m+)^#{9tC4~rO!5V0&1y+X`HNBM}i z@!#_;(Rqj0{qf(CPCOA4nU8#36;=}uO!D`!fQ$?_cJ)k$Jf?J`+h&(f*!ZQ=ge{}8 z3r7{Y_lak3`2hnJGH~Zz9P-g2FcM)3RX+NQMts^9co%tDm7%yl1 zQ}&x}hDjwuQALgSXD&X8aj4rGSSL0_nw=6{2vg{5s#anHXo?-=meaHs(1FngikNnY zIn|%$GNE1`@o${pA624SL(e=%5>U0xTFHUM_t;)^Gy0sItZ60QYIDe&zyW@k8ta(= zJ_h04tW|SkU5jY0Aw8}$PJQhLr}Xh#P*E>Vcd#Xo0b(p(UAxosVSb^=hB3$+C~bUeGXcAmYQYtE!B zq%#}}?6((B(>~Z(`4P~Fw_0>;AaD^bG#OXJQLMR`$~0|JB>|>ZM6JSXx(_v+cGeFI zfR&+f86mXAh#~I5Rgl!L9eI}dZJ9sz5>c|rFu#h{)8~Wm(z13_T*mnlB@4#TBf+>^ zva*`SX%rnL-jxsQw!rV;jCbN3HWXW68SGpK9@`H^2t-#m^!$G3>Gm=E;Y^0A{x>>E zeoYTKQ-spoi=~*iu6AsaUteu94n?JGl@cg%>3sjfLaB4QODEB>Ka-FpQuaIGnd=iL z(N;Rj#&wMR_ED&Kl}_)T#gGT+zb!Pucl}~p&bWC#CUE8K_*0@+OJdKh12 zeo`+?cUMo(j@o;Bg92(eAeS7gbR+NM1?PDcl{xUs$O)&XLO10jp4x$1!Z*qt@UfeZ z^7{}kGkz6~iT8xBB>Ff)d&8Hw$cD!orF43R+<`7VL)`_Y!LibhL`qDcx}>LaXrUtZFzq(47BBm zqZ|jY5G>ygj<%eO8HznyGW&||8haq#bj8<$`G1=}dj3MttQ~HETFj!7MaP-NokY2Q zmBAG=iNN3~a`+j3|50aKC&oD$jCJW8#|TMX1GzjRxx^j#-TxG9K`3zud?Pp1JVIQE zX{jN^^(uvcUHZO~Z z{+^O0yf{b=lrUFIv(yF=no=k7+Z_Qd9gZ+2lD}{jX)p*oI6@Sy*Nba9gdB@cU=-=Ri_s&g4UE`P+k4kIwX`$-GK&dZ?Hs9qb*J=1qFIZb? zt4Mx(_nAd7Jp7U#ss20#NU($8l;e?w9zx{R+@OspriOdXT1FSHI^98&zeP2kWKm>^ zF+L-q!+HuW`ETH^j-}R@gO}5%g}fu~?%_EU-ENzgKRp-pXQJBwGIP^+yj=ZjDN*fK zA~zOr^hD16h4LV*2G37tg)R%Uaq8|x9*qcigA{M_9cz_9u)ajN&VQij*B)IIOsfBGc} zuKODPaFs^78@4hU$D^p~qcP*bMk+Mh+LUwDd?HrI)u=I!p8yf&L{6fceu5PBKzS%GKHxTUEe*WIKS^J^t#vWhKxY^N7*LvifjwTVvsK@=!?S@3h z=SM#m+Qd*6NGI)>8we~EtcJl|@|!DXV(1Mj{8LxvwGH<7Y85R4r+$}-&QY2WRvHEz z=bvU=x#CBk={%K3SZ_nVHSVQeDc{-&cZgKM%bm#|QW0litE^FJ#QH$`?Uc-&GZs32 zk}YT4iNBv5$g^&{EiGj)c1o;qpFuHQM1#B5V={!s#5g*vAtzRL|yz~?= zk!IY_?YV?$0JTRaycOUvcFsJ^Ejwaxu3qN&oXYElbof`6$W;>vBq6|o-3kf@F&>ALC( z7+>5IwO&8v9!y6<+41?tC9y01QfPq>1Oej{iQSwZQ<@on*Or-wC5Qfu28qBOR;{IW zw8X~7)*>*vrH&Wq;;R9|h*xGc;*a?w z^dT0Yxb?b&!{YJvjm?kuiM_}-IsKnV8uNc3X>1%E|NRbr_y4WLX!5$Xf#*Uamxg$% znKuz?S;AYR-LRo@^&H4q^`lT1$s}1Mplp5Lx&BETJTAP=gBjmW--ZnxB#@u2nxgNW zqWSSmGrMOPWjHg=#H?;QAh;Bxn7^#<+i2wMH3A*3o)Xl+#AjbjN28>qQw0ibe5KmAMey;(e z5I#A+L$AAqJ!7V;nT!<)jWc7aT1fz3$~EbPN$bsN)7u9IcEcZ*ORiMi?5?*hUBaS( z@KmzCJdOLw*8b~PR`r)1K z>|OMoup9F_*(U>k8!=M;s}X~_&~e!V%}|8X7XjaR>}rQx2|Dn2J`$f?2@;+v!m^Qq zObH$35`9~(&hkbmh5D|et9toWVm?FHbpEYLqxaz|SKy!_B*GzFfE9iZ z(sr(?qunO`-$|rd4Gh)~3_=pzhMZhL@)#bcPMeK}lN#2o?4x30y;UhPs_ZLb?1{!4 z!!m2uG3K^?AOP3^X}qYN$$R1*lo;9mds&EREuE`1DS=OXKm~;(w7^}H2lw%A;k6OR;t=IF$?(=oTrv3LRRyoC20qU4$VvBW3`P6gBRG9&l`` z)8tSNvr_dGlB&=^)TPhx*Feg;Vu7mUGiR7fJK+mU71s7BAi@}b86$kRO)6OH%{nbT zhW|)e>SH>dfy1&lU^ho6bz(5MmKgcNkmbN?c z(wC;f#Ejf`sNoh}r}$Z=>dyhR;JLm?aqg(LvUCL<(5!WtgL@%SDrM@6c5f20p`AuX z$AXv)-*?znTp2cF2IE44BN$U74n!y8Uej0svh0CgHXbyP%(?*AFS4BP*P$hO<~EGn_s(G4>-Hbk9LEe?w#Qp=)!R zof~Ygf&@A1-chEfmhCz#wnDQPh>N#wt*xsB=-Y+0uD`-OLw}mF(Tkheqn`?o$=db5 zFp`bTeJr!(kbcazTjW9#!z6=qZeY!>o{_h5*fIbNOh^c8>qvR&lUD1|u>GxTN64xq zK_zg8rl6{`j|b6u&kQnv39b)Ce;;-!d1lYZh7zU81iR}~{Ep%>uhv8G|HarlHc0}m z(UxV~c2$?UY}>YN+qP}nwr$(C?JnFpCnjP(%!!!$A2KrYUC-Kk+ni@*Cx8&D_9F37 z%GzLi&9hZpNJ-?ap6AWTK`bOPx_4SGz?1Y$cLg>IJ5SOG7IX&K0j)hAH+(W?Nk;{x zpW+g5sN~lI&`%%WW50E6(O5V!?Hz8WV(D{ogWuaWS~xdS|6~grNtqP3hZ6jmZuy9H z7luKEMSgK;0aKQ60fOS9^4(63a+Ls>iVgg07O3)q=ySn%?R7R8HF!`JE$teSSft9YI^t-Ozp)5h6-VEz32v8nB`xZ``p4Ds5o(8VH}yh zF4+XZDB>YnE(`q)&4hKenrBGM0jxWrsDk{B8?86=Dvss zjgd3TB}LfwlGW9L$h(csQ4;IQdkj6rp$dm@ABe<`>5L*7DpJCaj8ja@7}LaTEX`m60@<;}v+=~T97!xbTk-+aQt(OQU-X4}AL(Fru<_Cm7DDL< zta}-dV3dha>^ZDd$i8w_(kNy9{ah9D0JhpK5O)yCa`svd9UbLiO=QoMZoXL^yw&BZqaKb+qocM~OijM#Dq&fV3k%UQER=4O-GbdGkT z4UsVdPu|R)t4nEye9&}uJ$@HQb%J^J=EnVt6FsBvAY`6}GvK}cf7Tv(4D%p|7B&gL zSilf-*I1F*nhIBkA7avr1;E-XkkflCGsW%;x{BN;k!^BW_*UPs$DV~pbmg>liEK%6 z0YpYnRss{dfG)n7&b2fUCt-ts|CzV&1waTQ@DVYs$lmFinO#-njpQj&qhQfcZZe zAXuwZGBMTU5_6A&-NguE6Wk4_2#g6`@g@pse7)cA9mwMQl_sfLLxH`08a{0ur$gaR zTyg42AcHasj4FPg@!+;zA*IDw%5(zNM;=Omx)>aYQJ%L1)`~8jmYVG%XZUL)`C>T^PTRmR#7(3eDFNn4D@R z^uA4Fza9maxi%KaY3Yey?fOH&by6iwYG%*k`BGy1so@7G^USpRj*`dW0To)UXs|)r zENu(ti_xzDpa91QB+b~g-k+l1{}N-To^=&T03AJ>RQv>#8qov2N?|um_%#JsapYU> zOGp@IPFT5ZryQPqzGX6zR*PFc6&jQDB_8wc;ZigsPZNzp!COia?|4Nv7SCY-XHUyb zcOy+-J7<0*%jx1$CZt1@VR(U0oUphFN#<`>bWHOuTLWG9^8`}o(AVC_C9v#` z>A|%1H*12H)RLfl#EaAXeA@3kuh%kK6^Ka9U6vKIgr#Z7Fi2I>&LN(Hby3#Mx#%B%I{fJR z>W!rf;`!JU-O5E%Q{W!d{2(DfQv(2bXNK$qXFGpstuAzP5+{F6HgRoH)~A-WH>V?=pcFyz7CY>IK-6{hItL4zM&!TqX;u2D zcCpdE`-FWe<$MIxGON6dnV}M&A8|fB`zFN{25vxzlLE z=u}?kA+0~~&k&aHM_}x?*bNG~z?s8`Mm_-cru#Xxm_Yg(@1y97GF zJ2g=p%eA*;{R|5*`dBgqmyzXz5Ml-(OK5&8ZwR(`M6#V~$UUTfc)k@x_wt zPRgcX>rY>VI*c$kVDDirMO^Y^r|HeaeX<=Ij=CaM3W+$W_1L1ed8#e7-P35xrPi8; zVA>=s4$3a`dme&>;KJDcj^t@Yn$7$oW1JO93;g7$?zcTt+y@IY=x11Qm?>HBxNV#u zbA0FleHI!Bw@6kxF@A2F&EI5UffHuW5b1F=iVH4uzQ0Hi3R!eBSrM{~b1F8l(-m{0ctzs36%OZz94`qE{NB23U)%))wM6k(@s_9sV_SfDSeh}$6cFtj^v@xQ<7 zKPA^`33fTNubk1!|C}z5IWbb1Jkst}u97H@rB~q!(qQi#0V-j4yn$-jDv`#+`M4u{ zG%1nquJaO^wUZSntW%yrU{7~h_hE4cx+qKjOuqa6=)}DiAp3F+QaM5ec?9-T`4BM1 z0iE(!qdwk55U{^~j}Yi^OQ?$-1t@umwF}ZjBD|~a;jbRIWW4ISCwf)A_Wyat41h1C zaGV;cf7~alhE7}(AjY65c)A<4; zpt#~Vogzs_=ux?FO<6Y^`Zi_Dja(Da^U=PGWm-6O?7gL&f+V7qzLfR;hUM`-IXD(+ zUNx61lg|Er)GPY)p-R`xOnN-J&RJstsRi#s!oYXo(7)16WQ^7uN-i zn@ta`=B22LL0`83*Ka?B6dpw1T+2=oXqGAOs4Vyo@GAx&PK4ZjyX3{?23Hv|Rd0^%J8^ z;T=&_h>1AP^P9t9h$V2S+MUkZ9o0M8M6Sd{oS)&|$7gw-cJ_~Rv4w5%PextY935QT z-en7~!P)jpr#izDH(PY%d#O`{#Kk1|m3lkCV7$qL0Ja+LaVU7kr1>*5ImW)|94F9g6VoqgJd7LaSfzQiS4^0b7+D@|f+@N)edCZz5N*ys4 zdNIY*x^*r4}<^PMa9tq0)Ey~So;@uFAMu-Sqdzh zUA6BaAK!|s79?=dmzhmpDhj18p_N_z7D^zXrP-=Z)Q?aiOF*cEht4k`TvQT{yn{SM zo&!&ba#fu$(x}=tvwUQ_Uc@WloSO!4$16hTFd7whrwCf!JBuTN!|l^T^reW5qGjH* zg$wutnZNZbM>|y$g4ID3(jKPFrR{dJ!s@h(nmDDN69n;qiswpDR6i{f zz_P^eD0+w|B(WaUH@n(#ya2C&HqWEF`0xYetf;l!hVJs`|ti{Sv}!F)*g~Spqk*nE~Bf zHei~J56Cph2`(I6?HpsuT!S#a6j&l|&ycr@W?U?lGcC?^^fvgemRf||`5_v`;`lc@9JX{!5_%YL{@eY$Hx3-VD>Ot{19BZc1_l1B3e*2L z!JZ-7wCLQcz}r#+DdO^mijSInSp!8j1`k=%^FUICjUly`^r2?T5)pxQDtv8L8;WIQ*3 z|5UK-q&|DJ4tNE!B?A`_;vCKZ&prPOZ{;F|2(i>f;OWNc!)u3dF=J(xSA(ts?jCiW zWEu#6lvZuVegF+{5d$clJ`o(IQNi#VpUz4xfRZmZ-)p9wmC<3?0fC$t8UyZK_0p!QR_0LuB&pYk9X^2$BXeHlXdOMbFVBNFL=(MH1`sP+C8W-5I(6n#i+c=I6}}1j-0jk4TBkyFevD0gRZCIC!B!U)}2#I!kw^rYsk2!|ZC!Yaey6MDyW++M-W6u}v z0)XG7o-6qz`Kx0*%g{4$o^xx{>7a$?@1)?V?!*l1kJI`Wvj}?42*g~WUPd^Rk?B%}{5M%b!I%cAmHr%qp&sQ~rhAFaA21%r@1?(o&xA_B9iH>0#cy);CL z3%E;5`cCozS`=;wWC4sCkATRq=-za8mv3BO^YhWe>B&@$JO`x2fhM$gPK9xZJS`!Y z<5OgB@rW9QDy1EaL_AJTs?rNxu5&+`V{-knU(&bV*&Q%p5m;FALTXK10Aa= z0`ctzts#{rv#)+&CBcJrD5bM_TT643tJ2^^>T*En@&$!9Kc`qUJjmY_Cz=N}%ITFW z4Olvw^kAy$Q!n2A^N#!?1hFpczKM(n#5o{56qqPYo6|udk1}+vI0vf5xAy73@#KUs9%QQi#>}3dPVGCA){4d3sfA+rZ{9i;yCRNukun{V5yC0KP zBIx3J1p8Mta@locd8WolA*FB&Uflt(a0PYr&=wzMq~}3t|Fu$caiQ;mZ~h;p2fxA- ze<%oD2c-&Dto?Fi#*58|kThc%45BF>a_#RiHtfy=GWY@GP7q6%X|iUp`>Q8=FTc-( zn_)2AOHw5$GTDoAOBrKs`*pfWB_J`Y(J0u3b6nB(AEvb!#$#q7l~#4}g-`RUZSIy=APDEWBKvst&*za#czknz+M>D^ zK=#0gy9FK#G6qBP4()A*7tzpc_1z>8P*4$n+6_0vqCt7=hO3y!2+Z_siCrSl^%&}J z67iO4R+xeuUjtCI_q5wqYr~o0`eN%1eDeUTLfk)JsUKhyI3ub3PXAF_{^PUys8a3@_K

(ifIqce|imF)^8e@ph5q$>! zH<$9m5c^34?gOZ>r*}zvi5=i}`Mz}AaKSM{oU$_+r%|xm2T!|dym?QGT0`K47l+Gt zVqB3QmAd{=b@zQIG>*hA9?V=FQYsPD8XV#(xSfQPpXIug2c#$Z6eu8QgzwVov~wCZ zzaZ$xUF;fPRg>HLY|s@4K;t9l-3ZpIe8vw|_t8zK7UV-x4m?N>f%4ipPW&4eK|_WO z(CwIicyvq>S~LOzS5jlIJFrC%g(yTD@VuLIIsuumt*X-pNA%2?jx(biow_&HW{Dl2 zJx9{0ElZwFswvUZfiuvQU)cAl&nv;fJCZ}iBp2m{Q9G@%Vl9*xC%vQm1>|C(@_* z*?JtHf-p*VmoS`}SkEWw#7cn+_9>NGQbAGwQ*iRR;MiFmRhnbhNT~)EUD$ zfDY3QxVRGoZs3ubLu(LN|A#7zAM6#cZY(+@x_oySOtr_yd-GCCq!yPblrZKvN;SM- z!kz2BxMh%!Aw}PL;t<16r2dS9_lc24pMw3nh$yTdzUpM>!JwQ5%{NM_(kLD6za1z7 z1Pt!9!^~*|%S~0d68UjTna-ADXJ9{=>3JjaJ(~$0RZ(d&XdbzRe1*RP14Xpx=EQyJ zxbZC&n`QfXBm)GM?oSkuBy`-)nl+sqr0N6qNTtIPmTcV}Tn#UT9P##d*6_-|OY!rx ztG^EzOKOi!YN4{TCc`lqt!>Xi@u#U0@6Yw<<*&%(YvpIOZx(FYg7CV{Yj7`lkpAYZ zDVz=m*i)vEh$y?cJ-ck7Z0*?sJl?CZz0}FP%fGk!wZ1#|8nQLwyYDGR0eymTX74rC zN9`4#0YIKQ-LW+yGBv*aGXHQ0$K>gvG_Pp z^MS)=n9{uC_d3$r2D}6`VZScx<{C|rgjLs2e|n1erc~0*y9H!2f2C2RxG#{ElYIb> ze|e!WN;P)8DpewnP^1b7J;b^nPJI+hqjN-Q?`Y+(O0C+s$t_O4>CkNokBThtA)WA- z`w3m0)i2g%un?>k<~-hh9loZf`wDY7l?WfYqpIqkBVJe3$4(kwJff_wnAg95A;$Bz zS_gz!wQo~60beECs|a%YcA_Ps8+lKiz6zL}Fv%R!VPf|=u2H9~Pff3j_Z#iAN$hN6 zN}s|~3ug%vhyq~RTnYUcH$*$Xkt2$h$DD&CTt<7CA6dneW#b?EwS9J2Zt)0JTD(8O zW0CCC0g8Jgj)chS6jW%?*q|d=u$TH06e9x~p4l>Fi3*ruz(&FydrR+CTT5!di#hOr zX|cwhj4u$f+5nWdt~T36B9z!7V*xkp=Z6+62;P*U@ns24-VMuE9W#;u$b1Y33cca! z&JlHiF{opoN6)Fx%gcVR!L~o+_@EsJ9rzz6)NTTW?WMxA-;hhR_K=cSlVX&z5sXV^ zk8kq%q(SZxphs@uK}c+3t0~WUC9Tnik#EvaFCa9`WnPa@30$2)RXFZZgsofk3w#}j z_*QS3oB@;`)E{Ehin;Te&!C~HqN+bHaDs>M?ZdLZM*u$AvU$fast|_ei+zHujxn^M z3_!teg1gt^;4h0FY1)qns3`H~VQN!h_lmG%VNdOG%{g6-KQ}0g#)2AS497hlo>T6x z7(fCu4j&gw$Npg+6Jt>q_K>>bRVdY9O$rvy-{m zrW-FW-m)}mHYO`Ti@jj#KU)vENylGzbJEKK4SgFpZ6I!v(CdKLuMY-oB<8uJeB*^f3ggzuHR6fi1L~(Ip}iuH;6C@Vs&fg7G{nxh)WY-`?^H z@VkZ)pCr_iMgA+ZEBN%8ID*RUq^Ko)<`QMV?es|86SI{AHB|)-t_gwAPsXpf#kuR5 zDMobQBU^hW@L&qg_P1bYyvFkEu1c%=MghH-w{q8dZT0Qw>2vy$wQ(^raedvmE8G+* z^Q&#l6pC_+OiH?N)a>swC{r5B5^*mSc(Tu#uhy6Wk^d-%RSYv9iB^KcFCe&ZUS6v7bo;MEHrT1mTQBUc*yN^%e37^T-9f?o?TNmQ3o)4d?~oJoTQ%um z9qpyEv!bcEdJ!JP!V%@;Jc%{CR)K<4Gi$oTz8+EwEaR?;LSE~8Zsg&c6# zfHt3~+#yCYB5XIx=yTUd%K=`l{jvhoIXd&>`OnOR&X3M-N1QAfFHgbjmZ$8RZsn4)~qjicG^LI76swZwDQB5<| z=VOY8*Nvb=$@$BnmKMp2mkikD%0nUX>hH1i^eBfLZ6Y49tew89wo~!1B#-GE^~fOk zc529q`eedA4CHrKBGy5islXsLESbNP@oFoNVJGH5t;jE8Ke3Z*+J#}kqXavBm1!nR zGWMZmN2fDOY7dSSxp(u{p#2cF_M5a9Qxb4jBRi!BPl2^@s{oYE-w{bfD&SAFXiU4z zP(joISfF*fIM{i75?9Jm`r?udO{ue|q-ZqN;{Z^vPilT27z$OJtOAsswET&0pcni5 zLj7k*M~+@Gw(>$qN9r}J=OlZDZ~32FE{MMRi7pCDmAN^^Jb4kkpT(K$2O59Q*GMzZ zm4KRrkNLTHBsoIc$a(=+VF-Uu%ZA4TJ;2ujRg8qA=bC$TDJwB@t1djZD2;fOWZ}eg zKv)YH#5I#HL2-JsZc0GYSpCyRq2}fdAm)YSpq&#%7c*B5NW{3{woGcL0crsm0ai*f zjNQf{e2bc>kqs28bcxA`%3>&u%H1ZI^DHak#JFvOcLADM?mAxz1z3x4Lz7SlcF`75 zdqD!g|9G`)G7Z?NR%z{^nvj%|mPp}J6lwr!(AH3T!9;%V^9L4nX%;%nTflO^n>Ybh zEUT~O@Hour1krF`lbz?4=lAdXPe%uv1b@GI3%+s*GbSJ_dP2-J7U-M8N=W*l;DoP9 z)MV?Wh%3;z^9ITH%DIeo;O=Su@JhIW8vp=-3Q7RNIPRX(NZRh5wam*1YuLoPu>TF7 zkKqyd!MJe{e*)j9fecq+j6)=J9%)h?0gh_|cM~a+k)C;>`FZE28Cmq2G@gF&ig$tT zn{#EL^YYU%@BCo_!~vB!vg5wgRI2-^Id;jTo_t}&B;tqQwX|2-K|_uvvnhI0TFyj8 zq>H3gGcC(UHqAr3RQGqnWF8zOtl=~Pdk~-)D4!%mxfk7CVDUM7oA-Fa2@I)zx0VDm zO#!|YrrqcF^%Tu_!5NOlZV}{?-f}u}6xUikO4+kcO%>rW*ei}9KL;o+OV#HpmKHj( z!&Me;kzIQRsj&4xkS+M1bFEs^?`bE9*n1nwMAd!p)qlh7q+Gwf9qc!)w?S)JmwT#=~p&-gsJ^hGEN~W}ldA z45Uxf!QNyi3aE$E{PHv18)azq#FUPQ1ngpPPoqX4dJ=%<3>Nz=`e;!f-??)lNeEdp zNs&-59f%0R5$$N{@Zg&6s2XOW0^1FylfpeP>{my1p0`bMw9sBfW0n-FFY=|flgJ>f z|B&8_JNnt<<>LG(+L$XgCFg07_ewAVQWg?ffJJNy*MEf5{|T`X$Q{H?nNEu-P-How zXtYab?0EG?e;zf#P}v*LgBTeZsdM8L)8k9f1Av18Aku56bvy(E44McJ6np}#VKe6< z<=?ArNK$AR!K>=Yl1gwP83!CmQvVB|M_a-DhrQ9}Np0O(zMMSph#;dp0HgkodDpNx+T`V$hNt8GKl7p?A`c zA3(q(J`-Y)tM0UKpGNY-^$1_p4BJBaq&UGIP-HeKPvP)&LjiU4IQOwGo3R+Z1VGO+ z34alPD~SF34g0BJZlPn^0HiZZ3R1B)7p`1Yu9-rVqd$-Gbgg%tw~BW4#LB7}$7##% zyv3s=doCmPVSVhxyL7X9Q(z!S?_O5RdJ?Xpt_0^ezn`LgP+_18in5_!*u@9Zm`8ep zdYSJjGu?E7HNaNh^lI^KsJ2j2vHHVkM7BxkmjXlevknI43nu@?wWmr-iQ7b=`QF?} zrN0L@hkf3MbKs1Eum-a^nLLc-Sgs5ab{pJ3*Tc>T;jWO1W=c7myX?1Zxv4mJZSYdZh)dVH0gLpQqVu+8YB!=3}=I2~!SM z-FwTro0p�eL8u_*OXL*?M(?CG5K-|47A)K)JnxrewY&1Jf`R^9&=Z3?i`B$VZzg zIjy`Z*8un1f;XRMyIvnNIrKB)5nX0|@Qv%LNKRYm8$Y`As3A}=4-jwgelm-3I`~{N z!Opx{Jq|1M9s7}DW>CB__$}yWv8LFM0FC>qwqrcpkGW=NBrl;&ktFwJ`dHj8R(8y3 zw7W~ODkJc>=&LnCR>16nq@^ca@kXFJ8oB!e3!N6jG*QE&qq&`$i%P)o%+y2u5)}@{ z8;all_9( zXX7(Sgi#un)qg4>Ez(KA-8IrlFue5qF?exQx^sv!W+BbN9@hF&)Y{hBDt}}XKTd5Cl5+b zP*wSw!nI7K?)eCz^hz`CHZKDHfpIId`n2yx+7g**<@_c9%WMcIe`18g&9`<^%;}lR zxgIgT;VgHJr=C`;E|tFxOEKX|6c#>{%K~$_E^cm*Kr2Dxoju zTB+bP2_mTN+P@$y79m%@?I{r9q{N91EZtsAM-0wz&3ypcA$wd-X*Op58_~56*X|3Mv>I-Lay=8qENd&(I&F9%U`aE5c&GaIKl2Sr zNp@a(45Uf(eX_yO6X*ye-(skXrRx|OS%Iuj@NY$BoKUn64?ZzRw(KT`1KiO{xX|cz z!2mRp&ydN|-H2wJ1ccU;_cQR3 z3>R}1)8l-vPWn8hF5^!qw_h=+Nn3zfxG#-7DwP;hF3I@F$C$R(NmNJsF6;}{o#VmW z+?*({AJ=ZUO@>Ilw<(t5hJ2p*;y&;$$o>ordEw)Ty{0oQl1+nXEha#nYGz=9XShRv zwfhsxXFT$lD!`J5)j&#kdQktUJ01P;tKvWO@Uh=4|6I%xU*ut^gS z63&H#v0+d+5CO$hfM|66?W5a_@pCSfrtih)D)pqjUYK5j<)FfXZjF^6I7~pvJ5rTU zAl5MVQ+hrPER|?thKLB`|JZXr$rMhPffnFQ4_p+B6|3fn$?i;5Pv!80C#0?pR4p2= zS|@2K!P(K54h%BRLz0z?@}}B&ceMHHx zH#eE-AdJ0o;H<8l`dF!?PhVHDEd2-izCwtt7zXg-pSNnDg{91jG5`5?fg(ZGJ15B* z2?0OEJ|AeN``wv#fJ_OD_7h?8DbbwIHrc+P8h=}qY?t{hpy;9@+{Ll4Xow7GAJLys zONk-{gZz=<(W-?bVm46yf`dJ4kkUKiz@(w3dl~_GU6LYSNq0giEN2Mj8z$>1Dc$o? z#P~*OdM85g^WwMqpu2rV)u4W7a$~=(KqMFET)a3QsRXn+T>5_~B>E76m9;C^wtr)?q@w*QtiF*;M#tB(}(j~Ri2se_X*NY=_^c(P!_0*M2JZ zx!YP|Y(6SXDP8zvTLqU}v*YUhyVg?B8Bz#ctKw$nYz0(#PZ8^2il4|4Q}Z14(;)MY zDBbdu2w6iaeR){_>AbvJ-I>G@q7iIrv^ywKjke~!S(j%uc0@=XF9E}O%TeMBY4bLx zC8ITyruP>UG8U26)%X4iasksoQ1}SlVJ_HHMfJ1(X6BI84n)B%h)*-$G-(0WK+V`w zWFt2~0BYk?A-I=cEu4U*=%Tz`eHERAU|=#xd9D3k zn1BN_wI(2UHTerRdkO_iM>0}QL9W__Sh>YGDI;@!UqU4RnZ(j;n((y~OV^p;4Rgz}knu9)q-W?-OQP>jh;3w4Ja zK=u8I+AXslI{W!0+}o^eG)ot(jN0D<$HCaW6N?wsj4qTRASdMPS`MwGFIb8fXg+__ zRxT(?1ip|-vBAzs7WLgP!vzDZU$+nH){o1M z5)h6UG-FW(qEk4YCN%IOSZMRnxpVc{M0x+l-g?53?Bn2yh+A)fA+`_;L}&2S27|^= z^@oXYgZiIb|6Mf0WwnH?SE+drH396zmu)nPR|j7j5vWG>iCDj7&(I<$R97;Kj998- z55gZ1guRy*HfKn@j#;nyq!8ON$>iI~9l?qk8^GSAipgWdfGR%G1gR;u33<k9Y`HJWH(oiFLVS5{^+aI zx5u3yfb44SrT+!MSpL5NjG2z@|8i3Nw~0(A_C-V5LZXvjw%z;y)`*dGGXlTfaCF66 z+JQ>A*(}`NRZ-t{wEo9hrM00gviu+T2^RoMr|`JU)o~KKO#%S~PS&IMF1w!(2PL$- z(n0UjkCnX~o;J>>6T$h7{D*A|TZgNM-D|OmXnmMWd*R|L_-6{%qRvC0YZJAA9o`NnH!s5Le+gJk-AZbPrq`d0@CV`F6lLT1u59j= z^C|Ry2^fWoDfBM^6IIksDvnCY$vXN&=imQzJEZ&H6s5YZbjedb!G?09O{c4!wFI~e z;;5=NW2uWUMPS@<_H{62a=v|b7M_=9UVdOUsiR zYtz&SQaM2KA5ZPq0wvyf=e*B`5&80FXr42t=w`b~Ni>H%teXLqm)9>XxzjQpYUpN8hJQ6B?;zJCx8)uk*}1Xe#e=_J z2|5uKm5SJD3#y4?J^I3`e=W>F$X@AUZ(v&>*&OX_HI`4G%++Mgst@Ot0tl^+WLxxT zYGGV}R}OOLM!#Ixe}6(UEpl#ziNBkydeWpwy|7?myGziU-N00aVn+#v-M4!WhA&$t ztNTQFADM_??0eK%uNwqdCv|;W(g(X6$CvV#4G?%&8l4k&`r!=}sd99yGld^}23~uM z!sng5*f_jB>^^jN!_kev!x-pPoOh~Za%6M4dEIlSd7r4Y@bQhHasvwGKlq_l*$TaX zXRI=94D7i`r}ixjNWI3WFtEM{))NPV^ZFn3)UE~ODTQYmg%mAL$F(Il)SRW5aePN?iYb`tqmFB7nH|q=BM%fE<6C%2&4#22l7O)R(6>mXLNlm#M#DVT}6JM4Fx>G&}@opmcGW zxjhzYd^UBur3#DvnVfXJmbI)dN-9ML)dt;|gj5(9L1vM>T6uI=s4!aU*{;NNrCMNw z{oqQvPY*O@?YAiQOq(9$7bv?b^c9Yle$%hv@MA+b+HQc8Bsa{BHR`Ll0cgyAVP78V z+SL`9p>gyF(mtvY=-Qqvwdj@frSf$n#joxhpi=!p9|wj`7Pya;dkx77eMozJFfc-Z z7)#rpHo<|WjF)_D_eJV5y)_$4XgNF;8V;>Udm!rIXynnR7(t6%dALU&{rw6Ypgw~4 zdPQl^@nQ4~5L1sP?tqs1++c{qxz7=CFGr^@-;d~S(s&Xwx!4}|X}(y5#~%qgouFV$ zQf#i$)WAkZU{D?ETGlH;5St~_4$waIhWJffmMN%e z4gByxXG9o2Gv8IfcJru&Nwmr+5Jj4N;9|p~JJ=XX9r%9Gn}kKt{83AwLAwBkFf}%q zL$k=iG64NxjT!{@;!eb*S&C3KC>;L%!x-vLX(cK0CpNhc5)T;yepReit7SMF2_{h7 zzKb=MFziljck*;dUn6BER0X9EiV#)mf|3DYXE08TES!c=RNKw#>h{XobES4@b$Dbr z0X_Qwt9f!s>&~mHc?Bux&Z+bmJwIhOet%~negJ*|*E=Jldit>7FKqyKK%k#(kf=p} z5c)ms$gV+i6;hI_m|w~|)dwXw3#w7lIQ=f1Mo^(v{F|2DHvuS68mPav&jwMl;LUMQ zYVKJpaO`MIQFI6a)i9D=$h1@H0FL2Z=!P_tp34pdH#rEh`^(ab`v)54=IL?E%UYd` zz5#Ii6$4uB`OmDyd!CP4Z(4;FPfrdI-<+DyvDxk`GH5G4G1|&>{%%aWFSgDBYoFsB z9;&FG=s?il(}P&mQnd<#2Zw`!rq)k{XyZ#pEbfr8@z42%$lQE-jpIb#X0<~5!Nqwm zGA=}Mm(rwp63lr7o5X96qF_0LSux2*An3=AMR7hFG?Ox(-91P?m5Tv7>C|RGBdr+| zoLQr(_Q^drr1oP1y%T%zm6@w>j+FVf8weEVJ4@d{(yp?RV3SrWf^uALSEwIzLMEKI zqEVI1DrnlP=h~hi3~h!qsd()VbyVh25wlkdeBPr$BH46;(+VokNy$HsSqs99t7pHt9Xk1ZB1~2-X`|!X70sy6Z7!^{$z}#VlR2!W;gH<9S0D9>| zYs-f?f!07Gyh64W3zDN0IS0m*Z2eA?cFVM(9-y^e;rIQ;W@_ILHEIe(1(n>jJW0!1 z_k&DzlXDg0M$BD0(&%HlQ-ksdWb9P!1;nWGCwpfG@VYkzzzi$($jL-Zbe_~Lhm`?P z=q74)J$N%%bkg2#eW9?hLm_DGvDt}_gPC*wb(>+s~LVepo zq=CPqZ3QguLWtM(poUxN=QT>uiox7FHHd)@a77v3`5*BcVKrdC-M&F}47NwW;oK3ydG=F|)FOi&EXk%&~~`N{7S;Rl8Z`Qc92Qxj&$!va|6VG2xb^;;ujyNy%G>k{s*5~)Dg z^!;m~(}ds*`6h3NIdCYjSLP$@=Dfr@Jh$lcj&`|jj63UiR-w-y5JK=tf3!7jaP;Gz~;dq1l+V?iUM_QNO0Qw z3}q;Z6a1~ZF?B$#H$C@$KeTn_RCLhx*t(7)GUjsX%nG||+a`#V;0WDiD(@4c<)5>b z)Z7FrOsGP0FaQx-=CqU%vvE*?t-NRjZBE=|K*kAqObWmVa@t!d&5{BfqL3&_p!iE2 zzD31wl?o-zWTcQl?ZnSv_u)?+v$Jp@>-inoikgCz_@mHB(1ZJCYxLu7-}z%ubKpW5 z1p=U_xI_d-e&!}#1<_sm(`>b&rMkcBa`IWHp{>0A^ljpS{r1GUC8MGFt0pG$lap_$=m8e#8fS55D+TR2IgTT%s`k6i~YD5m&s?Jqb36tO973sI?zbv`|a?wD;wrbG|| zpu`9~1>OeZx(uN>UPn+pljL`_M(rwhc?EzS%Ck<8cMwv5IY+e!i#FezRb?%;KlxeE z(}SEqOCtZGHZA%@w?lphO(58@z#t6U*swh=GSsG=7>ePlJlj>Foy^kb&!2Qw()7|q zG5MHe2kD!ZPJK_fOC0V$w6gk2mFmvT+vnw_qg|2E40g7etOY;*+cB({F8y~CUjPo% zv>IVAlfoUK>nC3<%P`~#J#A}F=D&Hxhy1&ZCL3aZzsrXkx%w~uxPq+U9Umd6?I&%8BYEqZcY2q z%RL08{oJ|Al4SN`bS%$Y%`d;DvZ_26eth>nA8%KyKh?XtOr~)?)juz_wx(JPZtLsX8UH z$uGd23ZHAhj!tcdss`Rl*P3>vQCo4#AK5+6M z&C7js6}ZRSwQ;t&n?Lw{7HX0fgC7R(kBc@0=VBf#aQ2n2-#ok)#s4vwKeHF2 zubv_2HZYL6EXA~*fRju_Z^MN<_~zKtn}{8`u<{eV!aAj-r|QlPb!}iK&}g0%%NG6p z!EP-J3+IN4z=f>uz(MJ{9}^Z|k^;eKUD5*Df}hwgB!OkCEr=idTN7n|ZibLEe0JfS z2{@xQS%=}9=qIxpN5cqF!*D~X*H?WAMdwrv|7+qP}nwr!hV#<@Mu z#rr4r*tKfaoJ}vmI}7Eyb$Y4oQ@wdqcG+)U9UHG`N~T>@l^;AcZ+aI6 zs_~FfTH{A;7G9PvpgM)px~T5*h_(S-{*tpQ=<^gRt|u7sZC_JlvK~CXe<}XO@3uKQ z$VZ;Eo{Ol^fZ?2)uJax;X$y27M=6e)-QS3^%lzD_*r5fBN@qey4{^;P{ZeLP$gVGMXeeXu@vcpW0e{l)j6RXSp^fvtrU94$_rsfZtC4bPM*c|BjGYWy zdMtZxzhNSvcE*iHJe#Zu&C> zaEJA$fnEWGZTPqTYc(oaT2<*BC_2p6W=DZx$&%hN+uB8eNNVEO&8CnJnd{7BmZ6|D zvKGB+*$PFfz|2)EgW4do$i8K4)p$o&B9)=qHIND$^OvJfT$>1Z<@p% z{KcgE|G0I_j->}uAoRuK_~F}>a;r+GD@2@4Kc=l81Hmz}7}!jsVgm_Q z!#!f-^+7Y^hnzTt{JYm4D5dbE$H0Nd+`tlmXF1?6jHiOv9+~58(^#`Y-En#jTw}22 zJsms~I*}rK+|A%eiuo zVfnq}93?bvuHo{mR4Dr{gM^h_H`KSJ0sVW$iZGP32*f_k=y_`WaVO9=SKQ@r1XVj6 z(2*ee*HxixF3l&=E8eN!^`6K?4);e*jyBNtn2B5mWjQ)CtHFU<`YyV2KZKd&cJHf( zow2iY8WYCWRy>8fblrX7*kw0{BkK;Eit9)~LJ>&xhCU;&`S=|yAC7ZLPckz(ocj#Wi$>Wkk<7+bAsHJ2sL_qFE>d?MTv!gQ68K7Vi~+n+b*48N1N>;~eC}FnsFmL8~mF-sl_>=>m70ey~+Zfa8~2 zDpjQ30r>Z5|0wPs+?>oz;qxVw)f|ab>%CLYy;C}pOy0zz7cR1Or8dWW*v<6lv8Yl4mOrL{9T~6>tUMX<&j=lc z8&}M~zEL~o5-Z^rylsw#x;)+to6O$-s)c4bTR%wJ1j4Y__yYP~vMzkUp6*bxfk}SL zE|Ga^Sq+TKZNw4}$h+0$MUP%e(VR_@T^%_PJH0}Lxv_TWPMxfwxd4IQ4!vyvfwEx$ zqZH|^q>|4OT>4lh9|e`Rn&`Z=;CEd}9)}EuB3F3moP>!R5M3oWXxq%?5U+xP(<-c< z6A><9WB0^RBz7O!LKem6TEZ|59eH>+z29GNL8E^kSEe>acECb=|L*8B*_M#Or!pWi z-qyCNeWv2gpX@>QMgJP$Cj-d3iUwF{07#$k0W1op1B!cl-ujwAssP7L7}{9fK&1Mt(HV(LXV^3(rYkR#ja*m`pgUD)cnHEmwy z>R#t)jWwtlg(8~VUK+vq0o!pDA%!A%uB2crd)i(y2+h*X%*n=+uIBBNWOWz4pX4yy z>#E4Wak<12OKPM~TV_q^9-^zvhR8svM!592Z>dm?X>Uk7(g+&9gz8vGOYXnxDx18| zL`%=k93r~~9Q)E;{gEwJkZ_T&B2&a1ty9)?tvCj({(p*dv#mu(ir1dn%436vP z%1I_wWHZy+83#gr;uNZ74upKfG&8dm%_B5$~l)EAN4JeI+;+7-rugNG@4eO1o z`Dls80e{5*^ zRA8kiDTd0ACTDcmRq{wcrV&UTTpaEL*{nY8Y}GCxq<0@TiCite+JIk_xG^EQlmhUZ zXlu&_u?LA3L6;db2AT$3HQNv5j+aLb6eb7&n%AN$Lb})~^eePQVV$7u zpoco=OMkiN(|jXWTHW$$D1+cJ$1mS49pCJQH(aua zm_Uzey{(4xnX){li$GY5Fx)c3?bU@M>{>4`y&xy{>H|VVAmoS28FO5uFL87GGazzB z>;m24e+HF0Yc_X0yst=mBVG5UG%vMV1^j%hfNIW6AiV08Z_O?_uL+hChn@%;5A#M4 zuZF6u(nXgS2?5um9Ti(!GUq3QVnF8?!1rR$vH+LB{;M*C_6n$zq;gT58rdfc)j?(> zNHYf1H%H%R2c09DH0-p{Lq8(U+w&cm(c=HCL~VkXK&=c`al(x@{cUNgBR;nI)6eqz zVP+w)w>D4$Hd1~N!@K(BA_m0Mm~=*kdzdM z0K0f4mDh_!!53f%1qS!N_4y~YI(XC;pl+GhMbgvG;S;#>^!BXnc+-vm`UUnnj2pDn zn$hg>FD+wU08_EQF4?qO8Ul?PzV+7?+ta+RpRQYgF(3y&M|8vYx2$?AESR@8%f$Qh zbs0K5G^G{ZK};_Fwgxs#gK?_)hQJjvEZ{;Nu$P z1lf>Nq)e26XYiKrZ1na=_9Xl?a-4$SLeaKs!6gP}(aq;&YHNeJJRMkBFf&T}N2oLX zHZVvC$)G{T%B|W{((oXC?47ZoyLz{T5-;p-X@k1k9yoqVlZL5wtBY88i-|!jMV*G> z9XeZ8s3h808u<5AM)`jbJQZ!q!5o7Skk53gk6X|S5*-hf=o)4br5;J*!}-@$X}X9N zr)od3H!PkV7M$}8*SKnZPkX6OadKJt&pRwQxZ5(HCXcE&wPHN#gHmUVdoo2cY@SZU z{SznbVpX1U<5fbpEC1+m&>!vUh2u%eZ~}6?;@PrE&rFI(d)y>%o}9$B&{lBlgX>_R z=uieV5PW;|2lh*<1%Ys7r#QFTjRFf;?vl`2&#Lr8N$jQYq6Zeet98Q2z9BW%1mxDn zK$nQ`bCVSn+c(Dmz4+#rl1Ez9ltL)T%VS&|4bIKEch+m3gWIQ-<5?Y@;%J~;R`Rn1 z^dU&iJ3Zlk6t-_(Qcu)?bGPzmozY#-lm=H}?a~l!O`#+-k5+Vvbj|xX|#fEl5 zEg^v6%`wuzSIu`Uw|@yNl}umtw)o3>XT@*!+-#N#*Z!PtI>C-1XJL7W?Q9Vjn(bz# z$|b@l)QD{zWq!VCANv~ic8b=)AfgJ9ZIiU>EaoJC=ukZSNXOv&^-yH6I)NfmLFQ|_ zw!9CtGMz9gw$pKH>l^)|UOREnT1}g~4u(puRGR%(0dI@P`Rhr^Gr{UK-1}qda^(&c zTDb=!AsBEAf{AVYL|qo~K%iFlb!&+k2K#Q%^s>eqk!%W|i7)X12lJRZxMvbH{&-0u zb|D0Sw|Ef`qL%&UtM+vOp{Rj<6i!2C=bfNk14H#IiZ-Hg^ytsdgzoXrTR8v7*qz#s z-nu=YbV`{~Nr}mQwZ0oPYn3}><{cfehUX5)Y$b2o*JqSR$}3}*dWS-bYi~s-GLG@S zGiGmMF3yWJNQ?5HIL*B@e5t!;V8djzu8|u#nu~J-_GTfalZxKr#TSP`>3fu-g z{BT`JX?Jkn)|28cg$)Gd)1@HNHI9^aKkXHEq4?8lFe?jZDoVAzWi^*MXg}r`fy2>s zHSZ}@1M+)4K5gUJ_Oar(=r)vQtm_0X8QswLc~p*uHv3K#9T zJ(qB!Q453^>yg60+0z2~jg>`hAJ5O&p5*6*~Ao%zwht{L#rPe}SBPd;grE{s!A_ zQ~$Su)GK6c({c`pdJ*Nv;G|>sk>RkFfA9DnBs zF+B}_N?rprbJ^g0D&lTz3_X55_N>GKuPM#H<@iKS&N~3)QH;<`-ig{->9q&iY0(U* z@kkH$0raQK!=ogE%u`8ePsPM{wcx!>mE0Jb;i6fS<|Xv1bRnVed_Qte;6mKR$K`hk zPHUu40c&ydCs@gUo2Q~e?gj++W0$R!zt_k*>Jtqd;X~T7M zFAl-DP05gpd#iGa9Lus1I127^Jy+o;xe7;YJAW0EwU9mfBd>r%=C{Y@jTOEg+^|5M z4viY(Lp5$|3xM-b-clcZm@R`p=(6idXP`QJcm)+;oZ9TBoBgm6e#k^MLy~qi4w?@0Tqy= zv+s^zwP<+_Y^3n*)$#L3^!@FS-0+t?A5rd<5Hj9YpLbWL&sY~4qV891xm3VkelCr5 zxW+LdrxtI8j`?)zuTCh}(A-wejZdfuv1oCpG*TNH_UTwBFU_Llq`!tap1U=^f&)fD zGZk~q!%C)zP|hia9`Z{$VV2=nM#lLiU5oUTn%}kNqUyvNK$+kvR z_`Iu>QsJ(EunKZH5A8Q)!_p|M3zOm33ldTFhtAX2ZucOVb`docP-Xh#{NiUVgnR1s}@LiwQh;&mfr@6;J!$oed6lZN+j9L+3soC|rr<*VY%l^%l4 zQU?NSQrx-7l&7n$M`Vz(df030I59u?oMdgtbj*IxqSd@rJH9+adty#g<=;ygM6K5g z4D2GF_T4fces%pUI`k}+hWl_w=4B%lZchN1ZhNssUJCU1BR80Sg?&|Oqrc3n^)#jg zjKeNp)%*mEMYzgq`CA5QbHo*BBrb$}UQk|hau8}eOATa7h*$nmIBKy@(N-1h!#z81 z42jfQK%@Tl8n?+;Na^nOc25V8x3=rx%*@SzMWy%shdWAq32A$p1Wc~e4kFUskP2nSAZ(0}Gr z@ZTUVU$N_Nq4Bv6JOB=fW^eT`OvQ|9 zFg&u8rz+`=c9y_I`nm`d{$#`52jVay2kUk5%T5T?#FN9{a{+} zJ=Nmx7umT$Au2gwL0WYO{?JQd`M> zNbHP=*LUz-BD3OG1yl--pBN{t=!uT;R^G3@a8qn6JL>x3)X)(OD(rmY0%G%Sj~m2b z#hq!vY!F=S^$C2Lt1Lqz0vAPg5a}+v#h^zH7946fUBa%pQ&9W* z^bP@j9|@+eLjP7+Bpkfu{Y5ve`p@}{E;3?v9BPqx71X;%05>#(sgp>>Zzq0;Nr2?t z-tjE$PZ!LZtmiZ#vN9~W6}{z~S#ukYSHPt~e;Wl*j~(3`(UTx8 zX_-nJOaj$tt~6iaF4ghAw09|4zChj~-46d1C9?njQ6e)t6U+Z{a4YG!G1|0kd83W0 zbrjGd1@U$LER(Fvl>Uk$S+<$T#3CY3AItzriO+#M)iZT%k~Yp7nJLPK%Uo9Xx4NVK zC>Sz#ZtIlyglp$b=EAms5Ul{{XXwXtqQHQe?|maKx1->7$BM(l$KB~uW=wuGFv-Q~ zZ7*yGfe6n5qY#?|%fWc^)WJP5tB1?i-iNXK6H&*Z%QbPds|9Ymqmy?$+Yz|Plg7#> ziVC^KEl1h#?0z?JvU+$&*QaG6>zC=w$U041$CZFZ(uFRqq=QU|?%|M}-3$JB-81uT z$}KNpd^5gr7X)^{W}$7ZGKHgc(cj;0MAq{Au(`Msc#%ewYz0#n7Q1I@<9^+V34j;K z)&}fg_eDACAR9Vh{hc6EOpWzi+E&Nb#?~?gtboEZf7*6d5xExXb@s{WLff9W89@Th zLhU!M8Abx+3zA=R(mEu7AW%|?DJprhk^KBqwbyS#_mkAEVR}TD@KDk7)9?w3N2L*F z?u|niFfL=8ZqZd%?Ceq9a1wdN7?VM$WcynhYGY!x-g|Ip&&-T6A8MSI+2Kd)>!)qK zd~?#?$M2iPui*TCX6#t4+!fs44z*&H>9HxMk5sofZiqa)cadLuFs%i-{pWTObj*x6 zeXNIx1X!U5BzuHBj3wK6(A2(aRdcgXA_XRTBD!`Ql5U=wG(k(ZT@23qp=50{PZX!E zGO;*YaXF*x-IbpPR?r}O(a8EWl_+?;eb&hO<^}|RFDR&M81vxgO<`*Ij$Bfo(Vi}? zKhcNEgxSQv%}jFDRMC93*U9E#+(M*cswBlrPw(WK}1(Y`}IVZr85em?OpubaA(T z-n^`3w88_a)L0c`p&R_;C7gxQ>sdZ(Olxv%?GtU4CH36d1M_fYusDVv09Ie!Jg2V* zT+u+h2EsHOY+t<+@=Vn8pN94L@3y&4m zWKIPb6?%pQ?sI9uFQj?zX$24JRcLC54~nX9L_`dVx_^W!B_?&;1tX&U$T?V5VkYtCvaIfP6f>i@&da?KMP01+`7%5Qc=0L<$4_gnl#G^<|f`3-$@aP*e?D0D)sWWYp zSYQ`6);+v1K_H-dfGQgcH9<{oz^`6mfW)W*pTC}d1Qs))52ub(Q|Y0eG_lVd>vmka zsnUnQmAEmt$gidh6LWpwucB?XhlPcOxER*dy_2#pdB?CZkEN-^jueBE5#7vAXY>Fn ztD+~`dWTk7Y?2L75SdZE@JA(o`VWI-g%0_8TP-y;Yy*ccV1#X;-M`X#;7EaofIATKe zz_$MqOB6YD9-4IkR?1I|g>234!)+(d+qJYr5`%{*Gu+P+iXWucs7n}R)FI^qq7b(& zH$Z+F%Qt;A_^AZ*mW^#_E>pO%i=2nW0IN&Abb68q#5Jl*Olop~%=@Xk3V?G>9q0&| zIru8`gjApA6!)4m9Urw$tb?ZLSxfN6c;vX)L~b24y)Deyq*#F303>p>Kpt{9xfkF- z?S*oeM5PDa>teL{HYQ=$n~-v4J+9uL4WYzc%d0p<@NHbWue@J}nxX!>Ah~OOo4}v( zVab!cm6)Bo()g`MWW6rwXKHzT?vD!n^0lQ7{`j(0^-oUn)hWz)mjE6e)+(xz4r!xg zoW0Q*3~GV82z^@!8V4xdU*U*YJgRBHVX*8#`s_fEwg+(Jej*Z7TJO)e!Se0+KyRMM zq^A+RS_l+Dg2^G@Fq$Crn~dmgeQ%dI?j8Vif;}i8bbd=gRB!?NQ#wDPy{E=AUnm+_ z)EN%GQ~kOlGa!d=3Q%IJ~oK7U&drs78dmLh$5vfksh^3c<_3xhRM^ z)6iy7vz!7}dGS+N(}V?6BUH`qAHvx9RlWs($DS{QV`* za!L=gAZb$HdOlKjOGBUNUtlWx9ZT;aNXTUMSIoQs=$>&U-%ih=9!bM*AkeOUEkIsx zDnD*4Sxwy6#pU6|^uLIN>GITBatM(#eIzz)2mM8o9Av zSN~_h>wtRj0abukMjNv*kd*bkPZWnikVpXfg|*@ccF4s#IS2F4|LCuu-{;-U3k6nC#Q$K z$OJ|7Gg6XY#!F2ngR+r(0z`M61HqNaIT#U z#eMBFM*w~W*X|Vb42KSeK##5t!q>aO{D7+YSC;ct4ElEY^>)YntcU)Ml#8u{d9|g} z0=XYSr2O>+z^ zaeH_LIg$c;SX+FrGthnv}#)c@=;{Fjqwy4lRDbhPj zRYLXxi5{Vp5ZE7ju!*0CA*Ik_6PFFR;T;DY^QXD=b1d`>d;<6mqVOjmNI4Ulw^!Da zZlht(ZI-yF;RC#IbDbPYFg(mg;twie;0d8SAd~G__ARf-F63I|Wcvxb@_X*7tWK}* z=%nqS&1T!kA;7kOEwG0nmX369(&jx9#vyp@u%2T0`}Y*9JU3H6u*aye+&3iCR4tqKkiYXTMY>&ieGTE9s>Kw zQNj(8hJt)Cs5&TCHFo-iH0)%;_2ixHJJ|%Mf`J%q0x|M#%6@YZg|*r$7r0@Hi>LuJ zfEp9pL-jEXb>xV3qjjZ8m|nSROoel0OQvPe`$5GQP+P)E@G*hHU;YSQWJ9q_4P&3 zr&_zj%!Yrd<|b!sZO}bONSS-pD4x4kr-dLEKTjBzfuM-ZvI9biz9bL3Ey`=mB$4*a z3n*?&q!i`!Ybz*5Nr}38(D=ADDo^p0xVVA&lQ!Rd78I_^BEld(&kWH4aSfSldM;I! z76eJ=Cg!N59lI^q5;x((z{5)mG*rP#vB={1$L@~^`C6Mxah}8QxN5xDxZ@9p6FB$i0Pg+O|L+G|c3 zYwY6=47F*?SH~SGO#_(Rg3pEs5IaV@cyslAhmToSbrs5?nH^Efq5pnm%l&hOdN3Ab z`PE5V%$SM{S9cN{AxQ8ct{;91q<+5?S{OKpD;RcA*Db9%z^AP*MosBG@ z-}Y*Bm&1|j0W7xG+R;v;Jvb^^hrTl#4K<6-M&bo|fe5IMAN?WJvDJvvsH^NGyM6`2e@W$k4W{x6Ai63T8H^E>hO0h<1H0G?O#>;vFSeAw<8wU5q6NN z#~8{Q!wPZi<>ykG8|}fU+|^afH5W+lVQkQY=_6Hho9_FywqQ-Vrw*#XM&UAar@i zo4wUPEBV>)hd(I2=kI#W<2rMAFKv5mBwEtG$MME?(ee$!=~@2F;~uK9;A#Gm4uqen z{w#upbGX-SZ6f?9qDgKpAoXGap)`v^oeqUZNn^N{1QN7qP0`I z<9H6;|Ec@=f%xb$I6G##d=X2%Lw^JKDC3}i;C*(KM0I;*Gvs?O1}zu$yrOq z!bpd1Lh;4ZinRGn&BRBD=wn49!i zkqm;o3uqXlB($kDTl~-tT#|+k^kA3`NDcIfT9`7@n(a{(s!hft23K$FILJ-@&eM#Z zynnOVYnr>p`|FXkH$nq0Ydj1m)_V6X_JGsKxxXiBCfez`rsnzt_dK)3a%Qmb3``CjLOUGTDjt0Q2h5q>Z zWeX9gfb(w21raC+mD+p_V+otg0^YNubJVbO-k!K1zX?OW*D2*oSi-nFvW4!;{SLpdEph)}X&J|Vq-9KO%>PST zChdqN8sGI$%@rLCnYZ?&U4%tq&)rB8iF$VTS1cM9!78p65NJ5m>TT!R9^P{`^#c3Y zSVg%&AYgXXXwYU*iXB_fRaIHl_{j5(vEog#rrM~c_~*W-;7nkvgY)%>|9(s7+m!i= zhmVilt3sJhWsYKn>&GtEV^k+BDMI3@Gcu{U*_D?Efy`D`CmZ`|##fMR=d8dOCE(%I z)yEw_y98|ANuzaKL-%5FKMVaOsm35oLhI3&`{#YxuKRwzd{ZrR!LGSp19T1Rw7lB5 z@$;~iyAO!-*CFw93g5WAZ_RWA|0*IdnS8KhbqBdp>{1)ao=!K6WTU#o2%*lmrB$kd zk-c*F*5?Y23H<024BErl8&zS1Y)OUYxd5!Wp$59OoraycWoZgoVGuWc+1OBBfk3qO zmgPovSB&=o94TOVHgxgLzaQGD0rTOEQjn++nb8f!z2h91RJuD_xWBYxw&W&nm;dH% z=WwU{v+*h8$}MFybbDN{C2a4*lL41fRD)<*(Np7-!>qiGT*if@!m`GmOvBw{Gpd&= zHM?)t)z4pXN4`~lf&WXXxm%Jp-yggL{z~Q!UAY*u;C6GI!(-PJ!3%sX!>Ag@`>^t@ zObJo3gJ2(>lg8^ZSo~Y*rU85=ZLD|iKRC)-eAbyNZDLji+Pu1$9fD|kTyf|0$#ono z+G@Grmr${0(Al0%n-k}6Y2TbZLpIgewzsk(6S-t}N8#E#SyyYqfGZOr-`-K$S>8y+ zN`SVHqi9D!QSlF6zj&OHu&SI-O+D`%k)+z8o$nC^W2M8#=g zR)#}Xhzx zjyV*(=Cx|gUDZd&qK!qesf~Lt|5^o@SlmYCGuxxjl-d>JjTIOgbLviQ;(fOFaukWd zo^cQ{%XeXhb#T9xYWx*JfU~V#8Tvg8YGS>2w&6Nl$YDisXwvPruE$}ZJDj0o`vMhGC3lI8FLXd7- z%^L)>X4}~ezHJO3>%i^lTloDzO7}7K6q3n8e=jsP-~N2To(48+E#=05u73BNB^jYp z==`HYm2Okoum7qVdKh|dF<~@McFTwuE68R`sJQbOcZIq6x*!0rgwQEK5X1#?To9%T z;C4?(DX|Lh<5j7oBRx2Bc1+Vn)=$isx;c0SnWvd*9_^W z>9^k!m?!nOl@32y#uM0JkW?e7NweuLLKM;uPDO>_GY$g%XAGdU9Z+t309=?Av~(4- zA-OsrQ@p`PqgISDp_64%fFyPQ4cE43_MmCc%_C>z!O#xMlY2-{&g0_l zb2fD5R3J1wWcfhdfC7$c%GU9Zr}_lV@KDmJgUF_d#aJO9h_AOsbjl{n5EZ5%r=3%L zjPWR9MelAwYxY0Usrtr4@)THb%y0|OjuJ@GzqR|B#zPp0MC}oZk_wrBN#wbuBAC=b zhr&CegJRH){lzaW)r^mi8?2Mk{$Md|0;wcg>AYgzWH?JW5s-FKrzdQKrEh~Xz-EJN zakCEg`hoQ}?Ggcxlc1UeF(3v8Q`omQ@;d{GA1`9M^VqDDT1qire=M_)518K2qf2qU%wDSJHl-{ z#~CN-i>Jhj#D+yxGHWuvHtKR$oSnu>W9x6u(Kp58Y~J){DTcC$IFC{w!!_@Aj(Tp_64;t$KgMoXcO3vm z<5dF6-X0Q7p~Cm!+?pCMsqRvikfA&PMvTjX&~k`1`c`my*{AIIi=Qv#$k1-Mt&X4F z-tv++Z*K6Ud;IujDqSIysHk*enQPusXjgaQg*#?Q83$WyNT0Pvil#L^Fm=l*ytIWJ zR^cE4F$pv~AcE9sEOl%gqz>{GD-9Z~S36slBHzQ8mB$guXP0|Zi3_Si{#r9M-sIfx z+U=-%+eF01=Mlsq9ofr5UQ40mP-0)sz=YY_T?m9PniOI|pEwVGM1v0> zid2X+4|SUWPOz~^hJ7hTNvC;bEkau1N&&TM*KnB&0T6akT_+8e+XaiL_u1H89wPJG zR#Zr%#4{3m2s2}n18p}Y&)*Ees`Y+VT>50zv`Ul@Ua{mZwoF%`FE=XfJ!o|{q$xpm(O*~vyWXv~{s+>J~7x4}vLpbMl-XKLJ#DqY#< zxXeVJM_R;Ft)i_Ec;RT_{+PB!0tW#|Q{HY-wb3T^(IjaQaflpMDu+=Ukb-wRze01w zk)WQAgf7fE!jb~AKyp1cJ_zv=Ccf5QrWK?uPPtMuSY7*MF8>5K_r8Cc+XjK}k%xme z;M(f$;ogJwkl71EUq&++{$vkj@B1oy)bCyBfHCnXz5N`W! z1D{lB>}fu1OA8*?gRNFz5Cm@>{f<(EUMNp0a?T=4P-+n57FROV!tM3SN`!?;vxE~5 z=kKj8I?EaVAY(hx^k2hfbJn|QY8A}g0V^6EO_i*NJ(QH={DUr~dG4`y1^>YbV2bHu z98c<8)^yOAcncC=^Wl`K!+g(W`y^dKr0;TdDQGHzW$*KZpgj4F@Y`k{olT)kNXADo z4Jxb~lQI07>G0CyplFh2mM2opzfeY6;za_|=X}Oq2z4t-0?mtQ%959V?pJmQp4C=7 zoyo~q?sW`a1el7t4R}wM<&kn4gwJo=LS@99l@u!HT9S@@G?NXk(sM42U0x%fs%n0Y z7C?t41WFS+IjkcU8{Rzn$?>#cQl^YeEXq;RG>yxntFc#O2T_oCbQjgHHF)hBa-RT4 z>edaH+hT*KqPXtInAI*BJ4q4*?5>So@9p3RE->5lIHdGLVo1&JuY}ar0VP9g3(p}( z#_F$L=$vdgG%LsJ#4w#UD9Pv+%nANX6p%H!6CN$ZYe?RmI88?mc`u#hBWFow7oG!b zH{=5^=BOUz6cR3kpX(+?=^X4blnRPb}R!DB}Wi(q+gi zZ!6zsBI78`EKJ=|w#Z4DojGeiRnPZX@e-`^5_$6s{yh;codJYm^5&U1S&XO5OMpBP zNW7g7cn}Yl(p!nO?>Ju_&<)^Sjr?yYzCmSDb!h3q?8S<2j*7t*^HL!fmX zYc(p8*<5ABHMK^+e?o0sl03aC!8oa5kbtFYAcMRtRK8bhXA%&luD5`y&^bc?hB(+8 zsoq0>GNr6S!hGqVz7Fy$^#Cpscx-{cX}(bD{FPz5lyTzp4pg2S&BxY14u+tL6Q}7#HOofrA#0Po zIdkenhg9fibbP1dc%^QKo!LTWG852rf2Gr&_Tp;#EKjN`yNil##9qv|@g|@(8*gi= zF;mQE(bh7=l%(a#7+iToOv6gIYNiiFE1i9EEkVT|i(CFnxh6;gIt4;&r=uz2$Q22# z4Ln-pO>Uq~+Hk>r?E97hGpTU=W~Ov^GH6>&P`7a;=IoUE?uS0U&Q6du;k|~hH$0XM zEVF6O;%_)ZbLORY%TXV6>3r^PDqhAi85=#WVZLs z_@OyMGkA{;og|sFQ4~mGs<1p>3)C}hqtXTwi7s8DL@R=~{75+vdm@Ww^d|kxci7kx zLOi}g?NrpJ@lR$<@i6HB2QxIgC&^js`=fB< zi`jFNXJ*d}Ip>5Rl7L)2c5%G9g-)FrjqTP$Hz~=Y8af4?VHC71B=G8#46R7w&TBqC zpp#U+W;l)43QYYS=(p}47@(W*MPAcFjO**U$GBnA{0t}hc$W^|&eDOuX6p|4{0{(| z67rB0AdONdSW!E;x!Kg-bY)Vm3ciOIj72vemV_7{3S;AI(>e6iIXJjF`Wee;OJ=M* z*d>X8y_CMLdacuF4Vy2L@Hg8DyG=*mnb2o6` zfa4Nb6aNvg{F46Z@ofz(pt!m5>7wZJ3D!JMVtzzx$yeEM*1p08v*OzDKPMV$tv$HywiQ)BRbx<6P1(zTUH7MdC}3~ z=q!HkljB}*j18|B#F9hA1wZ{g4-QE8dlxWux$eBPgrZXVCjoNMxUB=h;BH1Anic*OADz(t}o_lxKm z6g&IY3bIAg+IvQf8Q)&dR;KJ38u~_NnVEBTq%0Ds1`eh3)JW5}DZi|BEc7nljL(rr zIf|st>gwW9ZuvOoi(gUajN7t0TpRI-s&&_)i#ch8u+d8>c$)as1t$mB>lwek-x!(L z?HHg+rbixn!m*8myB_1gyx@~LtR)GwR*kV^XD7~~|3bG7c>=-CHpf?SLUXXlp-j9ZfR_;PzI?P=@cI)*6BXEnod$`QI53vqeKI$qov}j?h3O7euNK8%;wB2yFzeOSi5)v7 z%fIXLM;tRTrK1PTwm%+J=;6r@+&B7TR%s~7NtRycxn&#)ho+H{uu>(|ep$GA3}tmqqb#D-=dBZk~A2 zt`ImT38wflm>_7U)gd8h*Wm+Vtqewxl$|Pqs%!$40a^MysR=RGi|{+|hU66E7hZKv-=St_%vCn&6lPLCx&)ABy&ZhjS%0ovM^LR1$E*&kiPc>7k z+|JT=a6UXqV*oX%oSj-eQGWU@?e|FU;ci9z5FjhfPOtX;*u9A8Q-C3A6nT8S`u8st z!H+{Cg_*ycjK?fLgf#NHAd_IwbD1=HJ|;^ikfMxP173z z0|&_|tj)>wwj!EUXZR)>d7r1J$6uBe{#Eqj1WoRqu7p4Su+VE$6D}i!8TU94Ph*~5 z0paqVft zT9$zb24;R=-Xv;}E`V*(Z1r-u6xh(SxwTXyqHzdmgwx!SQ_szYcdJpJDd-4>Lp5QS zY>n7)dXhb>RL5s`?T3Op!$`{N#YeBq4lh4nxdoUG*x!R~*Mbdz&7yAnHWN5C>wmBR z%}a$U1NJ;!Py~d{tom=mYod?MDY#@+4=mdf*(Qfkczc#^QQ6|`l`U?|-+}E_1nyyx zhOB1&YTPZkMnR6tzdG_xbh_{S;>=WW5{=0(bZ*)O1ro4o7=w!AYx!wx37=Jdh z`YWs)&!t*2BF+9{XmgF+wETpsEmhy4%^Xdq=Bc9C&Gxlzwqos|Mx1@xc1ZKjm+yZ= z+($m?n}$`yO>NPRBZ6c#W4l~fOqc83pM)N$9MS^@3mf53rI95m>w4%-5|(kdVFvaj z+zd-DiKSgJ>^Vj>)2lHgfHaS_!?aQ-r#?!x{R6;iH3@67%%UW-+hH)Vq*?+OzGrnq zd)SkTT{$8uvs-F8T!%j+)-Kkf*z|oI6xeJ04fIzc*Fm8k%`pQH@;h=Y)k`E=RFhj) z8GU$%gy<-v?bARj10;ZEsu(T)6%jZZC=X3^loR)|Wxox%gR~5&vwDO^&U(nLxXAu* zDwhVf&Em^iLVN=H%jJmXnz{ucMme8>x`gki3XZyj`rTi)&yrhV$(Y<}+6h?5)YLni zzg`rTKjP-Bn2AbxU+hpkzDz8oEu{ef4%Tp$l6@>Je#?g>RNI0zEb&RNvxrC)UlH;2 zQMk2{i;M*po1va~w(;hUfoefdB?OmmG9u}&sX+gr8@QEsV1Q;? zM#N~WuP}+C=&OcfVJc;lvtHg}iLnYr{n2jVoSVcI3*%bHX(Q%LnXXHqjv`e!m9L(@ zK0Qn_WE{wgD@rfSDPIWDELlcjA&PJHlNwG%vZ2A1MsF&(N;JOi;FGzl40&lU_} z+PL-&^D~7-McpU3Qjvn}<{O%598A@m%--Tf_c_c@Mb8-n)eG5jBry!NRPa=F+~duQ z(mF7wCE!`V=~8+s<;^u*z|dNUs@-} zI~!W}Sp=CP<@Pfa&Cm9Qu9dITuGCnLXJ!>pee!C_b*cq@N0?aDAA&$S+fAwJfuI^A zFpM*T`Mvitc^p2A? zc;ava9FWqSwp1G!q#+j+d*mCCx{o7VnTY-KRbHwv;5EtX)%$kK)`EFBHC>leRo$M1 zv%qIL7;iY~#-5?1AWazlDdfEI=!$`-*$@m$KIHl+uMTU|>cPHZ_4E_z$#s6o*eJ4VEEwAC~BjxZI zGO$EPaqdP!$u{J_Jli{??J2+r-`bDzznXZ-OD}{&5mM-XY-36|oMA1$u#T8VrE*x@ z$#anlUcB;#t4^M5icK{87%A@PG-f_oG2LxQ$G)(v6N3ezvQ@jYAE&;-@G1>A3yM))K2!v;? ztaXv8ZH~WCkV;3x_Q41+335>9Mj<-45s3Bxi~c#1G>f9mIi{NZ#@G-cfj~n4BB=k$ zKmMu)H<@bY15&X)n~3z_F%pjTU6XRT@q%JXw{T40{4u-jkjSSjGT|Z&K&6EedJF71gM-I@UaUdhFAfLB~c)JY-(}xJ@FB-#~m#VYD$j{N< z6lwWC;Dv03tv3I@fepDqvMe+bYYns$VBQs<{O6JGMe?-0X+0=LkqXbK2C9f)WQNgu9vgY=}h)cqN#Z_A286u?* z|31(>dp0%NqM~VmCnCC#I8U}JNmRNVFvCRc{@5c~umWVu%CVdXU(XG4o_g~x52JF@ z*$6`Lp1~R?zO&1L1KWs}L~kJRpacS(DwK^IwDP0>h_)04O0HiP0vjD>S}S$Xmc?xl ztNAKZ%+Xxqm}h#X@6}-N9&xzu22K_zh0zmyX+Qq$`?XqdQ;r5uupfIqQ`&j8+)a)2 zB{%SkuUhf;qQ0_Ce!=EDrLxN(`XlKTOC#dLrg4ZBUgX9;N{0gYHF#h)OG?aKAn(I1 zNf?s7)oZI5tXjCLCSLT=IPRb^pia_5z-V)t4jd(21jw>U2pyy&VWph^2~cvrv17!U zY#kg1dUJH6-y-Ai@yc49QBDul5g$+8e=NXbj0w}V%7gJUoOc6VX&&WoFu~Hj{=1RE zIZlK2Z~yFt9rC4n#5fJufLmsa6V9^pxSiJyb=5a=&IMrsQHWt%|B|<=#mCrl4)DA5 z`}{M8^h8+u_=Y(|JkAS+P182sc>c?K!kR4C4lvy9cANo`L$i5Ca{{*;I7mzUV%U|0 zQMtuqkq*pwGi3&+eKl+E22*C#-W;u}gJw4z6& zuWZ!^e!I-*TlnPJUtp3pB1g5H10$cN1TzwuN%JdwO0A9>4KkxfJx&3a4}h9vyrlDs zjvlJ+$671RmQ&p2)NA`5tzkDYc8F5lG5%f`l|E)6mq=JeH1Dxr*EEvx2C2l5&j^fh z)?TS@(NDt7(ynTZMfU~V_zHTgYXN10H?a4v`C5E#*01+?-Qn6HgYDlQR5-I&H;lQw zYxVEdh}lW6rsr3p_x}U31i41f6g@ z=xFTT#1-z{=k;Rj4&U;S4`X%4SYkNS0Uc}29MI1b<;;_#9cD&WQEgDq``=l&oh|z| zN&KfEuTa}m^;idtldMg&0oZ*~`Z#BxLry9>D zp3NlQ8%3qc9G!2!M#N1HwjUHaZK2|u>ub{E?r1wN5dFw3Cdw6Tl&~id-DAw3_$r*N zRXlbb3EN{pyB;p_!9DQ5r}i9W0z-k=h}rC?;<{~H>D?s6ODA0D@sz%rY{?D*$dAGs z;L0~=5^jlcxOVQyqjN~1Wy?-N$ebq+XK99R>;mtE6{IkK=hQ$80=TD7N3giYAUW{2 zJFLX$bvFYN&^mym6JSPL%o~3rFn?@0!N;h+#{beEMS@t}0Zr%E4i!NGsCZd$27f7w zVS>;m%py*d#+eWyw%`ljDyL`WPrQt@fJ^9EMNYUW!_&oVEhdqQOB(||>bq?#Sae+C2HXg2SV{TVxigK7sWuH5Bj zV+bF?hsq4R2^sVZK|t7+pD{;LhP4SwJ&2d2hE&ua9MQ9WdV<5WfGa=AZGt3<7bFi; zc%CTz!@r`t8z0G$dNI~@zf3#>^8pw5?e2tPiLdd83ZXHq4IEoP4s~{m3H`vc>gZ^A za{f<8(6Grv@ANhoTM`t>C>U5StQ)j40=AkCYJ%lN8}7QcyR@S4mq4~Zft=ic2Z!%L z?VbLY+1L^4($#O&`!t)Uk2!Rs=tXl~YFPFM)^Z5F2ks|w@ zlK!x|I4IY_iwz?hpx?pup15|!x_ay++CTUw#WL*6u{?m*t`zd%nf1Um3_U!?EP;1{ z2yW-Cp>(?4|8_tu#*CD^ZLiU~mm-p7ju&?ITR97pRUmQ7DsjteHX?WxHmrB<0Q|q) zE{}~vS_`go+m1+})>M(C2rtN-#<<(tcOEEbe@?mp z8>ZdMj#vYeYZm06MV0e<5I-$_(Snx?9QtB&R{K*_B!4w3w!`!Xza0aLzG z$5l^SNC6^FZUZ|xKi<*H;n8o9+xZ0g3_ThahP4|1;?!)hO69a$S^&||TRP80%E$FL znZmt}K85%spc%RtQ;|TyK>7C_IV32nN{I~Yjl(;T|KI1P&8Dx`$Z`D_<28)= z5!{4vJd8rfE>Jxaxb?_8*`1WACj}!n{VYG;vPnO&M^nH7}d7V z`P#(DmsrBm#&AvGg?&R-+2hGW<#Q_l{}6#tSvshuS(3Q9q}5v?;G3h)-$H-uMk{Kh zLlL>X1r^p8?Hi5lC!>`HWdqUW)AcTZ=8-w(e`2Og|35QjWoKsk-$-dvE}N3+y_YwO zQ$eFgp)e#Kr4H%W?vux2b`2Sk&JvKK!l)CvOT_26AGcEsmn8`Vet!miZ^_xE6DNS@ ztW5hMFvP1Fw;wy{@1s<5M^t42Pd`n^LfoC4ucJXnZ6)u0>&~vMe4M^@%T#0o)EwFQ zL_>tmsbQj!W#T(wqB=a;K%}AjkHgbZh0(==`SF)`%PiBZ$1{#RqR5gp!5n?K49!aw zf`9+86AceiW5K8vhkXhLdU!LG`r2w7m@2q46zbKo@f6%rDmZijUh}>KP~Nuz=mG!E z6;}@YoY?a3M6)y^4;pgtqSit-?PMLw9(5jHa1RH;oxG#ZU`1Ip#z-CR; z@5WIe^Ovj^XDWJoQ;EW>UHozcO?2f1LJ3GbEg=Nuy#d zhGug|%0!(W!t_kJCKh8a!qY@5kh9(WM=W3En=;$M;WvC$?X|5JU1E$804?hyYy$&Y z_h`(nPBL(pG5y{umbP?C=cWJ^El)a)`={SaZ5XO#fvI#O+OC&=wHg7DqU@=ivqzrE z%zP+DVM~#kawUeKWy-0aby`&DmtDCyTGZD%azp_1fMu@1Sp6nV1h=7SN;&p=O$<1Kb z73xb?521?8SoK{$Mrx5vdfGzCmVy7NlxH58Kh*T}V2w3vVaa|)(ojCkV>BtbB89^b z_%iw&s=OQHoHbyDQxqDIf+^yXir}Xc(>h_cxvK2U8!jaf%t!|^DMA#WlB$ZX`s(oy zV3xUF+zlJ#_k2+QM5PHVikG@B^sL_&P786G=tRvo3K*fYw?WEfd%J_?^{TGeBJM>TR{4pz()P$^r zY^7MET_0AzHyL7rGb&iFhZx(Vm~G*1JeuQU@5)GfZTI4gF-(NX+ZB9w(q+DrAKqso zlS=%tGPqj64ABA6^%HUR>Vk~IJ2Z_beq%9*DISz577qFH`nfCbI|9sgg8a&Ca1abo z8216PHgXsy4_f;4D`Z%IjTsNDg!{--cAD6Z6e6EA_l+BFv=RJ_*pWf#f6Tdc$XxN) zRG26qo4vLGjR`WAfUzmoOWyJ7hO&l!pBmDnJ}HJ*vL4N5>0mDTxon72;;JUG23vXj zsaAboWA>(ny~q~M({j;-OEY@a3bc%MKSIA82M@UuftM;65K-8Z@4eZ!P&1uiEDUi$% zJl0a(^&KdLoYU&hV2k1}Vdsr^30w+x!U**_C!cFh`TemPV01KeQ5o(1XmKu949hdP zVT+UGvY1G-bX0~|3&e%tr9c9UGj0uv+-+<(5!ILoNGps&m$=NtFd<*QxY65k<=6C6KkL3mhcjk8^ZuSZyoC-MWb#F>J!FmAW zr5$0G0zwxoy&R8udwRK_wN_hJ`60*R0xJ^Y>YnQdtbL)L@=FBT8~)D`>p^1U8K`K#pr%D0yM7)Q{c&uS3K8_A&KlG z|8$<2vw<=XArrn#p(K!utTouPHzc2s7LFJGN5r@`a;mG7*U#3EM|br#6JA-AUBg-n z?{IH)T(pQBYokW=P={xR&%=&Wd{%7`wc^!ATtY<=Aw(*XI7BJcZ*P1{x*?J%{c-qP z#CWz_^0PRXh5RH=xVJVvZM};=KY88`q!n9LFX||p*7sM9!3J>06+rYv(*r=w3_ZF| zlE!d=>hS41i|x7r+PfTLrKd*+NW`Cwkrs}{X5NYFsx|u@EH{O3928anHUV>Pr!|;@ zVJObP;G&}ak7&ib+Kad<2)M%-09JBrt4Yl*D+Vx)H5~qm>A>5e%sAdcenLU~WiWAV z_6WgE4B@P^;oOm3_b&}WOyl{_QJd%T)yly6|yMLoc&fQ5AlYXaopMTYmWo68~S z=Fj|I-S){8C5^O*inN?g#&UQ_pjVBVR^#QFo%#O9_gR$!m=L^0ui0@1pl9k zqli|0Nm7;U+jY9r+X{dJ|S75Z6WwZ0sC(${l8q9 zy4N%LyqN>g;lW!4u>jWsfyZ@CSmPN<|HTV*UISZ1(7Xh;BRhKuY! zhg6Y?yGJ=%S)=+#kq3|voL&vDo%(;Oi}Mj!_P$Ufui|D!#|I_X-_e|$F6;xqt}1O? z14tD-bYko2EdT|wvITq$?8n2=`B5rlJfKC050CF|6yL9J@DWl8vi1j_GG58X`fX|l zPn|J8=69eM@tP({8cpeF3>_k}dqWMy9wL2FcDq-Yo6Zm6&}fgwfzZ2HIZw6O7%3K) z%jP=A$5ng%oKTq!0-Sid(x<_K=cmRaO1dMPJbA-&s?RK<_azP;{$oK!dw|^mQ`E~3Ft<$(|wfY*pDJ``EyzdMIi253w;@!+qEjsTFot$Uv z^c{>l8fizj;C=ASj5&|BkjyaV+)9!E+JhHr){YW`iQmJyH1n|Bh%Z4HVTadzT5{PR zi(BPW%+tRNt=+Yni|}TWUufuU<#F01n!7tag*%Z^Wc9(I*P@*rn%OfRc1i>ejBDfV zaOz5>tPWyciFfb|Ocw(^Um5899ph5jV5GpRq*;|}NbA`l>i4GFgT9=)3E znfPdlGAr+;**y}fe^hGfrus7T!kYvr^2NS{3)~ywou~iYeFb-BO1auV|zL)fOGfk?^&wU4JZjLSz+{`Z6rO z)E@TjT2j_cFSxMnGMgzp;%F%uV{LrEWiBy>uwqMc#*qP=W1iUw%~bLF7(8N2Z^$~< za6(F@Jf(v}+IrpPWj(mIuo*&(v5anqv}?=a{RvG|#jv=eLyoaAjXp1iq4HO=LrEC~@(VS5wx0V~H)%V%z2NfAGu@>q!gPepr z>4~N)%jElXc(cvh%8f6M;%wtZM0=m^AdVrN@~B@rl>4~hsvD!o#|4}i`@IN^qbLOP ztzQD%-*qJ=PA}s^MR$<6>z-BN2xwxjn+z!>#dMfCc1p=1GWmeKt!jg6ICr~Ba^A@j z8UsIDmeJuupgEtm!4y|`i_+n@B53#<<68L)o+2^?okje4Rxb%?5Qp2le{UfFv=We zn4Rnn4?V8E9MqR5VD=Cn-`}?Vp`O!FVF4$D06XTFxJyr+6Mgem_3hENYWtkA1b<%q zcnSFmslobn{~dj1^sa6`4JF;2wo_}NoVJ&%`rY&{4kQiUWlTydi4BzQnK*d;cFTLBj@5HUX-=AKf zjPzm|(}aI7PqiFt?0Dx^pt6O{9qLPxFKb=QG^xAO6c(ict!9IJsKrn2z?y9~BC>kU zg@OrJ8{%|Fn6-mR;(t)!j~s9DChFr?#PQOitDD_4(9<}2h*Ewb_0$^B%D|4I&DetC zZ2>EmpEp~48)SW7{KaiFq%P31gbu$(k)FhS8v^O>SqoWdR`v!Y=xbh-FabRZ68|-n+DTH>96ffsIL@FFHw>={W|IM-z%YFFSmhMC)~~%8JbC6bN^yb49$oT) zm=ml*4YLkMv{8{)O4sTqQhur5eeobyX5Qfxg zedeRQI{q1w4iZ5?e%_{bzz1c{wM&+!*!T@y+eiaGDx8%x0T0dBW>{&`^$A2??u8I^ zC4%QF(4-pNN%p=6QrM8rJhO+-qLv*6CKTPm)vk)=32xJQYp~+=C&IkT^><$=;^g>( z)9@sgeRg{w*%G-t?+a3*9e1u#1l;T@g{ngJ`*XG|E_Rt|s9EB5Mm%v}R23C($)H(%mMeXIwTP_N5&UKIOWY_T-FDsA zv5uJ7A_yz>*tvB*(iR@5%lM?*m3}^mFm4LU+fVP5V;P?9$;o%h?vnA>R!J&_w3wCd zKxXR&K8uNm)^1)((e;9*Hfn6Ie4c4_Q@6gEH5|$iHzD+qSrG52M=EIQ*+5z?nBzxp zADZ33yF2#K93W6^T)59QsDm1H-SZ{VTD_5u^N&fbskh;*5qIQ!f-lc+{u$ls!&->w zC+7?A8W?Zk_nJN?u?}ihFUf^JvXM4dUn$xiv*m!umzjoHANVX-k7e)Om}Jz z=9}2qWd+^TK~0N?roS2@Q!p%PvIVLk0#Qv-6ip}n8 z)u|yb>&JkFOX5cv+Tcij3=;ZB-sUOt$-8{IZu9~DJf`AQ1R{87|5?K zrC`C=bb;#R^+;$aPQ`(^!?X%zEtvorU_k-Onc1Oc7E{2f|4` ziqJ_(X=?fQ!WI;NJM%ah&OjtXFF=`y!s;(2^>@q#i}Mm$#Fn zC&lm(cohBe>!#o-v6!Sx|$dSeUspuuxdnf+H&iWJskeU-_k zR!9i!?tn@E%$d^EE4p1-3)UnPH5J?HcNX62r9l0)`EsE85`$ES39sLlK%^>__G(j1 z$So2XJmVV#uCG4`t$eU@vwUnp4U%h>-+}3ixtpVZbv~CRMkuSb>MXNd?n~uq+ZFxG z-%r*&D{R)|^q%Y)&d)vVM%!$P9hVt$+KRvP-X2_%#jdbOSI4{BMHrKPQ}QPPZ+)L| zW92mzV!FC&mo-t_DuXj^KWqM4$4i6#AlYWesD$WO5g}Pk-b!JktZG*|RX|`hnLTa> zcu=TYJx2NVL5VYskd{7eE2L0+*?xSr8YDup--t*2iBHoy=H-D*D}B|&xU2b}{vJ~a z)^XWH%@iEyh883fA`Kg4)CdpU*4C**-I2!YtAH4af|*!|>9Y8qH_l->Iljru1hMbm zdT%@GfC9^991E}s`9E^X1M<>a;F;?|Y-YF2ODudoVcF~7u+}WpxB#Bv={rKq{4}+Y zZNpIri@IcYz?JX_K<~n=wF>tgx9m>n{NJe6ky>kYjF?gw$By?i@3{mo;zdErBw|$q zu;!c2;lVj;G^|Xw-Z#30#t2(B$Oc;F34SYR2X6M_PCGD)QB!3pk-!o_Kj;qoj)=)q zV69pW@}$6KpSe`yoe(o5_f+G(`tuEhcEZgbDia1U0+JU2AYjb4lo&d!CO(I9q-3uK zxt)udMmblgWF~*ARHjRh6I890)GT{;8eJS89+>NVjtAmlSy5QA6b$Wn<*rHO;}%Ltg8A<_-dQcdGD4J-6Wf6p zL@ISz?@fClY&$44;v}-(NRiji8Oy(+-HqDTYaZMP?_kxMr;(${jbcqM;aiKx)D;M& z74YXq_Ji>~nb`M(Yg;#`Qw9f%f$3uDSL%;sIqt0)Q;+}7av<|ST%}v>*n9r+o8}Dq zP)Tw^t9Q2%&LN3O-B1>i!2zJZ2G$ltqx^JugqH+4XeO=kgqS&__d*=fWNJbjic4}= zFoemBc{%-;k7QUC#TM|3;6OsX+&@naKL7hDSzMz)x2f{bc|e{)_+T9dsP`qoCW{TQ%U1gLj{Qp&qh#_O0~o* zV$zi6!zb`=xHqIfqQ^+?4V6ncC6=TRLRX0oB7X(+3FL1TA`I3Y;#{lW)V&joWN-@i z{6MiYRh{`Jt!~F5+2&*SRNFA=2rr;af8CD<@VAuOE#SF9`x?+SmegH=!UVhtjx*Gd zS0>On<`EQxCqoRQ!}@|%dR;=%@ch1deMRsKJ3HyZRq&7`M_z>S}L? ztp>+7wlHEx67mS~?8NqcG;VvxKMqVBBVY3s+)!W1fNl;v57=g8hnD~IN-QfAiO9LD zRIeFnJz6(+h7k6>$nC&5Bn`A2G!XeOLj1wuxq{lVp=wu!C!62ri}Ev%)LmLzh$F}v zln@#U1=)-Xm5PhSuvL#SrvqBV``nfEPd0HJ%Et9CK8<%|PKGu+5QS^LekrXqO@`+R z+_C&wLnI=?Y)~#y84r|fy#AA8nsLm_(kccdA4_1c5M_+V}`gv#gjCmu_V&|EH3(kx2JJd zdo#koh`*Wih-nuX@=Jo%iNYDZ1qle62+|8d=E|Yz8Nt55<(!LYmsm0$$qtBKA~|Mu z^cX_3PofzdJ#*Jj<|YM5N!$@QMdE^PBS#>&>qdwy&@RyJBXfE6lW4~-MvAEspfZlaS8mg7~8(RT7;@go^ztLgHI0HD~y*<+gV87L=2E7aYx0CN&G0k(y}i z0VK1YphbABw7bYn!t^1N7jzVEOq2*5`F)2F8rv)_Ckr~X9354UYhIp0I)=&eVCT*% z!|K&12n})Kg%U6Frw)`7A=kj`&8Htm>;(x#K7>3rcrj7qtqzhhe+N+({esUYi=#TY z5au@|&zDVJKoP_5r&op;{Mh#ET}*^ZEfD1}-C3zwtrq(d^dKy6sG}y7ZgDu6lxL+a!Sj`csJC5{$5%!^5fgBZgqyC7lS827iPPyQ<2GN>PurP=*V099Oqo+8}kw z8C9*Ql0d@0M$N{e*q{TpmkG+iBEmk?H1XY_ru6mH-5Pqd6UythgktA?sGJZo2dW_x zlhrp;oEhlTG$R7>zc;Csh2e@x`b;N%`^QN^bVM>EYzB7)@78jNA+q|*E&SXP`ck7R ze7roD!4C~#yd^KP^N-AfwEm`p-Gzs)KZc_H5T%pPRL{-gI^st1abbs~1sQ#g$@i9# zQ9jk5G(8ekG$4SX86LJ92UY&==>#FOv;3|5H8zy6NXhI+CibEYb~_|#llJ#Q|J~6N zASGbVqbetTPUV zr8O4_5hR;1f4kxDpr+B^j0#VRbph|+UjQ?k<5*9W?PE`pnGe0Gl#5pU?R}r{&Ya*6 z<^R$=Gx5TN6E_a)F8fb~!Es*A=rYLJ;eZxvMe6e9L1Y4R7N)bVu$#73kw6FKWAKPE z)~|`C$H2&2(5z9VKo^A5%`f7Hz9Xxmhn20bL!*EhkG>m3*glj2g%1BP>bKcAGDoK} z+$?j`3r1=VhQO@(E_;n!Ry+{5MH|P0o>9R%#?YWOaCeT8@hN0p`~xK97Kp4|NN zd0i|OFYk}*#svc+{C&N;Ct@yl)X@|aOkb1UL0AvSrrP$}kDx+xl8 zyny5vl~PwdzH~reNs*>OlOCW!f4JD3L;Y9bumtmy81M~oqu|C~Ut-3h--+VMy>Nk1 z13SsG=ar}C_M}y_T%I~=6QLfZhIbpD{*d!rINH^_(kvrZ0GdBj}hp1 zEBM19rbh>O0^MVpv~2An|8ICWcoFfPU(8zJn1p(LnY(%y>-acx-Eaxx_XgzzVr(ag{3WVf7BV|$pWow5zl@7$DFUFGp-`{U~b zdmAsfs{t+Cfq%7Bvkhf<72**Xj~T&VtK%)2kM&Ec>R_G0n?vRz1K^B4+Z2>P1RSHN&lyQ;S4(l*3}b zPL>b;blV1IqCA@{pz}D*=GQV~$2DqA@66__L;T=#e=T@$U-#9l8rer8>GjA-EdQ{K zr6QXUr{SSb_b^*Nd!kBjSo&O99S#ErLCWxhrX~ppcNfw?>X)&G-|NWL!7FiW56#s7 z)hVwLhg@XvCK}Zi^mzrDgx#F~^TolZ4-zeXnZB@W(!z*IpD<^-+FM7sVcY8n%wt9u zya!5LeJY@B1ToDyM*r(K7I1KLp2i(mbhYmihEOjOgj36*X+1ZXkT8zp$_ba6a^OpL zS`$%65zQydpwWK(>dnwIjjv!VtQ>a70&itDb!OI8_tdjvI$Uujri$`29EpC$6D?wK z=Eiir!e@4^*(?qd?&bA=s9*%dM@Ei5$lVDwLjsq6)h?trPzs6v_xr|{|D8;fZs(vA z&1h>CXH3KcM_1bE`%|UiD|wYxFy$4qOvJ^rU&K9$ibd!OqWvwKiaKNxME)~8)%@T4K?W}^ z`f5=f(01sw#IowP>q_7J?B!qHTXF zX{wmkRXMH({VEyLV9lYh1aZBz=b4IO4k5?f??q~DviOP8T%B*R11V7l_nGkNtvU&6 zq&AjV9r{H9EzD=L=cbcVCWFDdWcqu%-XYpt2kTL9^H=wE_1N{l@>s=spaiNjEudB| zX}Bnfu?I%JJk1af)_%nqy_6V>vA&RJ9TYkxbyc$DpdVVCvjpF9q$H3nF|XcE?Ldyr z2tBXhnuu=CM~e7AQrZ<)V`9Jr7f(1{@N%_KR449|)L>xcTuCX>XM1sjN)7voCA^a- zmi8I0gyld~4W+m66iY0~#q8;mJ9Bdxdy4jDz z$(SUSY!L%GN%g1+4DWZTPH-})$_9PsbtXyj+ke{JO*T|w>ZvAp>GKNDzzV}ir2Q)G zl)48o-poRcc)Le2j&n6F-?g|=dy!uDy-?R2sP5xXI@+y7&RS{HR^^CI35$Y?KmII@ z9%%N(u9`kC$HG-C@h$?AM4;#neOe`KI<62A(Ez^Ps8yR%1dj2s`LW)=*SN3S1#Y8I ztPz^@p{)G|w4JwDOz6QmXsf^#OPaV?{Th^_Zis^B7pvx&(N>g~x(d7gl){fd$={OJ zzmk+`OwmRrB+41qgXz}9tw2{C=ctJV3V)eW#);eU=zCF`<6AW-c)F>kzg^ zLe8*vF*MDB83xs&_!un_P!hwN)G~oPqHEwOiMWc{jB_w@a_4^iNsM7;3=b!_a=H{p zqO-~EWEJZ~mj+F(Fn`w+?Y&@51Q^ydv7Ce)Npm}1&s2nONVONMcm z4v=lhC|4AsG7T*V*7~a*4eWKErgx9fVGhw4FRLA0FsBh>z&z!AR%TM)nKVm`{3P?0 z`rcldD*KIlS-^*!Oc5YBHAw&}_H7!Fdh<=7;stuk8O!dQ+GhBsKd92p(B$;%;QDRq z3o7Q1*Gpk=n29_mF|ocvjYN%(;A>RR4gJj#{9%4Rhd*Y~?QR$;xhqqIG~4_%(Ls7l!jkco;!{d0=bGQ1XTRkYtle6MkI|rx;n&as*6gGAw)zRL zG77=oHw2aojwFbu$T(sJ0(R_3wm@BWvU3^-eR;@ zW9&FJcXv`-7Tumw;tX!xHydAe0N3EL0KAuDb@GkZAcS9FqqOSJ0~0Wj_Uc(MqX+gi zwh^jK!Cj-8Ue~Qd`}@AVAVSH-bjjWv3h92pyxmI*?hz_-3Veqc&CSyU1Ipit9-qV5 zShG0IAP4vQUj3`y2x*RS@|bh75X+V0@XsL8l@yzk#QSjXm1>N*W z*~ln1Ez*{J-CCrz@wf5Gl#=sf$}>}FKQe_}7R+}&AmWjvd4)Z{!2_T?PSZ7|{Qq zrO2&7b1!TCKOlBrn9C)J!EUuwnQNBaPA9_EXL0LAF>O%6f)G-^)tItOAVt>{e!Q4< z0IM(aT)snhs*F+AU2HHQ_HX%uav#0}|4PA3JIK%(Gk8j~-3l;su?$Gvy?MVuIY80Wg#0@%)SlkMv zyS_T0$Bsd1Uhm7zEj$lYGF#bi+`LP*a!2Lh=0&ANg;Yo5kaoAQWNRz!r3O^#+pR+S zgSa`SYo*=K!MndZ$@+q*=w>4t?Z~o?lK0Q|KdRoOI)G6c-QB#jqU^;zs<92aZq1w9l=Su>WN;GPDigb?)5WV`7Hbu@PU7Qsl3 zRtK?}$OW}LBt36H2*=3g4O91C;Jy0wulNZB^`uVA%7kyDLUwkdRl1d?Mi4fNm3%+$+9YG=XTBiiW+AV=_6|uv8W4reO2|OFO zq(01T7ti@9#U6Jz%OF&6-l`?xH#MZfqkunblJ;}Q=k7Fv83(I36*^e%t8sm21d6Vv z=-te5p1p>hmVqA)z%y4lQ23bM$X1#*z+*A}@&S%>xYHXGo*%y9S5peCQ{-F>90RL` zil#dLEcU~re-6~Lc+;Uyzal=D#EOW+Kx4eU{TNc9BLW~J&>#eNHX~5Rgd0o*k7>yI zHvY-4M!irThu6AD%oFU05ZY#~j9DI`pWBmzfmZdyZz!Gz5EJG8iJPNWF99a#rnj-3u80e@c$lwS^;N1c0tq1SwW@vcXIG{$=c);=Gu9UXF zJ{|$@w6C)#J$u&S)PmKMEBec?6Y0vI(r9qv_PfhZiAO#g*5k9Q-|<8!!iOs zA8x|=jv>$Y!*80s9<6oPvI|1(XGUe31AKeQI~U7)?H`^xiQphs?nP?}?uPhuHe9=@ znn3Y*d$k8{`tf64AsyWe^Iuo?ea9L{=pl94W{YvjgQ5}B2v=%kH`W*KS9FYFt^d6O z^T0^B92!XT)rWQe#L~`F7o+;bVOL{p}++>ZA-DJZ94q1fG!mNAb$KtFas<2-=hs_@mwg@Z9okrM3 z<1vhs0*v=YxYr1dGuXYfPQR~;Sgj2mK;*rblHiwA%6?t^259zp)BHWt=R*NubPt|8 z-mW5Z&jvd+X)_zm!@UTT0u-7R@j;H1qOt0a05b$9=l;mZ^t;Y~H+?nIjIh87#&%qf z4Z?VScGE{jTqa_!FKXn`D1S2fZDVmKs#}DYuF>qfKpwdUn(&Co*c(bS1RO^SUXA^v z%?_KkXXVo>1@=2U^C9eWaz|=?;wUo_ADN+4f@i*^&5D&4 zdJ8h~LN|PG5z283;%%cf+!`urnG!S=RYj7Gg40VN!waxuD=yvWeR!2!+g`(%$|G5%hCA;Ai$krF~A?*P0T|##R~p zKa8D2cP7xbXyY%oZQHhO+qP}n6+5Zew(X>1S8O}4?j7T`?rEI=u*Yu9HP@QVL00Jb zzu^Sk{}p|g5eIFPDv<(CP)E}+oogn!-a?dMoQThw{0X_>@6dtc$+QbRn(8vYnt+le z0}Rksim2}GY|lTt(NoL|n(AV@v%W-`!a{06Q8z;I_t3oc4cgjVA?2+bhsO*>0@I@T ztGo%=A(^V@WXCWQ8@2pVyRaRLrz70OL?$eME{EiY@S<*W%P(_$KmxTOPLbfdFYb6a1udt( zn?BsKG(zStJ2ApvLoFqMz2b)Fgw+?^H0F zbQFy64#MNe_jyng{$c@9$L{Rdvxl&NM!4=Af8BcMl5>56=~HZ>7+$ zk6DmAxlGfqX9;1kS`ai66twsFUdNOKjr<5<*yCk;97Gg33sKQ@QhHvnHEq|^5Zqwgnj;#{XuI_1N z$Etqy8HQZ|>`8VTw>_$ZD^wb)w@rQr|H#2}AfX_^@ic~xA=ieR&*O9?hmC={{SSH62; zFFT-*3MU))r!G_e1!kn$RK`2zLePIdC4Uw?J-!5yoyB{XUzYTxY0hdX{as)#I~QkF z>4cK$Y`aVR9_e%>`lp~V50kI?pPM??=;P(Fl1j8rqsXs*r6#I}p(eCi<3!Qkx5Acn z>l|12nM<=Hvk2<^H52cGeaX&gysQrff3wld@J}z@oyXmeUq-0WWKlFQ=wNiV>x60= zYWS#3t6z3&lN(Go?*j=4()hS_gYJV1(=6a(pI_X%vZ z$xfCFOC_0Peg5{#W&N@gW8KFUi`lD0>FS$R;H+{N-7P4kVq)ZxA zI7=4?$ngsu#>HaG^IY>aOvZHy9P=?*Uk|p?sy_5G#5=t-57#r!dsugJ|P+a zodEbQ_c^o&ZTU%KkP;<)B`-M-sQx7fq)?2W;LT0wCF9e~DEay?+y1ht*=u7d(}0#- z8-J)-E?`ly0d#j~UAzM;kYB}``l-os-(h7VpHdby1y-WY9(A-u;cgjRjZUIy&t28* zepB~%DB6hr>R1jnMg-t$A=HU?F@t~@b}QEMqstkXi;!`sHjBWSRyfR$ay~Z%$e7t9 z$}Fts;1Uo-i~~!OVE9P=Iy}doUCd@39ga)Th*bc3rzQ+Y^ToDzi8G6Fj?djgIdi2= zbY}f{Xw+ZPJG)>*lQO$3x(WZ?eJR%@Q? z+Oqd~1@O^k3B;wT^|Q8F`ea#i_8p;-v&Kn9o%T#uEe1S#duoSVk(e=$E9+_Ux{@_? zkCy{6)7;cgKrQzOtcAMc*8S^OAG1uyeQTo`j4tK&lP=OMvbxuxOA{*f< zhzWg>xd#nmWeAc?AT2@QTL^M$3z;#1Pyy!9a&1%UCreX$Tl$)7E?qh3#z$vA9-+i? zXJ{mJ8TrJwKZjL9D7f=ybPd^bOCd=EaT@7&nf#?2av9hN55lZ^z1YLW6ML*7mENA5 z&TDe~!>7r0+ffXJ)u_|of6I;q9T`ip)HpyDbr8A%F;ex9`gagLkGz=vVbJ7179@{< z=&`IO&Gl>5^|X>4#KS`oC~Ph-pkY#+cO1|M#rZ_k>mCxONEBMKYD|C0ZUyA=nR zIDw|ByE~c|Yi(+3yw_jFX5{7N5F#5Z(MUcyN28&>(x8qPnQR_C02ruX({wCICX81? zGT!(-qVVPTCPF-gDkn!`_ZgJ>ZTOKpX{@YQDE8vxsX9yhf;jofW+|Z)lW0hXa6i}hMU8(2sg;yK1CBks- zy58o3pYgX-l{=ubi?M@Bvlb&mLI77w*6|Yw`|}+V*`f&bgxLUZ7q??eP0Omy<5&U% z;ap3kO0$Fk{(SvdOG~Zfr0u+>Hu}DNVa;e!7w}ZKpm_-R%4u;@QoF6x(mfnZ4pr}Y zH=rHZi-Y$5`lxtRlq!8UmllE6O+CNx;$TiJq>k$>HNaG=G#5?5iIqc1t<;x2oKEvz zryt#2)ysik_EquF+TySS*aYt|>%2Elm3xA&rr6Y(lfcZV)hc>u_{pgFAd^&No9oSG z!PPM!gsJ#-X}6`Y_E&g!Ty}Xk`UQ-pRJu_FXl#GBn$d;g8Z=t-*=RO5z_W+Yn&euf zRf!e__&yDBwRRBWsX?O{%w%UbC+}Z_rEob^S_c2pY=#d(;(r*pNL+n4UP{OzBGahU zSyW)ZF+XDcts_$ST943Oma$!Cux#qrd*pYvGZs;*vHP&u5uS%L4Hd)6IN`nRzNt|u zUB&SkC#_MbwNIm{%3!c!&CJtG8duZ5M9TQfXJFIj{#Jh7qu64j{UC~c=a!`rGbUnw zcNU{S<`7d$cR&6kf=R3kv{m0w=Kg@wQc{_*NVe-4%Ma)s57dkqR3Q-$rgR*t_=SR? z@mS=K1+DD$x~_~)1#R1kV5k19fki5k9{v7F*st|pZ^=a^o=fAci$GVndijTb)4r|XM zrR*-P!DxJB64h(=ryN=jaI~-m9ugj4i2C{E<=pG<=LKe}Ugk6(kqr^DL-kXnhr1HM zbHZ4v49{29=ja!$XGA#7h1$%|$tCnFZuc%WKMQ7Rh;{45?nnuaim06ObsP zXa>`m`-WP-r%VWdl#JsYQuwrW)G2NjIk;lWlc3QYh^PZm3h0|J0-9IFARK(mu3rLem{ zgXnD*9}b&uLvOMnj&ym}TX&&_xL;2PPJ{Zs<~A`lbl0VanFy}3rqRP8f#Z!H~YCpCb zEb2mG{m)4P*7Z(LMf!8r%IUwUzq1r$sZad;mXQ1>%=9bjF8mj&|7$`>!`0eC^q)-z(>ZcE9tR~DENSP$(MkX}zL!82?!AnQ!U|Q7jh^G8CQ$Bsc64|@ zmlmuN_iH11BQ!CZQ7Q81)MZpzP*4DcNrh~Om9-|UYU1lAiz&=|@P%YX_-pB9_@}kwV{IT=l5*|D^@;YnYazq~ zM#M{X<2lAbeomqv%vn`pV1mW_*p?fK_ksq3&H@Whehxm|2>Qkd0hpTjyh48VyDB_>LRYWh7m{Kmoi?%H5^&gJu)_xaq_fBWgm+ubFTT20H7mJUdY z(nInQkFk!ZKhrj;5FJOPR!i}q%>g{$SgLTPWotP#ZF?)R8UQwASpHz0u#FRs+3Hy? zu5Cs9svWY6k@SwzG+3g~$_Jr0l@x-z=EAb};Lw?s`WK3v<$Fr&-y5cAg9DJQTl=g3 zc9G&TN`Z@YewjVn``XtvaHAOx9)cn-UP_PX&1|zNiCN=`*AdTsF2GPvF4v-oBC@nL zUUVv;TK33(>%5T9t00Lk#inAH5~=&ZLW(PJyAxqHs#V!s+Jkf5$yzvYcRHp|JEpdd=Mn3h&~MXd746 zScDmF?|@73J&2SPkfOHDD&;30R>X|N${C0*B8Sq97f6)iMbK*CMW#(#fp4@DiXp`j zTGH8g(9*w=Nj^_@nMs)x<9q4l(X>_nGHF;soYJZ#D;m`s`Rzfd1hMi#ZiyH)5-J`-%_dWDi2~BS*WSA*wVS1X ziHCRym$cQ`(QU~0E_BgoHgFr0JQdr}2Si!8JEUanfc7grX9j~D89DBw>|X4EVlQLA zZ*p#eGsr>J6Av;?*cYgIKUd*y2mbcknY_9Kny~pwt%nG7&SfW0BYWj}QgnLGW%fxP zW9hr?Sj1bqGUMsIEr0Xagy-!lXptz#9D|b~ooR*y#;)veCizz6!Q#&h(2*GTiO>=H z6$1`4H5LZ2fS&C|Xz%3u<}q9QpeI|!((;4d-+HhNLE!QSPbolG{;J@>HKd z&V6Yk`2xXb;luBoo6|;91ghZK$pQlNM_`%i%Y5R;aV&Q?HghslGgUFz0WzAy;cc$Q zMnh-XTXFk=*gEZ_l>(g9GH1Os<6Y4E-h7L*}2kVaDkenCO0J6y}Jj^>mp}N zn_PTFWD;!!L@t@zL+~e4)jjh_-6MZULxcsrt9IblKD+UFlk_mV6y1rFygL){Xz@w7zvB#>k3dk)f=h#nH-lFsSeik5nJbV8-UfFxo&{`4%!m^p6=I8l zbX9`q0M`Q-Y@*Z^ZCfAU`DI97&}2b4j6;Of((OE#mIJ-%*^!QtiS(S9I%T+ib$;jI zQ$6Qm!kcLDDT!_x!O>ioq#%rHq&oZ7??lJPfus0cqUtRy4{D;wFpNcrZ7j~+0+4LX zSC*`t9N0Jg`&{0K&i1yjY<+$iu@#>C5oZ`?`V8NOc!2@KiZIf?DbT<7A^hNuJ8p{o zxu8;JoJ+-s+>(#c!ht!+d~bN}Z@6h9TuUPM&bLq_J`5kvxLd#VnEV%59ZUOW&dnES z)kB1Wi|^j5I<-jD>Wa&X=6$k=E%KfH*08Ch_|qwfiBH1nAkcQ#9CLS8lhX)EC`$%q z$OdtJgRR0>ajv0l2 z^UDQlQM*V9-9njh*9HdQOg04Yf1|G>kL4_-hz)AN+y2LLY? znsdbj{+QYDJzfE&K17kCn?n^-Lf?@KdJB_1WC`~P{G2QD?Zz&~+EV zwR1}>YOM*9d)2N(xap(ur1@knjY2kFPfAV9^FE};SFc9}9O)ApZ1;$%ciul>p)+3y z#9HVUF3t1a$4%gq=M88Od#m8DWuXc8^l&LvVzYrj`4wjLv?kxD_tABDbk%cW0zv|M zZ+qGUwCuw*y68FNGlV9*>U(9!wb42+C%o+uVoFJy zGc?L*;D=MD$>3vWuIocq0;`n7b(g(U6CEYcp~Vl1#1v8InIS103KZi)*f^S@Y?Dxm zv64fiR-(Vt)UPjn`Gd@eoH67ZPR!iZD+BTx(nDlfgRN2ic@}FD?PqGcKTVt;f<7<$#jzQJo$?evt$=U{VMq^q*0#_5 z=hiX4-r=6Cl7sN)S*+6);ntyqn5fNr-kK<=Vz`CxgG;L%luSE0N|%?GUH4TM9kqXp zH$%Yzhknthc!+o(zq=G3How-p05qg|@Fto-OWI_J|LizK1EE-_&M$>&o)tF)?ngY6gkE?o}}ip-Lrj z`Z)2&(}%shZ$y3)47~LqU;->H$B(NX^`ex+qm|S1Xwt%LzfJw;Yyf;LqPhswF{}Fl zzt!{bjp$XVhTFj@v_PHWcnb6puBg73+QXnLIa|?pjX?50vjyKC90k;GfPv3RTPJ z+dO4ipS^X11|{u((on3YD+v^O3xz;>cm_W*qCRQ8(Z2nb_0X$TbFk&RgMROOX0nq7 zAJV?dbAsnFh?SCmv`p;;A*6DVqjB1t3d>&(3T<*%5VKdrx<90sQ#mcoebIB{PeKxL zj29jEEWr=-;X|tUl(~Usyq^8GBDLH9XAinT>7mO91zu9P zMYsgzBYED5NKCExweL`~GDN2ald@maHAexP8ent$_V&HC^$}w&5}=6h58U;NbYiQPHpE0$w%8& zU1IV@+0^LoIy4Tv-{%nO2w+c;DdM2xMgX|8q~cY&^5*{zlbyl&amDcPo?cjD&F7v2$G9XIT3r zbVW=o$~LQjc0J0+I9nGq%#~_h9&c6-{BnheORYGc7s76+Os`rm+DLT+*hLen-LN#Q z*R-uvcYIHi{M8~i%-KvHyVH+8Hs_Jsk7-~Bv}8(cMTF<6Y0SqO!H6vU#)Y4sXpaCB zf8eU7#>l9`ST-uFk*{V=`4u`4cM`s-wyuVk=Tc_1A5d4d`8zlW&(e53IC@W|Qkg;# z|JM&%q1#4rPJUYF6c@9m}Cul6W68TPYL^Rv(c zhT0up^_qHX&dpvS0e*4N_WQa{soBP}Dp|6=|0WO(zm=tFO^S85nrEImw_J;{FAC!j zuwkHOfV?czsTbf;i6Rd#GjN)P>mA3Q?P&)Sku_en(J6v3!kf=)H)?~cm2*w$BcVyB zoO?R7LIfRYx73A3kChcsn`;+Ff=5jQt(~aGl=^DMk0gj@#?Dx>c|L}ZI}pyVgho`k zw`0W2r4sUQ)v@ly$yrjKodm$1-`V)7%TcUmhPf3f^O^YRv(B+>Rf2)?{N*RFM6hSeHBq0xu0+Nd_rO;`NlX?b zmKB`+HC7T*uwXftdbV}}p}#BBMYYRWkMe-Tvx4w62&8tRsZYQVoD*!{e_VOB_Y^{9|{8fbkg6Ki$@u(V^KSgA79bu?1xnl&9BFFGZZi)0?abxwwvaf$0g4W?C*cAQD(SrNmMVhb_w!R(T>$5X#{ARpCE zs)h`2LQBVQ06iWa&bK4%-yRNIH@FPtVY}f`9gaEV4iJj&QpJHl**L~|vK|*46b@~x zVC?m~mjPPWwTxak+E58i8EVJc=-n@%y~0OX^MVQtw5CK* zIy`U;ARj_?AM2Oi?wY`~B#T*PF;`m|$G<>k>b?-7crx}tgNPJY-I|yAah^Kpb~A)h zlw%=C>5)xYH8_!Hb}Q^`j+)`Hf(2JBtS(}q7nknVbW3H%JtiN)LY%tow%K+esp`|& zY+cJ*TjP4lQ4Hpu~)> ztSbN_7;Ft(tWs3WPo4_qB@07u^@4lu^!Q7UWek~@mub0Yx0Gp3M?3ZN@_TER3PlKP zJ|Ax!EF*FpLHGTE$`ld|{c+yrV~8~JU;-#1@WxN%p@Lk|(~cKo6KvTY>s6xbu$9Dt zx7Kx(lpZmSMIl%A5yIsGDYufU=a6m$?Ihx8^b}r4fJmsc?##M^$&nVlyNxiw-a0rJ z6ZtklCDb@J4?5cH)a`s^<@z|hP{b=R>FJBo4B*n(!3W9AGMMq;&06aow160NA4w&hRm;OExW97PsXJf)f)XE0PQa&i0 zaLgUpM?!rGL6i{H&Ki-0yb*(f0K%hM>wu?WfiCg15ndG{6`($LNk6lrW{UQ7^V|x+ z!&Pyg>`;YBbBgbLa5@B)#GMNd@oA^cdb>%H?`&ee&kr_E9cbAPs_ggOBl`0}$^S zVT68}`=BE)#wUeQk`rznlT?cpvIkD_^6?=Mpw_$`y5{(5?zxfNEzJPQLpXu#>-8hg z`#*qzM6tSqetT)T1=0_cf;w)CRc`ln<{yiSH}1RM=-;p~q^D;;YDo%qH+>(P!6d_m ziV+7)Ur8Ph@+Ws-+L|p5QEGk_;d0#ZDwrNWzx}-_$G3u13~XIT3<8JSMaL{d>G-Re z@5`s873p1R_PO`_>XfO#ZuDs1#FQQM1mzEssqk2pB*zB!Rao&ECWnN^hhr;hx_Kkg zD3#CMT7qyu&b3OPrf4FOH^eSHWOXA_j@w}pc}`g)71oDXgK#8V(iR=LTcWVieMFYh zWYJh?BhC3==oFC0LF@aa_$tDXL)`a1v7Fh+Ce9X-GY7-!N<2PGa33e5Ljhv$t+Hfe%TBF*851++Ar{3frMbQn{QLM8BW3&wl%#+@a zR>q=;5?voCr*WYzM_&9w2Cx9`^iwMK2hpoDI7%zxeW9Nu*BWO+6d%Tm@9s}96@}l! z&n%A~$3jM@a$W&|O4EvO@aX~{^kH~l-(9q4vuXTw{n~7sdk*ps>5llb7EiPbq-`h5 z3}g2&)O;J)a8`mWh#CnLNi-`ywzi&Cf)|w_YH+@D48vf6^Dk68+2N1~GuZ|pQW@_n zweYjO?Elnu zm;;_W{TOS}Y}fW;Cc!f$Fi zfvW)h;VM3bEz@W$i4S9LhfT-*#?upbdoTz;5NG5tujh27o7f(`LYpgoIFNc3VXbL^ z6%_LG975!%aob@%a9=IP*-w$#U{n#AQn3tw8DBr0E*&D3fIE|E+&&&_i7IDLyC8l66|r&!sZKIOFM|b22CB9cy{W z=#Sbgmr=G1ZCkxd1Xp#%km>0KgjXJQ{NYHp-av8mO$X;CA_rroi)e3#6I4OKnmbmP z^c_Na2KdR2Y;p6=L1Um_Z8UdeFUGdueO+g?u9LXhT7aT|rxAQLAAD&3JZ#;hv0JQ^ zSS#~$R;npbN^vqKpOIbn#vW*dv|r0dLn3N0g~dH^-1=CEXI_PILHkAcQWYe<<|k9T z;N1{#e!(dJa@ucj%J-PYs`ST&r8AYt;o|Vi)YY6^V|)~K7s6tUJ&}!bZwz`vQ(U?c z;oHa>PrMB;#-t+&<(}wbMd>H%FG5`yE!L6JR-@YVeABb&yaier*`*h;SG(>HK@Cq-ijG?OIdK=wdd`}! zA>8zkTUoE7+mA=5^rJISX(-;aDCsXkKi`2f1cVBXcIJCl^VCw$o+=S;s3ata`95$m zdcJY|h{-Ty8^?sm-PPLxL+So3TGgb=&LY}4Hd~p?;ZqlK{m(hKc7|l5WxY@aptOuo zNMuj$xKYFVUSNQZ#$mU>kekF?&(L5J3obhTvI+6iz})c9I%Cd0>3QpV(cl8d+ElA0 z1{!*q@gRajjcLRnK!KC74UFiqy)8E`TcFci?J7J97LH0~`Pc5pmnC>e!sO z6JU=i1CtP*R@ukz772tX3(t3i&T>}Vi=EPYT;<4|@TT<4BOF|b_hR%#m?zqJ@d$b1 zJai#)iVs@|UU`;LyTKZ=wb+L!=O+zS4%C+8xR|$k@7t!AQXDK^B2uFGh&Js-VDIce zO}BLegECUb$pY}E@DvCFfV0tY9L)pyp1$db%86(e+z)^g1>-0#iyPXBsyP3C5m79D$qd~Q0`edzy(Db+*{i3I?j-yiI4_W6lk!L%`BGn`C2>1S zhWZvSR9TJXH>CS{9DDdatNUFBAkEB|F4N;8mx^d<@!W@@^skGQ>me+(X!ZbfJCN(; zs3_s{%{Mwb{#<)22KSFn45tYgw5H{KnxyyvUX0NKIloNorm!b`YzC`j__fQFc%`PIE{mcc)A%=jK~1XeVZ+9OyJ5p5ik3rMBk%gc2jTP+LM*F z{yS-U0{!LR8!AJae>b5DeWHJFW`66V(Sb1%X4*T>Ft4UFvb?UO<=nJ}jyFx;)`%dk z7G?jNzTrV8t_ClML!v7)i&QJQ!g*7BgW&b!1N&sxn&^fLi3Kp#74S)?MSXAJZ!f^j>M06V(1hBcC282lQn-! z=`}TRn{E?Sym<`K@e2I=UOzo8zY|Q{9g}z?_@Kru!({xmWUY|wdk5C!(SbcRqw3by zx0^N3tzrxvPy!Uj0ln?Gn8?#FT8uj-(X|hU8ykiCmYaCL9BX3OLnaJSPOoM;_99z> zN4_FhP<~PL6?_G&7shAEyR{Sbr-4sXfq*oPCKl#JFC_k*7l-O!{&>BW;!C_k zl646i&~wtid(TuSf@;?JKuNsU9qnM#pFf$&iwR!t$x!uHt)(mOK1+N)@2*hg zon7+BGsjjVFKMRq8`er_EZwKE)kGLwPfts&pTlG)*DqL^I9gH=%AoUHx={K@=z?ko zKi2b^H)kiWn+rd{9K)TA{|OJW|6e@F%E9>G;6d$wp;yl2o{MX~8KE{YyLQV1oYQR6 zE;}!~mJQ|U%TOQVVCn@m-KKHtWV5#Sm+QH;$n9ycS0%IbF_wjW#|p_{!o+dk8#9o; zymBZ3g*$`Ku8q9XlZz&UInHgqgaGx;hpDb6i8Sz|G6s1*)~d{ zeDTnAh1KO9@N{r)?DVpGo1EKC;&qVIaJ#cTgw+KLSCZAt7`4itRyYDsrwx_sehA)X z_;!CeuHebmtQE7JzOLuVwWPuoFqkBj>-h6{WcC9_&e-u2^!3M>s?_x6VcS@Wz`5E! zrq&*RU%M3Ia**C#PD#0o^l$v#la%_>C@#ovK zZ21O-JiNzb(`JhMv7uUKiBgk$y469q1W$k=L3`V|LriJ7yo&w00fs!;j?T=p9B^Ar z#O9@*I=bG+Swr8xraE6J%nhqg28NoRe+kYdJHwQnx|&ooN5)cvXUdxDQ-g!R zjJ@t)4UfS+>)l2~g8hTN%Hav7?7sv7y#>Jq1VcY$u$o)T)~b=`?>) zqYSS%#p?TNVh4yDGapA{5<^DAyNl|HstEGgjvIk#EOLc>uO^30;<`P$AiwDPK;bZi zUfXE3c435Gs>)XUZ0U4?0+}|2kF&M2Xz4UZsxh>y`>@e8*65n-F!Wd8+l+d3L@mb| ztuLm}(K$Wck&mGY+}>e-@7*>I#4L2W0q~)nb!!#0zUv2e_X9qf%I`L#xr<7jl2;X* zIYU@aoSjMNc7blErFl*j3Qf47l7WNv0j$QW){51Iu!B!{-NrTI=*0`uZkD~fnJ_j8x{kWf?=Z%BE5+dKt!7BS!2%F-_9-JiICU}do$7-= zcsNoD144Q&ti&7(=zh<|u`zZF?n1#Of%0R86@ofIVT?t_7AvsChZ2h3U{j76?ODda zkhYMqdM2H8U8>m1d^TzANyjWV(qG_|#p{{FF*krM%XgRxe0^N?h9Zujxgvj-ymrrz zZ)^bF55M@5wtbiW@#-kuQQ0f=`xCIzuxYHLA;vT1f()XBvpu%4J&0n*-5LROFmMgY zB$LM7I7n~9Ojy=!b=Z^oq68}s`v;K)AfsnKgFUni!CG}sJRM_0vF_b;dUnRooCdD? zud`hR1=dq4O6oiuGKu`r$Uc{~0)meSZ+ja=SO=9DQ4?t0HFIk<+p6UYT=Z|batLK8 zD~CLL(58Ppa(q#0Xk!Apd8BjW7!59C=%wqGz2Lr9Q00I?*Qn8McPYqDf-eQycujvPFvgP1%6|7q6cUQ$iHM>IEA8A$ z{5A-MO%vk{@tgR77|%MEia-n9wVd=T#FUmB2c_R5Bzbw|k+|E}XwXRY1+(TIie-1jzfv7{9Xbv$l#ZJ)8 zk&$`5ZV+ck0izkU?lnGLH`XI4#p5ttH4~6K_Zd`CGy>I%9x2*p9W-HUF1NiZ2eWY8&G$m zQoBq(3%m1i-)0*|daPE$#Xx1!xdyxgv6`6mrEbdtL}x2*U$4f#SlZy5_u95z~mrU(&-v{dqIE!(zB@zbl}le0m4|q z>!Wg$4ecaP91k3=c1JO98hL@woiIkIG%aha9$fD94WrUW>z}y6mozF62LxhIi~!Dj z*#ABdMYz~JZ!d1X?nMD2EeaeMpBnOqHWG{}1qxp19CWdflpUN$U@j{Xu%@PGHUni4 zGA$P;>7@jY6;|Qw8VtOoRyEc; zi^-^!$;V1FF7l%|A@Y`~^~}IUoKZ`D+|jCdv#-@KQq) zjxgO_@F(f}l*n@~^3zqKesipqCk+jGsp|I_rI0l0;|EYsgOArjf=($nu1>5d+tZb$S8!h1Scz1F%7=Ff?*4FEwty4vxEFo!O|Zo zumnRdcJ;{Vf5>hsx`7FRmsp2s$49pK>C(on%gCZ1OiS9B9-ZX6j*{rM2qMg$Snx&XItHeXv&j-byx1L za>H1&U`%CP-q_dT9kSNJH*S@CIUD^vgzj4I`>F zqwMYrXlqJ|G|~!stlh!g0b$qNJF*@8p=vq!%oQpEL`UUat=$3Je+m>3tJADb`<%jT27fZsL`PfdI&7T?TC zk>N=UE)a^OI}}NriWN_Z+!}wZM4Affjk<7T^8+cC!x=_wlmT2S@g(_2;xE7&csHI( z(bE5e%%&AMWIhp+g3Q}HGNn;7;(&b$Dw-BX2v5*wDWpjuFa{oU4>7~g-4{K$1a5dfFW>70izdRZ!z=nGtF$Ojs}ha{Sh z0Re2o1&<)%)V~ONPent_h9s8A`I}sx+Eh01S)(6T{y@<;nd(lEMiad)fZE|q3UDmF*B+^h3_XNF`3=O>%c!T1`wIa}mRH#^Wtr3&NN`J>tKIi==jU&q&=LjxM zZIGPw>v&0Se16TBB3V#Sn{7Kp!Kq6^c0@X}gI@^&&)xi==={9#!$jXsMUQ#sMt(i) z3nIEQ`P;i8nfKLR#o_ETWYvhG^<($?g}EGp`qX7I%9S#1+x=~ucS>0@@9TXDJ zfy(5jIO+M%@ANBkcDiPMPA5itZQMG3duUI-ZG8`;RRf1Q6{yTR)Bx+jX_W5Y$fa*N z_`n5LA5vi<2u7{&{gBQ>b^(Hc*lP~xf%CvK&7+KcZu}(ebwpfo-DHkqX|7|BZw3H> z*w3TOw^zT8^FjX2L(xX8ZB`uSJYx4p0mx5zGvrk^ZMtgW4sjPDUYaPIm1gRp9HlSvglF%#B!7yt+*}ghZfJ6NTfZ&l$uO+ZoQFih;8_k_2UtyrF@xn0yXLo^BW3O1 z<-*E|DoQd&_YU}#3F8#B1_sf7zlX$^E)~XtuadKDZqyNX)IfQ2xwdp!`^2N4o+qP}nwry1Gq~fGv+qP}ns8ii%`1f^nPxfR@*Bb2m zectCrdxpjjyu;_&FX-RsBIq&|KwmDOgGp21>S93j0v%IKYQi?~N|6;$^}vY?-{UcX zlh>s0)FV5lv? zWCggwcaf@vptw5)Vnk-u$b-feNQ&6dXVwYUz+~VQg_Sd&e^!vrs+F-K5Wdk;LzRwOQ!CLRLr{g z-4GaHm#vPS_wN;yjQs&&cA*=FJCXx1ckk2<-ov6r2Wh@Q6(X!pnS;<$++78+mKO+r ziDmv3hr{fYVZyNO2$nB+bG$j>yuQq~mfgO?CA^u#uuCi|RFlvw;lmS*>f}g7>z;mr zw=_q7HmWkYDuiH;jHtij4VE0+wGGhzca1#s7=9cHRPTzp_s3g$zWiP{1==_b3l})1>?r=YS`Bt=XeXRa_`H zLjU02jfy~OjJlaFr*qH>#|>>6&c}KBQ1A+&gNd7-tEiDD)q=%c7Thg+0G6ZoBBsOu zZcAQFqC0OI-&pkaD-`tD`Hxo81Zjv0%xD0JicJ*k!c`Z;Mi&G7+en}Dn#XVtU1yK( zD0^_OYHNLzeNiq^Jh?Pr(~3UBBHwfn_izio^)@8d2`xXJm0{)LvVr>p2=Y_Un8*F_ zT;rmhQ2YAEN9-frQ@`~kuVZjh&Ke#l%0~Yif{aYDMb>I70r$-lglpkXm%mQqDpw}! z9{>n!_`Nd`NhS;9}f$>_CX_`i2r&$*G^or8G%s_ zjV^n^f6U$b*n-hGWo>^|4MVER@q#Hv;9XrGGUl3Pcc!O;kuLn^O z3!Vv1YRCY0eV(|gHhh(d;bF~r578cmms^T-vdkiUt1ie!-x$z1W3ZiLX5GEg$gU%u zMru5Xp$HV7x|c0wH{}My7nUwWncb}FP#an%Yq2z|rXA46=N1m{F^pzeD`bTR zO6RL^uqmf$FJSyxc>9*y*+IY88KGbGQXA@D9s=sQY<5yA9B$v-MuCi-b*WkR5-I!Dj%s4{v zoSQf#Ck~zJG6-618^0v)l4Nnu+Fg;;`^X(-LTJ%{5>jKCe4r)*mmo|8mhaHwoo8o~ zhV%UNui%mvW&uOMv^cKWGQ-mJYK}rjh>)CsF6fsu?;*hDgszM6hFWt(Swi&B_sgmb zSrz8|vyInP?EuZs9Bgh0^?2Yl_yF4Bs&vb8*I~s?p#w}!)y@})Ds=G4`ir&NPy>t! zgtY?_X2_0O=#{QzqFV#jo+Gn!@&w z6W9W3JQ?SYAOjR~RT|7^ze2x|y&Wz{Q-B^ikSAxUm|{K(sD_|UQ)Z+Pr-L-GmX%M1 z>4NF#LQ#I}0M)c7ax21x@2Y$ORQzZ$5@kogZAsclLR^)b z4#yk|9MPF>is9DWG$-2L!Z;HBH5k1@k~O7>wm*76NCAVSdrmP-0&5xFBTxzUh>_)r zbka78rm{sp0Mc3;DBm6*C=R4ALAl2>#?wXk4m9F@kVTU$5$6Wm&(N(r%}Sj= z6x|h(U12)&9lB#zQ>@=|hDT)ZU1m2p5ej(YmBVlc1L#IDZ?Lh8A?IErfbH{I&){Z) z0eGuKm`fJXNj_nLmWz;i!Fx`mL|8}U_Umw==r^hM%Cud)B~1=dh$q5E-l{>%JG{1W zrBzkkLv7(rxE`q!X21T?F?qRpc<#_^ps7U^0YGm<%L`AFJI7`}VZqvjR?GXs4?8iq zVbmifN&nL3}66ex21p}>jv3@Bq3Px4p9&T zxA9AeZnrrS4X6j?39rJCnpWh?aQIGh?+`+^OXT}cC=1kP)+e~n%2bq4Al@ptvb_S_ zpedyD(@5$#q+=bk7%HAFYciK@peIJI9k~=hKKrB0lCMWf@ob|GjlB*CseX8}adEM= z2NVV|aA5ia#ENLc83j>wWfTb*4DB6zx>{wFUhJ%x0~3wynvPh^*FX2x0jTq${*>{q zKX3QCuIkSwx+kTfPO#;VnSA(&RiXf04Y|e=h!I_<8wz$&O@7{x{>trW<2F#(5N_Nr zqj-Ceuv;aAlH&H-v(`A=2I`GFxW1!i071^G3bdv`6C{`o@B@B7k0LzRTlb_N4=_Hb zA&GJpo*HBTiSkuC@-`tR8}S-U+BV{mIgs8?piuf56C}5i_ti*#KCgDL1l>U5?@UO} z6ACn2jVGNFnLfwPyOsAn2Erm4LLh;?6Yyb)JR)WM1Hphk&bF2y%m&h?U)tElw6e@& z#SX7K_r+R;xf(xDNCdn%qldt8k2_Jch5~CKkrpcMC>fm)eGL7s>7b9jfO_}=9-v3A zvleD9Lm)}cJI}E9zNv^~7OSlW*Q_4TEFb4p;-Qq;fgy;!^Iz+MT480pN$TG7{CPwN z-sCPqEtTlnhefDlk?eBsa5KTeBqh=HhmfLhw#QZxt{b(8?~ZEWLB;(PBGdqRi#G^E z2lB~paReLoR**#s_4Y6b64r`YqhS^g*uWs#<2C&B-{VjSBKs(@f=O)4Zih(MTL0c< zl!=2Qdgd}Hc{dQ|ng`y!P)U(vt=AhgVZiE0puBSAQ;-`X)Un-r>6))*{B7|`aevYW zg$qmrNJ+Y3g#ik6hP>ht(}8OBOl1JO+amf|&BEEEJDIoYfoEH)iY~q_WfO34A*o4Jwy|oNg73!gt`q*XTmP&SpAQ3>_ z4?#m~{Iv-wO};KHir$tZc#KRS|GcT75u2}VS->?iMjnBoVa^?Z93#%jNB*_7g{_w9 zVEQH7mHaUT%giAG^fOcLinQ$WBpbXS`3uUXH?;h$ydIYLr0S1;_e?fpxZ&t~j?5bd74)ZE+>@JEu{fw-hy+Z`aK9nM4RClp zXPg%(RXYAOVI)jy(rM!;M7XuU!y*^`cAEnTKd<108Vnq+B1U|+y{-=LU%#C%d|JSyix9@QSZ&aP!b}+x#zQU2 zA6)K|eVzESf8{yKj|P;s)@KrJ@OZ=CUA2jt4uwDXjYVe9C7Xb))=X}3kY;%Y!MecL^lG=&&Ilxdq-Uv60^;yq+}>ACv598eMcPCb;f z$+`Xte-gji(-ma#Sadd_6N?(V>JSR`7arm0r-P2;9%qmH1c_v+mha3Q1lS+Sg ztY^3;)E7HFq*{(7W9{USgd*}Doy%9 z-}PA4u|?k76RKMazr4>;)2DchY0L5Hp;_6pv&5XdSzREfBxAxLjbF3LCHu+Uv#YAa z4{5>Bq5OlXDNj?Zgd9!-MKHNcgDR@O0?B?vW4Fug4%ZwW7+zB-1g|?REtrIX>&67{SwyKSB2<-$w7TJYKY+_{ch`>(S+@Vp&0Z1;et6)a4hARO~(Mp;uzBe{GE6!}FV#t4K>OR_mWz4_i_-$y$;Td*__W{+>2{J{K28?8zu zIFT5^O0U{yt4zPXiN9Wy4g_`i7a3@s3^k2>&p(dm5{iYxtdSVU{bZE7_D#G+DDxS9 z&A2JM?YpOkFc2fd0Ym;Q|6_fkAGc}?H)gFv!;KyoGLne^94S0@UKc4JQRY9vIPu ze#umay$y^B9dRrPlSaA!4lC&9jiyl1D5UUwz;=;WmXMfFm)rf^^qv_e?Q(dJU+H7% zH^z{~fB6K;66_K{NT#{@4hE)4UMEEXQYU}nNRe&*Aw$o5R06nnIu95}Mp!lxj{BFy zP`!m7JvIT?2y7plo90-(;#0Re8Id{P=a}`23zype~LHrO?!ZL8AP~- z0Hiqhw-e2!KR;136GF~X4Oyq?-elwp34f83XHVkDGEnKubLjbLE9c6QfU!zi`cdP> z&vpu%wxW4gME7J7{VCr)5BZh6VM&Zz{A&Gi-KA1*W&E62;mtB|d`eU$1DsGsqcm^C z`k7TJ>sk;r_yMVSZrlhEwB?Ou4Bc15wTIu`G(NNAT^NO=ZH6!>MeeHcHndqqjSu_G z^zgbPg;8DJm18AWMZoXQh7uA*W>}qBkd5!5$yTQf-XX>&mYX|zY44?GQy!6f3r2-t z7xGG$U{xj4l}!n8Ek%x1(AaNDAY7nB=r-1ty;8&}`?4g# zMyPR!4}|B-ma-<}_?ZEiDkVrqU=)5ggjI!9{AgV&RQy?>2m7UNP9B7g!OS!J`<{Q8U><% zpS^$oC_H}}H852_&F!MU<^waGx#Jl=lx5ekMxW!E4!|2|G=<_E9rRK8`5q+`yS4^0 zspMz^XG>M*wkI85F@(F@7g^+mxObRUo5;@^Jo_ZAn(Z_2zn9$RPi~$d(P~7c4d%`I z&#Zc8k9b9S6e2VC6ynhDxAyKId=C>r9tI& zg#6kWy|y-OyZ}w{^y=F%Io~Rn^A!g6hm1|vO2-_;@a(5Y`(T76w|#Nd@VInRg05%C z?91|(36HUl&pJ=lU+cB!5%xd4HQ_9%nV)d5*nIWaZY)=D0KR@%9q(A@qvQLRt*)fw zFzXqp1`-!XY}Qy+*gPn++((VM$*^O|)`I@o4^O*rBpb`q^w${@uuh7-9N%sFa(w!m zWFw{hKaoPt|0IR1oE(h*2U4ib=dvZp)PlX71X3yF&%lEpq`1VXGcck=Rc8Hi~ zA#Yc|yUU=GDvABx;g7R_>eT&qbqTLzp5u7>T>?jdSL$)pL>l)4y+Tv7M8+kkle_=R z(UqtBJK*d{!mmm8&y?4hM>{Vf#g<^U5^RR%rOGtjPPBEIZ?Z)@vPNUC0e>e;jzU)x zH7j}sXO3L8dM0nyC8?a5{rQYl7bpW`$B+8gt8HyF`swkstvp~tw6Uh6tUGL%+^wGe z;H|92`kx9`FbBNevCYUG1#K?gP+y#w9Ifuim z(xWfo7@UAU4`Mtpo-wTfo6SA?x=hnaw-C{T#GfOH;r) z264=CCq*@C?!7PK!21^EwaSMwiY~J?;Pn*ZR*g{%%m8E~6!MjK^w1LNPsQk3#<4Pw zm|7O)<>ib_+PaVphgKnJmKNxzagyt{G2rwg?v{{s-~odLnds#N+1?Sauy7dmz;!)6 z>uz1)D2t(iQY9Wv$bkBM8RvXhs$J+$`E>zV(!+(%*lZpeKyD*up!mEmYBjb4 z-5y~BjJpf{z7u!a)-hd^qy(_Deiwr@ltnpQ7ehgfWv76lU<5EM!*i3ZhFB*qb`HxG zvz?{GP+4{78viV&n>-av>S0xjCfa4;V+tOefxNMT7rg2!%J#kOjgdFSv3Y-A&}q$z z*k@3!${BB+p{K5O7b3y+zMV>#(R7UA9$Q*(qkS8dn2w#*D>K5-3ExlL)86pwF~W`e zG6)0KmfwBRl{!3fuBLrEQP19I8N6H|eTN!Wc!;E*zTHF-yvRUImvx4jj$Q3xq<$Cq z5afvv@`Q3}s-<{F-Gwt>7Qoly=BHd<&#t9~U$H23<)ol_EC5@x208r+@x+_^y80{Q^T=BFhXlt_>-X|aX` zS2-R^qAyo#tP$!}(AZp<+iFR`o6+(#zmz=cdeyq z_V|>`dcytD7!bohg4DsI(dAB&TT!D@zD+@az&(2Dfbj7&8QBU@@Xzj=w8Rx(-M-uX zAPtrnDQzM8kzoh^yL7@I#HI~1)@%Hx6dcg~Osh`F%ytOdj-~66y#oj%?Uq5}TszK9 zt@*TT@*>Ku^n>S{%Gws1GtoMzh6^RAK6l_1BcF$@$i|qIj3tr2>GnLE%vqvMD%eSZ zes$`}2s`y+JAXA7eo3~Hodh%$Bt~*D#51%KK{*`*ySvsB-C3_W7v^rsgI(z^DPsx} zeK$Q8DH-H^xChFsQ5$IoLJlx~@Qh$%ZJ9Nq;Y}hZ(fdX7&E`(AR+~w)fP1lay|m3a zTyw#v!qSp&pB>c6{Z%O?jBIcO{xmj0?x43eil6e=7pF$8EA27MHw0y%)|~}sYUs-+ z%97)y$W=NK1Z56$-dHbPY4GHbbs%+HQF!?r|DHm2CCIZaU3eR9iz zUKDe3@VOBA=fd=penho`Yg^MWSx}na&~0bEP#D=o+&Ft6-mZyqt|(8?17SGW!LaPY z_8;uJm^$~vfu0tL%(c|vQPZq+tKd*Mlqv3R^F*Sw(ei!E_YlYucv&VmWc*1(2M|Cg zQ^K6F=&-ROw&13e^n|;#A4VERxrXP2WN}7}zC3egZb$7$hc3uGpk_+&Av&SgQN?)*W7?G>Ls zlAa@WM!-Njp)j43oE=%d!9u6_L;0D*1!J|+dR=ZfGAB<3J4dxzNQ?rzj$ylBq5`>Z`Ethn+jG;<}rI^B%w;u zPA8?F1vESJ>60Fil~F?LoHpkx62BlAidQ#|a*eK5_-cXH7bRCe&UmkC*&l&(w3B29e0;p3V|V90hl_vDa=Ob7F+Tj$FPjv!`24}i z1poY*ZmQY0BPQ7z7J~$$sS{dGga{8ky%GMiPG8w^b4kdu^P(1;tQ7T@l5+oxH z;ekpZp_WInxl;PH%TU{ZepRr7GR?rlk}S*|U;EZ!UncQ{hsOHY7eF>F7cC_uC#njn zl$T@oe&egXDcSrjgw50V+NCH)m0GkS2bTVqy1shVFYaM_&)F54s2()IP%<&Vrxb%7O^npTQ z;lf1Zh}d6>6>rALapWNzh8J#48qN z-O(Kx6^_+b&k#AB=yBSiBQIzrQ^;F<*}> zS!L>r*)Q2l?`dvmNRBgR6rBvonZn zn6Romp38;5xc`;|FdgFOBIM$BDA3QB1|JTK-p$bE&JQr=ycYiQOeOaxQ}Ov5_Hw-= z6$}=8z>;+fKTFH6LB$B9w&oK;C7LO(eH_0it)T;5&ta%9Y@A`JA{rixo>Lg%&jSEk z>1I*m$ElWM;#)6mmNu@-sb(6~^_#P3?fRBiMJ&EMi60s(^Wp#a$9D2H2d+u#NJT=S zxGtFypr9CInJpP!0K%Vu!ijZul=9du$cUZdBY0=>1#os>njg};>lG4Pq{k4=prIPh z-l%UiN2HmKbmM7^jmDfBQv$7zzD*ws41ylTpdN*5 zR|2M5OGz^|sOM=F(E`K~*3@G$<75*NKl{n7x}X`C$V4(0E6bHfZ_w+DyHSqq$0IO} zM-A+cf9!pXXZMmA9Z0|8DxHoDEMjTy{|3##+i;esNEjO~Z8|s79j!liAnLWFXOVul!VQAtcTCbob<)DHQHok5a_V_wbK+bX77$e1XrM5{$f~lg6Z& zuI(*;xe1hA`)bib;3HZc5=yvCV2S2<>$Kd|8;)xunRFZdIKg_}+hI63OPq+eq0BDJ zWvPL?cUuBw2s*AgQGUgg=}xTFP}4}R!Dcr&cHR5a`{_%cf+MYU)DlKQ>bm)c3gG}^ zxjW^&#l;(XEl8bS%(?;CA4d62xp`qdN<~md3}SH)=DbyPS(C27YXKPV_^!|8|DHrS z+=4REcaC!vKaFnpTtHtH<23;*Av(oE_5sQa_VX$zPFN16E>ESNDo+9>@=m|qd~p9~ z7kv#Fyg^>_7K=MR4bd!lEO|kH>LTjPW+h{K0~J{>ct>4VOkpt{65}m`2rv48t{dDj zsI4^S`)f2j!GzK2ANV7B)Q>7Q2B#qEE#r11$1#g2aN` z4oez=Ac^PwepZ<{68lGazp5)UN#YlY{bYIs?^hb@uy1LJAX$X(trd~!9QAaD zbZLg&mj%OGt|40A$_U{Fi_boFK{6|bCO=~tU^CY&{04aNq>{}8XXOZ7L=w>kui20-fW`fTC+&h-@b$`OJnpuZckyjc0bI z-?oU1!bPWf$-DmxP$kw_+TD^)`PC!Jk$|$TQ$-jB zZ_!e8yv-cH9xgIlPKc%%;3O6V4Yx`Kwbju<&!Kl!`pW;*N1J052qO^6L?$vCA>@%l zbndlDA>te=Y=(sD^B&miEVOw0zFw8_TUOCIa(qEp1#$eILh-EMlbR0EsxO@A9uuSs z1^>Z{Q!Z(@lJ>-D%SXMDfiBTWg5b@`u3w-~w5Sex%ydG$ITWU6!GMaqWCodUre@+# z1@1%H!q<9e(s`j!25RlyDqO?b*=K=h#6VVT&slcs>b*+e{G7t1`!oT%GLtV)@ojw0 z5qlOg3}gCF0v83uw46(5gv?H(2(~XuOHd$skmc=~AhPjaB7am6%hD`@Ul@PQ17@+p z1Zc%Nnc5RxNq>HZe+3MBC~c(7s0E_r$T^%zr;sgVL%VCd*89k7n-Y3tE(D?alya?? z%!Lai`|Jd`Kg|;DZ{39#O8q;)b?+zW5-j3gMi?g?`>T?XIegs|bGh z{^^X)zYUhJY{r)@fH4V=<8<4_qBWMK~vZ??xx>(S8Je)3R}Yl zTA=XgsT#RYT(ME`HdK;cf}8p#K$&o3^~U4~KZ*fkt3F zY|ROJWCKH0W9PnFEIwRobBAjxREK47`rdyW&>a>`Nh-wa8eGnv-K%%|p)ns{CLqrz z$LHq_IWagqh$YRpa=#jo;digLkc8FMj$Q@-xx+meZ1Poc424SwTnHYJM@>IV@X&e; zU`jnjB>%0+{PPuf%ZsTEu*P}Tft~nxRqP7&x^XdcTTWLJ*J+?FCbJom>ipDQacjRG zPDQ3L^V7QyF284BiCK8{Rc0tea9*jW=)CC2zAlac)EVWwYm66Z2~@u|8qIy#HCIP- zrC&r2!5uv<54MDcsuf+m*;MaSmHfP(E@YUD z&H}kujh=!fx1@p=PQVK}-w?F;(_tt;-a^~PNq<}3;;TrPHM_%RL!oU(8MZqzWwS;9 zp^Y@HJ@Fhc7$&VY;>_31`*E2x0+Nz2x2GVH$ayc#MBXrL>Sn5}LfU?FOPfI7^VJxtCI6-fkoZb=t@6VrG9)h3mnnRXjljid}mi$D&q}kRq9WYPOf2uwy1V zDVcqBc_reewUi0-{g;6fsCdl93kq{JZQl3%4^x_^OqqK1#RP-iQKn~C zcl1O`Q3Xfi>2cS(>$0VDdk;d9;Lv2YxtjgJ*FaRvO+d%&BsKIY&9mh#wL&cOI~WaF z7}Zzd_fmS5+WJiGgj|>#i_N_~4f=A2^<=I6itDTpEa#NReruEQq7Z?05*wN^|FMC5 zRJQZt@?$;Rkk*J2qMJVb;YZk-s@Y{Z+C+5r8#0M~rJc?qakis^)ckhSv^x{Fsy&Zw z_KxW0(RG`z^ZVD?1A~KTyDqEU_I7m{OSW{ET$ESiPwO!I&1l&#paw5k)BcS}^L8&u zZP^AMD*Bez)=6u%8mR%+>C`jnKgM~PI6)p7c5&ZebP05_TDhDyAvk15w6?RqD&Qm_ zQ!=GKkJ|{Rfqj~d);qUP4M%qd4>c@b6G-w%W|T;}cxCfMNPqj^@<`yseKWlB#SdpRXyt(Q3? z=B9C1IaewE!uY<(EHIu1ER^lx3NSB-E|M#^{r=Bu5rJmdZN*S zCpu?C*a>N;1#M|Wha-ada9yq9#(u|O?gr$bla{g;*{SR2W3T)&xgyG^0H}n!3g;zK zjZ*s@=VCDx4AsMg%jX#(GD;0nbe|dpn^|yPQYt&%?JdKt)Ky+8_=CxL>)Krmb9WI> z4|SFuM>t+l&rF;TAJriE*lR3e>a^s(ZtTp*oSg~~w+$%~SY&6Pqz|R~hSzPJz#zzS zNTy`IMFt|));IRb!P-UyI}BI8SZI(84Mh%p-lve&Cq|@**IqUYu;fTKuj9f2B@>R1 zh)5x-vr^4QvrK_IJIVU6#i6U!X3MoDXmm*)+#2(ib<<}Y8tv4-34f)+y&9|p!a4)r zz6qtypT}He5qdp_l*2~Ckl28P|6)x@7RbjQpQtM9YxtwLdI=ssdE}cyHWsTfwccTZ zQ49(OL~pYvfiXh7@z-~Ax$1)@pR|jlSY?2TlN$va6sHhvgDt@@ryxv)4xjDPLyad6 zR^yV(l&E+@3{3-MhF9S8-OMw0GK|NS78|K_cLM;L>_n=p&EN!S>{PM;JN<2C(3lwZ$!7Kl!YD38sv5JJs~9=+)-Le}A;Ed<+FVHPfS5LSVXJ7&(;Obdd5c=6a?0yu3Uw>iVn5W|r&339L= z5_Q&0qv|C#3Q=GcQ0ta~*9I)V=m0I2N~NQlRPY7>VNQrL2cl4;dkc!W0I1E?lUg8E zr(pPT6QbjBN4+hlssY&sq_tR%e>lWi@c7_k{a(()w^@lm#sK9DSLOHt z_Oc4}bY=^Dl)wi`W4-nfw@SMQ&c6e@hsxoX4~x|qF#QWb_hT|RuiGfa2`B^h7nq^P zl(mY2l|-U^kaO$CV@WacuO->T2NS(!t@eP6%#>P3oLM2-3P<9Z2ul0%kaA7B-`7I+ zmv-u?W+y`#58i*Dhqr(u8=jnMAj&6BjX>hlHfB#%?8X_1TIF8=D$M#DH;9zr9m$Kh zk7m=AaHk-k7bI*_nF7bhP+5yx$_UvMgn7ZtCr)gqEL@0)%?A_KEuCaB;c zHD6pg4~gvytsRE|A`c~`o4u2|#@GwS2yNYl#PElfaui|!elHaI#~z-{WEs118u@L& zD2g2Gf{KS}-PjlF?{8CEq`G#k*_aqTMHB15v7r-4t-MQ!8vY7~e0L6tPz*Sq4 zssUk#VNkBj_T$df4u@8F#nE!vyKM%4=yU%k;cb1Cw0EhQ>raaP%x8lmedF7@orSa-pMU! z$$A#~f}=?ybnwYf2`3=J9k&2IGSH?G18{>b zS;S{XsQ(2wzUJnILyhrM7_r8A(?xx%_ubFJy8z24ja5k>wLkys!5ZiGf+zIOaJzru zeCISudlIF5$(&_apE4Gf$Chb@GB{jHWuKUO5$i&wClP=5n zv9}-CH*@4R8P(mY>`FWJ>5k|%b6adE%s}&|hr|&|P^w101W(bNv$|{7M%|SMx-DfQX21c zItbB@CF-~HXH8@VdW|{eUkf8}t$t^0S6}Y7UTNyi+_^FR4+UJj|(g)XiyYd+v|er1kT*WQE*fD2;f+ zPv$C%NZr?w2gGrd>Z1>UqT$G^vZ(uR7kpqnSMF< z71PgQ4z_uab7`%gZ#&6tZLfAc(uL*8K7g?ir-bLSwH%e0Xg|vs0BwLf32ciZ3QSQt zVOsxw7I+`6QAqcFQFeLd)OAmSwi1KSw6t3qeKuu9o{QC4>{-nF5Z=a1oKYT@yDKjd zY|GlQG7f`dq8+XQ1&#RQ62+d0u+txLp}~6j4O^GnkA`(3 z4t^csNyL=$4!7-({QI68R39nX<-{6`W^)ejDHVW>DbY%NTQ8vHn-B#XA1seb_x01!d0CX@GCK@a1+aB(VXFD(9GC zUxw6g+X0y#(ti->=U);uTU9tDEERUYjS^wYrEBC8K@)XXFakJYpepsG^7|f@N9Fk$N}1O? zky$WKm|+5|CDJE4Ca}10tIbM> zv`22!m#DM|@a*m4r`Nij&cmhJM&vcogZE@KF_o+Q=@!uFBPo?w_@YP#go-!QX}=B)aBkN}3uLoQ*X0+wob=^2IjhJ7xrL3hc;p;^t4$qx$o z65?C+R;0!7)^KV6ASNW0fQ1S3S=?tQ!c%eFr8|u-&lBKZ(R^ zsBe@?Slz9Y{l@uQBlNXz+|@9md{umO)=X&6f&9RJ1kPEyjUggYs(pC_V#~&x5^CuW zyW3Y`*s*wl!%o<1n%-W!(8Tr^s-o-5Mex(LaG~fuM&?(VK6zfRrz>)8^tVM_c|6E( zqOh6IugZ0QC`j9NAN;h-hmFnT+bO{Y*(Exe_MKpFP!=FRu6AEUFC+WFO_sbzxP~42Sh}nM7~aq6ABZC$)7x^R>1WK| zaHkhMm@e#7hNT_)J+f%o(|dSBS_&!mzA3uwTE)8c+@EdrchVmfg(l?qeMNy@CdQjs z8Z6;BqRD7ze|q#2dWyB~B!3iXQ)QXm3o(u*$Gr8zC=|E6z_Y6`Y~X#4t(nnMX6D{ED=I(9l9H~?D@E(06^fpVo5ksv zp@2Zqfq~xf?eAB~{Yce)x^pD~d0v136)Erc*p9?;k8?Mt?a-L?a=~$<+b%L8%g!wl zW9iVizG|;Sm7BqzD_&DaQqT~)1GIIhSi|{n<<%brZV6m_WYdQBQ zzWXw&o}@^29uQ3TfpFNmcQ7PP;thdW-*&*()U=4`InF=6SSW#ZlO|1yb=tg}wUaw% z_|A+OT1kY|Oo7jDbil_aU_fMO(HQU2!dt~pHzfZm{D_qfJ@Ri~qpyqKywDp?E1g~~ zCA|qGeS(B;pZOQyUP2Yc8Q3ke42re$95qPz0mRDaTKs36jgyh_Kj|AAE7$)e&h~Hd zb#o+r+vaw=3j3z)eIj&@K(P25UFA;+XEfa=5?dMvTCcPaDlIWNG&t$8q=3cGy6W^n z_wkag1=H!y4TNCWgDhEX$pqNknf+GJYmSRUt3Ot06mIF0!J9`5ucQPAWG}8{4-d@W zsr2HE49Ex!?Ks^7=d=WR0sJ?Pqu-8E&eCzVH^{kW#goTDoSmFJTs+tYKd~8Cdw6sy zMPqb3on8Eh41Sg-NoDhJkIt&_L-ZWbssENIvFaZ=&Q3rM$>V7CTck2CJ<}pTS$=k(2b2zdwzO-jPQINFru{pph}p}FdUdygq%)K-DT!L) z+HN_Fg_ua|dwV<2aV;8Ppalz8MQGe?p<%|M_Z-ho*kR`$ZbhYD8$qAv+$s*0qjs$w z&(mC%+kTUG_8yOpaU!?sW>Ko4u}{aQG8F8|#b(OB{S@ZRrpt+aEFWWx{uUp1l6Igj z4gDZIinxcZ$;Vn|6S|T-=QAfRm2aA2baSmyIrkyiWe45$*9pL;sVi!FCCougl(CFRr-oLzhH z@fMa=|Hml-{iRLIuX^C4k8|1jBts5cX&&P~4B5dFi&pUeF?LQ{f<;>v%u3s~ZQHhO z+qP}nwzJZzlEaLK6|?sqrxQ^d z!mjG%WG@ni*A2t1mtM3~aG7V=4BlsOTpfvcJ+5I3AFyn`?7q*y_!ud8t&HYGblaI% zh85eSt*VDDc!zxxKYs`meNKYwe{&!1w}ef{1NA&SwF+!u)#{Cg9YlaW?n9|Q*fdqD zF$$)HQ+vYFcHg45Yr9LDQ~#51=?cJ@UEurbT~kb9nzPxTAS0@0HH-}1)={x>y3YkU zDn6T*8Q}NPzQe724S0_L$o_nZ3aDO^4$7SbgdV;NX3q+~SK)>WpN~ZHUXm2<-C(iu zh*_P8mQcB}S=C75*Brg0n$4DFI&Kx1FqDdUicOmaZVl)id{gEi8oHRz>kO|;T6n%W z$N#!1$DMvmk+b@`Me|LICy94o|6+Cvz%O|dexS|Ap#JK6wLWXjSywl~au>p7KSktU zFa;*szB6;#YH+##Ep0MftM@$@*FvFojPe9N*nN^d@$bFZ<-p~={uvK{UlzYjdck<= zMj4`%JUg|Xu(x_`4%#)N*}B+{Aq&O_)dL+P2$v)FT|lo3G`$yTvvuoViq@U>7Ja|C zr-DtG`Ofboq-eR4678tN8;l64-3@Jbh3#=Xxjc=1z(2O~a?h~f5Qj=8YZxb~p#eh- z^Z>3G#tE|!*icaW&vC#o%XDB?zimt!G<4Iw30j3WPA9mxzWR;vfpyc14MbqRD(Z7I z*IFD6y=Uz;{>H;roKNcp7y_k?BUZ~m95J;LbSaRaX_iMH{#$cuT(y0(sZJ3NP{=8L zl4rrHb%f4a28n$3vR~%=D17WeFns6{*1Mh0U?h2nxRS@fRJX_)2+{jZQ>>f0kMkbR z@Lv(1foI>(MO2Ld%)$}Z|Py~5QP~oAgrmirUD#YF7 zmG7&kr)IAt7mv)R0qooI=*8?QyU@*+1wgD|dVMvYW2WF}I!s^bttPhY>{=jtD zbEp0O;!Xq=GO^JS&0vJtKZYzx7M$r5XbV-{jd+=S45WC`Hz~yQYGjN|f1IFJjck8_ zPi!X|h&{o6^aYrfOF)fAP}i=Vx(WU37r17-O1r}G=k+5fW*F_2#~At$wQR3a0>qGX zoymIcfN7p_nKjWyc$vF&nc-*1uj7V7R-)Quv$her3Q>WD zy7C(1!cJCz{|*!oJsZJB%=gx+y@WTZ_lIetzq%+d$Xr6gJ?b>Z%B}~-C z@%M|VVkM$^wupFK0(#A-Uq&L4)k}CZ!Hx4TiAJ6iAoxFZL5r>%qL5RYfDwTE{?IV4 zUmK7g->bVQhak@(0rC_J9o2*2dBuCzR^rMyJI2ai?Sg`}57N|v(`i4jSvvKdMwM*$ z^lY8)Q9lRa_o-69fv!x(TVUO3cLcgq{yCQ`^oZAAYh2W2-wR=NlpLu+#3UAzLEdBR zUiLzoP!%(j3t5TSmXcr=iyS2@Oo zT;HAyC82OYdaG4}SB_}}C{yf`cl;pz;@$-iM52^+A3d1jDV4dAgABPux6wa$B^|Sj zV29ZPS`q+%^si{PL6$U=t}#Fk1TYpU?ut-Ip6W9YvhFg1-F!`$$Q2xBJ&=O4kgWVV z`;doB{Yv4W(14bK*wQR!RAQI}H5D5mH%xU{osS=1n8N3)uC1}VqnYV*3HA&9YV>wQ z1Y?h}FJ}S5XJ7vi)=v<}Iro1GuoO(5n{~Ep=5^=*(TeZUZ-SBxUtLrYePNA34wf~f z5N&l@wCA~at4WM-{??+DxrrB`(?Ilk(}|4f7ycv`W5&wGed3xYe_IMjrQplpjDaMX zVy`IB0<<_6JT>W3;_JxdjT&$C_E?eQ(*%(Bi$DthaK|s*?{u5-*a)YQ9v@u0HRSGh znIyhRHxWm`wScwOkxC#x`6X;9gw{1U#*qdvAPw&Ux6qjzYN#G*R`KNvGt+}vOz2PL z3p-mPs$2=382V3%?UZLak4<>-xWlw(@Bb4D!tz$)xmX4maC16x$o zweyEq+nsF56F~^-D`3HGpSrZHFI8C{<27Y@;^^6RcS2SO6F6gB~dI`KI{oP!l-#$ zC-lFq_|LD>I8rR=f1xiq?z}Na1hr}2MP(*r6Z2yqZSyTFpB|^dyAz|5lF>8-L#m{z z>>gK;!5u^4F%k&n{Cl#w`KUBxTFUJ&O?(}0WF}oiHmES{ zWw$<)d7eDaN*wqlS?IAi*zv=rTs$W(Go$sW0lA*=lAsmDO}Yw%2;m}3IwE@=Bf`23 zw+svyLb%M`IdJU_(VQ#c({%07tK+iQAwx-?I?+1KvDux_)Of4D2~J#!v>zB-_6a3^(#QfiJ5&kB~ch{?NAGlT>z7kfHNYg%TUK zFQkM`=t-7v*M2#uDbuBhPTW}fWx(kZMoIHfK6)g=jPt8q9gY?Kq!$}D%+78K_k*-f zAZ4M&;S^T=v=5-&3BJb>a{3j1g+1lZt;=T043o2?+@Gi?PR)1C8U>F(Gf@W2;lRZ= z2%~S1!?vlvZAgT_ap=E&zgQ{=9oWMT2(9;2k7?w4Xsc2`)Z${Z_@zJDg0lMj>nM=6 z*at=OGt4I`mmyU!nvtDc&etT`qj4}k4RdamXyI=V#Jk7?&OK#4l2Os;IFM1k_vEiT zI57emcYaB1@(E6=lUpp?J2nXNLk#n3y8hy6lI{|@W6m|ICPwK=7xa{mF;B=Wrn7*E zpP47?No&lC@~$T12sj|+y%M>iPb|&c@^id2u>}*8??tE88#v8>1bk8SNbZ zd1+CRPM}Czz+Mb-SFv5a?7$Az&0Qaj%w%Vae~xCUo9^!GEVmUuU=Uv4niQh7+1Oew z?$9K$6;^GeP5tlJ6HGd@SevSqv z{3G5&t@WT)L*h5{e)U&_ANNiI9_59lb~Cc-j$(A*dBrAWYO?Zs4F{9qz@!E8+LbM% zN<+EAN$)SA;q^|_;P=V(z$?GA2d~gnkf4zi39WakaLMxLCA-z-YwbG;5pAM7Ub&9X zOV(btH2E(@-k?+dx0^(-ObXw@LI0ChFIV0J2H=j8S~dEvwV83+fSpI~(o@zeoLd=AD!`w176V_fX-T;rpk}6}UZ&`?xmo_Kc-h@Cw!)#*c&>Wd^v%s3oloEM2x?Hxj#>l6p`#u|Y*>d$g;!a|4 z)cav<@FY9c2yiO+B9-4ph#826+f_!=h_K@ZMV6Lal(x&(lH4E8^B!*1>(_-5V%BvP zZ@XR$qEiny2i|Qhqm>9&o>geMg7P2aNWNe~l@IBQW~$9G-@3vM$-wF%H^JbMjhod? zi9$~^zOJay zBN*a%xk1F=EUPR3-et)6X$!TPggehoTSKV6Pj3dh4V71H_nrz~#H#!d_s={h(AM9Cb3dcy=C%we3kvNQac)w4u!~7B-{?xc*Ue$tf z@)4vxIF~U>dZ3aMe&Qp)_c+fQ)4m7nZ-nv%u{6Yrz%o{yYi$(VYvfN=XUxGauOcbc z%cpiEP&y}SbP|>B35V@m<>zeiP@4Q^Pi4AjZIfsl%Sf;~FQdJRxvJfN&y*uv=N>n@ zYwB|lYg|}?R~O9$aiXLUu)U$P8o;RJWfWI@NJmH_bJ(<$PgZ5Bz}gKv$!vTWA4B(W7RU!(LdAr?s%++@j4CSe)@KHUDuWFQ&-^TZZNl- z#*!W!hw`W5c^j5jL=VrWdqv5$wY5ck-odUJd-m=O1)CER{87FahkgI3-A_tyZu+t6 zzH&YE4%AXF-pzMne$e++qzUSNh(}DLxi-4e;;6PqU{lItm%9i(mH2EbT)`tOlEJZo zdP%|?NsS8hc)L*li;>-{#nzw+LZA6=lRNiE|MNRgR=H6Y%7ut-n);I%1!5x0Qlo~| z;1XmmNxc(E@W9;Dy8}!D#ixAJz*!c9=}-JG0JqSE*)o&Ow$rUeu@iFY$E5;GUtTLo zie!>tr5H(`Hm>hpyC3>1yRnPfTf+cGn%^FU54zC|UE`QM9gC^pCkfJ9S@h7Kx6fArKIEuz zP(Lw#>mW2+u&fuGLn*j|EfWtS$m^lTpBBh$VbAiH2)X{2%dE&+VmW6O~^WkD`1C>hQ*#npM)_%ltCco z4ZM49tq^1!SdZ+tr<|L-w7eCTYxj_$kJ=wM*_)W%V(QL+38-fmJz)3tZEC@m{-i=Z ziE9@NzuYRm>kGCuMPvD9vmqLkB^jJgWpSWX8EUCJhf_LhfO`dEqR;1;W~Xhk47X!> zth&Pr>vHF3@PN>;l71&)@h38d!&v{$!9{Onw!1Dq4;rwkmb4I({g};hR-fx<*WY(j zKg#~!`5@VACdsh_2(r6xk+c=;*D$aTCwT^(Z=bEwCndiIkOnFgc3P%fe(=RrI#@V6 ze1lfC_|3_|jWo1)y_RPeoxm;VKhUa%oqsv4tKw9mtVyVcejyHAXQiPn6y`=}q``54 zuO&hz(n_U^@eV(OenAvcXOjLKyk+<=@RpwO|7$H2P3(TG-NZ$A(s%;K681rJE;83B zwUJ~UZ!i_gl=t6;$5M0)5_MB5{kg_=BYNSxnyelWo6Hx~rEt~cumy zV4phE`{L5VF%J8OZ1KQ-z0iNBRKw(IVPol%xu+JlB)#09U+|yc!+7OPXpak;JSw9( zx-@Z!&Ctwtb8@6>cM6iQo)ie5cQ~IqGx3P!1c6GxXsj)xCzZ|bDJw`9O4?^l9C_YX zm*euFPZi0cqhpe&FlUNICFo$wD5n_{uzntzzXM^QcX}m%I5H?p5~Chm8IL2p740f* zwzm#7%upRAWNou7$FCQp--M0DzJsQUPh5DqIp3c74g3e;;`f?>s@Y)#8Aw~+_b>X4 zN#(jPcBtWK@sgYTl^Z0g#Ez>Qmrgmx-&GzYO(lP=7+KJnk!U?qWlK2+w{}QRn-)7X zRG3f+gj2i9>;QL#6{}Ah3MP}FZiKlS$B*}3>5w8!jxxk8KT_s_@kklvm@<+lO)%wz z%l9p7h*6HsF&Ha$7wixlhE?x@DD4ha-+74r7#P@VwT~p)1%4w6l9o6#WTFSE zS_y=^s8_5~iydQ!r?3jZocmH&hh^F+S2v5uNvRZ1y^)c8=V)bVipxY=Ll*(l~ z5g0*eG#(5S=P!U^rB3RQ9z7q_Io`$Fv#}CH$B!Djwo9MHt!9%B-#{>53{iQ-SGXZk zz%N)>C{n15bkmwUdu%U+mE4O?EV z@8K@HrMPN-Ypp9gBUp+JiyKsADyw;_7;{=y*wM3zbG^4ed7dx2HY&>Y%`nCUB zH4<1sgbJR@1jQr&GXP%2dr>DBj7Q}2?E9nS51w}e`KC-&O}AFmx*52H_6Ts*j-!in zzCOv;Q&XQg- z(@DPXrviXl$~`(qv|fzg(0`@ASo53Jm;922v=zT&(|Q`zU9pzz>(@}Ry%l@iZ&JQ2 zSh2dIc4K8YaXHAp5SW|bhlt(W2mbdRgWKw=IWtIWs~VIh2%>nb2Is-PY}EqHX)qMpzJf+{z~vW0=I4*;17EEEz!HED|A zVN98dbiqbGb#ra1&0?GZ$9mg*O^B ztK+LH7|FWt_9UywpKr+BFxZeJCq`PBHzklSj9}Bmo%E(i5ehp|O?YioFkQ&~a4pEe zrWA}HAWLZUh*lWlZqP*zG>vPeT-(DW38kPES(;vW#4=4nLaOzN;IG)*7yG3=R+nte zmaNK~sc)#Jm4HvSpgA|$4xU41q4Jsdm{Y))RK?u8Nv4gJW8sbwz*D3+OG%-@SihICai*Z#14 z!jAz$i9$&G-V9fIP5D;2OqAaE6}J zCoM+m+}Yd|mORSX`GLf>ejt6c!zmFB_-5G{xjxlfKF4~~kb}ExV-F=Za7DG^wm{`N zlH*W~c9g{BDw1P&sbcJ*R|fC8lqq{(qm_LGSyUO7A)9*BsR`RPC}BjV(q95;0ALKe z7r7fs;GNiJzm4_?{-mz9aYLQ&9w<^$_CnaFC<=c}En}YRT)%Na7acB`9-&wB}`Bu)n6uTTgQ?{MFoF^N;c zhz~h0^h+3ewz%T>C4_n@dbpV6-6=$n_`V=Z{uZUnE8bbO4Q!~CKMe~KQpD%$-Z26a zAX#8J*i&mbqQ2l5T#?hei9Y>Y>8erSB;tX5DX)0&rHL*kb9wJ44#=IzK{EiXt7K>s zeBuJ>qe}5=HGL272)3S50=Wzt^6mWTG|HJ%7D_e;Y{F{D{FngjGh`H;lZIhZIY}xw z7zzHL2lFgnyUtp2yR7mn@eqC9+>#6U54N8Ji?$QTn-<$!{)Qk}wq_g2^pSI$MqM^6 z+P2|LXIKsaB2xoH-M31sG}57l$TKJyD2G|DaTDgDL^HVeYMs4ewgAj)tizr{6tV{- z`N1p)`)~7YUI$_o1Rz2ue|GxD7^n`{X7K$j?gKT+HW{x7+bV+s6#gaJ-n!*Ox{%0) zc>xFB!NHZ*>QIVmPn;7zFUzSAl3g35fLY{YA_!6cq+=h+U`pBO3iMaGwHLd=)s5`c<><$~M_^uF5wNUWi3Czw z@qWY2lmbeuuh?M5u%sy&3ma-UwfQ7YLvx!}r6;3c$T)v0 zk9+gG<~T$Lmkn1&QP*Wz^4dqTeJeO4-i`CEsDPvGsn?gvw7`6WB<#*X^7&azOp^i@ zH%rsM6pG0kVkAi;YdKzAFBZYWVS|hw?Gvx9O`FLU+9b7wB;ntqgDCNsj1rOV%5@zp z(PWs+(7V~dN66kn@O;J{5uPWL#)wqmN{JT*C*7UL3N6B#j%9ELU$J4e@L;sOY6wvAgKZr#ip> zx?bg{8bGm|CP*Pczcgq&`eCQOs0k#Dp}?PK%;-jxg5tKfoE z;fdU2WPn(v#v3i5TyVrvD6#5!H z9!HN=OerR^zY>T*Qe@tO>a}Si`#OUlDp=?)qNwZas3_L?fF;It47b$hKSh~jBf>-d z#?kjs2u9l1r+TNgVZt+(;6o`Ozd#pox#LDY?Z`JrYk0WOSC8r|@xZ>DX8tPegpdnl znMXAJ6Hl=DT4lH$oBXa{$jt!oYq9~GdzbtW#)*ts_CSf6%6DTLOeJTVkG4EDSP9!^Oq{oDQtt9!ZXcdg4io4v8PbT zfGg7|mu||oX`tq0_VVe+Zg=5&SMht!gmBT~2|NiKP&NeM;QMPr#A*Od5e1r*2rv(Y z3Pv~TVi2;V=GWh{tfhR&nOhn7ZDN(|Z*da|I6DxXrFjJi8=-x%lT<4-Gm39k>E0P3 z%n<5^u)tR&K_a=6HJ#A4CNB^XPv54_Jm*+@86i&47QdTFmuzVA474 zo-|sGK|!*iiDNu~z`GIp<{oWjHYnQ6@~QftU$y1w^L_|)kvFYcz{ueKzPl=owiy6N zn>w`LckC~NFIzFV5}+!-7333orv)Q?gKlk(AIL1;8@a$_ZJ z#C>GrG+}bFTU>E(*Bd7SzCZ`NoG+yu>4q*xQdD6wY9y$KD$=EKV`=w0z>4l;e|-5O zZsW7Oo#jccz^rgTdTL-pFM%}=1}*&Jjc9E&#p0Zzwh-aj6kRgQ%JYtRmN@$B!PStx z?H-)(RA8o!LpoWu-=|98YF8pgGEUg7lxPETby2;RF#*j1BEJul0E;}YDz7#7{66e!+;$g<;ytj#YThCd8yNao|RF%=LZ{Xw5zAjF^F>Au5dK)w#zW@<_C&5 zlPcyr4(H4B4oJ0)YPdA4v>+)&ZPC8aeiD8LpfJL#>q`0Sl(NlI;2@Chqf4cUhU9|~ zew5m87&}gzs`Kg!e^C_#tUmmcaq}jBq@95Z2wTtu|r!2K*QI&h2J8ZYAInA#^PB zs#nM;Jr_*f_zPgAB6UXa0iX)EOG_=D_)QQm#P)A!c8m_vhqlMj%)k$X0H#dHf zYl)Pr&yK!2_6%InU1_!$s`97T3wsW;j5(Wa}NdZQw>?{zbp)|*Li5TGOorZ~;4@fA&M#DTkrJo)U89Z={? z&~);hxWvU7z~^9^8qg7@hC2+R&uWZwomz3HiS!khS;8tU{3T{qE0f3+95QMBX?R1e z#r!3}7fpWXhToN``(}I80Z=i7-rRknh2Ctwt{%Fj*8M0+)a$W01h?(LBAMpO{%uAZ z!-ulOJnT%ocTYCAiXN!vD*^=7u0Pe+kMGjWVPBaOwW+^eIuQ;)Ucg& zZEI$NX$DiTj8tuqNm!;%Wq#@1Hp&A_Uj%5PBV?3F)owI10|UJx%*eaHC)73mXd5I~ z4nz2M_@%XlD;&%U+J!~#<#GO{wuQ^n!@>SGeloy4iI9b@ z!PVWy1wQ4=>1F53)%lK;Jui+^Fi0h#_|Tz&3p`!u8%jczFt@?Pu(+3!pLWJP(Wni& zT+gfhPpAY&$)v>!TYKJ~wOQ2c!rpd(T(SL_uIvYjmEPl5>h96nHX6d0at`HEwRqm=t0HW4ijW?{)eiUptYVtza=%l8kFHK3p|o^;u&**+(QwwjqNY zZpVeG>K;nV!Ymcp%*LLGP9+y)_?iZFFsY>Zn19ok#g%W{b%9Fg_j!s^6JC{=j^%@` ztFk>*xFDf@K-QWzio+rc&V(!nf3;CVoDBtx#59;yqm~=220_bac4zv;A(J1QK_fKP zwSO{0O?DTAKr-%RP{iV_`F_z;p*<X z>HN5u0F_(raPEjSq`}w#`sm;}3C$FV?n@wBaI^yE!f>hC2^F6oH@+05VxW&>($*fH zDM1l6j545#i$O=VjY(*;-l4uGs>gV8+#*C|FtUIZ-IN7#o`)_G9h!2=w zI#Xyr!A8eHoyJ|kHEAn5w6Dwi3>F@R{vulZ18$ye-6R{NYgn2FoA*HCu)D_BKc-_`|;4SA@y719-Y?om+e01=lQ;D4i1Afe6W?X zz|utKb=>@t6d2KGMl(n+Oa0f}d5A)@Q_eQ_wC*fKBMHr%CWYLaIp<|}IkBE>iXwXS zzFUAqNg!mbB*tOl@lSmUEq8kZ>KcjowtihAY!X|PHwi?I zUb^VejJQjZHaS!>YowEQW)`ETri(Al`<|UkaG;an?VMI>m>;-bR9Xzhd$XC>)wH#h zKu5o_I`Qqxq}$EnDvpuHJ!Nx9zh}FMO{{!AMt3ciD)P|ADhlqit}s0;{UqL zLYQPa9SqtvL??Bu*-n_+bt$zAI>(4D^KTj#d9tPV&yd?3+o^B<+qo=2M8sTo3umZ7V(HI=2Q7?re>Y*VTR(xI)iSmq>dNFTpPTy z(h!>@;cBC zt(-Ey6RhhNmvf{Wp^{$7Cgps|`*ws?%)=7qzc@ZlW%#Qm_1MeY>7>D`=md00lCEjM zY4c8$X)I&gBOqaX5kA*@9`u|29`@l!QCJeCW+l7$DG;6pog50!y%{e%z>4SXL(=g^ zBv1$k*_xjK$d1(accMIk-2iH;+hxgkb*|R5lyX}D+_%BZLr?o|*GU#o9%bUQ_)tKH z5R!yId2iS_Nn;@M>HGBzpKnw5UtH8U@p`IpIF|Jd8Li#MI_>~4vbQxP26fS9X4yC> zEnv|YY`il!r8oKx`7O#@Jf}FP^;=b&MoVbDo?gU)`|pph7{$R@u|7Z zDzlFBAXubqbYctV4y8*}M^jK4`IayE%eC9(=I-W~?KfP7O)PYP+VCp++Lm zB1=J&8D?UFdi;gur1-eEc8wKpci&vE z(O?n+zxuq~OGg=fDXxIIaTL&dEtPF-pGa1AP08-(;^=)d6gBQRxf^Y-3gGj~RkNMg z&2PnL>(`sR;+&n2x+S6blT=8NI!rIbec$`11&cZI8mvBU!~tV%r1P331E=sccyjLHrPOhe zA6`!Siv^>Ru%F?jEf~f)*8l$YU1`7zlgi&jjS%zQG*1FyH;lhg&IRy=`2qQBj7`>mVg1jix?!R?d66hVLIUB zMre>5Hp#g%Bv?wPzG(yss2z1-we|^_0c`ZHGlFM9fZrOQx!xn+Y30f*@) z{P3M7JFKTH+mhj>#Pqq6(2O-PlTY{cXDbHkrN{1K_u8>AAxobG`eQBJzDDV6nVUWI z0EQydy!@J!Y&tQ_(%=Q)8jc!Q*_zRUJ|GHHXK;lsN6R7 z(O7*|H*J)B-iEg(!otBvttIR>N3KOOevQcAyJTrjxj{^Tk4bBQ-9zv2**cIW!vN1p z&j~2QU*#D|3{#)?q$(}NZ)rrbhJr5%n-(g+8Sb&0nmGzVcq=kp0{$c!_Y(jnqCDct z7-Yu*GJ~idMM9N~2F(X&Y~@){PZWb!T0vukqJgw{6S8<0f)Yw&%x1|d8G97T0%^%o zXhX-Y1gCsaB-xA|vxkd??YMC*{TzH7tjq+(nW@D0n_s7P>TXTxZrL!5wox^VkR=RYY zCW^uFNXz(zRE=Q5&hkc@^_e)X&c8Uu3|rjl#vTi;H7z2MXt{_sNO7fi!9KpPG_Whk z8WwsY6q7b4dhz>7H3eKv_?H?eCF;{`o}oQe!}r8`Lyzl{jXxA=O~X>0d`^D*UC)_e zm0dV%q>50$?JHkWq7uJhwW4izX*3~I6O>EGZO$o7dT)Y@>{L5Z$^(DZiHlykfUfMjb&jz$0gZpZE&Tuq$ z=|F>wsI=^UKQaqX(A#r09xx}OgQ>fYGD*yEd2Ua5NF@VYtgD{`4U8Qdn~{(J9JL;u z%#8lA2Z++a^W#A6vd(@m{aYylr~HaMfIZ1DbDUW(j+^k+1@PE=2rV_=K#JqM$!<{p zGe}()3>2$cTM2>(Jt}zPXW5oy^5O`1<=LGCxN0;i!1@d6882Y;tQ*W*XoD^EXZ{_I z(=R^QBJQ=!vh1W9Az)#}yjcBP>wI=CA=OGAK<24Y9DR7b! zK$_psKV9ZJ>7gB|lMdh!E?WWuaSP&LlZlIpYK+?BkYhKq9WS?W!dxJ_Y{@^ra=R_xc3_i!VPosQX2PsNFT0iGb98LBVxVGUIL7#7p)l5uGM8If#?z zf;!+x{U0~qT2vwB!m{;YH7CtV8S#%rKDdCdvELPIMznzYuo9qndT=2?`-6zjJucCJ ziRri5Jq|kT@)thvM5XUQ?E0C$_&dtSa*8-G8G3@IwY?t6^gQrSoOC< zpxMKM60>jFgQwEm;5$9#q-7sOn?PDxs!)(IO}9ON2MR_jnVBG}2w&2iNGZI}0}2ag zhd+Cno$o9VYsLP&vtAZvX&rnaOOPZ#3j)9oi_rv9zpPu^<+^p zwo;IF)b9Mz0gPy`H6=;x)o~(GNMMkTz5Fg1Q}Uw%QvNh_^n6Ygy&I-B*%n?4>l7W5 z<7D6UV~vBkusESI8^4KW!ToDD1-2j%>)=yxFt4O)U$kTjH&(mEtRu=RuUoMP!v@go z3qSB1hwV;?;$|gve1X6jZB4g)p3B}#Yinv`ctVFS5I8|9hy2im0OOY4l+0L0*5H_X zo8GiVazC|>vlz&kWJhY*6=_&n5zOds*I5^xTkt?597R<1)Ek3Z@O(2Cuz zM6n}&BwM2Gzyy%j$Eq|t{tqj@N(GjP33Vj|+OzECcNKmrOfT5X-&iA%vZR0a&MPy3 z@-MWJUC2+3f4fw?H+I)x=Dti4@i!hMSlLTx$Ue+ub+b{)q#KLM&0@ON$tQeDIPCH* z_z?I_=|>PMpg#2%bifgQ5u|aio&&76%l7=`l4xUw`HU&dvK2My(xly~h}PJqrNvBf zA|=7SULux(F@GIDA72v{8;p@O*`0k6#GGem>b5O*GBC8a)K})nBth+Hoq!?kVT}p}Q6?H5)nXXFk;dMErIvxHw|9W~_ikPmEJ4$8X$`AUT z7A1eR)nw6Sk}hS7(It$ESJFAAePNk=VM4MH)e7D%JX|Yu>bbcS(qZflKc=*@S9toa zJg{ut%6-=b2itSsl*m|z#O|>an~?z{V>CIyfZQtUP7QFUEX+YdH}!UQH%qYUoP-RX zh>q&gM(Z&FVA)|mk32bi`#4UmeZB@u6EH>6>dHLMMjfg5rY$hAc};Xj@VZKMo+d-N zq;B<*S{>n^yOeV}LN)I89uCAj%XcUnk)SL&F1<4QK|E$B-HzR~O3cjXm$CCI&lY}4 zha(yySqC7|)9wwzIlA9ddEm;!o_wF4V!RZy{O95^Siz~R=w68;#kE^@g zytz6*pmgu^T$-i?>~Oa_v@b(Yw)xZLA(Pe5?I)^+j$oq{SDOj< z##JVuCd0xtQ~~U+;7sBcFGW$QazQORUt^%DS&&kcK6{p?cF=mZvdqo$sJm2=5+*8= z1DcH4L8IwQrAcI)H>gVvr1WnWuCkQZQCZ}I>zQVO1`A6fJtnG)TMfk}*eRj#psm8J z`o$2Ghnr71%wtRVbecoHX{k|bPrW8cx1!S8UN3 zFR+vXuGVxqzwwO@F&U~2Sf33xGE@*vKMHTql<$PHU`w#F1a4)=I(%_AU3%ivGy|gW zO^3e1fsKN{nrhB`YX0fobPj0hpxErPxX-dC$?33Ay*f-xsAAg$aP!YQXQ;FcEh<;1 z>9DODmDLC|Cuuf7G`-%%qYVrlzq}S!^@e!BNNMA43yIoyn%3`8|#faz0f1UX(nVx}cKJ8=S zrIbjb(^^8J9?~fMu(B5PV&R$Q8tBUGnY0K7o6Xr4X<;_6JZs=q%={uwiPGJ=4z10l z7D)GP{x%DoEK$e=;8t<$F1#ae4;7B*37PgmWpD8yu6h$-G1qhUX7v>_L>$Q(WW$W@ zB;4L7KL=Bft)=9d6TY>-n^t4c(O>PQwT5wxf!+eLv*{Fc<=K~iLiStkVs!O{x zTp=o&%r{p-mX_ro_u&i>7-6o#9(^YfCeOoab-UK>KT9w}U{7t|*R!v;bTgs-zxN-{ z$p7*`QHH|U%WwR0WsO;7t@Iur}Bg0kcB*45cRO#CDF0qLPu0_Gpr3xm~9sR zk6OCO@l>~XYeuqA1D&9DLCZi6LmExATwTt8p&<$0ked4?n!0wYCs`575YZkZ!x`^S zh_xiMOK`pr0!tc>Vh#Wae=bU`^PZ(54pcN~CO$9^%r+t$yCrhNIrmJ~X`@gqQ07Pv z7OB;BC+9*$>y#t(1XzthUp^ROr@=>e8T^aDdS>0R5WR274+3`r(NhA`XNNlh0mmqL z06g8*Xi$$W=bQ0Lg}U1^UJdx|H=SQM`L$;Vy8akb-#mwv4)6YxfxG3Y4kHpZ_5TI= zDDSb@u=MG?x>c_S6PEaQ2h@!oCzSu*MblZUQDRRW1NK$E?fz9jVvHqhV~kk_b@sq> z7VyIe#;4JSH8gHd1>XYnb#~7X9cK<$r(t)iJH_EDVb$^%D?38E$+K03QJQVG)cd3!&(ExL zCn*bH9Vmt#2>mX%Kd!InX{?6cvheeW-PgDv6%3I3lA{Y8)`9XKl8jmx5wr*vlT5pj za1fXOg5mCgAJ_iFYmlIt|9Z-2{U{wIr1dFR%-NefgBGYS8ykhJE)c!6Ain1#k8hUy z2jGhB8mD~i)Az)|5Gp?ntA~OI{v)wr*KTvOd-C_grBbL$Az0!bFYPay^S|pEs$ccP zloFcGp;gCZgnYnr+S+@x3+Nzr8~1tF?@wyBGq$U_;jdO~bc==<1b-n{=R`Abi7FU| z5e_ynyQ>KU7m9xjr{{ZCqI$r8VhABgt4`c?7maN!A2l#p*%9vv5io-}9Q8KyrL+&9 zRt5PluRry~8D8Xc7zv0Q+yuJ%P5p1qk|N&@Ei2oq%l|NT4$+x)UDu9nCl%Yat%_|` zY#VoM+qNsVZQC{~=AXC!jPd=Q-8^IL);?>lIj_02CHyoC`vdF-@7|IKy@@AvNZc@D zt8MpSxh=<_7zvy-@=oTc5QjyZL@Gg*kXeVwE$)niWw5!W&esQuQcb*TZH}?6$ zQu9-jmkba6@~2ORiV^LwplyR~3sk*CW$>!?UWFBW%p!VQu{_~L^jKXuP+WEWnx+&Y zy@HbXT1%~f=_>KYna9^-aQcY=4mrkmDEsSLlP^o(Z zgd1W)#7BTaY4b;b5eN$KBD%AYF>e8huH?;bLzICvYDbOyV6?9<@E%v6KDss*R&%gi zuhE)xbyfS)h$|hM)pEP@^L;QV%>~;=69UmccquTZ);#@E zK7YEsu{TKrP+lemoVXKc%|&jcP!x23iHQX5hq4`rqw90uKAO z+$>KJe~DP&l_aP&uK|nsG=WO%;N`!GapDIbEOj96TrdGACM(wK9v)D09ojA6d0=!E zSo4JV`6P>>^btFPJo_ZM{o8+J#fC~D53SHf$VhhvlLE5}R>IRX<%)S5+{k2h6(n}UN+~D=o=|U3 zQC^EeAyb2|CIS|>rsd+&DAg*r!l6!&P@6ASh@Y5ncO- zy5MEHc8>QoKs@>f0x3AJ-Q8~^TEF5M0bs(WYW$Lflm&hwtubb5+FMAL8NYEqw=6|#)O0x)z3{1B7*k^%z~pVpbWAA z*6XAP*otyUV%T2OPy|-Zq=-DNENtN;$7v5F!q%Tk+Zir5u^G>n@|7jNaPl#WP3WRt z6K(V}gx(APf_Rtm{w*POqUh#q~W$yFRu zih|3E*9vH`yx^u^Ela#)GBD*6Ku`Aut;Zf6dOHFF+}vCO{1elNhN$JZMngUud8o#M zJ`&GE9|*1YJ*oeW8PEjL2ZjG^Uf&kJExtCFfq^1`2X~~41&L?`QCXuaQ1AFV0CQLU zItxcOfK|v$$E{G}PF&^kp(l2)$rR8@1EbkONOC_(W~HpeZ?qhr6w zPAzf=COmK6%0bWE$>@l|!2%&)py3~02y98-2Qu}5-n(%I^lUJntO z7^4XNaN3d($^d0-mLtL%d}WSH`_XD|Eo~p!u5rb=Ocvv+0RK;}^b%0~D)lj5zVY&h zup@%gvWvG4|Cd~D&!W0FjE-b#OUuDzPABT6QSi^X@ zBVS3LaD*(#_dIiu^AZir!#XbWwpd!k-ty*W{;7by@ngynelHdkO=4M=4J9xM8Q-!# zh&rag$U#tA(!&In7l-gm_Y%%R-D;fel#??uG6@E3-Ymxf=#N(?(LN65I*tR)2Q;`! zusjQ(s1%1#e>+SUS6FX_RBXyKlF8ecmn71Wd$K|ASIkjN@d-kLU$I$oxk=M~qrL$I zMy^^c-eU+l8P;}3M_7m?R9l3p+$UdXAI(~>oXy_Mfxh9`CL&WDy=y$|tN zn_Bf2v(L9c=znBbl)yOYkVqeQL0;P3Ql`a}-ih0QT8g1Gs&7%VJ=S*S_C>-=`;Qv4 z0CL5fW332mq|->I_$$L7FM+YoGAho-a-r3xobewlx`sXpFz4hy)CyyXRG?BDYJD|b z-I9NcS>qTB!5WgZCb-PiHm021b#KI%FJ@H{CR;V-ojSv#km1QFGP zrZfisb;Th}zxFdp-hv#x9nmU@BN0X*rbCs$5#!-<;SGu@rcx5T1p;TMuMQv)uwn}{qOpdLsNNw(EwNRW;I56C|-X-H1;hwkPt+6-JOpAGn6fpyu zc4)4)Z_{8x@WmP*kc5fZkxO3Ea`z`;PlUNWxYdNa50j|p_1%)6y0bWuDG3V^K~rC4 zOxA={Av&ERSs#di!=_$`U)5D&saQbW7nA`Zukt37IIf)5i4-XYel%s)uePfVNLJ&BQ- z^gcl_nx0AnNY18SL2z9(e zISxUE%nR#n+E{A-1xVE;wa5xiQK?W~^udyDe#_spK7`K+Aus#PpMx?An)|>@(=xrO z?;y#?4CX}>Eq^|Ke5JkX=)nIa$FTl~9K*rL{C``vuC!;IHrfAnU(_skLEm5xbn6zH{PUP~!AuOx4VBsy&->@C1d$YWkk^&6p}qy7ckj z$y^YPVbfjWxZ>lj+yv;Y?%8NBEiz2B{o(a%_CRV=AYHk8-Oha`_1@iOF~42Hm&^h$ zHqXZPvu`MMx?-USP4T?BROEA-rKX)Plj57;b=oe>@iP^O+&u4@KE~nCaGKZ#_u-lz z@3t0FFl4@A;2Gg4vs1rcZl|u=8S3zU38C3) zu)!&>?iFaE=Z#d9{c*ZpeQxn_QTBlbE%Y4OU`U!tLY3;Nql);2MYXmi!}+q;XL|7{ z3(erT$ukj!43cXA;?JGE$~wUUY9*!rNq;H1npqYj))V7-Up7*waI&^h!oZV2kSal) z*oskimB)P1FA*p{3KK1+IE{gR#&hOt*`@{rWDPc9K)i~0L&pOupre~imr+byzuIEb zM0){ys~StqUS(Z-+C+Hct=h~{_iX%-vo(Y3D@lXVk=F;y>Ndlb3TI6li2=xHz(@gg z9Ls`KoL!n!t|^)QmUU!v9~TnqTe(d>xqNKDf%R5-2)0{U(dBwc!hKJgW|?xC?S?!L=S)StlwX43W0eJ;-rB{oB#haqK|I6G+uffvIK_U zInv?^LZOmqCT}8{-~tYYrLcHY|JH~vbhwTA!g0gdu6w>PQQpxz5K)j^Di*%XL?$?woxE2cs_a&%=8mtJS#Q6+0O2YmM1)xSI#&h%ucWc(-M~M zCkLW?P~``t04*$T&OmAN{mVDOY3e;p+7_m3q~VF71hw=q9KHv&ET%&03UX*{gPJwG zB58&B%>t^Ie?CRxK2$tG`hi=mLZh@` zo*JS^q*hi^g0eKA2az-Uj>Qdmqc@>xmfj%q)-llL3EDakD5#Oj=of#!FVsdq&A@h2 zWhDdwayXioGSGbfEQtH?CsKJt=1&uD6|grVhKdB*e=pC*jlW$^R+uF@H}HQ{CJIQt|T1_kS#clCqNJLdo-rROMb@_57M1}LaF5#XF?K%adH zb`~3`#^VdN8j(@mxc|*YS5HM26C)*F#yrd%7=aeed{}aj;yy--pO$dlpxY3DASL9@ zN$EMwRD+Fev-uhw-4P$%!$2R25}#a%=kY{6uoJO)-FV4gUD@p(vW4fRPZ#03w(BKj z;K%*2bjkP$VaHYZvkw=4=%o}ajQ0k5rCD!36P7R#K44-9-N87nz95Hpnj0`zf=ZmD zB7*x|R*P)EEVv~0C4S%@4>?|_2u76QS%-iez6&hq7zpoMExZ_rAzC?Kp<(t-)^?@Bx;l!N#TvcK7H?P-*});6%S-V1Zw#nbYyt+eY-6b8qk8hx;sbJSp^)|rS)H7#z4jwydfZq$! zp%SY4vVfM89YE;fjDLMZ3J z;m_n7#AL^E2pVJbV6oOBTfqWE%IdO2q+91-MWCeaIpQlbV6d{op(_Q3Wruda>QL1HvYM0v z-HHJ&%BR85hPA!%5IupmLzDSA$lR2nIp}Y_6_fg5v)>TN2|6;hbEHIP*7H2m-N1s~ zmL+336D%4xloPT{V9!z(MphECTLzr@bVKZQ+0_})O-2A{XB8^x-#-DY1jb~o$@yBV zI^aW>@{OOOa5c0Vja)1uYY2{wy!WwNBcP&(g}ot4#&5$e7Q;9K*yX1s zKZY`DwEa-S9o5vtGYCfQs}B`t`^?i@;QawOTA^zZasF|XURN<{hu%fuu}}-ZL1FYa z0tdFxmcb&O0|e~F39ad14EwnIUz55BRWYD~e_5+%EhgGaV(^T;yDIqGb+Q z88}0VaaWu$@CO$W|2-33Q8OU#PfouzB*lYW=$-;wktVm1E+qWE)16c|RrSZCd$CE< zfzXzZ!^7id_}s~4v;2JnUI%ju?k_jQj*X{4pCp~FhFUP_x9djXqx$W?Be9n{jlYQTQK7Jg|L+`MVIkgMqowi==FJv4O@AvTHB zh-lL)Jeo0P(u{cDh)5`%-UkHySqSMgsMmzsSEoo+jB*T`*xO6Ev!8FXt3Nvp4Bb3U=HNIyZzk#j{xo|iI= zAE^TbO`|0z7nh|G6+H2q?k#l`*p(_ai~(R4w#q%zKCRQ7QVJN|ei>9_@0WLx6M{`Y z@HLSOa$Z2;4FS)*!e&g&P?po)r30G_HHk5~L@9IaxGL@X3D|mlSh!h*&h;B^w#L8s zneh4%;889cp8c=NGg>GpKipc}*0?3H${%udJ`E_M|C{&M*scZ=9%P=s;(8ys< znk3&V{cc}_T%)q_rpePjxVXJr48q2by3EZO4QbYoL-B-^$bJ@uzB)Yx5^!38b%00Z}i%U!Q31v z7uc=$3DW)8fPpjGS?C45#w+GVXfLX|ZU{*q#klW5hw(`2m0hI@^#F~r{;~~|fX#k1 zj-Z4Bful(#s)$e57}@ysO=eQ_#v+d&ROy0yQ{LA`s{FAl)84rEumf#b9cN9O-(KM}3oDm7xFwpVqHpb3{#SuErx;h~Wo#^qW82W>EWIC2aDzT-4d4SwS{l#-?WyWc?(-`8ZqVjO(dv)--qLM^~$cu6d` z2`|j6nDF-S3EwflZYQATQF+|mG^tM*=dmQf~71AZHOZv%7xCHcm9r5 z9qw##sQtyx#)8}mf&gyX6GLxzC>I9C&JgWHvA4QETkT!2>Zn3h?~649R|X##0hFGu z2}Kw2E*z;&=GmB64`)$9E_T-z$R;!4(Aib&o9BKhd-F;NkvQ>xIj9syVy#a|R)$8J zCDj`hwl_HBd+7F+nJ}NOmbxEm5PxY&F7&GFLs-qb#8AQtgqcZ&PnGjSz$= zQqxF+vuBC3E>obd@MXV))`*#euKfpR@h}*Be6D3Xf2z)l;7y=K77yXKRN^J$K{my1 z^0ozha<6Cz(M;)w2YOOC52m+5Hu;~Irwqj?AuekMcbu^d($%>5!EuZXyypKPG?RX> ze8uB%asrx7X;pjc{A`AV`R@eSZ@X*!2}NnTuirPWhyI%DQb7}P5WGx{DlnIg(kL59dY-;*RNLLK!` zfQqUCDCPJ(#ahqN)W-m=H$k`rx{yWx)*SK_NZiaj=ksP}xjy28D1OPy4@^v-py+=v zF?>nS|G~soLT73aa&*jHPEm8qf6`BF$BI?!>s$Nx ztc{u+?{6h3EU3GR87Q_eS9?!Cz3y%rNUED!p;{V9GDgEIs5UCTU}ngt1nKu*n(5US z>7|NMnSFS9%T%znR>qe4auHbJ>|el1U(SKlvg#D7WSRY1`RmFb!xug?m->zZ9a9KZ zSB*`{kE#|1Jz^bw8MZ4IkNmRnG_hEd+gc^<9jLAqvp)A+>b1}(NTub@=gOu@(?gEL zxd`m?atD0Xl5EN%*=cuW5%9d})zCje;JMNa_3{`U<@+s#6=%#|#wulKLz%8k(`CJL zek(4TICQ~mQph%~_S!*5s_G?iOmzQaqsPTD)>3)=8zH|AlfxtHRF5^RA7qNpzPN@> zv;HEsPi!t~HTXXkt&3)ejV`+l7li}v8g2YMAb*tgZn*)AU+wXdGBx3Fg&vf@sFi4YihR>Jf+n$7DcB0Ve6C3Nt|_k{0L7cCB@!oKN{uBFmP3iQ*=ajznmJ83JTg7m zoK-D|{9!60L%&pfGF7?52?%>&e(pJnc%uOZ^FpaO>4mW9i_r_FLM^s2F(;!B?#Bo( z%_2fa)fZz=YIwJJF$%no-7)hIj0w)(X1r#TYbY#QA$z2~wzvO4@yXh)KplrdceR-K0gF`+{uIQ*?pXXXm;%T#^0_Gf6*$Ymor~YQ}uI6q4t& zs!1@;Buq2vQl3ptD{qAmmq1xx?BXq^9^F?by&by=Q0S;qLIWqLz8Zc{fE#AzZE==g zpk@P{Kg4poAZe~j_9W8R-}-Y`_|jYeG?f zR@oFUN&f!ZsJz|clRlESpK!a{*3&rgwCzR8orQ@VmQBS(OiA^9#*D{r&5)UL0|dR@ z@;TpqnTQQk)xrSTsxwgVK}^t!KJ6;?-6B8DO^#HV_b6fzN9~H12 zZod_M_jHy?5x&P`2S;oNlM_)FNA>~;LQz|r9UcFtf0OezvkBe8aK20e2>&2n4@@RN z5flX4$d5k8EP}Ok@Vq3$1vrm>;|S+->`&|1V2?9ir@Ox+fI>pUb8Q(YdZX``DCBS!|3n)ufSC%1&@0{v6w=TJ{djPiHcScu zXF@KIhWAcpUHN{=vgnlJj7(lX-VV(- zQJ+&AI=XSO7Vi_E;Sg!7;Is&XB~-5|3O&%su!@1af0IV#ejjEwHhWBF3wflv6|_{; zMtLh?=C$gj+ouH=fq*d!;=v?|-J8ZxxuhGGOPG<{Ebky-9=A_h5aTFleJ;DQ6TX6H zmabC6y8Bv=O@nzDwrU@agJ=I`+N;0FOe+Nqo?zfYlZkcYw$xQi(rX{bj@4%Y=xFvEhhOaTRStK|bL^h@K=qXA`*` zR2%b;wQIzg1+c-qAygG?Ipl((ZQfw#I;)K?P@B#m1{M82gdW(u>r9AkJ2Jpw89naC z!g`jF$ma3coy7Q$#r7Y8&{zeMLRq@_7^6Dk>{jboxm;TeZ4IY2R2arHR4pd|Tvi|m zD89r_L?U_FTmI}A4+7EZ(p{feot&p(Y6*aFICi6Y%Dj%%rQxT20L_)?esO214np=4 z&QSPD0d;P_8quP$yi#_tg!xS#$}L)H^&6r0lVOU{;Q#}K?x;vR7ws93#gp-y8NgR# zcG?`hiY2Z}I_W$tX(bTzJRY-qttRa?9u%@55F!ClIMaFuRXtxUYc{0n@-;!BEU!7l z{towHSpikne)+&yQGH`(6`@phU%(WBM-YuxO{|HoL(!iou0@fD(5m*lXG_Hd?*%^Q z<@ee$t_)eVIQR+-OR^GpdGk>J+n)htg0@dUth5XVkoQ*v$%dF;&} zJ9y_%b}#KvkixvbXuG*??ko$V!-P?dvL6*;KKFJSLndW6;~r@-ir6Vumu`Z?U{7jtsaO7!ttx5aSuy7hN z_anN~IE*~@37fGb$%!&eZITcqnDA>1expNpLvpD`#*u*_=G_Ra)z2CSP?>KRouxv# zr32|g>gV-od=;I)WbpR=|B$ppRqpq2yBtD61L$;A{W0na!>2FPkd9<>uDFjUROk9( z)<~2JDnUrHg+xQTXId^a)ydcUyM=%Tof{X@^Q7V&jvIx~L5TwlwtwEuAy+hZbwp8O zXwm^`o&rq|=V4`?NYlIyotg@lq0^9@_f$uhXpBIA+`3WZa z3elaJMon6kro`8%J@iDk4DqEfW7c&Wl^t>F%`!iV)np`a=gm8D&B(=$%3Phw^xV(6 zxD^|L2$pyT&v-@tOd>3mB*E1$*~4S25|E<6L7;vL?vY zch$6zYkcz`DU};KLzt=$e z6_RvRr9%rM+WdAv%I&X9#;txJ-7E#KxoXj1R4c`XzA$YhO71k&2Cte>)k>d%j0ASx zFLUpX_GRQ{e=S`0zP{Nklk4PA?Vo7ay2^x z`zlgU**&F@C-HLbp#Ev6#5F>Z=;=<=kPrtM#LApM_PjwOS}>$Rdu1AaM`iG!yaHF+ zLpL#fGDo!jQ_!+AMO@g~myeHoJU>C_R}-Zze)%L;9?pb#@a=j^`}L)ivb?b|yP`Y# z0Q}ZSM^g%4puh4(5lG#kqjxEV0ee_|OY~!*c!h6|9To7p;OmM81QpXVAb+&OYW^h< z@(bIf(TDb`Ta^MQ#;uvAA2|L3&P7RAkG)rksr&c0Hu!Pu@InDZLuX`=phaQl0TQkr z2?HCQ^IKicotet>a4{{my*n=anpJglvP7OBBHvxPWze_r&*HTM^3I)u5Gl0I{;yJ- zjL$=CN*pe+6cqn(_MaINB=gzlWsQ#V*S%|lwgNkbe>60XNoUc>1xkRM7#k3|r$r8s z#JX5VuHA;Cu5TphTGGFCiI+r4|F{fJUJCl3z3VWm4h-O9rXV zvw_6@i`Ms&SCETV{bW(HI>*EK`J72RQ{1NOZQaHJ2_5{mSO+-pTQX-(tOw9W_229^1L8|6Ux`=mFe{uUfWNTdL>AbD~u|_ji zDhOJ)o%n)K8RX4+A$4@r9J8bygajm*9>8W0L2{=B33S8V=xrfHi}MxLG5FRqXhnMa zUSr)nRPlikGBZSd%?k8P+2u({lK084$jOs0>*d>aNFQ`9T=p{`Vi`eD&Sa(g zDf3vq6$G}C^PktgonVFD@Hqqi9?0yjvbZVq&f7R6DdCx28C;+psaCtl(Or>REcJl< z0+H$(7|JU&Y5(xT%O! zE6m1@R6X`9`ByX`{277~L*tHWa*Tw>E!G^gh-+$dp!108XXV;?_mNM3oE_+N%CD3J zWxfl!YdqB1bJ5_P7nE_}aVgi0Q3=SZJt7_9@T;{0;d2UksTyu*=>8HtQ*|Mw+CW6gfXW;*Ku*wQp~gZ&d-tV3jqHT4X_5`&Ia zi3Md&aEj4vM#XKT_ez0fSDB{?gjS79cyNb^5C~i<3$!u#@u+Qw2R(vl3f_^cxfFDpPp(IFcN$pR|hCet{W53vZnxqfd{m#H7=+M zuZjD=Ci*|HRDF*)or}Ul*6?YX@rX2?R?z979>|EtgQ~(k<@-}xcPKJ1^S`_lTNcU? zq~=dkmJt=uYjcUT^_}_Jl4{L^zs=k|y|N>I%n5OnVDBxZU(NHUQ%<*8D!31Wqu@6SFh zAxT-^n+#{x)*D`afgWnMX`#D3ZzhKU*NfBgV~f6&T)^B5h@Y`9fgbXwj-4jFifM?< z3PG_3ZKbGMlwnj4*`SK!nsk(%+jmf(AW|qoOtF9e{LPb6>B>jwD>)9r#_148r_+JX=I8T`KU8{1B3S@|LYWa%A{U=#+yjisW^9(udcZaGO@c!CneF#NNYpf0 zwf9N8@zm_`b~pwvIP}Dwu6{v5mh2SvMs|K z5m0v4ED~V={5t59<#j#v@kl9qs3`rtzq52q$k)#EL%6@+D10$%;_`IxxxG}XEaV+y zNygLd;ejGKjGsYGrg1s^%ljjt>DnO-t{tsaHY5%`3NVh!4x6XD=W<1Lbn_uG2t(SC z^ZgyvNTwq;uM~!HZ;ATQPL2;G(J&SEH|%moT4rFj^W*V z_&Kw07g@ZlrIUgfNycy}wxy=KH{ngy1xk!k!ocidaTF5ij2U4>Y~Ph$pMzE8r!LSvRq&OakHEfATTEVHzO z`8!fd`D`%rukE$8H6zOh8wK=d^T&IqR_z3NxMPxMyLbhvU4!2F%5z$OSrRD&`;cxE z_41xr#h`2colP5?QHy3A?mX_e=6$V9@h3K8=1*87OS6qk&MK>bV-d*!lB2fzOg@%IiI@02ppK}pQ$6(` zZN)n)o0gV5Uq62ipjNXb;TAXXl_b@ZnN%+my>9ZaXuFJ|P-C3E2J4}N=e|Sgkg!Ku z(Bu9=$@X?aK2^%B#r$_n4JN*nuGlHXn#Om=L(iVOSc}oo{MH~1KEbjpvSm!FT$cl}IH=JG$ay{soD<*3xVu5zU64Jaop(UHz z4v!8ZbuyMr*n5sbHl!33HRB~l7D3a(e{-d%nntsw_q7!I(PzE9RtE5=Op_mlN*Rw7 zOHC2QG7H**JE*hcwWcd=+Vn7uKbKij^sa5c+KOHKS+M$Q-uRk{TfrVf>eD&p-+aY8 z2RUw2_XD>fJNk|7FP7YenW$H1Oze$$FIusDvl8HV7fTu4#PvZMX_k4G z3EcLk;*Z~VPCD3d{I@sxn)V5fDk`xW>sctq3E9#oM1M8JCQ)4cuwvWa;qBGDLfW9> zKk~#zxJx1J2CWQKW6M`L*HS>!TAm+4T?*WZYSf5w>dtnqmw}ba;iEBSeH!RFoJ`9I zl}>R(4I0&ge1r-r|0QzI zo(PXmMgx&$R9|&s6sZ(XMSq3A9# zW3z&WHZay(S097JDk2wochzBpK3Sg&;{Q-X#ZG6`Hi0WLErkmTr-5$MDD;CCXJ$*z zHSgHE8Y#Qx3nUCNIZT9z$L7lz{JqW5PwoNykqXM$R!)2@)MH&TNDGlcs|d(0KgLb~ zNZA4ZRub&Ho;^JP{w95RC>4`CLV`ysV#bPUjz>qRXK|U)>kN6ux~fM#A+P-Ea7egZ z32>Feh;UHyNHmX2m8WCP)Hody$h2&t2gYDmwMINq1-&FzX(GkSn<2Z5bXX|?`^WXC zR94Co3UcY~GvI@H#QK1e04R=}EGSzhe<;$}y4BoOsFWJB<-0>y^|5Nrs_OW!Pl58N(!MBYq_A(44~)S#f*zgKluQLBay+`ty&|N5SQL2)KYHVhFPaxJp;A-veDgaARp-PO+h>Wk6H zBz~;C(v$(!fQC4Pr(N+xt6Y6aqy&DxEApX!TTV!;C6Y7|x;6DrAR2za82Cr+cQ-GF z;rH2bLA?5BHpBk|v`(_A-=Z&Iq8NS2c~oMRD!J#20_(1kjvKnteUEbegiK>j6CnH? z)smin3ZQmu^lA5(0C*PNDd$jVv*Y9W#<|o#WEuYx(B^GM__)U zL0$FcQ|v+V;1yi2HB^#)w24|zCL4xwtv!XnGk_zZ#n%Ks)Ij178Bah435_KgOR}UMoybEN(HGy75 zv@liEDd*Rm!Xo-PANrou=T~U{tg0P*|nKOHZ8V!pLXLm z&@nP%32`8@yb?zVuh%OCb{Xf*Z(T)IbfrkPgOJYQe78ZyjvE5yY1>1u^b@Me7=$l- z(G*&!7YZ;cF2PmhRDtZNb>cyeuaw9F9_|{TIRQ7=atLsa01Z>1sdM2g2Ebly$9@W- zDY`y#uKC8fZ3k70h)~V1UmhGv-t3BnNvlWlw~mAzI3)Ki8*iXQqLO6GM+zhDT{Ni- zlxzkJ(@}sZXE_7d6`KynoJ&a%brzHJ7$*m!hL(POQ-gCRbxN;^TQ?-2uzvhxXlqX4$SfTc9|8%9+<*2SyCr?$fx8Dd~+yL zDtu6lD$KXD7zvPPnc<62etkoL^?>;%^8Kz9d-j`vHEu0<>AI;DTYK!e^UX%0Hl-^4 ztWR>~#rz%j;@^vi@5_|mnMGf+tklI85|=<-`$yDJp-|O{96E`?*n)dl#uo?{+rH3_ zYzd+x;OJ$9?R??+(Nh9nXO=5CK1NfbO#UHqMp!&cpJWjyk}wHC+- zH1YiIY$8?Q7R-8Qqs2fbZ%pE{1)>!V7ay6U%fOy&{`2>CnvGUN$)1Sv5OE{Yk)+8ews{56C6!=> zeM{B)MoGtEQOn0uC%k&qHuc9n5IJ&u@@y%#S@yv zUYO?ZWe;DvQS2W;8XoBGjHy^w^kgOL;m^);=Qav&+F=V)2m)mA#(*Lm{$4cNieMiN2tQQ@wrgc z%C1McVhG`Gq|oe@fOatCb460vrYYUrW_uTbWx#1m>7O3zq_i@PQ4^6Ghfpv<$kt4Z zaQxA?SL@(RDNI!Xw%d!ESeQ*olQD)Q$G@VP6jjJVqw_P)5$>J9{9Dn%Z}RpGjji4& zNpboWABu_>owhd?&lL~^ec*V>< z9=IhSO6-63EOS=g?S)yiI0-E7@j&k?@Sz4#OJs_vxtrhpG8}P7Cfv7W4IcOhhwS!; zKn?c%=x7g`QF&I6176Sy=RWQ+>$11LRaBNw9tmG8Mm$!=xQee#@TVu)g55Okt!ku? z0V}$@lL^TVU)Z_ zk_?#{n<(@=NJ`WyL1JyOr~-q@t6N01?G_`@+e?Wr6M?K=JA#dGnEGWg!mQV0e_jOw zZp(9-eA&20qe?|Q5Q1i&kbD{a=!AqeCp}8Cauj>Y&+W^y%FC;5p=+Hx4S-|I+Ksdq zxW{iRF^?F~^*!SHBDb5solw!`Y#A>~%uf|k>~xLyI|?i!I*ihYGPrBwBJCn8=-wW~ zG?MUu+Oq_F=6BKM5KmZEBD0v;+H+Wv`bWG7#hiY(&+u+IylxD%mo7zK&D|5_(omAhHL?(u#a3)%+7(zDl5iwkVaRcN#v%qkG^ zVx!<@5K1e&iQqQL#Uhcw^ArJ}orS4ZF58k{xcJW^?(8TiQCa~j(a>=%dw=)`>XgximyJj$-lCpDqZuC#FiihxO=Z2aO&9q4Fa8xY5#0_H#d_$j!6Y5Czpq zHG#{8Kf9rie;Ww~cm~{nO!LXvc0K!Ir zm}H1d&z!Pnsm%|16Iq-P9oG(E$KJ$gvL zAK@aEdMt{=nIPF$k_mO5YjG%}Ot!^5&>{LIB_+N+sS)aN4!6$#1V^;=ul470h6P@2 zuwp(GbZ0$xm^~#qyv1`0edVon@5R2P_BCFen`epy5{ar@&>G!AU7?D*+ojYW4zI(}xzEd%4A zzCew11cLt?pW^s0e2R;M?SJ#B37rk6^+Bhuqs~jul=0$5qC-TmoV2um+^x-hJS=S#7{u;L5)9oN zJUtovq7iQ4eK?W5TsS^08T-E9pC0d5sSjx-lhbf@y*liB5AU%?sTR-I$<#~gc>Db0 zM>(>e%E`%@tme;8!hM)wmYy5P+giAl{RcH_AOfiyG6+s(9ex{HP-B zDbzV3N!>LDhz@T-TA@Z0cf=iOlA-nKgwbUPR`7H+&^>Xe{rjeAdGe4~iRM&qWpM)k zuUyu5#>-_b8NE5mIO8X_00Dt&C0xzf`RTE}1P&OUe<-*Co)ItsGxDEOB7RMrjZFj9 z!Uy2=_*RHZMiLz$J6@ROidUp1?&L?gNll%UW8oYnWbzHDw4?Vk_)i z>bB?@DN;Al-9^t7@^9wSy5|5lHkEWPnUoel5oC8G^14ctPXnB-isK4m-hjJN^s=Gf zWn97?&8fWX>5QGK4GC$IL)TZ^$_24^%%39pi2PQ1cZkC}B;Qo&o4nyQUzKUth>^l& zbhSFmF)vho?7FQK0H7Kcjoc+~VvqJw%3J10MB`FHbd-Fi%H zR#Lr6_3&A_O96%!S<#s8^4QDD3O1P$dd|SYcu{_a7D_AW$%T{t69a2tx$V2Xayy6T z8pcGln{5&q`uhm&%84wwQe`pZ@631e{|!gyy4|>Ntesz0MoWIV-Oj#QKvdk)R!GOZkA_NGnZ&DKFX4 zEg$|Y_pekQ)ZIHyy_|LyGs=&on4=pnrm#RU@fnO8i1w#MViOlqlaBNWEvqVw8aP7w zoZXWX=kRIdb<3OlvZpmEnS^w7B(woPPseXrk;N{y9J0y2?7Hhk|BXFosGi zC^%LkhoHYl+^hUSdlDXQ!^R`$J|?=}LdvX!2{1G*N1Qa}REwe5q$I3w+m`p?JY#2G z@<-Ev#S}z;%48~c1x`#xAIAqT2vKf|>2Ts=3T5PdVB@LT5qlU^x!8m<{L;+|XR*LI z(+xBrsAJ8Ey!4VeDl&@cX~1fDEQ(|rO95&VlK4pqW$lO=nh3s(Q5{-}2Ezm43(22S zNum*s2106ChKcvqUzPgyPlDn<@DQMxN`V5d7RBHCG~y=z?(ARFK%Aqo+Tws*PpA%V zDNRLti|CXxbYS)Mw#9ZG&KHiI&9Bx+RpGInHYlR|NDv}2+p$9M9#E7ba0<|{b<$iV zUTca5A=-vqGE0%Q(`2#DyrUA%w)5)7VNfdAh5`@pxuipkDA3cI!{aI%*A8 zd)c?bB|;JqgvDCAJkD|O?7*n_9t2HMTV=ZiBiD3~{EpR4@83r+fK0KOl7O7O}imhj1 zzw$Lvs5kf!bnw<-0~uGrZ#r+R34{9gMD!Za0L5uD(7w{%?YK9 z$U?v`P7+O|OOF6>PXE7GJa+qrJFZE0k;YjIdP@YNZ|zyG%H^W#mjW+r08zTf2f2So zfpibZQ@-{QVD0p2M*IP?S{tui=}XRtefIXOX6eHY&?wxU8&{8c35xz=}0Rl(R==IK^U zu?@xO{-0?PSMKz~1cWTy>zDi~9}I>Lwo6Du?}BI;;5Vnorv{r5_D^R%EH#ne<6Rvd zN|XG%E^In2lr%ef`oCMd$TR-ieS*`a!VSN66kbKyX)m z=+=IHcqxZN-{~&aaGn-Iqf;7`w|U@G9J?d(@x_Hc*_6@=yJ&irN&zrTw16Bt(-NF z`+C+!9%w!KMJ1#uX<%6p*y>IQ5|S1*n*@^MO;x-HHvJWB8Z7U#zxZO4Z5it(hbO?? z|A$kE=-# z7OGe(BA7v72b*dd#B@(_6B>Oz;%^e;-7N^*j7py%s9n0I7nyG0`hx%Wd7z9dNN;uj z%Tqecp;Ey%*Vj!}ACQQn2wzDv8$aOtb8p$09nfax>^+Ek}@oy2jYd{~|=agoe7$i$ELi|`m6xE)i}Eo0nS zJvIfcMpX8n_X{t{IdtE^;j)5CY}jyv(`1IHlnLPw*FCv&8E8CXc!S#TQ(9Q+)MJ`; zxSS#oTt#j1=&vhp(T0!yo3ub2;}QSfPeLPJi}XVw&!fCx2D>*4j|W#V(RylR9c}#5 z`G;kC!;>F|^~|2VWIEiuTLNPg+aba8#b+Chdqw`r9}lb+#s*f0ymQ3Xn!ipJWeOSB zLJ4O*Z9)jBj!TDciyS*l8oDW*L|X$<>k_R}i%P7UH0v3bcTA1XPxT~%XMDStbJ3_d zM58Z_m>LvO(I+!!qwSiL1d&+^1 z?&Whxc`DPhKEr+Lze?=GU0x*--}9c)yZraBeKFlc&%)?riw!8vVAib(=Mk&_eXbv9 z=^l>!JTPT;?8lfcV;%ZT-tWWsTwXRyreR4)CJqBpZ-BK2Uw}{j*QhL!|H%4S|jXva7TeQu=ik4Mw0lCi@ z>M?AObbyaCGW2yE&goya-Wx8^o#RTw#%~3iA|2%DS}VV3GfS<>LE6Ika19aB1Plx6 zLzBuxSZ~of7&Dv1#vRe3x|ygnVdn}0DBN$t-ckmobc~+y9gPkHLwxAih#h?XIPVNg zd0q4Csk}?~D)Pl(5HTn}S>BY}MNZfkzWln|S_?eqX#ZgFnckI)LxnaNiJ6AVcb__Y z05Lph_&hQFoR6?r+`6^zZ}-CRD-nHpX=IB{?ZwXx5YanmS6fkQ=+lJ~er2e+^y!y`}rg^&ZfJuQTN?}16vcfFd?`Dm$pyW>$;)a!LVyWzzLWQ9XGm1{Aqn6;CXVpS6UV4hb^NU*3Y3~>xTL%83>7}T4bZErE-LvR|2_!|U>MDy@0 z<7FeXRuI4636;mOS49$P*#>ppzl3p0&v_$QtPSPU=5bfpU2`Vmj2s>Vnk4H=8lj0!^o(503A;R9j^(c zT75oQy@VB6@vZ*a?B_K=OHx)2#lZ5ult^3ytCBGlZ&;C8n##r`lR{E6FZw%7=MLPk zsAWo2h2!X6mumnQ``!|a)XjW3WDB^-iCV*!h){)jc0E(-5Bvp!N@yy~Y-03JvBgu%wve1J5=heIrzs_KF~N?tZofG2?wHIT$Arnn*D&e zfRW+FhEF&RT|lXbbUbHj(6wsQVhgUO_cV?8B$0RCZ)E~h;4zeBn^O z=>_Bcfee^nlScR{J2`4(mlAV?K$W%|spD2I#h_$E8eQPQneLln19a#cR!Z~L5q@GpKi#@7TKxQX1IIB$mdTi{Ps z)!7^L)ca(YcRaW<^&ud#7-5p8=PAo}(LBnv{vHF`kwilPW_s647<;gg7P_jPn(TQO zi<4n*)|pjI-dDf}m;ye0`~N-vfDCDSN1mrzkI8%3YRB!@^MJ!<7&2XP;0je{_MJ6< zwTl+DVg5;&>n4B7FmkjB`{4ITCUsFjrV=<5O5tN~xhB5xlZr09;g}JMZUFLI%u56P zkM9Kwo$$=Ory_jp>6by)2aBdz#zai9FHt8#DD~X_7yHD(rBd)mL@XX9Q-&EwA&3dy z;Cw9eQLtX36y^1u75l#c$uY7^HrLPBTj0ri*3u6#seu<$h^UUPlrXx5plDjQz^y%R zr;>#YyP*_5hDfk(p$|ajRlzICi1TERcM~N;_?!w1KLNRGrjN6Vds+^jYDcxOyWt;& zzs~6R)m`J%dO~p_9W|pLtO{ZV9J(4j(DBHa8FVI#5b{%);BtdY72V_3ljw$88?Zjo;-Le@u3CQH#8i;kr$#ElBQtOvKn&=-QSsV_}Y*lmZ(=Nx1CCz(H(G1Xcp z*YewqB$l?!cZAt@TL330+@LwEN(RAoSeZ3%KaT2cAEO-2_>W-mv0(wnzC50=xRx-= z0zv2%sBYh_FbRhVvkd{4JCAk5kp6-n;V4C4QSH^?UArfnRakjN^G^mTek}l+vhzOq&H4T4rcV?}p>W z;V%w=FM|VW+i_WOmEDfbKJ){Hl6MhA1thMqwhjr5G_*rdmybV z_v!cIWYNMZU8K}i*HXA6YuY%v|EVhBNax&dAm4{+V|e|kbMs(Y{tv@`tbHVza#Xal zdbQ27#~DRuSPXBIZ6Rn)tZqCc8u1khy&`1E)v5IPPaH0r*L2CVJ&Ul`k z5WcJOP)jeQ*O=(H$#+`K#n_V2M=m6QmMWr9X11Kz5fyp2Vi@&$uwj;)1=pF~-8BLd z)(?a~HtA{*mM>(9Ld1;Tc(Ui2#8hE4cL2sUVe@qDG3H~<|0pjHn~!{|?zWG7xI}qw zs9s#4>^3k9NzmG5dPymy9=w~eb!Y!gdD^%?2;$!zCkoKuWbn!nkzP75#(!)s0$&zx z?u0Hu5vi_2D?$1@lg_L#F!q|7lN(x_)-86wc1?}Gu@W(x;McNDg&td^r-0`os= zvjjmbIV+^vQJ_;QZG4d)GZCI5Kb8yEE&%Nj#J}gtr~oh(%&A(9YMt71tXvb%sr-WI zj{9dQ-D16RQiTmVTR#tl+IzrOLXn3MHSA2I9$4q*7GenbrE;rm8#4X@iV5BCHFLSG zx>Z+mM_-t-l%&kWI+y8WGSzOK66ePtV@?5eA0pU^k`~rFBa(29oX5W+0{7{>e_6^b zCMV)8Cq^fMl{u4WVcp{d3gvE1IVnSv#m9x5O7z zfGCG)@cX%_Kn!1vaTqlh;kzhN0g#`s642YA%WbE>%s2O>R<4O$c%?fiHu914ZUX3( z-D$-(eEgBbCxK`7ZCdaST$7~sb&;{*bQoiQ%G%LEIb3;rCYBZ|_0fL!{Tz9OtK)-e7DOW&;1vg>MmaN}UaD2)LMDqM9O&mc-h=W) zQPx_OqmI!mj>+cP@mukAl#_@Xi})^`SE|wgAq!}6VGa=sDiwL-er5A!L5;DXu=?t1 zLJSDVrbskdGyT;Pcc}}G+6t2l!$ZdHeZ%0?aRS`u*@HRARNl2VVY`S_VD<*MNCu!^7ty(oVa4I zLwV(yS&5;elEqftNxV5cQ_t?D{%XFWSX6Q;-Yr*fP2>ip9P~5SL(8sA| zMXu#YN}={iP|IV?VKxJyvf8c!I`!n4owTQujU9nF+4}?d1D;D?SmfkBv)^~^uc=>IiW5G}7>8izAaGjV~}I{30BWhM3dp-+c5 z;-<#JZk@?FG^%uO(6gK`Y|^7rsfMWG^jFN=jqFZTR@G$D1~4}mQ?@FGtZL!(Q#eu4 zH(QMLSaWk@mdtbR-QjSay+v-Y-4Gqd%!kN+=IbWgvDH3n>9i5#GOSZr{NZE_$2P>? zV~BWM4o0bx#@1Jxa_{`TC|!4rA}g-IdqFyh?Kh*n0i?7@avTveBdMI304>pP zHDuMLP>JnvNRoBd<^^daN%g2=d6GO0lGIPa_CcZ3%OKYm00`0^3SP3RHW8T$@vkLg z(>$fzU7vZT?cKp94Xbe0hdBK!Ho8xWCIa8{WP9Up+B?pvBD`5_4o8;k;ZP{vkZfD2 zHOk0-*5r^r9kM=D9HWUf*yw2I@9F%PjR=jFS;pQQtb>wk4mAF`%emJ6 z-t1^G;G^KGa~*)xl#5jUkY?w;CEZ^0dx)4Glo65u9| zyu(5j{E^wK+v~|WO`?_B8Df>U)KjRmyjaVSQUbP&fPCzW*%cZ-D$!g zXLqu$<>0ZlEyo{bLL?yo=HXnsU%SSQ9EY!T%;qM(9ZQDtA<1O5eCZDoYLNrA%h2a| zbY%Q0D<&Qbl=Ol&d{}g$x5!X#a?8b*htkDNa>^~Z&0;~cKxP|#eyZrb!UWcj9qXdQ zH`P=bBMeqd^VkFd)R{}Y~kC>BFJXAPr9RfPhO4G=!~4+gTTGx z@}zbwWtlig?J8FC;kJq_trhzhh{)Lxw@@6H<~5?6UNk4E9M)!@hlu!RJxd%Ih0w5@ zU9y;agQ$0!T}KkVJk^9EQ6t;4^WPZlFITO#d>w?QOSJ)1bNxNy2g$x@%znjcui!#{ zh9#@w5u;5lJdN|5jBDO8dgOAD(q6(dk>4+WN1uUGAi}&#;0q8F^bt+JQCYC+51odM zsZ}dJFt0#fIBF4l$G$M;Or&FkpfCCG57tf*_wu+kl~$9c#-uK(X!~6HFzGHjRZ5Z* zS(f&J#uuP=uP2&|v~8c$8mRZ z{YPSx@{ZCdn(T7RLRdE)x*E>6))6wuB)9b;giJlgbvH8%9D_lhZHHv#XMlU1jYfuw)jk@c_|Q*$G~%uruJ`;$;THcB$nm;#>iW5)xf?K9THEsKWJu0PeSaCTC;W zt~E>x6-$>3^B%{Y!B{i+!{0$8)~^dPC(*w9!TJ znB$_%S8Ae}*s%UUR?`sF80CB6m#3fK&Fb%PgU#u~E`C0~2Zy>CN~J_#zo|vgu@jQq zgA&LFF0|V$jG2DjV&#|Rh#qBDi0^>T9%u5*%M^sxrUI#riSX`+OCk%Giw{#8|4&3} zg(tPo{PdrG4ASFBt_;>{G)fU<^^u2pqq!Ea0HJH)%e$y18`De z>d^0S@d9^rGcT!SRl+LKkpcMI9BBE%&_DdFf%0UCo-_pNc!4-(M;X zyn)z700CLdh4OptKm5{X_{WSG3u=2|wfo(@ z6@klHa3RQW3_a>-1l7K@S*&mPV*`S-iOb#zAnlE6LL9}57U)n&<^j8~9_h;g!TfUqJ-B_)&}IChs9!yN_LNRoL5jN>8$OoX?}@H{xiSQ|)FJ!` zcfxj=q^UzBGr@k_n3O|9_E}kdhfOEx(L=f-7-mjq1U3#|b0tuE2;(43(C?^nC%*|B zG$V^O{gLvzevZ(5e9A%%|=qzOgV$2Wq;&Y2FUQCb!DW{9I4J~r2m z3~>K#>8C*KBM(KhIPg+R+0lx)%kfYn+9rBF2)iam9s1&jb3+%uck%JZAEfhi1WjWOWQt4BwH`XV=o#Dn0ZK9zcIp8 z>~hlHl3iPK@>*_;B~u=Jjk0!#V|Gpc#cl{kJ4KHF?bENA<0x>QozLxkY5vYdl}-( z=7U@#AZz=}YaUH7wDqxe=2^&ic>ugCLVjM+?j*`F-qn7G>23Iij|@+&cvCAqd@%}G zs$EE=Zu8E&($;uj2TzW-H$`t$U~&a?8>^%yU2tggwpy^@dkT%L3I1Te4A!|6hV|=? z3gAx7dn9~Q`U2uWdy+r@xu||zx+E*YpuXd9Tca+_DZ#5MK)I-Xii}O|79hb2kM`Dc z^Gl~P2}{gv?MfM^p==fxWcG(6Dd)5 zAC|pRS>*oZH3V7c`>Ll}dG@2Q0XpeS^450ZXx*WRdzPJW634EEY9^tWl>0BdOIG#s z2r=#=!(&i{lomC&tA6{HsHUWq?wV{z26};pnCd`+E5n5g+nF-KLA^_gh|kHA8|K92 zF7Uve;j|jM@^ll7WwLEHhG3=aXY7c?Y2|PREdnvJPw(EW4(3&q`3bJs>Ct6kU-U}J zPXV-WX&l6;w^?2bdY*1yf>8@j_|~MQU6es$HaE5A4SGS6&)@3K1Q{x>+q)kddYvAY zWkm7v61aqyt$%lGDo(p9ps4|_3!n2Wmn`7%W&-56T}M|E6NeJlSPFXC0kNaNj#avk zVMUb{eCI9&%(P`r4Wx^8pCUZ3Ds|VhyX0!raE&dP4vvY}^cQ32Aa;Hhqm3mcGA7M45M9^AJO-oV?%!T@1Zsq>J(XHt;@5Ap8665k5z4 zA*Z%?(Fx&XWt=EWrtna{IQ;v)6z$9Q zB1S*~JNPIhb$eX(8R!UWr{UMD!M2f^EC^bpUvilz|F@a0|J^^)(sqe0b^kB<>A0tUVKUZ;aSS3XBR}+_AdY zFeTQOb`TQZc1JFR@jlz)ey!jyP<_(*&t>DJXOpO+Mtn!Fdd;-HwN47{H$+i7Nzzuu zxbf7{EJR`7&>*7^*mu~a$O$e3^!CxVeg@kchmRWCB_o~)EF1nx;M#jj%%EfP_xY12 za~R)kDu7+hpo0L;q5*wVXZwOeb}(1bujx0m%zP%yBM;jK^&Fa?kv12lC1jqdb$g!d zn12jN^yC$cWGdaP?00WBspeIcb)*g}_$>%&Ow`}*#Mj9AQm47uN`32~@Z$IqQu5b| zg#Gd`nMHAbuVf)o1trSs2amfi#5>PTlMe?>nqhS{kzI-ovrAo*IB+REo~l!8 zS)^--wssE(yXeJ+>cuWLboxgrAmH<2B+h-qfIV3bLez38PS2z0BNO_|rtfq0!!|-b zd*ZW7I~nm4neYp(OVQ<@@oSpnN{7PYtNB(7P4c+LB~NW{{j5g`noWGCTqtH3!`$iY zP%^YLuw|X;=!dK)<(E#Q#>voUlbo=3ucgiO=2}yyj)3>83xdv;u2_TDOQRDd*tT8y zS$TRQa$+IU6GQc1_5-;UBBpD9o|^`3z;}vAs7G22O@}s52U%%+>>A3d_9gG@eQfg! z%+81s>%Y-buK%Q^tQ^c-|2yY+p)KdUF6q>HH@OxXOxeEL8>*;Ag_es8+)y!L(N1(^oR^zhGxZNQG>KQL;_hqz zI_~heRH$xzOFwV4PkMEUnrcM-j^;zExZC+ciKTDD9ingYLBdE3*R2(S)=_M=2Gr@q z8;3v#SBBmfdiHx^+=5bS5y3~bJ>opAm0^-p6syzZkh->y;==ed)s$0#owxy9W8F0Zxg3OilTdWm(X zjRrY;Di`nXihLr)i;^U(VO9mm8-srj=6Xi)o{2?Rztn2V^N+AC(`C^oF0R+g-QCSN zoLL)|%Ju%qarulbc)tdCI;(BU6mYVYpO2aA$yz9CnCZpZVy-E8s4q0FjErw*Kebvf z<_R6nTk*_ctx(D@*xvQ9So?BM$QMq&B4orRbw5dHXv*7N;4*%+mU}2FY?gNi4 zJgfr}=6;Co^_D&8t!VINhzK8U&o3C8Gf9W0Z2hav<&BK7KFbT!xTwjin>m%c`@L4x zcUU9@iW}NtjN&52GXu$-`|DSmn!B7%v&>9x7^jDMzxrD_x!5q3_J@v&@1g!6x0+k+ zv7_E`cCH+b*KhUpkzxaPF+KEteIi5DA%>rz(WHo!*}tQW>UBeeQ!Tpx@j`u%e|k2a zFC5}wFv{76L9%)n8^$7hc=U3eVbtiZd{#Y8r<`C{7-p<{eLf!KEBEC`WzERCcDRTvLF^3s6XZ+>wd*?-g z0|+q+w-Gx44~)6eK`eBO8I0(gMNR};21kjFIU%`rKEvO-pwP7vWnuS$VQ@{&wht&g z3l3A5{xKM~nAmQ!;}ky~Fsu?6$q0BcrJsa5CibW%+|1GGCv7W@Y`LEe?V9=1Wqq!* zK+!=Acj-Y7vxtZwG14}07LzQ0J@>NFS|wOm>MpXmL{k)2vG#Dk^e( zlFh+1U&u#N*s;GZjB?X@IJ5rM@r#vT1+pYkOikmg9B#3^Gj4YI!s|TE+Ba;*WL2e_ zf#abqGsg`*E1%}Lb5F~U3R8_yP>#N?C#krUFyQlzPK)%lm>&>=I3{ls_=4m~W9>mW zV_%{_=Ktx3vq=m$y-3e`V|B>uNzvr2A-FppUX?Y1d%V#HT2HoWt2yGn{3KmU~9EfL4VnH&1?+&SPe=vg~a z8^JHX|Icy!3fLG4ywDdd7#2M!g^%0I-SjW0WXn!WHFz6F*`J!m?xMUhH+B-@Ia=>H zhe2i$gxav%o1m&vE$A$9(%A|l^AjLQvrm9%oV$jFRXg2+tJszNdBL8Yxyp4+PV zhYA;&yG#Nkn;IQRp{3>?fasoXU*~_Nqvpg&+0$rm=Qti&tu~R<>=yPdjH$fTwv$wf z8zLnZ1;%$&jpJ2qpa8p68=={iG7U2gjftowvp+5Dnv{_A?FWwGEDj=JwQNNWh4j7R zjqZbH-p>&L!#%0~;?y-bai29152QYoV(xCyTY&6sJv?lHtaiO#gq#`I<(=mUMtdH#rc7g_ts&xiqjk)QPxZQ%z!-x5$W*t1TZhx^!G_3P<< zVPVFKP`C`ocpji(m)_dbYzWca7B)Cv5T)ftRlMxYC0UGN3Qs=o88_S)Jv{q>!+0{J zw>Tink*?0zp<9(LeUGN&2P}mng>ZqPr8{}^zcriQ$o5VKtaf9~wSH-RRC+%>*AM(w z%KpE^(SrOCKSjOa2g)GXFIBBp#hesbQ4z54k?|WcL+zE4@d%{%MHq$|Z!TnxAUWx-AE}@f* z^_)d}HMf>k5U_GOyjyheW_xj&yAC$U=pu9n4r_vtb4)rLmu=cbnt!^fXnZXw0#w(_ z&qv!1)Khgr6bR@{R36?7H`Ob;*$$SBOiRm-N7tsKeD_>*LFbv22^7uqS1a|B}*FFwASQe2wMTMtq9!Pi7V;~6?YHAb=Bk=5obd=*)$2~)@=H$!x)%egbt?3kG<)VJD~s~4NoBND=! z#6I3?0?6w>r{BaCWk{pQ`dZgxOGlxoX^bV76Ru+h-k;_fq*YK-S^Z=%a!;g{A5o(+ zG~&n2I8QM0RVDD$C~oMdt1}T7SEZipK1=iCcy;0fjc>36l%r=P#w4bZ#F<|aF&0tY z4hbP*>-2E7!F32?nGd1|72~X0x<0HG=d*@l7gLO32Muj1r-3{%>nQTJcGo0`bAr{Z zc!3_ZVbRY+Qj7nPBS2QjcFOePc7>r_lDVna+@OWc%;t)eaR8@HKcz-$>@VK0J{%Z`@3Op&=DTIu#9-CBGC zA5?^eb&L)r!d%TI=Q*auvoz|qoalq@T{N0^{Sysi_!(J&B84s9bXcl8W(}NvYUqNG z9U*vf7T?W!K-)e1B~rLD)getG5{jbW>F?e;sta#aqLh`FfzP+)&e_Oee?RY5gKg~< z7!naDHn24`u(vJ;o1BT+&|d@3C@L4hRuU;&8SK)?&Bzn7In!PY(N{MO;q;j9(|(n3 zWwZO4miR{zxgk5I&XClR*p&gZBxwnaxK-a00!bWwLv~f;cR3nG;!3tXK_MdY72#;t z^1PfHiq$^FO8|v@6C|rJsA331c&ne%77)$WdA@GEwy|+Dzu8Gvj$JPtJh08ih!YU>}mz7{pXw7K1i}R*|DSs*pfG0rP!=2wI1avm>+R$$PCsdAG{lK@$gP zvFc9v=%FI|mlQ#+U+nU>Z5PJ<=~kH_Yg;T^1BOh$R@T)^j<@ehoOP0@Vcz#hvjK8_ zq&)E>ghkqkN8Id44u9rywq_deavUL|bD+@?!*WK_!Fa^;8(?xQCi`4~!(10+TDGCi z4f`WL=^FH{`~`E%JIFvM<%^9XO}af*d59rTb<~L@G&i|pgD&TJi835Qc5`X@>0KXF z_(^jbxtKfyzhRe0O9WM(wbfZ*c!Kws8-rAa0Yqd^+%nX~2zla=9d!`vq&P3U#~ZlV zaIpkmQp}$$19p%NibF*n>7_cM(DVu^XuOk<0h%j9=RibMKOs6;U(jyJ(prJpsGH*% zin#`S>OH<3sg`HtPi#hE-0L`*buL(1JOvM^-`bM~0yg|o&bMEL73lxQTP3OOPJzhN zuo+Lq1tQ|Bq@L{^sK(pK1FIj$R@c7D!I`-w3yJNzwx9%)V6eG}A(1>K(?NkQti}`7 zK3vx&>dHOtiM6DnH9i3ub>%Ik3n8Q7zYdFI7te2NO9SW~6Hw^ca^Q=3TtbW=(NRFl6V zA7`vPj6_9w-wSpLCd4Sj?f<4yl5W07hQ3)_zX(FwbrCEsNo2;A@jQ7yPmW|@jyzHt z977R9g+!J9oR=DKIKRGE8gPK*!PsZR1uY%SYtl`WG3}&m=+#kR1Y77^EM=WV?1FXq ziWuD&F(yIi<;rc8FupsQNy&NKhe>jL{gveCRNf+UAFNQ~@(P*LMZ?Wbjy$P`KAG+m zt>n-Ep8E2T_ti$h=gboVfpw1GNs?bgf%Xe2VqtuPU5%xlg9Q_4cA(=?Xlyh9S>{e$U9*vl1Ib8 zQL-04^dQ!$uWsQUbkMu(pMd~L4}@L#V_)$ zd62SBw{dr9(r0(O<@bYT8avpKy(wLZu!7_rFg?uGO@6AYpRG{ISP!kcGQ8`6@;#AT zma5+}W&s@+2uNEsrF=OB4WVIOdo(?wqhB3qC;O`t-0MBJs^vb(1Dl~TFn4M<5ouMw z7)+aoa17J2HzMCRxIO_54J^`Id>S|LH!91ESvOEg;)4cs!d3}$J~O*{5stmst;#Ul}aj*Yv zhk*Qt_Sa4smY;cfcN*Q_>Y?7p9{BFq5uS9JOo4oxIUSq29p?2m+EPo|xB-F#0$n<7 zW8*`U9m=SDz|(;}?vL#C=mG*47KG$f`r$Y5y}*r2HVAGKizQR5153krs2s%S-3Zd&l|dH8)zw{|P2*EphIpEnlW z_nQJRx44RV27RPhmD~?5iiwTRsL;QRKc08BlYzY#HE+UxFW7f;^|D}BVKz8!^Tn%b1V zAHui7CDZQcd9K$V?0u+)@4qmtUrhhWu-JaFaQ<(GRiz{2{GIsgx~pEZ16OEYz$FXu z0xiO&AHf;TuS7n%5!H15(N4@Rz9EP!7**^)dzrJkcWPk6U;cv?!im=;tiKmhkxKPo zZ;X*eMu4y+f>-{g``W#gPdd&%){j5Y<%#oM%Py|FmD}r8zrrNPqh$SNVTjaEgy4u@ zNxy7w{6+NtR(1JMj;^K*vU2FFKM@$slH40*Z`aMPHy2JNefq%~Gol?L{%Mta4k+Yn zm(ty|YV3O5G*>J=6^|v#YNJC&(Pm9js8{F`K_FB|_Ot1t+gLL+kK<>`8VL<*0Jw-p3#fZrR5 ziV^uQ;%;b-6ri26@72=4)#jr(iBk9z;9{;m$t~w}r8FrJk(1nbXdk5ln^eq zBT%@IuUQ!Jq%2ME*So>hqP#2C*Zpmp6Ps&N!>mrBI#t7hGH%h{cVPLK->OfA*xiBV zzQ=~20Iw7RXV z-JxISrp{_lj3ITxt@$dd%lZ@)oy1YAmC7IF!0M08xM(ZqAHAc+R^7&hrb^6faC7GT ztKIVAJ~f;tBzh{N;N%t4fwD;86K@}L|*YIu@lu2@i`m>6I zzTF|Aa9$LSvVnFvr6o0A)S-s;J>H{-p?cuc>^!y7A`_H$WsvV=m4P{b0!@^%*qp5M zEMSH(a!P#&$l~(eqtR8G+aQhW-vqT_<_;r(F53c2 zd$`?G_yi;5| z6E;YB%kbT`#C3ep#LI0v{~5qSI})Qn;29#1`5y zBf&b}I|QGE6=P3z6b`E)yq(7Rs*@W1O#G2cX4!xo{H-cZHYd_;S%?q~VwA{IGwPE6 zkRptoDv>^Q!HLfKeggvxLB=6P9BsFD@KY%s8wrrZR?A15Xgm-!)0gmXVE~R#n#$Ixf`?|?Y7I?{$Acfr5ACTlkRaNDQ$mh$gGp$UMIr}DN{ zjD|_rHo2;ZGYQGt%=n9rSkUWTzGq@-3NQ^FOnQdZTpjb>wNuo+R8japZ`(x2@uq*d z#t7TdoKA)q2ky0 zcm=M&Yq^oZ0H{|UgqDSB$;8kaGdpzYZfEc;hRC~O8@h|)J_bOCB zU?j3+KOn%HQ@c)}3?id)67G>gRN{Cv9;y_4KTpZ^nnx%!%i#?~F$c%g$P*b+{G99P zc=51FP#Iis5287AC?SYnOL5)b%RRF5uA0N^N|u6z-VkXW7rU5s74{+UQxs^T)J#}8 z2O&8I=%KDdFx2u`5)0M~WZ-$QwVM5=rlzNR6qWlr+OFG+kq^|}MUxXMshEX4o%StSI{elYeDouN9l> z&hshM9aP;7e(ju36mu7XGGAUl_+dxhkd-cu9u2$KXTvfaUH#Vg$r!g5rbx+eeq85# zdbdPaXeyWhnO@Su>rS`<7?7QJ;Z6;)!sbVI;@wA zvVl|tBJ_>%T0RKRDuP+9zr!7ou@d)D253lV30kbddmB7{`pJkubHWxq7M-N`C7MI` z1d=e>hM{}K(Lwk!0Pjyi#c{bPGiLEX1Ei-IVZ)}H%~OJj)?NYqg@Sp$Ewz5MWk&-7 z`l+swHDSGutsRz|Rr>adgPh(N%$O|T89zl9-opS$fxymdZ$(#{TkaMnn}|;xS|0I1 z3;|y4r#LWAX+#Or^MLC2gwetV>gXqF>W7OdX>6J^G}4QaFG~Ukh(m(jbJ4Q;NZzuF z;MO4&b9}qYQ@&e3!6s&9pu+;y3Jn&K&`y-L!*Q{A8a?QNxsMzRA#Cm}Qkz3p@0$N}`l{ODelk-aWC%_0SsW@!!t=*!UnnHO` zs*(WV-R6qvMm!@sf-Xvn{EC|-hUORq>GhjN>sz)c zAt?S;Y}p)&a3*j-Qn+o8wO(nGbH9bs9`cVjTZn%B=@9&F30KgO5qmF9d4cTYggCKCt05j??AxJ%eEcXgC z_0+Zix8AVht~Z+i<}N#F*x*CV?~`d}@FHjF97o3ur|wqWbz825Sit7Ihqh5Pt@uN% z2Sb}ST)@$6P56*aOC1%J@j;Gb=2WcmDp>d<8u`RDL6^}ozSfSXO5euOg#gQNRGN7wBJFPE2pdd{AM`2lhYw#9&D!`R5 zh*N~ImTme2DS@zzv=YnU-X{~Hf5uXgeBh>Fvee;N_SCy%IU85%zC7WVyJGYKQ2?3r z!lfjkPGP7Gs+BOBZVFWqrxn&04m+)?uzI^d3p8g(cW8i>9sQz~xDlw2btxZGG9M9V_vW|{N1rrj@^rk z#N`b;ppoLo`nitVw%ryy3ghoH7L8wR$C)6IBqAe5g=Z#V#ji6vstt;eQ~ne1t+@#9 z#}S#%W*56Jji(}C)<#qh-Go=LYEM-TK+6AD`~|tZ3}K;Wz%FJ9C6_)6PjL^xUq17N z*nBBZlRr+V*(Z6JefSMKt?VJ^nkk4wII9ohh(RTf^nuR^yA&MH=3r@1X%jSJ2R&gM zkUb~DM!&UP@~5dO6Rz?sEJY_@~b!O%N}c z5}yhO0yKf_%wGc@1TDAZR3lhW@~mPE!HYnr1Efp!O&fX}cf$ADxr7QibGi|pXpWzT zMG7>iAPRD+FfQD~QLsHE)l8bwUy2cf zrhp*2%m)@lkIx|IHqR8V4*Lr%W?#<{QWJ#BQkPvPHB*-ZBc1WH0NALK%WT+g&xjFB zy)4Fp%d3PE%$ZP?@Nf8fEUk*)VNm#kaD-SJ-~-CZB?pD7mO-3^{s7p7?phNuYW9pS z2JpI;pI8Rqa)>&+x76RmwdjWgfby%=C(r}Jq$mmnO5jHCVA{bw4 zW<)+NP(L{o)MmUME<(N>*%I`nc()%$wU@%R?kwdr%ca0nNhinb-it&o}?XZoT zu#1qrGP2NcCzT|1VPSSkqI&GtnRfl*?nm0P0@{{N+f9;~8J&Ny4N3^LR96h$*3WOS zfV>$VS$9v-_8ZS987QD-$^C$@qma#%ZaeGzc8y=hqf;S(+>b5|f4=R8E6mm-QS5hP zjm2mRMT2p-ALaUkHDkJUxf#7B634bg**nR}c&_t#AbqO0=PUDaG898?m1EJCm*7dF z){0LQ2&rSjru<}gd~PIAO*Z=)v5&IXFTpS_wt#$0P)7kZkdue(gH@4cl^m%-ZGG_I zp<@1}a$exn7T;8+8dfVVZlW~M8kq2r>S>Y*{Ay0-5kVv-|I!_(8iKS#-R5xD7{usj zF{4X(7m*l5V;p*PHG6sqfz78KJ@ZfydxpILxUoLSHUt$Bn+s3|Ap`&T!{+;Bfz!f& zVTPbs2-AV}8;S{1Sa(?Q`tHMlm;D^9Cr&5EiaQ~EtHK#wTnClogrq{o%QcK&e=x|8 zwGdkZvn4+q9`+MVz|+6l&WS?16miQj6C@T(I?=rvA{;)PR@Z^m0TAxPYJ;vddo~=( znDem$3OFA{0!zr^oMP;|W~AFp^_HG$QJe!2w8=7Bq8<{ja;k0(+4rZ+B^Kuc9mtw^ z)q8p9(r!Uv55+`o3qjso(u9r}>sMS$T;xe=2$rMKv?YYyb7q4bM^3C1yINmQz(o#D z`BLl!{AJ;md+qsW(pw+n4IJ^&kN=xPC5O!leBEO%e}gw8dRk0SDSClp=6Sim+ake-kyP3y}C~a zE~wZqGvH^NBn;sk2`8k-DzQ9I&U^UhAFY0TJ~@?EW`_urb5wloYP^av6(#<_(6bQa zB+8o$gjds5k8{_niEgC3`gqdR5;{D;9!Q+1Di=tmB;&u+B52$>ckPg7@~ww*?&#y| zfid-;7NW=sTOz20>G^YY58^KPAsI1vzMLF1DH$kdf4Um8KUH`6u+C%68U2LpMF-#{$HR- zh%QKi?AfpHm@q+rCX^bxx5At_oThKCBzEIcSJB(ho zlq!!upN$~-xpy~6X!-Z0Na?>j4!cl&_SL_cI0n~zyz>v?J`?4^ zVdwp9xUSr;nqrv9vv&fL^n2YtelI_!J_(W4Z)Y9{6h8Jtl7J>ED(5BO4X9k`Q?l3z zrL*BtIglqUbu;$ z^PDhA8~l8JMFe7JW_1VkqrukrAbC~EY~Do7&g&1q=~Qh}{EznPBq7S?GZ4AB24}J> z^>1qvjT?NWDzq~sKgB_+syRk)0J03|e`j8lC%-s5f5=#7%+!o<@`*V$AUjJa>?Ysg z<(d=du=^g)ihseyDnwcS6I^8c4{(v;A9AD@v$SzBb^7<(7`m8>m>S!gm=f^uK{>lP znHt(cd2H5d&$%y;rtNyq)imB)DfzlT&1BZvMD1CSH_x|0{9=Qv5B!M^nJo~L{1AjPp8Mp&a+Qa+g zzq&a%iViMb)51Mf%&U67KfDRNB7|?^QR+|%XPY6mojw?O70Bu3^L6s-ZhyyjX-wwR z64Ol6dhTlHLWtJ|&Xv(BNn~VOts9{)(xSjV(X1OjImAbwyZfjAqmF%y9RYBa6=f7}=da;%|$FRqe>tquZb3NTvlB)xGAs2kuDae8&kA2mj0S{Tq9QL;89EppMH?h>2YMa_}}VN*2$=+qHpH%TH+vQTa2$l9}2R zS4{J7s-iDKKSkz-A5A(~(AaXs>Q78D&lpa}yEEYfog$?PdoL~8Gezlg(N*;bQ!{1w z>WfjP5QX_nU7sce4M}wDdXK(|n86iWS*Fs%UNjA=J9O$1I*Fp&GIaJV(nCpiHCuKP zHaSZL#U(W!N~8P;Cw$hJ63R~Jdb{8Aft{gJ7Ok-No{rQ*O-(qQuV&lb+B0*SSo9u| zJ&KY%NSQIlp@k5+?9Pj=;tI5uo*mKxJ$-cDa1;$Z$;ezZ&ScxV0_4n6f^_OgV#})o zB_UmMSD_xtjcbRF;lhbg=18*!wO-oF3v4W|bZSDy4qJ=GX7=dsqS!pK?NSu=W%P%^ za&@D{+EAz6pxx_oK3QVOW^SzbtZB)PH6o?yYg-osNjJu2-DPh!HoG7_hTyr_f)7}l$kLIZBLKyvAwoM<VC zvR|tzI;QY0SCARH4+l2N=7d~ziW;>#JNJe^}oJlL&DR&}ZtWQu!qU;n7m#P2c zP_QzMpL!TdFsVgWbGKpW3H0b;2*FWf3bvHAXe+MVW*SXTIjwchHpj>QB2YVO_1SLP zE~v}ikjg$l<+X=vp3%-BltUqiR3MNm)p!c9CDv76TEk?O)wHktt)p*iPFa4$dkQ}= zCeq8xtHS?rk?bQ7w7&tAYEsmC@l`k`n?T**YY9z^5%8uC^KUD{LS| zK^plpxkuQbiOhm_-WY-TIlsT?5eLoLOYl`_AhS2|Bk@5>> z^S;t~1R5@EvhSe1`X~Q8T$VGM1nQQ<=xan)^03fwpK5E%XlKbjiwAK=+bd4fb*u}| z`5A?1$DBHjeN9&Tgw%$3fI#%ZHx*RaFXQLDQ?1c97p9~+|{Z>FU3 zd zcR-Y=T;K*BfgLfe)q73Nm6@<-K%~J#O7?79#kK9Q?kSGseaM2TGn<%IvJiy6JDh50 z?=c$!4jc!7Ytfxr+{&R>M5IYB85-am-Y=A#6ERR_uLFP(+{}@3fz8b<*9?G9%N4sw z6g`Gl%N=BiLzPTi_KZ;k${u6yl?BcTXQcJU6a`T^4Ev@HTf;%NR5&Gq zd!H{z;@rf4O?End`Vm+Dz`u+3P;&Q;g3qrSp`-)@>*yGvF)WTL&m|kXw_ZV+EPtji z)UGBY2O|4&wpf-EG8v5!_ms14o?6?@{N{#?t4H|~)Eoy5l!E3BPf?Rd-Pp}zURtE7 z$Dy^@3phF)e=-1H%N#wY_ENFUf-3H+euj+B6S{ z`-BvFqzs}(1YoTNyv?A7SDEnlvNgy(4Qi>-Kh-Jwf~u$>-Sw`D&FNmIhYz5Q`fr%iq5du;=hg?UBc!7`@vsTH^;JU9>paQ&ch12BA&CT4^*dVG%-T)f zjT_Ihi8*c{&N}a<82I1f*-P=GLs*J_FTWN{LUwL$%C?EfAt0^v-)-2huSNn38urvO zKK{lF;~!^^Tb};~+a*oBHeussxo4toskYjF{=O+LCO+S%#H4A+6W=Zy$GDh?E>MS%~+;ZH_S-z8fzGmXR3`OW1Ei2;6I^c1|R#k zhe=U@Wr#NQ5G#Nzm(S;m)miaZaE(=J#B(4L?0(Rc^2?@WJtf<*(;pu{UqyFMEgeF6 zNUbhMfmEYD@4_YIu;J9J_`*CG}#aOW}AuZ+CjZy3!dkujmYbsjAam%O)6FVWW zG$+aG!B{ihx8*fCDE#J(_Kgr0&1~_y{*UKT89IN`d9zvGW2XTZ1~FQXFiSzdY2*HQ z6A#xMk#XZdr=NyKlov+T#)^8VxJR-Fz3EWXQJu`;szbMe4yCN(0KcgjLi$5So8j!;%bNp2JN*mvSQ=@UD>8q{-W)EnA2AMx<-#IlYFP zQOET@GU`>%mm~^bK$x6TsjO;iA1%y+rGf3A9^Id7hg_Evi+IxNyGbuT)N&kYQ{tc@ ziB;c|FH*BAc_D+$Gw^O;9!}X!sE92p(UutR$1*&{Xb|cuWp-N^p}R~ zemV7hknJP`&$XT4hrXC%j-Yu!nQoGK*~}UZ&XqCM^F`1r_EGG&Az@PK?4n~FGUdX} zld=|}fFBgm`~Yd(>U$+2aq=EtYG?8u4H~JJl90-ZiTN3;G^i|!Wn%gt85o)L%7P=2 zafY^((wddQ$bMxiTyaPD(#pIiohL|AYoTodotyE%!<)`t<$zUkX0LvM4kCetvwy{M zcz8ef;?9mxND|LE@&&=5RG=ke=!PNRhk!AwNrn*>HlM1r(R~NV!;b0ha_*q;+gfsS zMr1N}r_KzCDt{1&@y#Z zm-Wt+x+8=rc|s|bXRoNYG#m9t`FfmHWV0*jT?!52WVoQ_A~kf;vf)8vDnCJXq>4>I zR?1yB^!my{$*-5_85k1`kyJCDvD`l|#*o%7AViR1LJ7`&^$EhKaI# zUV8&#$XL&Pd(BXAseCi*7U0TVXmiFY-eV>D7X?i5F{<%DiSu~=wz`9nXcE^bQDbf_ z?}b9n(ki0yj6+KPZ4&nJaVWGB~gZ_#4f@Ov}>)<_suBQjf0c5g*G%aFE8jpq)(; zjpmc~IR~O6I~VnQ6>i#$l8uwlp>gU)6$Vl0Lr5UI^(=J)UuiuNvf_tYkjc#ME907K zzAhLRG}(aK9nkNYkRf-CqVhHb%|e67_1@u0r*~3}tMcGQ;lbiVoO*l`OEB%Zr*24q z;3!O(Zv$TVl|5Y}e@I_foam41|C<;Wrl2Awa^xhO8L?95hY=ygzl$_rahE4q9`di+l)acTKRl2K&jDW)7+HG~PQ3}IS82pW}IMgqHNjB!$~BY$IeGNFClE$oz*R!UFRg|jOn z-9BDigQU1~Dh!OX@^Ln}-TX_ybg%%PIC_mmF1`33N5N}96;|w3CwLoLFu(7n^lOtn zCEEeIr9noCPaUN4HjrwUO{4=y$+l5RiVl z|BJ#;Uc7%*6}w)+1G?&HvXT@?lF0O2M9KgX<|haB0z9+i@NMLX`g`4excG1zCz-y5m7;)})-LYJz$6+IZ z<;%12g1f2qs$rfC58db_ixXvUsHtv%(Kh>1-y)bbB%LLg;o9@g%dQu5p|F_d89r>LezNNV zKbc$dlVm>UI~{(Vn2by)X{Lro%WVdq6sWM1NnNH{{PE#1dV=tVg6nK@Nf#TK(jRC^ zrSQ~~iQN-vZoOWX%1Lr^r&Da5UXg&~zNd8hHz6rD-bKFwx}&;yo-=cI51&&{38)Aw z`yJUcoG|XZ3Wo-_>zA=sindbh2&IAKam(x|Qu%$*yCKy)Nql(I)m-7_H9X4^>pp@` z(3Yy!Hr=O4tn7hSF{J?+a)V1d>A0i^t-KUnS6K+L8lK_$bwiD8n$3=d6N3T?3+A089!V6FPq{ zHQf%cJQCSO5t5!9Vd%!My)RNgbiqAUpTKnfzuW+;ptI#KxqTpM0NdXKGg>6JXGJAzgt9!ucc&| z_T(<{adCBoKRc$)cCF&6U5Y^8kGUbFJKn^<#M&J^R#R;#ioz^^iC{J}Q%MA9=Ucv0 z=SPGDO%;CemHIgsZE}w6!GGbAqVNI#6Wn9^-*AtGorCTF0r#~2xdJ7e{v`*KZyETH_QW$GPlr)qb z4}J^jXJJE9gAi_ML|plDF!LEOQ)k9M8%NHbzhCHbA9I|@C)F+&gH@^n&7Zlm@5HsJgyF}Ls)*xUZFx)+(+>@E`#bBG zD^r?sP1Tn#V@W(zRc2A`1-(WkCqRGw-b<&(LLb#($isI@FG7Pd;p*mmU;fXsfa3=U z+sg?MNnDOnTReh^Q?3LzdQX|Uy?|by^2!8o8EVNk9 zNi$7FxiYDtqe|1WWXkX@9qHD}D;uKyRz=7yxtO?`m8gL(3JnaF!PzD+zQ}`C7A8af z{^Dy7p!hMRG8^X1LxgcV&D{E^NqZxjt|H0GMH*^L{?G5!+bWgDnpz8Vl!=Cz*8Lo5 zHMmuseXlg%ifAiQ{sY4yzp!+|Ou=s2pD5$5reUa3o3BU5oCcr3@->lN4G%7bk_IxS zZsyt&-9hKdqFgi6e_N=lXsVfo0F*hT7(14#e$!WxrxolYk4}6G76O9}IjsQIk;I0J zLzLi!&92hw9$;z%!jRtb^4cC4c&3`VXeYV&vWlL~tqU}L^*fu9rk=^gRGQIJ$6Sap zBtz&c6`%tf*U#Un`lCg-PQhA?T#_)wmXtR?Pb8qA>nx!|4~^C{RWrhR;UfaEM+s4+8>Ks8FOh1T=*u{=xlOTe zn{(aMq36AjV~obSI4}YvgB1(xsRac9xJV~LP01eH1o`n1sOF9Po1Y%3{3#x1m-|F! zOD2LRyIsvVPVOy$!q`t5>SH>#pmf#Yblz-RLWI}0lc46>s;L5!f`GMEVd2V5q%=Dq z&;j%^pJGy)ZaSTa!_>MbqgQjCiN-DRUf?pth*ks<(|568E74JDU)q+hYT)E*{zWw( zL{Y5MJsHFSo6kBZ6jj z^YG@rzzVFVzA?B4UfdKh_$Ln)TLEcD-3ITogF&#a1|6F)TSTDFoo~?hInN zHAu&8G8L3D!z1^zaps+-UrXbNO_Lms4Z5iamCepneZei{wBVZD5N>r_KiX#}ncAZb z(zCZz7igY33Gok6uno*_aslGxp>1=JA*j8{6@baxIsACLy75@&Q^YAfR_6tN)qeA- zBC!@t3f{5*B*Uc8^)GDyEUn#{qqF;QQ8@5(Nk|WnX<$(acu6DE{0jCb#$RL`h2jdt zDlrOPn=VW@p?zVjg{Jl;U3pf;@QT z)l+Q~5DIsaX|h0iB?e-ui5oH)){~P7qV?bNOp2nGHNGWghgfw~WtA0#D?v4aq_fNg zg?_&kjSv0zdzxs@71}OIt{WnD&-M zm(ZR@sVNc><=<~1TPXPF=@ant=vGJ6RhmWLPOtT=B1!lF97}0e4M7qCcV;;_XsYqZ z%0`S*yHFRbB-z>o^y+mXu^PXYZ|_<|?< z$>@f%5VdhrpEC%hZ)e>FMpz|m-?GoLV`A9ZBLn-8ui7Zf;8ubH@1t=DGd8iPE|7PJ zO<4rBj181;x*`ROyL1a~QxJNsGj3g}{Wtsm1)R1;lg}U=*8zUF+}R$?ERPb(;kM1v z!EY51wY7p44l+`(gtcA7za9>Wc}vaV)j;V196`>2?&xi?`3xY?0jf|`0&W3jkaxsd zfMohySGRX8;3FV9Tg@O-dt23Mk$$ybofOy>;t1Y%kKK;f36}#)LiG%kkjLXv55zgf zpf50`DA52|$i(tVC%|?UQ-oAtrMsgP1b~IfY`~#OV(j~zOjcOIIs|%{(d^b#v#J@q zCNi=y@jJ~e^AiJxx@5^^=E{QB=8+R{;$;Nd{T(|O{nibsQSmYWTZ3&Dz>z>P>K*vT z*EN-mU;wJj3e=l^UY2C@*>dD#hK+T^ER(7Oa%xyHPA>>n}dh{@&FR# z$Msw*67#P1soP81Ld1vUBdt^e8xa=_kOI61b#zx@ld##h)3LuEs0LQdF)BhQcY#Yt zuVtHE1nG_w8G+n*p3^-KZ<32#w6_g+kM_;T!}Rt;fw+{sAMb}F!N&<%3vmF_GyI_w zB`9${g6M*Ik*>k1Bdd#G(a+s2y^l*U@OF-NL^lwqP!@@YBt%L1LX_@cIfu}}&>~>S z`ZuJUROdZIag-j>vY&vvYoms7ex5|)#&6P!C)8C;nG*7Kj0&gAE8TZT<@+R78w{zs zhUT%s@62mv!9&TIm@Y^gv59a_%EhE+TlMhe;{ty5Qwnr_{>d|>@!x5&nBW0NVIhAA zzKmyakvR34pyRk@`b7ez2sk%hA%{#JKVCt<$3se&N+qFFC4~|GXKkx&W^Jf;t|_VU zwH;bdx5`eq%&)Ttci4XLW1&Qg@NmS9q;fa7UL*q@GD8DiJ zMtJ}>1UYLmUF%}#^J0*+|L!OVOQ;PHtk@XT(&Q4#D!R%8lzv9=EB)T>%F~?y(bf(G zqX94CUZ%sPIi_5J)6HCcEDxZ2er-h}%%m5ZCRH@&szJ$+19i51TY)P?ioaW2xS=Eg zL8&PB-aK(p=>0lLoB$rn+VXzf6x>%;oU0GvZXau6XbhHYS;LT-SG&NP!*L7^tTdpN znw}0^p!0QW0Q<{ZH~`&^obUKT5~tLga{+_DQN)_dFGMaG?w;e1h!_su5oj6Iu-10O z2i&RCSsUW88V(-$YT#~ZbT=KU7g82zL7VE_f~^c-avtFvot}J%TT36j`WeDrvc?D0 zWEt7HgpusR3iB=Hx=x94ld?RZj^>`dmQUzd4QP>nRXZki-bHW1!8xZwH#$?IxGxzb zj?T|ejwZmmp^i#|xhBmY*j&wwXVZtuEb&}s=Qx4#qvPL$38v71l4FOdTZ3!DNBHg! z#AhhXKm>1|58d8xAhgm82o`OY`maCGd>&{jpp+sC_cJdiAw>Y|XNm^mklfs_uI>>% zeQ>vD@aLmwL|oNs*iabj>kdN#ML>-B^f5u-RGDzFc2oqAOU5>{e)s}~Ed4%QAxRj! zKq3nTru>~wWUCN(QMY_13M*79*slo@!YueGVR2nbY-!@1TCU+NbrQDWo{&38d5)Sp zYqY`(f7>_e);zV{=Yp3#Fp53@1WFX7B1+Q;@VsYeDz_0W^!d@Rq2a2k+ z8;!laRR=}M)$ud<#W4X*Z$Bu?wV+u38t{Ovv8a{O$E@@?Ymt580O+xyG!km2i}A!M zu;F3!qoN{-65MJN*&*;$I+^P3^TUsb9TR1H{we%p0HqQNqI{o>;Oh_ay8f^KVxJcw zMxeRK2&ZsbN5?W_3ErdO;19WeF$=N|sD?p? ze~<2-8*j0ds93PZf$Yu-Wl5Asr#gJ+dX2NWu?SW;>F%z&VS*RTwecKLb)I$+r1e-@ z-I=JKk?7VQkltEx$6hcnkN&TlqmS?B*m1-uAzkzjWL<@PSn|@CFB1xA)7Z`DZV#i_ zciDQ)+dA=R2L5Xcj184+%@3O_idGN8=IyF|km8_XrN-k}g~>-hg+JU0Fw@1Ps~Uq~ z$)bTwYDjcV#Zan!qr?Tlm8{;ka<2V11;(Z3@vt!JH;aXGt12g)k|h@A3g)>gUwH6@ zBRCNvjD%3$!Rlex_*-wkVnn(253$9trUZR8QxBoPNS)R- zZ*u%&N00CU7oT}e(jvl3n5d$p>^6A$ePbKTJL39L`BnE;6X3U%rb2(#yUOJ`WMCG8 zsj=Kkk?=hz2i|L~byQx?>|#@Nzub;rPTDpW&?iwq0Td#Fmn*|q-gut>X*J$!rIK;| z@Emh%n?O<1lGew{SEz>gs8#9ipBE%h*VS&WjxA|uO0a#clR7<;_CI*4`R(VWTS#5L zO6wK3Qq*Dcd&v$RI>X&6_$)@Y(IS?Yz9DA*GY289qJm;!RIsrRRnO_Jm?EOT9=6L zw<6|Y+)HLZNG2^*$EN@i-H@57#xe3U3e_4~1tT|j4@ev9y6q%_HwV5?mNGFbpfkd8 z4UCYMth=&r)-ELVj;R6Kfvo8fMsR$7T6(9X*KP*@e1*)Aq|!U` zAv3&{c4#Xa#SsH~=+s7hAM_1L9T3c28-}+ABl%zlZ`FDOv}hMtTZk#=4k!|oDm(2h z=$OIv(!%q2JESc#rc!Ek-gFiNVW7=QVy7aTy0xTyzHx94>iMgKM&>cp8U&~hHy|2pv@Je(GrK=As`(qaD#;!w{hZ%p&Iuy z^EW{TKLtH8H_Y-fTMVRALMxr7A(I~Or?t6?8s&Vc! zE({szMOM#vrI>z1IVqzZ9?JnP=VLz_`vsmFh608jAQz~yT%^-Wy8;>jtk0#Cy+365 zc7#CSnUV`Ml!dcYi@mq|xV3MN5FOnwX8M5NhvECL71a86lM+6W47o<&MO8444Y#5b zN&mzjfPY9D?1iF#1)HklisT_5Pe4n%V~}15%x!PSewME$3ViK4C^{$VldI6~?^ads z)HfM`UHUKgIR9h4AwY&>+(O0H+?Ze+-?6*sH>U&?TZ<-5U`gC4zXTVt24TiUc#=Tt zU6g{P@J;Gxr(Y~umYYL9=>-D1QmCV$ls(ALJ0yjts(m?~MrU{cq0ZU6u8DXw@b>cy zu=qHLLkIx#OabnP6_adiQ?`Cu#AP!bh8#IY<6ky|bQ5yV%)j;t_h@;-+An?ePASeM zZ2ZocK)w*hzAsL&*pr5kEtn2j{@9?@NGq=`Dc@gB=a^P(U=#stIJ_5*9Ru%tJqxFy zjGb3_QHIQU!MNo-xPP2jwTvb!)jEe#cw+cQIJ~6s=p`{5;dD> z?AZBg0dD`4#EWl>)6MxbP=!{P_tH z@7eC)HVdxNmm7{uR^`xDcr3V;!0AYodQ*aD*0P72J|MKu9=XU#zg(;1L?X4)Q+NHq zq!oeq5#aRd6t}MJKUJ4`4ss1!^sGN?jnX4w+|?hP4;XQLNnb1O#i2RQO#sUO#5n+7 z+#%~X1xMhhWLcbCj!(7r*z5K9c!J|Kbi|&ne3~BkKrz(O0UVr5Yp^&Eu~l!SyTV!O z%7wgj83}R`MFI`8IIwUh>B!ALAb9VPpBLl}zj}XN?EeA))Nm&JC-}wuzu^}v11r=2 z4ZmizWu3Pql6tP{mTm*b4nrV{dlh>ptTt#{NWDp1$2hJdVIx9Bnk5Bf170V8-QxMK z{nw{!r~pz3-ek7IW8*9k1$@m+dwnufe~gpJe3B0R5AJ+=(&g;le%6nk-H?8!*OJlI zz31pkGTMhR4axpR-Fw0FP3REiA|hGo61C*(R=ux}50A@}hb~s`<|p7jCGt!%|IFW6 z{Tb7H2UCL8$`G;6mfAbyr7j;NqjL_@mH+j0{WRdn)~FM;n67i+$hDwk;>(;Mk?mM| zKH=~SgoWAdGylz)ElIg9YR{gvD2iE;-B```_E4(?>}VM3lI7h zjH`eB=_NM|m@rS+ujbWoJPa*zHQF}mDObKpl5w+F^G=ta%nVjBOcsls)GU)c#hLuz z^`J=#*gA zCY*VQr;$CWExU3wnK@`hITnhJi7ni{<{S^&M<->)7Fd}BPT4_iZ~qQuE~sUSKzzd0 zC>GMGIWzxIRT0&ebL{niXCM4JAQ&P%DS5VFv!U3_x6Ss6a%>}$AK)^P>s@W9Kht_y zL}8;S>#Z6DutKF6K_3?%w+9lK3oyI!30gtgL3MdqV@Xg^hT_0^0!Q3)K-h?k- zC@<%%;Ch4YS6Mf!jNq0s9Sk>DG`WFbjVSUlfArF|#hT{>rWUp2*^6BkrkXw9Ary+u z%%3OsXYF~E_Meh)C=5{xAidgWlP8wrf9f7kB2aArMvQPR5?&OV%SX4a`0ldE;Wy z%p`XJuyYrWnGkVYt;R2MPEbGGF>JMKoW24O2q9T7+WL0}#iCVrmJQ^z3i<8>;*Rzs zv`k*~LB(aU${}!AKX4d9#4{_d1CtRgyF`}nv%Hi7WvIA9d{Fg(d2&`wLpAIjyKq5@ zC`&>GS*e0YfV4fW zK5;A5_(0@}hvT{sIU&VCj@g)N3J*sv_3UpgiL$%ryEfd*QN36MFCg&@&WqZnOdrQ{HO1++z0s zJ1hw_a6DiWX@#@AHco9BtL@DU=9ukOP8_AnV?k=QG^o6hY!|K;SKXK8|f6V?aSOIYn4@&Q)%xZb6%P*thQ}+gd0_2USv7M0qn8=XdGgz6jWp&b<3;n zp5jaMQ&DbWYCN5F3RdzCBnH$pjeoew!~$~uB;fYNhFVCUv6 zMYRlMkBV)9$zkBLdZ*fFmrWmuuW!w?W*Iv_6BRf~5c=rzXd$S-Lt3?7leLGBC-YPZ zN%v`=s4lz(D1@qz*ZGNmNAGtr_gr-;i~9NBlmig!w4BuRiuoNFm+e7on4ZUcVgHp_ zT>XU05dRpt^OSkY1zKqlzV{A5q63^q=B5aIEu_eK6a<8FFtIUE-F$ryst`G#*qjFk z)4S?uP`U+%*$AN2MIMrCtM-nkt8o3dP7-Gw9FY`bjvNa|v(lx5&2!iUwp(@qBa5+? z#Hg@!5Arz`tNH^Oqt$e0xDN#OM-EOM2^Ya4TkazQ*a_Hzvo^gVb`aeE%(H2jd@xDC zb)mGMQx_z?XrMIZp+%G^>?{NEMVscS90{0mI9^xa)mn9T+dG@^$;>{C3ydBSWSU_y z4>O=5ZFW6KarrrD`k+B)7;n2eZTLtXh#{_dgDCd=5s(OR@p^Om#CHkJI)(~0v0LAe z3ne^g9zNsR86|>D=AR_pelN)en;Byn)@Nv>!chXwaFyL=h(ru>&GOx{XU8MTf`*bY z^ys{*KfI`#3i~+|<=?hQXPQG&eD2-4*M31NtFV-K2$gY_JHaEcvw_w>MNC{tBH+V* zi-;L=^$YO|kQ$ppt-Bs;zxfJ4)CjCrS!p5N0)G-c*l@{S*goV+y+#2fw>bV@(JUa} z0HPkT@GxLPdmbR!uN(+u${CR3^b<2rLn|CG(Db*LcKu~5t*>;3aMMzKcKuajM6@Bl zt#C$I)3PbEckUcvMR5tRn92g2h1I=`tz)vf@oacAWX77Us=6u9C9XR$GDPqQ0dHAp zSuihA=u=h?*(gvQk+J*n9y2r}mzMgL=xwRG0_?#QdHi|7Ux*DPPP7`wE`fP)etE#? z=fIT}CuIfVdd>RXA1?1i_7S4i`HZru^bt2N7!-7_e0RSumCju{;v`X_&p_z2Q6o(!N*#uKuKB12*_r6gkcF$<5n7{T0r`LLjw zj>i}7S3@;Y9pqt95Y-31U0k9TfV{gfjCwvt5KOlvUyY6>11S{YHFM9EWw&VmGPbKp z)@EOa>F<5+_V)1Hl6{X&ObB{2w96b6M&@Kk<3Vz85N^#ED>C($@Vj8r>;YY3)C~GI zf>#evk@LJtc}&Cd#OD-SR3&X&({lzXGQEDVm&qOgACOfCI`-`;TcH4iL}ADVqz#x< z4YSm`41Bv~>c*Qv$2!>H0GaU8`>AF%@Wvjw&uk-4cdY8rJ06IBY_ z_7Y*~#t*bt{jxVIWD&``dqrdn(16MSrvqM=JFt_N8i0KU5rr~WF`?i4y7mhg?hPvT z|5UHQuBA#$hWRdn%@yg`NxBOl=X`V@YY`bHd@fxZWkxDt$QBF`Cdl%OFaSswc}0&z z2E$dK>E1!PubNpqFCB_W?}dXW@5TS*)Q?-jj26zuBIs}qJUg93kg#H#Y9fyZxl4n}9T?NVma$@QVN-=XTq?&&<5fM|9P4x_Jt4UjXLn4H8w6 zjK1tyc0fmc)*BPCa6nkH($JQYFRJ-*++x+6+BjX0lKZUuNk zpuACmH3o@pgjgosIhI(Qq&nkGO+WmZOe7*6UAJ|irc$oR&x~AbT%0jg@+XYdG@%3J z+6~@uya*L|nV*C$1uDX2n}9-uj3w4(G8v-?lprx1Yt5+sRIiK>Hvvao8zOP&dg(j{ z{8{@(vftiz&@EuMor~-_XYL=`nt4sPq)AACnnf#aMm4OXf`bQ}U znuHDW#u9Qzy=Y(CQ1S0~p$1>TsmPJ6TSD2OrG8&mxA%=*Hlw)bSpe+YI8m5D{YBBR z_Mc*2+U!%C_U^wT_HzBU!><{@Qb(L0g&e88WT_B!vdj2Nu#vCaZY znuzNQdLgmNNWVp=SStEr!pjY5zyt<@zr~p?EWpTQWI2tTA6a7~&X;o2>`s2O=4Ph~ zud~c`O(b#q9pRE}1sP*e2-R?UXUm-s%U*7NOs}93ISjOY?^G0i`Ai1yT9+p9PQISY z2-C4gP^)JXI-zIi=63XYJ@C-{Cr@5OZmI_>0sADuV$4?g7}(X60Gjc@P84Sg0z)_a z8)~@Y-ye_YGW2z}8QglauAfhOhxq(0iDYjN{bFB-$yeU45z^iKD1o4^8B|Xf)`&w~ zLvbdLzdvTZ%$$E^J9NYBw)DNMXpr|kExu41x#K!_?R&R_j}0(@w$>5nAB>bw1aY_! zH+afV-YU)vT*&PR% z`-eMFP)$df@>Ozlcst;IICk62^j}htR-Ys2eYYbhl@KXT%27;&<>4Km+^|PDH0WE7 zLsbrk=PUy$BQTmXiOPOhzm2ra;S8B2{UPszl~Kh!6_cp@LEbpjY%N2J1ld}`6Mcvy z5m6k?x1$6O^ze&3T!l}MFJy{GkoD{iPd_GC3rHGQKti2|q;!yQcvsWan*N|a90mna zB<6zV1}SQ{n;Z?(ncYyBF_kz~6YZs7>(#2FR2%7bDqs#!C%X}(SZ&VgrnQ|4gd(62 zK8y)3OQxClmKW*wyVQdW1CoS`5(8LKb3Q(TfTib`C)LFD)gF7jPX&0n6Pa1{R}6}- zd7*iNv_HsKuXw9MC7diKK|rZrjD4d8rFa4$LiE8yHz4{ebSJ`ne>VEzx+E?P3ul|t zN0IINu0LI?rQh8o9d%aI!rcr34q(8=#dk4=R3AxR$>XmTi8~6SChgUZ-Df~{rZk#1 zKk>s$ao>AHR)UB^%DbUu7* zmx1L25t6l1lL?vd&R3sDOc;Slfpy1};hs+)IB#)RSq&WFPvd?z;3EhG5P(Z?Q-+;U z?nonLJQq=htPcqy;~ZCB%5%E7zOaHt1P*6E55ASV78(SOOVqQyuZ$~L!(1axYs2N9 zYyH}m+6YCcVTS*8ohV(T=t6QED(8CiRPEnNuQ%K}N;N@zf5-?v8Tf=Ow_7_P7*n0A z{`jL_2GIqJF7-Gm&l+FJ6=yi&Z*SNwm;$Ijl{?2o^qgj5El-~JGuWsQGcAY02)NR(FwDF@ z2W3??kCN*8+TzWFV@)D|Qbm7TtJq`d2SHhZ=7Cj$(c*wcGzGhfc6f8w(cq=LPa9HqnRxF5IbVYGJxgY=mG>-VjP*m)F zGs0Od1i5;^?;TC#3MSykH*=BDaGy%z&2fQ=dKP$u;`)`B ze?H{BBE=PR+!`_l3i5)Hrepf$1 z(%36zIRY{Sj|ZcW0E>BiS5r@0{jeU_6H^a#StvAZrx^GE0#8N#x`T5>-dZxu>I1K|`_TqkYIsM4R?YF_a0 zK*=?N0WiU?WBkrn&KpB`Y^UoX6Ejpk31LUOxFPT;fdalpCjH(tRessUV0oo4^xwR? z*u{y1voc~idGhS`GC+81fsY?gRc02}b%lvxYZ}PeS8wUQFwyo#> z2Lo~PrcDw{x2!##{`Lld!sPZD{&MC>QvQ4M%sC4Jv8w+pG#6fPHA=v)hU8hD6cxIO zj$`g0&A)_71r~bd?BafWiSq+T$QpOYh|?U9Vhdc1zKvXq6)zD<+er^+AA;OY^;9&A z<%-->EDk$GIlJ}jLd7q2OBtbZ%C_qho>V%CgTjARE(-_Q4d~0(+#tVfIBIisPug5S zrxADCMALwi2FtdVIoe>_}$Kiht%_?N^PR4Ugkb^m1b{nIpKzwlK3 zD}qf~eZSk8{t3%*O0Bs)pyBzf-T0AzJv3FD`(@A4Me|T5b#Uvzsm4(&3~^1ult-@f zS!+zvRjz3)O_w*9g3zGiuKC-ie`V#aR&yxquGJIe=QKfAr=L4LYq)8k+^u%i?PI!D zR;c;Xx1)aqOGn{sG_;kncb8!#nltn0n(J-%&j{0O<%0$b8!D|lHft!fi1DTkN*u&H z;1N`E4!XLqIG@wxl{%VWX0tm0_E(g+dX-aS?ar>(CgUm9l%nD{EGt;z-#>gS7I@&i zvJ1+cd2S<~0l?hCw*H@n*`&s?P*<3>nZ3ve; z@VC9`kdC?g2V+I0HtQ$QI57lDu*`=_ROQ>9mA9Bdxo42Q!x)LyAVq1>Lv!@@lqD6M zRZll*UB`J_I;Z(ZZj^?8m|VCpIxG2`XOT`D%`>O^2n;;L;bN-%9bjb_njSAt)83H& zT&K6$j+#rAIhs{FGABLFgMCBFMhH>!R^ZDa>ck18)iY!_BX1abhs?XEqmD7i&!+g2 zkz$~Kr#L~V13Q6~v&lscz~D_t7r9gR$hud)fZMFnY4B}c{adG8fl679cp7S0YqVtB zJT#xXnIy@oLy{3>p0FvbS=-I8)Z@uQAMi+e?^@|YcRM55k{tPgr8O=*{bu8aC}B94 zi)J;5i_aVS!$AUjji{|tNMId$ZXo3lSBf%%YYHGk`Nqy6m!_1=l5ES>9Azo$AkA=R zpry#v3`ZFly5~hn)H)CVOcRu5ke*}aqG@xE?3(!0#2u0WC9kcUx*|L$jvOZiT6!Ts z?W+-1(?pd}v9Zeq!~{7EzCe7*tn7ZYuKTYLo(znz0||UC574@j+=v+bRr@DW(*-&= z0ua{BH*YTM;4qm33`qU;o2;tfvk_al>6-jvR8!#F6E#m5LN%-O9~FP}^n&4rgn!6m zm(Vtfb7`1d-3CBpUqHA%(8JyS+Lx_wdI}c{4@<{fU__Pqz#Cc#)OP@Dm3M%mR0ucX z4o$$)$Wh|MVu5V7qhCrNrWol$@%*N&9YuJ_t-~%PYga1LkW?VdVd;lAw<)KA*lUOo z$%>Hy1_Hn|h+9^I1kzeCmWQ(coiUgT7$c+t;Lz{^@bVv>Yc=p0FvqYCKZg^K>+?q! zC`Cf^!)oOGH1Agcv;^70oF*~ySK(0{6omn}bjH{N$GAf}ltSF}MEw;>YcDn-TiB^G zZQT@V4ROonFb!Ay0;7~$BeRchC5G5ada;8@#xH}@M7pw2wSDyZ7$MC6hZhb27P@WN zk$QrdxY^G@kNz#%wVHF_#%LgOCM@P+fE~SW8sz}q7JzD$gvBNouyc?D)=a)&PxwM0 zu7ph%l*C8{vjQVoBzR~*hqvm8clC?-(nx9Oo0s}QW@rm3F-_hSO}Vt}B&JD8{Z0Nc zWx*@I~}2T=;{u4NAN&Y#Q-ur!Y>FAjiQ z9$6BwMLgi%dp6;$w^}mU_ClA9-5vC<6^ipeU#Bf_xOyzym}LfXlSLVC33hfq9q1=` z??m`PhH=m^^D0W|4QL1)sEH;Mr}?aQwBMm1jq6uK z;4T&eyRiRFoViH4nyG?+Jeop}Xpkri^^*%yuM43NSd*o5Wsh_cz9ni6KA500Z4s{tobVf?o~N1HX)6D7BB4)o~tD zt~>(l1WLwhJWe@mr5JG!#ke1`976z=-hq(|UJ6jf4ebSKVqhQ#^d& zBA$pI%aP39drMU{33yHF4nm<8?r}LUNQY7yMGDtNvCZ79}n=5%Z5 zi^tGOc!qHEd|?VRUQ)4iIzgP_07Fw$t$@KFOZ6+Y22H4R8LJ&75xjoWO%VLF_c_rZ zMv-@khPJD#r>CcDXO%BtwMslM1_LFkBmb?qh;1kWJ@CeTTIn25ZroyPz)x10RWlgz z;~_otMDtUZbh#R`!)}TTct-M}7oFQ1AnjGKvy(sF zPaz3co_hW>37V20U3 z|Lnd*2bQ2h6L`tGL2>aW4p&@r7w;-cev!1|F7zgDws=mcF*jsS1{qbjM%bvdipy0~$K!NijJwn%l!@NR7(wn}XjIrHj5{bD+ki+r2PQPed7M?J^55dquwux4Q#ZlFA zM%-&w5=mekSw-KxeZ9SZJ_SlLA2hAD*^9vdBr7rv2EZkj<1s;`0BRwlwu?Xt;w=oQ z=mvg+->$O-{Qayr19XBHE6PCTx8(aPY$hAtVcu{}oO|J=cTZpQ1T+&%!CyN!@y2ii zLciOj$QPdldF#+Ff!MR|))bt#WN=e_TqU|TPYu3o?s#bS2kT5f!7#2$Yi*}+`o*RQ zklR$>8@@G~Xx({nStw%l@Q0x!Z))8npx)u+T_fMw%n(tiNqGkJlna>{V0>Wd06jQ& zZ8r%~_#%0*0#k%4bkA~QS_uOeHl2%)T$KO$nGvK8iDMr^FPs_31*ZLzhUj~*h;N1b zXg~~H>+cL1Wx|<)NWM;(g4*f$QZgHQ{3wJ;esEg5!@+UJI<*i|~#v zvKgTIuM%s(yaa2UKc_sGL!Q;wa$Jf)sK9yS;U(Wv~2JiVJPA4{Xt1wJR8PQ z63(US-&g5M#R2+v;ZWR#(d-y%cR`MLrm}&HOf2qfggyG4^?USmh?9c80-!T`<98w{ zinn>&U$Y^=LK!p;umB5i?-^Qz&9T;_bA5IGmrcL}k}k?T(l`=~D@X^1y~Aro=-mgg z`E8j7QczTuvBy7wyinDtwY+W>GK()7AQR2Wu^t2sSs39s9>s=l(b8jSYH0PX~mA3i9EZR&Iahf zMc&%@Y~Tqj<+3+$ml!N3L9_?HYrc*6$jg@i+-B3)2br13f7QxBNb6a$itnzOlW6am z5Y|YZx_o{{O@KRc72sKP0n$dIRCikZrr<{R*rhAUCr!`z3OYCJ>83ed0d#TZJ|7BU zKW;MstG`v(geph&G)d=rOkl(AQ{jh(2My1E(19I&B|f`tk-ZEK4}bV}wr>a(Sc<2- zn5O)zVM07+Ri?@lxfE-tZWBAwNf;|{uHl|%RN8YpOS4J;z#+W1z3?c^J&nQ}-_HXt zc*H;m6u@xPAQ0qCF8JkqN@H$rJ`Py112$^}2^Jva44U?*XGTkRyS3B~zLRM_zDw-P;tkSd3xhujLD5DW$zKD<(O z^AqS$I(j_|TQ!2VX}D{v;F2y|V%*AmY!T62s??4%P@HIqCj#s1b(H;%tl!}_U~<*I z6-;^wB-Xc*x`Ch+Ei7}qQ{1WuWR^mC`I&x2p zO@~mVVM(2^t1+oSNWZ%s$@V3fSs5G*A`(}1<0paBlfbN|by1SFvv-jakHZZF_=Q~{ zlt!k{R9WsxI*UB@R+Z_~pm8tkuV2Kek^7_%s#Kc0c0y%@yCh7RrzTA*uAtDPxwAME z1>e-q=cCc9^tN7+)781?_X^E;vZu%;ml+#n^O2u* ze>etlq2P--3Xui`~C<*h?%$&U!xC!{zqbi?7jjKf+J$~ zkB+?v%Ujx(RZMjMmaa_`y_#Iq0Hq-8)e+rPrRzD6y2N@I`j24iw?x2=7vrz@GyE{w6tw>Y#903)5c`*l`(G`-$+~T^HEvtq zu4IL%In6t5u>c+fJ<0uq^OJI?>ODAKX)0_tXF9M~)UUfH=P}==_4qy5q5>q=G_yD* zW@i&ON#+BfRN8J+5%l5ww@u=_@9x&fDH%^6_uG-={o3Kpyn)l#(c$S#xHLAZxPB5( z-<$7&_sBtyq*CExZJlT$nzMHeCqf-xkJrnEsp}0hXRBDAEQzNL?wc-E{2;-_WTBVQ*a|FGX=7Ql)Xd1m9$-btRgPLDP1x_3p@^xo*@2Gr; zf#gpZ%4yXDoySn163Vn#IV})sQ(8Sr6VX0KN)!=Y4QrWXnDb#dB zJP>qv6a8>hrKYQDhw;KAO^6chw3rcU=^c^B`#h)Jv$dZ7el(Km`m|r>F_zPG^V{SD zX{L&{6`jY$tz?-fpl;juYA0cqRL-I0 zC6=JTK7nkh27&Nj9=<#^pN>jb-Ja{qk(c;IlD}yuioV&Z39@PnwQgCFXhN4OLRQ$x z@h2>60h-!O@y+2%`m8=Htfu;HvXW3~pxl2wR#u7$f3CC3vFkil`(s+5D&0EJA8s<$ zsh4C_aH|9iZE2ECIC87s1Y+_{fw$H5col@#?=N*M8#F>j)4z7de9KIg^;EMe&(Cpf zl95lNI<3%p%U15PgfE2s%#T(4$jl$xLwPbW-VLonyyHF2`9aXIXexh1L?X^XAZl`PN4q`!UOek(kxI&V~x&J z!T*(=FCd$DfQlnM1_1j@is_YboPmlb4V8~3lZHdx25wOFgD%@VzW&>5JP!D>3<7ld z!SzH0P1_}@jEz7>?xud=qfXoP5UIz(@5R!ChZDA%*x{Fl0iqlasI+HpGK%RvTmjMSQ-Gjue-8u`aL=Fb6^ygFY#(}&k?ypmARQ5+@qj3zEjGQUm?KG{qe z*>r98pz1iIl4DJkzE!5tz&^0QfoK=E#SxNgW(>;twu`>q#qOlRJmq298msi5-e1W2 z?E&b{-eHn;lpK_A8sij92jLlSjJm7_>Ap%a2UYb@{1AA~E#Qlu>N+Dg@>Em(*t~Hg z9xll&j?))*TTRs~3(6^db<dfryquYp}*Pv6xAsRe|hm zQU$gqvr>O5D|bi1kd*4FzADhKD3>ART(?`)bvxF#OrtP!T}(0}$ZNcQECj9uXnlgp zu1`EylggSFfDE#`Oq)?6s!2n^mMYnfBy+Lvs;iBQymoXJ4mhB!hKp|K@Id&T)MSkl z)o>h2jAntbii16>nZE@tz29mngh2QWJ|bHu?U@XkId5*{fQvZ=Rahfef_g|^@IpYURkcoWs;T+?!t;-_3ZYy_^UX-feH!Ad){f#1FlBstudEEX30qFq%4PQxTn-Z9 zBhs*3Ye1YMv-tigq^K!a%8hJ+sO(^vNK_} z0R?Q$IE7mf9e@Dx$*-yl*g37=)Y@?pJO-ozjddH8-$!8hP}9X!hH}C3JoH1Y;qcrk zCMQ5l1m(Uh;C75~F(z&+B69a`HP?LsIfzz_<^0s_)W@;X>xm4mxwOInF}e*!j{q85 z>AO-rDd2?AnRkH5aE>cQ+-~;WJM1{doQHwVNYP+xzIDQOFn+wheT2%~|(4G%DZh0!w$@1w)1c-tge{X1xJYv$XJF2npPRz z#B*Erb;Ude|4#}Ehrh_gS>@GvVcUAV+gd=K`U0GER@HgsSJk#P?=fWTZ<=x7(p7U{ zAk~cvMm-#7JgotYtVKWI2+~;>n#6bnhxXvwUuZR*1|-eeL2d-B9P9J?C-{N;ZV*D~ z@?&z9wt08k?bC$#kmhe>x4uJG&l3aLiHvWL&j?FG@l-rnsAZ$JVjv7_ug~)hO*Ev_ z6#ND>0+{6$3Pg$V#&-%es3jvx8bwbLl2L6+i9VlUdr@aHxn}9D@@e622q+3DdArEk zz5Cx6K6V+-jnP2m&@*W1ZS@EIKwt56>Xl((@5qR1Hz6cj(){o^`YKP&8hT19Pl3wL zDg=B9c=~_U+q-LbRQ6sr5XWcx?ul}PyVMZD0xWQii)>|ZQ7K2%|f-2D;IbBuJ&#T&3QLNd=P*QTys zJ!LYSs%`eo-Un31${f2hX$S)mmz27A0@|Bf0r?7u-7Vgy5PO(htkwI@Z5N&X1*Ilq zr%YPr&_?`>&7(DRJ+K{S>~K&8$(Cd`Pr&Jv?DWe81hSpFsZn6K2r`Fge~0V_45`g> z?}D47uk*(tIXUcSd=#@{rM%#G7O!m1Z-Ob8W;=H^mu>O{Sjqcv{E;qE{)!EW^qFQ6 zr+Uxx(v_)3VXBhoy52nUr>max>Yo%yiYhYyzTW^(^HSe?okqDWpu1I+rjBYR3X z@BWj%(|lvh8WZ=z5l0tT=taKpuZ_5R)zHA=N#@c*zqmNDW$ZulXQ3&2VVSwvI1s_t zJDzO#BJL-%J!=GeD2RviBHlG^JH3?HeNjLH?2{L~eD3yb@;q&UMf8)T>MVmlSd|yH zZNQxs3U4Pqe+y=ht#=0Syl1i;Vh!ZJH>Gs>7=u=A+*IEWs7ore(TxO2RZYVeFB& z%y28u^EGcuK@F>hWQ18F!c06l)c1FPbWGjJs8zRWJYco_}wEJM~Wz)x2YAZXu^DFKmU3|+eO*4t(p9>)8F``q@!ld%lc z>^A?oQX(_;=;g>-+ehqYJ*>%DRc4A;{bPcWg*``#(-VZ57iNWU*ARz)q|93k>$gQp zveuNx+@Z1S#*LYPD2BiInkmklT{UHeDSSgKofY3)>DF3)0Rak*tT6uQ+Q6kk=X(Zc zcpcU_05k5o368QBlAoxhaLvo!_4%^dXZ*w(<4pHcIQbC*2r>#Qu>9p^@OWV$AZ$fb zNc%M%(E#Ntso^9e%LWvGscNY3YED^W3FS`jmP9c1698h|9Vd7`DoxGmew z6+>LhA~$_B#Wn?a%q3k&FL4jVxlJUNm~^S`ml`X97h5!ypy8qAZ~td#QfsYH z@*`GMq^FnjJ-kRCiJGM$?PaB9ew2Z*%4*0qPZTXf0=muWk9s+gG7V-QV1ql&dGcPG zgad;GGfH?5trIJ@kbWSZ>d^V6N%PVg#QFHt_KD|;D*NUeFO8fB*0WVt;R9OU@`fa( z2Ui8VX&YVE#9+nAf%@z-=Y**JeJQnDGK2{aZ_DfT^-bqVCayJQnG2K59hY&VpO&60 zp*gCsHYHuV&+uwX5D#kA-3La2Ds zzX2lL^X8q1_T5jVs@t>Mkb5jx)w#i(mUrx-dI1FK~oSWz|V!m^lprpaV z>lGF7yXXl_j3e%?pXL8uQt}40?D-+0+yTayhW`}>0I#FFW&BElOn-Y-EzLM&tsN+#9Qat_n&6J8f zdk%x#Zp61nzFf9Q%e;!;&=V$4+T;=e?c!Q}Lk??-i)rWw+<%q<8=E6i-<{{vZw3z; zE&v^gUL9d(vD?pN$B=XY+K2@yy?4UOH*2bFs$;e94eP^AuD-vRcEpa|w;bC^5WSyA zckhZ}=I=^Ku8q52Mq12QrX0)LMI$KXTZj_qk4TJ%0F_215$`jrhNH*w{( zp%~zhZ+UeH< z2v?Gn2_Fws`Q!{!Ri@u{Agapo%z2&BB8ZB0umJ41jyy+|PEW6LeDT^}F25?#9&9wFyF&LxpyzKRv)> zZ;4p()LB&cJpsn>C*37#?ybJ_b!jT0p7~0cem&Y#><*26v1JcCOqB6O%FHI2(|I8X zdniNWkqnOFPY$ZO%KfDmqRP-H>cOo0`0@OFTPUjP)W4=Xms&2`2m|nUR;Ryolw%1c z%Q0@F!*^LRCopey5p<(-Sgf4a z&@pslYtTA=FsV&qN!eLUO1p=YF>wAhzi(UDv8LrqackoAp+I-hc5!inUnQIAr#pD6 zt@yi4CuvG&^094Spk!*HLthvB<5^q~lvHOhpf1jE5FBG!w3H3km)MUgE0fWjg%Q_`i@GwCtEvc)~Jujcs z)kO}4^T(Na$?7;^;TmQRpwx1Lm*J;%xH-aPD_bbx!4m-61Q7@*FElh;|0%Z5euMnV z4weV*og_tq<50OnV1My}G+6MxLrB@DqFiv^!&MdRHkrtoYtw~HOI5HO?(C#(i^>26 zfwkdFh4o*gVgX4v4UKW&MAv;T`O|>wOkY#pdf%3H)dn3Gz)E+!S-aKjaA(XwWs02i zYEVo`ENyB4!4wO0SnZbOSYI$9!_p&Nz42(QxNG0N)`VbRnK&daXX%PTpQ;a)$^Qhc z95O)}XFZgpoST2Yc9i_e^2)(!l?No1eo;?sU(w|0)OHzgaf(* zo*jhQ*XRWZjV1v@wA3kqyVj|hyO`KRky$x9x<=XmxnGbA zM`X~>bE$iDd{S#w35qoyFtCuv&y)hZnG6d8MbIdUnWrU!`gLMIkEqLOJYr?C%oL24 zM21zkf`^^Gv(5TU)A(h#LRQ~BL!wCgVp~I=Tf$(%%u>zvA@kaMp-7xe~) z@g4SWf=u;-P$5*U%q6%_9QMIXo!`OPu=bF9#?l^FqReh&Dt zX5l=BkP@()x9JV~zz*j64MgIlQO|F`-wp=$IJ9fXG-9S-a0h+c_KdiYG}>g&eoTUO zIt!m)uNPij>g)RJ%iD@_0Jx{sSA-~^2y#h(e1s%)03aSk=F}1JLhBxt4fsv=AL$n9 zlo_$zl@$x#HsHF_BU<-QavRrW`2*1)QKF^=Wjq}!q4R};W~-DQ+^Uoa6h7k}g6x2tR0M;K$ zL;+w!!G|g71Q^L8dG@6k+e{qm-GQz0at*_)BL%~)U5Z$=@7rRHuMk?#hqGkbVo(88 z`lI83>@4XR;qLIF&`L1akkf`n!NvplYCxDu*>KsqhT3RtmS1JrE8}yJjTvEUjH5UR zU#@F7AXrVvikY;;6j=OxbTYzvsR?OZ8`S*#M%~&?6R@ArQ7s5{41VD3qs_iIoaIhV zJRpY`b2RpmEw(S2=X3?6q;?|U(_pZg`?nhX5V?O6Aw@5AnIS^~PBRe`;&=f8Jzv3Y z3N{KGs^-Rx5MvK`itk17pGYuKj1b9j(Yi1EDsBCBw(C&mVr^6!5Rgmyz}u#f89(0a z%fI;wo{)Abv>%5+P=Z)@Mk44;tZ1(vn-0P9S2a42lOOQ-={FRsdA5mhIa@k{!xqAD zR&r2QUFY@<)3*;-0c1!gBu#)G1vvUmc8cKu9d@8xomUH=fSk-2?yY+L>HGGhJG!|7 z(XaLs=kc=p_zM9a(0rE{jvYah0PYLf-YFuP>^9~;@f}rS!?+be^|HS|+m@2(nLVB| z@uUblS2SuC_IwE54aJLuM2Pg2a-hsJqufcB?01+!40Smbrv`<^`m8l5*o~EwzTQUP z(9qdDyVdWTcFcZ!-LZNhhck42V*nm_9*!B5ng-vZmZUz3XLd zW*#6C;l1e5(@^JSrR=jnYT2!1Dd4}~gQ+m&SpGg{O(tw}nA*GtJ87LP+%5@P;UsE3 zLa1bzHNBpmcmF{EnKMz=eR;7B?cW%Qva3-!+~63~lDV!0rV$z&|4z~3qCCGv28!Nxqs z3OMzCh@}TJyw9zsGwv4}69){;OY8|81+h_f#l15aWS7~Xdvf=ae7o|RkDLdvTrKX! z129(HR6wlG3c&i~3z9M5Q_$i}}5U?*1)o#4xX!m}7noW@Mg1C;z}{0;Y%@Dq7xas7Mg+Due1tbMVZyh#GQYsP>n7?2d_9@2 zQ}IBK^;)7d7JP-xaPamEoWOBPm}=_ljg3r=j0zEGtDTCrt|8Tw^^i3|)Rggjcv1Rc z0&EOeyM;k}c7?66+B~<~+m9(QJFp3jLDCiO7K42tUdK2|2=8app*pa?F;Eu|y(M<* z9cqo>(k6UKIO~+>*nU7RXDW*C2m~>N2hnBnsB?!mAAYFIC8d?-B27Pi+%+c_rUaT@ z(>ZKO*Y;F?X&NFUrD3Cw!r#NSYAhy?Z3T7M45bLnPoFi|XBjPy=(6U;d_~skA}_(d ziu#2h?ZkXE>d}CNbXJ`46A-AWtATdWrwS<+AZ$pwBPlb=s_p zIl97SU*^BGhGw)x5uwcjbR8EDGw>xSy7Om3Vy#O*3fd?~q=t=0_5(1v$*#Dr7(FN> zBqI|`bPcI1P#R5N+O@Uc&VqS6>eueYaNY_}mWoU?!^)n54QWy_Q{1Wwdyz}JQ| z2KxinG!SS%PGJn!Dkv21D;iWb zp=w56_;T=_9fSJQySqR!=&HR#1ep_23#w_mwVd38^R|3HWq?=#*1*yL-- zY&HfVZFK>%5Kd8V_U(azm}|{Jr?W!|Y!JX8FH)j^?5qrFs~C>9ArPPlMHcLT`fgSo zHRGUdMB1W3ck5G*B*W3nd)wkik(ucXt!Q9@iV92yOdg{p1$RQG`TxihqAl5K%a=L? zSu$BIs;$Q-`@(Qqib?SmhZ`bO>ct*;htr#Qe_?j>W9*W$@s1DZr)(nWgtejuS@P1K z3Ubb6H@duaxNE>^8TZERD+#9F`MTZ?mm4qhfWBg>Kd_Q<`P)uU8Pmj z8^7>b7Iqh{IIRcrNq5-QqA-d)U!BzWCwW#cUI6mS+k?120L8`FO5mJnzduQ+^j1sH zr@n_#BLSy@e?_Z#0E416W7+YN(X%z+)i%WU8Y!c6IePs=qSa*^sLqB#PG9W`BpdYN zw0P*TxDbR&Kph%5s#Xb(jDQXDl&_ws-2^RNk+g;&z#H^f@;+fI|A^P=qiy=3IRM;) zN`L-!p?B<6fqpo59`6|R*P~+jG3^1nCx8i@Lq)C} zaJ`wlZKAXW)1poMrXRMPC8)cooi>Az1Uo=a*@6Lzg=!J6<>v;24E8mZw!qKDhe$3w z(=QDjQX?Fi=PD#498D~xgZp4rlp!05QfI1h4ijq*3le-SjpT`N$4+tCcL~D|IL@r> zrq018}kKIg2JgcF7tB@yEVtoH#DIJNmN9V56CA# zvrR)S_GW;Y!O6>pk`t1shIRAS2v>-JqzMp0ukBy%ODcfkR5C8h>DVJFFhc2CQa5#Y z1IE-mgHA?XBYnhA91nOW{j|b|pYtG8EWX@Gi^4c;#%bgWfoPmte-n&h>jW%V+ z9Y@TGO(vauv)-C^cNwL6dik(fH(Eb%O~Okk6sJl_z0hza&`^b%UlrV&3 zB`68{8Qsw%NI|Be2?{{cwWPHApcJDd?Gt?{-O}q) zz1#HQ{*AB$qg+&ZNydp`&>@G?TYd`??P06*oV}amZySyj_8spU+9VzQIdtrr4D6if%mm^3sdqLOtnwZmYC7JQ|MT%1(4k5HI4%T1yj?H< zVcP#IwB!(sSOxWzMFmf~Jm#Ehsqyw*l@3F!5f1+LkAhfdS*r*x->uC|xWjCtKK#j% z3;p2Pie&bZHjD3Z76|gCxb8C)h|%8_-amqcS}Ojq{ctu3gpwWh#F7;QkXY0?=*dk@ z(jucr_RCKly@QYu?N$V{6qMLAh(55gb}$1+;e)Vs^5F!z2g_B^gFOKh8g1t+WLQ8_ z7>cVeA{zoTMH}cx3cMqw=fFrvrYS%DUO*)1U;+1#^Pyv#cPYQLRJh2J3QUI`lYRUx z&~S+$d&+6J=7Z`BZNgB#vg%Otz8*L% zN>7}j{W!bb!;gCEE0kTb6kaTN*$A6cfzD0NO#C(o-@@=OC9FS>8y6QBABfscm$;Hq zHRbn}CY)q)aNQ7IoTlbu4%V#<1O>g~#&26;ZiO~nbpBT6H_^l+>r5U~`wj4R~ z-JMX_8Ix)Pfq6CwpuqhBNg?{q4}oY{8szx)dhB&%XI}enIFaMO;6!GY|AT99_&=__ zk(BMz_6;c=w0pHi&H(V!f7pXz4@sL#Cgx}e8FAG7N-3HI3-F#3#Qz?)PH7^JZ z&D^))v#M?mUw*N_$3Ra|kCM5^SaL0YBi5vPA3XlM4t+NN_no3F3Gv)Aoip`F@y4r% zLf-+{7J{r{XoGiI_dr5PPBq8lB)Kj9^YHyi-q|xjrc&6*F?xD^$olASzbb_lb$2lX z#T~4y;qAA?-Ir-aGe+vsm#H*9mr`wIiyeQTeWH0C{lRHZIfg2|R52>E@5+y2IWucb z;zhUp3oEO`Efme`6G0T`zF0l`_-DVxIr@HOVLX$JYuplNp+Ix6&ZhHhj_1%(Qg zdVbxk=odjQy3efGWVL0+gj#b}pZ85Qa;DgP%7da}c;RvdEynUWdkXu3rK1i9;~E*f z8#SAqxEo_KBg}E3TK zWg4fIve5RE!|#!?@8Ki7nBQ569ktSds+dxH0~q|9HQ5qkBg`w|z|Jgv@T%`bm5QuP zOXkBP{HZ%tKj{J`M@)#$i(>fF&W2vuBAlIn>Ii$rtRlajHkv1%CntDn7y8TY9?!hD zwn{ve*%DM=@8p)Fu3bO%;cSk7a8=`| zqpYYV*)WoK3FI~&J1^&hBVBWx)B*R$P5A@@O@!z*(M1L)yCPJ(| zMCuS%ZZYcLQ$EM#dAB6dedx8MpRC}N?LVO3IOLJ;mcRda zrGfefH}2%%^ZpX1qp=NKxvPd_x=*u%c- zQoDT9Z0qV>cATLz{q~Dh6DKqi@3hfH^e~gQqnGOwM(;kgh_kwxp^agMQi%YgbLqpkT4no6(aBpUj+I@ibarw49^;W!{eh(IchP z>6Q=_(;&U)Ar>Ta^M%n{-{3d9BLv5+FcPIxEXEE&Nt@_$|{Ts_nQlXh~R zzXN;1KnQrbL@pNMH^D8F_rB{%mDsPzci02V;DaDjB{k~XirSxclyg0lD`9r|uU3?a zLYkDB#w+TGmx+u&Ym_!qTI>HT4zI@i*JV~Iru4Cqz~nhHTXeDbveD#{7_}Uc6{r3^ zfx`y8BKC@`4T(o6z0?Z5`si{pgly}5!$bMel#$V+%&D(cD%xcnM<&gF7#K8g7L&xY zM5^)7N9k)xSkOgce&jo!8m6Y=|0b}d96xutM-q=($tcGDdRwe&?UcgsjAaklxM+%5 zogH}vyks%&nlOca?Gf2+0Y(@{QQ=6cCFW@KzL$uy9a$taaL()D?$pkGQ{0uEP?Trz zLJX&~#>utiSq1%D@jOak=;w+U^54|Kdp zhZr+^jnY$U67`a^zni0IfP4HgUCf!k9UOoU`DwA6bbW>X&?T0cWPw45hZ(Nb)}5wy z;Im%DX3TcHb;!eQ*au?~nRGVTQq~CZ$0>b*kLzSq+QORysnltT5>$-aT)Xo1P*Zn3G8luAuYSOIx0C0TnT_D~7%f^-hV7mWx5d=UeeVT>)&;PitWg+x_0l%}{y24$d8 zXED`o`Jw8~mJIEqU!NPxV1MPpRB1JxfpXNoJ5Yx-wGd& zN--krPM*1 zDkuk|Ae0nMPsER4MUS67)hWxPflox;V*jpaOwO5(JZ6uUb}A|DUsS?oP|HEfKDQ8wbeVn#hqxrfsqv)a_=#8S^gJmf}4s}C_t>6Z$%f4d)!n7i^L3f zl)MEZX(3Q%^XV$>%YcXV2d znQ^!2IN~9Dlal?FmwW5%fa(&Jv<23kTBjHU@oP&m`*4bM9v51)h6Z(^WGe}CSA$6j z*uXCh2n+40jV2%*H1&cuVq5fI>SM&Etj}fCd&%aY^Nr*m>3Fo zV4(GeRf5M_wVDp^RXc{TKgPQ(a?C4qx)#yfmLnj9L1pptMRP@^x$IPa%1Mg7S(UIC za2W!7DBAKfx{C3Vl7;fNnNRq?hCp!<@P`ELmgaOwDAU}96^iBW% z`Zt*d>9NbG#NYEI7uOUo0A{70k z|1dw5?G5X=we%*B{!QzXwNmv&L3r(uyKQsWPRM)Yu1Flb#jmXq+_DC*UpHk&eG}%n ztlHosyGNl0%>4r-0b0$+^~W})oo$TgWOXpg9lI3V($03MQ8hBQ)Keh^K3~F$#QoP% zkV9jAjz{duef|Kng$#PfbOe`zCjdNa$}_}O8$Z$vsxghGn4xqD!os2NYnp`!vvAy1FUh2-?1-b+ z+muFR$j>hYH@PQI9Fnq2u&9iJV+>(5hvn3&2DL*gbbf!Krw-vsb3r0q%rxJ`aau&B z`Q{rxvD{nnvY^|`xU8vZ&}RwWuEtS$M)krMo6#OSvjCDqHDfm12_z>vqJsxvnKSHAsGEygU-LK~D3AN3S zU|YOIQ;0!0^})!_;0#)%OLU#=-Huv(V$A9*`iwp^;_a&|!#e;7Il(i>jgBkcxXj%L z^I94&1zcQap_W$*x~@_|He|`PEg-Q%lg8cSR@Ow0R94-u;`aM-8MZ9|R0QBSMD#)O zqy1>6P?k-z5Gszoj(&W?HSv)rP8a+`Xq*=2XL+_7pJxIQ7I$Mn=)?k$3sF*Xv%09j zlz%THJtlwC*ZrU8t~U5&G9X z{C0$45HdZWYNx_7{H)iJc%1lT{lI5mrvnPvK7lu;KN8>XP$eF_S>1n9Nk+#y2I5H` zNF_Qj1>%ET-}t^OzBT<2Co=epxzmjMIC3K5q;Dqq?W|;}5350&^0yl$ZY{EX-x^05 zT&KbDq%+|R#f&e%6H6m*g({YUbEz&}-xYqiB9O%koA6@Jp@-AXh9tj!Ae}_!%|nPZ z7}ByA=8*~DYN*5)?J@0mZY5}9a28Oox1GHnb_IT>`UCG*!Pgd0TP>CI2t6@ddv#RC zX5hp=_T{j>j+Rq||D?t+Xs+yI5AZV2mROol3#^XQ`?m$|Dav2Gs2EnMD2DMG!7)k% zC6zS$F!oIZ`(Xusd0|rn&8{p~k^q4U2G4K)CU^XHp8?X{*!DYZh`cKbLZsn@T-iRu zmD}dLRcORtQ4|7-k^Yg_>9tM84R^5Nam-wB~!{b8@=m%9%^e2ZBnudXPnv!v){p!x8t7n$1CvdsrH# zGQ{~%6@qI2EHT-}QNlqPGrmM}6!r=lNw2CChc9mzq@Vhe1dhAdpEq(DPyvG`Gp6eo zghTm%{#EF4T`XI7vT*uew*J^zR}Bzb_nNkEpL}4*(+zHMhsNA5uDyXi!`JQ!`r7+q zhYD?X9x4Bp>f9#cyrdK+Yka&mI}^_`d}j#OZ`_S;Z$0r=sZXVw#Rk`P@*+S&V$^(n zY1{1;pcQvWBQRNp3-PK^9ugl#B|JLxX7f+0 zjRy$o+|9uOS@uj$dYY1LHmG>o7I0?)-45oN=5o?HbVH!^$@ytG> zIHPQyhP1bCpnfug4t?L+*L-;3e{*h8Z^Hafe`ozXSvU5Xh{|kX&Nhrbg!WyjrY4vz zmL5$2Twn2{*=Sbu#1*}Jjy94myR}b1$gqBwu#ypkMUKy=JQO`{5q|CCcCZ4dL>3gp z14MuCuI|S&b2=-BX7mZY{pzp~;smYOWe3++g-dR!ZMY^R=`K2p9&X zezN(gvAtQHMMSjkr{>@ouhJ(_&uW7(fDkE*pmL(hVt#!-IIaulC|&iZ)jlF2VbT)E zlxr0a^u$)$a`V?M`ZERT-}-d5^8?~Nwdeic&?V>pi!NC?xY+*xO~~s1G$F_0|1}{O z?QCvamXUvjz0;qHDZeB0h{p9jw_-x+Qb@ViMo53=fScb@J6-FNUsIL(aV zE9}WV!vYeLSUq_jwc)+QLXV|A=?H-0VnLgb!olfzGP$T~i`F{F|e8GmC#VgBCxeoF zr%_><>6y`&NyeAHI%u2%vQxfxe^@IH4?Jf4^+=_A{L%01TIl>01vGt2#H`*2lccgV z@1ul~DHg!Nihr!7r}idi38$mT!FqRAxQODWJJg=rsuwQN+CJ*%be_MCT@u^ zntlRn)_%5W#VK!ds!R-a$c>E!d<`S_S&__donxV1LqSU*%cC@dmrg6EOxHEcaBn zJUE5TYBfk6Prm@?HZ?@@-6d8;0B#;r3$2sfEF=gL!rml-7 z!+}uKqG!|YtD=c1Mh$V0Cbjbk!#2DsD<~b5l#ex=2x*pZqein8HS1RZbo_ZPWT#Zy za*M@w>{HG9-@nDQzfz84dR4MIhQAuQWwY%-6>GbcfT%F(y5x(_qKtPm9l*sRtMQTh z&64feMd4jSKMLa)kXUs8W`4G4VrIrvf#Bzxl|D2+NI3_^x8Y~vp0u?_MfOK6)XD;? zp{)?}`PhV!q#*P~!OxRG@?d=Vfs3U4LF0ty&XA){UjV|oSxf|f9>*ZiV8VBtcak#i z{UB1HSwi$IALbBF4^r3g6$jMnbwMcS&5CCOP}Kjr_=n`4%cZ8jX9W?}^_+!kK_|l0#3rJEwhc@%7Pox(nj}&l==*SxL9UrCl*^~U9h9?_+6aL~ zpvU=grbaN!*eAIln;*s3ZkUT89hCBb%ya1Ilru(zsK;6(*l7~YRHz^G2aTk{Paido zI9KHz{v|8tDAU-Z@7d#sGVSA_c+rK=4aKZzO8ZP52lY$hSj}5QIX6dmHxQ&EC7M!Xk4v-aMc#|jFGjnr=L6H~19S2}PHLeqc< ztD9)`b||SZLT&n`=D#x!EP4{F#J8EEth5e zI&8|IQ8myQMOPK_+SWq@`ELh>u=MLeT74pOA@zds&!;izPke5OSn*r;8MBYfW_z}g zE>kvnQsFyjsUb)(hbyiO8uzhin5o#~L8Iq>dd{gn93CS~V^d7RNPjDs$81V&z{DI@ zlk?wa0S?!X-#%!cF(Kc`ZO(}D0=dQ*jIoL%oeRJ#cOKo%PB97K0IU5N`0}TE(4U zHqmXMupmxb1_G{XyNi-b^~43>{6&eO#fg5(`1R6>evfmqBSFW^d+zd2HZ15Spl2$i zKWIddP@cp{JN(;rl?j0Ngdaz#TrjA;gB%59Wy<>Y`OvL!C@NwmV-$}bLA!Xzd0j~r z;s$0eBAFP_Qnz@6;l4f8n)|1}MuR37stjm%SF5+&mI$hp&S5(D`zW)MKM-m9 za@B00W|_3FT*QI}=ialF}%bJ0|NSVQ2~{4TkvNE2#eJ`xS~a^2q+;Fu@}<*bJi z+Qs1kba{Km%dVTxPvhfq>)V$Y?qaEM{bJi&IZJhfa%wwSLzdW<1J@FhOJI&K{;&d^ zwCD#%B+)|Y>V>zUmW<0Pq6OlkU;8D*sM115G2Lo=(ArR%q@0-YE+I)oOo)N+<$5Q$ z{^9xpP`V6IGwO2dAr|Z=x#G2Y?`Y`vf~(Do0WA+Mc#3tS&;d|IsNSLuxOM^P35gH0 zj;u}&r+68zKRhKA!YtD}YOX)GA9p_B|9)EG0atAIj#092>8jQ()G3|R;&ibwP{FFQ zJ-m~&5r{2~;uU>ByY-ts${G0*ge_*5W|tZe)C)l9meYvME%7jhgD+i?lBOwg0O43c zWLsD<;E{AnFjhSGm+CS}HSPll2Bx!e;}f+Vfd-o#jVGMb%XHQVnIdl#9qHau?@k^k zAjd+HPTF9bZ}%wtGYAMxJ}hJsi%J;0z7njF*Sa^E`OX+Ij>z^TC&@}tX8sUzBI6}~ zFt(f5zQuPnI5(+KZf3PzdwfSOOc5VB~hI!$93Y;E3Vod``P;#=j3g0)mr&B?uT_A&CP;5ody$)5 zp(2$ovuu_v$b=Y#<0gBy@3$Elc-hPpy&Yw59uJJ&iMmEV)6tzp`wMXE?v18dQaf;R zY%kzu&h1CIMSKGwCe~U<2K4p{l__@}DKBtbh~C9NoemuqNM`+@((jy8`|rmI*=nAf zDL)-!A4z?U+D4|Hmp7xcJD}b8aEsH zlZ#poJ$kE29Gx|3MOeVHwSRDxqUs`I&~5)z;9>u2(OUXO!F+e)F0z|Ya&hF@stt{a zpkGn~qV-Jf!f^uSgASIu77GWGCi?(=JcibpWY?pVA2JE=XZW>s`eo~1G*j%Z+pki@ zfBc=pQRJqq>8GLSw214dwxg`Ci&Xfonr{pdD+V%Yq#0x{Xe0O(PyLv#{O42$xj?yK zfTBCguOnyFu54wW+!I*vlDRPz@-g7EyOs+&=%~PTUxjxJR@(s=l^?BK_VUW{a`h(J zwdcd&S0u)mekA>^7DNylu}?s{e+7@U*Lw&0xVLaFlo-)OuCdOY$iIDiWBLjz2xJ-J zfZECk51g@OHkr`*AnS?F?QN^rgLRmmAN7Qp(y%O7TgAVCi1?=Lw?l533Aht#-?!1R zjF=A;KV^oKl5oI-s@<W#+eJJ$a(eRMKUwnOGmHq*!jU#yf z-bgO9%f)DEgNiEN8tQd-qzNoyab7BLD5MfTCV6@QIYhO87dXiGf>=dz&LNZOLq+`C z(ZXl1@=OUbBkG3`3imQ9Adwr2n6q6!o_(am!6AU|*Z^qA3Suo-zTP)z4I&86Wn#Ou zh}YOrru7-Jr?3xuZ%c6wL$>Dj_YjN7uP_BS z#sO?|+uR6$24f!4)HOhC14Xk&is-p9H~dyFTFz=%42V zA)b8zvguaEQg?YR($kvPM(a_ih=t+4vE@nC!-OAvc}R>#=yE6=3ztRoq3p2#RzY8J z9*;N&Gb zJ>){z7>qw`-+JNL(}g436e_svAm813!}E~71D|Z&ESZtO3$K|eDkKGM2IFIuSLo{x zYN5$sMx+IF*%46nnblM30PWs!znEwfL!~hu$o&zMZ%a%td03bgXiaJyb%*U57Ey*j zZkWmakPYqubvQ1@;~m0vDm!^0wV_0K#FiJ~!o31kK>U=zK#Rz)_V|}3z=Bo6m=Ppv zenTk=2#@>9yXq&pG#K*c5p|4ngR(A*%hor~h}%SCUnubtB%Kh4k+;4WfhZQ`dO{dJ zx!gldy6Mt(Q}(UakaQATx%=B+>;L)1i_uqcU~h(t1EgB>QGhZO=6>YgS}Sap^YT|v ziKtv^+)x-)XkJB)Sx5b!Yb(Q?-qaY&*F5imY+Kf0=+%*=WQR3tof-M{H( z8tLkDL&ge8M__YNdMDTS?m1><%7Q1$GYPuUoR4Pqc*9i7PoWUb-+$i zI_FX!1!Qcux`m0D2&`|8TCHEG|2~1H$?-jy#CFJhL1cabj4)64UWm;6P(#1m$(Epoh&p#yP8~?kE z`tRf`6_BVL*1b@si=+{Zbcy40PnB;1nmP+Lu`Roq2k?e&avCFDsTo9EsnkChRp~9037rd)H6LJD zzd$w|z8od&+P~|R>BqtAj#OQQb~knSr5u^n?CcWN#hUzR@7z~ysVH=>@tyqj4Vusw z20X?X#O%tDa>F}iDjwa7gShRQ}G|*iV<2?B(B2z8gOBT^Oj&$-%xXgjF zLdpSe7eWYnz9N3H;Bh;gTLyK%sH+t+Ee6%gg`@-ut#;aN-*m=Oe>3m@NptuV_|mhP zdHwr-naUzmocL%_pWfBRo7t=8gc4y`oUPj<5Xm<*6e?{_;@|c8JmXN~Ffg{1 zrgz{_!YJ*nz8b9et$X20uK5wr-1fzhM=J@P>iCoNwb-O$j!n?i5c@{e!dUsUZt(B*KO%o&WJX46G|z?py|2Y zl*_*?hEnD*INlbdC5-X&P~HTds@(8Zvz^rG1q`YUuVVGwu(7fV?{~PA(tKm0cYE2| zs-f2{E7xv>+398tMnBjvT(MQ+D5&3Gw^~0z2@!qk5@i&r84{^?f(_nTRw91Dp5Hc@ zyJlE_UK(AAF4Xwsy1CrBteQvLWRz*UbE(*{uqIrlzZ!8Si;Dni^*ne^CXK*KSGNtr z-r*|^Du_KbbY)7Omk(WUvgD=;TP=y^uitb5BVg3L#jIKb)Ej7c_3oxYSAsGN_S|lL zZt$#jJr4fitKwgz4Jo1SniuImJs4PYUP+HggQlb2KGBG|{jIR|fQ2UIFM=Q3o=p3N+g`PB%}bnJcxefrNkS-KL0JAESZ}fXo)sCrYyT<)Ekn zItp}!MsJbcXDA=oj}v#Ybp>qMC1V0NNC|*L!eCOdT@W*kjKrsc zz$5{~8JZUXOR}tplP!J_o2U*`57kKIw^UOoePH11pm~}QkU;`))iQljBMh8SgHwjK z6kUNzzl(tlClECf^CGJzx?|^(h4Ax!D+sTWm2_H+*1;r`{7TILuA&EK96Qn{zF4UF z1GGTOX;L`4Z?oTeLAT!zFjF zBj8!vJ1x7e2UYJ2n@v31v{BK)e*5+8j3`78Guh`@uV>GO3Ue@4$k^1s@Qi{YF`v&F z7vBI%LsJAJ2SM>pceq;00FXa z%?rAwYr2a8q+ijdmYIytj7TS6Ro(rY1-h}n;BKt*zV^>)I}j>b^_$T;`GZ?ZJU#ps z4_NT0>if+AYx{R7(1#QVhPY+p$}UGNEkMS!zmY#--bX)3eZ{QazySTIpsnKAW6hAB zm8Udg@*3n;-S*_7o<0$IsBB($o-YczAt*h#GH^-U?Et;`wZ%T3lULap25*T!&I|iNe}L=63A7(xI)mnV*tdsmyV|Y8tzyEm==ST z3ga3PLxcty!)S>I0C*-8=R9Ol@`z4ygFvfcsUGhVa>OwS*`x^IFlU#T5;?iR0$-R8 zexl?3;Yb!F68VOpQiy4xNESAu;#CN!*+5H8hge6b2Ba1#iC^orvu>5;fMbepOsX_K zVvJ24u95vhZeiy!&<6JRi?W`dY&&moSc^0P{a_I~+%PkJqtQl9f|Hn`poHOWNyU_R zQ*ZbeckDBe81)d2tWb~X$NS+a2Wt`t1~Xjh zK)%qNDjCL2CMTL}Z=nV6bAj_SbEWHCzP<=u_z28d%>sXoUilK$(SD7YIQnaBf5TLS zIUMgr(X}25wG9PpE#zs~&lu#ELdQJHV)CClpa~a(yQNWfj{e#wjRf!~gz^mhs-Fdq-bj)mhJ_=R*}vIbJ4Il(nt7|ume0jg@@bMhB;gLu59bUawQKV!8V zT~p*<&e9u5APv7DBD%{Ow5*8NxbBkiMXgC!t7A?#%c^STr)+a78dN8I_fn zWQr6vYCeS=Q|Qync$fxZDF%EzVqTv29P5+_)#$ibkuHK?Udig;b!Zc^P1%RqOd&%} zdbczbXj30(Pu3`l)qk=~0DhQAe+=dogl@0EGS3{h^sAAG1F;}Hi5~!WX?sbA3((m> z>Jp4b2Uy4lqSsFSL#5(^*kcY@7KW5i@;?Zzr!8no;}@{91_cY!vO9rqE0tr9xFh8i zua-Q+nLvmi?9nl$(ga{!HIF6(co_y6oMv z2+}muvC_IT>dheGO0<7{%%T0T+von?DM;zG+)0YUZXKP{F=ZD!rt`e&wIk_B)<}sj zo{Vy5bX|i7B9zk$S(;lKW{xh2!HQDW5Ggzv5)oI}=U3?5m510AxePw(l6ze!)AW7K z|8bnfPFo)7EUnnw5v8v?T`!GfG)#+K_IO{%pnr|OY=m42z_}x-#VL9k?y?Tk#@}#P zB%d0M_E@%cU))RxB#)JO&zfq8`=ct+c)W#N5kU~pg(vv^_D-!gO2sP{h!`uhK-3&X zk751=|Dou1rydR2G5IONRlK9#80)%hR)_O!tY?Qw*Zk-r@N(5-gCOY*N@!SRT*79) zM}q>7R&$60_*}QItf}l^SD-}M3>Iv*Kfmy0v;69kbamLUpRjvrEqP~^=7V8LTp+x5 zl~U>Vsjuxt@FG?`~r2e z^rB)7$Xq8GA4`U4(`QtejY@c0uLn|Hd-s?d-`sk!rr$E9Rl`pBm2u$~6*s_dR6)8E8g<-xHHGnjoOvmO>1a zBIJ52upk{wWthxEMJhaVk+plAVDeU4iLC4(%R}Z#fP6ojOf)wcUtTNMMfS#XbDdhM zM6l?Ts}j2jbETtzenT5eWRy!9>YEa|EiyKpl-~|v&R16&$UUCB62ElRBfW5=8TDJe zhg&#~3MN`_66h8yLG$=K;KCO`u(te?areKjj58=^`$pT4A51r)K6~0h^W#4SIf2uDG2XEJcvIX)8-|?KOTz z4}|^}GRivEDq-qyhmTCNg?CwW!BZcsD%of z4_%SxT<5zsljsO*5ijSGQ1wVoUVVhe+hxC6sF?+C?^! z`Nv~pYKj_3Q_xbmH|~G-a6xYq`mnW^VZAg&rb)dwubp`#!N0r#Yv@hPT+ys?rVJBJ zNOSkO8b^M7*JuOh-Eb7)dNMwGjuaG=Mq*qPk#EK@`z5}QS^bdeJQfy(OF*X!TO6(P z_^REr&^Zs5kCM|{R+y;W-}6Y!dwh|&GBX4V_!HeGD3`J9@8K;S?&M_TQmzxhQ00VB zts%9GGXme>bUl%&ouKZ&;o}ue!GKS&@aZM2SU8`^hNB8=m*cLw{yFEGuj#cVJ<64cJal(cZbJipDBC9cwq%VAnVsVcBY z`Y1iz@hKdb08z@Qf?^}vQUUNWC=c(Utqxb_D)GNjb4vndTf4WfjgyMFyhru*yP_5L z_SSlGigm$X#hdznQ1-W7mXhmh7_TYC7!IbNE?Tr&=(h8m;d30~ zBktsiJbsb+NrPBT9hPc$ITDw0Hg{(ofWw5IGGqs!WI)|^1-!bp?y70^F5txOZtHcu z^7q=Q?2$iFUwq0=!3CeP0FgnC$e(c=819E9AQ(jZf;tu!ofu%)(=xSjY>ZkUfI96; z{4UvZs12(kzD4`eftd7*d1Hl}A58?ts#?=Ohm4sfIK&i7=TcMn_P<@q1sv17!bW@ik5+P_Z{_ z)TVz0oFg73=;m;2mGb7KDRW#Y5YhJP`)*CJ>x zq6hZC`@oa_TlcCQMB`*%+)Eck1wd*S@F0@iccV~DH6Ysy4IFYmXML%SQL#2$Fe_MH zapV_0g{mb!c(*O*^Q8YL#b-}#F177}8b&eD7?laiLxrQ%MAAX02#rGMm)hR*-No*w z_Iuv@SOY(I^p`Cox&!JLto`_|KP}JIq3@SKc6LdmM`c6op+jQ?#2@kuFY zQ1s?VTJypCV{9&3tWqfWEuAk>@T;t!J%yrv3+Hg>Qr0e6<-CdNczbp8A!7>Y^Y(gh zcYcHOPe|`>i>$;AZN*^1?dmYY*lQ_ z3=*U%Oy4~`{!JU3Y91~U3FTSg9Nw`>U+=-xv+5M96fxOw2|j*sk{yy zlB#DdZ*G?8rgpZxvn~>0AwKN?@Wr&#WIj!)* zgv59X{XXHr!5sQEVZQmZCBVfOZaANxz}Ct|m85BCU-@RWZ%1ocfc&HR#o`J3YIi#uRm5 zLhV*5l{s&f(gz{Vo~X5zggRU4HlL&yPUqq^cP|J1Fbh=MQLzRFeM}!eK?sc>6#@hk z)3tNCZXQL7?$@EUEVo?Ux)zv`6^ZCNh*%DSZjhSK5cb-m=9&DYjvkVvA}(5M*ii{y zO}eaJHF@Nxa;0u$%o*S==i5lts^$nZ!bu`&GHc3u*sj<}wow}COL44MNr^uW$3>tkV~;Vs@@7gL=*soY^p3pp zfRW}cpdT2&bhHF3FK2b_M<{}@6fXVame)ma+4lp?!^qtodc{~1S=jGa?(CIOp;W81bTwv#WmZA@(2wr$(CH8Cd>+qSdW+ROc`cCY%r ztGes0r_OoK&55Qh1TTm0M|Jse-IHiG8K$4^eW46OzwUk|UU>pa(phb?_+QKxeAV%$ zHPs#BK$JtN=Z~IQI%1f*)MIGEu4#cPAAu*3eN6=$`**|IWxiJ$j$^|AZVTN+}j-9p^u|% zR1I|ePs%BmbliuWLxlif5jR(B2A~rCR*Xw)=U@4sj0>^VSaANdVxWAEPYNn^=RQrw zumh;$N_NO}G)9N5dgrkg2)#0Z1* zKZ{Wtmiw_QBsl!jS;P#W)f};m`5m!E!_4H0^&NrK%d37 z2_6C4g1S8sq??I?aPG9e1)Z3a#4Um(ivG)Tt%wD|IqJBi@=Ie_fjJpO$oV`Q(ctYq#mv2CHTp%4S#~3>i2Y1<} z`-$xkfc-_0PqsSls4Awg64rg>F^>*_HMBHhy@>QL>jS)m&xaf3nh}z;CcL$0OI5}`Zc(XB+eKW#?lK2A2k3+Gex8~ zz~&in_!8^&MwL}US`eoR1m@^aNPN5D!fR8pR@9RK|a>)`U&^{~|kj}-<-*#~Vz zKM~Os_a+R&!PSk$=%Fid--8(`?vJ;N2SmX|@^6Exg)?ZYcDnwLVsm3ytl%Ibjz@|C z3Nfx^(T`DS6qRP^BudSz>m;8#arRrdvB6~6^bNP4DWZ*4xJqIIEJIV}>R)2YzYHD7 z(vzLYu%@J_{`P!3eL6@mXWKZwnvPOdxBx-ZB1 ze2j4Wf^<=(rT?}{qDEh&E*C!X6yZ}NEf&*PC3Op~?U1Zi4AXLY>@aT)N8nNpy_YA} zp#O;mw1T!Vxzie;ryXLbh8Gs?d=UyM*CkEG=(Z3`9wBXtAP;$KgYfdhmUY9*6WIeL zK%WvNP2KC!)%_I~E4yOhZWdYc1mETM&>a1i-3Y&`kPg2`3q3=ZWRhZ)K%ga`e83bs zO*^!8M)GL~;`W|zeRoW>4;XHlJxH#Sgj5gmC`1yDs?v7WZr5Y4fCH9N{D_YdhKHtj zw)JbO=*JZ?2*bzvGqL3+X8POR?A#QsTh)dG#u9n2h7!3?sJ!`#bFP;zwcm~hYVqmT zXO8G?E{RUNUVJLX*`#GnW%D}8G~b`^)jJuoyTTxjt&El`44mpd6<(Joe9g`xM{a0^ zM`LJ5DZmC%8|o5aFv3$I8vnS#o+a+aOs=93Ta1j$nw`6cy&INjQt2^$o)r?CYN1y4 zReO08{0}R?w{fDi^9%GZ@=@_?#=ej+ngEDEgtS{$r`yU^YXV4jmtkx6Lo;@=Ank$& zpIGQR>{9dqA+Jcm3W4(X(znvfWf|{p?pSbPP^rp%5~LAoDEMnivYyYssv>6YjBO5UHH9UG$vcW>Ag{?G~Kb)g8B$u#mvX%e2H@XSsz#!Y$BAR0t&W;+;tLKR<1DGZt%2R}lIF zqxYAAH6kT-?c%#shp27pxilW^ta)1B%^FApCLtUF)&)%B5mKK!GzEf}bbJW@&*Up!YhUwluJpdI68f5GDgN)}W_{ z#9+=>(m?59w~zllqhS>^;$nRRcof0}ifQ$kaukw-RFufqQb`UXr;;X@gPrQ~Xm6*! z6q%nk&>PF%(@}D;ane-Z4m{zqfe6oELavJ!1kcmy;p~z!r=}kdSjGBjcrWxSSmw|I z(?E;cZB#(m5;sMs4#6e>`K`$vHZ>yK_t2h02#cS0U;rVwD3g|Z9espRg`CO;Dg`M* zIu9>EI%3J(C+x}d(62T*^z2ndJF-&yjfKt1pQW!GvgDpfh`N#HeS?QP(fM zo^$2AYg|v0J;y#^*)9ON>D#AtN4O8ufYO+jHiE_q+R%->fZ*7k-#4YR;9C>zzE!n8 z!;`EwBm?9JRCfvQ64l8iKBkOTb?yyl=bCKGPD0QA)Wpxp?O)C?NzuD+muBJ(9b9hv zg@9KWV+R0j8(QzXs02`$EUHZ!aWR)P;&=8ELPLD|r}P*wYTSF+gEX&QfNy-525Fus80ZW$_f0}N4kL=F(%_6KR^p7qMB0A@ z6mau>KDm-X+v8G2#8QKolVD;FyMQz4It007J~#c#i;Dpt?^N^*tkJ`;QAfyl6KE42 z_{q?y*~r99)We8IAfu}V)XD_t_eIGhf3l`cD3VnbLKl!6jSwLI`0`Eq5Qe%(qKD!; zPoI84^#lT@nGyoyy#)?;dJRf4cV39PxtbS(c+FD}TqOwf29S~B=Cx>7M5URURubWE z;VGGGVqdIlG#+6KLFk-KW*kXla3_OdiwVYhzl4B@Tm}y=ZvH>pj>Py$m zQ=lrJRNaH~5s{%%qV2mtOUm<`g+UvrstV>7e-zL|8sEh&$_;CFPOk(xN)51C*~RsE zk+Ev>5GzW&{ygseLpss9FgVuMcpOuFI(=hk2y)kBU2pxBs;f^BAI`E)TpLbJ^z#F{ zE5K8mw~j6kh@nA`S){yeP==nngOpa7dx*+KqmHA)L(d2Sf7L)RcbuI2K5;kzyK zj&KtYv9ur91TMk(eaU{HjEp<*=Ek?=f~M=$E8*C1em++e$$D}}vlH{ElUd?j*H-K< zybG@_xMz39h}~W!+T9kAR?_%T&gxSqWdZt(_K9mKVOqeiH zKS3?=$Wq5eo%61|j6}@61Ji?K5^svKXnSNNh=es)LO&{yXM#;>dk(m&fM<4r(Kt$P zQ8hvbSnif%EK2Ej1kFYcqVLZ4VY*fUdx3m=+EY`gK2^&5P4b_x z1z!VZ12BhG4j)+y74GY&X$ARcSXDv1uEv=4L=~%+)Sp`j z>F~@BALlkL-Mj4wPR>EHzF{fsfNyJb>mQC`)`JfsUZptIh8<51H{~J;yL-umPll_~ z2SD9Y*|Dg+2GA(SKp(_yM<&{J|8{Q#VpAU}wZ1KG7xlf}!rThJbDU}=N-wo0BThZ! z0X7pIa3>`zy9Ot5p_4RnNi0pFWC|2o&{_IF`Q{+(ZX ztRwzlpG|1yY>Z(i&Zhf<;rY=rC~QlB_}p&)7^bRY|B}7Lsu(ONM9e9&0s_1lZp5o zw*%gB%p^MaI*t;}+2_r(bf$7OwkO!Mgvww4aL8xgf$BNOysp=k{<(`t!SZJ%4>6z(-H ziGhNioav5`(nzms2MSF z5XhMq$tcfsR9?>8`B#i^Kq7a2?kIVgsA4KqI`$bPl~Ul6qf7YdTx{q+As)YTPpaWC zDAd62VA#mrSOG6t+Ql(hRK@T|Gf>4aX(sxQW|<_$PZX)ey~dfwFL#C(E>ns@PtFX* zx16Z;d%2&xMn(v!P9rh{?UiBONO=cR6nFeDthZnRn1hdc&0VNK=u|}2v;armPhXV$ zM<%AKH(+aBRknCKRh;U{mUX#@%Fm6Z5_M+tE;(2RBQ`c*sCZBPc_+1TI9;0ErdzGi zT4Nx5?~VNq$&g7>*P&f6wy_I|xL!}9zoq8V(V@+DNtNNQz5GeT?goBoL=hwDj6=ba zK`{WK_bIEnDPuOXt0C3>>nQ7PC|RYtM9^aXE#~F)0iB=Iy?&2$N?hJZZ`q|`ieseE z?yU|TKAxM7e@|Pd46VvrIqkmCf}DG)DUrrOPSjX#m7|!j83t5iKR%6rl2~ ztc2Yz$wZz~RQcCn2v}{O)&t-Q4z+y9^5`kT&fsPazmx%3!qpE| ziGr%rsXTrV6g*ZxAfBvHHEj8`byMY~6Q1Agp%hP~UQG?^s1R#-7W?{5AS?=br4hlmGi^KT`MH9<);ZN44Q< zrDNA)z-FMIpY5=nM#ZJ8A4e zq-lbecR(S~?{1N0J+mb0x)k-9tpl?j=6JWQHR-V+qg-F5K1$Zmn!g1dkN`k!aYXj) z*D7i?z|u#u&3ZtEOI5nMA_H8Ft5JA?gkZS^IudL(8wT3yc|g5t`$CB*@iP}tlJkjI z_u6_jCCi;R!u+Ko6?usK$OYN#R_IV(9UA1oj9?c{6Hcf= zyJvs`AynXKlQ6tHQ6>Or)AAOUbre$m0{C_l$lEBHRoK0Vsh_y^x}2*^P1FITEsLXp zk7#G0Kchr&P20*-jrEXD3|@M2H{&e~3x5w6p>ky}TE8gKCN5PO@_K-pfKd`F z6^j%UuRU{WsurCnB->(kEtcukULy8j9@p_oHfPAnAL9&~n<_PU?(&&J9|PVKR34=J zuD6zGp?Y+p5SRcRAzl>9asVDns-<&VEOJng*R_(Zx*jd_&4pgfl(2D~9J7d)R6XDJ!9Id+f&Vewp9$_P6Wm z8%uyy=$xZk1TWlr2efE+O{U|a>adL%e!=4IH8+koAS6$2i4Et>rwClce|P!m+0^gk zf2x4jKMAY_$Va<`U5U2>YCwDF1d<#;9pLqpUbqjOuzd*uU=^$MzfLi;g)e)H?o%jfE$*rkm|l#MkI8k{G=EOO(4z%@zR+?n$_@bM^h zf@Nur@uC4=b+OMqqNieH5!KTW#cU&F8Sn{iQDBt1R&0(?q@RNM!zz?d|EX44!!F5) z6uhb!si%76VU!}*JPKQM!zw|=y?_$?CwxFUNtJnNj8-v*lGF|4T;5LJ_)_L(qGz1j zLccCXBZLiK%1y?Corn^puu@i$HT%2Rdhmw35od0dYKfY$xBB?Ju9R-(&?TVW44+b# z0UHE(M&$q{*-cFF6v?7%k3O7Q9lVr<7iiw2^ddg*{4-qqV$d_D%j-^b*jUVv&jih% z1vm**P^B0HS%>f@s`XKIjKLfy;+5bQc4~`7$Yy3|KzwHH0-xnIy?Z}ZD688m5qTWn z%7>J{miVqL0~g4oUj*W}&_AA9T(ZBW|JIyR@c~g?jqc1Nq!$O9Go}YC<{b4{P^InL zReM8ALSpM|12R{l4SJXw_vWEj` z1kO00xymoPM8N>iu(sZrL61fkdeqCKeJq~V26ec(Mmh-2L2i6uts!1LGaRMkX_eNJ zEZ@@b^&-t&Ie%9K#(&DJuIVuQWR3D6=+~ei*!=E|mkgzYGgz;WD(UUbMJgI)#i8uz zp%X|3)aawz1_7iM_H9%%-WJiH!*1B~{6m))$zGigX_^nQ_(lC;6LnAK6gI^R^zn+4 z+rqJa!Kb*bG85yF9*k})Oo;riy;PVZccH?l=;3GDwjs`o-cA06!2ZmFj#IPDP4ce9 zk1^2niDAvN_T=L@Akub4w>~h%Ql1tvSs#O?(|v8Lp}i+=#pjB(eJseK8o{TvfM_u) zjmHVObpKP2YS6x)VW#1-R?wIe4u(j1!T{+#9Az@<>3#EJDTFtn^rn@cjg*)%x0||( zu22b~4APaP1ettI5Fy6N79F;|;zajQG`F!^&6m+guumI>`*e*b+OG-2bJtr1eibn2f#pt~!w!4HA+n z{(^cHSItgb@AP)cyQgf87c`9o{KyCTD$H3uxUoc2`7wIVe6f@m93;VMrU`1r6k({ov~^?$}c>^5v)Xh^W(D?%s3*#NdPJ})jV>T^fMqX}1!q=Wl|QqGK4CNY96{Eb!~ z8kRlgYreY`3!+S$|YR;J7tLQ00#D3V=M!3jVxA6uwQVhYfEz2fm;*q z-WTu*D>=*}+>t07bhuUAbClx-z8e5=_jP9bOjU8~>Z$?T);1jtEA_9tlecZXxP|+i zCboTQh>^Q9jgtu?gR6Xa!dgbY=iE@=guf1tHn)L(ziy-NEl@-i>moKZG7{7DcZhY0 zqutPW;vcDQyEPcjNVbg@S#L7r%<>GfUGVtBk6nO^J1V#D+AW9%zIuiRys@=;B2@*3 zLHF9rdb$m}JkqvzIkzMh9?5GFhXkgA!-C6IB;QBy@9ci{&Yn*vdKS==)D^GdH6a|g zCwglp$E(zIK~D|f0e?~}>*9|Yn(kTW&$Z?;_QT~Pte$o9Ma3Pb@DA(MSbesZcMT01 z@f{B$k^R*Obj`iH;Mm*ypyoIAZrxf&QhGP#wxkf;X@M91vf{)Q1A+UoGxZ|tg^6#q zKb;UPJ+z@X)@ADr|5<4L=P&NerppBd#8M#xa$p8_+6XF97DG-KO_al)>ys7*yT+-t zytw-+I&n;t%)HDM=3M(cx(5nhxW|S))TWSo=%U#YC&oLGRmyz?jpS1JmZbOynri@F z^7A_UM<j`8U14=D;2J8B+6yYl}8JG(mPU zDjJAz5z=#0z(|T*=-?g& z`jlnjI2o?G3dFcm=$;5VUcm~ZlAD9icgK=2?YoeMm`7XtGqK|?1FR15Yn{hQj7S-5 ze439CgQ$o8NXLClnWvj3g+tVuY62qXQ0qF%MDe2|V%7k$io~v!t>usufo%35ZwvQ^ znHO{4E+3>4VNQ?$)O=nJ7?qxpn_a&XIG?zxo&tB}zJx?prz!FQ>m0(SL4+XQ`e@%_ zPS?MeJ5SdLtSF$??6^sV1X#TP;Q=Shpd=!>m=NV!H1jj%qhUb zQ)ENR5FykQUIM?hu(L1xz*%g3j`7mM`OaullWqs7aq4!G%=5vJ|3#{36sNlP7;ix8 zRO?4NsWy)E0T$PnT>_gd^m_ANg{o2A1h|5Gxi|L!=>|95iuzYZ1>$`6*ts^mU*7Nr zM$3I}d{f1pYvN9bHPT?a@AFYZEvB=Lt{Ouyh#pSlOtP;YZ~O0|B*Ia)4f8l_zA``1 z_p(CS3o5=e;uHZGf&=NdClIaA=re*_e`40}jExRAAdKvgXb@!?A7HTeD2VtL6;3^h z+BGfMn|9`F+Y4_hKz3FoKE?3KB}<>v1bQB_IzmpM5RgP;??)J0bUbsFyZsXO9=(x6 z?22?DI6bO3A+WkS#Z!|SH-aJci6_l2J zGmo_=QH^OxwEXdO?%xpz|GqAT#_9Qu4VUbv4KC=_jznadPihUl=q7J2mmxVmEU37; zQ-9n7-tvFd!$x%EI6HE!rf&FG>T{(Pk$h3)6cf;?VnezP8$qQ( z8Q0)VC|+%K7uq=QDyCajn#FhEuOgXTsH`f%8gu?8(KD9=faypBJo@h-M$m*k3VbwV zi04mwEMZzgOTK#c+iciE$nfhkQT%vrz1Q{cwz`9y^L|l}-eYeLNx@7*V*pwTU%G0F z73x^G&JIVwYfC;{1=ZgA{X~QyuB#0NM`7&mUk(m1qRqULrmB`DColzp{L8a>gb7D3 zO$~TBC|!U9X;v=%7G7u&-V;)=!jI#HE^EOaN$V|97@qctwN_?t7w5l$9{?RJ{|W0c z|8K0v%E9*kC>Yjg$vWYR{|t`S@?8d$MGOYdU%b^S4Ua<5{RHWh63JuvnpTFzP|Xwg z1|ef?B~rgH369$v{_ROE$)A8|MvZOT%MoMu8B29&T$-5CyZTy^1w4HC{!BT8WGa=@@5l2Ikub4%gu|e z{R##6GKx)-fE_bkS$dN7%HcO-KiG$~^e$D5Ar@j&>klKPv=e+jo*iyE_ued2ZfNSv zJGa(rfqU?-h-&^Z?>MZs8-Rq%{U-Y2#j&axS%0%-J;DTUu+FmWcnj>4JvG2|@NBa* z-Ke87L2CT4d=H~Z+5JI#_$Oar%O8aJ=VYC2>^8rtpdyaJ6SF`{!>lubo$&uf~4 zl`j%zp{}V5n1E2@x=E5@3D+;+Uv@{!MgioS$G0Y0OUhb)@2c!+4?z{)RM9Ps99Jfr zs%?*+%gwqh7KCx51;drbNOo$Z(ylO4wBl;>-D(@fVK8~RZFG9&%EzV&zDaz!w8w5u zZoDZAEHP*;CyT+Or^P{ZK=R9BRl49$=xOH%_?6qii;%X>V zw7uE%AhO;nzho}^tseFm ziVQ4BOEhZv5qLhy4b8LlbF7a+y^?99_2bo4r`#u8-6G;w--N$IC<@GlkW}Uj2JasU@6Dlx6*K7S??r_4*TkFBaH@N zBI2Bl%8QZPL=TA%X^tk&)fT+~8Auv29gF7V-d{s2l=Oo%SJ2sF zo3=yyii^p(t;d1l=(=TTCvWXTdJ+KrZ59a%c(feQSDS8!mk;;=PlYW1Tk-DY;PFM) z<|$m?a4jBoQ`5pwJh_y7!OOmYhLX&edQv9TdVwe<{iwcXy5=g+ZhRDKOrNSzf}p4a z^39zp1DGgnAREh3W4S_+cFtAW?(cnX#`+e=A<3a9!C-4BwJSj@JX_L;nNog+n$N*5`hjk3UXBSEl9Tx@GMztK zN$X6m#zeHbRNZMjB+bsBH3?z%XHfAc;;g3;g>n7|!voTRtNK(iPM5hdy_CJhKsGTa;$h9&kUmxtNk_z)#aJ+GLn;yc zWbaS9QCnjm5Sui^0g_D@Jkn3(U$TK^7jC3$ zerqv@vdS(39G_L;_9!gtX#D!jWI?gHK=;xI}0^}*qtldHnP2PR-vgWh&jzN+(@bcAEtd*98|oW9*^HBL-klRsL@kH=@6f0Im8u(b9@8q!s0>7;Jl$+H&_5RANU`>89N7l91c0~CAuWUUybnwHBmTE@7+IV#TONLO~AOHfzO` zF|%qphV#+a>S&>ccP>bx$eFE`tKjM#d%_y3)|gAyHquMK%rLiaqJF^R)P*<7ZUSBbkHFsub}@%UUC32K1-bx6Q33vxn>8syfc%QUd9zO~+3e za*@rYC{NLeJ3!JeqJO+#QGANIks9&rlZczgmTU^}bJe*m%YAe5V+jTd@Z{w(N|tfa zZgZcA9X!4F7iBu?WG(e?@OywU5K{sQ@t#r0*Vy}=H+j`?n+&#UFe1h5jR0&yDF?&dSMn>M17VIe*1Q{W$%CDnzUL97)N z5kR|YZA%BM6&k6l3yw6C;Oqp!y5nykjdpzJnYSq1@kTzwxL<0Zx_Jxh)K%jw{oKmG zVODehwd0UtmoIuznT>%M#J=tZ;v9suB082Z(?5hU)tz14J3cl{aFn8D zlpXdWXd7~9n?4W}k|mrGS#VjAfSCG|B5Avx*l%scbXKnM?3JD5$=Lar^G~t3SaX49 z4%KsEFYef5hA_K(dAmFTz|temvG%zk3`e}%u|QX*vV7CMaq+N|mjND|MESdcKi!{Bb@p zbKaQ>2O-#RCIeoWIv)ug;;s!_z#<=6?6k4P_)Oko9pao0>IYsrnu7CRe3!f_l+!HO zS2|LoZ2h2>R;hh+&*jNW6G`Z1$%uqeT1C2xV?`GFSGD+sSbY8%cV3hup_bkYPz&^=DAhuAys^+dl0^iAPI|M<#T^{3bW25ry3T+8GCOo|Nz9R$h>TkVC{0_BxqV zJul;_C*FI)X*_cDz5Kuy`hufFBBx;<9}gl26~~?H*#aB0W?|ye#jcv9Iy1;k3=#89 zHCJUhr5YuqKMLTX_;pedTYUKFC&aQx=)xO{>}0P~yq}?qZp)9EUwf7D=I55+KEiM? zfgRKE*&uTF4=Sz}tq2S1vhMsP?5at{pSL0rS}pYPHs0Iy^XWxvK}oAHgeiX$3-Ub3*4`XNslk*94KZs zUQ?e86KMP=xLqzQ1Q4`@1T`kA+f$XVZ0p5i8vQEbae1r(*~t=d`lK+lqmIf{YE+MM4n@G5%mFsf%o-pw32mW$5?prQ!*qDonq!K0L_k>rCj9i3;?Np{H^a zdtUA$@4YdXq)tQ9KC4pnL*qz0rAf8C8{yQ zQ9nENVxKuluTyuJsA1&Qxvz9$n~Q>cX&r30`y&;UK3u#n1pdtH342@k5TpxIN!8Ts zf>r(8IY}6VPwXsXLMBLeZ+Bmx1Rw>^;sjKFCzoW(y_Wf6FF_>Jq(dXEWASVt>vLx6 z%aFuPf~b|ICsI{das_hwzUoZNEyR+d)~w3&coJjMBD&Xqi=M8^vS}7jz z{-p$Y(_g(hv=#9)L-}RB73;1BRTGAF{O(3#e?>+y?MqZxRrFXY90vE{a||4ypzQ-^ zFKsHVaP-de)&`(%t4}e!o1(J-)|eU|JQbn4Rdx@1Xb-m~HV`x3FI#AH*C_<;3U_KL zI^pEB(j;3UC&?Orn>PH>KW6J;-tS*N-uk|}zC4_?l1jIFo`vV|yQn<27F~prpDWlC z_bAHP_Rd_nD9t&0%TV&`zuVq3X1}^A0UtZE{MdAj&$8T)rY+P`bV(;xMSWUqdMkMNPKtwO7Jz0POf9PhCOv!{3fjx<`dZf`WmO_v(c`))2%Q`RiDm%im}mD z&x@Ua{q1XeQEKfIgIMyKf95NtJ$k95x#s4I5=-BGFUt#P@vcLd^+g_eAXa?6+WG=t zJ^Wz^pDy)Xc3qd9T~D_Y6V7TXNgfS4kAUN&-wR|H=t+q8ezUPrHqEwXJR&TgZGLM= zD4FbFWT4S&0F)gYL*Y5d!m*ps(Re_lfTr$nNj zO}QKFK1@7>Uw5kco$E&uZIcJF1jw26Ah`T^;nnLpfh?PLR%3OTI`^gJc~s)_K27k4 zD$h$wjA0Irf`g&8wm?s>j@YRzIf*aQVRO`Xk6x=YC=>VO@zaamSg}ifsIQMiM&HH~ zpi$dS4Z21Grw@NJb&EGyH!t5h{jK>dI2XLIyq0CMEM+M9(X2p&&mJc#XD~_Dw_TY} z5vzmc+=AGE-@5eyLd z_|t{Q5G1M~A4}iV2XocKyJ2&D81mQsH$xlb{O;S%GpM8;;-}x_qDA|V~7Bnr=oTMIYR*oI)1 zVwAgKO%!X}3sGc-V-%}oN_*}P*?R2R-58Xt@-}t|8TuEpOp_4i)DupUq4nm7)B98Q zyA}xl&Yf;q*(2%8nYJhtS+bd5p2p|RqJroMFIq39s@Nn+jEJ9K+y#jKnd_9RL-6ID zW9;8Gi{F_9)wm4|F}fs?Lu`X&Z#*TL|IE7Exl{a~hA8__Lv)@*QaSk^@)lrW^39#4 zjm4TM*S%F^>IAL5v=-qF(O0b4Fm8HJ#(YCG;+1MnGM1O+xI6M|)BFsrw3+5)Pqo)7 z@+U=n>69)kJts^uQ%MhbssW8S8s4aJtjdwqg@#EyV#417Zn1=hprt5kYSkIu)<^+wpmh9%os!VM?Tna)4cda}}`a!?enfc8u zZIrDZ)lGNbh5Y~tJqD6sgf?(SGOJb(@=c+nluZ~iYe7pU5vR(iJ~2p>)qgQD2$bsK zsp9;h8_ACFcTZuyeiH24QkFQ8z0^)D7~@g?W^C!~>ogy{g+fuP{~d zxT2jXrGyzhXpk8REYjUJ5v+lz(Gs0N$$vtY-<+~86)Z}EoN+K%>31QSL5ADvc)t+5Kbf5c1G3Z@H2CE2 z4&07-gM2ToS#?qMCDl4=SPH_DCZZR+$|;|`Kh@qcx}Mdw?#SB9Rp*vTO;U5!>W|(T zDU-S#R=4B`B_c_FG=gximP(`T8Y$7gFS76d1d?Zht&?WQ0b5XBN+H!PcsPe>W((Nt zaK`#|(c3#p?@w%6tnox7YR5WAt3_6=M(W#7a*q2E!zd{lHdq45xK%Z~vvhS-HsZNx zLs70I!sgHO;8a-aZESBSFs~kK;<0D~9*tGnRG6P??s#HuWDcR5S^E-a zl8|E9-~8OVWtiB8pOt85K5VyFR@C&Jc-Dpu%P@X2J+PcNKgUyl4-*f$g6L%Fu#%ep zE7&7C>6e4!fk2?Ig=m73)Wu3yD{N18JOq-!gJOb8l0?wB87XQ!`(0$)Kb*~Ob}as2 z50XO?UW;sYKwOmr!%`WCo(*GpK9kg?jY42qL^B5<{Jsa8V%sE;TZs7Q5h`fZ&-*l` zkQGc}elO1=s|B&G5%R3$%$5j|y&BnBOve*C9`g-Pq>=ayKkE-Ih2CprS^r?dc zQ~>}9ea?GD`BdhaQc)kWZ2dVole9?LMbTHF=V%PwmW`2-JIjY-g!g>bv!dgX(%QNz z4XmWI1Ks(P`gas|b~NBM$hbi;D^yg_gi%v)qP|FgB8vg$3tJ$kLD+zsgAXP%4B#p+ zs0J)b*%Gk$s_oz}o2QLTN{6L3LAVlNVF=#pL{wkiPkw<~4F=L!aH7Fy=y_4su*hdO ztw!_Y=z6EVKkUu&BMuO?Eqen7q376QIQ0^9jtx1(rOwdx!^9#}N^pofQjL+1K#g_* z7nTqHjCSz3Cpf2cmO%%wib>;i@m$fF)SipUXhD7RRuJi#$>}-U=xouV>n8XKBDR27 z7<(;Ku*GA*%;_MPN1g#drj-?Uu7~bv%*@&A-=&Q*B<+}-btw_h2d9!J-prh2ebYGpwlcGFKZ611Q!c)bz!`E zTrF(F`aA!VtfcoN`?Y)P4jaX7!Jc^mgd>wh z^4bEc9dtj&eT8nGHX^)#*YW0Y=W94dDSXLHxAOAUJdF4B1d~aOePb*Ez;je{!+8D;%S^@s9hh z8-8+z8bRhZywV9fQ(24ybp*I75eC^!C7FINnzebLle$zF=pn^m$ob8XB@R8oOkTC# za=R`gZSR8iOX?ks<9fvmh-|q;k$fdn5z$y%yU9RMopD+v2hD9~Vi0Qj3+K3crH!R> z#uwHD=!W<^js&8*qzG@2*LvI$oErjwj88dGAGU1pzii-4zf)~3XXLGCLo!Q7J_SiBX^4xZC8ZcmLpFNu z+rjc~*Bsz~Z`*%nE9QE=33AkZ95*jBEod!QZS2?fil^2vvSsS-IzoX-te$cPM{0u^ zcMHR_e=^CcgpA54z?4)U=;i8Z9M~&_FYprACP^%j6)N#{R}>vVpk@bC;{_5s5bcHs zj!d0;5vKb2^{2?RN8U^y1kNgHZj;p=CWfHWNRJ4?xI7Znm8$pR0$K{Xcn;VvJ`CMB zGF8teXQ-s+fondIkT%2YpP*FCB}8-2yEUFZexttdIOZZ#e<`8=8=Gg=`yB9v zKSC9_imZ(L4b;Kt{970zU(&H`y%20H5~WB>c#~EV7h$+*LakjI3L|{EuEA{FxLNuw z{q77Q%4d84m4;1bMD**)rctVzz|GC$uj^|ozURyk2eneLLpr>;HBBRIZg^85(^#mp zg&oEQxhOQb9)6DQoBI(*3nY=fA_OtEHjRajsY&Z}Wr1nRBO$T*=95Ej`P=hOHDg;? zaCR#Bp^oL-IDn9nXuZB!QbweF1rQOwmpSa%$0+8)EGJUho-6jn<~qW3cp>I}W~hbe zKKw%Fs;yN(7C}Z4@qwBqSPE3{_H+~aYevo~w)uj0zknO|8ouqT_3FPEJE!JQm;l?x zwr$(CZQHi(oY=PQNz4yb4{jwqGu(SMmpTl*q6Q~IG z0U1V2a@&&%XSuW0WuHWKiWL#mc5JnsNpw$-XbM=?8adj)UHmSH-_jyJrQY!{6}@j| z+!e5gRJjKwWHGFgalLT0C2~?wIlmU@Wo9>OIJ=Ww5Ko3;w(yRXZ-kRbth#JZt#%Ne zBLuaclCba#fFKxr426Yb;PWIUkr{tnW`h62+4_+w ze>2KxK}wL%KaU&_pRWhI=F(L8Z%-N7du|*b9`L>9zIyUZykKy&C%`8h&RR_#8Z^*)t-Y0`977d{o!eD4 za{XUeJB~I-n?SeQA~^#AS>+>mx*2$_dazO?Jjh)kgME|62Wh>m3>0} zj#dX8_>&+sh8Gjx)BL1M^SUT$fLzOU1L@fZdJ6uH)GOTS=_?O5r0$ z8%tZ-*u;?HawAtYkF^ip(z2o>8uyoWU*E8_MWOL+uAY>2Z^CH~c#)_(X1hz{1c}Bv784rEilL0maz6(*;-g3toz)Oj z^9bSX%Q!JXk_gXx#V2^2uQD_vmxk#vq8FH%_55W{bLftP;{(InSq$NLsI*4n7#JbZ zO2R-2b}JvkB74p(&J z(|oRIXXxT~4`djxvEw;cAJ^V4sjX^rmd>}F**Zi)x|2(|HCNUKPwqE07UZ^&2AsXI zOTw@ejI)b3HL5!Pq1Si==2?!quqQe>;XB8jtiz4Dv?M5&pXhP(VX>GUZcCzHSVrV; zd7PM)pOi`Cq_j+wF{wed(ALsC$CJxD}#K_TII%A}W^l6oYm8JhU`BRVrt1Aj;sEj7k z813RKA$N^4_lyh!MtaYo*43G=^k4*=b^(MBm9f*y%ggN;2KGCg%Zs>ZmEmu}f6h!2 z7B%&}#&<*~YW2|d3v1uko-12onaS2-_eAE=#g!ivP^fxzQ{Q@c+R>HR7Z&zJ(-e(| za4SY}taEQhzl~j$(<|j=AGS9X=n&&1W71^q&x7Sa5HykLFRi0{?i(3TKxV4<#HR}2z5;G#^;`sj z?;NRTljzUh3}Y2{vh%uaTzzAU2TXr;d-4>xsY@3{Uh<6*1vQcO|31uez|$whO3}Gw zdPMc*bG>Esm}MY8 z_TFb-Eg%Y|6ucxCTze*uKE&?aBmM{;6(8%8Lw5=|wrn1$&Y6mDv*lwn8+ZytV%#!krLqHF2Tjp#cPOd_Q5T28X3_T3nqn zSiRSyT+Zw9zo^+grGgp>5y$|Eb{y?qoXCpUfbn4$R%_AMY0oPQ$)e+v-_L~BpWNck z`C`8uM#_o)Znf*TcpqAGfJK6@>{TO%gOq|le=gUQP?-}B=G=M?{5^LVM{0ibm*fLT ziunFueU>aX3jOu9@@5M&OZu-nYP=k*W^4|TY_{4Hj69LH)+R>2u%Fz{N*~B zdaI9cQU0jXru$l#7|5B(?-p4eAsx}K5lAM=3)ItQD zy~wx8EBmue?Zd_GPY1He>V%@}G}1IG2fj16jK*Un<*Abwr8TRpw!&aYwH~S0-anJo z;rtUBP;Uj@ZwyUv-w#5`w#%-v{7e>#wRMXMB}0!KF&^s*tz6RNWX~4tQfQinQ7$HR ztBRmD3WvC#G7~&cTz5`vv&<>d$U~UlQw=Dgv)>VTGP%^3Wqe0W$I&wQUM$bJS?4<7 zocB48>symkr|qb9XEh-L(K#PbZo*nAr5xI`EO2}p#9U(o`|brp!)3PWH9wZp_9QzR zIN+{>ltP8al^2z6o6whgXM~Wq3s*!%yi=^PqsKQMbdNM^2)*7tBS9%9s2-XvQGZtC z&8df*N^9&hd)6Z>i+1A+hep#dXBy-Gc&n`zzi%e_k@|HZQA<+(@tq;i;|pPx4TF?W zk>GY<{Hr1vN##bic5+&kkVXuc=1 z_~YumX(Lh79c1vA!`szUP@bhVFy++l!2?E4A5N&?9&802N&{v+7DR|;a=S7T))Oxz zfvt=@%1)2vw#)<;r+MMIsV9AYBzU-HzBo>KuawqANqO-)>0-O34AdQRN=SOO8pos z2rbVHlro#GA>-xnZH@?N+OSv+(D|ePfI!(ee$Wh%cqv>0^F`IXXRlJQG673W zRWp+LY#$aBL|1qp)+t~ums(&V+0=QIa+?(JI1{EiV^Hx^R~{%3fCMQdpozEPHgwB` z$w=nUQ|_;Alg)-n>_|LlQ)+SECLF{3;Q|p+vid>GmfJJYGtsB+4uDd}=JJaRn>$I9 zKpT1ty+-0)0a;_Rswpk9E55#yqzx}iB*{N*qb>Q2oK<~=$=$yRW+NR}E zE1|ij=usXMO*YS^305pJScTy?0ZnV-3DbejjX4rzY*C{>mdc!dnZq{6}zqt^gh!+)0?(C#d+DQ|?K$ie13T3p8d9xOS3}bPk5< zdUIXYdseu;UerH}3)A|Z%slj$FbWyTu8Fv?fz7Ls3qG}o|aqCFO?uy$(?en-G~ehcK3 zDVoe~|009)4bqtn9a=Alf@9)J=1*gANxTS`ZLeVE%{>tiQW{$?ivnlzm2R4K@1aIc zg0*J1!DHtLL4&P+M2%DC95Amg7=)YRpctbA6Ac#LTRZr=lHC0tF@V?r=x!TkA?PT& zy6n#t(PK{>6S$5kz?ryXxL%y(h|HIZPe}@TQ07ZNzO^aI6)^&A_%)w_0uD?934nlb z;Zeb>ii`r1*{C*n7Uqe-a9+0o$G~uFAQYAGuS0ZbK!j+{o}1NDd}5a%MwwTPZwg5fFpZ0V`?8u?iu0kgj=44$->nJBJbl_k{S>XV1IS1q2M3lRD73K3OH; z7VDSKZK(8g=2y$Qgl^Uh1Z~rW;gSBS5KsuxpWJpyFwiC&o|9yLaf2bCoMTtNp@Sj+ zjH|qt^cn0Iy>a7(7GfkN;2z?RO)5xW{EjXU<`@{Pr~c?{DvGvf3ArU=JXfgrH$;MM zaEXR>hX9mK8>skf!W7TBys4CzjHZtn#1VlJ54>Lk9OHY_I1I2H-rHBXA*9oMjHvdm zz)tPn(VvljbpZGH~fkkxj;B<=8CX|{e+<9`K)Z*bBbEWXq91#T1mPVEBg;h!Z|*d0{9SN zawPM;Ln5?cMz)u)?(GQBc{Y%t-PIfk=w*y{wIyU{Io4A;33d&e6;uGZp9TZJ=qMg;;5de=EJy;2Z;1efc+Fmc)3E~SG z8Zm;=PFs>4_@UkpuT;*(OV9KV1K_TteEBI>B-GxEB)}=gDeaNP_NfT(;C&m6OlRb#O=wGt5&I4 zKFn(Z)qK#^;$FV>*$dm;*2n_YEOEos#mpmf-U#^%s`xcUY+5ZZrX|Jl=5XvPnBnu}**@RnhZM{}iM zq{I1dX3wiKF|@xym*sYx^i-!o!UVn8*~*Rq;`}O5^}h6PJg5&5x9IsdF{yAa(+moo zGQ;b3bOe+-=-iA25FoO3ig9jxDAutjQ7mUX(k1r11SV1iNRY;U&u}6`gg6PemZZI; zfq>gU?A2JTQcKQ4{4pQk(sJ>a93|TycKr=2=_%eJ9h`(u{goxWBv*U)=hK4r0KR7S zJku{p-Fhqde<~_lvRfFa zwBHHmyxX-0x^)zVQuwTgpaDUzJ>iMo6JGOU!+Orw95elr)N{~iPOzA{W8JIi^Tlaq0jPg z16{kKSXA&8FcC{ZljG?7_?^44b32aqr9Js&F;()Ne)e=p-T{ z+iY@UM1RibRhAsSW~MKKN}rXVMp*I&Ewq*=vKnO}5m}MPh>fT@PF&rX^{8LjRn6lY z2yWt4;&k)Q9HHXGilRqW?W*D$+UUYp5x?*Etd(yhwOl~)y4)@(wZQfr$j!=mdiUYP zxf=)dK%b3}y-lr_BTs7zN(HsH@D{rTu}*54zm8{9q{)}eDiy$3NPCfU)%}#QlXL~SPqUTVbR>aM)V+O#N+KU*I2+R%O&&x|l*dcI_VF?q6XO~VRj zBo~8L?Mb@h+H_qbXtI;d#sOT_^lyBcsNdnUUGevCvc8QYfQ;A@CM=$|;d_XkZ7`%u zx1dI=7(*gK#U%c9uMT3&Vc|j2pkM;+AW^hOuK+lr0M;n6b9kheo`c(2WC<&|+lsDH z_5|WLj8Z=TB6f&K;tLf;>=Sehu~~Zj=?R++wH*+MEvN*r@*F*tdVMRa*3z$Q)QpS# za-0aKc&3GG35X@$!S`)(ukSbMQ9!0!kzZ2+@mSte3c2s5bacjNbRs^pv6t(1ZnKvA z$w#AmO}6$4Hmz$sQjD0s-hAoUY$e@}apj@Z2$MH|+!LPKyi>T+DF&HQ@K?=P-Jhw+ zStcX37bMPvUxwOWVH<$(Yzo1`0TELl@Qk1T#{78us}^*U#xl+N%ilaYiB6LY-a7Nq6&>M6tb#Qca{IPGKtR`thJ-o zI1teSiVUdohg9|w`o6Hos_I+47J2wg-8a3*YvJZcY#xeM#Zh-#dHl0i&w_>L271m@NLr>dMB*@bw8k@?w4?P zO*pXTbi_0D%elhD81RcCY3u7qR~Z6hY4J#qU3(39W8X0} z7dL<@6U1NzTG)?1s_4mpN=ajSUo{3d><|CW$GrF4g6~O$Iv&w!_nVU_yA>6j&7p3k z2We0_@CdEER~5VOxFy!yo_r9N#n&Uev7TvhTz`MhUT))RJ>Jaw1kv`ow(jyJl77Xc zDO4t+sgY?Il!FH%ppkLuBgOZSBPD7b&2SasEU^esJJ4>7!adUM&y4W;>_G;6R~=HZ z6+fNvJuC$Q2;1w^bC%zSz^mQz9L=zO6{&xgWwZe<)mmKyyQwr!XLzUv6t5AzxU!y$ zX(vA06sD(K*NZd|XBTzh1LtcK6O}#INF2gj)?a&u2Qz}7D8o&kP!sx4>$`SKfZ5}p z6LMgW{v^R(YO1st2t5(h8BC_$h}*+&{-5dSs=CKE%*D-v$XC#Iq&@Q<-g8(dasap6Z`wWJZx9xk@$!3Ol$r31KDLr&aQ>eOE zG&dpP_geBYMT((Nm;Q~NJ(&6|S8uGZdqnXuq(m8~Nv=94HEOXcqsI+Yo-^^rBdsB> zbKtEZ#?CZBqr2ag0RNUuWckk+)PCaj%u!J<^dCIbH^;~u;OB;Qt zPME8%8iewo3>#lLJht%k$#e-n^(ZnC0X$0W8g%B(Q^s6ZdV@4@^!T7F{dJW`EAdW| z1Dx{kxec$mMD4Q(xuJ>)Jfv%{3W!lhKD6=lKSV57#8}vqHRq)%-P-W2j1`;uud?WK zm7jJj_-03*?g(!Bpr^JJhL7TMX1-=)<~MYP9(RrOnjLTz6f^HoC2+To8y3bzCZq1t z@>G37fqSY*w(c*NUnZ4qWgXi3H>m*dw1a0KC~E6zz0Mp?@93Yf1<8^n?empGzeHRh z<;S?5ZHHq_0r3SCcm^fBVqms+m`>GuVWo38S7=YdV_bMJbv0`qwRC5ib~k^Em#t zn@UMV`y!(c*MPB4gCuIWQvKF<_f=2z_n2k&$WrdN#NWIs>C46OAq{!lU4D_jY59C{ zGKu{5FBhz)!$gjrA4i1!*dmb$6@}}C?%3bvva~AX`-AN$d;M&Ax?f!S_jztDGtqsX zozCpM@Un(LwjwmTrj=8!mN&&EK zNrU(CrXNT$hK}E&x7KkLYuX)E_9ROGa)jy9o6HxqRWkSbD2`ssOVZX!CryY9{vF+a zSPgHrX)=jF$HK+o0}{>0F9^|mo?^v}F1uFF#!gu6Y=7qA=v$FnS_3u7{9#w?30AGx=L31U^GPXhU6lyE)B3tZYc(b>8*}Q{A z6GBPn%T=v;`A;c1>dx|kYkO{fm2I<34HT)~H~M;IKHT4OK6!ne?`Qt1*?*0z=*M1i zHH^7B5tEl@s&)sP44odk-}0}Xnx>YFC3vzbpjugm=d-IG(?(l>{FBGmrKNt=N1kl> z3%5Fvt;jWF;z1OfPO7e|o%=Us6B$z;{#+?%!++LjyDeSz=9g_bt3C>%ZEI!CU~O{# zB=+~^@}UhuW2p+Rz#JTO)UfO7KF|&ydUb5oDh2N)QnodF=;<}PHq)H{nF@kG_{5Bk zYTXlGm6b$;Nu1W39tUkeur%?rDc&*_d-pT3-;5w_;wH_EApO*&Hbd4R3lTDEs44v| zne^<|>kB7SpI_!5+QkU+cxh;V`=Qmf))NKYm84}%5bw-ea7pUVs7)$>z3D3}D)2=d zeX+>`16O8ce@X)OWXEC4K0!Mu)buac@>A)HUG_pNBZl>+sbTQaESnn?GU%JMRcF6U z3{tfoSMIP-W8bVT{`5_jQjGH}>R0WiGu-XadC6lx_{^%+Th2R-3d0Ek+tamob928? zh=_AUj6uejt(2;b0}q{iw8;HAPQXvR#{7CgV>c;P=!EbXW(%HnC(Cl3FDJz4fJI-$ zNwLqY-IycA#;hM_P_Q6g`PR*f2BwC0-)VZoe@FM*fQy&zKd;&C0m9+qw2vj>VB zdv5kOff9){auC?#>>}uA6w0V|e4a7n`tumB zbYpxNOjEFD?ZB%3MpH(PVS`i3!o^(M$8euWwZIXq7D z33MHt@$V>FZ1i#yC}UOkAX;uV2T8b#1>#pMt%GcoKfg9gMzYlqS~TKHIGjNV%jQd= zCQ)wxU8_r6F1r37u0leS76*3^puj-iA&sXm8z$G# zSit)3c1^XaNHBqIi+M)1VfEok5eg&-+;GlN(S=a_;cP=e=W9h?@^SW6-n_cOp5Kb{ub|39G@e|B;pXnIf!=mKF7EbBo@3j< zl&JX@nEpt*boX}ZhI@e>#F#xXOGX78MkGVzvBZQwWcTA{YW*4r#Z4kWd!NUtVGW!T z%%S5@_DQL;aP>tK%%(DD*LOe4kXGqvnQG`A)kVu^88({q>1#`da)VdZSYhb0++QS1 zD!(j%jhEs|hzS%ipJaUI84vmUVJQ<(kYRs}=J>?)jnGcwV6iAN4#>l{utbqPgih%6 z40WsDo~?QPoVFQ&f&sesT4$g7G#SE@V+1QT zRlT_}*9sRs1Ot4d0b?1T!kD`RZ3n6Vnm-_(j-=~;n={DA{Vn|~QjO(92Z)ZMpH~Ff z7__cW*u|Og!DD{Pm05(DL3ujLnW+M%Sji#o%vq);0cWfirxTu1H%Y9QYu>qQ|JG_w z=CNU;c8q1_yyGN>nA8y5o-X2Jao9tJ?woCyvo?4 z7TuJutLBC+OI;9;qvkZ*;tiRo_~*)%=2a=37g+;gHcwZ_Cr@vT^>rXA1?2K_Nwb=Y zpz_O57MLB%OMdYvo3cdY-}_!1B0~2JA(Vc{9(IMqRb>9>{M2Wm3=+N0xKYab2`zU7 zhn85fyMnLeX780BCPHZO=Yhjh7i%)ta_386-biIZ(?;GxBYKP<=ZH9!rM~ZgIiT5h zDs6;M@XVY29Fq;;&8awK#s8muJVu?M8rzc*FM0f=>IslB^*t5>^5>j^P`L|+*yTa& z+9@bOE0XruvH2(lZI?Crq62_}xroM%VwMqx@>@5ROZY0E*E|lYd~vAm68FflN|$tk z6Ryu2G0qAXuN=z6G}AV^rNhGG0B&M7t}b`W0LN@` zf51!eRjo2G>aP9j7msP+q92Pp!Ka61G#6QgT*XV&?e3nGx0^2|{KZcE`@kPWFl~E| zKYVb_V=5z!I5}gMHLmwmoHu~PXd&u?c#KBXnhmW?9#6tDtkgY2U1JF~Zthg7YUm9p z@l>ib>anVoFdJH*!&fe|3_Q>rTmUx-jh?9!&H}xiLeong{qCgP5DB}%386v@`=qG(LTVWAra|@844_k$^bC?AWhY!3Hwpfs^S0;$ z#ob4tvu}CFs|-gh7X)1eR;@D=2T0+4rrbS>`f9Ef-24)@p~Fb}2= z5HnhLOMQbszkP#ILtS4vy~r;$UN59Qydu4+jfMHcWDE9PEHb@psOgirZbgDrene(d zZP0&g!T7|=e%R-))=i+6;*R`^=SK5@DYn$li$Y{TS?RoaM8-P;dYbQ;WPLCu`>tN+ zQ)f)%b|pJrh=YI)A)*cu1w*gvVjxTa;RFoEB3u%Of|0rg8=iaB#zQNw!TZEGp z8^@|-va(nF3rK(J*IImkPK|-kqSiQXWqEqr`a+9Dw4(gihpu))Kb;*T>Q}U4pGppM zv?o6nj-r25I5F5_(^@Yp{f&mIY*!nk4oHwHUu=1mIy*(JMq3wOD7`*O)&8wr=&MGt zDKVW^YRtB}F1LKQGOVj08n~{iEwM<|v4nmnBV6_#aYeY;1g=hdskSr>95BvY_BN2I zbsjJ2%!A$(s`|hJFdo`_eRauMK!PgPOwv-r$U1kq&$0)YfiJ&YWd{ZO1|NY1jIC(| ziMmZKC`GhX{W3&xXEz)-NdVx8;b-mSH@G=Ch)SZoJGRuO9Q`B3u=2Siw*DUGNbSav zjS9Wphm@o}N?WS2C*cl9UNM*<#xwh91sRjPb~HGTcb_fnutxLI*8OVQwOvo95qlhv z%I#joZXw~4i_Ehh58u;cUBOO&b>8vq9pT!SiG0IGWVEw=Go!Qp?TwjZj}1*t_1#ux z%qD*IR3RO%qap!o4DX{d^e&^I&LZ-3h%Jdf-V0QZOI9-;1 zkTT}_z0<&XWljRjnMsz$0@5ZU;hacP2MgH|e<`bW5H?TW83A|i9+__fBvi4*Cq8P3%4DNZPaU;^t5?ER znR-W6QRnHJr|mfFCx?KJ{IBcZo$WV6r4WuLtJ5{dMKXdql^}RYFf^rGx2B1IsV)xG zOu$?9drP$GAanG@-Ju6+NjtcR>_Rz39R|B6w@qeGZCiHJiwuySXUoeoyI3s~xp9V( z%L?M($1n@d=zud$@j{bdd|Zi=y{S=c=PZwUXmp7!{MUx6;Y*J-fO@(MidWROW3F6r zSRa5>P7)vOl+95{WQR*r)!H)0O>i!8pXgVM7=S@sw3L-Y!ZQLRf0e9e5Jf7EU{%jGigz!HJz8sH)!_lT(zm0 ze}Bikp~j@i3DpcXojBYd7=H>={so=Gx5HmewQMW7)&3@I@E7uSoInK?+^!L9{3&&9 zZF+aL5u7+`)DwNZCjYrOdcgNb`PHiHC;Qgm!{w0$@^9S;^Oj3$o5CcP?PW+q=GzBU zZt;avs%2*v?r*mYQ!b%9cInO01_2$obIKMeDg3h|tC1r~g0}A{wQPV1x%lD}H`NL8 zBHiPDsJ#0(1D7LYTTjI{(+Ps>DGp>>nKAao3P>22=?g{e9D*rR0p>XnoN+=Iee2m% zo9?FwlPmExM9OD4+POz*NtvYYm+7TZstHjr9Lqpb;OlH?gl^|B0`nkq)nRTxiL7CnDm^Z z^bzf-_5S-CXP-yA_ftnUte=;zaz8QOyb5ACd7R;!RZol}N^i}l!aB~#0R=}a_lu$I zLx+gU=vV$O57no0CQC%lljQ*=*pYXXhq7UR06CAy5stSw1l&r{o}d`6gV!_imtLg! zN!fTtBcNQ~tKO`?G~p*Mrlh$AT#l;iVN)f27HeAaqUBdYV+lWCzR+LhuWfs5n4X^! z`^uro@pxB9KQ!O8#gBP4=2rvp5}TcqMxTR|4x4khyoOEc+ppJ$Gzu1qqR#htT$Nma zr+L8^^K5Zf7F2oHCl9otPLJBSnfjDHOR5d;S)roX&%kaK+0(6=kHPAvH1dy|h|FuR z+?kz{#bzh>r_`r;rW_f;gH?O47joZ8z6-d#O=dD`x3b`JiUG?6^g4Z=mkW1y5#rhw z-+)Kd^uv38u|z4enM_@zHbJzuK9Uyo^=eJtio+L-eqK{|WvXL@5mbnmDso#;;H0+A zFFwvI_~@ZNuTZv6VnTdWQgVeZ`2tX+$Zhr?mxo@0<}{K2B4lj;FCk-PWdHvlWHs8d z&RZOb-IukScrF`)q8cdof&fplsjPPIow^rpvmM-A^mmDv-Nc?IHEee~ z&ky0ejzzK_+$HduYOSK`?5}s)!&p}XGp(|UN}UuZsrPU5cfvzL z^KEo>aXq|=bN%Dw^?Oak7>-BLi6jYf@@(Rg*~JMd-#HH!KlwSF8mMFy%f&gSIO{B; z-2MjmAxj5k>y|({BU|-R&sQvB!5Jd;HehL#r#g+LINP)4I9SvLn*V0IJrvX6D=MRy)_V#QGhK~mr@U3jIHaRQ8KJ*q1Y8^I5uxOYRlEufb51Ct~xO@wjI>cI%~_ReNy|t zHPqp8eX+Dy@xxlfRx({LxA+5YO>DF5w*(fGx0AoqNvNZ!;FIv_BaV0uR^)RkX@)Hqkj%W$T0i)W;EMIXmBam1gvewVni294@}ai)I#%xGf824N;j`~kbJw`l}B z?~?<2$#60$>fIM*VG~Bl8YMOGT-GsqsW8#U<7B=$ysM8Iqy~U%*B^1^bc@>_AP<9Y ziyf>ToZ4uSf}%x8$)^Vm`ja5ZQ^NrwxEOH?H8}D0+_zuu*Vi9KD1TQi?l&L4nDQlmc8SBjA&bgd&Fo_R}sPLogsaJ3+%u zrUHCiwGY2Xqj8w3s=8xxy%_fv2gT{Bu`!|T+*^Bjo))T1BEVq|q=sOQI!;{Js@p`) zX2c`hSuG8>F^0l~Tp`0WWBh?`%jkR?I^i)KNxvIx zG*463qf|k{BpnF^-hthRqH~-m$f@5UgRpS|v6*$Z6>ld11pt%}q<6$=-U6svq z!%xc_Ye7rpvQC9R0+iee6nmhlBQT`N_ZLE(sb$+NCF6-T9k{`{2`+h*)!UE5NGo%` zW@L^ZRU|Wf{Ah8u4+{1#96{h=fSW*EXn*vys9hI@-JO`17JjMNFdD<)Y&Ln+J-iK+ zKYmT(vl2}n`t^6{?da+f>FRe4@Uc+J&M}20WX<C_Fgh_x1oG{pyMi4u!9u3;R#({7 ze-(+VJ=@oJC%Xw4&Shs-fXs$#vI6IG2<)9UOyDV1a>TApQbFet zK)AJHs@S^5N47M`;Y=Gynp)LC+3{SObA%{hBKwhIz#sv9$VV+mC(H-@qo=BQ-ocY1 zP;BajK#y{%mE@{t8n4f`w-vRr3!==SZ6yMB_#tast+|N$+d#ooe6)cepr?u}hdZ9I z45hJ2RfpCDQbs?wqP!TS&f041R}6CR&q~w&o!}UblZErB5k^J{1_z>j=E~pwY4=L@ zLDc&Tk`2~Ptp=&xvbe6B{{RVmflg|ovUlN`L5~OJ54=yB^gT3z3jW)Q9Kni7zG9OU zJm2j2aLNR=3f7%%SC~eOI-7MT(>Xjpf~4pHo{qmTip3Oa4|euQ^-!}Xom;N~JX8=! z!hL>`-$7LM8QLCjF7$GqiIqs@)zwg}yGB)vF9rp`~~NtC{pk93H{Hc+g_O_ zG9i|d#pOaGavGRGFi-l@0?vQF1YKmQcs#CxI)vd>hSBNUt{(zD!^$5_0_-ZFM0%uS zkul&x^vG^Rk)hEqq_D*gO^61L$98a`9+22!n|Ung#d-p=U{LZAX6c(Hm4TY?Gca@F zkvF6l;>az7G27(J;f59xIwaukEylaM5PVP)n}4Fe!0abt>7zMmr|^UFM6AB1Z9@sI z39->>FF5#1SY{k$o`#NY|Fvc?O?We8a+8FX?8*gYW}1WAx=v-Eqo}xarh+r5-wSqj zP7AU+cY6I{>|wOU6nI6D)pf2fx$fMN zdgQx%R8r(k*)eX#g$RgcO$%N>S`H}KJk`?Hk*gI*1xmi4HovrGVOz-RlEoTzH^cSs zv4XWsS@{It1~t@+uFdjk*WQC=NdrXDeK*%2U$e=1ZWBUb0k;HWE}&;Wl8qPMC|E0W z3kIG=uf8T@CXnFHU(_ED;7Sxmy#V&syUmPTNLn}oO+Jgi1MGrZ;JZYS!?l$k{ilAt zdb`G1Mi)SL{>kj7?dQ(95`u&GI6UtqB=Jj8-FpZy$L-KWYI=Mh{)-fR8*^c!M{DSh zgGqeSJ%nTZK4$LV1k-N6ajGe*732ZNP24MXy2_ILMaB~{2YNKudwzHVh>w4vdQ2o< zhLFnhe$7SSn%+qv>nUwPRS5B71s;W)OL{d-6Q23|xn^d{mg5`HDl!1~1B=fblnvgU zU?sx9ron0QK+G9D%>gKQgCkXJRqZ~qCqYZN!{(V*^Qf!*moD!8RzT=sr&^Gjyg#d< zJV4kfg(_dY6Uhrisv^N2IQ=V;{sSOc@0P-a)!{)FKaJpxd5k2MPjL@l1&WL^Ke8HC zKL(%%MZMo97#S=U?6RE#Vy|0YK9MQh9I-5^m?r1h5NhV%noT?62ZmFg?@_mm1c*vI ze0h06C>rvNnfMGF*i{_iCD5e4@yAX}^53)EJIpvg3mWjh)--U`S#yO`giNbHoCx9M z)(lo`)6N3tFf6CG`u;E`xL{?6zy>l%(kpMZqyPbP0u_>&S$HLhyu7ZC!e7tb)DwkX zx2_Cc;jMgYAWB@SdSZg#*K}20pib4 zLM4zmAYJ?D;#y?;--Yj8bbNJ9m2UdMgxk&d=0n{9d`h@vV0cda#2$cu$9djDIdKDa z2cCh`!}x@safC5d5fdBBT$?8{$%OfcB;P!53mTKxidH!UJK*KQvokt?&H3S183H}e zrI;D17CVaUEQ$0>!VnY@`{o78y~+Myu}^l9de($k&vfEJSg zLV6*w3m9NA!|iiL;&O*Sp?mT$3#$3wMZ@^R?Qrn>0c^lgWN*#Cs*H zCjnqwv2Uu@%}5(n3BIZV-)G(675psAZxSC!8nx*m=aO#+A*}mYRZ_IBF?z?>7 z9H&K&3C(c{fmWiFS_+3AkPb5Eyt~Kor*>+0M#v#WqYH5)@(9nc$)dfiw*~Lly3OAw zhCxu5a$&p=Fako((CdVRr9zA@fY8y!lpk4$YWFpdM4q5CJ`dVID2cKM;F_(8bV1AD zWUN+*_rz2STL=emz|~xe33vG?nsQ3(EA=n2Q~Boi1Khz$WJ3)pFwX2|YU`UmHnC7JW5vtaA)-J< z1dDn=8S5<5nHQlV9_rkr^)wP9%NqY`T?BPifyJ_f0NzU-o^Wr04n8e|nG=3R$Bh51 zUJgW2VN5f1<@zA*j;z{>ao6K-x^|PkHmiK{`p2ljXb-`!ko3`b`2(k#MZ?fsy#-J? z1fdP}_1Qnx1A^~V{)MTf3#hz4rx*F8ui)TOx6C8e`!fwr1JEFPnqms&kpcg_xayh~ z)=@>eel1l$&;$x)CcE?FZNamGyGW9ee~&N<9Oh8Da;iZc8_KE(p>nPCGr`D{*WWuH zWh$rYGCTWV#!B@2?*v|Ab?E%e82&13#|9wZ9bAqCdJj!VkFXUVe06HygcVLr13nz8 zAs-Wpmv6$=BM)J!?Fr)V*V$tRuzi0%HdVmJ55mHebOZ^S7hrkacb=?t3XMU+#&+oZ0 z2=vSLT-|i`M6S?yXps9tW=nP$-MTmeX^ldWc>Ea?_FFN|S}T6Qo2VA1|N zHuG~onW}akO*23Hi$})&wije}DGvsKJkTz>d3@dLh$66$22j7#egDE3(C!N1O7N1j z34xT}wCq1Z7!auj=p)?1%CBnU?3E-9+?n2}?q;4#( zB=xW_9G{Lvmp@0>qTS11io>CpvekGS1r^07nB!VT-vnD;$5jcD2uun*C?6pHuCq}{aUQ`vxet4V<|w14w|b#;9awk{GE4hZFYF% z9F0bwt?1i_7S4*tpY{?=0j(d5!R=dFMwT@;g3eB0_1Ucn&#mw#2#SuFwNN8+% zF%008SF?mETwMrfe@_sZRz22ru=scSFhZCrn~=ZylzV|@DnXC*6>5*xf{@X0DfKl) z@L=%yf?gvvOylFjO($Uo{|v)jB07=kLGDOj03W~itFkaQ!y3dxVQ3;a($KeAZs$Kk zF_*|(K0}M&Fp2vx#0oit(X)+NFVSQN-!g;2F8q)klt1^tP<>(m2l62yq4PzLkQDCa z2i_1>he--`+1x`ELFlp4Kt6R%T`-$4dU+J&4+yqtt}K)xgyQRd%coJUJJK7+aIf0< z#G ziIX!Idd2cn5Qf4U0q|I{OamKVZ2J+~Eemw+pJ#R|guIdu4FeQ~bi!7nP<6|lAPf-0 z@Rws>tTmdk_pKBuISB&}tklh7y~D^JE;e#jy)R~?!%h$|PWiKE=gj9PTKqwyt7Uza z`S#p_y4TJeWu+8jGhXh>21gEF4$jJ@69lIM6?aUHJw+-fX=fpZd-{cLJ(QMBx9!*J zpemOHxCV_mJy5XFB^w`hvJl~sanYpfD&4S$YaF$|+rfxXd%NAV*|F_S_S<}V(PBJ6 ztaU;Ve@w=6iexF#>fEqt7A@OKYL+kyNNioUq)OtI^`}Of2649j$*pDMFF6ER9{er-dSvba!3^wkQlK!0?B z8P4sVfR=5wx8^M5A%v>`=Kc9=^jLPx3aFw-0SN_JUEVTfyxy5AYj7gYrQAdl>?)4!k9OK|L_O6AG{Qt<+Vx9CgemfK z7i%B?11j=KzV%=EBl~~kj|`lg|2v`A@;~{b*#FoWmTm&ewl84WLq3EGT-NhV)(dGS zm?*WGjpqQbz}7Je0kL9B!#>8G*Oq<)XX2>=bE8kzKa(?NgyvdJDe zMt%qTs;(8id|aPC6N|gX_VZR=-p;#<#>};FKZnh!xc%HchwaA?x5y`{U9Xt(EUe>q zZR3Z}*U#Jj4G!J?gVVQ;aA}iy=5V@S?R-gcoq%o0Xd8#M(yi?7tIE+PN%>x-bZCA) zZjO6ybPkGD3mQDax7UZQQl}0}K>q_HT_uWK(h zZ{TWWGXr&pCuG%Emgyu-=}bOmJPMRdE%fGgasP=|3xW^f4F+!G4MricM>a?v$2N(x zwDOL;+LiHh_$|yK)j~zh%vNW)WSwpoX8WS_Q{7(bqDw;Ml+TCTqD`Y-oYW$X3V3ML zxE-}?3Ls}Vk^hZZR&-KzkzlBvUFucH%<6wWW!F=qP00*R>p?;FG0sM#{e;oeKDJP{ zxp<_Q9k)@~yNp`q4S?-q+wr>E&o`nww{Cv8;cm!+sVR}`!WZ&dHnV*7-MIgBDYhfs z<9{KeQTTOhYP3@tE||8}x3kfUE~wi!70sqou^TwJcc&O2U3M%rg{HR8sikx!=}E(O zuijrTN34*^*>*x=ty-2uDGQ@Bya|ytf5^V=qRFQ?JtPJ(o@Ws0D_M3XFfbm6nmW@n z=V*jG2lcArk%e~s@!h7wid1)76KyY2YV=wd9#e{X?Cqw>T+@kadO7hbaC*CTR-Co( z3Eo~}qPqvtn_Z|p_-v%?Djc=eNkgTd^i{w9unKJ>xgIs$963Em4-HpvGRSgLuDM`x zDRw~n#50?rU5MNX;=a?eO*;nSHJ<%Ih=TMmi+Bt%JnVN(;|edAY8sl!^+=e-@rs{i z!kC8L0p>&P78m3Iza1EDu4J}3;+09s7ER3C!#W1GBXZ5v=r66vs;52^jpCK82z{45 zI%6PPiBDNb(`pFpH^lLhsT8?Ghg6oV$#lecO@%9ZtZ1^p#}Pd}5)Ur2&}f|NomI3c z3{&t>by2S$0kCJeW;bDY1^}F@pb{ zZr04vNFBr71Tg-r)=}`g1XfsGa2F|%SjuT*p`KVU*j+`own5C4HDvJZBcT;-DejNz z7e1P_DC%+Rx`3j7+-mcIz*^t@sH)0F9d|Yz;k26I33ksOYXN;0mg@}st!UNQ%)jv| z9?`lFmp{gd03Or%1~>eg^&sqH2hxQaB)O?8kNu5FFGHsCG{X@I>|Nj=XiGHe0ewbD zxm%1C#2+b%z&3fp{FW$xcUUa|6jA{kMmvQq1|ko6W4nU>7y5GEGKRv9mJ7L<@wVHe zl@P!9gscR`SIK^gmG+-l#Q0laZ#Y*qMs9WGDd~+4^4}7o&vuBVzjsfe0DEg7K#K#y z<_Z&xgACJw>I6?%{_Zxq@bXggOz7=zMGgZ@0F|R6>z~wdJ%i3+PXD=Oe|oh&_Ibz? z_DspN2Ka3d0_O}1`xn;EOQT3m{I7pQjA$cuxSE``G~S|ojawmqgrpRnL(s>{X{Agr zu-;vF(G{2+g3jy9-pVO$G0h`>R}YBU$Ea9Xk|kiGk@R8DXGS_kx~mn$68-Ux=-;o9 zo3~+;9e=WYGfA@)e+U4I8KA;Gn^)cT{fJ_MC`ULp1}$q*5_#2*5}9Ai>GJHyo93>$ z;W$8wdq7B01ozb{;O7JniI#<@x&`!-IVA%i@;7eS(O+ae3j^wnX_(?h5mPFz-J*^_ z8j*Vn<WKzQYT(j&w2JQiTU^94v+&)N6w+Hj72>v{}P8ZQd4 z=Uz?Op_E-kqbb}98G~**N&`wtU{-@pFw6O<%-w$6%1uAXy+kgL`on}{!y}3uGf1Yl zZkcWKLsU2I842nqF9{)$xY&aJVVRiuLQO40CAZfSRk+XB-m4yHn?=4=PsRTDMV`>9vs+F--BV3`TQN6YmWJw?r@2@tK6}Lz&kr4UB z7cZFB2^2nM!3@=xjWWqJ{1~pQ0{(v+@#|o`s%)u;X}pc^@P`}Vc49{?hpgutNrtAR zWOqaU6qTY*YY9?`^r9?jkZ4uyHXrpsiT1-TNT&X*-VHVU=zcZ=cH>4cF;F-a!uGK+ z-S^Mliv$zL)6qyjbg#aOXt1WM=3qiPiTZp!9Ij*(7#yJK{B%c9@M!@V$gADYXLz{yxB zAx68qUi&ss%O@E|i<|^T3@-|Y57FVw8IchC``DPHcTdgq>^RPg;7;B(z#t)V+y)yW zdbyL7OaBe@%eb;Vk_SM7@Gf1~mHc1?l)-V{zL;t9M#LM~ySr0dKmz{PgRV;Z$!)s0 z6G1Pw%k9c`Oi?q;ut`ITJzV!LD>OPp2Y6Ky67Pr3rTXQy1A~e>LiQwQMNff&u$sRF z6q|MgtbsU6z~hI>%XrZ6@`5geahzoHdB~Cd!bw}OCa!68H4wUqFoZ`3zk|nEGT~=8 zONs7=YXc~`yRCD6KG<~)X9TVml(0ve|K^q7W)B;=s+yU^Br(-C^EIL_$qcu`q_8S4 zjW81coYYYCl_Q0BT2__tK?kDB69}el`3jMWMx1P|d&{aINfILxNv_O1u-o!*c)xnvBvCM}myVMM)VB35w!=&jAW<2^7}?k$U%F_iCVB zH6On0jO3qr&yh;6OjlX0^zI9(9y5XFV?KBhY?7ObVKA*>11F+%#SCRzFs5ROT?t0$ zw-6Q|5|u3b*cPFY)24(_B#qUproIEiD6r~o>d91pUrpOU%?k!0IhMAA+Re$vj_M{T zjUizSz&ZM1$fh94`Y5m}S>_`SIHd-7jCCpiS@EjfvxZHDwRorlO=C+czAHP+FDmip zp}ywLlG1#;qz^&|UN5XvV`3R=Z*pnUI-ZNyD?&GB+263aoXPQFLi~MW!u3)n6>G`L^&k2VoAKJyF2dyP?5Te`3ioW zBWCn!x8o3xK)x6&P}nZ1H{3{7WqMs^F>&xD9kyB#gep{K)|mu0X>yaFl7Uila^#`Y zN0WJxg8;oL{?HeNa}9x%$HQ`;1M)5hB*)VS9_Wx2^%%Oym{v^vKzdmK$}3vQi^1BA zNXYEDkufc+q#4EfCY0cs=}=h<8#Mbe6XO1z@TpnsKM2!0=9aoIMk&wLS|};i9g_v) z3XDw20N4T7d+pS)6~N&g;HjF&AC7(zwtzy9(E~06!*XvV05bw<`s2iLp*beUT70qD zxPKVJk+Vb~50L0wfwxNjM?-@D2x}gSSqLHqrkbIWj$whC4{%)Mp0M;eJ6hxG=Fc@@ z{cFE^LjLq>P1EqP(;!F8ap%6Lz5zhB`1WHZLC(#Rjt>UlJ>{iO0AWPNAsBFeT!x6&bl&a*qF1l-D~8lE+@T4x&g?Tff)=3|;59-H>5!P*+%lUgywjOeFh-@pfe1hk>^Y5cAxcZvaQkC}Ml8KTun-6Y@Xh&15U}|$ zWAumseGbq}$?8r4Jh81}eOO{Mj{U6=@Zu@0-pb9O;yu zXv5xwmW}JDvs3yk!>#Y_X9=%TYfp=ksg|K`kNONu?mtt57pT`Qo#cU zvOGjuu9OwRvf~&K*s+rRVmUrxTMYin%r$sk-aW6bdAq%nFo;D0_%mFZLx=ckWq_Va zIuotjeUpX*7w_O_gV|{CP9ajcvLy<4K4^nb#JpsQpcyb8K++IUhb%;=oaN(PLoWfo zGiDG5Zca2rS!aFB(N~^KSM-g7M~6Aq$%i!(WPTTX_LVW%8yA(7WZG?AdJ8B&Cp!=J zD_=xzaICLJ3~9Px_kBQARn`WCbsbdQu{slxr?8 ziQ}s}^I>ly>N=xPW}C!7L_@;Yj~zv=Eb@(s{2~p)s1cuIYLpjvpZO|;A2Kx(l`@S0 zJE_W+LX*K@6TKIIBm+ZLy!xG-z!EnJDS|)dXBAWBP6k@Rj#dph3hDhPH%z22%zdVr!apZIH6~ScN+W!W?Y_!lF-}T72=3 zKHvFlTDw;d%*CoG*Ompi%vBOFOX_G8uHpE=0K|#NWRWD!vikEfyu)wRq@e1dtlj7$VhcmmLo2|6M~0^ffWOKh zA;!6_d+Fkvoq#GKEJss3gS9sqdNH9%LhS+41U}AcgJF0CWj@2?&&(oWt2`65vg)eR zc#EuogXZT)`IO$nltrfY90Ua?blQ$~w+&dSv^XWs2TT9V(^yNT>>PjP#rC70uo z$-`lMwUEeKY=C|P@vSh*&VKV>KxGF_wNakgs$BcIRX~pgK1$e*>a+`Yqcgi##Xbe7 zt~w)QjdxRrnK7Ko^1o^)C3|wF^H=El!8eRsyQ*$K`nn%FOqy#yQ2d5Kn|-Vl@i?XVl^yBRRUJ49`^*Or6Q~l>flu>A zZ<`;P+oxnhuHk`cFHJCyYi_X?)p%>o)g*8>L%8+1Ev%2nndAGN_~tPBGNS?NxwEBa zTvKQJ=AGif!M7ZDe_(!BvR2QLkEIbRQ1n=EbW^~5H(}VXz|ydEJO70&(M}Z|$|*6B zT!=l&v`Wh#J8&Kz)n@=JbMMw$wwv_!NYdxRD=Ir`@-%Z4>Kws)l3ib`U0Oid=pw`^ z?FOKTx*GuDCr?ENN~Qtej-e&EbtedgaeP#{{d`8ngDPzqzngyG>5AC}e<& zfO%vyDXf^#6_~{-!>{I?bII^sf81}harD|J;?w>iE109?KDP_x8L-zDf^ZkS{k%z~ zB0w;t!w3QDL2&^!SY8q3f$Z~h`j~8Aj=Zt-7kpP*AzekkB|atV1VV^l?Q}8zPw$`ny;n`n<`BxGiVs2Z*)JsZ`#(9`>2Ph%P@)& zWvyXHqtTk63+CCSN{MJ$71=L1FHR>t&u_cUC}D0!15LC@E9@B@>fi}O$%iK5k@GLE zKf2jWGx!$TlgR`UDcbh0smvNi-r6#LFa)5v`2i=x{xUBaT;;$evp>P%@X{@v0c1tf zN1j0sCQK&2G!JX)5>VNYayw6wUY$6K)2o7nDUIZgx~nd#cK1LcFCiF=Xw+Lh2}P8> zFJ2T%!wC$ko#LT8Gl*`jT=cthze{zpOcJBY22*SD-`k zhp|$uf_SaC^$^S(D_EN{?SE)7WT!e9T#Sp zV7Dy79I|DC4DZnpXNPr5k58DlGf`}xkLW9|Cx{2dH4^CBv1lCzwu1?(^u)O zy_-B9e=pEWN1ue+^^4;O2!70#>Sd+I!QX{5_>Q%|=8YSK4)b`zrf0>MeP5T8gl$cYRJgO@V}>qHQG~7Tcb(a&UGsu z1KTR{<$}x50+!;uG)rkF+D9#QL%ocEZo)Pi+XC3zvU(rRC!1!>{*2^Jl>jWDR6cdM zQ3}B2?rJ9!$(kd(a!Vq)rBC!)thTx(7_DIeNYv z_q0cr`$)$rTr7ekLxRS7Af%?78xj_X+_&B4DqRm+Snsy?*n$xkrnY3Yt2K?raTOs1Gi+$t0D5!~woI(1V{+gsJ%l!i>CJLAC@+rA)!dh(IEJKD?bcgI zqAKO1YIkuwzy0o|+ZxWLCQ(CyO;_qmGpwr1T1}(?+3?H1OD9U-C5(h3c9F=C>Kqw3 z2>RT)rmovjS3Z=DrW3(VdrkK-%Tg+nG->%RrlV%s_D3+V0gPC!K;YiRf}q>j*V2(I z{jFsd{;p6LpA8{+0klT7yA}(p!R}pq)VtpPeBFAtR(E$+qoY1UwHlbwCHnpIrM@mJ zOJUto@kR(V-MDE81I@cq(rq@fa`UcfC>QtTyPmVsRAb)nr?A9{rSLj{P(>FbfQ$8@ ztF}i^J&R(zLG1!aE8A?~3CVtf34i6!7?mD@=RWDf#rzD-waLCnSBW^=t(BQ%b7+V_ z12CqnKvdnP@ZWj5Gnz79#OrVc4U2o)wwENKJcP_3rh6)1K(pELatf128>S2ewolP=eY3e{M|lQ%x`P2g@FZIi&P4C$ENOKTun6|_ zGwrfPcX6{zF|$usRkp`Gv(NzuCEN@!N-X2mHMQ3^S7fMS7_`J>k`dDbe8chs&y}s^ z>tb990c^3TsV*(Szs<|VR0rXhCs$AsU0N&At~7*-Vz5v6g+U(Ef%ixb4w#tJx;^>} zS%lObmF%h>k~wb;c4;+#Wul|H3XZs2>wz|v+&&2emz76^Sv&8{s2Xa}=kBgxpp&kw z_36LB)z>Uq4a+BL!j_w00zTTQ%q7iYa1;)=)0s^tuAJ_#Kikq3^`jwuuTwD<>es#Q zWh<~_7R$dRwo!5HR# z{N+&4g>xDboA9{4+%@~VT7`jcm@W-xt-1aL&@SJTb#vS7^`P}g z6Hre>jm+Nrls1Nd2RiQDUTl-V03fzt6Z^>}Kv8U@u{TdwQ>?*U^;aDsuW_ax#q?-9 z<*A})3BlCM2SY8Q6pJ4mJiK#P-fUM*L7RiAfw1dXk z|MC(IlIwtS;jmM2U2_|mfG78$l(%+1<&G~yd^OaCKbalkeF3mZ@L8S&TgFVr1 z1@u0^7uGd!T7wqQ^CnOB!OR6#J^~lfJ1TvL&#Mn~Y5E?W)l9B=^N&IG@x$K^6aHnl z1+VKVkmoE~4{wcrCLIAv#A*W{<*~CEAu`fKU#S3A&`q3iUwbz9ba$gOV472cxkpe$TJ*AuMH8V@24tg- zDeV7q2$&TlG(o0XQg5@r7ljvp?v43^cKNsqiH=5PJb|=LI{rA^FYj5R2$!p$6?3uzUqN^- zr-X*tRf=X+qNcy{4Whp(+apg8b$WAy$iA!?+vFrzo7i&3p#KAr4AvVc%l+*U7xU4~hXIDMMtCdHNZ?WL!5a~M}Rmd~K|{5Zp!%D%N6#867^ zSAhmTM`2+~M4e@j<*isLZGkPB_nxR1kEfVf#Wh+Fx64!iKz@l%tPqzk!0PR8^f*2AN zf|{iKvoADuyWiD#avF<#v1foE3jr*Om)>9YEzZy(zL0==O+_Cb!TlI=jtCpH@HiPu zC}dnQFNCfS10x6;N_xc+4j>*=h6%-I0Uw_uYJ?D8#s1=Xs2rVcujk767+gXb`@(=> zJ~Yg%SiFpxPp2*{(+IJ5Dm1AJQ_i;@CwY)~V}zn%@k%31D?`UzH^7GW)4Ix!qHswJ zgAr75xB_bUf88I;76Gz;D(%W0;&psekPO79CL4PZUY*)?*aL0D&f~A{6XX1t20UX- zJgA2rp&k*u(oOfKWKX3J@R|M-K*PrwBG3j)P7@V?)Qg}IC5fkjh$?$uz$a}Ms3xRW zV5gs>GY07KHsXvO?=Kr?87T9#zs)uNneu4!->gcK2Nf7Vlovx!jda~_ZccB!+v|6q zv&R*NWzzH}jW{42tEW~f%D4Mt&{NkpOsHQY-5a-O^uuPMF{l0KWO+|#r)PbKP-=#_ z9QYlI{k+$~->q70nyoS?Tynyld@wd6?s*Kj@3oYXcXJ`&)Jl;PV~u?haESxZ z&;m3teYeZB6CsF5$UO3&jO?H2)l&sixu_mH<4Af4?)Y1IW3jOi73#s^V-CmY5uO|& z#`Y%=7Ct%nCfN$(Lc=KA<;`NYu%SR=-p=k_8{S4hr43!*wRw~11nzSyIRVZw=zGRC zdTuce1q=KpJ~#_`W3%zdsVQ{y0`fI>**1ERwX~U*F%F6H3yKEvNyU!%7>Y$;>ne#X>)dQod&M4)uG8a~@_8 z?kmv)@-AG0fp?PdM<; zV7N0*v=ko#@)?+cxP|O7FI$v5piSXwH-93Vf!GKa8aY@QYFL_x?dTctL%xwAmARQ3 zs2g{-mM+p)Y-Q~;2@%p-sMC|lfMP)*hI70$uP?GjwXqbo?R(6|5jKa+QYNZ`Y@L3( z%ZBUA>6!ub==O_FgoSE$pKI|-wrb(;>%*;yxYp*3dVPjly5hSy@}VKU`D4(yJ+SL5 zCEYcOBPaRGzBmC2<^Q~P&P@hj{6+iFz#J{(;-5VusddoQU}Mw=pu|0-sDoXpN#({tA!(D{NA*Js64O#q@bX+9u6-K?i3=0~Hs&1AEeXB-K zeNyrTIVI#y7|)!}S+b2cuN4(6D+YB}@$+~Z^MvmRK>Zd$=)|3O{}xj#Pt+kd{swNB zo)LsDvNel9Ecuom`qui@Ud7G_DDHF9`h!2JijUq$BMv&XHW_uwObJI!)~2scY<~^S z0XpsbzBc)3dC?z9|Glahfx_(n`pWl3aQ`KT>MQl`qZ^KXHM{x4QF5A`)#}N`A+7bw z5vfEQX&jX3W3_PCI+Blm`rUE;&>B+)IA*57KE!O3xrEl*W|_y{&12Jc`H68cnP@AA zKas>*Imf^tHXIm837GjRqO5>Lp+=_Hed@qe2HX*t1}j3-Y`l6Rt3ENjMoEo$WissDdttylVdrn4v(kut;hXx#3}3qb-4r%|P7!-__+TS6%n)e$li)RUG-ic{b(7fvyo|T65YN`KxfS?d}XEwz|>_rCaXd zX!siQpch5>#DgNVQO7OE^Mo7Vn|i5XbJu(OC@=nonAzU1NPD~bPXT;uglB^ys-x@o zYKZ{oZ&X6bcgVQ_ET9Juj`Gt8((ZhjEGL#U1?^5s<>vQ&iDX0)4!2|!Cs4_=!@*90_55KE$ZiR@WyGQ;&DyOP0)ftU>F}!gO3-=T(pQ z4J|>)Q#H}z6Z9LU5la^qqcl!zpt}oRRI>M~G1Kpjn^%gVb`(7&E#Jy4^P?*gq(fo3 zy=8U!!xUFdQYI^hYJ_#k`7p;TIeKPps<}FG?a*22nj;1BR4-EfC9j}w1dSHBVR0nj zt7?aolL@JStxxuu~MvoDTe;X0{@;E*e9pSN4#mnQ$4uQI&roG^Eg6Gtkomn|84R&#<;AO_jW?|uP9 zl)UCk4>-nbD$2o%?WH^s#FA<|j>JeJ;j4fq9Ys{(6Vwn1(1|Pn;sB!(=9l*p(P0?WY7v zjI~O_4R51c(5~1&0LLPAr*&XS-qcT^Y-#>beXQB`qcK z_T|#bG$+2(fZ0h={aNReJC;tItw_jd83>?+OOV9r8Xivrq3M_k8o(}-4#cti4;aAi zV5pQ+V?dvIprOM@4MLYK2UWip^{fR3^OiEpK+5C@kdnq0ICS`>>7)_Q(DwC^C$Q<8 zP%ZqdxoVh+;Tg#mqW}wb){AA#oftT=G{PeIQ_L)UW@|?BFA@BpIv6Ju;@;Wkc@@~< z(4hDYBo+Bm?yAmC>bBY1Nj_d0{DT+AVX!JZvDwuvnc}b#n143lQv!jgUcX2Q)x%rO zo=xM=6ti#ALuUoo?9+q|(~`M{LLXUI*F~$W2bW*c?~f#JP9arJj#LzDjNez1!afo) zzQrs1$~P`b_S`t2^3?VdY5@#1ZtW-^Io@SFQXm$C$0YkwD%=54eS4KA;v(wKj-cd8 z8}l$2T^X}12FY&1jM7BZx59a2%B{Y4&7s|QPLq0DFMp>svr-C8$t7Nqvcv9Zy<($j2m=d=RIY3*}hBu9r?aEYV zwla%wnpVWn$47M)0;b}N)$16J1|K7GvS~IOXC|a4m|(@qojHk?KvpK!F#q0k+l)lln53!5RR9F~h-{YKU-JoqVwWb;I30VQJ6p_Sfl3>*| z)&6O7-Z~uBd#%ra3>cwOzBrFR6yx3)p#u1~lfLbd>F2*a#oV2{?|XYS>PWxcz54DR zJKy|%`)kzk`{i?c?oIjPdlJj=dHUD80DRY-4_Pn5*V>1yWrtt?rs42)j6NRT{PZ_O z&TU!zVVcvpVE26m&l?X*g4N29sm7L0I2yh>6Ee&56tpY*+vUg0oFhxKR@7#C){!H} zk`iY(Z;EEF;ELqkVET?HXGbf%Wyxl`hs+?1fl8er6x-3l{i5$Xna$Qyn zJF28K!zXKV^{*>IG`2nFZKCd>v4)^}O+%e}X)~%TTPeC%dxrhU7z!S(beC((nlC5* z^D%Jz(rSJ=`1(JWeN%PI6aCsG-io{WsSQ1@SDzWdx6X4>wK!S><5j9yOHHSy%GFUP z0o5?Z)&<)?*+YZP@ZK6M4BE5~>uEi9pPQI0Q_k0Go&ES6JQ6QV(XrV^{A{!>fv{gX z>Aq;Z(Nhg8oMT`U@bSG{`?X6pQ>Ar$GlO4ykSCd^cfrQDH)<=5A6nIwXw@}6Zl{9)>4bA8`O7&euSLS27 z8*Enivu0A28OVp=_Fc2V{=qjNs$NJ}ve^=LB-Z?cr(1_FO-2?)8i@Q*FRfoVSpm}h zM0A%<85^2we+zGuQOA0z4dj&0Oh#bzEz74EH6$>^Kkqw3d3$Nku$z-i#5KQAcV2u| zSJizZyRv%=Y0=rrXn80*<1g=CJiT_)g1>0nGgzwIEyp8CIfDXvgrlhp*H$dx8Gb&4 zn{l|c;JSCbhL)V{!ysj)nqu&NuE^UaHioPKO4z(dpnZ{KO z$bj$HaO~(h3g?5wk!j@hn#&!_XFF+T;&sx3QefPTNHV|%&04Cz@WO^878j{2p-=oU z0;j)5VA{Uk6SxJHwf$m{5jc#R5C_mD+9qI^is)sUYn^}J7=j&n0cDglV(!iKI^MDW zgpB*FsLl1OWK(~uSkZ;#!Q!O>D?*v0;?L|MZsZgLo2rf)E*7N=`Gn5geL@^_j=eZn z$dWmC;Ue>=z2<4J`Nzstt)mKv-MJ5l`wm$2k6PlV+v?J-x-JBAwiMcSR(^mQRD^?j z04YohBc$wQ&mwB!4X^I4>jpKg={Jo4=m8ao0$M&-0DJgJmPZ?FAsV7@a@u0T5<&89 zAYO&dUcninWdUuH{n>`;viMvf(UCMDp!8WrHks*YW>eM)c+byqSnwO~G8*m*m!R(` zK?UylcWz#Jai|`k7+7>B*w|D+dw0`$JJlIxX&w3!#H#-fd{=%O0vxYKg+Gj6?Tg^g z41Xo~v9P10g77~BCkT6-Gxe2C5v4^Hc&@y{VzO5plhczEf)KL00ykU}*;%ARp0wfK zlWb_O!P*ivo(QrZ6jA--w!q4;eOKikrdyxP?p$Y{tvg-n7DtSH>?6V=Weqf)+}cH{ zG|dhP$@1Btp-&%Ox^6IZ{?ST}`xaE?D zPPfpgVZpO|iz6O~bMEscsT;L?T;PlVFxo(fU==}Ess8h+MHF6Jt||>DN0pR0d%JbJ zl4P|l3;_kG7E*;A7r`ilXh>_r+aOm{i9BzJd;;m#2%;kOJnX zFm9iN7Rq3K(Hte6 zVd+eYA<-M5m?zwVOOGgD{1~QJlVu!i6blQ8fU=xpJf&KB6Z>Px_kwC+*pc~+Yv*w9zk8J!Jsd36Iv=78r-NnC-kN1!+9CL7EhGZ>J7nI!* z3p}tC$0W65J<)tmU;@EJsohK4^|?&@v+gr zvngt~)ER(vKh*8(z%pm3ATV(F!~HUQpe1RzzFMe{^qlZ&uML^|(4oglTOc4}rri}; zvG$jL(qA-1`#2(zakAzCrIuS!uT;$oW#cK9y$@qUMcad>S@nc18ze?hgpob-iBfio zLa7Yf(U|3i(h57fn3R)@V3bkn-;V@lU?c&MIW`Tj0m-*q#-Xo*7K7J%x^Sze2p|}f zts^(3=M^Qp!v5vziIg;0&T_ORJj%mfNWQ{P)|A-31B$NOx;XZK@zrDiLq+r@prHfqE5L7UH{Y zB2pW_23q6|M=#(Oq7pA8#?%8*7p)gl{bcF{B9)90cZrZV36> z=sckTEyn2E_8RR661hCJ`IG=-RL!Mo=&q8`D1#-l1^@LrKjS}OC_n;JiB$z|q^Nob z@NoE~8CWs?sS4s*5k2Zrg8_k#W4cRv8p1GMC=7yPc~>nfkI+AQn?Sj$30Q?TlZbm2 zmFzMTr(3ZS5Z=uyhUaY?lFq+3U_@XV=A;R#cqwkGcav_Mq6k7ZUT$X~c(EBA$juKc zGr&3pS1_4~b%+m^I&8vlwIZaz5Vhag$BC;VItNq>!DAKL2lxlSOz`&sDYxQ7Www|uDVsVk%vLQhW z%fOVIBrY+&!SC9wWC_y@X+I%M;@?$5cw;BkrL+*Y2!lXZR^UFUx2oHeLyR-OGgo;z zZDc~3h^zBmqxP5sF~)nrS!JgMBQZ2mT+fJ-It6k465j9@YyB!qx^Q)d;imo$`v{Wt zUKlhL+qdgUBFp^0(^}Ek#i&D~wa+vSIEhVcNyVt;_L~O5axNse1&+8lx{pMT(0bp! ztwFPj7$#aUPqWbViARp^9zewiuzapWoZ;Hx3%p%eWUAf3G3M%`#ET#gI9FrTIM9^aP;hZKjEt-o^Q%sxb8Y;^}jaVhneM*PG&T73LmFnzK?eYZ^#v9#i~*PmBQaK8KeijyMRi>&n%+X4>P6jl-v~TIP_L|t zF8NVsjFbyW4Ir%v7?AfV98^j)RCfat9r$pdA5=a|4zvYYIy=|7sDx_%nRXcM>djC<)Vz0IF9!P?EQHCSELI*~EmeX_5{OREu160QEgcR$*+l|rmJz;Lc=YQ}{v;%d zdpB&vkcK}CCPz^yhc(T#UDHUU$sa#3<%f=R|B?Aa^J;ecC6E6n&v|l(B26b77H{5Y z`Zq0k5>nxwbN)? zU!o<|W57=~=^7%ylMiU`(Q5sm1xWnIvGeE6v6K=%1fk;o02ql34W$AiWM&xSX_D^w zq!>A~l`;#w(Wa7&^iaf6JhI*~Vi}<6GA>E+Yc%QZW~xwon&bvs;m+0~s%~(m&`tmb zA>M^UurIJ}YejU)AXr5>TSgBQ1nqA;s$p4F0e-)6CRQ{PiYN|sO{9r#gqnf!P(bgk zH%}+_B$G$Rpb}D`4i2#=GR-~)VamY|Ta!u=Q}y*lN{n;2kc+xH2myRvltcIctE+__ zJ%%Snub^=b*zgbX9lLiu=L}GdS<4X}A^XuunRPuHul3+? z!mKB9T;F_p-NxERk$wp^TJ%T&Xq%}nI{aJ8#IjgCag-IXoU;C0bkzafu@a|lU0@ys zX_O}h2Zt$l}lm(tHc?h+e{E z@}6i`qubxz*2~-{EGCy_(KSGa{S+w%JotWLt>cK2wMzzs>*ZJ14v?*kT2ufZJE=hN zyfEWwN2;n8Q;5MmtXJpB-39gl#lD%{7PtazI`k)$wz)zD$>LKO4*(G|HOeZj9sENAX*%NaaOtDRYEs2#QYZvRzo9i*}T2rc~_0qgZv+ z2b73RaJhsK)oek+N&{(j)Iym8PV3T4t{nVu?qT&NJ~8}??`NImR$RYQ!7heo5jhp-2;e) zZSnf+E{3W^_0SWYnkqv1n4cWZaa;eO`_6gDcsL0;s`PrEEzQ9B%vP=Tu@(ChrvW(*z$!|GYRYKU2;6#r!s^_11x+4Dm9;zleKD|@rH26*;fL=Io@AJ4slxpmx z)xz^=QJC`OIJ@96sxvlSrJUt1`!gr@;{oZ{_dhT#fS6-Lro`5dA7phlD7-UN46@D0HK#tHVbrlYWvMlixRy)!CxY%=_A9xEYZ_tx#xM zd6Ab#VC-eb{GEv~kL0r4zf5m`0g@Q`x&I>#WM=rUG?0`1{{kRt+i#91`MrAYph5G} zYDGtw?XJxLL<5iAn)+axo`C8K@^ci%54a<*{3-AK-_Nxm}b$mHwd;&!cACz5%T*N-lS z5aSb&^T?^xtqvM_7FOkI)x-}*HqqzwvSsT0LIv$u#HNNnrw7`rMiMF%Y@k;!VIvjEtwB`+6qt6SNX{ z?}*hLhW|vpChsP$Cdygldaw2uMN|Joy@HkX6Vo^;O>)S2rxd69!}OC9kFV#`gT)mt z2e!_-vgS(FhTmx<UP7R_~Ge=TWQyKvFT5-IA9$2vbwdE&$39f(iaPQDJ%KJRJk{z+J67+AiVVE za~l^+D#7zk)*_KpbN9)5*=$YB7J{DKl$l;_HJPq};2}$MDNK>zrhg__CTwSya+ld< zwQbX@;6Q+Jl}Ja98w<{L>FkEIA$RT7P+fu=cFZ!+N!j)p%?^Z@M*Z}%(XBOt-gvy7 zd2>ujFzEegPdBM(U`6<`Hf>r_WpwtGiZDjSE4!tbP75&vJm}IHqEYD$PQm8>A7kef zomsT4+t{{Iv2EM7Rk2es|Jb(eRBYQu#kOso?EAV;Yv;brxAipJ7^~0mjn4eIXWQSS zbhABg588u{Px%?bpeM&yEpA@$!WlduX<8y$N84`t$SFzb9Z7b;>vz{t@Wrd!OX=26wat>Q$a8u?C!vpgTUa!6aLRK5Pcy?V0n+i=O1Te# zDDc!#qMvo`7*RU1qy1x!QtD;M_0FnHrpPYR?9g4=vSfoWuF`yaVdi*{6VsJX@+B+* zN5hDnBS$_xz1#a}f6c*Y zWa6iLZ)_ZPf(#U|dP9(V{dC);<{w+)w?%^`tyHC51-lgJg|UmPB$f-UI1usa>OdY_ z7rDJ;J9lp(XcD+$DYuO_=T<%+wpw)KZyCh+-ocdXxM#)CFgKbB7I@*{rUBtbTH~SV z$f}q4ieg%vQu||Hy3$VApcQi12J|Ie@U?@r3|Mm$j0gy10l%?=?|A9f+E4c;+c>Yi zaBtx=rRg(VG^Sy3$i~7Dv><1Qyv}#JfyXSRfnrjJa5@5J#6P-qG5Byc>R;gsPE83e zX^eWhN`2Ny*>Ph$w>w$$aHT`0&Ko%x>;T37Y;*C2jdOnaHciNLy|pA0m&jkLKZ`^C zg9&+e$Yt9~rp5Rsz1FJw`rxFcVQ6k@onumdm{~aw7KTL#LM`@lLygnK-S&hh@IXV7 z?7^lRrNIQ0yZ%io9@d37_Euo`DBcEIw2}2d4e81xwVzPoJgZ1r6O`w_2oL)P!!8q4 zeZW6wNL#uYfO?cIos8vr7y#%_o5@aJ#cKinzO*^6*UTusYu&N?}=!W?M_BUu= zfMNgoVi0-CPko&w-J#(Q{fH?8?cC`+B5?P&HPw#pr1X+$L4f+hLu$()mn zA9xC;U&K-thNoBxV}Jd9QWW<}c*jTV)@O4bhrwrsPbBBGAjlKsbxsA2R*C2){Y}js>A`)-ijM1_H$jOIVm}UPodnZJ}Jpz0G!GGbso(r0_|k ze!9C<{iC7|ZW0aTS<3O9rMV@)uV2%ZTr8x$XQnPHRbHUWdz*(+*Zb)0Eu?8~adQHX zFLu8pk1qmAZLCCN)?;~M(48Rli%kB7-Wz-va2(SW*VZBj(_0_#a>-}F>vz)qGuSf> zt)cq_EI0D4bfZi-{p3jd`l@7yC*HBmK!&>eILC;ZLO|dBhY+SaEi+x(zE%{LHcYH7 z3%oImk0<1o0_tlGwg)tO{le_Tan6;Twj9a=8Zi;}wdQh9N-GS?U{pr~dnhqF0-6VS z)6)lq3Z`wgHii5f9zX6o?br`!=F-L(=x4cEhM=M?;;8Cf9F=Mn#+kB!Nbr!~C-gZ( z(uudz&XtGb%$Z{gRu+@NfwaHcb9As1k+-_tDc2anZ+5N_wGz*O-YxP8QF%U1$gb6~ zXg)4D>1_t}stBu|Y~Uk!HqlZ_^ycrZUOQ?7Wxh7+eq6@U2d0rmA;S6zd1n|s{f;&Q zOw&!d)|<$&F_R5OFojR!vup%O6x=@4E{%|q4;czaEZ1R1xi4L;IeiMqZD$SPq13Oj zIpCC>)ArmO^G7qS*K4wNqTO0O_gu)C8H`^>@A6m zoLC;<@O#@ODwKDuaRu|FrNUQ69cAMRgHi==w!J1|YaK`HhxyCt_lL6zh@pn)1#bvF zVH7vtlTaz|!-0jo6hA0TSOtN&Gz=s|ybaopX)j^3aU5P$8Bc;}uP_=b=;;dX~#GvOsh&_7whmDUMLr z86?KTSU1NVpel_Jn=b&wyGQ1L#wY9!rQj6=8NU9K7LWqeU)ck4gWyg1QrM*y|L377 zD76Eg@d|x5Elq!enM+|VEMCvPFG7sxD29YzUEo!FZ^5BfkmL!oeZ!7JAtu(|-agSI z^`iQCEUrESbHv4Q6CY9;+IghZ?M`L+u$i;l0#_E*ZG@t92cm?jPUtHSZs5DfCc001 zd=P+p3l%))A^hzPVsWG_VmUz7fRor{|h@RJ3Z( zGz}EYXz>3u36B*Ci0+q2?(hj$p7wWY)?A~+u3`l8(yyMNlkBiJJ%piX%|$w7vcszg zb#xN2gjO|5YK%fVdtE&}&3qw_RvUeA@THYuHXrtdM+E0If3VH*$YFj$PC4qwWlT!s z&*9-i#KI(BM;#W^am7y-prpm9q2dW7=FMUkE0=qZUwe%+_>Mx%P=RZ-;g_fa%&n)% zvm|khQqsrwx~lN~QGQMTv?4%l1BuHF2Nn(jzRDR%8YZy5)UAdbRrm(c+;^?txbUmm zW-WWPY3TT}+M47Xoa*`;&k|S`2qKtC;(V2RaC=)IAMr|Uxt6)_5zM{nRcV$zDNyyL zPWKgQdZ$5}4@3!6Me*ayenf2jOREsoo(S0$$kA@owsTT~O$60vzs%P4j= z4i?FvbY=pBsj1Kt2o0f_<|sx|qn$9KrqJ)daAXcv=>5xB_YV`Jjm2Dw-tGKrfS5Ue z9A^o+QOc)OwS(v6i}i;05X2wj*nV0*CbzBvSiX?_3eu)GqRv^w671!oU}cco^DON~ zXBuVmseIjO@T4-+^qMnVN)uywZ&|NJj2ZjvV|~C9KmB()PVOcC2r#>eikktMjE?py zRx6^qJ8}2eQ=A6#rn=QAMi=JU=R^2BlIHJ_w-qF!w3<{DCn6yhp>X3cmT@R>+FEE@ z)0%M#(~_7OUrwL<*gZljak-WI8Lf^p+M)zPhGXlb-Y zYtZx$Zqz9MZso|?2e|}?{><69K7t;X`|oTU74*Iq@%D-);|cZusWR5jgJ)s3Q>=tp zWwg81e6S6*x4ymP8-~X)3BlD$1$Ar^rwuRhq$Lh;+Rsi`tknp2b4y#GR=22~pWhL~ zi9Fx#kt`hhm%YDfv!>b2@poI~qKQEP7*o%%U178 z$Fm~wO~JgmUkFT}zeoAO&LongQ!4z~C`JKXo(is>N=Pzi_%fjo?onW+NCg9Wc4F&+ zwpukCBT_AH_muEoq5`+|it6Sh$T-bJ_=P^su=Fxk zX{b7q5Jar%yq-_jWN;(tqUl;|DWHf6cFg~(>=#Is`xmwjsgZcg_3sdf?qsR`UCIn= zLvLbUHqGfqME5K$XD*L3gw`zDjY>^)Yp58KXi|xQ=pcVxW4Inn)3$Pn@UfYnaRl*; z7w4iWwAO6bNSyI~!5w^kq`ZtKbIgNA4|cJvDmQK_jed5 z(Yyt}jlGFl5IrmYmpZznDPVYoSDT7B*>USujI{T8yls^2)UN2JaWWP@9!iUiy zeIxvJr;I&YrfUXCIt?*{Y?v%q5;OqO$#gMBJ(&Hf2A_;^AVw;kLy&^X6@>b~UGyLk zabR+?(21S_{F77KXK%l9?U_t=0@l_`@)Wc`QL+Q36?i7h4qcV{kOX^b&O%(0qonCw zpb^hEXd)a=pJAoEeca8&h1}4WKpmUa@W6OTfSJ(GE|3)Z5$Jvgx|5-|9hXfsRbsdO z2mDRof(JcAmhXcEJTixJ;a6hvKO(~jMk$h_7~H`9*&vQ~RCE7a>LL+nBSv?*WkJ|m zF;VLkyGp=rzUa=o9thHfW0bZ}S1nmlL@M0{`0$(sCsW=)j|7*^i%<4&^ZOzQEq6RW zgRw~HE9~@bBD)(=OrW|e;;lL-sCtWES*LH>o21X zH$gZW-*rE|?io082G1V$Mo-zlOGFhds+BZMir=AcX$*Q1(#{lU*9hFmTIkmyLCE3# zXvFfL*fL}iD?u(E0{$YnDtdjz&$`1&+OJnqxtIS?!|%u`CC_^9w|nH$Qi|Q4p`(Mn ztCI`yh(y}4C`Jmm>(~z{-@SU#-@;j%iDb&>i5*kO9qnzNo-X_y?*tI5b68`j$5FwK zZWAP~yF8L?XGSQ0ZK-`?TdK04Gk?>gUidz3ZcachtPVQS37flyE}V+W?7ocYl3R|m z$l2!%anu6l?prQ;%>Gc3bx^rBaQTeZwbOO9ze2B3v(++OcrPcRKguXyVJRqthDZkg zYRlQg|J;@A2csY~g3KlK9}M3{`-_^GZ4={M(J(mnCFSD$e!~A}MNP9vDe;|lVbB!& z>R#tboie{WbBawi!)e0K{GN4;`9*yA-)lS7J%m0?9IeVlW;#plThyUk_twF>srwSOe|7ElNz{#GEpj0xF#i@5l1c$m&K>M3B2u^}m(=XpW2ZbTmrY;fN!j zO&e(>%pyr}ow>n3yI=$lV|XQ#&PNp9N*9^+m*ZlQ5*8M0j5xR{t|#pqDVcXFZF<27 z=_bpsh%CHUi<0V%9o~^Oc#?hNe+i-AoQY;be9il7eNx>_yP^qqE;1{wczkV^AFHF^ z8Tp;!Z$+t*AGmML`Y+2u<(aRlVt5|SR3;k*;dgA3733{Zca;es^AJi&NfV#&QLoJu ziVjAUrxblSj}uN+dJ<9ORV!N)|HaFKEQTEfM2Ccsl40Tj|4f*_|?5*qXn0!$NEkf3G{?9Va9Nh0e}7oqX-Q0hmbzt5S_qEg5XkQVG~0u(2e;8M~RtI zl690{&$&0spXSjg_9gG`1oFMOZUacRTELa@jNuKF!Njm9wVlE1aZaU1Wg2y<0%;fk zu2mLj6ld<~9BP}xTPb?aUGbhC-KzHM#_^#-oQJmhK54{U^+aw6iw#hyKnh-4Xp!yT z0H{e=^Pp~qds}23!H=CVNeO=!?;>KVOLD=0(05g+zIIRXT2S69ONITfnD8OuZr-2U zL+lU62y4z~&`XRCUxzEyIakNgfpCiIn4`i9FTo?W0_u~N_80?+Y!Ol7ph##W&MIah z(~Q%L?FuIFia zE^yKlEG21C7&s#U?4RTiH@vmeZ!G?&n4ZOG;B$@PEDH+MV(@U>&NQc;71bMK`kxcN zR~-xjs5ZY9k3zrwZhSBXXrFfXbcYIDbe5ZyDUtHdz#dnG{H{YJp&OPLIA)_QqClZk>?)Wk2r6?AcKZG5+FsQ3jTXM#Qul zT%+Op9{Bk26z`JW0pY@4Sr{`n3+-APRhjQupopXiYIal3-lTbg3o^|Z`=9Gm)3LiT znwd1jXIFK{#Z;4M%UqmzOS%1JrqRkb7>%lG8e_&{Y;Q7Au{m|x6W zO<=gx<`>IA>1)zTsLz{GMxN_Wt}RyC`1`@2UOFG4>TXuz zq^3^#XJa^uH$1teND1%KiWBL(qdRcGsU08hw^)hAkU$rmbn%A9YC-dN{m_VwlHVAH z>6?enq<2KFO}uaA7DhSroLXoPyjX`vGLa+u2FvWs3jPQhK%EbB>#`ytOrC#mtW1{f z&S8A?%~|G*1u}F^M#0}#fJnViq(C=#Zt#X83hyyU^WO<+KK8|W(O^me!ZcUZKg3ls z$%kAP8RU9VyM0)^v{>l%#NPW^|f^z;tGsEt> zGjq;e$$Lgh&2PB6*q63p3R`z#_jiAME z1pRGC@dfzKspXd<+1A(pLaib=GgFj&Q&yCeXiDMzVn#%t4>fs--y|X_GrN)cwR`1{N#Lvot#Hy={nieC z6CBG=;=jsM6)|F^rx7#`l5wpWArv$h5o){g0Pg)(wC$}b^ad|~@&5Luc?Mvx))CxTmvSn2Qbo834{ z2+h7fG4$DR`WH2L+;{rO#Wlj>?UTP3o(zx#3Oa)?-<}Lh{2C~w&>wMiJYli(RhmK` zZj2s2r@#Mrq=mui75cVb@Q6-t?ee-z9H6LLk)xkl#dHYBpJ4v5h-zL^8eX!<#|DNv z#0zJ6HbKxao(3|LEw61a04wsV12cl5_M4(nS^{{&=<;6>EHF|WK2$Xn34b7ruuz%;kEw_l|-J)Of=o~0P=%9mYkB{Z(5Ij@M>Z$0<}WU#4Q6xlaWujw#04ZX-N zMw)h1WH+jAP&iv~alP`m$M4PAA~m#54KFI33WGV&--E!cvR{MF6ylR|%(v&Sz}_K@ z0%m2wanM9EA+U@kCg0vNAxt}*g!OI^ra(_s(`3hn!Bsz7L11wd9HSQPg^Q~F5FX?| zx6_1-N6)*F^=|MACYxd3CVQR?qE}P2=2EER$mYWosfmaD84N|t44piVT zq*j1=^^ZlCkBdh~U8;M9)TM!&#AFWd!CuKzIGJ|K-dBU(k0x2J;Xy-DEnIaptq#n; zVs8k8N9`tc-`ZN?zB}UuxW-;`H!^wV{%FBq&h}-JR9I;;sX)9nD{&|R&9(0FP zHq|rb9Ga5f!G}!_eAz!va(XGi@65sQ=+D(A_q8`HrneLv!#(#`qfMVFV(M~;ov#72 zpMjOPh?f$c`s2HYe=WztRZ67j5YE_gktby)&y673U~6Czv9yWd!rF*;4!TqS@huw2 zh<~7TQIL3L_lm4va@-f+E|K?DyudY3;a&hJs7jzxlc5?7l=0K^Six3SGJgY5uKl=a z`pn#@qkqh0<$FU35KxF(=#cMR`$Z5b%oEuD1*p^w4EVr=@~%BI>HakQ_RGNBH>l~< zPrq*W`;W7esgW(r&&l4{3Wk}Jh55h16dNNu)Bg>oejLDV>zwtQUb#zNu-62v6vkl; zAfwZc8~7BCdv%-Hr>T?aA~I2f&7xM6)Y!+*Pi7eQY<`9nFhEkP&fNCgr8Ab1q);Hi z6*?{SaXo}3&;qizJvVO6{E~>6kbVL?&d-b=>3$fg65B+reB7J(4NEagXB1+|HG2AG%x+>1!kg0lpR11`FVgLamh!|>ENqoebvn%t zMbiegC{`X1YBU(Goy}1OnxLXr(OpYX>4FbKRnm*7+%j^q1}PZ*`#zs97!u2q67jbCz}~!x;9#lqc_PY+hHpZK4bnuj~I4wa=AGMg~rG8g@FBX2O^R0 zRBDuGc5CIIEq4u1-ppR=I}CI^GE__ZT_S$@L#4PwU43V^V9BNoC{*CHOEzj#J32FR z15@PP{!pcB4yX11@E!q)I6bbDUdJkW00 zCXZI3K1V0a4_MwO7ENWD^?KDU?ug`uEKhZIZxF|oofYF&Li;(x@o#N7=a3%-qZ*ee z-2-{%hY(M9mahNkJt;zcL%e+Gs>b=ysR72JN68n;LlpQ1=5n!Xvb2olLzUXLBzm6! zke=aScHyR(Gzdy5^&wB5y-+dj5O0T#mRZIZ&ZrryoLL zq_sJ)H6-rnBpBzH=RCL@%eqpT&WR9R&E8UIu-ZDDP+Lu{A&rlr^2b7=Vu|3FOCm+h)ewy&5`Bo8&})y#JA-u_%e z719X985X*kM@W3Q>)mS+AHLbf@nzQc!?WxZ4IPr{XC6BXyhMd3BQ>gpM0Xi)C$q z5L_Y6KCF>V8O0LRIN?Uid~&xyv);Sm@f3btAlyJOlNiC5(EkIz>BnwjYHuPMa zqdGJZ#`h-YJ?Rb3%exdbBZX<5L@~PL90%+h*?&&OXc}INqlUL>r|_%NMp?!@lbJsA zoCx~(69ePNo#Non83i-|r^3dJ`BNG?ZU2!Ch%RT(Df1_cr~?F`(fq~U6D7CMeRdQ^ z%@K;xNFSsT!S1=rcE)3tWzKW1q3X;g9?g>-xZVw6f8h}hWlu;z8m9jYL3~`c#N{}KZYK_Uw5#Jx{3V{l!!RZ{PUW�|{tLw#Q(bdU=TSL5ssn9Q zMg}%K(Rn9SXrZ;0Tv=dk?Wh&FbktqoKR?CQ{87m7*$1NodD)NpJy~I#u zJ>u1wo1nvvA?9lVIulRaXeqd)9B@1rYwp}Q33F^t6Mo%uTTQlB;=ycFl5q+ZkOyPT zJ*awtc!Mf9qWmm%LEL}wt_!W3oM@BnMZl7Pq1J2M{);=;bqvkUcX|IQ>Tjf}7yoV_ zPhaeHSZY-PS@BBjm$YP&i1OLd12N4F`GzR3E5^->>RjIzV~sPgvMvnfvAmr=t3RvF znU@O`3hb{Xy-fv9shMC2TLgL)<5TSEQ{EjEoa5h}XTdF~bck%2;lMQ~k@}J};c>2{ zc7sHAUv5(ViPV!fOM4`3Q7nLP#V4!$u}B6d)l@M5D2>Dup<#J-EkhIrS_u=T$-Y({fBCscI8Bz%OtF^94oI$gRICQS~VZ%=k>dI|cDRH{emy zoLRC{dhqby=k8I zesM^EV)Q0}1m|d~1tj|g9bD*f6?)D9YBp>LpK+=YSrXjIIyQ(Wj$(RSX7GlZee!4r zJzN3qcSkRm*yr*hk+w+L0^X!Chrs5YO)(x3V4W^tRbcGC(RUxs zzC9-E%fL&(2=v;A-_p?Np|Ei9t5LkV3C9=x{aFu8xTX#`2J0tjhlFQRb&pK{Ef_Yg z7|&SDWO~OZ8t2Sag|C}91WeKYQ6M3n6-Z|JE%}QK^0kh92$Yk)b}v@WalL1MTO!vXm6DM44f+V1tslL?KHJl+K^(<=tPFTwEzwmCDjE<8}QFngne}G{_$an zIRdleg&*i-nfea{#6;)>v{AbG$)>4)>Yp7HO)bhbs|Zq0oQ`!C`wmgtNENYu`SGEI z1HLs=#peUQ_D{ofq<1D(k8Mu%mYmJmzj0*cwh{5FP5BIYF^x*S;b1DxgX)1s8UBDc zuQ9p@FYT$i_O{DwZ^Mz1YkA8z$N3tm-;2E|U&jO*AoYfmiqWt^@7<$iVEg_x`w`-8 zF>5T9gl&yp8Fjf837$mJ7#~$O`y1Ah;o_T_g(u31AHSGC26WW1q~Nx|Yg26)9q5#_I}+ zwRQIGDZ(8#%F$N+v+I_>Knj`OUuc#vh+@~OMnZ^$0~7?%d|zst7@9nvl50H$#su>( z-J|$h+f{cM4eyl1Db!3#e;%=~c*t$4*ug%EgYdDh3v5z5&VQ@d+l1wyC;75?%L~w%e4^?ugw;f$R23w1$$RvdxZ>gfm~Sc z?8HGl^=wJV3w(OBX723v&;DI=V^@($>F2XEj`Q#^yK^L8Kv@awfZum-x=4K4zawUG zn6lAuY;sseTH{j!z|DZjVczu;(5#Bli|a|Sj_JIOSt&qQC+hQ`vK-rT6GbBVV0G!v zUAiAV(ZzQgw#_!rHuHLZaM^xU=j}7wr5>(C<8CLFtA2b#D>Y-U$NkJKcGc6p@v?7t zwje!)-`!cNXlOwN4L6z#u+LtG0(VJx<+clE^92_WMV0S!OvqfY66g#4y)n9k#j4u; zH@{5(GO<=})v{QiJ-EF#e|MsG_1A-M4}`Z;9<&9iK0JoLWyjq;QViQ`wI1T&G_i?z zv|BHH+rM)i-*@+?FU>SWWq4-WnudNE z&VnMV9HT5CFfTm6Ltx{%qM)t^_Romz+{(6&13k-zheXQp#0(S>i|@o>O{ESY#`FCO z6uBMve2T80f3Tz@)Q1F;8^s!(@COkvu zk!&?~*f{&-5)F>!bcTsJ?_FF}({~A89QJ+L5If+IaIeQbfE+ZZfWPUhMOcMj;jh{W zqfJaf)JVtesT3d6>)fulf_?0jl|F9SJ;|T*#xMmEelP5X;11Cx^TzV&hlzjs+w0wm2jlN=Vc`r_>)y1I$BbXOMOMsKPsU@ zTP2)PE)=nZ(jvjKAp`Ak@Io7qVc%gcE}6=&gYvE+Siobn^QXu);xnM1Gigh26*%ot zMx%fBBt!?oQ}zunXTL+xGmyZLByhEPf*bn`r?2Pmx0rm+j9J|7UFT1I#wz zD@h_1bEq0P0lYkOx-4(~GM=es6)5PON48Wz@_CVh)VZ2O*F}u#D1sFY;eM@YR7#_e zGIveXY*j;bl6EY~;V37H%2H>IrVGW*)~L9nEsB{t2)@ZrnLN&K;ZGEzY(YT!tpu}L z_J~6jsmB9~Gb)P{wd`uwLXM!miSN&EwAcO+d@QfMT}!k8^dUUvQ$Er6MRO%{vn)wP zumvXTqar0;*h+Ii@2`Z9Fpax(&gh0-xCqXcU6Ix_&cHKKmQ z@>z2%yozhl?!o3vg_T5F@#UR8TuIO~-hSPjBlGG(pBpb=1$&o%a$PA0u`0c+dpdi9%>TjFhVRFl9LMHb98Sc$9m3S3 z-tmcf^;|&iU_{=%c@nw0NI-dqdbtu?faHp4g*+iB1q2?*T6|bNgG=*(!HAkOzEZ3e z%c`1;H**YRtbDD==)@YLCPFk z=XNHpJkn4W3r_CfHh`1!l7{zi(5rJ31l(R^-Z4?3A3O+SUDt}g9jFvO6@p8=A(QK$ zdasZE$5#c4gQiRNVHSel!EG*g@)ru9B6ioTmwMwwh8dXNbc+#+AdCen`$vh$lKw-K zRY;rtRL4K;nqaa;#lq3H(ktj;d_U6O$N=t<20s2RAu5gSXp{XqFiU^y1+9YZ!xedt zU(un$*uBoziK5-ujX$~0FOKPNpe1z_4Lvx>j<@?;Cn7bY=MsNZL$q^{8wVHt3g9g;ODm4{MH9xnr8dwJmGL$sCMORj0v zLmJl{y?J%W7dCa*9Z7}sqqzNld*e%Nw1oUkUvbFTh#iK*l1%q899{cGQXRTIIK^<&(rsjRw< z9biN*UH1(BM=YmHp$sS2BpE5Tu85Py-u#<`V*F8cFGYH={W}Kb`de#r>x+e<{+C6n z)Y%+bl9Inmv(92HI4#zmqm5X*?YPU@!JqczTBU_0IjzGfr_C*x=(N(-Ab4KzBDUsolXD>msf8rYkn=dRwo*N6VC0f9z|NWYeYU^BX903W3`kdkz(NzJ?*$mju{F@I^5lMzdBF!=SIy zqv>rK@^c0tLF9X6Guz)O-a69dLoCJZ%l(octkL)+R2@&SvB*XR%y!FU5vv3+1S9N5 z9DSgs>CRPkiD!=O@vgDBl7>_<`%;5I%3JUlre7$8NmvV!Ym%55*soxJJ!o$n+np|m zh~>zJU;YB+h-||lHUGlKd!<6yIM0N+GTL1%6nZ4;-6RSZmbx^b_iw1$2)=y3_>$3* z^XSA{F7REY&q*kB8w<5rL^VmCW-L4m9VcAqbGsMWH@7>2J zrcHP9Ba`H(rB@CCP2P1eFRrfF)p zV2n9M)%x1kj6Ng@ZC#DmJ@Rvcep3Z(q+ve;BYmG5@y-tSNB<7=d?)E)tPcav%FCTr z+*P-IcgEJ{oI;^7QAUgFs^t1e*bqP2sHgKY6pdM}qa?9Jy z6W|`Fply!dIhgc*64j*PrxOsQfTsCirH_?oosWL6!m-l zJO<>`T0GwM#LGQ1c#wHk{K23SmLRkmpXYF=Jztn%GF)J&iw8ImGGDDXWJCJrfoI=z zaQ=nXurC>Q}@r_g~?079l6+HaUqv zq!^@`^|g71@UvMZEwOEoCmJ+eq2^mfYAl!BG~x*W&a_vFhzd4Jx-_suS{^7(r6jqh zP7>;gE+aOcD8uY+^!uEcq8Cp4fS6%P-3bIiQr*2RjS--E5}<68jV}BvZ(cAh^GYX? zNa^lKJKXIEjEk}==JsFk2a*Q~X>*!o&RyHdbr5x~+1b+u3HIDqXK}vT7@Gnry?|Rv z?@gP9%qpKC)92eLhCxTiRE@v_8b;B6L5s9we#DH%tS8~dv+iF*VxU)L;e_qKI@&rsE_K$jvO;xu+i^! zzdf8t-RVq@N3p`IJ8FRg0+00#Rc2*^3t)A87-A=ghaItb2*u(ciRz)@Wh4z6m?D~x ziN%xDyIJo{4%lH2;FF>uix-?pQ#RU+qO%k|=3t2#BA2QiQOrs!buH8^Ym9*ZVseyW79l<(hwPa}X!jKs>!336V5@bl1h zPGhB$oR$4kv#w8<*=#UOKWo=UHac~&;B^&CEV9r}6AMQ_(KJew^goQ9QauOywr$(!vTfV8 z&913C^S#YoYvv_0f5JJD5qo3&4T?r2c!DE98fxQn;gYhVL3PB^t?XE*bDY@DhFb6N z*!Mfk!i*5d!cXz(8aA`plJtnjf)Y>&h}R1}H2qmq5S=3i#hV0b?6%$zIPcAN2b_a7 znh|95u~udMB4(0LPB!Ig5m*S~xE7`rR&9tzEz@YSE(kHvUb;pm@n*66;YJI~nx-Nj zgI+0#ggi}VUgFFWkc^*nL?EVg5Bof5Z`lalRi^dv_+b=dh`UZBl9)=yR4uMHxrD~- zq%J5Ydsd?Ql*znnrw3KY1!P3{$AGBszd0uZ!$TX}Uc0fJ+o#MGyGdFlM3|i7& zjG*2~8PNM8MKZpOQiVPFB3YO+x2vob=r=wlyH@fn!9Y;g?>~)4A0?*iy5gm1~Gw%lgYGuIT1ieN1JW$*Ss_d0)8vEYg4TX{oU^ z_V_FBPn6m}XlNWWo8P6)kJ&|am!s_Xq)hIvw$Am+63iG^{_sARdSPv{?X!KK`5{OJ z1;m#j&NFpl@rs7vl~8+An(QbZ99M7{ibULEVit<~)LYTWS@JFRv4Gy$1L*^KPeR*x z6N#jgH!gh?t#^ic%$i7_F2H&XeKO{&bJPQQLfKHO$mhiK{Nn!@d2KnS_(*7%IbCHS zR`l7^;1X#bF}6;jhxbgiM~%z`tJ>6uG+Ku5gPlp0sMMV{)>nwYLQM2H5)r&=h_nH^ zsNQN?0+i~lfI5UF5xK=kuUR#Vmo6sEgjhJ}6Eft{^bARgN@)%ozJ}n;M*v;7DLtHR zM>Ja`T)h`Y@D;^LBAum-4!sV?Y+wb^?T?IM#{l9%$*y6Fmw!$B^F^E6`Uvw25F@7AvKMDnuD7v}56apS!5 zx_l+c72oYZm7&>hw+!{zGKKBa?VqPtEZT8f45G>lhKH*fsph9p7Gdv
N?(sc~| z$Sq-V7WfUW*oZ5id~VN;PtWtMTx(<(2FX>L?nenNZ&+8Fx`|6W?pvu)rC0!UfeCsV zPSHk#-9jd5)2mQJ!G`+&sSzWKZ}AxP_=9oX3>`my_^j5O9ou)ET{&r~WsJ8FLA!@~WaH^p|t zN~`v>t~$CV<(6Z?D42F%k^qJPm%$MQy=b*mUQVAzu()U#$*DFle9}~%7aeZ0_|@}r z1#YT5(<;~EJsxjYMU(Ruud;0zzOXbZfVgArAtSD}?`?&?JID zqwLYv_&(dtll}xB!EWoAt%(8NX*Cada(qlx0}08s;XP+>d17?Y&8&h6(?9LN8$!r?SOR>d@!lH z1PD`mI<7mD4+izbmvu$ku5%IYy2h|viA>hi;8hUL&o(;}p?h8IP&r+x@z1&z-CxO= zih$H*;R^{rUuxoSU+m29EuKEx0-u{Yzl#wP9?m5VT9;%8LtvFAY%a2`$D!(FRo*I4 zsuJbW-((B-S>F!Km7x+0*Lz^Y^bD4sdA8qP3y%4wSbj z>Kp#OFvK|an2&p4=+(=-{L5$Q1oME*eA+3=p5@Zqg|L8ZtB!9;;TZ4nH$5o4>L#Ae z{QTx_bwGe^2e3fDY;~t`ygsx1SQwuFbw&CbuI)e_4MAJ__?Wh=>-K&bXn`SeZUjH& zh(3JVkM6Z%MRSeGY|y$ZvY$UwdClWO*lVX_t!rXn0{ysh5I*|VAWC3^B`70@EUi*< zQy~%5u+yzKIi0w=wrluv7JZ*s?fHEkT&<;jFwr>VGk(?5TQIItI5d)0TyQrYkeKJq z*}7_Y%c`xfelsD`hltAE@r0J1d-UUdIUs(k!T1tYfJZ<&{MN@Z< z%G}ia_yZ}^lW3VuZu}t#Su9#D%Nb(`h{dyRd+-3h!C~QXB>x9mU?%pzCIzS}V;@8C6b zz>qKJ44@FK^stFxX)W`DMnj>NSm;^3!L65lPN#b#b)i|ivV!Ivlze)0lB;6u%znWg z7|j(G`pvD=#XG8&KL=t9m>FMbuij;UYqoF=2PM_R~6x)K^+%#Jpn|L}n{|EQ> zs$~w){<%oLMe5JD>8mj{$a~b}p}cs&!ze-dLV�xwCbTl^5X2itTCD!ESoOg8bJi zKt&&!D@m$lc?mXvxfd=$U+M?GsSc7SpHf0&8_{r7=HXOdMC=z|0aK6X8CW47(2`VyCHY2HK}>Gd3yWbo2tm|pt^ty>xRtCP$ANAaSS=C(xO$paX!aRVZ z?X&)M6*EZWpXAFs$4!nW1gadrt4+sLi-Sg()4NCeASa zwp5zD63JAD8B!~BXPh_6z3SskT~2c5nTvItyunhYKjeZV+OaEPN#+&WOJkij?KGFz z$!P(SFxn5!_d~Ys`Dvcwz*>LJ%quLF&sxZ*n3m9cF4p#=e5XL&rHtqR~ zGcmvD${^0LR<6fX=lAiflUO2=Omo{p9Hg~|KAT2mbsebe;rc4orWy7(t)zgina!Ka zuVqJPgL8^w^3;8tU>kk?}ezSfJALcfXQ7O7wOBK)*oI;hwFvR^-m^Blq6O;qN0f}AqZqa33xP~~Dy$#iO=^j|wllHG`1o}BO=1zLYz@00E z#Rg1$#0i8-Kg~`Xu&8poqSm93JFCpU!mrx24|O{!&1lRYR~bq>YFZzF7@=~fL6jZ5 zHdBf7)^}#8q3;1p{0-M)mC=|y#AIIa&e%Z1Xi*Fuwcwv&k%I!FGR4}5m}xs}q2&de z4ITYX^)v6#V2B&Qj!cSt??#5w5ZFpM2Qn-K#1KC5Td1~c#ibG~bPC|7nJCnK3rP;7 zEHfrRZcb=GNG34vnzoC?B=;&HmpS*bnIl2PgLnc{(AY(>`2wT4{w&l+TV{=f*j}tJ z?2aJElKrkVU)ArpyE8bM^W0AkGcb&A1PFTFV>#6(S=cMEDKyAxmJAkrmSresl4B5L zP+xuXM;O2^`T-nCZ;cA|ZH#+I1Ggn>Km?KmQb#s%$w0p1q+Y{?00Jq2#S?3?5@XMZ_>z){kZ1i|Xnq2JM-<9UgFV}>v|c4Z zd$;MK=-Du?3Em$B6S! zaiJ0tkpqg1%>mq!=&Xh(pBEAQ6NxP}-V`m_1T>gjQ9d;}}t(9w(z+5=#`D%+LJzc%^F3aE+9~8_HqJ{C747AuQ-P zC?L~VABZ7aA0oKmN$``W=)jX`!3p9Sv)WYOWBPC0(`2bP=;mJjU6Ed*=ptDYpZ(wk z)xP=Kx;Z>OLc3%jVTfe%jz-I>zAM0fUn$Ng@EtcmvkvvA0?$^EY+*>TRMYW^Dni#@ zk$nGVz>a7;e$il&6s~6gh^*O)7`+4{2JG)Qngid1{~!Y!vjBov3+y4oO}tH=8454J z!(6>XXn*-P2V(RF{iQC8k+t~sHy8#f4tIJizPZG?3 zi+tt`y>e&XJm|UIrPQDct1AtkMsB@{8ggydK7T0`}<(VNmFzA^K-(31U)!*-@=) zW6USKfDrD`7g91%2mmw~ac2Or!3&z57&@#ER|bX#*#!zr2VM0SU!n|D-=>zojxVn_ zSjJY=_E7g$Ml8XPm#aJ782z$x;-kc<?BhIhaAX}Y=MaXb$i!K|O2#w~t_!2KS=jH8*qM(~kN2=xjv@0vU?DK=4jF%6G-qCHlqOIc#Q#RVfQ;LlS=t5T zCm`pm1vpmV^>(ie<$^V!!G=`ahbt7^o+Zk3G46c_pOOy2Pq){uY;gyB57QdPA_<2n zoc|J;{w;993}Gxvuy5~9@H|W=(E9lpoI?yg@0p!$DO=w;I7-=`estLV1J8q_EJYk2 zZ58*~jcYj)CIKcM(~MI+Bs=fOZsNM39(WREh~|V@5 z@VT&kYU6MupLAZtFSzYYwSQ#UB|3+3kMc`;o2I(~);uF215W%dlW(sEGg?!K6g-JH zRM}0=)k?TQt$L6}uL&7WfDjORj_VaJ67*EiZq*XmWXKtD%)k*xza}e`W@Adv`y-OC zoWHKSzR6egxv=3>Sbm zUrP_A5)VPDd#t`3^vXj#Wz+>9MUW9pnF5YM2qp86qeOx1Sn`Y;k|Y*v*>xtzz-giZ zR^MUa?x_1iz|(Vcy*cE)1w{1w6H>+u688A9Q}u*lfm5I=rQXLwvw{bfpC)~4RxK@I zEXd?R!~~wk*X9QpP_~&)c#%!=#o-xrgPi=3E^k}SWA%jQUprx#(o80G_G%2n`u;|L zuT0#vC%2}c>#_7rM@2SsOFnL&uqR&W@izIyHyqB@((*J60}7|;bAW#U%3iz8?EAP_ zKV^-&tOj`Dd*5dH!;_|;tYy|5B^E+8v>9;TbchyuaNkFiHc)~^_o}ot8SvW1_{@4Y zQ2c3(S`v`x1kDS(7&*!bQjmbvgEZjpBRH?6nF+Fl<0LcuF zz+Z3~MH#l+L`m*Qo-kZ7_y6@~l86bMir;(|Y%ICU^vLp1W9_luA1XhR@kVgc1t7Xz z&mcYXIxV91p;>c4+9?k8A%HS7D2utZ-p`BbVd|x_?1fu1*bB4>rMMY4e40$Nyaw`K zfKz|xwZ7_b%LJ?Nc)U~qxWl{o`S;DlgQ0U^2=|1P&>#6kmZ$b4m_nd)-7c@@>2 z^(#i18pjprTeHCVT-tP9+*VXmH1Rz^?=)T$++TmpjrtPudUf%(*9!xh>>2lwJUM*e zvV~qdPGZX})X!iX6FHXvv6^+BXR;?=MUTT1b==a)Br>s)Mv`B%!JJ?K8sjU5>ICA0 z@5eqv5%aSY=_nu(cQO2RcpV&Xq#Y?JL@MjVdNpZPALz@0_TFu4=mPW9M0G&x@=cs7 z3Y@fi<29k!jG*!kc%kpZC8{q(+B@Q-WCW8-D1-**^#GQYRS)*6U5)bP*=^{J2Kf%* zPrwvY*&2%LnyVH7$lHfw;+E0CEZC<0)7_uD*;48At$T^7i#q5>uWrC=$k0m_-8{K> zu~#hciTWV1t1>;*1A(Yv47bn(71q=pZK`ZHMl!blO$Birn=)V*n~ z)k{=bGlWQL0O?u$q{?hKeE$xXMfi9!iIY2Hptg8T4;jfwpUM?7BF(hegVxz7Jk4w@ z&`gOMIBgp;3oa?BOh^(M_(yfe-7o(`odQ`KCg;wsDgc_n&Z;MXK0GfYcuSU}lU{i> z%!>VD_xcZ+>%g;I7^j`xL(A2iOa+K#H|k;yq%ef+EFie2Qwty96lXM7L=M4cBCoRV+~D2$2igHe22%cK4t-Wsa_rA};#%0A0PgH*ozG+F4Z)tG z8!lkrC9?Qkps-|+E}r03+kt*a-RJjcXliS2d_z)OL-e+hB}Amd*5-gleSv{}bcQ&( z=V|lZM#YZt4`>APQG?KC#e|F@p;sgrvmfwNiP#8(qVm$(KK=mQi1)Zo^n0j8+QqVY zaW8$KC5vOQ#Ecb#I!1i`R2!|}I0imXAq?H#XCI+1I*$y817Cw8}@6+xakb_dvW47=!Pc<~Wf6{8h)84FpiY9p@YPXz43 zBO@0;{~^xjv(llYPXWUvKY_YAQw{u`a|PyF`i zzM0t>ew|$TXOul)BR5YddkOocvTitT+wZ()j|g6dQ|j}dkhU8oM->PZ7iHSKcBSAn z9KA|RpH9FV_SA-ziX zr9*}n&ns!)gEhUY!&GsY+dAvbDD#7zP4m_H8|Ph=(#4(;_GgOgDHB%7WqH}d8!3|2 z3~`xG@|h;LLa<2jw7ojhZx~di%%~7$3s{Vbcc{QHA#5FUpN)3yTqA<;4X+_{>5*kG z47SO=v%N<(vknv%bUi1c5!KwXlF_58>jTU<35wD$ZPRs>E)!r#SG}PqO&Z6)9oA?) zg>eX`CRMGiqs3KsLJbKP^ug_{F3&e9MfJ&OB~lu>mvi?+=nu^jW6@PoS88(-nq6|$ zp;~;0>MrP)AXI!&H>;bluih?=eIkG?G%qL?pa{I9ZIkFH_d4o@agM!*t2e_`4Jq_m zxIw42iR125=p}VI69i;7W}F?&yhY8(GtvxCF!%cNlYpE-_;8bNJ`P4v%#V& zKWTBnstcX!(ACP;o`;J|J;?@!5=7nCpgS&Ccs(<)d^78Ksh__8(1r-388u^dykx0~ zi?f@wZs?LnnfjFDbly%yYaQx+=Ee7NA{NY|gsDuJlITw@M!nGJqNj=)IlJ)}VEFcrxhK2J_ z75m`+9OdEHhlA7extpx61j_{BG`bBX8UjjHkoSv2dtaVkgszd(K5$IQ)Ar$c;5~|PiExz4L1~w6W+Ye3BEG*kSe{fXYpTMZ ztH5}kXKG-hpS!j6Aj$m%K00!D1?QI6-A9T zLn1~^Eq}r#seBIO#~GzpA29#oFT(L0BA?EBwzg7K%;+HcG3nptqIAi7%fxlorKeaH#Nm+unxt{H7WJ)e|MN?Ko z<)E0hdTdyV<5?``ok_-T9)4er`9SGe)t3n$JQ~B>upYiCpI{Hy!2Jk(0f#xZ+|S?g_inNGvTs`{OM5P zMF$1B2LX8wMwU1yX_JTT8_p1QCcvg_0d#Mc#vuKsV^ji~$y+dSyLm_Li%_JI2O;VT z{5W`d$qRT}t#2jw4-9HR!bY~)S>|@x2R}Cx86+g7BJKFy9T80j#-{%ek_#hF?Erxk zWrHrO&GxSK1?b4Q(tTqlCX!eI$V>s*alaMX(lTrhCE#W$=dUeSqgCg2lS7>7tjbmE zbeq!$cHGt#dKEK*W+qIX8=SmU_l`8zt8C zy9tcxQneel(x0rxCjui``DW31?`$MhK<6=QoF~|*K};o&uhuiS8owe(ra&0auhG&v zu78p%16bnxDMKynt=teKT>))$=zn*;5r&cje+3{^i^!WDv}!C#p#@16V_SHW1Ejvf z#nx&58#Uq#Rm=@o2-TJzIL<1X-g?WEmbd^S3vv(wUF(U6B`0&%etLhg)#DrwE~iY5 zpE2pFle}G>c_hzlk&7HdZr0)&3-|-7aRaWZrb~-O7n1-Zb|9~8ROOFd=3svV?C33&QA`vz?Em1E^v5eeT!@c2J`3(FZkvqBwcx>cb4K*LNmq`X?eR zQ=YS+k!V6@A3(CUxyJNBGRu@9y5F^pfH!Aat)7x3j26*&9uNSQGJg{^hi^KYOzbqp zgUL~vCv~R6hp9`|=c7WGHt3L*uPp~ZaePCmX4-ulJz@%WLt(lvI7v6#6v!7cXlRMK9AZ8DY7IZ`6p@ zB8j*t!#)r#=VMa32#}<)beNL5B>U_vSY$!q)&V0am|Mczj>^MyQ;U;#_8R?NA0MYS z-p50Eo35*?$G=-e^%9(xPoKY5R?d0|f^;;ufi8-uvC~Pv&w9DEh{p~a1Fgw@Cj}9S zD%#gdITBSBJc{-LtiOlErnM4Y!%=)t)1>WO3gO*gJLGu!tQt34bpj=hUOg@?&=*EI z{)tUeeK$O-MAsBWNc-jZ^Mf*7-OuS!#3Vw)WLCohZEYG3YKW40h&su+o&Lo9`7sb! z+H4`K`Ivb1Q3E*WE#UyLUH-3Tq5&f)h~-T%C#(elj>Q3WS=bvah>GrgQA0mudY+2( zR4d*V`W{J3c8K<>W%r8L_*I^pDSSg|O za)SYJD^C(k#!M1w9KQ@SJlerLfMiPP>dXOF)`iN&U?AGraotBdR}BS9eBr_69e+TY z!!&=ho+xN3NqGCc*p;ZW za-)S@v&l3~01mKva=01pY8PdFQ(0r1@9sh1FZTQ&bx>heoc zoJxJYsb*W~4T6T*EJI_HmA zFvCUZ@ca9@2?k>4hy~0BI*S#PkBzKH01t)4*eE4?>hrJuyO7L7%2tJ^&cM)TePxRuUxH^Adg`Y~qb|JD$!qTK zZPFX;#_F5B`hI*SUA6z#xRC|wu*2#OY|ihxXpk6ungpb9uSWX?_7jqP{3q2bV&;$n z`~fN=%vyvU9av{?3o1!Cy*e+RNjt!%G8%_R6CqkM?>Zpl=+jzWd?%GM-d)4x*)l4` z@y|-xldg^JYO>}d|1*)kMll70$Qb>{MTf0}3c@Toap4^`DKg{;An_l`_Puo8|pW;kF^@nj&5sVGC4 zeGg|D@DtecPzqZy$wOF_=hM+2u$yuL%x~30hvI)K{frDLLfj}vFeGB3zz~szyWk^q zaEW!FI*n6(H$AuZt`Dm#$JV~Q>{NMjk9dGc!^L=CrdNpvJJbA5hV@(0X^C_K6GD;| z=|p814N|bKJhGpTj9$3hVnWU*|tCyqkWjgZqlFlKF`^TD|zSaWssc? zli-pExQUnTL@Z^)_X9eh;{etguY0HS)KvPrbhAO4qtJT(*64ukK~s z*Uqzws{A}=4Y!(HQi;g9qK#8fPgVUE-Dw8vC+WN!0<WouPx3>CpSM;Xjn$1fOynk}OpBsr}XIroW0VY|zR04v?Tr~MX-3J0`m9Xb_*}J-n zX6z{x+!N#C{Clj_dF*#|4>K6cTz{l8jN7SoKnuOvcnkLu8BvMs7pvIo#*K?e&8TgS z!TJ5e$?4Z0J8%hj1&s83kE3xT`kfPZ6n^5_E20~FW=tG;m6HkZ9O3(QdBGA2|5>qC#5N{PP?WOo0!b)5*h z|FQYGd_VfB?&499phh&W>)YQ+M)j4^S}_9o=rXmu4tLrbDyiaQhkH6{(S4|IB^j;` zPfX(!4nS^?zoiA>3N*C{^HLTI6Ic0XeQMKF&&mr}Mcgpxg#4mSxZ+{0yJ{wCG25U6 zwrti!kfG0G{jWn!VA*+(tCM9KpBgfewCCPBCJOjViFzNRFCM5dMoS;Vp$*=o)J7Qc z=5d2vrddr~E0bcCEqnTy9J4V)M8i#tZ*oICd7S@ra=QRV3>VM8agxzyv~%QNulDPs z&b{KN#S|%}P0LF>HmdB*j(o?fY^O1;7m=8fOP8kn!)22O)YmM(Z~jzvB42Y7q~{j` zWXxi6OBDAj@9_=5C;60(Xmgnz;H2QZfL@%A3SiDl;Ib8Vq&v%3*@!_^sEmg1-ma-4 zpRIDy@p+0FlQ}#7>bNRrK)_yc-Uhf=c|Pj;s@R!YTY@Ha%~}Jasd`zpgfa&RC1Ewy znMC+`l-KtoOdS-F2J-%s)D|NA2%UU8zC5*HiU2Jn+0UA*-B-R&4dB^xr-`EVk>nO6 zC`&~H;Y9}z(J7*xwoyTH_J*{0sWT3W2}h-yeBE(w=n+Olg1?JVxe3?rI80ILCM(t3|N27^ML zy6<*L7za&z+HBc&tOQhl$)}A1A^dYWm2jg>bY7}keaPc{Td*FoDZj$qA}*cVm(c8v zBH_r3ed29L)3ex28Uda)|d0tywx-^RR+xdHptYr=P zTRJd2pUJ!HYM%rBFv#|>O5dk*lWyQXF)a|&ja>^-WPOD+)=ATJ&0i9~ zN*{S#fTL57Fvp}k*>Rs}+Gu}3T;yLLe%l72GAI+19F9ry56~TE9%I}NJn$vA+Dmcb zs8Yt+zr9Gk+ZI(GV?x5_aeq5+-@{8xwSR^<%W0GLE^zaVa5ew1c|QhfXE(6&(tLNS zO+Qs;`$eZ59=6?F4&0tQdTM@N1X@O-;P=?6JwK7a*X$eo5|@?yNl%67#F;e7 zn{BrvXVAJ7!+hgrEQ#BDlQ@Y=t#t??s_%uv0J%w#di;f_ptDRR*%N|pa^ROp?T!$h z0f||XRn$4>XNCQBc1zDbiHgWz2dE!Q&UGBA2S=L6K-A~(VVzfddwoR~jz}O8$cRgi zqT`WT-qgR3Y~HVxEQy>Ez#xY?Juq;ox{NEoyza zZf-1ggCfCL!?Iv|`68V6o}Pl&lUO#oLHnKE*YsFsW5`6V@p7HNP8ZkN44FC_hQ5jV z>Ng`Bm&u+|9cZ@?&CMSErL)a{=dJtAoJdi+DsIi3JQYMQ+pI57<7!*w4s_HH+bG+f zbq7a9L;f>wi`TZqT3uhBe-($~;&_Lle*1zD%YGAOB$M18e;KnsOq6+cV-rt%t!?sG z4UsJMbj(4OXF9dd@iXO@zJB93a{|tBG+1xa0OM~nk=3Q0bk!JCC#6YOzETZ5M1xGb z0^S?W4BtvJR4WuoQbRXQ2O8+PiYB=Y z|N6t!LdjRolG^tom}fTcJ(N+X>odmhkYj1t*jcotnb%l9 z>o8#U()d7Ui*EPV(Z?xvQf)O=@Sw13CKaZlpuj12bhNtJYk1akrkaJ*B(e5Q!ptXU zUz(1NnTwuYotJ?gf1j$fU#9PhHZL0|3R!u!;bmVYf}0y40*Il?!S`(Zx4ii zJN2o%Q}#64QuSC5A%Bc94UQWCW(T-;0Mc5|qC_9Zb{*yF(M8*-*j58pjiGKlA1R&y zoji3}v`FYiZ%i>iQWg#k2yl`BP_*@}3|27btnHbKH5d!m7rEsFf6G#&CUElFVhaDs zLApd)F7G~hrw@cl7gT@b3a8@_Y2UUt<>UU=8{wvSfL? zTr}Zz4wEc$k1?m&)J*|t5)5!ys;T%6Lx>e@d7-hUU9)Pl56hK-*$L*pG-*1K)RAmJ z4J0yO)NKG!G*tmZz~R+ZockPujiEDmc>Jj3$yS_MA;A*EN5H5+FD1*zS)=zE+6 zVGkt<4J-f@fMMWA_xuiT#S|Iy612>yOs>o8TFieTU;St5?A~3~e~yuOy%NVzA0*j` z905|k9jfoc(S^$QO<5g-NOz9@WVNNMYfm^Gzg)$~EoUscAx6b}oa+m+NMA6W_ zZ#P^OYE;uq7av7+K4sD#cx6UDHuGFKcLeT@PaeGIm6(@+9-naT9IH6C^$ zIraQMdNuB@sj+aBDl!6sz8%BxgMa=L%qac+<<#j_epQk6Dy98lS z!0_lN(ESHmt|Gg_HGZBTK#SGs(cB^bfuDJJKycpp^FAA?U(a&y2^%LV;p1f@Lm)Mc z|5S@;S2SA!Tgf{e(Y@hk;rWMYs7@@T?+URW%v;vJE2J%mLgEk?nqfgdynXbeD(6oF zh*_bZ0mO^5(CYrS449%fP0bUWg7y^}?)5}bOWR>+!^(brOQQL?iD*H=%u9b^eqUtAg z0J2*KXT1hvkB!E|u>M|hji5hj&e@5|DS-77$~c zAoV~lF%EObgJaa-0mMiKpD4vxa7CKSuU%}i(AE$ziDah!fn|Y3ug}(j(K&0xSQJ!P zZX)C=@{_z&9tkni9xFwtqR~U$>|CpW#*a7N5@zJ&^#dSt^9(k(@Q$Pg`3a%)f6TDI zc!LT5RD>uXfUo9d7+o~X0O=A2$sm%MxY(~{Q6XmB$Yyy}6A#W9+Yqfps@jeKyerlx znb=TsdCwv6FCT4H-63mECmW^@0>-Roz=z!nh@3Jm5deB)N{H=U9~{cb?>hq8&4ZS1 zzy8twa*S zEB#_D%q3pDcPTylrqmSoXtmG;KpaoXTuA_6+oR7M{;fB+7*CKNRA?#a=E-VQ_6P@v zlQTjyNe`&NfkTWaah~_2nu>fu1)`zR!Mbj20`Z>a-Y=b@qiM-wHhy7VEAlHzWqjLvh~E&=CC6> z0t{IQ-&g^@2HdBwvmqYMpR^Z%sx-VkMj9j44d-xxZVJ=7m!dc`FfbV;gTut;3*#42 zZlMv**i4jFzUO_eibDk&6vQntfi9@6*9r+~4FWD|` zhTzScBbg`317lHzO!?VPcUMHQ0RdJoK%FO`9GkTJfdvplAgsO#p=T}(?5UN^Zk4zK zM$7J)BBB*H&Rp74jX_kVgK5xQO z^)={+=bVEgG&mKWBUZMrS)=Z})-*MdWIVHHB2DeuM(L;&U8DcU~^q=kap)@Cm zuLZ!smR7(XoT^xlR;&60V^%;Tts~m4{ENgp7jYU!#cFmIj$R?CD=Rn<%lpmRt>jhN zc3?sMzjabIH1h)zCnY_E_hwVVKGOc5*JOGh1tFOlZtxrd z%iF?xD48H;>LqEvb~AGTct4Q~K8Z(Cpe9`12vKZ_ocX&&)Kku*o@7qQKeaqY&=v8d z6i^U%t1Vp>p&ip46kXLFb0FG>rZ{L6pAk6#$#Rnh)ss5we;F@8D2C&mJXT#XsLNKM zqb;sfCawvX1SVoM&3D3}IhXjOxk9x|<1o;g?D29SfZMaeG1kW-a#?(LW4SGTGw1A{ zioHmvsT?50O_9-OzDMmNCltXFpKd^3%HE`Dr&zl>4Z> zwr5?`MGuoM7dO;&CM$OxHkUh7H-y)KAU9h(X>J6|f@FGksk2vbgv^4Fd;OSIXz76) z`ZofVLG}bk{B^pkA|n!GB#?6^yrQKkWx;01`iBnM;=X~ZEORR(FVmf&lH2w;{nRQc z37@RyuBnoSNWCgb8xQX51lfkYdE0BMfWWO&8vF=%Rr*|baQ)a%G&4fAtpZ=Sb+^ZE zyMXP@^&~HoDPDA^Oey9*mL7+5>fV8+5Am+M)IgK;1N=*?vC^(>q!!{PUN)F7)-5-~ z61R10MtcMM&s>nipvZh0t@slA61Nh83f&CXxk`AmPi1>%S{+=`8m}xz zAoD7mw`g}LmS`~$Z#MwKx!W(^NN&5jJX7=}H9U&`ps*jey>Dgc{KIKCV zUK{a&3m6W8mDDgrshN{QL@-^x8u=%Q!N2)M^d7KaAyE?~8VK~8vB+qOq=Kh|=P`cm z7~zR_ieyfa>|<<~4oQoyJy~Zdf_1K;=uM>N?$4lHvzJxZxZzL!lx~Bay5R#MB}tM% z+5gGlQ;&OD)_dDq95{pt_l2wgvk^7GL`8{gCuRR_1d zW*`Mj!8iWtagoPUDuLW*1JtD8&Oqr)>iyW5fr@=q$H{jKxv zdd0*UN*~^AqK<<+8pq!#e23q3ZP=h4 z1SEx!Hv6UZpx(0}mlcB~K401c!NC);aN-6$N+uD!S!5GmE5?c=b`1BKHzh1sIX(iEf7Yo_G0-x@WyMC1)0nSstUULQe|~S z+8_Yqo~wN%?oL7~zuK$=J@g*`FUHQPSF~{3vdgw@+qP}nwr$(Cy_db0ZQHiZ+Bd0( zO72POyv#o^l9?|(y|u1eGCPl$u-K@ThFVf^Tv7EnXbK}?C@mvz)~p80kOds?3>Y-0{VBG zy#D?Pi$p&=-Zb34BxO}$n55WOeH2Zg*4-A98{j@(67?^X*n4DH@BoE4eR(LL9>Yg{ zC$yx%YRW3NePj&N`c)welyB(e-$hhPO&rsW5c|NN;61vZg|Kwew@P%^%T0Bu*L zd>XTl-4pf7K?g9J7Kr}l@JyA$Z>BGLWU83g4C8mu)pvcBw(TOBnaP0Uf#3}I0VyI_ zaUdyDO24~g1xw7!c)de4ZXl>Mf%%C?2RtT&@HbBVud(C;jh($83Y6+B)Xd zv&>W;hbJNTpI7LvZDoh3mu@;&-xB@U7o{jIC?z8 zfPO6vk;ewy`+o;u1MX5ai=P_ijZ)kKFB%Z35NuWaHyk-O4BI=%xBG}=!kCDHCdqc6 z&7&2slr9)*{Zk~pMlWVP<~#H;lT4Ez9D1zRpgM1T00+kM@!dYlUC!Sahi<&%fcc|N z!4^GcCi5lX8u={*64YM$?tA#I%7)p4CWQRuVHI%phq;Z^aD1HnUE@qhAv zwMW7HPuU;)|IYqcI9UIS(a>6LY3Hr6#JIl?GmCm2&Mm^Qg*&R9iYTPYY@~5x&ueKu z0NDiC%-aI+0cAAbHQ(ac4e0n^@IGr(H3P+x^{lOID1@v5UMGi>^i5K|IHi#MQup#- z?hSkr@i%aOoO@3%6hBQG_bK`Lz5_ufvG*4ZfoCfs5z?EXOGDR)2rWW^!sLY1Qx7`M-O}%=5eHn6O zX;z8Y%+xt@1#Fxf|g#vLcCIHaYr-x(;mdFj`-+A5~5!_Db=!XvJHZ*VTQS5>7})K@-7nKnxjlkf_|BC)=utmznnZB9r@Q&XNFKz)HRtRe>^(S?Ror1^NLRZXXn3r$ z5%z~GDsEk``h+3+F#LSAzJNY}%SV4gMZ4wX9-p`}1GBzgkQ{8hNq)S#b_)&a!}Xlr z%8B+;0|k^F#>3m@v5 zo7z`7*;L`=OkBu8-BvlNrg*9iP9c>NYADW-)M27I%J=?LG=qV}YSn|CIM)aY+xE=} zuFav6-9^r4VR_7hD0=2e91pLB zXu>ts1*aom-B4(Dh;UsJ_mMUXuc^x#r5sq@t@PuwO=6ns4-Awv*EhGvg(p{lbeCZl zsy7;YDX4rzI^J8p2x*kVg+7}t#dzDCDiMMJKSl0P&0+b}L!E9Lpxx{;Ko7|%0BMEJ zqrDWrYJ;wo)E8|o{HviT3;`IWfgKoEhP%r(UbKGJu2WUtI{C1asNUsPtj-mvTiY0} z7gb&Q1Om%VXapGKtQU`VgrEfkcHOmB&(QA{3;skF0uh4XNue%@7a}a3%BWxCh_<7kP&Wpe3$oY)JB{$UV;^>=rksEn zP##R+TYm%3l6{JBgvG*6P&A23R;7>K>`; zCT>|TlkSiZ!?H$JV*!7gdt{9BRtJ{P#*ok)_!>s33ne4lIl4BwNM5lqYjIWqyjQe& zS>!X}&Ex)eDSjT0rUBV)CCNVo0DVpB%PauRTbz(rtZ7N2l_pxhbxADYeM`;N2Y|S} zA{tjj2Y(J$i1;o8Kbi~w-{x5x${+98q3TsIWFCHF*n}!A;GQ$^9hR1aX6m&tr2PH01@${Q`zHcle;2nU&BFP=F%X2z;SWg6+rd!o9a z!PN_BgdG!vcn@hCn@riAqC;i?T4(cLu4(xNBZe}r_XV|gTYCY-Y!DpciEqerN_^<^ zIO{|Q(iJ za8VcuT!n{zBku-?ecX`unhQQi!tni?K%o&)Gyn<)&jpyoZ)g3eRrKF98A1IdXpKwa z>&4Pluhg9i#*#LOM?n#-EFl1NkIUmDM<@Z*S5;8zQ*7(8EP8K8#JJd5slX!Nwd-NL zfH~W7Np|}yKaS4#3$a!;sX6^5)sY3g5qZCOZh(J~=*y0WqOyeCe5?RIEMSnD1=+kZ z0ND+JAqX#;{^uWpVt7o8+VzG-Is$@)8wy2&d~_oqm0`Uu)HBar25kVq zYVV&XJakH>ffM8s*J>I8{MVzmQ^smrBjz=p>uA2q86!r^2sG^8{^-dWmvTuWfMR6g zk`yk{j1b;-go?)^9i-=rot$5L$c3AF7#M`2^SDf8WTXsVV3_AA*^K%d=QkM4lc1XU z;4Z@+@|vSSW1k+qir>oXe{N=d|-4(k*2%b^^avsv%D~6vCQhgnr9&9om?(Ak-`1_4`_QhoE&Z zeBRheW$&?$LIq+UljZ^^2K`roiSlZ)UT}lUZ6|WdAjop1W5$<>zic=`bsu4f+IvF0 zxjcYP$xh_Qe}ypy?k>k&5cE+laAi7cCOmtn7(nCb0RgiAjzm3;PB`kUN+l)cr{bnF zT+y!*{tg`=h)f1a!H8h>#n0V1&=P!*4n@umA5cThc?mTd_$9BfqXWoj`TEmBq+G*5 z;+7H$Js=gp@DuoY=`K;E%*_mQU~4(8Sw+(=BU)LdRP{^@3vUNj>F}aP5(|4NyYUWH zq=}V^qrZL1LDpJ-MCRt^iX9T~-u`R+B7E`| zSMSnLT(*G-V^W|x-H`PJ+J++q64q#%?6DRam{_9b3F*@?hrFy`)Kx{hb>xkB4Q>Cbl_&t zg~s2Q9aAC4)vpqPh4Iy~eH-FXAe8v+<_rEaH;o4>wKma+A zftQLGX6#5OBWAl|N`W`>YPzy&WW1Z!4{ne;Z6K#c~M+m0>7=wpt0t%(XWVq8s0uOlB1ERl8{O}Wd$u~+QrLF|u zfKgDnOUpK_S}e&1O&Z2DL{}RJbP*yK=A@iQ&I)fj?5eV?cJhi?JmA8cCx}XKNWgvV zRhGZjm^8Krip8849|K>jHKF*p9kn9`t<6rO1n%dDU6!h9+*Zfc#aPcKLRN-HR@o{x zb1xp4^7=MGvTev=dEg{L77HmD2PQ!pVlH#%Ox(3;snY zAQYLV2ol9cBu2vw;sKDXV!Ktc@-Tf4&gnc;ERVJtvPQ5InMxR_hXVRd0m+f6?`#R? zD(q>%M2&9hnjgs~6i<@EH_9{^OQ&aN9phqZbtnL4*#Fs;QcA$g{U^8UwKH06X2>Tb>R_JY$Gc8|qsF&@ zOIHl40604`q;>OopLQzhP&yxDNv$Q9tTv}mN4jx{^S;Pl&!67D5bGhFjdX(`EJZ^~ zC(Tgk@~PNz2>}`Qn7!Vlh6PPZ{4)5ER5KY{nidy%ih;Prq2*f%3!S9JKtX?I_~y@b zIxvAR_FglUT|I3B?rL^-m% zo3c$OC#gchgI$;9ea2=PC2c~G`f&_a!9>~cr6{eOZeac;RXNZpGA%UOepj( zK*zx1!i1}YXboNsPa1pxdsJHz;jGlwoJoi%{lro@M)O*8aFUyU1Rx$ zUf1|$-UaCccdWY%O4 zeLA~qPsIStJq_+^|N3SPOvGR9`feSwe#{c{{Bd5?Yc}$p(mn;dk?;}Q*}*&t13@IB zr34f{5oS3e$bCX89s|r}4`ael>T^W&!$H9Q^Y;j-p7?1{T)%NkEpp`XYq&E?q+?(I z@M3V@AYufP^6+vB9EGj<=SqFS{UYmkkqB@Ie2(ock*sJov!jfjQ2Gn?d-fUGgQqtK z;511FyoSv&g=u9eJgSIF2!g^LFHWVaz6~3xGN?ZH;>E05cp^U6NMfzGZpz3h~l-=I93F&YbL02B^{>9T$x+;-zlgs*k)aFSe3o z1Rw>HJVgM|)&E`&`s(tCM@F!)S+t351Wd)+OW*w%3dx5Y_%>h+(^{Q8;-X67pwA*z zDiRyJ{vp>85vSr%+|p9a(j5c%5O$!tl?9)Ah-wv9G>&3fp2B{>(Zj zRl(!)w?rh$E*PCR;e9+O_XvNgp7{ax9{)T#j|orKSYPuP;zU_Sd=smCWAw1fQ)Ug{ z)qbAw`ZCZU8-6$&4v+5aoyDk3BeoW4|iWsuvA#f6rStp;;1|aK8VuzhPZ-DwvPn~tRs+jW^1q!$Uq>l z9a-LsXfXnebXub?DC}tftS`dgXk@p3TG?6}(VhxZEv)KUhOulxViWuOVE@Su*$uOf413#GA*ZhHiZ6P7DULt3^oQI>oasjYKzE21e7IU{~_hvI`2{1juUVUia z_18J6Kx5{>+{rJXW-`qj{x@tTN-8)~ymssDr=EI89zZMdtXN7BcL-U3`QSfM2<3+* zUX&f}zFAb~Y-L;dETGnChA4Fv`1Q;8CtI}l<#-m5M)T})@TBpwb4*>cO=)(A-YY{x zZQUZv^7Ja8?sb4in3dY-SAG8jj3MUQ-~74B<7>auIoj;OjJ8G?K2NILO!wKfy$^YI z7RQ9d#jngWpl0;c{AoBIZte`!X(bo9xo*7}nPnQ-GS8h~lt*w7k79uT`~K=m^Xs3q zhej`;fIJYLW8(*Fq#_AThug+wR42cxW`7ywpZxwIza6NvCvskM?pTxjEF}nTVDR1D z;*U|p?*cmzUy}WXkMH&*+ihL2Q6hP+^YZgc#!zHNZPa{5{G%<2TvwaKnUbV`@tS0n zl!+_0RiDexo9-zycPVd4!2ST9Nt=?_YHm?Gc|?I@WQtZygh9c!a4NxFpNB>Es5ec) zk;LZ!EBgI+DtFxF(jigH!hY4Z94uYJRI+L7%`KTw^ekw|wO8seHarmm8{tnb7p_81 z`R^+bNj_u=ZJ}J7&8F<*!TV!Zk*?jJWYHEsy@XT^9$iYiD4Vzv{ORR}fATiZ+~9xZ zOsmJsDi1d!Qm7D^y5=sVnpx)VZ{*)eK)qwk@AzPRxF2|@8@dVCF3r5hT`B?w?y&&=Kz+ z>+4hMXJl`oVCH0$?`LlyA!cQ702v!sgqV?^k)f?rnX4bCprou{Sf!_@UzMJsU74hp zYFk`Vq?n}{SAdYBm7}GSmQ{3^oTXO0lb(`Q1?Z507vVTPMN3I5L8E*-DK9}gW@KPs zz+z&6U}RuqY6ST{+xxU!{QEuHe7$+f?8pC-^B%q+4>+HSyRGR&x6_;1HQB{kb#Piw zd%?Hp?gBZQ5BHLD118!r=FuwrE#Darw4Zz$ylq zXY!fLlIzR8ifqS8hxYi+lVe?Wf5($KI5^n{03g8c`2L^H!}5ljrntnzcV}8wV)ey&i8!;u6XW$5aZJD&0lPYI*-^=0;3bPx9;K$=dl15b$FZo1z6h zVz@K&BkGUBZNPZ7Pif^}t{sXi%BRdmDNs0azi&=QdoR9ENmnbGI|?tZPg$WJovWC8 z(m_wsp4vjQF?jqIzZo(uYeiREF=sx62U)m1_pJL3HI3sQ5zVk|dX0ZK>VOeGmVXCN zr0%=s?&f^Gix>wj!YAwvL#pM77HVX79Jsg|Icu8;y4$Cbufk7g3RSI@Fc&+mc%nrG z9eowQDQc$v+A_6du%zqhd>J#Pt`**E)sR<>Ig!3ZTXAP%mCp(5;GRr7H5HuZfKXTQ zey6?BigrxZb1ExDqb}!=G)yM#mEzM)oMyh1sV-(;4;g)IQO?WIM6PKuW^2euKdO0% z`d?efTPt*J7N)JCiFyht6=)~E2zC%YYkt+;a#>15 zW7b3cs~9I@&ijq_C(F0xa5&m)+fGR6pnBR+9!-pPwTWwjXWZXK9!#Jx>Zk?t{X_+v zL-Zzz%O?W5E9|oNwu;s>N3ogU-d1(C`j~c$dCrco^%|Ewc`voSqP02+M&Bm~q!Gnh zgqtg&_k?hRIks&g3DCflEB!wDNZx5r-h7QxP35crSH{T7m8JJj^>J@9LbExgKJJD5 zTTH3^hfp_7n`?)i5_imb*C2g(6gNhVa5R7FbDX*Jl_mY9RqvnG<*cp*tXhiI$JXG{ zueN*V`H4T$NC00#Smar@y!Kq=%fN!xhs@{Nw2NvvFjuPSmyv3UTnf6Jo2$3vSyYW9?(qOFMR({!(#Dn9-4c&XyeO4DgUa2EgR~*n#isJ8GZYg=c=C} zJQ;gSPOfq{8Asmski2)Or>05 z(>I7Bp$~~Ij|;~T7z8}boQc0{zo#JxYzBf*50td= z2}2R5$8PIa73y1;ZOb1wby`W6LxEhrU}PXmd^Od$@o6@ZHt<{UFDvq;>o~KGZd&l+ zGBb!GEjnzkSR6!ipZ!aaXZdkJzDNV^&M3#RjHyP;DqN5 zi+$|;gR;VaV2$(!%jf5YTP&GQ9a!LUSm<6)A~7d@z`9~*F)N_P$_sef>O;x`pB;AP#AN70Nt{a%Rq8weJz^a5D+{++o^K z;4$nvFv)xw&!4gb(oHT+qkHZW?U+DyrPRpg*3Ecr2T&%MAX`TA?C{I-5jIbo#|X04 zx;b1&1ncw8PrEC7gM3KS9%#bQH9Dz}8kV5(%R-)FaI^flRszyQfAIj@6)I!V=ze8C z`WNsCCyPnsBspqY%i$bDir{P6!~`HO-Ni;dN)R5pJWB8)#YZsUY$)+CjWM+Bfre}XE~jMgZ2 zTeqGA9vQGoDwahn;na;5rML#JJ2FQoK}hr?kD5#6mVt`+Mga5P>!j6E!Ib#5MdSx_ zye7hwaZX~ggZNnf`x|`DS3^0)CZP(}sZRA_)i$;?nBm?lgVgLcl%WGt!;`_t;86edM?@MvnqLw2xl9PSHITW4u zZy7VD$SR*S=&^sAEGj}v!GUEbycI#73)A4BLwgq zd?FpzIze6w%@^n56>KBMFaodWH54N6myu5=#_mI&Q=DqeHZ8%%!B$wk93E)|;dB*Y z2qXh?3SEpMdT7&DJ=d8I$LVe+K-8aBU-;_!TI3_MjqDbT3h=-Lq|?JkZa#Dyyyu*j z?b;K8Ac(Hjb?%1y-eM|KLf5}~^E_cAbhf^RZwH|&XP{KUZ+d~J@5WLBc8^Q`!E*S;wPA!p0~jBGsV|y-AZbFP43wJbL4~RZ)OsC4DVs(i1r)K%9HX8$3jw{- zdq>P>iQgz;qyp7#lr(n1rZZM^YO+u}rsGp?yZq*V$xl1HNBWov;*3k@R7bg3Hq#sy zPturzdkpg2e7^;|r*4#^HkEmz(BDwEH4mwawpE?xZw-caOdAj?IYJ>(_8l_Q+JUZT z%*x8B9>N$SsL}R@t7|fk@~-7AXEa{1*7QIkNl(wWrbb)-s*gcTZt{|Z74n9RE6yy7lb@ISHHAGAN0x#AM zjyDTyRp#azAastBI3wipfon!Dln)`^Hkeydm zgevrc7kra?iPr(39F2J26rbLejI$p9I#fUyzREo`#B0hFcD;wVuM#6-xZ-zYwR5TGzu5BwYCOV$NGpMH_IJzL4QHQ)T4 zbxl8M$AzXdNX<5t*73#4Y6r6 z$@P#yZ%tV>A0a#*`0@aaD30p4#LV6hrcqJIU#C1q(Szd}m#Z+o%W*Nr7!N&uSi;iz z%u1VgCr?;<&4b`p{2(d5_?bo0Lo=NL-8rhMNYPG}G?CC5F*hSa0ivg4B9yXl91Jl~ z8$m#i6b>d9ylJzEnP^XEe%$)SlDE~roO^?@G+h!=eFhurEHVq%g;$a*legfl@oi}E z(&cLvBt@BN_{&OGkL`jvRp&p-7YqJ!e%3$63Scve=Q9h;E@?gY5ZBlsI7J952$LNy z95?jR>^D?LSq;-TSi4$XRm|spY=HnQ4ELTWYxuj-jhfaToh%>Y{o<|5iceS7#WZ*dE6~WSm3}S+dmLMEAVK%56E`TTb1fe$bmZoy zKpbe5!#FGvJH$UT$Dw`xH3;(JZFI?Y=SYyQRDnG7lMab`4lV!EN^@$;UQqg>%Yj<(g@u&Ep5OEb;)kN!+)^pYUkRj0} z$^5R_7+-^Z(MyhGp+fkW3+5!jfl4AV&drNgC)(u2GQe-9;f0s`I1vDb})&yowJ_Z2hon-0#kP9?)&G>UD;=xq10~zl!QMU3@zQM z7Lq1TYaPe;8ydzIkwr|M@?X z#E4Q6@t8A;VP3@tf~J2ga@Y}C;a~@lt+&hut1hvOC`NL0eO$itfZszZCu62*ag-<3 zy+R0j+wS$bcj|(Jo<50x)hmbbu3fP5p^m#T)p6tPzU|;)__1lyl5SUSBz}PeXq3T< zQX57NYr#_M%H@h`gHMr+E86I?lZT(*o)Q_iCOC-CR-1EYoc=?C@h&l2bPG>GXPAcG zOHXFVl%&fZ?YWUGBw-%D4(`&aYx7<+Wj|P(wug4L>s)l%=61u8p*#NE2=udIpIFP0 zPyc>DJj4_SBdLO&;9Oqwb2Iu6{APX^a8DI3jD)u%#p)tDKxyxv&qVra+u>7}dt||{ zhOGlx8Deivn4^#Hq8<9`;wJQJqR)5+$r`xK3HX>=(HZS8>Vq^2>>KO#X}rI0fcL0|SdZw~rGaJ=0DuFEiWYAW{$nWci_S5L|C^KRU4`ySIm7c}j3lpimL z*R5qB*>PEMR?WE|3y7)4I5ObUanslI$w5&MDk|DbO~(Sq>eGe1%0=k9fi$qsw!;ML zL$ZIR#t}tIrvGl8eR}cQcP9KA@?@%KE(`v?PL7`Mpgr7}Xbmm7WzQ&N*=l3UVES51 zPzuCiAp!lxNqGqKx?${zK-E`>Ctv7TQK9%T@*hW|h=dQtZ3x{z&cFOFG-ovZ4ZSAl zEijlCCl>4u2(54lOOPR9t%B%{q7S0DFOzPiekoo$Rc-Rs-;FJB>Xtr-BlL zP8IDDVucKOcU^^-o8RG?crQ9ijP;AwsR0J1oLP0Bya0}$9?!tQt=|apPg?9NI`M5e zesBEQuA^OkBhdScfhPRe{Wq+Sm&8ium)LjCsFb{6qV5JS0*i_om?}W)7JFFf|JUnuAgC~DDNJ8$bMN?G* zdnk>Hx3(#kF73hj>T7I+1v$;cwh=n^%90?d>{_My9XpF;)F>DHT|`gJz|rx=qt;}Lh)ij*(V&@ijEcRe-L4!ENlrOfgyK=} z0nGY9O$^hZYV-3ymR_-X_LADmg9>^=lyLGLqqL3)H!Usd$E7-^PPdR6_k5x`2F$p&mzE_ zSSRHN3urWtI8?E}YuFo$YPDdWH^46aHJZHD z#=1dk61;yVDHS;zLkQp|H_h*<7QE6J9-{S*#np@m?hB^xr6}g0`{X|H!C3Z&UUuIe z$Nd2x4l0EIL(0U$@V`?gRwmB>f}x%MpB|l9QqSYe<`9^@GH+p+0zD5)LxZ4HvrJ@) zX%hDaq&J)a`%qXYn3jix@b|`caXhy7_TWh|lt7!HC)w3t7@Q5j^V9_WOGMC>k-`%+{s^PO0|Hqk(i=4Y zE?-v$9h{*Bs#T4x1BZXJrlb!hmB9=tm8JBjKhQP?4?p1d_|c{RTyv*g!%1XPBHL)X z+dO?uki`F7bK9(QT zvaU%i4Ua4dkClT7|6ti1x+Do68WkXYC4@fQ7U+FSH2Q3tWZO|j{nKSGYH5C?sSMfy z0@>0OEtP1}pD@@mqgY>*Y1TU7=-bb}+~TEV-U|iL#F($j!J19m71ZjQp-2Oa%&13s z=#);S#EQut>We6`AZKR5j653BDO+W{y-jcq1z7O}3(5zxciSCyQD|Vgv_wWhtnvfj z?Y&^s*-Kg&nwu!9;9ztNZOb>ek`(7_!5>f=K)9kGiS0+TF3IdK5agW;tSD zH3v-jcxBL$k;vVPf5sk89DS6;hiy6o)9C@Jj4`W3Rom8q@y&+u`s*bSvIV`rTmrV@1kw&eVxpk( zD=KF1ptqWd6;JWF-x06ui0FH%4qoiX*QoF?fTf69x&HhK1%Qe~)1T=QH>o7H;uhI% zschcZm5O->2@1duSZ95ya!=E_9^AZ?Ci4beLWv};8mGoeWQP@rv9 zCO$Q)dt0T*PHSL)7D)pcmg465zy&JlxQONmfAZjiDh%1)b|BB<1t6zNL|aG{jA=0& z#c1WPt7eRAG`NG7j1BU8k?2ZMtOm5joJz+>W97#Rv`dNzRW1w1-DMNmLgOghT93rG zVc+3di?qwvAM7>fA5bSRop%j4L!=6Y+v;cbXcR;+2yoFEetLujD3>NXz5*Utz%Si- z+dIfoFoqdTyE$-#L%niufc@aPPk_QOgLp^)4wfi$FkqhQZ&I)ye#@VneYRe!cIA2i zwb?#_GGz>lfIzrKj&p za45%ARNZ8GEq&ANvN7*nt$_sC04oc^VNcEG%uhVB0~-6RBB}~Op{v7eQKLviR{lT! zb~qpg?`onK0p2F{JB>r{ha5yLqyXZQQsR$P+P+thZk}4T+4<)!9JXDZUxPGhrZm90 zYmwqn%CcOoRLlaS_!7;K^?=lBn4TucgM}!pI8aPXR&X2n*KV@8wPe zZH3pzzl~H6P=?SjLlHB+fU-UGm)rr!pI&WMe`8m@ZN^E~rlW%VtdXdJxXN8?onFNHLPGqyn3$X5*dqO9^9F(J4Yj;JL-Q`7p6|3cVzmg&3l0wklAf^0w( zBQy25D7^)6ld<>GZKyU-Lj#co50eF^c$fFsYkNR0iU!Fm?@N4;3ABu(RDcCQ-9^Qd ziU@w%P(zLg0DDnL8)mbZr3LWQLQ!m^vV1eZT62``YF2yz)=hV z+=Lf^wpxc?K~(_RQwoDH$QZDAYMZ~?bs(YMiRLD)J>LsvMfYvVux#(9;~$aXJ!I}* zN>qPM5=whQeb=*j4^On1)u;&IcI^gAoN}H7(gm&JVqkKaJn)?lltJqhLDl6e zvy)PA1gR?liauK+#y$wd;D$7ddhP4?$=kGa_Q~vCo#|R;&{){ooH}R9{#8<@yh4NV zq1SHEws;`sXL=RLg>M3D-G-8TXU0-L-|M2fgOg{g--m)@fRn2*t87{d zr+1%V6J}~28+3$nUkzPUndRk*HEJFkNxZN^A+OKe1W|x$qu85c=tEme0f-!+|4M|w}vc3P4cyXq2xckivrR> zNVX#n?#m9`fXx@;Y|CvLKX-4S6=`oElxrD%SNZO>a}J<2=|bd+F|iZ-O$12$Bx{l0 z?=lV#v1X8&84N`>!kk(S2bL_8yWlCom{1zxQn1K3lhcaPOF+O=@^^xC%2M+64jlRZ zRKIf4qN^ZDXbn$_KxCu{MW&z5=KG(D0X~Km&ykr#4Y&j|4g+&^P04cQ4@6NkuxtPd z=O{*Q^ThUwgMxI4G9~Q6#hc;;_ZPTVxRv9)VXcj(iYu~6>XrWR^zrPwP!YX%2i`ea zN}SfdT_T7HlSszWRIX(A9smnY@69M>uov|)prW&z+iyYadH2)OA1|ya&=;?fOrE(D zt8A?Gg`5oa)zNzF^9qs1`c}1oXmcV_k?=r8_`hjeAj_B;b?{T#;$aLqt?u%E{%s{E z*5v$l{SNI~6#_`wFj}Q5p5}8C;qBB1Yv$iaXV@sKdM~x~6#QV?3Lw_Q@ zLUs8KFo&o^QlPmhB}Ny6OAX9bZk$>4UYqBR=Pt$YDIgVGbrNziK5m>&+4FvbiayRbByW>RAIbuAM32(OQZ!lD zDFG{?cDQ*Ru{a|$pG3wFGAfe|m!dZgfZTCubc4dnGQr)&ojXMsnWjzF`P3bd+ zV|0#@34uwk2;W^Ihtj`7RS2N&o!JNyyui=+Jv@g$sTZ7o3ed}D;%GzPu0k>m6VTd; z@Fw7}Z>z?LZ)h9`LtEQsaedL zU8kRJLJPCbA*mg!+uy5|egnroQnTYuJYkyj^iX~LyqK4h-Rh#r)nZz2SSFp-O>*Nt z*wOrt5GFG*Tc*Rb$tf)<;-5-{<`nF@QU6`LMNnweOjiFYpSxJPDaD4IjuelY(N7A>YU zeAOeVJI~dZ7XLiCdEM+8fT~<>!Ie{xxaNHkRP&qABidHIE$!9eC<9yiqoL}eFO^@ zcjE|@Av-ofgVaAewm2_+yppa6nK~*2Sy2h{QzR!_!^Dc1Y;z`l<&apfXfP^{@WnLr ztvNMOoZqKZVp&1*<4GE}Rh>t3DniMh;f^M38*Azv%Z0^`^fQfaF$SwW$q!|YYFQ2% z^LlgTwImDoEd}z%caS)wY@-e0J@}UT0jH;@!pD;D5Gq0?wox@z z-{uLY)m)!z#t#YN$_H7>y?55zD6hQ7tST&(1a}vG_~x)^Miu!_)e=Sps@1BAtQk~U z;k;xXBZnNhXeJsH^(@7QzLb9iv~f;g`)4S)ItM3Win3}v_`yn8)e|L@v3KPe4Vj|Q z;gP7)8RTwyh*kRU)Qlg?z!n8)B74e&RgjaK^3q)AwF94K;f?oLaVSum8-N3-v(iN-=gM@z5R>ByhlBF(M;#V35NEE&GwSGN<5-75)Q{_W$F<|1{ zpnj+K5gJ;K%h<5~bz>J3)@+&BRE>okA0usV1HQOjvk3`cYNlt~ubR_b*$M<6;Y$;E zD027~9Gh*jRZQF3YHXqU<~Vc7vdfT(RfW|UO)tuFb0Uk=&T1glC<@g_NlWjnyW0+! z)L^+|ek{|0g)IAyY zW(RK%Nss+QqsZz&aNN*7vty3JE+NO;J9U#qME|`}iqZyZs5o(;={v)m%0umaNlF1~@QP31KD6xkL4_rKuu1M3>WNKy*T#RWTFZPurNr_~vfnDmtad!U|x*8%zgP#yr5`xkt0XtM)+FF&?Ts6F6!ztD1VBC2C6As&MOX5i5UzpB`LGzhA1S-tZ!x zitQcWU?)b>=e#;Jxj}j{2~JLO-2FuS{%1?Zws`;Hb!^pX7m7|Whr#qxo}C5nKvoA3 zlA~SdL!#XB7kH2H@MF~D1zz+%%6#z>r>P}{DP{wjh5E-dIDa>tT2D~VQ+5{R++3{j z-2>>ny;Km>H(Ml0X; zT)d8ZmMj?M)950S?1XFYV$`G~B!e{{%mICck%^*#`e(Qh`#h361?TQ$B!}XRrWuR# z4aSx?R~TU_4LBm4d&0Si58S!~%S4NN6}jH=4TcWS-kHvn42nYhh0>|$ zufRX-!b5wq)Eb|i_H54{rOEz5+Y#TczB1-;{a^Q?z3aGPf<6E#*iyXjoNPWTl+LkJ zx^V!Z_(;%iQGXT+AYV{0E!Su;%b34XtC6@`i#5b`0)5t*dGA_8NVQCM;cQTPyQ&e5 zD+4WjT?`t`(7})=JE9-B_BTh!kBKm6h-s((=bo{AZbnN0gh-r&Lg z>4jE%ez6op9`%540ibUY5e~eR717_}wRH-Af59idy`Per3n1CJ`U7&P(|-=D{{n_sx-{ZEM!1F{_jA;%)00M#3lbIR~|* zS`I0i(b=Tr^g|QCM264|5Db76mo9Bb}t8-RB31oOiGd87-2;7RwsnX!W9;N|p(ovZ3j# z)yrEx47yk)OVwQsoka(>MlDWHo>fJh$*G6*$2(ByjBbC@AI@y++Tmy;?(9bqxuyS6 z2Xgada|G%Qowri1ZqY@RA~X3h@KtDFrd@u0oZ@*~xwv`*4Dca9b`&WyT!&@&Xa5E-}=jJuenjcGP1%zU?aV zcT?(K+fH#^vQJz>zE2IT%7Ml9>fEl|OY7aZ<5u+r}5$R@b@iGe-CR4f|@obIr%t zm4~Ec6gsQfH=P!J7_T?20`19bHPw|ZTZXqv_Ha!M#j>V2P|a-AIdc9s9lWAdubpc5 zq$gS(Wl(D#9aaF}#Z%|$_YlVvO-Gg=&5E^LmqJroomCm3V`BV6st|pQYp7mOq1#{=K&jE2;85N8>?j1$ z6k$-rKunGG=z`d0e5uHCF^ThRoBUp%a#1SU(-NST#$DSO{qW#Cm0Zk7l0`FoZh?`O zN5&Y>Zjiui)h;zKX*RuJuQ`QxBa`!s<{bN~axoLlZ_={NV-3#jqkPB4Vo5bp@1*sl z@0Te=sut>0)a4STo1xSn3*~{PB?jSdX7BP#4RU<#J}VQB3~RR-twqiI%E1EfuzTOT zYui|}0c{hgUK@gE-_)E(HY?IV`fkqns3))1nSK9ABP-FKi*vc1BP5d4sK0oyWbs3* zaHryFd8H`FdJSFy#+l{hwa$Gg{u9+~QM3t{y!}BdtndPtT5Jm74}J7}RG*TW%xh^y z4=Sb*d>B2>Y1)ak8W}~M+f9Wu`y%a@=Wa^&GIkYMFCFF>R#5Wt}ZUD0qJf zQRg7l`(PcUY&<8c`*ERZ$GL(MMLmEDUclflQ1lhP@W6;_PY|WAcy5R;=GdL;v$S7z z$;pG1i3m4@+VzD&=3HP3kPE#B(hmJ~FdPX4PSS@Ps-5D%ZaJ!toh*M-d@08lJxq2` zRdTbYOCgqKKpy0VAJ38RmqYTUf-Bs$@cJux=g6mcj3!Ci3{fz0Bqz!SGl8Ho=QU~6 z$Y62RlJ9BPRQmAwz3+=W&8QU>4f*V0;}@(YLDa#%77w3!sr1ff=iI?J>^~edJS7_= zrk2)lZLdIw8CC4vQD1*8C7R!hCQcAm(PEnz**k z_?8gl!uTZi6vDCQZr;R}k);a0XRz+)iUxMDbL*|GJcuz;>#uO{lx7C&Q71CpdYGa2E*s zrO3;y(N{57HIM-)l={`NZdux1f!BoxkXA)gh6EG16cYgq%yPKZrVMLb%4S91%5{xO z+}D%FP;M*MD>hJW1|m=DOewWqyT3N!=$BiGdo9m>u(LeaUtL#ajdylGflLg?j0Y0u zOg5LQ$aih_xD9Tya2&{el)9eswBH4fO7hSBnSvr#m-zLfP})b+jti(6JCS6V`9@GU zBFVrmSLWZ*k=u@7g!TyaRT>hkFwx$tYgWwE7=0yqS-{Ax)sX>)^~80O3u}W=ez$Ow z4SK9%8jraR73`;?LPLu4)u>@=-c@Dy`CJGdOTJ-4f zE+$lV8FJ?z|N4F;_S5b&uVD6qp)pd#gg-XMbF=r?X*twP&=x?9=Nv3D-oAC7veF8t*AsVFh_){snhc2CA)@bzrNNH6Z%fe;;t&H9y~;@S|JLH9xO~AR zbkHQpCb+u_i%E?1#T*(hKoz9d$*^>^-rUd>W^N!P%12W|Yz;P8l}D}y562+i=O9UZ z*D)rU5V~u!Z=u&6eS?t*2Ga4ta`TUTDaPMF_jl*CLw!|VUR9$!_^9$eL%?EP!muvi z#Hz7n^bg_hp~5)S3Pmay*`vk>>34UxlFLcP%y zd&qTrfrvqn4>G^~ki&9jr9AQ%v$khXh@g&qSJ)wMbzo|3s()fin|LpuLpW6HLD^z)eqrrLVW~bzM||syFotDkBpUp zE~hCPgRTDbl+#&NOAj-aW9G8*S>)bI>Vd3rP>oSa7NLw%k}*fndq zy6|skug&iBDW+Q{aHnXqCY)BRNt?4Uc{y{PhiiQuxrrbBsBhQf_DYdmVSizzoQI%L zk}1|^BMQ0WNYXE-y)q02Y5%+q_TBj(u14|lTuvO(b0(5;u1Yx&gHT&^qheRK`Ktcx zuSSSZ7Fkcuq$|gu-F#w78%I9bjU4OpkZYJC$zJx*WHq@DYUS%u6m{mJrLh}HhrI0Z zPc5Kcg*N5fBGh~qP#pTTdq=_4$bb_tdb{ZvFri#w+%tbP@?W=ZmKJpNbIjCjbl&GJ zAeqB5IZ?tR|FR%xB%pZ#foTzBN)kSScSISi=6(tcjRH;skzpH9mYNY!6Qlj8q|sNw zWt1ABe=geT*|$U?sY5F?DsOt2gBn(ND1}!r9C({ZzrdSwUwsZXn%>>h+vmDO2dZc097O|(G8X8b5&DZ7-d4ra~8RqBY+HO+k)_e zQrCPjFogV~t;@$)UowVLY&{55ICv~hb*zN!Q*+jYw-vwzGoDL6lFUD2eNtG1r}=)xWtB&rF(oH{;+Jz} zuKo^CsESlD@P2jDvR-c5=qOadG2{I6h!$1enevgRj}Gth7o`Q)5hA*_&1`q@AuC^S zcOr)-d?_g=4lk#7l-vhVh0~EqHa)_^^~KL&ucCyic!V2U8e;WOIRzi-7vFOQPad1I zq<~%}muCKqDggGrZpxib%~b82h1kjYnnVO1MJw3dDRyYyM=)XLXKkjfL*%iDibHG} zj})6cdMf0vqGwzyeqFomL{t^_)IHI5Jj{$b-S z)zY2o=Hx^tNNE*yh0AY=S~31dDryBdLS$nNIrfB@Gv&IcZDS=bl|tWhxrt|qvFW1c zc+)uxPp#VM8XCm1XBT#&C)U9PT_GT*8Op3*bWYM~&5t(k;K!ByR+LJ|?hXZIPf$O) z%>X8(?i%qNNR>8tK4yXd9~y~mwsjnk$4y3m>KV65wfQG4e~`4j1>fLZnc2dL5ct~a zE^R6TZk-cY{0-pMovB&t`~|{{uYDdxvx_&{6y_+Y=}K&cu)+E=0*orJoh6P71+?*h z`#ka#J4eb2SM-&^eT7DhRu~%sC)|AKLBRew-#RoHBJF<*q&GEvbq4ezT}+`*l5-w! zpMb7{xzRJATF{8SEz^|ng0Ns?{@Sff-U(8DCp^w;$+OICw-gRokV}j=jFPX+Yl5*t z#*_OHnbs)YfU_elPo+zUh-k_+ZO~4QY;Hq_Gaj=_3Qup_@N+c6vy9^2@t4& z$%#ugPo@})2k(|;9gfA6zN9=9Fray@s9=7L9y83OPbS%nysNY~7HQeuTU#0qTQ%DV z+dE`7W}>CsnFxH?ICBx%4o+I%-%n4KUK$eb+I@k84QtgrJf3D}K9FxHk~LxuN7=jU zjTYu^#Kz?GGH<2IcE4#6J+H`O!qsx2Dok2u-b5Xv4tZ-gL5J3kiq-qNa3`x0flU6~ z<wFfj!pqWoGCF;I_SaWOh zh5bimk6?(j!yyt3x%kb(^F^_OQxjZ#i7<);rR!I6lNb)4@BPKTDRM)}N2@I9NzRUA zFIs&VI(IQB6j(+Ud|z+%c0iZ* zf$*iOocsOTq3@tqD*V5bKL7)p9s$(OD);4cMwS49Kp#YNK}h(DS*%Ix9?=y~oE=hQ zJZ63<<=(p+E1icpKkq&d<#^XNuV(u1o#$P_hM!KWPRGw3K3u(Q>kD$$KYbTpq0)G; z1&*94ke^yeA&5Y)W@^TYXDE?^&R?HL`w>ZVz`h+CqdzF@#ma5S-{j=dva`zeopzi2 zyj)t<7u(F*%X0&?n!dG1u7f zVitCu^z^b8fGvA1zV=)OEje%~G2MPIcz1r_YDC$Box4>F7NV>g%;cD#)x$}qJeOf< zZ;X|MN69b3=0dUSDi-~&6cGDH3Xptuj?-l%Vd^T@3CgP-T)rNeSm@|Jvu`ENA{^Vn zf9B7PW%^zQ4!io48;qx>+Bm!0Kwv9^C2#KSU3hV7LU-A99IK6-0{j7>%xyMw;(Y{} zV494!nlxm8}v0`+XH+XT%7w{=teo^pDXFpPgh5CDU{zJk9v#|I=u2=0W=)c^fV^OH2hP zlTgNx{x!f$XB;S8=TQI2`N8)=F3yO{Bs4tYqpmnW9FsZ20Y-*iqK@Z~aR%NqD%S&u zCP=bZidr1|DOH|Y4*3<=^%YlJGUm~hA+g#t#z+}wy zr5BubhsSgg?-6zMn1&~8iU=a--Bu}r8Ih!ryQ>}HqD1FzE3xcR_35eX9~dQ{fvW)+ zHj~r(7e5gl-7wd#KS0-8GyU&zVzs$jy1Ls-h^@NP|5lEiX{B)U#(hW@I5@C z9cS;K8g}{9K?dIsFQ*9=C-$?f=%)iNPrNCuw8B*)>k_PDF*1FgF+o}uUOs!FBl#q& z*Adrf1OeA!BrXJ3r_y@uv=3ZI%OMc*mn$3;a(5k;2`A4d-ajiR3!q6znUExF>4#j; zf`_UBFF)%-CU<9?qXKc?m}_Ipx&JNAWB#vc-v6V+IQ*}}xG9;`b8)>mCDckUpTT$@ zT6r&*#@6zgu${KS()m0DAu3F|McM}}w3*e*<8OkHov8*MfM)tq?d?8@%nrQ7Lh~@m zz%uDCAdl*o1IWLtMZ^Fj2p(Id`5 z8YOp*TzY@C_8pqp%@yDi$kg=>k~g1|xF6b>66~#dL-q#XNU~Wbl4)nFzGDz*av_sD zu0Xr-y=^W>2k*-ow_y}Abq?&i6qVF{dQc~`9?37JH@<@Z=SVj2%au7*xxszdoVg^D zwyM=khokkjdI9cmK%QN$y~s^!g6`mK@LQ-@c)nw%CeDvPiLvhvGhwfhMAH!vTI8D4 ze(;ni-z-yoxlN9c?c2h3+#?TYFreXe?($ewEn+xoN1Fr@%Mf`0t+ePb=rlqbbrZZ#d!)cjJ zNfB3u%m4diluwK9ywoQB;>xrVj>2(!37)U6weyL*D-3*)+@VF)x`p<*Q%#G`m|)VO zbz-jzrTCDx2aGB(0pN{g_^*CzcVHfFL02r*wBXoC>F`(4Xq_uAn|otz(ZNl8V14?I zZ(^x#v7;htE{4V*}BL~P;3KXvG(F3qfmD*N3CNiJhIdPK7!&o~YP*g|$?V*1tr zOj7x-ZKrYe$r^H0<;YS{xo^P}ViM%cGjc!zEe*|(njRDwTbJvzGTWea(o3WiJk11` zhpX4;KAFWt8Ne>L5J60UUqsBEuRtMm6M{2IPB_BHU#QITD=?OLLMZ!cF*uPvt^gP|rjt(n0Ky@i?5 z^yi**|DH+KeiGc=?>aLw$xQaQqKrO6LYDrPJ3iC_ULRm z--3|)Yga=>8O8UY2i(jfMw0f4Dz>O`8;nQnO!NnS$Ps+kooggQXLDFce56zb#J5C~ z5ql(_;vz_|NC@tGBtpvnK!_)TK!R~LyAQo?L79hKTyN;}6?leju{pvVJ%B10uPeHN zwOcdo!_(tY(iQ^lqum_OnhxIquVQdtF343wzKmAOydvr-VZEQidB z<=6l-34MTt5^PT^I0krRU9 zR5=<2Rd!g=xRs67mDT4;dYDu|xb8}a8Zrx}oE~62e^{jy90v4~@IU24O@H5X2pBmJ z(F`r!vj|bG!&E^tw1?3(9d+c$9t8Xd+L9+6;9_XSa)XgEvcGL!`PN!!TMg~Rsg(oMJ6(77{ycuO{%EVV|HOb0N`1s%n)Sdy z7Af{za4Q#Vu7UY@Gsy>sEJe<3KhvN81K#xamej4#D7>|6qOE1j5i4Wt(Dau7+6QEN zNT3`6y$b4>Lm~B30S^bgy)Yfys*unZhHkbf5;nG9l-!+LxL)A+h0*e09xN>=u^SIx)uE4mab=?Ywwxc}T+LCnaf6{Y zYmz3$sU{hbHDLcMwX(?EP|_B(QoL6z45a9}jxOHu5oiX)|JSdRAXS~fkpWpsAjOqg zKOhIy;fYq!WS?Ary45^<^Q5a4N4m7``~V(Th;c)eW0Qudtp>0#=R5D57asZq9dTYFJ&`P zpEX-SAF)F&6X_C^d@#IA&JPxp&j{CpO9swcUl^=C&;jQC=MrowBH<+XYQ-{B;GISk ziNe}*j(P!vM^Q?SN2dX#r%kU{Q65%tUdDp(!0e#jj*z)#ZI44x;ot`Wh6&I5j^{1N zf6p012_dTsW6)HbNQqeHIm7IR84ire0VIKh%8#%gm|0HHSVF8X7RYV*k#?z0k@DO1 zVT@>N1ZK#`u5VaF6GF)NMz=zyl+L_Rn=lOUjFnWxY{;CE0xTQd>Q+BH#qw*n@EY{r zGMoJyy@`)l@FU=mZE5eC7K8*uR_)M#6GfJiski_a!yI2y5ICc46tBrs9t+rC(EsjA zyATj}3JVaC=L?bkbjQj~aePAK7G6}JS_5$JBFOXI^@Hk`uw)iQFWAD$ouOr!chgqM z-ZBHWS_`SF-!_6tLr{a$hc)`n8e$nCuf?dU+rp-j#wwD7E zl5D$VL~S4kA{2(o`mDVf%wOB!WeRY$vO-N$HPT+~2*HM)Is@4iO-P100IOk%3QT9)lb({V?~B!Ave`HVmVDY9rKjTHGox(Eo+6Rx}g z;RJAu+CCX9y_H~=AyAbuO;s{;dllol&Zwt72$dgq5 zNbrW=RD3XTDzM=}^>ZBy*p9W$K?;>yo3P;Q>n<2MO`s{4C43WSxkK%KL23Np^JiJ zx9&=~K&>;6hoT9@od$FLSs6$i$ONlXt$~zzlB=7dY_8&-n0O0l;1@-(hSJnu`l%kY z=yRqfF;bZ8F5{u)QxxG0NS;89h8HTSnPBMA26N;Xnp&DsE_nGXDvfbJ43u9lF*Yy+ z{t5T&+!#H1aw(AR*^Cvgoo(rOy0x>F+?`-qt`aSVsQKB6?{I^$?XrSGSv&fiT$DvFQ-TQtUC!zu5hq(P+NlAwZz)_r}bhHTrh zV9ktRRB_2Pl^xyS=dXb(;zpslX!db$ot-S%QlI-QWpVN6G59eMeMm8Uk0lKrX!}US zVV)4vj-UD4@qR4>a^}r_2pc?S;Rdsw7+GypQ`NZwAEJ7vyw9eu_2vR${dsxao6)WS zA`*x|THFpet(sW?B>_f$**%h9-)t}d&F2?q8!~k<ovjvY`Pb+CyUA`drw3Mf5YttkNa}u<$na{S_sLq*;#)$?=R6EPRVIsI z39#NhmOjNJ`=vdni=nexkThsS+en|GgXnSY zG(C}zXcJ2mlbxshc)4n}XF{qU!GdF_6Bf$loGn!5lDJG21s&OkjY(7;5uh-$)6?A* z5twD{&qpu~D-=@o4&dZ&oMOrDT#k>SwxdN|7^#!mO^201O)RgU2Cyz{+QncOq(7)ld>ESVFy&L8wEg{2f9bD^SE^(}ja zKF%bsH-(iKj7NuR2>i+^m~c~aR8gD=yM}?1{z$rG<@jS47VLe)5>H6u0>5f@^uVA# zmu9fcr;%$Uzpd$iu!j=h%N{x8dL~d1;d=(GBV;3RssF>xr-kvx3y1I>djHl_aemDq zIQi`-VzWN)PityHeGJHnioSUjpr>T19#)y8ylw+S+ub?rr?akzbQ@0jE{8a!JyYq1 z?qHqR_&DSchg+E=OC}RH&SW6YC)>wafhDic(XCZ~t2PPlRbJpCQMoH%%Gx`wbE+8r zX{xTeEdqJL#{7&e-HR?t%RoEe`)r*ou#+dv3ln?zX#GyqbniNI;+u(L%fg8806&TC~Wh2;s7URpm_HUu}s+e zJ=fP<1VVGbK$5;g1!Un^<1g|CE;|-3gT79R9l$-nF;x2))?SS4@~#i=BQTV1oL0lX!MB;HiH!X_;_F&6Cj3tFTvdgTyGMN1qJaxfn(F_wbdqvZe1I) z>9cV)=^y)3W_dddkn!Ov+nO~}&n5ZrK8;Asvrg#IOZ`QVX6R=p>UmQJ;DOLph{q38 zJH25;3`D^9ut3JK6SYj-oe$I7N4MV`-Xy~<;(Yni1fWN3yy`3R8s_C`?MU$PT{;`& z3Rb0K&r8jHnZ1}wZQZ>zLTJMDDbBZuJw~YeZBEGzsIag;oL+YK%Y5N9RN*`NxTVt4 zm}^AtGzz3ox0Hl1Ps=(ihH9ARGpbugl012l~6~?hGt3X*_ zianzK=!3i%`SFWO>K5>JV=k%z39Cz_17Coz3H845fI@i&A~ls18!?|aN_t)9#saADWZ%q7qc)Xy={HxG z=V2$;O{(QTd6>yh6jRFU5>9sMqJ9;)E6hnz;NI)YQTPSQ_v>A2-ky(En6icPQ%+MU*4Zj=)PIvhU8Kqr}qC(Q|$4(11WZzH05mK?*!qlHim+pkv+^=d5=7 z+sRW{!4l_kxQ4m|Qqhj%oS%sI*f%NtlZXOn1m6Xf{t&sSOvZQao_Tk>dOjCi<0yPx zAuaZ;Rpy4?lTg)G36Vc94^EHZNCctT%LDNiTv2?BI8B>LT4vw|R8K!SQ^xjeN1v+= zJBQ<;=fnBDMoKB~&-L6i?_RgenguS<(Jw;C()L(X?hdSdZ;Q(YT8!b!3lOF82C7!u zZms9WTsZU2>djI(3<+>#lsZsGOnA4zePMeqj}#ehUJk zp1j=S3)flybL*uap7ac0S^>IN7XOD(OEq`AP~}#vt4N2RC}H6)3~>RbKsP0K;K6Uc zw6$j`t!ZB*9Q{XdV_wh*g|zubCxWxkwuiU3j~YsMeb>gO;iB=@0}dTviKg+ML@N-4 zyvmFyFQuDbqL}$6Opz_&`gi2CK(a5cKx`2r$4AwV3$1RM_wD9cb{EXd~r>6v58`Ws@n;o2) zWNL9A4e+%fM!=P zTM9PYR4dPV7eJvjtD1*aqX>Aq%;%WBhEJzxpHj7?wXe7n0Mcz*?L6{1wrC`d_c__MJ2-{g~5IKyXC6#szX1;cbgrHvn(*S z+^T2mjbqi znV4bd{`yo}WkC(oOo-YP?IK&XF$F=4itKTA;FH(eCWV(x+nGdN<5WxA1ADflC6XE9 zqV!2*-$QT#lxsd1*7B0kMsEQv)NKPsQroOa$h81*+4dBRo`-sr{kSRqvYzx1sHv&n__Aj8R@CaI-;j#lgn*u^S&S=9xOK!ljr^N?XKyN zQ(NRDWw&8f$>ia+UGX+UkACeu$FpC*kBr7~u{D&MNw_ivl!qn13x^+6@b?Fn%mslM z@3`NXgP~B-hFmhqX(?A5Zi(Ae`YofRMCd60s5%~SG-+A)hqd1r#rV4N_%WiQxHTbM_8ZR5a@Ost zkuSWo*rK(>g_Yz*QG?!kjW>ZyBds;y;l~cUr0CQoaNkJDQA z$Qm}a=gNK~6L_P2#4<%TE}ne~0Rr|`s(FGStb-sFs=|1GMFtUG^xgij2EC|Kz)YFC z;FoweR(=*HPsfHM@U?F;c`Z_8_A?=!bLvUnk-Jg;yFA4u*EXFyt)Z8KK9z-|{pM5` zZZof=<|G`vB2#r1(;qCpoc-3XU2lSm)QajIV#sW>%hu!cx?h;BG8i@g0onp8-QkCs z)}zn+G4V(czJ6a@i#Tr~Z;a|t!va2MBxLgGqHjv~1*Tvb(^oJhs6)1%v)j|u*^rx1 z&cp{L)K2_KmIeyj=qYnE2jravuKe0=BmlEm|JnI@P5@2AVory&$e8Dm{r25ftz+ap zfzxE*CiL2-6@ewwPueu2P!6c?W95jqc;GF#Hfy|aj9b1|LINLyUY`e+*VO=f%Zj6C zF-E5XXTSDUkr!T1i=I$H%ZTz|rbg>^Ow4A)^hBw*ch7-a8}h?F)`c040Yq?etSUEJ zVy~KxXgjGr)2N%NWx#nucF>D6Bk>r?X*hgJZ$T7@tl|nv7!OGr-5GhhySsg9LCD|Q9a0%^(1RYIVXDZ>DcX3bo@U-tWbbj8>@Rn?X5$LPx`=h^tw z=I3Ur%IP(mA<&QF&rzwfNGvlr73WN+fW^OS$g8I7(1TpKao*wc2x}^PsxUiWm0*jK z?p|cnIiJb4&0o83*7>dnUh6pT8--l;K(t?|mg(TU6E6z>JqkemdVHdc*(WUdqU358 zM+Y52+249m!`w@WyJwhRJO+pn{^4X$WN=D~nj*1@)}_i^p+Wr1+@{Qir14gONrql8 z5m8~jKv}Xh#}dL<{CnY4g|Y?zlUivpu{Y=tqC1$5d+LFR)f}}Ob_?kMS`(8h9pt!1 znwdP;^Vqu}@ONoi7FQnfZfdSQ<#*OA!cbk$T|G1!GA<}u%>jPe-*+`Uagk27^=)}? z8veO5v(?%I+z9T;u+@xX!NAS7N?D)I+Fyt}4#s>+M!DD{RX67ogWOi%=fHm{-V-|e z2ZahJ)tYb4^=7?nf!>4ttjiM{UMMwZ74x`DC2B7`KfKsmNzu)Q8J0niVUiUpO7_CI z>=DX?%}`ybl(pT!eBR-}3k9PSxn|+P9u23=Q1|8~ma+hU&FVV)*JQ00ZG%=z&8m7Ho7hva$rae<4NhQ484`COBXEemN{ntMK=n8E0{ zc_|^+IX8yQ(lhnHg^~NP)x{LjrUMc<{DSSw^7~Pld;x;FV*{#r3K7KhN3@gi9w6RZ z#FO)%ovlskaMfxkfxftvp{d*C}b7*#3hWm&l_Wk9|QMLL&m$GjD~AzbY>~WP*8;R zTwTb0>5TO`j1q_B)=Tx;pF>~M9lSbl;3osF5lo-&o#GUi4iM%#ZnUHN->$PBJA)v-# z>dLFfmJe|PNPnDw|9}h0KWRab&Dc6(F$$;8_)DF092FfU%vSY)deV?+Y+$x zh;#vS%!QK|pQ9cs;#*^JnKP!7Q~p8e>Gn`(N9LQPWuV&B$z!L8cR)M5bfD zPcvlLy`(UAePJZ@AT;vB2@Gi-0wc}DNv~%?7~A(~2X%6&rRNkv3O+$6a4}ESBmmQ5 zGIPLHD<;x1dr=x@e&PwDPP_XJLzAZ5x`@PgP%s3QmgHVl<<%ZFh(c5D*6BNX5SVt^ z{V?-7<@Q0;*}7eG15$_(onw7|jo`EkCX!E0f?Q~P2qQHGigO!cVg=0%Z5zgWE;AdV z647MlC41Yk*M%&y8`BtccZVUoHV#xfHwNj3VUc~`R}8y?=(X95&wqB7Jl z)EZ;LcI?o0I#%l8YP|=9_$7xa1z?TE#-pseXJC<0-$)B*T&y3tcCXh?W_TF(u-Aa5 z#I5!>ut(@UcCUfA8JXM*4j&55oD6q98dQ%tZIBD}=P(_z@o|mtUtY-_ahy=@?WM$) z8-ot>6z@84A6OsIU|P85Xur0BeMYR;0p&mJ$>9n)!Al&D+|QBR#K2CF&{icvI6YEv zNh{9eRT39}AJ_i`4(+k;d|O=1l8|owtRx&U_m-MI)?7m}J7JR8QbJ0?eCmqehIva* zWDBB)9ilQRp7rG5NSOs;T}~M?Urhz#%zHTM)3rE}+O@O7eEj4Lk({#Be~pJQJwQp7 zGJUuTQf?QfpF%Hkfyf~KM!4AhK*~*oiL$t=xevr#-%BR_40~zQR3~dCF@-8cX?tk= z>ag>D6l;68J~a~x%QExiJq!55ACBg|7!<_m@HId~V`!F?jGb(hZd~=GmvBNOlYMn3^?=&^}(~ zK|-OI0!+FiSVzUKbAKP!$`AJ?_`rD;S$JjGT?~X?IBmhrZ%Qy-wDWlZ_+PGt=A-`27(!CBB*a?AyxT>zl-z*Y? z;i<_UR1~L(vzT0Nqal)1v6l;4A?mv)ub9}|EO{6O>j<>jE7ng!LQ^(NeFjZ?42MN& zNTqT3@4kGf;Sve>NGL&vx7q58PbI1W4w21d`9{J`cytFbdz6(_Baku|Hg-jI`Y z#)mDav*$5yppfDr8@17;Bo|4KRZot|9eVTOiU%7_$XA*I3+7Wa4G{CAPj#J$jpO>l zH7KSkJTFZaGYApmG!u9fqvv}Ra}{HQgE#`!XsT%Mx{B~Fg*aC_aF}Pt1HqZV`aWq^ zppyoyUKIK|NLfxxe*1Pm_;MW@7P`IdZJh+wjbEdN(9t7q(Pa$)NIB53d zOJFiT-pvHluWh;GcZ#)-tb`}*zIS{WeP!+nQ}ZJu;j@c0IP_kNq$M^HsG~Q=^y8=# zB@D)OK0oDITk=AMcqTY-9;?FOUW}fzuSHKk7F=O|%&7QI{+q*gcW&v#X8G*tVX4Gt znq4|T%=k&raec?fiLrxbbQGAsuG$1v?=lz&gX}dmeUw4?;J?nuAtL3s?m*p1`=Ui# zucI&6fjg{;4kD@SSXYe zkr5*&KkPFyF^0U9p)HL+ z1;1s*qwSJ^>fmP_hq$t-!LRumHoU;Bwz|u;6Y+(y zjo`@o4NN(1Wb{)0S}Cg`*I17mB}1XtrsfI+F+|%=C#1anMg{4D5fi697tBbOhJ+6G zVZeO$H}M@&j=A|26@|}}`*6n8pdbKjNtffe9>Omr+jHv?Fsq31;sg!LFPc+GtNg?x zK)@7>zRX)vin~asCkO9Ne_fdU$3sHCridwoFrXXH{;=zyiUW>L{Ps*KFhS)@k^Qb~ zBiskWoJ>Q)z8lH0D?2iSH*s%=V8Ap@j>Um+r`n5Gm8gabxeb;!9ahJB@O~U zD&JyMDoEteir<%p*f3qRi^|mcUM165?HfEjp5Fd7P~LuQd5pF!o3Bn6AQI4J-Lu{6 z_c_ezWQR0~ZULjjE_x&Md--K^^yFmIDf*SuYEKl^UqCrB)ZZJQe3POp5%IO^fcjk>8prG|Yr#gXh9K*7BA3Cc3L&dH2 ztrr0NH~oNuvB>n^dmwHDa^WSwz*4Jep5WQ6k8fgn$^P0Wj{nd$uO-_JI^N(VISaqX z`uW0#VB{7VA-(=IR^dPwR8UxF;vP{$uR8ekdVZVYAXNDPjl#U=dni_!s&G_UVQ+=o z|08g3Me>>KJ8#=>G!wK5?d!Onw>J=1mVX1-pBf$dp>&|;>*tyVnYWTsSNB{QgDEg$n8(rez z7|phH__3k7kXi0}sI$L5P{59Icy<_%#0k77go0qXndH5fi(E?+xVkz+GL2!$k|3SV zi=Zh~(sph0Ax%>hsa80{ol>s`;+P9_+Jv5|t1zlRsHV{&YT5V=*5Hy0!};itqI6sk zCn>sL$COJ1L|4#Dx}D2w5Q&|vWo7pqSpk}0Wl>eIHAaNn_{<>{o`xZtb{Tve6@jEY zm`3G(Qct>cLkfa}#3+Vw;Oq;nbOfIg8DhOZ`!FyDV&zihhC6^g#4a|g3ogX2vv*&@m}l^Z^hfE0DV6gHS8GYJtFZ(0P8X(f6y9y(6pH8y5k;*!nV> zP!h9(8%URal$r&k#zaq5rEmE7`a9fGxBv6RxbbhNmyr0`SlzOR(sFB(EV=oRWqF;N z+fk{pn(N4}M<$@pbPw5k9Sq=9mbwT^`(5Xk<;ZnGxcU0CNGeK;_>)ersJeC-Qg_*> znWlOB?b=|;al%qJTGR7aW&viS>yPaG#X}8rW}cTO!*WJbk4tesv&aLxHrx zD}iQ~<|gMTBE=6l$-%1Be@iS`|94`^%E`?9|4A$-oi-)YwtwHKy@*<7#xNQk!e*Z3 z+EnPdy%n}B>1`Xi%L<~A4v+(*OLlBLwLTDZXMBu-3I0;{ET!&30zTxq;XB#NKFg1P zki|cFW%$;sj=w+H{F5I~+RYc|k8yEekuIUI^V=1-L9rb0dVd`8d(4D@89jyDO>C^G zc{KfkIt@eo3HPKkHmxt0jDaD@pBI850dwcS%^n8Z)5lARm{+Ac zE!!F%?4H0E#PZi2USdZLx?V&Nql5o=jzY&=J^ZgHNfFS|-aeQf&LCuqF3J^iF;0Dg z^{ud}i+=|3DPYGnt*(-Bd_cCQafXEeq~-zgPDsoCn=M<%k~LkvD|4FcGeJGtDb0dN zlB)R>gYyuGAf`=o3XS#JnI?^H3T4-eqOx>L4G+UvQ$}_3aWldxbv~y)3w@c5evUl) zgJ|2My3Tm!vD01OjistC7UWl;62QJ}xN-oT&JaG8vy9kNpK@ORuQ}uUmSCdwG?{Q6 zph(ZRB3s7?7`~0ru1$+UY!g!jVB5ak54n}e8$tb?fC)hULOVIZY~r3f+wA%1?UjAN zs%@WmkPWOK@*G+Z-i}~s^J-TOwbbZzWuwi2TToKD6*n6RIcN`ouM4ED>o(HaHTcWR zQfN@U%0gq+PLmO=Ij4II?jq*w*yL-!Y1`aC3UbVFPW~NoCSbqn-br^wl%`g<=( zj%;7I9vLYOYS3?WfYq?R6+qTc9tq3%aolq5aIHB8c_*S~pixz2$y!&Y<}QYgl{vCZ;n$Fc)-={i6dNV^F<8`nPuJ=?Ib1*y--08fSx8v6k(* ztux=XB-ao$1BkJNn;!Som~jeX)%Hu3*y-pxa=vBk1HQp;=?)Z5V4N35kq^i$_8p*? zbwCMoQgi{5_Ee)2qx%Bh5p4SDmG2iEYEY-cMVM-)u?$6vf2Y^~z266aNt&rF6ftlu z?wae-TDg9gNZYfdD+Erz`v<%WYjF?bFTw6XIk~U0SL(WO>9k)D=r2~rDa~%fuvs_~ZU>%?NL|aEOn!M#~Yh>RjeX4*3JUGuf^BCpnA_jSvaJfcR&NtWxX)m~rMePnL=DpisZR(( zKRLZ&oPi$FRl`^@H~b|N?BSt=xPag6OgU;RAuu^cZCkuK5mp+=teH`wsHx=L4eCwh z@w&T1DT>c7^0ROS2?#^RqYhMBu5&7c4r4S)&QU4` z?(Q1p!O#kIZrT8KXQs!j$!j+X;Ux08=%sL2_9-WO=6rqFExYlpu$*l*>EU7|>_an4 z*Yr7IY~o9;U7en!YY9L1Xwm6|qO%pR0JuMuNhC;4x+sn2y1eQ;9}v!fL$TW4i#BXZ ziLhu%2lIoa>)R=Ny z6=XdjHB5QOdwn0f{sN(0k1s2$hpxi_=V9tL7E@niyjL}uQ&#L1zr%Q<37}o)Y%@a6 zH-rf=MZGu&!j{ez-N{1B$e^NVwW6OG!2U7OebyQrx1@Ri^*var2S_d+JI2|KCJOw z>j3~)_KNh8ZuPT@g5cPPT~gdD>ST5Oa758N!U)=J!p&LmTJq1B=dp}rL5mqMmwtiB zz9Y)&8WjOEL!iWviSHt_KbM}l)zuQ5k54=BX?3kFui6B{W#8eRyZ57jz5H;dgjwdS zqF(+P6ugz+;>>%7Gs>v2TaUmekk}>L@@Ie;z#)RRU)`=a-j@hG)&JCrh=ggpc&{QK zS}QN^&ZmgjpcMeWK9De^Mxdp&qZSh%kjYA^+qE>Fu~~uqRh0CWUZJ z9(o~5WurJGe#tqQ&Ij-zq^$cr7zs26O0 z-=o#Uz-nfaZtDO*KHE|8B{O6F7+BJX|Mx)~{+dA+Z3V1;y}Bjy4L17KE>jWMsIzLp zj(X?T%4$ea;h%WgUR)iVAh^3p)!h%Gx=nM@xlD3tp$j0?1|AJt1}`+9ljf<2*9enf z(JzzNI^b)3-51Pqsy7-G%cbwTy@kvPl45E38mKC@gCexiS3>FH4&tEABpXtapYzI&j`o2*7pFKgzjLci@yBplm``_AJ5g|rpl>JcxWdYcO;P8_bXmk9 z8ZxY>Ku%DffY@ENW`3w0Q3H%&wIgGu+TW2vrB>yK30h-Vz$>bvTx=sU*ybi0Ro>jm zfKq558R{^c*;V%YlrObNCqiAs85Iaos0>-X%J2~=cvCj_*`9MTuK zFhU++4VPosMp<0&M%i~9d*F81sueI-x*EShJ)bG#DjvmgANQHDI^NJP2s07ukc38> z^UrX>Y;`;xo)RsG=oZ)n1$WszxH!hI(qE1G!TTUY`v0ECL^h?$%%m!@7ko=>QZB%{ zSOBs(NEU=0(IJGX;m}y@UCs0ixA9o#IB7oU+7dbr)mN~sS-mRULY%2RCY@6&l5Hag zf?JTW2Pm5oEKrO@>xPbAk^UT`-X@odN+<~ySTi#tdd55&o6Vu;>_)?(K`w@!JFjXq zsXA=q2w)~4lnM|^qKl^dJdfs8{x*O>yh4x|Um)f=u!CCj6s8o6p#A95T231!2`{eF znyu0pWNs|6%lR9b;&A@T8Lt_?i_nR zNnuch^?|h5>b7a+vz(ivfP{}70%yXg?y);2h`;+m8)|SRdye=SO~);2z!*0K@{&&J zLEoCZkfHB64v}ye@2n4x>C*%_Nr^$95+R?3p2@(=LyvQ`G&qXC!ex|>!K!sRo)kV( zP0WumAojB%9s_AsGRsduggIjXz)l#;(nKyyqk%x%HvcZp19b2hM>I+^zH<+LCC>8; z;C!yFK#ce%UgTdq+4VFk)-Sht8%rK=2VJN%XAuH;>2S{oPCe}>UxLA*$|@0Av0s8S1Xov^ z+a8nbZmh*+StIzNI5Mwz^=qtt!?w)(0kpFATIaz8VzIf6riw_#Gx|63vFDxR``B^8 z#EtShBroi|mITKH%)a&C*V|=N5Hhmr6tD2w+qpxDgGX0$IeABfM#QU{VvxOS(u;*1 zMo$HxE!F__pNP=-heDM%JIo30=or`tomL+H{+qznCs!?S7e}LuQjn-hng}subdbmG z%?b|L`82aO$`o3#(p0AFn+r4dTxO6@&b|@y-Vze0I z7cTK|*%t0wDNT;|1Yh+nlV}D-ZA@Q}QzV*S7&K9FI>QYMDKilHSkE$+=TX3|T9zK* zWy6p#+X?RI#t*+X4T9N)9`0S`ZT$ zzo#*2Zwb0+@2l*@)P?s3%_Q6cA~i~0ea94R7r3d*hTlGC86K&0`2%;(M3kK954=;k z=sF|+SR%<{_e=@D>P~QzWG7pGx^>gtihlZ7zI~U*@8)pt6bVbpw&Hw<{KPNWf7Y;I zdT1Kxa3`!YJQOnEo6GnA#nU{+gU?R`g6EL11`y%kOj)ag=S0v&OH}?DG%QB@9SijW zG|$X+drspcRI66v6sSO^5}%*HY$!a-GoZyDRNI+6vVMapyeD|1NAdElS!(e&ph-A~ZsNWu1bejzuRe@o@XGvt0I5vQTb9Kk-8a^OJ2>}*otsRlqbwz94!yismE$fMr9KXT-Qx9ifL?!4Z&#!LmJ2KEgl5z`|}p z8}^?^-!9)P0HgVxNH7rTH#67LIZNE(^Es=>r8Ja!Qx1#bAki}%w1yaWR@x`*l*pJ! zpN8X9yTX~Y(V@3YnkKOAx1Rf9FciTN)XFwNry5X-C&)oy3ZZDp-Ww3FaS=c65kG(R z&SMp4_^$hg3lrr!JnEYBrni;qPIksyeIl!_`V}nmNjo?XP7^9q%^@9Ra{A6>XTt3a z&uiA%1E4RLXV$0W8`aH=pK4tZv?)qrl0I+9m=!K-RHn;E0sK3Ql9|Z(sW*Y6b>(q> zgy;e2Fs0&8@-5Docm5OF>Q-&Ago*T{TAO z$i7uC^UrlvjlJ~#Gh4ot&nvG@wxb2-v-AO%BsBq}Y2A?!;U@d@N1bqasJ7~B!DU)e zl62>0rJGR)R-H{54ZFF*(Q_A-!|tzg>usJ=fCGsAa%p62m*hhLwWFCD5`JmXseO1o zj_yzuhx$avxfvQ(tj;?WJR^cQi`OdSfNIE@)AVYu9@ZO;9EVt9e(^|&TB{5cSS|!V8V#6ZzcgC3(2ROs?W2k>SyaW!`vs~ z_vsH^s@ML{@{XO?-`$%|{;1=h!^iF*RJz3Fm(W)_X%AruU(TBBp zMVwVk-_u#zuc)-LMGxju=JIayV91Yf*A+kY`qz^NQ}+u<{^ku&4R@q)z#;MQeXNHw z;t{`oFWc?*vhK9Pp=!+(;hNJ38$>gEp0!)v$SkoQ!n~Q*@c^# zbH^y#G>pt+44X&$E~7-Vs^-`2O~r^?GJ5$WgNyE?M@!8`m4()2&b@M3zL7>zRDXFP z+0{~IUPmg5VQMLj^s@>zjZT|-pNOLEy>cL7+W`zazU|O#te+ zWOKbjBiVOP1Xmp+@z}sYjkm;JwmzfOR9~0%hoFSYfYx(##~O$cJ6kSA1e|Qz(NCpz z!(|m^ZexwanQe_1KPKj&Z(|5o_OImSpk|dNXy_Bj18fW#x5l&)l3Hq|ks$vgO@#ZO zR99S^Bf#cxxie0-tc-zihgBJx9Fatc>@`KS& zIrG;I)p+{=vQ>q^(+lZI6@OWBFR~&NDW=aVX%3;=Iu>_aih!|?KP6-Tu?F)~Oqda* zauSherc9;#AfWH+LIG&PYk;5*Sq8Zm<_$)$-7-!tOTR5OCjJP*DXi?$nn>2y7__rZ zp1RHK3{w^X;P2T2lg-74 zbCEEprwTeLX?Tp*065!X(dfe+z1E$6nbXQ#_BP9l75I^%QyII#lnV2H(<*}e<4sO3 z1L=u|PNt(TdtBqIhl$4purQVSZeJnCLl{X7v=zi*WObBbqM33>*(P-+*pW>Ux59|m zbGMhzIi8JJ0Qb2=7d_pT-goX9?CseCNg$Fa=l5Ouc(}OzxMT?JjCXw|cfs+vnQl3W zSWm+PhULRVl1ELLKd{FgE}MtC>mLpbUYyxq1!h{J zAjpPK1Z<)FT4>n*10Z@&jGN7CO19m!q=s!a(hL9_*v=U=;ciuKTJf^P4!&~79)M55}0uyRipFEkl z59RwLM#y&SQ6HE@9S>N{%?fL-M?UD@vaJZ)Oyz6jie-59bg`^}knN3A$)uBy=22G5O{WX_L2XezfJtW+i;^LSWjE z?DPKavJ9)rsUxlJ>EiDGw!ltMvD8H#DF;Icgl>n{0F%VY&(BUcpyO)3xCRb@$tXyU zW^!ilPn&*`DJp7#oUqEacLSWAgWp|K;|_F2Sl_{L0y-IJ!NLQ;gY=WiQ0pmngMR>W z*><*aHP=o^wrEzP%pcv7nN!czZp#k7s?yuJ+NFS_Ues}ceAd6F`R(qCakaDz!t}5t zf;O?V)xE=&h}@Rw8{7|ZX{ZVu!I%f9kVav)ta$9*+~fYF4p;=Xam2PAS6`yR@y+eR z3sWf$4vx>~4G~#T78q0CwEb@v_M%`=5pT{H?yOVJU|4t^#;2}B?;A6^2{Wm5+$bxA z_zD6EY`7E6mJIC&h>ep!0YtValGPc!Cq(CpZt|z^E8X$l$E zN&pQwqoi>$_Gi61;%HKO&f{dHCIqC|^mT6y)68gzw6YmdNGEN&P~4=^4CD z>(rY=bNjI1v@R+)!c}f~(tR_$h_m2}hzQu32s3l1$i%Lm?OZ%&+ANF6T5=(*`S^tgjCK$H_a=wkgGB5L;B zg-CLeJqnO9Z?;Uqe_*hG^rHQF+cLoe4{QOJDHAX3in8Z$xa3EsYi))H7LqU9!u7A^ z%2lIkm>kUl1!~x9+toz{r@NwB-2^n;ozuhe9s&F?vf>2d<&qyJb;vS!#>~q(!U9d_ ztQ0TKWqJlA5=;{W&i$M>Ga+h%_zqpkMXFXTzi$4d!fK)+ z3B;P@92_nTX^#z!DY+N^A*9?)b{sLF7*FQ#(XuDLm0J-XSL)-$v~lsy80XJXS-R)@ z(OdK3vv$zH>1aYN3=c|e3##M$(J2`-nDs8nR4O)~LWc&`v`S7~URIf^gX#<4Ok}d` zBRJcnShAc%lvA!52JuPf9hN+&6M;&0+s?X??&`64K_v!nY8!#c&2Pn-IvuOn9uy=7 zwZF&8n+cXdT~|IIsA@Gf`kUpI8L_z!SeB~d;ehr!}6iQ!ISmp2bBJB;m#v3G=qCH0o~xs5eOOt$l>1 z3TwH^&5ADsmg7Dj%Z2Mcs?SRwF0*;Ibr2<9qHcdrhpHCTHT6$ORR~W^q5d*(0KfjkR`yWTP*ET7B-HiQX;s>3?W1ga1NcAUQ z>k2y%CQh85EtU@bf_NN0A21+0R4fo1ql)FB*F3MrP*IQZEaK1%hUPx_{Zj?48h&NW zgoDsSNz4=}0_s;*BNaqB5g*71!3Y{b26C#gshu{@h<+BgOVaN`Sr79+HrJn=Km35d znse4ne4Mk33(L-$VyB;<0eP*lUEbK(aRN&zwU0OLMZJOgsa^rQ8j^^`%nO;q?RcPof)fYl>SVmuYV5b75Ckar@2>;@IA!m)tHAP2Y$@V8of4OEV&`G}nR*u&r?b6(nzH@8K z1Cqa|n|br9%}Y6ELdWuU!*5l+e7X{#CRdZFh&snggOmOH{XM-S>Nf$;R;^KgW*y9I zg+Lx?*m;1-5%>e&!l{uBX{?_Kf?(0;3hII2c}S9~NI3?H zi8^W;H-I5+rw{vgAx7?i?>YnZjM6ThAfFfDtZ8a4cjW|kvmshD#0g;`B)3O{FST^n z_;Ax)sX7opMXud~Hlw4uZ@us~-hu?#aG~=qPNyrswSUxH@g#(5=OCOf$Rl>Bu+yW6 zM>uWBXnerhug3%N`3tC5foCn>D9kPtlUx#>S!&jdfLJ<*?RL@eeok;yCpnd8`km@r z%tRMaOXOy83&tzKE$>Bo*03_{*-CK!qt(wP)nl&_3KS>{;~xkHOkBF(hk67`L^_}6 zl(KIYdWD7@MTxrg9%>+2W&j+ZDPhGT(mzjHr8^&i87p<#u+BQ@L3SqS2g?7-6H4Wg ziR!-CHtKBLjuIkb4XE2oS)!{DV1?Mo4uZ#4Dna58gl8|jv$`%%GMg^q{}N6JUfbeH z`3u~fv~{d+4dwFdu;7+x9as+0?Y?&ZVeXAt%N)*g1 z4tL@#Bip|)0tchvS4b}8EKgdzy~vxYH%@r^Y(A(d7_Vz*f&P6^co_bZ!G_;rB9_+~ zgb1Uh8kg;+P9!W}8D&cy9QupBi{H|YN6@hyK|tkUixdMlY((JIeA)o1 zH>$%MiCN1j2r%cagPtOfeq{d+tIsG2DoyI|R36;88S}?Z2iy9UyF7i)>ga{>nH4YUcw8ni_9h4L6R^xo-Fs1mTNW@dY5;bc|4&-+ zk+OXaZ&_Ws|74qj(%qfVfkJ`k^dD`PYCNp%+KDz`F)+8k14|KHADQw zlT48g+2eFatJ0OK2j^~Hyls?FuZQx?e+A3ER3=7L-fPKh0-qLLt_3xz{NK zc%gYUQopqp9Ip)-hhg2nz;&}83;!tsW&7U|C<_PU|D&Jq`oE233CDlNa#A-Uk4Gi0 zHNvyldCBV-5^2oX+Pa4|N1ttYjG|jm&`oOUSF6`X5qndm;x6L&6tIQf=9rXX~h=B3j# ztuNZM$U6;+U_{io{><<3iap=sLdBX!)}cLHqZaCeMMWb|TK_`Q={Y2gf&H)f{pq;M zA6h;Y^(4yfGKA@p9ky#$oy@HO-J!$E7Cog-nzWGkcL%?GM^g*E*;)7={yB!edGxxy zN#Jyc5wUtUr+o_-11IbTAvd>FaP!v<%|KPdgqi3K%~MT8ppAF)`$k8m-$@fI!#N$# zdtaqJ_(~BlYKS%x>Y0j`uTht5QD`Y%jP1>-HlmErDm_`2DIN`C*GWh1X~SwTI}pX1 zRn_*Y)wJ*GIN2z43ngMB`QyCPmVfE!eL`YEd%8?Ean!fgbj$N89e{pf3 zk27K8UbcI4YztI~d&U)7(_1VpSYepV#eoIYBh6ah_bfSR6-DboiBJ+nSQNWKF{fUh z zeCG3L#jsnb&>&vTezlZy5XXGg;#mmCbyvWy&G2GNsN!W}Os3lK$TKMY4d!guoz&EY zYz(l~>3D|Rr2JK|J8a{-XcnK3>&Ga8Uq-Y|rwXU7V8e#PlU`l_X#~_+h(XM;?Qjmo zk;Cnd`mt$qzjRgv#h0PIrsu^F=H0+YgOxGx+9$s(RL}}sfmPb^vBO102#Ah|d1A>} z6<}bZ=t)o#5A9Gr&fXhj4~2zeU886_fKxLojD@3c&hy_KX|2(9bTY}i(u!^G4zbdp zN}-{bwr_q#FJ~lrUxV3YxU6TU<^WJ%31hGaouNwd+2c$>o}4luoATi<9h)-pTXs%f z_}U@uHmd{DH{BF6t6}Is;HkXo3!8S#V`A7VH0Q8|N69|kK-Gg9`4UdJBOSs%Kp1z4 znu1@fOk$geVUa~B|6a9Yt^@RVc{KdK)^(eFWqTG=kswG5N@Q{9RJ4|%ki1raURL!u zGUYh5F=z%+>p}y@+{+7T9GI3G{HaKC(gdQp<~B1=5fj_=c@+d!nFMTNs~Z283v8jy zbaHBA|2N~hNe!b9+^r8(0Zy&5c)oBeO|vGZ)?{& zdA$=ue`er2A>y&EFCu{KxC7fh6&jGe?rQ)Q94WgvFDmsr^1-x<`LX3Xlxvmv(G`+$MDqxozas$ zG%pCLf7M%g91w)z+8MzFBgN@StKjjlfShQfqsKja=fU9pAW=&B7~qSh8*phJwJ)tf zZaSP=Z*IXPZ6*OvvGGB(K%Bgf)v@uTY$St$ zA}k)7%B;LnNr-VH-n)*;AGRJg_lM5g3)zUw=)M4#-VK_IsO@M)P_O2)b{iPDB`g2@ z$hzT1hJ*70tvJuRzz}~qnN*K;wdF6Yaf!x_4x{-plN*uEx+)g2uV%e^ch(UEY#J_2 zFCSD&+#=Iu-k1o~b%O4GpI5?$@>B{_#S9Gj{MeHTrt=SiRC+r-{rg8XFMGW9891*kkvRt9F0(;Sou*0U1KdWk%WcW1@f374(_ zTV%>of)>y19UVK{mS6!KXoxo#Z6a;Jo4_|E^~OVsXpV$@ZMTYNx&fB8-HJ~@v!mNE zlF?+LB;k2-;^jRb`?|E{j*XSZ8$$XiMi1U!7;NbeRHu%mW6Q>l4V%d}pC#I>aV>K$ z%_tgKEwG)Zuk%H>B|v6=&IwKik3YnP5DgHVy_4HLzgH2`qy$}fSd$-#PLrcJY5;!n z^2lqn8g(1iv7c4vv`7?g^;vQ(*QJ1*UElRd`7EbZ^6G|6M+%&0ie5`Qxi7#WJF?=e z@u^8CV?tn%?x)VqJ4ae9);doA;PNByrwiHa-%8t4OW!%hFPk78cRF zhbCZSP925s=Feu(4Amv*$DwVJWSB_Lea$5potR@gfai0{O1YU>?Y7A*26nm_iC_vp zl_XK!AjUh9U&Js4x&%Qk|4zZ4F||8R&_lo=1@Jm0BLw&rQ?P)Q)DcgLS=Bztk;~e| zddAPaJ{m~KAxvld#IacBg+;!QfbLyl*wj1+Ij@!tW5hm%QG#$bOyoln>{v-4QmFdS z;vF9rqJT2%J77@*%2t*L_9|BsBt`~M z#_(mFB|j3}=ui`^{Z_6LL@e~Si_8azp?(p&($#?dkr3LxUu>3|=78K6+5!eLw7>?|)S2d&ymm3=)Iv+7T*cUjK;Qf-N*Z zf95x%{rdilie6YkDk68gg*KAk;{;KS|LH~t8Xdrh?_4q-R$o?zQ41p}X_?ZB4 zsp;)|QE&uR;n?Mg^hv>a3L)Ir#RcpU$N^WIKR8~uIEVYvz;v1X@k#>JN(_-10%(X9 z^Ns$B;{o@bFvH7*%gd|7W$)51T&gdjm(!qIY5)UD{aL%(aKFYS0My7d zma3x+<;7m2rAj>)2!W0a+Zv8~En4p;vmCU{2ZUGB`f9QeAb+@X4->;jzpevp;b-Hj zR-?hGW;u#FerT}#hbo4&vJBVIv*zT_p$31tR1q564_R?+%Zom;NU`=-Dfz-i_LJwb ze64^MK~0-#*VHBVfvXM?!X>jges^T7IKh<06DwT(4icDKbYQ%Zy09BH_H`Hi_%}{R zL!)g(in8wBQ{u;v4lf&&wV?l<;uHl0(%Q?=#@dLoN&2v=2SN+7*D5X_vGVylT=B&s zsXAcggbMuv8EXeO6g?a1ulwU3wg@EBaL_Ms6`i2#aC^VJ+)&d|F_`s;k!)=~J!(_H zj!H>OnZBuNYfqz&LL6+S84S#l><&UhYcgv;gEJxIQjb{)F2L*1OX%k~j6oF{rjXw) z`GBk_D4xL(WsV3%v4{!1#b5*oq%g7Z=3hGsxa$B92w1E$>Y*{ZtfPjHvP!2F-S`F| zX{ZP9w5q9XQ~8E~n@?rFeL~;2qwKsl6KXnHzrs)rA(u!ja%cao#tUblDTlsX16uYJ zksVu)M#^YCRHEs70)ZFT#I-jEdpnH(*H%v7sCNrU>qFS$KmcntW-pYO5yXM~j8$hR z?sVas5K|e%$Q-(>=L+9?5=}{CY|f&5cj0%vHe&}@{q~m<)RBHiS20yqDjhx*)r1zs z;h$C+iSrduVJz{m9168>T9qcrs2vRj5M$16E7An^SOyWsbW63KtN{dJG(f9po{1QM z4pzspT`%IeH<33hC1d4i#A0t8zn$D~WCB3iSfcW;Us;{jm|J5%^S4i)-v~}+TSi~^ z>w32`bE&q_5_c>k2QZ@_&cQ9~{_-I->ZIu~SHkwholYo1zoz?*2ae3Z1Nj3>+-*&< zIhnr7d@NBaaT(u^Z4aT+DyF&*vfr{G!Jj+uu~m+(UPg<8YQnqBtkX5AAxlR&Rr#*$ zwx4`5%2U&>6X&eM4{8G5C(|GE_5;J&!agKSlfyFH?eU>vvWUUwT-*aa>RP4Da+209 zt;x+iNK37sPcQ;k7B{`>C#+(a^bbOtKZ4HPJO3|D&M107wb20kiCg?0_bqq`y|J)5 z?W!|H-q7JQmrqfkct_`C1#hhUy-b01+H`2HarsHIRbxqQJRnnrCZ;Ct^&OjZ_5If$ z@Wx*+r!Dxqo=Q@+OS4wOCG3y9gC?AYm##~Y*eA1ByKS78h38$@zLI(}eDzLx+;NBO z6O%j5^_FEvOgn+8Pe9II=)EsZ*r6Lzv6>2Q$Bq*&U=wTx(%eUK9(dnEpraeFuHLvW z1Q-BFMe+Bz&}2S2(C2}=KJ*lj94vrPWbMJ`x3OxJsfzQ=F4|V^66%b_dx{2ae-`^fR%%IStzO5BAdhY`b%zNOf6$t15Po0P>1fa+bEmj+!Uds)PZSc7 zg!`YZga$Tlx824dwUjy$QasR;|)|FXf+6)0qcMdJfBLU*aQP|^DtzjsOuN2$Frj&B#`kN zqDf5k2O~I)yVP8MQOH|3AN^M;EE!LK-_PmrpUB(D-R|+{wxWMkfNxwfGxnQq_{DDz$Es_Gz#^Iw;enESHkE0wvq(|x_yZVpZV5}j|<#OGz!B}|UM>ED;@#Uh-R$Nf! z#Ag0LkC7notFYq9O?j5G0Q&Q|fRcF~$PQD!Jwf8jFWP8qd3_8t^eNx91Xp3JhEv-& zXpo-H(i3{xY=W`s2`-Wn42y*q3U+sxqUApT?ni$5i^SL~RrTKR#s`3?weOp6Bk__4 z|73Buw5hJgb{^x}rK#yC>|Z|=t%Y?Kmp4mG1{t{x6y<-CbmNlzZV{ttqxczk^;oDK zlXXu-HeUPL+NG*ueHofo8$fr{_S~p!odhM?^HD$^2HiFG|Y zz?O@ScZ%$LQ;XEU%pnqU^@X%Y*lud&yLc|Ax=9*Ot+E z!bilH0T8?4sh2j^U4(Y0@pw>F&1YotSM^tRGf=H=A?yaZsfb#8Q1$y>h@bo)!bU?Uk$S7WxbfCZ$Gl z6=_J<1kDf>wwwM+x+xQ`zij-v`VAQU_T_k0zRlZAB-KB)VHgmxRcffV-f+HdX&>*J zOVXG@e$#)swew1vv6J`aLj1U({WNaP|Jn5TaVK1kktb7HO)C=-G3a*?}VK-q_i(54WsM zn}$}a(^)xf=v~p3nyXET8F4u za;RAE)tBZ*>NZyxS(;Lrdg4dzMa`54LG}?4%>(5iphj(eKL=H#g~zE{m8f?{JRYa2 zN#CUQ-#4B)Yqof3t981++kT?w3=2`AW7h$dV3>~-N3T0Ro4=$EA-ea@APrdEO;3?h zdT=umpp-$Mc@r#5}qCt$(xw44Y*%`n&a zINWH?-vucgZC*6a>4LU#Pl6_7z+bl!BEX#yYUmdVJmxor9tws==NXedh(Azf(HnRz zAVc{bFPe7QwR6(fJJrXtw!X^A>G99A??F`fz3}5U6aGC|Ja~Wh`<1_s__Po*OukiH zN`Dy@ot>Z+E9=}L^UUy@@VHf}uzuRqDPaKP(|$Ja#0Rg0VAZN|w!HF0@>4_TrII8L z{4=@TS)s>75+R0MS&*u?NF?%9F?Ojt9wlee)Coyvl&C@WVIP)rA{($yZSMYp?j+jS zqM?#)2i^B~2XUgXF>)kwqhllZr95T=_dBK<9f|r^S#dF#B%wH5ivj30c4qgH2J}ig z+fr-8kbW!--#8NlmOy~nH#dBhYhd434rJh7855$kadS;JS+O3`^NI}KUol3M%f@`A zq+Wqf$Sb-5p9+NRZ8zUWUfyoPwYxVqZTZFdlDF2J#sNxxbi?DE%qaKQvnNq(ly2}T zIx3_wN0?#^NE;==bYPw;K}if{Z;dcml*c@b;LgMfAPy_w)UoF*qjc~6k7Z$8H-6G* z(Hc5beC^|NMUl8r6oGSn611}q|ELmUP%_USbv_~{QsZLYqme`FFYz&-+hF{UMQfff zrM!@(p{NQ?F9SgEPabcv$TLkGto+f*Z1CJ@KI=#`B%dS8a>o9s9klln>kY?qcJ0#o z#Pxl%-?h}thG&Z@gPagq*xBDrkI-4Orh)B5o}PTSzYv5SWz?b4P~5$@0m{zVtPIa~ zliwwi=H zj0T#zhu{ygmIeCIyUBUs5op%6w%4_CH9v=^XP=vBU9S;^NasLWdQT=!%F61{LXe@OXW^J!A!yf0}&^s6Q6+2T4PIv5e9gCNpfk-G(W3JC?2mbd0y zCt0)(*fC1HXwDo|Z*0FNFhiM~!tjKZ7b9|MUn229M)PhX|HU3Fy~iQvB!&~be)=}) zoIxPSGF#@$FRr^UFKIa20_(R>Idm)GF`Ms}4O`Q1#=F_M**?3lMG{d_<=u%HG1ZIN zkYn}|ntMz0HonR%Pw56C*p`Pq^LZ(E&}@$p<^VQ@jjM!kA#u1EhXr(hLm{KOWW?m6&fQ|`>tO5&z7zCLAeNorB26s5~N`4C7E zr5G1jmt`0a3ki+8NtSw~8(3$2wqx!v!qdNfW@VP-B0KE9`1_TLAi_|8()1$V8K(kc zYiGJ#e7wKEx5v3)aEGvPFl;YxB`o$$5TN&Sr{4bUin5*1nkkNme6s^76JAkTVVoGP zVnrpwapjyIX=2VdC-B6lK~8$}2Kq_}ZTf(ClyvoU`B~K5GRE^R%mc9o6Saq53Sz69 z3Q`2VNC8%Apj^uILwwd+$9ooLQ(1tfZ`7zv2vw?D##v?aoclwqSrE5Kd?D%OK)%A^Y-gGVNO=H8 z4WIokr)q*}xaQ)?F=YrQKk^5Tw^qCz`bjJIHf3-YZ{n}ygA<9{p|>WBWS=6`BPh;b zkh0B3S3*G77=>MM6Y-vFI_m{MlWvl~NKx$?v6n>!CBmj3)WHx?lq1qmVmB(>vpqs8 z2vk%Zt;v-^20z$-vccr@IBc!+1ZwtPhFQzQbz ztF|!4W~J9-`a(w}h0oQ?1twN*$(PFRoq0{)+l;?#8rVkUP-3;+>Xf_}Ek`my#W5+6 zDll<5HqBU8V)RX&B;#{C^1@QZilhmJk8$wn{VX^tg6Sh6sp-yZ`{cOhn*$RvIPyBr z|1kf_!;RUom?|AGQT>E0h2m$OlLB*L|k25_vvkZ&%_ zR!{BIHAZ|I6@tVg+uK*Y&~6ohvo%=T&^C5rkRTVFE^Dwen~8b)JeTX~=Avb7_|p{o zbaJQjQ2zbgWu47N_h@sGRl`O(E7V9MbU3mQ1=l0Txzs;6xpXJ7_h~m!)i;0nd!g#_ z^+^0KN5AIrGLnZO&Q4vYLNcb=@+qz4%}1W?>(OVf2y}J93~MT53^*I2#zje1fb1J; z3?tKJ_`duhI}ApdOgNOQh)>8~zNtO(+-`JzCd&P3xaDn+pOPZqx)gd>;-Fe3%%ZX< z-V9Oy4_wSL;7!#%>o>2g4Mc6T9Xv!$Ss9)yK0Y=S%LaV+Xe=Yhw}$=cN~!;bNKzij zULbrA!E9d`L_$fNXRD435=6a9nW9QXuqg+NjnGT5OSfr(GjVu}NZH2u9RhM+kU389q zU+>7j_N;ZSCrw8|rm&Jb48vyJ*pF-OKnhtQN|fhE3Fg5l=LB8QT~V|E8{ponfAgsk z$l*s&itx%h@~x4W9{?tDD1v#vo1tyT^0wyJH;~vgYb5{d4)IVXJSAkDc;nAg(VEAr zFe_i&AloGv5I28JAln^<0Pp=5W9Qgh3ADA_*tTukwsy=;c5K_W-Ek-B*tXfRZL?!L zx8Kj_RNeC*R;^hz<};rGhSCt4@NdnI(X|5#IzOcHRyX1BW`l#aI6uEkVU0na2_AcD zTvU`*wTH9jXGTCS{%irNiu+eO0Ocg7jt+13=gpYTgxbUnRL&5k@yJ(lXj=nAqAKdS z7b8h~Jom@gI2pr7pQ%F(P#H1`w3_vj=0Fyz_8cEUvNDmp5tL0ZDR?X1IF>(OwTySD-xS5GQN~>I?qu5^jLrAp^ls5OW2Jqjrjr4>` zYrnhcG%=#|bdvXK!npa(->I!#+yiJ(7f?-0cxk<$9S}2{sVWJtrW%cPxJaYwg$nv~ zpjB&6eO1iAvQXZwkKyNkQ8UzhLV|x0Uidk-Whj)aZpnOCl>jdD$$R?>^P3f7a#wQ~ zilgk(Q&DBb%=F?bK66eyy&xB`rw9w^#?-mb$!C;FJ%kzLaMFdac4X!V+0E~akjwhq zWNMBSO1PxboG6+ygeQRL$k2ZTyv0WLj)jVu$aLK}81lB?Q1%EzaUs}P)-g7I3P6nt5PsQhb!3)Ex6;WrMxRQu_nhaDTzE+l8z!aB>87vkp+C&!mD<6wS)^yOsKM}r2Y}j_D^n_W9;9On%6k%LtB!z@%8oB&A z40nWV!x}eZMuTlE>H4d+sHl)r_%3&6Su`RALC^T}A^LioqsOX#-60sz_M{+K1KX}< zjXKH5E=VVUdoX-|zh!!hKu71~v6Cw@`Fy@);P)HM_4~9|nc>vmT8!}RoyoP5ywR6B zuysbbO@?FusdlOPDjx2Gqp-x;^)r*r?KqCU(PMqZhs^KrEc8OkjGJe+dIe<8minkmoCfg8 z5O*5wbe&!SOt)>Fo~J{_$-h0dWBdqs5Gn3NUiH7AQl-Nx+_TI_ytW+e=qSIx`Qlv|mk4L^^D z-vT%KoChSN6W`s(sh=7o(P7NAgrC7<0(a5dK<@#n`T&=-)~j@SkUUguAw0)X#Uyqe zfe{ZH@-;Ib)|+Vi>KM8Na!y`JcYicdbvW~GHSnzV!TiB*a9#X9(Vh3q z)4hfKJ-PBevf>kwu6ofqglqIvD}4P*^zRW$@Q^;SAE1>#em#08`~v273-M3 z3WbuXm$MMwfXq*;^$sLUWnr>S)2e5>zQ#uircsTGb=hmeH9MCG>6ExDWE|!zJUNO@ z7-`qUqZEMlg@+@+2N*m~_xI43g*!>HY5MTiMIco})!8^;VSBJq4)Fq>zr$uCdYzKC zHMW#F0y+s!(z&~R``uq@0OZ}k2zuu#Xeeq2qZ<9-X~lnCn!Z5GzbPcbcvH8UKlFeI(U5Dfj`)%eF+#`khep zGaS=wE5QvMks&@|kST^s5h(#)V4n#VZ@6*Jeqxd`GydFzV}U?QEt=7scG@h~%;3&@ zyO)^{??_|-o^`PND)u$kDWi<%dg3|sUQbx*7dgMqPq#0&DtsxS znQQ3&mmq&$oNRGj@Fj)Ui`s;)Z~%jr?|?#l*K%Mlj2}~pkxoiA8<`472i0j zg0%#Yx;n|$ZoxK7f?&-kvRQ;^*VBvhe?4&IiE9Hv{f(6LvyON9)n6PMX)@Y+EL1Fi zx3eBbv-DHEb<=}>ApD#AQPL%h{g8Y@<%qA~b-%A#s;HZjLQHUH4=`N6ON84%t+RF2 zy%qUEqgTGa5}^^Wce9pd>t$B(}CIsWAjobwM7P zeh+_n)G+}?hn=+rTt~Vuir_SSCE=ZNJX*GoUoU$R#=;?~I^NHgeW5iE_6wnw$-oO? z4INbR5$NaWCC^MLdliQFsY2R8TSY>{u=JF%8_MUi)1He*7X^ia7|w}F198YJUju+I z!WHmKqqhV|@h7l9XG6y#;e(YbAd(moRqD34gchS&#Z1YoK|;kY#&+8Tn&79%*hp1DoPJth_@*g+|gzM!r_ShWR{v>}PA8-kQV z))k#ooffm+K3UEIEoeVcDN8c!PjP1%UtX}W=_eR@=-7HN zryl9dI2%IpKN?hk3n=l5NxVihNFfX)f&8VEMLT;2P)Z#C)2o^P03i|W8Z*uJ&<%dc zbfST}9VoE1k~!2kW{e_Iw^>skM%rX~#dXy2VFt7AUJc5<1JNfVAk){M={iD~6GV_z z;P3?ghKd&0nDC&h;>l(RTu|LKc9kOWoa1&zS!fcH4pMo;@ww-F0tx=$gKgxcI7saZ zZ1w74c+(_n)5F8T@@}&20H`4sbf1kRUI1_v*9NY_SrM#EMHbRiEa?Eo)ri^m$MR}+ zu*)R+1%7aO#1{*uC|wWCyj4^NwX|?j7-Es1d$H4u;%I_yLbxEhe_nB_4NsZ<2w>Py zX>dj1`hSk!oXIM)Ma@quhRg>_k2*$^P?$q_Gd}WQu#1)D+&+Uy2#t{wz5k^BNJ>5M zhWLY(>L;s)o2M{mKZK#5!4U`O*HS!4K0&k0G^Tcs6_F$fGp7QVadHMHXsUsRSpYX; zEBZ8K_)QyC@pIcDPchF~FliZGj8bpS4>4ZJ<#NG0rod9?CLwUoa-%)Xcx@w|sV2Iy+j7c#6tGrWh}{x}-o=IKFT(5!bwmtk$&?4Q5j zXYTmCNafLdw~|RnS-1v7*n)DRWvKJ-Y{wD}$^>ozdwMEo6-!an0u2s~5iy|dl1{%-xOs_P0eg7E`VmwI`_J;cD_YVyPsL^)`n$ zg;?8J!U{>|kdx}K`yLZ3L3uk^7$vx0i|F=gEmII__d^uC!f`oR!&O*X_K}r7+1WsD z=Gz)^z5nQ5To&fd0Ai4n8{B{hkV%{i@;n~unsPDVBCV;u)hA)ZAXD~88QHBzk@@bH z$Pc&#Z0q20sbbH}AqaH;`uUg1lVWwMfv9~-5`3}K+Q2M(J4-b#0HvRxMSz1aasnA#fgSt1z#>nC^XLSI zrwnIM{iuGq>xzMUw?pZaq~fDKM=xY5>eqj%gT)-4pVWDEOLezG@b+Qg62KR%vKVT> zUk4tDNp1DLO}_wXiC};5e8P0V(nC7w#Jg60*6#3FiXpx9%U5fS<@fqt;|YjEYifU7 zh;7#1LIlB}+Ue=FbYM8G^6PmIJ*(@Y;k%3`*X8m;Uf*_sMhJ;&CSj%j{e7AyiU}L` zz3?B4IVT^Qx8j7dzd}^va2l~sq}ziEzqv_PT}|DSYb}x{wwsQObtF#B11ja=jH8sW z?6mCPO5ZcwlK3+XNaD84LB3B?III*722@KzxG1Hpp9uea9S7)EYT?=#POZKWK>IA8T6wfIZ=tLx@@eEdkJGmVGR2XSAC zff^w1BAn0YY)lAhD7gxpvFo8tqfAy627$EjXUBN)`AvKd!`Jkkm0kpu+52?Mm+Td0 z71PD{!WVNvzA0@g)CAhvetPUhLGQUU3wZZL`dSl$@02TYUd_%)CD6)v+YmfcO9<){ z7WGy(KV|u&mzi9lEAj*6$7(J1X68!}`fR%}qB&+8HsZaSJ^+58lhq6f6jv#BgLNt1 zC9)iOcQ)07Zk8bL^ewA>@$_b@-}r$z&~&y%!nspvP1kScyF z@$}JArI(>XL?wp;A$461s#Y;HHU1aZQZgFiL)DrU+uJ3x04#LY6)gHV8vQ$6Y~i|r zl8Pzt`NAv2f%EJOHs2Mn25p-9{L+tN@fy#D1Ux)0PiRaabBjd$RVvH@8!7Yb#lba7 zqc7bV%n0z9|0$G{7|kHp^Jk~WdH2D8j?1I+7>fO35ya)xkEx{mZkhW}(@X@Gm`*fG z5p9_i=Ew>49v;fE;L5x)kUCHV_!Mqs%gzm!uX)4Z+by%GGD!wm??~XP| z?c+Q@u#X)Ab>MR2i|Dk_lg(ek6tScc0nuW>{6Z^ z6<4LDB40ffe;=9cWHuHawb*X#?>1Tf{jSCyW?qall>f0_mDC(}p*a!T5_5q?)HP*&wuQUO z@UbjJdT(U8g^$+_^7gqpyPs~r%3-z@iKs(L8{#Nc`D{cEk#xUTQGqMe70$_UwicpQ z58He8hq2$pxo3B;-FH_sb$<^DwOqnk{h$m(GkR`A$b3MDjZY_^=SW$$F0%WfumW

#fuohg+wI> z*A2m&VdzmE-O>^1*E72lzfO`Ns`#Xt-n33rbQ&62om4@AN6B{;fu){B+n=z6M(>aR zaKju;hOU*732*Sz_g_ALSNVazWa%h?x2DX4;5b=iS`3@e9g}P4%V9}o`nin4qLrVG z=;{G(+#sc8wKwNF`&?(U^0HA`A!u#7TaMspx4EdeMVnn)=P8kP_=a<-x3LcI2G1JQ zWN7ZIJl3ggL}s^Noy-(ZP)_=$jZC-rH20b*l5P-jA>$7|r>BpBI{sS;EdDe;wCVF1$70H7k;(Zk6Z1GxDuP0tO9 zcwy^~b^{250J5?Gpb5YnAPf+>Sug-l-5G_x!9oB70MG!9b_d*!Ke@q%FMN;y5Ev|R z3p$To^Qp+-)`TQY$v1>;wPu?`w zrqq(!+MlDoKi&sFf4Zyzxpt1! zW;|acT9$58pL%f*^RSQkf{#T)JTrwXiP=bNqe2%cZDg{A+4Q`d!?K(>s#oR`3ZpOS zyHA&iKug3-Y+M+Vjq473KSt#5dH%420P3-91%G@C+Mcj6l@UgpF8Seds-G?kO5pI* z$+%TMIm&iQqH7^Haj~QB*rG5q!tH_f(5GM@<9FhLUZ8XLt*(Zf_XaG)-)a(kp1nM? z$~#X_o32XMLcbe(cwW|>-K(!c{G4(qTS zoBSX-@ZFr6Ni+7nA_uL+%GrHHGRusa;TnBdv6#u^r0m3rdcn!zl<_;Gpq)IIhet68vpO5`DY4x~^(WxQ*Xe2hrpG>Um+vsC|;_tu zE5`VMJB{=^=qjn77C}~|+btYyCm0^b^dIIjUhj_SiohyMcTmW}f?*67wmq-zE@qi3 z_F0!s*(se^&$!Qa+vCFQEAAhelU}@n+KIvR9TW+nOyg`A;4NZKu^D+M=?E);L^QeuQ=V6^T7kxs=*2AF*GN z0ZY~N%tuSoEjCKugQwal1S)$94i^i5xj1rE*V`-FaTC0R5^cAXcY~Q*)_5-9g7FZw zjIOH#=eU7j(hsZmZFN+MX3-GR-s+yd&w=Z41!G}FrB*r*vXunSQxYD0VmQYkHoPgb zhL-s0XvokjX?y+dh$)uHaAnzxc|v*nz2OaiGVr%64GTJtx%G2YHw{om;6%=vl`Vke f|1L$0yN#Q>mm3mAA}k^yE-D5g;o?%(RU!E=MtvLP literal 0 HcmV?d00001 diff --git a/testshortbib.pdf b/testshortbib.pdf new file mode 100644 index 0000000000000000000000000000000000000000..69d297aedb343010f08d64a05ee686ba2a92013d GIT binary patch literal 1274235 zcmb@s1ym$WlP-)7?(XjH-ncsq?(Q`14g(CXgAVTQ3~qzFyZhkoaCzVTcK2I3clZ4N zt<$GFsxsn<$co6U&U%tUNlb!)g^>f6qU#~=9hQZdnHXqn4a?6@%=FdM!Hk$m(b!tm z#g3Rs&dlDz#gdqlg`0zzNy5t3#mwpR*Vf3zOw7y#X!@zCK+MDSSCf;o3o$o4Gcf=F z>+IrWW@HEJv2vz&?2Ol(@X_ht@1s+!@91DqyCIB(gPs3~3m zvxFG8J%ME;{Ie>KklR4o2uR(H&^$;T9B3cZkR>dR&^-jsWbpm3!oX`p2JJi_5J-Ar zsB=%q2FUc@8c?UH9u3U6U~#n1FUd&E{BE9B7&-D3V#)y7D$f#Ip@9%39(X8sVUu6+ zFxX&JEa_mER@_3C{K4_4U*_C?8~yMYMJ>;@iciKV*!T5m$T=DL zM?=lRd+bgj+kqaWqj+|(zA(2!7UHiUrtnoDeWbe%aCt_ynNQXJZ{^0-=~f>lf@d1U zu+{f;_eK6Qa?JHR9b2;ggdaHTZ-8xN;TR-?N(sBj=wTP}SyhB@8k+AxtHEki$b7`Y!P5v+<35s zMLNz3ahGG$(+fRx>e3C}Wz~;6r@1R#yP2z+9bLtWa#goIWj=O^1?~p5wJBHt2ilXuuPAY@fnfg6BM&mS)6jnKs)*H8$941OL z%pL+Zsy1cFLnfhI0Wv!SbSIXMrszofTk#&Vb zq~#E5J(dOFm^%7#wDkq0vng_`0n04^Sm}oS4S6kS6##2B%ST(2(g7 zb_Xp8A=StPn%;6Dw{Nk!fv38GyRyklK#a@|X7R`UruGq_r@&}Zm_q?lqV5ss5JEv^l_QPGVn!Gp{kT< z@RMsL@N5K0vR_GpMt0!TWxkZanwrTJA?F_9*6Gxt(qUUGv*Pp_M99aFMHwr@hvm~8 zZ}Ycnc^u*}=y|%jvuARO^3%9xIn)@$ti10|OSa2trW#t0A%$Py4!w$;MO{hMTHI$0 zA35<0jn19cqojcXRt0vaw{p_J5WR2qaHebl3ZPq~Ojs<8(1PcR>5rUz9&cVB8AR`n zm$n}mrX1Bvv)hI)bBKzx_o94jI0`GCJ@sz2U^D z)(vdQuDRdl9nXt7;j(B7SDxx&@vDieRI;Pw6jh!_ql_rQan+z0!(6?t)Mm3X74okw z#Q1RsC2hk5%oUgMnkF~u+$$y#mknL~HY_A7jIJQv*50RUOdKHij2QS#L^*B45b$bI z-t}wXc7E0$c%(b+K2D^{FBLrHMoBE%RB(;)qz{W1~;@ofQNwFs~3sUAxt=@dw&Pu4a=Ho#Zw9uJ0n z3Dymo=+4^^eMP%H|CN6Jt7__L#qpaA-7ana_^!43;N4x%yF0M+yU*{wgl^bxhvxY> z>V_7OEJX5kE1Y=pzY5i61okBr2Ii7Rt z>GQq8WM~4{Zu1UPrk!K<@N~UP0AaY13-m@{XA-V^Ht}Cc>4ywRRLMIxpt8{Y?dh(eBaK-nX$MqV6m%9uxD>Ccm5~p`C)f||@^;oXq!<{r2~tj3x7~R9&XDc{eM4BTY1Cea-{n9)GGluqej0ly zTaB(%Kpj#oz#_iMz^~0)HkD7SvyV06A~+XYva3+oaI8e2uD)IEw^bwN&79J*tg0D1 z2&*@w{;_7M`2Ol<3rOkAo?0*;F({J?21u^ODp{k6`MnpK>`k>#=#>kvTo{|saqH0U1cU5KN z9F5r=p3E`p%!VcQyTN1*>@mQR5gqDH_YA8<#~e<$`=k!}K*sOfm1(r^kPE=h>QzQT8yhepDKLdOT!- z#eC;UB3FRP;1IJ4W44cM%8nVAiw*ScFalql8ESFVXjio|rw{^ny$at=kEflKygIor zKzpwSE@Nbyis@dz^9feI-FG#9+&Rs}a{7UBU1u@=e!Ac2$sj+C9>8m}iEb zqj9j-b(9tN!uQ1SJ*gzpMB!D+@6*{T%#@KKp4>%GZ*~3>vwkG>&LVVS;~uU}i6|}X z`^{M=RwMS7CU&fzg)a4miWc{+UuCnvl$-e6ecjKibkbP>}QrNRi@NI>O@;R^FALg8Ia`_Nz@rL zcILsT=3hAaiW~i;y}{U%c)x+r+o6~4On}O^(QSG5i4<`lm%FkD?eLij!QxJ*nngM3MC7iBdY#Y`H(^3Gd8;+EvDmxWB^=w7UYk6Wu zXGx1G(RO)k?K#r^DWQXdtD(xVJo2$)q+I_lcNKE+O@_>N*kVIT40ynzXn%q;skbx! zW|;A2nCpvwR|Suxhto-_S{-}Z+VV!h)!5bPS^M$lgsT*$tL3dxA|Py)NpQHF)NviR z*^Sk^l#f+UV)&#wYtPZh0E)nw_eks5II@7HgkI22idbaoV8~eejnm0Bh|lN;R`S?2 zgg**a(zERSC{59CWO{#P>@*K&So4>L-@4t@Y#KFmk;2M7Elc`gjn{-fAQ!Z89>f>^ zpbU(T?H`T`Z|&;fKyUirJYPkk&uJYrv|o308%)D%e_`yeM1Y_h-j+`(wsep9(H9|; z5#zK(^w3aBV`qEKtWhe4BjD{c8<|yHPH!|a)`EyV+eE@7)&N5fDUxGc8Ut<1Zg&?P z4U$aAg`SsrnPXw{NBk7LC;ZX@hVb`k=sj;?C|0`0abd}Wb6G)J@}R;@hkQ7iBWbOZ zK(Y;}Z!C;L0gYJ*L}Bg5@FZ_~-Fh{=$>4~-;3zLLAX5&rh#yl^i@14vu@3>rByjARKdgE>Ts<|?o`?uxB^yRS*x>2tA zvCWT~huvmiK6vwZCvrV4&-}=9%PK2t9k#&{?x;KO9cL7`E&-%gyM2UI2t~)*g&2BF ziWZ>OTT&QHyuhQ1yG|M+?fv?@ zzNejq&@3y40Pjx)Cdv;}WUjX+!e~WHta+BARfj3LrHP3NeMMMKd&ZrtKle2PHpSUq z!L#O$C4>1_%4`!E6&`Dm7(g+M;#GiM<2#ofY^Cql=%$#~rG*I5Wru=k z!>rn~QMbocc%7BEMm6`TjQrTUY3e)4OWZxzsD~tXf#c*_flh=KB}9&5B04UA|f_TS-7cXmsU}2ak)X*x~=c6{m8g**NxI%<{0T? zq>EMHeu1%VdV}O(js%-3R_NR<&xNrei>iXvTyAU@X+m&6`xc#IxZ(4*_adOZzr4O~ zY#3xfkcuH13~X;KuUAcSJ9L-qrf-`!U#i^(9nOn>uGk^9yM8zHSSLAv z_4K{HR%7~s7X-+Gm53RdOWalL@8>aB{02y2hGt4Y_@xD*G~w2(Nw9iSw>1KL${MU| z^M(@q!;YDYd24|7r7sN0)!%_IIy!Pxl$}s;A)_V|<(grv>)m68@BwOeTkenqnV2@n0svdp+J7V+>(Qot(-?g%7p$ zw{HU->EC1zOMBMLE+=1F^rS6;}jRj;>ooJ&xg^mK9NO4;$UhG>@sw53JGcSSEB_H1BG6 zBfzO_jFfKWxgm?AO7K*w;FA0?yja9TU12V444d`9OKMDRWVV;EI-BzMoMfAI4qYBL zq8LdB5SAa`FuIu1wZKRYK?)lj_hus9&fAS!A>s4xT%>78H8`XS3Or|?9!mDSzCs?# z+-$gsZh`#@al3z7ap0{G5^u@NtiBUst; z!DY)HTC-k-=C15JIf+K@IZ8khvBB;{!qzns^$3kiO@6D1BH#PWj(`()LN<4BGRR)({D~Z1X98448bB~~AR3^ubAuxE1d-zd zq6GM$${wv7fTyM-@n41?lFYTW`0@%VO-8Ouzk8CpQb`^iR4!zZ5okv~ETLF`Yb2sE zr5GlXW(dT(3oOrHPqJR+P*NwfIB8) z(5h0CxMK$*LG^XLZMcMdug^_22_B<{4JXVcJs%PI2;DHh^rYW6ucCt`^KS5BMUa@l zkhn*&zBw}6atK3K@~g;vXwnzYk7;3}tqH?$qbeuHKJ}7BsVJ(>^z{4Rbf9y`9w@>j z+E2}e9_~MO)G&2yN|wNDn@WJAqavF@6Bk_%mqjAhPn9xv*<}%SM&ktfc$oi zxk^*Tu1()Jn-7gRfwdXJ!0y!QtR7qxVg3Loe6evxit}i$)nn|(^Mn{urLrbxI;HZq zb$Jh@)X8wfgPaSdh&wy<9GbU)QMjoXEm=d+jFo1YIC6Pb=zuWiQv3H z__s>IQ4i4og&D9-&lp!EV?wW&ym%KDJ%s1%mRw_Vqy7$cTpaFhTZ~|~Cb>{{mGj8p zAcrA5bb^!FmEUt@OI;9oQIo-2SGLmoT%-19@Trkw|Am5{Je6%?3e)xub&PjoE@mO* zfeoj>pjB*SI=NR%1Od^mwc}6aEY=}mF?yc?F2NelFGrk}pys>=)#=#~*pnz)uQOfR z?{nFmI1`d+U!MxVc(J|Q0=8BdU24KutFMP6TsXax6(uUnCCStqf*KuajGuoq5kyHz zR%$$X^wDMmr6kKgrJl!F{ylJravxjbpGo@riLXbNx3okDl8zy?M7&B2_`g#o6n|!` zdPv?DAsm{`Iv1F%+(#+@D)3(%F*qFHx8=F~J*$Yrzv~^JtTbT+_X>iCh%R7jgSz1YXIhmbw+ovpny zuZIBD-zL1D@?V#ksEGeoaj_Ag(v(*sb^!uyZLC~~Ss8g4Sr}N@=!wmpjO@(ZflfBW zER38~pO8+b=DaFm68`}FJQJX@ba8RuWnyx7cV~2GV+1-`FtPCP@Gvp6GO@BUd}=T_ zd)m7gc`(>JlYRR8I}j_)e>1apX8b!QMiZbNla(eD3nMerzd928?2eO_gNqf=p7<}v zM#ex_7Xebzf1o)0llT7s^V#r;!}f{8|0(h+IRQ;wP0XACl1g&Kl9oVcmw$mJ=3(UE zU}WZJ`VUI5f7{5Kd46`=)EV$Et^alk%g^*zB>xoIKW@rDUHR`~{Y=Dv-{sHj{iDzP zpE|tqKvOGoPq9zN3c$+D%Fe*d&cMR_m6?Z^m7SNBLz|!JKPi1C`O`}DlfnArxc=`} z|D^O!RzRn(pIHaU8GSX=BzCv@WX~)=%h$=t%=}+`{cZIRrT?#eH*%1d|G)R%&hD>T z{7a8l3~1tNXJ+prEha$a>S|@mYh=Rmc`LYB8CcEPxEWYj%#9eBKMw{oZf*`9c6Ls4 z4o)K~rvF_UOilhR4hL5!+rO*T)P%{**6c5r&YziJ`3%I=gx4JCWM|~^IbIwbY^_X; z{wl%0%H*@6J{w)ETx`t%e~$%WTbF+kY_0yPN?uzddkX<74+c{+b0b$&4dU zAHv`IO#hDYDVhA0JfPFRz^Ixz{wv;pHvPNlFVlZB_>@imeTe-34Cp@~{WnGa_odIF z|9PXC{%v9bz!LvII{x1Q{rhbCIX??fxqr@;|7FT%VP^iPIs3mQk=Z!eSpOF!G7rmt zB9V1%fOvx#ZyN@+L^|l15B1QnnhGHhQre3$XCUH+6(x;QIyEXT5?$zmT(?Olm+A{pTf%I_j zeB2W-LeAFOKqii&+TlP#6TD*j4O=1R4dFc8IMF{MyFLfWOTSWulcb|Z|#uCNsGH2i6p|q^gSlt@Jq?V{&?G?#E|fuH|@{r zu&=w}XO>s+NzG8njACW`l0UQI%=x5G#2uSSFsT)lCXZ$j$NtFbyr-~b8Z^m0lUZ8A zQ=6IVl%mT-&%|a-P2scP!7LLd>?Dw*c9zg{w=sM5b-N4_6FHW8CB z6zEJFA;(+3+?!~WImwXEZG(=O;-N6oPP)&MNV4@q;TzJP#7-(WxhPc(`;WFgsRK*u ziCliYqfqQ^Y6p6Bt4=C(eTo>sjRGl#nQj`eVUGop-m5d(BnD-tm`yTy`07*&=_w5k zV^6_cF~;X`MyltVpwGP)SvLEt_rj1r;G-;l^S$BJ!LBv!A@ZzHaDp>q|Cvc8Ym)1b zYqoh^;Z_9HeY^&W}Ggyr1|hW4MZ+UCo;Nfi;mSDgwqQFwQd%Pk#5 zMqz7!92}d%JXu3r%w;@vAn?k=N01{Uv6%q5rK?u$vVW*OS6x3m>r|0hZ(Yx5ZFar* z(|NB^O8N2JU!=2T7y4WfcOwsVpM85lm*IO4X3fCEpklO9kEvQ`^@Y;qE^Rd<-P5dN45Hh>%0q?duZLRKA%X>K*6+>#I-`jC|GE&KgB4;X3*9LOI zP<7x%nNEw!vFnnRNPw6cGq7Pa z(I(VfIqW6`P~iQbst6_E@)j9}`9nPTKdygbD<;*ULH4mb<}V>mfAh4}g$BFKzk)ri zp=OlWHp)hn=4VdG>^d$M@eYJEs|mv)5cNf|u%xpH&S?y*#Ux(gk!tqaYY@%I`w7|& zFCyHX%M(J?-k@)Y|5EtA-sZm3x<2-Myv=R4vg%o1PNzkSiR2-Y<`|=X{BE9lgNgxv zDjV|RF6IuiLPK*P_UCip%Yp}}gJmNoiqK5Z({|0smlcUles5UWK$tK=P?J1~%B2#B zb&L{hP!Fg9z8H`=9B?PeDb9d7ue@J0`D5g<3)c}-xWp(TNKo*tbs!cff<}D?#(lB9 z0jhZsjqVrdd+oAi25N!_y({F{hQR)xuaRvGBW*M$<=p=+(qg3&P z$kf5%dl$OfKzojQF*Ric-z|Sr<-%|9BNhlPf~Imf^g|jh5OaW<)I2#R0-)sswAcw_ zVpw`bE+CeRWe(QZtqU;8jn@Y;F# zT|i&%4;C(x0;uj1w{`zC8FaUu<)HW~cX;`D(-zvmN-S=;lJ)fs^-eA&H;}|DK0p$` zBAHEe?@u>lS~My;RQ?Qg-vkd>sj#Q8?wK&FGe#f?Ls77IykZx%KDqvR$LKEdJTg!1 z1~if^;MYYeB=9<)()X~H502S5?Mv>!pb~o3s^W^L{|e5^Wa~#+X?gK++2`lLP0hKj ztVM5rt;*V7LqHzuVs^{O0C%m>#5*;^3RJ$;!-`B;)tpkePxgjX)K#Z^>YIR&^>*{Y zpeBVNkSW_;7F2-p^;bz!SlExPz3xgE;-o7R^N3*Z@c9=7c6$4&$w_bJ#e?>gi|6^( zy9$G<{o&zFMatXw{a@Gs;;JwYD->eQ37u{18I7d+c8fo&^WLc*bsYi*viXru&^)oh zB{g?ZKB6h;R6Qy}npGejS*JaAzA;u{Y`}T)gj;DJGl{~ficwPK4hZi44R~caIH=7t zJ>7HU3;CJrL;Goi+g={aO-Dc>iwt(E$4PEj&J?~0&T?&l7&EP#M{$MTgQ_)_K||zX zeiWb(KYat%|3p#SPMkkN@QI}WNn*2m3Cb%mc-lgV%+F~!W;@}@ddgazZ67~I+uScUJDgVU7tiLa&k4L<{HU5w zXsGy29UiUPUh6UpCO9h^`)E=}aNhIzlMRNe_t`utyONpqWS2*rd6H+}%M6C#mWOAg z+yvJeG_D(c1U_mLt4X-8RVMaMi}x#ZhFW~cHb>|@8L*sK3|#6`1l}3<>vYCn;GTkC zw~KUChli4NM_nJta4Qcl_Mgl2Cif)l&C9D2Gu%7cR#>s*=l4JqveQ`Bfx_b4)UfU4WnxQRhLC3naBMYh>dvtC%yd!@ThfB^hwdj`k;IZ zl_4wLGq`U221DvH5G9y*qW~QRTlrOA`tSm(FO2q?j~a|I4C7L?SW*PAVk-v-_c0|% zqVJ6Eh5>~PblgVJBpINL2+w@#=iRAxA%pzT_KyYPG4ynyV<3l~;l^%Zv3lUWHlvpk zh;ErsRz9>iZos-4!+Lo+g&q8*SjardAr(f()&07kXn%MN33CUo2%+`Fx)`LLT*ogcFz5a3u?7=k&?ePQ4T0N?VmQZ8EBdDRVR z7Xnl&Knpd&Woo}986!~l<0?gPW3f_ezj{P`#urJrpUENJs-0b^SaYN+tp=lK=V4fD+AkpNy3Gw;MCq< ziKv2E>Dgeq={IT^vg4iuV&Nd-kd@}h6ExeR<=XgbE|!t8($$T#S2WZQH{gpv&Y;3e zrAf|KQJS6_6I^mW3)BZhkwK7KDpG874BW&ral1tj-}mc|MkMi3W~P#v_#mB)Qt4a$ zw(p+FM0t1!biQs=b4g5|wxPgKibF4!APk6g_d(50E@@xu-rD6v4eax+dZ0n!f4^9C`mi z_My>SSNkK$6jV3ZoetRl9>h4VeCYVCJeUD`0On$iQC=0WI~SXAuNU*BOz;Zr>m`9q z91ULLZjY59kAQ~e5>0gZ#6wGxkB1=v9Hez$6Z!9Y!nrgu-+LP^(y+7`#%Q`rXOWbo zP!m%q(aZRw!DO%cU{-$Z7TpFLVIZv)U)6@~WB7ORX23coQlJBHmyFUUEUw4ePdu10 zq;~bs_M4x?O;?!P;|cFDVVxVd27VG9;s^%RKEpc+0REvMqv?&xdQKJJbK9v+3z3iD$5OMBGOu2oH_}#Pu7zaCJA6yvy)od?* zFN$p}>jZ5a%lTlmAvNg2c5AQ7H5?1YW4>Oe&)_?#)c&R|P38Chx-jM8{vQ{n>@0t8 zVg4U>H}f!a|I;hR&)v2-;lnibH=ESf%gO6zgmb zPfG_Cu&wzW0Wa9lH4Y1~=E|i;3Tu#)$xyv6s0EB&(sW(*(5tN<_8rm<@r9b=AcmiEL%70P@t{6-cWn2#W-{uH&%>qLu95-bHtB}c3ti1sX^1Nxx^ zL;E|ble)~(G%4vTdKK>ot%}H~?bq9D3>}Por6GlW{3*UfnT1h;nzYM2lj;Pl7)K=< z*l+kO;Fj}_D$CIR%mt_XUJnvx16UI}$>1fPp#w}`Zlj;z z-o(u6o#I56G8>{qsji+#zQ5>uQKO;JDL2J$h3Ojtq<;j-d594ui5ZeU=VY?1gaai?1xXkyLrrs^xojbkj(@U5&Q8JtoH5SwJ@R!9gY3~X- zM0Jx_O?u(O0`48u6F8>McAic21VHvg*JeCdHWTO30ah{5OnxFY*aRk%|f;gB|aoxFl%#ZcdL*8m7(aYlKD zvw)Uv^X&Ys@d^i5{f*mv*B^~HpPA}GrK|)uo^X*V%82e!h#}Ud6~p5$tZgPkii+DS zkw2PL_9dgdM1Sh;LtB^9w8ID-`kM1P!hpFKmcNkf@ zq8jBI$5!i9Xp8K0TD4UeT7InFhX5m=in8;uj&9Fn&*}+n{eOArET7abDwTY%RQP5! z$tTw@fe?JL7@2`#UNm-z#;y@%=6UbanA(17M!RyDEdc8O(u~(Y%=XLd?&iEsa0vHo zA@aKvg*k4b>@IrL9d4_-IQ<+vo*IUn3(E5WI_$wAvY=&(C29D5-bG{BcsMrt%jf@G z@cOKDA9K%4HxdNZBk^IO7FDTqL-jyb7w!hP?26VFeU8!j9S&ZmK25j83zFb2|EE_b zgBTLLR3Y&7Zdl=B5PkC-vaIy#~ z1!UfZb=cai|PKd<)@LbKw>?PFSZDSfSJQob!L zJDA5Q9AEDgJv z#`KFqxweCKVs<~pY?LXdvU*`I?+d!^a{LklK3@RNk7RYOKfyANylb>?He8T4)@38T zTx8Bdu+c7!`PF<-)qd!n0W{h!%GV#}&tI$`d>danS9}dbojXGheA?ll3PiU-vi)H2 zi8E>wewgB}z(EP&criCs4GaBZe+t1H_MYnO16A4*+s1@4w4)R)h68Ztn>DkY_IqA^ z|9yJuL%NV)?AG=Dqn2;)_YY|WvNaOty_{rQj|9a!^1@4RJ9j4shq3$uSuk$`s%qP1 zhL?Hm$>Pz*EV-r>1~%2%aYih)T5jd%t(!Sbn)XGDv;M9#1^W8! zxckSrZA=sB9q*1tXyP(wl@& zz;dF$Iz|LC2?L^70z&mQ0FAc0l*ha3oUyG{{>a*6Zay^>&fQ-6P!#m1!w$o7ajZT+ z@++as9=aVq)g(!Zo&I+Rv1ROpxt3q!ACUuu1jiLDL1WR5O)Q{JOG~~YUwTaLq*c6O zDdKPgFsTPWcbi`-2lyh$rTfBw6hnIjBJDWdIN=Abf^35*@8+Gt_kg$|{bLlB=`jkGe|B5|qky}T6ZZ)og-xv7cli=UQ0US7)>z}3KC6!?sM!Md zENhSU@}lsJXx0soOe4Tf1jCAzedJ5yqD;~J7VF)QX%}=m;+bU1J`9tJAK)&ycP6Wc z2;=EYOGa3J6*KQKXx(nYu~CbAp&9L7F~GNADOoWc!%mJkf7k+;Ts?GRw$xeC{+B0| z`&p2fGqI8o<*Ujx5*Q+@JOB?1{)f*`i2bDB3h{eJ$K){W)P1>VZcR2~CecHkd!CzB zz#pUcRxK~6Y8CE_Ldwxj^h+T-!-!jSqMrW1G|{POIcJKoHJXM(CscUwyg3Jd!xwjAh^p#i)sgnWGMf z(WjfiK44Gg=ue18sbFV_aCLj)-S^V8shZ_zgqTt4*PN60L7wnKB7Rs-#&8g5y}reM z?jT?ReKJ{A<2@Du64%B2B8GK}YnN#Rjd%)qzH*p-`uMBcN8kGkJtx=F!`54w4%*&s z4*eewu9~}iZ%>zJ3sweihDGx9_YECbP85sE?+&Kbn!!R=(tRG;6X#u8v?@kG53Uyu z*Dq102JJp(>cwjxtpVbF9U!p`+NEN4R*xsQ#?Fn5UNnofqXr=SaJT0_&lUv_nUd+> zKc3Ww$#EVI;Vz)AG!-E@hZRv9Vq5o#UXJ(Y9bhEvswSCQ z-4?1B&+g>gj_m=A=O$Lgj^Cc@&z%sc)qmYPfY|M8eJMa#CH~3p!}{}%e5`7g?4k3h z>V-)0VOE|iLxFIN$wENiHd}ttM$Rt{F$&6E-~Rce8gzKK4edS#1pj_Xm|b(qTU^?B zMr3d66~Zwz27fnR59Sq$0d+s<>{qleRXHZxAy1N_G^R5>6=ts$f!TM7SW|tNa`4{z(phP8LQ3}q(GK91CmEZg1}J?Te(g=-@ly!Fteod+x(|uo zlKhaF6J>1}5SVBn3F@`kLK47{9+NqH%T}_1am6^ZW0#}`By<@_EDzLmH1yIwh`-}6!{C~%4pW=3Rc zss}i0^Lr(WePAF!Y4=s2M#E25Vj?plN_GR3U;QL8W?MAsEI^rKd6Qrz>D2$JHj{7H zF38Z`5(C=l-LNxjW8=~G*?KEQj(BP|pb=fF%u*KADi?~vQr4rE;I#_n_LfL2XmIqu ztl(IAIRD2L95WZo|At!pFDtm!7`#Cg-${*Ilk4uRD%nl=!2tTf@L+6Ya;hQ7o}fI- zjJgJ0hnBk62cCIbCll{oOKPWfJgFe|J`#)8veq(%b*ZXl$iYVLvM3W*m4rlO-7z=v z6xE)1J^7G)?z|FiWyw4os_E^Zj%0kZChxj@RB(j|lZdFqKbq)3lphhsRD0ZP_Eb7a zHMHM5q6?uFn_a)ZeXai)H~ZVIS3!2*jGi$x zd{s+t&k`t(rFK@iXB^>{phv$Y{p*(|+|l+Ql^Qe{Mn_OA5k{*zx>FSyH*sBBx3-c| z;nT_^*4H7YElCk)^yI4B9Qc9XjIz+BQ_QM-@kkMl{z#(=|jV zn;LPF!4XO|sjjjW3s_F5TM_N519%0JH*K_Yux9ijsowlOwHXk8E z49;+y;)ldgYGZ#>+C#@mt*X*apo)z|K;=-G;Ng8~1yZ z${G?1He?(=MdhbE71n7{MHo^J`F>oUC53IYqznon#3Xq=2v{K=5w1ex=*;i_H&=u9 z9($)B7WR*$vrKE3^GF)9x;nNRC!_0yn|^KU`o8S&OdHQwC&>oDP8@0vSP~?Tl(aG8 zP^}U}lOn7nzeB*&Z-neBE~X!N4~m+tZ*M7hgw0sQ1N!`Wj7=WV(Rbg*3FkB7jU)Lp zJc6I$EBmOj*ZDU&6W-ffU(ba$_V>LaT_vWwRx0!>4&D#uodpdZ4@bA35B^v`=IQx{ zsv?rt3mIYefwE6zv|R86B)rJE5F+|HLQ(JiX6p<$f`3b2hSKDx^Hy!eP6)FIZS^eg5TUWH!!^PenM-4RQLM8SptI2Z-JsB z3&7{G`)#Qx4s~;IKbkkb-5<0VR2aCMn;V*^ZrUzZFOELhw)xbj8tp3spu4wq;^K#( zge*2eJrpOYIJTTw9F_vH+r=^97tvNvB2LUu5jjVzQ!@7VjXecIufc!@2(R@>0XiW; zo!XDYOE6xN7dJ`t8{PTYD=qyXFV21YZ!Jdq@A%kS&pbh#8=E!?KK@L3^ved#kB0sN zZR@J{B5Wd6@?pK_-!a{qOp1qsCf4I$l%JJfSnf^!46en#dCp}omL@6Wl1!X0W%s@D ztY!DL9j2}=rmXH$l$egUWrQ-NEHx=E0$=Wb;|k0bbX%P3z9;fDIi1}1>gZPkIhU4; zNV0bw8BFai1ePnXCVw;~_a#mO2H)v7?XMta1J>0A(){cBgEJx=RM}CB^P7=TrsJy} zDqzberyAjRw~19r^(yk}opSRtzlO+@i70Z@_C`VDX^g_m!OfH&mmd1aCB=iigyFqYh?Ox#o;n!K%w&?H($q-^^I+d7DIKIhUS(7LOO(f`0IRRbu66@@+BRg88_uB z(Ux$`h2`XjZ|VI`{l4p}*M@jp-{$)}5>#Mv>rHS-`1SR{d^r)%bBqFHiFI;e$fb6a z%pT;0Z9!Pv5AUO8aHDeZ^B4h$WWpdBB;mraoEcH2xRb~yacH@GV>nJF8y4|9?aC$?JjJuRKcbfv+Cw&SP+t%+3c`B3d(v^H z3>zIKzNWYpT#-LMz%1m#r@I*W?5{#40gSpAn^WkRj|-7Q0uAGf9H4S z{Ks}m5Qj+w72Uw$(T%k7=|`TNqa4qU7|XfS6)Fj1*u!WwIM{&IwapAS$PenU)79EJ z<$?@-O4I!5nJ7~NG)*}DKD0pt65eWNPl1tOilr9U2{F)=xd=#>WT@_LeQy<|>?Fkb zG_gOj{MFa5&~39NLq9-Wr7Lsc>MBiO_rfo$Q5eTZ@Dt681~n@q=3{TWy8)Hn)VD;a zC)6A1lbMwZ@pytsu6bkBQmLToiFbl`zbbHXrtMBh5{O-BAAUZsCTX{0=YxUn14 zt2irt$a)q1ZRZ*-)@-*L6;3;dl)Lp378J&=@9kA3Ah4K~WurLRSbVrLd$Tlq<6+@N zsF<8J2s7)fs+A^=!1f!fbYd^a^Xrqmv-2d5`hYc?z5^pj`71vh6dI2GUJN~cQ)Pdhy=a_m1SF~57M0@sNDywy2v2mqC|otl z4{fVqZgp~7_;dsbQyIN2YBO~vOWnGFlB#hKb?6OB)JWh>$S^PME9ebn>P$ERoXKGC zQZ2I^ebnFLTrKn6jQdU-Gs{m{v_v@`wJSl`ScsHC6S&wK34_{fm{!K-b6%}Ms`u7! zbNP=vhE(>AD0=lq;^cAAZDmx54Tl-cpv{%$YOb1O>Q9eHb`3w4K8&lH*oQuwts8w8nboXW-QN} zzoABQwN@WQUKl=g`M+7SjZiOBvf9135pUHu)e?hz#Sau>$;S5d3$*v>emW6kG|*$< zT9hn|q&S0Zus&1&&4{s*HCaBt96|x|N7)=YXc!Dbd@g-e&I4p>c`3~49pny8$oCI& zHt+9Z#H&(KpE7r0PXI>JA+wg(s7fhfEMYTD?H}1%U|Lj0HS9u9;X>Q6R|%MoS4N*! zq|t@q3urEiCq(EU2~kvJAb7&vT$5XhbN8S8gR7`t3Ca`da;d&AN8cL96b&Z6?~r63L(2ipC{U`~MWMC3)cb`(j#FHfQZ)S*IOd}YnxIp= zRf+~%ZT&Rb%8R5oCansiIt*vPlanGHI}*1 zQ0iPw`M-~jHMk#D6xX)jRi`ceam7scEEWbrmp zje#$5<6bPsdbSklU6-pL;+D6z<-CfwWR~WHw2!7T%xqNiD5l3m9@U)K9BzCXLKrji zZQ9Mru)#DilNnTu_^{Oa+l>ZMxa)f-m?oM;9zZLpV&eJSyPNXQ5qVgv(>9-LynvM0 zf_`meTJP)n79Z}uPnIM92WK#Iviw)hU}FBC^x$wZ{NHBu|DzPK!~OX73E?B-R3}JC zCP(T87-!{=w20PQZ0zVp_TP76mvvo+T%&aO`N5mSxXo|wG1c!t+?m(b}pn>qcT4x9_xEruH@Gik2T~ME}AJx4p zidmiq{ni3RL~v~J9A0)TTQMJ%MOJ*zT?qNJ_Y7FtkWPeyr?WF!r$K)@yY zYnV=WfnvmEAHwO(wG4=SuB-AcGF_OYViHzW$c)Vxw{y89Hh{XS$5eX(V`RM(1qY+= zk;g0s_LsK+Tc;v7J{H8=LkUe72v(|kRv@VHB~c&&Wo zkpAtlFcTNbTmXMTpYM0eIWGF=wc}hb+t;S^_Sc7ZtZCh}p?ibR^Uk!XL(gx;~S6Jl~Y>@cyEMDB1Nh}v?aL5We=a<`{(Al^xEg$kMlsPlwJAlf>S-c1zVA-bz|d|2|qZ z{PE&(7rJ}0`7#m`$lnW5LVJ$hBFo6f3a=XtW1*wtV_?k+_@@_a36i^z~d)U&2}7sLL*=myc@Jx`0?+a zw0HkBMUJj#ypEQcO_skXlRNvWX9{{H_)i#m$6|(m?2!fJZpnX8VVM@vpSc#}(`GvN z3fmMDXjnKu553SG;Tm9@;onv}Si)UQsj<*W-^630kk()%a!Oe$F}rWOcu5RZbc?MCLnIoTdvl*>2XHL zLtYp&j5jE2$y-}6xK%kVe$MGBkKwSVHyFUDBBIuq^%yNsV`elKdAFV(In-0EUzDGs z|9<%_i3>c-vHVU^*$oAB_R-;;{07elti>&av)zBr72=WxONvNjY6UbMjWq%T0`+oh zoP+&=k7HH%WPKj))YF+f;`V+8!aWKr)>7ZX!EloeX9QNX6gJJF7uqaD zc%O#OYp;52@wwY~$LIAmiJU(7>-lJN(Yljdz3F)U;9|>}pRN{9w+5XZK9^4Ia%Ljy z9Ln%8^(pP~*}A2Dheo}HOgZDtbmKSNPPTdBl^0vPk)Q4_n_X*t>SZ2thSIScH|%9b ztTs`2j!a!&CjH;tr?xYF@Q?g1wSApwzjf<6wU4Zuj{{tkR^^-KFvxRudHT{*tE9pc z0&4I$uZ-2#M5XJ`)9b0Ss3^`$5>|+++1@O@zkcL3b?l2#BUkgw?;|*Um%X~xs|c#N zm@=@rJiF8Me*H=iZDYRD)?w4S4$;*e)U%&!FDgSEeNERDAZPpHf%Wi7mFyUID}HUu zAB+4y7EnResW_N)CFxae=;&=V`eXz`1T@W=$;%KKQDm%uN{*at>hnQ7R}*3F3&HyL z>^cb_uMS(1B5yS|i9F*u{Wc_N22b8D``Mrv*st}IlsJBwJ)ZO*b^(nc9;V&P3x$(S z<_((1muGJe#-RiQKw1KU1(?m5Hu58;l&mH-XPQm78B=-G@<2p#{SZAosQ1#6#KWXn z>sp9RMo0iO8`R}QdtkaUDEkl%?co61Z{tzWXFs3h{Mi(@3|=_hZ-S7jx@w4b{g;iE z$FAW5uZAczT?jYx8~byaH`4jN1m2aV!t9}6@FeJ1j?l0jA)#AB!nOs3ZFBM3=Hs-@ z(OTvb3~VGAx7Fmu!i1d?hiVmwJV`rV`7iA-M z;j_4ZdX5>!LzxHd@%OLym3nU20gvW+8MFXh#cQ)GM>su{VByw95FfBk{0t13QgwrO z=Y0hI3E3jk$J_bMCsbhnIzQtnsqK{2t?uFT2&;lIMpKMjoK%>DnwH9M_mASU_}LSf zxH|L?F-!n6@-k|&W^ih(C}BQfZmZz#BgDZ~`BO?i41vI5 zXK3u3*LS|g?}_6^GlB<$6RSMiLRD5IBN5)SK2FualI3u`z-rX1?f_4*SiifAWl;x!a(XD)ntcUHXS~F0wVW471 zN5YDYfE5`9DOL=WFBxcB*3+V~zKf*um(8~>MgenF^3a%EQa-lhTf7wI}G?AJ{p&{yGU@e^2)+YPIEx!-d zk__||jtCHMI0vMZ6A-{3;#&sAizK)&7$x0o_CtXg$Q+4pDUYdWG@r0Q4B3b10i05* zq!k}aZqwj^zKXBloR)WG=g1=Ba$u#JEJBsoO0*$@r4Z9};28qkL3#Z1NOJb8&=X!K zfy1%JE+T9$^QI3p%}m6pMvp1A@W?0@_U~SLpj4|@kV97I;wkK(dLOxHwQK9h(U)b> zY0CD@>y=A$svd?Ty3O#`^@AdhZd5~?+{u=wAb%f9SD;_gC9LO+&p5?lP0$Yh?J!ECXV_1i1lHQKTZ zICs;00!`*yxcy$d%~s<~UYs=9QpUC&#FT>9PYTV6+sFY8o)n+xYVr@(=JR!Qw(E8K zOdO!UbTHyIQg3Y`#AAOW-}k9563ZiE4+BK>uB`hG=G^rtEx&i#rclrrG9MK9r z@ln4;f_>SS^9K@{4Kaxs9iosNRbNyGFe1ECFT1=%>$Q!Ay#sPK?$u4cxpq{2?J}@= z_erFs0bMNV+GBvLnX^cYdI%?!xThZpgkYfmvn<}wNOAlPwV3JP09PHRk)(sd5d10Z{>Oi3q5ofAu9XEr`vBLVrCm5+ zIZ+IxYIyr_w%7CF07WX)j{@~{FTyncldosc^y)D%VX4kM9B{S<>U3uT^#F*89{?`)h$Ahvjm1R`Y-5}g=lZZL|O z&yhn_78yE=tQli&pVTnK$x#jG4a^;Z`IVwM{i);uHO+H!PS_;F2vk1nS>CE-VOVJ=Dy#=3$CMx{Cw?U{QXgC^n zFvy?L8c8w83JlBt6ll~xT4W6acrlHg0jx6-ZPY&D7&dyJHF-*KB6@ZU0%NF&2fQ;T zEdl3mG(6L8IHPwU$sFKH0AfiH5;jGLj#DF>h~a=sy*!1#f^4{gdYEQLxhT^fs}|w+ z@nv1rNqOS0eAUTV61nQ*P~Qn{ZA{v>b!X?($cQ87Z>7awUv;x4xf!(6n0cs$lam4U zUA~Wd5sP==6U!Q;;Y1bgpT>`qtV$0C?d+DQ3w(|Hhu76g&cEd8vDw+phopaF?x~DS z+^3yx}_H0pOaK;I7r7OT{!WLDj_pW0w!7SF9C4Eaou;!THOSlKNgMvihA@&~gf zh3D5iQfk2VaVk5x6TRlVcWCKr^(yYK8qd+THQ6ic$m%jhKRb1s7aca?3wXptHI_8A zrG%X*NmGnT8x^sr_lqF-0|;4|MkAUFLYr-czppxJD|>9CI#bwW2y1pj5R0F?7__^4 zV@*abFBO8ew3uZf&KVIIe;swJe=p3Z=2MF3#}r~4a278#ObOKdTxh;C7W6YB?i|>3 zTa}}w8snKY<+a222d-WRZPT&?6Jl4&(=*2vYFuSfLb@^bm)RUgf{a&uf3 zAG;f!+TyxI`*?aWXzSn9h54*M?+df&&7!VUAC-4wR^Sa^*b~zNW<0-^-TB?jrl9w| z#&2&P?MNPXP#bot(5m@xQcAe>pk;)?11o-9iZs48NYD%tvkvFKHe0#-H1Yz4i3;g6 zi%ujTo8NoiODujpmV0nzjlWr=<~7;zWP7Wz_8Xp*$k%F4?v0&ygu$7h@vEClz`oz< zWa4JGe>*#W+q0T!u+&iWYkl}Tt#{F%la9?7V&Y0qlhEMVAMl?@4X$)W1I<1pcpIC6b@vNn3RCdIku(ll zDIU+olFPQ=kIOTrxN@57(q_CJmv>VMNF3Tgo-gzO_!8-M(eeA}d)?MP8y1-Ths?U<*=)$Lf_*Ukkugb=MZQV9SMRxH)f+@lLV=5O+@DTgzy>?%?<)K-IFx>^(= zpG3E+9o0ajc}8&X=X}b1Y8X)x6PG|EPL4YIBONy!SY9l-E=DaEV*d3THE(Nj_hh=- zEjush1Z}(*UVol=>cy!G3OljpO^ca*U$AP(^Mdc$X%A?8D-cU@m4j4Zq&c>01{>G^ z1)l$%;g{Zo&9kguJ4qo2GqC3oDugrQmC^q!W9(OMvgAg;wRVVQ<-h)O4 zJ@GK1yyATnMYHd>Oo=l@7_ART2IG54Z`su-%x*aRlz9bP5Nsh?l?Pjc;2X5s^Vwi* z)LW_(bge>Tjo!ybh&NN*09ZlyVTM@scw@!@#~wJYKk1Ow>(Hg1ily8&=rmCMD>Fm` zRXil(yv7|(he&VFg7~yNYFMFXpFy%J7w&GtKTZjDEw*xFChzPoLZ2c+hbk{TvZ&BG zmDH~5k;s}5_DE7R^3ceOEE{4I!ar|1hc)F?1XU15L_W7o^wmB=c!fG>_VKxX!1O6p zeZi6Z)gs)GF19BNU_^SDF#XR#^k>i5BVn6BEhFY?(YF=(vkmzeaveY)cPbPi4VIA3 zH8=`)-$?vTj{CsPMEnZyT4Lv~^Msg#65c~GgG3bafpq=E&}jxm228b}{UCU@_C{9` zpYOxd8-z5%680=cLTE`(cj6!dV8IC*8EH1suG4|tOMi}(lFYt=KsJG;+R7C^F)sfk z0%du;_d!5|zy$8VX1UHhN>f4L20!7bH(^iC;R{&3hPVA#XARL;M?S+|0 zSd`>faR2r^p%2d)un9rw98Bp05m#SCy>?**Kh9p1S15|Wk

>u})NF%C2rQFq$}k zLYsf)fM)&M$|749EAi2i;y;a{i)uH^Lpztjq5YJky~VStP8mWGnk6*r}Q;1Zl+! z)ZAoKTF0Hnw9tlaOZshfOU|4vJl~yhMVZ& zvOo&j!Hl8|GYqvLMNmOdMOg5PrT|x@BNQ&p6Uz=+&d5WNng6!-+o(@j7GKH>thbnP z6sz2DN3Z`n)vW99^TGq52;Heecw6gcg;}U3lt3U|6))g@i8&weFED_gx_>vdzz#aW zEp+@V==dklkq_WQF9_KGnWs}}Kfztlx5HWhKL&95DSbH*maOa~x1mZ)32G=h;GZwZ zDcHRGX)!rr>xfZc>Lm?)-HT^O{N@G9a&E+l;DV5ROlUy6UB95K)KSsf0|6uHkx^PS*2! z)TIT6mj_jo0IhWRYfi8)mO*xNP(3P<`#u0b`}%6y>TrG$(@n8=3c1O16(+=97Y)%$ z^eMnWrxOAzI3Z7SIJp>32nCoUKos6AgfP-cV_<4@&3Lp&8_;tp8QZT~H)XqoYl_U1 z-D_u8Hsuv?a0TY@-GiHBXr*?{b@nDDb~=5oDc3;Py!~t63$$N^i%=j#;eh%Afi(pJ zYYGH56bNjn;22P(F`$S)0mBVX=(k@i4A)QxVr~NA)B_Y>Ju9q0jaX89mb}xW>0hK# z#Y#oq^YA%w;i&bVD$Xwg3eW;2aLLUz=cKU2d89pUqY*3ibkH~;DEv2%_yLZ$em{o@ zonGdLUU|P7#4j6?ut?!h630N{TTcJPvc+E9S-n_2%b8KQAzv{aRg;n%ynWb0 z5P$yE%!6$2tu`wA>*`v-ZWjL?U2;qIi*S&h=HI~ zDs%cb$1%#MODk4<+SG0Lg_qT$k&=tjIZT zDu@_I*}YTz5DpPXW;79X-?D*}x^vxkn#~k=SshAb*OKYkl)zv&gzb~VG;6?x`r-_^n|zM*a5FLaRGy^jAtOIFVR-W0;f%=SM) zO9r<82xM)J{;xn*1AY>`&Q14~HX?|5A{fOSQ0O+0#pMCPC5rZqW~USv_e9+-^O95= z2s(+1$T3hhaIkcw&c7Z{zQ^h52P0B+o$4zJbes{%X^zUM1p8<@t3k#NDpy`^qDJVE z1dy&GE~QwykUff!Xt07hOZ7Pm+Q?9j(rNSZKbQO=XBuH*j@e6C>kf!QMI4ZC`cTV| z1)m}V5h%_Q#tWbJKq-*jM6>#ikWfME1BNmFE6`dGYl%sQ) zab%+sD1qdrFx^_hI@9Y|mX;FcYw@9@_D1pr%CxTV!!yRP5(@dnBm>6)F=Q~5PQxbM z^i@5&!^i1YC#GKCJ1%leJ8q zvQ)KH+QPc5Eu=BEri+AR4J$_chG4)Y=7~k5^EZPyV62k%VuXgR$fkKh4FjSw#39jU zMfoEyG-C3_^R=SQGIt=$2&i7X$+9?sxSf;H$yp z`Cv#yUGTREY$gXWBwG2PTY`y=+hJ}Jk}uRd1%fqw186ozjJL&5$o7Y! ze9<8&P(skR;l@os6BLm#z&*;*Y)SFaMr@z`GDP|F+z4P_UFHenYA$0lM*wA?K zH~f`(n+?Ox=fFhLRz{%?hnnnVz1`7nl;BqPeUI&NDR`UVq?iR;F~aAuyKJva{f=Cd z_stLW>s+HyZ%)>1FMeD7Klrr$7MDQq{xMvX(MIS1f0kegw;RT=QCIJ#z|fG4Ps9z_ zD9hK2byIGiM!M9enZ7q9FKzF70eDtK&t&e59{u0ub|G?7H~S`5H}6E@iHSd)xVTaS zlw_Ak{q~Jl;;V^!QvK!q`reqMg*)YECJL}5&CJf>hr#Gz0vKg<%joIRMhIf%4O@

&?E=Na^YcNao&&2( ziS(#+dlD5FtX>em6p{W+2wy?T#jVNY-$q-cXD$;Q(1$m6<2I0qAu=nlZZOfAsj&`753<- zsyLlZqE1P61L;UCh9k0O&}rZodlR?IP%!F+}mwtSe!=S88%B$dVNe2i$1&jer$ z)gKK}DA)D#1k{3(92-!>AU7iQAEJF5(89Q;PsIzq5IzW)aZV;LL@mwvQ51kg6vc}c z7yt_kp_uRE55n^smj*DL9tXC>wF-JGr6owauDI`PY27@&aB2H??b7bv zbk*Ui9^jaa-`AWlk&aR4+!Gs(JaQ6`?^{QRdzzaBy!)_3g{=ZNnyX%H^5<<)4H;Is{k@XT_bL|@+wVBW2YQ+WZG7xu#!B7#gZgn6 zoUc`dgT|1iP^}j`^$~oE9v)vJI=Vpg$D*Ac_O)ccmt|`ee9`}2$Gt912AE9DoJ~LI zZ1n)Jm*LnULC&0z9%#}iM5yJA64nJ!x>(2=47p|v z2j`W-cx%^^RpX$?OLkuRGikbuHD=*%5H3n4u@rL-k#L4OI+9W`YKac!= zwsPb_5v5cia?%Xprz+NfgjWqPz+R~qF7j$(L7iSwdq0Q6cqVHUez57X z-0&N1c;#E?iP5~4I=qWk)p8V|Lzv|ZH#tg7F)9LqkvivJ%7a#IDFwYivZYZ1{{z2kN zSaI64nV>zq#I{kwr7}g-|Dkb@ye8$_Y_zNF_|kft*g$i=C{}c2`k0pB#pc)dWvk=E z>M`X_{pn4!4!=vA!wSM%H7%|hq8sQjwlatlf=sF z``(8SwT^y))ZchDWy_&TcT9=WDL|2|O`)vypJT3~^>6rFo|U!#uk}0-U!F@%Qu!YE z`4|>C<>>^?Hevh~J;{Nma|PVzoW~C8)8^vv(_K*$29{0sZ5jaTy`2~Mm9G=Xnd5D? zwNCPL&&D1ODV|8zG-FdA|6UL7UXNgZ<35?o_{qXD^=(#@&qbL!M*FFF&j<(ArcwdG zLJJsw?6y}+BY^}kYX~sLuXN9y@cSdLrcn<*rUX1_{i#gdnqvN;@vJ&yvE{}O?iAOu zcIqu269yHSidWD-Hi-xk_}!9P5NeALxO_FU{8K&r3m%2->H8m`!p8C6n82t8$=!+H&B0YiVmQM z)NX5*LYVjZn4ufLGM)P~4dWY+z`^+-*f;7|4<-5l5zgn; zatH9mB74HL2aN%u4iv}dfq_WIjE#4_I|~B;zB?PjLgaIv7FaT$ErP%7_732yMh6r! z9vvlMecET)ffS5tSHqhJFGR6{S8KbEiIpI0J?P z7yz4pr(pThm>vTq@)WRt0ByhK?~;$lWfa(qkPivA03Js45zdqbRL*Kn@^gL|7j;n` z@0W+N%#I!xd2zWq1@=nj)0J*#yWMs3c`UB_e6dWey6RGeR#kPeZ}d~B$@(a{sA^_r z7DwfQRHUUxe~X^yt)v2vR%Ovd^%K%T=hsVL88!q;ksK9ex^RfKC=w78()3TT16oAU zxY1F;lNpYZUrGBGSU|u28pHK?d@mGd@8kTQlR+gUJj+Rp_}ktZpC28@kQ)SN@X?e> z$99BW6V2A^S=i!2T6~?mEq%W3Q^Q%x6pIuPGZ_KvgaBA7j|Qls6@mDJ8*M-hf*BEt zNeH}Pm|-jujEL(3!KKo9gdr7quF`CO7oqo;OVnbYdhk*lj#Woi{VssJY}dH|29fx8E_g^>b6zD{AYqi8_~ zz%BOC9^VOY(Mm_qRs(%!eR*JV%$1+mx;q)Y=?iUeaP8xc(%l1ZXH#yIK^E+c5#Da= zsL#_TYvCx_aip`M)2U;-f&4>r;=?R%b<{fa-U#T0PviU7+Bsdhe}dzz()*zuZ0gg9 zd%RS;T0Op>V8V%o-92!qX{>7GNRpVQ4$sdOj*pH1)m)II#qj*_qtZj3HxymP=li`4 z9@Szz31A~)nlA}KWoSp#!45Tw8|E8L{E8(>p^0wazeFRJXa`IQ>j!&F`}hgYx4Q|7 zZ#7KzbH(4iH9byF2ySFZ1LKdR_}E{316j2~HoQeVzgHt3*zqQ(9t2d1mH=byj3`AM*+47k^jTG0!o&v- zxUUW-l>H$tJOG`>eZwU#xwm`AAa{w{3Eoeo)Ad5$*wJ$Z2EkW&3=jD?nfs3yO?Gs> zoczC!NsndlE%sneV!N+Y+K`ia(DL(C;yV zB11WC9|`n*yOMob;qKlr@wPo*R*QPNcJu>431dAMj~7vuwPefHS}IfPomHI|&p!6O z>)tw8rvnL^KFbjsLyk6Yj|Np}L(qsxA{*bnk9XZUtce~Id{g$fRUal!Q3CkS&(%5E z|NO?Mz(##Ln=I!I=CRL)PSQdPVmO8joV_&TS} z4xeoi>Kd}H`yvUqKS#ImWKt+Wd#f5LYX%&86O~)%XQvS2^~I;ADBmLf$rNDolu?Et zWzp!m0_+vLQR7JEr&LY_-*x*-b`C{~E%bE)kr%atM5wW!qdu5S)^;hg1gClY>5V1c%4#W*-RqHU2&ou1Q$7p9n78|n%ty0)* z-Qy@qI^PeK`k?!esp+AOWcbS(bhAxza2fhCP)J65V2yZ;s^?>EloiSe$Hp*P;1Xjz z++qxUzt)CAw~H>Q?p?X@evk3*X98+KIMJ3v1kA2Bt!t+Ov(0B28fq{-8n-~rAG(xn zaUXvF-3YVp5tdU6u^tg_Js{M+kFRkbH%DN?8jc2IEE1f)9@)}sZqxK}qo~5)uBh?d zrX?#>giI2&;I`zOi*E+y5q~qMge^%m(Z+1&+op=5G?98^iqSUN@tmw5r8>`fPa8Qo zdlxSIF>M&6$P1O}uwF?>4&}_q>pMl_G<38}B@g0|p+(bmhEwv# z%&}?yhCFA@x+)q`-X$A-4@J0(#v3gPV;rpgf?_W5)R}8zanCP9(c5fjh=@sBsWD@P zt0)^}70!3fLH&7GreG){@R3yRi=yNn#BTdf=AZ{oX_cCy8t~qGr?Yo`w|7BAqHx-$ zjcHO^gwjUbu0z3+njtMy#^kI^{V{baA*utAd?tbb?&|C!D528t%9pK86_w ziuEu0$&sGP?Q`k$+9A_Rdo=IOO3JZ4MYY3^vBsZ?(#k9ydedi1%3$h0J?mB(X#2>( z#cfl~b7@7L<+t+hbXqUI4DMuvH_3+>^`CQ3Lvkavb)T+$&yJi{_P3O%iQJyc|+WNi8PrpV#-8pXGmb_0(l>fddFwQED zQp4BQY_7G)X|R5X?*Q=LBo2=^4A(WvYs5|Nm&tibHXmRe=?6BRuC&3XPlzgKn(g-Z z70PwbK}m@*1Y`VEOf`9)?SU~!AmlV=COR`?Z!p+A!6CD`J4}{;M?Y%wVc~$#*S?7r zU0Ev2M%Ox47JiK#jw=yx0~9>NHvW&ly8nKhZRXd6f}4(lk{?(MITBGcD8hu}s6z;F z>u7rw7K3G}6tY8^teKrw{TY`j9I9CtuXnu4srwi5i8(Hf>B?@y5%!gmL(Da}H78)c z!WES{8(Y8HRxA=pUuKUU+bTe&nDLtL~(as9=NP zjtBr#Y?7`@e$P>RZo7Izxwl}}{)9;<96D96+uzP=mj>5r;LzNi6W(oSTB#>nyP20+ zt5fY~sphj-fllqAU|EA)3Ki#aYpfg>h36k=amp&NVt6#Rt4o9ldl*P|N-7Mm@6u;4 z)wJNafg>D&B#&8?C@)T)14z3Qy~zsYlQ?vCmE`D1$7j?zQ9F;2!V5aFQcBPe{w!nI zXCu*{Xg_4l1_9s9ElPqcBx#H=VNN@yz7G~7jHYSM?ih#tDfJLOw_pv$DBz`Q~q(fVJDe1NTcjM=I z%|35$cpD6y32>H zTN_A2f@=#qWRNL}DbMv~k>N99%HfmZdz3Xv7+G?qY8dEVMxQoy#!}+N<+$=3 z%b}>!RPXwt>BUjCPA@GAJzZM)P5s$hOX{T#&8(9}b(i*XBF;p=ZfTi*xhzi(z3zl* zwG#chx}w*YNPhX`ynRur@p>GtI)0aS9WTfb%U5^*H-VJfOt{ojfB<{p&IupJ>6AFf zdDhj4-L~-N?}|aB$S@a*#hSTXBbM(Kb5nS)i;Nbqlk>0)>(0%PJ8)={HQyBZ`*i*L z*Nx5`Ah5=%apk3cSko`q12gsbe}Ex7$A2%RVESjE{O=E@|7X{|Zk%=-qR(yJ-k(5$ z?rh9NVYGMuLlOZfaE7a9^F@?USM?nV%{sKi6<58FO|Q_xqv3UyUMtB~?ae#AohnpL z9JteL%)c#YE&~>nu~LIAT#(Bauymwb>W&+=SC-T5k;F(=BI1LQVvflhNmmk^w-g_h zD-^N7o-10oPK4@0_BhK}M;xc&VBB;Kg@mzHP+0R33PQx3k^@-l5s5yREV4&zl`DR4uPB7;6QKXu}I#2^x^Q01?Vvdr{`3a1&NnuN*Ex8r#Dw}PUQz-X{ z;Ax_Fh)GisH3M{3qpJ=@fIVojauplMXHxm&gXw-spAcLN)#m9$NFpX#v?w~mKMmjg z43NqLIn+=oDA@3RbK$k2Mtw;-CRt3*x_>6k%d0qzASi1$;)s?NA>O0{fG8MF6c|7U ziJ#ChYm7TbaY1eR!c1V6jl(s>2bG)a zkj^)LK%>ZXWH=rqp+vswHDH|&lMHzm#3W<>_sDT|hXKx{DkB|CR5F410B%M4lZN1e z@xmYM_)1B5A4InF76vex!vBSW+xBDYvthPxmqB&gch|G(%skw>v*>AA;`m?ndipg# zJTRw^&+x0{KhF2Y>Gc1uo##g%2gVwI@n%z-rWV&PgkR6gY~Q>ac(Q1>`5UILZ~M0F z+qMN@x;z-Tbq{+{W?KU!ef-hd)%0ytL%jV__tEk7@~X3FUQg2aua7#rb*dR zy=K!&yFZm_CN^wR7_@0?(=-ZN7xW9dPQW+KUelr$!p;dme!l`zeVn?%1cqUF^ugDlP&aHR3V zXkaw?3wuqK>{96yu=d_!d@~`aPz-)1e@FH?B;e64$jg6ESig1hzPY~Jw`R|(4S|BK z@g5==DBz6VgqB~yvT}&iGZgjW3GmbVce~jMZIT|4@L_Y(Tm+<~=l8et5HWoHC|sXJ ztxtwBP21Al-TuD)U6y?%0L%?2OIt-p?{nmu7tasVT`rztz+t3Z9}@y!dA#_zsHA=& z%*3IP0lLR%ks9q~brZgdiLtp39syS}x9IV5xsVLaVRr5j)r zmsa=RN$nYIA+^>EwPF>P>@eItD<>b+qZnXE zFh_Lt29JmNYK8s0ku{H^;H;mWCxQUz(1MEIjKo{BHwwupeQVrTpo&>7vAkNm<6`s7 z5#%=VKD0E&>*22Yu(TLQ@m_9j$pXP?d%EmmFVOvh;WN>c8x0LNAYl_cb~o>mOyoJ< z<;aF`Gm{)fL5$UmY|msxzF}9wb}{M)v_Vlt7~QhqLetX|J06Js~=9WmLt#o`B&l zF%c7>7~-OR#gV{qgNAlWE9AL1;r&hYV$a(2(cIaeLr}!P4gyaG6Vttv~|5w0|?DWoePLO_pp zhe~_GBguwnij6XV`!A*vw{5u{fhMK@6OYs<1iVti5gGs(7(FOnXO0E%#;fJs@9TMQ zGS>Z1(UJadEvl_`E$=(o|7SADc{9kLOO0`EsWH)q5R)AGU~mQ2_Ric@ls042QD9Vs zA5fnr>SL7ol;~{nH=VPD2ZS^WBKtW@>oX9>hLHPH6mNklP!r06Cph-0d90?eBG2`R zva4H_eI&*0Zy(1N7e_)e;CMs%h3e4aR!!LjMA~5^i_Z9sXGF1*BtN457*_{&BQw7( z3pS>JMMC5c2J0e>o^dzKu@+=q`CQzut z=_OOfadU&PBoKt5nzPecH9&tu^JtILE3t-)!e=+T({{MwX~B}u zD3kwrNf*CYd$P}N&WEfEvBJ{+W1f;=opQXEU3!{6fU=J!=`efG+m8cipcI#=44%0Z zK<1J!iEFL|p6UE2LTU@Wzu0lB{LK(y8B!v{_g8Rc*DstQ=Ar$}`u}0<9h)?PqAcA+ zrES}`ZQHhOJFC*Dv~An0v~62&+Mev0zV}8<--zyr{b5I(U$D;F>sdZcdA&9FYU;W7 z{Sn*v%S!O1cy!}!=SDVz`Y2k&j5?DMXSEjJ!X#Sh{PwdoyKt_N#VUUr=160 zzy8f_-$)&@tSCrcC`NjAfNI9aDm;O73PoO1=LlO8G%fWWobOx5X|bemfnNEznahv^ zPjR0Z^=h2=a}d@oFTWuDnsDRfoIwT~BnJ6`&k?SmHmtT5CdbyXgq2busC2>BYCY<_ zh~fouJXRFz!vVEXXUA7*j4Wd>t^@NTFqkTQyW4kvnHtb(GrC}1ZjA4NjgX;G2pGrKs-5{=%zn| zogTq;y&;l^r?ktHwxGm2rDs=cPj_xaKma(LCZZ)P_%BF2tG6F2 z1S^GL^!sM2;v*1SB~=M7D))zBLQY{7I*BYfeqF*Uyfbw@d00zZL>J>IuVK%Ne@Sy- z$6t*FQg_#TScp>nk@7?xSy~!AWZ~5VR}zACAe>Q-{RgMBk8c}%QEgq>m~6Sulo%)t zW(BKMfiGkVpSOFktr#4BW1Airc4G?SsftdzJD-|<>g8WQLic80mmUU=+zUJSYH(V7N!O$iLa5J|o z8qw6K`>1rZ*{1iIV~gC->#ORk-I*;Ok1uBWwM0JLW=f!7_u{|CBV^h;?t30X3pm?$ zZDtG~1R);R(41A~P_4}LeHuPwCBZ>fJ`@PsP=1awvsa&6T@(w=1~WA^Re>*K&AHin z2oxAm_j`?%3Y05OCxKY2DAFWMgND$saun|fr48LTFJZ>t5WueB&HqJ2adQ6mv<)Vv z|JfPf{|{~BKe_B3lC|b=Cco8R-V(*iZ3SE z#&XDBZE<1KH3bh0ix~C*=`Za%0FPr9(s696R8!}b5t2oCG1Oe^8KMNo>5g2E2rfq*5J8nQ zj%N9rmQ$3M!kD8*ZH)!4jIy4mebW=lPTc>4!-zDVA!m=OpA%uAG>4Uuvm{x^p_YwE z8-~w=JGNwN>L2S!p9Kei>!~*NKV4KG|uF}$DG+rH_7wo_?{F$oDLfD{6NZl{&Cp9BzAToSBI5fHghP)5^6(G%Ccza%| z-m2A7n|XA7Q768&V)VOsDZjccV41Zo0Nrum{XH$Ht2cM+iq5AvolxgOa_e)Qcd#b9 z`cVCfl!>n@Wn}`7j<(``!ycAduUx)oRDBN-{GhsLSrs1-%rCiF=mS!P@y9P^Sz$aFB0 zpq@T|WL2cSb^@m>OU|&r!5I_R1^V`<>e^Z(^Ol!2Me4)q&NRizXY{I762f8!V-{jV zIuiKQWd;&65v6HRnpF(LjcA%6lBvqweW^tdtvE8LD*MR*pZ75`i!r!k=n;_5(h^wF8zR+o~I_b`AbwJiCvC?37P?*}_ z7^9td=-u|;dFb@<;Vx|*A?;%HAmBzigZ?SSi&P+zOv8jFvac&s8i6;V+`T7Z&2;*1 zD?;zqoLL?)zeq(FnmqX&vVL?C==9}d(YZpOu;<#eHP)Ul|auX_K4B9$BK}M*MIcOD$Oon3qIU_kcu(SNFDKX>(x62Q;;wDRzO%a32 z=}M%9ZIy77KvD)F4sJiyFk}~+pqhIRn);ddqKPyxPEWx8RGi0fbl8+0bYhxMH9l&z z4hYK#iu#-HFSF%zmm(k2$0rk$qw6X=$*p9{Vu5wvMWS~gY~m=4wUR`@cXP&jGg8f} z(=EC8tolMfSWLBpUgK^?LS2uP?5UBEa;EuF)u7oCz7GeKK^b zpng6%`r{fpmdpcHE@ug;a3BiGk-U@;CN;>$MU)_NNB~QOTcCX(0b>E_AZJ)x&Ya|% zU`O4voh_8Ajvh^LP+(yn(34+h3Hn zGZW#p4W-`;AL4z8nEsFn3J1$=9M)OrG0^V-9#3ugXEGd|}9kXh+m z1V3ktAsmw}tD`1Il7ks+L74(mW5-bw>vT{K{7LBZkK)jN{VQj*r|8!MxD!}k-Q6@- zx##2z3nW1xyk0m`q=M*zy0ax~IV5S^c7u55HstR>=pDleb|w6H+dp{bdN+7+*rSX#4Dy2py7ep=r7tEh z78)*=iSksy;d*2C8R<()*b7rLuCq|Hh_e<49~;t)M9d~T>tlk}-MN0DL$N3Cx7&@G zKQzfvRkdJS`^4LkAYQvTun!D`5q=w~$n_P0_y4>R75o}dNqOeR8s07%O9cK^8r#X+ zhrzjHua*B7-p>=SV`~@$8}f=^jI$Tdu8bdwkGY+Nv3(I!$2yvhZ8!znL?))*%EWdZ zGm{;>EC$%xB+&Z={jV!6d8x%Vv2(bpF)qi`U)hP1Ju1#1JP#g2%UE~bBVVQEZqprw zY1}~WAz9|QzFaCVscQ^-N{^J|PKwR4OM1ta_(Wm18!!_uaQ$rS(`d-Gpa%Q-9%GQ* zMPnY7)r~nG@cu?a>&=JS5q?6~`YgKRFKebuhtH+DHry1Bi6NWbuJeGPri-b%t*73q z&R6^3XO4@?C4h6M59~7f@D^v<`9$i*`y1!1Lv4bCM?Da+@|WkESLXqlCr{ou(&rSW zBK$V{LGA^rL_a`$RP>oUm^eHUvJlyBr`F84PZV%?QGWk1)6>i=FbcBeut7_qsKZzS zNixympw+*0Ht6G4@U-#zFAbbtux6l%kB1$2vVFiU-cD}92v#k9!qfPM7Xl_Yj!P=m zw{{>sp$kug4|h6rX|eFiKETxLJMPHJ#ASU|*WQ+Xr(LsZ#MtIFD{3eOI*tfxa(tYX zNq>b=zRvw)D24v=^kglzm$d@jeFV59dtmjXTC?JW$st)hdK@6yn9awXzJ0o@UofjLEYez2%hZ~~Y=fec#Ee~z}H8bdM-@hWHg5V0<3*6Ci z-CF8f@9s`LzMK@|v|T}MQ&;m|Z6H9M>ATg*Om*5Y;pYN|%yPjXIU?X6dks^t34CW_N}p9lv3lnw5!d|`$H3YmF@2kML1BWEtUU8U8NK=t zQ>QGL#0LT3`J7Uf)!K2ety0|?^)h8VmUX{_`ZVgY^prE6Ex`odJ%$_I2fkxhG*sLv zYlCgc39gMTbeB}qfhZn@{t}m`zBG?ao|Vi-jDif@M(Dg^PCGu+xRbtZcdN=BjD)K3I$)CQ+FYr&>f&{hkCXR6>JRZgsnheo-KU2gA7&${cm{Z1UC z=ugJH_p)TstyRO};ByERcL*aqk zJR^DO+*v;5Y~3w^W>*RsjuT*Z8*KXUluIx;jW8{7ed7rIhPGEsKjC1)s#jMHn zvnu~)_q*(QpPtI8<-mxFd8Z&8_@Zx%hXUa>Rc+xBL0{T%BAXG#;6I)$`)m*G^kx4rLCR z*Jm);4@>Gto>+HebkiXM1~pWKi5&iXX8sZ&($RPIGqzh;+Gfu_v#ZK}Vpp#MY$(EA zJ%bRn?ZC*S_~WC~e)0m9HNO?56LMgN1Gi>#>~7;-Txkw`NxqX@$2ABHe=;=}PAzyh z>s8ItP^?A09)K+T#FMN9vIRvPc2UPt*`;cy86S04AUfN$cI@kr?ZgZUI*%dO9ctMx zUN3&jeFB!A8JKdT6m))Epq9*a4U5z$CF$}N<1g^lhQE}Q7jN^@&kDH4kdo+frZqYZg z`!NjA76kdb35z(rXg{>D(Rk!Lz;GEj!tSsL92x8_6U$73@1J4D&>2Dh*trlrf%fNv z&gTw~3%8SHo!KS_MF&TRQIVQ1{L$oAPFrpR`;MqO(oo=SN66yY8>^%%VTe9!q~AeN z!Qv!64`_1|{G1$qtHmD=0|FTeL9Z;vBT4uso)Kdjr?3=N=$Jae0i?37*h~Z#A@ZSqLsgZ6T%;tgi#T1y+yRk~ zIC9F?EEqE@jj0mnRz%NFDRxzG%gsv7iO16Lrdj_wn?wQ55cf=g48gpk+*8Z@n

~ zG?A^x)FbB2Ttx*N#l4mDE|ilkDFgp>#m zjYD8bdtqAU0WzJuRr3cT26Wu)s4$W~-TXSQsj`tKpj)5LaNjTrbyzBw&sKfbK(2({ zO#o7K^DSvQ8~jWsx)FkjawDTE6)ol!%QX^M1FH_9m9&Nwbot~?TA%@upK5_78`NhN=fe~E;` zmyjC`3aMa&srap|yA)H9hmI-Nm+Gn88{R3n`|oq*b`Y$B}x&5mBIKPYnPeWDWGBFu5ehvv`>U%JeGSl( zshal-BOFPAB`WyAB?Q?8r5@oWsz?=MXH!oQR8X6!BEz2xtK zrTRlOEbCTQ%4#1LTcJ^*@BE{ z_nO!PWXRS&C@_L~g-&d^f2#r?=v%4JaVl_*i0c1j`??31|0Z^em@Szd9S__u459*` z0^$eesPotxgho4?eTQ~f*0|=|Of<05X!UQ-sb(NNxZ~qpdNlWQ;xWi!s7=|JDNSvi zh#FdQZf?6>!EhKxJJGZ<4c;g~Z+r4XR?K(2sm_8LcF}NwbEIpUIQ?T=D=QrE7)LO3 zw(&T!aJk;Q<==+=Y%SNUI_j_auJ*PNcW&RjD<{34uEdNC>0u0YNwMrc6wsitwB7lLta1%tF zf^l##_yucQv#?bII#HB?OinR~@C?bR;<0|%>I}|a@ zK`DQj@3v&>pGDw)A^ug~0(0BkUaD)FoAbB|V$1tl+o(U3xZF(aUQE04xEh;}`7jkr zRQ$aeA#^oJ_zbIfg_So?SLX^t%U7(*y+E05m9%YihL-}7KD@U*MuXD(T~BibkNih9 z+%s>%35tdl~5Tp&eaD|ODy}JqE^1%oXuAEth1)i zXRia1L(9ZcOKGg=NO#yKR+zH&@48O|c0o2$>Sl?yr7E^7~GEMDfn{*H`Jeu&t%}BBf zw4J$o6;)F_Q=;vmcc}^jZUZD1-~28EJ30Obe@=QgCvVqmJAX_J0>20E&hA%JO!|PQ z)XLum)dp}(o`Y%d4mVfBD53jy1>W2c#@5q~ zc)2PdlVfSpC*Bdq&FM2QLnFyjc@8o-YVjoiTRr&hxLxDN!JeRL%1-iwI!jhC{IhVy z5z5$)h`sZa>M=s&-RUYRlh}`HHk5Y0v-CsfJb{uF{<>N=^WU@MEMpx~KuFBIBcc=2fO8@)jFqXEiFMfCU)h=Z8 zjeA%kwPjf{I=s2lmg_fHwxhpGFNvXy;a-@grap1$8^IJ${KWBL4`1lt^XBEi>qzIo zDiLTjMXFH3_^|v~KE+YF!3Nk~zNUd~B}u%Tptx89tZmZr@xJcUq&a-uql>J9T++Up zzFB_-(MNZ}?eKox)yFkgD1c0}Qg1;rb_B=D2(|^q0+bL~ms360Pu~@nuIGLG4CnvR zp&jh672FnO(Cn#cC4wUo{gT<`hGt%E1OBKhmQER^sT)|`1$?nE5#wdgN;Gc{g`DaVBHDdBa zP^;r}KjhoECFkhSNbq&5fI6ZnCftqhC$`tr)!p${?GgM4>8df6Bjbz;JVB|z28wYw zXj}HHtE6a+mkyC%m`vlIl3Y6U4ooc{JTzI^#3d40pAPUo!b?O3~m0u-a=U?LKjQ#nlfgvuSQeNmB!E|ID7$}W24iN-H1g<2I@)Bs# zD^Z6aOl@;~xku@d@0uT?2_Qu03BOE!BLEZ>8fhG|+{q~tgswg_lrRVjzTq8}4e~d` zlW~%pQNY&8@YFTBpO*K-)uS<6+ltraKZvNVutw6z1bqRNW#z8x7SBY!CT>yrj0CdnzpI}jaz$5uUm1YDNSrFZ@UW@AjKh8a2jzQGjn%Z_{WraGN2u#_2hs0}2ClemX&TKy| zAXjqawru_Sol34mFg`{X0Nlmkl{EYdv9=FfNR@N}bJ}EM5sEE7YkNIZxUwOZW&zVo zy8+IsOj~t>387S`2-=h)^gtfX{^d*&M0;K6>&!Yna^@QYr8+WT z+|FiXs!X*{6I}zPwdrgmMH#nLo{e5E0pBrYD})p9Zhd)@X~f^{&!+=-kOaq{55GKG zJw**X8lP9ytUiBc<5}rz7L1L_<}2m2DgLeZ{mW`|RN^|f?*{iMhbCZ!&!|Y$)Dff| zRNk0*Zp7syB@f8WyMW(9tLl8s|2AnCZp9q7!}XV_61R!XP4WylN~U| zauT7~7L00Mtq;^hb-V|45+B7wE*@W|;8?bUD2!CuEp}R8a1a0k*)vZL2}DwJ{3DTg z*-a#hV*T3}Fn{dLRESvEt2=3+_r7to|1{6gGH@x!)6cl6JC(Zd2-5y|v1HBidX+hH zd7e3uYLY3?Zw?f%R8Nfc3th-Y(LkABnTu8G^=rs zJEigxA|hLrFXXFni-dIPt;XvxecKORuI$dZ+{2Ci2Xhs$ z!AwGx-e|ECJxZE!J7;WZ%$9^KWY;1U95aoR!8+@7C6R?Ifd)WNp&=p)mrX|N!82#Udri-}k_T7L;>Pj?3?$quxl zFkFX9`&p0tTxkg6?3eEOhZ_+F-5s%eyk^Iaa2r}jN_;LTN%*O5xIma#fV;hvBgFBq z2CRL2Xte}8Y*QouGF-k|0+-dDj`_IH^!k1@WAzYc{A)uw;5*oOB~)oQ;m8}rZ|@z+ z_@*~&=^X(Ygjdw7Y!&Ep>><|x8lp`6D2Tnd$7XP?H+T4ri!m8o1rzvuYb0_&o9uc8 z_P^&Po5%kocBlQoB!R1K)zk=^JI!QcyjRO?L^N=L&xEm<88zMO%rpd1YXFUrNa#V51~=2xXFRaKSyE z?e50BLlZRjAulJEkFrYIgNLMoKiQ|^UQb2{;Dsdw-B4m^JH@&ru=nr{R!|VB_+PaV z6UToKPGDyGpX6hHv=Q6?rHzyvNLrEnuWMIF4uy1`%{fV<2nC(x0#U6TISt$$`Ss zz^wcjF4*Ud>qmfU0qLKbQJ}QTno!g0p-{9MzQ$|mXm^6J8q$`pv<~G8@Mr^1*bPzM zlk(Vwlr@=b#9_?t5^sieej7BqX5XsMjrsUhnG5fTYP5*-k-cW)r@uw|)f0NDmHdeC z&&&ATbZM*6(%DA!tv1k)AYU>=_LWfh zSCXF-+D;kx>uAqR$)!nJrA6{T#On!T`uU!O-VWTW34@9lGPjkZ{lms3xJC-(1$gxU z^8Fx9__+M)o<*Nq_F`++NC?#?_dl)W%Shx+Y#WjBG14+sDuOLr=g@175f{9x-ldi} z>9*;}wC4)Be#?&ru20`uE8m|>9D3hxBbmS}=`pn@F}pNsunH5bO*4+6iWd;e!X3l=QA@N zeD0*+N~AFv*Rt|JaKvg**!dz zzU`a6-FzNVQ*OPR^b9HYYnSI;JMaRq8%Euj?k21=Yf5fYV_v(^fs4tJ*}##h-wL-H z!{7kCT3Pl*s|H}3M7Z6I==8B?=1tsYi@ohKKA5H>_pQJ}J%z3ORCcS9(HeT-ti;a( z=Xjs|O_(7LBG#acKDl`EL2^qJ?U~|7-76pVZ6^j??-r}=s^p5REE~fp*2I9mEMnu70tN4 zPkg=ANP?8(qYvjaR436sBy(nDCrvmfXwA5oo>mO`}H@k|$TThv+V z-vb~_S)LL*hDs2tqi@b9H5xh|jd<>zU$Nx@KAk%V8-!PvU*2riK-BVeOv%9qhF$#a zKdr0)fFwmm0KD08zv|#|K+>b<3t7C=ZfQL)a1Ar3fofABzQ%jrXL6>)^Tgw}kq!$z z*R@1FUQm6=wHD(Zs_5nRixDTi&DZIY!!UyM#UiFzuudbN>{w8-eC2ni$MvT1iTw?)_hexp8zs636@{!3zEb#iV#Ym-Xx|%tc@M# zuLgHH5-=!Cw4_9}An{4i1MxZSofQqDP|W;D?bpCV<9Z>a;UeOgxQWrH)y)D6F`a0l z7rI2fU!N)XK)qa_qkiySqaAp3#NJ?R8LzR=PQWdjez+HgXO@z;g&^8)0@D~$+kjYB zu@+|mUR)s0#QlYIp+_dc({*3JH|C^cd-lfG=dpQjm;RTlC9S>s!28-~u@R3o1xpFu zU#{R<&Fpm}5=3w^6^7w6@TDgRB4&AOU5p&;54}3mGp=1Op+1xL$@f(fb|C^N;Y9_E z(4qx!BA)B=&~yZ+V=ziiYDx?ygP9aNx3Pe5K|y2HIQZROsmGS~pH{z7U~XHkA=E&^ zTdv%j#>+>XR;8#Tk)N>yVGGKfebTk`q37*5`Df}wU$jGV6St4leFQ*ick-65P%;8h z!e@?&w8QXA+uqXC>azuHOW%Rj5uM6hx_nf8J36Yn^V;_B$c!S$K^7P>Ea!zCSZUi4 z&hy=T5W6Y>G=ID@JUjvy6G!Mz zG$3a_gZ9jeJrE}>eYrwp86vb)M1N#FVd54Ssv~bk=F2Nj@|xaLCz1QMMP7QgIGL#f z<^N}04}D0xC?3;gI9h1mCakykTL3$dq^aqKyGzb@18T*{sw|jgDAvp zSbs9So3&x!AD-G{Rn}YO*7O!I*e3IAqC6X~B7=&IwN+k*L;A9$zk7`~Ekd>mR`Qr} ze44O&;6asr+0h&dS0(Q2%Y)x&mc(Qknsc9^W?owX3%$fi3`38%Q7F>}k-(`5rp7_v zEK~6LzXj8eP@7g~3Zo`6e7)xtgU^E@A{$BEf>F*1dSlr6ap@v%H=4eEkOuvnfg`AF z?r0EN_K%9FljF5^FL(Q#>`aUQcU3`gMj1Eh$)UJ^+*bfx1Rb(Yk3fhbus@?qLrfNk zt-JGj$p(#I&sy(BIlzKR`ai#^Zr#!E#z~8zQd6AFg~q4 zUXk(4V5>5;y@ij4`^ji@89ER5LHB{l&Hk2B_{w!b?UghvnnSa46~os1E3R=kx7aH? zA|;>xR?CJmuw-t7iyQiH9DiKpMmVvcGkgoZ5=|2dFSp9{A3VJ?J^o$hLnyQxn7Sxc za>Zh>fcX0;pyEy_w8KnSiQCHcX$!j11A8D}7G&c$*d=0Yt@B_GCe}>Ii`G_GvB9M+ zNZ?;kyz%*am$g2t_-Zo)lc+i4p8@S2jU&>@f?X@f!d`Q~{?i#Fgo6&-b~uozDewoo z3LPPdat({3W1nPYU(4E5V3Q8MPF8C|^C-W{>FAWM5ybR<_q)$x_B=vmAa2s{_H*dY zyJz*0KZQS9MzX(P@)UN4d9zLvefEZ~;4&_gs3hwRt}Uz4wd){B0=Th`=1B`97~@Yt z6=1Q4J8EdH)p&-{10*eqE)`XWo~lJx*%s2%0Q-?8ZDGu?COHTf;`qGBd1K@AsBWu$ zZ@(qi0mC8HL*#q^(zh9VIdz~W!0*XleAON5uw0A(F2h(IA%Z+*)asU07hV||VLPr{4T zh;)%UO(GD^A@n~oX?5Cr*$eq#kM(_3JEw{kPBk1<>kwf(#j?1Pr3F*Sw2$(LZF3(J z)iLEtrE)x>)Vq9R*E^O6lGCq1pYXbK0%1W;9YZmwDuwaCja}hAGr`^ji6uZy zzfJJK2oGhdZJHNd`(zte>mB%gMG7R$yayHelL(38K8AyBj59#-2zmZF8AGrHsg)*X zKNxRIn~!_hde0F9|9c>hCra^r83ua_NKkOOT98enb`a#bT5xnVL4-RsZ&m>9ufEv9 zrgFo}@tcD3m9zE=Ud)oWmZxAVP0Xm3np!Cm?{tyjJxuW&eMrL@Pi)Vh>|I$|Pz5=x z7hAl%mJ=lj{C<2UHx_QMqj|E_2 z`JeFOeq;~p|0R3u;&A@INK3*%I?)+8@CP!ec;;f=LLLJ>{9U&7#BI0WJlR$1*Au@8 zUaiyxke1lGeqq#GClpWrLbE$DI^)^m|Qr$4118S88fiK zfm9+UfpIU*K5?mWLPyzR&HxEnOXsq0bF8!*DhocjF7lGKvMZ^r2MyP_Y4zgajcDft z{cr1THK%emF*UvYdF?TJE0scY%baQh)^VXcap-;~CnhTH<0M()5o1tkBChG?Ewd3c zl0T&K1r1W?q*93uC`L@>X-ex3m5I{s9N?u)+7p-++J*6>R0EFj2`V2J$UIZ#A*e$~ z@#}hS>1Ch9os;a_s4xIH5y1+23${&4`X5|)yx;dOr{UAu*>vJUFCtCC%7bsRF)N;TP#&^Sap?Zh>D%|=8f01!=j_(gok z8qpqiQc1klsK13VlY!N+OdFbqi(w)hg%vRXh@l=m!{xWR;6j%1LcE=`3cHII3#oOX z1)D`OIr|Qy+`N6okn8RiryXi%=^{3Zf!uo42Dqj+Mlb@r9e{mLc3#xQs>O)Rg`Cy! zGVR3BdIR({rV@@uD>Zme=fYEZl7XmzW3MD+U&NH6fFrGoFKWAp{kd+NjXj{z#%@`X z-GO>aV-ks#qmsruWFeCqi-2p@*p_^J(RVaONPY>MrMK@bdB^uD9fyGX-S=!*pvUug zxc1wrdC;SDW9ew%WGhJsD$Tl3>ifSnUpAuHY|K!uTSp{ z*0;~|{SBQ>!}TCTRFRyS!(a1-Udi*4N3vNs95o!6tziRMy4B=ZN& zhc$~}=4X^Yeh2kD-u^R%Inry1 zUpIrX(-WVi=3q7kecUlNQu?L~Bze3}WD3IJ>eQq8-~LWXA7{jtWzwPesNU+MIeS~z zDm}vu43Nwg+bh|z|Ly7BKumd2B!nS)Y!53auR&kc&(&JPrmlXi(Gk0j!rW%;aZ~2Z zo-{+M`O58UVkzVD-2l4Wsw)8S<$nCf2Q|nfun&noQvLN~5x0j8jC2{(Fya3{) zv|}k(&L8X^7~sQ{HOwrM`RrsA+ZO0+f&oVDM&uUMIcp+uSzq|Hsw9Kq9@SNh^^GUz z%&E#+e`wZXi7bLg5OC|+N{*$(~5NR)j;+8KhSR>@s-NLOs3xCmk$)`f6|fq;uNgMra* z_3iKG^x01zMpKn8*U6-Bw^NfE)fhALb@}x1+4KOe8G4Gc$7hJe*Gc4wa9?YN*eV6D z8JIw3;YddM+^K{YfCbo(&G$YI&hl2?EMKz#!HPeuChY?vXFsk!tj2dy-AW!l@+eKE zh(fsuVvPvGx^Eyx5CgKJRZk)foP32f1sgpa8OtFX+DWK4OMgh8MDP0L%o42L--UE)5>1XDJvYcoQusP|Zn_?Z zW(p&{mx5**m~x)QWQJS<66#mUi<1J)sTo=s#R{)(oIzDWxq$d{zg?6lm7i5j4#`HA zE@1_s!UU1%yV873NDaB$!XjjuD#?i5#NhNN$-92LhQ5Ud*|vDPc2nq-q z1c4Pi?_b2rOJGxkkO*$xN8kfD+`C%3&29Pi9V2{OQD^lV-75&aSUg?9Aq*hK$dJ+u z{>bKt2oanx`GWN_=1jdxtH@J7-q59m_E)}yBb}=ucAaK`rr%1FLDzP)rqRtEwNMFC zDs*RwdchQ7w3J|*Ek(T|%;WQ*8TdXjk6c7{$)PbsXJO#1g?dV4%{Hy-lwlJq2n z5q}Rd`r7hjPPW@w^?X^(#CS7~iRevBDtGOKy#Nanh4%Xq zV~2NJt0$1>Na}a4rB-U-mJ^6BF*x{>c!HW^_V}FJ_uVav3f>-2WvRR~E-n3xCOo#F zQQbb~$H4g*w6qD-QN4|wk;Ht z`1WDwa~YSEb!Suj7t>m0G|!4wLmLQz2JMRYKIAw9D%lW6%y?hwqD0PWaTpXDpqO1U zV+#9Fzy5^ckW!B4)9=L7T<+6S__s}=T%J=G2*0bm8N7#f4Tc|e@a4<+aKh`bWOTAR z$+euHGuSaxFsbU>+*rcZS}4zj#*@pgBFps2vY8a-jQk9J9Qa;V4&lN^j(VZD^2MHq z!NaHY-TkUKw!$fMi!Wpg`8O;BMWKoa^c)Kck>q5?P@i^jT9bvBnxNY^Ad!_tu8_{u z6oMN(up7^S=x~~mcf|AIKwiKwihYX61ia)jMbJw8R6Z{z#p!9XT~zder{?|AJtD!O z)(bLAG_OnjL+&1LRKl?C5+b|h)v#1qwO^MSva@OZHy$c8V{@wG)m#Ah!wLM`ypRF+ zsSJeQ*-sMwp*W<+7xm%DFpq&H#6OyBNpT!%ZV^^CThdG-!q+WFakhBB-Y;I%ve*=-yXq zgspOMn#a5h&{K4*022>$QQ*OSe?z6w5#8&St5zgBC7wIuKMZ!XKwG{6!S?8NQ+yu( zXS&z?zv6&2;L&&QCN1{x`TSy%YR&Wg*tL?lwmMl;B`Y+B#A1$f9czai~{x#(Q z%^Ax<3vY#&Mh|j6uwy zFUMIPzPdaa!80;ph^3j-N=s$6C-R;aFfu^Iy}fF zZBIo_tg=BawyjyvSg7bZF0N#RJV9+v`AcjfAq9OsplyNO>xfgvF9S`(RTwC2Ss+(y-Vo+G~DCuDzj*8tBW`wF#QrKfScY;q1*F*;^n9!KgmOmXAH6rW>=64f1QZ{l?(z`cW4u580wq}d`C26`u(m!( zc`}5cF0iEHO%?SF3JHd1Q$;w13w0P|G=p9prHbSvO~odn@_NbB=}zj+gXkU@bXSiH z#<-(G;`fEhL{SEjX2-<(^uJVJ(7OAZOEhOC_tK6k6c12?$R-kD6IFuf>iTKDc&{=V#v{!r|i3LV2C&X8V{RwUCQ^OW~foF4aXwKG3W zabFa<|LwINlQ+F%NO%^>VlBz%QMG42clhrXnNBswrZaPgp5js(I;!ln)=lSVv@Y#_ zT(eVLVR|($uktwS=`Xe}uLsE8XJJZD?Z)MUG=h2M>(g@7+xz19pRCz8?L3{j)u>*! z=_4K=uImSHk7bP_!yg5`rU+F`bV&lyKk?&5=8>k?yGC8RW5WrTN$^xfdc77D!)fp` zThugU7u07{2RdIet7?H6Gh|&Rn<# znuiT8!b%JVv;q*>cMA#^_QS7g>7j22gGR>NtF97 zT=_>X(($xE%wIkfYPJfrI_%u6ieN(ekupWcYyiPindgAt;YE*r$vgWHj%85GA4ev2 z?3>hbX3)r+=|k6wmWy^l3UjP=B0x=BLPGhl>$Q~l+}B~0QG9>g)tQ&g>ZM1I)q3z} zO)+$?qxqhjDVSGXU5!|_>C`CX|AR@QKJeSFC7-lw$YprvEpoVX>hh3QeYWcl`K6t) z4!f;eTPL=K{9ybBMeQa1g*E|B-^@ZB=x-r=wPNl)lVu)tV?=p>iQ&YC)DKQ8;>@h) zy$(=ceMGK~$mh2lnJ(_Ism-d*U~e+prp^3)|E01#R$crU&S>Mzji+BDY7DbDdv29N zK*Ef3_7L_px@fn-t+}nIGSS@|XTRY!^kKx83rH$!sg48u$hf^-7jQL~wK zP5Sn&XhJa7`RdC;6%ih`@!|ej>!tEh>rRti8S?jycWizH>RG2xcHk6SMPmi{^~6_q z34iVHz83t(Nuryz)E7t~=b@=YIO)dfeMx@i5Ymf1;xm6DyX4TC$Mp7Z%O(lZ+dj=5 znA~jn88n+DLE%!PE*{F~v6maXU9os@@o}(jJWOtTu()ESGL54vl0cG3gzRT(kHaM` z80z15%orF5_)oJKljsHMPuzz>*dmI|kh>JcbcIZ@P~C0 zNC-|5usexPV(sdF=oEOZI^d#KS|AaSa(HOmuK# z{l)9-x1$5ZA-IqTP4X3jE6r)~41H)n{x~Zq_c>`X`e`u!N$P6y5XH^G>JaKOZ&${M z6fj)pcc3Ni_uSq1oIyNp3!}z>Vw_Z@w~zGR`FPj^oSys(R`$P*hV zFYjq59EkcvykFz`mMu0E|M7NFnw(*kGjErg(o}L1K{F5%fY9}v^p_A>`2_x|<^>7C zHlnvdFiK}Sz2<_Gut5vc@3j9 zns~(kkP5G}0nCWEgkyoWF)FhPzt#)9cQ~ z{g_Kwhq{AzyNPV=a`0zecjV_bhimF{T8b)WhKciHbL;Xk_7p zeR~>U3CEa^H?L0f*9xHT`D~r2$*%`D+4#T@ znDJ6*%(@}6>O_mb$6U~u|N4x!oo7_>0TuJ?-_~p&_z+Fturh@NlsOV|;VHlEKB(nc z;eS1>j5_IE(nrRnCM77TQ5%!%z-V|Kb+@s>X)hi`^jKtJ9F1^HHUZXKs>~vRTc1s2 z984HThmYb00q)VU@+&|E5i&=cNk;2*ckgs26PBC$=*-ah6p*5RN@JuHzR?LTLOuKB zhp5UAN7;E?I=1cja;%{-XL00a1mf2a(5&Ih$M`Tc*|V`5%(YF!sQ6KV5{}dKYHeB6 zN;xfcGe4V@msL$q%bM?*J1;+>7~wUG7}EXfnD-SfD$y!r^njnoQ^E!2JPfnS3b}$~ z%(;&~1K(wzgzL`TC+HYlF$GRfj$g2yk*$$>VB@`2;W==G$jTd)hSJ_H&P#)1If~SI zp^4mxC%}9H+IIXN#IijRA+3)Dd73|sO5Y?0SnJW3!_o;Z+_XL1S6-qz z44d})je`Ls38Lr(Gv=HTWs`zCtJN>%ROD^vcFQ)Vw5N0T1BK(87jQXlYYeM(4&RJ~ z(E-}eFm9Z^SdTf~x>7Yvh1!v+14p-15PCT9=@cYsF>5lB<-Wm7rdVdTE)^zVLI zP5Q0xk3NpDhBQikJG@3PYGkS{5n|#AFC75(S_9lI`d^WTl-?8m36SjLP@%_(vig5>W^U z^EG@s5d)FWI9v-^zoa9gZ!n2D+@o=ohFmp)6uj+SF$wDeIm?%@K`*wUsGFM9Pi>I* za#odlN*x@_-ppyez;%iQ608t*wUJqV2>%$&q+RIjiX!6DPtPS5jmWPmt$%n;#V|I29Xjr?5%?%K3wh$j_H=BdyP2??{ zw#H+e!Xy(&=shkj`o?2<5y0;(arv)h9V^Fw4=-S5`=1cie#$z||Fi#Ok30Gw!XhSu zO9UPc5hqR-Aw)t)TtdMp>{gryv5@>U=R(!x3OD7;;znEo+A1qag(%)&&qbpm9omHWaRQdkya6^cGBOSweeA8 z;Yt@qy80BcH;d(m;lu5yuF2^!6S2&uMUh3=%jM^s5C=LzKU@R6wI;kD42LC%ljc29q9BBEgtux-BtFBvN(p$1?W#hJ+^4sbTsbfP=j~JlXmV=@$V~eU0GPj-U z2<46>zOOE{6inuIX|-ZA0@cC9D6`MMnBr(p%1nm&EDrrd>($^8E*>a1HO$GFi*9YX zx^Icv!Ze7mzf?;%-69LY&M-Oc@&r*qjv{fx8Y{=#|9fiPq&^K4A_moYe9LYSVQpf{ zur*nK=i*2V9bnLRA+@y-{+hWwYvt`_-}A^ z&Rc5UcWqmU+27{^ceSPzSeo&wH~RL@%?gPjq-r zHn>)2Ish{*&b2wtm)ZY9o5a518jfwubaiy58S+s|Vt|@Qtcp6?}1w`NJMInLAQKbIem zj?8L$K{_-qJ)EshM^)5&3B9rYjE94;#q7bxPm*w4a%Nvjmf_DkN%mkvnLLKJ%Sym> z-G_(X?ES($4hkQ6VT;-mTtjydi3eptZZ^AxXn(S`#x~elM;&ZL--)Ls86sr2&lV%| z04{ZZfpM8tM2>60f$5V9=H7>k-U<2;q8N`mJOYPe{tdi=tzHE|sd$J)Ij0YF!-?9z zP<*NcCqx7RwdT~t<#VEdVeZ@T;foH;L)Tax&j}?|5|KL?PXWyU#zGNxK0qVQL(M5) zgF587Dq-MTdk8A5uxfB15GN5Z?-zfK){vYIr0i&8L2;DxrU2{u^!DmD+jT^yP@c~^ zELDX8wHf(pGOUOjLjXO09}u;D32zal8fgS<(WUVL%_wjUSQpRvJUtPhPbKF&2u2u4 zy1-9r6UK0*350bUlnMQD45#_9c45x7I1wBn!k{mbQ40iS%V;7w+p$VYstI>0?WTt7 zJ7c*^B12G9bI%!8FESz!|LO}UV3Ri!_6bbrozJN0QaJnzD19iz0I&&Nwnc^c3 zGsx2z0f;==#DfKntseH89GU^E2S=MjzDAOcvz0ay)dcP>2U~A}mKc3W)%(3SRI2cR zU4#~C5{afDqPC_p!g&3d`=j?%!&};XAalL@?7=b5=puXFz<4o4I%iJ)P(E4yVf*(* zcNHd&P|mKr%S?lFY>x8i>{^aZiKJ9QDj>#8_Aidz zWmF8k?w*l`aJs9NMfee_v3>?m<`RZQ3Uwo%$HV_) z?K1bLGDgz0L^w1g^KIBGC(F46?AUy~&GQQRKUI0NDp1}y0hw31+5(x0VXot1a3)JV zTmF{T3H5a6z>It_v@F7iD_HYG(0M@$-bX{jfkpB9b)ei-)^*gg{JQV-wderwc47J+ zpRF^$FXB*-8h*9X8n!?JFF-n-@di|5(&+r#i{RNeQrC1#eo#W`{%Wk z9Iy@pf?ce7nC0?aq@2O6aeB#8tHu)@r5Mkt`Hw`koChAmELMoYK2 zWKL*vnp|cU{_n(zhCN1mu(VOe#uGy##_g|b|3K~{X1vf_F@gb(RM-84@eiSWR2QR^ zF#dB!gk8lBr8o|tE4qgD0|x_9GSlu|ZdRkmsN?ODP4MV;2emMrxErPEdECEmEct#Yk9A(w;>8IA_O4OFfXIbwV+G#cBwD-m=*9MiD3 zNQYWYF^7Gd`Qo-)T|PIyy(|||hu$FT{VX5Pnu;AJE5>&Tn;6J0eGwLOAKx4ZP>^1U z(L8I-5jyj9@N3X1{*9A~LK$j}f&@U^>Y};brMIL0Yq{dy{xDh}MaPx?-TDZTtkxi` z+QF9aghI!)LKIrfI#(dSX&J9nJ|I%9oouHj3bP!=?|r*-jj`??|7)nU^`}{A+xeU` ze7B%8I+Eakmr2^m)oymUd0}WJo{PK`k=WQ9hNor4Gf-yBRJ05IVn*U;@5qaOE3s-w z=qJ;L?fsrE;m|YmK5vi+0>l{{Y|B@Qd-1(%9d27xYn#-oK2+ys$sE(-C7p|y?0FeG zL@0${AZ&a|zhBX%-brB+_D(<1XlvKyY>YCjF=5;&elGGSDo(MWVItTFrkR(W9kk{C zebT@0gERyR^Tz$@0{;@$?L1)`9tOUwQT!d!6!5Rjg{!pG7vDc!DW3?ozn!E$+XMDn_Dr1 ztbE^GZPbHrI3J?R%sPrkvXit+9W1`~;rHO`@_Ap$>(-=Im%E61-CTbCIJfG_Zb=Ht z!SwCwNhk4EsD06_Tpg&fXyrpOp092dxVt@dR1t`(^Y}{&7qOX}eL14Kcxdr@b&)3* zRdlcn>$uqmo5py_!};Dw&P}F~JMu?DNxb+D=m?L_hOaLEBFRzTNzZRR`8mEt6m2#W zhWS!~teAHD3pEhcQa&Nlc6JAa^cQZJU?J;a1dhY*1pAa+YfRpV?L`D;gW0C_J>NP? z3g^3}Zv&?;P^JnsFEn+0%6+A6XA<1g)T~I#)3VgodHe7oKR<+Y`-Az<8xDHRfjbY1 zpe5$j_8BJTbRL54spd%}(WM$l<&mtkHQmpd>60#( zIu$-&_KE4*m%s><^OTT`6mP*Wdqq{eMo4G0aJ-dM!~ABo&@xUe>B&gKt0BoN=KmFs z$zJ~%JcBy+RYC67u5-uNVHf6p-{j&WLPMdHl#o+@EjLPEmhxNKbpu)}OG?>gigjn< zzl(s?j43sm!WIGALmQb)t&q7Fi;QAZU}nvM17Ai9->Ba^Rq-LM|FuM7nWP#aV2S8)wYKq`GJZib1JbmRNKcU~`@ z4avVN6iLw_l(FS|-51j{GyL~cSJ^g_YFMTsP7}wX{WN*R;7UXgFhV7YA?1-8Fzs`H zWT6SR)hc!e>qDZSj2x3499wN@d5&bu&AXlC*L2X2yerIuoD922sMFRg{F!TqJlSdN zt3Wy^uR}&7mH0XN8#PX~*#tp*FcrpWm|_)Z3WqccP?L8FOUi>O2+7E(R-YPw+>C9? zsDFT|u!{nNtm3O-go{L35sgr#^@Qvdq)>g&U4_I>SW3zZ-LXN)Bj#6{G>2KYl~id1 zRyB(f(Imzz5Bc(}@h;?YiXN$lmFrManE%+XtFl+*#!zqG+9+#!^JV>x&7G|mJGyw$ zF;`{64LPl*#)OcTPqFekt%8w*l~R?$-3Z^5j(EhRXJS{R+gNnhzwW0ZDGD#kG$}{u zw2m|^TSpkPne05L@uqxiRLHr<+;t;*RgI%6cGv@mpuFwKij!J1hRIX&@Oi_e zn(p|V2#H|)g?eArDb;NbD`^RM*Cym-3H>HP`I7fhVR&Q5=P~6xkJ>l4{ELCXSMf7% zhXE3oXGe$r_@F?djw9rej<*|?m8`NXN*&6Fg6td!L4dY3;Fo%?PVbhg?Uw`B zw@2B!chA}`mwxe|m`@y^S2HdPbJ>*@E8{KH%V7vSVuOC197dYJ2<%1r@MUP0BVU?0 z%~BRKF9&(&V~Qf^{yurUY`(y|!K`7Ji36Z1Gw+PTE{rm~Uc>`u|0wZ-?j)BW2U*@Mrfel0x%z4_nAoLXVk^ z1o^dMsa)J@eN0L|8v_BATa0|c)LC~Stk{#x=)Gie765xj>iR?<^Xiprak@5E_F&tA z!>^s#pp@SgSTT{+XP?EDyh<78YVqX3KPyhf`Yx405tnrN>F zK!biw?rm(07L^dsW9eQUnU}AyK@CbZ?}MI>-2-b&OkeoZF7H;uu80VJ{PPiug z{FFcMF6+0WNw~0kzHjUNRNSM33XELGijP;-)d`B&47dzkgm_#*w3BJm2k~+73wChP z*=`6FK^OU5dK_NlV(v7feL0}pQ0TNPe`zwR?7IfDkkHJ3)sy}@9C!$4K}IYo;E$h^ zzk`u7ps42|E;VDGr%X+n1w`sg-Q)E#-(Z3PAi+!$`*S&+qup7m?aen8XTLv5F9SBo zSmyjEK@xcbDR$9b5qK42^FQ7I?V5sK^(tl<9@uSiH$NW(c+E@O9+-qw`ho}uH9g5+ z-mhS+B(lG==#ApfnK!bqlM(@&`Qy>gH)?s^d=z34YY7BzdM+T*EV16d z4xa)Faf1qz#U-UlNK23s6Zys12foI{`H9ytn|59^Z;olY)_F>p*}$BpK)GQL#LOOE z!!x2m$Byaq);Xnoq7hbytW<-&Uj3Va@P~DJ+KtZmeRWSpr|(&cT2$r0FV zu&p7GyavMz4f|1y0aeRr!0j5{XgVm~D8Qhpw4 zCyeXHL(=u%b$=5rsQhP0@`o|bw&OD+{1J3&fotkzhhSqTl?v$B);b~-Px!`B)7~48 zk-Jfgj0?U=btm@J+iyWjY`<9D$3`G&B^w%A~fHE2^hvyI~$=vdd$Gi`p!wT6_|r1nCITG|JUg&y3h zrKx-adiYm~H>?G0>!!R2Lq9D7Xi2AiUZ{SVcCSk@ znC=~@x*NOO8E~PX*Z|}hKZxY!5)daxL4T`t{Ux?dC&Xs$Pl^@t6YnPC3C|FRSMgZG zr&!pz0Yk#QxNnb-PV3)_ak+wr++ZC1ZjqG-ID9>T>ShZ^X44qiN;*fJV!QPpE{eGq_z-#5#lUUgse1H`OY zz;i~3QZ_?T@TDIxmXRblBeF|GQ||SlE~mDM@oyYa z%UUEpLyM*TWQKS)Q(Tw?&@KLG8fxuEFq0y1qk0&8x(C!fj%YjWk@h;n9JTv6>bEgf zFKfXgf!@YKJuS2~rmrSg66-h(2_&hG8j9&<6tkW7dHhMqARgf?e=z$3Gtl!iW=Di0 zh+F)PelZnuOTl+uw6yVG03w(LbC(-|Rrnp?FfEq?Ug3CdFFMVQ15BW*|K_%W!{hyO zfge7@bp9(NfAeBdU3x+Qc@V>qLpFJRk&%aLkOMYNedeFd&?veCbjU{|fE?P0%k&6X zoJa>sbnW1_va6G<$?AKa8qN%|njS8hKFYQ_ohXvNlv#trZSvm}uj@?%QrP-z50FJ`3>zA=lwWm&K8csL8dIg`9Z0`&Vxp(`tANn{N$R z^Rv!#&Ys9R_gwpdISJtsfA!a;<0AuDQ88WrCjvWN4519k*aB1|Yas1^rUgicbLDOp zOFXRScv_Q09~2RGCBNPY=ZWy!(BB3(RG}J%B2)~tcY<(`#zy%}jA=Z~Ks*$$NDA7v zO^~Mb`*)5)$}vF$SL}Ywuu9)vRnM`fB$|jl>tn;LoYtbqNiy8waU6UQ4)q^iv*QM*p z2H?n&WzSeLRSPZa z^>62y=wT=4*QYyumfUmn`>M%T#dpWfi->7sg zUPUEJ@RBm-Rc>h<#^4t+qSYp2R>HZhwTV*nMt_s&2?FV(VlR4j@HWGGY$I=v2L$Me zO(NH^`@htU* z6w6fCk;`n|M?U|HqWSb>?`_9^@Lqe+*|_R|O3Hasrdj|h1|{nF1AEqio(%#73cKkm z&iKj^vf2LhU+X7!_Wz#j!1RM``0x4H|6eOk{{hJU^6d>ISVWj4I|*SBQ_6SM&hG=q zAnho0OAFfLv^vl;0#D5;{qC7{5F_jBuuxAxhc{qNF<0+AjK}4_hvib0t@yKRH1R@d zCG3)21%1fc#&h9h71&;RX{ZyTEGtd#6isf5QPPTW8Fi+ugYHF|vP;=4EbSsHO`br4 z#2$f1HbWj!B>V%IQ!VMv!W}hvOua}iksof7?o=;e(57PJtd&c{&I#TibrN2msJakc zIxpWBm3ddoIW%%zIxJ?Fvj$+dNDWgdkg1nRyOV_NF8zU0cG0@%z@1YGwPHS%ENEF#&rQ%r{Bu7g#wo-P@Sni4GH$5^^cHvTCkI<>DIUM_8HavBuf$ZLKLTSnMG>C9odoe#r6Og!$V7zE%>n|zx&-Ig;`rJ zr~++>VJK=>c$<(GR-9Lak~e`Ew@a+@)@4DswK@XbutJTAt&D9$GP)hDY*}*Jn+?%) zWWv=#GO~v@^jMP2nht4S;@R;U*YgfnHrcU?QLVFLa4$wYNBNUW`205vxYPposakx) z!i}0k0jscds*7#wi)xaIDPP^o-zy;(^)~yvuneyy-n@iKD zv%NmIAvnEGPp?BU-`*eHIC)=BW+7fDs~G<#&sKucR#)x$2}UulKcXH(^z}b^M|O<~ z@q6EJQw_w^p>$LDo%MasM(S488PwFT1o{wvrr_;(EivY1>GA7+B=ggneBw%+FVXx< zg!m%F$lX{Vxn559AAip7o}(teFCTuSr}{;yamu`A_s z%Yj2Zeg2$k1i*yYiIDt^TEXMbRLww((>RUHFlbqIi9U**Gd%+faXX>Q$Ez$B1mGOr zyWA39`IQ{Wg~@yzH|q9t(jD%pBC&lC=ilCrpLfR{0{Xr947K!nI@xd>b-8isRe5Ua zRg?G2X0gX5?#!%09FS8MmCNd&zhe1;9=;Oy|EDNd#`etSs{hTjZ8(CUP&amhOPVIt2=L~+fqV{i~QvcVHA01+V=qQ^60Lsg43gP8|mMM#1Y zApmu`C;TCdPYw)+7i6AvBlZs4nZ=}N#}my3Q81Vy%fNw=iNDLEXJ@EWBM?QqC+4X~ zX{QI`z_1r;0`?lXP!`HA_TGYUz}FYHjn#U(eZ5K8e)DyDl`fc1>ny*W*#Wvxu6M`J zSEq}o*IT@;`S^D&T%QupJ1K)qVNWAO2o3cf5G$OEI)guK*5k$`(~gNbzGCg7uhLmp z@ppBprmkUlTAMctf1+tE@vJeIj5wX@hWdu~w==%`O0f(qEJRMIOhbrX>v(FwGxu1mg34w_$p-=OgTBQs}n!*^lO886|a^LpX2E-g)cs6|KShp9t(4i&LOWI>ff*hk9taCQUp3wsqKHLw= z{#o26WduMW6gSmDEAixf1k|?{Hb`Zj4^C`1jY&VJAWGkA-X492`wvk>9;`X51!%T7 z#*>*i1=;FXb1|bJlusVU0zWhyY+6E?`-G!jw#C=n3P(P_jJ8ALrzZPih0Minw(ZJE z=p-y-=-#aH0?bIX#R`B2WN@(0T5S`KX)Jj-l?sVrU8|(nV=`H>5#0&NT^>p;9kH?s z$jWhAqAgnv3VU~-4;+>Ey$*pbfQ2`% zu7r_tDgYG}_UDgDC1Gfpu)Ood8Gao2D251vHygtQv}*e<{8j6A<4t&#`)D8yw4u{9 zSRCXOT)DdtO0irN2`g)o6wJu>quph`PU0k{lv{X#W;F;qBh)W#-w4ExZwyZhDs|Wj z1NcF0Y^rM~ZmFFz%Hxex400ZA3U=bcH%~U^<4G?A&rk_?DNZ!aQVWVgu0y`l0#G(g zx&2ifz*TJw3wnpnn+Ddj9V)Z zA`h8hM9SiDkR!KERc;veSzqR}9LCkhjyFsU{*sJnvc4cT7@{^n-By#D6{a@0Il-gx zfwl@G1tXLqxox81{4HR{&P?gmMrX({AJ8`+Vpm9B&_l;PnIjFq@CG}KIo)7SV)wR~ zvU$)>bq>T%2Hwomw96bS%yB22R1$3u{&SYw*Spaj`jRa8s zKO0qTC;v*>$u=M<tT43Resf$3FS9}QB)apL(pbIQTEPW zG=F;U|9zWuX$-n*z3M!|cxYn}4gP{m*~GA_MEVrqlL?2@t^54@!)AAq`q&tx5!uJRKM zT#t#1wvYp@(2QaJQ>0NREef4&+-vi%1y>@F0Yh-cL<v|P7~uqp1P1k<~H2AM~xQdHeUvM1IHH-S(=Uc!em-l?Tr zpMX`0O^Jano5RG?=n>^f5G>d%eRca)c6E}_fiOmfcaWoGfLhPMwp*1mQ@>64)t{$M zR3K1HnUe`K>8*u7iXVJ*{J3;xFAA?Jk=fkqayb{J@&IRw&W@zN-ka;w+u@)NTfEN~ z$JhPsX43z)7U5w3?{N!^T>n2QbDMTJ639OU?$gmj!_2>~XOqNG^-_q0ee?zz(DiAw zun}f9Cu@Axm%;M zgBDR^!s)B9g=(p>`su}I#P`rQe17d9T5^zMptRxKqzGTe?o$R^sk)+8@l}9H;9kZe z<56us!*ES>&aNrNj9{ik`~h)iS(-CyD_CM1?N)~o^2 zK`QsrN53afG*ptfiFfWq86!=fAA-sq> zNIQ-UwZP0{7@_F+7CIm*5wxU~dRW=wgyHCrjXy7HOHUP2h|cfy#u*w`B)7O5mZ7(} zg*H=-SXf!ps4O&VS#)NT@jnu^S}s;YEQJG&O^h3wWAfTvzv|0rsN4~*bvHR`T5Dyf zTnJrkgr}>J=D#a8nVHtR>WSq(k;A6X23{n}IOc)YG30^`UcWDn3pI39lGb=7vWz14 zEvZ-!8rq^v5vUoQbBNpuC8vI6#$o>6Xn9rd6{tQcd^=wj^0(OT^jhq5#_a z$54+ye7R#YKgXwU8jkm?R}QX(tc?;MSXIXJ4z((>+Ee(>nxYR~r8* zUHJ+88ce_IPwzUr+SebQPZtC07A^bRbjSmqk;7@=L{x>R91V5jARyl&vrj&mQ7k;4IA{T*L^$K6NCDaO{h+Z`Z6t_mhh2z5wza8+mrc3=u}_3 zX7EROZHLQd@D6$Hr5|~e=+A)JTY)b_BIxt%x?q3qj9r_wdM1;TIl$y*$aj8jJa;T- zjc?keUe6YTW-9=7u#@b&)gPu;qj$4Py2`VQ+XP8J6C_txq(*xcTyv}Dnq$t%OIzQ# z18?85*Sl47whp~?b-k&sC$tg$x>aL9Z&coXwpO2xf5k#1_b<`t3S)43SU4=#p%Q+W zA}}h-63IJvFn1{NJM(i z+Ox(2;`rO%yJgF1e1aR zatkLWxV$i(^&GYfS1LTD!vI4DA@cL238KiC2U<_APMlkQj zCt0gPC$|w;Ss7v!^1`Hc0aiR|t2rKIuaF9|A><<`$ER!9Y2->sd}<}^?VKTAlFU9@ zsiZbL8)Pe_&%yvkBwup*$a8Q-EoodxgP%IkfS%3Mi1Z5x$lzy7O#T4cI**G{1$&r# ziMB;36}Qo~zx%gvAJU237y5vG>Sosat3s5mv9AwD?{=5w_C>goUga|65VJTbGIz18 z+MRjJo%7w+be2-R=_Z1|CJPMPF-S;SnF*IrWAkNd>XS{Ise&R!JIVs|rRcfnmFSu1 zUzCPA-1muwx-NUj?lhyt4rQCOw=Cx2&;S{H;b|(kJCF)T)P1yc-sVC{6?DAkt17m4 zO4mfvccl*k?n$hIMF54;G1*MKgg06a;q+COi)DC0vCt@*H28#gjqWg{s0_J|$Up40 zRZp|aYoj=ci`!LazovAbMgA9nbITenhFcSdd*2?*Qk^@KpjI1$A7%=iB13sWQIn@9&-XtN zA5{lha6(QSuT#D}xNenf_->js z=6AElK~K;u-I=;QE2!#($#!(CclH``==y)0m`0%7M~zJTv>ycTvg=Lj_}Z(PLUISMVJwA$Tl_;6KKXcBl1sr#1K;dkNKN(?Kx{4ABSJ@}P_1;&2=b zC%_p{(sF6JmhMXpTaL7hkvf~IVF781vk385xB`cZrAneP{MsEP*F9|XvD&p*^a9H# zl^5oXkN&$S$4;WyeW}Idm<80*pH^*JSc7%L*Sxnbq<5&|smHFByc--A+1||fGS(rG zI@kecfw)38gxY_SP0(tI>H<~%OX5kxGi;!%4W9=!dubKY4m*c@u3#T5nuzDZUwB}K zDcxV$&s^oMjbNF+M{Db{HTp>%4LCEmQ?R5aJ1 zcGQX0I319S+&ncL6e8sgi_3%|6vZ@0zPtL^iW_NBNPbf+8Q-!+Z-i+(3C7h}2#?cV ztj2-=E;6Y{Gph|z@-)mb?B%{YCi5DnQSe_+g78JpK3z_(pm3l`QU`M zwx1>a`jwQ~`;EFBJkdvCOVFR061yp3lvM8r2J8zxU;~*@O$8Oj7JyhnPa_P?66~LR zb7#9n-v)h1djRNbR}^7dF)-tP09u<^_!Z~*^JWxjT{=aXn6*$rlmaAydjvX*&1u0? z5J?&p&v`R?4UA5GU~=&y{hoD-mT^VVv}D8 zN&M6{3BU4-5f~OI)6H21q~Mr9leCmj6t~uVnSI4;QC7@=`O!xWRmH#1dD^J~gi}w^ z4B%dcw`~>HR#BJz zHu1j>4C!xy6%6V8E*dr00@?kyDn68>an$4WjKeC-R^sF(2fLEu8;vT_%Dn#1>|`e;Ue9vSCJFyeyrP?)p|W_QR9~% zAC`Wh1q5-IMT#c^`0wBFF2P!8311@=|-gR~`q6giZlV;&N_PF#-% zvjy!@hCkOsAB;UO8+Qob5T6OKm7 z^_HBYWuzSW7-^!s_ttIGi14^**{_TT`!_u zMMbPwF@K%&XT{n3J9Dlvnj0udD_-!zrW-TXONf`Pfbhzq-ZkdPZxSo@{DEb2IsUju zUUX%cPlMBM+`U{@q_fx>vd*>0Y10ql`pnHz2D=85Um@?YL7=!OjF8&9i1R}ctz!yV z0m2K&;3REmhh{33$(|DoJiLQ`=Bb}+eC8=P39nOAY1~Yh$8?u*ufqw6dG)sui`N3@ z750s(AXhuYybPGrc5t~hQ8P+((`!cIT%gZtqI7#jyi%B68O+V1lme>cw!W@=bPI%^ zYcvrn` zp5XY|_I%Di!lO+*7+8G6QOx!JFnkSq`lmeYt}*Bwq>*fdGr$H|4|nw?HHIGwShHH* z4kcYQy11NWu6>SVdp*1v#+ZJee0$Ne8OaC2vk|ie1|2}WBLtF8?sU8EerEIPOu7BSZ;bGPYC?u#Uy$(2DIT5t>8SJ%Ei9`&vbH~|rTz`stgDBXu=AVJM+S9dU# zLr+LXYDT+^wyk>y?|TbJhMr8P=1+3HzR`pj+tm)xxQA2bJlS1Xe2zdIsqcx7;<#_~ zkbRk*&gbG&K;ht-g3yW|Pz56(rY0shK)Z3U0`nH7!_EXCW`ch5C<1sO;7%<-p?wDe zVf*B>LBoJZYecTt;Nf+o5*3|00b)P{0*Vr;yA3i3)JS7Y&clR~yLu-A$p4yGXD=lB%vnU6_A5~X>pjp!plh3z8$r2EJLxH(B$X*{cV z|5-RhB5`G-`5@oP=KC2eT(HCU`8urp{zzK5!TWmqS3*}*bLh*)kH5VZ^>i_zpiq1* zEUgqg>(8_WZ&g|j^o<8&}F12x~!wb2pQSF3& z1O%L15CiyY3<<1u4jn~aqUj`$PchXsdbmGwfh$X3+M_ym?b2i>JA-CtYFjxiJ*)bC z7e@o$gyulKs!w=89|b!BL+KF%RO}c`_SlgG6MfA^GC@J<*XKtkB6AoJL8`VlNk`b+ zPmeR~tp3vd$PYlC3gAjmPrV#s-^nHQYjmxEAfeC{N4!|wIe5Q|LM%?klcMJR8~Rba z=WEf{+r2|CpYsxSe`Ia~YF;qkn>BMx&rUUDU_0JsOtOZNC`sjDDYHN{-ZR-~{R_J@AZLJaD{ zf9bFCrj*a~G(Kntw40{D^e9$`J68i#&30;ryY#9`e?-F4G55_ET|NY8gvcT*)<1rr zFnhi+URQjBtgQ9~WYO{_&2TQTcU7GhRn_QFk&LkB$p-St8)RUh{wodW;W;~D~o zOEF5S>SD24>c;q!*!3=D0(8#Trca8MjiNtGCe@OGYY1_WKv4$<+33b^Nn^&QI|x$% zeSIi15uerbVUw!PqVfQA7?J7_GkiFodv?Jb9**)Owl}zxUhk}7vC~0MY@74rIB2LxO~T__#I&mhNCA4qZeaB5 zdreA?sp8v9tT7Rh6P3=gKJTC~_HHZlzIv+L7&>LHd>_^NF$1i1=v0B3GCf>OOym+f zfdG~m5+QIg3XwCCjIb*;alwa`4(5jNGIqCM&Ug+h@G?)05|M6q1B9Cn2?@L$t zSsJt>dw5uKc4Vy@u9gumcHc_Vn%J)>?;|};5;hMRtP~R3o#%k1>recvIk$4V*B0@; z8!7Wvx3DaI8o=Sllbyu{{)u_yCg51Y%@ObfxHo-CbD_F&Xbtg#b%B9J%|eKp5f3FJ z5kgiZgskX%uy4AaX-(w!&r)&H!*thT?9EHcx|lvBi7U6^r#D&O4)imZ#- zn=GfvpVST={41`vV#j$_lA=2V96*xTNzmV%22$+GFv^fFf{hMpt*nTv@be{6xB8V~(D4e(YVCDYYp{4_b?q zF1gt~s*Cp{fjpBVo!6gqJ|O!UwHtZ_00OJLHFm5b(;_KfiOXz9>ZLhF%B~A4u+ui?IF1E@Dl|wu1>S_$ zgN$S6OJ@V%)8(2JQzrwTB$EL)vf#Iaa*ud1aVbs(F2z)OgL=@5hNSDrgd!*deB!~% z9XW5UWcaB!FEX2SDIVHRAxEbgY$;qepuUfX_IIsFZ)HDvw5&q3@7{ge@6ERV(lZ9! zcsyWqVPe}24`5{!+FZWP)}qm`*dd~SJtURx@Oa4YkE#VceaU>? zNPxU+IvrU{vJ((ryny=#0Q$YJVF~-U=ZKj=2B45I+6mGZ*T|4%_H#uKs$cip`su6f zOn%n>hF_=frYXYtJo#6$t_QC{?ud~g?r@$6o!vA&@8kQZnW3RwTk?Al1pz`K9KzP2 zT}cG&vs)pf#<`)So`+m1tX8gS>NW@g9lRw}9V!@~1k3XApA>e%A4r(5>Z|{ng*e&% zd-4D?)Bh&<`oCr&oc|jrSC5a3GI3-dYA2d8U?;UOTy9k?=}*QWTiV=I<4nGpjYQtt zGjki;kW^8o)~!`mrmwYsb(J`{cbje&+E{|qRu6kb@T!2-vv=Bzh&85xT@Tr=Eh!@< zi``~j$cK0r4(&;+D7-HX%?~7Q;~>l!-|}cGxqcr*TuD4M%h7Vwv&tdKrTQlZy4*Vx zJJmTaI?lS=HNMQiET~2M*E!tu5B6hn9Z4c*yfsajE~!sEpGYq zaXYMBno7vjvWr&Ec*Xrh1x=A~!JX&ni*ma#*GRX%s)jQ~kW<>~i>#r^YC`?fhf~k2nz1w+( zG~$RLHCr9uQ65_6El9>}#?~}65R$^|7?*jn`gpfIs|Ovg06R+d^92Kzb?s)sD1y1IYzejA1*9U61A#PCNU!_&CrhU_R$H97#~a-*d*dJ7R*8i zszOFFv-yth2V$S@7VFamt5erd^~Yj__{hek$G8?Q{%0fVjpy@?|z(nGj)3Etl{>tl1&ZL zsF~(!2-U~g!d|HL%fi(05uS}|#%Is1Iq{K{(YR7_Jwdc&NvKlg8G~|=GO)c60(dk`m$*{69 zA}nsH_s(yrH@8@Ewe)X}A+4PmxNO%At9R;#L!qj!U_mvmmd=x}(LAlIiU$q+9ICeVIKvINHdb_I@5l#b7Eo-yec`-(qnQ`||YY4(ym{z|HTh_M*`1 zQKvl_>85J?oh*gWyx6gw4XK$>4WKtQ`zj0R9kM2NQ;S+#k0-YF6PP#8j9K+ZY^i{_ zVLFajibAP_b&q5Fs4W04>|meWHgq+pNr~D9xFP=F5c3V|CiN$|S%p}JjNPKpbw?W< z!&mlwjhk-CGK5BW$#1ST&$?VSjZ9A*qEL()heV<>|nn&Qo0uL3uImfH`)t$4fuM5xls!Tu(e~Jlz1=&jXR&Uv5AM$y@^*FFvs5i-lAko9wv4%6mVSBmVAZfqETveX zQ}(2FyRA&!mjfstt;T*#*>`ZN9Z`b+e2W!dv`@1W5S`{s)_kwRo1HundoI2CECLqr z*>d6P-ZY=y8M?NBy*%Br7JrP~l%pt%<^K47OwHNPHPZFT7JFVzteN~!i+cB@>%csS zM>R(2g1hrot;`A~VAlClk7Rz|Uw~!ZbxU@a#{b~W;^Ale=WL{j31WuV0h-zI0CoVO ztD9a@#_eFHxR_&vuC;D59i@1W*AyqnPqJm^d8h92RSo|!)SgX~fx)=TAJQZ487&u` zoF`h3e=b#M5FXAm=4gbp4@taj2LVz%Jq1Sa!$9avgbHT@N<)l!{ehVuyCUFpu#vEN zU59ARZ{ESz9D$^?*h8E6<>m-es^=#{b4O}31c?BM2l}S~>5yt53nT}^_aXb)TRKN` zq-cQ*Uw6h;{ahayBOK*d zONI4e5k*lD4w=PU{sC6vs(P069s#%)SnGol8b4@?9o@3(p;o_4fUyUuv)U&zKG;pf z@3RL|&9MBJZS^m^oQ(?YI5MQ&(BQdMP=6I0o`nzb{HX)`y5s9+Xv6zZrlaF=D(Hij88!e?UX#p5@sSk9loOss0MT2tfSr zce9w{NI}C%ybUQlCm^fv_DqA-4QBM2d)0MwGa8TXFAM17ogC|VE^1UK7B1dQx28g$ z9e)!$nvd7qjB1pD!}1i`Rh_xqN9B}4fw@#bIF~ib@r3vQUSU(z?60K zN)%vfZOw>-cNJ=&zzwDqTE)n1Mq|Y1PhnVQQ6>j6%X}~%M1R9ehBq|$PIz2gRZh58 zffqh z4gShS9LZR8N{( zNc9Ah@H{WKJgq34Hft0raMHIA_JQhowL7`_f+_YL_1v z#8*~uUL+zI;7`@^BcG9kxO(VpLcpPX`=EV~S!u%L(IpmQkYlj#lj5s8SRojIO> z2DJxBPqB)Ac!vq-HVqbHsand%`(=yZq2!##BkRk~KRr-KYC5G2tm$IrfW6NSouVM7dtG&;cHZoN5yrGS@pkR(Ta7iH0&HUBPd$nTJcV=v~&2D8E5_5 z={!Hyae1ie@KD?8sJ4T3pj4ZS^fz<2=Tdpo0em_(UU3+!5`UL1yX9meSnEfL^~kwo=56aB;+5{ zzLDJ>cG-zPYL$cID6(es>h6AjB_SawYH?Ed1|XI-sh3?cVcDl;W>T#yL8RBqnqi2bSZ_apDd%y2@0>I`05JMNC*%_Yk^m) z@9!koSwn3vhlkYWHxJz_D2q8|^1kMxfMPz76T(#W7zpGIGAwT4LFaHR;=C%;ls}eC z@C`anirSaq(htKUe}-u9S&gjTj{(Qm*SXvd&RSuvueT3+KWbo7N%+;Xw&0B3rEtg( zOBvY{jn=tF%OX#na$gdZyQ3E5(sjR2U*;cmAr&Y-s}>`7l@t0OP40KrMV{zdfNN_152)Q9ptk=gjh>S0-!t~^n$7gQ;F<^JxncgltyNLKl^q`uPAtIB z#5HxLbkJ06k@tPVAqrZpaj8z1ZrQEE#Bt7eUdjG4JuP#z93&rngZTqNS5MW@N0KO1 zk%}@f=9+8_-lp;qR}{NhDtCT#p%s7(QW#Qp_peg`m5&#D-QRyhzDwWIYu(a|Uz_wP zrP4n1z?FWjRoa{k%4l*&0(1oa_r5?u9`98??~k>5u|z1I=YuzqzXyJC+iL8E;z7ni zDer15+0$ecr(7BO2R@!4uIJH(scuTKdY=-ViAA9p{4zmjl!jQ^8s!|(Q2Y@LNkMA- zsd%8Rnl(%D^{!(PMQC}n|+=5S0`IYw{yqI)^&>js|x60?{X*#3KZ0W;(O z>Ft~U6b$_bw|fspFObE5rhX&O;bPt5J|8l=+182tMOKkSC-M20+by-o8L(mcE4(Oh zwKj9LOT~c;W1i`%8WXk4B19j1)&2`XW~>kxV_X}YhSG>MhY}%0a=^Q?(r_1WelX;2 zic83OU_?~R1D2fkR~JkbDplkOk8bS(1`=4L^nes5_xzC(lS;C5NUwk7U5qh+Dmfcc zh>JoUX6%e2z0bpk7<^DRQPDcBt%oo){k+?p!!w(fe2{X5ys z{vy5b^)zHw^0ALU0r*mq`P*~qe(HZRZTu{|pz~`=jY`+4s^Pk!Q6_M)lFGMivl`BN z(4Xp7%buJ&WA@N{Uoj{WDMGVP__%cwMJ3iT18yhacR-(9uAo5yq7RRxQ!9 z!wlPrG4kbx#cNN7P0fvNWr|zl(xv0F+_kN)Tjkz{0qbPxpeM`;r&gBqD6SSL22iek z)cU;tR?xz$Xms!G`7pyHbaB1KqB(}9Ez6{pHOsWp_FPeH&s;?#w`PoPC%hTTx=&xQ z6l*>uu@!4pWa%R~s_YenrHDBQyAT#4b^=wtqvj<0AWk6=tF_)`4lKgM(Edm^M{#&a z=hUT5jgaTI@FVsD8g~`C8}+5g)?6uuE%eMed~Pk;LTb=>kJtB^5`^=sLlB|!L`#Fu zwjONnGtL0K4v)`O-k9V7VT*8Hd&qMOHs2`KE=TztN7|VwD}Y3`;TZ&euFd*&uZvoH0jv(@YK>i79QkB;8 zhSqPN)-PLaa*F!+7}e<^vfX27v&+y{myx|TO^Mi%cs!Qkmnpk*#QA2P(>VLaLoPH& z2XJa6Esi_Zfrrr-jv?;Ef13(8S4L7FA$t*ct$}xPh8)e}Dk)))b%WcjyyWW-v*YE# z0Hf-os6wR%h#89>Ur^E^y zI1b7x(J;Xn0aims-9gYD_+tdH;z0J2h}cE0$NTd{?Th;4bJc!wEnEqmNMkC9iA5}r zkPuY5R0RcHafQ3|jm&M-0tUeaWCB(aRHm3mBKGA~vfG0<);?799Et_jN1MJ!xAu4` z?4@=kQ-y!)bJ|RBlwCHrUKfBwO28ASIeZLFhTe3y%zgb3>u8ZSrLtcP`YzfqqmOo1 zKtUa<#GuBhyxn#xIt$Qn&`&1ItnU(u3|zw+pHLsXll(ST)RJl;AT>v>HYwiHDIm+{nJBoN9lSWzYMDqWfCP(k* z`m;AZg6mg+HpCOY$y1x9#AZZ$+SUDJ_b+$2vJPu5ZzFC(Scq7QSc=&GqvLz&EM9bc z=8pE{$xwJPi=8QN%Q8T|_z8iJBxIg0Ry}Y_NbCDE{Exhq*00O`4a6MN0R<9OuFxIa zGBQx2n~cJNL#M9F%0OF*OIpXJI7MOitBM#4x?7ams3;bOkR)wb0)o`RUJe`R<{mb} zk&~vKwjAwFQv$6XNDYr@`v$qGyl;%NTW)kZ1{$ret~!Q;Qy+0QK4Uw8 zlon8M5OaxJ;JYFd{LP1~a*G6k!y+sWfoPJrb85qvxwHN9YRlV|d$X%$@F@H0ro(~J zq^e|4t+BMh6x$LzrF9+6g-&B-8Rjk>eyrKkLAA01t0#C;6S5obRKk)ZX6LtHxCHnZ&c4$W;sU%WvJbX~)p$Ir6TM~pI@w~;aTom%6>RCA<5 zCa}eTj+{$BBlO`C{SU)rP0a?W*b=DkZ1?M|Ggmv`8N;@WUo)Emh=Bs`=e}+MI z7i=SqN?hC?Kbxx93?g-Mz);<}v|gqTT5kE^mtlv+uUc~F^@Y@#YhZ!H>!3h7(og>SF5 zsl=X`POr9w=xxEME)IBfIK&KAbgd3UtY(QSu^XmC(g8?a%V}-8M_C;^h|6zS_S?W1 zigGU?Pnwz{&%|FSMO_I=Ars~HeCGw%@_X<5uamzaut-4|C$ZHF{4$!~IrsjgFgs6j z+GMBlR9yVQZfOdHZiL(!VA!jFPPdYlMfur&%mx+?OUZ-oL8N^`&)9;Vx(+dY5o+|x zC%l=BcXL;J65E`kA?$o;sk426>_`7396@%2ppv{jhM2&nlC=Mw9!_^81i#;om;C3j zXs4M~$;QLEyOB6O%6$oyIwok&0F?&&FQ2Z_s3;(%*!!^PE_X8 z<>k)Yyl0NNxeaRKU)2GGzG={-KRnGQ1SiO3t%CRi&(TZhpIe0~jf?raHSkwMc|>k` zaw%0>e0+1gC%B7yhnR*p(;nIa=_{#j7Q)JT8^TKEYfXPNKaH7CO~+BAC4!Sg)i3uz z2ijd~48dw5mm;^S{4~`@UyyhBD!FcPDhB~`@shkSbm+SJxH7n93D(ND<>|SZO-C0NWH0%5owIVs^prfG{0-)6 zv?}Q66veYnMlzdd=~jm5rvZNc8*~81tLeX{D@NA;o^8PVo523J*arV|6HCWVi~aXL z(QlqX3vrj6wwFNL0+Puvgk~|{ZHoFj6^cK?7LTC8Oxz^-xOjmfrP^B7IP@f~?qrBu zFT9|hj>pO26oy0grWsD(9m3QLgxZ#uq5?toK^-l^QPEdY^dZN(F;bO`Ca0q!Js|c< zZn%q3FD{6&#D?V3p0^|ELb9ZO51IoYVG=a-f!CY(LnSI4yVyh$8mI=WkB(T*#v+<< z-U#H?-IAJ}pQR%H9IE3(4?ak^Xl9(&-IBFtPqGxM>#+Z>!>BFvB_`Br)jOw5Qu=&k zNVJj?#jIe-AIZ)b)e>e9MS8&wg}D}MXh5v(^4DC%(MK-Yc+P=xj0F9n(OR+Wi*~3F zr3ojCnQ}NUk)6_lKMGYZq<0oGdQ= zg$oZ*lTdN4Rxtm!l}rp0PH6!RxX!@wq+=eG^G??UI!_1#xGq$+Jdo}!(bEH9Lzub_ z)RgeRC11Cr6i?;(B;!&kFe^4tk(L+3RL=HVW_EI+WPTtoLvFs|Ddq1vgH!=wljY+B zxL%+%0)YX8?_x+Oq*YWUHc;ptElWGN_?}i_k_Je_R{AhBOV$I&C7hXuO(m3RUqRFq zu?X3>KM+|V9Eh(j;qpz?M5@z?@asvlfmw)gQ5sov@soHLWS3xXw0~vz1EFl@o0ChkBL&u>j#*PJdfl!P76UhB-2HAA)dC8=!lfu7JNnz}5JIf0tqI zD4SzU1%VzrP*~F7E8nQ@lOPLM*-%DtwrLX_-)aSApH6{45q#A z7&=A})2t9cHu77GJ%+(^gt$r8jWI8ThR`fLk!pOZ|Aryn=UfnyIcXRA9ZGLHSo*ZN za3f=pi7MpCv`f5T|Jxg%);1WfB(;ReviAVlSIkk(+b%HmHv@dABD+7bUm9qsn6(md zOFQDmX1EPpYrPsDhS!uc&2t4ZzeacVm$r|qW6$+0Z-3Dc1~@HK71=P1Q}*2=8Yb|e zPztcuIU$U#Rg_N4xj9e?ROcFS_JaYii&QV)44?uWB%gWDkfFH$bPIqD zra?{H0TRKD)XQKJKYt5=Sk<<&_o#Czf&v23-)IpIPlGCU(dtKk)lz}VO{T~HXZh2x z6Hs5Xtww?eRqxBCI(G_espWdT)&SUROY%@Si8e&RuwKwrO{FoB855=9_e+VOc{_Wy zJgd`CW+gG6DnsR?)piYr52!t<(gg4Zd`AFbCdU(Aq`tilUNYeV zif*N5Mh&d{kg-+>c2@cyX=e#g(H0CLyNUj1Y8l+n6z@GqjmS7DA1{UVT z`On|*38mhU8T4<0`$%zsYcM<(2aw z=FmZDRnu}=V(|OFdF_vkRyDiFviPrTGaMOZ;0r%*@qY~68$ymtDmTCs$l=Kj4T;>b zS%~6M{_=aK>rs{2=s)I;#Vp=Vrw+4q9y8xAhq$@hd+=QEp)b)9mQJmO)=S9wG`R?- zI|jn!Xj?EvdOaI|B?44LHJxLY%$d_4zmB8$iyYY!`UHn)e3|91q1OvvF0I1yyM-~U zN0-kl&rf#`P>XnJ6Ek5ExP+uB|GzC0dI)VnKfPxr+*kkyQ{&z)mEaWR-LBs(T8iUj zk)rkkhELygWkb<0Q1Khb6lBzAJ=neYo<)b6qpNUZ&CFQ2c(4CPmI3jibza4&yopis z6C!6NM9qtcSQHVl$|33@=YyTGrzpS+F$r_)7H`nQOmEe1&F%)DIRk5(H4};I#9*xY z1gbW2gmKzg$mL&7ANYCS?qH~zvq%}xx4gxL_Q~B0rBLRS*O-elP)BcykiB}8FK816 z{ZB~ktq5E0R>Nv>xaHd03Q#pPsPDk*yq_H#oQ!xkpBcyM z8iqRBo|wQKUHej&CnHMwWh>3(>a^IZGfi~m>_e~lUx}740(cX@^FUp#6Y7=b2PO`@ zU=ma{h$qi)48#Sh=b3ih#_>2nCJ=jN73I~$eOlBn-jZF1xUihH1l%NcDeRw{_%C7X zp9Wqwu3to*jgxL$Vx{f4^+AJ!QVMu4Q@rD6D|zo9>s7y zLewAiq4-7}qDk12D|v0Z?JR@ySLM7^YV=+o%h&mZe&0g;Y+OX4yK(j^EB4*Y&{KDE zIuAWcay8`=$q|@(da1{0|9y&2CU*KG)#C-qp!eta*|Yr=kwJ2GEZdgfxB=zaN|;_a z*!yQ{=K-%MLQ@&tyDSfd>GLuVe1F`lh78-QJycGn*4En3VS)GiyYps->$YV|#3Rn9 zy*&fIGd_%ucw|QvM(a7dN9r)%%^fd~fG(2ju)d@23$@ zA%fI}U>GM|_^$Y55+Ln}$Sz0qiQ~q}P2#a^Io3(QX;B;lwyR{~%Y-Y;pbfIE5J2j( zi_DU`IgJpHPnz!JTBmj%xJ8SU*b506#JspXj8d_6I!zWOe z;`GuyW0h-*1O_cnq>t)og4LnOBbOVGH@shC0)E_MRz@ayA`e(DLp=l8&%OaD+-dxi z2_$NLs-a@TzR9s3A^e_zj$brY=XACv*&ELsm)wGgO6zinh&2Ghf2RT33eK7r2;ahE z0{j4#@@ESEYZUtrOA!NsouMTZ4-WyovWLA10ll1|rIND^0sS8nTQg^K0%iuL|HDR? zPMlUNqR&n3uF*Y#ZX&ZVTojAwK8blG7DE^9EIfLstD3`Mm712M6~|A{@6{@{K-P0= z=?XkF=*#KudD0WOFtcgm7l!@~-e0PScJvihiJ z2oeVcp`?bWcGDy%tR=E?>HIp$4OvAh^T;(m+POVdM#}ag!t~%8D&PlAS+PUVezm|6$5gNgI5|z9PGqg<`_-0LpWs+7+VxFy0Q?0!8(#l1e zU;0e)ODt5>NHp!52kxJFq54sq@A46{!4~}W$(k7hL}WDzBrjQ%*d()B3>w9n9j%Jn zup(p>)hL_IDq$;{VU{dwo+-I%v6hMs4Iv`_{L2e7_DokW%v;ABr~@y*g2s4}w+5j! z+!daxrsYtL-t5sL%wR+|us1~c^*pL;qRUeZTapoJkDvrSrA&?}UJ&KC{xhYQI)#N> z7I_Svvy97U$LRNzK82l9M-Sc(w?TmTMx<M&ua;tZ5g$U3C16;W=?S zBa$h&_&C%yL{lGeN}D>1Cv2yYX{IkE_S#Dc3GYvsxdhscs<gzBocH-^#_rLfsWp>+q3;X-`rLq(TL=AHA9d1L$y!rRhR=pPpr)=Q^lbn5ub-L> zy2stYuif_j)5P~O(=R!W(aXjh*@r3YI8gOc(u!`4Pluqj-)CBbC+46@wP-bt*;?+bVxv zJ8Y)~KA#p{4V@_)&0yn=uYa}K&9eD`n3K`7*@9YQ&^`$cF0dSnZ^ki^t_k>^1|uND zb7|14kMhj|O2E8$RBtk`IkawQ_gsjPCB74usa-7^7^8p%LF0+TdJwNCYiMQ=`-{;l zMmSLF>4O6zQk6MC2(V5S-X|@{oC01D+oA*_9(-D0Hkv;v2$$KA+g|`dLD*%|_dpUV zK)RTApTu+gn+hFocCF_A=FhVHzMY>tIe8DeHBm+sLBG})c6R)E8Y42W8c=iC)j~Al z{2Vm9#m@zrcRWt%u#x#a1#TY8OQ8|T;;_DWc(Bm%5um}Hf744BtrKq-8+&K_?;m!< z?TzjM*dstrgV0<`3I`E)?)|=YA#Rr>nLD?`b`?$N61nHYxm9;twQiY^?ycHgp+ML^ zVY_2t2Fh&xs9?p}>)c}Q%bGXQw0+_F$`!BH5k}M79OK03JXq+P%woY@+S)vCa}uJ@ zqNE~NsH7C}Eg z8dlq-=@LK~tu8V%{b<+m7mB=Gl@jSjRpFgiHynk79RX{I98O#PSeTuaZ&|1}Y~%zC z#N1MhZ%|qrQ4IV<8>{n3nzMJ$@|}- zYlkv!8bmX(YgQS|9)%iXRT7U<=a+~0$?y#L0#(nDdZATi9Qv9;~~2|(Bq;lN{j ze8>TDEo>dqBlePx(;2;TV@2)aCy{x1x{5v|bfpAz%?p1!iYZmEx1s8x%c0@vn2v-K zkSgPs8^da1}s=elrQQ?`%>yT8miFP;GLJMWNjfdrW{00hyz?KlF(c5CF zV-m<15-vGFD0I233El*nP99eLQXiFPgB!Gf$o>p@8$TK099zl%O_N%u=0g%3GoK^% z*x(QV+w7b(zqn)*HIY6w{WHY%~YRmxEUi@)taf%&xZ} z+a$C(3wFxm4;(u(SYO-uL0L}=gr7Ncz1UGs6}C=E96&{4rU9XtHaiySkAj{e;Hm*- zbAz^Nd@FjFW6r8R4?u`^m~2GEb4IRTbHx`rA2zZ74ImpZ7|?eDh)6w&555;z@#%Atr`uC|5oUq z;fp}QGM{T>Ceja`bs*n1yVzC;f^JUyb}dA8DqW2sV6-1Un5A>4{zp#LvrZLb^I8`(9&N=rB!qd4?@P0& z+!b=gyGs8x-D8tK9nN4wxK~f)so#bs>7R7V-_QND_#exKLk@RZDePf!kLS0{Mk$en zQj$Awwfq7*yXcS4*^gj$JI+WDLzKpidf2d$>!#}(dQ18n!Y}u*l_4B*u;IIJGp_~Y zRa3eu@i}p2_;g{J)FeLHjrFYIG-{Z6={m8&$?WLL=k?@vyrY`;F8gTxnDSVZjJRQp zxxlwk=m?TKx5eckdl|BjPTkAQ%cJvP`q56YXOdX`oLhr|p(R5VO$3!SU~1CSC&#xs z6<}SS^Y1=4|HGC)Oa+~fq09dgh0B+#V}ONJt06K8i&@TnQ?_|G3djsXw?K@_A;v-f z-dTPSyT2WA(&)LD9kit4aug;R*w9 zr+<=$*S-d+XmS2GitDIAw!KV=?*A=c3CEjxckLm@hsaB`0gx?RoIfVKmMDgx20YZP z5lGygMy)h+%irLBfSOVNZTYh0S#aG0mie;m9X*;rSR9tO8)|`FQrLn#u0c_F{SlY0 zy$U|4H|Rvmo+WY05K%ae=-RZS9e&0N+I{YJ7ZBWQRQn!ef?wOa7~Ku4&xV)hiupDW z(j?1I$s+$`y7S$TZ=uHSd?+ASU*!n?zKM%Q+lhT2K6m1djz*EgD?G6)I5`oOyjDo` z`g@2++AjW|Gh?!!Zf`rL-==)L^1YXT%3ndPHy&1B{BcAm4<}8Sa-MC#TN=-S zx&ciVnMcy%?sMc(*nK*`6GJo7gEgz(N_mLpp+r0sPI82b1KTu7&Cn-D7tr&ALKQ80 z>a#Xw#rW{G-aG*>PIXcZAL}1c5>L*C2X)Ex0OJKt2suPCfAI*4M?}vY6g;*6;-MFh3~fpSwxt7?b=GK?`4(HqpCvRQl zzY7O1c@K=!0eHRM&B*<*bneg@fpQZys$r=K-9>EUlf52_{`>&>spuj8*X+Q|`rmRH znArapiXV24|LE{IIyn=tbF%&aC8%1zScdlhF#Pn#U4c7rdYd`${wh#eY_5N4JO%U= zfz3@R?DeE7oekNa?|i}$QWI^b}-hfehRA z2qBGcQb>Z>5V{c}DWFIqk$#8G1lczaV)3wVTGXSRG@Lvv{DWTk$*?qgptUrj`PF18 zx5#F19)KOeeZA1HM1tTFni!gXR>6#+OG5&~3O&al*ZXsU=wv0{v=Egagyuk?%Uk!33m{i>nL{u(8b?k%|i~#g&YSsCyYeswP&kTUerId#|)+jK@)^V-9!=+XI2sk zM6GqDp$cTZlk3jag}*6H#l{~LH3vE>Pw(w$J422V?CBFsg*5`r5*#vy0Ol%b#KH;h zc_0cE#^Dz*WR&c!M)EH85Z1;TSZP1w_33j`TGpQKRiaJv5TY=qFtTe6SvVlEAtDjF zXnp_+4H7)`)x-2fmsj*zCB(jkzk(OU{(LDZ zlfMC@?82K(wT+lTNrDA~?n`KjL=k$2lDpfHc7pnP^mLpy^3iXsZB*Mfu62!Y=+$_G zUtJvecx5Kn(sx8=MJj6iUxdA5b0<)v>>GPx+qP}nwr$&-*mg3pZRa1`w(SYd>~l`t z`&8Y%>%LnbV6j(s|2zi>W95?SD=9)a;`EZB)t@&cImf=AzP8rpvlV;JxxK1{6KaZt ztdrA`Zi4@$u^*R^A^Ff-1GJ6{++V1r&;T2FSGLJXPBv!clGNOPQqkToVs)3kJo!_# z_mg0hD}*|6aL5jWTBvl=m&LVP^rdbQs`x8Z`M9{a_yjDDi*F0 z4qg2IK`b&fvIiq@OHz(*>;r;VN_Y+7g^+^a!reay=Eez_FCn9x^PsR$i@iZe1Eq%!d{ct$xp#aql8w0Om$L3=Y^<-#X_C}&r}X^xWPy(vkPeH zOgMa(Tc+QX(ur4J#aM3auuYq`-EtLSjhC#c=1h|@gUL5_9)5xYO6`@sZ<=DgKVwPD zyhp?0okhpGn!&(Ii^9oF@7$VHlj&G>ghR*TD$zwS{z@@sRWBJ5b*39MI8rtxIU~*> zUL$TNej^Sgu2wS%nYQp_89Hp)DfasxYvhRGi zjv@od{Kbve(O`d{)EI9+{ly^se6jAL+1hb?52fGLl4!8JctXR6^Hl4t(kAofsvN?L z9S7Ww0c@112)71>xZLqEpzUaZ376S^Uk%u<@?x||()+{>xF~g&B1)KtX4xF!zFmx8f$;ueEgghwhP5F#Ou(+; z0g<}B$=ZD1xRG5{!-V0(rvk^M>L3{oy66VK{FStc7%@6+Aphe$*a=$G0&fZ#Rw;%S zG2u4U`k8iaEY-}s+j=wNz)DEy=p#*H4V)`pIMuI<&-4S8K>)in2J+k>=tryge}+uF zp?B^MPy~nfh2Xau>g<`V{MQOt{DaL%S#;i0=Y;gQI}d8!`w#;lVxmQ*wwQ zrph)1=uio=bX7Q>mPR1nsIDw{X$pJeV%ry=nXN-JzT`=y%1;hVKfU|a!?wG=-P~>3 z4ZRH%U1~`Csye*zd*-Vyz6ObwHqQJ-eClt*(sH+vx?%xjseASK@g#SWA8*VmHhlgF zFs$}{04IX_Wc_VvOT3{AT)k9jzbv?$vQVz7A_dx#)c+a6Orgfq*MdE+%!wBe|L-Xj zAvk#r+w9qNZdvf0e{Wb##u8oEvTjGwFS*kMH!eLvU`VkSu#8(0yh`%50e5&kIsql% zrVTI^KDq9`h{H*WIJsU?qjXUuB;nMH=7zVf#z8rW;MA_ay;l}z~*bzrT_R;G0Q$3(B9Y?~tCQ!sfd6QZ1g^gO5#u1Ne$8$Y$ zH&$xhb>3U^4AdF-w=-^2xjWzDXvYJutJQ&@|v$dnj{zQ%ss~KR~ z_z|BaBZO!7u`nAa;8fzJ>KV+DIJ?&7#*8nms!eTmH9dX#Zv+RErTALj7eP7>#=;~3}b7KL@@@s1A1k`Cj)K|rzGD4KJAEe|h+813y{XYrv zbD52U>fk0$3x=W;LW^?l&3*L4Cy~;7)k@p0DNUP(E$BI*vIoxlum32#Gb-*x6qcK} zYfNpIo4OGFN%Zrar6Lb_o8EsZ%bU=FsR3^P#fyfXa+KVxdxur^=>98T0Gwoo^m@)+ zKtRB2fGko%l&keqzIiGuhfaDjBaxZU9XLKXNR=@G210p2P>)d&18xZ%)NMe}E@m%! zFXnEqRdo%A_rh}ves}Q{Lz;^q1m8UgNCjm6?#>CPh%sor69XTufGuKoZd92IOIgyt zJQ_eo$M;hKMnLMCA`eQNRo|2ZlF`4B6e01X2L7IH0=K48=BbC^2|4;G3YDi810cG3sD zo-;axYOCu)hj;DYf#x?;LmQm*!+ES?A9i$_Gq8#TV6cs=4Jqfe=fnSX7yG*;IA$mZ zcq;m*7LrK{rhT^j7Aa=bDdt=g(lCz-z zLgci1a%eh%Zh_;vpPrjcU_tCOm30|u$V4D>#>KCX=S%!=ZU74x;GE$*zfS%E@CB%& zNUD=R=zF(*xOh3bD$ahJ!l=yH^5=#IJsQSLrqYqu-Dk18K5}<^KydnQw@O>gT;ZNH zqcZj12vF)^UJTgQH{usaBws93WZjq530aQ;$fjpDxH<(@@UJZLPPR}@4n`~N$GB}^ zH5AT@+c?Bx>KiCnd_}(B&9L`sOh`%9F?S>7rY`&i$R2A|l>F0pBR*GM&1gDCBs;mTTOnJgC_@-{r|VL)!RCu=cn-T`W!4!U-a0ligs%Sb`+hxQ$rgR=WdO1jHHX zczDPzyLmdkSgu&y@Ku7_FWI^_(LP;SJ6uc)v&)wcQ>HFmXg2Jc3`!wM03);MksA8+ zPS~5j^{#E}t=6JcjPFaMeGNzAUr$R z{%Rc9h5lD^BHHyXTd0YsI}*|NXfYP^^M1nXKUum*cn@!@xNG;ad_uJYXmMnz=%gCd zk!V7`XV)YWCmriM;#9#MO2n~!0_}~u{AEc-GC|=I&B0N}ABv+nE<5-mzo}L6G)4A7 zX?xPmUxvU}5HFLZkT?W$|F8cNx94`jt>CVE^y!pwg!r(P7b1@JO#` z=Y}0Ek+K21yV279+MH)XZb1ZNd;SV4Ld`9bL(pIm>y-Kly8xYUU?w)y-RPreMNHY)~);MoI7Jhlxn5C;^1zXLAC zb?*7E5CJ8k&K2pKO0nQlY3IUZD!Z&z1ra^qOdjjPgV(*=mOR*MRcmAb} z@ZAcS2E&inuoits+?==(p$cZ3mMge>I#)<*rUSnoAIq-lR?BoEebbu}Kb3gv{RPuw z;A6^@*0GB`PEN*zgi@bE)4c-1a|goL)mNkfA=I9%f4oDzF8*r*p8)@jO%Wo;rXcFS zdN7Ne-2<|;$RTdkbM?C?WOT^L=!KRnB~c=$lEd>_@{RL~x|oGB90^&atzvU2whf;i zP?D)QK#-H5yXdNh)kUqN-p)y+Ou9lNM7r06x?Ed=mMNXH5&5?_mDirbgt^v=n?YHN z`pg=(m&fH=Er0qQ49ax5JV%=lGWy_KK=0qmc3SQsK07SG1+#cV@d>+3D&dA;lp;Ey zVy>NMOR#n&fBPaBZ}U};0+quEN@oixVjprb^=ieRfP@;Z^L?K@JS5%O2Oo8JWBPz! zs5qF2xnS=p4kCCtK}CT6rcttfFIpM_>RFw3*T1f;Am%E`eT#hu;`oqJSX_WG#|E2? zs9MZB3i4)?+;IZ+RVf4e#!bYK7-ytLej((5G|dTr5~q1v+fF$sBPM=4qG6mP=UYCu z8`0@J6HkB_MvXs^_0A1CO6I6FFeu_=GefGx{jlti2mhnFS{f}mR1Xq{yDY^etZ}um zUM`2zFZSTDD!6RiPV@_HooReIR9Fs$0`(z;(VovU0chQbNfR}x+EL=7&s!8pYZ|l~ zc8OE+XrUqWXUTpLBMF)5RB>XRt?32cV|;OZxNqqZRa*H08wLy2nP&;nNL!5Ua=TGK zwy}=+57-wWx@oN%rQoQs&_HJx6+*}W@>Zg29~vCt_~0b)n9j5{)1ZL=qkM1Rxq^wk z2t?WhuJ#SiJ`f7tFsPS!cA{a$z}eGobuH%(BN+_Peh zOe?OpFcB8Z? z{DS*zW|@7NO}x9dZTgGbUqflb$xOwStU-GD?vH($*Hz@GqF+;q3It0O?m@e;z@4|( z(Z4jJSk{r-LoNfB!R>-x!iL&%^?TZ>Bz)vvVjdN&2b_X;;NvSxL^Sk$)7+Fbxg_)nl5l5%P%-mu+mDuhLH{+Q_t}8 z1ebyGb(~lbI-!Jrg1{83**3_wS8&nWrYm>U_U`HQSa2f+91h{aafpYg%@Wd#2L=_8 z3{?gmm|fYNML*@fF8QJ9exIOXY-pI7CvwLlV|GQYlBpEY(JtXw$7Z2r?K?m4R@C3u zgJu6whEH_cBAkLL*28GOtzttStt5eE{yW`m^dY7C6bm9kG%Rso@OppyO?4zt&fxbB zvI{XTl$*_+57P-LeV>78nnwuCb(eL?u~LOcYT3$cxuVzwasJlogUYXkR8$cS{_N}z zl#XxSsQhwtDvbM?VIHDkpn(cnTuQ4yZxM5^yu#w7hJ*#ovvt%N?~#k*0zz(qG!Ex4 z-DoZqpJFUSiH`7WJWKg&5uMhWxXIe{iL%^fg^4`ndJ_}U^U(=6RD!l1;Doqy^BYaq zXKa((Dp#prC^DnOO&s{Nap4v3{LxNjy0*&eag6I^mXGJHa@#=|ROv((D@Ev?T>pE~ zgfd<#*^ZNfS_|Gg&g%@x56&}evl?nJ63*GXnU>Yc)O0-HS9U0^jf88iFJZbzw3v7eSfh(mlh}Bm>HBlsb zi4jfB#Z9yE@^h}vjm82dRo3!h(n0s(u-*g;uZ88fwG~#{vfyAjmUN)odiXtZ^W74H zU0?ETq`iv(Ej*P9B722dan6aSA3+T$MlZz3b{pV|Ir-xck;-Btr-@<1b7&&<(8!?r0R z-54d61lHr>ZiG^{L`kMZVxhgls(*&yDRF2=y7Cpr7}If;AIK9=AIT0{T4I*H->U z(06gthWI-F??qNm#PWr(rb9(=Uj}(Lyk>w=4*F7gA~{;^mVW8uNL^>KIK=!A9SqjQ zTfAD6p)!N93h}~-+N&>cEdMpal=~zzf;@<68bF+04B zd@0>?M|uy%YD+X6>+1`r-TNLG2OEg|czF3Y=l3seaja}9H$||M@F0G}v4M9##JNH{ z(G)-uPSwh2O7%Kds$U4R-bpkmXMX8a$PjzplA9gUM|684Vu@D%vXTRC)>DroJcp)G6J4Obecn!xt^b!|D67JoePC>eRV*%cVi6qhV2JcJ@qzAG6oFr3{t=9v#2j1-0-m!YVz^h*`ZC^7i5_2bv^1WN z_k|Umlh|q)Ho!D*o*@CPdy7ttEkt``RbLMk2FX+Ja2Zaaig(%sxN=}9zOl-fohjn} z(&EPn9ok8IaeaIV^w*DXAV)Z+g3@=p5{(*iz1 zle`MXcB%E%qFb62fhBOP&q@=l84}@Q5?y*OjDjVYU9yV>>lSH?IK?7xq?sUr2;Pqt zosK!osz|fjWmQ$o^<07pu*$VMV71}5EH3JkzME8OpCcRY!u^BxbcTT;F`-`b9&A0i zSGKNB0(2b&YsPf+I_s3_YiA~ZL!jQE=qv6OnvXASYFc7!!wtpo4adQcirUt53%mXH z3##uxc&rPjK5twGhK9*OejR;cqK`NX1CGu2G6qbKW*|V%>n&MR<5ct(UP^FQ3{i3k z!UdDZQVaCA>y4O}>*?E((CcnaBsJed)6{~+1Pxna^i1jWzQfH;_V&;>eRf+_Q#jTV z(I1X!5H@yjbS4L?m&vXsxm^FK*L=KyxNpG}Vi2{h0qH-F^wjpG&doLjJw8JQ3f}~a zYj4&6os96GKXNK=#;*U{(Dz@qJ|>a>$I!>h_Kz+5-)5`+Q&$Mn@X}OSZ506KI{m|$ z!is^LWpZB9W?i$=sx6TNdBLkz16n{#Bq)q*E_{-M#*7%wNI?n)6%t7tz0AD~A04eS zzp38Svb<8*+OoZGcf0CA+hxc9kpr0)h53L#=Ca65iVGN?G^VtJGhAreK1f9p^;B_%DJRyryrE=Dmj zDm7fJPEKZz^0OHaCe+QU)6Gv!<&`IzMZMd~$H#8}K9hQ1C5Ez-!+BlZi7~J z_hk9GBh4$pkMpL(-E?xa5D@uv)8h#X>7Wixq=PS`%6gCXqbl~jtZ22NX8&${RF?zi2yI+41{y>W>fVv%LO>AR?#G)q%li(AUa`cf>d0C4NY}bQwAgWT)w&G z`D01C^i%+ELyJ>OL$4eXx?K=9D}>l@X>`xRopf0{2(3cb+r}=~K7;$>GDVH*)glCMrly zSu(-fiq#QD1z*)IdplqGTIdvEByNR!vl;6Zt^*HLC$QMg2oZXDvj-~wNwvSP23xrL zf)z+%;1LF(S-Z80^;RVM!yaIkz+OLVR~{7fT6fasWzfu-SS7qe>XQY@UyA(Tze?3L z_+#IWDCUUjbmSiqP)l@XWjoaadh8NCBASh-e?1n(}Mp^R8jtl%*ZTLX)grY}xRs76j*p3t@ zGQiVjY*Nf%0OcAtY%sq_!i&buWJfD9{GvK$D+ZO0h5vU=^o&J_b$cK4^o;LPI-fRl z2y|Cd73m4GNl4kqEW&_Q0WC>Nic+4$9}P8MwH2t~1R9!~i|8v9qICT<*$Q|TKYz)c z(t=Sl>=DF3!{HJVvjNl~UZ&kBVc5hJq=+w4I@as{oSU4+=lCxL51n2JPF{rS;yp6~ z2l-!!$noQ2d*S%)LEB&Ce&0GSe>t>&rdPPD;t~@zkstMjdr1+?flNxl3Mf#UzW;)& z>=J1a%r2BxCFURt37vT?ke0#+w{dQx*C@wYB>Z*eH7?1{BPUM;L#8lfW~sA@XrgDc zDd{s#N_r?dxQOuto&5TyYhdx+N9e}SjM2ERTY!>OlS3Jy#8H*F7c>G_EjF&KrR_#; zDr~J~9-CKNSNU`m<8rOr>wcZZ8~(*t58f3t?(&X7ofBPO7O&W;^YL%$FlS8$^@$Kv zA<2|HTzN_Y@uIxnfPV9QeP5I;v#b*IrP8sYH->GsTKA|A?M0NZAhDt&F$*mwlB8azZJodcOIYxPO|>&S)5LA^K<9(YMCixU}#;5sjtAG z?=#@W+Cp@3vnUL*Wsxl>oR7CS#&0g5oirnT;@8A2$L&$C=0GQ??-@_N}a%L2E;z@Pv$xLBVU{b$nV{|r8wugmRjxWl7|hV=JldP zuvkyL48l+X1C%J8kL3gs(A@)e{ku`<%3hb|S z$xuY^pgiyl&opmz!lZZcw-9SZ9^6w7RH07owtF|^tt4l+w&0srs}7K$E;vtq%GTj( zjCLl-4KN|A?LfUi=+*>3fJ?p~5gpHXsJ3s4JV?EcE-sMhF2MCfK(Qh!B`YHO`)`{| zUC~3hxiI^IfP9H>w+`ba{i|PxbCiBAfk@GYGJ9H-&ZmA}oQ#ksE<%rxJDkchdj~&or3m zw+S~o=UZ??cL_`k2SUZYqo9hzW4U>La#O}IvGnlwW{O3kE)<8irFYX5_65C8NE^D~sMt7<(&aTvQ`qEeG_vACy9Qi8xjA_D|4Xk^21U37^!N|Fm ztuY!NEEzICw7?q*=sK{`0Xn{ruR|i_M)>$|qw`d|ql|1``ZbW2EbE{a@SZ-h=JkP@ z^m96FS{vKaHHwX9#h!RqfnXMm9486;w7$aGIl*~fuf6En07&YW(jv1b79Bf5yOuLq z5kgpl_E`8Iw4B`;LON~fFU>#*e~`%FUE(eSx}tZ3%*IAJEt0p_ z{O~=TH%QN(S}WY_Y-d8u9A*mklY*Bo2&2?0-62O+5K_A3~exXgh$p%P=OF- zDoNb3q$UNRq^JW1FH+Gct5H$IuIe=U5I<|rZgDH)&O|<;qo>mWITbWg*{uERXf8AE z>LEI5!MScKwV8%rCykgD^U0fU@?H|wx@$)EF+eS?jC2D381n3S}>^Ub4?v;KKhArWmKCmANNBD1W zc_QKp`Zv~ELx?Z>mV$3O*2bo7k9GQHWp|N8>g{g%B&w-t*;%;|1i|f6tH~bp!Yg?` z{?>}hf(>OefpadWIrp?y72HCul{g zB+=fhBpi&^gJXk)_GyjV`vutzDbY{uz&Lh0FLWnoLUGlPfi4aG7Ciqvk6J>w84>=% zL&JMfT}Q`p$~&SX;HoCX^F)dT4W)QHGub4Vci%oSAkqOC3&&!8_AR^O1Ljh{FoGAX z9B#VH_x_|K&27uZ$}i}<`D14ljThB7E}O|59J!}Oy)V&gP|gNB8Dhy*w~=3p673&g zQzE7`jlC@0GYz3JYX=IWyrk7Z5tyUsnb7PO2c(-^BtDf9Vsw+~HfBB6iGuU62V9TB zmv0C5grTG1Ac9x4Hy{XhuJw)W=2g5r@qxYK)-pfySiBd{Q*heI`X`t#4 z+PN$3SstE*{wYS`F|vXt4~+jQlK1V6>`eU&NnW4C+?H4O73W-e&m=o-s)E)ZL*o*L7rMUbdQ>Ck<5~*Y(Hri@5f?sce;Af-Q@pbey^S}kX@ zd4J~|^9@G1tWR6G(g)w;FXmK%E3*XcF|bylo5aQsNjgN&aG*Qf1AgW>Hu|zaVgNAe zooj}((~Q9^99g%2g^!DvQYa-`*D_tY>2ea|QXQE)(!(Z?Eiw z5$9d*i;zV7?cXvAU`u~6^{DT`*hklqpR zWj)va4!T+wq9Y6s>i}MBFT)hUMr`Msj0JD{yjsE*r*panc3M{22W);Da#EhpDxC81t*y)Hn%J3IQXw~NnuCF(i<_5i9` zzrWAUh%Rk^&hKbA3SFjm+4xv~BSAZJk^IYwWU6>Kvt_@Biv`}V>WDs;{)bVB$gS35 zHmY{^z&s5~$N^^SQNy?0tZL10>C$a$FYRE8UU#**ME?89OO(2mFb=fJ)912|GG`!&WS4KXKQ3QL9<_8+>KF zk>uN=jy8_wygWqU^kp6=79y^&gAHq>qwlaVHYk@IB(EEXVW5yxwHN?66c)8uUJoU2 zB<90Ung??}Y zRtN-#rfIchbw&&dTG44)QNGnr7dhLxJqwlpFq}De6aJnS)v`ABwaaR&-nyVPg_zX) z{v%Dw&Sfpt4e=)*0vC%C7qNQ1m8@An=#wLByHV}G(~Q{u!+P>RX+}(pOdS9Jnvvdu zove--D&*a1tK+R1p^Glg(zFiqsZ5!UpppsedWmA=2sozfxF=pIEzH70r?PZVTFOF1 z+9ro~8zwn}2BmA$IfN)XD1qiup*>uBMj^|FtZ~{n6o5(}xtn$r!QXc#IcFE}g9|bu zi715V^gi*ssQKw>DOW<;wT`q?A)1Hdl;hFZDHxSv`1kj<6v#S0p@#!+6e-vMzOp6% z3Oh#yj7G{5@^rMyNmr6wghq(8I&L8-ERrHC_YHhXoUz+(_vr+VO|^7 ztl0Kc^R$_`*53;L?A-1>2@C>Cp3Gyz8-((LX`1s;ej)VDqG|O`ZHy z*>h8_az-Bmvl=VMpebgBzLk#kJ5}ea&eZ9HF{fwm*}a)KE^?XnE1}b;S{Ubt?FvUz zS&Jt}d9jLl_%ln^mZN&rNclW*x0`dX7uyPOkD8^ay=D!8AixhlF;x6x2}5 z%IccpD!jJ2tbtCc6)F7>*Wd9_3eRejJ^x^szpKdieCk znC<2<+gtN4DZEykPM)loN$eP=FXwZZ;hjfe+$$`s^K-1Em=p0FOJO7g3GzrM!httI z7{~%G2!rHr|8#^o$jia1au_%QwQ<;)e*_LG*jdnf?0Sy?=xVUFUjHu8*F}7k)rzL2 zw(RwfiJ}u@-N=}59CJ6KP1auNM<&m#>|+LlqYw=8k*5fKu1=EE1)kIJK}3YvsPz6V z5GhuLv_kD2;rO;Zd@;oTIhJ zlkRK>5V{PblyW@b+@PI|y3m50Ibc+GV4-`NrlXElDnu%Fwl%jJ17COk?I2*!+cgNg z2^PvzE!F!TM30kR->osnly%D{oIXsyXx2FzaNxBhe9U#q72>gYz2kTGPe0uSJX|dA zvA>~rr}KUC1x9K9RYA+6+yDqaBeTW?v?*jDTHM)c2aZpoS+vy-&v%$*DcnDbI`Ank zPyI?1uS&i*-1zdg$?!On5F_RUo?`c~{<}@(KV#+p;vtqqPFo( zSNf`tzHUqyxDyyzL>NpN4Fxscd)sVFXGdkXSSGhwV$*6au%29NmEA0t&AMIaYO-tQ zw3djZ=qM(EgnL|~Z+;SBuY-s?=G2ST=kVRP$;QR;n&aV@_u5SXbh!=+3k7x7Vtq3~ zhEgV*(f`RZAFH_e`*3x~2rS^Q=4mR* z8jpeAs)k7Q8x8?h!eoeNH|oNdtoIR3=#=Ih;dX?P*3bO&L+f*I9h&HtbOvckE-Rf^ zz%0nIkkVG^Vr9f?UuJgN{J1BXjM~l4~^gPHvT%zf4#difkJ|=JMU_@Zzw1%=qQ15BWy@J3K zf8X#8ya9ygP6z~c1_}o}8n^@O2O2R{fnhZo8?!!U?dt9K4t>x$(K{z2;Oy>HTwwL%bAsuLi9$c)b+8a4)r2Chhrjlxz2-2kkJZd~ z?yWUFnIeNcnmo^Qgu`m1Bu;~)kA1I2sSb-gMnC`Qqm@IwhbW`bapKis#`m$QVy5D6 zS)LDUx7z1Qg_TQLADhYg=4|)Zv+QcMhWXcZ?!L>()HRa^zr%X>&OIYa5LzOaIf z&MXrB4z+eHzZ0yi$KF(sK;ke`oi-X&0r{pgexY4cP zyK}RSxPUlP31@xou>vORKAWZ`1knUTW|>U zYx-x$Sx_du%jN*8s*|Og+D(az|O|Y!K6As{Mx)`9$g{ zM8x8F2RX~5>W? zBxvWr<`>^R!)nP2sLjA${n3!HOJhpj zcmo1Hf!6HSulWz`xM1W%d`RHs#~Ii$vZHB9Cj+MZOkdxP_MSGrsFVgobv=8Vv;wx~Ut>6%< z6|^-A-YYSLa?Kj*7J0TDe9_bTZ0MC#gspl=(z+Y1&$L`-C}5YG=M%Hr{3$wZTP zPpcMIA~q7WJV2~|6P_N^B_~XUa9K(cx2~B+QH=QzU5+b99W-j9_~4#3$nn&dq-Z|E zP-t&lx-YQA%|~jU;1Ax9!TRS5)h}4N64Y{zx#V+>vtL`T#QFM=V~dA-(nraJ@(}sxyC53l|=j>&5 z)5z|>pU%ZXGz(V~g-czf6CgRRX1TW3+I?QR9Fh6XsC@s1ueCG`Fw8OVOY||u;a4S0 zmGSfW8C}PE#L0pyzbeg6U{Jf zKEMd#XELxzQKre1E>uL2RAD<}X{D;J-Ty|&Wbr*Q`C-|GH$rn0Cqg}aNLc*!lQ}zp z#~GucT}gunpAY1CHQ(bH)U%=b%UY5%9v`TDF%h+jl1nR{85h4>An?~2@TB;hP(y<= zekAY_ji+!J^*v*qj&E`OGZW%*rBcNRjeV6@f|?;wkiRp7w7O(IoYxq6|7yLR)Cl@OUT)+%T|Fa%qp9#6#8&$oB)C_Y{o}jE3d$0u6 z{d!tWIvE8kc&D%(FiYuna%QAkH^fJRM%){m{XH1FyR6BhJK{3`wEE4LG%>L_775a2 zP|Cr(1L8CzhLu#O=(qgZi5!mkB_`dkK^0}7#+nBKq!z>SGR|JXKk<-JC0n2|@6I6A z6aRp-sg-C!s)92?+TAOq45fRBpR{j+Dgx6;X3>liN%%PkDvHFg^;_N<<2Skbp8il( zg%{~bnjpx@ueivIvz0;TXQ6=;L!{K+G8ywx`0^bVb;#mNLkA0L#+P z2pYY*wQj|f3utnezXtSps$B@?o^Oy5 z_4a=HleKYnNvxm`A1!(YE-ui|`*hO-WFS1dDVFd}bmO2BfMN{TEQMA(pD|&xOEWLj z%Q!SmId2;1e8Ot%E?yOCAY|zn8aSAk1rF-HJ0_IC1J*wo^n<{&h^La22%FjS;c20L`T0LYoZ{?v&(k9E*TNqzAb)_RAlC4Iz0=S; z!bCL+OkIy*9hn~)NMC?W%(r#AMwjE)*=A_jEzk;CuemkW&iD-6W_)(EO{pfRXe*vj zRapQ$U|2XaW0S76v^{~UNP-E-R8okvhYT`W#B|B&lapx!7F_Ei(Fa$?HEL$y?2Sd} z?^dVDe}4j@w?Wz$@-LX8A+E*GKg~kptoM9Tao-}MM#b4C>J(52Me^Yn0A>}%6rthi z%P^EAjAy?CtEn044jMg|0cPEtL6(~~`S#tehbKV_uvDPPkkUoph(!Zc4n?$_RpSwj zi?VWx7%QHp9pJ-81@Z0-)i|q+wca^e)~FbkNJn&FC%P{B$7b???i9o% z?~Q`AlBjm+BjvPm+T^;8UyEn62mGs=%TW4>d3XhwSexAX{7Rs>m4h|nrBpaU%yhs< z(wXEqNJ#S=&1{2in*z~4oHydHda2JJ_JmVFvTcSGwBZcUTOnEep^54Ap5gWEL;n8o zT?^YODhL-r49i#=gR75%w3*_vi3-3jt0PdzB>n0S(++R9xt5RDuga}8$D0W#_;p(z zk((0;CU@d{XN9A5grsh5DVik^Ej!n_C)0ILr*L#oOB264~U`q zm%r^H4QaYS*p)y9DY7#HTt#oOy5LS&w zhDx$dwn{ORwBmoT65c$Jv76$qm!HJ@PPg&PHOk@B5{sxgkjD1&B2^=En2M2PY!YkH zXYFEmBu2%(W{fj{#KH0zj3pBI77xOt0DQGEs*>cAr&3X0K=vIbcHA=LtvnZ0_+XHS zf1tuv_V2Zuz2$5b5s{8NIFPE>7TafWHBDbtVO!4$bwrrf&g9D`m-)d5sSMFI?U~!P z2t-Q;48Aiv@ZcJl@Xv7x#?D)%X9c(0*MmIyb`l@07vgP!w7R1ZPr;dfmBuUk+P*{} zH+h+g0h1>X1r8fSM19I=0?;q1ULo4_s>ysD!G>zWjh4K_%v-&qOaqo!lc`pz#L%W; z;!@4Q(-QlW8qF|6k#H`WwyhiuC|y#TX5L~_dZ8pzn036(k8S$5r&8+5_pTG38jqN4 z8J!(7xDz-00a30HAW~8Z$pdg2WDWAV$abVJsiSGgpqetCdhl+Y7)pZbh}p+~VX9C! z^zUhf29?G|Qo+uo0 z$kmcdd-Eh{H41iuc*GBg*=k!q3B~Qe-0FAj8X|6mvMkRWyjKX{JBnT|)Uh4` z8zdqeLhT2e1>^Nwv}RK}y4x|?N-=z?ELWVw*@1qlqeG45|(gcV}slg>op@~B$`!mlTfhxX*HTbCutCV zk^X^Nz6Z`5`!)@3$Cx3KV!tz#2>g_M_z-i9#8V`0>XzwF7ZSu~<6>qE-?0G2;)@>^ zKj^~Wz(3_vSvxx4L>y?R_cBCec!4VZ;DY|H`r~v2b)1LE4hwzIiP9J?gk=^Z|Het3 zLTLz%nf-_&0Tkv*^nHT6RaU+^rSbQF&ELCpPM8!cA-e2^5B2H6nVf#6Ze(@;{l|v9 z#==reg{P3tOF2YBiwfH>5%m{|7w4lUrY52!s#l2U43ZT45E;ZS`Gs;gttp4^vyR^(wFkOl((U>Bz-xx zfAh|#LYcAy1rttYuhh$&GGiD3k=0yj5IaSJ?eCfosIOt zjwQX&a3tu%9~@Y5fXjzb??=}UFg{9wQL0(l*`}~~b6iCuVZ1nKJNDCcMF1pJ2Fnfe ziSvf6sE5edaQcY)CHmnm&H;DB<8f-%b2eHmL{SRzndU-NxNd44FLSiem4pXSHb#TR z{vMD>6wyEhBVbwu2~I$iKh`bbAH}C#a$S2e^D=s<1XX52(+i^DefY@bxcU6`ldDs+ zEoW*n%_N95VTF`Z+EopebN~E!Y~Yk;qn9XDspJ~b@G}a2!SjIPykX8CJaz|m|M7B) z8EYrWbaO0(p(E&H93B)0xPc~+d-IHkYe;%F`jBRGK}O~DOkF|!LXN1sXk5IiorK(} zpq*}72WWz(UNB5`n2fB(qj)-r0`r6?|3;kaR&w@$xW=X&InZj0Lx?O{41hO+YV%xX zE44cdZar~!FdF*T!C|tByOEBk@QT9<7|&L!5@^M=f7GZ9@oDi`3VMZ9s(cJ`0*;UW zQzF-stUtDj%xaqYHvxerY}u z87;P!gS>d57yQuIjn0wccjuFw=Xy^wv%@J-AmLc5oGAyZeN5IZ6TkQDwo__j{Exdg zP@EqydFNr_cA#HqP^~<3dFIknS~^WfQmoB|HHwA2V{M48ak6Q~I}$2sY}Ao>(5Zj} zM{(!M6%EDQ@70VT1_*)V(B^Zp11qj@eQCaU3&lxj7-RaG0(qK0hA(nLXUFOw)+N=o zQ3K3EprcECqqBq`7`Gk{`8gmH*QB!Uwd*<#%@&5vgl`QKEO+IkCq`_=WFFU8Lli{5 zdT*e&zC1u^9_4Y^=(P|f>5m2K$-4DQJd96`m~N_O=1Nna0kHAPIQni+l8$F%$h7G@X{gnzmQc|vn!1H-+VyClgqJl8_uYYaaJ z`DM`VEh=f&8hJ!#QcKGtlm@_wcnWM4J=xT)YP$6<($dM=L{jXWbe(qhV0u6tf&P{^ zPH0JfX?^oypdudf{-CZH|94#~-jrKrw!quThV^n!83jBNtmnbEy8;FWH*!E4N2%4}Oww?+^2-xA{1DgJmZrm&7^Q)RB;EA<00B zZbOUC63q4(-26%B_Zgm9oSVWZjI0kd;nRzo=!|EmtbJ-ksh=cEX)4tz)#bQu27h#B zEH+#A8Q;tD)U0xA_h6`6-N4)F#rt@2NqycY2;S!b5jRudB%r5;mSCprLX&8#wDIK3 z>&Z5iWZanT+&xT$tF`yQ*;ImuaGf#=DTY=WQC7w7bjb{YaHT8&)hWhlr>B**nj?4# znNGP5np48QnHUm7t(lw$m4ukC=}1=keHF=^ui(en_?#aF-N@pb`b$s2{z#xImu-%LTr+uJua&n*=q}|6kpA zp6zEXS@Wh*9nERW9Z$zOFPH_LP=c&-1sEoQpSF|KM^7qF$h|2Ni1VL9;zB0y7t2F^WfMhtI^)UMScW9QB#QB{oO~ z0`E_#E#)JLE7N1X0(-sPHQo!QL(sEvyb;Gyp?M4{Xu%hT#yk&ae8rnp@UNT(PsY(T zP4K!Sx0jndHUu{r@z(5)s#r>plWrh~z73+*68#eOk3@ZkM14z1cCyr>$m}WFIN!t> zPT)LFz&u_k@$UGQ6dXf?zv#M|Fs1`l^b)O8oMNdu5UK%zC;#?Lv{7GJLzcfIzkGyE z!R>=aXJkXpzAs?tyYgD7#fki366W?-_V|zr3sz(Q__3AIXV(q_$|rg#w0be>JSD~IS>#ibOy$}0RbvjWivsU%b0VBoasE*Tv#g{X z7!DeCOzK$>+dE5n!|bBoh8qyw8O9&6Za_xU9RlXb*hlwE(VU^!IN7+^+L>I4jq0WL=FcMINtmJf??dX#%`PN;RBV>Zn&Cf6!j zrD~+-&TNap!xw~w&jzN2Q-xF~{v(ndX%4)HS{ygCiw~(DkSG%N^L;|F`>zT?Y!Y&jNx2yMou_mk$&-HWu zx~)hDMz>&0c{YaF-StBQzuTJdx3KY<42=JTPVFcJIUwaQ<+HG5xuvQV z2eQ_ME#VIH1F_m`oo5L@4-#bfWCO5}m3zxXBT+N4WEvnsa8r4*ch4?VS$$?k_y=U% z7MAY*Zovo_8M_u2>^NV_wh7V z)s0-!Nk5G;+*L1r>I#WcR`N?p7EuSbf9(_5bb_yJzM&<)K}X6aZmFuU!0(wDn4ae? zZE+}5um9pDmbXPPlUzDMwo>}F`p_2V=LJ&1Xzc$w&AH1c&8_XZI;Iy(~V8Qu5uJ^Z?#9&vihA*FAt`4H zM?tU&E9WwXJl}UDITX(yM<-8%u>upU#}j<6ll6&iIMVSKwatC}r|CO@X9tbf)$u5% zv~tHbi4~Lpl48cHI&LSvdCtGj`tYLj>w%tLSHqDY%Nj&RG^xf>)ywA}0~8U%@guR} z{cXN4`~`I5%XTvogk{$yYl4pX=p;28r<3IBCKf&+AX1o$$ato^ z1AqTD_8)j*G=q7mxQtHT$Sgtd*`fzamw6gO@!VP^AV>uwjY5)YUz;^rh%b)?-y9E( z_KXM)ytPZR-O=dfU!eygHufnjSg!ULz`yqCbQ+@fVf}fz1#+^0ktEe(2Qg2pomAA_ zJ|z{);tBAh)yv$zHDM2o_NJBfLMO4Nd+R*Uge5nHb{r?d_{L_ReVaGIP4i(5eVvfr z&9hu!?1^~k%i4rg@)8nAk9s|?SG*f<-S!6p5wjtIF8*}a^4J4mEf*F@V92Rb?Mua^ zPY=xpGOF8v&7($uY)b+bv{uPbwYp_07RJnLruLxO0v2)^z>I#uQl=`f>qk-rYL_V_ zzIh4QC7ypLFlJKP&D0&E@e0%KXr7m&(Iw4kV~)zCLa7KdfY;g|_l=1*v^dZ5NkXg% z(C&k=>zu&-?)s_%NV6c6xLzCcy{8ADuQ`9SR(TshuH9?RY5J*~5) zv;}D^5WA5iQ1dd95a&&Ob5&AqUQ$w2^fUCkL-Lr@^yK7AIQ6)o>b^ccPP0o>HcbTx z;zM}%f&^WDQ>oi&<%^$L=GqVscBky?PtGxJ2*7=cCksCVI7^Z|v}1|f=(!Z2E2E*V z9*71LdjGu>&kcV*L}#*t9B$_JmGBL~Za-vsZKe+}+$6mas@qyxYI^)Qtku1T%XWul zJ-QpKR0P|0-cl-43v|H%Lciha)rxgm9@O{Eag;Y0Tu6;t z(d!#(Di&*@q`I}t$O zW}!F9`fIorTiX5gO?k$R+qDCNXNyKcFiWnM2Fx1fOJ0+8G)iPQ8#5hm=V7bEaMUKLJCuags-z&`7pqwVI zN|aJd=G@6GKR<6aVbj2Tf2U3QmXWzW3Pqd%Ztdkwyyy?4-Z{0UMW%^7At4=mS`-kZ z;sOAn@Ag99H8$@fUm%|CJqT&v{-@h;bzWyQXM-9=x20GO5N8|c*I1)egJ+Hz8!Q=fUKfr9m z3#Yfr)9zxYBvj*&9X2yoERexk8P#%tI!ec$o_Z+GzjFc#D$rKq;kjZ15|? z5$os1o@yt@T2@pd+H=ptn{exhNID2eRfo$D_R5NiPtxiDfj#ju;}#R7Fnbv92^l+@ z;&9*kR&y@L=7ig37pPH=`Dxw= zie(7Z)DNOA6uYjgC)$WbC6cExCd93LaGJLZjFkIj)n4Dxr=)>dya<#V2B7nDEZBHS4v7hnwJ z7QJdjMR&r1J7=U`I5C^q%ml&aKA)XOcyqDH`#|%CkcLi2V-)0}g*|JZ-a?Wo@Vt^4--Q9q9NjHv4K!%@|}PNCm< zM^N?s=BVm=1yJ?ngDDoKJ|MAIH>c+D;#}D0PwmaC*iA?6yH&O2VFyeHcDa-o#5*G1 zSbq2WJX@+Yt4jtDCEWr-zB%lCRHcID`rNip^VvZF2wsjUJ!LPahQ+P4Bt2rW>9%IA@K~$9VPN2g>e`NVDf1(Z((vH1MH=d>X-nXO)45TNP zkwYwnz^bH>jvu17Ip9odxktRaxSb#;P%%jmS=CI(eF_Z~rA`l|4T5N(IQc;VN z7tj2G3+l-*V8gm02lPNYfKCfc*ut>^eX63R8W0Ih!RmJ}JiN)qu`vGL{XWw7^SvE; z=d(ojrfF+S?*pUPWg^p8lhzcvS=opMM$lzqm`JNF4&GOQ;g1PMFE`;h@N~zkx3(C= z$uHNVM@ilje1f?7O&?j$xCYBLfKTBuDq9yZB!q;SPicEco>6BN$f-}1YeDkhw@#z3 z=)t~zW1hcaWAwK@Ps+ao;dtm#!ar`hx-|Ze*8L*E-#YklEaW~2e+uMxF=_tIf+X)C z|LwRB`#g^gWhhgQ+Vd~(W;o>E(1jQu%Gk%^?9ojuJ`_ChPpCX5g$p| zMD~t;8zknH%eBQrj}=^B%s*!ltl{l53Cg#qIUa&P#11KE;SQ4ijKIjW@q2^Ja@IAb zcZ-u1vYA~OcsT~))k6}2?~Vpe-!pYi!}VxCqqqmm@6pL9K`kS4mZ6e-?2yh)O-bb7 zF1GrM(RJsARgRZL3c`n=dPOXtaye#SECug@WF)xDQ(0Y#aOmK#d{0NH|-xW92p;T-zswtYO#M3(~#f_LQ%nnw&lig?Y^b`4Au*Xq4?8aQf{3@ zYDMKaEVid55C;>^%inz#>T6&iZ0nX4pPiFSXK3>rRlqT-t*JC*Vz$|-oCxNx;eh6hRwh%tbkiP*=X} zxcEZP`Z7;vL*mU~+41c@`p>x^g=-aSsY;p*ouGo!ff6QDw;+MCTzXRc-=*HCz|x6y z5cZ!ghb92IcgEz4A(4NacYqGWhV(zs_BZsm+E>=A%KyDTB+A}B_(;5+9p47wuk#Lk zH6j8sD1j6{M7Utgw{>V3I63>kcy1$RVdFm$-0v(h&NE3z8sr^N!n1_OZx?ls|1Uug zKZhYCv&gS%IL+nsJ&<;-su?h9g1)vAzIx)2h46AEZ1bK8Tc#G4_LUYE3|KPa-#p38 z>Y|nos1oUuDh$LfB~~rgAugoi8M@|?G?DZXE>^JELhJDt!FAR1reA-{9;(mAyF2Qn z+bWRd7&j;?x^s6)Q+*Tb9Y}#X@*hu)8q%MX-#y!V&UcEF50d?_$H^?$Zlm4GSOf6C z|6*v+f6}*C@W$!Fv+S4`NIPKy(G#1W2x6|?gXml9{9*e{KJtzG-CSos&-u5$LA$tB z0<6*JqV=HCg8sg2C7&FFgLerLE1z+FXLnR4x@zzR6@-5|sp)o`B5nSlWpC@Vk;NeFs z#AtWj2SyBn)BXbByTn}&ffq&dW4}2^ygfC({6dWww~}!EtD6b$zByu&d(&UY^DnWRVlhIEI6pdM@RJU|PSU=f2|R>493Am<|D|r9dSCij=fU zE{12RzQGDUy6fKDH|5E{G>>-~UR7_}jT;u$@aO&1UtG?!LdFUWF!?t9OU(P{-xjv_ z)^F2S{&P$7&lbWZ{5xgam#%@Q%e1?+df_q52_cvETQqjxga7`?`}Dt+-<9Jh8CJjJ ztX-Yws*6bGDH!5unbH-5@`|rw?;{#~H4|>P>~~P`MQyjYppoG*{djV?`8fBm_UZkd zFl3vr-gI4zeJTs^?_z71y}X^B{+}bVO=MXh$gqEU;2Dp zb*vP#QbtbnUC0U0dGkfUVuSjb=*%57Pw6OE-b$jgIz-yMzL$wk7OCuN>JM4-arV#O zM|iZSg8)<+gzsDPnu}(J!+aTF0Xg;Dqq#f2Om&0Tec)4VU4g%H=n80QALgjT9=?8{ za`k{C?Y(cWgO&uib}#=dQcYDrrzh`C0kqyg$DM7qUz+G&pE@-?mEKV@{vNo9)Ro zdCNBOlXa}-a?M&kwvUrD$RN%zQ+d|vsYmzvP;77jalw#;adI$_+sVYnMaM}=Vz^Vb zagY#_;Ye*%95n8mkc5JSjDz%ruSWtg^!tr*huX+Tee6~j^-EyrdX!btI(hwm4GM^@ z1jJqjW-A5l2EU2ZWqP;95PxrrxJ~RX;wk3sd;YtYx9Kc4pZoU8U3^Bb!EuNmy_fqu zmP_d_o|oNav$fv0te^dG@S@AS*}T)i6OtI^PXUjpbb$u z;?IbR5jbP)J;@$PPsC*idt*?B;5!0!WafyR5h`Q!Jt-f-9+4Vyb0q5sju9}7@}}Q^oWrW2BItsIUOS7h|&=%Q5wb}!-!$UFoO~N9OzF( z)O&fb$AWh77Opwxp6k}J%dltblc*D#`<{0 zcpvL4b30A@>$+Gk+WS_0J-vtT_ceN_)@^%Fnd9{ME;c^jVJ##mGP|r(PKzhg!`KM> z52xwV%B6C=+&1T*_v|10y!<`B@3ojq<}I@?zGVjMiyF}8wfV>FQZMV58%pO1e@+XZ z2=$-Sj72F6TZf?8KJ9~r)#2B|I#cf9m+%P-1Xv>)@&f8|k#m?*lg2KV66>ZX8T|K! zSGG)07PC}Em0?W~Ui5)K(go!eLN;L-qt<^lVVMDHG$C*3DuLJp3mv$c02qz&C|)zP#=1_z^W(~s_msLBDi~2k3e_> z#FQ)gIiyHb|6&JuXY1!8h`@LFtoluzFMQjQ{yd?oLcnN}F+PiQ;`9=U!<-15G}96y z#R}Owzdu}lh`9EU2_nN=;Cl>+9Ufd5$vB{%I3Gx*oPOoO`@|FzEfF&IP*?;+8@ebe zRuBG_qrZA2q^F2h0W&S(CtA3+AlkY8>WpPHwBSVKft!2$T^(y z>d0q=Mw+)L9vZ(D{02F=&nb3j=w#`}NeSkrj;F7YST4U>76>eqP`4=1Y@bWj<9grJ zxSwn*uC{hd^Dy%U%z9y9R7}2LyM0Tvu?Xd^pi-!WZb2}4Y7uT@gpVTAK z%DejX6N+#^-vH&A;;9J~@3kw@8Med&?L0piLarni;`Zd8qodLds%HX9ZQw5%0Nh2uOauc+NKimfN!X~kU1(FTk5cvBHap0N z66aiFaO9jmr8>M@@v`APM`4%)AHpUwpa7MkF_ZSCoS)y}`uJ4GoKr;A+(_3=)s3VA zxFuGr#98d>X99Ww=EYcB7NAq7Kq=JAnb=L_E`FdNr$dxV^&n017@j-V{*iQwDH&n& zs-G#N8{(|1seaQp&(^$Htr>ya2m=fEBB2Sp7T|t$p1?Q{ zto`R7nPQV*-3ou}6aZpZqr)+_IMt#`s;XKSb;!F8_dEkuMSP`jme|+QsTb;e5Yhe{ zSZ4n+2lIA_D|SrzsZwG%y-p^V6|mgX&DAGt2qI=sbSMOMV*Mz{XWPB!cV!7$q&68S zM0|cU?QcWgGJeBu3#aeCjm(XUe-r|lIMo?G!>8skv*mYk5PAds9z=q9gL;N}#fkcG zKc8)p!y-)ye96MEuQ}3+Pa<8ox|ywFm(p{0eRuj9=tQe5PI59>d#`!{svoXB>j_3o zQ6}9CbNSvCm*t>CzW8Nc?8V}%JK8L7z*UuTNul->ofjlep{AmmS~dk+7s%Zo1lEBG z8~mgQ7q6O=QX%P-9g-ckYSRykt_VK>oIl1n+P8%}06wEEtJlHFV5S`gZ?Hu~*FZl} z?|wwk?Q*a-G1_qklB=WX_8CQN)6&a7^!iztf6Gqs(7kQGZZ+!V zb13;ve9~zQbuw}t_-Btq0^m?c@Nw^-Y7FcW#sSC^R7V+CLmks<$96~fEkRM?^ec)N zSbqkf986QU#^E6azntMmuMrO5b(0_96Uwa*1laAxha>EZ2tE6B2Gx9uSrDskxtG<^ zmNr&(4SBoD-Tb(31BC6<#BND|JT837lk>Jy5S|IMTq+~qul($Lm6CfrY{y`d@XBUB zA8#a(!%zt7l>lYIuK^^EfFCG75SV&O7}h%`ke=r~eR!NTq=z}w-oGBU-DTXBqWgBr zDl8CV87So^s-Yv}VpIgnv&>~3N#%=2?pmDMfK+eslZDFi#s$!xFF-rxl6Vq7SjSeg}fZk?4 z{9~YUx$@J&Wci z3yN~5M%GNCM3Eub_%vz~BT$TD!e$psQcSXas=FZDj*>QlK^qmGYS*>X*dF@yp+Uw4 z#2mo&nO$44;K>5^Lo3&g1EGoT1N$`s>>V-(PpEK@w}C?5x*GX|M|0xMCwF2mAod8O z-JVz)l86=2)Uj32ylEbY3wZiJpNZ>#a0Nmq*={#dS9fPaPaU{MxlDwY580r=i#@26 znQPg0T)tVI(Z!EX4Waza!$hciqP6sShGpe87Vh40Cqiz8kp0A7j?0Ij{JBp9Hl^?> z9dC+BP*o`N5;rATrb-f?UN-!5I(2-i(yDcDVb=2wk|8HIUwFEay@H{VxQ^W_Xw^mU z7uc?V`QJ!&5HV5eeiOj+57HfRyv*QZSjLpbU-_tW8&l`_+@`{E8g=J<(+m7^0q&>(=6f@p`hx##~ujI#_u?tIPA zj=7xKp*d1#80&6$U12!GFLMG^z$z(YiK86Nv68#YSY>QBE~DH{Q3K`@oICtlo7#Ka zo6<8b>{QT?(3Z$IoMSL!p+7{qqlBf{6K<%UDWc6ZV*zc9lYZfL(7~cMZN5j$xD)%i&F^X!iu7F?QD@z!Ex~Od!Q1a zmRKu}wMg-*^A{Gf*7nu)&TL_*4ghB_O1;@!(V8!K!T9>_wf61{Kc=p`^57^UJvc4; z>uuSs+?C|NLm5>qLnDpj?6I5}%jVv@L<9zv%Ay#27(I7XAq%R18u?z%nWGeKn_JtRt*rY(N;NRL^KEXV`-A^2HW6Cu!+O z>1;h`JwvyQ8q`$g=;1(shrdb6GO60+ZYHu`({;MDmA=upH)fMs#ZXQ_fSjl3^8!Jx zHZ_vK6ws^|bOF^g%3|+noNdaji1L7V(2(H_>b<08e66+ri-kj2c)b#Sk)_@i6ERR{ zV^=&T4|}7EKvlxV#K=a1ERVBB+)r9N?$Ln9)Z)8lGz=3?k;j&@4xIp0DK_O0rhNLh z8D{HFA=cMetBuY%I9^_4g|4taN^F=BspIAE-~#OaoWOgUZ}4m>oKjuQE1wL9RBGVV zOK76*6Wq);|2jbtg6e?31p@-W=XBDeMf==1Y;FQ8N z5tzo)KfiVaIcU*!h~A(PfcE~H$EP&1U@HEZZKGI516#6m$jbP+U5%~XCsg8z>;)>j z5tCNfx$pwa@pu(GC2gw{JemXVWQkikh0K_xC`ksGKsYV){3buv0u3wseu);p>bDGo zTPUsBDW(|~nXfz$_tVgY6A;ZK$A=B2?`rDnT|nPmgUvC3KJjeeijn?8Ti1fMO~=F{ zs6Dv61%Ivl1bWuOs9&_ozL>>rk;=K6qy1-h0FZ@3txj%k;!n*wopT7zehOdfCUy?TAQaP*s&d#I_t&#UyurWHU6@H=h@@Bi?)fvnP!={f53PjQ)WXZ$}IU}no7^CdI(yb;aq-+i;%4T8}gDYh~~oW7tA2acmd|MA$)mKq9D z{yxAM%2ti37%XIl`y2F%rL~*0At2a^(=CNqdGgh*61h{}d`VvaWxN!S$P#X3#v>`z}{=&Dlp85UGWONLQ4Vbni3&5NXv@V}7&rwh6 z^%P^u5v2-+3iKe70!hSj@lS6^o<@TWO)P(bg&ID9*F{@7{Z?P6yX>&`bY>ya(o6(g zFEN~f65BXmNQlOJQqyi>}cDuR1FYD<-;TPrqYU2gw{RRHyzSpS(UrR=%NIf}Xk9~vKGemndi=Y&%7;=E3KT)5<0=7&3A)Xkal|w*7d!{d<*+CZGMmb=Tr*AHGiPQ zYfANuYs=1KVXb_|VTo*5PE3yf;}d6jXuVP#fGeABueM^|7O?ifXrgkK_WG3;{Lb2P zO?gCkzxhz`oyrwMXMyDprR+**Fa=j6(9hH^Af(}BH>;q5fpg$j&3!CBy|!Ts>qWt| zq{N`BMuGqF)CVHmq&c{Rknss%OArtsEuw9uc@aG|9nc4%Pdf&eARCruxJD18iGWIK zs(DF^m~Nk<__YM|M5?BW>JpY}E0*+=c*0FU9PMRo^06L04B# zO2 zpA9X{*Ohxk|KTZ@GCOvsR-pvY5CSxj$K(f%vHeTfp{f(}iyto$?t>hw4w4Bgb(c<^ z5at63>N$9|`Sw4bD>p-{s1ujsjJd*md+@mYv+8H%- zv9Yn}bSwMd`vLaF__e*yKdq3k_5dsp1n2jtPrPfR{jht5gGE)Kz1a1$!6q0kK(qy% z-IEc}e@U=8bEi$x$U!_3o}RZxclNK_ky(Xx{9hbqkk`!sF@Cn~qBvoiYZXC4Sv>2*m`~}0&k0N+}(EW9RzF0?cY`= zOrBOz%6e%A^Sh22W4tvoG#H6M^mE<6x;!N%vf{-$SH#^;jBdLJe(x5L*pmq1*`#};*eM){^05Hc70jh z>P}6|A4K1dtEB(~;ds^kZtRz}l zL{le1l_)jZxAqd6Js`ut&EpcJQy>NA!|?lDh0O@Q6h{FUHD}_R3}v=6IBdPIZwSSGSrR1`JlEBbfGFGZwpjeukR2QqG4jpKtLXPIAs`;PIj zqUefB^^hqRWx^9eac>FKx+!FHi^fhcuoEO@sU{_^;_AMa(a~FdM~=(?_Y;pJ$69_s z>K-Ky9hnwE;&VW=7O=Hy2A#M`2x#RrI>iugrVmq_0+--_d^zu|p#>>KD{}iA&)^{fFY2nnfdVl+&SGmFZy4%thJqCmiO~BYeo9}G&8b=Yn|oh8@t&?u$%b> zviU@=`30__kC=(yg11E4fQ(J$JSr#5b^?&+g(K4qJ#?nxID0C)@lF7HX5oa1_y-s_ zuf^AZ!)9?&ZhiXDi^~>Kc;`F?_pLsfw|@poP`^JxnArc;Q(s?28z;d3ZdB>w0u(V@ z9{?fg6UPG~%Fly9YPpsXHR?t!)i#eN~^XQ~h;8pbr~ zrO#2nxoAV~K}Kr}HrFwMld-fdU8Z~rh*)Jr=#)nV@(rf!u|+$>&5rS#FWNxs=kG)) zCo3x~cw#-m4NIC6+`$~-3MR`D=5g+!EJI-h_6;fdq1tu9u^;riNNSt?Ad`tztBwK?+RmfEUYK2Cpt9eDp}Lf0=&B^3s`gVsAeV3<+JKk zr&6&-FmOF35(X$t%WVEkfp)$Z0OCbRZdp=O0$l|B)p4IUs!cUJDyoTT*z=#esMCg| z3YK~c87M#YmqAW}(ADd8USx*m6s3-j&nEhVa((npB48o}#7;tBrT;p(M_M+BCePC@ zW@OKhII!|Uj1L`gqLEpub<9_=WsRLj(HPTLtJf@6?7n@0`oo8nG z|Jpcgdc8W7PbL_HaJtk{h^HVSCS^>)Zl?wTgY#Ik<)tl@V`am;hO^(DHQts01o8-T z7&KNa|CpAG9XKI46+m=#HFX&dH&5Ew|D}@@c=yI`T5w>1pIIv;RNuebbEW(vd=Ic= z+s;?PTiy>ux2ORigDN1OGH2>KeTo&j-EywHi3M$dhadFx0q;7y)t1Ed)mQe|cEr}+ zY|x)?DLuD+6>7l7D;WYtY$Uz#BW+aYlMko<6(EZtCE5lL3a0=l&f#bK3q~Dle2`Gd zp6_QTi0pgQY-5r1jZ7u4J8~{1!6zXjk!B;kQL-^3BU0W}2*RR`dtg{zu%07TyE5P| z)W*Aa=72?fQF36SfrA;})>Fggg4PdH{Ui_K5`#@9zFqdwXkXi4!SOI>0I4 z&}}?9Kh7oBC;bN-3Q5J008qOAvcz_u&{}@Gb??rLR{nwqN=|sWEAdH*g$;yLJmnAP z-sZ~EW~K(t^h{Zw0)7a1cR?3&@B|JEl3dX7p~%Tl!gaYr%zg)x z&8AOY2+BR92Nh*4gMMmwX|SX1aSI@!acPS8SQ4k=VOWnv-2?$! z6pU;rDr(11chPo_yJ+~WAj?UZ-#jf@|MKtB(qnTWLpKCU5I%ZNeztrwp0Cbt;-5K< zi-hfX0TJPGBH%VY%XbslRbtX#4K3X=@QW2Ke`&HW)8k2TEP z(jSOf-fP?mI}QPtx$)Q4sU@IdBIOpFc=B zYuQWKJ@MD;y@ejc1)BSJV0=vp8dyygR;sa0Sjao{-ST>jS+Qt4m^J!NC>{AJB-fuP zs!bf4cx1cPJJ3A=?W5%yb;+jFtwsdnRheUMkb+vSp`A4K3ELXp1mjQ__N~or4?~m) zg^*IQr{TAg5FmX6`%UL%FCa& z2#Nh4_cBHGn=e{vTG+=dqGYOv+=XpO=Pb!G};cDIOg3((}RwPe`$$cKa9!o_r@F$|h3Pl3u1tBv%O3Dz!3K`F)d1!k zZWuc(g&r((Z*^p*!6=tu8gEZt-4dgZ{TS~ALMkKnBKLM#{%#rj&|T{>XbsPm4n&^a z2FH5k@@d203zw>^n>k)BT{B-TZ7b1@0pWYy=ESs8tF7)%AE|ALH)^u8Yzn0FzFy65 z!jkiy&+q4A;318xRz3hPSSf)qIf;zD2ZaQ^2bK89`3RaMHdzipe^T)IL^(F!ClTv_ z<~P@x+{O-Jd$Ko4!pac_er>3{JI9T2;rQfl5F!ucB@7pYZ%75w0Igg#uT&%7_ z`WtB(;$Ve<-@0+^rM^t?LE?=cfyt;8H(IFlXwWW3FwVoSKuTT5j=p46w2iQa+{8x0 z=e+7Zug3FLUbZ>b>sI&!;RS&sUAzyf{sL!9?Ep64V+3trFUshVm5K}F(S_6p;W^Z_ zJwwz6E^q#%7%4w{M+>4v%*xhERx8~B(>}9Cy+E{F)sV89!qR%$7T~gwqK2P~TLKHd zL%W89|5Q8F4eN4@BhEG18-7wF2j5p4uKtL_v%6|+8+H4nP!%XTDTXime`xuG4E>#l zgVdqh(}NKAH|mBHr|+2+5L{)Z+|Cw?V?IB`!~e1XC0X)-P}q*uo;r$OT)P@u^Ic>$ z{Q0_9V0k`=_dD|3_M+ezaeAW*5C$MI{|(5YWr->FQXyBqb_=gUIvUX9YT>J^?XGD> zg%cd9yxhUNdc&1d2s!V-GU!9Ia@1cbdoR-@qNX`8G0pK&C@=156UGdu(A)HAz+t$A(wy6;qrO7CzTT| z9)0E<*2p|gbpM)gTHtiRj}-bh&LehrE&?tB8V*(l`C3uot0ioXQi^7V?vcOcKz?b& zT&7EIEKFP8^gH6<5Y>n3=&^Y$a^H8E-&Z;vT%~Hp>>@{=M4jrS=BObzY-~jAGj?Q* zh3WWxxST3`+PUHX;L+A4I?RdW&5vq)JnOCN?m363GyiA6JC07`=7ICfhuBHI$ZiAX zMqHxC1R{XkD{)c;MN7xDeEpBwT0x)iAaIAe_c6g?FXx~7N~EL1-eS9?T~N}fuZT8e6h zLITW1!^gC9-PEBxVU56U!a4;e^*bujyUZ_pH-KUGAI8pcOOq{1fMwft%eHOXwr$(C zZJS-TZQJ;&%eJN;VAh&>g#444XYV)>+4a!Qukfimx3s{4sBS&k1Ek`dGC5l^1b-_4;>M5dvbV?{cvJYj>j zb1+KNfhI_gQ^dOmD_KXkz!_m25uL661r1aOZh6Ld2yz+EK_@Q*U+6S9i5D|g#qz8o ziwqVSK7;`}iwBMk*Mo={Dl&KJ!;bNa*`ZW5nowg(?;tQC#;zkLRo`AKrpPRWnXA8217gcKGD0k`n*&c(-rRW-WeKI?A<}*d8IPpo9T{SKics zE3me$ovmR+MO{rf-vAlc5;o-<9G>0-`qv{#=Aq-Rx^m91)$Q8tJnf#VhoE4*QOn{t z27P~8XyNPR4Ld$awk{PEdxADyb3_vfZZE?8@1l+?Dfq&dVGA6*#ITuj5thnZEv^}G zyP}4=awg*Vpisu4_`t~Uqh7~Ccl{f_`@bHu zuhY}B8b$;&_K|Mp--=2~DjO$bAEz5d3nupB4;}TL|I|?C-$*KFGaqDRQ3WV~v5&!N zn|%Zh=yR%s2Y6T6F?S!Nu{Jg=6qPfAqjOqWp@WBpU#9x z!U3XbUo|D2TpInCi`vfwzy$#3Y9Sv91qg+*59|5y+j0XzVaP+ZvLn4TS^Vz}$~(o3 zkS->I{?%cfa?O8;AHA+j6BC*#`x*XuIi&g$Jx6twhw*L6I zdcj*SNvQy8ML3+9O*5r_uEsrTAX5Y34PPs1<7fr0R7=zTJDIvVMELhv2ZEjsQ1A5i zgBszhIqeUs@6TM=XizJ9WuQ|f5P1qz73#A2i|%ygS+{>%a5y6%Llgb}S$GSj{M3%L zgZYgO3Xpx^L3W|!O65vLD^lW^B>89{z=8YE@BoIztjezD!2mbdOixmUX9s^;=5Af;#)%D@6)hVq%CWjuK1@0!R z$NjS%(DFL|kZuK!^%n7rbI@@gaGjVl{RS);O)5+YOOsXnpp-Bpf7j0ywW?0Hjl#9vg(JF|edezRS4>OoI zD;@DYR5a8ysHxG2p6dGk+jfT&a(A$?xY7DY2$v|v`%Ml8HJSE21e$#8#$0s!9Na>C zQ-0Lg2t@suERt8FszMWjmS(EBX=xK71;-ht*xK|c(`itDT+K0)pN`?6xW?|P_V0d+ z?^6W)lpmq;c15p{uFVtY{=Y4iPC`XuJ{LWl9Up=O#Q zsUhjZDbkM8Bhuui$tVQ+=oe@O1gqtLgAh6Y9}prZ7c0~Mtwt~NhICb3?R$kLI>SX1 zV}S!JJ|hf{4kTO!Spq9<7c70n5N6KW`y@@-OqqPzqktE-Bqn5YO^lL+HWpj7c)d{W z^tthPTl;*;TByI=eZA4y&mPcbv?yqP?f;tkC561)fu5e6Ja00TR$BZAxgnVSE^?@B zP(7;FCncG_{FsR2|ILQ&a+l|0bJXJh`jF!6-6l{l%yV{zG00WzOQS%kKYPtm>$B(U zuT@BjZPy>#kNr7g*f+#tcM;y)MxCa(Pw&xU$`TM`RC`Oe%=YXueTF;F&T49HKDjh+ zC}=ji8S}~kIOr=Xx0o577uSu_M8NrZZ8$zhCY6ZyKU~-}Bw&Az!5%}z5^%ne!Cl78 z4(gBQj{=Op8@)CzH3MwT;zV729B$HNJl9S}EW`J=Q!Ys<=x;{%S9z!$_9b!viJ z%hQH(Y|Ph~uL-;cdl`prtYD;If+lXF*T}y?vcaOMaphEt!wxHkO43NemBRCU(Zcuo zM_1=8D6jqOdFJsz>@8wM_y|*3=4a>hill_ViFfODvATZV-S3v^_IS-c!99ha=`Cs` zWjenn*I?*;`ky_$z z>*YFyQx?#IDu43j^#|L*=0bv92cLe~U$gH9zSj?zTi)N1$0W=<%nJzD{}}BB-vit) zyit94KigaFIrtaYw$Q#ga)(_wY2$}PCon6x{B!6dQ!ur1NYmf&Y(aI%VYGKtiprNvr+j}X_d^pLMLoUhPVd_wPQ&@fB6~cIV-9IJE&p(gU#e`oi@kz4d-VFU&q}hR`i9AGIK3^ zAY=f~^2_;7cg~(o(P!eNxgQYPK-KiF>6;+h(?njyT}R#Alk#>G2gV*r84zv}Hqlsy zkBq|PNg^##SYq;robzG7%()!>Y2?UH^ciIpIe;-T3mYnd(pY}>#9FDu-LhlRF~81K zF{!TtQ8Q&TLaGwec6tZd>v;fI^KO;~mt>+}qBb+5a)}K0^5ngeCr`sC@)PLjVfE$7 z`<3HCYr{0SoP2-GnE}4RTT~Y=T!|h9mWvMe?g3uWx#Y!Efm}KObu9ZkManQanQPD8 zbgX?#px?_FX*|+_q(>k5_xOpxLsWMnL^t^Htc)ba!WJprHf_o5L0J)G;@_!+-WSK6 zoHNU#zIxm~IW@KT)^zihn_V?6-0nO#W!;CfK&U9V1#eS%%Xn4H#?JN1- z9^dU8PV(%kH$MkE)#}W?dLKzkOXR{4aGEs3SuXJSR52sUIT5OSH@(j=*b_RY@n>%i zRedSE^Kp$9%#^JUt=67?zqd~<3Bt;8$fNp4@!KZ?u>hhJvft9XVuGW>J!U@@=!9K@ zRC*X#R2nr=6^oj*s~g9Vddblchu5W{de9BcU3E1~m*S6{Ws0&`2QpS2#7dU(_wLcp z`$irm^^L%XL)~d(EY;m_9jHDqWPQ6;1Od;=n?47NQ|cMDg!Gw}!V4hs*mz!WsLQ z;7ZOFjpD*g28YT@YE*o=2jDGHssp>8{gdjEERaj@1uY^3)ng$ArZq_lh3uFMxK*5` z@hZ6?2t2|#D-F0+>o)i3LxbxFmXgEzEC;6Mth2RaWX!#cIz71|{l(rhk$vQb>n5@V zeIn|Ny`L9!!`z zFwFCZO4x1~BY`SL!xB$niL35mOHojxeaR|ClU|3NCx`+kjX}hlXloNBNaF2Q=s-AY zR8oIZ+Vk60!+j|g)@x^iSMp>mt0D%GXT zfX#vdtZt@^K@zq3NKi?XL%g8V*en0AC=P`m=x*hn4&JQT2pp>T-x1pRaw}Mw42FIJ z?vD8H$@f;g7eS-xZ#JAia+{HpzEhm=Z^>g&HMG)bvR zbSioHobsh9PR4H_3#s|=)Noe_fKKExm*04<24|XQ z@0NMTFMuj35O*DWn~!77@?wszkhp1+^TK9Z@3H){BF~46Tem8!N~J?^Y&bxjS(l|4 z6TR^Ce<2vb_*|(Mp0U%Wd@|A_+G>SL@}G&V;tPnaGzza z?uJuy{M})6uQa#C%I5rFR9#WtBl$8@`VsP7ryyyPx%79V9%m`Be(1{VC^g5E^?uwS z5r=|prhZPr?zJt~ZG1v4WgIlv7U8HQYAan-U2sXdIo^isNm(#iL`HhvpIvO87=>C{ z4I@LkgPJ26!nRQKm?|uINU=GlTO`5^d4er>Fp{4R_?7GQ{eCrwQz9 zdTFbQ+PXFo){Dlk`?Km0@q>~1f|38QR!SF#J98@u><)1mM@%2l%7*qrTM1LwC{`{M;#^Ba$SK9m zY|n7Nz4Z4V*rPEFuer`4*1}DF6D>zgV?isbt6*TT*7KALK)nQo&Cz=4~HP9f2ki60u1S>JDn z(B-S~5|4lt+X7|HiH1R-$0@`?QN~GFH)RPjC+#ox3^NmNV6){j&r_H#FgORzORXOMC^QgRa%bkc4fyiBRHDxFAnMcqs2TtL*Hu$dCi zl0(^rJR6AMd}HgIu-GZ;kzL;R!_3@!k))QQB)tgsdNIy z+@y@U2P3xHoG^GN(ET;PZO5l(sn3!W-I>hXVknGpoV!Uzv5_(Y*B@kYDB~iK{<>Hh zu9)i#mA5_Sp|*bC&U6A#vamk+90WeNXB(if9&Sap32iwtlrpXlNV9M$twz9wV6))8 zh3t#;J^T)aJDLn+(5G-KMgSavtngQcpgvjoQn>(Ksb4K$Uz9m8uTl;3F4)j9+Oex! zu34&IYA^&r!pM`@!V7m^FBd4~um4qq;9{V`I4Blf0<1~+1 z8wc|GsqUHG_GuzB9OL^- z3Yz_^y}Qb@ncjXd}w9C0*$ z#M)5-Ad~ah%nDI4iqH$WYD0|(n~fdIVev26b6TgaTDoCXg#vq|wP{84Lf1yLE63W3 zdN$6_VZE@`c4@7J7;KIUr*W%dmnCk4nnAl(%_YmN2T_R}Ny%NNtSKQbSLC8`VZ>n= zQAr&u4f4wVzFUb@`={^sp$u43Q#;8B+)6T27c)vL-oQYy^dhMSNXX)%$VEEG%06x$ zD%w$C8aOkvAlM!GIjh`kqEZuylJ6i=Q$k-hWIS;jG{WT$ob(OqHi5CJA8OFlBFqBc z4NrCv13kM4tCg#Wdvoc_AKltA)b}-sE8KsLgZLjQ@&D|qY0IlZ5umzl5gHJ)-Q7^p zxrw_5Nq&GXb}Z2pW>Ix$mZUaUIx84PZ!=FKa@0)a>%Rf8uP9vCC35p66*7 zf~EnaWzzbJ1v_pXzH7QcU>@vut*M17Q>HMgX4tO;tu^Hv9Fu+63n3xtQVB@~X1ZV~nB~`m^^A5<3 znaN|1ZtCIUJPIFHR$UhmH1PcFv3F5*^}0jaIEWa zp|CHOnZgdHkf`Jjx1!a`f8Xd%f=9}bfM?nnK3B#z6UaKUH-9<2XESwdQZ#(%Ue4n| zt^zSOA<;sCoY+V~iz8;>Asm5XG;^j48pHM#5FW4ZnehT;_pj9V4HQMN7L0$GyUi*v z9+%&F2eiqWzw@HtH+o|~ALKh!gNLt=hW8(7=?gWsLQ+TRX3IR7HxpJ1i3opVR2>?p zsJtcpI@x}b85fF7fBPbdU{yh^ef^xs1#BJAL1iyhoP)&>M|n=z8ecZp8Xj)f)jZ`L zv}#d;y>W0Gqcb)}2l%U6EBb}@NZr%G+;Mv0pBBa9g~&zXm8vKo;9e*n-B7({uB_k$ zXCp+;0$9YuIR1d$0$6*NdBy~Ui|IsJB41OmE9ak(lPHX>b0>(vf?A?#6Y2~oTX2jj zbs{RghAlAiK&Oh9>(XapjzoU12N+IJ`TBZ18Gc?r&R3$?LT9R933`uM|7yMA*eY${ z^bqtDQjnRtEL8e>f@S+3pu~u~;lkCmCX;>s?bibcXcuV573UepjfHA%Ubdc5HITMK z)z{wTu$Z1g)Q$ly$1(r0HA9=Iv6N>Y8*QNT@ti?>8ELeXEg4(%msKFY2`8{jhb&TB zWCv~YJs?2fk@o7@m>yeqjeFAXh#Y|koD>LWU7i}~a-u0gplzWzZR}r@rv}-}zdj)_ z>WWgDL?G{+5?C-=Lt(Q{J}RnV-Z9O?6X|P`8!(-oGX$}$P*={`UbW}}9uD6;Xch*= z-mCdT;gZEc>;&{L^+EuPorfUAuFF|gR`X5a{d}K!Y*;kzB2u2*ilPi^_!!gbdzh65VvH0A7IAu3e{t4e2*n0ILK}VH^dKvjj9KPgK30-W zCdo3b6S6UJU* zl1h#{X+<>^Xi1W}cM%%J)$v_6`<(1(&=PPZI)Q~-pZblSuAu)QkJzeAMp zgeYzDmKw^ot_Y43Fp_%o6NsFUnwy##8Dq%Be=J9_Aae@Z4TVEWN}NZnRU)2q&Sx^5Vsk^c!^hDS6qIcn_0zDq1x zSi5$KJaaC|;|Vw&pC8$!x)#)I8yX*j zi_^-hb-p}Hf zXvlv~;ck9S2H-4yHKqW2`q3kfuVNkivt64}lG*bX;wA8SUK`6z;cJs0G#?l@Kv23$ zMnoRjZhWVXyXaqcQlp~YWP?fb9&kU0s?&c^#`GC|Rp^ zFd2!)5Ei|;GZn&x!a}g^J<>#A!pc~54qM0yY=mRd(m3tPxdOB1F`ait|E_Lgvm4BX zCEOn5Z#b*^U?leYy)WKotmPG~iXv{#!=(XlMDhOq%%Nh&+vX#e=bo+(?FR1HCx}@} zy26F&n&y#H0!Vz6#OQ3ckSHM-f|6X|LZtA3$qBO3Fn%;aKHq?r({hzsoB_2;Rqmi_ zop>Ycz)d5x!qy!eM0ZujZWEw8Ia7?uZL$~sg7M~{&S>B#@&?9jmKTThCi{HA9Gc`;=nRaAB{ zSQUN)zHfA;iifC5o^Pjop}`6X)Sg6s3`CxspUoY?SdrSa2pkTqiKRUm0oI}6~HGggDV(IO%vtC9c9DpZWN$~Z zf@Jr3-?)HtH)kqTf%&})lk&Xhgug3|EOxIJP8d>xYpxZej6IJ3dHhho617YJ4h>3V zRvfYrwBz;n<-uIFT?6|Fj-ZG+f`!bAGGECr<%T$w5%3*~kL$L|ycBBrTAJ4%34)2H z8Oz;ofx8UqHol|+Jk4-P1%pF1e#f7Gtu2>RC^c+GrQ3t;I3p=Qek`TW0gJK;vUGha z+BD^v>yF?{e!+!(_cwZ9M>BugYMtJpc1OZ>!}aoUbrXH5xWK-#)eSMdNVhfp`cXSX z@~@L@z^e#S1j_tnzf}zrHip|GK{>eCvumT@f z|AS_Jjk-mp{WX%HTRXIKD!xN#Zg4aLNH=Q?>@BPs0v{3Kcfbe%y2~N5A!~ zlXCVK7tcVirut~V+HMNnp%CA~%0%)!#{;4pKyEEFc2tBzkR1|YTimK- zlP*pvQ^Zo4=x!cnQ7@D;0?v4ra&sl=i-LyAj5`n?eQl!Yh5JhS(-F%0Q6D^l<7l`KYRFwW6p0HO1ZAYj#J( zS;I%eDdcRgOP{y*V@-5sXqHyilz7No;0y}ZkrlYxFY@Iz3~<-l4P&%`eB6rt!O2?x z%sRLJa%+SER14H2=n$_SMzp*LI3v?wE#kY2n3M9Q}+jYQ<)4k`<{}fY&9wxqD;c4x~goBv2hELH9*)B&4v;xVVB~*$?Lc@61@-D z*(5EWZ|CQnZ;zRYzJ1ma;PM4rj4TeH;c-X`)L8Ktm^*F=hh=_l?~MV;<}-JvXzv5m;$!o5T;NxfJf?Kq>-%Xdn)<8$CEIVbQ7)x8g7 z5;0vj3ss~k`SJAUB>z?6^(Sm_4~GxiYFh|h=+M4;4kI>%)KR0GM$f|}MQM;iQOt3S zq&wxXzCO1L&6bo)TT2qz3I^Bts>u4DGqP%^%iFBP8pjjyYHMx5*n?YdK*V`W0*y#3 z%UWucyBg-qQlplSsXvhgKgnU$@8O#-&r3QPv>|ssOyf|~zW*G|pd{Tq;=vIb|M}Cz zDNRFE;q#k?_YhIEx#jJ@FVKV>6P!xAH*&+|X|I|JbXwe2$CY9w?>HgY)Y90E|Cq?C z(pfRn2W}rTZ=*gNf|KnWDPf5Wh<{!4!UhiS1UenbX~}Xo|6pXn@Z?y@NS>yVP-w+S z?p3y5z=ESJn`cWQqjWp!?R2jlP?pqM$wQ|H4vl&p z^GN5;qR4a#A~f$c+Ktv}5laateX5%uNS$_^`C=N5Aqz4ZqN6Y%a{(NXNu&xUT|<%&8-RcCaVyQL8cF{j4TJLnUC!{ zjnf1T-vcMB%HF<2o;q#jBocai@gca&Q5DkX|u9|p5D z3`8|zUY))|?nowxtpG!h<_cphmX2U)@HG5YLT40#`~8AG!Vv z+PG{L_4^@sKioP(^i1ZB6|}&>LBhR^kL=eZ%5}m&)(_5cRN(tI2LZl(`JL}~YkCJ6 z{N@pwle~SQQrZh|Q=g_>n6n>0UCmjYU9LG0{2b4RKJI}}0OWmtT(?%RZ*^5jps@md zI_Qc0O5R7RT+eke%?Qo^Ip;eJwI8vEL-^v1ACm)4xMTNlUy+x}v-%iE+29Lid0%3a zRt}fkUIM3sPLOv=qwlsq5lyo#OCBF+YVrllKTwd13T1HaRor* z#Gmd-+MJ=&la_YccDGax_4!<0Yv=Qnm`45HFEbQ{@GUb44E078Rj;(3!ns+ITPDy~ z`LKCeQaj2d24L(dIhQi~LX>U0S{XoerZS^1t*KF5@B6+!`y=LCir$OYnKzC=ImhNW z8og5FniLvplQcc#7g_Dgrt;48E*1z!BSabuVSxTH&s$P72`n)(GG zIE<38T2lz8H0~;K6m^PoSqol#kBZ=~!k+LxmB1V!bjJPT!}wz}{GhXreOW7UQ@tD z_oD~;UtI2tF=_Q9=2l_)Xq*u3k8igMj=W7*Z;^&V{!(acg1Z2)FgM<#O8&@S8S1l{ z-6r?3%bX9;l@z}vS%(cTRYlYkk{|+TFkz-FnbV?z??$K>pgSg@?!4f?(bN1H>v3pG zqUR|&vjfKEHyWFVu5)3x34ZyoWLdw{dFI|hh~ddYf}`=OyxIVKSCQ97Gfv+BTylXuiULC->L=G&y zCMA@Qc>4x=*fx_$WPv_8ofWPJFSJl@;uUJ@m#Tu`srqt&euGO zr7Kv8D~g~hhmsQKaS9Ji49bYW51N7aA2EEhiR_7?9;i=&t%%NstsnhbP5R#S{~O!h zlR3K*2zaE;b)@GIF;mse(zrqbe&|e-ZU4hSVA4p?fZzA#pXtt~(;k9HI)geKHKPb9 zv&b}>Ap~?5WNI8_HdmndAS%;m?8a#}=o_lo>NP{nh$79LK<)=80A)%DU%~kP=f2-9 z*TpQvG|x18)iBc7e3G)w&+kXX=DbP#Q_~EH=HJ$E=PT?fiu~TwDP0Z1xPBh>fFyUUh98G>r9OaBg4q(Yv1OoNG{+X*cIkUPE2Ip}W*q=7RXpe(73BB* zcy_!m*V~=yrzbL<_jkIoKg5-!hF8qoM#j`{%Jk*>-no+8_gw5;@Z51uSgr&Y45$19S^)2)){s|pbLu0j1@;;Cobz0B zE(hno=Q@)GXmaY#<>VGP(>r?N;+!GGhq8nXpP9Ps9qvp{9`=k+9`%e7adMu4RRlMO zxPYE9pNY?5&qd^_~vg=c{qHW0D^VY_nn{DI@^EV*?+HJcm3U+3ECej_;@$* zzIa~bk!9bRKTod$Iv*Y9Dm+$qpHJtvqO2eMANI>{3fg{ew$2gd10M4SeP&Z2%`(I_9VIAG5HJ27j}6?~n2r_S1iZ|?dGC8bA2Z)Y9)Ily zhhgPn2(*|#=ds-qoctfoF8lb#{&-l8DsJZOQD)^L6zIIZ&&a5Bh@9hZj2z!+;^_9i z$SZ1Cqv$xnmxBx#TMl(9w?;f5?d&2%%qAU=hhLKzVjbuz$pl+Z-hcd6lr+*vB$k!b zF(HiAZwP8Oz4KzrHRF~tzyRW6tc0f~PtkJcXpZiT*V;=oXm)D-Gf7+#gvQYUXJ0Un z;0rdw*f0v&1N7Rt+qcFB!zjw#bcGzSfy#Z);2SZ1_yA1dsBQIhE;441uM4_=Yq(bs zif=nd8qc29$B>mo&zA?iy^PH}UH@;O$cRZ2h3_|P#dAN=5=f^XE zicnO%NXi~P875mJCUG7}zfF#KK2(U~$&EJ~VMHh3e?Jn@57q&X1E;!1Ta`tpEqnn+ z4O}_i{+xE}@9??~*%PX9wopacJ11-zgOjfhq1r+P1yab-#Q%=^sE8Q$MpWOKSy zr?xE}8D(aP+#cDGjOQ-(kG#u~WC&|`oMUGA{LeVe-HM&ky9{0R*gIpf*@gq&9eL-= zJdhq1f20HLx{=%27vCbX<)f7yZ(ueP3@VU%ZN>QSTB)eSe+lgP$oOX}^eLDfPCn_o z$1Xg2-Q(e7?s4!5@i%n#ZIt({O2w>{v(si2`cxj=z}W|7fYid(P1MhK$Qhu!nd(`?_47a} zh7Ei{DspLeew~9?w%pk3@GMT`cssy&_&UttC-MF*ZV#Nuo2w=vvLnZt>ec`4(UH!;p&*D(`hiKXS&Vu5bN?JO6)dm70QRpiy#x`2!|{VWAy2*@*n#f581Q~`<|qSZ(s@Ri z5#4sfG)Y~$I*_-usiO4>!Sv%d`wggDFwQtU>*-x{5b~O+TuF7E*VR~i8tIh~D=yw? zmNu*Hr@E$)rI5duj0F$bz3L28EJ(MF<5!+&j~jXT2ZOy35k}kWFt#FDj&$IvSg(ay zO<68a1BOK{9ZWLU5>f{P|7sR@QXueqFa~zxDtroVymlPha+Q`+kWroQ^S zEs)A5ad;4f`JitE$3R<1Di;{fz-{?YCvI5RUPTAlqdD4xvTyf!`;Nd{TbqBdKgB+; z)8Sjr_UjeD`w~IPyMF`8UpCG-?mBQwDFC;u#X5+pJa?}Y-G8xNIFce5iO@JJQ{W7A z1+9O@LiQ?`aW*y#=#L`18~AtffC;`hAFV|`=6MhAxvpt|=OQ!hVj=v_0!B^-EDrS? z|BL>~>n+_i^8n5mwJU_DjHqLS&{|};2?!h<&~EIjbN1ME-=)L?$8(KfNektoAcS@h zh`6=fMBj{4He7n?E=bNw6%@DTGiTT2@1C`I#SLcjTX z0l6bi+-;z1%BG_PcB~t}u>T`oavb>%Aw1I|L(5=Iy_BA)!1!yOta+hYZJVbL&Z@4P zP=unEoK9P>dREpG$;gY}O#!(VFFkVf3~GP?dr9iS;h3Yfkz^(a+KYB7?%v$BP>U-*a?!C%w_(wYseBDPjzoVJ&h|u_xdwfi}q)MbM z#1`Aa*n!Zp%Nwt2SuTU6==>svH)PZn2m<#oroeJk8&w_JN^C{+{O0F|tB>4(r$(uI zZLN_qka*8jsTxEb5n2KX4PgV~%W=*>6J7(P19xC14avklYSIe($S{{HP6=o?7F{8O z1tdS)>(py(SW{6?+72bqVE`^t2I;eFvo+A8{PQiJU1gmM=--4Dq1nUA^e(UROnQ1*G{*ov|B<4(TzQ)BL`hL2GHg#N_EzD zz~TUfJOK@)V72F;P85P8n(PQc#&OObSBU7jSPODQanc+`a}(cDSb-P3@Q4f~bjQi3 z^z^7!lGA2dS2mX6qJ;=M#W~JR(Fs5%1gkBsHmHZeGr54l1zp$W<^79f%r;lfP)n-~ zp4_F?tj`M3bPT?EL!R#(UcADO(Z>(y&;#$p!|Nn>Qu@)VqY<+ExhsezLMC_&_SxeI z_aYvivW~&_V(_5zqwmanh}+($f<`15TIeqe&@#dZfm+Qq+3cobjs%sB;yd3s-?(2p zKivbzJw%OlT6}v)gdg}@GHcCp?V}s*G@GZ)rk1e(Z6$pbF)T!`ILy=JwNi^@15qKF zH1*Rcq( z>eR5O(zVwL$|M+Br_Ej#^|~W^!E{+Bj1oX>78vq{0FuS)uq37O)=!`}LHV;c5Mi(F zw@h3twwrvtE99+HikOiYDpUA__jGJgo4sQH5o|!BHA57j^nr^bTD{gA#m-oLPzg7#NPT&T9j-(-j%-yU zXHil@bLUIPRx{akke5QCdDhH4$oTLi3PTC_JAc19>}Cm#qc1kfqW#94TnS5k+F|44 z8P*8D$aZVqANu>A1}wJlOlnH`IPX zWMr-Lq~x40wd{#ILMGAjW)9Z~1cdPB?;-3CEkThI(bO3lM|WDb0mKQyT-QZIN{Qc0 zYpV+AQVZG>k~V9cb3A{L z9Ov`vud87YiT+TU3I=n$=N#c3&Yj$!R>40yPiL$kiE@V?VkG}(eOVhnSUTsc(Kor#bZVUAub5zj0?;lUcKVSBygKB zNq|K;sFQaj$_jSeFmuJW!6=HW1=|r8>_Irz4Hk+a{U!2goWp0|a1NQNUC6+my0npw zPLEm-b~6-ld9Sa|%kDKKu&mh5n^gB4SDFY(p+|7TmA$fvavA4Sn39eDC3a%R4?uQ$ zn;(poYoN_zashUUfj*~{!VvKtlMR~37r@))mww~*SGBh7SQl;s!aonkA;oqS`M}RF z($+Am+#91k$VoB;llFB{*mF9v+a8>rjkuDJfaP>bm@`ULjFS@=&edqO@5OImVQXja z*3NZ$AZ=L!Ce;AhG%&)?nT|U`5oJ#HMEUoQ-mEEb5K)UNb<;gpOT(eQj1`4WB~=E1d@m>R`Hn?YTt&~IYGqTPqPOzw{_DEEqFYTpP^lWjT@qyq3-JoPo@*@ zkFsh3kh6R2inDuIZHb?WHp=@%)SB)(Q-%$tH{mZMx-cew#rD!m?UMDAwKPC3V1rYCRKh#9b&i__!dZKHAqmVYd9FimQ~-X%aigfx(yt>;#?_u zm==N)R&HnfxNZ-ht%owpN@fSaB*Sq<9V;(6o8+9js-P!5mNAUush6*^haR2?pDOkI!Hg~n0_Bmi(;epE^7^eszIfm|tV|qKXhilW&kEW`J zsYO*&Dk%$RoKPl0Gsis&Y#S*Zl9S>mqJev2u0ea#x`W$z$4)$9Sup;NIru6M%36?1 zIpL`TSu1Df(O$cK$H(pTUjG$8G_fdmoG%Y>PsE($p~M5u7`VHfT-Qzz(}&G>CB=nC zt427!K`D|XZ9x&lU-4cQ%94pB)~RS%UYF4qWqg1eh}$4fFbBE<^Xt=Bn5XAVaFFmG z1dNjWms;|r~yl`(0s(LnVpqa{=x-8^jaRs0eAfcZ8Hu{8fM zQKA&3P|w1)PP7@|8VwB$G{D+CDD1$)MlwsQkjerIn9o2BUzE@n65OvC3&(@xe`8-pq(+dLB9uanPturB;5RW)? zTd>ykCw0;=h=#V+09VotOgkn|~cKeDhdq2A63eGG0){79(|8_5fLh zpi+y@hf|?R(rdK&QfC-gw)w+uczaJ?Y_hocHa_C7)vA*CRc zfv`u$Y(Rj*OOBz|j9dT?Oz_I#!m6DtTPTC9{)s3-hpM4ok59YzcUmSx;EF;dBtLKO z8uL&H%*>skrqY$jDla~)DNI(lx5{!?JJD{i%_`z^LNWc;^Ud%P*+8Na&-Kvf@zlu_nz z$d*aw+>#qz*wZPnvn}P>9Q(=Mg0=v)gFLZ6z&o+Sb^JHRp!e}*V3BF*+*N_qk{cwM8r6U-=|E=S zX%z=4l6$00+{XmAnR0~I$_!)ksfB%FNCT_m`v#~7fj%PTKsggUzRK9H&eV=Q^`ETHMtHM&6j&$hIfyeD-17f0c8*P&Xu+E8vTfV8?dq~^ zyUVt1+xAT$D z=Ei~?nY3{E-XByN&7W;|z8o>|$6#JUiCX z00#~?J}seoNvxwWuoEbVvKulih&n{;4tYpUu7$wQJp^oEP!%hGw7QN*1Eii`Qm$bty8`b6VVyF=U z3tBa4ai}P)Vikp$urIf%)4bi@1pL`m$Ae=~Nw$JE89Sj^pT!FjKkWt-D!26C^D*KR zOB5p*>1;avuioS3-kXf6+%=|v)4G#86>Zk6Wo>lk8t-SdYxZX~S)ml?DkybB@F*lE zWX87msnMJ@Ld@dUqLQ&+gvptgXrJnmvus`XnFS<@P3H+oh9e%~SVZP6_A5q~L(s>u zs;MFo8SveT(qW{QC^U%t5O_jLmixg`?@ss-UiK$Kc3%I+jeUF$fLhtOVdHUttqMLO z2os%|lTZ~48`RD+h@jk} zGdRRl_gfPWL0|@kACI>Yj}Xt*Slzyyk3tnoIh+}>{bJ(5`yc)7w~HR8+55IdOEA`V zF7Gigd!qKV*dte4!Z!W%b$oOuf~LS#&(iOq%ZhxKD@dTOPIav}rxW}O4^f(oq}~1S z&RTkj(9{#3>Vdp8km%dw9qogEe%1X+PvMT_p$3-C9tsX|5?s*v>lx5V=Wui#e; z7y!&;f`YZ9r4QYV1?+pEhO(p%NfpM?E$WT8)97a%S$M5}NN=>>KZK6fP6#|G^(ejo zzX%W8nFzAhw~^Pw+!39BUQI<@4kIx30>SZ2g|Fws;_bNM4(;;uD_FM{1nVnG%({JB zAJvhJ5z)j^XLUZK0r2jXJck!Pd0#cY4o+FBmMw*3gjhdr*JdeVmMZ4fWWZ7Bwyjnw z=v_;OVn-We8qKwA^Ucr!zHm(q>Xw6EliIt?-RLeXFBa$I+vpAAYxpny1`>?q**c^t z+FJ6-s>jS0#ogDCv~WWq=A6MBxt!Tx{j1Z6Q>|R>#tKEXsu~bsLf0sZtyHvs5DH=k zOfDLtOPXZBgng^}tZ<0`u0S~xa5olP2fZlkUya`+_^+La_4#(6b zvojn8Sfbi3g?^Lr;B4*B&q#OuuX(FRs!LgI7S*-O%WKNhs6ZdU{q0rpa53)?JU_Rgiby>Y{y8QzF{_%zn zXb&4@qid+;PS|z|9RX=xEXrwkhVl$l|`W^WmT1V zAw-qiTV!?`k`N~!3}w)|9-nuAmhIU9&@&`f86E(5pxO z@0~f(KwnNyXo$Rp#F5A;Z9DiO-`wVAx9F{ZuKx!hZ3EZc(Ua}T%LAoD@%tI<2FYcw zM_96=hwN>urG?|V%}bs!bL~DqJ|BB7Q~O8txY3}{ZFJj8b9deF3wTDq}MSd$G1X0Cu-AtD#QvrW9Dc^5Azz&E9LyRF>bq{qYO zZsU;g-i2rGU}dh748d;}*NB$v3Ejx$#N`HtPrRXB?3oCtndWRFo5~Dmnr|u?^p*&Y zZJCEuitQru&e-Q;PTSW5X@(-N!=>m&AljtXtY*lm*|K0$l4RP$u3O$b4p8pb7F}Cj zB3)oV!;w1YzUMyY))i6iF2P?luOTSMC*obhslC1pdyzG%MFUF3mjpt@OsOR9pQ!m` zzwk*+r9ZJLE-qng+%P{z-@|iZ2ma5y&DVnN9>4l1f)1Iw5tYC60?^ZLo=up z_*pP)1UQnBF(!QTt0n>AZ%bQk+K+ys3_P<*hw`{9tAmcMu^>BjW^Bc-VFdZT8vIM( zSMPKG2wnw*2~`;y!fr58x;D}^98{n5vi9w@r^J74F3Y!B&&Cxqu~9=zMYPq++=6e| zM>j=E9@Wmp(f*tIhi_p74yeYSdsAgzJx%bZ3g@ zNLjNP+gu?A_O_XMn@_V#m%3{VUv06$FlU^&Ue-W2vZ3CA7_cD&afq3Vg5Z)eGm%B# z;|s^Lugk{+1sx^*X*ON1*<50`M7`Vj)$?pM8WK8+77j}7qn1uh>`ZxB_$@AHFs^ab zU*c(XR127K^>m@-kqd|ue;c^D1#Ad-c^$q<>tgZ)9Lq?O(fI=lJMbFL|3YS&g9E*f zKMD2ev3K9wJWtyu+;n|n%hN0 z4EScg5yxVns{*Ibk4l|RmHf7_2yL2Uz@OS|(AJM-&4iVU)8jycGzj8eL%D?*pmzM0 zGBEh&7hhK*@C@yn1Y)BjXRqcXWRSrs^+V)$HkP+`m{gXP#&yyrx77Kv9GX2gvTM@z zA-=59rd=5){%S5)8I7h52|5&@VJU_P*{KnyuaCE$&68X>cxa$pxV@BZi4sf|h{zR_ zHBpFGw|!pV@nO~98J)Bpvo~6p^s!y_DDj}%w@^MM?-Wwxnd_~hvmOU@+32Q*WjyT^ zGW@w2B^>oRUfgHknFZg8(#H&)ee4!V*rvb@M*#sZuf)OvioCg-nhDFi}f9Mb6D zfyP`7_(|vRfVOf>@pg;W!5=7x?9s-GVf;)M8bVNDQt#g@WH`NDpl(mOQ&dHRMr>5a zunm2LhANXS#muPQSs&X=1(nNb##$_DxK;c&m9*uN?5?7DJ}&#>rwxt8s4CU+;HBQE z1zD7l?5dXfUkc9B{lbGeO06j}$=46B89v$E{tT%*X%;z!ND#%DKG(>nJ=7e6m|npz zo{gdz_}ju^Jv>sz%M;cJ!ThORQGiaY?VdwmUjX){Ny0I9+)M|sr--2f_H2@Tj+$cm zdP5QP0ub_VYj}7*U*3ycDUM_IeRD*#HKhWig@H+%O}f;{DJ;)5))Yrv9dJTar4`2) z5>1EyW>`iC+Ao~}$l+*M&}ly>n>VQ3Kkw(xr<|cnJtWlK><}Skv56EnNs%K$eh-n@ z#T0PFzmBck0QI&#Ywsv|dPWj!uAY%mTM78@Ob9uiz zc#G0ETP<}yi$8y@PNUSLSv;_ETZ_!@wq~5_Mj$u%cK*AV8=b!4A^}IjGM(`5w;Cdp zHl}3V_(#Yq6Z%U25;XXRzdvuTx31OLj@f_tjMdsm^X!b_IA8y|5c)Xx{?p}zvy(l* z=jYjl+SlI)%Z;n+Ln)-}c#>w!Z6V&deiJ3GTj8>M9a}MTUb%dkKLp3EGaPV7i4fub1P} z)hmg!9WdU2!ST2p4tjboX*t9Al5-4IyD+y?`U=`wLo1B6u!E@2;k@JWzan}6IK5pV z=o))&2=uWvrXszI`EO0^GDSNu^2L+^F5o;5znk8qWpjWj9deXUojP&hDdy46qg@N@ zu38L+a2^6h*_@Vmd_-5?_TAlP)cLZ5>Z}R7fi65`o+u=Z8 z^OubPdY_d49xhCdZ0x-8a>Y_UkCr7orUpjHWc1Kjbh{Nhk!{#ej$fi!o@2g=cq1Nzgw zRRygD91>z=85E>NaAtYeP>&& z>vsV42GjdXcc;HqyhC99^;yw<@sd3p()-fO?bCs6M8+pv^Uc|%-uu+cYwy_$xr0As z-h^F4>Myz=ttrR6Cy%4TTPmHr9bua+vlFMVXVVtJH!?qT9mHFI&O1yx4hH}CnWLlA z^2o^77oB#bI#sF$6%0(nmEf5Fm#qV4jecCa?iT%M`=?3H_QX!Md|wZPmY&4XArSpm zmCG4fM@)>Pu!b%jHE2|SCCk4(>iaC_$S zcrIan;Y%^KEmJ3MyjYaa_XrJZ=KW4J@u|N?l(%xf=zRD8GRx=|(=H~QO}(Q6j5umY ziFDgmw{skkuH%i})_LvXx1(JM5W>2(;&PX?0OC1fmTC$U_a|B7Cjy8&1iXR}Iz)&A zB$e@vTG}+MNB~$nwTvv<%B$)5fiv)-XG~8vMs0{s2yjm~Zx9cH{?m_rl(DnLv9l)a z(O@u|)^r*;K)cKhtr^qf9~Rl#rE(lLts>wAEMOy^%~I1(Mqfl*D6XO#bZy){AsblO zaaXE4f-5Cio5{(2eta?}X<1}f@->A}+^HdY(#^+Vz?cl;Ih+P0;SHpx^4CyRL{Wsg zWgJWcXJgzL=zmb1l8bnVO_sB^hlk<3)w-3xfWM_PdvDmwy6zFznrYnJPB}wonTyxx z%ggokI)2(Hnj?r`i^MIv1pqrKcMj*#`hn)rs!`+<_Nq1y}m3gQpG(JNpeyv=2d zI%8VRNseU2z_Om5P}R|aMKBG7!%(Y6k%fO+uEl6tSFoTUo`FoA5q!BLB(!6}X0okg zhl zW~8J~s%`(-LhE$3Uph3!;JL%b`$3#vLy4f4zVuBpQTh zz7=7t&fd+?N|&QgNNwU0dIhu?8ryr8v}1zW7=dq;tqm*fUU44cpw8I1KaJrE1!ELn z*F8_}&oSEHRa<6d)R3>Hd0XdDuTNtnqt%`M$33<#BGsrTo8# zwI~`mCHe5DzC^E7n;n@vpCgf8>U0`<8NCSxb{Rd{88GQ#_ZU2-gHm^~414l^S%lei zv1)UHy@2~5toM|M4=dpkMQS)`T=cZNaz~%`T4Cm@_zY))J`Si?TtS)Rk8(J@*OU8| zlUPBFJ_%(pZH)m|eG(r+x!b@+J+4{m9F2~&SI);il&NMNvs%Scdr)Q_WW`~(cGoLt z#IU{6sKYEzr|lny;BEJXMezvDk`)w}lHvBOKNADMw|5*(NbcxMXsYcZ1?YV*8oZub zyXADLXOL0_#R^&$3Y5d>WOmaFo^<#F5n#485Shx7xYui5?r&@u6BkT}U4vYPi2p*f zR}m{39xrOfqZttTbtyoMGaNwm_WATa9@kI2{@JcX%eUzI3k0ypAoUY(E!;fO>~(4lVM(4;`5Q>8W@4MOsKZZg>8e=J?2N zo;)`S7zEM67RkgzGZwCg4nTrN|55u;%QfuZfU~BVu08;f+_(&&t$js9ygEw8e~Bp3e4 zz$kcqo_gYTPhNd9;13e$QKGC{T?Pg!e`0kZwWa$<3JgNtyHS$=t|~~BPwk+rc2f{q z!p8Rp4i8Anh?c6!;yx9*M(jz6pNj9Wxs^=v(CTPO<#> zoFHX~!#<(DK}{3t4$#EX=jHt7`0pf z!T@8YHaJDKg)*1po1reUyb@@oGe`UP<;4{iajQR>mYlH7oGL11B4fB~xQczXp$>6` z^g0;l($(3yP}L@hZS~i|0#Z-lXD0)Le(-We*B#49xvlqely(Q|jjk@WvId0htdvYD z@YC1cUfr*fiJG{*7pK?KY#eV{#Rso^;8OdD&WBIe2lUm^%eXnA?IqA!+f>S(6e>dn z{Og>@;qPt)e_NZt@rx=L_B&8CYioXRHECI!%$m;Hf1ox$Mbh&U5kWyyjm5Np5|F6hT;YwM9>?M~ zaQ=5cCeJ6A_Z|G-$x$uMp88(+ndCa;%$3eSQ|fA33UF5e^)l{uZ}+SacfGqL-Fw_~ zxXF{irp#LR9Y}7p(wIN2WF~;fRybFI^hkWM(K2HDj7`TQI0`s=4bG$!=-skc$$Q250B z%*RoYGf!&LqoK8hfd60LxdV5Ca&Z%_-Gm)5w1W0>@zkaYBSv+fwa^ZbQCLk40c_?b zsV|cPaEW5gU_$JB-tR-)Y;7Ol5a!|e_vH=R%YAsr^6=IL;Q)=M9%_I`P*-d>Y4 zI5xD)ULfZCWMH=8p*m{KJbpw?*A(lK-JfpOhuPO%BXbghflHnSw64-Rcn&R%+<~e2 zPssFT&t^a5?bxEjjm9&!8 zzLU0jAVniaG-?-LSDkHshW{%GS(w5(zXPiG= zdzYSL9>;i1C?PmlG?znii~94IY9>z_K_16>@V)2*b?%$dO&IxtI62@P(ra%!EFfG= zdnxX!K$Z4*D#h=Sg-_;)9<`rUg4@myQC4Ja?^{rQRvmho>nuuTPNlCg< zFWk8UDt0I-4e^h%Otp}hYi#tnuC}%>ga!2kKP~d6)8Hf+IUIOw@Odd)Eg@M-fvu{r z?*n0-07aOsaRaJXZ_u9*5Nk+A0DzP|5ju-d6nwcVVHTgXi4cRlfoG0b)w2S>qY>x+ z1}D>P7>E&gRK#_1#Wy;`KX{gLI;Ey56=Tsflgj9EM43gjs?#HqmN%S7Ll2R{oYOG+ zUUUjqK|SD|W$-4vf9g#&SYextNpo`ssCGsx_tVf#1Jn|+){}z$Km#E$8``g-u5HNP z+-DJigg5M*&cuW&3c;l+?Wn3LOYz|?Z5ujtN1AK4P7QvVM4MGbaG-|+*(5!DKM8Z= zDZUil^_GUPcEL|+Q|3>0W{Dsm=S7`MogyjCVL+9n_ebbL9 zNlsKyS~vb0v3#%`u8wNY1*xVB$?lpFUp0XE}KX ztrtlo(DoPi=?F`yn5QnbU+s$+OMO=w9skW7HfvC{K{~wSa6qd3u9*T3F-2C6T_>Jl z#QXQs=E=*~AJiby3q&<%XSGNPP=Z6;I|wJoj`<|Mxs{N*8n=&9$d1#kIbzxnpvz^A zi-E_nwEZJ?`l^9bPpYF>{_*0`V|y_3h25Fj2C-;agB>$_V5;l1ZH(LYu(Vh=!5iZq8NU`nc%HV6<(GMH5sx#02|!RYdJ~7027YtnB_M`9H;7u_Xq< zB%H1j|3dkH&6<-Z!iy?@83K)Ki=j4y06luNHFX1t%DN|Xd-V#T_Awnq^n)t;GjY(= zrBRs*Kaq#6yq*n77E_%hL8F9gO1)YE?BupaE(+6e4Fr>s1*Hd;<;Rz%Kyk#(1*D6M z?m9GXAPXn)s(-vUW#fW(QEj_LoLqUr1UmiF8LZL+OhC@o?%NNS7%t|`)x!-iXz4v0 z?zYBv5YCfz{SUX2Nu&fxmu!l>$Ftot!v;)A@}g@YMq8#0=(u8w?)9Lh8eJls4{neS z0x3xUV2Ea@@WQ1`^}CoLhy%j4e^c8p?TZ|oYdZcgTnUfvj~`5w%vU?+^pJbytqP(C zkUI6!ZWuJ5*^?D%vJNQe*P$16t882(-R(^njPPRwq0?{ITtNaV;$chn8dKZ_wCW>FoseA>%R;mXa3n>^)3}q2MoohhDE6Iqa5n+mz~!zT zx#wODUHm4ny<3aIJ>|kD^^}fzIm_}gqx>D(?0?rvQzm)!ZN-FnaFh=Sb`>1^Fc3mK zp1Z+PhY6=klGqm<_(^cRp3V011nX?L+^B0fd^o%WylMU!-V$5$GDRtsAKK+T9$x!B zzp%@;!tD{u(tjs`=vjsiiy0>I%Zz8oQeEORV!n(3e^^RjG7G-v+QvN!A2HMVM3q*2h0IW8^)2hR0gRP!)>byKfn^V?cYY)^|c40J_DTa zBtd^cPFIQw8_4(!#)=92lyw=?MxF|amZlUR=nR$XnH9TlRb8@*r>CCZ-=5!3718Nz zAv83!Y5MJYC*$#p#KeIpG>_-#FXLgtv|otVd~ftAnotiG2m&uqx`8V61F$&ndrG#V za*J_$@jm=6e{M`lT;>wcy`xx7>3v24%(N0vc@8OLRa@oJuSvPYhu@fY!qpnL= z-Y=u5fWQrD+oBfbk|jAS3_zG?fM?`}M7Ly38M^{bg?YK`lC-oepD@)p92MAeSmCM} zuO@FVC%e=7R8$RbcW+*ILFM5~Fq5*HJ}+JLBc-=$Ln2s8 z5sD#B(myft(LYa?vz9KbxY7TnJRl#1!ooMs+2=mrvg7X(nfJ@4-lRx;BL#-@p0be- zdZN)U3am;>yF`5(MSUZAvnI3PZ|jMhS_cudhR*PW)H#- z`bjYq-n@(>B)>igR013bF~OiniT;GoQr+`=3sCIl0rs-6KL1jAt&~jHjnhw9@9GXj zA1c2Iq}&u!7@ji~XLYdFqdUuIAwYJmLfM%_xr>N+b7Ps1lqGV=i!cN~6(X&p<-1*O z<1tKrQ`Zd|LxfqJH4GC$^xH5xArYvS+vLaA&n7SHR8WzD1bOvrm}Tn*9RR7J8Ko zZn{?2@y!h&6Ut*k#(y<+5Ldq;BJ5Dj%#PVBPE5)&(=mX*Gd(?yBlG6P%(;8Ol2n6N zi?`DFQOuePB&bvB&a#b*UAi@&U&_GwzpP|;_PbC862$Cf5a^AQhITgje?5_%DsWwa zaNGsr@O_DSsKX0nOgF6;e$8(Vznys;8#oo70~1bt4x#WJxZ7PK|q1XUB z&bc&&rExu6*kk5|20flq#jMoPeN?4-wRY8-KgklPC0DYx= z7g=sp*Mo%co>n*5!hjPUmV~cT*}sR;-vL1_F5BY!EWp;;L)HN`oZrk*Nn`(|6;jX8 zFSx45k(mqKNXqds?j8(EW2c~gO=DMQZ5nBc2|5PVzLbYz=jeagb!X1)y6?RseT=7Y z%_FiSK}I@9VoHa@>kYnCy}w+16@V;Y+=h@+uj_V5Xj&b2$hw<#;O&I`eFE3>HHIo0 zee^BCy)$263|AUyt1T^R*?*S@sNeW_UvP-n)KkQhZg^FBI@-&dzM8du!;wfw9% z90iv&WG#0_n^{Pdrc>p}!kXFJaQ$3Nu$2f@$gRd=t;1yQlI`0gIsKGiRJN+&h3BrI z1ft?iRU7P!uTH{N`McQrKFIta10_{NZTTf= znz62`U+(4?`24kWGkR3=)usQnI*3L?-*eJ-%VGpm0Y!=h>FT0DRNx2=JD#I6Ft)6nYvO|tvk@~>u38^&@7?Z!1ciO~$HYShfL zL(+S+`#GO=)zuoFc4D6Ca(5t(;SmyLfH44PSDWQH4ezpF1n_Mz*CFZ^YVPbiFs-qf8 zO_4?ZXPYbHG;Vkd+7boUGq;K0dd(a4KTB@>5h<8SVDZifcZ%fNZB$`2;P0-{B2ISo zy^3iicKp|y{^X@eW)mgmH1+vFsn^9MqoWaZ82ui7iNSFVzIINRhAABjt2zgv*;cvz zm4(_WTWq&e;uOv3@a}mhn-6;iW;g<>hL7A>5439Vz;G44;zGSx);zxk|G_GqQSny- zZ%MR;C#&xCx;i*VOxHncSdBr*i&|?#zH8RD<#9T)*BG8*5G%<$1cad^pAZYf>+9q9 zg~n9gUht0D4lz5ZO;Jn0_ixMZq=-Hg@tYth)gHo|mr-Sw{g(UViZ#mm_*T}u9n-W( zXtnzI0CQptp<>#6bQwN&{!UyqdR)6HFH{X~Gp?Q8rTYolatp}jL`$k=b7l*rt@28xPa z{FNMxIdsNC|yxf{|QB(pyjq`T<{mTW=Q5+v?sQ*y5zWO&KgC=ArU&^kldhe0hl+miM*$ z1Fj=gJ9IM)Q`E9J7pp&w3kH@>%_20-L);J$#m>N?iXBbbgDV|V*cQT@^z3>DV%P%E zd3#tQZD%N78Pszm&u^Xzg7YFa4gDL{|C;K?3QhgqQ*G;?1W%8~t1_&8(%RP?j9^C; z6!vrZQ~5wI4Mt~iwFMEo1c8NzP(E_cBA^Bw1;H{`GcW1jvRx$}Y3kv=4-WcYXQD}& z*9S)4eNGCZVN%tyceC{KZJsYS3b0(B(ie&<0;)y)CUM!dvNE;T? z)p}J>YX4$*j7sq%hqgpZ0QIJTD(ztwWi9*BbHtUu?BBd$A|h9}WwEJMy2SggINT2~ z5C~I!ucnL;f}}_5xGIE0WH~n?5mn`rr6b@e_cCGhs*|FePQQl}lnG@fhhA~cHvW$i?PpLqt-TIXf zY)|SSA$#CmW$3x3#KMI?k9?T@t;FPca7spe2^+ej!&Qzb5Ck2dG)uD?tE|OXgx~_cEmFib`&>6<*u&cEYtd;-oHf09qRO129x`Uxf4N_0|EAC0*WHg!2rMkR z;rb0S`E9M%d#yI+n7jIUegkMQIyLhMBn$pCY%jg#`9Is>32>ugMA)_6eKw`f4z~#) z>63S<`B|#ZmIb-@tM-ZAc|K+|?vsE(3_yHK`sUm;ck^Q}?S#qM{=q^KQ!uP;& zt2<;#M|)dqm{2~Px|xVKX?(|7TMo%wv6oX#M7%*6T8&k~c+ZnUL~=0OM$srjWJ zjip_wGGMxSZtG!;v#+Ehp1;DAOSWKzCeRP*f zr&KYDpu8*Bf69DwAq(un$?inxBZKf8eUO$zG0|`eLg2&MUsYSt(lNkN;hz#i`;jkU z++60|~-PQjF<0RYS*Pv@9&>P;6{KQj%7Wghn1S<3s^|s4{ ziPLUC{BYYFF>Vm>c0^izw; z%g>1HH=HG8!m7wJUf|2&htQVi5Coh>B;kSqLB0XIYx;_!Nk)JShP>$Z9Bj<()t=&T z_LHDSce7kX#QUt0MGjen-9=I7J>gl1Up!>?*P1Q3h8d{)@kv_u8@mE7Uol^>U)*!N z_oYp0V-X2L4+&$zleQ^3Mn9X%%IrVrL9mNE1ClZLvX#ZN^gpk6GULqY40i@`?qy7K zIxC>85=vVSNqMj_XN2KnEKppoB;I^X@#?6&Pt@MDmBkMjDdeKFkJ9gnY%4=dcOKor z;y&ad4vmEWINIF*dYl34>ib3^vQ8w%(espXzI=79qdx0%2HWLi!Czf7x*7pVU-ts)DfMS#z+B2g_1B>{x09f**4pu(OROBp7K=RPk_ zi5X8B%~Bchh{DK(PsD_T`H+!DB1klm9S|J5-M{=&cb@IOJD)Xv-@6~Hu$oX!r`g?k z-#xRRytn*BqPKdYl9G~n%P=Hq^0`bN%%7v-dvkjI6e5T5=dH%C$%grUlxMr0ALA82 z0sVS$gQ_4se$i)W$jlZ?yuQ*}u4*zP^~Ya1|Gk1i;&-Y|Wr7d+K*E(-DP4iIUPHZ& zk@s^LW^}WjQ~;${gF%GD=R!8#q+?N9lxuCc2cN6yR^{7R@N$H8VqxNHX?5rK)zN8g z-owC>rGOj(zwyv^Mw8x`$yAxKxqy@3;u{48FTnJmxBxbeP#m!ig(V_kNZJVfp0y#; zJpofp6uBs3c?i=8jwv;qIG1Pz`7qpN$lZvKDJ@;lhVTRlB_d|X+X%TKLPL=EFD)rK zN*&THQdIb9#PATR5!5|!Lvp$_ElDZTWW@Opv=R6{`#t78CsX_$qCBFG5h9YD4jFN{ z-jD%ta+^3TQ7E!JvON;(zsxfcSdtIqJVXYh?+DZbB;pkJ5cgzE#08FGeMvtf5b&)0 z@W%M@ZoRW-{epkpc^A4^>~z%`TcSV085V);PZ`UncDt*ezMBjs`sTl6-#qvg-p(uj zJ#l;w=^qU^acOW@qXnFr*?TCn7Lx%9Xx92Y-lS4<4CTmtOASG&2I&r=!yYX9lS z?Xnn-1t&)MkoWN2&lIbZf8SkwjT|02B+HF5>mQnuVHy%PA%=d=D*N{d7wechB zFu+<`9g!qt5`LcjoUtJY<21CYi0MTf`W^%oP&#Jh$%J@4-h#(;sGa3 zRDw!#yW|ASk`81FqNji{N5A7oNGR9R=Ry?eg;bpmQmR(85W|>&*rsYVgLXH=c6Xt6 zXE)IkJN}N}Nq|09Hpx{i) z(hy(%0BF0HCvGIzt@SxHB2&p-S~{)Gl2Tk&(?_E~LY5!NJpL!r5!XHM`k9U>gsWqr zjEj_CEMXRI^!i%Axi7naTk(&VmSy3%Hvk3$4jt+Gh<HroRGG z=>LTdZr=7v$vP|Q>NBTQRg;!ZifRPW2G+)I&GsQ0C}z3zM=v)~0yusTK>ET8q!J`0 z5&gH`l0_aWH~Nz!nZp>?kNChG<&er$0jaXc4vng=?j`+A2Y5TCr=q5!ERy|C=Ayiz z!p4MDW9?dJ7t>gdN4o$njah|gQ2mX02^w!AXD?oHXA!(p%>n|$m^Ce))JbWy+P@8v z#r-X;l@a`V> z4_uMnKKx@4tk^PUr|GWud;2>I^2=zO)R>&XU4M3Fsl7u_nk}QXrsL<0TjY%IggIdg z$`Rln!7MP7Fvf6c_ec{$L9g5#2SM#|N@rJ%Yi8ZS4ybL4wB!tI<1PX7@qC zYcsR}mQ>ihGJ(F0KbsS*pJjQeXVCjs69+ddm zol#X27uF7Yn2Tln^a<_oZ z8|n(`3i^r<{MZP}Dw>+Q%F3E`-MIm~TTk6JP)DlH&g18YC)@-<`V-sC)_?sc?N%Yu z1qq9uN{1>uOEbkFNKj!12xPdCTv!DSRk(dpSyWA}U~g-a2HrpX0f^u?F;KzT9XQU* zVLK#2zr#M?bEPNaf5d=TVPuu^0nO>cJo`FNv^ofR!6d%HD9~)K88TIgOw`h$A2QfM zp@EtPYL_EU+B4y`LDqZOkHX&-EphgFAJ~2Udm>9!!1Lr0dMeOo!cBw2LF^NT4&%XD zr5A8A98md^Q@{=x0-VjhgjX(bD?Ugw=Qbvm`RSE)%lnjnOA6mZWnn1@%3RDj;EtMD z=OPwK*cKOtSO(O+nasgHnY{Rb2YW)~`{quGBH9uZ2k>~@%cw8~t!iCYI+%=sx()Tr zArN{5fm(sWn0>yaKeGg!2`|Xt6mVFtlAnqojg{|O1tssA8J?88a+FA6t0<CayjvvISX(M0xbI1$2}!Lwr}IYH>?43jwsb4CAlvE|lMC>J1Nq{T1j|0>>i>S{QXWiM`_9J#o3 zO#haOiCVP$n! zrEMn%yW{bO3v=kA&6`J1YE6Aj-oASv9*_)+u7b_yt3sa}a5Cl**2ACfSo`hKpo_W+ zU4yt8jaJFa6HJF$DAB{Y6c^gfOIa9y3pTZzcQB`+sU9| zh%%b5{&`gpH@uL>kE5jh>^eIRM;s;XAzS~Sy)3P@Eb}tfGL9Cplv-*r`SO0Yk0aID zoka-clTEA=MYXdAR8+=n9`>MV_9K{yf(9B#v<&waPv%`}*JG`R>34G8|JG?Rp^qxb zT(!i54DNn-xEKy`r5Lmr)xIv*^HIp+r@&(FDm_BAdj&wKbJ(ncB9qe(z-3K+W~Mgm zS=(k!8xV&w*-f}fB&f?^60Y}$p$sCo&pd-{3At0HYr=v&Q}7U-NPTAYAUxeXe_jqQ z+ELT=ETJ`_rDM`>HZqqt|26(WV+aLilKy#l(&%>`c4dU;^|ClTjVL=x>+`s?eO^^( zXu0K5_rUHO?Vm2AnG;0WSw@~pCI1w$dH?9A>!diS-wM|gUz{tInogSS9_eI#4a*Vm zkJ<}+_+VkhF-5Xjz9jIU{C5uDp88*ls)RZ9hRK85Ih~Og{1#{Nj(mKSRS^s%RaS!P zbsX*Q^6~#!^Lpg1qsjm_D#B!0!)fHADI%pnV4j1?gUUDGbF3_zVe8`nsV^v?$*$cR z*k78KyQy`kToir0Ic!$NQ5BROCAUH*LnetoEZ-o=FBY5{aS*PwML<9wpec{vNXXwF zzgw4=!isnH$uzNTaY0Fp+CZsi@7M<4GXk~{@JbVYp^pJvSl}ZU(J06O*tN7U{rs=! zNgp=$iaKrGx1C#4yH-lb_*Br!SGC{CvEo6nV%QZ3htsu^Pj=r@i&3sXY=;4zxSYVW zU?ZAH4%sxc@ggM;bl1Fc5TdA=lrV=nf1Q*87i!f8QcN?*s4%GWmLsJq5O+cC(2EG$4!3k|&ng%?2Na53#5 z_ouesIngoE%5K5#a2ugaFjo;&^l9M=O4-eZ3dV%_dASE1pQ4?OX$C5aSwJCiWCbvz}=kRx55#XD-+in4IqD`?AbJ9 z)}7)jZob-}-k`v)&j*i9+jl2Sk-$g2JZm2`zwFTEXYeV(_`JdMypIcoi|id~U5p+1 zC08xDd!z9S%@8#p=IG1n>HU$@>@+&7f>(EfHGbWRtc7sB?GwzE(v={+I`kj@k@XL0 zVCiV2;I!1w{i6>?vFK56>ZN#@E!m@Lvg8;`4Sb{!YRjB*rU1;k2;FX5r&oVcgVNv4 z#o`%`hQ`u7)+p%RUf1;SmGsx1Y|NKEoot(WB~tau89G_T0~+SC^0LbMo4sAvP@UB_ zdab+Gbo&NSf`uek0Ws;pEm0gkxa$g99z$3{2ypp~2w&~d`C7Kt-=K5toQ9U@3qT5c zC9v}g1EbX=Fz?GG%bj91%V5?_QavA^J(Ms%I=M&h`PqnE)svSxeN1Apa1Q`ROI^@4 ztE<3u_Zp&~DXl%=>7vQ2)_p%3*3^ZbfE%t<3XwRP5#Zq|q%32CGRT#b8=urAXt#Z5 z&4$=?;CpGw8ssPD+KkaibHc+eTNK($+x=b z4Ysx(G~e|Uz+?$IR$DvJd!bYqb%P{PNk+(`K_x?5)?#le=BV21xiATzqc%mTY8>2Q zdi6vpyBi@09Z*lb73g9Dd+7$maERcn$6U`7y!Mt{L{Hn_?5xZPgqS`R#eac28s^=v zb^lYhN4(k^AM6|G^TAVZl?B8QxQIDI+aeNn4D({tsi{(47gitR367ZL4G3PTttIJ4VM&I<{@wwr$(~dT8o)9qYBLG%6{bsFf z^928rPAP?ZJj|gLNs=o^K+w_@CD)IqId}$7=QX)Im$%Sfb=1@*H952z=;B7muy4Y9 zj#c_1oan-ZLP$7-NNW)w&41o!Liia*zigF-(9%UEi3WwniU-a5=SYZ57e&fpe(@%5 z9S>r<8$paS!#Wko4&&;)T>yVQzTaHma!sw^5Wc_hj;gI zSG8T-Y{iA#^*A{!jP7E9*>8Lqnmm%da0kZv95)=d@t$eN#tg62CH&WM&RTcbj52yz zA)-0*WKNL~xFQkFPm05w0sE0Z~Jh5 z7V9If*hBhM(L>LB{>NYOzlK=XB!sJg<<U5kc`jAtSf5hRCD-t?wC>SOW(&m z)Le82@v+@owx84$)IgVef_~9_0hqyFue{aZ_V?JE>NN<${g(UPd$xFphS9-f20=E`G$UlVj zcH6|>lAYD9Z9zUL46{G%r*kCj=8FGDy5sdiiPz8Pzdd`YzUkRN!k!XmrH~C9ub!ME zmqM2veuh88?bEK^&e^73Ew})>YOD*r;MwDGE^i*xvZ4n)5%f;7&Q^jcq@adftQ@mJ zJWTUaKd`*T|c5;FmUu zd-m10W3JBl4dQ@==TJU;)v}U)3W74-8DN&@XmN|HS*{SicXV<8wCXu3#Wu(c%bx%W zdEN<8vpoTE;^VDg3e!7Q%SiaIM6TjA_w#c}n^X7(S0~<4zmth?E-YR$hiavEJX5Bv z$D3Ln^|>4T0O2@9^5{RmdCRBjotWGwOCygKa95?7_MfjOM2t^2b>k-<)8Py~Y92yh zl?IvPtRn23>fCO1tdhBWO1O4c=!B^Gz#>ZTLHs+c&r<^8aNte1>pEpq)?6iUZdq@I+CfSE(Mm!w?gCI3l- zH}shhI9-|y@gL*75xmL*ZgHxL1s{t|x=EDqN}?fU`cAEmJblgzkQ`x4QHxG?e(ChM z&5bXz%gispcD6;8>T2XwoI{JFxxN_TlYRBE;3c^4PHMj{XU*mUrsy-Sg!2SI-aCNE z7!6u37twcF;>i}6_y4X|S~MfpZ`1J7*pxXpzJ96Q=-YIG`hq!-6)jlK*}aIf==SfA%I?wV#^-54<({_RNsDk>Fz&v~nqn=HKI&+h z;6Li7`E2}?mEo7x={rBLRMDTqrM*AHzd*uaqXO~Qa%za@_4u3Xg?fs}8>EnuH{6+TP ztBNd%z{%zuJy;$lDEl4loALS2*jgXNDDqTiEcRLVjrw?UxoXxa-+JG#Cq5psNHnV7 zoQ4!^cXi;H)4dEDqlZl#vtR+Lu0lfW4+L+}msycq7G(S7B?i1A!@S$kCm^E%UBP;W zoKj#s?H>~G$>+Abs!Iz3=SGfj2l$iwBOTNHgQ`}cW$~DWrJ4Xu3^Zqz#2W?lLAUlhfZ$mvi4bIy|tyh3?|C+88kj5OTSkz!8l?j)7g#W}?&|MPX zCLyv2+lag%rQ*KcUUrcJ*zU~#XQx_#3WK#9Wcu8uxB$eBn+{l zdr5j~z}o~D@r~YYy*2y$35e$s5FJ1?==2h)lSg?obqLeZKC&!wJP}YVcp&8rnd_dY z<|{&S<4fN}s*O89CkOY8NJm~^%#;jiUG@&+Oxh5}3%af5Rq_jU^6nU{JsZu<75T@& z#DH5*m$9=oX1GVTg5kI8INmeq_mfL%1gbLi8CR8^a_*|FD|h6UShi7%jOyQ;>6%FX zCXm8;qdM)yf}d@cu-^eaZ8bVSRX~Bg%Zbcc8@~<^ler=zA~v(=M}EyTFwhENK7+Ic z^QAdpQWcI}sJyOq9!4ca70U9A-dGau<>IU!Q3u6>x7(VME~%GoO^c*b#hH(epu{90@VxyXpQ0cA1lU_Yjx= z-j3*s^y3Qxwm>jZ`h;^vz0s?I{9@|sq7FdPe{sx1RK?~+i5*Ngp6Q5D2Z+Vr6i9kk zvu7-&F2i__v9~_5_SR^Jb9rS2C9eKpfIOEE%B%G^LP< zrPIkNg&&gGGU!xBsEVEUx6Y943q*gAeJ^yX_vzi~bqyKxJx$AMRcJ@C`#0AiP_ucD zi%?dUUH-~kcrFR61nFFFw);k}{p`4K{{~0{)_(TfDI++m3>3d_J~!CAGyK4944S@T zMVZcFc0SJ$kmikDC(Y0Pdeeb= zF^(0B9prZ5+$$;Wz+%~A6^yxR&AcipsaM}g8FXsP`CoDFNGu;jGg_uKPHOEecxTg( zpcWpYUx&pf*4a(RIxORljX`wG8C}c$e6_^+u_^dXIXy$yOB8@r4=3Ud z4d7aRt8J+Rk1>s7%R%e8QhJL=9yXdDt#+ODPWSE|;gT%=Khq31r+YYRmCG?zlc=5D2R8`ur#&kz9EDHLnqvg_+Zlsm z@n_r?$b!a}gBv0F28UOXlmgI9YA13x{xZ8V)Np`1H=B#|a(0jMSJPajpAX|U2uK5c z^Iq%=%L&JkyMrlHyBzor;n!gNXxu50JT$>B)zX(J%{m@^N{Bx|X-u-P-k_M?PRk8? z#=Clcos9Drc{>n2BMJLJtLFBMB{!Ign$zMecj31mKw|{8e}K9e`Z{AVc}2LJ7|Im0 z88KNiRMcsg{UKA|xmJ;(xScS~{P5r9YO{5|kJnOI{lD*bCsRF8q`AGODM6$Vkj}O~O?Q=X%@Lsb+ zbr)YYZ|6{tV;5OyXx1_S`Zra#oJHuAkD0l!7lF_YrrJ}zZdmv8u?}x5&e!$FhjY6h zo9)ZSMOps*TiwQc2}zjx69+}sB=|hQ^ls+>S>Uri?Qu|x=_CE~Z13cfe7ML2{q(#+ z0t@iJ8p%F!cIA~&Y$hHmMq<=uRB)m8KWNW^i{7<}Gg6iU8Bf=6oKxKHYkqsH8X|VY zMmvpKe^&XS73P!FvxAC!Z9{6R847rlUx3`O2%yXiv<2LapO+k_v$VMqVKdAUexcuo zYj$4^savX0gnkt0gA4@eP_;~_QEFhL(NNH8#r@4&6Jwbk+gyynDHLbOzvl6_JjAziu zu8>kDvqJ4Ns~u=-5;a!Ns8gCtq{&sot-bO~P0&HF1~HExm|kSRqTevf&r0|tDyv??O#mN4iz2rmCEYt*wqe9DfVN3-77)F2O63?!eKuV zzW0E}5~50*CwB^M2W#r*eiKF*giFM<@CeftuT3X{cK+qUarspj^JeacaSQiJ1K}7L3};A3vjUWDOR??fpxqnE=m< z81b+5uk#B_`GQ*Bs0L2bUlMe&}LwSdQAPK;-|C?YH=+Ew4lpnld)!!O*Yf%ag7R&5jXrH zfYIbBQ%#ncDe9LsFG1m5h(p9I(&{p+b9<{2^kaQz9lux4Vw@3Dw-6#MIOh%A6RbzU zb$WRsyHl=Hc=-FeIDy* zoJ6`MVhc(~O{kbLW-}EonF@hITMeM5+AQpvIWp4E0NC9o;;ugOYCdpxh?D}q(XjC%16LdgYFz`|aK3a~^^te5Tt2wYhXKEFv5 z0ZG>wXbYLv=Ht`49^O!|X8qr39NMz-cr74L+MV_fdnVa{3&me9kD`q+6K&z_v8GV1 z4ws8{JdtB&b%7v?saiAw2|m`8e-#wco`-20^esPx1KGX);r&*m0G-HO_s?sqYXlId zQU5IDD>!X8Q_7aCoBpA8S!Be(GpsP*)k?fxIVE9t?L9uS?L&|uh4)lFQgvi-zLvEc zcHD-2mwksd4H&@e6f5b`ItrViGK>DQpMp7x7cbH=k9)*uRj0*hiML4sAF}z_D$+w+ z*+!@EY-Buh+S>2iH{d(xR8pu{ah$S8snVxdJ6dAF6?(XvJRd!qHXoUFvoDR7{M_>X zle+lX;h%Q_Cio;=@ifGbDud^L^EV9?*#obR2O~W^rC#*Tl1#3c0CyysPE!}t@a|el z=0^Xd6VeNfDake49BqBvI~uTWVufT#+3qd8E4sXna5v&n*AT52He1wucxynhfKH8s#Ks>5fHJg zB%Um1J;%X>NBdNP!xb=eNb z32Z<0WfJg{4+7uD1__`2ObhMp3Gh*|3A3(PxuXNo)|#hE0nXIU)GU+dK`fQ3kuP3s z;sxW5Q&Y!0N0iIS*!!)HcJ1JT62Wi5B8e!QQpk%(ciriNVVCU}EL4Bs3|$jD8qE~W zh#lhWP_tr<9!-4YwXLMZ#QjIwG+{EI^C8DCekA;7(V}L5$dXr2I2B`dTYL;hw;u9t z7Z`>z6KNCk9MGC^=z&ks{nY`2GJ776f7GlOjHYZ_w6C%omD@7Hy!w3a960{E~> z!+tFtbzAwlS>k{acG$mHu-t`(2v5Duw%iG`du3tZ>RXyti-e;u6m-l*0?C1&4q32a z!V7|`T9u)omW`J??)Gb%(9pYM3NjQb^o0O*_wyUVrZNP^TksavCG=}cBf<&V=lKldHJ8ecgEMn!^apg#w3krD{axu4)Ig`*gt#`GaX7(!L^kfe$x~F>Z9(N`q@-) zb&oTL+12oUO=cqFnkU4;gO+JK#VJ#@LW2n#?h9I3?WvS5SvYrKy2a^~=EhSmLH)r* z1ca>O(#3ZQ(itXnxg``R^eVNuoqgsvvBXnMMcdl}N)Aa9r&iqCA9O`f)o9vO@U!tr?Zb7| z;?rDIdjJi>@C$Q4l;Rqt^#boH*U7C#fm>pi3G2N^ZD9kD7sFesD3y#zX#7O9^h_z~ z>KfRNjFfqH4)(>Zj5u6JsTVi_pfBs;y&E@H=v<%oX;XX3hS|w$>b4F4Wy*~mrCxHz zf7(}zCRM3*?PA>?pOQd2*GqrOOu)N1yd(wm!A4HP)Q)DpvZ8^gYmR4<+${?zLm4Ay ziGvrFS7M%+>uFiK)rRc7>#1q7ZoBfU*Co~y_kM8-xpdn2$vFC6Y7I*~-K&%Pp$Lal zZKj|kES$@He!Qv4s-6j_2RR{@v2Q7_ z;(g5G{Wxn5eBMLq2ZsMyX9)V*9tGFb%3BI0kvXXE!!jLGYrFrsh7bTk0V ze?l;_{1*sD7B()H|C?S`rRM2^_Jp4!x69c>HX833j>99e^vpK|LuDaqORSV@peT#P zgatoq-p_V8!m4{$bW3zT&0e|1l{9TtD zilE2uquB1}uHOyzeyj~WHzzM`qJynu!el5#hxC&(Iega3mH$p86fB7-3klJkymPLu zXC7?C*m77a%=A|qE}aH2-OlrfiQ#1NR1bZbG*V5C)rD?ec1!OG>*Hr{x~|YVIt-Pc z8~A}l#p>#6D|Xe17*(`-`pVkPc|M6I$DKR8vPd(ij7SsVyd*iALY<-}sN93BgAIei zM#y!4U-{~SN6Ax=6T@YVNSMP%k<7>?`(=G7t>mR;yhz`!O-@?d4!lNwoIZVTp;kq* zLn`*llFFzxRA7~Vsr)2yoaIQNG9aOjRHtyoj658i>G3|SEpYv~K)6E8$ov`7W66hI zoOK2&nXRqQWtddFxjZ-{iKOasG9&jz@tGH3^XcvJrJv$N zi(`q-&toWzW%5Ay1I|0gk(`5`T@G^NnDcM|qnDy#lUoyQ5Ool{pI;!-df>v+OqPxM zV40qIX2G+V;HJlRZ=D&Ki@3`+lQwY*^OzvV<$3+$LlgK3;H4CIYXDu_TN4?;fm(fhP(u^!n%MYK}^cW80iK5550WZ=V!#P}*xN z;EzOn!#ZS#)xW(;{D_HtoYFH9l`|v#9Teco!s}$JM;u*~ zax5hq*wY=7cq^KIZo`h*9giK^=nCJ7MWxl*nV%ciCe+K)iBz%+>8w!bhcuGs0kIUC zcC`;s`>hRpFdD1#6T|v?WZ~#v=7ipqUsZY|D-*jEy@`Q%E)@1yZ6s-1w@+V=QOQ6@Mit%{V4AG%n~zz!qH zLN13lGD#fYS1I8|^;Fi6KP@@xszO<^?}0+wkC-*w0A9tLG>SX}lCfmPjf*Dvk?;bl z@jREWfW-T2^N0F?w8gpM}<}C>)XG$AiFi7v(;!ybQX` z{u{(%LjGuOLw#;vo_Gp`HIikfz>%=?Y@=7`&wT0cn6MvpAGh7u>*uYge{yi^NR1$n zR%)-&HJ0vqc3{7F1CWeKr);I&`2tah%&QE;hGH(^^n~*x*X|%QYvDSROf_|}E9`IJ zBzSBB!))NDkIW+pNz1y)VHYbRvtbE&-%U^DFuSJewwmGla;UjbM~E2O+0JtgWx|2) z6zrUgXOYfaK7LtXdV1>|Cj2_TV)3F-OlPwy^mppvk)>3AyZNSV(4B~2ElEU`>OUh& zTnm=1tuG+PfPQPodmPbv#;nI!KJJuVNhyb!)AqPw$_eyie$-p(;A@;RbuPfbQf}wg zhBXH(aCS~=6N4-fpVMA_%o16pj+>ACAr?F)B<6HW z30{GQW;>hvcw^Ih+^ca&h`I9zZsfdwzfcrXPFAjI&*}#1&r8u%-7`#E3Rg#(YhIpL zt?4u1Fkyhu?UEmUQ&kGpAMm3UM`*z3-W>@#3A3bJwa1}ysS2o&EUYVy$2sQh@i<$s zolqYv7@{nbUuFJ`WIr)*M)_#5o4<~R0wYq}4hPrgIDBmuIZ2?ELiXP=gpnF*AWhQ@@ zd69H`FO>QF@RmrA_0(jcBet#pW6PB{=v^pCKxPstLkM64n4kb_An?1+qe&&JO{?SJ zj=~^igN3Bt`BPXrPy4bWc#bQ_tW)Q-#P_o2y{w0ash2WWKiX`gXCpl z!hdewgjqWmVX?eDpP*vo*pNpJuzqc}hb0lZ)3#c4quRncrJkrLINCLVUcYwaD~1-L zv}zwI1li^dDD8BS1D%-fyrZnbTANfi2?~pKeev!xJIs3t;(8xH)dqqmqoG}=n^P?v zPv^*0Rz2TDD!oeyljBqDea09TmcT#7=aSs$^*@V7IiXpVtoCWzC0_@FvAJ-^ecmaw zkgaQ3d|NT5P^%^@oA~M3AD;M0%&?D#et*>A7%w*NV(K%kB#fTr5=g65vpke;F;Yb# z^e>&kB}f!KC?z{KA`a&c%RD9oda5tBvdpau>T@?LXl*acLy*d2Q(f9Zy*SM;t1;fR zMoDkCLAlKFG+qc+MxvmFO2i+SU@g?G+Ap+?28p{7nq8ic&^F9alwlL#+k-GZc+kof zvMR@r#cHc&nt=Yij_*V@mZ@ z#OH&D71x(95yvy;#6!JN9?r7kU#bYpJUw*|Z)Rd}YlDJs&Fx2WS_=~$y_-R0k2=r* zqfF7*p-6T$sX?@U9{s6c4>i_s_2>Daa=)nwC8juVV(j{*Q?mu`nk&@HqGEEqoOjUaMienGGjJh{J~%_Yg;%rP3&%X z#fK+h1XG;2IUA+{T$NpF0fm4C`(Hz%Rh;ox!w+5l=2?gCWgqjWizYWW6op8GJwUq6 zg=YV$8)W@o-JqBzY+O4e_AKd4`tDNR&5>(<}@O3ccGhi}Xq zIiAYkpQ1agXUe~*P*XG8-I>0pKNjf3jZcuK0b8iqtF4umK9V?$#^od#x=G|sOw9ad z_dA(L^l@^({|GS-&R_&B2*QNS;9%KwMgw6#JuMtb3uGB2RIux--oN|QIAmn<24E3U?n(_n#0 z%OUVB)y39Z7`jO8$<8fddIP!;;Onz$Fj2H-#iX1-&Di)0zO|*5sq5z$OXdca=vdq% z!EACWBHuaPfJAyVSF^8i_RXB7<$_&+d7^^O%t*I?XS%$qeXgeuu~AWI<9^(^t!5wj zX2-+IkhWLeSu>7!WltTkbhS59u>V}Gy)_3PgnzqluOdqmy2&nXoo)naLa}iNX&p|{ zFvp&eHJ4I<^8`90M1PgI!Oypy(LIgm(pZQFe96O0lUF+ot58| zJ5!b}Gc;vW>ZJ9z&St+*BVMRolQkkjPY3^@`Hd(qVAfFLp>s?i6UFB;wXy+V?x1S{ zA5IKc8j*Q)a9N&f3636(KH;}=RPxBfj)r;WO3N^z@PX3> zD!(bFXry7MGC`%sF=df7^7M`Re)%MNJ@dDSyuboR@_&4WZdNd#V^5dHY4vxO(v2*14*1Rc#`{*RziPqWo|RdKnm2 zr&~^@wW|4{eN!!`-4`Y3s^!w+tOO;O!3>Xa$g8fT82mJvvp!2ry9g29>Ea0hpi3PC z1RzabU9}fmr-xuXh}CWVBRrlcv^bv)@=US#lMZbRT*(2ZBbPcWS3YbP2szMtZVUs& zyd~(QjRLOzo>Wk=8bfBsVsx*71PC6r9~7Tl{O;QCb~YlMDrVQDL!u&3vtC^`nn5f5 zY))68Pm>sc#f0xRo-B0#n!=n>XX-%I44#ou0uAzX?WpO8-5VN3%7jCG{2{x+&4Ym zG+4EZ6s*vCY%6?b1!sjk2xEB4|9l0WHOdb%-mxR(08*@fVX%Q#o4xc?!+dZuK`fk+ z?EtrgU`4dt7m-m_Y7s^E5`FrUxjy%rpc!#0F5w7!(l5@P>|7vY2j(?i@M@XWC_EWp z1Qjrpf398pH&vU)jQ!&8!_ZTvSfE6P`=U^OlN0+i##~U~G}qCzf`YQG&!*S9!HMO0klEK~8pN3jQh#IXwCgUY^nN1pF$38FmKPvVG9aTSXlj^ceLv)Pvdktn^DS&GR-; zzG4^$g6;k7*5ei)CgCF`QVI;VUqv6#A#ulP_PGK4e^_(|c~{ch7f) z6I|uipcLe9OvLXDV(FY!<`EWdS) zQv|qa#Z-w>B4Qb+0U&_)piH~vMz$u9O}Ctd>bu&lW1mn#wA(Jq8P@gQrnf^kegvft z95_7miZ0=h#!BNx)wQ}(IFec7|Cz~%e{i=HI*>N#HBk+1 zYB*8GX+#qeOiA@4__0lj9iPLt>5s&Soe3IT`UyoWR0L-TNBp=r8}XOtEL~gAZ0fg( zfdIPuhZ?hdl92g!x*M=Uxf=HX11fH>=?ksdkOjyD1`6<*b|#~Pz}0xhK(;!d4kRn zOqEqP{7hD?JcVGN7$=%!xNMc@08`{$KvFTPM1tnBYX-k6N*oRZ*Q+}DU8|-lrz`V* zhU~B&!l0->RpU2n+)TwtJR@#VQKHXbJwM-6Mx8?KqI!g zc8j#&eyRbi~0`l)Ov z_G2rC%2t&$+Qj1IW6LWof|X)ca5RCxI)Yr3-;OQN!ZFn+dxMA&@O6YBMJkh4cSsrw zN&gJq6_||w2*cbk2L>fZluI9=qk`;D$kqJSfF4ZhU!W{xFxL&L_CTohiwX>3!Q2*< z&GSQ17}6zl8Idf1!V-tYOMZ9BU4@<~oQ^11hfIwi-#5;Nt88W)Q!Q&O%KeuGso=Zx zZ;|wXN1BJ}7f2gZ7)?cPSGxOT9vEQ)Tq!WOZ#NyJSCA5Z?ao|tmZS1iwYDHLI*qFb zjLb9Bk=;j@qFRj)E8U&md2FX;hP55L30oMNxbsW>N(vA;EO{5CN7E@OiiD$L%fnnT zcxpCY9oHztnp_K)oh_=2TrdZNhA=(HHhrM2e&Vt4MF=DtmF=@?U+4y(jA^UEA#iO3 z-k29i)(pM(ZhZje8%Hi>63J>3+7)5?}M4+f_-$b%Vv^(nX7Uev~GW%M=bA~m)Sro7}jWDEVesbILG&i^lj9 zOH8ETRgJG@WpmC@`Ri&#el+If20JxrbBcyHOF4rLaIcWx*3qh4dkvzQE6XBIF#}s9 z5)Ks#uXFCr#i_GfP6R8oZ1j6_;0t z|1)o()sf3dx>HQYxyMLyUmD7P!`KAd(A|c4zKdr@V4kl4Yd8c_TZG%yv^eR}UbNA^cvUx%(>z6qmbm9=76_|%|D)lx zKQDkyv-opawP&tZ&SePWdDcwj($ZB9H$W#Uiy&PWKNsHu53G4xg~Xaa2Fz)=-wQDsqVDc{B0qj5wQ&%>|%b9;D+-W1|kCJe@%K>Dmiw9^ACmom-oqUxjs zR%RD%$r0iX8uA_^FnVM`bDj+tFLv5Z#qaJ@?JQ&@B5kuIVpnd-=qKOp-7m@W{^a6B zzm6glQJH3aN$uI5c4qeA=MIGOPp?iLDB4DsIW(&HdLWeTCQ735KM1%V`l*7Dg8!aN z=`FoX{fxD8_sv*rnr6sOovwX|dl#y3TEj4!=0%dv`Y**~p5$VMk`n z9$ZGoVw<|0ebqgd8-m&$uZ`vKb(X`SBx@=KA8 zD4AE@pY=8Tt@1dI1Y{l#6Vg%5BBa&uDu+>AQf6IP4h*XhUic=kAc>p)K;W1s zFEc(EcP|2fZp$UfADAmvjLo< zJ`2+lLX_~x!U`)s035xWBro0!?uZo>``LS2&)98bBq2fowboCNpSG9kCX5YTI;iD8#C1|JCwO1wuntXV$LkJMNh`AWs?u)<8lXU zYIV|s#!d+IXDS%&;s%r}UuV1RQ!R)G-7*w6zt@EC{UhpT*=N#ONW^X?D*u7ibUQ0< zmOHjA6G11)0$tyqU1sm4j5IBBmzpy%|DhP)Znkihc!y*wGQ`?`%yHNLX}PI-`DRZ+ zsV$SOeHr&H2}^sbilnMjruN8A4YIpn2Wty$2@f6dF4S0BmlD4B<{Q>UPG}gPbfRQ3 zaiz!)NIT{j@ggB+Ia3MIAvP7(Tf#Z#BkpvM?_fvCh8@XyN4+jgjI0toOerfTm#MqNCGpc1KfWVnNJfZ%*;GogGuN=9(cBHTHjc9GIBJ`hjPHg{c z=cM9lV0oVwT}e@DH0DI#>nJ2vIi?>7xi;O6x{e5lfetwDW736TAoz> zN|aZ+0AqRI0T=O!#98NIADBj>uZIHhPisO`m9u{DjBq$@39buSs#%Zo*u0BlF&BiF zAsSD$P}n%IL2%lTM2NTO%1jZ`e_PaHhh+N@ z2#|K+S`FiHc6s3m>KO)&(*AC>+16=QIDsNO5K7WFdmUA|*O`B3${&^;(D6w_{+hD2 z@vib!-6YxGG~jR#;zZ-DQKzi4&2dz|BBq+HX->w?ku zLrZtP{}7DH0T+F9AITnWa73Zx@e1~B3;@IHk=I7lke{MFMj;dG{O-M7eElMVM#k}&#J1fxlaRTz6(v1DTNHbRM|6`_6>PVZGN&@+czJmd%`yWCm)_VX7E*)_CZ@nZ{lGl{I zsVuJ0hvxb4vGqN#x1~@e$36+Uyn6ofB$==$`sDIpZKSP09Va{P39czj7GwxfBMNYZ{sNXOVL~Fo>%ZjCp6`UnfA~)r;B_<5JM{WkM4D=6(%)lG0qb(~Q!$C#hx~ zRsmo`H!$W?WaG&f3e;Pz*gqrS;it(?kC%ywjwB|$vj(0%h}K^(#RlZ;>nMp73x-zT zFbqlsCCs-Xid3tz)zVHI77+LoHqz&e{HLGh_+LN$KlOu!o$J4-AMt#4159us?peZ| zOkbo$L4D*}ER9s_dc6r{y&@(;j-lJ#WB~WmTY_7y!-mhz%$ihdJ&uje6?WfwvxS5c z>DVVN(lx;r@{Hz&2b30l<=B_9zsH|E$u&2^DUeiGpB!6dZ>M^lE9O_-=qqUirV*+o zKBTN5yMIw|#vghLmalM$@e82cpT_dAb;4*)MUWm#L7<%0XG%d%xqjn$sRGxsB9t28 zfq6R28oG)5hW}A=!`D{=>zy8DmM_Q@eM3+Kx{Lx=i*ll!Dbq{rBJ&nTCuPp&dMX^d zRPMX0nh!Wf_4x5`A@Or{9m#B{}pEz7FMSJ zJGxD3-YzOn_&t{E4guMLlapqnU48lfg7l*cWsvkLuhp4QV-o@ zU)!AvWK#u}GTNu0wP;w36@h{Bh%}>nr+6D+YeAmJj~>3zJ=tv=%U7*gSyKYd0@oMr z(_h&;HxFn?XO_W&L`X;W>#DV8pI3Q@HbVE`jfZ%K>~7n9sVruK#Yij*A2&G|Ss@@V zxuMvlMilYHNUAH?w%0zUa*5?u_C?z51~0;A^X%5@^eGGz6FQ{7rIo77*Xb)#>=Mdo zC(v2ge~4447>hXWm>_k@U3GF|>*K?<2mqaIpwsLI-mCJQf+M2S81lo#J6foujo*uKU zYPpP@je7mNLBt_eC?i(pSSk_%viopdl2}b?I!X?dkU?1zG6QhI$wOeG`S6FpoT4wP zD}M*6_fHp!gSn`r)B7aaLl-BH^WQ4$QabAQxW7rb1&AfhIv2jPkwE9$j^{nNec(KE+b==s{b4ltj7Ee z9$)%Zxt{*6VuL|)NygZ&=-6;1rqSrmXkv$gL&MUU-!%`r>rDcqaPC1=_D6f5r^kp{ z3~BUueT_1|KU2@&9Q3^mKqYI?aZjzQmUJ0mUT%9ymD!PABG+uz@C-Tg)T zBWWGAlc(S)v$5<@A2rdjZugg7lIMw|^h*504xh>mM5zIcLqunSQWIZBxQLB?x0zM2) ziR~f?%Wz)MD$8JsLxF=*F4qxm+;3lPU9r=JH7OuScez3;m@_A7M6{JimW(;B6kKp1 zh`#xg{t9~jIeQclz$YwmrT6Q!Z=8jL?+v$O)o*c+x9=c@`2AcQ83VKI_MQ@B_E&_4 z{jeo3c^V150?Pn47cQipL*M{DX+r`odP^OJg#!jIh#iyWEMnjUqlMN%9>V5TR0vw}`{MYwcCzp&+=H3-wdH+W387 zBO#|yC;mc`w_&`!;2B zj*%SQB$7ANOxEXZGRK5FIjpWIMvth2K$pKVf|(K?!B_h^ThG&eR0;!Z0R7 zqVQG}APX~cS{o;S$Qf8LKGhQl=KD$-7@FCqU@47m;$ZFkbA*l2vUG9W8iRCnN5F#v z*VX>*E-P?D24hOXoE~DvAUoRSvK{cdyB|t7Rrlv;N+1?*ffiAD0Dbd85%vR%@2_Oy z#5%z%MgyV+Yo9`$ayzE##e{!l^OGATZ&zlJ zn+Ki{R6nG`3*Xtz&E+)?9v(iQ<2}1CwaEjvJIg!FY(g~tRCspZ>P*;q=s1jwjDCXB z#>DwKjrM3|hciV6-BYihbv?&ybMpMT=||lv8DzzwzIV}y`twf#D5gE2#3~<-GaG)4 z2vZtXo9*X;U$)uUd}WK?u@83)&#<0Mjpo@!md$8%cf}ezF{|iy#c#_1+kA#$t_c_P zieF=Mf_5L5?nW3Fb!^BHm?bx975{((Er ztk?RJloC!49NmOCX62q8-2yh1^rg_PrN3+|(wfC=?(A{adqJA9tQcH!=Hxf0?aFPA zLM-rENIl`02nQ}b40;!!+wgoOj3;jQQ+v(7urzc=aq2}K)$acvW9QT)2(Sg|wr$(C zjcMDqZQHhO8`HLJ+qSv$up4n7?nc~SP;XVI@_d>3ZT+kTs+)^c1m_HfkF9_1YbB3a z`-b#tc1X-&z>q#QM>RhLgFCOkc6`vFKGRv`2yhGoF{6WC_&5F!&M|aH&@4U{KiNgM z$NiU?09+)ILU{1z+fyCFeuld)o-fMmNA;du+Acoido>^PU}JR1j#_V-2Yh_i-`p>j z1DreDf~Jekr1t;;0Lo5Lb+mLQhU*PyOU;i;_mibya^@t0AFw_nZm$0}5pn#-MD+iD zr_4aB?YJqP*mGICcq3RAb0C8J=Nphh%-UMKaU<16!&0JqoA+f1Ok@B_lO!KZ z!izPIv5rCR?=3^vH&Yc9h{Z%~?9H7Zh6lLhqd$)2&u>J54kM8oL)9wY^~q0 z)cYPTZBpTwW0xE6HnNzne`^q;hM`TaCI3Npxfqq4_I`33=J)l-^NO=aqEyv@zJtu- z`jDl{SWtuPf-={noa_N|4IJ3(Zd9I_QB6x*HQpa1-ExHMLfzJFVt#Q&YQFHntgPA( z7Ob+iCD}^prQ^oT55yVN17tgY_OLpcE`b^uCO3xNrV1B{f=}+I?kaK@k^)3)GG<}icA|cq2ZB`# zMGi48Px>9|*T3^?hLWd__f(iM%%sR^=m)`m)Z!D7gPSVbE~sp|oE=6(D&AB%Z&nhC zgMlIABAqxATQVkel@jgYb$eRygVYEsn;jnow_MiIC5GrgXwZ0<5Vv9osd=vM-gl)g1&lOM=( z26f3{v_0(AdO)%1RH+7!qrmd@NdfBB^NI zs(+w_lq4f;HGV&f>Gl=D$Z{m9W%*i~R1gehf7FkJDvxnv#x(fKrW-G?(45C#Dzf(= zakL&Cq}Ug?)aT>lIs?@{)Ln?B9+X1LGsnjaN1MGk5PZCPtN(_#?ZZZ`?YM@h&|=!9 zMq$dBCigp0%EFxs4I~FRLFu(f>4@f*Yfn%hdlCkT zP!9vy@!uf1&>@#Np;zx+cOFv1PheXaM>a_%Uq}!=j-b`P8+zr!6=35mVhi>UXK{OU)Ldj8r|TD}CHFF2fR+@zF}&rtSU2@j zys_Gykja!0R<}YC>@`3w@^IY}kjI}F?19l4%-sy`1)(=nO}H)i%`JS4M4yl;E@;$D zfMp1vS1VlxWtdJcRbSFFyh}?rgET{d?iUFFdkFl2T>#<>(~sG;>=7IQ<(&s%^)OXv z$z)1vNTIzA2k~n!2mz5wsJXFdRlF#vQ|B{TlALQc%y;$#j7&Jt4Z;Bn?_PV{r>Ga2 zZDa`y@5&{AtTE=^6$S#*lm^^X`ba>FN$ab}$^s8=L$(V}&OEG6bCyM4Qd7F)<**Uo zJg5a6{n;)dBnscGsUXB|2!jB>UAy@XX@q5kfIrIQZzhcz;EzENj6G+_DGg2XDPjKh zfGY8EMRx)E2E=U8*wO7t)OKJ?M0EM5uDra9@p!R?CdH~DR|y0EkHt`evWB&hD(?cS za)k!^N?6}y<(2T7^BzI{{cRwUBZfbmuY9?HI$RN-4jBMzml?2YRnP^=8&BYI>%~&) zMf6EM{IW|v8OSbXLF}`3fAQCyaVrI-G~ocMe;h`Ar3F79a+)ubgNo~GQMm;^D^r{i zy*tP(kt`ZCz7P{WjXHA|4$=Fx$$?v?t6jReswntM&r*q)EiOX&OB4oXJ425sdE0kf|>1 zt!8Fgv+cS-ikKW8;Jqw50r{x((z@Z!3^ z2ZMr6BMXC1ccRtS0Z!6%9(+5O9 zofV!K!j(T1 zLLA*D!cUiXXn!4jjrX~?eDt(5yaGBQu)RNIW#ye+eZB^2RVwANGcRiFcup9$!}1JY z!pgv$wD%Qqq$|Uf4(54t@MB_7({;r69W>zm!NCQtj>O8!`7HM~)?K)^_t}cFjJ-A> zB;Cr4s`Q{zP+fXmPIqKb3B!U#N26XvBm#Q;pP!pb2>OI)l>~cA=ob3@s)X7z5tH{+ z`f4eX+(hd8%F43kD>4$^%*e^Vh{boJ^i{U9&4Us8f&kZm3xI=-c0&u+Rmu7b(Ms_h z_rn!{)N($Q*VC4|uJ@R=yr-pMv0Jf))-$%JdSmQ9^RsppFHXBD>?VJ!^B{@q-S%r&UX06)D`AWp4jeQVPbrT6 zjy<_%21+sI*ecrRN6ReIPlmp$mWi2#0#6%st1tBkiL^(TVI(61`Nt-{ZH7;-oL;qJ z^iA8Y05+IeJ-DpzRHIFnYSKWkkr3N&zwM%U%u*oY%I#^YsHNTaAk$Y$bl=ekBj#oD z8Dv74nkq8yfvalFkcH#F-40NFm-)Z%y8OLPtP;CZK2b1OGBNgw#fs9xe0zPquB_yq zAV<(zVO2G1>|r!?jCG^$v|M=L(98PvtOAZyPeZL0%XYCM<)T~RqzYhJ45!V$Yfn(H7FZJb6iT8^3!vp8!*cP#I>FFkY2x3LzvhWeN1f$`6_ z|64-VOjIt4rHM#`TTL7j`M@YQqzxDS$1Y*W;5Af$=TH(oF{QQ^?Vc-L+5P3&u-rhk zqyl4t$I*=uWst|=EAAP~%p|CduP(EsSIlF(_+^^Vkqk}h&DhPgX)U-llsA@f+#KA$uCOk%N>=joS_CGutyL!5>OlAZ|f37 zPo144-H@?$xY`e2Q+ZXR#bJ@aX!`?b*@wxHcdNGj=M;nU5T_6MQ!PC>-DE$;3mqQG zx}S%T_3w1~EygE0sS6GUS#}(kBPM(?ftOg=kf^oPhZ&7 zn`lW%iwdtWh=so#y2z6>rqb|Sz`FE%Hte|HC2psv|GaN1BtHlko zEZTjgmmDc%W2lQ?@C~dwf!&G<}v%t_qfS) z`1o&uf#{xEjTuVhO{~lJfX6biQ0lHPHMQy7F`r11zw7zI=1z1m5!A+v$(C9~_=%Zl zF?Cm7gTDqUvfP#7!-$I7s`#6>JFAAOO^){Z1pqcoT0f^=;2UmsGZXSK{|wfc`){l+ z&@{`(72;4U?wbTCFHQ7^!sL9b!Js^HGoQsyjLAXM6W9k-Y&cmRXND$7Kb!%gPLzav z-G3yTKewP-t5?T>8_*WmHRuq>D?-N=06lhfjAcYQ_EA6NZ2m@rdweNlD{^r0`AxP5 zMfU(9eyuI#t$^mmX6`o#t#T*4&}tmU@!%2<2MJ_#;NYt$(-zWbXekUQ%H16O8R)qw zH+BTZ+QTJ@*P6wKGrQb1yS^ZM+N#DC8Jl>8fDTHIWs{#{1jkt0*UyFCoyq?~ zjeJZZAn?}QTKx*Gi+wMWSji)G$JGnFnIU`o*y&@}yqahuQR8*<7W8T$AgOo=N`OwV zez5PUt4CXhDYV)kXO5IRc$e=o0t{NpnR$(2*A>D@w66hr;Ce+xMcM|?I}pbJH{f3MY`9v*HbwCbD~vhac4a|4FFGvi+3x(A{ERo?vNMfx<<;VlMa{)* zaHk-bdqR(YotW-=y`%->szzPlnI$h@YG}GP|Ci!mdfJC>mGVJw6gKaabbIQ7uaF@< zS_FsGW;Km21H{a8X^-V-cSAQbkiba}PU~Q+9UA$ zL3m`t(}nOV^IlJ}YE$~ThVQ?LhP^M&6a2`mho8Z-U^cxLt6p)km0Ze?~&wsHVjNQhGTxwT#FbIi*3En6bkN)_g8+CbD}cEo`m-_slKL=?#@Lfz?R(MP69Rvd z@+B_K?w3YWQpLh}6MU)DRpZ9ueB1H8-u>rlU1oEKK5?_PTW4`WV>!=Ih+xI_5@cHD zc=w~u^CRzEZ@QT^K`o06j23&@O#D}Ql@hu*u`YWHVS<9yCiV6d(Im=v)-GNGv5jmKB z0(Vn8axo%6{3cavV9VSks_)vTACL`~&3_X`IsZcx{r|0r85!CB2bEMN!ABa|3}^J# z{Cr!xhF*(&`{^b-vwr{~r6PaEz{d!}wvFH)H=`L4s|Y2ANJ}5WwDB`h&?Ou|)Rff~ zdH2z4k`_+@T`x<4mfB-XkFYlr_ zv9V_qa)(4m1BaL@-p8YL$&CT!2su^)994n~t*`@BoDsHgVx)+}o<32vG}3kObu~qk z3OO6DuR1(Z8)d92CN|1+52oYm8cLQXu~4tU6%-2V?U>Fx5^asEH5F_(Ww_qemO}k6 zE3l9yG?2o`_XHa;L0W+=Yq=m3RBdV` z%ft8}M3l!DUIJt^dT2FlAz3)aKI;H7hiUJIWtuj`Z}} zXcve&{s770lcbRt?mAPz={EZ0jK5^mt;L$xqph?H)H2{npgZgjsR%aWb@WjZ!P?#Ug>v7yE8-PYs%?}&fzcvbYANq+B~z-RQNH&=6j;O`1s;hH4~3?&O@nogu>glyDu50}V z1hgSF>#p<;kp@7()4)!gJ9;4B;09M;V##}k6x+^9LX_x$JidX{G|Vi(Rk)Yh+qsth zU0B*bRgGm+c&qSR4BCN|UaP+uAtz;CQIL&7BT0mLR(xzdJ$yuBipa4M^ zp6<|jI39M=2yE$iiR@vDgUYr3rge{Z8JT6=>g-*?4DgK zJU?Y)2W*mP5`@Zz(??FISm{Y&S6XuUlp5VBTj+5u^EzH%xe4%LZKM2s?n||aXf{WM zRjpcte@ zog3oh_>|blqx&isG5T37@XeDpxItR78Cz4|B}HbH0pT~h(6Eek&Jp{O*2~f6Tc*p4 zGWZ4Hnq_nRFSi`)e^zn--_;Kb`~MV;q9tXABbM0nSi4EnXT-Kgr_#ghq9LDbd}*>y zH2NZtE1W>nFnIy4vrP;8Z1qz2PELciwutugb@1r-%RTOVa+#@$1yi;p^_}{M)iJ+7K?=e)#xT%p#Nru0dX*!synppoW)E z4=19mE|1UClZo>eHt9=0HdPd2#%Sk#Il(<0%$yrzAIi$7VlkY(FoQyG2qC2{{p;~- zz2C9tZlRJ#J$uoywO$j{gJ)S(W5T@s)M2$~`9A+U=FNm@Su<+&dD?ga8?-@_Jxj%F z5?9n>xbEQSps;3Fu9TF|C3zyEbYn@)LL^$S>r|KvTXefNYx!Yy+?XoHSu%U8d76iw zG${U~bdF3wn;_K>vPe7VvrJw7O9b*8KE^;HGX)D9|YkER~H8JSG;55i3u zC3T7tjE5HUSPjD%63&Z%mOKDrowilkn_S%dyQ_kXHD3aQjj*kMoipl)y4{y{Il><` zl>R#ff&_tONNA3QNlatdU)!5}G z0_g{f9v3U7jUGCY-HLA(4M#Dq%$$>>EJ0gk#gC_u7;u*RAexN;BT)r>ZPd)}xza3o zE2vqu(#@QQ(bK`K;aLi@g24m)QZQQ8eKs46LB=Za`MYk0L5TZ8AMYYr;kZifq=}ugW;5q`v-FcW^M#ylItrt>)fpl)~8M)GDw4<2xS*- z+m1Z<-VVwA-thCY^@P)wLVGyfo>B#E;`PC?wR{71dWq+;d7b4MtK%4jmEykGVs(o? zBTg8>8GZw@eUPR&DaH3wOhiHTbrs?!n*kBxkvP?+ssy8IQyBbIw*G%|Ec1 z3CNTTaUweOb zIdyCUBQ@m7);_50>|i|VRr8|Whf`GPsqzGQT~xC2fCT?+#JHiTTe(>%E?Irm-2Qm_ zy35o!{r9J`Cd+vRO8TVoBVjUR>-Eu&9AOwwP!{imq3sG1=n}5}s19TvZiKET*+JlO zye{P-ofdLAW=l5%U!Y(fy|ZPc6;$Qr`TDNhsXDds=6YTW18Z~=rc~oL670Z&R7&qw zn_amz6+!WfO$FpCA3~|LZMy{8EqOzo{piRZ?wLtT<|=fA{>dR`RFH^5SCOMJX_FTw zrHJlK8Xk74>>f>;F}3I%kYKl+TMflTM(RTHM@-2Q6z%pemj{7erP@O=i9k@b#J2(i zxpn8;O+a@2QOJhDgj++cxs^2J69^Ww)GIrLBaUnXt1NBuIQRpPc^46f){3m-E5F?p zC$7k-tEv3YsqFm12E!)9YeL-bXx@0hAepmk!D*2s-9$3;hJ)!KaxOf}UbtA=@Ad10 z&}EQv0Fkn3V1`)rvLJ@_%Q7~hd7JGp;#Ys{b?0uLO!Dp|eDk5$!J3*!TT z``4K9z5Ksq2_dqss=L+~!=Tt9<~Sr-A1F!c6%`8}*3p%_k>%k@F;G*-JvU91Z`FYD z-Vm%PIJl9)=S1Ut0Q=sZ1gL@ai_gmPe0~Pc%bScN{&^k}t;EE&_^|-mY2cq^!eS5U zd=m1F`{%9B2W60$1U*>Rlh0~S<0C;RCu zJiAJE6H+cfubN9uVvy2`8!Y03I+U+X;G@huz-mL4bZh9VuoF=l#pMHCT;&!}PWr-n zbp(r}W4w#pl+0a>6*Ynt+H(j7J(XzbFBO&}unNT8auhD{p#<6y596q@F*BN{8$V$I znl!hp*-0+`km+!DT4Th;`!EH6isc{-jl#2ziw_qT_@RQm>~MI^2G3xTF9p7cPF@Am=#O{Ocu zAe1oVugfmm4-g4x!mH5YRAx%L-Wu!LWqv_6rzX^zrF|xk0s_^pxNzg)1ojdR>DY6y zCft{)@7A(o5*D*ze0u;Fl(TKN-fZmsOI?H&Wg=q|GFX{RFw{=>qJ&`b>(0`G=euQK;D^9&5M+xJcHJr>`@>#-U4uHACKC)3- zi`=U?C#v;W3l+fJwXF9lUilQVU~3m7DUh9AW+Vu~ zWJB+v{6I*5;hED30rs=x(qGPR$3K%9$2^phRBWl*rPrIa7Q9m_feyFzVqbEQ;vLE8 z5k!Gkle((1srvn+V?EIs3~Vkcp0SR2!Ikw;}#&C^Cw_6Fqvk0PUGekt&lK>{Jd~$USEC!$PCXD=}@d8 zUm;d4h48y{?wna%{;+mCG7nl4HE7AXrmA$J>?N08SP%e&ko!;4K{Nj`hT!hpIje1w zB5tewF(3-T%#%VXV%(_}Q|+~R{Dv2p7M#crfJ_RFGWp|QmCj5)!ZxUOLH9+vUv=GO zd`wA=)UVZVcUoP$tfgVfie$}s#=5-3ruysE$>q^F8FsoixR{hHuo*8F^MH{byGIAz zIAKQkcxpE?=S~fiG9#&3+x=qZg?31AZ{@nIOOsG#G!U9)h|)%!eaidoYtc>ZW4pZ% za(O}_8x(UN^;uQ+f5+rEO&(R|Z#JR0YMhFstc)_*n>wDojzZKhbOi`#_QGK8xv7;rdw+xCSMFjFvGd-5jLlr z%&r`CM9PkrF7c*hV45Saq3U?37utQ&P&Ow)Y{P~Uny5C;hjXeoq1y^9L!WnXIghI1<*Q$ADJ{)`dJEMF|6T;r{>#OP@?0mMVzVLwIlXg|7;KH+RQ&%4h@GRna-RZ?cLmdutIXGgusXtWyr&a*lF8%(3P zpne_5l0H|bW-Txm^0tb=nT*>6m_KU0pr*Z2HTi6!)Q2Ju6e#-wTH2>*jQV8%>Y$Ym z8QvKry`IMcT3NsH?CjTOS=Z)krq0r&@6%)@!8Dc+Rxm*8<~%s^E*VdYu1zY<=IiZ$A&7egFkn%oF#|{vjvijDx_ON1R`-~)XG;lilpqmWW(|$_ zWkT3WiSM#h?f+HlVrHDVVX0NQM8Ec^XC8yR}7JWNrDazoP$uN0wBlFQ{S zH@G@kB)E|hus%XbI2yL!y(0Os@cVqXGt_9Gf2;qUeEU>YS!Jr*i@jT?P(={`WE!nq86o)<4?^Wop}!?1Pzmv_}Aj=l$GPP?^?@7Ts|Xl2i>PsVi%oK z5}u_y$r=KCXBbj^;l?z`Tp0^+s^AkP1&qt6bk3e zD4ghEN`e)(;V168L0U$yR}yZ1+bR#?yFH)R((;>;`A+{28yUK?I$gRZt2{}nUYvuf zv~9?2bhT%1mR7Puo2fZQ!pP(8F~$U7g5e_N(1q05Uxc3bIJc*q2N7xJo8Uta-6J^r z$5Y=#WcW?0lqwz;$MGJ?^*)D59XNhjA{Qm>Qb)P_Ny>Z zk-~4P(7|LX)YkCVI99cDHq}Lyx@yU-toa zUz{ik#W->I6N%!$T*lQ>(ws(hnzGjN=403jIQ#R^b%i;_^u<%~#dyKEM zg&uPTR6`xYO%d65ToQU7NQa&^dPTmYqVl`@zK8=MXXAJTS?EM_Qi$6z)Ret_LO16M zfY5x}&(0Pj7gECnVUoqY2U?F=r=7Kp-=3n#>u_JQnn@QogsJ+{KafjBIQbyq!R&@P z2nsuz*ZXVim2nphGkThoeY$yq;7KqBu90mSIYYVO!|Psl--`08q_*c~rsMSPDMNV#JJ2_Yc8sdO|d$TnXdW&8N=g~Mt& z?%~bo9{-(*Mw0QQ|41b8R>K|s!3I;NjDCIx)PyU~Q<*HhG(7!CV~V?|{lyjpTUh^^ z@n~J&s>a{y^@Gjo>ZU-@-_8@C2S8v&<$WTM0_3jG(c6+UZ}kg!v`d+Z%kFn)fWT^Y z3&kgUdU@!$@eV`-Hv2i8{{{XLjR^j4!WZLz$ov0~d1he!A1t{wT3SwejArG5`fK6%{3@RM5iP z3p;yn>b%H$qaUK<9KYy)a_C|o=C|kR#j-<6mT^as7OmP|+Y&~Cv^Frem<2^BCtgxjOjDoAi zgiLom25v|HdVCM=II}fr{{zEx4xPCc6*>4a$7p8TSC<@9HH$G|cl`AI@a2k=uNt*u z$eZJhraLxRu6bFF+fDJ;$ERhqeDUNQ-za&}Lsf=LvqSK9xL-2ZESk+sI%_kJNAKMS zP}u`>sy8ZQl^jOhe~Nzz9ShE~W@_VpTtS%I=h5STqKMx%Xdsj?r{3ld#7{TOWL@np zimJd}h6IW0F^fXE9_WHN-^M0*oOX(0`Rd^}ew=vb8A@JLbxaQ<77~r^6<4DiIjd6L zOrq>it<~!dIiqyC4Eg2VU^VjQf!m~%1K zD8NmAv}M_wiy&9&m{nuQaJVe-T^lX-g!^(IwHp>a0P}iB(CasBtoJgz@q<{pfjyd4 z0RIv24izU9?&Zs3j;qSCr3# zb_`d|>()?~*c3!FuTY}5xZ%a`?A}m6^>>~xEmVH8iw=0G5k@#>^yWMlIt`Q8)fH-w zYs-VG6PSet-y}XoxZZ~(KF09{F8f-$sMqRF7%^nbe^}SF-BpWRlMIq1H#Y zldTIJ-PFK*BIxJn(eY4;|DLC30IGdwWcNUG^H+u20VTw02Ch>nt=VEdPcvbqw(|EU zxD-Wq2`f|}aO&4hzAjXc6?A?z3d2f^RG=20$Fx^qP#J>ea6@2}zF{I&N&gCR-E~Kh zO>MpvL?|B*$PX?2jR%R?;|Jl4*}de?-tRr<=h%O)q}e!EFEpBqoNfS&QEart%VN}fDVC!xTh7N(E>bA~wx0b7VEN%G8LgI0A- z#g%0y6jeu55e-jh?C0VDveeQsEMSD@irRSkA_#E1TNeHq6Y);ubtQx#MX$Y zPf4BpmGJD9z}yGHBv5}Y`ES0q0C9&T0+ku<3khm`>N#g0lJMN)T^bLI;6hQSJ3>OA z2~&}X;w^Vz7O{Y;5(5Rjqrdt*-FTjtJ=t97TM*a*TH0~Pdh4Pi5L?eqvuF!QASzB= zQU=NBB6x<5>|Ko%DE+ys{29!RSlHgBqvN{W+@-qMnfEf>zE{gL*084v_t zK%$d;Q!CjZyX-xGfFvl)g21Hye7NKt0Oss89+gZFK869o)qz{S;yr}KaD;oXe8g;E zn0&++MDRDa{dmnx@g>pmruokqp6&n^Yk$Iye@srz$#J1IqDl?+ZEA; zX%3kR#Q9h)Ad~6r9$xrj`NZyp6XCiR3H5P)&R!2fD^wMO$HVzX~>4+ zCZQN~)3#9fo7 zf*dji_*z7;_R!Jxzmlb#DRSMDi@GQ+ef-!?M9rSRUT+Ks=j)$~fB!{=ikiC-ScYrv z{6JbmgfF5ce)x7nz`~IcMX!Z{?S_Ry#ir&no>)t)K@hs2LEeaDq-VMUvwRk)`+X~? z=vAJtE&=N^UO6bF{FM>E<|V^$drb7SgtAW9IZ8&!aUCRqh5-)brbhD*A0&iF0t1IM zW=Y9Ai4T=_wxyr}=t^?7ThtCEQ#&`i3Q~d5S)FlnfLsO$Sj!j1ngfqDIoSX#L@b8Bb*c3$C}A zlwpN6UQ1|692S>$B!^M4Onf!UGmv%cHk&SVumMJctzg8fx;nq zoIa69(mCtp==j-%x84Yr1J~=brP>>ZH9lyAK0-~$t@XSImLnL`dg~}hjkeb_S_n6l zXMN&RQ*Au%-;z&{#(oUnyw4|3$ufos&@U3iOq|?ARWPR8588OzkaCrHr|ra0D>bhh*y_Py&Y{`(Z*aByd9#;A>$)c=@V@2i)}haAE^h8ENLef7UjY zyiHafs2$K%u=bnkw+DW3F)aoKF<`c5Kx5l`_%2=w4+tVu=N6Uy1^Jr10^|^uq^!B( z`GuU(A1|6k-n+5efU40m@`U=+s+3wd;~%$NX%m@l8%Q~$%M4#^RD0DHEl_Si@GZfU$$A%APge@u3N;bEFf`S8M3lnKGO_` z)gmZ)`>EbxLph^ijKgnC2y{?7*C85gwF<_FfNmJbuZ|0KyR3$e>MTE1#;lnw>O#RrMTWaMmaxinEcD!g)~Q9gIBwGJ`h)r zm;NYE0pJU$T8|B_oARpHS6L1N{bSSmLl=pOVaczWAuYMdnUG{64qt7A(cr!3vRDen zJrd0gJ~;FcaaxegQ167$qma9{aY-Z%e3WZr7j>rHV0PyQwJIm zTjLrq=rsdN^K+V=e+GVwDW`1PLG<8sGGxrvAgc1^Y&GA7*#)g^RhElkK-8n`LQ zB7Z64z^9g{+%HrJO5Ay4m2n^U2icDZKhlHXBY!h_04UFIn?K0rpBbUyFu7Kf3{q9r zbBe`WHr*$+ z3*`|>+7$NdAl?ngu2h=NRa>}R!dEG7*jnkBNe+_AuN)(4RB7cq*YK*|wyu!$X?Ud# zwO36`#Z+Ajw7u-0lJvxx-<69*8vdO#NwNlE#kpB>KT)MJK!s>HWe2f|yJGdxk!=vT zO1+LS7R{H!6|K_{;L_r{LwyHV7K2nVhYqG^Z2(P|vG>=R=741_{Q>3DVj6(BiLIpK#TJEWdW@ytFBMfW8! zI*i_R|dGw?)axpS(HL;m`?`{AG|UnDjuF~gbq;i zQ$9E-cep4>yqCZ~0QQ63I6S#$KJ9qG&5g?6mgX)SH`cniquB6CYC}flBQDtpXH`@wZ%c=H0F1*A$$Lgc3!$acUd2<{mY%<$JgqpF;E z_RA>yh=K~?^1wV|O!@qiQ$nRPqvE1Z1+0jZopU)`+^}G!7jd2FLuPTGccEO>V z>>kt&XJNSkwGAsYB?95!39W_WycK~5vr-;wIdfi=JamVF&8ZX{FFEb2QFG{wlC4@@ zSgY_j(E5V6oYOezDoKL_f@!B~=e(5!GF&C*x&&1Hw@DPL%<5ARupq+ESWTTP&(4(Z zE8Dg?yPzKdaC_oZl6hvQbmiBo46#1@9PrknR~+3F^I<%wAI=mN$lelhG6}h!GIN7$ zzJn3@#ZGmWSVfQvCl?yuljgt?@0g@!Gj&jGDpdl~%l0u)XguM>OAoEJ;ZQl%Z#6vp z)K>JWL`ld}l|OwlQH$9>^O8KVv`a(UfBYV-xpAF&rF8bQa$S(UUVQeT_O^j`f%NXI z)_^tsh;tS>yKax>7rqvnryZLvrffYG-9r_d#2olQ$(|Eq3{`_4o*H7dG3JNNM})(D z)}($vW-;l2AW(d+`AcpRR_QE5Ge6R`XgN0%qg?q#<<+eggx3ZMC9*>t0^;LaaG&1R zD&mt7=J1};5-ef&I6Vj}BHuredqPQLOxI2k-m{M0HJFYn0leYC$GXNTLUD1ztSSb( z@q*v?^YsNa$DiT|bp&zDEt>L3g~m`Y`Q`J8tgQC^XoujRB{|uW;1J34%I8js=pBq6LO6Gv%r3n*s2aGc@@iM@DObG- zmp}mmeGd?lTXP`qm_0jW)mrR5!s3H_3hk;%^i8Q2^yG zxCs`s-hr*D+%VB^9G+gw{1CqyFhp*|BVsQ{3>^>cL0}|8q|7W zs1A5gXhS$tFC_Fb0O$;pl)x-Y&VdeUSHe~~eKFj^hBP=t1c6=}y}%*}Q8m+?lPK*k z+=YzFFHKfi-5;L5VgdFVf|o*G>F*gZ*G7QN#@%sES4Xq96mRYiGkxb2C{3;I@15f} ztclhK{CjL#FV`Ri-nz0?F9jMcxbkY%fsQ)3Q@0hDnPlQ#lNKDB#{64uJ+tk_NdBC< zD?NQ)tTOGP!Ec&10Jd~p04r*gf<^z)I_lU@BpYQ3T%f23NSSWE|H`Jz;cDFJZ$QU< z5ab54j>ttV(j+OPaNb$ZAzYR~mPM@vo4w_tbCY+eAeJk&5&zd2f;pwnHmIxGSF`xt z-&SX^bn&+FS07jqCU(NBUPk%GQ%ad9dnl07+RzkC1~##gN4ebgbFjlyJr_n;kg8xU zDa=q#FLLO^(Or&RF13kP{^N7djoZUC7ze3}hnx$mZdn{X)|2CPbkE>yM4-*iL8G?9 zQ8cmDITx8rIIw4ZrjmiH7{s}NEE0)+b)!3JGisXV@@BKoP(bC}ToDMhfleIYDZRcyeuv$}LDPsV2Xk`*)tp~@M#!4?gdRktwL4V zb(Kk?y}%wSPgHO5m?fw+i7^y`gyqi1hxfQ2{+OwWs&+g#m;jnHd;GOA34|#!1}NgL zVUJ5n+&)b55&-U9Z02Ri=r;@seGeQ-C)bXx>e)>$ZzmvYX9qw}mj@TPC0MzHip(7o z_;09KGyDHx>>PYMfgU|u+qP}nHl~<%YTLGL+qU_wF}2&NZQI)JCU5hS{U!VUiJRQd zJ@*_%t*%%-hB9Wzcljr4#0z0fQ0&CDP0%+-b^RQg4@Kg{Q{RwY{7+0V?60l=Gz9+l zTG)RPaNL|6|9|^ajh3v_hQtp6H=Tv(vwnun8SeS}RAR1KE+g3_!75HP)81b`Jb1HC z*iE+d`!<`>a3(8mpeT#zP=^|ixG11fq~NFEemz0+WtD8^n7r?I`n~WfEYQjSvK&wh zD0=8U^?vv`+c}B(SClSZwsCUy3H%&igv5%Tg5!b7YIAvF=}R=dmF?r?%GdJ-&-1EF zNZ*8%9p?L+6LOE6^S|;CPEl=zIJ9}gKlB^H_x;)7mc79JeC48s_O3lg(;sl|%oDm9 zljetmW{<+9a<}iAml5~E77V-bC3g`<7{iWQj)}Ki0^<(!>!I667ad=$pDD1SNL4{~ zxU_2*gBjvBfa25|&7`|FHOg)ORt)DfnM-$Ri4OjUetTWsL`=|7Xe&dj!eR#thrp&C1*Ey(ob~2lA7Luq(o1!4i%AKY56?lN$H)L%|VL zSkb>H!_TYkhIjFYOy|znL6{>^xG&|7z^I(#18Wt&6KUTnjQ_G87KI?`BZuYB;3WY0 zrOMgAqq69*mCm-n)NEPMwT-R_F=n?0MUMYy|6b@^ORi6T_=kgmK@gY}z#Hz-@)Lci z1~llRM}e$bB8mO0e2#mfOF*O6!p7!{i z*SkhV?m=z4vVBGWR%fe3b1SqbO!b{Z>4D{Z`^KrVKTCM=+~?S$$w4@&kqrZ6&_WMtuRM+~qKVX44p?&-&Rh z&-mJsNW@7q&m-Pp^iV1z+09f|yhzr&>89ghZ?&3JgM3_T$r6jD>57t)c8`XQe}rG% z1RN=C88h;@)o?>&Sa3t(XKT35aZ~uAi6h1^SHi-$r0$*(1~4{5Xj?;@_@TfXq;lwl zRZ}`R;5X)}BwLcq!@&cj5DlR_**}5{hMEi)Y$gBAm%lrC=0G@N&bjC>AV(SJxlOy> z+SCp%uQ+n|Eh341m)RQphw4CJNXCl50f{oL2)-_1$Z;V$NVvy{QM#kvR|RELc)RSOrt)tF#nV;a)PR8#dru4>$Dc`U4~mb zt-Q<~8ggvj1gx3!=R!u3A-c~IXKrPR+*gq)ePcVkT>-%bS4Zc>m?}uHKutc9VeP6J z(}XOj>p|r?)rKID>;~wQ;DQ+LqhSwVqp}goL;-eCL!0^3^q5U>J%f8%K(*tzszCT+ zk3sO+d`T=Ew0>^}-20Cl$f$Orez%b3Vf)NCnOOQZtyzbL+Z^;+$kfP9FwW zhy@`~eoeMkd&I|K?5E0!GCh_*2EG?=rtus72;`s)Z=m5xV>0f&Oa+g|fS$I|s^o_) zeys^b=iGtU`ec1_cw_$lFy;pB1)B1XtEQSi@bf1=TFQg}&64;!V98#kM4|Hgg{g3+ zVuJkTt=dELgsRbo%W>hWf(L}cjTs#z$%t6WEACR>jDR0I!PZC#=!(-ZVjM(P!45W|Y!>Py2v-AU*y|8FWoLn!R{;A+>#-BMk; zZNb(`joDq!yk4FTxejMak90Wz7NqlJB_^wQm{Qwer!a#-F&`Vv%nJnfW@YFj!?`+# z?5&RGW^$8NU>2U0kcrQFImxIEqfE5KM{F+aWs>#Kh9#+6b*Da|#iJ0{P6jwtcBB>tR=B8(0QZP!>$1 z%3~vdisuPzA1Pa{6(l%A8J0d(xT;;j9}vALPX2hEhmFJ1^nxv6zNr+9ZRNP&7-#t< zP&z~mD;o0yZG?k^03)Lix|SigFYT%E+TA3L!oH#stKq$K&q+;L;5ANV?+H%w_!`F; zF=3QOHFL@crv*8oaLj*MXH^#B=&Tje;n}~hCz46}jO5t>0UxnEQAztwN;IP3KnH58 zzC#`ubA<6AB3oo&=}A|Hv#i1)5*rheg`3A2uxE8(06h1{%ANMoH-8M7Dp8a0NPWwWq7U zAi+4bu+a*C6= zDJPy$(?5&L*K=#<))QemuG_L~bX*b78-EXjmbc5wm$X-$C7S>BOusySqG^C>+xuj zH8g`v65VoA&h9p;fU)f;O{|0Gemx}TC#|6BVLUIK( z|L6f50X?DHv*r%6U_SfTdFj|Yfkly&$g1u(Ii~WI4^=3S{p*sAA}{0JsIV1qg`paX z^LI*qkWZx$(;p^0F=uVhw!ZB-H$A)}XRiF&bGR(lepzD}^EJ%GiSD5Gl`IcH; z4cBAJ@00ywP&HJg_a{f@aj3c6F6^775^|gDgu-&&vL3^A5C)q- zY&}Ss$;Q+>QnloGMSPnFVItLL_!%n-uPuev{@b;6)vUhVz5f_QN1I@*R~?p>f9CgU za+?+nl_vvl(uQ+Dn-094#o|EOn#nzzy%z>Z!1=a}hW`BeJm<*~sm(*e_uduf3csg~k% zlKAGM?yiXqKHS}Y>8m-^jSiiOWPkGbN+S}JF&0uA2*S;VjN$^^)4b3_i5?ztnqpW& z%4NlBK{Hq{^BCp;Q7Yx}`dOQWsabMr3Wr#@B8L69amzyABh%MFmoa7SS34iUNS~JP zZ*#up+;&@&MgFb*K{>N~=Se!rA6f;i3OAP1Wg3XgKFWMxHFfU_<&|`4{M1|xu+CZ2 zlgVMmBXTsj^;$$QX_aM3=~GM~0twWp%QjCfFr4K9HJlj7qLjg7Hs;w;5q0`BBUtr- zWe@A7e8S8^xF$X~@8gImPLtO$a-RFoH&NK~GV?PSx|1UZsMrV_hwAxZP_Y*^gNaRYzSzYLulkuT?#%%y09V;aJ zu^8aVeC$pbx8`n$we8?{HzpfT*<1y~vB{D3jMVdUYsurIYbc8{HD99qrOD4ROO;PCQ_~mPg6cJ#Lt1nVAU)26FVEIaFQa((qJ%boH9BDu zvJn|-IQ~{FwkS<7FEF9sbuYDR5YR%I)X9{_N7wt2!{z1S=*^Mvlt7Q63Yv28F+!D+ZBcXZj#=?i<77+&rb^wznzG%Qz@4S=2psh+VAOBeZ%HwUTZ2UI=6J{&iT@{Wl2X4Pbpeo z-&;2HeBdTiJIwgyb$y;De6NtN-<{-|iahkw%K0>qIDRFpbQ$&=6kb)0(rnmr24HO@ zqaICMUi9z^Y@@4u+gX+8zAa!^7#~%y>I`btzdrF~F&s;x=P8E80Uy*vLMN(-T>(Yh zy_GWn4*(gUka#cv{n@6rtIc;iIX~QQdc%ZKt^d<>_n*?PTpa&{ z-ocdCjMI91V$XZ-I)SrRIZa_i9+;!k1=+q__HENgqEAW}C_O6%CCX_5m09Aq>vw3) z+5)e&8z>QV#*;B8x3&C|2AMCCkl*H~3+LSknfwdc@b|#>@EcLi2HuB5-{~dUhh{b9 z3?m`}V{vx-5ViqDS{}j<2RmvO>7DsWwSVu68Zr%IWaw*y-HEof}p%{9mbD4J#Hb zHxHk_+;@auh@R`@j3ILUG+Mw|vzvOo^Hm03x^h+0)!AhU&22#wCT`xrKqI4V zxaE2`+r+I|stBJpQ(p5oVc7!q?+qwd=k??&CiJ%GGR1eu1Q>BIPX~{iHqi7jg`U3C zULOC=pO=Uhh3KqmMCVE!Lz8Dy7bZU?2|O~8C}vA4QZv-_AVro8ihtb6bC6H{XMPH1 zIHk@EM!gjFV4?d4O+}Uq9O516FzMqC!>Qu;Xo{o#yCxSl7chgloQHXbuA()mfAlDu z?km)0Y0!&u1G_8R{9JX|CaSDF5fMS&_U-skr?!%pT;>)pS)k{S9P4ol8iz6waPK`` zeyelZ7gE=T)Z87F_EBap*9*43x+JyL!+y(1rP^^E%C|67QjZbsSC?E*f?1>G^X^;h zHQz66b386pYS`(wiHX%G(ZD_#u+6;lL`~}L{+&rIjWZ*RT2$kG{qAz+uCR;ECgLrt}=?y`6;NLq4ujr9k z;ysS-t$waMLcnTMWJwBFtBlZlXu`PV6GN~mS*Dx@Nji+ z4<24_llIM0+@;mm*!{qmAgRd8B19?$q5qd37^DbcAOj9r7swRiqMW|^4QtT&YP02y zfML|m3``MbW+>95~0Z z(+{Z^4U&wrW&8|g4hiyL3sFOdsE<|E?hh4rUNV-5lD&M+&6xU)6>S8!nL zfGEUK7zeo%%6{4f3swQ4GioQOGM=+7|=9>i13>K^E{@n69e$Rn!7!Yo|~L zNV;<~>S`czhgm~?YLba3yt1S)qWbNsCF1`1K!{N>^7>N3S&a&M?{PRmY&LbEf+-eX zMWan2Bjrm~(+JOm)GeS(34{Ffpwx=VhUdkZxrlCrUS<9`U;*O>*0Hmh{ZMt4F63dy zU_#8H>sYtf96$P%u1#IPVaY4(-Ljqv@#0xf`?WVq4n4zBqzz1WeZkR&*T5AsOXE15 zI!zYY;9tD{JkJ^gpvkhy4p+(&7RmOTN4|jc!LMTBLK@!Ufv9LW+ypIaXM<}8o%hDW zosLH=tN8AgAD9RzllB=*p&gVBQV0NKj?|^l*LOwQSKLoN*tURP*)PIu8!iIyg}GIm92=t_K0U!Y;tK^}X7K zY0zur(X9`s1nLPH3_KrS0E71h@GNvlw(>BlH#Box+eALS`BQ8g{hCwGTtbT4wg^%Y zMEuU-hqtx>4X_uwnWT?Yzo!XyJK~S$>jTmTFpkm3&77NEBKpT-gCbFLQ3VdKNp*V0LQTOB|Rs3eDwA6^?_Z?LILx||e5 zO)|d|jIR_{$b|(YWt&xP8ggb%wSOqnOTvCGgUX*`cLk=5O&W4aa~xDIsAH{wm<_r< z)oqiki0N^&Q6IC$E%|y7(|+>&g}t_Sgo8v=M$Jxu#%@z-KBq$TGUNTz+z5!U%Eo?o zWd&P`gealMA1c$Erek{>-!Zr}c!T;*0LTotQft1XMDD;6h6tB5V^?_t7au(}=R ztAt*pv9n<*`a8Y{n1h}NwF*4`WQ=py(GO^%@x`Ioa%eJY?lGn6ZYus!# zh&RdmT6!?&qUOxYDx@bV6hqve0hJQ4)aM06&!%!!S;R?Ahr@f-^S$bsrv+JIugq zR&STc+$QpYzI;JN8d*ngA|~yb%~I&T;T?|*e!gWQ3XhISFdk|ep=LSM3dlV%CB*g% z)LwST?__a==7I;t*e@Hloevl5ghu)rGE{KkJLVR(5h2!sAVgHeLt|E>d5qJKVN(%A znm=WeC+otYBM)|_IXr-U_DJj;N$SGh*p;rh(r6i%YVaYWe1>wlpzVA?oRf+`7es&) z@1`RdEfqprGh2yE^Uy?3PEa!T-A-P*ST3v>&oa>sSq8%UmHvW+CP6DrL2Mq~;6>9U zYSB*K^Ndsi%oi`hFJr}J@A>HdFVLu!A1AAgu`Do_FR1<6I(BF#QTlpMz~sFUf_>mt zkl>|)jro&O3I$s@$7&X27MM1dYp(k@uf=FyIUEdCp@J|jnWKS-{vxrt*==tvTrtUMX&Nb6#JxS8*v?-L^LTQf8$%gjzWNeFl|H-FB+uPD%n6O&rHHe$Hs=q)XX7ZLSg)D68cS;-X@udh>(S_&-ibw9XDF(Bf8E%7D#4=eLz-0h zOX3t<7_H?~Yrt`1o`drm3R&0+*^G8@4a~gTM#FGp+C}$LvWJTo=~j-ow3KIAk>Q0R zP!UZ_nz?`X>lxqtBD`NEavuw0dc*$0+$ftjB$>OzQ3P(dd7kvr1!`MW+`%L8eZ3wI5`+xqsNBvrA>`%1AXl zQX~L(1>eN-13x#t5}4o}*)%)QU=I(+Tt>-{CdYc{b+LKtY=1SqFE zjZ@dilx87&kO4{(XF>8SQ7_W(#5J;pk>6FFu#?w(r3Rt``!buNnK%F~R|RTFf-*_l z31513S`y@)6*&B~X{hI1d?iY|s#8v;K+e~x*71>m$)@0}X{gBw7srHN_;vV2w?=`cUYVUWxgy(<=`=`?tr|HJH zz-WCX0jH48jqbP9@YKAAv?OM#xJSmbp6duIUJ?`R7uLdNw)09PWiY7eRaRXlkYEHmg?*ZM5Otjo&aq^IDJgAn+!&%X4eyu zG)oQWH){sML&;t`J~V;7$@QRIuCsZ28IrIBd(tXzuok{h`uUaD1Vgbvy7*S_`L@Qq zBXXTO73{!;RV8R>rpY`827db^di+86ZR5D7uh;2t&c8K&EL4v4-~sXq>_Au$W_0gCL{B z?j!{vv8Ar+0&~3%)S4VTU4wzHYH1U&^Y$9E;2i9@&?iROG8qkW8zDUX>* zOP+||-rURt6(rH|XdC9E!BR%`RMK?iz@4C>Bf>UfH=&DOJDY7n!7ItV1`Qv;K~h!+d5cBiQy z3VF+bTY1pIkd$2Ua~pNg>ITf?v5_uLwrQ^e!hax9WaI)ajmw*i=sUfwTa$FbH2^4F zD*`=@oJfG-DkUl$J&FH1&9M5ReD;S0V(P+vkxHIZDSosvl>Ut)+v5!;RRJ`qOs8Iqby+v>>3g7Z_^2K?1R(aR^iGCYXwfemdD@<+?*aE* z)5{*^M*Rzox!ttCtGplehB*5WQEjG@-6~%@0iJQ>p_y+l?=WUT+=^$VU8gt=gn!5Pz-dxC3M-CVYk0eknd1r z;bZohm-pwO6nsRmrlH^ym5WX>@o#rq=HpJ)hl<4pg+@4MMRJT-sRVNpbd>!!Dz>tl z`NbH}>wcs6X3qo~`}AC>-1Wz;Idf zzxTAiKOLZ}g`S#un=mbk=+}_`mqzl5NW#k%WxC<4zke{U?HVKYr90p6nK!Jez}?D&Fy8@uScfX0|Pm~PQT;xpqHDoon_E5H$;Hn z(fJ_dF=U*smCp%jUirbz#~o1@F`^0~b=_ittaot@Kc609`0DxU!s^_?2WHoS#I-qg ziN)T{*6W@SA2fV3ev1@|l}@$j4;GhBh55Qj>%jc_`te$GrF~GWreoxwu(>*FnKC|D z5xb}1`BMUVqx^#(d)13Bs8IUH>XxTBSMF7`qCB&(<_y;tX;{4rRT2}^<;#F2=2pe4 z0!7)_rU#0?OK}*Iz}T>R^t`lu+%0KL6(@tr!C42hs+Kkb23v6BOR0f{uPo_4=o)cW z#q$OEpP#G$sziVOKZyuIZNin6H>#z$%c+$3M1_|%@TIQ zTG&;2lqyjI`&`H)bH}}rD!Y3A+^xrMkB*B!duF@pNb@o6t6=%TWK=DUCs z^&6*fu2IPeFBc76EXtrQB0crO%3*ZMEH79d&Wex(ps7r!4STcmCG5GhSVY}fdhIvY zGN`|1+RlPl+%}k}eRNbL_H~2Kl0g{yyiB{@dq_Zyn$N`)Q-_zi+!_=%6j!K##rrnz zyT~L#eU2fFD&N4KO3>5Q8rr=<=U7`}$imzb+3jSn{U>fHQq5z*bA^wtY`&l?}pU`y4ULMipOzjBf-JT$o@j^ zq*p~U`J3Hra__SATi+Z~%^Jpy@7}!+q-r;))icC?Sn7UwD>PDhC+Fne9I{QUG;z_{yg=UtKj0`uz&Nlu*fKm}tm*%%EW+XFO6AB$8FYL9L?5+6g4cEk1? z3#fY=Ng%4s%<@gh9R2>~HD_w_k8`lIdq*G;^5Ke|kE0*;71!KLpxI?lzF}1Ty&|b; zE<7-Ax`*E@8o^{Tp8NIEQ*CK7(Hjp5Kv$c(P^o+1k`^J%uI)J?IF45Xf;f+2Ex;8a zE2hR+b(Qw35Fkc@4%^=G*D>i_2h5FmuM`kAqibf9wQSNZ9PO!-%vbwGp6g{EQ?2op zIJ27f+^1=fl0Ba{CCCGAb{XTyCLB2Eo>J}y+6{VTNHE|s#Rp3`Bm6m$nmhX@(;v(( zPfR5CX>_~Ge3C0N&l<7T>cMMQShPChrNbZ4`w#`<2TKQLNu==C^(xyYpU-b}nJiHU z=Hw3|-L)|}=y4pO$zxo0QZbkcm4H|g^(Sze2*_g}7fn8A2c&O*mrB3tmekwS}U zsvXs)9^J!U^hz&e7_Wtno0seMilO*e&jXW^ErbSz^ zbj5)SphyNSl|_G=*}Xyrt_gEOiI+hQdEox)eYpXcR(QMqRXd(Ai4UeXbyfbF`%(j$ zV($6!bejOW2ZH8%k;0o`=Td8WBD5`4Jacylh+#pctaKH2%?@*0(q-aa~u;7;~QHzUVFlA#}4;N->=SP9FQYuMFBmMw?* z^QM@%Z-;*$@Nqs3fG`69)c-(RR;on`2F`+0?9@{|T1BD{o`Pp%E0saVx%K%qU=xDu zojH-))qjUSUpj8WwWSy8qkT-9a4@QiYaM(#k)`Q6>yO@BLv?;u_uv-`vA(vg+e=Ia zbS~;Kf}4_?1oVRk5h?9?v{V3aW0^%<55k1LE%3W&3W7x)xm(rgIEV$EYm&J#5HHGU zsvdpF;Goq;EbIJNJcAKX*03M@G#3Ey))h>V!u@Ab6kg_TJdPNyu*Vq(Eba?^XmMwq zqwCRD^Z_$HI4$09Z5ImsCARunR=VhN(6{uz&Q}iM^R>uq6 z^^58@&-{+_O}h>x)0EWkn8H*5s;y_AV%N5~>Yh;yH9_NzRo8-HR-mpv?@_jOCKg9H zTe(|YC%UT{ z@7384hNV-#t3YCTB{5@@JCpZvUT)1H+UIhD_m><(AeQOnXXfVhJ#C(F2YC=V;d?c8 z^bl*cmM%*#_|}K}DM9kU5JYbZ0_PLEqM@Q$)vNFBDy|k+Q+J=iL)1BbC+Lx zd(lUUHkXlxDjTT-@zbZJ9oW{7ZIMzK!>JjU)7O!Na=)RM~}L(Cg)Q3a4(c}ia&$ye#tNDfG$ezpL^z{ z`ACG-Rz5A01EnMo@PHVa*h7$Ax9;GHhDe1BjoC&taMgzv_#$P-4e@O@YU?8y=H(i_ zocHT+RJEq7BS}NE4<2X&yrC4t47+K?G@O2$e_FAl$p~=BMn3h)*XAc5XN$~@AGD+n`jgU`jTEH;&zsomU{FR}uVaGNkImB=SR*u(4_dbfG zaYh5yE-<%t&|&{Sl?`B+#IL&6v1D3pE+R@Qda#R();a!n@5Xk}9)P0= zFGv(92&zp`<1*mb$Q`W4&^23=SJ7}p$%cCr2+>8>w#DVBk0a85VF^LTP38n{bu`?b z=!5l0O`oH>^J`PWNB5mX{Bl`;bhG;(nsvt;PeD0q*k!6$QrJjZPH5lTs*a_?MXfJV zcpy?T0VRyocCX1LzmnmUw_H8U72czJS2FnP zB9hP0T`>8z4tF|UAUtu|L&CH$kBnnF%B4r}aL5yYeZy{xPs7~YyC&TiTkw2mr_R73 zuZ5`Oft80ZSsJEd>QJmRZ_PL!I_$n14Abx<1CT z3MgR7C}l9H28%9bAe(sZ!q!ZCt%T<6SaV%vKaLisH6p&oXeR9+c7R9Z8hC_JJ`G6M}0#f*j^H6`?b%12zy zj56z4|7~j8uhTB1B)1~!rRjW;>$K&|fDkTjN~@$>)D#)+Z`coND}Yz9V$1R)Nni|P zvQ3Y?hqjX=qYgI%`1q*1Hk4c!pKhh@-%F!Z53Xr3Kep3AW^VpNHwPsYhi>9Cyo6ba z!0@;O8GFc`j6R~k=dP7SZQY5E+}u2K+>D{A($R^ZQRKeu`je`9^%s74xE3iUJ$wym zuN@R$Bpu%3$$LOg=%zO6RVj@U@p#|JIq{q|L52jFhgGR-U+wlh)JnH>pKL^JW$waO zdEjz&Qq_aUfzl0zjL1r(`G|299KFZFjPg*&#Nl58^aFREey|#lGTwB+lrBl5@@&9u z+zdz-3F@SI6qIJerIjj=wJ@HOLCrSg-oV;cx9r(=%=t_z6eXeT1@^sV)6pbelW1iz z;BTRu8IG+JEv7|fo2$ioDj%h{Cf->eo@dm@u9`sr06ET(3FPhm1!H8%5Vwnuo)}eTiyErP?XkHRQOiqwJaiEeQ%P#CZ-2BA|r*w1t&>P?mG8p zX*--eJ3+14FPmJh64$E$i2OI+J@cG}QUz(wx4wTfrc}GG-Pb5zG}(^9Ip7j2pE+0niVhci;pt)k)UALbSG0k<*ZTMgN8|EBZ2r-~ zHI$JHH_ME!t*wgBYuEJ~Y287-ARuNpJEGf%;VVyhNttkBO8}vRe?kld?5oCQKteY- zDXtstRa=3W@>nov1mn9a&^V{~@+_sL*Zc}u_OS@e5k5StXH@gk_WHEOkAe(vj0cQg}KvH<*N!sf4c zxT}{%Uz0e!`bVHPba2fzR4DMgjY_7WUNG`;zzm?hct-F*-81F=(2 zXZ}ya9Q*$(t7GHj{2$2by0v88eq?o9R&{_ex_0Yn=v`kyK3YNJ5otUty2EL?Jxw7p zG7<`C1fkW0@2gK-H;1Z9fN*>un;A=nm!l&|A>m1F?R8aMWrClaJ23^`xAb>To#M;E z(E%Z4V6MXa&mK2+KaoUvost<$qG3UqwoK^|oc^LTN_051szhZ(k#>&ybo7~(W|u3^ z=ocVt%Onvo;B~IQakv-EABUfe{ZJpm99y-rZ>FwTLK}#fs3ZUF`+7O&Tx<x|pG} z=UmDpr*OGmRK~5QTEZEemU?xNT)U&2w_CPX|S-Zu4iO-IJ% zlR3@xp(-pjapw4&-Y^b{?_DganWDb(^33$!fF}oH-k?!{c_VnB>TB%mV6UGQ#h#=dt18S4O?B763AK#t z1NkP}*Yrn1^(~Ntou1@0(L=9qgQ1I$=Oyb0`-1y4>tfO}5Tb<$i+NGjqg>xB*Bvr9 z-stGWdZJ>BQIW)X5ysb)X01Hj+Yx`EhKHx2wFGFSori{H-w?hDL1K9yWwj5&qCljw9dsR0!3mdR-ZYm1>xTwCQ@JW90od|S!ke=Slbr!vod{%%a~n7bJ8IG zom8M^i#Z4IIHc)vRx4Xr^o^a8&w^lRpg9NOFp_S@XxLMw?6}gG+RpclKIr@ft3iy- z1!F;Kh@ciRg7G6j_JHaRO=_3b)0_9?0SujO7hWfHmnMd~g$Brj3fChE$PB#yGu-Yj z>AU|-K5kW?>!Bx%Z^Qfc(>@hbSLpJoZ^ZR%L5mT4PVJ!-*|Ror%Oyk5serR6<)zIy{GM9=E4TrN+%79S-@t^vv z7bM-u@yE`sjc9s!9$eyKDC@xQzM(K`{rPMdvvP6dC}}3F!gSfE_ucVI5nSd52DA?- zjX``hDn!rc52ZU58c1BGFV!z80UvG@ zHq#NOQTQBUVEspU@#Kw|nrF^9v%8o$Rl$SnqpkjOw-;cmFl#3Iz&Ohma~p;q1YaR&MLxWCO>&RWfAUEVl){mrXHS0HFI*Rz zDM{eq%P+&wj&KLdb%@%WjT&nOCj7<76G+Aizn?~bPaM25!k219y&3joxE2>D?%=49 zoE2G0U(dsO2psGU*BO^6dq)I={M6>Kw5#GDdhlifKctp^9}f|TbXfzv9n~9-m9q4O zM_i=p{)U4i~j1q+ZYjAU`Np`6qp&_Kwwgb!XWpJ>x4b z{P5_z5x4!g7>ts1f6~qhte~GQZT?swX@!;l2u>$xu4t6E$xw;^)eI)qsPiEl4vi+U z0(h$2h=%2?f<1O=-Kp|y%6CHJLxR{Y*&l?==AV!fe$mHt-GXgvq+f2IV0a4gX(9y3 z?GQB!&;N09zhkVzFoxuYnT2LFF8h#uJam-iSe zwbftG6Uqg8Yr;R}2T$Nz5T?!L?Ib$W;TjoJC@Tcj^wD5$4g(%ofEb2pm)kwuA|nEf zQuYwZ#+1>Fr;!@38dCA2)lTyeCS8v9VX{e=&9rL83(6 znl0P5ZQHh8w`|+CZQHhO+jiZujlO?(4?5z#h}V-G<}lC6wZHu><0XR9oCM@`CgZ5& zh^GM9{vQybw3O{4OkeGOCWQKa@1b|`l>(q5N@Do$)3)#-mlaw-M=|88lqkoj8?ppK zL1e<47@~xh@-55WEw}5+Mo`HRF|h0UFJ()>_dZ&;{o5=Slyu}yF?kp)Q3^x^W3<$n zqbU2Yawkn5Wv_fm()h%K2Rvj6-(ir+VegQ1e@AX_T9%pKOYHy{r9*H`Dl+8+BiZG2 z`cK~n#!>)v*8x*6JO#i~=<&|i4T2WXl!rEY%JzbwgHMS|CHRD9TTA!%$U;g1D1SW4 z;kQ9fooKa;x^zCJ*7*fTl)wgn-z;VT~Is&N7j{V{A$c>HYg1fwuzL zd&@hGKV0OWik>5Xb^VVECP1L*k3FNM^)hR3YaRf-91mXKzxAi1wnX1un9Q|4`}3={ z^mRZ@JLT=|afGIL2Uyp-=13=87|end!cDizmfO0>NziD6bl_eZLq`PJ!d3MB`KxSaz8-eJ1{zEx&2FEO(}qt$?V zO*FcsbqF&1;-bLT>N>YjWmw3{2-2Pg;7dY^G@vE}pMqh7JuIdvPuD1y1XWAmb)F2j z2E{(_;s^m6KXL929tGXq@Xrn*&d_{--T97ipgqi6L7vuWnCFeEqI~sziW}29@MYc^ z<_X(T)SkHchie{lPer!}6&D6XsKONrl=sH=PY9UC2%<5~(8q}Gh*XZM1Lyx5p*?(4 zY}TiA`E`ssWoHF$C@c*#Y2?eG;CoGqENmH1i=bxK%!w!_2cp-uwRM53whOQptw>l} z8iAxq*1C^6_&}$|+?!%l7BMs^kL0d$ST>#X*1#U`I@%BmEj9323 zjui+Lq5mbko86eX@bmS^PY=rozkMvh_lFnkzT>!e-D2w)2Ixy$M@fhwD7KvypA8As z-8Ld`ruu+&fhwOYUk29(LReip@eWJO@)R1b^p$0Wh5;l%Ap}1s9t4hO?e<^3SBxQh z&$sqb2srxj`BZ$l7KcbYs_7__!3+(FEWCA>Zi`5OX73}iUe_7E5V5CjA2-oWW>!+cT#VKZP z60Vuc-j*eb?njI~Y@LY3N}{x-sx1CCb1t3*3hGh@#mt0{Q9_H30N#tmT5?!=r^TnW z%{1`9*D`!p#YrC3Lwhzg{c3L+?~czk^vdWxpa>M?kF<%h!ATQ}CjuRwT!}pcSWkv= zrC}q@t2&S)1cH}+1uvTnw9lRwiK7><)Gk9k%s|@Q=a>C)a1^Zs8J9GO2iED%EkKLp zcv_pyn|B%v=~Q`sv;fK;RO~Kyi9Amoi8=S0f{TuOm+fdJkrPVqr|R~Yo(9a32LFZ{ z3XHc*;88>PB-}3KMl=G;(?_Z9rXveCVq|m1)k0)O3$C-~bAa+*ZoTF#DtOrjQ%EXN z1cDg;dUqRO52V$trFwZR3Ems3PE5#iXw~Z9P1o*2iuz6M5QOhkcAr_k;v0uZN;sXl=A+QrCcv_A@*8+lH0Mxvk{f8w47mEGLBjK0rigXl=h2*@q zQnW2GwegjPBdP%@1whEnR&P&F&Br3+=10LzG-J2Q8HPSqnfxJluK@fA+1H4| zVB??J%mdJP_Md#FSzWcfO4X*3LUr~-d3z8;1-G*5*g}k5`@)z3v&@z2Oh+xb6w`*^ zrj4T~&*@W!E_5&PPXBRf$2E&kma2Iw|GA;2ao3_wD%-UJR5ip|joztmHYtWR%b<;K?h5#Mf_a4Gwh7P|B9F{5JtY z?chPcb;nk?gs#Ds#Y%l?wY<2o{db z4^EJ3$22A<`|!DYEYQTTpM{;}A^>2Wj4=b={jD#@`fE{Cz1!^3>Q?$^)b}00h)ooa z4zBA|tbyOTgY~YxI*4-MY|V6vJMEEg1H||+@_S!MmTuU?ef@!R&>CfoBH*;X9vKP( zBk%!!#rRW-Z;2cP4!zJjyEVv-tVg#q=Kjg6E5m=Yz=9f-u(YvokCWT6siP?)=?K~J z__k!wXG|{gyKjc`SB%pDJKEbFH$B%Xt@x?SM39!DYJ6gAng(Lx)T%AMIQ8J`^szB` zV=2$v3vpjY09;b7eva%h66gtzhif1SK<6K_1fW-uB*~p zFYy}4u-jphH;>;FC-{EIJiUH#`xql&E2uIY!(hHiN8*v|KN#5Ak*grE3}7;nR7YFe zAh{nniyc1N)mqwetp;;jWTF=Q;~ac+jEtBNxpj;7PlkaxkPF#n9&{4ozLjozYVagD zNF?y~TKHDl^V}$>3=o?LfZs;qcW@2SfMyOlEvhh??$J(9`RnTX3^AR6t28S;m8fbw z>pNFFsI}NSa`k$ZoI6PiHC6qURjtk&u5gg!cbk~dX!D_j!8*_N-It$w$(J_0O(`gY zhD^qyo%h8&ShHo7cBveA(QrN#f`(E zWnk8+wN3%(i|oi?CThKcXo#Ucu_lIuv~Y@7c8W0jXMERVd{f}OcWr{6^zVFbM}!tK z;m@&E_V_9X5GtG#Gv$lL!5eKXX7-F*aXHrTb?XyI-OeA&|Jg&T!cUt~_W+Pzd&%U3 zZ~&pef`xELQP5a(=;80)c%o}|BU5O6{^#?73xbhW9@#r9RM-Pd!|~x1N%-!jQ@)4e zM?x7}i@5iq#!2IEQ~KAyKo}nt2NJMME(8!KM+N~=xpxdhGh0=`QCY!t^x7K9=$8u-}xeHsx+>m8c9Jk1{F+{9O3`V8xCM3LBeY)Ms!Lej~@f_^gVZv zEOwz9M`a-^NyMy{gEi*IPI0Z5B#<&`44%K~cF0=}v)hX00Iw2>9|Q#%p2b_P@)aKU z?{#X{o-~TuUr4}P49}i8(kuie0W8>^ZVCK>E&e6x-UmCAe7x=;y6JJDf`B0xWL%Uo zj^0$aC6PY18aGEH7(+vkSG#yd>xEz&OU?4kqi=w`XY~Gu>c`;hk@0#ke1aeJ&KQ6{ zc01#`FyjanAC~>rio&H@dOM|?(Bm+TYAiavi5NZMg)YcVP37u6N!A<7Dj}m5BHd;E zGo($!VZ2m-7X(|i$xjngV|*(}2HwH+K1?&{4fmHAD@Xdp^L!!*O(ijc=6K<>)FJ%x z34VmrpYuID@V^nDe&ZHy^`xy%jDE?3opB+qPWg^LRP}x{qEe)7dY{U7Ydy zEwt$`97FFxVvaYiFD@n!g@2cxR%5C`rZnl!hf=cT@257hsD9(1BSXHif4m?3W|X3I z{yX5s_FsV)D+3$j|6Iy;tt}mgE$P_%I>C3Ye}(5RgJtM)!mW&6q+PUNDHez4a?7KJ zg|T8>(IrL_HAGZm^G)-MvG;0xwZO2ygohJ1ZP+k^X6g!p4`by2*U&_bkBGSUjfm*^ zv7F|(eE4^9@9$wbjt=g(H_6{?ith|Mu`+b=G5Qv~5hAV=F?74wH)u|%{}@UW+#2Rr zW9iVu4=1L`^>{cj^}ayx-v`*W&HX`=CU%qN%zms7*S6-EdUk)Ph{w)#U-l2t>DEO# z&nm&c6v$9W;Ww&%9e~YB2PFqK5InjWdB+Myy7uZUCrwoj9XV_+jG}E@$E?q?%X&f5 zwua-u-IRHDTI^f#AZW`B$yx`(93Srvdt8}D+~Zkf!CWO$iS%m`LlN2 zxy!Ny6}NH7Bi3!vR!+;9EJfLny&$e*`$_|3!}@p2Qb+dlRxz*G9IcA8`=Bn7CA6%v zkUpDD@*298=r1_$A3#eesoZ1GDjaRnO-8y&E_9G^Oq7Mi2p((ns_Rle#Nqd5O1K&9 ze$SYhQ&;@6oO1vy%}x>;4x-F?iqHNnug2x_JF&%96mzwPCHM0>8iP1i*XCYqmtx-h z1*GU4A&4W>Fq=|Ah~vm`}`Gs$h`i}AlG8eim;}} zFLEejauWc~;6ZV6O*nnSi`0m}`SW~@A^(gBbcvW22=*;B3{5FmTf z_tAV5l3}?&Ymt26fLAp?3?7c}X}{I{`ep!T0`m7Rxfs@;n#y);GmW~Z1O2`7p98~> zmY>R>t=QAU7n3z{pRsD)+q0YR7oPbS;5O)Ppsg7$_g$B!QfOrW5_KDg%^|#8Ep`t4 zc9(RRv2NxZg+~?M6@~4#ZI76(1i4msaL71K&h@{`{2Yodc4OaG7d8^**vpZN!0ydF zBGFo3z8i66!>66I{DB(HJ7hk8uoP?5F4639y#;b1JD0OA(bfn|yVC47U9dokIjYKF zxxhCacF&fIqWpYG(N?xhJwoz{Jir3Sk-lQ0{o$`qx^<7$NIPAB;5vV&+<`_VC_c2j z*<%yn8g<}91MBeXwyvpQ!JyvYk08hIjeS>Yvu0RfEQzSX(e?Uu8)Z6AEi=wu4>xSh zcK8F$xc#9w(?XJ{@)*h?J}{Rr$fun9E5>#K^N8O*6WN~Kb>y|5kR}sseNr|H;jaw* z{{RF>)uml?XpVk*>@1+?ioJE-p1i1JZzdsW(@%ilo%GsEOk5v#0QO#DG67(4+qghf zPXcc_e5>WfVlFCPj>iPRSFTF)0o~5m0EA>1!)PsP-$#GX%YVr-h~;BX?vl^clUJex z-v)(ZN|xwG(N|X4HuZ?@AufE4c=VX3u5f*a&Pb?YdAzOr=c72DkwDPVSPXHy?h+U~ zme21e;9wT$LUV>jjQvHOUhm~3AxG~Q{;MlD_QCOEWNuZqH!XR0iDc^Jl1vbOu&=FWRADod1LF=8;Qy zSRPvp&Ktr1pMa4?Ds8Nuv$&nKb{$BTc7GpQ`v*+iT%cp%1hz~j7GWoP^~OlN+>Tpl z3}DSm2*`{qdE3=rKG4yWUt*`koe$#zex=cLwD$a`S zFRkp{UXGm_cf zTgR60Lh;G{PD6NSKQN(HF!icOVrXjn-DXogyVhzuK}Sa?t_~g>XzH*PQ7+?Bt!bi5 zczyC`JGe#rWCRpJ0LYS#V|^plfDwR!o?6d(B1|7Vn@g@;wQ0W#r>67C%uIq72?;{{ zvGjSWSFl7751p9~ezIGhoAOu(+9ZpJl)w!0vN@rmZ4zX2LyQJxr2Vh`7VM8)2RsV? zU$AcQ$bh!{qDb3m_q)=LC{22fZepfAY2B7T0Zu0K4i02+S3w6Ijg zxD*x_Fi2jFAP|~JSo;1`j{^N2-pXY?hU>8V`Vp?+0f((n@M?l>rVbsYWRRJPu$i;| zl-6~X@87_we!6vR&xeWc7#wzk3?s3eGR1``Ft?!7r1SD}dfE5JCjAN-#rZY;EwJl4 zG5lr()3x~GM-w*oNV0Ng69$bZ(8xRfgYk`2$6A1>b!S?|$%_67DNI%8|7sAZp9Ui^ zBFg6QbIXkTbvye}Hbpt}&t%e)Il>rBGrH-AapCK#K>3mO^ZLLo|J0r3zq}@y0j_~p zVR@n@8o&iaL4ZFuoVESN0smriAozj8*gLmtp3}&pYtX8ad}93JZaYR$%PIWq4k(s?t$mG2JgVgQryhB@dB9b? zwcnxVbo<-m2LTHy34EZs(!+~&507MhaeXk^2gVzSMbXh-R`8>EYyN-}Gu)C7@cCq$ zzYlPJgEW^SZ0eQ~p>7p;HfuEG1SXOGNy;%Zyf^4*bQpO;F>s3MKrQ*BU3U=ZbOxM` zzkch(RC`8Wb%v$yq5Ey6TaNDEw4hka1=Vb~nDOws->U>>sX&HVV#De+LZ*o}aKKO$ zX`e2Q2Djo9I`W3Ir+4;N&_3v`k2~EO%&`uf(l2mPfYERM-`MDkrl-)GhAMr|8byvp zq4F;p_H&yRZf;;iL8u^sO57)k5xcB~x5dFZr639gjv<{*Z+TjiHCZe&IpeNN#{+8x0Yj?mS$`<)YW!9CxSAX<3S>4+L>NKQ2h6t>G5~8=K=db%HycF@5nq;4dRAV0vAqqI@yvZ+l~{ zZ6E|^ZT)Sw{xzI(HKlg-fv}2r!zLMnl5yGlu6lW!WPTZ{(%&+;4zz0;dlWcL+&h?(dJ2 zwTeqz1d@gbBau*UwJgY8T$=&Q6nb|B$Pip`lpkP($<}V5D)^D?6oSY?o;^E06NEhm z_m(IjwPKU`-aS(~4Sb!&RB}sHg`tM+9Eq8W_Ln!cCu>A6?Kx~20&#K4Qb&1MUrPS$ ztwIa1U2`9;2b^2eKo$M6ekGQr6eukrdzNfKN?4fR6FfLB{edGYc~*doX!WIT;pBl6 zKM)I^%``NCyarc}&g4DjvgxbY27)(b3Ukuj(Q|rdfPoxhcsG-G@O3JkjLUu|4N3_w z&-a7E8RAtYqQYAgtmd_x!nDiZ(1glzW=w^p2NqJEUpp5ULFN)Mcqybi#XJoFaBLwg zI|jJ2>^{Rg7QKY-9vuqfD@R-c_vc65#1D8-3n4|(w7r%Oo1bRxY;X%rr9rnC2s|84 z;zqX8ZkzzE-+oP&)$eYDCvNafO*sQ-)@kq2|84yCiYcmI1$yE6KFo7tR_p0Nt#ePQ zgE)@~PAERh)~Il~1$5fy;tmX8&omvTeBoWlbF z=+M-zl$T44z^1=+R0#nC3?Nk$NN9dL9PB2D!)9~ErA1d|$x|9#gzI85gqCLj`>Uqn zhe37%YTBG-iU;B@jD9SD2tDH)^ci@}J_(%Jf3`yE03=P@zwPswlMHo`qUUMfKKii0t^lqmxN!yYcZqo84+ zu&f&mXP6jK)))u!nQSj^(LDmxNoGnAvodqfS(Yl9h+DXQHx8Fgf?j~XF01NNa@W2e z23k!>*?{~`G}LzlJPtAn|6pk?9Y*IA+jIV(%K1619N=?{JxjNU18gJ&CmrQa_L^gy zhgIhrL9-{o!YVJSwXdXlY9o1b-b@rmko~b(r!y!BL%7%{dKyLts(!*AFtMV`HA*?l zI@@MXN*C2~U5E%KtBHk*$S=Wds0ch)EIarwo{Wln8Z)k46BrJ+0Hpt72GdOjmANZ~ znhy$U`a^1k(le?zPkJRWb;$+I*tWCcgy&)`etplsO2w<^ozz+}PEq5tY%Zk&y|2$O z>>E(F$331*<4VT!CtzS1sIad+~aO6-S zP-;0uoap3;XGRW$y)igG7^-3n8-JYuJ}j~bfVKe^ zeHD!~(;BV&MGIPV>_aE+Lkdc<|9okwV$0AP6G)F=SEc4l;%e zf|Z>Z$FXv!A~&A__esnnO(TBzklV6VOn|ePd~hhX=THC`wF~T2Ju#D2yHfioetm$; z1D_sgcf5^^E`>4aB@~`w60)Sc9rzW}qC)yi`l-DcGV5wHyX?Gt_)wG-XptzL0LT`B z?aJeiKYWLe0F{719w6S41N@XP^N;Dijv(%u<$aWMSTnLbJHz*s*J0G2?}A=)G_qoR22Mjid21SV*@u8Vl8F&av)`|l)*T970Ba*7`|Ka*E zfWLwPDb|EIyA6mhwCwE7PMT7J3paLZQSahRUv(MT*VXWAA%LKCOEb}he)s$m*j-Xd zXLL64OFtUP6n6&a3GP%CWsp><7V46M+It!EGHCjUJ%UHd?%GED$UUX0dN1v>_f4Y5 zo%9ArHmwc7e2z$xZM$5gdB;4s$l}lK?bi!HYcF!>FL0;x&$#B%!=2;gROsJ_?Z`4X zG+7`vo&~v^ks@V=+e;7z4Mh)2$xC+x{vMJ2V^|HolD$#xnW&WII z(Nn6x*LXp?HN+9(ZBskr%S|?`F*hOij&@7lM2XfE-g635m0j(y3a3=}>^l$(^&+8) zTB7mBVlzWFgs{R)}{krGtJXq;?OSw&aOEAX1h1&QjGcG#w?9OXMcK;C{ zdz|Fc~JB+)s*5kZjM;m>?yUZ8s@D=WSH^tz&AFq^rc?l`WGV7h`ntn+C3o1O<0` zZ>Vo2=T{#xH2{^(%#L)h4y)4`^nQs^yc3USzB7Xrb){SJ1bF@A0zvF@I|CI-_aQy!S;QsNz|08Lr z%`z&o)9xt%Q>0f(v0Wa*+gKhV88ZswZ2)u?wqF0vU)VWR;5Yl(zFMp@>|Ody zVO*dva(+0Yctt4AXG}%G>L@zi+)@;ElJL*=`}OxR>3*c{{Mgf{dAge(=i@&h@Mv9N zOE-$BY4wDw^-Xix=%jH!2r2ETPuJVaj3bTHLe-Ln&YUAlv!)wI);aZzN#p0C`um@) z;q}Mt{^&op2D6oC_d#TjhTY{Y7pE!w#_dSG;fWIym=7)uxKID!tKvD6moC@VK)A-Gbt8$Ta>9zDh03?ca@&g;#Nmi8xz=Z@bv~EgGoh z&1#6ElY|+Y@$Q>ln%LP}Ug8Mk0Yj)I{-S*5xS5#j*?Z!7%x@asRIy(do(2r&Ip}+! zsL{dV6{4Z7?E^q=j8CEM*c5SWYZu&;Via@mRSSm?eiD1n+vf6J*<8J!qrNJ)Yx54Y6U)`w0cYNzv_n=}QV5~k-8OU?LOSc2nlkk=^#L-lHpfB z->=B9VPBn6g06_^CVgR?3-2C@o#x>R{zA}2#K^Hv$Zoj9HAXH$Rta{AAcAoH>U7&N zl1m`m3kp|gn}P61SK^8q(+14@@7iqElh)kBl$ntUQ)Rsx0zKPH##J_7_nm5b^p!?b zVLj;Ng^ZvSQbvJH1tS@h;YU>8vrxHfZO@z}7x#M_>ADmYy|op?){1kUO?4CR{1z@g zoF4$~cj)Ah*gAxhl}I>#C98W84urK_0Iz-K*C6 zqeT7u>j&eW?p+&c&plUY_YP0xd!Qit=yElUbR=UmFtKdml3tIO;|b~;0kFEzSaGBK z)y!Z+nuLfKhNo^0m48P_wqbH zxez{-cej=687y!a1W4^H8@v)^mBuynz@RltR#!E>m90emJtUHT47IAtKMWX_R&T|r z-bzrdqv_3H#u`Gsh31TqdqHinDU@)=!zPxJ5D?b8ZNc}NU@^;gpGj**@-_Y~*VX=S zaLYG9!NKG$Ocn&v;OEv%B{;(e%;->$#E3?scloHNROljrO`=zAd@!e;`$HeQ zH17--(!nSQ+0K#+a^A^TRZYu4%6n%TB$Q2;$sH*{=wkXk)u=gx5gsF2U zC!SN8;)x)kE5j(&MxJ+()AhVlWDz8I5MM;;(+2_Bf-EM?*ZKN#;*uDlGo)#u8yktv z^YG3Pnx4SqZg5~do65yiR1<}N!6=85JeqkB=qSb#D^j2Wwc|t34!=rwd#f@iVG6dh2JLKP{q7N%xzkq{gUoqGebve%qlM5B@T=Rg zKpzTjYK7n2W9(ih8D5VY2XDUKjn_@GZ4EHA^v;`UF%lLujGwStJuKFS&Q{myk*p#E zZ|KExQ)9gi@MO(K6j(PRSWx4xt3z)qzL>|#T#Y!(QhOXo6*AS~XdXh}C&4MfW~@`E zWhX}2tX;a@ev*afcCxkeWRaFwp03}p0b9WmN}9x8al2hpN%{}qyEb0zVaCjXz zt}x-EycS5W;EX}y_#;)b%z3)iRE4sPk zaQQLH`k{=5*U$w0GusrU(Y43heWQP!oF3dKCU_{Z_S_qm0C4Ke`KY?vHbP@I!{7(t zwW&H4alP6JhPTcr7d^Ji`p#gXMKs@M>9aLjXI!&)aDD1SCzvY~5|UYM6*pJAqg{*r z*|pdD-6i3im^dF(i=P6c2C(cZOqQ9o~%`?wHy%o-T0QJpmiRb$#7w>nOH`&D=Ko9ANk64i`-zFvNC2bTaUKGSV^qt&}?xr)Kh{jx5h4s&#~XV{)LsE3ZyjF=TR&C5O+!9~jp|M0{hC&E=5OPM8@X zf&1q7x@U%E0wlH4DIG!$nCi?+^AL?I+2{`(#v4qhsP<~*#)}t4r*bJek~qby3EM8+ zQjHbSpzFXU%_~vS*d*R$dwG*#<<>J0egq8j^#e%VR+h}b)4&(0#SAJLt+9Da{|LB(_{n6L?|RNq#sX@llYp=!h%UTg6qF`C*==M!-m?{? z6_MdFm81{7hi}dvA z3+B1xwM0!sxWq}JqOu_Hc`%;k%{hs(fOt%AMh`%)SylGRvwpwDO(xaEI$Q1e3tKF%F>uduW0_N=@9nCt^Cw0kb#1RX% zD3p(YY0z$y$H*HyRFWZ1oiTQLOohr$Tr@A#wP`h>0jH$0lB-IN{|NF|KHvTN92OZU zfqN+KEEx#qfV3~ovUY>ipo~L5IXUG>nL$wcqqgxe?hm($YZnDMbcEpoQYMJvppNem{@t4bcG?_fKDVs&XiQcieMh+;gyX zD0Lg;UQMjaOnQ61ya1A6@z((GQX4>)!Jja+(^DTQfDC3*3YF=IKh4Sc3XB`Soe>Mc z&^N>sg1Htu@_N!xOQ8ms7Yk>FXXHk;mxROZZ?D_H@;#-~zxZ<0V_6YYsfRK$lw(`r z2l|Pr!DC)R;H#3lm`LW-M1L9!p=>atUQuq~!QP#ZYNk{eZ@A=U$l3W5MloueuWRej zQFl(Ql%=)hw9vEb+<+R;|HyQ>KaEf=M5r!Y=MpjP|H4;CUSNm(_MgyYSiO9VeDVG+uvIp(x*sz;ntstyFMzOIA)X@x$$YfK5=$Rxy2;Kg|B>@kG4d~VMkg+}ER zD&quCxr63j{OI17o}et75H)gG*zMu%F78;LKFZl*+nxri;?A$0by2y4i)%54*(mJz?na)q5Vbl7G(#;zzytCy4>(Ze85TirdOe?fDn%J?Gx^k4JYE!t` zc@zGeL(HZUEDZD2p6y!BAYW%PsX7*RjtS%ba3m_~L0-jS{CQ_#NG0RxS95c-Qe1G^ z62MBI_?;{f`D&#sBvF+P2pJKBZypWF=rK>-Ke9#(otjjE%~MG9zw zd&1(tU*=FlR@-f@rg1;yya}*gD_U zG4fY)kw?^hc-3C~aXQDil=6-2E#Od9x_7#ob8fz+;F3tshzG{iPa8TMqa)MaG~zmz_EUTC>< zABOk42aICZxnWCynEcUlZ8=NBRx+jnC=n-PDj!y|t2>gJ&ahQQyO+GFFKH>+!St)R zGPWEo+kg3Q<$^~?Iop(DiYHLtlFj$QCjVR6)Xad(ci^>+UQ}i2(M#<=CEsgppX|T5 zCIK@&2FoFSJA;C70MmQ+jm{<7Rnc7Vys{D&;pzDZNQYeH1&y`(b3EN-v`>U%lz0o1`WST#pW#nVa zJ$_E4-x7laP*!jj+p0Y|w{*mcQ5jwsD#~*gjzXc528gQJ^l@2GZ`z7??+z%_u+_n~ zxgH=>;M4_Gq=m+SqkRd8jp|e%Kj(gcxyw1@{~c80__&dy%cwyFWoj#G zs?O$6Aw(ZX64i?@(I?r)r`T<&cDhZhG)EFAD@|6W*7$yghb5EpvYDPM%F1N(7#|nL zAPzwblQm?zC6W=3B1ngo3^DH+8nQJ6dBnR#P!Qvgs3SOssSeqP{~UT45;4Z5%cCI6 zBSD8#4ht0qyGLq>@W@UR-4NW6JV0h*NZy5ag)52xGG<~3qaeXJz%yoJ2=YWI3&aPf zy^)P^nXeNF-<)sn1c}4jPGP5>0hH{+AMcfOU92N;9CvNCZoj;4v2Xjjo=Xo^hUwtW z;@?*x=D0<7AK|;rc04Ne=IM%S|Lu!IiRm7!>Wk`uj`cKGksk;SEGm)~8ynr`BV#%> zJQ)AQ5Mds^3$Rnh!u6OVn{X?joO>x+{84F@=1jL6C0w`lzQHT_k^X z=%EJm5nC@$j_kgGLz4x*&!@(-eS+LbDM#Um#piP#lzARDe$=Qng`TjRhX4&I1&JaH zbl*Q6dKI$$ZBKr}5Jz!tNK?$!F!)9JGi^I9*_hCOlL%t+3`-j)!Fq$>( z!+E;p%<DNC?7=l^hNu+gJb$)0Bw3>p@os37LfY?rbA=%G zZLlxgopfFI*>Z?aqJYrv15R;oP6$Q3uUk<|Zud(_{O?q*$ROIr2M35^V!@Gf%0 zB%@7G$~x%*8x=@QR?WEvvNjaOZ3`IQVW>TOYFj^f?^CSU3wIl21%xn(8v5m1C6vm#0& zF~Xe0qj&w-bM7r4QNeXJswJr@Ja}G5`z-jHNxG)IWF* z8F$9VWmq-(lHF&G17sM%{ zm|pyu1LcFdX0@F9_y_EK%CkpOeO#T2gytKSzc5OdKY@@JpEoSD0Q{`%ajrO_$_@or zqLWje#(MHC^3#~pGI;aSO(FKX}?MZX`9gD1zHX!(wF-z%qb1d z8_~)hdi9Ncg&M=tucuL0Q~6EENFF}wV&M;yxxEgw-Z^|7;8fPstWdD*=pa$6WBKRYp2@?tx_8qT!p+K&9MYYF&twTnhrpO zm>ugQOngM9m5HusGVkyVkb;epFe_`9NWHUF$2&#Ppz$C43d2&Mgl zR>XMYd|*;w9fpu)K^s?^OY9V+2=$d&B23#p{do+U<{*hM53hDcg}KEXt{O!>4o>1X zob`gCbAOe^qbLy-UBIaW>~6;nN!@bb#LiW!iZa3oyZSBJmq9Bp@R5w}tJ$Xm89cam zzMFNQa}8ajX7KE5#+0tu9{hR59JJRcfeMtSCj(y&c*q3B|^934c6rSJg?{CHn;8DAiR1OIb})O$gOz%e9`uEP+GKfO$h~1beI|yRJ!^W;%kNlY6HksY zKbrC*z4AvziY@S#QvkagC=cTVEDtOta?-G;QzJXN$$S}}po7D|A@N$*U+7QSN{IUN zTgcg@ts7S(&3^+S!ZevW~((6u3QqWWKcPP9Z=TIbImogd`@^RJclzj0RyX!b;SINiaZnYq2^s`31}b(!{cW zp*W(R-*m`g=wLW7*P0`pPJ|s0q-QP65ln0=8`p+Mhpv&Y68^Scv}n4owDD|0hbb%E z1BIoz9J?So(a_kCsm;mnTMsdtG6%8P$@g5m@@E?#jLeIf64442_9j%HI;#VHPN{wd zt6*L~I#ah-#k$-$_eceEpv!ps5?VZY*E7`%Y}t$%doZI%F6+>l%+~(f*LQa3*^du`}HGWBDJb9 zi5AB4upR2L1JyA&GYE^ZO)#mdlX9+hEC~3gjA8frig_WI34ce4;u*IrnYa=UhjEjK_-(dR_ z;TsrUHm0tX8#Tfu+iSl0W5Ud`v~OCX)pGJcEOP2S8aSNSmB_I?%fbB*AkPn^VpjDa zNh&*2<0fqkYJ4W!*^^1??)TsOCBNuk?U;nTju!>J*#h6?h!WHo2e^u8OPwUGC6c6;wmTRmH0U?x_}E^wOm!RiQMmvgcN)Lg5dx&xgi z<3_JeZB8C}KTtcZetza>a%#f15p7}_z6h|~NRkyg!?jZ!06vV1=VRv{qItNDAp+SY zm~BL}j#ce_+wuNpak;<8{af0pTp6k)%jql%!n}Rm=Cjf2vw7_WAh6^M<4NZ)&T^%9 zLUdFDQ6A;!)^1>i{HeZkJ2N}aZ#_68F``4I`CGe}QFu3|xOCzGo|McRSV>zW^+1^< z7hn!EWG84G&}cm+9MTzb4}s*Kdwjs3`#-3yXLg0~ri*ckbhR{8qXDl)hSTftfu1Hs z65%J%pRjA60Q=M{L>=0gmL?0n2#+4bj3HTbvEn`E7!YHdiI5#hbB@A$AiO)cbxcor ziIc40P{MA5_w{q&V0EKWZh>20i5V6NkSe;@g=B4eN+SU z;JBJMpTzm7uG_|fd`ATq3I{4{UuXYY@9)Nf&Deq$#b6mUQRjfy-1s{D>x zc#FP5rTXAk3b$#tdeC~IBPCT*nRVhbl5&M*arj&jlX53+cy-)ix?k>N_lhj9Gkm+$ zhZiUIs5b^rRj_{FNHJS6;k_<(w6P4-z|{Lt5~_f(Do2%yTmr{^RbO_uj}VP0;KFOb&H zS`R(Ly@W|90K?}VA;J>;UyPk&b0%P$rsF5JZQDu59ox2T+ji2iZQC|GwrxAvo~@dy z`S5Pd{)y|p>(qhoQ`{p$TaKl)pu3ZP9&U4kN@|7q{*j_;H0hGClstp^mn%%Jux6qC zbOx_$dmXgGjp*qu&aQBepz0g%)*XG*R(oPoJ@LcEb zvrJtUHI+_lnX~5MfoRE+(ev4sW-W4Q)VlRyl!h(o_~@I&D#<+iv8rWzLh)$~ZFfYY z^B*z#d1)UZa+XYE*rv3qIicd<`giX7I@@cC47W7;Tfj;ffuJ)5EY2XKG?DN%dipTz z3u=7SMHaM;Bq=yF(M%F?56@zLRFYBS>#f17A!}pV>H%2CGI?o+Vp4`PB z`AtU*f$f4N@nY)VBI`K28n5D2}LOhm_)sDm- z8v0)hKc7aH3DTi|)LFbj4}IlSjU=1`{zj9>i*5Sxwk|YPMD7J^D$)*Oq7EEBk%fDC z@BK^)5v8l@rs?*GYX4*Z=oHMEszcGBdPBLeRNC#YPYD;dQ-%0D+g_>X*>`SNZAl z>5PytdHYUPn;*=}Vw-6Tnmt#1oT<6o*8Zq3cJQ8Z9LILdV>#Bm8hVO{%G&q? zNXLvROAFcv?>FOe-MpwA?}NE0Xbpm^;t|RUN0f<7w6oJv$`iJWRiVaK)n6QCc@Zpe zr64&x|6~~F%uZ0e`pk_PF^_bSv(Krv%bs)xAD&o**xsl9{+HM;r`I`$SF8|G$GkWa z{1XVegYbzo&Ogu;>)nD=EvYcMH|f1Xra{yfpFV6d`E z;^fCyNHhgwK9Yrx&~~obkm_Jq#i~XczyI4ASa3#W%M~g<`)pK`_xPE7B}|`UeDvD> zuF1CROImEjU46JPv<>$5Hj!SRM~WfCf`P)DmW7SFld=f_r2wayjt&ThnON?bQ?zwo zgIZDW{9EN%*Hpnx<94*5fmcrRKHbr7530RivAt|*%Z5Xzq3*)9IcFpdHRvs_ zLV5oBlEUuJJTP1lEnyPmxpV*9x=gFv|*VBD>KRIu}#jmxoPO`moF)-Y56n9;M-_ z>&O)x`2B#1w`*qXPze5e@KmMeMCS{ds-QxTIm0G>sJlJ0PER=kEp=AAvn;d7KfvgHa3|Nkv6%%2lF^!l=$7T4xs?6tf}O*71J6@Z z-19J{bF8fBF;AXdwwMJiP_&MrzhLWt*x7$aZqtSIcMM>&q0cfTqALhs3|ykNTp8Of z1H2;1&OTS(rQQ>8{7gdTKpv^&%GI!y-;bIr^+I8YQXdALc5o;uUV#-jc`;79b(0Ox zXv7B3x=Tm`&K z(5*SChWO+VTph2CADDG+dtL-(h`@`=q0d%nb`-g9Fwr?Z7%TmyOqdxhVgKwjqmM4= zo!yLE_gDuUJ9^*(gUef_a#hhs;dGQ0aVqLDB&*7NYv1#EK1;E}#E6l!=7nf8r8?~_ ztOg+WdZ8+WA%wa0Oq9(F<7X>qC|58|7v&qzj10$T=y*QvmnV(RdN=u4XauvJO?#I0 z>ZX-(DzOrxs1*!Yn`p0|s9ac*+JiQGXbtD^X6tOhc6+J0tcZqw+Y4+u03d!A0xwm( zp<1-#;HsxG+Qt69J$59eOZ*!YOG2imY!}NhZ!{<|o8zxzM*su0?46p@r2~Rk1<{H| zNIc_IA4x*|q6vWj~!mALl zjDW2vmo!Er#J1+XCZ=7Yg>v-BcLA>XrJqvlo=nY@6Fln%IA?v#=X=ZVR+iSI4l{T> zA;cdjQydTC|D@V;{SVdtzacH0EdL){kkkLRAUE0nZ9(2_(Ry;6a@l6Mqx3|1k!rLY zz2?`mO4wdO2rZ%mfWqWa_IIq9gpbY$%fc4YuTE&_% z)73=#6$OJkW2#yy6id!M>4ZhwoK=KgEgJN4GxTy0v8>rut$F|@qa1FcRhv3(W;y0a zvmG@!Ia%_~nR9qs7Njh@c+}Cck57cUPXcxJF95r+JURSv=T02sEQwo_;fWrW7!vLEgZ zLP2cjw=7SsY3dvdIeZ1;lDd!^ns8&uH#kp(qXXIpfTpL|jxNSYRZhZBP|HD!ZiF?m zdH6hb zJ+wRW0yv#-=|U^Ryui`js4vtw2OP&e^r9ihCZoa!HW33p)MejYb)d*EeI%et%xSJm z@d7S}I|My(K&lRg!5ND=w zJKTWO1WLB5zVO(WFu^r00UL!m-u{qG+yYXpCgsUUJ!Q)22T~1|I&l}s!pgfJxtSU8 zkmIinF+C572$O4{*){gi7pH&1{EwRo`H5DyaUo94kq5(a?-7teO0@8AK277cedt?% z3o9*Qy(QnSVm3ErIt_P5-2r@L=v>wDwpt;bsbpi+LaC({j01W(iDrKJfeDlUR5)Tf zI{L(3B0&Y)<;gck9lyhbEQr#Q z1jP0l57gumL^Tzy9oi1LpfU`#OxLZ&BU+b+0_|$kCub?;;^9V?=hmiOIvts?nGgaj zRxTRZZ*X;)NdQkp3GC2FLOMSlMjIVAZk6@LzxtN7`mvI~@oPR&;*`T4XCPSGt2k#V ztNsh_qbMkjgeOaivS3@^0VN9%cFtlHiz3O?g-^IZQidy#5Y7 z-j`S8M|R;Q0XHS4VKm?i^CHP6$6I@by{!$Igk0Db7ONm@ixx-ypljF-ksjlb66-5*K}j=Meq|_kfY9U__I}WC8C;(V(Yxkr=5fpp=v?4jCo)XrBf(Cl}Zq!gDRC0XYxOp0*slWjLO_%l)DD!sNGU?1I%@9QX%WUKYhfJvqaZ z{zh&)#`!I5AX*KN4;j|K^XRLuaBaQlaiq3BQD_)zZONh2bZ9bTC z1e67vS=BFrp#Y)MY_dLAqJg5(N>7L1w#f_2^dE>WF?M7cSelC7_5+P#4wGkwI&3hG z9n1;|8vL7gjt)A*aG~IWgr*VU?am|(gU7~AvmMz(qF6(F>Rhg`0p$;`(3l5jca4KD z2Vqs#o8G-RP_h%~gd^t4wWtgpNBV71E#xT7DJDKHkqAR9#jm4pM7&IMy@b{3C(z*_ zxd&c|+MyADw-BLUwr-j^YUfe`&&OTF1Rl-SwupPtMPqj|?d>^9_zpjl;LV&F7uHT( zl|8oB`&{oN@91XN5d3EB$5Ln#jQh`z`Fyyw?pa;M>S)V zizt{^#&r4Tybp8hj+NtYNLy(w!IDrTZoIipo)QnMZ4inYRa-G{&IY z>z+2k`g=-Qz~c|ywWh&py--O1V+|6uTKs#Cd}^ToIi*XOI0D3(`ArJPQq={)D*hR$ zJ`?{?+`N~rh*I(`AJ&2HLX!XRwyXC&SF}6SF-Kk35x|&{%Ya)`{kEpCqOIQFP!au`2h@|nQfp7Cpoe`gTa2-r>>8j+HCm2;9 z&7m`p7;iR6f|#+%CeE4s6jm&t?zu6&+y5CEOHD1OAskh604kZ|JtsyKwrw;PpX-GM zLtik-lWtQ)$@1;NX*Cn&0Rk}0mcJ9=Pic%wi!P$i?Ayk;;71c0Kq)Wd#gm+tjup@kO z0AAP2?so4lvv#dH|J!1xE#MW6aae)}O=h?dPUd}X|h&KZirN{H* z8pbF$7UGsTQ?L)xh|AmwqG9DRtIhvxtZcQ_> zEZy^JN$SpsZ1O3nY~{B>fTyW=nBwC6j9MV;K(zpA`}6{iF?8$Rd;H*`qVJaNRPJ4=>tchQXRr^$fs?NpgX2X+?3&~gCNnp2# zINS}M%5o{@X6SC6s|_rqLYQvu?_8NoV8heaI&!8*oawJ8g82kdM_k1QEVOqn4U{$5W;w@4w(0Hc(8XMHR)i`@dd}} z{>{*`H+b~W7v9iO&&DUDB=je=imd~Kx zz}d+kkuV)oz!aA=xb!2t9x@mQ6C$1G6+V+CW(kms`*aJcuPV(d7_5^uC7d!RJo>E5 zXEq`rlj=&iNRWY|OiO3ftU->ub=NC37Z#Iy6xl9&h?MNeL{F0q3)237K==5UmrA>V zR0Wjhe>*cp&R!MW$HBY^{1qhWeH40ffVSS1`;DdUcAOm88@Zc2gW3s-bB6zwT%AdM zL8rBM8E++g6cmKkz6Ol{1keZQhUZDc|oq><=z@@{artg{xvT6=U>K+9@og;x`J{R(h11b+b<-?<}g%0i# zp#Y4;tPLeq7^ivVKf|1J!vYo}!giIkgfBGJ|)M$dQ ztF?PD&<$zbkAw8B3)E44`b~dWKf;Da*O5xX1XNn9^w7TN9ha25#ruA-KvH(Rf||vp zoeS0Qr7^9w24S>!TI~!TduUFWnw9l)@zgzg6wcZj8f z#30Em<(hlxzh_$bXw05J5eh5n9X>9U*j_yQy`GtEYPESDn#=t&KyYd z&v}z;M;g0#TRD`&avBdxW=I=fR2MLBjTgdTtIC@6uN>bfIB*MC7Q%P9w~N456mR!z z=kSy^r)4IcnZIO^M&<{2-`gFbcbFFz-iLQvba9d*>8olH*ySN44yV+jnHkc?ob=g{ z3KVw_C0!so4|g~S0cQ0aQALmvicW5f-zr1ByyV41WuPdy=-J}RP#GeDre`me?%JUn zQ(j4m_TQdTWa7!ud+YLCllHy5xjI9RV}m~1Kn2umPQ={71?znvVNO{Ec!e!@oZ+g~ zt&RDw?4&tUxmkW78F(9pOT;N^@dZ=onbjl>Vb-K|0qARl2>Y(kQCp8YF}zsG;1v(~ z-8waj28`LU!)!L?C$~?-9@NbHBITP`8yI8cm%BMbeJeN0P=w-;BsMv+Vn@8-ARPlK z_bqxFgqTBlDjrILl7CA$K*pXxf#PYi2e@m9Hs!ksBkpl&*zdkkmsTr|$pulj69u-r zPvzeq+WNG>WS?P{iwc}uU}hLz?D6~`^}=|L=Yi2%kh@mIV^u#DSv;jFf3{}8V-g)-d5SDbY~UBVR6>ye9XJ#~7l$~OO{mDej_Qf&Zm@1FB%8qLzwRmu8 zf3tYYmmwIz=I~PwZ=KYbCWtiK5{sX*sYz*EqH7k947?MTz8|+fhX(TFQ@2}ls(p*R zIYGs@aL5!LF#e{|hY4nX5PcD;pDY>i;GNQ_p4_kmZ3=1V@nKyu3t2cbJReL85m?XZ z%+}l;y@o}|i=@oXsN)VIl)g-*vnAeG1P z4N~XvXSjIv6-3Z$LrmI@%J7-*S6daO$B_ZVPyU>7K;BgtcV)f6a6lFc2IZXv=Am)x z;wdBUNpJVQEK4TUpbdB~#_b_&k-`T1#B&B#kA2u7Lx_9yMJM$gbLf&aec8~3Qgkqy zo$1N}!aI+oXyebXCXJD$hpwawqCP-i#spkS`092g5~Ko#0k*y(I%rH5kU2rC`=C2| zZKn_mprMz)xAj_!kHuR&7~fPvVaoxx0XY&P(B3LF9frB31&#?`4flM=yZ@F-=R@P{ z&Ka2Mc8Ox4emz;mB8k5Msmq;|qsQ`XR!Q{$ta*qOag%LR{(Kc@F6q}`#@&KK(m1!y z+#+p4?apUmt{x^zo-YR0e~~qMGolT%?qD0kR3Zh z`u=i5VfcQg|3eByenTZubXZR#VgH?0Sp!l3;+QoNt=8KE>Cj-&AWnpa_s4##-+2it z_@_pyM~8J&YXN}02h{aE$ES}%)#2#Ek6I6gVL|A>c~j9@Ui;+*Qw}uN*7YsBS+i34 zd{-#OA)wEmy5`K~r~JS~?9UEX(~(6;_S-djWtg6TZHevhTHoucc95}G4EexBRlJdI zd<}DwlSAb_3|>=wN$&0?=m!87JpS_^#4>iq|0I@iaQ*Lz)}?>F@eT2RMB|zbT=z`F zV5q`vmDtrFQvP+@$>+3;oULXGc0y$Id2%Q=$aRvR4X$a6pMHFHmP%+KsUdVV;OFJwb zU!5hj^>H@FrRV!frJC5_+LR-mp^u zm)^Fl8Eb;kQ;p@taV(#4yD7oi+LMctZ~Ba*dvXrUU_}94YJSK|tWUlO_LaukWrgR5 z19!H=aE!iLt4wq3Z--I$?=nBccKN1KGc^i7?tp~NLz=AbD3nzjba>^9sn_vi;gfYU zSy#JdiU)wnkl;UdN?|DTJ+o6?d&D2E&s)kugiBD4@7jg*P^O2?IyWT%|1OUPdo~rx zZn4?Pa7UX>tegD z&M)>mG}Y>QzqUXtDXdSj` zLs0(XOC7cRji4RDPz6ZHOo(x+iB{Ew&u{ZJktWGLO!X*23W@>DU`}9~5fjAy;K{go za@wl?$=^-+vXdlDNMmJe;pG5LPZa;q5*Eu3%44uJx$;Lh#!{AR;Ttsi^XaReX?uJ| z7}=m688$3n7;?_;Fcv@lC3DJMwf0F-rjx!MbH2VT#zcvYjncSiCjs1-zqx@syKw`1 zlJIHvhUU;fLAU;u6jAUuA>=z126oefx_>T^qNagYC_Rj{$rW79>r-8xz)E%r6Gr*X zqB_Qf`Ho=YXXhu*x3j%xqvAXvg@aiE58x91=x5SM(c;$}hG-^wRuSY)Jb`trDE5@be;2+REB;Q%B zY?n%4<_zLQ*=W`(U{7#g;9i=6?DTGrAaQXs&@b_--|xXXmu@!Q!28T!2TbhAv&qUk z3Q~-#Vbl_4>VyXh8B>P|B+3w1eyCXNz=1BVUY(&Lpsr<0w$d_OZGO{Nb8L_j=o7bb z(*C}I2V(Fu*?;G0=QSGk{!JCR{s@_v1bJo-cTnWBo=lfXXqsQ(r8gA}fgc;iG~UqT&;lbxfS~ z2Lh72A`Sc*uA*!3C495YaW%YaKL9_hHBqN%CodU8WFc9GsUiXw5{L1Gbgl7h3c1o$ z?xtlJqTjlh{Fr4&Z-;fbfjYsOsC2PP=G?NFc*cDjbwgw{O4Y|qlc_~$(nrgvLt9;xy{uF^zW82n-XSvEc z?C~PSF-5t~zws=B*~6wbDUEq9XFFH>lG58(TW$W)n5v^}+Jk(w=%Ft8o$-kqKo

    B|+HD3;b0*PeF!p%!TCnN?Na4=04xHGyi00kfc9Oy|L~r&;zD8r}e%M+> zb@)bJKsxBV0|`ivL{jR+``Q5-gm6tJh@Q#|_>M8;R=Vy~2;=sa3eH*ZRmN zKtU|Q3jS=NbC9L$g}?D`48ysTmX?yoX_5=lhkpb)W$zHTO_nH0%hxslSA4&QHE5<> zT*P18*1Z!7mJ7|nO3nV5#47d?J55rAOQUqng@MX>e1BrgFWND=&3wYwQp_hDx^n70 zdJgy-%x>{26~#&?{EgIhXcBpZ%oU0iluG;*LH7go>7>Q3gF1MIq>WX@)ALaCIm=Rv zv7{G@BvA)*2MY3shSNs40>(0joRL-75gBuKCNH^>`400OEO=0gMFyT3&JmezZ#UsH?w>?lXDduGgCW(Xyo9m4Ciw24*Pynijt3xp`>O?4t0Qj&j{v1Au zIU+gB^=%8)pid0NrA=yHZw)&^Z8`byNN>eMbjDrp9d9BALHZxtEMarPvFnwj@VCjv zNn!M%cv|oBiCQBTM62l=7SQRo3y!y7cNKZ_4VcuE(&fYPMIf@nZSoRAK?5$17ci^X zC8gtR{SPPw1@O=a1-THC=Ld4BULy4h_NNxGpD%yX{_HYaz1m6}C{m~{O-TG2YG+bt zktmB{0ZFsp;88Mg4UmcCluo1MWhA#Af0rI_$#WQkO%Sc4j9MYvw*DlOzT@;WO)W$9%gycj zexJr70FIy^Kf#Yh~hJ~kb zHDUe3mb8vSPr|*8AIh~~{bu?U?e|0HKTP~$4H+43HB{J11P&EQYxL}`aXW~F8MV2m z4|^~~*n|)(DN4})2l`9F8Apc{{`5Dt=Y#wbSo+-txuIu(B8E<9tcglq!5Gk3!3TaX zXV(@A)L^r6)N=Xmdwl|z5eo-S+3ylb2adbMB&W3}I&YQdJB6V`SQjYZ^yTmBzowj~ z`MMPP{laC#gbn3u?C&$jI;HWN69#Z65Y%wu#LNY?|g)iVlN;?uLuZ zsjq*|=BYxEf+bG0%w!PSqU{W3)qKUJ@idvk!?eWWG4Tq?3^JLI7x6h|s8G}cSIlvK zxMpJ)cJ9(#Heri&Z{G(l0^|n+Mefa{USRxjRbahd0L^p<>)q?`#9zjLdF^TgRYjtI zo>x&CH2#f$PDp*5z2uI8EY)5JiD9I}S**`gb`Pp|SNqx)40n?#X(@zXPQCh|Zm_Q` z;tskrnR0PjB2DBI01Tjn)k+bnnoL%cM{A<`4 zUh4KsbMcRp2Vg8e-Zap7!-9Vb3N$tTFaAXt4!RT-T5!8w90y3tj_NRl^k-UDB24Z6 z%E!3AXOv5LfhPK5(0crPXo0q}*4&8_3i!c9z8s~Rvs*o$&Ng=(9f8#) zA_oAjft3tt#3WX$dvf5+>yocen4EL@VcpS&DI84g0*w0X1DmiWu;wcS(Cf+DTAKkP zs`2xgoDmOErLFCp6HU$7;FtO(e}Gsi2hq`ktsifja=q3cxDsEq1i2JyZ}@%te6Q zrW-+>%EVA9Qc_$|EU}O#_{%L1z07)$Vtzg%HT~7joYb!=os9$^N@=W)p_0}oKb1)r zEW2Vj0Qd)(uvj$g1pRF*a1ZJzpp~iLx*IO_(0wEu#8ecQ;w6%N9n}xV;V%%e)QcX3 z3{=J5cf_Fdm%1a~#g=Ns!NihX9l=0|q48}enEzoPn?Y&hKhJV~1S}s0g+Q1*T(@>Lu+G9KhP19mv`qIr$cpAqs;%OeUAjf<_^XUOgD{48vJvD)_sroBXFvqaoD_OYw!k$51T$MMOsLz$++!Rbf{y3#ft@&Oa zN$8WM-LK|iU56&ATjYsf^__@6BL9g%!;h%Dig4=^6pX1hlG!T#W2&vsesS8eZPTG> zV2pp@9rX6wIW_s5B^mzi7EW6RBhe49G_$YdGC*yweIDVoyw8f&2i?0V^l=td`C_+N zUJwdG8v?u-!Tso-ChyJmIm~p0JP-2u-rD!+(!z2mPH4hjwJ)Ttc@(^3d!tY~sr4j{ zIeZ)p4cSXMbAOLsYz-WCauvGi>j)i80<9S^q3=QN$XBC!l-#IuLetF)4lcc(eBf|X zlletlhB|isK#1Xv9xaZMBU$vsOh)OSb?fCp%R#HKzy>k(l%!=-6SWrQ`Ehk~agUSY zcL1K0+aa3L{nf6XT4_OD*UZ+el)UTtCnS>P8)Bg2^~UhIg!+TrKFusBKXiI0c|~SR zhHA_q%r#3#HEnYnS+C@V9+}4y5y}($^v!Mce9aRI-N6+Kx!4dg@FKt-XAEUXB0I!< z{Tu9q|e#>zM z;nnv0S;Bij+;I+9e|2%VAk%iq5BiJ?&z54C7C5hJ^U0Ou`s#N5@B>*Hm=!1rI3-N0 z$zTaFb9t$SC7O=Cj~7UHK&4^z^Qfhh6Y5+I%qiVgvBcDU&su~Ere!nc#Hjy}laI7P zGZfqRT4DfI`zYVpzNHsRt#ZDSgJ^*`o7H(%GuWu>bUh5GCi46rT)nD9pmh8CpGWBl zWoH>S=rLP#q06;3h}w~&8@YVIjTY_0rJP+zk=FSy;>;asBb}!-xm4Zt+%GlKuk3d z$W0rs2J8@K{#7RZTPrx&zAkrPj^YQVb#{^QKPhWW|3g{(f0f6Xm>F6Acd0Ez(Mwv% z4YkMS>`P}c@1O)Vv|WuNQKZ#eyF{zN0JtB}Fd-i_1_5oNL`f;XE>wj?)#K-qm0In$mpZmgiWmk>-}}j;b_+5(rM{r_&wkpPyR;`%dx#&c;URPSt*7Xp z>%1OSt!~oN3S}*wcGB*O0%wVmdQ~lr3c9b^L<7v464l-nb;W$6G`uNUQ;ejUOH+1c zuBlGxs-gfQY2g&=i+HJ2lkyIFlcDOqfmR0`3 z<1h0#^a|Gv*NqpI)5j8KO3;cS6F7bXc#fY{*kA3JXft0MY}u_e;J#{Rjh!d*4LM7{ zaA*yoEzFtxaL=#8bo(#3Z6CNrY0a+wZ5W6=Nhr&t_~;@c@3aXt#@NHq^PA(D-?jVf z1~Ii+yxi=)0xMMxvZmo0%i+Q}hc5)bIInlWk~@*4fy1PAqjQXYXjxfqvW+MDIuICT zPYp`*Fy;{S^Ze$iU?Mu8r|+md<{#c{7pdTYq3l#1qv*QE&>RdLHUWldJ@0H7eJKs| z!x&+fgyL^lQC5^!fD~U`2EPZWmR5%nRCz`#oMuX*aL?0CpK#rG}5n zzLhx7UPxGcoO{?7=QhxqSajLH?_Z!JKA;4`#%hJ_)YD)VLF{Lt2b?^d$D!j>CmRO|`3<2ot$*1X;SZ zRSS764FSbmZB%pDFt0umC$l=eZs?r+8!+A1y1T^2ala)NwvB>fS5&Lr3ofJT5^F#S z#Pj>ik{Y1KJO7b5!fWCpD+*)SSy=5w3d8t*@jG*WCPU{)bLWzyDeaN;Z|_!xdB&FP zX~L0jXP$MqDBQk(Zq4<3gg0(%8UoJ@o%gHkY$|JLn^XBCuU^m~pvB?l_%rNPxZQ7_ z-U_VZgc1g83w9dEO=b@c!^Zz1_x=v)*UNuehI$4*Fe~YOX7!y(t34vr zhvpCT;(R=kba4eP0un+C*G{$85x{#OFG>perYDXUfOJx3(e|lX!hpf7za)62Ix~Fk ztG?5N)F#Tl(DE$c!9F_M;j8rb$U3W-rDH2jJLe!9(o?<^lD)3A`Tm0mEiq z#OptuL+1Z+4*xrq%fa~n0j#Rl)N=YqX!l*|%zfs$JzKPw3TEPf9kEL+7;?29t#7*6 zYG`fw=QJP%1cUgy9{PEGk3XA0fKf?OXq4_phWBfGdUo`FcK~{2{|d`0elmP!*T()G zf10lqOVp&;oAqYn>gMr$m%WIP^Db!^KPn}`C!%PVOIEYoXU;XZqDZYIx!d20y3x&& z{=a;7R?$h)Sr)(nJ%-qR@-OpW=LuxX9m-B|2n89`sjr(f4s2g5N0Z%W+WSQ+`HgKu zXV!+z6Gz4+Nvwai|7JeCidK%7eFeRaI+rw~-%M& zRLNR`xVz#NOPG0%WLb0ftBjzWvc+u%pn5!=T>IJU8ylD-W*pu8jyCo?E(elWg}5Gs zkJzE;EXe5vel9K=LWnd4%L=dB|hrJBtBLxBp6)=!H*HIH;N)U-boXRNWB;!$udi^1rJW#RtSTy~W-XD3AUQ75S1Tiw|fB8tv;)*MGp` zGS}ndA{IR@A9cdLLLg0jK6U=(3OqEU)Djge2c)H1&Vlx8! z9`7FU;Q}~<_AH`olOsg9YUhy5D)&Q6ZXtXg;#E%kiRg(TZ6ICskxh6mfjTUbVT<1g zogNq#OUNd<0W>V@@#}I?Emh0W$`JmRqTjnk<`rnqLitsc%Tc$x+3r|jQgg=*gUua4|nyG!! z-_xkajUb(=C>gcIJuGL<@8YZ%dp?%Z{))c7nT#bHo#&&D%el3`%1tJ%tV&UW{ejda=qll2z&5YFw{1YE{oTJ*u_|Vd5fO?o zz8d~{(-lccNf>ujfZ=NXKTYtM#Yjik$7+VB@S}C`P6NmkN`Ll(MVow{_#dES!Q7al z92@Vw7j^Uo67t>Kne6*k1`h!H-+#lECP|->6#Gup@el}H{;_j^Ikk4s<-&r zo{{y&V4Ps8epsS~H(&TS^Mr{r2j8=S-Bv&{0*VTIL;>%R{EYFz&9`h>5ZQK@os{o} z&lW+efibbTA1yXD=Z-FOIP#uWBinUmDH_DtQx9R&ibup**50&tOzK4t1R|wy{b)HR z`-F#=>M(XhPGHz}k@!Uv;PO$RlO(?!p)mr@TPfAV{g^lT)45%T8r##ES2R-wtg8umx`!lW7SFXk2}NPUOx|1EXdid zh@{3xT)8(}-?jW+oqymN6cA0~TWb9(gH$YQgIT=s{oCxp9++RKK(Cqc#r^c z0WRUsgMkFqH{W!^oAGJs1GB8p+yk00Dwlmi-o6rK<^z>5?Bn$jEP5`ml<>=FxueYx z3#l@up+7x4TXX5IvSA}gEr+54SZ_)9A(t_j|D}nPZAfj*E0{FEFTU4B<_pXF&kn7J zj@3Ub+9A0=pY^2Rh7Xi7Tdc5pyKU5xAjJYUkK(zQJ-9ZG_LwNg=UIw9cpBL~A-6?* zrtvk1BPP!luTHm-InW1nCy)R{X!9oQ3BMjbDndzv;+4KXDJ0XsB<9x{3tc0H=Gy_L zW8Xsm2*?|J0$AUw5w-j5@6EB_+hnkTm?89`JE$t_2UnE~uh2pxH19U!HooSNR>@%p z)fm<}+$;Lc7o6M>xY&CIKHQF=IorV9)fc}W5q4=3pDXd|gH{fjPdQd?Wqt9;d;k?X z43hY$h6vP%xE5X49~Cu~hG-@JDlNZS_X`+$>plVofkkm2A%q{LA+V`>Xgbe5V^So{ z!B#I^XZeH$6%owF`iKHF6Pc_Wx1{ZHr6Wp-yak zw24M9t*`(l`&{XP4#F2BX^sq7OiMILZ`!gZ&oF@TYtTrOhv&XQeOycUmk`RZj?-%xno{S z8ttGgb|WGb4wekJ!MqwkA8_8RdTwe_=ULy)fBZOJak(%;GNeB-e*qhI5kQCsUt1W5 z=`Eu+OHULL$^RS@K85X>!ifE#J#i$$CbCdSct?1kbpv5^X{*}!adhq`K%NYhU{Ac8 z*O9Dm*8mXmJ%HePYBO6jQ3Gy*Qn7nSU%0pggjh)dEGUKtB<(Jykqd>0$iTp}@=h~)^4yZkM!$x)fkON`Il16z z{IzSNImaeIcnLqFbY%e<@k6R2!J!>NDxcz1(LvoTp@UFcoAQ|e6gmQNOI&ic z-|k|^2{eDNaD_aZDYeM)xHF0*w#ElT?`IFVtcwIhY}ZL&meOrNHXv|hL_nf+W80f z7TA%R@&cxhGiWP&ZJ*D756RLAQCspxv1`cqH55q@qc&&whP#XN+X{jqTKRs-b)^Oe z!^*#W7;f17_K&RsZlbC>wunIqltzO<{(fe)JS!WnCaNXG@u29i4;+K!c8ZY(1-ls1 zAuv>Ew!v3OEBcbqTQ(qeRju-z24xMj1_CNdrrD1*!%{pw)cD!^tKAatXt9j=?qvuO z>h_sY&L;XshsXdq#VCoKhP1}mRr-^SEHx~n$C~@jfaC~Si%?uZz9T0x-`}5D?1+db zbNn@zW+2G-wGi0Aw3}O=n+8ihicK6mPDDO=Nd=0QWzdqF`QR0Q>E+8jf~jQ5t?&&j>DwrB*^%Y;unjK4o-xpby_c*G|j#a1|JI z*$Ntj>R`-4IGSbaQiLjw?Rl>0fWgKpJgIwmXR%v&IdZPitGE zFI4{jFm_JenMG})jcwbuZQHEasMyXMr()Z-ZQHhOSDaJbr|Cx>UK6XUI0T3JLUM zeC|+=CbG#1MchAlc7|9H>5_n)xYhTjU9FHz1`yDsW0`3?>><|*GC1MoEgTP*6oiI(>6a06| z6}&C92H{SC;I~k7Eso`&I(|lcaF6$EsMHC=xUK*XlNI}N#=QiTJ^q}d0mbI7Ajae7 z%jGtFix!A)up`A)UNYZf=O}=hH2B5QRJ+Ojv37JU$!TN3gj1YgfK2}b9JmmBKWj=S zXYfH&6r)LeJlWqI%Vv_L!BTlG$#`9=j0j6Axpv1WT~ z)kdqO+^W*zp&m9#J7<#_P{{i<& zmm#z$rq>87lxy#cUf)ubi2rH|8!M>OU9oE)C1bilw@UOa#>hA6qFgTc9XGV0peHn| z+d$>4$p%erC599j;$j_>L2vM$!?xQ{UKPa0TSwEmhCvXP3o*-J!ZQ!jr5iA`^P0nXPtG&lgQeSXOM5!Xjvq0CSaj%S^5!F$=FI??t>LOZdo44rXic2@E-T80Y+Lw#cVRM~tMUj9|< zs0SBkQZ70X{0ai^)wRJF_|mrHCZGcjLl_d42X;8kVQ~PQX>_$hgSoVKgRW*WCwCq_ zqT?jvg>T`FDI71(D?p+1VO1Wx0w{09568Z|#Q;I8DYCnk!7j2zE15r(HS+nxaVC}? z?20fxh0hQ>4sfa;%{O7<3xOuK>pYg#BIF*m5N&q~8`n}ANL(BgQ;S{RiDofbA;eX? zg8m%mBIQF*g5^sAY*p8p4toGzR;@m*0NbWUni5XZ zc3}pG2MRx@gXw3GH+kYBr@82?xoBBLhKtMNc7No)vIVS~)N}^)_vZ3+F%OGc*>{*Z z%N`EBT_z^l#_(5!m0(97drt_yZuj-`^5$CFN;259bO9ol0^dZ>nhND?bJJYz{EqzG zIeXD``@l4w+B>j)@EqLrR~+g_bu@RJuDTyCse(7YkS>@S>Ms#s*5Qlq2Z^tp7Vx%* zl)n`DZ+fXwJ{(7$=8n;^;kTE#9UBOSwPw!Faq4$g-fv2t$}<8AZzSADO=&ZFsaaf_ zE!=4x=Q11*xUr^@RPP(>_AC0Xf_*KFw@Pchxo#}fS`dbZ-~{z(Z$s~P!Bn~pUK&4r zZ^LU#DWZM0ttqm8#D11O;b*DA!b|8^_o>Pv;UZNC*IvfUelAPsYc3;ndEA4~V*SR! zKmHXG%hNOvbnzOMeq#lD@QWHm^nge3*4kaNP&6_73q|ByZ~G7k0ey4O5&_Mlf`*FN1(6j+@%8$dET!2I{e1XmN$ zPaF|DB=~XCX6?1%*1pq1F;_om)xIXdv(f99V^$c@l15}jnh`QeHWg;(`-PQGsQr4! zZmYbz5DHldH$lG^id-?cO7GoU2Zqzr{rjxs5AgYL9GjUc3VF;}JwGWgMIVLOz=hqMC7ao~>kIse(MiuK zzwH&QjkyLoz%DT9+L3F(I?pdY`RzD0eK?oWbA*!n8}lcd@$bX<^Ukby){eq~;N$Vp zow5B-brNc>C+XDD!IwT(Kjn*|B+&y-VcW(v(IhJ+cD}{JNwREN{StPSQ52!oNLRExtgVz zvAg_7dva%rlCEq$d9&w5VwNu}sxuXr;*K-E`VbfxGau}DCij)S7~m=+TDoBPyWFp0 zIINm1%o=MhmJt`vqiF5Gc-w3%#IquJM;=SRg2qA$?0I_DJ|+pKjwlF)-wkg2SIvd< zep!qtrg-zL3OF}-H?Hye-$ij{LD`Iy?PGhUbHBFzw-oCkal6KQGadU+isk+%#YV!t zn1lPy8!=BCiE?1(u)iMfz?!lDF2#Gfq5t6HrK@zRx0|#^591!TOghbyZGkB?9L&*E zMs&l$9Ls;Z6@M#jpw6Vrp-COo)5&!_O+?6AVTySfURQ5V)5G|k)Pq7$Vm#T<-H(u< zSAAb>a(hYfnYLP4KJ?6HVJqHil$v3CA%6xVUw5u9xUSxu!)d0oP-BiqMtWsBvM<|& znAxtslnRblu9An4NAPT`Lti;74ttu@W99*3h(BV5tj-3DGkcvE7J|_7vk|npDXy-- zenIU6dP6VC&U)45w|ju<6ViNDroUsiS6MTv@l3&*7ht8Z#1lJel)f_CrlHtY|CDe@ zb2V(PK9Gn?h7|9d`J4fL7)%j0C!?4VHXm?_JPpb^*2IiD*>9t+)K%Hx1;KL6&G{i{ zHr%nA;YEOv1AWU3r zHnX2*(R`CAuTa1*=?h=6ZMY01Z*W5(#{L9|lX`}Xw^K#J10+5@ z6;Yf+S0?mBslk~Hr$+r)oCqL3^qCS90Y?9wB;o09jl74VG7$FhhQ_>af|hwSLpC$i zO)&3AaQqVp0e#oWz8$s<*_(CgRsXVi(_@$=qe&-)nAq5$JE#8j`TO}b6`24hTxK<0 za;c5hVOQyg#yz-|bN)sRfqI5meBr`1>@|mZ3{^IJC9D_6h)nb<|3|9}7bIl(L#9}x zT7#ltA+ghA+stAP#E8^OI{Na(RKvnk4}o(Ja|hFz%Q6u_TQ)*FKE4VQF;BqYP2vr0 zA_}+foK*@Fp!>{E#;uKN72NGG=EXBH}h98Ox=9PTU}4N{V1BIWsY{+A{t*p(YcaLb2Ft zz7)m8uNcJ)xB@>xa~U86kU3V5Sk2nlnL@Re6b)>{5d03TJ}02YEgPLZ?z;G8*L9Z& zhBZzzaE51;7>MkAF$ib*X(+fnV6#R#jXU5@4pbYIkaVC+q)9P8%Tzl-wCx13GC|Ljk>V|o$P%R z+dwI-uc#?KX$66O6gR4y0~L&Hv$`4t6-onD&=Pn~U7#tG9hCfp%monKfzv9vtTLGn6`!nr5==s2`C z(M1Jo<{?R}Y<=J1U(RUNu^W@*(pR<_ighMyXnrv&S2flb%Pfv)b7t+NkObB2S(PV7(nl+=y{}LJr|6R8#5iFCM{XCxz5W|gK{jY( z*8o3C17tE-EGXnfgiGLDkUl^qH&S0g2ZLrECUUGv>Nk)z?c3#gEYroGI;TEAeT-@2 zZ5z7@gkjI7e#?P`ZoYOY1VH1Xr+UJ3%d-EJccs9lEZY2`>cBT?N+M-WmoQv*f2caP>XsLWiU!(_XeFzEu|y|5C(^RT8S~1|#9| z(9#HGn@l`YzZH5~S7L;C0bE)rNy?_hj*e`&?!M|#UxF{2DdN<5Dl>gt-uD9HGodjz zpO2eJW`M^8@oC_VsRxytiO3N!f@`qMFCGmfOUy_Y(a#u*<}$36cnfw4st@e^tXasM zWOKQB)@d@3Ls%2FTfKV99Gj;aMZ$2b0ZRMvdglg1E>Ox!XiYh%hO(u+daN&1*flnz zf|L1tYl0@{p?nd-htD&m@=XOMF~)AqO@U`4O&TD^B)m|@U*Haspd;YW+GL<3k777Y zGV(6^a6Yz5Ra}cHLT*{La5JecfxpA}va(XW!>h?PKOu*nLBwpc(=GSJJB@tp{t5?~ zzuuRzXGnkAikys51t|89Ipr5e#&Wm0{vZmB{rNgsBj85%IPp`}8yHrc>pjg%lBWV< zjXO`VF-XZ32;7)cNP=+n!a&3lswIRiSmIH(bvZWi(KXkle2n(NTNFx2+1eK{=45Ij zmctED%*N%>g#odhE9})I=Ss?~uu_F&3c2m@TtX23lZ}+St3(+%jJ%UnvT-24w7qU1 znzis6r`sqztOzN)BMW2~Z=0tl1Oj2~hf#o2hqdpfbkGGy1Z)bEFF@jB8f*a`51bdV z_X}lUGcNGecmwpacw%}EHxWfOH0rRmt!~UHzgv?4%1S)OE3?b@t`& zDbRv;;*<=G+vbvB@*_uO=H=I>@i#EhiQw#P32`Z8#oGtGv}ZhYWlJmb>p?4U5t(KF z0xvTB6`5F9-5{bWSuX+~;??!N6f)mAL!v{H1QscZ!0SBbk}S$^ez0>NDA=IrqN^w`6FU;nOoq&0=nIY()&>TkK6UaqA415^GQFfMiOoVQaHg~(yV0|3_I$v?4AoQ;qN!R`-VFvQ+D{%D!e7zMf;#5UUM`rJ@@|P z&2#?wql+*HR4vU#*s!4*re!@|KZnM2VYXfc=|`5)m~1-3l)#r+H?*jaeora3FDYJ{ z3JSe|C5h%9`4`41I;iXO=Q+XwSAiN*oI!9_f#K)Fe{wcRuUGE!F;1o++Iw6f68w)U z*XF*ICF@mYKvZ3lI0=fOu>32HBoMk--p3Q&1mBN5PXc_7#G>V*nK$QTPX+R7^Yv1CEZPRAk zNQzgEGK^^qGhpA>qg2CSTw0kG%u$+Y~MMXemI_S7rO^={07B|D7 zg7-)~i|DN-=n191dN@TXk2d4BveP0-d%D<(A$tF6+b6qQ;M=iF#Dq0J)G^G7K?Qxn zl1mcjK~e=BoNa=JHD0R*0-I=CqNE{4SKd3#FRb$V7Ly9nm~Qyki%`HhL*(MAyEnHP z3Vc2+Qwq|edONQH@|(C&7PvD>^cK)T9C|q_w2g8J(~Nd;@&%CBdQkJ*ln~`)J=&B1 znsrk60!>BSZ#;H+8ePA-w8JuFakO7bF?Q50rUzULWR#{D{xnRg`m5^gPR&d}kQm>E zJ@z3mZ4yA^X~Ca%ngv?7YsXDbBlKGYL*e$ERTLr{n9hh;Am#HNjH|xM8ObnuW>jCu zIz9PPR5)*+&8Cd62X*K8S8dE9;RG?M+Q3mEvk9U-5buHFxbK!D-U^Bl2IN3y=Zlj; ziUYNF2qUqf6hb_B;4}sTP7GGhSD-9@9!tHJpEqt^Drj~`bH>Cke#^{YV~=SIk^Q3Q zv6VfR5wQx7!#IqC@i6D9cnVDrCO+U{+;kBgz>*g@R^L92JEhXG7K+fpmAR^)``Dwe&90Xc*#M4DntbTSHcEs5$!qwX?JP2(_esB=YyS!hyz6DJ>7 z%2+E1f|$hxX-!l&0Oyw(RudLi^a@v|Y_Kv^94&_|0br!6v7q~C-C@g1J01hTC5wLE zU(e2_q0B-UR}_kse$EGrTa&(#U6MnoCv&NkVqYQf7KI(QHW~|6eQ*Q0D(Fe&S&8eP zJ3iAw>w4wO*N%|X!(Xut5&gY0F!GyfgVFhhZ4$Dx2zz2!okQQS5j4^#vHwcUY!SNe zgC^13JKX=>qfqN8$VT&)8??NGieS``n zKl|WQ!)eynMTKQ{>ri%O0kw^bNhTFTWFzQ|{?Q@Gj+1=`m%~bB#fn zAgBJ&0F}`;*mP{HD3IUCT0)f>OS6VOZfLkNJtjFoRu@J($v`I7w!UR46HoiEC;`d2 zqeM|INuq=|r+Xs9v2om4Zj5^iUJqFVAe5! zDzf$UF$UN+A%=HS@_9X8E9Zj0^zEg@SGJd9l2QxZl=6AQPsVRK zT)sWq2GtntO&F*#Iu4BYntJgS@uZN_=X9QFEmRMHwzJO|grzX;II&4p7~h0$3{y(B z&8e?XDn#Upap<7{GVu&z%A3pFfu>QY{yNMtqJopHJ)sAL@pE|cNXtqB$PJulMhzq* zvVUraydijfLb}SngKD?;#o5@A^WY~gh@m8%5Kz#0mqW6Lx=|Zsw}KXvxBhy1AMeYl zs1WA73BL9YZUWO0xuEvFvc=5<+e5}2!S&8=hOU-VK@y|t+b@VlJr4wxIAMH2T>Eav zdJq(7&So4uinc0)?-*X?J=!`(y2k`Vrvmp3WIZ(e5(hVdkp_CS)-4A_068nLc*LNaD9LxF35kYi^6brN`AG-c>=ve&bB86-8jMq)yhWoP+OnFamg&y!+G`|wT`;UFvLLr zbq)7tGSHLm~BVxc*?B&15F7*%sHDh0h_I6{PuI@BY$7&f_}{}#w5T#j^bWLZBcWWQ|sd!2-;hkXwr zJA$?3Y1==*Tr^0IAW}K(-e;`Vk&_<|AM%^+Z*48r_wXfc1W%|)%R>H>WSeLtVDOZC z+l9={;kKYKre3Yv%OG?jM3e>V5!?gbmuYbfg?Nv^CTqb?&$AGh4#iiM_|R!_*SS=% z9ST>P&5-wF-C|+jxriw#DDmPE1v8cwSHmak8BwoxKPZy@mjE;0Wh0s|{u9vQLnOnw z=bYnrWDSon5lT6{K9Aa_lR zd*EMHEMVQ&L!9xpo7C6IogYxL`(N4ri4U^Nmz`yKR!Ecj-)Kl zx!O77+6q4lyT|WTxq5_Or&14_+V73JvpEKM{NAO{Rbw!vTUQ5%%AX=LBJt%S<>~I? zCDqw_Rk1^r?zQ?|JUIrRfw7-wxb%%kS)=Zlnvw^eRe9HCp_5h5?J?{-kzO@E8o?re zCf{Z;cL~@|SsANkl83?BEm$kn;~GO|9&|BMWEDD5s?7`k-VXPkz*aYVXl~D7ZYd*e z>TTO)GU~=1Yxp3Irl!h2TlI|Hm%)~2l})yEM6gGYge&q%S->i@cY1ZcG37EA#VhGPd379+^5E!*+>niS zkiLR~IGpl`g~HDwU>Qwj;$Wb2pEwN`d7{uaw*F?wF~Pz}jzvlFT&JLQ-d6(#+$vD}8CL(n+2e!UCC&6k$FH)*VB zZ0b~rl{7d?PpBCjSbYRB3z;vV+3YNdIBeEB88&M#glA?kqyd%^n%2c}Y%S+p9sIbK zeAnA+#XP0bR*%%v(bueppm5Z$J;foVqwCSTXIN~*Sl|Y1&0?S=){EoK-kYl?SFQyx zpVcA}w!-tzb;0BzB=g@f4A&la>?G0G#b3}+|C^*=PD$Mo%J|$_0Z#?brHI8hQjPah z13g0&K-+%6n@tuSl=(AVqI*53Y{!g>ZjtSM^E7?Eh4R@1sd1_?Ulz0%e~-Bp=)u{V zPy!@$OS2S#i&Nz+%%-aWL_V20?ns-6ZQG%YOajxCftz3TPP7#ErYVD|n8dDlxM974 z1Q4vLM&vU{mYsFhr=lyNGn{J@pAHq?x%`-s^KhBt3&VdwX_Y24C*sm*|>ZvnqzN>;k*TJ4y%F-DQ)zD;`(E*8XxqvI!*l&92 z{ftC@qL0XJqEx*SJd^UZlkE8PlxuDg1xUN^=F6lgN&N=Cz^O&+k7E+|z+A3v9bz;c zZt4YEGvDEFDaQ~}dc@3uNA4A#s54wD%kWM-RIIDJ^fPv^Y|*!2JcX4!3nczh8IJ9+ z3yU2=%?*rrph>Jb^|Fbex4e+)+CZ`qYV#zXv1|mvt{yVKt|+1J#1y!;hw7R;)qam| z{v74Q@}eGo`9YcMxRH>ctDrUh*j|Kw90(o+95!lF4r#^ks$i`y1t#W z$QOy>Hx2$mvB1#Eo9sN{Q74JLYp`B9rbxj!fXz=%w_HNFA}Q6zX;7s1+cG>xzqYQLJV>qww@C0n&X?|pX#Be*QnoBmcsr7`iavn?{^u*Mf)>7 zIFn5=lhZ(Y2kG&ntO;zN_BZEg`TU$Z9R@osq;-f2@)sj#Cmp&B-J_ehqCgD!jVQ5I$yq0txE0DQDC1q>N94P(78ls*F)HTgBEp*ZfM;#YT}VTfYIS|%iBSm6b?|$@ zR#4bIDTKj z9yzPDoeLqI_aQXTpxv10EA8w@`?}3$d_5ill~D3p4v1oAiH)oW{~@&g!SL|kpG%(D z)piTNJKd8omIsw97^A{#c~crQpk?baXPpEG%9m9eJkdep#iz2%#U9NdnE zE2}W0A>`0=uqVtPg0!ivvtl@@A8lN_u?2(99h4kk*_bux!LAA8*i1l6L zU0!k;u5g|LV>E38nWbvRB-oH8@AgPJM20$ix5hy43K8}doe<~hv#(VKcq)#v{V!A1v-WZ8sZ{ykv2tv25o1GBvS^6IOrY|#UQ2u@} zz@r6QX+^sm`4t4sYj2re?U`gIw|zE#BH?8*ru+7Le7;WaN^IISlzWB<`$&AamjFnw z09z>;I)SdEj2Q2B!{uyRtQ)8CjY6OtyNm0K{5nM71aX&++$FnvQRF7eC6?P6#5wMJYhZ+*{3q^a~s? zWm8i^2#DTiWu|Q5Q4Cb(N(*Mbbx|iDP^mLEH_>gv0*7^Z(fB4skuieJGOEzEl;40sUzsFz zjrfw4y7H2h{QHk-kyuIu3B(U{&TY+$+wBcs=K6{b;j!T4eo@3&8lDHh%Sl)c^x?sB zx6~HZ6E@f4A>{AY!v?myg=VHj*q1}mwc9J+2RTU1F7Ab7FkF$Ixu!R`3EgJ88=BN5 zncwmwl_=YA#ZTCPyosKp3hB8P;bSY+zkwuV1xpny344Wro|7kg2Z~>ivP#NX@st$# zh~51?Bz4Ck9B%VMBRtQdCxVbtiIgiTwJ(x*CI%kDEniF2Vp4$l>PDjne`_+= z|I>NB9b>J{F<-K2=LCi2zKp(4x?}DKV~p9(Xo>cO8`HiW*WeuScIc#TTay-XC<$J# zFBmF2;oomzM5-DdV)#jv-;&Cx)fUK%l^RJ9ZtIT5cdsjyI_amuY4-F~GZ(e&EjJte zmXvomg2ZG{+mZ@awYGR$O1n4e@~xEye$c2e{uPx)Czl%U+`dm8KY6+Vi8UIs#TS<&>XNNf{*najyBEAT&s3DGXbPp6 z2rg4_HSlFCRt;ATTts^nzH~*Pc>4XXJ#X|hF~b{XQ}s>DjCFu)x$EfuCl38}r`#bo z?u(fCvWy@RUohIfE|Sj6)v!!33=XKrbKXh6NorNMYIxO`_sGY6vf~Qg3sF!gjnfy2 zUuJNInu${N8ah46Q4FDq_QYbH@9#q+^XN??^1kT3^`ZuVj}tDwy`nasxszjCpT+c- z7RHsi;bc3YpD0%A2FtOhs&d+ChhBCWno_^^#H785xlcxktvpZ-A)w@oD)~Ig zgn+--RwzO}kql`m$wSY#swY?I-FWlUTlX84{|H%9>15xYRX~)Rmz6bxb;<^Q z{KZ%aYE6~IY8N*_9Yj=`1iV;Wp%>rVg6g}d0twy>6WoU{t*INQO!!hFMt1NG+ZD`<8^@P#NVn$*m_7 z<>vILM;@xMjFv`$7mlhs*;0@H^9eh@Hib8JXAu`kq&KBb?d7f=hXIw?pZxyI>(Wcx zMV-->)>>K+np1aJEh1t9f@{0h%E?l_2}qC|J5@&=2QpalcCy=sbpiDvxrgvuf#vZC zvzgP-5t$gkY$bXl!BZOQWJ`o8B%wdLw};LGFlO;UDSz8cAP%pfNg-0lsD7wwjf?5; zYFq+gE!55sd8P%oiM@|i%2QQstVAo#u`jwQwH?DDEasrvp>g^4nMFX!r{I5|#Py$wHtVfSOTBu?JXhlDIY=brl5X^#1lH`#W0ldx=qby`-rCc)>oB-rW%sLmrl25?GJTsowKE zFjV3Mja%i7s4*Ki;5LBHqD=^I+8#R`A=t%y~!qDI2fb&foZ?_os+i?1ge%2K$tV^9`kxd# z4k>{jk|DiR`A%Zz6L&rHSc0***^Y^`pU)ByXPX=q)%%*ILO*)S>9ySy9$s$0d%GB_Z|vK{fv@gI4_ zk~zAus~f#`TeRTWPgjtLKneQfpXC-{_@q>>jC8N=nO0#~fQ0x;xrVy8NU}|=itG=Z z^l}Qs1aLp#Q}jeyo+F^*^=g&k5xvz_MxgxmNtb2B}XY>pQ?v}FWy_NkgD zM?A1NheCj_y@1CvT@6HToqp+{UcH2LnksS(LiB$(ymNh{&bi?3;d+r|@jMSks`1Sk zSP=slM*S8z*>W8X4td>2fCI&UGF!OJ;0ub#PP{8XZ7|JD*bLz3x;UQ z=41^?hA!cf1FNved{8anHUXtZa*~S*?C*_J&I9G-igJwA!-6prx3mdvGS+%t7s*v3JJo z0Jp3z|Ef1xjW3q9>z46QR#5vJy^84oX9et^d|t{#4KlrWZ4Z!hE30Qiaeup(i~1iK zA<9sOlh2*V%zty~zRy?d?0Np~>GZ9a7OsO`hK7XC>Us=y@JVvuzjvw`a@o;~z|q7f zmcDP;zJ{cz$kB_|38B*)_AU?ZPXr4Nb}HNRYYAh%2@@~v#{Uz>V*d{qi<^Vv|BbP( zb!Oc+xe|Il8YpvL3c7_K=|W%Ok51h9tSB85{_ua8cWm1e(lc}LHk?tE7qR-k=DgIY zg9D08$Fp4;T74JoNfL;FR$^;=Dgym8kFax({^&tedjl1->kivpAMdP zN7?^lxIKIcdju(F%?4u{8%NI;tt3u%g1WujT&=I)KRsxAjK}Q6^EuCGdE7cV_W419 z77;`Wx!;5o`QGgd^^1v|K`cmt`WzZV)&P7w4%n#oia^1-TLk2l#%mK zaohe$|1#l8RB3QO{&JCk<<@9b*W}DS7f$)@HXKu1oB43)k}_EQq>H72uGbDl`cH8q zwOz2BohYfkSYCGFIgG&VkGajX#x6aGx&O3m`eRzil~di_ zdlkZaJFYI&Q_Q}G2oL<=X(y1K2lQaEDoYXEuCH!Wot<$GFR4%QvU=vUF1BP4a-7ZS zHI^_W|8<=)a{2L4^FH)%&KJ8JB87W)F(!LA*KtvEUV8d+x~icZ>kUhtH0BF*#i_}1 zp<~cq^71P*$EbJ^@>*8n9SmB||MmI$G5B?c=I_8y&*gz1=P?IIfBuHGXkmCyiB;0I zl8oQHb=U|)FtiHmz&;IDXbFH@Bp9-F0|)pfJ;H#WKNzA}*mg1ulApUIm+W8Js`fBu zIK#HU5)5NHrjFCu_gG#?7D`;F?R8`|(DDE4m3fq?I&!Eo8UsaZ$0hSYI z=iW~?r{4&FIwP+8>m^Nr+Tx+IyItF9bNbEI*LR77%U-% z)PD?y#cO2OM!OCjt#qb+np?X%J$-HGkHUi*O6mGFmoy9sA`3NZSy|Ode-9!z%;Tm< zpK93J7^^C#&=d^PbZ?5nx0{zEes68IZr4`|G`=n%X|!hNLRJ<2q`3^hewb^+Hbo~X8FhT&KE0Sr>eK>xeKX zPR|v1J4UAGS^=2sAB7P_Dab%UwrF2;B_TtLm%?O%1d;fEbNYLs4jv zzMq4V!mzLaq`W5G2}Wohpy(Yo{n7G; zEGO>cCEM`KCdO%#ply(^**{mVo>G8>9h4-UK~GfO@u@Vp6|2t3;I=)A-HW`IQyxiT zByQg%alj+I5i%-kQjg!=@%9F%G1srmIBbqa3ka%1DiG{Ydr88F|6twx84tx|6ms6k z!(pej1RodHRDi!hFJ^sVTuPrRbXhxpWB7}LCxQbOKnw8>f3jEHvnm z*+QZ9<+r*ZWaZf){&Mj=K3r3bt`&63n<6o5w|FeUcKSvvc0p%1LVRi6*ELas3H?!D zg>j5>h-PO1jQ!j0tep7APoD$GISE-fV??0PiC@Gw)(C0r2m=!^OyvS271VZ$n5Aob zOuey@lW!dfD>l2b5(j5sdJnv}HvDgfE@NAOUY(*!mc%Qyjl@YvoFcpkRc}Mc3q+Y7 zc@9BRDa%NTcKpN{e6Keg(Xi0{%Mt{@0|J1Q^@jqo>W&6Lxr6)8c%k`0xAFHO%24W}=JzVOYKBDpo^;HlW zw;&IiId@-gxcwEvkG7vvDD{y218-Q-P_T>^OE5@H@G#VV;t~rPNcz-T-Q9Frv2X^Q zusg~*wh1}_0Y9d1yRM^{2bjg6uF$=46>~P$!b0k#4h@*uG8LMX7SmUrip43~s@xZd z)(ci6c=C?}B+d%UdkjguFuQXc(30b;ijI5jNSAG#OQ+b~GP zuU_kX_<<|T6zzf5xSYooV*1$jYrb0=FkRdQE&JeWYabowgrqh^g{86%X97GTRO~A@ z@uA^*7mjBz()9_%F@k#!o1r1gg0%$F?(|3UD}O@@7hO@2p|p=--{t)v*{mXH$gku~ z0M2#pNl?6Vit9IWR>Wjw@^=nn8d!bD=!tQoAoj>%eC{bZ;46qVS3Q}uv}2vluM+Ui zIJgfZN8>`}xt}<480t_Z7|q6dd+)s~@)CW4Kh(@evj&%2Y0^N?X;k*Cv@ahVW4~dc z{&{qQ0ow7aRMQyOcD)!{qw-D+x0}|`?I$2C&FaAS&s6Q~!PgX~3hhc)E$?Z8n?Zf1 zER;hj*ETlEw0Az2KWuQ(Webj|)@}rJq&Ltnkq&uldh6RO$U$h5VsSJ{5`qDb!;T^) z_5=if2?^~zv1u!rFLjbCTP{kDl;fDG@y)?b!jGVO zL5lFYIiNf6b}VlrJ=+uEq@Y<+-9V+$k-*?r2;e8K z#r=IcjS}(4D0VauFx_mrs@`!i}%~?Lko%;F!nE* zxqrDHzT6&mH~wWVo!^yN;YeEDkTUQ^OrH?C5nN1megKkKs(qnYG_ zjO4{hsIz=vh92Y|iVM#F%PTX$WQQ~J9vC@PmNL2l(<=2$@IczRO`rnI8mOuYjtzIE zPQw6NptqxOoUo@CDs%sKwqF@YV`#etR>BKhYf9ypRYGpqo?CHaFzMvJLZOsifO#?8 zrG>FNWIIL%j8P(>wpLaMaF4%NLNA6(M{^CRix5FHGR>2@CrOb5bDFFEMv|gmY+sw_ zKBMKc>)T|OJxF2udxEHqatFf9&~7==DStsFz1W&mx5hP)y`<0LW678g9{%rv*f*eS zkLr2X)CMR{ab)P(#Z|R>qTx?Ahj0uoTgV%g$Hi|4oj|*Ao!l{huo4m5i<)4?wYd4p z10D%^BpQGh|MlPPs0p{+FqQ-KPCjo>_Umf>#z;l=%@SbiB~YCtdCMwkkkLl~14Y1{ri% z*KBvy-+eVGvzU;!RV+{j?z><_j*s59ZIdeo39cTu{fq3~uBvd&WrObgnlo1AL3idS z>991aKs2|Bj_-e8dzns|EvDo>hyyMzt>EZ$!>s%WPP@-!r!^l;=xVGhDNz865khg* zlRtd>5*I|>>uGnXQqSFkYfnJF+iq^w&*3V{pkP8#IFP}+enhh8`{G{FuPaTrkR*>0 zCkj)vE~+Y$0~I;C`}Q>VH)fiZ{7AO_O;h30!xu(GD*0Mvi}N6?N;F;|p)cy6q56Sa zZ1rW~ODwDlb_Gg`l{q9DkP^F#gt|&M`GPqUdf)L!=tU;@shQosu6PU`oeDK&*ZCTL zIAlp}tS41|2RaKKThUMpEaVP5owo@_ULof%>M{q6SD}guzQ%cHh5{n@lml*4mtd+s z+W!naQX7+&4ntI-s7UHSQcO%k)PtS@m>%HEt3Z|k<mQ)PoVPp&7zCvQ}Hc-LA`~;y0Is3m!>{^YeGFx~%Op`Kw9g3iS}1N(*x?o%CGl zql=*InEPQ>$D$X)`$V(_hkK32)bd1Z-K!{BVY4bIFPU!@_{qg7P!vS_ii^}({zU@c z1C3oB7(tNk#fL4jGo|i6iX=``L^$sm8et{a~U) z@PYYnPrYS#1JLsds5^O<*5Yu3fl-|5UH;8?5Oy{iG%$AvTw z5n|$aaG5Zc$xK$>&tI8@BBOYzBs_8YD#(tV55AMVJ zw5~j`8A8p$wQ4%2ooR2;z?L%a`VK8<1mTz6Oa0x$%vuDqkS3V(_BnliU-N`)OZQy*Zp^V4Crge+j8fJ9 z`P10UIJ-h1cuKBoiQhx~IPwVp*NBoM>6D!le%{xZ_0`xTTjlwPp ztW=^Ze5V7(t}Ht?n&>R5j4i;}10>|7QfucREY@P=!s9|s6qC7SM~Jt1#$s{>u=Hjz z-&+WNdWp!+$|+)YdFj`ecPHT&ip4c`U`tBBkqqWb8KJ(4l|2MW7#9V8dbFeh>8alE zoJU8Th!?kXlg_Nz&D)!Isx`oaPo~|n=TIA2WWKzMQwSx~a*qr3{D|}ZbDfkBSSLqa z3?F96@J^Y-Z?Nm`vI39Fbh*~}jqMFFbxCDvjb*8CD*E>_tVGNA3>U);2mMlr4d#T0 zF=cWdj2J`Wh3VI+EfvUqc^>Ft#Ab56RUJR;r;*GOD`HrLJ3(nFiaZu` zR@SMVi<2LhQ)BG9P(y>`FY4X+Z7UwHyk+9BF|VYdGS>2zA*s@fFhMlzH0bR74+XFQ zT<8E{xwetTNaMeM-sRxXeROpJr?QXolo!N^4%Q~xFO=UBLVPcp*1AN@s{4s{oi{zc z??>bCYfJAunXB%K6@7L>E%MD}5!f%>$V~r-v2$n>EyxyS)hXMyZQHhO+qR8Uwr$(? zDciQqSN*292R+*VV1Jphaz$VRCtr|#;UzZdJ*T^l$|fB-hkF^3?MC)#{7nnxgG-c- zdtKat{YH~BJvIApz>f9*0y`FZ*8ibuT&*edpHNHp#r49qKxxc=D0n{atwN7P!w605 z(sSDR)m7ceub>?~lXwxpBF=mDr-RSHsMc#WRYL_pTqtKP0!C~MU!Gjqq}%(AvGSWe zhR6eRf$!(X&-cmi)i zWJ|TP!{ZQ4AuAki-;mdCf-cb_I(J<%IkI9kS!!iqUE6z4pd4Ko0S!N0SxF88K5D$~W9)JRn7a5F%0i+ST4m_2FwOQiPc`PrCmRrw#+0VMC}A zZooH<`=7}@b?S4p-iRXltBNOQvV$?wWMe#QDtBV}AKfC5DH7g1#|zHk$dVwG?z}mK zH2=BDKmp4}Wp3Xb;$WBb@Pg`_6|d;wA>$3@%{!_l?j``4_L^4$RWVHkYlz1bX}h7A z8*WjyM#G*R+UG*|d>8o$h=(42H)B$KL#8=q*4G8!?n>Nc5*|XWNKfAM5(pQY;Zn!? zN&pNvVB16NvQg|oBnCb`Q(=dxxmz=Z3>J#d;qw?W7=T~7%81?wxMPJ`L)>FP+sE&s*^=n&R_B;r0 zi{Cmnn+ji#NUkqIt)vVAot+o8-wygK2Hw*(Uj_Vcc(5(4q zB0={nwtazT*Str<9Xr$F1B{Xr>l1|=7eF#8fKpNcQ#7u5-Bj#wpI=3xr29-^$@L@A zfjt{_t~Yd)7&wZALH9kqK?vPJieZgN~ebZn*ZxN$Pqm8Y!Q~FXGU^l3~;d{~}-;F-n3ry>S&@^Vdfd(Qr zfgER3VQ{M}6x0&c2Z0TSHjYh!yz9w^I0=?bNaF(qQzsbj3JOeq@ob~&2`1Cdn;Ci# zA|DST;6jp$pgjH~xfS9Z^|MnXiZQKkLBr-l-?4#=6-S4?M^dS!zu9bx*a>=$!7BCH zMz0Mzt-iB@+bgglO3%(93Sa80?*#?2R3-l{*td?~&_GcrL8BJJK!imSeMXz^e&PTJ z5P$EZ97FT(463Tn9pX_fVOV#=GVms&!EZ}ICu$ZG5zUQa9(YRq6ujmP?d9i|1Iip%gu%9#&reG>`?3K;kr5Ka|jS0 zq2-!?&@-2Gm){%$SZ6v4n!lO3510OtV+pN zLC>xjMW%NPWfqeqRgWXq;EXSd?&wGB9_E}`OSx<1EbT-891pko%X@E7OK=N$uZ-RC z&zs&+>z#ahCiz14Pmjx9j5BC8Kvw|RSIwtl`z#Sk==wZTg7&@+s2LxKgs< zlZ4kPnA%be%?LfD5SB?Md zpVyMptVXsD17DmN>pWtTMm+LH><(NXH{s&}HE>PQ6YgQPr=B}D>T25lsnckSdru+V z{2?CheFkCLK1hf@dHF3=E#VlM_N_Ms=uwDLMGsOj!I&a(%i`eV1h@hIP79lyvmsY{ z61l&u=TAJk@(^a45fZYQrB2l8(Wh?7hSZc?GS;XVrQ5$9B5bC=64o`M8N_Rs-lT8&?)GIm?LutZAV_V z4Wk^XbNig<=&0CBHQkr^2J{oi?U7UX5MB?q?;ROzA`jrXGRSwq`5cPwU_K3P4WU8f zHP#f#4G>}5vi=QKty!jxKZ9v;*CdAfjxY$Yw4}LkI5v>REk-5cHpYw}gI!cC=j1np zI%%1Mn^gA+@OWNgty>)}IUjotrzhfFVc5_bGwz6_e;lv;}g z4e_!*9sfj;G|3X-(@K#n>ZatwPH<(QKQ)erVeYR>p)mup6 z#8rf^ljqDMfP6h5B%Tklg38!7U2%MU`oUtRrpyTdv}4_tkU&7_Zh)ahA(@KO(Jvmp z_SDfu*U3P<@6xTatfhwlp~O>%rYKRgGzstC?{z!PZETh#c(UfUmYy<-orTzK92=%I zd{4+fxdXn>e8i2h9}R(tNCe4-VkKf5Gnrhet>|KeopkHph2~XMXFRgwp0QQP}_aWEFFI?7lxDsO4*)Rr(8A_hUWLXxAM2o5YyjZ;234x)3 zvox(g(ykm8we|=JTg_t?XT@PX<>)!8>$fAK$P?UX2*hd?keW>M*gdhin^jzj@+ME* zZ*=1zAu+6ZBheyfoXj~l^H{$Nz`O^P z3&VIZpL^1uGh-%~g|DCCY4mZ8pbD?Hp3#P4q>u|x5}}su-Vi96d2^7Y-nH+~KB{Rs z=ghhKHhY1&b9cUTs+e<|JKx_k&}nC68&BjmNFSkaxekBV*{g`WcAFO<1jZ}TO}7$i zX1M=LEaTKKsnD@Lvg{(g5@eapgi(%)zD)q&%Ofr0Q~}cjrwv8u3@h2}!Kmx+4|Ksj zCKRiNPX^1trX`7s5>gJa8zgY3<=>pcQo-8PCRXxImLM00*y(Ti2} zCIfU58~$WCc#+#dv$68qV=uC7|x#1-Rqm7Jcfo!J4(E%4!@3A2en|cb^qGz0qJ?m2hDk@rwr{4G^Mh z_r1u_)NzIOQZHq)_F8J!`LsCM4i%Dr6Sn<=>WA0+i}QR#TkxZEGnT4;&V3I_RCRD& zOpUzP4MbQZ6VF*oDMV4K_3q9mgjwLdTaR9FcA|Y=rF-km(?Ap^0?rZU*90clGZ57E zJ0y)V_en2e>U^3_QqQ%Si{P5@>3V2Z!3xGIwyft4C!~v?-?1B=9qbs0rT=;vW054k z7u0+3RH<$80X%M<2r)mrK)VY7!s1(Fdo#JZbEKeVe@5!Y9u- z%zpk0lC+LWPdiY3-{N0AHKM zdX#F?Ca~v^(vd%_gy<6Z|-^`Lic{UKQN_ z4WrsAY!JwR7PM5qo=JM_R8*#RUf_YZ7EMI(*o%G!0B;o_PIZB5;Be}UKb((U5TnpR zRB+BXI(k}K!n&uJq>|M#y)2oZcm0RN3~L!oJPo$px6|CCy+=s^p`tyN&%@yK2JLrm zE?;&7xvAE=M|)_g)S}~q19h_1?=gK7D{T;r_Zl4O1s^yJ)~R#3twkWqy1>ch%i5dtyCjQ&&B4)U&nz1+_(GC;U02j z1lWsdof>V0OWtt94GtNq?v~{RhuNYt$>K?D2rHQ1BM%w6$F&AhjQO<7Dw-X{(O&jU zhh&Zw9=6*bNNa)sRv04w>aa?UnyhgC7{fr{80W9CbYC<}Fq(5|yd*p>!Q&#}mA_u@ zKA-#7-AI}*<+z%g`3GjRK3Fc%HuYKStyz=1eok1dS$R~tQO{05I%2MiQK{q6i)X2X z3hZX(`Y_lE%;O(c`N-0~pUts|8^1p+A$I-TRi;OC2_YW>Xvu)f`R39%`3L>PQsGY9 zB_OO*-OlHtEMh|sIaa(^0nT&IszS@(q!uOA#>@>py)Y(f=o==i<$FjE$@D2G2(@O7qN)i#Q=7wvo$D*GW2%*2kk8R&z4rAh9K6UXo!l ziUHTG$^!UNqjB?MSn+>mezKP)4DqbER{Z&e6wg6w7}$^P;@3)6MeE<*AyD+GAY&FV zR?wETjbR1ru!|`V+&gZ9EnB`l)h&~GGwsR@im4d3XO8?O3qbPxsYjj~G?J14jI4Y( zW|B#J7Y2iBUiA9oGs5PHltoS{QRcLFjr%suHWxiAhzTtU)iwIyiMSCi4Kn9LoHA<( z>TWqmA5S+bFJgYFt|NA*n$N|eE8GZn+jA^8!uUv}Hsa~xavf1iE$>Eh7Xx;E-4(l0 z{@mJ+<_9Y#024TsL;Lb1Y*8Fp-B)tti?o@2eW4m*bkaxx!8m5gjv{HmFv)!eRBDSd zTW7IsgGOpK4k>uVRK8%5cmaC3A+$^e9KHK^&#N14e$*GMcO>o`YpVtkt%=cb%<=W( z6`>8rL(=++*&GiF9yh6faxLFNH;_LThz|_R_RPT!7AS<;!x7UP+J6E7V^;tfIpr{AGr#I!b(!aLvbuEES6E2Ry5e|U<-4Q>VRQ0Z+09B zX{0=4<_?-jQ@YB$r_cOG!|_V(H>b+MkD`ADa{UMs-dlBzT5TVKGC#xJj70;Vz`_c@ zS(V#>2gyFllFBygnQBn2EI33M4}Cs?XTM`~@LRL6xnQQQZWE~J)MLB4OBX+`8Y(sC zOUY|-y(yI&e3Jhz*7#ODFI~-KhtTXPPrddb{Lt@wkuF<-XBI^bMc>8xmyuFG!H# zx9ChKYGLhc;`sZuHgGl(HZig@Ho@oRg>-UuG%>J&bYHLjZ8u|y{YO=oc;atlAY|^s z(<#Z8kYF-VG)2IU#q}(0&D>8qhdMiL0iZ0-XYrzDGo*dYUhk*>ps-}lkI!Spg>6vK znUk&YA5mR#Z_K@Ptm7x$cQ#Gj{oL2wPAqXZ&M;r&(dps)=~D7?VLV9b=IQ%&xY*cV zxO7O#7)Ufmo2ye#58|1cne8l`xLUtZfnojVl=PPxe}|P;vHjrRu)^scqg9l)Q(Uev zkE#f&xBR~8=gP&4*8++LL3L#IFpKQ(QfhK)VZ`rJDyS5Vol=x)VgE<&kY6;Jrj>&V zPc&ImJla)-ZtIqr<(MPYH({)uZO6F-_xPRy*ngH%U0f!41aJ5MvYuJPtF%^Z9Mc`# z31glmvHPB$i)2Z95#NC*t}1IqT&LtesZRTQ;SQPWfF*C3xuY zYh=&hEhmD&447tTT2z=#4~xS>;S9qAba_mCl}P`rJ%|YTHPuDHt+^|@t8=nDDAMrG zT)?_UWsaK=Y%5V+Ye%we@vScw6}4e|r2!%nl*I*OKr+ybP7HR1AjkcXk18@hLhp2! z^7is683(^`85XGXs*r?T>NM$$|WKC5y7Dfbbv)&q)+B}4P$f(r_ zG}n$If%(W2jvhqx=hkg0r6uL+4=y|%KQr+jy#nqdO~uN+8VGci_m+25U|1zJrW)#` zMlsJvwZ<87V$(}?uI_rCX@yCuzz*emApzENpopKE!Azm>J$|8ogHjZa;DYw5Vcy|{ zdjmq!&kI(f*fp~MQT?g8TwXNR|C78}acTdlk^(&Jr7wVz)9R?}Q=vv564Me)$>I;! zmizGIgD(PlBm89I==#9-Q`Ixn<54rrb&2JO`2tkNx2y1w_uY^y{}8Z;F)YCc@Dj-s zq$r{>Jui^CKrC5R8C%lVTK7z|&UVQt*|PeV+Pqxc-O+>Ns}H=H$~Q56*BJ3r6hJ2|3A$w{20e!_NGUPeyX4*^*l$|m_b-5;pCae+D$LTv}7m^&ivt~eRn{q&I zrnY0!KdH(Xugf?-K{$ajm<{m@Q3vB27T%zRv9~hSx>+le8QQ%NDblaTPF{M|aQRJ5 zyiOy~@*Ux6Qsg)=24Yaw#k$xpy)@-)U%VD-`u4_#ktEXUuJ9AY~+ph#E!_BV#gODdjP$xdZe%11L!!^ip7~lMhrD$dNB6wbFDpbx8>*kL5c$D*T zAdZg|&wt8)Foy8guMZ}fL6|vk*My1^CYM( zxu5F@A>1Twa;grRH0xXW{dj997h$B&xMc%s&SY@H{JH$pYG&j1Qc?S8JFXTZA4et| zBz6h?s~#H8uR)a(K^%Gowc%PGN^KSJSke6H)6E|UT!}D(NZF=rsfBXXk(SX$BFza= zxueVhPU_=`PLQ+VXY~>(*de(%b(W+`C`GMNr9G>Um9$x5sw;`XyC~7P%Uzx6C&R)- zy037CI%HxDd208)W=E_STv%?Vfm%b&?6bejizEFH6~DK(fA^BHiSt4pu$4H`v}onwXX2M5!jnMQ*IZlDN#)Lt=H zBT$woxtf4-%_vudB>O}6LK3HoVPm)-k!ZV}#_eb>M zBDYOqk61b&k!6BlLTD-xpe}XTCRnrF^lQA+a%wCB@B`TgkHWOHVv9Y5hX(c-*bARu z{hp}|sSD^CB`i_UjyyNL{(3YAvje|TR5WZ(UUnAu#$KOUe0a&zl@gkxjx$xSPd z3xK_HM{@}rlUa>HsLt+)94$bVp6!vX9QUW?qe71xF@^%0011iI+utWkdD^;%x-;&z zA}@K{AL;xYOk3Y&Rx>&vZ4o98!a+6&5G)b~pW(Mg;Q$K#WE_~8MeZ#G_8uV*)NH6N zH&|k?jD?!d20BgRGuU))Sp;P9kGMBbZt(~-&wZpJkV)Y_>~We5tXIN-m*~Eo+&qPrsiv(`Ussw@`3 zbwL)rWe7JNmdnqLZ*8O`Lqj4^M3eoYJu+1I>vo{Nwcqn{21XMhz-40;MQ~Z)Y8y-` za!-D>yu1P;p5-t}j;!M5$~qicL9wSHRJX|K1Bg^@tG}pdQy(&fF16M`Ft`{#BRFy}Z2efcI{;cuP?96>6ed01LV%?6fC#mPia#xfmZ=ODw_zpEtRQUUpl{cXV|T+|801BQd#k# zM57OT&mKOV8^2wp$Xf{iVsO@N`a_SR_s&w*8@PU^Q;YgP2b@gmrzrP5@-T;A8JEoX z+qAaP;v4YP`V=Y;8}|r5S?K{6p4B09pDN#uGmEH1(&tm%cS)t1xNJGydvRw=6dXW0 zLqa3^ga~@eDLY)B{Z!J})K%wBoXlHY;1!wkPj~}yZ&(o?QB{y?BujxdjQkQ>PVArj ziWOu{rcBytSt9H#JvM@)dd3Qd)18lHFz>UDay4jXrQwj+3*zkj#?L8Z zs9`MPEc!o@tHcNpB((Uj*GoN-+59)a(=dRnDQJ0Iq!ZT_HDT`i9_Mv%%AESIEJMwk z3uMCDh4KI|2^(mVisBh)z`Jb^J3Ika%;`6^Qacki@2J$Xu)E`^AxFOYIKYC(t^V=B zOvP7Y{*YaN2a#+33@%+?z*nzM?qd#v;rf?#5qu(Cff9iH0YHAS*cttEcDA15d|9#+ zLo`um(l?FIEM%n+>VMdgI}`CRZRrCaMx2(iHXuI;5XR-3@St!j(hSGVsJ7&zR!vra zY1&0L>7~YkLVz;Wq8F41&~K{uJCI~AMO2b~Lgs)pze8`-h>jIDCFSD=vt2~PU=DV_ z)!YP#0bM8+%p!T*Hp*qSzq+``@Lxc5uBU+vFb`T##^Sbu6_Nyn$MYpAp^x=v1C`>A zpI|kO!g6-!haMdrHG3@@=OSU|&!_@3hvx!#5R}2=4m~qHAUC{Pm!rl6R?^Ox&?v@kL9hQ{ijz)#Ef z3`Q_p!2OVG8P9}^_4}qEeyC4v zKDw@(%F-twTT~_nRKUKh$Q`}MGcq{27zMOg5dW(;BstLkC*Xfa$W$-B)Ndtn@a0Kr zY|@BHrzM?oG@3SB5Zu@A9Hf*kh0ET81u?4i{dY)e_6ST zYF*$1f@4X@9IJD$((U`lpdW9qIh}lz30Y6NNhv`us}=dT?Uj^!X zxE>ID%ne8qpph689XyWq9JHQ>^dFB3Xl;ESKSrHv_Iw5+vPsiM)CN6GD2BRZC$(+K%lt{%0rqWrmTYm97;@w-wdkjMY1>~#`7KMG zC7Ir)w~%1}e@R>8RS7qQX2E1~KaQp<9LK?@{-=)MRqZ)(5m544@GH3$re7ZSHa=k_ zD!6t3REYQ=e4p+1yhZfS_Tfn=X%OT^e4+e;>e>PP=QpZ0NLG=bQt?kVR`I6p9TiyX zaZ^Tv#$-gL%^o6oh)4XfsM!pv^b553czUu*nxRW*t_0%Lpq-FVHNwxcaZr*2JTlPi zpOwbfNaVy&Ow+k_s^+f3U{o5-TUi$RS<(b(Py%NTUj3QTTi1!^W-*E){>9b&T&=JJ zDgfQ8fgzIk*O9_;okAg`y}0)3qdd>5f$?_TEZBLV`><%1$(6fe4xLU0eKMU|(KBAn zq4~#40{W9CCDbPvicPYzRwI1grdt6+Kc}R+e@y$EgHp0$NIgli+6NsOn$}S2i?oE9 z=kSVNaA$7>SRetR6FWyw_XQ^H4CX-H24&2f)V-bqQmI2q(C-@K_#Kq|pm?jqqHPYU zW&DBtNPMN-14e@ohedrhhug@~&=XWiGXbeCMWobpPu;Fm7D(PJhqT_14hWT8^Nhu(r9fk@PsV;?l! zTXt}|v>{pb?b?J#PNAPo@c{RBg7k#LN|;Z_Rwdv!2qqa*$}-*w5p+p7$W1hmWayDi zI11xO(Y=Oq)Ko45?64=zB?04~Wh(WxDT0Jj6pm@?GawWmVeD3_qUX7;wJGZ{2B01f z8eFYpxH@b?E+IHv2&fv62GTvTHPm|qKeElKc@CKY9L;DlpAS$zezyC?; z@JH93?IaHO`@;N&jm!P^<}@BQNiZ)}&39!uXx8TSydVWfSC_cwvfcci*NCFK)Z`cVs%_{p@!bs~>$9S1IqTR|nk|=Tfdx1fR4s@ zsvcnX-inVbC|&!ww-%Qo0ewz@I`kyz{hgsW>?2s@1<$%WF5kZ2K|5-qCmFv4q1l-ve9pkaDm$!k)Zv6hg8pYR?P{%TqF2!zwkP}#lrs^a%BI% z$dQ?Y?SHAbYDu~MR$RBOYI)pRJ!HZ2bNRiTn+>x9Z3_jvXBt?`TRY-pHx&nRwOdUBqV=^dmol0INQ12o`mnWBtDt+WOTQ4 zvUIF??jqbKWNi7egE1TvK7>7rNz)%9T3j7FIT6rS&}!*sN?UvN6|@=q8{3>?8@VN%3f6S(S)p$|~KIYW@YtCeLWGBnY+iVasl3 z9{p9I*;&3jd$+$B>e9N(w8zL5qvW2Kbk6RV=<*J=yJOi6sXTeLr}0QrqtITbuRu)wAy6t1y?HN@4-U7#Ap(mx%D0JtfBIthbAH&X| ziXW89O0f2eKt8h^v5{YFb|2YP;PKvCxEOiX* zFJo*r?|0!547+q=;|k1PfUE&l9}Sxz`j*;9uBc=sO5NGilib|J3zX~uJ}Fo!Fl(+~ zf7q4gkh0%lrNbzbi|}X-OpXaFhPmvrri{8f%#?sMdO!Jel`GR@TsVD)vJIPRv~o{r z5#inmTE~n;GRjIgIna-BlW#TRL1XGk>3G(N}2p9iXy~^Q}u!Js@Km)Go}&fuE@Bkg-Pw^ zk#D?Wy;G+^B$+_-ARGnSTYadseyq9|Pav7-{0<+K|B?xWTWLLn88X+}FnmQk?_MXd)7G;$So3h5?JHA~nO(PWbFRy!W zY-8!xQCx%1xVI`prN1eiEu7wKjtg`2+GVkYQ(ts8oGA!Ku$eTIr{e8JMHeE+T))MMWhhZcbp<)o- zcX17gfUD@U0xQMy4iS<)l9(9%qOm{COm%@)Rokc@lT@mH%~CNL0gb*CzM4iW1ZjuW zWQ}WmzupTVO=G)%ZT&+f3x*0m9s(eM&AINJ2d63co7JtV(XJ$?0A;%?w$I?POKXk%2~vbh=4eUVt(% zP z#E|C}%)&o*^uUh#wcMxQjMJ%+H-y94B{H7ON-V@gj*!O4E89`hkfFdK{7q5#wg~s! z`F1BY(XV`w*pVMjJKEJn@$o{UEiIaNItk8DIc2<+#dV%rt`mVZSCP}m`;Uk$_O@8S zI>MFi;u4fK0l_R&0t|SnG`83Mn6*#JKk=i>+6-S#So@i>y`Dvnj1jX#a?i(Yy)pRM z208g2;#AvxfLu+6-=Z>9Eyy_ksN8Ba4#BisFeE8FT;HUQ$euhHQtTeN$rw6Zm+@Ay zrrG`yZSWWNp-vTqXVx*p+-71S0TYnYj3?4^I8Wpm?3Ec)%A7$}L%(&0y2yb`v`VCV zcs-yBd(J-FrI7YLyWCWSD2hzpbQh}p8)q;#fJwWB;8>7L@`mHqeb1cQUt9flpl`>! z^Lx&Qg`~jzCgUgoM%XR-U34`kkhF1(tdIJP~?IO!LMw5;h*H6RGy z)nh3KFZF%b3TAjp@ZSspkhC<_0G@{X@VewUoKXocQ$%L;UJjLNF*24Tk)6 z#7UjhHM?!%!1}aFNUsR4E7D#zA9^$sv(MsDXePI_oJX*8B$)VA4RveNY!c`BROyX? zd->bVtIEV4Z$y|GGs#HRK2%+0n})+x*0!BO!QeHQ-IlUa69TlFo?@SqGlG`){jTem zdiOv^!x_^fXL{o!`U$(FKcvqubOsTPS9Xb2Ppl{K>-x*auQSTfBvD0qSzfl?*qFEBs$0y$DA=-YMip{HTkV-J1>Q-ak=7C?skq$`6NoOIsbv8Ca8A>20%KZ_*s zfWkWEV~XXVnfySCVv@ISFe%m|BTd#P3xY73B#iS4EFi1f-$LX7bX|EDU_q1KH&6gb zW$cIk-RF&+zTYj)abl|AbBN6xfDGRad5dK`xEF4Ql`@uWB5@bCPrr70Ca%X&p)$pM zJ!+J;zBm%fgUe_eR)C+>PB=Q#P5%Q`I{p_RBL@CrG}xaUCaYGPu9BlB+sup{G)gxP z{fjUf-^n*~sz+)cxbG-)-#w77zboq+<)wil>K{QSnP@b(QX)8j6>wD2Ln$b_aK|O? zCO-=i`Ct%p7g3=qy2+oj5-7(zJlEkrk2$>Ic6&zuS__H&w6?0%jyQL*?elO(d?M@y zeIGV`0E7HVSRKv;`gnG%Q&gbEL(>Jp(=5qhrdt_UUAwr2*~LP8)asJqgY{$~zR+v)K5K%3v57}siMuD^Dgq{?V%1W=>}m)?0E)BBQ%%H4 z%4ZxG$>CR|Z9i6e)@e2@6$FSj#?qNfEfAp(0mm=K0LG*)Jy}WUxgqlFRVeWnC3L18QbBv`B?w>AVrQ)pfFYN$dx!o$gFk#$xZph^)aZY z@Fe%x)x;q&I$~_Npua0oPtlUxc!f=W!J5H@UT_FSI5POv^{zME-~vI3i)b-P&XjMS z3K`Ie3rDR`|4jmXEcv+_-gx)zJ{mJo9KObOOfX}#uREMN^_lgMi~;0vDq#n%02X`f zb;o#RXeg3r6UpHrG8IT>16A$uNAE4qu-S-lI*8o!9G!|KyLy|uJl_AT@cOgHTf@pT zSJ8^1LhRn2_7j|2TY1%W<`maJ1D7(xNY~LY<)AAkpmx2qK#iu9s1zOGxP?(-nqPJpap(ugG?~$lMfd-m(80<7(3IbL}a2i7ebu>)Y-8<#iiP^e>me$(^6Mc5}F7fBvO|L&ZMV$ zTasWK9|NxcbYjyL$TMVUjb1$*xk?s5U`YB|U3q;jTt0aHugiM9Ad&!#D&!l{16UxC zEt*Qi-MZWjKKz3T8umbi^rvjNEOda?&o3fv#We&mjVs;)r!V`R;;266Hea%2lzIZj zT0TtG$n?|V*xF7>yLX(EIiInf!%!(q$4E&03Pz0e9fy)=8pl~>76+?CMZ2O9s_jm0 zi+R_p>MW)MVl+DFTu|tXH0ixOsxS1N*^NGjYCV5<&%@>}^dpe*Hm)K4NymS+v&0qn zo>4fTokGR57seG@9A7(ggYJ|COhM+udEW+KMftw|9zTW1rdy8F=0q7QiJB93e@rW} zrtY|d5OnIkoiLQc5KysK-baQjwUo1&h&$k#i#uO;WKO&Q&LtN_*=)~1MNGN`Ps#y1mj^M}7)<3~@;y3hy-{uJKzv(TSifumt{Gg$h)$A*n=W`=b*RF`@%^ddgH7% zMR?V5w7*t~tcluNQr3qn%Zy5y1fvJOsyI~m6>q6K0#+FPUQk`#8XPF1E5TB~Mu6uJ zGfZ>1;x5FGY`cIPUJK)pI@Wz*exyYIjFZx_(hC#X0fLT%4OZ$!?BX5o?PQoG)(cDs z+&6t79?VA7x;^fns<{Y{jfbEu^Pm#$9FdSg^+NjNw#3rGdQ!=GWpG|+=g@-D3q!Bg zjV6(<#4u{YHcf5oA8>?pAQZ81W9{jGsbepipr5wisqj0Uz)(*b=Xm`$cbXB9J_DX( z5xm2luNW1il)uiUdj-+eaJ1zWY{V#sC^!kGN4QMg~WYdf29Y9D5wGmySE3+Ub7Fu}v0U`qcg) zU0i-$eQ`oLf!TL_PuJW>QkCx+xx)EYuj}1@=E%-j&4d{IO4!uc+w-3e)Vulf3PQwB z8xNMxw4T0XN{tLBZV`ceY z9tuvn{_YNMS>5Q+HB}npAqn>U-5zh4AW1dpByst-xZ3P5t}d8Ptc-7-dTw`;^%lVK z0OUl32q1u|sMdPnP5?h<#%;S9F8$+~MtVH@q>b5zTBgu~Kh#9PZsb-8lW>)QyD? z{+|($4H%8KVYQr#^_^}K(oteO2uTeYpI-Mm)(n}h<`OHa8grIRm1;V7);Y0^N!{mB z(?`M5-tL#zU2bCevZgk&TOx&AQiP=rtBTJ*y#J5xe)mFYdF4tde9!TdkxaIcA65y2 zrM-_uII)BkY9}q93>ecTf6?8jn@hWSSVDUI*}@fEGUcEbXE*2T2B7E!Y%l*i9bDZ8 zG34UfQ*ee<5sT*wl;6sFS^rbn1Oadc1|#G3ezoh2QMmo+3$9|J{ey$oMLv_ z{|Gkbw@TdiQ1Htjk50rOi(D#=aA{JaxTo+#ov~>WlYwMxGq-QcSIHI zf7yVn1=@}@tS7f~M~)I`9#eBV&oA4WiRW8>Gu0uk-Y>?rmX^qTopLdCO!uS4*T8sU zs8mHlZAdLBuN*EYJ1g9?QTVW#V63TB}i231Bc#Ez^Tf*><5`7;f2^iz%URj#q8S z3wAT;91pKmSlxtO=@hg~t5D)hXU}h%m6n!10yUOmZeHb#V4b_g_;sa~O*dk5ILLnJG&)JHKCSWc{UvO?-WLp@pD!A{#`yVPvTo(*K zckY%T6)Q`0UoA#z;8GvG)V#dBYq#XcaUh^sI?jHCh;~uii)ABa0eEvFIK3sgVB~6o z+B2DdXt0$@_}I+5uxBe2rp0;OGJmXo$stJ5hP1xp`>=p&k@09cq`eN9S@bk~S5^ED z00r`2n%7}{g@bs)d^fwJngARm7(!$9>Wa`9*Z;@ZIRuFoY-_q~+qP}nw(V24ZQHhO z}x^q_;Z#&0S!_21SY`9H#smf^{CVi^8>1;Ll~) zICIe>a?zlFsD!~up181qHC{n7QHh!%xMDI`Aw_=Ksx}r;-*`VRf)FhE1+0#s{`*BZ z9vIp;@LBj65z&;BRHhnf0Yh}w<#BdjY4ioX0Tw%|-{Ik2g%VY(1Rqx^F?7AR&6+#9B!xlilKrSHMuv`7Skq`o8qsy;WbUzdqmvncP}N&& z8bV#mh3ceHRhtX3?Q)ezTcu}-dS;f9%(aV2H-W+ntc5oMh^$HQCAyCB_MDV?2>jc! z?OgebD2J;TWn1c>bgNC><6Z}^OcC2XK2H35=l-k?+v)1P9?L=I1{Gwwd=W6buXPNg zWx!K}>TWcw;^~WrO_+O;tD&GC+kUOLBTqC&;4piw7MYE9YYWD<>!#43)gu>d-~{~{ z@Jd1V0Z?EN`E5DQaf|^mD$m9Vb(0s4imlzbjlkZ`G#y>j_qHEV)zu@xvbwbyr4I}e zJsqV3uzP6iBbDWbCW<^}?h9rR{s2lJ(tJ+_$fyjCe(Ys>o>@ z;zrHV+4GxQz5f11Rg%EaQ^Qy#>$3g3f&uw6BanfLGj^@a{PC-ZL0Rq&9Dt) zk^aQC!%~O+QV^Ls_B4WxiJMl8F?z?2AEGXo24e;^ys2**dqT}Lu~0jO0_-} zlB;$;?EbWnJf{Rv*1R^z(~YNZjB4UX*?Pb%)ZG}+Vg|9&@`fF$NrABNB6e0_u9Jie zhJ(jSi-rP5=R&f?&7NjRQv;L0aWQ&T|6GA$Q4J}zXTAz(Pk<3{cIl?{*_HzxC!0Oy zA2RT14|bN;2zj73J5$(b4fuW8|wIrd-9M98af7lVvpY#LCUN zfI$Q%9>;4*x=+@9*%BIWI$=N+?v4Hd!pn8$W+M`&lPf$Uxqsh1ue*5(*fI&-a)-F| z)OI4aB{G(MJ;aqph(>}U7|*Px*BI}`JyunYCxbDB$qfD&T#DEMj(D*}g?whFCqG%tDJ+!7fI;rit))Pf0qDtLZuxckqmo#G< z(_S2W8=4c5!yW(QAEY}9)mZ#k)_XQ2xpK`%iTf z_4A+p#?7o_NG|vlJShQjotIB>tiX$o)JF6#+@%Q4i2A@>n{A z#?z|D7J0blMwyZ3KnP2^QRh!3tfEKVnwPQWl(;!2q`}H4$!oV!9R=FKS@m;-l4Cxg zZF z;?CL%Ww(b6ZT0C^rKOB-xBGV5Sy-;Hi}rp<>H5<}0z6`3Ev8rH@P>8@yL@aHmspZ3&j*v;E{;L<>`(8_Y!SfIWSZo#uDO zy&5(v2PG6&NOfLB4`0E|a>A-dN5Ra(TK&4POf1$DvtJGPxF!y%z&us0vlbA2v^Sll z_7{SaIP>q{go;ui4#xP@3fhh26>XsBs0+oWf{&sd=^c#eGp!X{6hq6BZiQdYBxbEt z_Hakucnxu3m~T(Ls!otkLc6G*^tU-1pyi6bPZM$C0+ zZfyr#!iN~kn!T}MU~I-n-kiMjnVIrv zir%BMnW*lu|7?^;L&LYHV8jjiQX;lEXHqK2f%mi`S{mT)c_L>*;tJnxJXq_aQ)sGXHMvMN%g31uYf)vmigba7_Mja-G$r>jd zFu{Q^;3Uga{k7z4MKK>uQ!k@`&x2Puuoa;5SX*ji>8vwF_-Lpw6(L@b<=5WPN@nMg z>dTprRLgYFy21gOuk7e=D1Ar6#{w$NrMPO}HPyV{)bMd!UapHhV;;$^-IzzK+ehgl zC$0KyrAXsFq-wux)ynQvJi!Gq>^td}!iEF8>~?Mng#InkRINI4Y8{v>t7-psMMh{& z;9^^Gu{H1Be>y)HRBqmVJYBzbezZTg;c+HQmu_`QPyxblIQ151adp6-MIbp03OJSj z`WZMT?5vDLDqY+aRuWOZaDHodM(i4?m=Pb~XU}ZWM(&i$EaVlol0fWno$6*dk)|e6 z&z)KbB62q-C*4}~@EOzZ6oQn0K}gMUB$sE`Ah=Sz#|Zlk6ZsTxmgIcpyaWeta}#|>mFzCQ^^ZQ!j2^S zy&w0^ITxNy<&Z=DGuxq?=O)y@BZn>xbi@db=Jx$=P7cCo>se^*?db9G^N)!xL+pWh zbjH4S695;n?g@;T=ACqdskt*6dO-@sQ>VsU+qGGIp|Woic-rb)XL{Ri)&9CzxS{8^ z_~$EXtT%QJ>bl>}uwD$u9mYr#VS*sP=MqLX{MWZUy9*NY(f3+vZbMB^(5MDGLEW{T z66(b~u;@+V)Fp$f@c}PG#(3&IXH}lW36NiM5YkDWedO3#7iY1tRo%gWi&rSd6*^gE z@s25fTflP1-Ci(N;e0Q)%37wFOLP>>j;kJM^sVrkOPO3I+G zbF+7dGd6C|o=H*>F)|GH)#p?Qh&T{=%y{IiQTeZ)P|GJ$+gM65r~XT*d98@fK6I2IX`6(*4$L1I#s>@I ziPLGBPWs(2Jj>VG$apTiOXL=DxYUYT2;nrTpv^As(G-OU@iNXLfuI99S?G%iHQ`G~06!7iuw=Uu{xLB|@X9p~qZHzipE`8(X0 zM@x~*GzUzsMQ7E9@G_?*!!FVBfY%FQod|%FLwt@Q+^*5hlHU;lCd5=nPsjUI-@jw< z);4uV8;VJ8OpIss66=Uve==SVKlcu{*hkW6V?H9{t1J1RKqCpZKbO=C-D#h9AoYLU z7SRH%s^-(r6RYp*^YgOn|7>fK3DyoRmG|M3g$PUZbnur!O`sH8)mwfl z`d$-KFzC;$=B}kE5FogC?0(Pz7*DTI{J3BJ=HR`W{sZS?XZR1Ci;4Ze;9OVQ(snrG z3BQ5Z10bKx%MoYB9r!H?$(6a4$)W9*08C>A5^0l|&MuB+*@=anb^U|9y`s3#IzAf_ z6E#)WrdNz41_Sp3;3_VBR7}jf5%h$LJo0V?*5h!UiQsbH~HzRP_}FEm%E3>HMM2TaCKA=mYTY@l}&5J zAhWfS64pRf8!lVbzUgx2LeSLf1E$*a3o3)m1AUl!Bx#u|k5;`eo4o1c@_a$h`Zohs zQRNE8P}pRXyhE}S-RFLqm$*Bf;y|t-!3|tHMfp(hQCS=#d%sp-pV?k;-rwmrW(M+; zR3GgrJ#+#+22uN5Ts8etCKR2)!drD%ATG69`>laNh&W%m=Tb?9*;sbk)1U5IfyucL z=+TwgGUh7UOEB*KSOOdSOVOroo2n?QmRbicwGqxh^-Bv@)%+s^md$3^F*tg<`lc%7 zcQqsi^^zq#o3^GFqRd)*&A5#FdhUO6W`XDf!R?7|N5_>pSUND}6(ZL4~lC-ag_0)7lk%Dk2!vO*eMURq1Yi=G+ zJ1wM7xN@B5a5!);I|Y;&^#JL5;9>gMuDbTq3j+Uu|G8e`+pfE20=3i?p;|&eNDjnW zz}Y}ty3J=Nn=X5_RBk9LCjj|}X33H7uzjidhN@EPgc@a$@dXJ@sCK+I)^@U~k#0H` zHq{_~ZBfL7r>fA+WVLSV(5xC`69Lf|e~zdz^p%&2?duJRD~@bCU}{9!0(%NX%;VIu zJpxBs3KY07k_D!Ph=q4qA}6xRLlmUD@iW{FJ#w~>G8eep;U!)YXXx6IwJLZH%cz)< z+1Y2c6Zd<~^Lh;&KiZ7@z*W_$Zl?Aub?D8Z)m)c|v8T-E%MEa2yay&!YpPaFUpmEX zie4!-+}}5f9);x7Qp7El!%)l9*Teszyk_GC5}~a_%ko^w{uIH2-?jpi^U-z58NjN0 z@R3)B*@V+D0`eWMkGmVHpEARAIkMR!3C?cs()+voDgHG2ZX)sVz{o^Qak(Wx71L?*ct+|2>^~8qJ6c$Ln2ei zl<{x^WUijSgEN++?6BF!R{f=L{_qu_QOwE80N}mu79F%eqSW)Y8n5Rx@x_I(hS6-ojffweFLFU z|Na~{3?vv5ZLW9?VfT9Y5lSKog0g9`Vk5o2d0>%-HNVsGj^u+}B z?S}xwx8f6kLvLy{7y;yds}X~Mg6sB>V}C_cgLqACAh^^Tp|(NQPQRnbh@d=A{H zJCl>8G~hU~I$&^0Go@YB<0Y*Zq*5OHkXOsViG<{&=$)*RRwlbLR@F@WLEz99jQ#|Q z+=s8<_H>*ef-SN!s-uiEF4Tn!S{zL%OZ;i3yT5+{K>H+=xzf~`Wt|N5Df{K()qN}v zTqAVWp`qP&4~HC+opO`VdOWyDmWrchuDZ21DoE@EFYt~0jetWey`{17uuVW9;(P;& zAMQtM&$%+llUAw7q8m;^4?IfawB)M#N$sD{Js4P+-j_cyR_G0RcJ~xzcO&2Lv^6 zVT@g9y@sT@@TZvGLmpmZ&?z5vJZ567but&ItM(U+eQ-^nNpMKkurY@R#R8k-X%AsaIgd8$IP&DWq z(`LJq(p8-y?T0uEN}6c3lzR#0rDNHj(f*JZ{NKZPIOMizHYLk_j43)eh&{yB^fZTo zt7TS91?@Z=_RtHD7v8=@z*+RMYfj?YsxU%>$plhh+^p>H(klrl=5|~URp9c;D zN*xdpNP}~8QQHk;k+eVD1&BbL7pW?t>zFv)Ef^qDb~3=WeK-JsaDUPQG=`p#i!&Tw zKG*$#K)Ef1Hi740ZxJp$L|_L$nfnrO^IcZA0|cOi9=}T>fUz}peyp@Xj@;<3aB?yG z&>gv?KBoi_IYM zmxv+m#Q2|#Z~-VNP()8<36M0$^U{M~B7&HT6O1E{5efDC-8kWhR)xn37pXYnu&6g3 za7gq9q@@UVwob5pMwgE99p2oLDspUw?EsrV&mcQ+=1;YW=^X%Zl6E4r0MU8`xx|^M zmRYnlBpHpWuID~WK8-Nwn}{w=CM}A%K&iX;0LO);2a3ua8vqh#Pl<^4ZpD;?gA~)g z7qfl@kKpa^5JHn{e*+o=D~@h6i(#Ut2QVIixT1L=;Zy>AeUf?)hAi|##^N;=*mS6T9r&1JiWy0)23;rgW&{N z1#YeDpa^dC0E1$j)OJ8_AEdLFlCQ8r z!&Xgy5Oni4c$b4jcGTkBS&FYRQ{0=Msm@4`YMZ5>ju6)mcg-`FND#n82Tbyc)v@T! z5iCvK(cev{#%kDz*bs{xH?_DYuxM)pUG4R~<;!Eb1|B!hY~Naokipm9^7t)4lnb-A zrx6fKaL^!RwKPN-SiHq|A&>GoLkV=fv_dEpd=gakP-qQzI^Q`70mwEYA<28|=SGKp zUEH|Qk)^T%CU=yH@ph(-nIg-Aj61SE<_@m3Yc!=%B9@o6{lE5njehtm{!M_1!+R49JJmf zU0?HS6{+8}SnNW022lA8yz$VNiO~g7$cX=<10<*%$X||F zhIWRiR9qS+s6*aW*K8tOQ~-3Fvnc`d$aBykPn@=7ksm@{Ptvhq$q*zB4J%EL-0JZ) zAO|jLHH-XY&H2jVv?}IFfcs;WG>8^5xQE(Bk}T{ z_fun;I61QB^HUJA5j&eUOL8n1O_!R*;=2>#L6kWrx$z`>e;vu@LkD5-qc=*DEaT}xdpB{h zv10p(6u_Sp^3||$W)8#+Y|e`l4DiW~7iPB7)rQnTJ_<&-V!b(mggroRA$_yvAO3tbHL_4)~ zJXc6>Z;^tHyX|Q@&BbCObq`8{LqbOgd1dl5q-%fVlS62@K^B&2muWChbE~FVu}|lfccj@ zTml0#I}!J1eP{W)o*#t&{5%`1^S(R>-e>>hR%&?TehXZc-ucGzlPcnoK~c46BMXL zHEp9PV82wVf`4eLCfl2)e|3`%6BGbvWvy-zK)ms~Nv>up)wdaUrXHmO#u0>3qsBvk zpIf9bjV@OaU*M`m$$wfFrvJ4p$}UFE9`>dL^eVqkMqvVa8B<$xXA1&GM&|z_!-yO2 zV}tC1F?zF7lRcMQKv@ga6NMEF5|Dr)5aEy^C<5+a;VlSL!rj&I6!QH3_tNFc^I~$b zrDv}m za_B*OXH(V1h|f!H%>m);6+f8uj4juvauu|{iV z6i|ZYvbIi6nO~Te*cM$YeSW@sxXXFF+-}OaXqt;=O`4T-#Y)p!F)~6i@?+O!d{_>O zT5(TJF+LW%Ts?SlbHt*Y%BBr{T|eL^7DpeIwwjY%+t2XM^QR{i#G}zNJAZa&KDrjo z8oDqx1YOoWEy7&-EjX8rO`_>Gt}n6cH8&Yyw4H{#@BD2aJYhCNa0f|jpz%flSD4R9Y%$C=~zrfLcxTv3Bbz+g>!P&&tx0Ln`i)`iej#8s3U=?_wGx z5w@3l*f4OhOiA52+?&B%?uH>^M8Fi5E)V?)SB@O|jnxXqu~quexPNIx3D3(jmM!oEav_2$FWd@C(p!`Ob2r1R1$nu*hU0qv0JB=4AT^f7UD6NS7bM ztNq|$%M3H++x4qZ$y=l-IyX}B>#I7ee$MEoAgO6N!$RXBiHuO!N6((bVuU)SjuZ*x zyI2q}6xBd&mHxf8JKS>x6b7AZ$8a7=_8ICsj?F8;S7vemg z2X|dwF?%(nN}<5*MjGugsqi7+7Q&GAg(G)~jRORto$SGWsnsbqziDpC22kOefQ@GwOo=4G$iznN;W`sF zt@iUD>jwfLJHUPd(PalFlv>fMdYWQk8+3m}(-aITFkt-kZ^oP6A|j$WZ0D%H0vfBm z!zQJnYLoO8U+Y7Weru0Vn#<}_3g#FK#R+B-jCRs3H? zzfdxvnJ+0Z=9!Q7E=7I$(rh@~K&l!m$dp+>H{_{2G+ocgpyTsQ2E+^DT&^bu@g1oV z(3e6aDGonjt(1yjExfolf>yrYv!i#k^wK4U$$=9zc(4e4gvo8}034dV_Fh<2Mnsh^ z=%{my$cQ&5t(=u~oP4jgTtZu>y`(e2D=A~ZhQ7gFNJ!BCQyF9aUuEonsul|aJIDWj zhVx2W)(Lw&q340G>YMjOh&oNij^@8C!E!_}DkxL}zA-AO!Hqug+CQD1k17hK2SRsh&&nM8uhi za!wf%BbdWbkTYAPRZA`%`c3Fi0?PuWgMCZpjN|WQm?od4zDfnm?3sP{OBru73Yg#M zdEdVwo2cge6*M}yW3Sfc28hl#(^faf6FkXF2(SYgx41&OUmMkJ=df@wuP60(rG+m} zEX@XuG`SusMwTi@7W(XB6UM_2soB{`(=V~slC=O!6{C*Lf>P^#SMLc_i#k6zJ$+pD zQ^W757Q7IXK|ASi#&l|u%C`CLdE=;x>~2wV3$1OYEmmwHj1FTgBqvWDN}yykvK)>Q z_GCL09!S5iWQT477nK)_yX*OU1v(o%&^M$I|{@uD^C-Dgn$YmwS?0c%6#Lm>rf z+?6r-cg(;2GHlsV&o9MDC{6$gicf#0p&aWkLNw|m=o1m6Ao+;dW7=6~Wt7xiW*^8b zS~|6;wTL;MUgSBYd~n#!kin;w-Dg zJq;ubHE=yadIM__xM_xZbwkkdRy5HSoVDskfaHFE+AEq$Jw=*kmo?PU` zx+Np#%7>d;w%UjFl<`kE|6(%Alq1LmSrPyf;QJTapO=JxasZfRQ1Ffyun|+6bI%$D zp&&>!VQ8o9n2!=&3LqB-4CTer88l$^P5Opu2CKCj&0m~?2^-+$eBhBLlU)rO8v+s= zkmBAKhKy;`_B1tRlCm44y7naLA#gZO^Q)hrB3}GBZiG@rpy=mEqoI^%6|EdvCPh-9 zo+Qo!*wn-|>F%7G`m$<8C#zA?Y*?av6J(2pM882m?h$x2;gD@!p37rT#hMPOjZG77 zldgf5C*iublQnZuw0%RrXmM1~n(lO3rz4B^60;^#UOIp>E^@ah+Xt=JT3S!tQ#EKD zydMF7r!`VyJ$W^3&j%Y-@vhW`VIeU3i*xx6#UaB*i)dn}boDo2G|oYw3!``-eAgH# zVfB!REnI}UUO(T&VHrCR4T-B=TW(^7R7D@wj`9_ciXs#;?*kUvDo2rleUMH=2gRCp zF%uCVZQ0o>%=mySD#>7#D(G(+>FVK0E7m+3jc^0JKA0d`veoAN5UcbvL=G$|cuy)g zoGC@mG>AKIPM=8Cg!Er<=_J`|cug3#Gep3luEaSR9 zQ5y(wWYB0(<}cNGX!I5urgkG`1`QsyFkD(jldmo0y4a|zU86H*sbmIB0Gt&gRFp@cO? z2m2n)!BzQLvZi(VVbi=xj|9y<(Ux32n#zAz5s3%Z3#$Aix}c$uZR0%4V(U|LQXQEE zl0ht1q`%fD@TRCEtIm$>y0E?-HUQ|aBIhu6;7N4pZce5KlXwKDdFJsO+T7;IR`L7V zGNM?IM!H=0Mil__>7??+)L9U{0NZ6-`OQnWkEnA8@)%Sh{Mkt}gKNFQr3C_-vEug<1xs1fWjF+Bb`-H?%iKCYUR7}I}qJdjbD?J{Znkn9$;A@zH+*vDfsw5YbF z(;dZEGZ;FcF9Tw)v4cG3=@WzMlV%26E-kZ-UzQuq$-akk|0euE@IH&IrJfk0xQ8*Q*D&4Y&}xhC+QA*nc(+VNqGiE zb2KQTgpq14rfT3uyYnVmM!;5j+Js8O)gF{sK-Y{tK8c-}4*tueP6|V=L`t~dW@5Y1 zK%=^r5&Eyzba;o}rQpnm*aiDr_)p&Oxh^7xh?T6L%e;u+8E%O+Otgw5W53$mn)zRV zx;axZ*#!izvb9TFYxUTf1@qEjlh^7(b&_)6eNSy@t7K$sqO^sBQE9rU^kGmyv@O*6 z0JflWQD<$AzarJPrpxntg!_QnsnJb$b4_as7*L9i!RXmEE24CW5oc-4zA4=;&BLo% zMsD_J68$b?{f)d50(UucH=**=^#R@$1K{71D`N^4E6=yWw=Qw0`BIr$mf$OiZ&|&s z0n(fJ_}}a_J$im-aHib}Y&;;LgeY7#Pg)9<h7_Rpbjdiw@Fn>i-)BV* z7-&mKw9cc#hVNJo@1UQLPnHfVV(yb~zMA?(_|lp{~9ZJfjidpGhBiIY|fh;dox zW^3uUnkT?lH^OM>CPxKOv^X#t9g_~RjyT6`qan1h%8E>nc@Iw5j;pi74m_^ufccqg zwaJ$Paar%@c&*0q>%*@9DlZ=a?i}eiDWM#6hFnQd$6dnJla@tUxdJY$?(K?4eP^^9 zi!ofa%)>~qQO#K*^jRB{4zPB2*PnugFQ&X+*FV0=f=~*4h(?2V+}&T1-(!kz*hHBx$6C*DC0hweRK7j#j+YHLoMb3sy>e4@uP3 z0MrH{$4LMkBdx`_Ay=8(z z-aFcG_c!ElmWydx1eXmfP%&{AlW+%9e11sAIa`IC*BB5uE-ZN9X<^}pJlKn?9vy|` zOdp@xb?rvy{GklQVJb@HVvcC6zIFL9_z7G|sePzU%{Pgt*E}qddu53T$ z{6fic^6IKv^=K!m+BfUx@8%+&lv$}8zG~BSrK=B}4j7YuB^d)sa#5X?=YkUG36HaA z!3MM12#)h4v?H^bPDsYSYA};(Iv@R&cO2-&)pI%RcLgRvM!@@KhOmE{vg=PRrs80f zL~f~$elo*d*=xNuPuZrUPdyKfk-ozosONC5=zXQms)L?7wdzhNKm5PP_7}u8OpBTa z7R)}}&730s1gIFFy?M4&-&sQBXHkYEH^^|Z2V$lLcX3!f;IL04Qz>{7w}@f3v3SY+ zR(XtJmgCL{^u;H$Mvf{+@I)-$(PoP9I&8tRC{KV+LE9}rT5o<}p;6YD1MQXA*-70! z-~;st{93lfw@2ei#?2+U23>SF!Bol;DGy{Qruu?x2*LzSETCf9rItV*;ic|;3`~;f z89lTPxYoPrr#DAe7fnBrY5rq13bBs^8zy~P_Ow4+_~(28HMTTDt zW~=0$ypq<|WU+nt*_(2n1kpN7gkFhouq>xhH)S4N z{FAN7{5P$x`>dMwOAk$V=)=zXr=qfm@@~c7GzVq=5RN(Zol%TC<#JJ$zjvUT1#qiJ zceehUlJ zK~8L_2avX_7WGM!O6VulTpV9};1%kX0rp>r0e}O8(dC)<-U+qkyRT4%MA9m4GL@J4 zQMMLxeqI$;Mw;F&&%T>gs$%W0nwiZ?);El6sH|g=yld@(V;NXZlS0_VGcwN)q^p*v zq|^NiZh>|32GF&+8z;FmCyTg+BjO?w^G#*dtFgq9#$!oqb)TRvzGNTKh|2mXM-%S| zT8%JxQfi^B%_A3k?ZYpcrPY+TKzRtB+;_7ECB?U2@yl_EDDq33BCbgXdh+72=w z?4sJkT;UvC{g2yl8#a%Nw9gQlD9d5R0daI200$F~mW7bs3R2gD(YcBKPfxv;r%pyi ztADPl_+LP*-&-b%OHUtZIm|^awU!CbA74cj{KMpU6n}H)^VI*9d~QesB*{zniM`Dy z9^1qyI`e)OEX>YL53S+T1V~D{ermoaY%?=v95(kZRM`ZSTy2gRzE>f2imgH7IJ)OG zX2c*mRaZ4`PAIA;I)oRc#vr#}ryoToEptNn)-NgF?rjL=-{~$_KMvZddrmAqP3w8_ ze}@LCyX@>xOWH&?x#?(Y%)FG1tMo@aErSRLp9AU3YN+C$*1l=BeDh@8af=ys2R|}= zC3J&@j~o_O{wPR5mu%2|H^2B>|AGTn24?2}1{_E`{gZI~zm^Dg9G_~} za=|6&@`adU4y2aCV%0@s`Diw>eliiGU|ah+3pokD(I11Kd9lmoFj##bd%j8l=ce%t zjk8?{Zg&EIswHaXzE39PuSpWwLz0ocTfYV#3G!9&J{$>8F32CW$??09fjzGpa=yEa z2{Lqi_*Ym?2R=o;3d!6z5pC`cjo%pPi}~zqY$-bbko4OrE-ms&BhGeLI~T&}Kfz7K zIGP4FdDfR#MHS@~@@)@Nx-!3B_ZDWI+{1;+IZd5id$tBG&i}c>ZOIW!yE!vD_lM%! zhig$Q>gDLlSqzzTzPhT`l3!UkQR_|+M^~@)Xv;Zuu;y9Gw2-RvKGMtmW*C>2sYx&S zs?$mhzjQH!g-Ln;w6*4(A=z1|R6Yzlg-ZB02Fmfe2pD}x);D<8$Kk(4=AFMnmXSvT zZ)4@N@BCBDm(EW?42cvlVq!)^T!##vm37D!0v*TnuJ~E5{Bd$NHJEOh8L>L<4ECoU zj#h4WqpW^!DI#kA$>FWUW)sMjQ2A=)?KEZeC&U~=xl1SNe)972zRr)n&tBe}DsovB zLn0O|#c2D>?WKe+i9o+X{6bxij7?n;JC;qiiKf%)cki2HaU#$D+An%4)m9(BWlTL1 zHF*R64r|ZeOmicwRljju*iNh~ZVps+%#6I1URP?i7aZ-WCQ>=*RhQ;h;i2w7TwWbH zr|uCR$$je3b)MT+qW*i=Nq&?I{ooEMqtLt&Rkq`r3j$gW4@=%J@3q~i?9!~wo0qcm z>B!RQ=4*dN*VsRPZly~*iUgB$SW68>o>E51Gs@{W>)d>58XB@*8$V{!WSJbQ5Q(rNu59hBjKs7; z$+V$kdpa@F z*}PCQgiQl~P2@5?%B}dBPMqRnt$EU@g(ApM5_;`VD{R#pgc(dL09T)>G3ZA63FwpA zaZ3QF4FHhU)Y<|IhGvjhsAt0D_Sj(P1cT&ve!`uFr-{kOrHHu!WJioS7&&c@?8rLR zIS{+u6#}WHpsvD95Xp2Z!Jkm9;}OIs7w?`9I&CZC<}z93qBatf>vVC#FhhIfBUI1) zr^WUVZyMJW(u}FViarirPLD}BXRjizhMCHFF%L6)31v_(Kp7?}#s`KE}bw>#|kIvGoe_(Pe% z8A9j&j#4bG;_MC+lW+Yn-FigWIHn7vTgPZnb=fBvG=h?buznDeTqGiY`6nXz(})Ls zK!J_2hGY9*&vR&x`X!eUJN2;3E!b(Z-BB&3ruRU(X&ANaq0 z_|c?PdwMLPXedn_;zZEgA=-#<@T?UA0fW;wAZEGt_zc3>c!lUPL4%K8mhnG3vw@8$ z?j`s7cESK``j*0A032Zg*4C1Jvd%g{pVncg?LSqY`osa^?}#$~H$`tr>m36i+hWMK zwsDXOZgu$kWIUN)w6=vRh3^V%ka9kk?*sd^Qy8=RkF^zuxg=ot0AbVju8X7}p^+L5Jsiq(fMU!xzL(k-3&2KJ)y&cIh znU~OQ7PbSjbv{@)dbmmZIARsJ#HMxL;6TH(Ia-b5bNJP)R)f6Lc<8~EeobaQO~_)# z9Y`*Sl){6MId>Q?EuvG5lq7L)uT9lSs8~qI5TTG`AN#J3n`=vzptyV9@o0|AmBJT; zNJl&tMtpIM4P*-;7S_cM5=cb93FSzh*nkl44BRwT&AP5et&v?qnC6xr<40;bVwJDo z7H0FDe=LFGAO9SYI8?RBCD`FxTyoH)V6^P!XNiuz-*O%}eA%!ilh=G7*oK&K0Y2sV zB$dp)J4mS$6z$-QzS=5#lN&F8mKG9XPRS%r&S1Betrge!xlM$A0sEYdhVl1*%La^UrV<$jvy>h_?xF8!dH2}qTDgdjOE32>qnKybim?N5_lFyet`g9NZs zYJ(RHq5$L$ne8Z7lTPfb;#h+FcA=}nU0e~Dph~EDBwTv8Ny*ij5A|qwMVb$bpGJ$i ziB7urRp)%JS3DUPzRYvZx#SBo{*8jkOJoXhi}PS}Og&}1F|2^c39Lv1xy(*OI8?99 zKU>GS^+iRSo2`^+(V}7zS%#;zhQbMWBz120FenU&&j9{C)>)JLNC3vN5L0oi7#%Oo zO|7Ts9;51T{mo+c5Rn5@&U#BnSTY_0EI6nV19}+Av4JaqU^%J7$%tUG#}?n}A0{Be??n7a+_umhTiGf;6Du^I=&xk*-z=#zZTy9%}VEB3NsNx+{bb z^%yQ)SIFM0L_`5d6ayc@t4vZN(^>Davt%4easN~S6Imv2G1wkMyDLHfBW;JQ?=bWp zB(_cWCA&c$ecAnj+)|kgck*(sVvZcs{6pdZW~8Kvw;06+GyzBT_E8AbO`GF%y`5{g z!@ztuxafgleZP)PdcieP04bh*XS)G4-qfJ&%jIzdsi)*SS;cvd($OU^o&SIr~@9Njp&)jOz)ch6d)As4P{X1o$FsLDteI+kkl} z-7wW7sT$yA9h+;DjkzvEDw3ff02uUb)tc`LG>yC;8${)2YWf*kwO4gFmsE9MZh|6u z@t)jAM|c$?_tRw8Q zNJ`Pzv=|D&LBXbNJ8k)=I#?9fo3>APFGcYQbCLbq3^k^FN8`GAVAWL~*tH%k|h1M2i4N!xNzl01Lz6eqO;&^4EHh z--2cf<^%TtH-TRHo(x{D1Emh@^^4D76RA}={*JsBV6Iw#2)kX5;L*Z( zv#1W|q`0$Ot5wKOt?e44jrbxsj=>9STpl5hgZ@i553n?H1(k_7IJ*iLuheym9YD%< zXXpnoy_{j@Y`33Ox+WW4Zs%~Prj1qFYzthg{!w7D4S*_&oXmq(^b@{w<%|^BHd#qfb3P&|nuC8u! znagB#0BkNUVbL~~3ELb~6OA?pEVJ3jEZj^1vYq4e8c|IBb5Q%9Q^I&(Af|f?n z4vy9t2b>6mA_JM&vSN7SZqU6MtJ-|X(uU~y&Z_Y{5-ckTQH;@II?)XlW5<=$Tw`3gBz?M*av|qlp^(70ipMTOmv8P%xM3V_H7hr{2HYY ztTHTOI??JpzoRW?A?s0qx^3=rv#pgAbYA0gfqn-Fgl%H2_zSdA1aDAof1|4O`;A5zWQV|r#vRVfDZ`vDAg_PGwlt&TZmEtj-Lf}SnPoW@3zqAXIL<2Q z(*(pIil9!_6xi5v894Cg3@K73hQ@VLYB_k)ei4pYc@O9yA!jH?E@Ma18B13CIykwu znQZGKZCg`RBZFeT9771-bU%Y=I-iGhm&ucT77zR-I2r=tAK^B8CQY%I_e&b#Pcjze zuM~@)QVvLvJCmvKt`R-HpfbX>)tFu$z;O*++vP#T-u|Wao3Pg-H-k})M1;Eo8mxe% zXUM;pyn8np$Ap0sk;OuAXYw5?T>s={5Sa<9p=Z7|$wgFbc3+>^{TSHyh=LwA^L#*dLAXMp}HFC&@PMZN$?X2!DO)#n@;)GatX5sMHs|hD_-` zP1WI%rn2Wn>jU5g@1R{{yyxnRDj-gN#wfz#(f9o8O91L-zRXMmX0WClIAA`B?N7s| zl=Pl@mt>`3>Vg#_{>G$n7LDTxL`gb~YRpcYL(x>NE0amVzW4mOJpE3fK{GD3=(vTM@@{@5A&!nS|57-gs zlwwmfA##tB9%`n!E$8qaGNith50oCo+{}ciw9u}#-S|24!6=EOz(;)q_A;%($o&Sp zF7T??X~6Pb*%6-s?h8+U#5)&$|MBqja?puRZKW^vQ8g&K8P?A3Wnon7(GEzQSWQTD z_}>;YfLb1AH3-fe& z(kVYJ0Wz&YJcftprj-9*38-EF=!3uPraD>odAig5hZkHEB=djf>-NA@_Ap?6Ov25P z-G=@}oE&+~tG2?gw+o^-X6qL*Uf>B@w=Xaa8qc6K!WEr;5F+$ujA{FhI8MdcM}P`M zG-!u>9^fQ#k(8bSw=pNa@Tnek^&Z1Rqmj@`?}L3pP46nbSTj7NNyvZXTTuismf5pc zeIX}`T8Aoquc{1R^8-%y>)JS@3>@Ws<{A#DfU6Gag7;8gI5)p5^WbLOZGIzS)#80n zW)De&As=+}wU@W8fc%Nr@y3(}?f1FstZf0IgK>a7`N_oL*=a;dbR=kPMuM>iWsi8p zW?lO_?d;;i{e(9l;BEQem>QY$SHh90-DQRc0aQv{JOFTPPNG?CtdWw3JxFR_v_z%*)Bc+sUi^iNa)_lC9(0X^7{LVn|w~Saf?>T9d0I z2R}S=GBOpN{NMsVFY)GSo)Nm={nXXllSH=Nzoi?EwsGYoL(>akS^bUrHzzdx`f)+$ z8uE~ys#-j!`re#Wm)Cnj;|tCnHrWfFz)T) zDsUn>pC?-v-`f#{IC$Qa{cBaMZX*hQX?<#V{&Vb@)kMJ6)ZfGFQ{d(vE#^(fV#bn5wB{N24%7OIS{(?1>RD?2On1**3 z^RIn;@s4Dcv2`ZFcHV%i1BNBeFjjpD8+{y# z=~Gn*{9o?dPVU%!>vr0d7iVBwAP7;>yKp?OJQ1yA5vHq=j$aQ3cRwZ^NxcNZMT9Bz zNQ6wc#=Iaok`5@o7;?Jlyr|CZ@1Fe@jl2u}V?$a>rHk6vqKjP|ZR}Oeu5^=Y9jiT{ zHmW3xn|=14Ew_o(XbDk8AX@DR;L)xK&&*vFX;T5*-io635AC)icuuq z3JMzq9^z913c~Z72N&-i-*bYkHtn6a*(a+;FmD?1t>l3G+qx4ex6U?Vm0*EDG;(yc z(lx;Z_Fq*EVd$yw)@NHsZR*F2ASeCI+O&wJs5wOw+9qNeqBI}fMj|AU=rqwB)_aF4 z2yYzhYAyzii%U}a+WJps?QD%==C@?hcvECUFVY|~Z6p^03HN?alKgXejD|-H7|51i zi1cj`&D#BzDy^|lVczbL`M;?Ok&i6wyNsX&>{<~fIy56#g$-Q55>j!ucxwd!6Zw?e zH7qU*IO6&y&nm4|+C|$LvB_jxR z*zf)LZ$BkgCh>b$k*y!}LgebQf%BjqoQ~j9=5b7shz#IAi2f)dP_&zM0Rb91K)@9q z`|1~<)P!4&ch&Tse<%>rZNl(w_hFYby-m zd-{P=WTIPQP7FcxHY&bA#lKGbED3>1dU}xYab{n;RFb9>; zaEs66Ho^&F_#=7LNt?t<62^B!CLUDYZZSV8(+N-iU7d??YGIgftor$`Z2ikYFpL93 zK}q(}2^c82=WsX{;6zA5&F!O-H{5N@r)$r8&`mkT+hxOWV)Xi|$pQ*NTImXul|X@2okhaeTi0{SV_!Lxe@dOxQQ)PdX8#AR>T^p6#ayr@goPl)`-dZ)4HniVG`GvFp)m!@YNyz#9#YkES3tu5Gk_PSy^YX>4Ln?4(X!6eN;a%q+ewLhZU%;rsEzMi>6VCmR0?k$ zM%fQ?DYmn*9cvX{a~qcei$|Mw;Xt|cvr1vMd0di(9Ex_R&G4^3{d7J3_wkV`;jvr7 zAE>r`{-*lQqdiaZ@RPaHs0zx@q1hE=SDnLCrJ$yOUoB$%Wv=9+tm5()3`vED#<2Ok zDXO&likbG(ES)6o)DklmF2jWL>9jM;rPem4<|lG9gk1OcavT->;VlDDqu1%TZ!!1* zLOfLbqg72oL>{!fAL7SluoUD6(_mfV{0^h~(EtW7e-pfkvj`$I6l^AD-Wk~@iFaHs zphWV59R@vB)SB-KjWfew&<_uVHY+fBa*2P6I{`W!6m2 zqxndS-KPd2UqX5VJi1$-PYR=Gr3)VwQ$N^mG;`A-XwXo@Lj6iPV*+|N9##)8#Pki~ z;abn%6W!Q~xI$%#O?4--C32Iy(Yi|>!3xj6=yzZ)CZN)UGLW&e_2ST|{`I0G1*%|a z%k^onxmpA46XzKDDau57?I6xzdUztw8qNGj?5(CpruQcs_v)|fE41~a1oD>5qB>DMToG5wgn1^3aaZr-7vY(-THN@&FX;It^ zX*vp7A_=q%ZB*9@vjplrFko8IE}Bf9)%A*kACfUxctgGp_a1p6t5WsA-^kceOC~!3a zKKPmM)LS+hkL@TV$al6DPzN^|B1RPdZcsV^N+lzPLlu=QLEjB(s{rul93Rl;1~MZy zbxkC@KU@A3OkS}ojy+K7P~q!z;1*w=Yx=~XUf<1&@hhV>y*}^l`mRp))jrAMUOI1> z=}J)b23fN!h8sYge+rjmnEFhDp^$t*Vhyr<1eCARwRje)O!PGcm;rVcGh00Ibd9e# zvypEOBtLmso5(aq_{DNocSy7Eb%LY_@Bx>M%fx+YXx2b2M?M7aAK2Ne~hep3QuXuuijLJCnkB)4<^xyLKtFHXkl7;EgpmCrLsuS=Jpr zwOF!bfH2!F-9&`8ji&(VI)QjzT+eOPwi_g`57tz&H!1M4Jv}`G{F|i%ay8F@tY)#n zh8RT^Zq}E94k+j!1W<=ji`Hp&V}J@0HBT14)B?OIAV`m*V+c9rS|%HQjVH3}#a;@k z<8maU<30iG;1qio35)y6WeRC%VswW^r=ZAg!r`gnrc9Xl^GJ?!QQ0OxQ6?~vy5zvm z2~6Ru*2t%B&EKmj-Zj2kNT-;boik6(-2$XL^0pJ=kLOwlD%F!VaFb}W7k2xVHiE$t zgy7HXN!{FzO4O+z3zQ1X3CCc_Ps zJ~&FR(dzXbrMyF28pUq%Vm9;SMjjDZ7{Kunm?WzK`lil;a5xv6H!-?0v~nE|i~l%S zp~kCS><|@cjG)n}em6xS)k~hdHd8`vsBVMjaJ}{CwC-+9``&08i1V=WZg0!XE9KA! zpD8)TsmFLX`Gt`oExY!$iJBH{)rTs#Z&F!K`icZj8VPPt=&gV^N|jVIB7=|iTPP!^ zA6y=|59HrI3{?6bw8Kk4vZ^xWn42O{5D=dQpsBPduSs=mSNKp~zEo4omn)Z`a7WNZ z8gmq|OY(})nL)ToP(j3&VxVu`HB`--x7xlX)`B3ol`0O5J0+B9+)GRiBh(czKg|CM zS8(~Hgw0mgL=LR65R%Q-9he?ZYP?ORS#*ng*WjsMIK*rLn}%;cshHNz)`JZ3*aSRP zohoLFP0)}Bw5f}|=cYXaiU$4qL2aaFP^6~n7r#-WQlT=ML`jHKvjpQXIO5qTJEZR) zztb4-S+dzMhnJfS+)mWHrFHN7wC-zekR{bY8Fwdvz{ zf4o&~6QPx6I@v>8UCVgD{~BR&2H$PW+*6Jv#Wzs&vSk#!v+W@CQInNXB7k-vkJQ2< zhh^=d?0Q3)C@V@S`dP&H2y*lKN>`H6%cOAeG$m_x7(4=PwHFDFL*5h1K1@gUYnM1+ z(@syYo{3oNmrKze4Za2}=q9!hqcd+E(? z26o`$hDm+QB~k>#P1L;S8E&gQqG_uUL-|bk8lby zuB3|QH148@s<Q zy6x|6)7e&ZwZv@RS@^h_2tqi^WUAF$r$^$Lqzq?#XUla_#pEP1)O+vsF&YAT<)^sDRPN;_ zV!S(d5AV;0M#=qk48#;}!Tq(I_H+Zn>q!Hjxn?p5DN}5%NqJr6rQ1(+?I}m-_6;SD zpL#TgR1`=tcv;;{IbPhgi0C~eLTasVcCxv!vNxG-CU+qVLq#5!RtotQX!3wad zOs*gDHyI0*Tk0yed#B_e&el~$WUX#As(uan4egGBXvDwR*9IFU+{};$t#t>j(RAZw zXq)-f%kGMRO;f)vC@CQPV89Fe5pG9Tl8kDGz28wbn5L4x4p#`XgP3cLR+$IC?|4T) z0D;I713Dz0Q~co3Eay)y=Y36)9eq3#C2EgcaL&oAL&9w5h%$o-2zgCeYK_r+Eydlh z3dn8zEN4$;gp&CuEHVC+*Jxr|iy_ckksizIwCK`fx^=THV5>_IYI2NC%NqV(!_ZtY zs(9eV4OdGIV+%J+4)GG)q!oJ{i@ocpgUFpy!y<_bUGh?6l^MTzWC11y)cf@mJATs_ zmHEDyweoavR0PR_W}24k+R1`>eg@FdCE1d!`a(-wgv_K&)FYcSuwoy^$MmKTY);QE zN6~fL+rJttbm*4KUQ;0o)UxOaV^I7}V-ROj6IJ~}nq{d~@!e6544&#LOa0CE5Nt|9 zi6yBNcRcQA-8ze*Q>Q)Q8inC%P2u=mUGSy(#a<=NQmqenB&|(+!xr2N{MIhl430T2 zSrMWTeOZBU;IxdDIA9-@oB08a&;8MLKvs@#uSAX&cllNK*70&cnNFhC?nK?_9PEKD zBx%5PBfjEawC0Dy}-?}-(KK#qW5WLB`#6eyI?*_J$OO_spO$JnWqa+ z$}s>dA*(d=9B^KFtncN|ZrIjT6&1i1$XUH9>j3!Z&=9l2jtZS_&s4>mIhvRk+5z8p z@8pvJcPG#5ncw1S`p>Kb50?fGjCt|SE<$I9|UV~ThK*7tA!FLxh`Vx!bI6CZKl;wqcS8k-(P0Tj#U zcwAjwwzscO`VhY2BO-l~l_R#AkY>uxQVFv4hA#8Sj(7H4e1~D={unkHh4&h_3AazH zSD_)n`3xOexNj!_p5n&Yx{q2oz2^80`Rc-{ytH9-Yhe+4U8Mi|F*QSF<^L&uBezLP$w%g<<%O3ZNa_k zDz9(0S{Y+M_Z&I?W{EFQs1orXeQ6n((g#+JuehhOJq!jNC^K`FFsjeUw461-&6rXA zM`{I_dfAo%EL@!&;mmxPFHf}(P%nfhrmD%dWUeg;4}jwrjy6YNe8Z|;PutVp^JO$j zt`A5ECEF*5mSU#3={i26=T(fD5}_@;$7Wipk;$gb>Xj!mz*n_=8DX&>!17_s9r*66 z&!>X5=yJjavHz@*7LI8scyO~lKszIwk84Gm6BuU5#o{?kZ2mF6$e-0cv*PG~)n>f9 zam*S8-{_glc2QNJm3f-{Yky`D z;s9vcLoz>@_!=j?F zMFe&VvQ|m9Q59jVx=${Z*jtuwi=*_%B{9{SLSn771M)Nl5nOSTC+y(9#U{e0IRp*N zoB(u4t+#WFHFf}-m6mZKYB!rCfI4k=THkgL#lHC!+v*XduDfCqdlMr7(sb0j+$jVA zJLKcOVk(kOW&2SjT8M836XL^JM|r0V1`RyuYAWbmA{0?+aQoRTI2w{K)zOM8$!VQnUINk5&2g7f!+mq*-Z#lD4bnWhr+{iu9y z&00w58V||W+lTnMoA-_~#7)ijL)NE0E?S*1os&juYPGCoCmc6r%J@VL{_hI3Tngj zs!cB1RBtVwFA*nZjb9+1*)G7tV|pXlBW#^M6p+ z!s7t+Alc_25+#QWqw~QS-N%%js-9nV54l3IKHar1a=|;Xwb}A1;F{#6<3}9;EDMr8 zCi?`lux!!2ZlZ}HHEcj6&< zk#E~vH(O>~boOiV;+6Gfwm#$GcI>2kn1iKg)4^E4^~ZCYg;#obdGZFwtR z*Bk1z#S@7_dDI!FfRf-JHW(Lt6fTNJ8o&)q=slwY zAENci(+C%=pTV~9;)JA#ASvM7j`7w?I}lzVg~+wjc~G*3bLX_p!!4;X_UtAOB$hie zLYwf!?8b@$6Csa&Za>|DE+xm(i*A3o>g;@@BOPtS{RP_&5oy8oJ>4m!u%P9S7J`QV z=*4l4zL4*&dGZHwC_A9&CeOWvIy}3?UwFi1+*)!8aZ^`k-Fn)a2pdHt;14re(dXJ- z413fYAK!Fz<2M#@t~~YIaSZ~uwc17t%%>=8iB^x1<4xgmcL-#bfN}ggz2=8Ja*@ud zuz+9&5@OxQ;y%Qko@M2vI91ovK1Qdu{2b%jkK6?j;eQ?p5gb;rNL18s?EFgsW97B> zQji{V{al>5DDv1RPho?nQwD{7cU-xJ{k=6Od#k607scRv=p6RfsN_goQ~>ew@(fPr1qcVjQfs zo0i3*mA08!+;|T*$wJCDk9;U7kZ6$upM(6ZyJ;MM$Z_H}w4sXFJ&|FpbwXis@byf; z(0L;b^9MV|yJx{Hy{Wu`=T)r~ZSO9ZCl?o8@m{&qlj3E`kNUt=XicqO2S`$T;|Gpt zvXyyqxb|QXzk}7v@;U)2tg#gl>OOXka)prZ=`lb9R?9Lw zK|TFJZ*{NBQk_{`|1m@0L!CtDReywK_-NYNL(}^LZzLt_GGr-wt#UoMJFeQkj$hR9(c8&kxPT|eBa zLqCF~p%k~YLqG1s;YpKbQK}T6U~!diu=ZY5YPiS-GgQA!KbK&mcAs&zUN)#dYm6rY z!-WB*UJQLe)~$>1MH$$(zEu=7)E^BUHtKYETvlvsA5OPq!zhrJsf8^L3e-*>b8HU` zCB1ZwrTnownmHyp1<&`)RN}o5*hX^+yN@S~7goThPHKoxidVdJIA7CqqWAM9X*8y5 zoUYBl@Gnd`gK3SNG!iqW%dCpPnlhAUmG`SE`++(ow(S~Ean*bN0nL1pI%LJu?_&F; zh--&2q4PJ=Qq2e?@65E^)U=bL2bBvgf*GT^0p!^N_h1y?*UqyX!l%ESG0v`>@n_6f zndMjV$eS(SY_Ij3@45-%sxs6(wM#O(=}`IF=*f7=?7{M1A2Q-P5phnmhI~*)b=eJ- zPnOp})AoE>elxu_Qu&?xq0J=xK?d#RwUjzNqZWyZ!^ z*a82O(4jdYB<}iaI-PAu$h`f4C>U`;1}w0@*>dx#-=Y3jM2rtf#I0RSk zZTI_y--@s*r_Ju*nD`UhH;WJaGsb3NBfKof7fIAKqu@#^c(X@~DHgurqVA+Ic!uYblDtM2XT;5tC2+!PJ!SM!v| z--wici}}gLsp1sH+TjOUNJ!)KNg%It+PKTo)=oxlJ> z959eAFeo3od+6sq%A^w79QD!YlnCyi1 zNdyJ>x6X&EzH=I_(j&V$U(RO0Gp@+D>%c;>=FmkEMS{65UbBteJ=DZxlNfGE7{tB) zW57hyWKTs#vOJ2Xo4xh12Us1b>Y9lYy^bQ{br2afH}5JJF?hRwbo1~iUkz5Paa?5GE+XX;OPsT5$f*){D4Fym8Xt?+sV8N6zDwRn~c z-^m6T3aK;hBw(h5p)1?0sk@T8UeDRTm1563I3%Vw+AT{1MZ-sF zY6nG5(5=e>DP$og%4it1)C6y{&Z(^+gJy-u$On9K@Ad%94PTW0jGxj-qSG{JCSp>p%xXV%jF5ePJ(ilwQ59 zX%g0*U9bBWdTexppj{sm-cEWFfb z1eM0y&-T7!X_&3pPbRxJUH7mYJzaS!{=1rcS8C3gt&8rqb63xl+PR{T>LRtRKcy58 zwKGifTTi#_&pESOJz{Vxn@)UOv|gcO)QgZ{G2~HsiMK zrG#{hUeY*tW?CqO&ZcIy-1IQ{=hiB|BW!D9aUsd>bTYP^*s)3bo#PkQMo@^qWP}yHb*&|Ko?c# z(?xKhrvM{3M(kt`RJ}60f*D(OmS^8ivrYqC6BW0$+^(&R#V}V`LAJgRno)JQc$0p) zc-FTEoEWwl(HR~<^$*|mzVN>~INzCzFtHzSc=;i%TZ(^bL#=vI8y998wXqQ)Xcc{F zx7HYCM#0p1>>DYoVlv;5x=geWxs<`vL(&Knlxm$!?(o8Aq%}^p)S%_Ip!LCcBkzm- zKY$jBTb2JCxnldDD$989`BjxcoJwy!}`^~!0~Pu6{!7JnbZrTwb1+SJAh^&NR| zNN9A!cQg=^LiQUT{F#ckhKmxB`m~Z?8#-NQjAB%zxwCg7DX!GkMqt2`gso;I`**B_ z3T;SUGuQ|VM^$)eGuR3hE{wCvKHXaQ%eVSj~D@r{Z|P<0b5eBOJVti9Vd1k0hOtZL1( zCmMdu^s_Z6c+wbyi9&XbrY-Lsy}3AJXnD{-Vy98e`kJH?Z+(ThaLIuzRHVnoumcTk zDZ2u=^fK4t<6@H{N=ncHZXAPX@Yn+(m5j}1bQ7RHbQ4QCo*;PpRN!f0g~WYJIa?Bw zSkfDw9jK=jZ6@ZLuav=$3Ow<@i{$!%X#~nvPSgSnL@5&YA0-MLAn&p}BFriUk9JT!)#6VgRPln||m&#FEg_I$%sGmTw!x{V0#X#piG5JU;03|b0o^)j6 z_|0aVv+UZ;QylwVi_)flr-+<~SaDrgBQ5NsZ~ir8WF~O!?d#yw+#O%G z0IL*ptKGN>WZ&__M}*el?gfD$jFkw~k;a)#XN&@T5tmNnqM0~D=|-L<9%;QO$>QsS zjd>o>n_a`cnG=|bnrOZ}l=e%V4|`8NhlUpAN;SuK1#JY#qsj#aF8>G!Y5Zs za&kPQdTT*;2zN5gBRYMIV^yG6x{`Kz&@g!k)Fo${1jkp@Xx>&2x{>!A!1@3~=^mEh z#DV%qdX@@4!}x!-R)^2QH~@UH?u!4~Mrb_Iy!o68-Euo~xn2wwUF@FDR9woGZE-=*@ zXTTH)s!Ie+$O3X#p2ZBn3dA|3*fQy0F#b4*y%}nYwU6lL2|@J`w@EGxqBb7G9TfQ% zV>aZb*i3xMsRt2JybdgQ1`N2hNgme_C(N3$Vv>lQY0R-S+xT%-FW74xTZU&IHk$`WB<&AE^ ziQy79iJ9RiOaNkMJJUpO6n6J;cv zz)2*}H@kT@VZh6BoT9&9ZBgR`nW9%>%MZW=lkknzMn^;po=I{)>4Jw@P9#sxzSg0EGn3C;5a? zQF_dE4^PT1)VDZS2?N$pvUV)Y4t8B?fwgMSRwy7e^Xw6#+llry(9lJ7moAksLjRq%A+Wp|}hBs`2y%$y_((A8~2FDSBxTSRsO?Goegr z_r}WDRZ>w21Z9J~FxEm?Isv5R!o2lej5@-=UOA*afrPn)?1hcTs5MCvz!*I#q(|Vf zLkdgmDPj^pt_$YJioep~d(E&zHl%{Cl5?al=*mHl_Mzfsl|w;k6Ll0Au~EEyv;Mh0s}jb|{-|BiQS+gLcXue(26w~3B&9Abu*6usgp4b`L; z!L%7e)G^6(Pk=SVV_FhE8B~|}w@O@GdT0nC_3*K(5OtR|lv zPkdWpH_W+`Dc%npDbG5iqIrZdhsHcns*)Bd*eqNMsSpVqD(PdLmPzYlmZ^{SQHq7w zo6~0tmrhitn(x9Yp>C@}|Iw13w7m*actSfU=N3*Z#P{Q_EGwOK0p#cX+ME4-gV~tE zb2o5?J-JyBL>c6q*&twqIW%8Jq-uyZ#EdjAU!oqTuZ*?xBWL3+hBW@6upO0@FwS&% z;|^ysdBKs$&8fR}L?ck68I|EDi(w%^wBtsAIKst+AMW)de%ho|cDDcrSyf$Q#*RA* zZlF)37H>MZLZAexqvC$wlhl8GMngJSS>LnWvvHn{ZHhj2P1n^-5+Vy_LttKE40(uE zaDm98&Rn1z_Lug`uIkfi+bSVv^5HCMJQHepPK`ro3a6zh!BRR4y(9tX3lc2yT=g{U z8!x*g>UCVGtVe_I=+JGPHsY$((=4{u4I538)xBjsg~%LBJv1LrE^f5Ur7Lyk^KH8c z4jaHD^iZmU;%NzbCzA=Z^O$_7&n0k@5Q~e>l5(a~V&43&0PFkZV}MfYurs@mQ!=+( zUdlQwzvR>}zC^&F*&mVs<2>SQ!pj?0M;5C4v2Jszqo1BB zY--RgZcbtCLROL}$opbI6+#Lt9wre3;CgEDeu`JkAwF|5DegXLtn$LhXOx%xYVlUp z^s|7IifJnsDlO5$A3G7Rbwe@PtS0`;c7<>X3g5}wBUD(@VJC>$BGBftQX?4hH8&Ky zbdV!*!nvsVeCY*6CeOPmmRKLMgd(1|(heY@ZNPn-e+-wo-Jc#4>6Ixfq0PCt>$_x*7seJ*`@1RNT9;muk7)415lz$nzScco8o9PCk z@~bNrlYi3YZ>DjsH#27}>b;9}okG_WB<67t-K!63BF@YM-va`W-FA&R-_siTsM+fb^5)Bl;JR_x-}D2gGx=Y2nP|RoG%PZY^DWy%T;XU8_u=p zwn}uJH)tV$_=|nPKtckf(b(l;@nX~qIi=2biv@HCTtROUyY@XR{x*q7A`$dM8(0E~ zZg6!|C@IB0$g4X+Z<$9p+7s1W8Sf373ew7C3ifme$P+Ew!mM-{Of@6oAEz1l z$`W$EumHL1cig5`Lz~^2@bkUQo*3jN}I$ z76uwaF}EV+GYaOn^8Ep5By_R%x;NYL8kBfoTxW~3#Vda)i+F3;iTNrcg#{#A8yZQ8 z341)6QiJANXP^Do%P(H5$*Bxn_~^(`(#=>Gd&oZG3G-#zQ$j^RBvr-lT=il$XH8?_ zdJ6plcLUED781zC@$po@g#>DyW?Tlad|=^QEj07IgH-or=0bdvhiXXq=&-nK3)$_3l^yC#6Cur_ z`Mby7)|fjP`-i0Zs9kPLK9IfG$$v1I-MDH<)Emy)$_P10ty?pl0zi&1(ePL$i_(wh z&c6bF4JJ;SZ(IX>G-Q)9Cwrg^q$5&NG6W=m0Cq9RUFQi{Re&rO#-=3S@7C#=F zG84s@kQxmR!M6P;pi<hPh<%3E#4pGCkg|9GiQoY_&?fT+^KT!L762x+>0{2l*7DL&=W|PNqlk z>%wg-_YjdhdAlA}Se?GkIsAxEEQZW?kkMFn7I$N_^9&8Lkct>}X5PWuaH>m4uMJ8i z)H-f&Z5D73wotqcN&l6{19GT@${Ii)28rVLB_+>poD+dRjCiN~(Z^o+_v z#)65c8%q_Zt2$`wi^R|L*VnSofOXWJz7ise3ljbEuG*(qZpA!8eXxC-_#^X@{35NpaJ^F_Zw zw(LC$nEMPy#2Dxy5H~vfNKmtF{stlbY0<`tN@--VeTRr0xprcRLJroWN3>6Ps4fah zWGyf8z6pbQ?q^#8P)P!c-SVm>-;P+zRp-9K%sy_rl9 zG^nzd; zb(+kYZYbMUt{uRDfk~EwoF$&`02I`@6ohiZ`Gyy?*pS%t6>a$LU*{tH*369o6!pf9 z@-yzr6~UdOWC4`BWFSBdG5Y7Qex4F4pFuZ=e+eTE8NJ&ZkHD? zuieH4s#CsB#b>zk|6%N%q9Y3zaNF3nZQHh;4m!4Nt7F@?(?Q3!Z6_7m&duKU@$50q zd0KDvwpLZm^?!5jow^Bau3g`H{3Kc%iHX5OEbO13_SuD;oil(n1gfmM2f(9$$yXVL zD2ZlEHDu@tRLL@titWDrdll+2Cw~z{2x9A?6`n@!q6xu>ECc~PI&N|P1nlZR9d79U zFXG4bKZqYQ^Z!Gl|64gH2c{^7%ZNeKUPqhCqKgs`WQ@vd{1(A-gr+c^sv9~_q zf_?0xQ?;O`2UrY;-liGRKSGPsw665B~V&qqn{$1W3oH_qL^K<9Q`496W zh;dClX3_qMv$j~cE_#0#y-ZkCH)C9n11#iV!IkT(O7-}aMN@y@CULd(=%)}$9N_t# zQKg`&&n2r~AsZtRN=ca)QE8l4>V2@`#FQi`L?i(!GfOH=kXZa1J`0>oPqOs3@IP*W z%wQx5h2QHCcdc1}l{aTZN#28IZKQ*5b!Fgh`nkoCT|hw{$oa?pP(Po|pLHsq*k%#o zP~3WZPwX(p(WEn_0!zGnf7lC^g1KvOAq}*V)DqsqaxGRD6~;o73DYO}wFGcUib8dy z{LiE^if--CV%PODM7QeuawJ3D2q)?4v61@Z45OX?JW{GiXi>vtqXnruKXy#the2?3 z>*i#pmCjDw@O0n0%!iIb=Gj2_lcKD$lSWfOlDz6nRB2{YQ4LoH|6*08XZ8UtC=0IR zFUGG|!^R-L$Dr*Pgym?jc*8v=`0yd!57 zv-0#TWe2LBA%q(g4B{M5<=U)$@wX>7bpH-SX+@@fNm*}?-_vI0`m4$52LhVgXjX_a zoh986Z&tUTRJ)?#v!?vmf%1byMIO}}n638Gai?r>i3Z3j54ls2OY1=Ei|^7L9<<&^M%0`t{n2{uXf*^#Q%ar!|2Kh@yyo16#4|N563L;?|;0(Ko$fL zq~fg0S4d<1YP&M&@F>f4E*#XaBAOb3rFx2ieCzF(^tyxevYbopn ze@&G+u?O>uesf&>hM{AZEn@Q`&NLc@GvGu7b-4514Ot+?A$L&8#lH;7^zcVn4=@%& zJ zQVDT6m9~y{514&#FS=u?>lpv6;tHK+fB|LZ)C2VI*&guzRCc(~p_Xr7N?xGgjUUnH z--Y@7V{!?>O+|-ubrmqURW@i8PZ83u5`yMWfm%YXGT~r?v-W+Km` zWR)n$d8w+fD9{F-rRPQ`fwuuxHuwy%p~u<9<5&{b!V2YBw6xhp+6y5@;CF>}^VN{5 zhiqDEh4aU3qa0==3{#L!!%oonCM2I#Mg1Ng-VOo>wb>weiU+)AbYgCrnrIV|SJ4vq zv3*Nb1uXGS4D+LKz7!WV4*zA^!f1z6AO|^RcUML{z4b%LZYNHY= zyQE+ZKZP+Ad;CK&m=~+2+MdjOMr>I`nPz$tU~?N}>I>q;DeP-h60IWW^fu|>OQtd~ zQB8W8tAg-3oqG8ys#e;$V0v{C_kn;MgUB>#CiXKqeBh|#287QBYl|^7c~aDepYGZy z8q9MkrN8816FA8;wgi9$%)Dg<;WKjWBEltfapuqwGYDm`V0MaF{Hl@WWgNJ(x0~o3 z+!&uq5_#1%;l?e>Hnop_>Z{CEJ9=Pkh5iW11XWp+6%D?Danim87-$ibFRTBRB%yNA zSQ1i9FZe7$WA!n$(?~2jk?PR;j54CJb3^WY?JhwZz3tuolCC~yXDafw(vF~`PeJDy zVn<>0g0J=c;9KbA(zU~;W9UH&i0(@#CFw4$38!}%@pr<`uc#+qTv1E3)f+9!MC3Y+ zB_F-+)sgMkn@>Z+FQISZ6_$*JstMwoU^Sp1`Ay&+A;&{(QI>qw=ysmRzFo;GM;v;sIF0K|Cne%q?`Q0T+t;i6FeIGirxj1OfE2xr?X(yavT$kMngUi;af6{ z4)1-RA9=}Lk9fttKtKUFXqXJpQ4?7OV@1T;@{m6xG~UhqgyO;d`}a%To$AVp8up?rinCnO3?}-BuT*4_g~ZFU=(1{oA1~bn z?}O$slEEZ_)RrLvgbY(*woB82qw;5aU(J)0dq7K9oyN;RJSR)F4(0SgpLgCsBxSoP z_P~5~h~Xko(uLuw?Bo$t$k3S6SRRtvAYvxkRQda?73^!Wc%sxVHrEalq$%SJFpYhe zi75zxSl8y~MjTR=Qer?B-{54jvq|O-$Y*gZ5HVo}r38Lcu#P=Q;mt0Sl{QhV9*{OO2%JQX4Okfr zY*~Ok;;!ds*qLxt)tr?`v;}44Ulbe0&Jq>0KT=>0qp5=+kJ9E@H5jzjhuS>}{ueEJ{Wq`*sKT zL2`2v&ossPLZfo8#K&daL}zvQFg9VY`vVEpqeR!_LX)T_+Kf^VMf~!-2#QHKhZyFj zDK+HzlZ-;4Psf4QmxKCG@tzzrn2B3YBa6HHxEM>h@kiwbm54D*oxfp#Kz@O5E{MWA}~n}$iWt~TE*~u zfG2S#q@%h&)$#Mm@UEcnZsj6K{?L7KtVAcNejJK!WUtS*t| zbfgw|ptCai1B>#*Mv`!x28L1NnKvJ~tR@v@)C77ux=DVGd-%OZogTSA2-k*Ht!wV@ zNh=bFzOhuihnq&5y}~^{_(A{mH_DQ6`aF=I>Vd$4s=`xr>`Oh_Az2YF8Jil)h1^9G z-O)i(ZN87&r=orCmBv!j+aS}q)BNgKVG*?aJ4W*FnciVG*qCdfP`?J$1ZEWppxL|8 z9-@IF@eK~{UJ*KVYtUL5LVE+N)!pTIxk`UvNnxXYSaF0iDG`^&ChHa1TGc%g)3ac( zTZ{~lxLz9PLY;VzQtjyhq8K4#4h~|+)cW@8PKF@pN_ zrTN;%;J+R=poBA)6#S0G-wwY7p`RB{$m?0Wsce(!LxS7;lC*gnYq*uAEqk*UbGh62 zRD3@hk3tM5aRf~i12D~3G#3h!V^6K8>CAPf(i&Nssi`9$+q2!9t_WDXER5=c;4OKB zw01C&@+k7wNWr^Hv%mTY0-ZhVdCvuVK#)ZL9(1+(xqC*jn~L`#se;HPoq(5mj$t%F zRemb)B)8!F@gnFmMC?@V0jCj!2F*G);c^-O=ev7gc>d+cjCq<5dp0WMlzP3?qINT~ zH1kM?+q|fEWZ>R*cjrE+6k|$P33Serv4x5`(z+tLZvNoPgd|IMtG#lye(t1PD&k$2`==C0wdW0pE&i_+Un3xm zc}*;wECOCZN9-Ep%H@-N3 zH7diPRLwsnbN?(=eR4$}1p(_Wxt!P)B-H0wYHz|L8+v!Obe_K0a}d*0L{xb1GBw+F zHcnpQk2hy&k+~j*R{!$sLp~$ui8raYXV#+}lL8m^U!7k^+VSUSE{ z<2+7XsORIG(61yXZBZ?qFIRz^m#69zse!h31(@lTgX0my!F>D(J<1@CN8h#|*UgSP z5zg)5XHw`;a+208oN?sx#U1Mj>^c|*6KbK%>+?Y~U4!eojHmCi$KuJlxP_ zIb+gvg!kKqQv()?z%`FC%H9r5Cn>6>8)#hkE_0if3`!%vpRymAYtA*FG#8f)Ts27* zPx2_Z#T=l_#*Br2yC9K*=FK;bIB3ztY$@|}#P20uBl`4S*`AAtc_l%Xu>R?Wyu|WH z+`Saghi^&}W4h7U?(h)?mg+qB;&#+#lV4IhY)+8!r@2A+&E=qS%>&ChDlE^Ss^#(z zM~002Yx%Z4p~?BwBj@<8--Z5{k75LdhjUx`(_QJ}K_|7D;^pXm$ui;5Md%isx zFliiFd|>IbCz62ZCRc_LnSUuCHF*^d0W%>SS)Y|KkXp8bkpe)vbw+*rT(_mwrGSgv zBpSNj`l4i-sd?0!0MLjO3cLK|oQt>;f0dTMdV-VZyCq_ResgF2$@flRCPn^oG9h&m^G2jc zy?HOrJ1Q13v{jxXkt4Z=5Guc*xd2hsH8B zdEr?t*XkNV-|^!TA2_9?$CgB=)|}5$P4Ra%Q29)zax%hEi3;F=Jz0)Q`6D&Tk9%WX{BjEiZS?a8 zSOEw|pIz3gVLRJFt8V7ix|kb0voek#4Slkrb#{3S>YkEd)J)LsV_kmQf2J# zE)<1BB1gHl)a(+673XOhU(h8*kp&A&n{tEzSDh(eo2YRR$k@BxAK6xrnpov)<37np z#{S_+DjXOT>#Bg>43dMv`uS^L8l>w^!fm4`v?px2l~bepyMUL;YULM{@4)gei4E1( z9Or+;DHVLMpBuoZz571QXt^olL5NO}Il!Z%chi^)Q|;1lJoS&=BB=!d2G#9B>hO5x zTb0-r-m4wsZxD;_JAh(#RtG5mM@Ggu?u>=zf0Wd5tVR+YuzaKsa}oAxR_$OIJr+ap z0*A_vP5BAM0jGIg=e2)wo9zJ{_G+0ED?XQ;IUI^rrCsRJu58qEAWCM_;N)Qx0l1cI z`#Q5fS?aP{pQU@nYlbi7>;WKNmS6bP`Rl;lc%UD~3v~LW`Jfi1&$WB_g#I5d%vseG zvNTkRXRnTFsdO5TF07WZRga(ht>5^s*VOMo{Ea|pFo+$(%X(Xq2bZ@uc)JJPoJYi4 zO=R|yparDfq$Tl*UC7tqq{P&{+UHvAS?U9J;j0)toO13i(FZUOr!Y!RV#no~LOW!I zpLMo`;XWRA{q(IwhtNq*Yt|G$D*%_Qqe}i5==sk(H=at<>8?pvxGg1V*BGuE`(IqNWrzR+ zG|BG5w;MwgWI_F6X9J=~i z@|qe2u*GV7`DWdYA~PO#3$*zUv3tjt{Aks|Q_Gif>6CtZdfXv$2I0DHL&vl(@P_sS zkk;SvhOHcaKBqt)$08_uX0^?j_w>NDsvcl=CY@=Q+zZE7)Z96q=!ZMriV>70Q7xi0 za%GF*k0La_RDsV;;t>D5rKS>mvhGSyAj*7vT8KN68I4-ZR|CCOfDmuK(_9pNPkXQn z&fA}ysEn@P0JpcIOMG@$&^olSz@xWSmg>>Bqi7|4K_{QLl|K5lGD2~WE5y}=@q-@F zyY_D6eXvA@9#k9OJD|W5_GvE%ecvFPu%_|^`sM zzYLUg=%>Tn4zT~bagZ=zq4_#{c_rW z{C4*Zcj9o$5$Csx8gq0 zL&>IubUx+NVRv;ZtgL;joUNiD;S;s>0rjDQvSDt+11F^ z#@8;1;s!?XUxvu=!w^T$Q5Z()6P{$=zI3t~@#F+9F92zkeaxM~wS%u# zzXwFL@Mf_Z(svfCPFHD^^ZJV*Y9mjAF|0djWNn9RIJaVDv-6yR9as`tE7yfkJ8=!l z25BCzjo%=y_W)2EPXE$HpToh?#hU-fZya(zuPCOpJK*E&Ed(E$zBKe37XEOX%UB5w zoMFq#hGKFZC4W>~KFvP_@?>oP65c%<^tLZs?w8>6`=|#0wN__&!ox~(%8(C{3z@sB za#lcBKl#T6gND#oW>LQK!*Y<_>RrEUJfLB!bXI1WHse%LU5)e!QQoxro&g1#?$%^d z_2xwzaG(h2{b5rS3f~kBzq6oVAr_kp^3v1&BMXo5P${)C7yeHb3!efa)Flc7Y{#%q zVR>s`orxITaboe{h8U{p_wC)iSHpmB)#xu-$;SOZC7**?T-@|WUDg&3{QBxl8SmBfDogHNDgvK*w5FE;1#Aq>O z4_%sd0)5>m%O#j#E=J^}qp5fsC_Z~p|E^!yO|X~ta57ay^_n&0!w%KitL{jJ4oov) zVGJVjDqt4)<-*qe%DWLO)6n_5R( z2S}JKLl6-Z+1J!ICsKW>&6q##<0y|WRA<=)NFA&V_tKH9VY`_myFK_>KXB@h~C2QR+s13QjCdqZCj@qOC&Z|pZU!I)Y9hYQmE}8{W>t| zJev$@smb}=ESGiDW2C=mUX|BG5?t^p$2S+0u~wzXIG8aIf^k)xpaxe_m#wGO3Q+}b zG4gFDgX33kv`MBAYcG&3s^-nJ99%$2J8AzHS|>2Hf$L17cJ>fT4k8m;ImWBwhg^s1 zc{#9Os>m2I>mF!>rFJ8K|H^x!5KiGJ8T773#;kdh0QSeiAa|oJw z6s|+y#s+|R?6%nY8LxS$a>YOIN@$m&JGk!D5=DL998^B=z(rpb%Z$*W-s7YLW!HjQ zUIb5|Uq0WAPHejl^L~J?%mCIE_Aha{(q9}E2nrO90v4ciOkY=naOO=Qx_Vx{z4h>a z5u`Z~pkI7-mijM6J;ipUzhRt{X_KQJ^`UA@AGh4nUZMLwgSb4px#8q`g$(6iV1NU; z-F^}?!i`>z39BO`WXUY}fem_BnKO{56es=3pd&fJC6|%~yaAv)RG12`*tss5=wzkS zKuSn~W1!K-&p`ReD#VvP5@bdF7j*gdoKTi(EkttD>l#y}`SIOMzXS{jiXu2uLF0S% zc+=TnSeNZXO^EYDJ)qLPH%u5o1(`^=e>9$#T~OLd7i8@y!aHSZTYWT{LlI9h z%pfW0Bg~b$s5b)t`>*x_>$(ybXLVLZwjO{siSb%`$qVgOb4-dRowY2!8$FO4r4$P~>~{j)uFdU5ab3l zc0c#(QZhER=AUa_TSt7aR4hKvtuOf7R97M_{zDis8C_LYC<;$hm#$oI*-&NH*f@wmfkHaiblz8J7td< zSg%RL130w+eAh$5nU*)L5W=yp(H>8Cu@qZk$I7nr>;hO8UyNg*h_;X#Obt~Ygea|S zhP?GfId|?t@{>?ETz3uC)F=pZ7q>v76@U7$ssbiKu4eC_mDr;pRuF(DSOo!(MvuMNdIGAfM6rC8 z%5GSR8ju#UKmg!tqi#h#$=?a0Qk(9}b1*$XB-mE{e#ovNM-hYlj4beGbh}Lk0qx5GF)VFb;|Q)!OMCBdL8q4*VYE$YmO1!b@Oacfkoy#r`#O zzB+0jItirnvApaL;@nqY_}o|u;-4#G>99Tj6Iz~$&F;KNX;sz!^JNV zmKFHeTfA;xnn)IBp3Y>j;CB+|%O1HaOHNM9>)%oTu=YzFnRNA4hlx?tZB{;z$i17_ z7G;l6SC`L+e61jkVC~8kx{`0owbk%(FAh-lgXUkmg)|ZupI&Li-}KNa;DjUPEa9}a z5O=5SmM;~Xth*n^YUfJw^Sl=80|Q)62}13t0%8`5siXsUHF!iNc-4qpeo81VuPHj# zeda{nu@mCi=0vWv7f&4SNGI@0Fw*W6w*AidbffxA%%He?RDlpAVLZjjb8B%LXB1I(Jn`HRb1B^7_ESaB>QiO+!9b$p+ILz)mgnaGL_ z#n6HGElM#fYK?t$P1OOn`s+@e>E{+RHv%gU#OoUs+_31Xo}msZEln;%LdUW!tv}CS zaHO(r;#<6&d36(E=Ge>5k|c2ysnIh_B0;}rwOf_iG9{iVjjPOiS@^4Cx~I?2FG^El zYl_9+pE5jnbl9~#n#n4T<|(X9kl#YG10gqe9ITB>nc6O@RassEchdWT4s~whT7VFb z3F|#3lU~u1skrh`lgaV@eQp@FskikQd&FOsh{U3^CY4Qo#tKuoix2+enRez}E02vE zp}}prm*caTUT2zkCRfv|D|3ZyO~@AX05{}R>mM= z>}%Sw5STi3Pg-s~wgL1N0=V-mF;95}YQx|!mtR<_$uL4fClHgHJ6k|2Zs^k0B9bkn zBJWC(ZSLP|7Hk(|Mb<0Rjy0&g^Avn8VG%IrjfP}*Dnf@Wvhb$OQYVFJqYmO|9S0P7JD^kB7(c*i4KKk;7o8=l(%^2#FlHGr)EMa-00q2HeL4F zu~rOYgivkznW39&ZGvNgS`c8T_DF}sd6k7mK;xUF=j`+UG5OkS>${u}a#9ZU#~IAbpoDg3o>5=#FEVt}qBkT7!HsI~l5ar?K!OJW5?UZfc7FHai9!#-xz&gUYWzaSrXE!pEQa+nW&1@QDTCtFoWiL}uU%>B0Eo5n< z*CJwvTTWj{NSOD(=<$MJ133F9h~D>AV0NE6gP^z_Sz}9C`tEhT%W!KB{UYb5`VeqJ z_)~l`=S<}hP@(o^VA#r96=xvtf+C($wr-iFSVT3x8rIT17#@7HmI)fZA((80Psk`& z?wnk;H`!dG{mW$N&}Uvg^xj*y(yg&4Fe2Ry+a$^)gTYJnz|ZY zJ`e{n30(H*SXs6|8(2A*EA1)Oc@z~W4bAZb^=bueS*I$u6Rt&%pc`ErJ%&TKnByrm zM_w`#)^l+z0YV)Ua7wRrSfM;`J=`W((o9I^s%AHC9bFrI-fs^O*}GRbbjGhP)5E|= zM7_f5sBD_WCgB1YD~effs#`?7w%zoWCy0}S9$H|Qp{#x5LFIO5A9;+uQB4hsC&O>| zW8a|UGsg!113hsu{SWlS$;tG8M^81{|Cy}*{h6$84Z_#hp^?gAsh5|z|Gj7dq}s$( zYPJQYwOh+eLCAsEGMwBS+84_9 z;qvr!VrzeaPuet4AZER;8R(3v0=*?cpqexYnTmGwM3(eHdz5OUd&wUJoaABJU(+A6 zP{zmv4}sI0v#3%rB94gv9ACroyQ}n7DIe&&CLH>X9u`*)M_ih+n8fvL&{@d>oW&BF zv?1pW&H(_G?f5ayn;610v}aDLxi@4&zA!tx3}S0Ft$=x7$xCqoB@0>|V>edaz{1k?GfNTH*Iu&Tm1 zV8oAZox*BI9g#;R2!}@cv;1R+A*_$u&_~uRY@9*4qdk~+7>FueVG1KLFOve40aPP@ zg2iu)61Vvc;I_@jP|22Q#MXAPYqL4_BPpzxFR>X-JV!UK){Q!>bj`+k|6+9uL1jD^ zxP}o*0M7(53|evX`uRtagQy`Ij)y)$a|B!P4rn4}$Woe_QzjidLTq(c36x=PN2IAU z-ZWWB{MrA*QRSrPA6O&Z)ZjwleC0q_0E}8FCV9W+q;6k}c|NB(5v1|u+Gb_n^ip6~ zduBE)7{1`dUUj0O3;#HOEwhtVUI^ynDW zH_l{)ZHu50XPm<3y)X2}CyIpMzM15}j8L2U=nO?{_-SrUos&q%VniR#%w_;CmJd4^ z!Pc7K*%=HS6A!x94U*$M@Hbw-SsW{JK?|m*AKY2`V1%aZIkltq+1T|NJ9l_b*0IGe zmo7N)Q>Pp`JaaQ&FJ*zoww^S8!sQ%VIj8~B#N)DoJd`@`3CwKOJH3D|2=^$ciUa_Pw#+KrQxY>f`i!{jH|I-Up=- z>qd~{Es@B@8eU|_O0AHlKW^5B`71ea^6%sFfobD*b;|;LTMo^j)*mW>^$@Nc)H`uh zvryYBeXuAKP688y6Uddw2d8c#SMLV~U+s97{GF@11JY(NLI`srCA=x!Jn`|Ro`yX5 zrnF*$K2Nl?v55$lUp4U;NIo$GG$s%C#FSK|=|>-_bp$$hvj%XkiliKylu{@Bb=BZ| zmV8%vPcm+C{bzU(7~!M_(vbP~+2)g%?xqcAefTmm*3Qn)%H=ng$@l~B-Y(=D_M0$> zuK9b*6AP;XI;h9o`d#AkKL;af@om)xR>e+prU(aUrUrlfbfPtEgIsoor0H!?RLi{2 zRC&X)UrX4e8D#ILg2C9mk)5k+u7YM#>FUy-=FA7iVDKu1#MTi9jbn2HxRwS|vmHbr ziz9_Bfl^dFNRFUL1?higSi(PgS3=kNqf9}7=}Up&6@d|E%a_Tb)e|$u9eIK69zaDE z5IhW25-}BgV(EOB=KOPjQ zQ?_WR$#QbWEUP6Z7o_4FGo=Ua%ifEyvald#Vip3#?<$T*V+&Ow!$0)hK^*y_L~E8n zdQo__4&~4~tLo5CIyPB27k_+)`&61v^_IcTsZm4OLiIPIf}{gyQdD-CG3X?$@5MF> z@j=6Vwd5Pcbncy`dXE)~Eu}v@HxR5nFYvlu03wo+fwK(j^-^`RiUMoq;Ig7=hM z!h%3jPir^|E3K2y?GFCgkf&)UFYHm6@IdI$y;UiJ7?~l0gm=s67s?%&E3GW8JB!nx=@V|EJXHfS{ME z5QGticCIRNtf`g?n|weO-B-1i0~W2JhLNR2fg0e_rofr*EjGrrQj3((JnUq>j2X4m zsqMZR+3C3Gh2(@aejs+xTP=Lp<-Lna+;fdOeXRfCD)$P#kHZQYx|r=P;W@(RZZd#1((tQai0H)7?8Ba8S#3`JsE=xY>YiRMIjBExF}4pR;EaRrxglP>|~gw#hV zR$c$<{YAJi83kZxM+K1y+IYNnZ`c947m!~DKgnwzKz^zFyP8Z=(y6P*d2TvDx`)kw z2G@5|xn*UZvC#KuqPjHoMa39U{deF>ZQhwS;WmiXyz4|8dPDeBAZ`o5B^OatTEWEL zf`mVygqa<=cccd%-=3#5$??~_A_ZL)K?FfHXx2WIqS@y0pGc!asqYXh1#7un1k6~t zATgI)AjudlV`N=Ve+; zEr<;qpj4$Ag}c3gWVmC@ujnK1DhAZqt=ISz9+kL<3Akn&b*8IE(*j}$F%Oj~v>Z=7 zaLvMh4U~H%t_$F-#;pZNeICg78;U_XcyQ}aU+tf3oyDJd58k%a3j8BSv&t~dv0@TG zHPqL&V|jcOiKMF#38PqijnTwDMEI8G9>LV!1@tzMS|1sRZgy2KFR1$_eb5;)&qZTc zM+i+;`I9GB4IbKcRbadal7LBdAv)nZVAKld7UZJk=-&j&5NQ7rLaT~L{4B(J8H%a~ zw=@cKKCJP-u4BM6wb?4UNY>jGHLc4!vT6%wl%Q{0sS>#4o6jp-ER$~tQzlG{lG=$+ zX3%(Csy@Tgki0zkC@FIDabIyRxks=3j&;CH&>TdMe)-dmSI!B6eYVAj@5vZK!f2bP zG^H+?eTa~G)LslP@L}>Py1D%9QXM4gv8CiXQbGq)xjFzaPvE^}|2P?Wc^ah(){|aZ z)DgX5UR<#jHfE(Nw$7_gXE{tg=U|CsHd+@(yA1|5DuJn0O1N%#Rtb4u1mA;1I^K6) zvbh?*J~MO93GH=~lw8!^<2~FsFuk{KJ}dWk_esPEM9*}q1m20FwQg8n^v^;miIkDB z89=_mPvZ0Ax^F`4B8!O`(OvZQSsyXljPsaqi;;P@x{d6gJko}R2_Dlo+J!2 zvmT1)=zFLfPlm#@5{$8@%RB6NA6OTm=5!y}9$@lO$D2K=mN)D*<(7(vi+NCC%}M0y zQTM7|LW0S{kQ>kR@)W~%&{}@sLyhfYU>h}`UA;h`^8tabv$yz(8m2p^O$fuZm#rIS z`geZ`nGnoP;efk2XIq@=h9~DQPd|ajdL`)`Ijlq^dm&D=ui%yvdiI;Wj;I%nRmfkc zTQ`G~%>V)!c9Xk*uKRZ^af_1^E*v~{`%jrt3NWT%>~Rsk0rz*IOU-%PKSj~a3og@Q z--?nT*NxQ3h#e>}V;8Hw(l$_bRErDCUq=}TL+)%^K_9UcvPdH<0HfXrioElwG%InN zF3_(&%Mq;R!24H%M1H<0Z~49c>IDSZgGDaKRrcn+8uS%KOTFF51Ns@+OUJ_UQEtMP z>_MlX+zBV%gWJtiOI$!_f{EEo((Z?S-*Ja(C4e03Z99N;8lY}o4+~;eL{dvE`c)%% zXYNuqrB+=zld7xNT}e`AN0@`)ef?5~|PFpvLd|J>HM`57*o#uV3$id8ap z>N1##tS6FOLf>xDW9N28YNmmtZdq-_QvkMs_mIo%WYr1V%wi~2G?Zh6Q09|_XqZ|s+VB;zox$k(L>@fouN|a{ExLvw<0?Pc@I2<09R2oUqt;fF zDs?I;`Rpy%-IPHRqNRMd-f%D;%?HF;aA*D>G36mrJ=D9&Cq2-7U00947eeP&Suo@> z>SDkg7hg0S>Xq``p^u{!REq(#aTx8ht=|OtFKuSDs#iVc0N^a&$>@R7cSbZxo!pl| z;01eu=1RDn%pY6IVn#H_f1Q1Omkb2U6$pcvkyJNyE`bl+8}yZ4K+sn>IF8YLt=0O0kGsG-R71YXq?;>G!#mDR zu7JD`p%{S>sYx5&|VW^I){O!ol7WRFLEH7ZqJ z$AvofigjG3d1>|9Qr)UUuUQ&Tf_tft3zC>LBVE;;jH|tI&YY|{Mc-#Io!QNAD$-_? zru;8zD%hpn*+|SDb)e7<%`(hQD?luzi1BO9FBm`{iI0EUz3uY-17}y<)pe24s1UB=%?#;VTTaS$@>Ki!j&{$FI z6>4C=4M(`j62qw8*7D7$mlDwJs!X@8p!3pPr(=HKKFl5~z%Qcu`1WtPD#MGU#>U$) zjbr180K7)bkkwf};pKA94f&!+p;sC>Yk9eG! zAzxbFoHFt?TSgx5Yf{@-9RNcMw~2T5S`Kx0b>^3l$((tEQ`wxO*UBUVuXZK^K@7oi zfu;p>I+&ZR3=YNqO@2+Wv*)!l2EgbysPOEa1=dNWnVztHVmSglY{FKn2gj?Gz!$kn zZvgY-vw!pvHq7bCZ-+Z-RaS=MVZY5}$vI4BOEe9M2mE2%6bY}$ zXNjdK=CA?tkXGy+|@>yuXO-M`GSMetnn_H8<=Ne+6@-;$>%T+j(PU^99Xw;NWKB)Iy*8RRdO3G>O$I>B zoO$>`UW$ska+iv3jFSJmpD?e2VHbFeUfFp*4c#D(a6_pra3|*`R3=S)4`=p=Mt`IK z^1Jbmdu4k4U$Be$f5EOF1jZ;~Y2#w*^mDZ_bTJh(HMTb~B@z&TadvSsHME8CSpUyj zb3^LqoYmEWg4|8h?`S6C#g1`g@BG8G&D6^l6L0g&2>9?2@+Rp3_;|Mt|Cd}!)+!h< z@=?}~g%N1UCZT+#xJkwyp924!L+rf67sfB$>bSeH$GJMm>??)+ydf_qCr_7;U*i!o zoycZRudh^{eU#JY{jkgp!{#dIV}Q5toGyNEAC5e|FA&DHKHfz0X>8bQU52<-X)K~g zgOIVNmNqz2fxlKXHb*&_7gld86PW)>8#4gfgMAk$WeXXp{kLOsW}eesrId>EsMHj} z?}Tgq^5!ZsPc55r=1uX&@*1{TPrUr2PE!I6uS@E`{c@(AUa0voL)C{`bV8B0d7n}^ zESteNbpr_dpUP0HJs+-ypZO| z!5-NY%gkZ3UFg%RX{gs2*p|x)gr&uTM{%fGm2ID9wdeQNt6nk1-8K;Kk}IO~#weQt zryAW;VbW(tC#Wq3U$X?2oWZAE|2t8b*oKEE0)=IWN}|U>C9>II5(N@s?WmI8NE_wG zoI=ElNYma22~#j&H?I5Jb;D|Is4stMMjqE6s?Tp5eWE1*xam#qu-qd6DXh_wFx6Ls zF|&Q6MO;1UjDPZyp10yGsCOS3Ca7!Mrpon^S8jRHP;Hvq;N;kyvJM#I)g(7-;_|n4mrWb*A>318YDCgV3VwX<@%(g5RfPI z{0@+vLj-m?;)uEf;yW^6W<^ki#RLaANdSGw?yYSiend%ZQZBzsDkas#N+{}H)AAYw z_ilM7Fxcg`w!$0iYs!Io2^?GX#=$3hCmbqH%Oi430p(sdxTLs#@|DC z)mU1&5j!xW%c`r*YFXc|J}|&E5RUU)Bqy1q+DtP$7WdX&5~kXGo5W!TgXYym*VXuh zmLjrpEQd2?fj*YP9ZUk`26%@`0fRmhg^d*J=n%?Tt+>)A=CVaD{$fbTA}GB_6z+G9 z4U=2Q>-OVr_8>c01^9a~Lm*nR(m$#3YMQI+r0K4Q8|qX*_BL8t8&us_2xVA}Da;%k z0ms<)2JvP>(5tW@{QM?FgA78+%#QixHNSmr)=E&o1TF&gcoCpzcIFNjXrr4z@`P5bt0b zV#Wa@U%^_kkdT|=8llFknwmsNX05?NDd2n8taYixGg#??sddn#W$(5jRmrxvH}&Sc zrocX*qFh~#=`ba;nZ(Jqzm~MHu;N&NM{O%ZD>KyUi#3(*3zJ!S{`-)_f4q1IeS5;e$M;QL#7Jj!11SWD=BP@o9xdg1tL@q8lsRSRvJ z@PC~2v(5?v*Gu|dG?+~hqX0a)q+?Q@+V4Lh{=|Xt1`txc znxeu|DaXYpwI3{%8c4J@?e5)M?HN`C`}Lv4hR%Cy=HZb=i-duyDUmHMzzm-1{Vd~2 zstiy2q+DPlKHa}_uB0G}{n}D+g>)m3vJ1<%cnOc|2Trfc# zTBa7qf1n2OTAV;nmy>yd&kwn5a;y(HG%$o{oL`6`91mW3a*%0eo_1j7T z@o&_YeX`3bnxz~FOYR#%x15UQ@$`}6o7Q$4vB}+~S}?Tnime>bzzxg6f@zS9 zid4FK-a`0GMZQfF(bJrPLqxm4+T79<8)`RPTV@*^y961mnJOK`DL- zh@}rlv4x~hC9}(qe&85#zGDzd|F@HkV(|aR*gN(L!$jMb+qP}ndbVxbwr$(CZQHhO z+cx?=-5>8s`u>JWDz!$(8!3XC6eP+VWd!8d z8sU%EcHf{-ZSWgy=vg=fXn(9ZP%@?}d)0}4&wg3gg0&z-24&rw!MQRq0)H)Rrv^Z! zX;lAkIiCCElgo{B&}2ltgxuwU^Rs1>xFO)(^71t9hNiYT4Nsy4hTAcOH}> z`jqY%3(M^V|K!9A@s~ZeHs1ow1&6y0fI99^3%;XYQ&pOX{@=UulBts8&KUWH=sNKr z32E$+FR&tG)rk0J&Lb6(ROt>d!|eM~RZy|ze!l@%9+Lw?L<6xyH}*VHLnQ3l*##DZ z%c8YZazx4($v^I*V4Y3=Z1S0YKzMPpei1dmY);Ha+hkV`Yf&hlDA##W;v0T-0SuET z&&+Bh+~#6?z~@e&5{U>9L1KXp^L}-)L;HGK!fmRVN(2>i;&so=T16ZNG&{@$&KqBO zyMX3&s$d^VLeSKyI?#|*b31?5<>EGyv&cxEP`F^i%gS$BBw)h<9C_fb_g$N*7xfqb z(1T?_dkzn%-*9Sp&UXA^M*?YdhX7xWwC~!kXJGAYZ~?+z`JI7I1XL|yP?NF=K1l3t z0<>NIX;4HohGJ9pKkSsb6hKY9HJ)<>!wl!>v6;@w1p$Ilm=AcWrhWM1c^AE=B8<9@ zK`5AQe{}@n4BaGQ@L|QC$*CyOHMd_oIZ$LOxbykKC zT3x?JtEO=pL-{CX2)RJf*j*-Az*#@+bx2vRayvLTpir~XDd)jko+KC-E{1q;Nd;Wa zQ6*qs0Dg?x+fW&9{QZ4H{bltmkvE8j1&Jo=)q8lWb|MgzvS)QAT)P&nrKD^e>GJzQ z;_Z{}ypSUrz3|a_{Q)P`CF6#xheq7<-4j>&)+T2fa6p{FK|&*IS8B!U>&kp6aV*DH z+AaBrvVZ)~bDNFYzH%noX`F7;u-7SdH^)`Ef z9vF@(j9mwZ$jkuxw#DuMe-Au88o~CF-wsOK8~^BHp8$jXpb*d|EhmM_cf%$6ZBLA&Q&QW=i$Y28MUcj8@2VLPtFGp*SxaRC#!Mvm84`Y>ln{U@d?kbQ{K)-n)~oL zN*nP6Nu~!oEyuk9$274b0|_hXqfkF4>S4AwiP>fO`&Sl`A`#mOT?WzM`Dg-sjcqnq zyId7!;kqy;wwm=x%t6o74sK z0dT4n?7t#T>Lv6+hm0Hmg-0I!3d8e*fkwI3=iNWFPuGEYmFMJ6-t7Ecnfw@POxSg~ z(bu_U55dPzNYtRHR|$)NOxpdKXeSJIj4-&2i;OX_pry4*hQjGX1%ZMo3)!Q?{?4P? z8Kqrp!?~(Y(hcBGAzFZlATE_7atX|l9Qz{(Mx}u*R`AkCVH`nddUhjVhJ9T0D!j{0 z z+28=fbX5ZsudLAZU;?$!nW8vQO}Of?mfOpBh6o8{ulX$Gj1_*tJ84*mm}cXvmh}|j z=B^lp1Zj94s6IuV0UF}-4ogY95g$RlzV9pdBQc6edUW{jUCFrGeQd|c-x}F>0`)+M z6z$VVO~M_hcHKGlb)}O7p$Yd}NRLl}PI&jyZCNTTe>qvj65 zIPbbNzU`Rr_$7p#dwN#mU#kq)E(mxL%5AFh4b$1oduT##R z#?}9~NN{{s@GmH{Dy#R}T^$Gpm|%{+jr|!FBBX%>Fiy{zLW+Ly+*09%kh&Ns*OLZ3 zpC)yR$~X;UutMH82$d18aU8Nrr;~pDtXBcymSAl8f~y4f7@i}DrUWbY3+FhxtldKO zy_~*2I+YR4fXbTOX7kZ$BQ=@3?4y&`-xgO$lIA$}tSx>2pP8TfnS5kduQ2kX=4UM_ zQ$0MSU}P9YQB?UzTzYZ4$*CVZ|DiuyQk=@@FfYo4%%n4`aVOOVd6_hO+!gibwfOFf z*--4OMPb!`Z@-(ea_Lw)Cj3W{EU8+?uAi*q>!GtRTJJi)C_rKVvq+UZen%ZQ7fs|Y zZQNC9vhQib9d zQiXlBo=g*l@tk)qCh>bt9J=8_(~hbi&@T}YwflQH$+Ox8x^>cB2-DCiZ8yn;tpe$h30{^2^YwwiV^|R6 zzWI|iuh`BZ{Ib|(ouIie-B)+DnoMy5!&-p(ZZJHGKwS`!g>^qvm9zJ{ z*BDdpK%#H#2aU{ajltfkvc$lB_bxny2?p+B68eRVFu$6u?d^nx44I4FOPv9mRDd5q zUP&sRFCwNS-HuGlKv$~0q?}R;2`nft{2n9i=Oh?!Y^?JmxjS8Hm)q76b`~a@+ zZl5RK($WUyu<-sXe5Qkg#h9*93=*wjv8aa}wT78tUZ#I(KRc1n(wN@z4A9dUg=UpL zW*WPSR!X0e5cNwpNtZZrfA%kSP>HhT)+p~RO$`eT$6HQr7cc*zT7$=&)0vNJZ4R`8 zZ@f@|8_mtNLp~GAMIjWdwkL?gtjEJiZgTJt;)gE4dZ;2H!8*A4jAdJCO2w{!p3Om= z{B`Wl)fOHM?PRg$@06O{Ed6NZB@yrxE{#mm=W-0Bl)C8*aG1`|J&NGG40IUf7KwV; zFAHK3@({dmY+l+*qqdHMz{HX7)GL=yQ9z2P)o=T-?y>rDj(t7X*+wJWsxAYB&aFJ z^+DBh4`S#`-$J?10OZbnl`o2!-4d%QkPRZ=-K%<3*VsW&eD#F~V)B5feoMQ1;iG zZMGVRZ|2o815eb;4El@=E-56*OMzy}-do_$8^$nUj)cQ_1G@bjfSzzMc?UXV)a!%xAl$gfrB(SW9&z=BQIR5Y44F?%ROF zTO8q9>@|oJRM;uYHYA4&I-h2N+&5{GkC=}<@Ylz=U1o-9&HskInEp@f#mdCM@&Av# zYP4n?x5fV}DXUFlzvi`3{m~$V81OmwANiW5`LRJ0bNtVz`RCTc6KTdS=Jzl0+4c0Vy0z)9 zHa6E+JTW6myK-@3rHuHN&}>y?l9F{w5uJ2&y0%s9@O$Xv=_%=Bg?^qA`&q)+*d|k; zuj(b?JrqCQT%RWs)6DjUziky0AK%HjyH> zg!%b|##>_9`u?}*J??`@0_mZkJ&)d;i2Abi`tlUoC)|+O7ma*vP3D0?Z}^_sL9wK8 zy325eHoNIC04)QHJhGJb>hkQvo3KSXg`>l6G6r8#lkn*12OgDL#*3?q>+=^vw6ST< z{*yjVzYP(;oKfM@bc+}n8-1hLuZdIBS49+w1Sm@4MP8IHg2vi5m6Va2BY8pWL1F5; z?kOaI@rVPJo>*NgRiq*DFgh3ZS~ha3L&jhr1L+wlZ^Vt&X2Yc79ktEa;iKZ65eCuB zm&5vIxJrjrt)8k%X#o}XWqw}gc!jH3a`=j&0(EMDA3TS+4x9=5T8VN!twF#lRcANF zMy*y8V^m{35h?Ey%r8H_gu=Hc4?DuhGG3#DmI|Z82P`bXJe?&V1wAEnDgsFo2)ZawI!Y|M*Gkd8dGeRLwBard@2iyJ!eqOY?$lL@o(x-`Qubd zt97}5G4(@zG5WZ%#^X_WNIZ)H3S+odI0pn#f}?u|2VVf8LOrdX36dunnf;(yR5QVn zgBDp9dR7x#=*wMa@36KRi7zz^D2D%NS@atdpR`3|6}1Mn&%d&pnpX)P=|IjYvC5^; z_BPfB|NZ5*%*gh*8gcPJlNLfGVp{a~xH_g$dT3&RE9FHDoXQfM>Jyg^)ny<6?ANu; zs?$i@OWM$L$FI3K9E+93>2PEli%Vr*BW!SH_zae^U^P}VPr)iIvIi!}Ke3yJtpE{b zX%2SFL&^5*HliIODZqU1`%Wjz?s2jX`b2oJ_>S!<%K1=?sS=|&#g0ux<#toIE-9cG zDZYC~Iez)=-@H^HJS3M|VJhD=vd9(nI}Y-C81sZWu-eNjuiYe?is}H>rQAEZi%it_ zZ*PSKf?4x>%QjG}0(H7bRO{}w`bvin)n=iU2TCIa=vV(m8#jpmQEc;KCQFMUFxz^v zzLv>t*5UP=P6w^u)KGi*scb)0E&E97XzQWV#?E4>Id(>lP7hD#=Udz%vmGBWGM%06 zo0)+9H4GKcJ78GjbhG+GVgFIMmaa0|&<6V2Hhse3R3x6k=hmT~eQ>a_>mEv8WNEo7 zXz!A7X`U-8_D`|^C#Fe$({&h;s}-@oWRbE}MG(Or@Hd67=fHYlKkJVqxz%~pl zc`65nyU2VQunU*Ky_j}gWoMBH-;+-Y&hv`?!s;`0hqdK>BP15Jxx`N5cr=+iXPI!{@50UImEtrQr%W9CDdRX#)EPk z_OXl(>H)F1-}}RuR=?ripj}APO1HSj zPym~4e0=}@3HkP4Hhwqu+`)pbZ`nreu>@d$nEypr?KE_rV(;x7M&o*}^nrD}&eMQ@h2iGuwtZ(^$aBpkETu}cu#*Nb;GkrPpOJxW?Al4CR+pHx z$1nzm8fiR|n;oRGdlQE-0J`BTIN3|(VW|57Bxh6n?5~&K%dAur5(?Z$=oapS?dq z&jfO1O!p7RrG<|_Vx2!al;RVi@>S2pIXX*@zx}BMj=&BEjZi?SgXu_-6o41xb7XOp zoIp0u9o6NohYEhVw{`Z0$AIp~L!5p+c4LqKYt;wK8YByBKsd>mheEInT_BSCdfl1m zc-wJA(as<3fb8rJzOU^(@E>K~Q4@ir!EM23R4aPjdpuygSBbnJk6DjLy>Hvp26l!3Cild3Bd#ozoh{a~9?t;lc6! zFX*O*`lYuK=pWjZz?mhH2a3UzPY0aDNr-q36AK(214%KglP0s!I`+l?SQw}8gkCRR`CMfUIapgN2gHUYJabq;OGaBY8F)4Yygl!!~uCiATErUa2u4Ggml^d};AkkCS-T)f} zcrkX9r~jS##X%>-uvk=3GWU*QiaTYf#Qz-{I&wtHji(@xsT|}WAkIG;hApJjaSc4~ zsZ=XzsEPqU=+B#1Y|$|{^_|(lN>^|mFD=gyp@8#U&F~BvndjyNIVCOcgADsGn~4A- zI-8{rS)r1L1+P&{j*y=v1LCJC9zh7}ZI}!g zI4H)eNm88;PsE4{GadDj_a=Td|3mr8R#A$1y7VIV3u$Q=#6&nX7=+5i0|pKN*@LY_ z_$16wUlE>E0qfjfb)-EL3u!;ha3m8}%upV2sD3UgOqVZ8#;1?Aau=<0Vqf&2bj=TdW6gv6gtG?dl9RV1 zt5jG_C>Lumrm(BVtGA={*bIxgPbGH~s3%_>*cxWpFE1cpsPxYAuCO`zm@2L7izntK zLMc6tn|LDJ2kmv_)Nj6QN#J9K_VXCwHi1YkMV=-wH+P%Ci9Nv`zlP;Qb41@;b8jzB zoZ4lqzhzfNCpI`lmC`RPCro2}gW46!{qXpbX}&D6Cj->#O)!5ZONj6q@}t?*q19MF zS<^A{(*Tb`4OX}Yg0K?=2s6x`8H(!1knv6V%bze1pQtG>}=!-{g1mlkgOJTRdN;5NA8hjLvJfqKf} zHcI1#r+*)3t_4131{nepIEKI2S~XesP2=)xWJhO?hI?ep=-ZF9+lGMAr&CMlI%j=s zTL_*%5sHX6EttZCeGl2{Jsv6keq)4^bBSHy-CyAPg`^zFUNpS`dIIInZy?2WL?+1# z-y&W6dme4(cg-XZ-G);qj6>6NJ6Rd*pq1e#Bg{g2(jJbptP;z~M6`Is>ps)Jzu%(z z9rWi3VCy5lKf+WMwq^VYAdk-%r?x~L3QNRFLs}_U$%2?T=!5dmEcktm6xECc6}&g3 z0^z%ycjil>yTa0$Ja@?Ml{4rCc9Pr(mJ?rJWM}*BQqM~~kq80na-X9uc#aY8@cf#| zbzop+2DlvNzGQdoGpWK7)U=|}*_bfZokcTzOpQc1y#{ZiR`vv0dg;bVqUb)YNj3<$ezCXE zSNy1}o9a@hi(baHcU4Hh=465uJn}q-6P2n}ByA6^E0n6SHw4JV%JE&1CFmhtDM=}Q ze@Ri&Ho(hVGCMT@`JN%><>+&D7B~%dMWKGsBG8ofJWWw3cs}%-mr&R+ytQu#S&F0Q zrF3GevbSJmp-R3`Mlp|`aOGgR*Nz_*7;7rL&ErMOV|tI_>Ou)GiGY|HS)-5R;qm(U z`3GbaP#nuj^m&^;rKbI_V(~Zvi0W$3-;XyVY6hlKdji!cvUg041j1c)1&SMU=biEU-r*`!4Sm4pmCts z2Z4<)j?jdUBCdzB*IU$C3>fS&Y}vDttMtB9DxP3-+)y2Z=kD_>-=oe7r4{;jLws-Q_APgao=R$r&>&PwAmnd&MTh=-EG^{+kU;1N}l>E);}N zXZ_u~MB3viQQ z=HvDIrwc)@m?%qD$8;ZXDZ+Cr|G8fPO%;=gBZjBoLox-Be$=G3jwze2tP&<4t|AH8 z2clpKznY1HvyzEl`(_G8HBvyHiHez|fCM5jA;S{}Y6Fl0V@#w_ESV(c$Vc3aBIKqD zwn2_}=0i6Yxozu_+5;cbvDqu}(@+-E26jMJsy)pRDWvf(oyhwf7=t5ZAcdm0FM&&U z-o@@n@4kctww=PU+x}pzhzpRX)SpfV6p!Hzy@N1g}p~a2r`?CBL)oVFliuT1OhdsaLyq zh|L*YLOiEjB=xcwfBT>dB%X5qZTe>k=bvsKQ$tojW&ru}u~C?F4ky#SLUZIoK}&ra z$E!A804N&%x;NROa0lSr+8+R{|2`|wDtAyhJrnUQ-N1+i=fKVlv5*4c1ZO=mjbFl@ z>tQ5z1p7y=&4;e$CH3)mWaa1C{*GFCIUO13W>o zR{|Mj(&CBCNnH+inOU#XcjM~@I74oxm7zO{6)|SP&AyCre4iIY+_X&;6dXP ze}s=iW}Wn8I_=iuC{6VjVxP+phIH+Nj79m$5eBZF0>oJ_&MQ;sd?U2C?4HX(Q zAW_Pa++JR)eb`A#XPqs?hEw-sTf4F3-i+>PKm!euF_g&s0(jX9UT}v(ur$!;&boD- zxU4%&8+gf%LFD(`uK2xI94tiRm67PXG%N&3w&6JzU*JF4tBcAoW zlKqp9A4^5nRuiabeqhp$>;_8m2f}787YvjI5p=qs?gF&#g|m^CJ|LmCFmOG($wJ^T zJ>?)_$#z=g=xqeqCm9+kVUrI%)&oJX&WDcgyVqKnp(C9NyX~@m>JWeCqDFb2AE^a& zV8r&Qfu3AIUcRk8G-+wBfyoVk{)qb#qE4%;u_9sdcIX+6glC41jC&D^>a)Za%5Kt4 zEv?>8+ES%oi$D+mf<}qj3!c-fh7uw6Ns2$}R=uO^eb4G7KX9|QBClv+r{34@Et-lyCDhcO+N18Srw>{qfSw;=K-*IWK zY?G6!wH#n%JL#N8S05F2(A)=HlLH40?2B?WL8e}CF90Cn(uCZ7MBpS{{djB@2h-TN zi~EIN-Zc^Y-w-A9|Ar{pnf`AOrHm7fm}Ae8G5+~dC~#X%#6C+GZ&xNdID~+9B*+A4 zehR6>&_cR}p=LIGxml?8BYDkVGVVc;i}U+ske;c^%QFnt@u(!Bh<#^gZzt2!$<*}V z+?WShoYN=$?>`NklQH)HLTwVZJh6Y$Xk&7KH8=wO1JNqHwqYpaXuLpiWNtL zDH*M=@0JcExjNimmm3r3m*g;1o<_~n-EN?xIx=y?N%I0*y6|ZlTE|?^?+Pr$D#j)B zDeZVaAD3A{7m?NRsw1<9k^d$3Rwk#88fG80Q&ab{OV~AW-!}bF{Q=ALn6CI9qifiXU^_Va?5y629pS`vCQn9`ZZ0p)7t!>%&t*BcqPTR~s!p*NEvRE% zBy)7wWO-yudJ;Z`b%Cc)5A8X-x;&2&Pante?LX(?9auHl=dYU9a_i!3Y^P?d=}di} zKk8yih=XJ~l+qU|n&IKQX`kcpNhU7D9f}p5c#GA zX%Z*Zjq4tY6O3TNiQ!{;@!nX*&zvZunbf)5b*o669U!OmXvJ9=m7*UhwXK>c^}xrf zMr5tIk_ZC{9)ChttUld>B~cNWXW^PX zq7g{4n2-`T2k3Qi>9RnZDT^d4QFV6A zWw&5x-IUETVU4pw(QST_(W;YJSX_7(sYP8~FlmmUNl&Nwlssq}L^t8;CfAy_A@#{f zU&{j9qUMI2j9A&S=i1zNRz{R7yEoW34TX0DBhsG9VpYaVq>n~rwk(-zJTw+Y_JVL@ zYSdI-@_b0xwfy~Dp&(BRl8I>JM%G$7DXvIMVwmUmy&6mBQY)rqcoMH*0+d zFu~-nX&AftnWwz**T7OTWuuRa%0^LUbe2_{F`_AuKEO}7_eV$SqiI7s%?Kjq zdJ)~JYA1ukyDiwpqywGCcWpLA#kKy02Nt@(V@sj;hHAJWyo zkvFRf@^N@(1DmxP3IYq>Xgov@<-H~&6BZf4s_`^NGN&@0|n^|6YTM*?YyN z+%k`fcOBq(w^&VD6gE-lU!c8+?lvukpB~4Q*ygrIPAzt3y6A^9ey?@eCFW0fA?hh84*oxo)I>hfJrF@>~c+^A)xzh zSE!62^<(jU{mJ)N_Pd<*BDYi9!=Y!hcGbCA+R?o^1A5G{M^{jhWh-Z|Y~0`+-KBKx zMaJfx_GI?ja9Eea2F5^84WrLszTP^%J3o!C!x)Mu&2^DJxtPvN(GL0v< zDf84NA%r#~bAP6{k`$WrpUOGqKK|(Gawj&w`Dz*S*(Bw(asIy1m|Ytgmm!fxwEGr8 z|GOAUR*gzN%O2;b{e&lnj~p4dDMnBr1uLIEG95P6uSHA`<~C(Sge+zzM3O_2oqipB1W zo1fc1=OUR*-EMbnd4PfC#tOFW6btH6#ieQ2+U#jLv_4BUn*VGre>;iwJQ&$SLqHzaq%?1czCcT3cfmi1sUMQL9yx63K#59M9B`X zy8`?FG#$K$i6<627%tiV6IP$cm;gDw`D-yiWtugcYO8Vu)M(H%JXH=aW6fM=PId;s z7uy99oqtid*D0FC3lQmJ1#eYf#8^7(7rpq_g_9-^Xy<9dr2{4dtOlQo6mS~ z6|-N!pT<#7kh+9_nJP}R)GMF6io;jE3IAdB*N@aEJr7UpX}JHLA1$`ummP*BRU%I( zN)VZ8wnqx2D1!(~zeHP&0^_c!P^8Zg$IEX)NHQHpP&7v*P})N1q{U#6@McTLgTP%b zrpewRc0OROfF?%9Feh@JA3Ucp&kkdC`NyhvY*R_WRCB{i5=7xaNqn?WS;T$n|HxsH zuHZvvXs_FVs#AK~aqAxSkCFB?oFIo=f_v6hefuqVZ;7t)uWd+g zqt}@sHyn8c!?x6-b3^*umeS+!+@sxb(lXW#JE&TNy=fiu{BH8i7}3O2po|{!O<1Xp z0&_P4WQ=HHX$^1A#*%m8thtj53TS1r{|K55aWtCw8+hd;*D223>xzwd1mNps>Q)dp zeITwtLU@=sqd8T-TUGIOhGcknkG$*I-%~u^(=IoK(+30C0OwZd*GzeUqr5JI5^(nG zWN@|1F&Gjw&nIP-pXLFw62oMw#Ai*UUcw`jM+Rlp*Ay9i=*0oUohO5v{%z=>iOYHa z_^0;0#x~)aS6?3-BnU{G4Ku8d+XA3lpkUuM7y!KKiuF8G$=!+n1fz7@6)d&0E#|8< z8U@Hh=r`kc)B|C4DP6EK7NRWy3Y%NBwtmb(!1f<)2AdCU#X8n1lO3;uNbtWQU33T~ zewj|1fkwDcI@*M2pA{WDJG--Hc2uHl1VT@W$7;*eD?d!*`~IjxV&yWIY5k-OrQ;hx zN^%@+@#26(x~rX~m1;O!P!9Q>#4lo6Jr3>+j3QAR;M*fp;*gI(R;EFBxyf(7Yo`^u#TBRdafY$}b(eR*6$ zM-I06_eg*$eC#l_);Vu(KZXwG6p_r&LcWZrq3eQzuYV^quSVh!LV+)d|0d#JBb5@2 zzG~mTT#y}uN?H$h!UaX-YU^@r7ol=8>t8t?c0uJWu*O+!>x%2=H15$1RO0;Fy7mnZ4HF$Prg$g^4 zq5R*4dTcl`MP87q!igf1?lULsfADI(&P2k3?Up274P&k0j^3(tj6K`V}Q< zYm~%9lCT0YxhcGNTl=azyhS4*sj@`_rsNI83v-%hPWuSmX8pWOQ8OP_!bPHO!U9kN z6GARXUo!9lS#o5uFdj%6ec1BY+pK0&|LX;Arz)eohxT<`yLJ0TCdyvkfu7VphKxX^ zN6TJrnj)w7AS@USq54rDyo#GD7fxjBCtM9cOUYXx63Rq^K2Kg*p}7ucVAvwx9&6B& z!x9$A-&o7%(VKP9SXjzO=@Ee?s>4ov@8Q1TIkVkoo7(O8Z!L+IToksd8>(K~E4Z1y z(+-Ho!nW0ANi{fITR!eOM9*!Yt)c6fdrH*J=c^Ga8_p?j4m7-orYza&;a^|%mz5wW~;Z34^FMff_?L50b2AyLKDPGK9NdN8Dad;4kYEz<*RiU6zQpl-&pRB6=RVu!-vyq*S2sd3Y3vz7N^*D zQN;c3+$6dbUgzZ2Q!t&_?8j6wi>w>VVHKBY$6h2gbD+^E16Wlh2+OrVw;d4WYq%h` zL!_yCts#=yiCsRza-!jzdzKXaveA(P7Iy!BFLQp*@Y8=UjdT+DZAxtcFt^Iaab^6S?X@tpg`ek}vSFH5!#`yC%2C7!uhq zmAna?ia5$)Da5AJKk%`^XDe!cJ&@usl9lk#cJ=CShOLi7IW289P$P;r?G27dF{t?F zHAX6+*Nq4!lIPWDgy1sah%xNCJC;$UJn(h$l+M{`+?GNEn3Vf1XQ37%z4-zR3T{o@ zwha7JMNVCF+ohUO3u@EB(rkGpTvQ)L2GQx|=r&OM%X2=pY$$@};2+GMBzAYh zWtCO6X#>y2rNOiwEUiK;Z<4ACVhGmsvF z;{f-XW9a!bZSo{^6Wc-lbbjm@r(HohD_8xBpwYBUfvsfYkEfus+vNKYNIiTyVu-c<8>-r_(EamD1h+Tt z`ULqHB`Thfqvya@n!gzChu9q$54wM1-O0I41P-P0*I_xFNgK@bt@!+}CL*y7(a|&h zJaI~aYj+ODZ14@2^Yz5FvtWA%p*Wu>*!`t(6NwaW8d#sY%l&*K(e_pP^1-F$)R{Q! zpKdqYa;&(q7VRYh!{d)^3$%S}0TS+}4S^z=HaN0b*HJjSP|qGc%F~DsNc)^3*hIhm zn<35vmw=UjPS1S+(z@(R5?}U?OMnRF%wRE7m}&v{Uw6&w)Cz194=T}0vDvqV7wXsY z5$GRSQ*CPAdu1KfyLV^hh0jIE{IuIZC^M{nv3)0Qf~Tje=YQ|ZYgiTYReg;`e6DS! z@27m)&bhQ^DApU4)-k!Z5^sG%mR~ek^1R@ytb4q&Gs>-tyH#NLmT!|C5ixwlW#i$u zaHaECT{j_>h_pR7NI*lyK3rMqa*2Zqa{|WSd`H3QW3XDE)L)(p0l^mi5kKAq+c z-{qB-HNP!0G~j=X_mz0pUUuZkq^Rabkb){3;QF{~O|D`9BdS3kTc( zh2-HzOXh!6u{}?R$UY520a?(k^Ov93R3Y@qsnialh0R1pZe+!f|F10;2oMfHvgl*% z7u`2QWyRc|){{l?y}qFUqJgins_E(IWU6|QzNj(2DDk(XZ|N$?(b?mED&V-P=-JGZ z$J5*Q{=Q|kNFJnY`}})Hz*{iZn?hOMQn^Fk+)C_h4gB!WbS{r^l4V(e9}xuuD^I8Dt@Z}aKKfGFP<36@PeDy{o8p5+2%{OjF@bd4ytc|0NZeCvjvfcjk2-2FgP~sYxy`SzYG^R7$I`M_N(~mOry7&B3c}{G&)ZOb-1huL2rWBH- znIb-0)seuKdgl0z1$pc*mW%_}Nonj^ubn?N9v}-qa}=?o$iNdq^M@yB`=wQWxFrI?Yj2H!*xmF$l$|&(mVLg;0hao^y#Z#cUxJqR0^kY{CJ% zw2V!35&#b02GgQLT4_&Pu+rL1VfQ-fA0&8b_GxPW859Wg24eK92{;$m?l8dYmKTrf z$ayRm>(I#EH90fa!|x&|-avCyY3`alA^#L|g)8zQ@6P!nh8v@@aKP@IQkMxun2YSS`Nk!yKP8pr`Ome_&85#I zDH}GM|F~?fID2LVm4Ttjh6bvDyg&mTqWR3N}YXm|NMqyd4H0xfq@X4- z97*n=uOw4k_~k$xd4WjT%E;4tZ0&a>pe81hnR5C`rrd-gYXH!DB_mPiQ`#%&vsiNyH(KIH@g^C?3fE)-Y!B!655NG%M@Y4}Cu@eANL2A&49E2`jr_d~uPVR0Q${6rio8oD`eI;MVZw-4^5e zD_Wo=ABoMXiAc0TNSaN(BJPquyucqUA!97?Q|$+o8a)=sLz)Bq0NR>A-*GAptlHdkB8V-Ho9wC5tX6 z*VBlEqM3Oh)beT6UTQSuy1KSodKiy*9F${AFpLPrb77@--prsYMa0y%1(`k&-Dl@V z4ce*1ZD5mpxo`YQF6970k?5{9X!gef`qenKTk_eRX?jdHZAN^^EjBw=Y6{ViU= z;B}S?pRMyDGI6|Cx6P;sk%NS(jzeqwe_^^%o8#xfM1MViaA#31uy;aH896Mjz}L99ZUQMhptpWdAC24l)t2Di&mXpR1BQA}9{pWUKZ>TUDRX^LluMmceF0)Gk!7 zv1Aq0F|uH5sem37TFK^jRlt0 z_o*4Kur;ati^4Ytk#`T zboS?Hj&hfl^x34zRN*0K^-`^M_wJHFEaK`k%`bK7g_+)%&DSsW?6fGGA!q|anbeW8 zf-)J|mSkS3Pw1)?4~wl+O`Ebloe7objP}SAQ^A4$A^)AW7lCwcuq};FoTiF?FYqEF zKtXGIHQe3)%}i24vJq3sIAIK{&p1&@=xi*yr^Po>jj+rvAZbV`x3G;iHFV_D3d~5J zY*~AVN)zJ5tr%49Oj|mtl7V+kk`v~nGKi^qWO&2m4H!>O`7`8A*vokb$y3#B7gA{? zA|eJ23HQz_h+7wKHSWeM2GJ!YO*jdI6dg3sFh@!L_c{?xjIfd44Cika9REGRA3i@~ zhLO;gmdG3Ke-W-u7qV{Hu)Un%KUyI98j7Xo59zk&`+XH!NhcepP^a)Cd4|M%CQfQ= zOMiOvSvk?sEL|`cAtZrz-1T!HBi9T@@`daXlPKq(O9vNCJP~1BhFQPXOsN{;eotGNr z2d=6TsuA5zz)xTiF1}CI71OTGDpTE9ahdK@_%BubW30+F0iVrKbFQrI31rV3F|_-V z|HHRFmLq7LVfXg)A$cK7_ZB&LC}C9y0?M&giq;FlvSKAb{~saVU5!wTjO-~!*$AXT zE_8$M4K82Jm=0mLrPqQx;)gaQG6|5Kk+e!|1peGW4E}=-Y_JY@xC}UzxgPq*h+`b< zqX_x$DqC6u*lgT?b>xC1r4r{&yHdZaj5Auk)4Nf{lg!J0xiW!@G7-D1TA?Ck3F3_b z;kM=VY0MECSlArU_MNB50Q(>9L^gG?NmT&*TWSXJbCCBVeivoCtT|1zVDxkx6o@Z z^B=;Iq%g1QPWMV_^|L!llft354bstEUcYZhZRxpgpYxx!b@#SK+vDDV2YRBFgAQ`D zl}X#>*1qiH?8~b^$?Wl6dgb<^%%)p)By4JlRJ8rq$Z32TSz1n5w#i*k!j#t}{#2t& z=IzGfn%BZgl{+qh*2@)3IfmLofHH1nC)-YPAB-p$+Inh}%FAC1dR z2i!Jy*}2D&*0tH|ofXw7;c2recIYiM2-N_xc`v9c=avD(UjPbswt77Gn?KT*B9Qn- z69J+DJy_3@2xR@PwRfJoKeUWtwa_%}`rc5bYs9B2FVCQvf>XW)vZ_3h=P1+F zwey3hzE0U3e^!UTJ;`~eJQV1{Pf{Tg^%YWIn)n(hZ*=5_dZQHhOb=kIUbh%!ih&#B0 z6LH^U&vLdSGIHfw->S`;&a|b{MPA2qmz`VuUOD>3rG;7C&djJym+8JaAF&%M|W2zdqs z6!oG#?4NJm>o;a0mc;jVW>p>WDOR-8tr1b9r8q>Csbsg2$X6H|Jw#0^Lv6bWI_~I> zf+R+Y8In)+k*!u#OD)m0X}|a8S0BY}Nx1+A%EWp-&QkK}s08ZA&E%L9OezWNu?8mm1eXPlrk91@vz@7okK zE!yiJw+~3K98}qF-909x=vvxI(#Lo;ZP{FU)kl-lPCt6J;q_(?K`7`0r+lrl@~}b_ z-lLvq?hDenBZi8fG5JE@^>$|r{i3RFj@y3~o;KvWu0S|ArHl{~d&gZ#_eM;<_+A_LbICRGmT{rGQa(~_z)KItdv251f2TX zWwCa%yOQ}pQhrql1A<7Qg#2ol9bpgwVDGg=u3kyqMMq4GQ2ir6>k(HPk95d3(&SYy z@#RboyMsnARnQ#mmW$uuMT{qomv%^H)F|p*7uPYOvq)wRV4OD|39$t+T`C{GPeOu z1vb>p1Qu;I_DfbK<){q*^(Yl!_QC7x5GcY*tE-A}Qox(U(Xjy$)a;f@v${#&+Q z0!Yc3tX%%5Ez;1i2oNrSq)epP=1 zCVassLc0j9s@nC=NNcMQYms+GClk39cp2FmIOhNA*FlMvI3sukHGk$zK?w&t>!z3C z84Dyg9bacl={TlQFSa1Ii89PzGY3{h)gqc7epu6Zn-)j7SxA^y@`xKfGiyRrwaXo2^$V@+l?F01ENWH``^$16=H*8R9ucRx_AMt(<@6;U)34yCr z9o9M=(-K*hju?DAZaE@wPDr+p@IF^WDetc>DaOXe z99Dr{YG%d=ksJ5}D=T!XA2YG%dbJH7-GueSXZsCm{&HVhH&6*&TEASsCNRIxtyfB) zW#>+_KRU)3u2fo*Lebrz3YxgP`+lxcgn!IT-rIs>RW;XqA()!pkl%DaGiP9J#DB<~ zeVi8u2X6_EjXvKA#grCAetx=g}c0i0GgM_S8w)Pt~B((SY zf}T6+a_dGz<=Z=Hd$iX6=LhsB5<{5OVN8cNfD6Vq7#mzbvPV(9E)Z$e0~@^3N<6Vl6= z*qS+;6S8nJ|KI*W(O6zausSe+EpO~{0yfrG(-gv#za4EHeVFG>nvt@Ihx4l%U zmTO;dcV*%HJ=C7v-WsnMCFv*IyP}B8x0-qAi#S$%LRrDRvjUw~Z037K`b-5(ntF6K zgVJZp^GduOw|&yt^WTx!9@f4N88#>dB$lXb?r^`lKUD6EQ7i~Y14*HO+&xJF`2XHZ zYcpi~%!ljG1dN@iVR-h?pz?fT0#4AKv4X4zrUBjzs((O?bj&^MjZZ*~y?D=Bw_cuO zk$i(Tbc>8m9xf}~R+zIkEw!wr7sFg6xwzwJkRA|F3s-*r+a9s~uRZ#owP9uAWdFbS zX8K<={(p#&HvwY@!H@;tRo?6Q#x1mMWL`AStzO$GIPqXn=8xi_fsMzj>+I?5nc{0} zm-uY|z>q*xN;kG2_47zJAmzK&|B1&-efh+jdB-36-S1Yd>+-d)e|+GVwFSM-E?e8X zH@MxYO$7Xh3~7IFK>xG{ddj>Xf~j%HT=BeT*g7V+lgq=6g+T8cB-gqs{xHR9RIoep zjL!`VSA5gLkn!)8S1{tG1{X5j{TQeN|Kob^XvC4FSu1KiMQ6{EV?l*W5N(`#x@Gn5 znB6PEc-G@Ly8|cwA24J$UaUFs82V(T=F;CMqDj_|lo)#Y%LiMIiG4*+I@qen(r$R# zZnvvg^#6b%d#ceN^}>db)VCnndTXjIa3i@#AHQw{kA&uPW9eVMI{_pe0~GP!kGyL( z=wOvDrl0z)Wlq=36r1deJ&!qN!}}uojAGBTe0gPk*uFusNiei4O*VCF4Z9=XzWV_Z@Oy4ELBtokM@1y5Efk32;fa4M2|7 z?%`E)@}|DJ!<}{yXY4CDDtsfn;>1nWn<_!9OTw#N7M)eYFhd3a%415=@}{yHb1{U? z_uE4;WWlwAiBeMPLA-7LarwH_4Jml^r*-KnJx~Y+@MF;*5y)zA#K87v4m~Z9IY=Q2 z;ZQs$M}-5-h{39|F7)au8EnnNq6Wlq|4_2k*qT?nZW@RU3^dE@;>XB~We1JeaRO}-|K)J?QoNC_dkq6j>21Qjyd7Nj2oly2d0jGETv169tI1c zE?KkS%%i(YNXqDeIy>d~brJWm{ZVx9i#bFR-G47a26*1buaDcL!ADtZzux31#KdfB z#7`80$CnCBne98yNQZU*PhQi9wP>X&Ge4LQ7vCyayt_PoHBb}=D0$D43)@??2YdYn&4s_P z_QBA_iS1|l)xbMWA`!tgWPNP!zq5!h8TB)J)Hz$;agYapn! zaFG_9&QYm;uVcaV$GD+}q^Y%+)n`dl0b@9AswAqJ5AC+`Z$4I&Y2s?qdO%-<1_B_n z?Gc~Hg4#_bmVg?|%wL^T)@TRyPg-zZHXnO4_C!$&x{vUQECnoPf{WhkHZ&D#E`#CV z;_76u{Ar*#FEh5=$<7W`%I3xNpO1wPPI}DN?)DP~rD+jHYN-aslO{jW%NO(WlQ*`x56@dSH zZcQJ$gA(=1l1vE>H^JIjYkG`}McnyIdZs9Z&|T#9oXA*&_!U`VH7p#OB%w=2qu19K zB^&UZ8V^}wNrJp@wMvvgmK-Tz%-LZmPza5Oa6#XBUbhz?kaOb0W^|@u+)tFdm@W>6 zAKQO;6e`2)OLY01?KFPE3)C%lP3tA*!>hU6XSV&}z^f-y2ju~acWLxG02fb3u%F1w z&&TE2Gs4i{%;N8A$9!!}OOBr-)@Mto0D|2{{who0tL7_JNxw$F2f0qnWhZ#&^dfD= z$=gy!2qZ?2{};XYI6+`we%y_BzaWkphC2I zzo?EFy0PG)5KjEG8PFpwIjrlfb`1ic)qt0rS!yHI`U&!{8Du~|Dvhk?=KBotmk>&C zelEsy*ToH$+Z#h6bnP@%ADSWq-L%duV{2@|5H#Y6TN4@&LhUoqK?bbI7CYK;^9+Ld zv!+*Yea!d6Pj(F@O>6$l~;jDjGEx~8p@fXD$EE8VI`tySYYVlrA zFClArU=vuZON#>?JAlo<@J<$+#0cji8um!!VyGQhtF4}lHHXu6~9d)1X5xJc+eF!Mq#th9?b93S#S!=#a(E7Lu%TAiS%ziUgllSyxa z!3?s>Ai&XWW^RvOUon?w6F_kB2yC*1DyBTv0kFuS6yF;i?I!PpSbhgDML2-dUu@a# zEnuC>V3B}=;j%qEe>s<1|8~QEQWA&bVr4UW?bVz4oG2QfBny%!MTp$HuWa7;?Jang zYLwKD{VOLwAV<)jWl0b<(_psGEJPVkW9cu+q$WVC)t1=Bz+LVLKTPfXra%p_QG`~X zY@d^iFby=R8#w~fn$!Cz)gZ6sCe;gU^%bUrkKl+CJ2)xwTjp?P< zZ@fs~E^t0-v{{{?_M6c?l_DDKxNw61X z7$k2j+VJqwXVo#M=K!__f=$}ci+H0|lh7H+gTN&#rv*7z2G&0i|RTo((W=ttHX-K%ihc67Pkp zYAaWrQjhny*G9|8W7plYhbh-Fez9rQBl}Ga7w`tu5~ZQ(&_a}=-F(t8QF{o~!M50c zwyGXBD{9Pmbp<6(nUVGT)AXvMPtM^%x9fby&1kL2{?^h@@hCC|WC}`_ux-!c^!n}P z>~g43JwD^lwKC}EB`8r3wE96Rh{k68Fv~&SbpqD<@;_?`z9M-MH%vYs77>+Lo>Z>b zO_3auOs-mr_{xw^gljPV$cESsdP$XJAcR_5MjPu@JyCt#W_FX7@VQ@=*(viyzch$D zr{aHmsG>jPnr)gc_+Jd3iJSA7ThVo_NgnnTu*{-k-_lk%72lt&6KY*Kwr+axM90Fm zou-m>xU1=kl7}V!yL?bSLn+_N+Uk1K^i-guhj3n3P1M5NoflZzCd&Z*=lmpPWlI=H zBP}0XN|k6{d=`E{{2vAH+vbPq7VHRMk61K7A75d>B0byRXU z@o9{$UvQZ$ZdDk$2zzkY*9ybA^%@^~0xl4MEg<`s{L5P_ZcG}7EW_$ClP2jz!Mc?5 zy@WSENga~Z01pESIWNUr5oTz-R{iM+vir-KIcpxOvzL(V38hYwK2yu8Cy#t@A#fe) z;_o7sOEE+LLUM5dxx;a!-0Kp$eS)kxCC2Qx5&C#r}vN{KX z0g9dgdtyG$>DH5>eV`7lsCNri=*9heAxiMFl|*QvyHh3Ru3rSXBoJ@A2?&3^C%348 z@kJq!JwWWyg4d*d$E-~DNO0OFF30Vuq0N)_3#UUquJ^&Fiu;U&6ik&e0-V_n4N1g+ z27UHwa0)C6nU`e_>>lD%C2#JjUbWYPRqm0VVmp409<&1?r+%b2_0(`g{iBb_JCqRW z8JAGOtYJ@(nmk4kIYQaaE{sySldtzuqI&z_4Q2%j~ zHoL|1rtFKBqx*T=J;q=l&BBkDLo}C8?BBo{?Sa%yX0|IQg@6xm9S2x&9!tpc+j}B<% z^($n5!f_rrcr~~(fR)Q&E!F2bAP{%Nr+crdhhtv#L56VxOI5^FFA7Bve}Y1Te~I2OC&4jFa8tJaGeMfa_Po6lnNqpTT|(#gH$r{H_82c8`@* zixMghDyEQf5nJr7jYm)9U#8A9>g8mh<1Xs4JLa2@X21(=&uIFB>l3#2yFeYD4x(RA zI4o&EfL<*gYjdTHN zK0k+kU7@5e<`YVnF){2JcGMP1$+ z-a>kN-Wvx=b4#NKC=8xK%M|ypgIiW+clo%Yp&;xu;t9$2QOlH1__=aBHa)M*drK>n zZS@uJ!EVl3j`-PBH)N~@{^s1#3~QlFH9{XvT1sR@N{>W8L1BuYRQ^o zazIR9*r-^_U9{J7Au<83QHVn1J2BqS@Mpcx9@mJ8Aizv0#e&SOEoi%1p zxndFi4w_wx;0}ybLp5w0dlsx?lMf157&jBLe-La39^oNNTCci$e43O0`u#iC$|MGx z?$oB%LA2@Et^h{37Po#TS_uVtJYr*W&ANh9j2sMx>+wh8xidHsMDBpY&TuL~!zCr~ zF>a{#&Z2KRIOz99G*?&7ZLhbqWJjW1DFev!NknzZ7aTuQwB zT`S6N@WJnYf0yoagyslhG8%CQ)r$=Jd0{5JH75X{BJwUcu^g>|hby$m!0{unklQ3z zsP0|s)!skMSE%j=0eoZru4!U&V_1d)Hl)T;meal)^ zLaOzcF52t34#u_{P1A^e5&O-+I~Am8RSKd+aby1zJf#IqRKA4}qI^Cq{won>m4ruE zP|RWbb7q81FdC1)7R))^F{;}ZJ9M+&soM|S&@B)hS&-vi4$AkaiPE6(NT0RfpYS}~ z6X8>Km%nAelkU@$`i2(AL#UC_E`4KeQL)rdApUnc%Kd$o3#^#O72qE!mRM(u0jn>4 z-2R||jyKju5xTNzS;H6_)oHFrkt5w!+0R-jh`;smM2UGCsa`LQUA$$hL;(bZY*Ub0 zh|<_13}d(23;jqS?Kt`#6*D8XiJ{%J6ChwJi1vRzwe3~jWI(GfWjjH|1v9*=#lQGL z7x>Bj8Xghe6utvNTZf$;I=Q=me{1TPszLPH8+mL?}thBrqJ(<5vEvpK^W! z!`7zs4+(!l@cEkrif)E*Z{o)tr`e8-_f10CBf=5C`=9$OLcA^f&uf8&O~Fq)9-Qpo z__#j*S|tF`2t@ zbR?8(4{XiBrfXo6W?g-iO%_fe8vu~ihWT}P_blq98ZT68X#O^IV{6dz;KV#9l{KkX zc_^fsx196%$$E9+SkQ`!8!_Z44CpRTJ+7?mQxHtlav+PXtGhh#<&fGhgID5IJgG7q zWEHwQ$R$6wMilL#sd`(s=SUReB8$av;f&Fdq&I%IeiSa5m}Bp0=XpH@emVl=89eLZ z_pg$7=dX~{!I}ctT8QpDeVBRM-JRn|uOMYcGj8y~bKGCs1xuCPA=}0KB6g!6{jg$b zo-#`1viC6tmCV#4j5!>Ub)_AY)azMtO?mJaf}9Q(b)Es$hI=0^e@>H1dy(l4wP{|R zI^MZxFGy{3mbNQ2^Zu}~;Cwy)gIr2hF35c9Mynrx#ji-37GlOWE-m8nqbIQ@tFowS zIA}b0C}WFA&eo%8BGTvg1isv9$ga*A=N^_SFsWe0Y)zHIrK7_%$4-n18APU()-Z^{TZ$_RZstV~SC*Y7Oxo)SWiV7qnKjV4Hxl>9 zi2zAD4VmGYRSJZ`L;HJ3T~P6dM)5Ue!HNVq-OkL`OQ33%C^^u_&C$OUrMnR7UWGIr zE*eG|_=6?W45Str-wh20G2H5>y{^BKE9=0{0ShO8`jaKNL~OEV(oiXFjy!+d(g6b} zm~h9*=f1MNm+rL9?afm%qsZ>uR()0QdyA_;hjks&y8P^u0QOr6Cv#KrA7@zWOo?Y2W}UO%|7lAqV=BB_BlUWYj zzKj#AIbqzoIV<}8*D#merayRBp@4#)^H*(!NTN;Rm=Kc3FZK1y>WK>xk7cnnPG1!% zoKNWYOpwfBb_SK1IDweq&F)^>I|{=Z_rBbg8IPDNvUejqD@v+LoY>Se^*vPX{fC79 zTC$H$C0wHWc?CCuf8T%9X;p=ERh`2^&aIeg6-b0T>z$=>`9c~6w&pi3-g|lEHh2Vr z^3p;kgFvU_w|XcL#@!$HobFURQ&k3A_p^(EpHxFwcB))87aoG5nZKW8wNKo+JxUHl zTGb-?;L2HI{CyUf3D3q(u+Emxb#zzO{HP;NufPeaNwh)3vrqpTn$I^G@Q+o+lnr3B zH$3CJ>KWxj)c<>9(9VGT>&v*NN%e=*QCalT@~KgJU&5({E>Km;@x1+)*mnKb?;`DI zk5_Y3kMho=AHRet45$ksuLO05A3G+4IsmaQ)?oXUrCEI7WuR=fULffLV53Jjmk`|^ z)J3Eag~NAzs-&bCdCK(W)jurkwMj^xoBt7T4qo>k;?n0L+7=Y|-!F=PLt3M0w1wQj z&P(V8iPlIupfB7*#89y>ecb#}eJ=x|tge`JcGHFP_slTPhya2Hs})pu;|R1YPHXQE z&IiV$L?+M&MG{Cb+k5FNde`m#x&^=W2RN#D!KV0qY9-9D^Dq*4Wsfy-1moD1Q6qW zLia%+Ap~Xb65^)&RUkLbBBwAV;;e0iT5ufVB8+kudV3K&gPB0?5#2E zt12sJ{{=AR7#qMaC|U@~|HUn|7h7c2qNdM7%fM=7;rz1()OTNM2StSHd6)Y~Vz-yS z=d7Rxv91N?L9et%>0CHj)9c<1v>Os`^cv`PvJgU$Ia=?_ND2{M%xSlI!W&RXX8iu2MB-Hv<%EV8P9 zn^61F$hh^-P?NfyOtS>ujSZV@Hp`ZYqC_T|#X2U~vPZ}Y8vCnJYzVx+RW~6->_;lb zffQ>iBebw}|Mn~)cur^)eCu>Rw$M zZ(+XJF0Q%9`__icAt=sxFrzRMX9!&o4{JTY7xhdJbYhmL!$FY*Ld#@c*ZG^d9#0nP)iwEe{@;^ z6<(0$)gLC}P7QCiPAN$lYbglJccC7qmwl1wRBDJ^qm^!u&a}oltF82FG8EW?%xB;J znHjA-DLBHRVhEMFJwcX#AdB6LE<=h3<{ZL~n*;LOTm4}o)d*Y__3#qsJVIsg9*Mzo z$WzxN9(8AkFYL9G{0-Z&gi$^A#>XNE1xaltKZ8-dL%mM7cxkDlj6U7jziS^fo4Wwn zY*?hM??@QxSQ`}XrbdVW90qInn~vQIO!s`#`*PGlp)9D6wYb%6qaJ)FnD5US*HubA zi5shTZ6MDQ)8!rzn90{rU?YfDSvHRTD3=ogSQ1xe421`=4Og-dFHy6lb4a|)n1!1) zOgySa;DjqY><+3PUl3e;R^Wh4>X&=*-BFe1DmQE^CUh?%Diy3;)Sr-Ik|O1R>uXI= zKKMdChajp+2gXiQFS9_ZR8ieHtPH}$UEa;T<>=cn4gT;lNF^8Cwg?TrnU=0uzb8X%DGIm)|DFN?DYNFu~!a zAtFrxWM%RE76N2do>>cQ0iH7h*2w9M`?&Yx`srD03t8(2IISdLqcup`BjVUtQnug4 z>(YneVbhqlgw@TI77nSs2xJYC8#jK|ZmW~FAYl8cj8FTD{B3dKiW?1tC!aDuSg3`D z)NT2~i3%$3Qs(|J3MgS*Yc}omTkDXdoH+(JqXJ#Gu*$w;12UuQ9nzNZj@d>^mLr`` z3$JlJxed!p)C#e&T0kpy=2!0ooa%SOUhLbZ=Hd-XM&fgD9s6FnwB)8~Bi%VE)-!5k z(d~dPLa!PELfF^Xt#Vn$gEtT??(R7s3H&7=XI+%ygr2-dP zvytv=*O2Ac&%ey>EbKyy8uJe(Kml&uvNCi=E*i6oc>yJ(qlS#r>KyiVS1cUMJq;+K zLPnBq!rEVM8o(2V@ml%AK&6U`{s&V=3rux>w6}TEX@zW)(U;Gd#j`6u(%*$}+?j_6 z;qJ(T!C|*T=?+}hqCAH+vT^1o5y9bYyl+2;oED)}Up4ocPD-yz{%L2i3vJ(L_HEq`7%|nc+)X5QJrT7b3>&|+@6TmIPg1|S9=G! zVTWmIZY7}(Mv#-h8#N$7%bWERBHk177SAtE(#afGiP-}GsN0kQIV_=ye}(j{V%{$d z686xoPJ^^nL;O5bsO{-9% zJR`!^k5>z)zY-$NAXYey26UF(N0k;lq^j$kyugwWy+2EECr3T1VKz!1nd};z-c3|v zIjJS7S~Nzps2wu3t3s1>>~Bz>&JamGvtt~pPiu!YeY<|w+?L+s$s8A^rwGiXhwoiT zSE@H5LN)3f2!11aC#AEbdW5%y4Upcm~f0m{txpTRd>~&JMdH(EdnHtLgW$@FrL5UEQ zvt;IMDk}CnEt0%)bQ7&27l7&0%*#)xn$jt5xQ1O}sBo|f#FW8QVpR~-<2uicG%qv2 z6Cf5ei0gx;*D>XCL1 z>h?UxfZFz(>^W&mK~*%TAvlR^Qpg=b$)#dXY<1_G^=hn;%_La}mO_H;6j7NnY7XsX zJQYYglR2M8q{KxKpXe*T_P*Ax5ErB~0Y$euMAI`!zT7>9XDu0F+9`WnbhASe%goBHuz)Gy7(ewp``of$Jb#8*zEG8OahI112$XqfO+cwz zP1m-DU?AD8h#q>z1^{&@*TMv`RU*jDU~kdP<#esDle5dMYgVGt#3Wb;dUd-RCf zKlVT`Tu*;NyY8j-&lveczg}C=2xBX8;kcr>r2C91<7ZR)q`5@%HOIYz${xAvNr@VU zPHp%=Q_F6RYRv573C~hJCd!gCEsJPVp=3ND&#S5z^GigP2ZD?N0K=5*H+^4`tM%4n zbhLu8_StAas~jp?d{(uKAi|o}bFU!Ed{E&8DQ^fc5roFTEMGvk3YvXcny~NX*~fQ+ z;@7#IB8PA*WfHWAEy(?i%*K8)!`vYo<7H#RNqHQTkfu)$g+GPH2TQdH*3;mcW$OF~ zKqsdj^W%IwEESXC+wPH$pr)&)EEt@nly}fUgSelLC>3q976$?P)fywklyM8*{4&qS zBYm!S$gTzx_V@w%6LX?{x4~3%t70EE>Ci{4K9tKu-$?Z!1Zne;^Qz9BlV#_&sQao> z%#)>J)iiJLI?K}zE`&6$(TJ@H(Vs5^ElACvMXpO>#UC!Ao*tT0Fr8y2Em42lPbM8f zZt?q2;tpgwdeRPpz@S@06s!3`^p_ZXU5BYt`y*%lBfc-ut0cfZ!K%f={ZJZY4tmZ* z7y8w+@W!)FGq_=?dRnVR9c!FBI0_U_Cyb$S@@+>5 z?FlnZgy+8y2m_r;F1b#DWgaHV@b?3DdjvlNG@U-Rfv*p*+gwo7sC)EG4?q2;wSS2)XWh(=v`CQ%F`rlc$Ib>oHlI}LHs@*-4 zi@S9XZ(2HSyG7F-QF21>r@^C~1(M^FEBH*2^ib=UN+O=!YCr`!2+x8$YFO7ds$4bfZ2P~CON@m0%uG#leYS+bWD32ZgEs$Jc zMoE(Nr*Gm#>1HQ#?2exCV<{8;8|VA&e%D7%Ky3Km1VxVjA}BJj{C}7!NB-4oZ;2)L zKF;(^;ks{LK(eab3Z087Cd+6xnJ&fwS(!Ejf zw}UDBh}m#k1H=jxQl5{~^i0wGxDqnB5?_dJY7iXc7F48ape5|0d%86++wC2cOcyDJby~? z97Bx4X}wN$5W=;+DypF1HKvQx+LHgY`+Cgj{0VgAVnttF%bCoc@8HVFl#xQioOR8WL>!!LTSAJKZW0P&!UdlTYB+qgGC1+=Q@i1gey}u$-(o-I3!+DC% zcya&=x6nm7(OJQE*_Je}qIulH(Nk4rav^mrG%o%Xle)2xJyRFw%MqCSkQAr?g(1$c z84Y*w`{Z!^;*TUo)0T4|OB1z#D0P({DSOwKm*07L{Y|MO z4SD79M9FfUs$?w5zQyzsCxpc>8)p_KsF`s+rkFvMWjWe4wwJS@wB)cU_*5=8$Cs?2 zbfVO8nyKUl!Zm?cUZ@Xl)lC#<$^F|gW6)r1=}i!KlNlCFwWl*%5omX8E?&qnrcG5B z(o3n=raF@<0s(Xn3bm%&9O%d1ZN&}g=Hp^@m($aYzgb^l0!hOCEC{gc>*vi?8Dx88 zX&ryjxqm|Y_L<8zw1u-7Z@dLRMCXWw2*TufSf$v+JXc$qJ`Y=*Eov#+O27Ie=zI1b zZZ}d>z>n2AG0Gztqo$3xD%L_bPOM?}vuIVlz(5*Rjq3&G`a3vQv_R-7j=f(WMZk^m zwc6<48(0RYdsG_elz>N_0Ey(uKuauvlIDV_MLRd61NEcrN|sIXL zOj>Hymd(UKU5>%hSy*eeTY6^Aw1*ItPR`l)TF zMS8D))nd86-2?Wp3?}`-{@t*KK9$XI6o(RnOZl$HM}#%#eO-ThtGf_G3^CAPa*nQX zbnyAv?Lxbkj6Id~>;MK)j3??07*_S;O#Q`L!745D4qN^TYG`C|`{ zPX{M`-~WQIjU$QTk&dw-^&TuzLHT*rtHRhKPRdJ!LM(b^f@QDV5<%`UixA}8AKxyX zr6@>_LzJ=iOvuHmy*hh%zv>>h{X=i>kSX~OFc_6eT|ETCdw*GBi^m4)^IoC~WW1|M zJvom%Z8cp2FH`5#?<&)Nh7*q(Fcd+wgUs!fYGLTlLKBI|>Wl5}+928R9wFElDYmQw zpIWLYFdQP}fQ9S>os@F1qlzcPcF48e0oPyuxd1;~E*U|o>*el}QnI{VL^n3W58a(% z?Ero78OrMyNbWKsGf`rVcQoE0aKAze|9PghFfp}4MOyCuy`xWWPH^O}?Fl;sVcZKO z5gaM7 z<+>)g9`AemCszT8Yf-q62~!mZx_MRz|Mr(SHzY*bFn<0UY+}rOYVxpjr)dEsQDKJ7 zTzT@X$9*Ker{` ziNhu>EQeLp7E%Hmw&NDGyr4oHI(3tQhY;-QJ&>b$5;qF@hwKWCtlERo1^I)<(X%W( z2WMz39E?6lP>m3InKNuIIyXKzqpG%>b)h#p4Dj7B$&tRIk5*>S7v;0v-_N;$%o$-z zk7=j!Whld-p^?v>8f?YLnk}C4&R<`6X6l5yQt-goqcYAb%{77UL(x$Qz2{D~ff1O0 ztXnZ>n`KM`36Qcm$LUJ^)xS{cmKs(Iys8<;BO_w_C7Cg%dAE-sYK7#GaAS@#YeB!h zn-`mD2sBQoCEBRdmkTv84lx0-@U>;v%BYadwuPT>bo=QvYApgI+AHYJfL|X52MK=( z^U)@CplPfcB~gqh`)xZ|LYTh~s^lBt$7>ewwfToekB(UI31rs# z)H#N03%M`o{##|H6%sUNB55E*MlPgEQ#>+CcDhibX-TxSln$ii8cfy&Qpmkdrmh@4 z=oKEbPm-&0Rf}uXC;587Sw?kiYH*mqAEBK7SJwts70c4#&UO&Orx>=Mgj&%n&^pTR zKY-Ue%7=c^jb89CaG(e#T=A4e66&sqTeBw}aZ57=h<6etRm}K>mto*F$NuYHU&ZhH z6qW@taL@sXz+M1uSn{zvtIXD)pi2K>LNjd}P`MJnuw+Gjg=_ij;M<^?i>LdN6p zL@0no%?qxJUdt+?VHo_v7HazK1IY;f;4Oy|`8D^d^K|7t_&hh}tWiM}b(jC^z;SnN z`vDG-2JPrJbr`JR#!k&eV>X|#nz$or0llVfH#K>U_uFuZMw}u+P%(=L9jY2x)#2p9 zsikT(lW9NljBxZh6OPT&2hnd};WjLxVes1&i@xFKX}w_Hbw z3$hE{fM3;JxR7^!hLXrhIgayY@5|Pjt^OHRa*4uEL&Nbf1zbW95hk;Wvha$U z6+gG6NNy0GAZED%RNogWjC_na2vzQ|RlBJfgpydskzD=~4u6X!k^$vQ&1F?sDM4^d zAZNuo5t)*5n|tk+7@h8-c>WsRa;v|=Yz3<-I&;~g-=rf2Ch0Tk{8x-SwyLW2?*bcY zC*x{_!#g$l@^(<+>^7|)@(l2ZO4BK zducB`-id4!jU3R9C0F6iExPgznaB<$A{w1I6AHAge=aZvf^F-oS#S{Kx;Q zK#%(@PjwoOo8Db$HdF&`URDa2OTg#)b#Rrq82vxS&iS{~FYvRqZCg{@wryKe+qP}n zwrykT_FdbZ8av7UvdMln`~4T5Jh{0!xvz7u<&z@^IE>x9f46;6nPxl>N3;~yh$p*4 zOS*b{+`*5)=5j<_li{y^hHdq@1W<)|ul92C3Ja;(9#dff30K2Mg39T>dIeF# zD^n-p?4QziV}fMRSwuqCyzPm77(2;&0rS2WU@G{zbPitM%wozv306#o^Il_T>>(?7 zi_e*LQonhGoz~_Q>fenYk{$1gVJHAQ-|uRe#LON+56<9?PX3|npQBca`c38%Do za=t3L4f^skuM#B1bw%VHzIH*$NON$?qn84`x_Chi-wxf7l2b}KazV+u%J83DV&eoO zOJn&n0T8|Axi^S283w_uXwax&NsL!SYp8JFf)6H1S{Y!zlyg*pH`Cip17L`6KZOrX zi7nbj`&1AmIMTS#97uFMAC&7c)h{m=P+9TM+NB17DGdgVVo4_K-SJBYQ-*gLvwKXm zjZf)y6EZawWdsXRyrMT`EL`XO^wzY^+9-kkYDYI9-TgAyxtyr4l#{Q46gwzkzpSI? zrDrnwLPVn|U*Q&8GK_O>bzgz4Xs>N(%`Sp!N8;|;0PSG{McHLVV9i4AaV%!<^Qxob zx8FEnw@38*s&zj{i zoU+9^VM>&ISPxA=0EsArc*GE1L@9!r{s-@x0Rt{1?aW(ma4;9BJaEumL&3v6*7_?l zq4w}5xnTw0yU|aBSECrdxk5`TwgX$LiArn6DhU8Nsq;%w$ko(J*z%>_6-oPW!1==s zrIfvDha}w0p}$zOCU&0Ip$DR;g0TYQUmrJUoUdf)87Q6i@rJ0xSI6%i%A<%F1amH% z(&|?WP7f@>h3 z6Od{jjl*WBGpOZK-?An9itzN@EvN>=8)lJ87I3Qr0g6|H zt|czj0`ui5YLVI|1qecqUib?b8Q^UNTD_;u{^q+h@Ms%e>N!bz3ogHQDtPx(IK^fB zk<;KgtA5OWS;r^$vcgUOnE?Rqw6>0ZhtH#73B=!AwubaZu~W`L92{1?1TY5@^i@GZ zdTj-@m_1ol$HnfG_GMYa?EMpl^lqUJo8_|h7NukPCSfV8h7{p&g1p-fz8~!p~$vXkEfx}3G}?M)<;))LMq9` zOIeyd?DEn0ldzRO%MhJqWJ<R(Bhj>kum0m=y%{)2csMW3ufD;|UnOneuLq88U z&(BYoVvo*|IAlUxno6B2C4sPX;dsw^%)RCPJe53hBD?Eo(KrY}RG0h8h zy5C#9A@rOCX~~1N@ASKOf%l-j28Skp97)ZM@B3plUT67jTc&yma0OG|+__3Uu_$^f zJYji0DfJ;*y$JZW-{_C0)V{nb>LFCHEKH!&P;0IcSPs@*zw&i+@zqO5Jd6y;=axQh zsFj4xC=GMONC-}6%#!u%vhsR1cKI3R2E z1TFvb1k}fV3fjukPBsl^veMGA^oASc9sZ!y=XbTELx*d6s>){` zR?nQB5~v|GtdqX#va!61d<$}MWC$hu18a?}2LVh=rJdk`_@KU?n~x`dq7vUJ57YXU z0p=u0F9gVd&rLFOM$9K^39;F@V z9!4?#6s~=U`mQ(*_d$O``%J3+K8eLsq95bXDM3^uG7zHxHwR7slR(dE)zQS!%o(qU*jLPa)>70#6{^ZLcratduAwEbJ zF!PI*E9u?LQ{BvA+e~rbIsP;sNl>L;=dy1u7~$KoK^LZvo3cT35+rm*djK8w{ylG# zQ`qZpxSqxzMWbRb4dM2VQxq(B<=Ub8~s^mmnnD z>MdF;9oU2&Bl$pMv^NE>*>f0jS`1MT+dxDkBjPJEsrU?-gi~3rz)!upvfZ`m4KX4e zQ5)S%`xT*jq*LOTz)`$fsYBkc3x|9e8q0#@A(nfH{NBB~H3PxMm7K#R2bgFgJ9+L#X%d!s=i`YR#A0}Haeog=v27(jd z88AUxST81qE7tY4LC*#ES6PFg`%z$5I&*98X0UGjZ6cBY@0Qrt2!+kG-na8tX06Y$ zE{Zk!sX#YFWFr-l_uUtBr)`p+8%T?ELt z`!3-k_0Li^RDA2Ed!s7_nPK+|!4$2lLscsd0DjV{%{wJI7Q*J@kyt^L$yAjS#1-zr zd|8rge_vJj<(LO2ZvxyA$#zS|>6y18s$k$U>;wt@=5M1WjeLu1#?O&uyA0Cb#)!qn>QC}+{DclPUYP#a ze}w!pIf3+sa=x6XKg{s%Kny6mLl^@-rlGAj3GfrsGqVJD>a1dPtu&223fm zOYBLUM6yjY>?E`Nw#=ZZ#hjtRpvukMS)-HG+k;x) z^m=Mi4#g z)d9EDUAi^e@KaL|4CqQZ#(BkFqUM#*arw98ypmC}xVL$Ec!d2shrtOP5bI*MAKI(k zgqBBwD)v)q-D*(O`}-_bgNjS$;7;IV&L|zr_gp*R=c(1`*P99SU#fl=wTEktjCg!> zA7TRl_E}=)zhCETe?AR^fgHa-gmS{sXC}Dwbt=Q4#sLsmz@m04mdZHZRH^QQgHErP zsf6um1^vL|jOr2Mu{EPHAp`&{M&$LErf4187cJA*|2_ zVc<3BA{MbqoUVxy-iwJKIGW!cTTnBpnoB39P<-y0%M2hPzLvmyKxc4>u!k@U2TiL78Q7tW@pB$IM-B6QdZ_mE_^hsKX zupQBgaF>*meFa&W67jwfB$+~#YaEB0_i3yjM7#*ueK%HCeYK(Q0Qke0vW%k4ahN3e z7VZ7LEyBZZzsU0=_>{7_yI=xGCMTg~?KGbzjGMYhH%EPbdp<1M^HIcs(i33VqBHhAHGEVvTyun_m^`W?~v zq}6)64wGOU4^Fg8T!m)IDGUNy4Xpm-V(;6dj=(3PT$N$^PfAyNx>*yRDYTRSoe zt_tplmcMm$G#?*Do|UAZOP0FI3N%R@woUPVF!xE*b^t< z)J`K}Ud^$^DZUL{4|xER3vemytcNZi9w`kOO|1u%&8sobEdzz0ZiDgW~z9g+?D^Jg!|~Cy;PH=3dGFtKU}NA zER^J|NtWK-@F9@NHAJG{BV_^$b?2qKe>nRE9IM+nklgi0;Q6hAZe90eE%@rkCW-mc z#KuzH@a`{j!5<`m)0(KKk7f~s50c6l3auGdL8lr*A;0`PN^wdKbDVSA$K(K?_T_p* z4{lm_6zp!F`37a9MZVbbiOYUL*Iz}s(1O>efBoPts(2I{j<@FU`vs$>pp!^L5D;ML z*IbRkGvdmeD9zR@|A?Bf7H3j?#R^7Lia)L}&d3q2&gv@-{|RDK0Kj^T0L)G^D%XoN zZQG;hdagATltUh?q9RI<$co4VND~ zalqnD4(x)|NjFS_@tIWY%yXm$EyLa*vcVwZ?X1!vg|VLUTe)Ryj$iF6^ep3ZW0s1N zxTFt@TbVDa{W5N0EQ}Hbzblv0NRX53V?yp9nBoEk#imW0NyS@FB;bufs2|-nS3Y&DMHX&&tNs3YtSF48XLVd*}BQdHuF81KA zEMTjylRb;CWRYA#=-i$aI!Pqr;qfkx*3ufwIyw4Y@w>@5zs;6I_k}>shHJPU1PV|T zj+YIU>GUKh^an52eBl#dfp9>w<2AI@)Krs!iu`IEhh3`lv+m#yX%~vnBxp)v$^F?bkC=-Vb5sG5vRvZhOrEw9OO=iC zH!Aj4K$AC5{?S^JgJRiGF#av(nD5k+bdpK|F9P1rV8N!?>xT#urX z&(QhvCKVH&c8ci!{bC9>p}wugn1zpOcw&7oUKuPsj~JZ2f7<>aH{)X24eYmIO=Z|b z+;?KmI!2SdJFsItbiiAbwT`YCdCJidlGkoqOv?yA5Sob~n@yEaYf7c@`+b}Px~_HF z3s+e+KhbZy7mirY05enwEipN08^{TE9fj$QBacY7t2T!Hn!zAQHVY+^c{6(mtt;}t zUni!&Muk+Q`jZa;8-~ln9<)0{J;u~P#kB2~JSxOl>>oOR2|EyVb+PT6C+}^Q_FCq_ z)um0y=G8Hy=_mAv0mY;NwG5XOZx9z75h+|^?7jVV=gxm!fKFqg^jaEV31ON?;RtiM zMY%Uu-*{SpO5`j9;V#|w^n+&L+r+Oh9`y)^t6wt3s>l=BikB?=Yy~Yhc`08l zUfXdo;dHX*-mn^P0_CSgrM>!f{eh_yH&QULMfJ^Pr0Wyx$L57^ozLC=?O zHRXaPq842j-8exHthZTvnH@ZL5LkS|)GRLB2+~Bo%{J9#ZFNid5Jk+}q|8^GGBDFM z~^-dzDQY5ZCbQ_H=I=G&qq@eQ}@3fpQ+3J+Z<|z&Trtx}+s5N0u z>xwHugfR|7qH?}GD~RXD40q-1g0)`!b6`m#2vaq>$QlsY526@!>~E7F++7fb%W#4&VHii{E#${7!7SmBA*|xJ z>zB?-aBilMfARISOuV9HvUhcteqHjT@2oChNNrE(-i2YUx+yF5F|8L%FqPYNDap}! z3vbk`03m;TWe4oa@XiQSB$#cpkI~-A#QiG;P3OIWbTku}+e*26#Pb8s^wa$7G;09S zn*i-pjAVJ@@wi7D?V4qO%T2C|>ZjKB?}#$P*)!>CaR$v!D9zZuPCz;@ zgINHDvW*P>QKQl(hBJVhS18o@#`R~_0(M>NvW)n(m{-uiNq93D-(&?hoqMK%gFg0e zqiyKg2hK`EN6>lmXhZT0A&-+dB<-ImVCb{U?81CH|X`9&le0gf)-z z?9ndg-0$V+;Q`8P?VAfq&QsHY54%uYXRgLHI70LOlu&!R7XIbvmrMHO?PkoY$Y;gw z9;B68Q!o(}OTOG^aBv#@1F9aiTlydPjEnKV@EI4=|KCP<{qF(Y|F#hxir+!mWiVcZ zRz1iind$8K$t<^x=Yf(9l8q4+LI`zV`*$OMK(e27h0jW$Ln)?g?yz{TM_lu>hI$nw z{PoTJSH%9|o>XRTx&m{cxB04)K`#g2x;^G$YxD&68RPd#LoO~+>B_O7l zD_@e~YV1@zS!5`{VY#Jmc0JaW*|o^jnd*Zlz<^>XlLZD7Oj3{51O6d@gPN}aCCsQ zw8ZHwPRr*j>w_s_@_RC;@cg5Xju#D2j6>JaxITq;2-a4878gx9(J-9VYU$bOHc+pa z44!QHxj>VN`HeoZ>cT2-mN}NnwK%x^-5kDe36uIiV?>-j*%yO(+anvE@CJHE;jn@T zg)NeMtgXIhq=1lUfzs?zd`=BDWIj>({XBrF3|WV#@LAT0>lD_j^JV%v(oA4WC|{& z^U_w;v&XuGZ?i{=%QNwj6#_dK;Oe!C3X#Q zOdo-_lh03gbwtb)_c=PKVvoKhyF8OU@#J$yl~$>5P zR1aU-I;zzw=nEfdKmkFcAo+2J9UN9lR2Hc%VN5^J#5NHiqjU|Q`dVv$gMzEJ56qeg z0g<>i)6-utKY`ovQUCi-&}}SSvUJ!mh)Y~*3|$^c1`ir0y6At~D4X&$MYbhn+*fG3 z+qyOqhsq#=V>QxX9G8WbT{!tGKJ%}#!+fv~SV!;y>c!P?ZJt$dz%pHdin_}nzPXxH z155Coi#jLxDJ~A5hz;Kf)-$wewU7nb#C||=tCwphAeJRX&f^x%mv-~cpqeE_lRKZt zAjg8Cig{0~Dhd7tnVl^XK1!5tCIo@M0fURf<&^%Ops>OIrt+}P(vGDvlW)_Th^5_zj5 z5J^xmYV&n4CjDbU2PgT=pNV>n5s@86owttu7JL|Bj#~GnYcaD-WbMw2m>UT-ripq{ z_Oegf<6FrjpV~e3=%KScpXAh7GL%a<_UNh_065&*60l}C08Mq4`VZDJBoJPua`28I z6ZyDDr83|8`fjXfv%vSi7-u6@O?hN#L^u|#iycV8JgTZzL_b@2Cw1HB;tr_4=AF#{8wlkt!OpK)-W zyQ?iatVrjLj}vq{3dDpxU7r{HEXzeWCP>HP$gjcP)urpM1!&M8z)3R1$!D!Db`Mrq ze1ImhYGomsBE#LmRe0*?&5}%}oF=`FzzdFJXE-Ssz8F&=%x38#gsObGy zw|UrHb)DkhMCnGUa6O|h40bzLHNB$f&EF`Rzc7#>yIGPg-#^VyjNGARqvZdBsiRr! zLONnj$`Gij92Ef<6lM%W9ba4u4R3@UlifHIG4>zql%*#xVge#~<-zOIOfuhCX@e{0 zpG)psC)q(d=$%*fk%o65Uz|_l}`YO0~ zZM~4_q?*t|=!s4qw%})YJ|!hUkH)Deqp}L{7B4oK9Hej!=80Y@wYb|%ANCC{&mtK_ zHi_iPEU^a(!yAKgQqz6WY$#e2&tS;Njkt#`<;zH0yDRFtD2q}?5u#DTzFyBRqPb^~ zazsMK{H{&bh;Fh?o2{~k;PjA?vsGIC*f&?X<8>c^pFSV6-*3AJ5NCP~>~!9o&d-v= zCD7cIm#=VwnGa|@HWwo|r0~z)=VBLbO3LBJxTakA;fu);$+{-}rDp!to~8`o?u({` zT@jE?u#R~516LU?29l;|8>y}ck-T#xWeHmTn&GFMm7C)R4iQ;)YdJ)l1^C-@UJaqA z)w&&vXc_mR#rSLpbI!sGO_f$~x%S&>C=AuKjqn*O$L~V%Gx&7V0eXTn=tYm5 zD#duNoBzCDi7u%H`k_r2x%8e3GfEgH^o`FOTjk2_52#YgPeuuFwGm-M z676G&otNIjW!Wa}u%*iavF|`T!BQizSEt+>ZLnX^ovI?8Y|@i2`(5*+923LZmIlUG zlO0Xg>|5aguT5u+8hhE}`Tc&gh9h&VAd9yO>iVzL*r!AbOtdNkZp2jvcEdA%_5w$( zk*?lSGVy!8Pr=r`=%24B%?OcUV}DsQq6C@C&sa{fW?)2veV2l0ay7%qgaOQgF|oxk zNPaTxZJ4aK+g}k{_b=FL%VcT1KvV-Js7R70ovo#&V&O5{-Om^_>iq^lIDt_7aq)1DC7S=u`WH>Y+wx9_@tS6@_F~ z9=pw1Yqci@L()Sl&A>#?6w<}}D5kQxso!@)m@_Pq-P00Ik~l`@LwoUTQqh?e(w(9) zFrSbY)4?JQTl4)$mL~Rj$cA=*LObW`P$^pu`|qG`(l#5!v|9LHxOsQR!CYWE2^od7 zn2dqx#IDVRo>&^Qo!#?&L-}n+3-ljVVZD>VhMliSt7w%sBi4VIOOxeRm|U`^E4QcD zrId9SJRE==_idE+H7u!t&J;^baS762*@!n&Y*$D80NL?}HUK?Bw*6wJoR%+}v}GXu z49Q))>Dm31OlF!QlbrK20WN`w6gmEv--zGolv5Hiym&nTtw;Q#OZ`?ToiyPr6ps6f zv+1?H3fhLTquAc6wqH>X$e?YZM@^wU#@JE7X)@m7f?M{N*7*6anu{1{A@LUnN#me< z0`v48Q)1;0HClnn{f_Q%X(ql?EKx^L=Y;4t##$};z<%DrP=Zt(q~mnxjsUds10+u_*)d@69hX7)U)!8)QSBU4PbttQ8&pjn_iyeRW zz00Os>Y|H|{vQ;u3WnA3LCSyK2!=y)AbLiJiq?o|ohF6_e~)WpG#S3m)!%VfJ*>%4 z5sWWH$@yc{QSg^zm+0ZX+kAcm{cvbl5xuTIbTF*98a6IR3A-;o|1RW%2`FiHz4*8f zqaXa==SrWTYuH6}tT}o3y?|l*f@Rr(N7-4Y>K)pjbo4iJitd7c!xu~PFf~8o7SDo# z?9-uQ`%mX%uazP3qhib}n!_2g5JIG96Ok`nzYiE65)DYWF74aelMLEXDjbf_BBIQz z5}_~?IpJu*$IRv|5}(j}bdP_oHxIIKHtPpdpMM8YA5>26EjFfi&|!z5_TzD8jurlo z<=HBR&!zn^Zk|v&P+#6tTqeK7@Qw{Wq~T~>G?j`T5-OG@{*wDq6if+=pa#s@tMsb6=yyU@0YkQFV!9`PF!5}T z#>LTPew)yEp4&Z=Vw0}cg)5iauwDAHdlwMr*PQ8G7hNkS7{4>4;@`JZU3DkNZD`w} zb^!Q2C;IquF)G+$hg4qc4;E)y5%11JrZ`wy%c8$5Rqlj^+l^#sYeamT57aG21EQD` z=a|+rj9|KI%aDl-|1tvjA3-wpr!m+_Ha=d>^<_DUWw;|)>q<_sSSBT9)2c0`%FYlg zw^09zmpCLZp!dqHVr%HOVzWu4GB)9^yq6zUM;?LYWyjWf;-ffgPNb6ivtToUyD8|^ z`jZ5yDqNElZ{)EYRB#00q;NvDtpEyjHGz|tqEyiQ)>@4|PJ!&CC_$gV1m_Q0s9B`j z@cl@A@XDdtXoMuY7;+@Mni2tfrRIco$-^VUu2ENk&C2-bMI7qnmy`vo8RIX(ho=3& zh0b~RDGn24*)@@N%zm$H87(4p(v`Vhs5=BBH$atxw`u+6J!X&3Xe!!dXlwUp=#VfF z(ZgGTf4sc)%z>-0HMu)7!^ZYGJsNG5M#ia6Y+!uH5HVU;T&+RD<1Z;na4x;X0n`XZ zZPR|i;XxgPM1!-RtKgyG3VN`S^W2O(S0=@dLUVZ7^H&ZW0J8_GnEbNIQ8f9TQ0j*h zK*y3S&HBEa_@*vvcQE!Va;SLW8_eUjGes)bML?fG;6uBm5$%k#^n@9SNP^OIMJTA} zUlBmzp8d%V9gnqOQ+<#S;bs%=Mp4^U!DB-M_Sa(UsxVnZ>V>(@GLL5Jzl8Hswx=7W zI(aI)@L=QfrU$bXSz^f z{^&q-J?mU(eCS@@#&gj*F4~y5>|Ixpb|^uEd+>0O{TA0)W1=()P)Qdql9fSU=%$|w z-JesuJSGgBCPW)J%v$oDJ?0?PVTs}4eY%xR>pP!5-`$ZG+8#%nQP#8oi4Ar6BRE$> z`myhnu`cGavg59ZFGI2M^R@&Zu>J7tSH^gPt)pK6SFt$OQymL%%%jQ!fAr}o^E>J6 zPrPvpRT|yaRdNlgfmON#)rUE%YUXtA-|o4Au;70Ha7dZloWW3oN~^R-t-Ncme^Ah` zT;?o*yTeI4&EJrvom3eS$ho9eIb&ZqL?gaAVJ{RHC^qzcZ5m%$dUnjvJ-2WN4Sw^# zA`z8{zCQk`j^n%}>>2eHn2Cf28uzcopY$Of?fs>rNJV?D6x>k>!rX<;-gGSR3&ix< zLRrUi?WY3NAo)c6@dvEwtn2dcFeKs$K$#u$+^pG0@xBmU|ERLA$~X)gq;8Bn`~5&j zCrs4w%f$W9RbDD!u`Jk{4vW8Ft)Xv&6P>npxm02&tKZKms5-gunJL9pQX7 zGqxs(*_mqK1(O&Sc*-G&2x5|g8io8Ud)}F9?~Z8{K4}N}cehTy33GLbe!PIEOWMz7 z-T62M1cW_lF6T&=p&9x*gilPLv`$e@V)H*8qSk!fnuHC_?H2R%aVMC2!xC-}u}4yA z*syzgH3;jC^+bP5!lYW-DxcVP8=dHF4wxy|k3YMk?QhsDP^Vj=Qa8Za%TejX#w=hn zOMba$Nz0o=qBT$6-mbSlhNdL-(y>2+yQVaj(R`Hk76*~7Uh(mC_1HusK8nHAFF6`% zX1B1Z$U4iotk8xXUNcGL7vx;2ebz5Qh|Ei|0_p39a|D*k?-W07V zaCp`8)BkOvrf3DhV)jqMS{m1z)AHG4HGN3|jXQm83IK_bGNsNM%zj*pw$B%eMi`qy z=NXdFWp-PtF}fwI(JW*AeNmJboXACqeu+8m3HcE+%Mh{@$?DYxU=+FbY4*?HIa`7r z-}KSjo-52$br{cj4u2E2x)N-vE*r%JM|G{~|8*zUTx`2n^Vp{w7A4<&DyE^orV1C# zTGJV2b~T$L4yE7;yQeBc5&uEa z1u-w?8<-GEPO>yLuQh*+s_tdIAB_>s$bx-#p|AI=KI;Ly2JEzwr-$=vS_(vOsRryY z@~CnmCI=3Hth;^b4CBKt=l=e*1p-@vnrGgCsAbHI{0zLk_?g^AS~n*UD6s2103?k0 zI3tLdh3==rGHQ3>lu2r@c^j6OZ)p`!u*shm(A~%Pr0b0plWx(l;sY!m)0a|uWZyTn zeBUgvMxKC?5iCiUsQzP6h-BSQ6?i+grpDywzHR4ET~&2f;dl{;kGbmFsW*PJz2bVv zYk=1v(C^Fi^Ob&QR<|T7&BL-u0W*>B^Hz4tF^_;a-KLxB*;Gq!MaQb>ki?KYeQ*~e zO~3hAL>GP2Kz&JgU=Jmp$sS-$7v}@a4~09_Y2FYuY3}-$iwGzpU78>NDqz+wOg_gc zepZ+q#K&%&HEw}Ou(qMg&a0`%VsHg4->0T&#QaX~Y3|xg&5ff(al&FCsr&AV%$p+@ zA?`5|e*K(cP727z3LLjUw#cE2kum{=v)Vwn9gCQMxohETk0KUSB@T4^9UC z>Q@Zw)t^Z>DZ>!Wui?9u0(r9SKuS8Onsv{__Cn@?c{xknOP8fCHE51=L)f37CN=Kr z(x}!^1Tm39f|Y0QuvI!J6-pmhE_5W6GrU&Uim!T2EuimiUpb3N#xz9Gx0esl4J+)G z>y~3mO|5dgPy2Zg5NAz3IL|)LFGfl!TceFCx_uVMm=x~GfG2m|iynKS9)rC?SAYZx zPqBpH5UCHED2Bb!$EJBSC@|Qcg6nR@43LL4PuI0n)w4tsk2WmlRE3kuLi2~=?0 zYCx9TXsh?E>4!O;da7#0#RQlh8x|Ph0SVU}_rd_lxP$wL=SeQ%%7F3R_T3YJj!+wy zs!ve#6@dkN7b`T}8+nkZ#2RmgXqdVGFjvU4)<8)C)o~5;nU88dRRQlJSRWA1^^gWb zmFTlL6n=~=-KJCO<^sHND3t)4fY=;&75_HhrzAG)c9#;{SgWC?Jr6P^ zPm*_Zk=>ebOYg#NZN2B}8bU(v0jg=j1{bPW3l$=Lo*)g*6EFrBBR(%B4+8 zj-ijamAe94K)gbuQ!X%nHYgbbRm2*rx>o{5(O3+{0`ggu?7K`Raw#Yq{`K2!BaN+9 zHt>WE!vh{PLlk5e28V8x0H=?BwHYyAv zp{BN}5Ni-lSDGvkSg2q7ao}LbH@aL4BHF!UdqY3^B|8R9>DaV~IXP^9{_1ow;}Y7U zNL8Ag97A6o=jDP5?I5LRk+ce7I58fVcP7tbxGDd;1shyNg0<2`I{5;!p)>Q&b)Tld zO22&+zt7Q)XAxP0lbh5ka!t;^`OYok`V+%=xqLB%ZX(CK#mPH4S#kw1^7>}KdC=~j zM_X&GYDhKhkonSxW9_n()FPY%m*eBY9cW#}Nd@DYXZk>lGFqTpQbXfEq%5S^X03_JTJulG)rpQGf=5KKQDrnvyAXFyV^ zj4Zg&b@n(bw|ooum!zezDK=-F&AmCo)}b~4dPlwXbF(7Me*g+Wls zAPP&~ENhn!**Wp>=o_(lYv9iqULae(P&0@mV)9q=jjS7#S^QQn22e`C6F&k0|%v-zPy(VZY_>&mwJ2M&W=U))Va)A%F( zKJ<7mP8Tv=q|9}BYqW;2F@)qs zJ_K#D1$CAYfAJxRe-wH&@apeKMP5` zMT8I`kB|DFHjr;JZ?ensCK4<6ok1_S^0s?T>^Dh{Jb&o`E5T_tSYTa-Dzxv|RH4}F z8>wr16trdux`XDQ+L{u6f@t4I>eav%Yp3cVgf-`rCVZ97`F&|gIHWFL6AglWm^(Eg z2NEhC_DPGFgEhQ`NH@T3O*HvUGqNRp3M3jQH1_4PU<`?6AT=EXHbLHKyaGv^l*WnQjGj=+kA+sFw z|1-7TH%2VR8zHeVw5t(JpjQ(WzV`@2B~WSH3WS-(Ma?E(N9f#bs({JMQMG}I<+fwB zik?Kh;nTY$DG73H7{gel%>S+BgQGEw=z9_=N8v}7ea$Yzqsc1rKe?ANBJk_7Vk!;X zfUh=0uQI-#R>Vm6OUQ7m>*lXdAElhppkVjrtrEoMCYzs7jl}(DXI&q%emZutU#9Ys z@Di30;RXJUdBVqH7@T>cKtx}PBGZP*O!jraEDGFd&Y#Ld@=J5XXk3K`JnYkS2|IXq z7)yGGL zPm=L1zcCU%f~K#a?S`k%-1aqipa1j{YvKlA0TV1V zBP4(maD2I3eA=?$I=zCsk|P(5?LZE zKq)JFr6w=$9kZbcAirJ#Ma$cj^b)}V7e(F=DPg@T>#_5a-TzfNj659$Bc+hYoJOFy ztN1)+#9VqQXtZ)n8B!~~k|D)S(huz0I>pf&Kct+$Fd#40emM3akpg3zsBz3XDD93^ zfoAEhMms$R>Hb;?YL)oING2JL@yM2P2Z*OBqBkbE`uy@y+4C^qa7!AT+CrDMM{gQ< zJws$Q>RHq<;&dq!77L*4uChSJ!%JI0Kp+U`@Izyk66@AD4;V`rk^}HV^n$%)T95`U z$SzXjX>7TA$dE1>Qxk^3FwydXNgV+OwcZ+P z9c*~_NW4BRQ;HK(sSN(C@Su>Lbb+6}}E1N4MM1C?r_5Itbef-YHW8%kucU7v3f+-fwffA3rcBO^5 ze66xv^Mr3Ub$^V}3~SbeBLTO3ogPdKx)Uz9qhM6P9x$Rt|6&J%KDcscfEiSjSfmOz z1iNICj_G&KEVV*QW-OlXoB{t#Eto*^crY&m<_Wn)n%aU)?>u^9#)S4DcXt_sCa8H& z_FF)+$6cZjOhNjPXFj|3H?WV6R~l_7ijVYjh_eXS>K4$8B^*slN=Dxg;UxR)CD! z!5jU1;%+J)qBEZvheFhCxNl6^e=$k={Nmp_q)?pm7oGwE=2`m%|9%eZU$uQiM&_d( zuXD&4r1H(#`;D8CMKOAMr(?>Eu&K;vjJyg$LBJ)4g}qYJ;M?^LOm{TSS1RmDkHVN}(g z2PDf+OCq)roGo0~qcXskM}v8uHK-d$>inN5%L3Bi$r*gvcyzm6Q}AXZ5G&XilQEB9 z`ak+!*x=u8NYiNH77E{JMcg(GSS`nNXf$4{=MNL5mSN%KU~pQVUxMO>kIzVQCSiFE z-1E% z&MJwf&&u-tUcA4qE^bV&r%Q%2N|Y4h8~~!3{6Jx=O+fEPM`syThxO4Lq`5>aK5#BD>z1bqRhQDj4=U)w(M7%tL!*hrepG|>z z|I-(W8vk`l33nB%-Yw$}Z`#)@n>PF2@H7KUCCyQQ;c*gomnBr;>bZ^qAQ(-c{~op4 zg6o z!yE56fHA=$A37uO~#z!E_;{GXxF#ur#kPHgH|#hE8b5Sa{GnYKcd zYFQ%Y2_{vjv(pL89#Ui)84Tu3J_AZ04e~UsbCXO+<;E09Bbg4i7aHtxDOad*sW>NP zr(k7DFQ0v$cCGwW8dXxs!>5@uYUqWqSs}9)G_(NQj@TnTVn^A1nvA>E+a7%$-_>JYb9>k6x0E_uqYXl)Cg}9MPPxj zRiwiKSSOE-@;cS{>>G;@iw-_&57oIuke~Xha1)2U^SUG10ax4AtWzqxjr){$!29Po z%eVHXlGvr?Awl4v^zy1A+zIKskV^Y;1n?|dytafP{_FvZAAWVRj zhxhjXV(grn1Pj(}TQauOywr$(C-DTUhZQHhO+=}~jBF>ID`v?W;sZP zt0B*vXB;d{5kzWJEO>c(evHS4l=m^mmNmdff>7srkeZf*@3*}MUoGb*L2ze@I$fGE zs><5sIhd)JSwO5qzdsacC0LUeE|rsYFzYfI;IJ)=q%1&B@98fRRB-Ei>=)d}Xifs8 zZuUk# zF2r~(_Gmyc%p^T!@@GYQg;X(^p*5_t$9t@qrH=3&u?$&FqWX00V#8{79?)-eoJ#&~E`Bf6-ClNcbR?VY92wBqdM&JPV6>ZU z;T*Hhuy74VXY_eFknVE04AZWM>(Wy4ta4i`GfXS~1BI$s)L{OV$u{M1O0NC%wZI-_ z@LRXf9dLZsWX(98$|@B@7DtAhJYRhjlLHOQzSyd6!W)U9CK4k{uFUXeuF(VN46j1Q zlvPqx1CWqvHJ_zs>O7}w>lL=(o6f zTCmo$68t!qP4sS%Vk?q~C>5;xn@g>>3$pWhQcT9s5lDnGcff|=i}Nm;+RRF!78NX~ zQb{VJ8D?1^uEhG0B(gQ7REhzBE^I_NGqtY1RpuQIaKvy;_cH~;p|gbm5zu$pT&vNj zGF0~hcA%uK;o3tFO?(x)uo`^@_SGt|L+zDEbCXT{_<&R z{~I}TdewNtEpfH&!ZPO8K)G1v=mn1MrMv#J+8#KnbCwx~l^M5uM3*wj7#@t32ML(a zY-8)@srPhnMfP=gedO|fEVR*eax&u;Mfd?^moJI@Pk&~{|KiWg{NEUYaAJ2P74@-) zcUG)N*ku<2v{z%VqM?w7rfjVA{c*xRku#eEv7B6%*Vr>8sZf|#TH4q@{i!y= z{K>>x^;VXI(X`ise+0Ljk#-zC0@vH!U%htQ+;Al_ae@@i69^LbShVV<k5Tlvyviz-pvb4pr2`wxmUyyS`oim8Y&Xka)qEe z33Bc#xNUA@OD$oy_o=>FJpM@1T2?=7h^{Q|y-G%+llJ-XCUBPy)PBkk+(;V)L|mvNGXj ztCl6`S}23QYKmEa2#YiL6GapH9Xdxs1|+`t+)JB})e#}Ux0o<2R)YB+Y~ngLiBr5E z*E)8w7#nyWc7)I8(I0iIm~Ss{yHAy}0?xsgJtf#JTy{CnRJrnouSfHVe|u4%G}iD%?d3p zjHw}|!Wow63X>LQiG<0%S@k{|gnqv&hn`?3(z-^@|4cs()X|Sxuer!sJ-7<0lYAa0he-d4(nE%xz8rXI0i(R_qdTVYV-IlAgY7TMfhwyZotma@k-Mk&FF&++AFRM2ec5^1v% zCqsUUEWQ_lO=7}C64+zqZCKg8e(%^=`CVl-Ze7!=9eh2F35cf=;Y~(!ljW`l|K#z4 z%p{;=UhLa(ub08^SX)*h%v^@P+9Q3>-Gz_N8b3W1PPdOhdHuwn$a8qNyIu#G#xN$~ z?6JrtT_eP?*DfB$)Hb)P9&TUlsjf=OhW#6z{Z!Pl-cYOeFkur-=MopEDIoUpUzGj> zhV4sL-=)4~FW|&tvOx+ubFb64B-TfJD)kU#6^HVPs-&;k#72-c%K%H+HqM~1RMStW zb(m31qD;J%GMBG5Db!dRF)Qm)1$;ns;Km$f-$bFZAOSm(WL~qhQzhYGvCfe7S`IhE z;i0U0Mfxs&P=-jbjMMkGOSTX|(aAz=Rx6?1(U$<6vg8J)w&(_b7s}D$k6)b0O4YZe zMB6dYrIbI&P1y%%6;e_V+&;tej2hw2l#}gw)nH-fLDV zIy<%7_=cnyYS0$@On`ZsyO{3-Uz)z5)6%m+zw#$=4-#^5D$&0e&qN+i-ovj?mB;jAI z3}qM(h;i(ft3ApH{A`06m89MQ>F;b^p-E3+w?3hN!2pbscy;LT^9UYd+4LGD-q zoDTVACdR;3l9{MA7Ps}L-SAF$pVrz9gW^3r&3lnyEPw)t4^$J*r-loq)GV$S^RaMd z1^+51e0%#C8)kvvd?+>%NSZaGvM7*{^)3uqb47pib%qcGP67DB)QkqxL8Qn^6lK3* zgaK8*iOBk`x^@SAa;{xyH2v(~-o4X$GoQ78Ii&o$y~ZX3{<_^h&g+5CYZplOFBlKo zu#6elyLFnbOPye*^hOALIXMzRg4qrTbZnTAtKrSYE8uSM0kSqg8ge_%OT`wE*+r9M zN{;`oEN7(W95DUKGhE07Kh+Wn3QRfhpa)F8K@IGmJrbG;z$UEl;1I(YZR8JGK&lyp z(753a%7YAS(UMYoIT`Ttg~L!w28cnZ$5(arJI9Olg-<$oND^#Bb@+54GCCtY39RZl zSlZmNV@l#v>v9lp1^2nXv$D+GORaYZ{>+x+1LEVreAyowiTrtuN!mEZhW|=ilj!G( zAey)D4&X6ixJ?g&{|LSj{5K9)a*P43?En_jzUPaJ@NyIvS1K~Lp$vW z`_P8ZED|PPZ*IWRZ!%8971M0wVi3g5!~l;868jf_jOR6(t{bw%NR&bddyPm2a9Rv- z`N9o~n33Ofv&3eZ^7Yx>1|p##g(eZeW>cQY5`;75G(&bfh-tNe&{x4|&cKxsLCWGP z>L8-ap3{4u1yztLW((6&&@6mwI`|NS?9QHX z1gV+D?TcMHEI6WL=wy`gn}PN>$HH3pCiKctrhfO;FqH^TV9m06AUE(6l{U#zla z_@MMbD|TDB2?EZB0Pz??29+ns+}l0k$vVJHsR4xKyb!nZ42KmZYgKA8seXOO09i)J zpbARNb)+uc-bOzOJ4(qP?Wnym(*81IG5-(wc>sY#3-Ud3JsVvBpuq!ks|0ZV55&Ru(*txDz{#6T zOAMb%b7$|(HoNfn7%wyr%rUnITU4e#76D#a6#%EE$NY=;ED&3B&{(fSg<`!$()ss0 zG2cS&siNal4#7+YLtT(`6?9UZWr%MUuK~ica1;Kh3SjuPi7CK6m+I%#qMHSPcz3Is zFA8$Sh7FKS*`cXG6^kSpc5L|6VC5#(KB>gVYXInaF$8QlyG47KQ#=fX2kYxaj_*Jd zrVBIJSHa87ICpxM;LH2n>D(rq=Z`ibfWtp?242M_S}HLXqu56E6{dMnD~*ZS(uGL0e2g-vg-FC>f-O9F2 z3P$QI6Yyvx7;^D=ZKF-?kxJ%uBro3wbfV>$OsS~W5B`J7;p}sR^dro8K`6(;NmGQ5 zC;ZHwCucCJh3q{V8^5?U>y9GPqx3X^N1``^3GpK)>8Y?+3kk*4G;W@z@2SOd=2}v= z4mOZHa9yi@eN-N>lXOv(1L;7i@goJG&6M>qbT$gV7n!?O-JXxpK1Nq;9P(am(Vn%z zHx4-a=G}Isb)VdlF;;Mq0>1L|X6UQlw>mnVX21v+iU|CNcNm)52bPxP!i#T@COABMAQS_yNwz>q$f^#-EXXoMK zX8yKSKX$T3Xg66J2Z&xg#!j}DEsjX6yu6&ATP4c9HIqgF!Y>1bnSPAAxc%}D z?qVGdvl+#iPo|ep!d~LTE8&d|vv?*LB?&l}lfHS97 z$bs2X02*ntP&%&0wLb#kyf6hGOsipWK3)pCYs&W`O&Qh+-ONC7JO`90MyAhmRP|^s z&fGE_jLd6q*^s?M1kx4%NU5Kg;jH+vA^rJ7d!ff{Zwe#5(m z#wZ3vOTpZ0=(-&z-fO&bl1 z&kL&;w1blq=NtsyfZR2Qeo)F(kpJH;RH)k#RC%wsndwP+q1?25=8*B7p9;{%B;Luh z!b9z|;n%<9`{5KdBzgx`Y36)~r(M;YHh>%(-0X&hX>~5Nsp>M`AFZbx9D?*H=gay{ z8p@!Y4td)gzA@3rc+V`I<($YXB&q4zUXc_F_H?}?n+^lX6nW2~^pA0(5f}wg>AJ5q zQ%KknVLU#)+z!9)&pvL{_Mvu2fOH5MW&5}cwsV;V)>Shzf9(*dl$+g!ObUBe0xgQa z_A9F2UDwbZcpJ*zCnLd@7HHnoM#?meKovi&lop53LOw)PZ|h%J7P){UD;$~z(h296FPn0}W@TPnXOCU5 zmsxQ?y6r81&($HGz9)pr$E(3Mq23U6(}Kz_@iH8xjvrC0hKiLHGdxxXMD2%>*yF!y zA*&OfRM8iBURy?VIt5u8f{M%B9nM~{2>R|3G9Pp@M(*X0R zDA4As3P2m1!Q}oaX={B4pKF%h;EpDOucy##mLeLVF9T!1W-tMU<|bbm+M|^38D#&k z%|Z_{^QPZtQ!R00*CZl(?ppRLGxOe07V0RD?ZP6@iNYOEQ}&8Jxt`_FverJznpcAo zOhJi&-Y+TwNsB9auc+Y>kEgf`x^Ulk>Kw2j{|Oy3|8I22#Lo0zTI)B5|35mElCo3h z&I^%KCgbWXAYU&w)~Fw~j2>u*uEip-Uy8N?Hanl^W!s_up^s!wQx6tV#J_5dfus%t zgL)br=W>@pecTYoK1=*n!t_Kw$$L3?K5oPxcaVP17d@Fa*<+s%r=>wfUEG&H$d+&nFUoL_>dpItryfnB+fW^>FuF5Ym@A%4=j+ z#%f(gkhKhCEzK<@dKT6Zrb?wpy_D9<#|&i|u?!3Idp=gYqa=TSSY6MWVwaqMn6b!4 zz^sIGRM_YtJSq-iMwYXtwlt6*TQ8BfNSdoI8ZS{@WN3FBoWQ+QQe+g_$`KXFZ1qL* zTKUjsNFTl^ZOmmxixXrK^#SXsS}Rs@Jm#2prk75Mwv%94IICByWME(^nVvJYFrZJC zPry>aa-!16Oa654y?Jjo<~2xZTdTY{J5D^LX0XLX*^@frhz3&{eYP;3gXutXR=&0S zfnUDYkI2mElSJRu50*KgF4$qh&BZt5eysKqugYV`J~t0lHnyuPU(ir}>2+BBS@Tlk zxHN7EVm-<4^Fii9bVeCt#@H>_*_y##!@QtF=i|G<+I{O;5*84YNvsfD62X84$7+(pzYUE5)oX|W~)zli!Ino^s+|7Aaat0_> zz8EHupr!?^*JWuxC10NoXlV=O;39jvD{cjB*r0sY|4q0YyP4L?dUxfTrk^;jJ3m=B{Jzv`Sy3Zu ziN13Z^m$ma(iuE7BX_;9%jg|Hy5o5@B9k@|9K|$TJa<`G{@Q9>!2ri@dKVa{~L9PQc{h|v_`NJ+)i2Xd_a^CkCN!k1|OGX89)?}<(_WZ znZkweg|BneNE<&*q;S1cRVtO96$iAPM-Q+8*kIVw2o*%Hmez$i2E!C5sxsSqCnID?QF9Km`c-vX=<(QDgcd?z0wnvN zh!pcX#|K>3Gd zz~-nT=Trx;$+~aIBIBmmc_AG+(io=2ReZN>M1Z{zB)sJs-}C%7iZ7#+((&!ueK;q~ zT1>#v6Z5=s^R`a+;z^kgFxz3SuQ%_aK0W6f^dZ)OwHf zB7>L;;gNtVHUFeHKPj&=X6fwg3{*YB3w9f8-r%-TEu;0EGlPP{D#QkT#v=AC>UelxltCA<{kTO)pD=C)^}F zaVb;Vv2f@ItyGQXU_Fh4culf;IDrOh5v-O?t2XbN@CRFV2q>G%mp#v zdFH47Hwkb!;#zRM$T~wBX@J*(47;$)IHbqHQFx5Tc5`B?-^AJ(a^bA(;(u0g@`NxK zDl}ZIpm&&$L_TxW10;q0>Z1Djw1~Y`5v|zqT8P5E1Yf5! zuDS_^gu?4FdLSrzou8E+HOIJ1TRNUx(8bNX-NHY>>#hfwb_@O(W2mI_Obgxw*B3yW zjj$M@lysM=dZ$8=qBXj8P-Nt#E`l2yEmbFm0RbxaC$I$``Art@x0gT0RV%FbhwrxPKpCM=*(0y@HfS+%8t~O zMEE$HY4+;-Inw*-S@aCFv*@sF%)9Ak6YUadt)Zr(UQ;?&+}WQgsz)tG*~`UcgselK@3Gko5*lxNX=lg{m{J*#@}n zNK*6XEF$VCbQprwJ;NdsA5@pYv;v6~CUE%LE0cw?Z6W!R8?mm-QR3ndeC`ut{8HvY9Md=->bZIH-7_Rf36c6Yf}OK$1Kp@y7Ro z2-W2&%u8=C6YDjaQ(6;KLOMss=i60A3uE=MA(ssy=EWy38LwjFUp5TML+h|f7IXm4 z&YBTPO@U%ee2%6J&@41la8LyWdx&G&0!J_O&xoUfBdR;cR+433RhJ-Q68&cagDo+P z->Gy@q*9XNzt~z0(U;ib$0fB*+C!o0#65mr79@aKaQ7|(K7 zaj>#FSMkJq%DqHrP9NZX1IF5E4%aFc#--9HG-SGbVAF!|YWyi@ zsb%c{Alzzmpx@>ZmrDUDlFZHYhax>?4_BG)MRZUqtzGSqoiQ9_r!3`Pi)d?F37nrI zD#>X&uGxg`5V$7^o{jdYhY&v1X)TI%arv62u{SPtGpSb}cgdMhPP0IAJ?`_RRUwn` zL>^B9z217uRok>|RpWkwAov)zx=Tqhejl0zLA5D8k|6YcQN1jYe9zrFtn`E{Enz5vWKaaO0$<$VU?3sZhqiwv92XCq|C;i2&p5MisX z0miOguJcdPpdYo*(&=x_j@r8r;0cBC$s8I+O0F48j7_pWY!W1c&2hEiv5Y^a<`RV; z+&(L}4T|M8(LJ})shc5u*ec$pc1zwpVSUpGQ=NTq*B+h<{sXzKBjw9Y^Y;~8)5^;F zLA;>B9SA{Ud3S7GUIW0P!NU=lWxVwbfg8#3hj8+X_z_?AqgaeBGgz9o{X46{&LqWz zw~F5-o7bx%BC208fP*>HeQn`!QDY)d^{!(`$5 zWTMPuK9-rr)ON@`m#hC zZRsOU0aa6;VZKw(A@M9)mzm%v^9TyXND6@7?0^X88wCC)%i_=b0?ZW)9hW>0ipupq z8yjP}3QA^`2i-#@IA43fQ0R3d`YjN7PtG0aCDyfaBy!9(Za{R%_jl!}WUn?|=(PDM zG_Nzm?$PDF0l6P17|jUXLCtN4*CJ*|QPv)T))lTtQj8i3V#=&&$KuTgP3SWohIZNy)dI zOPI1>z947j!|X?U*SPWIa)ps%X$TdNEi;<8l4ZfzMzFi($K&lwTi8gz$m1`@5^5t1 z%>*KBTu#}0wN=WVY8~%MBBvZhiR>j8guDe@k6#^~5@ty#f-5~)rW69Hq4{A$E{T4K zvc3NhekFu^8oL6uSQ(k#3Sb%wt}sDX2fyOVBk#mFZk4stwnjxhlu|`Tw>uUvf!Nw zJ%RFf1EBi2ezLiM-q~3}zf$)Jv}mwT>e6?|{v2WeFHa#eB*fhWWdGIvN5ev-dmis3 z2@Xjf?drYx9brPKgMYmW-Ubn_caUeBxve@l$^F;+g~JQt+%infYb+P&N1KFdw5u0_ z;wZ__JiWjBwJ)AoL7xcH+vipFP0N-lt?sr!>5WbPkjYm)YMZ}%%Tqng$s;+(8Y_7J zPNx;66&2Mv<2zn9f)NsT>F5m@$*rTG8Ls zceV$>S1#GxG|>oj-JM9@c|{G>t!)8UVgdHgP|w1UJWo6K`<0l&8QFIljmQjbYV4V==B@ zQ{tn>>LhxKrOloAqTx~fZs4DlgC3pkb*yP)8B7$cNJ;EzBN0kkQKYP+Hkc3#8O3Oo z0-ZO~JAJM>jmVcJ3$DEIPBWILX7wF;;bbg(;+UG6m8FzBqEK0!!uougmL=F0h`vbL zMls8P(&dbHb)D!DBz3Wh!8;4>nFG0b=rrCd9;IK`%EH;voUABa0-iVMSr1phmI;-d zAp@33Bw=DkocmVtWGoFX7a=4mP-LC(PgKA(mzB+-6C5v4_8jTm7~9jz-C%F3yUoxC zj1hq1i9&dyIem>0#KCA5>^t#bqXW?~?C>otc zyv&OOHqaAwjVl?t48kq5^zeh@%?wms=@3fC#RN>DGOrrSJg)>CIRs~7iJbR zAhPOp7BV?1IQ{?)t9n68I}S>y#Kjl#YW;lvdX|=g@W}=}pc(3a-G9z4KC>8}!yO8R zL8Z$0!tXOjvVRN4&u|_eG0bvKs_jPGT}Nr zSQm-K$eYMUu4ddBRd8dA5Ohmxg3|G2N*hkbFsVXw>`IqBwx)W=gQ|@F`baGt$#jyX zw1A_Cif5_g%LW0BjX%qd{A0&Uc`qu$!gFA9dN9gfCaOVE_+7 zTY=SdW(#KBw}K4BxAC*aHIvW5-n*U3{R-3=mwz5+@sV zmJy|fRW?!TJ%m|Y07L9!E4`8oe9>wcOujwNfP?MuIR&NQb5lC=!)@yht zEYEwpq+YORNiRy%&n5j2J80f^H-ZTv5e#D)$I?{qmY=g1WKR-H(v^Ad zfJ|nOiD9BujEi4f7r4b9C)S|39j9V#KwXMcQ-yEsu1;M+ohhlwCkX`?Lwt5|9EWM1 zVcJ|P9zFa0FTd^g_%`(G6qw_yT<7&)=!Yc!EjX)Gv8Y{qD5*ma0W>Xan8xY@7)PL` z3&yCzGXa^$zGY$H!>|K|M#W7svUOZo-0ABtq?%;37MS}R4({yN#95|g3mQ8c7f2%m zOnfRm8V8qK%8)hz5z#(0m>(KmdPEYzNRU@&6@{m$Y`1T`CKpYderw6{7EQsdKBB{0 zPf$M5)H+I%bHK*lwtS=?78rO}92^mum@={xP0DhjgEKh^`Cf_*)+3EM1j?TzD4br3 z%>*0gn2Vxo*$5qIvp4`W@l@RThy6AgxZTwGI4YSN;sMGT67kZ{d4MKbJX$t=$fAWf zlJ7)|EhM+$s5wBEf&Lcn;XS!hD$l=&C8|4n>&ks$eHZpgs7&)xeAn2)g&4ocx={<9 zF2Mu@##rrvJrC%>4<=Zz2e6`mUyCVtv-T9g!nyiL&WS65D6k~DGG*(V=$-Hf79wwY zlxYrFEM!aY&)RTtUw_0xajjP%dgYu?}oO zUS7!*Z|^KoYV|Vrq+yJHysfCp+@vcsc370c=pdcmr1A{$zZM9V-_4(Ej1{mkc+?bk ze8T2Gz@7u=WQw7+Una6J7#L=n#tVzb3^9u0U_>nS@8-Y|H#;jcD+imDHr?4#Fbe1# z22fU(Ft||zRpRgFm+2_EbT)g12&%h6p2w-V)o~RRBKKb+@mq6`+c#j3lwMLK{n}Ax z$X9PoW3Z|1!5t&YhIS*cfUMqIaZ=`B9Es0OyiA|O4FoiV0%ZiNTr^^e4+K*f8uHl( zoVa0&z;t*llZzuV;8H*}ymjmbLwLrBqs9noZ!JarTOmIYzvLajV;h8Q=iJ+a5Bm2` zfZci?5rAz8VDkUa@Jb`t$Kf9{1SszGMWa*e?#|4X8)Cmyy0+opwIL;f zRIpn-qDrM8NX9e%9>>nT_$Kq_TF|>zd3bp_!W-X zI~s1hoe9pX+SR(#LtrWK1BJ`D=HtY1tHLy0M1|poE|NM%b>Ffc6E~yaCg$ymc21Mr zu*rb$>7&g$Fg5G4wa%gV}NoWbM*Q->x( z+q$WP($;IyvSSL*_@9ey4azi_NqFP{Ex;M%}A0j3l zurH|*sNN2hszlIzsp98EKuRfUeaA@g-=*I}l8GV} z_XvJCzHeufdIEIt&ha;`3fN3YQ>8>>WuE~!_mp_-eG@`jBK9Nf58?YeMfGnJ^8 zuOt*<&P8l7jBf^s2@eP44+X!GtMTh%z?L&Ihpq;j=LEk1s5d*K>wMK@mL?-*#@rqZ zyy&;~Vhjf(M3s5bli-k<%5skIp!>BfL)rG@^#kovfe40yPD?P&N26yo||N=3@on`VIt zpi8n7xp3+OyAAU7`v4jVpwmfGk2(}Bi9eci@z>- zF%$05yc|0c<+9*HK6v(~X#l=4Lm=XNrVZj#}|b7m&Pc`Hw@_dc}TrOxW1S zcEI*xmkj+HmHlyKb9^W|dtKH&9sOVh$cb?d`Z?F{br*paNBCc|4SfGC04637d<5RB zBlz~gMt-GGHX;*47|xLl6IbpUk_RU6`G4*!sn;T8bw-0xdKskv%Au!!&B*L1BcD-^ z=qE>DK%eb}=z44RlXNEJ(+0=qiOG`F~l58a-Fx6z!UE_%?Nfj~W0Cmm z>XN_aHR82-y+e5cN>p=mVmlE6z}_p|fhm=B*NZE$0C#SVHA{zGKg~GgLQSp3IstUt zy?mY=oz0NKoRMLs$;{0y;nxVMV^u!;_WHL2VHe<F00PNmhznx~w5ymO>$0&ig5q3KfDPGcOJS8oGW)96kXKO3O@4 z{*y5P5&oyo%hONSzbbj348v?^$29S!tjKjgR9brqPB0P!LH5y-EInz_Q|YeS3y7DaEjePb=8nnoe^!XZ$#)qEJ1g9F{9 zuL6H%I2WWL?JKgH-*4H`D?CSYB5&E3`I5(9(_6ZTZlzLv4{hY7!r77gW~T7E8SkABWfV+y#TR1%poW;}`Vy zr^NI`h$NAD?8W+=#p;P8dD5XbPLEx{)#hxL z8FIoI5X5O8)}FI?Hi%n8>R07^qwF29t!45%aEs2NPuq6v&eC$1`fhQ)DCUe>ord%A zE_4Tkf3kmsp*^b;1{|)Qyd%JhB8Wz)5_<5%v{ptz6jOR=GhFs`PnM>vmzcjMVciKc zjvty$@xzajFBSo#i!^-8O3DC?m_TjCPXWL|9ZbL}yM7}#_|Z`}TDZ+w1G4&H$$W2Y zwLuI=YokXVj;n#dz7f zWM3$=Ay~e^c3V?_i~Yoa{!&D!aGlJ#+TWkj9vWr-C4I=>29T4g5O?5-krUCntY&pX zli1l+q$y`#C)~5Ontvk%XOQ!3dtESV2V@t^klGI-T3vHDy@8iINUZ>g#=UoyF<-w9 zUf|yxp<4spkew1Qz-w=4*&XThV8dF``FXys+hZizk2+xuD0Kq6F*B6O9w+ZqaDT)E z=;5G!UNB3Fm2Bf_&nuYV53zcOU1k;=Z=w!iXY0gyw1%Q#@jn~!y8G;f1}C;Ymilpy zj(2*PGufY&hj<$-p#X?*eKJ`uUfb=&m1^kAR$wKIa%T}W2@M#Ve~lPGPgc~D=RHGl zw0$Xix@Gv@nGp|fNTK~^jiC&sWygMv^=~^c)K%Z~ls$OL?xJL}+G+dRMD)+ygj2Wm zXMvX>*XVfh6E}K6jt5Q2&b(J_T(Or68Wz38z12RcBFJfcu}M6!+Wyp&YB`pU9R~38 zIG&>Iq&hI{LtC3mML`(%Ow(UwmN0iSTogHZ#8km`I0lW}yJB^`WO#7@p@_Q~!1M48 zZgrp?Y|x3*wA6k>9y@YwW8yj98lwXSL@_P%msW1P z=x;%T@A%?B=P8?E0n?2GlkD>=^JBkg0KR>CvD&KOZ%S9pfo{hzOck<)T~j0lL8d#u z@VU7qgWcTVxk(qD4ZRoqqSr)9dwY8a7iOPyiLey4&-{wQQnWkJ&!H`YB_3@XmM0o+ zGe7dp?u-tA2ZFr?q|DImdw?9(YdIeo=NSNf{vUAzA?|12KXl&4(tWg6CL3244_ghU zExSyvD;ROhrp)YMq+Mz&tj_HGCH~;2{xQ0|mb6DKKRvGD*S!)@yI?w~1OGaKdy$Ck z?L4DHteeO#Soy1Y8Jax#UMQL-m` zTp&i`J@Gq>CT4tYYR*xUq-68-{-oeF7+X0{j=wD4HE(t~ zh6YV^|8wi@;UV$vi1u|9zI(2|l;Y%W>*y(-EgGj0=g}_s|1ow>U4lSc4lZ@swr$(C zZQHhO+qP}nMwe|{(=)Gk9`27gYvp7o`C59H8^+=aaj5iyDajpJem;+@yoTNfrK%Lo z93+M|1}#-47KG7ED1wet(K`N}S@*u<4(tviRyDh=Y{!st%Mq(&wrW=0{>mq-ei6&o z)~?Qk-4_oR!76ertm-)7cm>Nxz;%n81(~mAxwd>+aOa8>wmcWRVe6l?jy~gDx3`giQ%tFtRpA`2mc*(qj+6s5_YyQQBtq^OZ-nin@#dA1Yn&_lQgSz8kW_n$2LM^udXkA zIE=@?_;Xsuxh=+UpQlA^y7yS-2}~nCV*e^K_w5C%7owV~++CGKLZy>)4#&s+`Na1~ zT}%`{yxYMPt&Li4+`~Pjx*E=>r~-gX&fRC+*|ZcCOFmt_n?KUCLeLY}l0Pr>_+C z7;N6{*e$|4FfJ1>D4%l63e9P3&wZIH9c*CNAX#|lO5dXRHOV!AaF#r`|inzfL0 zLgRLAm|ajBkt29KZ}D%6xGNx-&lVOVyV_e|_S9i1tu?+ktS<^POKUmUG?lDQ80MZs zjTAz!EDX}DQ}eGamu5zFb3w{+9z1R~VlVxH;h%E&H5kx|_=Tv`AI%(!Y&Kx~J5h+k zLPcO|pMZr|6AF|&K}s74hUg;-6yoO*c%eWO5cSTc^zAO8_=a@aNE~B*S1_eN2b%+! zKh(75f5cn8mPo6VGp9L-tDrmEB>CzkA3{~|6Z6-5R0oWgj@J4uc_q8en!d4K6{@#y zKkcaQJ9(^ga5YK1(ET>3vs`%We({ZxO9F+g){&(`TAXJ%dr5x5);NfOR|(xrl!oXY z)(jWArmwrL)YOW_;Q3S^q?oQZ;SSGe_FNNv1SDL*KRE+zTBk}i#Ii!B2uyUg2O~H6 z@;CVXg(ktW+cto4S&`|{1qZH(JW|iVV4C@4#GnG^cfYF6B>{%yS886osWpFyssH9P zJw!MqYEiz#BZfVWEN`a!Uxy|D2Pp5E73=k=svR>_Wr7xW_~a2u&V!KHfRi|jUC8@} z-a_FbvJinSSfPM&(n`&AdF^>r2E_ihgo8xlXS!tJ`y(W|6D1_}8y{c(X3&7iB(NxA zjgLgF!}**%n)aA)j)@aq2|VM!cW1N;x9yJ>oBld*zy_c-+Gc*S7L*VuRu;4C*>-wz z*pQYzmrmU7^XalZdrou?*%>#7mv3XQ|9yO~yQFHwd!lZ^S7#Z$H@m$!u8YH`BCmaD zd@WsXhDi|0ql*ymgp?VUK=4aK3Ti4<*q3;e5`Wq}I&*lmWoQElkr z)N$F4+yMmI@WmNtq4fj3(r{eHG;fiX;s-n6&ovjnsNQ!fV+u@M53hN28?e%(_*aCCav^>+ zA(@?{)h$Ed{>-*N`-`%yt>!A;X29O|EE7~y1^+C{`3@og&nAlbYxBs!GQ{#lUu&KL zuAzFVr$ok?Z|CQpDejPItS70ikf6=zz6T*@;1A{Cnt;xhqnLN%?z;kwx7Kr1+4Z>aD)BQ&Q##V{Re+Ntbs%*_%Vpi zW)vR7$qFUFAZ0XjpY?}!mPYr7Kr(%)brQ==nR z{tb@TCh0%_H1j#SD0fv_6}s5ak`H@$d?X=NPZT}^JW>p?Hj(%3p^GOz)Vr+~otNEC z3*YW@wnl<8e9`s!(kuV0i(1<`7U4b(MFWX)TI%%#qDLygGTbzwKH4;A1WnKiGmcwY z_rxvRf)>-S?11|f`tU!L*gn9l*&odxyk!50Z{06ntOhh+5W?sk9-s0gbPohr-UFen zHmSn+gykjF$5CB8ogcDwrJikTO-Qh6Qy`eGa0b}e-122oQcPz5;MRPp*1jlgFhCss zI=Ci3NT+Z3*Q!KiX{H;jNK+K5Gy`9_4(S4&zSemdC6OOr#o}1@{-$*%t)`X2ps|hW zRG(jvD`sqQc3>{QXe3W8B+Iw*oeJwQaJ#^-HE=kL&TMIc(%) zIa$IN!HYFKOHsW*Jb^}M-cIVX#bK3UJa%!3S-`^VZjCk=dmi5-5(SmhB0KjT>&+|^ z?BHP}mrPOR$RoiCpszrt_N`{HIpKmhry+bW2bk4(`HhANsGg(*A4BH^tY&vVDd{Bk zmMUa9uiofZiO}Y#(bXv(Zq{V?C*_631H~8U>+2WG9b40h4Rpl9w@wR!RB%q{gBi|N zOH)?5f$VS$5aH!3DtaT>u{}Z*-giU%I!gRn&`F|?7uLT}W0HQ*LRtOd<{12}GJP2O z5gK5gjny&iRvO1*~{OOGqROUmkuMSgRw<#5+ z4u_pp$Ff_}P>h~qT^z|&5hZ~fe#WF0DcBW8B_!DUOl>7$BNPJ+jtji?ryMTLF zI%nQg9c-UG3iX&;1>@6pPhrFBR-r&5)^vV)rS~{@u9-V`BJ)>8Pj?TSNmgx=R0UKb zD>?I3{^cwhk9xvWXA8}P#HIxyM|Xf3Qb(S+A|*7@_%;*pMryCZ2){vq`m~RM(6GN` zk}|4V3O9HWt%k_3I;O;1`;^+RB*#J@{a^A)W`{9Fx5pXN7YanOjn%S0RHB`^Hw&0& zKn4U126c}`vZw9xRkoos@97<6Rr8GYkmrwAke>=~8$NFXvqc&}aF#7$T%89wVQ zAp$lA0MV67-9xc3U-t^_fpfgNV_9j(yM1r9o@Z{e-&2=krR(yM7YF)7X*0P&=ap(F zH8&AMB!z$9whd*YX-kR$7?6NPU;v2c2%tj%w3kSx&_S7EPqap3zQr>tL>9Co)^ET^zd=si>Yst`HtZ^vM4NAeGb)vWr554jIBYGAp=rBA;n2P3&}BfJMM?+>t5` z^soe!U2Ygr8d*=g9Nk~O?hX~omkoU`xIwXQHy5sFuk}5cav0&wOLtX zyTL)b+N&0J(!%~5L0wGg2L84ZyhxF>D7T03kZACN^Sz3#)u=FHqUq4#gDyC)Ykx0!LA5vlR0Xg=*@U2$1;rB^@rZ?okEkLgFp0fXYb#ICz6E z!rg>_1pe#O1Q*p2ALqt!LTPDWSXchBX~nh2;y?ysF<>gSmRg%f(>-WNF|QI~b1+RM zF29c6bJ{f(VFSG4oJ}pRvd>$68j-FNkwb48KPZ+o`09V0Ez<3B+2 zz3<u&tlT3>C^%<}}@7;8xWnC!q@R6PMyJRI}f4zqz1Ivw8`ZAn|?s66#ifN*5@ z_0zld)usVIr`NY+7m2elO3Fe@npO;!SbChB(fRI;xd=%K`+9;pDf`RoSv<9{R*9cQHzQ=(nwGO&lwa_N$Yw+hEiBW@%& z|D*HaZYq9FNrDMJgwEo&os}H7#2Em zox*4LrAc`02>WIiadiITIW4B@{9D$0al5iROFUPBI>ABuYn}+FlztB!dN8K6F_{wb zc#*U=KoS_POE&?A`d-;8*XncU%XZfA==CaNT&#v@kOQa?pvdE}Sr&}Bt;gWSCWdJ9 zzC=R&eC7uq02c2oGnAwEYZ}5D7hE;Q96;hG6^5eR}nXv z|7EoITr!gFe)KQdIt&HNtCUmrvuUKRG}36FYU$|<&-%qELx=;OUfB(l!0N( zaiQ6a3A59O=7%HKyn0~cz?FL^SW%wYSaacJ8*WI%8y*xD)!B|EXZW?8p{$zhp{4zS ztC1SXwNuQJL4HwMor*Vl7$IG(aMZmdxG&*1AhM{^-AK(HPq6~Zy$l)QH!Dl_6XE-rdK=$TmTiz5Sd z8f`$CBjT3gmLh|#o6_j5##N%?oiRB`*E;taf%28UGasp1v8xK~DnU0F>FtJ&sX!7d zFv8|?pzj+m$``B%)0%D^FK$(O8>p9K^t+Lo+3>**~o#8(EKss8r@xlz?s+iC~}E zYCv8a32fb*_hJ+!(T#y=P3EA9Yz=X^-V6<%&A#RdrH1B;I+P}V95y{CLc{X4xEFvl zB~xi`J4;CjOThkK?F|&>s?pYDBw=%mrz{>5AGXYh-&A6s7cLYTTa1E) zjkn?OS7C^3!GkX*E8- zNXK&K<=)B1!@>KJQIBwv8_uRfMHRZj~1K5v)_Dra^bNfJ+GpKIJ z<$NjeYKz1+N=fuQs43eZrU?;Aq$VXko9c!TjSH7CUChRxa<1>s8Vd%>UV;w%jxIC{ z1Wu?Vjv@f*^M`3l@%t>EI$1(U)u-HLoH0D9%8G zRzibwnHn=9}9sQXk+dt0AJ58}zGL0iSeHv@~LTLzh!u7T}y8R%p=;W{qi|`3y=82I?&i8ZSYY_@j}y!kD&a$7Hz4R&qe! zr>Q#CXce-J1FKclI@2+e#MC+$oh^*5LmkzbhdjOnqjYd~64!|{&2G+lPF878sF>x<~UhGkf=v8y31IGJ|*d z&w1i2-y$vUvCI@G%A0n0yPqwuDX)$6=&<~#Nb2ucZ)-WVh%}nIw6#r#p$-|}VC=(a z8uHINVKd=_;=QiAsZf%ShQVaI>|qc_!`t#1gQ;~B#K@^KgbJz4*!==iy7X>0)L z9-0~Nur8`SduppKh_(*ewZN%pcO2&->xTqAtHu3qA^Jmm$ zjcJsa9e2oBehjT)(T~7qO+z&TE|0>UW6{i?ned3j`_KGpsC!ydEzX*4tu9R0i#W}w+%=&0HdGJvKI9ZFFX zrT*SXK^OHA>d5&BS$sY6qtYt$B^Ii+6{NWn_a=)2G%3ydyl+v$flt=rkrd$@DE0a7 zTb!ACdw64zhEW>>1j8Fe!}v}&OD z*CsT^m?p*5SYLOP6gZT|NL-?G?l1^*ILwUtn%QY|3HjtaBM$5hyS~r3^dL zn9lcxi#={HnL@wU?eR0$D{1nKX1>cNvdfkpQE*Pp(C!*6s3LT3>C)SGSCvai*`(Zh zVQa@iTwdU8UxQDc1mI;Va83vo0#lTuxTsD9k*n^FhxJrnu9r=o;P%aVHw94xFJ=R9 zNycT)-_52cvqhQ(Q38P*iH~|J^)2=Z5J3FU&-4m1Dtn2pFubXOR2OnXs6CnY5&msx ztW(pO$RDQL)M-(XK`rcZphE#4?9O=+?%qrQ66av_0ZmPa~lIh!TCjkj-9X<$HPBHiAf zI2C}kvh{8`+8dX;9)=j+Re8qk&y}Xr}ZN)*XR>eA2{dVb4J$LWy;s$~FDh_lg1B|5<_$ki3f566&;!Ww+gdu|IV)~ys1WCIU z0Ovv;gr3@y$3mEQZ0RvRE4j`St^z`(1Xl?)Bu|ZH`fHqlX+tVX9NSwibL{~Zoz$2_ zWrc=#W54azwwa{wZ>&vmWz}B~8xa%u`O%|WHYFX4PGz>#1y~9)V2V#?4Mc099f}kZ z@)jK7uukjEKtldh^ZQUPhJ(oS>beqbv?;8cf^?cDqfgt2w+5`hy9Z?GOUz_5wpJoH z7S9}ySf)Ul;F%zYQ4&z9*G_LeV3yvx(Vzuad2Ou2+ z(HYI!%^QE2qa95%NB$e;mG1myB*7c?8XuGLkhm?~ey}v_9d@x&v!< zt##s=QFx0~-h3pdQohEhTWynhPSay0+9H)vfd-Q+=Jc>hm~Aae<4`?xUq@%><(7); zMl?A{OaRWCA^7Im<9hiA5|och7fJtm5zwD1Dd1*zRo330WW$N)){6(j70;ht6L{CL zQ!>Z5cbUUWbc2g?9>T7K9m~p6?d$4bDfM0!TcWEG=~LO++#H!9Iq|(d0&Gip`vQZu zt*=OMh+pT5`dMC}jkqtYLM(y!zdTT+lO|ZT&Ke9MG9T4$sz9Y&r0C;Zyw!TrADWM^ zkWJv(`orhs1)Vi4S_Z7MFB(7zm~ItRG&cu!;SG=+;O#^9Be6li(J@lIbc~sRFNz8R z9LET0`mNJYj?TwNn4%THA^nHs%=}_}k4>ibo{`Sj94jsjpWph4c)}JQ^V0tPMmv}{Kq@58LcgCBC>ZS?poB94q}YAquFA>e;+)MNFDZx!~O8B z&JeoWY0VxBwIG}8G#SEPc0x>$YFZ)nyf2!lT#0_PikbjbF>E|m<%}y>V0Q8!gK6vo|xay}X*8<_agb24zUxZB+ZpbuAC%-UX_Pm*7FSHrztFQ$T zdT!+rN^>6qEae=-`&&QiHnpf$ zrhN5A-fKeTD!(g=QZ_X)JK1VOtJ>cnfY{0_&T;l$gUaaiG8esJgTaKC)@5xz9mh3v zCvsjpMnPTR>Ap8uzaJ^aF17{@Yu4;B>`a^arbud$q)I5esC#SLif(g+8B#$@Z(&QJ z_b_%S4~kR!yMtFZCJk<&6|hFgmWPstZzI%o#aVf?ybTq8#T}0Ve<5w7M(Zgnk)i{q zM=JaX)u(+$ou|tB_n+9%`9trGO>zOSQ9yI9oxkDJD>$~G$CW|02)+ixql-p^IG3ek z4`(g6f@vWxBrS%bcmgL#5GU6ADZUx3#c3qbhc_Tjend)yQ>Kt;iwx$Y1OHOU3fB@m z{PL0}5(SOrt&zy(Epz~ur?yE5=$xQ{MZ$-kzRD%(JETS83|-rpypN253k)}H?&8w( zSUjNk-T$11;R|hs#XIrfU(r!b7P$+|AOZ|7?V;u@qe`vnl)gDqC*tP=q&?(TZ_T}8 zqeVLzR?CbWy1mLxWTLUtJfC=XEBA91ATEICxpmf0)thx01LVAF^`PJATlv&)Z}?e)>mIUp-;dhHLT{U+Z)r?u3` zYNx*omkReV2{DsJcW)h|2*~8+!vI;|s+v*WCo^`(uO-JNY(!Z{orU4? zr-NEjh`Zwr(kyfj!#ple)JIFY11)ws@$n!ll3gfigOt`CFw)jOpfJ#7KGVoCt<(v! zq?soY_%HSMW`@t^@{($ATY@Dj=Og5YKzvR9$9xPdUDdJ53=)gX+L$BITu{jlXXxW$ zqQSwgOUpqJJ2$8HoD7u@`fYMtC_U?JGmZ%cfD5tPVJ;;~H;JB4O5Ebuc1HnTH$_!r z#<9&!636v0ncn4??Aq^GURj{ zApTN=JNrMBy62_47fNq`*)oWD!KmvT^{x>Dm_ztxkTjAv_5O4wtl~Qpiq7wuK=sXO zHtO|FlwWZ1rMd*Ub+@qW+ZCy4jN(dk?!?0_n_LFynbjcI|U&=^y z+>#>iA`IFe*VIk^Ln1V{k@v%*p&_<_65@n&DD{Y4dc038TJm|>UjHr#phCHGw#4L8VoxY8#HP9=Vy&`O*w|X6suHO_KvwAOf01h}>W`LB+ohFj z8gm7;H~^#vW3u?jKMzonm$}|->?Knr_>hUUL&f{97H)-r8^S6+Eq$pFt(JRTpKv3M zoz;cIoW+_OZTXZY)f$}Iey;|pg7(ViZMh~rXi*I>t`gBEg2^G$2)x+nP+&a3`o(vq zUm~@pDr+k#b0aZ*QH}XuF?Ra~+lI5s37N>|7dA?>K8`q?dT|}>Igo<7wVG68<5h36 zl`!L7OcRX>+tQ6cJSA0i-@71)lWnF_=W2V4RZ-v@N51D$T=iX@-PDf|+Y5zzd4ld} zpDyFdpCiZdpNs!Vk!61 z(=oMgf+4HtZ|XO@o;bzEKa8Z_ym0J_&Bn?Mj*qzF%uYknWPlS zp_XyRKahKZs9Cl7W@GKn%qdiDJHpk!kNhP)vRibSTo@i9O5}^PtIN|X5&Fao@8CHf zU*IKMLb=4CP8BTLJt<=I2sdD_gd{oAT@F%^hje zezq4T$t=Sp&BO(m%YmF^;O34(MnK;KF*#yt!9XIHRCB%Dlk71>24;F{gM?N7hc?w7tV-6|xDb53wY0xr~q+G)t z&hph;Z6{y$-8DPMCp{_G7^7;f3$~e3tmJ71hD&32Rc51x^=-1@)t5Yo(5UR$wCmFG zqDfbYOiaw@_0m+0F0S_HBUR@Jg*tGXMptJj1M_NZ zgRcILky_UAtUW$sP++~vM$LAT&-}o>*K3ko=Kuh|yM=V;tplgUpGY~fiag-4sfh`F zM;-BJs(U&j@h<5)5L@lP%cIK|gyPF-yfke{lyH~svU!SRJa%w;QMrL2 zVCmfMOkRL3@4nx?YA1mrbVd4phPPIYD*a(+we5G>aRA|`kjR!QBRVD} z)vN+{=_ZuPT{iO8yp;N{Elg0MV$g}^C4WUv2v@&b>(!Y!Gd4c}o%a)Pq0fSUDu_;; z3hquJZljHoUR{(7*2(b|5xV6WFe63ay?U?N#xh1~YLH1B@dES(I_hDuVe{ZJ<%}wk zc_0Ouv(6xC!8wuXKRjVdjP84*@2){nNHd{=1#4)tz7jbB-?Tg7JW^F~IwK_qYbA8F zq@zWixb5^35Afs*bT~GVGjb+*q^jpCx3p|~uDy5u$ll!(bS&AvqSu(4Qreup9JcIT z`yXJLB@e$HsCwxke$Kaig%|05v|p;Al6vRnRVl?yUJ%W9-0?-d+{|lWgwf9yK1Q+u`3{K0L-qO)P>@l}4@1Dvkt&iwp zp{6Nc>RDZYBQd{?V;Nk`!cQ3Uy(8VQwW>`eGG}G1Ksu;6JlE|?>&H^h;iVVi+eUS; zvOv^Q6b_VvahcYUk8i@hiMF040z!s|4$Ixs6Rr9_@AEN6R2X3jk(1q(`wTGVW$Y`I zZ0wsqtkHQEkgGcZR=dF``prlB_e7kXt-5~r@eehCvXzWaB4}{+Efifc3LfBr&#n&m z7y|~A@*=4~NjE)s*o~wG8m_Q=ggJ#mz5Hp1%me2ZJ1)uN8 zraYSziJCY=R>*55T+2$<{L=G*uj>qyMHHWek+TZ=0|52)=7&ouGJ6x!V(p06r+!Du zfxv)E7rQq2P*2xDjsl1N{^Djo_?2cq;4!maq&&Td1qE5pnG@O(kzHO|h*B|Q)LKG`vhz>LU@1!wI&S(DJ>tST09E4+(!3cH z%pabyDCDbNzjF%W^4F@JH5aMbI4X;Ml}FqrOuVN8a2%J4g>7?M`)dX^+#t;1uG(w) zQs>8Nh-%q9{OX3f1Zwn9+o^q*>tF#Qwbb?6Jb?YB5Z@^_y^=+{-`>fHU!auwd(M16 zS6=x$nWbfi9I6~UjOnNS0`n&rGCu^KuUeu7=iJh+*}%l8>KO8wENanLh8cqsP$t9b zNL#b%So?9@klE7Y%LCYOOhiLgEHW%?t{j-ZN=~l_wohP8Wb=_|z=&xecRGRrUvuzs z9q2hwcZeeUdQ|GdVhUCFKoUsgnXZZ*Z%jycdU!tn|Ex^@b3rD=NYho#x?_*#Y2LLZ99OIQV4a3`#Y z08rItL>HV8fr>uMe{vEha~jl*yyw8)fH8m%CIV2Hxc~%J^p6F|v-kV<`7wxdkX_0s z&TLE~@EPfL^0*2OzzdRkDW7`8M3(d|-qmp{;0OWulYDeXkT;Sqq6MJ!Tao!{{)v~i z+#Q|xgR)^whQ%A6#RyL#*hc^eEDrfRfdn>Oo0SITQovb&o21HHMxYl8U)0>bC%h5n z&{);Qh8wF;ZZN|BCy^k|ni%uGejm*ZFA%_@qW{Tm^ITx_AG~Lk-c#IOV0W|BC)PL# zpr_)@L0U;_e_xZ;v+p}1rvN7DCxIssF;l?5UQZz!F9&#*b>w;vJ}uOlv|w03W&yy? z1g0tQO`OZN_BlRYP;Xb_&k(s18_UB9)d;l{d7OpHi1Glb!>##RD4sPCRKWA|6-QtL zb36BAq@RGt5Q-Yj;gyv>;G8r7Y3gNO(>$L&RFsp4^9KNZJuFZdPC;)X2@uw9YnP2W z@1v39S{Qs~iATt@)^J^dOA3i*XOB`nR+NF{*MII^tNYv`)Q8?|KqYbq%*1}G$xiTHZZ5if%Mm9e_XOX%{LiTgWK9QgGEWtpnz zX+6FJauXVJHzcx4F3*Myl|!u9>i1jrX0)7|&B`ju>xs%mnBx;4B`{8w?dczdapWHR zw>E(n42#iUFueiD9kvv{A1>cLQQD#ahyxhh(Co62YHPNS_qIM|j^jhyhS={eMHRj= zGIDWcyAoYfc9G#L+}AQM_Wh@siH6)J%sD#!bbRkn)ce2z-xR)|fFUWMfk$+@8gSe` zFIVU5$+|$4#dSltnCRW0RsMl2^OlSAeFPgv77F1~T0r~oWgDCAp#81B+AlUi=!Xd( zNGFo@|he2yTocisWg-Pa&37fXp@9TY5cQLSEa^npn@+y?Vc!DgUm;0fHrk5D!#$N zpL&Y-Ls!8JrBFI*bj%ms^lgo@&bBOI2JlnT>$06NTOOBLcC%}y^eodc;uZwqx#9U0 zKTv%v(uMmhgx|y}ty=#p(;{h5JPGEc@y_bC7=h$P#m1BF5{N@%h z7fr=q+>(?>=i$DkPybwKwD5ZK+!|)D1M$!r<^&#Jx1fIxTI_ZRT!65^EZnSblZJ;u zzv_i^*Tyh&G_E6W{-I4P`z)gm=F#@FQBWY!Rp`rEO*I{!0lA$Im&`Hc&cRc6j|MG! z(!R`?ed4#elcp-rkTo_zP3R_o7YGTvc00IFYp^^T%d+hr!o>;&TFE7dRxFA|6)8}+ zM$#`Qh{3$Q16U(lZ7-)X{sS&RNBw1I|I0s35FSQ>w_ImstBjWh=)~{g;i{_%Z{O@9 zBOO3GzivkghUd}s&*z1?#uLi))sw{;c;m;WSF0zI0lxFBx^hqMefBH>5dNkM9McM#hw3xnaTq<~dj_N36aEY; zze{j;`y4O$!^!%UV!OR`Sgm$UlDtnZ!BCnW@BB<=My%NglR;m`F5IZsn zd`CYp67E-k$|pHj9xM@E*EUXhupp3p~xJpj0(vnb_uBx&3T=hvDL_I-Ac`CJuyH$_~ z$g~H*WRy6xp?x=j8VbF`$p3W7JcWiO=O{H2M&1|VGJ>&*l=qEyC-neISxjE)bZbdi zSloLm8}sh}Hqcp&2eggsWPi6#DWC@T;g5Ei0lMIH()q+YT19F_KU&JwvU_AZHtD2I z>Hn*FWfY5Zw?rsz6hQ!xRbMZa(0y2EJd^Ly@o-aNEeLHUH1npSBXqXzz_cMsb;?>8 z2@UCv^(WNwl{W?{&oAnHi0A?e2eu7)d0O=JW^3o>%3R=)W2HaiwmBwC?EB;9Zef5q zoC*v+;V|8v%;9%?6b4^LD21Ar8ATvorJ8;5+Kh{ON~_HtVN#RZEl`uCL_!W*-@?)n z9Cc1$5q4%$KeKe$hb0ItrARq1Lqc5vAky-^@N(L*KJgYOY=%uu^i!6s8ofqE!d*0E z!WIcEX45PZ7y&~=4kA$sEWpd6ZPo6Qk69uws1^0kr&>RMveES=d-YSPLe&%QyDOhq z38?C{+2s8Ro{%T}MC9SsJ-!4XpQ-$kZ@xbPlJ42@A036+Y=4YLuDJfWq zeuSO$@ZN;QM1@q_$~RC2fhl@WtbDzfVXP+5TC-`N5ciqhaqVy;L)g7J0f1cwaB8|& z#4QUcchOMu5z1RIw~Z>Wgi?YhNJfF6fArZfe9-eCjhH>l?>|W7X!7ekWhM=bp_RhhOq11~r3+e0bVgd{?ygVxMlwhrsk?a63e<=0kYy zFeDv5>1HLKU2jQ$hk;kw1TgxYldVyUXT90W@VIsAoGXP}!?%sRgjh#GHF8s=eEzi6%ZVt((*nux8c^^J%)K8${UL05g_+d_WxSPkdR}e#xZTYwLgJTP!w+wyKB| zeTUM(mu!`{^+NM`WXgzQY=qK`DT$@~na=8hOl!19%~t>}`gvh-5H#;!+gtoX(MU^7 zeQp?Q2q~bk``|mOI8MEA(jNks6xN73>58tmN=V)%z#^xN785rdS6>`#<7#C{Yz(>Y zYNLEr%%pdMUP%(QfB9U$uj0q|R8Wn@zXIKQ&qmmV>dJ=i*#kiYbY4Ua=wD}`(lN0{ z<5S=`*@Fr)yB?^>#@wdw$e06Rs48TD362G9D9M(iXRuAQ(lkoRKW8tL)-%LC!{#I6 zU*~g1HD?RPP5vQ{H_CS|r9$7p5HGMkeweI29wd#qqacs5fZ5tzO(9#vmk18YS&_#* zk|_MBA)<*KQXty*nt)0h?wria&puv4H+Faw^Xhc|!=mP+fl^V2bFPOy0_M#t+Qa|S zpo1m^?!K8!Z=C^W(~HX}J=I7fcCX1(ds!$H8l;F$4PsFVAAZvJ5l)f3)@OuPrlA>N~|B?N6GY z)-WG`lJvPjWFOu)-nTxI4!42Gcf!&^!&)lTiDR1jAPB=*(=bnrHs_U`w3}NkJ+Bm)Tm{`k>b}Mv=_aljO0WgYP?dW z%<+bq)=isCGRtwt8QzAbt*uh;>{-V*3Ldmjm62NQP~07^7x5_O&8Ef;)vS*tXU@Y& ztbtkFOiProgV_78=1&600&_VuHZninfP}wCOya-S5Lc{FKr5U~zKy&K9WNP4*I6qQ zUmHvZ_JsB5MIUBICMWt%NK{VlckZHn^?K3remS$u(HW{RM(93)bGhH#E})|cp*9y>S9I)O%GKp*4N34jUN^3!xZCYVB=L<=U9>*yg(`rgU1?NdaM<3C5`)?wG9s;*^h+)YF215vkSOg(1g` z($LV*B+Ri%c%=l%C9nsZY`Cj2+Uk)E45Db6VzcNW%F}Nu*1Mo*rsEfa&0yy0LyB!? z2Qf;^K1<8x8VJOaVkQ|Xv9g0~{dIBu+v%S z1Mn`76SHypjQlh)rE^%HGPk(PqrCl6avw0%aJQaKF_}TLW$`FFtFM1p9Ws3}! z9+!Ilkvh~Dq3dz6S9hoMZ}1oE0IEvrX8TpsUM0sSWIg};R0cv35SI?Y@0di!f?9)` zm5Z8Jz(2q&VWk5wC~NiW+XXjCn!Hji0HewwNE}sP@h1m7`#Gz$y$bc%s6{--al#im z##|R5VXN>bPX--Wh437#6}UB5Nl!rR)3pz(7x8Q9w9@uk_~jW0KWJ*TW9KiMNds_< zGG{ySvb6}isr){QMUB{NdnIb|@;=E6k6on=ioYo-KOKmgkWFQN6KVC@m7&ZS8BTi` z7-C!#N59xge|yzDoXwYalkYmw<{*~LDC)A0@>GW-27A)kcN0G;X!{-$iPIV;P;~E3 zkg+Lb3ZC@0HKnv@y_+u!Ymh<)ljak?5=Fh>WqLwh`X{rG$s^k8q1P0ITEG{#}W3{FVE0E=Ug=c z5-=4#@X!LN@SjQ-bM{s<_PT%DPdWo-FE(Zxt6AH%g5DDx6;oT({k)Q}^mkG!Z8f~= z;4Oi-eUzD9g%sr|yTZT)8UW=Vz&_M%U50e#kz+qP}n#;d-shj-9FAt(9eUKuN*3>Wu&%XW7w;h;CL@rl$NmeaQUFP8Qf z_{AwQ`w?vq2v&Ex=LY86?=bSZt}StAZRPGuJUm2<oq({(M;}loAcMb%O{dzndj%U)_?2O= zy2@mOxDzn<_7q=~CpJQZ(-&2XIpbxud5E{i_h>nPsu_*f@4Y3QRBM22_vTe*U#+!p z@mT}N)zTVqbDq&3hB%xMiqA+OOhEsUvr) z5eYF`36o)pf3zAz|C2=FOv-xSR%qM=i)(si4U3R<8UiU zg8;uDVZ@!3Xl}!k55wEPIi0|W>u_-1Ueg-mah$-^g&FbEn0^3GrhF#j++OQ@gH-w6C}{JYjw{M40;c!i5$=iQtc9L`i+ z+@|K>k-^V7>7#H16f>NYb$W8$<_E$`wp^RUiD)Fd*yWng3sI?JSjJwNA-3C6bHYL2 z#DP+H`BSK1M)QvOdsQ&YZ0; zWR3mS8iKt5(A=mknYFf5=b~!mlz4Dzv`W zrf;-BOLgrFI?vqddET~dFb5jGTeQK_^Jftq&qDarAoM2g04ihKP}kaV(E}I*e2^B@F2iyJk)%JdJER0y$s;;lYS%6 zTrt2OS9nCR-&+W`ovtdjw(T~EAu2dd9Q3tRps<~$g4XkwW^nNZ$$=N3aTCapAGJl~ z4Dsv2oCcOan9Fs`Dq$o(3VH@X#8b!mfN@B^84ov}d^j&yVZblG|2ps<>U`0T^;6y5 z4Xn3NiBTc;Vqt+-&gv)p1zI3`GWeR9p5(yJ%?SVsRUhNb=c~{1%MiUabq4GImd^u# z6Fgue7n#es{p-q+y9Up7SVY5pVU8-bU zg5=^8y2ZJs1SJBQOaT^D=<)+#U|WLPkGIz-A<rvVPA$yss((tf>Wy6tc+gtnX^odh@pB1QWK zbabY(70(GXWnkm=V)7ggV+x_V)*RCA{>7Ym2OgfDKIcq}qxk1%!JT z3}^?<_y=Mg(u}g2k5uK>Yj>3`6g>pHQG}Ji-kH$?Qo)rtVbCH_Ym}Q1W+4=*{6vH# zE{+LsDGw03j4elhG!ZiAD$S+Kw*98hAFFeNy`>J5$BYA{%u#dplNV62lX(f|mRG0G zG7%q#gQJnKq@m-_%eQ#_4x7FqI{5P7>5fbIx)v%`xzc-R`Z_m}ZukfpI_g7L(l?6+ z)#UR*Z5#1^(J6Ur-I{_@P%ZVV2?Ps_lnan7A97ssB1_RoV7NMJAvk`iB88zi5LQToU+nMA(5-7oa3@sDVhhY zZoCxOPQ+6ca2@o)7{S!hJ5>5t4ECo1uq@kmav#C?>{05-K}@syjBn|6!qUB|3>D-4 zv#9-Lhbx8q941r35hGp*GG=SqXFjV&=*dw+W%)q&Rs#!i zeq?n|(SJ=k?B+0UZ2jP15)$w+ban-4m;sT(z@8%4lu z1Eb)dM3=+%5sdpB>|+ibW4BI@53G zdXmn98i8M*T|U%!Wv@@XKhv&3iIIM4;Vlo3FC9t@02azLI- z7d3Iu7rq{~O(tClv1S?OmUkPkoE1T@o!CCKXDe}tdu)mPBIyn=HPHjE2~Z%Af-^C9 zzxAtY7Vz}Kn~%>IE9Fu`iZVm=a8@U+5Y}DN65gFM0$!1va%8EF25=Jx2@W`-4Drq2 zkE@E!ZRJM|iQif5YU#E}>ienXYq7Oz9x3p`tWA;fgvmg(F~&53bpRK^nBoJwK1`el z%~u)dZBUy0-4EqHc)8QP4oa*&TL@grP1RBTp1^EfY%7UQvjr9~z=r$rk%PbfeWLCX zY9|=@&((GQVod(|AoIXxJ2H}U z*9rZ4ShE{N^>c%p#xd7hK;K60%`D0}Ab9NfbA9&u;M&GqNYS3j+03B20Yy{P$O>tD zR&83wsIZ74$O0O^e+iWw0*>$lv!5oP^Zi}RMy-%mxL#(&-V|uB_;B}{YuZ&_N93>C zIGi@G4&t3PFJG^8e$&jA!31B1EL3s7v+cM%$f!LBJI0fZ$fm>#RLrMMtl!{=rZcWK zu9eaPb12cxBdPGYdwVmu=;bCoK}hR81pW%pO1h|0>9$PY=~W_E70T!Qq;G**QPkZn z{Ot#51OfM5P|4NHlI|<;IzP;14D%dL3JZhASP)pK8BgE~E>c)lY)e}WKW>KrI|?y0 z*Vuv-Z17Wg_ph?8*&|fUP5LFSy5pH*lo>r{Y7$X@#~0G~kP_*coJ{aHHj{+^^EPaJ zL4UZ*vK;O#zkE0-cq@+qHQuBWC~xSccAF7l-G)N_-?quRwWISJ>faV_K=|>x(+3cT zGFCZy$-j>NnKK@N;q8+s^#+$JlP=Yx^=r!|wycF5sp>RPL9jY#D)d4C-YY}1)k2gT zCmE)GRsOH6nSoyc`BHrQSi`hu9KuzP-v#D^cu<6bAdxtS%P{VykJGMLU<=5E!<&eQ z-WV}!=eM0S0)Ig4X$=R7$;NAF|J6H-(P8yzn#_QQn8rLMGWQnd(k>x~w2u3VBr!lt zw+pmIGvSpc<>RQXkRIuC5K$+s9iZFm1IYPK1L@mn#h7P#8FWo6;@;q=z#lbCZVk&6 zPwCvrQ8=9{@nFl>jQWYhv?dgueSsu{9*I0@nofAQ*$?i?9WR~p1`)Q6!y*PfWm*QZ{TYjv-3rb=JRUUI@Qh4wE_kce5kC9&tibM zTC5Ne0~Kgqoz;MaM2n2eGX+TqGQi|akbVdOKOsbpn$EdP;22gip=y0W?~=}3CTgyc zBJfj)W4$4Ncml7a_J2owKOcj&b}Ke^|AwSLiD3pLH#U zD@;`Mv%;?sUb^(Qd#-d$-$`>!#8d;_*l_lq^-1vf$6#1^f4W1BC!CeO-j99(UUZHU z{u>oz_%Bq9jphH~?QUyHIsGRkc(aN7q24VZScXN(&LEWj;eN8@3|!h7LYDs(Gj7 z9N)@=ltaoUYdVf`3Ec1hqH3(+SGa1~W$2weB3T&c4;}PWt2vOLhfKwW;ZdrEzC1l# zoYDOosmS#KJMZQF&q#$;BjMDF7;7uRV~1ZOXNUJw6mb>=bqW0;6BOs;zo6hE2R8_J zA^d?*{^0Hj8S2Q?qhQuQES5aTIRu19GPZ8z&;dB_uI``^ZrQw1ggQDq0{RkZp{*M=wk4Z2&y`GU%gLds#~&3o+S@DPlLg z++cf(D!WZZq`%2JdY{E^#fgCXOsn>U8NB0NmDe~#1kldLc9&mfPA`Pl&HW45c|zLkw0i7>Od*BD0C( z)J)CpUu^IhSSCzN1)09#589UpTv*ROWE3)6le=;r)Xu)%{i)6Nmd)bd=;pQKbm+X7 z`?0$$*qdJdqX5g3yaxM-ODL=)|BwGs` z;`UcfCQ+3LSY?sU(VbNIgU9t>*e0-Sm*F6UBS?TLfRw;~zWs4aUcI+5_h#JCj+?A4j`5WzBn08s3x zF3gLHH-_h2on2gVA8JS%l;_7Zd z-Wo`cS>u|91P#K6a(`Wy{Gbw(8|0Wl;MiL+3e{b$ckX>bR#4bbjE(}qw#@Co-m*kp z8e(B8uxP#_85=Lmj|0RgqB7CXr7uZ~GHT|6h=ZP^i&LKAia*SOqk02H{%}Zh2w2`L z@h^aK_Rp(_xYtuxmRPndS(=N#*zK)4fv!!wVZe@OPSBfWCX1-m0<%|@<1z5yj*79JO-`8BZ1_{&>W<`xx zt$vRBu2|n*w&+2vdu-Y4Gddbvdhyfc*-Weq5@8?QHa}048}sA+BVJCbFF+kvuB|NN z{@zJO7|7OKa$dXwQdc;1&#*@^^RS23pz7%#FTM*3LFse z!xaKjJp3um%VD}&%KYHfU_^#v!L@pU7 z=8vqq@EL5C$b0iGJA-*+Q+gJ(sqMcVNq++REkyGGgF?0KV)I_zO{g(jR)CoX?25Si z3J91Y)TJ5bYOZl07K#TF3UY@=akDKT{be((wv3__@~HP_bBD|Y@;8vsS#XUcnf@vq zvJe2%OO6%wJp3^(;0kkxjZ5z#-jR)CP6Ui?1_zXLL?yC+K(q!Tvjw*vDcpl;K@VEo z^L#!87`_PuQ2D(Dv=jvx4j;MTCZ$O!L>G!p2X_bNAqFE?7JH_YcNK@j|CrZwz7Ex$ z5a+m4_+o}{qe-j0l2@q1Nd9zm^XvUc=&;Ir+s0nx^T;1I^AwZ>WAhKian*@a}LBT zwz>-|A{Jqr<4ZzIOD3lC1fhC7kW#9-%v?G#&?hZVThU)F;b#IvL}gF#nl*)Gmun_L zPwo~Nz!t&$@fJd;R9!3xA)lM%$vDT}|3M0AQ-ot)JYLc-LlQ{?umH+Ln|#QD!}jM- z=^vE^jC%I*P_qfkLn_@wDr$JZg0O514ixO#Y}$S}>YTP5J}}nxs_)cY-Xy^=@GTA( zBGT0t1B`ki)nO2dAT!#Z9>Sg}#_lbY#?-v32l z2?wrA&SKCrGh1_z9v4~Fg4~KBW3X@0PFnAESG;QDYQgsZg zPsqt>K@Bi<1AIk8w|)mU(Gn0+I68&m(bKs2*lJEkJq=%KQl_wYK!H&xB%tNJbCYp> z`f;+_15BU8HhOQQKG~~7OIi&>t%BKYW&)G_Okz6lCWcUZ(vug~Vr_eI4NLN;i+I9O z67!gyevrlVoC}z=ru)z#ZNOXahBpvR2J(#kU7UD4Nx2L)4SBs4kf_@k+|#0n5R=W9 zbeLO~f@=5(4F`w%I4kiGu@M4!2;G0u;xDs$#5Si7Xy)Eg{t@!vvH4r&{NtxguKSm? zT;HW-fvjf2QI2f9_?@AjRb;HSRgbG;E8Q#?sAFQ|-?DLdVob(9)VRVXo!B zPzqBCBtzX`aJey!G!z~d?1dZg^I#x24H<25K?;@yq1rSLi|2#UZh%)+rrzZe9+-~` z+nLV-7TQoIjv{)_(%-F|qx@UBbWy%>WftDska&*tX%HHro}*J5M@ZOo^@9&*#p#h9h+8wz3rBZ4jWJS1-~H||MRIYU(6(pZ6ROFByzyl#?U%?IKn@A>;C3)7j{xCHs!KQ+77-vI zeFEqB4fp304epczuD+`8h)0Z%kX;7MCe>}0IGQuHkX&5M-DOxR)+Y%BzwEynBD)9!wU~R_oVR>o>d=bOh)?W+q#+)HFWijITty@A zFmlut3$qZX>iAV{+KP=}@ZbA9A7k3Z?dj_SOuVqVf()($d36zOh@eDS>ab*4&ABTO z#*Bu@zXcc~(_PqjbXVAhtweInqt{6-pZKzr$3w*+b3c*WHYCVwVI(UeqeV8n!4R%- zNvOl-8uud~nYcc9F1d|qCRJX3u!qQTALqo;hOInDir#`t z+It9N>`@EFZid}=AaJ}Ol<0mvSxd|R5>wv6x4B69LuM9omG`F z!xh|2t|x*raF=Vxcp#G)ZU;r-)h!4(* zaUIQI6~qB3RNBXUv5@hzAjj)=6#%w~IkuJe=Tq7hoeqK=sf>KKxi%)<9*oFOjxBad zT4o4AVh~bG;WkReI!5oztB9zt#8$~H3G>q0ZpP-9XA?mv6772FBk}3wz-WJ2xg~yl|nN=+i+iPV2>X$#3p9FSL%SH0x z9Wx|HGnD>Xd#3Ih2&~}o&mI|A_%tw#Oruh@JZlV8Yh#bWlh02Rq{l?;8Zcq2TXXc{ zQ3iDzl)N$Yla}tuFiPGI$lK%%q<&S4*!FEZXI&L)XpVk7yiPFO(fh098}V}hCAM_P zQw#i};8W4cfGd9}1(>>AabZQnrjxA5Rt`S&hIY&oB7Oa&f`tLldh*I>3Z3)qrh^Fm zDiC7c!P7u@+rdHj1G(- z7K`?8BT2|@;4NKAh&v8xH`PXIJ|mz8<={_^En}(YjeScGon1J`d@-dg3E?G>ac}j6 z-Y2STuz6Qz2pcw5+-B-qCVlI6^{o-lfIVhVjmw;T9y=G|Cg%#OEt1--(H7&t+i|Lm z+I=^JYi_QU+An+j+P4PmKYeM1 z?|^kf@_(B)-{ef(iW4kO$emjpPk(DL!;0HnQ%s&NEPx~v; z^lr(ahPi;uM_-`L-S-h>=2#eR0UW;vQUU66=;7!`*q~K2OKM7{_5ID+<E$&k`*| zseR|H(XRSI#Tg+p8td#1nDj44+rOAJ z+dUKLr0yVn+unXhB@vxX{x%A4Jh-UF^d2;?stqQ$%|F{rR5x_Oy`hi_%yM}&W zZ=u7IwCRbzV^2ae+5l%`oA@BtBuH*HEHG1EIlz{owxUQN2>!W6a^GCQO-mta9cSY46*&gT*Vy{H? z_K@V@2&~fA@mViGN7Z}yf1_KB|AlUGuyOofGG97wo1zIm ztDP?$b6KR|OiWbZ^694Xc68SbY;PA~k#)CSUs*a%4lUxz+Be08lAoq%yA z$5}E7*jU1Td7YIPE7cVBP0D`d0| z=@YUwyg2WF9*XS$I!b0ArAsw6lc`Pqhi@IPpC4W1q9?_Un5QF)q!?{2e~ELy{tw?8 zL$rES{>0^r^elK7LXrJ@InGz3dn=Pl*4{+vjH9eismS_=?yCISMx8tg!{dQlQ;Kn^Z z-KD^i&86{!JzcsuJ{__2vPJjb_ZETihw)p!qKUg)ZB2|X_ki^9IFIb_C(%Dj18!-P z(4Kf4FHddAiip^6$jFouzbFwfoqfbP97KfiP3=2V^*(BB$zYaYdcfSj zv5*A6FVbkfp(0Z$lg%y`=@|3+(O_|zIl`uorTT@86)BySTG{dOyqd3(ROx|D<$>!B z7)W`R$cS|P(m|^xfl$3vSWHDkP{&Ok1W!0aR57-C^C} zt)$W{^G)Ko*qCua0Ct^jbrUYP@ZCYtr8(mr9-XK{Xtc;C2e~||pdj`*^jSPJR?s(# z`%qzVZ8ld$_w8l_M@t41I3UrQ804JAu%eDo;|N`zSR;P`;H(7G7c9$qh0-8556~59@DVb(THF7im z^!xzqWo?^QNbM>>x(`n`VsXZq-NyTG@8b3XqfCsCxdUXkp7P>j!pU%OJ96uD#ckZ6 zB90IF;CNhwW#V6WB(YIQtJ;f9Ek3x5bHMnpaRc@B8HQa`q)ZqsT#B-mpyPK@655Ru zPe3Q$BExc-B%dJ!0zVrQ)Y!+XYqx)wX58t)_UEINw-85)P7_;gydr5uUFXVXoOELp z;0V^B+Fn2se?|R{RiXOrF>&#}qU9gzzKTQYTQ7hwC6uHt0AgU8!h~1OoNI$d2Hiiq zud|^^-oq!P>&0IPnfRR8O@z_n=XsYXm+W7pUuVkoJmUNlGJNdN`22;*$ae0e@Zxr2 zc)#U2DpWaX_Jvr=UewUV_VgWmsf{YDj8VmPVIzl$&4xYMe`5Yj%^8>!Se>QsYj#y_ z{QG{z@>E!`KxLhTnqPKXe(IEv<+9hj>q2M%qt^)G`!fDV8C}%qF7+@d;RU_&5ih6I z3*n63--nOy|GhEZjiG|`?EXp>0IAoJgz@D+W7-F{oUDdSdJx<{hmO?S)a#;z7%t(_ z(P4CGvmVimqXKkp4^>wEfp^fq;lXwb)6A+q*_2sbG0xoT2c>@PhJ73LR8V`|?>r>| z&~JCjW8%lfk>7gCEUf(f($L;oiFDMZS!s3QP`rwi5D?D{+IX_7e%mzpI{phgcaHD> z+L0aW{CRw~l^`DLPw-{U?I<*zIeZY{9z+z!6<$UQYG6N%uHYh|s}|XK{OXHL&+^RG~_=m`ioAyGa6rrt`u-Ihq;#~V>=pGOoJCc?yAGmWAy4je{Pgs zqOhR;N89BCRsxMk!cQOHVM;>xu~CnCXj2+JsgM3f{gMPOo7ELSCw$U!G4h>IA3Co( z%cW{<0wZ!PAmnci&=cqkw<0zE^rk6Fc=98+MExaoDoT{a$zPQA`#)-&AOstLw4P~$ zp%2=R!nObYg>T`u=+IHB*VR)*-tZYhf6i(O`vpj}KSA^|;58d119X6~n??OdX36AMD?_!F_Kp!yM z4Ra46ll5N1#b}S=ev63*!J^T!&cr}3Ql>!#x6xXVl*WV2vP08>HP2kunbkfE zFOGz2EC#M5#QM-`Sy*6nsw^Pl>yPN71%z@ztZ@H^0?}m~lF~^d2 zj%Dy~utJ*w8Dnq-Wg?ZGITXhiZT|zZ0p3hWwp+ejh#!w}{Ly#@43T|IO`{8YpNgAO z254;n5O&fJkDtI~-fjo>FVF9hiaZshy)ekMAUsTci^$ut@s@L>B@2)F7Z^Hu_z-)% z!z`GF5eEWMndOqDKYI_FLpqv}R`?Zi#z_sRv`Anme4m^X^|O&4{mU9X3U-6=_TGhK zp2E(gNOs7^5&dvbk8}~{13<596_(60`S7q>^^qkLVAJ9{T>;SKij_NyW0mNzvxTyg zgn{zI=5gEMk^%^JWe%HgeStj#B7yq!H~0`9SRkHW30+Kj#%{nwle>y1HxIXPHYY?( zQoTv$6RmdR-rg0i#oDfg*fQ+%>^GTf-)Kr=!uryv>)fva3Rt!m zBoB<&gIE6vqK%S(C&kRe%b zIgAx1e)O^);9^VE*o0YPXeiq0K8>;4d!6>DIRyEd^~E(~6UQ~t#+|g`fMUM({S=;| z(Y_A2TpAzl2W4Ni#4`GdM0L$hH-7cFy4i1*)Nc;6q0SCMls%RzZ8dkwUeyz+*^kGC zNA7rCC0P#s1w65E@ssV}v>Z~R{NS^$@>~H% zMa<78aY)xhiKDpiLW9%7S-u2PzO+D)y?}LT4mp)0BVI*W=8)EnsKG)6Pz||Ex=E2{jTZ_!$!s4UBcgFeC?3%g zK7QM*VFBoYGU8Xo&g+in#!|W4+uKBvf=y#Dc=%%^aFtH;3}aQ^Yu3wM0?x#sGbyGY za*CFwsuRj&28e}pX~))7{x=K3c1BtlgwA=GTes&DI0`2C?nuYQ5L1B6TS|^nn(I5> z6(shB61vP$4eyc{moj_aOX*o!*!gDx986$BH~?W9JVs7lPziU(dKnr z)z8yej%I{OtU=8WEN&aPC}CxD2c)RH34HH zU~r*2;GH4t_`srx**`eNvs$PE9wSY=;nsaAx0^FfwiPH4_C_Invp*B?Q>PtsjlXOa zmNPYGhi$vvA(RR+9ONo^h82=D)>$gE*1H~VuYbR zt455O??vuiJFj;2KJoEI`X)`Jvg^kiqA#?)%hWO<4j36UZ$3_M_sRTIu#XrA1Jxpu z*~ofIA%x!B7{Rd47~r$}A9hnKXMLq8a-AXmDEZ~Y+L9>0hEiggII?pZ*8wznlVk-9 zpTwTV=eN~}4RwbTlnNuK${)qh>Do+ajHP+FK*S3nG~|i|CQ#dE@Oth&9I8DOV(7G^ z7lNu7_JcI=r|zvh7}BM_D4fM6nuWu8%c%cd-5eGY4)-2v`2jd2^BF^kbr^dzfx&me z_r!}EAUzjJLG2xEXeL~A8AS8CzdmWJ3ntuETEeRc^D0Bd;)L_<;Yd4}U><;3&HXsf z7&uLPdMnoTT;&5@pA^QJ9L*$~#Zte^W!vf__-eSYE%akmoXYjCsCak5$OD}C;CM++ z5+oZ_3#G++TQyX288o)SBjdI(pwfMDNbejk$;@o&*hk?#_cYpr5A4Kd1q<67KkEj8 z5UiH=2}nHD0?zS&H=KaJ>ZWY%f#X5`iH2`|TZy<=gP4&56RY7xJ#am+av`uv=Aju->>|BR2TS0e#{~XrMuhwEVMvHp2rjuF;fV=i42ToT<*F%W`Jek zN}+|QwBY)BAcg3(>Ag%+#NMO08+fTy&a0fb3QXgGcfWgt!Fgd6(#M+sJ@Re?FPhAYp`MV|FR4n>}&G+e)^q09!i|8HdBj<>j_ z*6dL8ol6%D&4Jvy=-o+-*BkQ=&351K{X>#d`znHuc2J*ExSKll_@qv7e!jw&&usD` z)W}`0$}5)LX0f^dFS-OJLXiZYxfIb7RL~AHu)v`sI-9(W1$-Oa%9I0TC*;ldmM~Bs zr7H=}#0$?4jxwRt@*BExB$-|>5Fr7tM-_gbHXhd^IC;!ast*yc($!yhO$N47LyIF}EI~T?(cRGELu#NfnHbL`)|<+sXR*mf`Zzb$Xw+X`=O#-Jx5cy_w7w zO_s2BPGpkF#y=@*OKNcKj5hUSN-3zx;5#9z9e9`=jKS<=zb1=OsopIYi4i(-=u~>x z$D3qqMG}9wklze#UnACPZ%7&XW6x5SYKkV?S&=Qe4k0eg%fDt)TU39rq&>`*JAhR( zP(%3G^hzNPbJa3!Kl;5?^tBchQD;u}SC(h8l`sXvFmAGNJBmu`G>S(}z>2@Vkr+`= zgUTL>c{A-ZS?l)cJzGO}W!m;6_?2l9Wuv?RpWfqu8M;HJ{pbB8erwruHubRcFA=kp zN%0zQ>>}aVn!E+lkophFa4xYisXznv3j&^=4xX-;P12SWie%rxe<6ut4hs=~z)~eg z_r)4vm|pV1R(TIF+liaaReO%6)(7%bre-qV7$1Y$LpNQYb4_`FNq7!1HdL#Sc~u0P zt#6U~;>~7Jes_6vy2`n?Av z&&f;zWf_~1pdocs`X~)&+G(aK_`e^{{{oU_#ohilddKu%=p8#7)BlCu{qLqk4VN6! zFf5XwkJ>xwq-Cr1q^eZLd3G}yD?Vbv9w{WOxTL}F_20BbTvnz^C;$ra8yzlhgzyhx z)sxB8^kcu6hp1S`FSK7-G_UtF)7xCHteu#H+;hXH3&-0%?TPr=u0N*E9^Gi7ckz$u zcY;w>_vuTX&h2YPWT#}ca&V)mdirtK(ed0)GwSGW9Ggb9g5c#|Jkll2(a5YM0$(m_Gjfx?qA9K8BXvQc-A{Xv_%3bk|B2kcV(cO+9 zX^z!d6pSN_v*VDl)iNaSoCWO-}Aq!X4Bf@|5r8NcZL(qhW3T^ znM9vw$0n!xN29Bz_WKVpd<=7=?tjtcn5X`!nm$wfU|_?C>k8XbKh53&f&t@M#=Y`H zSCdXJR_n|rN=2%48ueMl^$|KPIbfe7TR4s6J20&{+*qr_SzR`uj`gpf1%LN~*Rr~(B zF=SAxud;QEoV3tPQxoqIuAKY$NC+?H1CLWyc&_4l6g)Y|VTmE%j$1)ykCu{)H4A1h zu>itF*wTNTnV~6XZT@+)QHotCLl*Tt|2hNeM|^^E_ENFHsW4v9yqW8IOwgX6Ie!Y^ zb+>F*mLE*b!Lx!Ej!36tDa%04JMd233$mo-o!&jZ5~MomqNH^8>O|leVFmvPn&kQ8 zBd2M#ktfTIobb>LEUGOu>YdjL841~FCY8xhmZPx*PPJC{MPTmt;0iFi9gM)opPZqp6LCkb(ym?%87ktn^O@Cd^h}eX~{%_;lMqA3b4nQL45>3##aMO4R6I zWUQm9bo?k;Oq|4bvT*dT#u4GRnplpQf#{X3FEnI8B^N;0S5ca_omhg<^YRquW1p(L zWA@>cnKQb%I$S*JUi6kZaNDsA*i00S-{}jWm7sAU)3JQKvO#TPoF(zf9{zVQYOS=1 z40VTjNIU=*B3rY-z(Q^LH_C;h+yAUZdo`UH z-{#3OC(-Abo8xNh)Tagy!NjmswQ7cuN;jOl^JMS$)YywDmf$_xCrm3@&l9ins zbLd}XnPGaXR%(upj*=5cw#>xD#D=OW<7-rI`vdNxun|k!0fMNW^B{4ueF{3wbA2>g zf|9GuQRJ0#3ky-4+0oMY7Y`C8MHCy%)WO?n(cfLUa7+)-@q;RprWdogd?0Jafa640 zoIFVP8}O*I+&5J~+7LCwS3u$*(7R(=CKg#O0qCB(U>fN+gkcb4>Iq{x(S_>;$U`NC%N{1>%+oaLjl^6bX+KwF5h}^IlM-g^2v2S8FOA!* zoZ#>`LD2p>$k|ip5MqQ-|6GeB!6b>@n~ss=8{yTG2OO!1y2Yp&SwtK|*8!;W>=CeE zbRL6YPZH1Or>P0cN@T%!C8hcL8Jr)Uo1jb_fv<`RuaD^611PThG;g?v4L=1UzkCP{ zuF{buf3(s7Wa$vh9uflaUD=XnB3trQicyHKUZ~Ol2qnY-3ypCv2s&wX8i&B7xeiGc zR01eS#$}NU!_|hL)0F?5a$sX@51J!d*wE1F4kkLPD+J~UNdB{=Z>M?u7P%A~U~j!s zk7ucW(;ISSbDhp_9%l@^XUd^3X~#CtV-HXwphuMp#!nX%I>*USH>P%0+?X84hlcg?^Q+QKv?2X}HU~U0JZ$=yLCB)Av(WY)^ zuSIjN5pq74HAo$J*2jsiYySmVT{k^p)6@~kq_xBhAS7Qwu>gy8+BIS(o3FP0_^_1I ziGZj6kp;Ua*Dk%i+!#*QHeRI2Go++1qR$B89w?6Y>CaCHB5|yvRF(1!8_2@ymKSv0 zGt4R9y@M}En1x3C)}!uQIoAy(g=IHy+IK-_;pOA*>E!Lq9L$B!vps@s{5?UjP-6Zw zA;F25x#VV!SY?zhd1=Y`+1iywZ7~BiK!@t$mTy_F~f$RPq z*4wJZ4+n*PW=7x-!kn}b!%R=U$F&IoP{0S!wQq|J%(4fCKI=d5?VoN1KtC7tlMqBr zPx&}}as#If*A$13Hr|JAhY$+VLIAk#St(+$$D4$O?{ksB%9KB-HFWv??SH zPAzfbt@Zo!Q4@{KbGVa8LvFBUkAG(3CI5lpDqEjr2QS%o*?P|d1X%A7LA3>kRXRf; z41tVkHZWF$bKQ}iw`Pm^|PUk-kt)Iw1#?KW{c1GVmb9M{B9IUn} zChV;j|9TPDM(z>EU+`XbkfjuEbq~{9Ym~v5;sYgDN-#S>?aBb_0m8n% z$vMY^Fwn3jpOvMQH8$abx#y2=9DXIZ*vfm_3lPa9Qx_U@d3AL#NN=U$8%cLqP$o}M z9kj)Sc+%vvBanW|3hN%hH8?&%-Ul0stuJ& z;C%h(GD-U~oFyYsd?=$TIoxG$(U+YWTrEWmD>MYa@ZG46W!AqMBGbJ8-_yI|lE)Oa zJU+pQ-L+U}7{;GLdCzU8OsvISA5SC^>b?P|s)fX3bBn%xNvq}R>g~YJ z0XeP(9`$oXvd|T$$wz0`M@m{7fl_>KBM#)c@()b}wxVv9grmzEv?+}xLQ(;U@Soe` zd)y$W*6Ui;(nCeu#6y38B{;xX(TP1x(VB?Yjx`WM?4r@mxKNW5~fJ|FY+ z$tA#fv~l^L;C?2pFyEj5v?ruLo?pB-5RA{y-cIMuc*IO8KHi~@SeA(;Eepm5N2+RF zgEpLIS3H^ui9yV@sXbkZLQ3zY@1)Sg!)jH&@12h?!W9E#BOO(IVoC#)}p_zF^~VKkQ64cPJvQ~uG-_DSA*5e zbX;kw(Tb&Vb%qrJZx56pV<(|4A%6Qx`+JDQ1)mlkd;tgR5m9Q#tq`az-xB{F4yjh2 zuK3Wz_!BI{8){d*-jqyd=q|SfI2h(mY;V>&JW&M2|3My2kqq~lV4t#dTTWb3VEB%i(8g@MWN7xHc3H^LEcz{HsK1i ziRfe^k>BKxwXOk$;L%ZA4{=vj9SP_WGtQG7Dnq2nMqhNj+K5bykKDg~79$i-8H`3= z^gCyT_MUa3Vh1Q?Dp?0L-|XT3La;p5^M?v;6|20lsSwrTP6UMrajKm&93|?o{<&4n+g8y-y>Jmyjh`!8 zuWW%DbM;VHk$o+w{3cMMS(c2AEP-|kMdXVNP7=1Yv~-7}QKaI~Y38J6uM|_16H!Fu zzl}q5-{T85Qy>r`PFJc1MPOou`8Q?RZSi;*`9MKMMof#q!*F$#(NSxmm<3PM2V%$^ zQkV8bb*!(R+|I70cz4pW*_VoPE}flI+Y`iki$A|O0co<<#%>cERvIDfb%X3s-JcF* z#pT#4S01h4>?rJK2CX9H)ngZ2h%OsWzFSom9Vl_1^n#8nUE(A#ASo{;w^XW+2SqD+ zH}#TTG#`1Q7OBDOAkL(4S>wK6>{MO{Y+&c-pDkqnM~V?=+@r)=;72M(uCys&WKO@s zL^%vHe2tR~eI}|_=pN3@QwdcaSp}|L51u|HKD?p1Hv$O4WdC|Ea+R0AuzcH3{XNWAs=!fatniEzwnm= zs`wq|0sFr3x_)#$*o+P}`gxH}S$$jI8KBHQnF9hj=EBvIoP=CuNu^q^gSI&2EIS6H zY{uSqAYqh|Z-9TL;UFV6l3tkRuM?3C?fK)rHM*%Z(81^xr*6do3$#5Wx;7GE$t%UZ z*1>9>?f$2`4s@3A&5DRPwESb?B{7(m3kbdki6&>>9@ZI<{2~DZI%0+^`>V}>1`1$^ zc9$vC`5r(@FM@23TfOB9)jnrtf5l`BX(B;bp&Dq8G_Ot?tf$(+A7uwq;-|0^TN{H~~!)rzn2t21r4H9tMB z$P_Rd!ThLg;R&`-tEs^08Yy#dz=Jms zNMh(&_eX<%dtgAG;=}AVD{!fLwJ;#sJKd1c^!ocmD7>%25^Vb%nuC1aJabf_&X2?il zggXCbpas?Iwo|?ZPM)qx(0046^KZ2s%(J1%l>7x_MF?r3vW2#Dy`o+OvV7$CWX52- z2ZXKzhjt-*9E{tsXWkJ$H^S3_7;1m5(@_k4W%W~hRK zhrD@PquZABL>~iW@D-Z@Sf5%>$#?8g3TEHCUHYCo_DyAFNx>4sV z-n`C^yv-J5=P8z@Us}XJ-_w+WZ(c7cU`bV)2$Wb>d!eobC-ZjH@-kU3%!A2x8=H_w zEgD>;$;1kHTtx6I%cKvJy)BZe>84{anSMH-M||=tQslIrehoWoA~w z7VHL5$NNCiWLT-8p*Vk)eZw!^pXE{2N;8&Ow!b;$T0*!6k{LosHZ~_UGMV|IqcV+D}nW?Dr!Ixq3rcVFo`u{nK^&Y2Xo~|t7gWHW~ARh zZ{@Tiw>H7&W%%NT=8JiAdbTcJcUQ9GF>-2zwd>>Kf1D8K$a}$Q=~HA%=j+SI;oZI7 z-*2D)KF)gcldv8WcjvSqx zgM~Q+B8>ZOir-xA)1_yQ(JaI4;>QU=3F0HPR-4g6(^97~4Mw|CkMMxx`VZ9$FSpw4 zgzIvvIAyC;SY^C23d%9?VNyR4;ymamT?f0VX9sCrC{{A-G zJ81U}6m4<(Z|gF)`};++mW>9z_v2j<>({!=$Z7V6lME;?U>X&JG@Suo(a-|qsv0*B zG_8F4^V$&ooTE^*HleIv8fw9h{l9x@v)r3-7$1EJEu88gq}~Sk=1?w$>gG>!0R3>zyOt-g1$Lxfywgof_iZw(Hi` z_++Zvt8RcTpsh7tm~^1cC#ia})22bIyURAij^Q4s!7H|nbp8qh5~fa<+4}=XDX^vr zd~vMuW|`ggq64R#YpyfnXtvh1S!OMl*ek=ql8YJFrjsixyS zS}vRB%fHtS%gHPBE{h3FWs73Z-=bp=Zziq5wSgWi?Y+J7YSS%am7xBpK}o`GgS$g9 zJ9tB4>6AFZMeUnPGvd9)glCo9g15UE7sQ=FXPfuXqr2 z>~i2{NnYi37z)kAfE<^sPxDRr6eK;4#sTb)EDePP{FAbJ#Fiu6u52( zxZ1?>W`E!z#`!2pj!9Haq!a(ICs9rxTIDT2Gu#I}=Wqj?3*#UT61=Yj*o0$Q<~)8X zBneqOu&pSQaI4xxL^i>7yUep=ZoAd`ayX$4i9fW_R1tC?OV215)wz||0_9fueUM@d zL2d}EHXh1fH+zctm}u{xgUEiEkzoBRv zly9-Ri-twT*M#@q78cR0Nf$jd9K%I%ovaCcLAfmod2(t9U$@8C#*G((_v~c-k9!*0 zazy9&{-%t$zL-36*H;yeN99qw#sfj_Z2#mY$e<-Ja^Sog6Ruzmo+dwnrDragKIj@S zK+FJ$5e7rM;K(w&2}m0t$TSKp{)V*~@beIU#EL{wh>0ls-Wl4_3Zc6by~%Ti^o){& z&t7$CO(sB0FmO*?dQfd(vrOe4_uU?kx5{%P8lcE(X~jmQft=TF0(5QG%9n!cUm*U? z;P6P9rjXw#jpp15GehgM0$o~iQ*+V0hK-l6RHQS}?0`GO?T}fHZ9@rb2 zB0U3ijK$P8x=$6nfh?d23ladJw}Ir{HMDIm)n*9XuuRfuJaNr4?t_7z0#@t8^Q$s; z%QKFVgfT#MmJtAwhHkO~)sALY1+JU{?I`!q3qU><{QEu*R47K@Fgg*D@FZX-=zYp9 z{=lS#Q*bwry4zZi2KUZdUu-=3L-n5A_fia+eB9ZN2JQUqv*e##){X-V>1Y#e&58w@ zz%)I&X_|X%Omi9rkg(bfhkGK733ycFao)qd*OwI!w{2?Qa0%$$m7co%<@rPp4OcVC z4RMerXbDEQ2y0tz03$Ny&I0#zuQ`s z)mxnh2YqkqteLo^E3%wt6r`2j2}4$VN?Oe@q=T&7tjXY~7p~&QL+J+c3RrvdDQGUb zV@l@cZ>W@K1wp9dQ#Q%E{-=o&oS%?Szd%GPOqh>7zp)1Pa4W62bEt~eQ()}bSTMV( zEP#uI-ec{Gsf9a(kioF}%DN}bl4tH(mmh+hQ{CkSXe7M@Ln^bd>ldG(n`5!)#~Wtd zrLak=Jw|;IVrbn%35S-_@YbvGOmxxSyr3pd;)DTD;oT)Az5R=SlWVG{XgP){pX+@m z0&3B^{zD~4b~MhVFj8BT1tLU;Wj2)nB%8J|*V*b6mGmn*M;aTL_0AQhp02}}73Wch z3wl3an^Gn@5ZHa4@&Mj#eFP!gfe{6h6_ldOr=!Nl!pkKP=5vn^eX*N_iUn-|8gr`3 zR5y69}5!o&~JmHYB7H->}K*D zNQy!Xa@XeGlF1*(>1o1UOgk$BJ&=sZwjLuE^r1q^Q$#)LV_dlsLro0Xoi=Ft)TLtJJML= zzsuEPzN#AV+Zv*kxDv2A+%KGUbcY;9PhYLwvyDF73Oj!F;O9jZ_}^KFW?rO;aN+}; zMO8>Ngi&Ra{8X9XpqPUD(}6uSaDE+f$a(;49}(WAGOzUatu_BLPM8F+j0fW|JG}LM z{m84n{jAi>WtsCwD*q|-atxw!92Fr=0bd$odqs+QJY06z@;dQAouPU!VKo>VhZS9v zHf*_K?vd@igJ%UOaTn-fg>oZUyl8mk_%j7tr@n!XCTsXpG15HM7Bg_?KN2~^RKOWP zcbwLHHC)sKn5~ugkOqb*gJUY4(^fT5NH!?@VHhNmE#p$4PNb%>nv1K#vPXUM3~6N% z&7i5tluqw=yrF{M01#c9!z$1lLOIt?a_5x-cEYt9pLf@2??EXUMur~RYVyefIs;y~ zUv#)Ik`sfJwfcy^_JLF)JPUzAX|p1sl}d6r(Kj4*5qdlM<9=H(GWv51m(A^-B#;?e zrr!nm4MrZi32;01<&QmCs=$u}WkmLjfg|;va>pI1bb*e2{axMd6W zaB^uo2?IBx(ix#RvpR-3aGL3vl8@$cC2#PimMG#ycAG>oC9hJBtwL!}7CZ)daYfH1 zx@dZJSw$Z%YWM-m5;XZ}T4(BZ7Bpfb%p|l4YDEqWFCqjgL_@ff2D{)Q(v{eH;LOB# zou(d@)Fxh4*vFbv? zj8IbzfHjhlKhmQ5fkD+&T{umD5lb{doy5T?+j@6+?$sxORAQoyjcfc7^@4PHeXnrG z6j~KUB;Oq0&!~lTC_n6{hrr54sxTV5TF7hxno#ZHyZtdYkCJH?KledK1hV z4&s=aHm;AA&Zgu^SUD?=g9AN+?@^)+ul0v&HElg}ir<(Z;e-&|B$?47FrYU=F!_Fv znZjW%tE2T7nwfaqE003J5wk`+(_HcwR|o2M;1v3cEIEI9P=}K8#UqIz_+r+o^=DBM zKA~Mey-!QoDO>+%5&M`HQ!?o@{!wt74*$cf6$U*jXJA>*N6T9_l{3Axd;vAR@8RY= zvcT3!9c?hVru|Bu@=uxwf_`0vtR;NJ-RCpgqoEd8okxUwZ}QRDN)+)lga6`e=3MjN zc&_Zw*%LgluCT?JeoWb&sNw|_L->$umj4CHrP!rgEwkJJZ^x6nhN1Dm1<%y28 zBxMv*4+17jG5f=RHAWS579*a`Uc~Ix<7mQEAIk3;)_S9BPIKyAWvhAszzh$25_c%x50Yr~Juez-mM@pM z34tFa5b`t|C=7JgiBzZMq7Oz>W^VcZ`nxt~)5gX$@$k2KhCz~3kzQc1CO3U0SJ)V_ zU44cA7=~Vg=5H2Sjid!Y6^df)pcwI>cQuA>#dW z9iY~h`Y0Oi%I==v@GbxNa83$a{Ttj0v#UcnVlT-wRIU>me@otIcuYyNwPP zZLHz9W9P||zJ}g;fmjlLJnehDDUFi~g#=8)B~-sml+;@-W=BkjoGw2bbFQXo*uD zmN1ko7{((Iz8jT{tJ&=lL^)xN~9fGp}qr6=<>!Z4&edv;-vLZB6`rsf6y z=(cumH=l1PM$je;2?`B*1eo>G9WX>&YzkStL)<>ye5i&vJ*Zz^K?G9ZC572bsy-uX z@rn3sED3y-QhQ5F6_E!k$_eO=wu~SP+<(?p2-c($lbz*oQwXxmfU^gM#R?3J#~0vO zz1WFY_X z`iE%pqKaK~=Oq;+ItV%kFf3_gG={P5O0LNiB64HT_a+@T_h}WG&fW_VYn7sZanJ|U znxL=BwwpI~Tlq7u7#3b}AU$-Czf3@?L;G^x`mCNN0=`J4>Ka`+fR`~qs$ohgd(Oph z#p|AB-$m6*8${uhMi!}~j4-JcdMHJTd#N`!@6|<-`YF}(2vV-_=9d+DX=n}{M7wJu zH8O(oyez_dXQ}29tJ?6+V82F`y3SfI!Yw^O22qOcwByPPg}3k6(WA!@;rb3-^!8qj zf9z@Nm~(ZUa(3H`6dzQFW_PozS6zHROaZ{)p&0iJ0{wz)F@YuV&BJ|N14+6A#U`F~ zyZTa8D&DtUvOMqe4$)6nu;h<`M4xv1Mp7!nYTomf3$XWBvAHOC$!MnaE3goh!)Al7 z%Vz^!54v-%@B0Zn($vr|R8V;n2`?HUNz>Pk@NWblMwqc!g+>~_Pax!ITattd}%c0c!<;bt7SFV`zDyq%Xp@AE*N}!k9!;i230hN8H zM?i;i(t!I(UlTVJH7B^C7e!moF7IM-tEgi9hhp1*dOzjd#3QwZ6VBZ_nDO z+0l`8UMl04-Rq?0y;!B<^{2#b*sbVqI{xJI+h7VuqvrCK%C`Uk(sH!a{g<-}`s>jo zRDtGHY82Iax>^yK>+H4@F|(DZhHaI~qa%0T1aY%pxV7h2_+g21V-J?kBE=IoIWzRJ zJwF#Rwht^t-)E4WvVnrtvdjWnxa&)AL;s$pKP=wF@T4q2qsC?QCG;rZGjWc7MywQg zKBxruDqjw~^_VQOEVsIWC%DQyWO6T>_wPwtAJmegrLnGUTI7ltO3x?1WDD}X%i32r z8sLor4eIuEx+|u%bk!PxBUQC-jrfy|rG71o$CDZMmZuu2+GUb!xwA<*mFa7Sul;t=s)_RK zq)atFDr-mC#2gfysFatL81rk^xNf0EeY^O%?9)`GHz$h|4<#7}iy(hk*$*>ynJuhf zHBZ`HU|_u`EoPKNKZ&o(Fjixt(^ihNnTNe}a&Ya|Fia>CNs73%dQA=)N1oU^-JhNx z?)AmR6~fU_s;8|PM0nyQ>NeNJdP#r|4K^j!^W>}^Vs@bp+BD2P>*H^xcy08sk~d{ z%saU)Xq5yBl4DZ>K#s{WjwH`SCvHcBvz4AxDRxzD+q`j!stv2Fl+Y=P=63o$5rfxs z6!JxXygaixvrYQ=T-lFY&aL0())#P}tb$sBXu1g=*xw+%Re&m9u>|zU_tJ0OTd`v? z;}ceU*vbWJaUNC=sRecJrsou|*NMH;jXW;gC49*jSvgz{(5)#+;D$z$u!ht2t@PLz z+xq%n&r9(Dox(~tN~#3C)yQ-X1m1l1 z%E^GdgJkDTuu)h;PRt=z9=M;M4P8WYS2dVps-NR$3rL~0u3P*K_Z(920j;I zGImu#Eju;D39^jP!e|Z0Ur_Eh0Z`_3nG*CcI1bo9=JBz%ijgPQC??WLv0=rAz=jxU zQglyjlZyHcob+VW1B@vH{@7@8S(+hcX2T)1a(upfqG=$|KJ;|N{w%Er8u`0oKS1qs z#n6Re1C-(b(KD@m7VrTHLrLNWMSxRqP3CDqTx@)Y&Cp-B1+->#nH_ZKo7A*iFK`@< zKoHE}E6@EN`BO&S_@YXPd@IIF1Fg$!gJl8r)vjy;@y-2~5ee>>0BHM->3H#cto25n z3D9uqU&7@?d4xWa>#TkRRqSGSjaFM3u!zTtA0{xxXo-rPN=7rZVL1_~Kz3xE==pW> zSZi|I-y<7i=(;(S2Tw82&j(q_vh*rDHn!@bgSw;&hswNgv=*J6zK}W@z2f=z2#~)! zGImC5b5uV9wjd9a(PcS>e_1esS#bM>XV33%6yS4c3>qI_`1?c^14$dw#nf53u-q``K!!o9Og zZaTfUZ*72d&Y~`ouDuxnQIUxp16;~^ZK_Ygj@Y~nVTi)2I!Ac3DB$auvc%6)0z=;=g`1RaK=`&%BAjsboZ1O2D3XFE=mx!+71hak^5x2<>2X)n${HksoXJX3gd3L>Pf07oP2NU29LoA}@Ko zx0_7&g>`MFno}THMSHtz@)@=Ef3GfqJKE{o*l^&xXVS!Nu!Q>{GojoYG&@yHZU|YK zu&w4KGOs}9dQ9u$E$40_Q!Ut55e)2eF95RpVbN*!CodZt|?OP^@sazHIsdum7zw$bOpY-dTYm zR>tZ-W^gODI3LkC+)+USM%*9t>PN{jcTmEl)Ab+{`n&CT1(m}cB~kLIrKZN!#4b}o zf>3Q4S)!&cZNu__~~byLJ)c!`HO6==Le6sXB? zvsK$h{1IhH;Hx0LB`c*- z1F2AwoUf3klS-q%R#k#24~qSf1O(P~!9k2t znmWzAIgfvkG`CLTHNQ@K>V!;iz@ngj44)uHtc>yuAkmA^1{O~B6U+f}H$?H5vP(K2 zJ--n1fsnkpVZan@hJUzOBE0ZW!OZPw9J0Dfkycnl5YC+za}deKJsbAI>p|FWYb6;%ZTwzTHD(aw zf*37VpR{u2idFoZKvOYW*mTXc&59{)mhZs+l$VF`QI$^I4fw z!ZoFR*u>(Rh&%MJM%SCuO$fCn$P~D@{@>GOR&3@mZwC}svvCIbb`uHCZGM|3-v}Ef zdEC4_oB5NUMJml#;e{!mS55kf)ON9*cUK=Un)zfykB7mtG#UmaY`uB8a0J8@VH!>9 zOb2N&ynYsc_(kp6vMt__bb|>0%mp@aH-UM|d(b+aw$7T8b`4=x>M;0+@gM^y(Xmin z@eNtBNI6qu(bjUwrmc}(IdBXtN%p1L{aFWX$}QZ{(zy}4cVDP?0UaFd{VMA`|7z~7 z`a$FIMbL0QoT@B>(pPHM-AO!Xhibhoo1o$?H|Z(S<77a-kw?5sJcYCa>_cUTl?hV# zg2J%EiKNp}JcPLq!pv?T{2sR2It@QjwzM|OUmjSCkxwsp5Njh{b>N@>>_5l5@|w3d z+9tGrnE+WAHP8ix-RgYC);m`yBTvnimva_!#ufJ5WlGd(%y%eP4P;8~tPm0u(na3D z`1Ty)En;;_E0c~hg`sWr92Iz&%g5b{*#)?9F@@`|Idka!`;jh+9xkftonUMxww#Ph8@iMlGnw<8z#~2ZIE;IePM~ug^`SCDe_4r?p*8mQ zNC#iNt|CQa83=+KwdK1fU!MCf+M;D$EvKli!5-L{Ivd{{?-X&hHOQ_ma+c zi8o3f$^3ax(qWO9^b-8PXpNG2O5?Z}@TnvLJZ=hvh+v zARnCjql|`Tqm>oo^14l$*^z;Uh-Z*AM@O^dYqW@y)p|JgmF_s;k|$X_BG_aQfvZY| z-Qu;5GinW{6vo3vB9&*dU}pZhQ?x`5l|;L^RRyo%Fc)HQvYgGd8y?eFB4MuftJ?=g5_CW zr;5@WQ-zyXEf`;qG%7w)ar-B6BP#ypHc2FVm4020C!tZ4CL!oxP2?ZBog`dY^%UH_ zuE$^OC~P5e~CsRPbL{Dvo}rNmdmKZ=ZYUiOg%}TGn?poThQ}Sz|nfh-I+)d znn>QwMqPq`lPox4*95ZrUNrMV6lT2eg<_)@QXY_ij(pM=+e78}*za zys|YZ?_TwEXFQl~X3@Don2B{4xg*meF%bX zk`v*sec}l5pM>MoRL`SLy6@Ilx-6K-;@|tq*%J=oLIpnE!s>T*~j z%lCQ{k>qrQIGLNQ&G6;_7M1PBMwg~~kku5pgfErfH>5Y84u2O=*N&g$Qdtb7X7c+! z*R=-zbu<%$+C{mz4pdh1%5o@tY+a!t}G6abonb*{=)yGr>JJ%3^V3t|H_^C zLI!mmyLQ~sH}rxdKM&VY=Z`-F$g%A2(nr~6AYnb-mLPrG?=c&=qnCPvBGQ07V&ox9 z$&_pmzFvwTyi9mD*!EfTj$`R$tzE6X-)z(~I0f*jTdEtJ5Zv%{Z$oYgvVYO>B15#J z2BEnk!kldU;U&U06oOM`bB~fbNuPQz@#4g7l1un!haVfMS$}(!c10Tor+s_o9y_z-r}Emb(DzC)d$nwe}$7kuAPyS+lB*xQ8r)Vo$}k5{hmGW zAWECy`W|xO1qOzawG11*WQ$9>H~WWgf__om%D@d22Vu)7;{p&Bh{fBnu&9Li)%AI6 ziI9$|KwK;&lhxynbJZMwE!`#0Rvh8D?<+N_?!IZ|GdpCsxHuq8w5 z1L}Nm#uvCdY$GNxpt~EjVdTN%z9>xWZmx^=(X@u1-!|r^5)Krai5l9^`+Ei~2n&M` ze(VQh28%;_60}SY(srs`w^=ksBzKUQ@FB#%haEZIcJ|sykb!5iFWVkvh?m!O@wbce zgk#zEwy7OB7JcT505^X!KI8cuYt)#Z?$M}Shi&+i=#v|CvWB01tJ=63ct&aEdAX#0 zQA+eV^0!Q1OPIp2J3(!tP@Q&%%=(h{)xgQ3dn&42tHibx%4gt2YVhwTJH7AkEES*2 zILD$yDYJv>lZ{K)ZjY~vbvV@C&R=Gg!;RFqdLnEtF1lza6Y5Z1S!>g^Im={}1Y21f zA2vOi2vtKr;xvONb>TZy!G+^m!=}21%>DPmZxH!mA^ZQ9bh7{Nq?46{i~0Z3O_X)o z?E+U{RME|Vmid9IZY(DjF9ejoHS(^b9 zYLaPf$6lR=t0jO2fZp64O@e-&)5^Zk4u22q932y8SHt`A?K-=mf6;O@a&SP@))Zy6 z4_W4$b38+Q!SOfJBguuQ&(kOTnzm(!xVSjGxLm*eI5u}5kKX>NC#vDNUmsujci+Ml zVK>i3DE+6NDDnU6iCZ(@9{0AjJV}@}q83xObv%D4sc3WQOcF~rsXQIAb0wK@w!X){ ze)q&HHB|09^yCH5R&F#@W->ewN};_E#oF0r?pyXq?kRE8v(ntxwB6HXHC{wSFHO`J zSWb4lvF0+kk2D*M#>uwB>pD!jd$Q_;tC+BsE@lSh!sk)&yu((09{^3U-@;U=6l_^L z?rx$ zWOV;5L-ZW#BDZ*xzuYp+>=V=x@(ryt^tM7^`qz&(t=zXDSKLh3H-P<0>lCnf##=AR z8B&}KZ*69L(_tX0^EBs44TIWoWd-{tCZ@zWyJ|b{ENjpnz$-J|>eep3^SRbN^bhd+ zQNVcShc*&j5Ww{ExR9qjGJ;HJ%A?WF-yD=xoc9{xwVc=JEVa&Im(YX0w`&wcEEONpMql9a=+&N;T;Q|!Gg7Qe>3*6v+c@@VXT8rH z2bfDuA8Z=RauH&@b64&kYdn!d^~MX+T<6LWBY)+yP7ih7+T$-p{!E2O(}LURvZD5}_Z`lJ(U}js_%&@-I>Y zOfUFALs#y|UH{K=WgjnkK~CDnVof=VtWHULj?eY8_%c$KCq$qWxvplEKuMh((kbpW zaz(__W}Gv9PX!d@j8D|}UN%9hCiQ4~lUj#XBXY9DLVE}4(hWHtv`+31c%z}SVbE5e zyj5H-C?b@5Q7Hw`CJ?*3Zvh4GLQ*fXF{FASS=T)S37_uM2-zNa8;B#iGt)jzTQ2uL zjS2m;P0ueb#HOg`Fg12$3Yk{F8@#;FkBB zQIuSd;Nlijfqnwnl9_kt_<(HZA{p?>4aNW(Tpjl89iiXW-T9T=xNLj%CaEyVC{aDy%k+fYgd6Bkkek(iog(F zvBC)pn`vsTTN7CUbJ=&o&+B?bP6%I(gvd}0^f=EqB;HcWV-e)_hf8V_nCaj4JrbhY zijzj(hsDFpKvu3#8zkr;3#O?R3t3vmn?}da!8_N@S*?2 z0(@mMr;Cj#u1QizaSGZN{SKECdIJ1w2bBAQ+tpa1)6jfR%#_ ztW`{{R=jzB{Ex(=NHe7A4lLOfqS(3no6ASP(}+&eJ;{I?8Lh65*!y{e|0V!29QGu6 zsDcL!JL`<%k;3?% zf-N+U~?>~S%_1V|NcgE;%5NunwpLqAj#WzXspC8fIOPW02MX~ zD$oq!KpYPEc~c1!lWuT6cDpdI^p$p<5$BzzdKJCuEq!+dfy)(<7E617H1LuLp=JxH z=O3-RbM|o$@T4FcWMpS@Ns<*^9YU}JfqwE9&4W?8cxzfbcYV5+%^ucPJ2Wr6$G%0c zy5enqey&(j(anhaU46N~o&h8aXSd{6uf1~4DYtq^-LpjH>wQ_NBI0#VqJinN#j~PH z(w0d#12$|$6t8AMG@PbTQ*t;gNMs(={7he#Uw>~zQpKVD;ZT2`IAxAZ;A6UZUn9vb z5?+ZUML3TURs#0QQFU(|$_-|L`1(r-5c}h*CpNt>DdR9_077`gRQ$!Qhk+T?s`f-6 zvuu@kx{(=Vy5B$QytW&3*x=hvUE*JsnLv1A7O<6G`iEOBRN~Xp*mDiC^G8uGNG@5e zK>n$%VhR~Ez+S*n1b%J{nmPT7%yYPNM=TeKpmSHTfKSgG?;HbH^ zx-OeZ)nyDG1GLV3C;#iqjx_W=aJ7YT48aI(!_YY&kWfmbJJSgSoTX_jaE|cv%uyST zI4L$hdwwyEl<}pPF{6=KFzu2qThMMVp{WjClkcj-SVtdtM#}{G0_rK)lOfd%7g%-v%*(VVw?nA76W17u@bKvU+amd zrb5_FYTd&WM1UD2Hvdtd8ll-iCdY%iBDlw&;1PJTqT2kP-85|Pm}dj7P`4O}J}>7# z_uH!bmx{}@EU3uJ6y1xxd$h~x+JsPb4Q-(D1xF=fys&B$?G)4R)G+0vg9hEd?gfrlC5{l{KaeHMUnLJmb*x};p>51LgndfbM4DJI>*yCf4@^5V2SM*uZ=kZ$hTl<20 za}^mWmzW76Oxj1$oFXX?9FX??Fbr2tWZ3f^suLJL(umJz7e!_*uE6eZ&bxn;AB+r= zrKU+YDG}^YQa-dF1jDCx(N)#Qbifi7ug~`CJda3a!I_v2k133`aU!n)@pJkL`Ag;I zgST|gO(p-c%ux?^g;_zfDGN8_RTH(z7*0IB=N7GKON(b;fx*eU6dC)FSP138BOU>E zbO{Sno^8pxzWvS1y;d#fkb4{b{59zrC6&mYf}cHV7BAUd-dVG~AhL*Y;X@Efty#bK zQhcw_C|;$UhlaZxAkV~>NPRu2G>&=;Stn}AMV_;x>$S>T*of#Co;VDGnL7#-qe|&V zX~fh&1O^wT-}hWySCkYBJUh!s=OM!zsnmy8HTf~ zf3oc;0D*O8BauIqDlG1I_r=>(Qq&xGSbnaJ_0&_-VMck3aMbyz2+7a*B)_r0V!Wsb zrC(?Io!pqzHF15Xc<;6`fx)wnoJFwt@h&mK;=iqUs2tOXUIbs42M^p;P}%x==LGE&PFr6$$iHBFU>y_g0%iHaaA5UNK=Ck81VkroV0a1tI2( zM<85>8CeBp;fuDQgoKwO{&&VC3D54gGaci+4f=`|wy&3XRyoysN zAqEl*E7w&tsNz>qT%^uxzKik2n{m}Qz}KiQHCfXwdMm+RkB`95D3yepCM?O3NTcQ+ z%3WBu2V+(QLX|u?vaC(Oe_>BAb4lyPk(EuB4;EbCJFiUlV^aIbal|Zi8V>+3^6y_@MWOHQ>z_*U>OF6>Aa!5Kib-c4ggvxo`gQhkX4Yl>aB<);I4>CiGB zSK-6~pu#D5zxo)LddP2zQ%km5=II-qF6c#G{mp_|dGDyL5R$`hg_YSdbZyit_vvf= zZ^KM)%Lt&%y;rWKfmuOm(Zltdj#{c5xfAq9^>E+wADY3kz?I%!6|^AfwPm1b@<_Y2 z_r2P=gSq)45{OH-lek(6So>uyZ1RtNwC*EJ3F80}VG|WdIrYi!rDRLU3NW^ONzAg$ zReo2f4$bbWymaS&kA|Gz7`~JD5|009gJQu55HKArXm%KJZsdPw+PbNp8k$uQ{wXhG0^yw zwg@$Mc!Ev-29}W}QnEQVJFB`=ulBWk@ z{O@H+OiH=COlx09$**S}whpA$*`72Z_>`{=vJSS))rSL<>T#XO++3zF(Oi9MA2L{*6)s+xfpawWPD8y0`@h0dl!sHPS3XGLJ+5M~sfs1~Ox9;8!$LQ$*XV4QpFi;=V=3y-p`Q$gg00fE^F zMLw=_vt8B*rFA`3TDIEl55U^LY)sG`N53Lm@H<>YPJH^5L00Q9q}ijgzfOkbZQoVF z`@MyP$Ca}arP|kJ31P#QFENQPWC)>cBVnx1z$DY24mi|f?%?g!S9PJt%->&5BsxhI z-&(+o>u(d+zw5Y?4lPFOJquwSP(zZNv>ZYktQWi*=L1! zUp9izkt*z=Zc&gvW)i)j7#7^ICFL7F3)FsDw4Lgku5O^AOAeHRZHegl6pn(bPUzpz1IIhHnS9QVpgU6}I@Nn_+d&sVW z51uNlLHW<}JOoGNz|egBJFLFS7P*Kg$lth8tFj?VyBw!#4>a}1)uV>mtk*27=vwRR-#u%eV$8S~p91}4gu{CxwB6JxUc-Le1ydTnC%xh=FYA*H z7GV@)zF)0|ZpPNcrC!$i%8WhkXT_?0$9oQFH0M+bw|c4uIC4aI7mKJ7@QtH1~q9X>Z}S9Gx10lWFpTnP>s(89~VDwU`8-e02o2h9Oort zErf?X<-3zvzl6TrCx1S%-die6Jb;uN`cGJmm$iPcXLDfD|1frr&6xmPwvBDuw(X>2 z+qR!{td4ElwmN3Vwr$(KnYnd8&0BTnSDdPS)?ORnl}UF#S((r|24t9VS2KaBnIlxD z4W-voo3TgW=7v4jd>w(sY6MMV3EQ?FL+(B`d8dZ_JqpxXj1}^7y7LXBwGr?7pCT~M z{~3X?GqE!L|0A$Tok^$vuIFCOKMA)`+ovH_zy#C0G!TzB=xi#lla{x%^yh$z{Uci= z>jjpyS#W5bX8xVtvhq<5B821QH6&o`EQSkqCdk?|@zLdkD|{1nlO(B$ZB^T#j*@&*G)pQqfb(_y5)rb0bws+*+XN8V zpFU!FcyOyK0ia_QHJIwhtd4=a-j4WAm;Jl6*S;cK!WNI}qGNx4KB5{vFQ-!Tp*RPPEqM8-Oq0B+ z^nSfikg!kk>V-i5dlC79V-J*_Lb`GPF+N-lgL|_}3xAW3h7=mbf3Rp&Q9L~c&-QBN z42yszX;$ZUiRL+VbfGs5kQITuVFa=6=OdPAmC(!0OR9fxY231UbTiZoJ&-<9$^G5C z?}_}1$bzoXIKNn7+42i?UA8LHxLAyN_}-S3QNSv6eB3rEgAD3jj+Gf^xj>AcnTpC} zz}A>7rvVMTv;X@w_r5{9iAllQkT|?-d;d2~cGe05>{q;*`iXl7T&^-esqe1B8`gMC znmL5eH|TJe;Zmn^A~+79RY!tm8a|Ty~kIpo*1?sjz=3N_Se+Gaz|7#Vq(! z7Nk%I5dI!eh+ie(K%*&`orsKSp^>`6Llv^{K+6w`wtPvmiyD+|~6Is_mvNJHc2vmWu+Ql|jL zg6?2EV>?r9LjlT^4ohyoC-6<6)@`v9{u20%#(OBlmN|WSZoEL-Hq~w$*^bt^y?nWw zdKF%lFowrrJ%A@4S?WTJB48cTWvwypSPlWB#N*)*4av;2dHo*&t0=@h1eqjm3iic| z4{u2ciYd_9WuKt2K?5YCRT4$`#m*s3s9sIqgGHUR9EBWjp@dM}GHw8&A~PR5A13&c zq@pnX2y(bhv)PK*YydZHov9}%Zw|f0a(XY%#7vI!KZ){vH$(rR!X1r)j?u=2c1n&I zQQWv-kbV>N^a8yS^~4C|3@`fmTcPnOs95|*@W(eMW-vYR11=?s;sxX6KU6wRnxRLu zgTDrrB0+qVjtEyKsv*FscOiG&;WAWfovwmP(X)Vj=l8}GA_6~o0My>qhFGE}K$Kwf z(G_jdKyfIjF-PzDAPVsEa3Gp_G(^@Vbs)APiT)vdk-|0VJL!PyVwZr`VOw^ma=>5zt*hSNp1(?TA?R} zpNIgIho~01lz?!c49{>2t0EJ0vYc}Hy*q$zy%b!)c6GB0cwjqNzD`1r?Z2;gkDsfW z<8a$(PN^_=a4nsMYKxjCI4#&DGY16cML=Cb6@4tsJIV?}Va$Uzh06mcHWT1;_GvKQnPX_u2yk4)q zCMhe=?z2mvw!d(syQoS3ZUFu`ZV(+oEQ#`PQiX9tn;OoT$QjJTI*D{pKE6$Ymye9C}n-+)Tta2l?PS>Y2evbRcStw};> zt=C01X7>jtLBWP>t|blS48OTv2QxFbK&^y=lrRhAiZ-zFZ0~*%Chk$V1gw>VbtWhn zZUy1UI5ouq1_-psdijASOS{G3~Bhj;bkTXE;ww!-U_0ZudfBd|VtDX@G z=8S*wqh!1e&w%dO&FqsKWBEFar4T2S-e8UVHVxdecW~fbH*TD$1bhSn4v0y*x@~}a zB~ejQ;|cdJ29&_4t>3nQb(s~vhFU-{pMREd>cI56fm)CuBzJ-=4zq@*jmLbiQr%D3 zaEnCP1q;ws=8f47Am!wCb= zGL=o5xBpBdnaDi#hG3N*xFf*V-{u;=x%C}=E-Fw{lk_}ln)>Ym z)%J(EOwJ-q@iMAQS`U=Sbh@jDt-&6_{lg^;7(R%dZHo8Is3Ben;V=;fFF^#nVdLh8 zP`vwIH9Y_B)9{xhEM_5D-iF`F#MMlFo z<5?&M@}ibaz8W?q4GVW{uyLnSp>Q^Oo_QXwL2jS_Ae^L3w_*))(5HkHsfqVs}q^Udu4z)0 zLy$987l}X<-B7CGXXF`ptT+T$)J^}VH0jo8_Y9-dej#8}4`CQub#J6?vX{o`mjQv3|*oqE-WIhR&;%+QM_J__T=$3ZYkAEl_Jpx$aR)<&s)1?py4 z!d4V@^C`qU?{EACMk^I>5C?5#SkWkYS(uuX7w`QdAq@cH7d7@6G$oa^UbaoaE9+e% zpg;fzc0{ObOxU=zI7fils8oHcJVT&nCc7x`@JRuMc>^-~y7W=9BP=;SG!6%cvafBI z2CZRvOomYAx+)c$(Su$o+%qLK_C>HX&sS)ilrs3+TyWvBV!eL6jX3dE=#ix6&Jma) z(n#Nvi5!`gS4O5G$%Oor)<%Q*i6ffnws$-qDVQ%K%`M2ktLK~oNNyrw zI_jt+5DKB=J1+!Ijc;?x5u0IdiWTE@3@%DiX zgKO+?ngC2(9Bmre*?QR`e*Ej!TUOgOrFK*t|D<0}WePJ(Qd0Szwn}{#n0XwZOF=ey zj_&m#Z?jLhFzKH#;2 zDBo6NT0~~?3aW2SrAuUG*S)u9;rx!?tNtZ8X)kPdF3^ssj4%eBy$j7HRAVm@S1!wR zKM=)ZMG;4q=Dd<)w5W>Wkckq{6gx6?jZHuk7QrG zs2Zp|Ueu8_Rs?)*Jr;kSCv2^-`0txOkVGU}mx`cDfUM@~B}bjT)#E^?v-_Il66)7Fw=8ZN1u zEgu~jAK%tR{BM>qOH74}g{UPsBF zBj5$O*(UOf>*?>(8;CpWc_h~5@fa|BMqh7L79X4E)Ikc)m=Hwd$#+*6QfK+LLR;N zW9~H_!Y6@_Q3n^aIyNVn%aobV1tHWTq-wlds!JQ&-|Xvcyq8cskz+-t-(=3Da|%lT z^#Nn|+oGgs0>7D`a98c$x3v)_H7valFgx`u4KNxZm-ZnH-_D3j#R4I3nM+f9x0Ved zfwBM=4X`}s9~PL-tG@y)*PKoiW}EhT$Ry@DVAYh)|M-JV^MF9^G)9^by`AtrP!qkB zGN>S?H4y4yuVvq|>-|yH%@w)&y!fDNnr(`cs9>`~vEn{JRz9WD-8z7!5h|9Y)jZA7 zG!p@hVjFi2nPoNN5W9ayHTZXBa7oTxZfg2+3K=w;bk#IFHUnN{>9@D{@sZ^V*T3R0 zl27YQe&;8e3UCD}MGuw{q9W${qMK|=H0aNz)G}pZ&tkhdb|#f4G#sxJ@FfWDkScd4 z2ws=PW>s`o0f_@KWTcFsGfNko#;u`M+U&4DokK_0eH!k7+ZRrHXz1_&p2{l=V=o

    #23Z zg#T5sNS8Dl9qsJo$8xY(U8JUmzV5Ac5oaIj4(9|fA&=_a-hv#pt-x6q=RG7N^;h#L zx3`1|p^G-Xqhz_R^Sl6_-8YikLFYlw-gWZb8$+KsSYnJ#EQx$Ox9>%Mk!BtE!po!52`B)f3tp7$#PXC-S{BYUXZ)HTgEqf8B;H zvgPr&cimsls{K}mJi!j@T>z!R+;wEh@D!mB@ifh%w2zvKOinnOL0>iuOcaw!C2YR9 z2hoo^V{JN+zYw(J8-oeh$S?3huAEGnhu;s{MBcFhHT{VR*ybH42#ApAFW%I$lv8Y5 zvgD@p=i_US9ZF@ncZv51@+X z%7?J7%t?O+ggRe}J6Ij#o17^S!w)`FZL4?}(U*v8@Es*x$X5RY{x*rD`sIs-Y_$_Q zE2uBR$?*@Pd98S>4!`oKA3F|vk=4kgsk1{E>_|kL$MCO$MlxY1Izc*B;Ns{UYP|4tfolN11;9eTu6r>$&u_As+xry+Q;fH<9}U(U-!=%>B^7Bu#>2b#8G+bb-Da!wbCYB>gdk>oQRyw>VAF}3IJO-p=3%4id)|EIu@ z>wgA*Y@A&Gi@@(vr^5+XD&_HX+y-=uf+EQHkXMC33P9$1A{?HTbE@X%jMj&uMG6!Q zAyiS^qc*@}Q&rC_Soa+N2qIizLiSq}LP_m)%4?ec*!=8SDQ8qE=6i1Y)SHn|@A9*o z^7NAGGkr=T^w{g&;V;1;9nYco@%3|S2wrendD+59N1Bz5#PK~p;O^<*bhPi~BUf?s z`OGpsc!zI+w`T)2RQIRX3!SQU{(F*qE8M!u8$(oe&-ml~HCAWxbwZ|6*$5CbxjJEu zeqdS9$d@`hm%4uxss1x`;kHM(AHJmBNuRn8?Sl5x?dEtqL-iSPqV+|F#Kf$;bLAV| zRgx&rDvOZw7$rA`+e?8swM9F$)?17;7}^hKo1+HQ>8dp@B*BRs#dbg@(~AD)5!m)> zgrpCimC*dO@M2nz!dzIJ7@YNI&WyF{oxunW~J(eknOhp<8nM zN;S2v{+H{c_B72}l2O2PRJ!P@IsjxGuFiI}BnKlCq=mR?@}C1hb>G1KkX(+iXSi(e zU5$U3vIzm21B3KRABi*48ju_TP9MSk$fq=9Zm5Tv-k&>yZ|ng4#1`S9MYr@3BwD+@ z?{K0j{OQuB|2t6=gTv^z9tR@AM6P%$?@HjW*M9R&ZNgwBJxIn^#m@Y=i5J$!BGt-5`jmc6@ANX#j$M3I{X=v9SYf|D%KTxz@#ra|ng&6diO$;u= zKjAcE(*Xxx*j7?o9Ta z%W2L`dk}a<>ZR3SOWsEJ8j7fN&|%QQkVo7HbcWi272UgMc!<{j+R||GBAK{Z|I}08 z!*7U=K3?oj`gDX|KE7!Ps@b5%&n7=wMiGhAl$(fRhckhpDk5^F_N}1IoK=x*DPIeC zyFJ}*CYg?Tt-rayb$pxd2`90?*jS<*CXGs?uTQeV2{FI9%nfJUU5inACkBJMb8eJc zaR5veG1JB|pAu=Z4<>Fjhy1u`#z4Q|5BU+?k@3b%!WTp^G7~ck;dTS2XqTsJAc1n8 zeBt!Wyl_9y^2^k0K(4{hT5_e0H*y#y0f}ez9BbT2;;+Bu5nB6zjt$s6L}boMbaj{u z7!%BeU@MWT;ri@d7|(-2<*fVz8+xSX@S$Tx2u-Ivh$}uN;AW(X^Fb++db!%b!Jm&` zhxg6BQx+gRDsK9YkIFA@L^v0Jw*;5uK<}uyR(7(5#|p#>5kPIBNgg@i9lyLM39X3OSs*V2j?* zu*B*Nc_xN956rcj&y>SEUK$WpfdzI?OkZRAz6}=3E9w`-H}^Y;4^Kg<^hX!{0e6A6 zXyYlT^AhcE7&x%QKVXvE3kyh|Y3{R0km~Dfpek?WxNAo11t_%^z2ekU!Ww3OY39P} zI5{vX1hEFGkHrlaBM%!^@Nx4~Y&ah;Z)ZnOnOh6y<4Kl;|2bw`jHp2jJ=v;F5P!)b zM6%z}A2WsczeRBI77|xY8VRzG$!h{u#OO$=O@x4UY{}dNTUQr;-_%$!XZlEA_ToWK zc}LRKz^7^G!eDPImGW5R#Z?lPP_x`Q0{yBqOSr^xu~^&_V7C&6fD1S%^59Hke`_LD zO=^0FdV?ZkELHG-TfiOA!erwf8h`3Qkr?z;nKDQBF(2@aXO|0ayaUYxvNipi7yWBK zRC-u0yln8kK+{0|`(wsi96S2`YRyr2Qk%Tth#ZTvx9=M*v5_&7S^^b)z~wa^@)lij zhr5wP5d5)l=66!rFh+Vmd*ef*3@DnmI?o2P11aJ5$>6Ty?Gv2b6!ep& z72Z&#$0`y9!<0Fusk?HSrkB(n>z{XGY zHVZ#J+p3wD_4wYt*&WAGM1Up({}EB+n!t>jh>Mab#H?+!(8}%`Q87-jM=Nd~=2qD*DFQ!nks-MI z!omLK8YDiR`N}d&?Oyn(uIwUH6AZ*_gpyC5=w;4%~}$&2dF6_oG-q4*hk<431Z4Kfl`2ihwifD!J=7t^>oc z5ksLZdfhj!k5KvktVl9^ z_8n%ltnByhwi-43EU^WV11B?q^ht^(AwwAGu0!h5Lw5v!34Sg>XHJ;rT$ zG6)4<8PZE1*mJI0f2_*r(LQ3%D&W(*lw07H9}P^DwaA&vT$=lvB!d$|{GLUIvDjV5 zj)PZE$lB2vRU1>2ZK|v}J)xP7J}2e;%)0N(-#4dQ?6x^;=|1I`9#D&+z1XFHYLiu9 z4Xm;GiysesAp7q#ygVg4s~6|F-ASVK`Yl<6B`$)oaw?E`cs6P|n7+rss;p~gXl&L( zlWB7IO8mD0hJCRY6}>YD4aaYsU-_UErlsy8BCAZf@@gjs(18hXlZnD-{?5QS+O-`l zO+`u3%1OMZNtCH!)WLJqBjF3~P(>lIZ;@O)0@C`l@GhU@?#t5Hnr?;HD#opk1R-;$ z%&$~U(o8_4`Uzw^JS!=$JCZzvJ7-c$ywX^UD2}=`tRi&1ETI#Bl zSh;vvXi9tfz2ym>jb>Sn5VFwxz*ml<4aoA%sf=W!`oj62Vb!j@NgO5~?Ly#16RpGP{ExwmbhB(Pb21)_i z_R`r5Vd4P6jaGL;>)sebL!OJF@bZuYrIC; z`F341)cs;p$~`SPFhTJCT2U3>7x=u^GPJZFh+$JdO(!NsGJZ$S(SqX;xP0%FMGv^K zHYu*rdI>IkLbHp1$TH9MxdBxaE|;FBft95B#i%8_FMK5J$UD#2ypE&|AU@%17EyP5 z%|5Er;S5>n6jQu1r6=LsmJ^MYfDYeC$<8_SQ*!>n6qt8{gR3hr)iJEsn}=??)t+F~ z909j7(vJ)f;DQ>fTTo-}=dq;^9C(u-f?%F`a7k2&Um`2#awJs;%#hWR1Wo2C%U|nH zVtDGj$C;mRdU#Ce5s8o~oAT$5@$QIG*a#q^WzKsjRFki2ap=OU>9EM&sCE77k{tH; zjKsp3Np*pN$IIo$@n*z^z|Rv{T(WqZj$ndbXHqoU!bSC>nfNyEIiyPLTy!y!t`|Cu zwy!Ae@B*l<;y?B5aF*FlO#M?)?TO0AN8;@eD#3 zFhWo_W~YBet&#PbAHq_u)(y^!tQ6|?^nE#UF3kj?c8RXM;qXC3DY!w%#J~0EF>dVK zvQ0Use1F$K!gXM1u!8(}^!^jw0gLReJ$3eZejd4DRCDDWRkZ%VAv?Cj1+_UXf1rx`WvQ~(b28zO6e!k~& zZdDE;W^+a++o}sx^)Y1e_noYAx2z13K*uDOkW`2Kd?FKsxZ}h8=pTVTDJfYOEg9&VE%blwyzONc zdcXpLf0L@N#CRB3z`BSZzErCn;tmz<5j}1S(tNZvYFq2lhz4Kbajn^gShKc{ec4db3VGCLB|K_<(EVj#CknUQe(j4d9| ztPP%!xX<+h8v@Tg{#6h?3#LZCHB3dNLWrEUNI-|XwG?-o+*%wuCAwe&E%@XYq(;@R zf2>PL%OQUwaF@>yUEEgTP81u^Y4Q62cgm$Qi{L7?X2}6Sx8G}7us?F)H_`7?TJwtG zq#{XucTwfsydjom@6L0=^HC{d=4Mlzb9S#cSP0izVdFLN>(7v06Z-syT6-}!AfL@a zDi0;(Kn?GGL?N4t6Y^DD!eAH)nVdV>zc)wQ{n_W&qE0Bn8bD!^{P1y7bVv>nsxT`t zR-l(|nJ7mm@tUo*YkAtz!{#`tCh#0w0uYH<+H3Q^dR$mF`5z2^lQo_84!fP4e1o^#o`aNK2K@ zG`aN+EP2mE|DR$n?*9~fasMx7DE{jFNbN?FH-Fdag%+l0q0%6|%3DhCSCwcNnZ|Ek zVvX(sNrj2I)hmkoCzLDT{7FvLm!CkKXyd%CAQ`!;XUhkN9T5|_=6}x0Ebp0|6Pr?g zTmItj*^;qH_4>1(_~E+r#od#`&%e3;r}U9dhNX1l;$dI^Lu5)MrA)Rw-A${YIy+nc zI=tU4bQ%%ysI+H?RParFoDA&vcJp*F#~p{(lDA?P#m1+?B>K4^i#FGkSY^-lD|KFj zck*>WrgCj;FDauje352kRfO`Vf-_Fe*p5KOKlD$H7iL(OL# zT*`V765!~nj+uKDi5AO~I2O+05D}?KIv(&c5a`T-`m2k&axwJPn>gB88n5cyX_bRK zFz$4JSNKP0m!+R|^=XAjAG~T&^)=2v!4VE?ZMWC3`j{hUmjbrgqjita zx>YDk37nYek-aA&?)9MG`MmUL)?H`GDwoamFOQkY>aXOF3YfWgBJvpgb1Feq3ZT5{ zJ<1G9OD=9Rpfs0KobIl{A*XTBuShm{{(XvpNsnoO5o+=MalHq-91R;^%TVToR2v$* z^;OBx-Bbvb?*24=U%XES1~I8@vO-lhfsL}WA%&ZK3# zbG9@+V=-Pr=*#GrwlH~rflvg>4KV_?$W%z7d6>0b8Q#;x8)ksjv)f$(K4GFz3S@lj zT$(W6Kc6%P(!@RsMI9}1Mw^SOZo-zL>>KFOu78utN#N){dj&MJb|fmNPk*E;JhLZe zVZGOzyM~#ghi-s$<=YF?HJ9t==~L#coJ1Mr?U&BY0D4tN5qvUUvj<=j6yzr8%YHfS z6R8T_i!4|Vm5sNK18VbCoC;6RJ!q9@1=4#2Qn{M7LkL6_Z%#c7Q zIv8{eHy%eUnk-1!A;`QbSZVV>&NxC8g?2#gA6+xxum$Bz&Z|HR5D{)Ohit;w{fXdn zaj<_!2FY*!h!?H|VJ|k)$IN;O{;Ri=EZ~C7vd=!%>ef%Klz9M76{N|a4Iu#3=X1sw zr6@rT@@a}>H-ix$l$6(<9(urvQR8jf@6T{JQ)!EwV)3XRVZm-!W!|-a%q@rQFoPsP zD7*ByL5H?FBg6$r6vdkMX9$C6$qgNdGivn%>K-;;t+i@4J-x#5GYFL0|s_n1)pb zz?d-A$dyhUWa?`K3`G(fKm<`c0Z|gJj=R9A+jz8=cR=Ttea@BuXqk5IejHuIxV9eR zo{p*Wdt8VWhPy6d^(P*1`{@!@v}?;ZO}R(@M>{`WE@RPaD;?(` z^hqV1a5t%UUiVv-Z{C^VoApYbTrqU3-lk;iwskxzC-oX9q@{##D2lWaLlVy8gHzlY z!crE9G*Fu2_{7nL+#~!)YI{nbZGR@7N5AWLvM7(Y-CM%rF=Yk)P7gO{kz1M@T>~}F z#|X&O27eeL#U{`^B&r*Av2Y|2O>8f`*wK>#++y*(yy!i%z`$8`AaJyk_AjPC1pyvK z^Hr6JQifoA78fh818u&`pG|c%#s4K3IE!cQIbRIZfadn;nxtmivnd z%6*aV)n59xV=_(U^R3sjTsK0kK1>>QUKujelqOwJr#vorG5Q}}#g;=rdNM>ptnm9G zWx3bw6gG`$sL5RZTv~b3%`4pN$i|wckEN#QrxNFW7j3%uBI~z`EPu+c;3pDi*{H_*x^>)nw_8722?~Xts@0 zi?>2G?h9)D4cQO5Shr2qzmN|O1_v+s6e~ks#*>Z%Gl_~GP(r~1T=3q-3j%kQB}Rpc zu_2&4H2El9WWEs6)vFkgXjWd4eKYfq*3G)L#}l1dtu!ez9OF}{f#zJ% z+g7&?-v}c<(Q#q>w^h*ePX4)XaLFH8M+LysK^eh{+0yWFBG`EF+N1qqj3k-2%MlJX zGHU&ODAtQKyWa?;3|v(aA|ELo*^0JAO7idU6sN$SgkmW^X`))ebSXga$xsG~SH04_ z6XS8>@h0UA=%<2i!=(>33R2@@ju$FAlBFogYuTDNM;EEC>%zm-7>ZCcA>3E4S5e#l zc6>MLuCXSK{BQ0;taK;e%UFcI>~v&@mTlTYH^E@=$4q3Ke2RhuQ^!dnz)o3NND4_6 zd|lO=H?Y3EgI9Vz8DjV=nyH$pX?I&pptVw>A|wA!n%=l!EMG7%m>~pq6V}n~JrZ&J zk;wa-g9tCiu760KG#I5jQ&>D|x`0z840nz&-rjGjR5Wxulr%h_`KBQsSI!pV=DRJ; z?%Rvgb0HvZ`T{IMZBcCW_YmqP&mD?EsT&iK@|WsS&I?U$1XG$bG5){dYpPtYzA9~m zTE=m!qoB`vmQ0wwff4SVH(6wcp=S|~Pf3~70Zg$XiE=*%qr0Xxwk+TvpYBuor+|Dr@KMeYNqHNnE zxu!OvP>ZT9mTQekWG|bO*psM?))>=nm|kMfVHguZ#6MDrq#e+U;FJbqrfdzqSNEZt zrm0Zp8|x(L&*bTrB0ymh(2~q`9p96&9hX`}Fqv)REtj*dlPZQ>pk?RTnA-W;nv<*F zxe*)Xx7s7qP20!t(Q<>sjNr*G+Q6hE2>@-m5gx}sVPTSK61ihX=%34mI#K=v-K$!K zW}HOQh>-a`!<*(JVN468AOK9Kk~M-C%`VEQ9iajh_u*!NhJvMc%SlL4aOho3GvkFh zK~bE0%hB)`=Ey{il{-yIw9vQeJDM$}j;qQbQcO4Rt&h*gg(*oJY;(Cp{54>?Ed+Q_|Tr4xp&Q;*e2SKTX zMAUb)mo@EwJ#Tp79erwzp=`+ud?o|SYe~w^gIgtIf+jTwsxc~c@mxRC zphz7UoflLP0-NQ(Y)Qb1%oFxm*aYP}c7sz~ok7;{+~E~x{v!*16PMVnoYushH|ir^ zx|8WDIn{D}9wt4UccfSGS@CoCF71jm564R|uUj_ryNpk2?q0Oc*sx{ktQje8L5IF< zkzT8<*!6m(y<^jVSrJzLc=b2S^qZ;ahgiAl=HwlAS5LHiH}&!~ONQ9+bpG`T5?qbD zJEe?{cvm#4y`E>qV)2Qv+>f&zDtd1ipdn4SD-h8tWE5y(*zqTau zM|0U8PPIJ)0H~;+*hfi-ow+Oz8&RUk=BW2{VmAAaXbdo)>O;yAAE-Qy)p8Prq0=UrAeI3xijo;W33sFh~vH8snP(^1!Z5{YZyjUK1ZUz%; zr)Fa3ZqzcH`%7iE#Ni#tSMt5FPLI@FCNZJf7by44ouf04^ofB25VyC7On7V&j_e7D zH~jbAbqa#_sSb4bVsHW!mh7TTo}sCr;b|mdFk?kXue3tmr?#I6uTG0YP_0E1yJ^}ee=A_N9!pvVdU>apMWI3Hhno1A|fcC zDSk%Z>;;06EqykD+$QSyZmv10%Dm>Bv4iU_-duYa|1fW)zGvt|41DKTXE;?K%6{;L zNOnG!E*N)>Eyh~0YjW~DF)cTjI@*ut?i$9@Cr(q);Z7-jZhX2CjD}oe6nOkFmVqp} z)qhjkB0w^%wa0WQ_>kyPxy#Lzb#s8pNV4@d1iOW|z~CI9)^h_ z9CqJYog-}3S8Lw?$HN|3%I%OiTh%rI_}dPIQ?1=z!~GZ9raIgNk+YPsgYiP|M>4~p z&+a{r4-S*%!pi3M669+!i@?tZaE0+JQ)cvfkv zGqCX;U!No{h{`&BlTrO(Yi0R9>>%%^(VphMq2Issso)gi5 z>Bz1r(Y!OQVr=U7x46iH=r_H%@!V9#`<3)n>a8FI(DeCT!31Uz5_AtZ6F$1g&d6#)X<5iDs zM<{~f)L-U6LUpfmx#L=mOg%wJvW# zoao|s`S3(Fl}Qy97N~T=tuMtk7PivH^THl{VVT_*)~c^a(looFp9t(|?{;tGzu>QD zha(A%{|A4uZu~d=#rwn)8u|l&CBBS8eooJ>h-7+P4OH5)xe`#IRJ-fT8y0O*W0fP< zp33-HFn+mlDBmRYH*ud*$ggY2M7Ncm)J87}0bscFidn(cIu1erMua+X$)CBNX9OG^@ zU{(DVZX&g0ejWlT7+P_k5nq^Dz{FYB*4F_3h`cHWD2x%j%)`T)s3wcEEBT+cd9O^hx)8 z&~4bfXZ>2!YG1r$?xD8pXYW@(UH%|dk|*QF`|Nz#orKc)aA$>fxh#mjhni0wbT12J z2G}lv0F?}!X*9q6)iHq;;*K^Bco0bI6et73{#!{+j!3BT(e3EtYHVq1Jt?*KPkSOM zL8s6w>>Ob&9wf`EW!|8N_!i4{7&Ii0Z$}Cw{_@|c+UwzOcrU*r2&z^sK{=bMgFKoE zC@dzsC!RhZKI)}ju8mnoLv@QLZ+?~`<)H~s{urZLth!kp@-h&dJak5DL+KO2Bjp9z41%X`b0xZRVm35#`3?pkJ<2Y#2S$FspWO!U zg@UZ~zuF`B3afFo|ut>#l_OlQ; zRq<@F)55K!M<6|T69MB4^UgnP_(CuZkz!S} z$#%~2{8LY?+^2e0YPLam6kwM575A)oc`@fTRl8LpsVMlDHQiIav%Ho%lRYI*FsCFB zUYS87yXn-K-UMk$O_pO`FcSS=DP%4_0WFFit%!NM@74TCsr>c;Ez~9?;YF7^J?*Z;)}}};Zvl=ILUDaIuy|bY zsm<}53CUL`GJ5MD7r=LL3T@}pHb@|D7>EO^4(l9HV>Fx;?Q7creC~^RhUB6 zzX#H(dQESrMB;hDpU~_@Ry`L2BDIhBt5cbA2T@4ziy-OxF^=sUNVd{^yEa6m!c zv*}fcNpb~=j4Z%t9zkSW*JW_e=Odun)%Zr(H@qu>)CePN9NJXjkgK(`P=mdas_zS=UD4B z^=ty@kDYN+qUgw z$F^{e-z@jWOmmoL6!J2{Q-D;pR?h&Y#aCk2tL~ z;zy0+ba3ht&Q_4W?YHqHx^cocZ;kR*W9y02-4{)lbB8^@fZ?Knh?Yk_sk6l(Bb;@C z-GTD)r$~Ulx{rB~yrYVB+SMG_lxE|sXdT|mg6}b7z-^J@ABOUGoN96@rjSd}_Cqw> z?29k38;nXj{pNRX&-Lm+A=?pgOPO2wHJQ>)#+uPzx=(QSMBQQ1YTEVWK2%BvI&rbb zW!%M$U#vkktfJZa?KMO}dkphblC7+Gl0};4T;-+%?@Z{WI7Y%3Ppqe8l)|n4GBz&y z#8R6rJb2Wsbkb^A^P?46*y=%9;IDUDbs-t?wwonsrO-3a?i~=cmLYlT!hq_2@OGn4 zSofg%L*&pTb)_OM9|6L5a1VS~plSp88c`vgp-Bx&F@l5)(cx#V9m@g7S^rq@oELX>o4~aO>GFNn0Q>Vx<eW|k_{-_F*4C83x7KO;|)$vIhqTJbM1nGY2QRy!l_b8 z{bjG4K889B02As_ATR8T-e^zN)pCT+9D&HxLqP3XD#y8+dJ*N88stz<+6Wf7_!hh5 zvRY7fnDB{IS5-VYHt`EQHQF5nBAg!Wp22QIPoaYYSon%ypbdfrUayJUFgoyzP9RJ} z5@Az^lz9+M9Z^1~|Kt9ycJl@Euc*)j$(rfL^Oy{7@)(tvc^wUxaG2_QvmM_@OFn*d zj?ZWFei)Wh2o-l)TnX&^CYEhb?!uh4geV*feKZCW;i~U&;TLUHp+IKoi}o1#@~urr3O=-D`F6 zA3&kNq@VIO?}U3eU9Xqa2j+l%DB6#`0H_HSl3kk(4lR|FHGMS`+(5|x zri-DqkQonjKPoo0))U3{zsZ zmL8!nF7zUW>i&-PZIvI>LH01k?^X@3mNfrT5@m)4%m?_mYr;xQ0IUFq|I;l($|@p< zDwDj^OmGPc>QJMumL6f^cKPWih_nHEhc#g4qZ+zPI~ziN`}J+ET$sg~@eeB{f0rcd zFCQKZU#+Zi*$T8^t-2w#wlqf4O*W`IF^qhp!&j^vVo>1d{DW1s-cwkPSctcYQveDd zjvRJ9h|*U}sqaQ#uT&?>jga&lmkv-1xO&s^WwTK?>`bvXx*nNk zNiNSt?X18&ib}!*AvpX}4@4qt4*XO%#>e^HuQtTf9RmYB<2jy%J%fKa;fN3}J66f! z+=MhNZtf@T2`cz3cd*t-3Ac(smH)Z0WNObsmBk;xD6S(?0&hK6oP1;bQk4C9Be%*j zlKoIE!a)Z|*2Qi`s0%_y?$qLVm+d#$vTg+r-`$kH?C(Na;ivDxuZYhge zK%(FHAc12Zd2BM>J{5N;!G323GWD!A-yVi?RXA`ZHy`SaT2E{vVnm~B#HB5B__wC7 zwd?Cr&QP^o4j5F+eJ`Pj`>STt29Nrz8k6ZN`M&pX1x;f|K?pe(H>}s*GQJWXlPZuk zz~TW%msZQA$NjzL#JtoP+Md~4EVXcNVST)o;9B(M z^;|-ZzK`Zf+2hRHsZ@e*vdQjq+AJr@+k}3wy8T?a$e{*#Nr={#X>_4~eC-xE?Ds;C zWW6oQnkq`V;L1_@tp|*}yKpZo_^2nUr#D_jiuL?UA8z{80&1ZT?c0e)n5`%u3VAao z*3fz>RS%13BQa4)ZFh&!DtLhl^Vkj7RIdwELic z(Pn4>r)-+~U4;bgDCh7}5&4D%eLk-rU+YWAc^@DfFuo#a(O&Ggx zaqCwYBtIeT;Sf!2rN$77eVC+E=ZN;gLmrgkof^K%4~QVV=DI}9WvrUmFN67?l(vaK z%!JSH%dU_Rt@LjcyZVViZ<5oMZNguYyrL+8qJD2_RI*!j^vAr7@T zB>(E@*IZpZ-=4wuszAW({QYK1!{kJ;T3mzZRURhGVQuFDn>cPa{Mh-K_Snq&VB;ZY z*JVJ{&x0Stw%pCG%w-{q@&RsGC0cxXz}3ZASGHE_L{WKx~aaI zvnSL1Q*-?{hzbPljD4pjca?-b0aabQ!sEsc+pl2F+P+ z@-xd(nbLAngJ!l#r)L~#A{qYPJ{wYL4&I+SYd_W{js(+~33M7@ z8qU*f3D}-d*{UyS6A9=F=(lI5o9h_WR1sadc+dlbIdsuK7oD^iR{h7)cpE>UE-evB zV!7fjo4s%CwflXd>;)6f(!1{^cZ$|f(6{lO%AXwN zM#=}iQK@ZrtAqb?M3eGIp9}tabn`_@dO!}~Tlw=o1e607?`Mbh)0jyyyXdAQI13Pu zVEJkEQaj18kF#Rrvf~}nAus+T{U2Rf`98u5OhfKs;iHSEANJ&|2l1pMP~bmJ7PHNo zlJW~wdFLJ_(drdDOau&gy;^nhO}k~k&0U0t%a$yYg%QZhp-pU*Z8_uU$g$w`{+9l5 zhsk*JJv!3tk6%lwQXlYFI|`bpN>sBb33EnPBV?Oy+9p{ol0z0Rp2^lPB=M0X~7-5+n<3aV(rV z1f@?4s2CSAmzt|T&J1zxsS=u1-~b2 zs~ZdH%G|J?)4jVg!Acb_&$)&e^s=kpPi>fMaBij!Lhda4Y0bXsf9W*;V7R$yKEs$o z1WTEmrT!B}A|!gEV|@F1pJ~}@0#OU7D%ZaZn4t5#fr?XEV!xF(YsbS)ma|#0?<3g< zcybGLm6b(91q@j|gSy3;X4F(amylN*yra?+xo5yBKB;1DRnHdJQ2GmoTsA;AGrZcR?=$e;SW z9=MtX0d(6GRC(koogZ5M3MLqyV|IH|qi|5f9L!v;$ptSNBg>&kDPneMo{U~PBTs

    LxLVkQri2Yy5!k&${H-PN{H&O(V8_t1VqoG9h>~LrS+K&wJYf8}gj96m=x8L1j4Ly(>pO&STn`#3MocQ-hAz87>V_fq}_|DEg;75`l48&9Y53?PDHaH z8a?buY7FAgaO99aO?~w4-w!(*Ul2*q8iGNPNXve-=D!|U>(@Lh^*}A`CKy#Uf|Ds7 zW1v~etKS#RHkq#EGB9Mvw@8o$G2@tzuB_y4vC`B#$GxrM5orC1!#_&ttKiGs(X8DxRZA?A=sAk*o$((#E zfS@nm=3H{RcvuU+;?`6!aX9Ex^A_||)7EYO^IKEh4EVgc;|+3GFIE2)q|#!j3wEK1 z{u7_n?#ZMl5NNAMadT@PG9`>XtE+ZYNSWq#VWuwtELu~4kL?=x0<_eAC0n&qZxKK1 zRVncga~Nuf>WB7zxt3#^olr?&?^)|U-Ia0iw_O#^8Fu=5G^ZGn9m03B^ivg?eJTX{ zUKK!JCbLj7f;P@JKZvPOkq^bDAJK^3diTmUW|nh><~d00uN~w@k3u5+F){;(fFnPe z!BnV$%`&s+Z`)s>aCrfZAKF1kU~r2EisMD^+KSz_)*4-u!k41pk|W`eGZ!N{ccn!c z81U$J#Uroc-q~sH>|XXdZM=rI<=fu?kEEDlH$Z#X3$Po#=UveYil5VjC^?h*$;$LJ zLw-)dMpBzgKT56U`2&U zXYdSIG}i*INu~NokqkFXi-C2Xe#dm}!b`WMdPQv@G)9>!u*IRa3JzYzTM`Wb>BEfg+`<#mpbwKFrsLfD+ zXwG{~vrC-^feA=XGcdZ2FSU{`R4)>`rwA!udco6m1B0+<~Du_VT?&P|KQ|gzjp8f-Th%R7|lO=z%}s(|k8pyFLK{l7y@~+PEG=t!Huk zPDKGu_uaft4?j*xgrmw+Aj)pwOE0IIK`tC1POB6>C?^NzP)PK)CG3P6R|2+jPwggm z!%%ZZ*4)k{C5QDg+Ot%aPd&p!< z4Clt7N*S6^IyGCt`k6i3Edj9Q(b>-jOsSIb5kwhBiunp^lglHAn83PNOz5-l*K)=vHi!i$e52qUoE zn;}d2YEOEfsd%ha<1n*F_4C~Ijk;}4z;u_=i8i@inM14ujtW&j_e@BjQTZ3;lm-k> zvpAsB&0oEhw~fNK0CfPd)()X&W+KjYpu8v(K?{req4S1dvDpH*Ac0^^MF$dwDs6c` zubo-(t7y3b+2@2{iL$>5Lttf5l=DpRyJ(Z;^mvJ*+?uCQ3=RhM`ew6!V@nCkYt1%= z)fPF7{oN9bu-q*uNQ?|otZ{D*H#;4(83&~KxHEqTR12w7A&!JjZ`iu*^&9{@u6nC4TLuc7w+xKsCocN$G zSg6u8}ZG zwP>yK^bHe7Wg_SrLb_2i<+7%tt!K`ybD5&O1#&rUv$T)qp73xWkkq8vYAt6YYoTYm zs+0|8ZPpv|k{x?pR9{E$H+ovm=d>-y|F}U`8h1l|5Uv6dXhL+f=bkouy0;wROc#M? zsgHtIgJZOVY}A#>XCtd?SZCfOppT_J`I2+wt8`PyPXo zrS(7hZxD$2KOxXR2E_P((Nu0YZE_^}t^o4gjD+o@*BdPMg@SIgX*O`yXg5eK+sf%5 z1XLAKHfWTltCGaN0bl7~3A~v=5@|Y4_BGsGMY^Ppl7!6ic`^chH|3F=@@IXwH$I)| z*ZMbKy%cX(ey`@eoBpniFZ(hVzUdu0M<1B#-o4?Q^g2a3N*`?QkuyzS(MNe4oZOzi zP8@(&_}t5q_%unyLyRvv4DbC^`+{c4Xr6JVY^$Ap%zvgzS)IL&!Xk;!!P=Mf@aX#R*q^rb4VUVqAt~*>vzcSgp$lx zRa`y2=jE+S!jWP(rK2LcmF2;9hdLy88k0Ou#i!~n>wn!_2qlS~DBRvc1|Hd#@!~9cyWci36GY;1vH)AFb-SrlL*16o${R%r`Bhyc=P$!RSqtnosO`_)(`wS|PuO4s?2 z_vIyLea)YH+-&9{dFxnbqa%Gx28PDEwCWB^?e*c|Ol8N#hP|?aWz+qV*sH4+-$WC! z(c}Dg%`!sdsithxk`@}&2Tk%JquaP|Onl5$>OUp62$i37+Uvw5D!sjt@(s6{MHrTJ z>u$(O`00~?g16%q@m6zGfn4gyz&}Y!bu4-44GEw+R^f|Yi)YghHg!~ z^|&r6K&16}+OMK{X||rawAF60be*OEzVr3!-!dU%Rk<>)*4c&Z>0KSRhmB&4EILy# z8WvT~inP!SFwX~3+7@g9efb1(dE3(vsdnVgn`(X%O|GeBmO}BQIT+z5;`T`md>SRd zEcgP^a^(xs!hG5CKF?^^0R$*r#fBcbp|8|GY_GO^Jv5>&*v?* zy=itHr5R#%*q|+BnliTI~)*?gNgZNj4JMYNa-;gB3 z8}IMOP7FPDFWPlLQfho`l@AOeiStM%-JwC9+pnH*oSg{23|BpeEn9ofK2KMpL{~ig zlLf{v!qacpdC=`IsNISUvV|3CUEo8^ehTx0%6EUG}hzpy3Y|e82=df|8{g zOyMQic>?Y%r0oemy3IOBI-Qz-U6VprMO&&6;X0_{(utq;lTS^uX@um^MS~lB1~Dh) zqgFx5beoCGGu{l zBM=P4_|S_5yvTUMbjNCh=cCq8Z!Dj1J%>K^9R%VTKYldb01P*ws%c#CR|XADn$jNL z$}E5PHG|VTLryPnNyBiSK6k(#f`e58?EXl_8pnLszf0>r^_S&v(z@KVYX`3daN2E9 z@a)&+FLR<`@BC)Mq)(u+#*s`S5uxPiNeMRQ;gi2ntR$pVQ-ayJ4Hg@sS+L z8*M-%D_(F+BsiV2QK$81dL1IJGJQCS55dUDaRsmm_8 z^RYDAOILQn6?I~CX5Ebcc2%>g$j-?+)2 z@i%QFX-}FV)ojcU-m$)5t`=4A3-M9Px&%+L8JiP(PC%iZOFrvMbiA25{&VGIOrbm| zn6QUW*@0KOi;oM-9%Ot+YR;0=r)g$W#?ZYph9+%3e@%~Udy2XrXCIdk(2&eSQKZOl z@P^@?j26{IWCyu#tTUtb@F(Kn;sKJu#!TVRqZ()b)xJSureo0FEmwue=;h*_{$V}d zFjISRE)WiR()#>j9@FO+rz8bS5}9fHl@$sgKRhEUmy3rp77?ap1pSlNT-}cwG^9l0 zYQX)kFxK|~0Ao;~Q!_**WYbcSP4!ZyVbR66hw?y0J=QZz8)z(N zeU@3I4g$!bi8@?1Ujt*0fRG%UlzaEk}i>>fwvr-896gaSP0(6Bpx@(j1XKd*}~>c4Q1L`vkYE5e-&6s{3AB@i_045 zVA%v*iz)Waax@dC4HbH@3vlw~$`@TqW7~s+gSkH>KRa}Cij(;SOFL$QE((9-A**OjH|C zku|9>G^f(hjSqv_|5H^SM3D9+AQSPnUxKjzvUoxtk`>{nPZTCc5#gBjDwC;mFN?CE zSVu!wxim2zNad!iL6JO-eMYqdnuqs-+9c(t4 z8j1f{IH^OjZ^1AV2(u(dxg7A_ko>Yhjzm%tS2(my*+WTf;kz2|u!|z)`CHp{9M!}6 zkh*2Bs4kH&-*muy^A3f3iIdz#l)he5kAe5Dmf`294iS-%;^8Dof51yt9@RGn#9a_s zPlOY9&CTl?^^DK#euNk|ykV3{dGYYIl!7m7L*_Y;=N{Z!Ak?%aoam!3``U}x-_e1|nVW)jz75x_(P21(iva|ud z8s@m0pWGOgciNy4*)amz6&ha0llsutOl=%Gg48F}BQJOzY>Q!{4|F&SMdI7dH?a63 zkkWEaZen6%=yH-KrB|+xDyu3_N;OB4Guc3X?SQL3Ro`U<@K07qjvW+DnT%m6WF7Z% z*r4HYmr3Z@df6}#&RB2+=#ay9xuDA8g zl%xw(N{Hle~Qr9dIP9rj^Pwl29KD`-8f3JI7k7qY7NbJ?zA6z*Rz9s zq{eAq53<<3m5lmwsmDRVvuT32cITeTZAW$|pyLTrwz;X;|RK)s)@Fm$B;{Fncv%%s=}cz$SzXlk%-k4i2F*(8kX%=ScMje)#)f;_Pb}ZCWd)q zhf?Hb5x5QX`ly3x4%4aBCHse;Ih^KM$#^4P73c<;XMwUrulYyLQumC*)$Dpn~lwn^F4eq zX}t{MZ(23_;HZXHWBuJ;mIFawm3W+Ntnmi<+%JpV8TMP^D5J!t$EOG5ZBH3-9}G~+JOpw$Mlz~qynymU0{ z&ayh68dFag#_)IpP9zkaEJU(YC^pP?Lu*KWmn*@w@O2iYt!yt%+TRF{F6ZXZP~QuG zFRWN|Lj7IM3Q&^K1S<}|| zTi=8=tC>k7+IE(7nWM=AVtfrR#41|K;Bz%v8c&l{m z2_oLoUPWVx=owoBT%!ku=*fbzop}0BYKB)@^XM8DEhO^SU}|4352}t1x&*Q6#V!_? z+1&%Dl{9C=H3>nU4Yi-Xb?A-cXdu1nu&iTX!g@ATSs9kDIJUX3-mQDJgr8#FYT$b5 z+0|95qwNJS*wbZux+>q74@0iIV`v+A?R}%pz$h7Tt+wG^-AU2IiEawHP$D^Tk19oO zs@SZU8QW;9yX>lR(V&6B1=x^&qBD)}MO~|wN%_y;+hy68z^>hGF=-HN+|SYcl`2@> z@?q@An>Mna-rb)5g<3v7KERL~&flv8x7fCPL5_UnBSiVsEjoFR+@L9}wntf(=HNNo zh4I!F!yza1jXA?wmd|&6X4VjO69JaLxgxh^EzZz3;UCW&T(zH%T-%*IJLXfunmLsZ zV@59fKLb|LQnH7Z%z7gv+JVBjHDmjZj_k##G%S^CU!S#n9x6{@POniAa%LwY>knsM zFEDELY4kz&L%qv>&i#F`WP!o*DR^f2#BTDFJxaoOYrgnO9ol~Vao3;iYtYXg*F~|T z{!ScTWILIqmwv9{`50DLE>wZCD8B>*X|cIy6MlPk+Y9ome9C_G~aD9-|3JNkCnjex6DX> z$*sG`7=fP5?qc}oFHZ^@Cianjqixr0J`W&u=V11?4YG|BwyxVw{GA_YxxIQ~k*@J5vrD}ik|qE*86HG~lg@yDjkpH@5Fo^p{07WS%3cbUpkCtj*umkU$0=gz zE!DK#Z5FN)LZWf^UH57Qo*AZu+uUoAiYZ&h{(DW;c6D9;`i#186sXO?vbfAcMAfxL z-l+75t|Y`;0oA+ZrdqvQbmNS9a;-+$FW2D*S%R8;t8&CfX^w)U`NY5+)dvNK(rT`Y zIbGQafO~U}TZXiH3b43Y_u>?b!_08-cWzzd#b9eQWRe0XNDmi283lmE}KqDhnqQ z%m0O^$~yn!sXhM$3Hax2HBgwcKwyHWO@(RbR%=TCcxtn*=FipJceY-tE7;_`C*89vnEqu9gtqGkO zEILJGL)K1CD7(~**cHF%52rq5Y0<^KWAAR{eT7|hon79mXwK=Wu#1!XN=p85G#Gy> zI#YC25%rBOqV>ZmDm~L8+wbhQs&<0ohfoPo^5W#NGOdq75&X!*qg!EerX$|mE&Q)D zD3YV**@W*mWVc`3fmV3tZf%?kk!_B3?`co-)6M3j4#XNVX>xgqQu~`)T*diC`5`LL z^$($=1(6K%*&e+T$zZVKMGsvUe^otCv{0yujlAXg!Ro(obyG(K59eEdek0GjDsI&` z4ylwRNMoeFkWR2YQ&3EKn0n$+Cx<%w?M$@}j|-SR=`3Pf;?bUHB?O476vcku^50Zn z2}?g1fRBnjiStO)5Git46~73ZM?!xwbVllhjIm5~ zrCxw9N6)y$YAeN!@ZJ*>3hSZbm?F$8nB(Y+ls`Qik7oyBEmmwAp`!Rog24hyo_Mt( zVQFZ$KV1SD>fQoJMTz3VnS1R76Rew`wIXSD_!bWuc{`<|1 z-Q#^(CHpkudxB_CgTpj!oP4|oB*ADqp%loay$)+**t3!1k(iv9Euo2$K0!7KrIgR1 z9$YJF*dwX;?cj9~^J#FcVX79=Stz*P9~2X5XTY!#-sE`(R;ENFsY%b64EWwjU-s<` zv%BY7yTrHnsw|1kPoIXM2vYXXIw^_ZaQ36tXg-_D6{9o3x6YJ|i7TxX>6kafuyiuO zSg^vWV28z_mn^RSR(klm4P36Be&{05B$UIX}dGq(EELY_Pt_~84 zqe3Umv}O#m1*34)%N}Y9>`LCK3ldlIXqN^!Cq?%ZDznBIge$c{%cH=X(-bM1;URC? zc=5vOj6puCA_H|P0}MxGFc8SaXzln^;VjdnFCLL=fRyd&6-dxPkjr$#pC5#VJ=}-| z&zMrU3Q{Q#4u)06%%X|lBrH%s)b&9cP(`r8P)*3Jr0k?N2(+?@Tw@_hfuTS)4N_I| zm!VgXFUZfdJKIu4GsI9_=##Bum9fEd(ch&PSZ;|clAHZb07iV*5Wd?ft0!T77++3= zEr@glI-39}A_gyn5C|zGZt>F=ObI{Gfvs~oM3~TWIY7WDMBvbtHL%$QNM&c@YhPUO$Mw-a32_%2^;O5YSh2Vkrtc zCc;02@%xVLdI)UTUZAh_6Md?S_8zQQ3Sxk8-Cu1XhP1m+r7sV2aptth`6ouR2!G;F z^nNg+3|Zsr&X51i)c>MyE%|iA=mq(``{_O2w0CE#w%SbK?8(bA%KY4WLV~9Gw>$=aqum(B; zD642A^G{$(nOl)_4s^?8`*7N}ec_XWm&JB9-SDIrKLvp|lcyq_l~+}^XD@yEqtG0; zGzg;yB$NhIDC5b02G3ZplIJUIfgn0|oIh4MW6ReNR}fRKF5*UebK+(vk%ko7%wUc7 zygFJ3iQ%v6Lj9e}+_Akx0GPeqx50=(LbXN(VKg8X^U9WTk6FN{F4hYuhwy@00a3j6 z1TWuOonuc?icLv6z{D_k#N&sF)1ApSQl`ll3ckOK|GD!$g*Nfn;E(Qko@~lFb}r~* zy1Kdt1jWY_N0omQPPXkA1UW)?%!nX>P^g_`8={_!)WPGyQxjq(6uW~A(N4BK#j&b0 zICtMo!Ov9Vg}pWITZ|>Sc|Vn`0y#c`nL;1Zx@KQm{lwJcAyY+`BwmETKRYKOQm&MY zDnsTGu^?8lRmOxtHOvc|*q`e*yB5k#Y^H(=AckKm>RgtHSpleijcQ`n%)2X8{DP@A z(3|9|xB#Ur>axA?$qtwdXZN5>LU9<8n*5T0Y3${C*2+?x%e$6Fof_%|`j7x&?59xJ z;=xw}L6fwUn+f z$Uym$6l4GSYBtsz#SMMKvWdhg9CG6ZAE8-abyek7Pvrn`iG*<&^D?LaAUoPh#i46d zwiiGDxP4{r%W|_NywP06o|dia=xq0U4$Q5k$Zr-q+hp@iY}9JrVY-W6_>Kz=Y7G-f zb9`v4+i4>rRK@F_a^7|?JsiLWgiIDV%5#L(7cH>3N`?hjr0A7g{?t#@X`}*1GwI8fn6S3UlkWZn9?Nvj*c}NX0LTigZp|28NQYY80j^_R5-s{7yz#+1%Kj2S; z^7BLJQn4Yh5!QW4VH&+H^${a z5EbBL@vuCc=2b z@(xAbuNENfPMU8$FhCOH3~}+sNg3wIr-GM?-zyQQX-WqjhYbDmZup#-eOiYE`52J2 z>w>ajd)1V>I0tVwuMMYAZ1Gohd_GogjZiIxg0R5Hh%_anDV}ecf^tFc(d@@wIa?<{ zkBX>VftOpsRg@QZg+#SnR7=g(YyPCW6O(xS;TsmYhi+$SBis@^yj*b&iN71*^>clE z`Myd}n_46J84)Wu8wIesQk-HHpL^U&sJ}kW2@7Vg(J3w~@aDDvBf+ZKPm)w$f+dqwN^VY$&MPSv&-Vw!-dJ^Y zwDwNmVv+@nCsX0xQ2Pa!j1}_8Lm{fTy`Lv%OMc#Cg-8KuxgeFzaXe6|Q;3w@g&LhV zbk+Ug0A=FY4G0M{NHO)_uaHm>hW$>Iob_@(b;UY;ujW8{XR~C7DB;eU-zM6$lj#FF zLJN7LAt3qhh%?dQS3f+;Cc7tjxL)>^kGr&nFVLl|Iq_`$xf7=NP9(*;J=UF&v+XoK zGeX$Mh%;HIWMqPW0kW?s2JIUP?2~nVD+qSjJVd+^nNdi0k$)bPi^_GMI?QU#V=jaF zkxsd#?wwhWPn_18VGT!fqYM$Zkg({im1OqU7L4W!kMCG;{Q9yIIoKN34oa0PF-Z>k z2lI)*JE!mQxn|wqDK1!oC$jk*#nJgNbCWz@d=0u+gd?4@ATyRy0b*{mGb|~9S&oZTKYnJ?0qsH~N`WdEpUI)NqpkNaQeIih3 zaDP+NvO2BnijChynFQiBL>iLu*@fJVhD9TXi{%mw$Ex7U96GaYy-yTe^jH|F{~`_T zsZt*)>pnq12=qCdOZA^UP!MC8f1xWKXX)`S3!`L=Na-=}UVAvF>z!zfNUc5v?)g_} z$8SSjzELMDp3kXRf+mZxmF{90q(9(Y{8ML1iB6@{Q|yd~?LEm?w#DjujV}dt-CM7< zvA~D{-1XEshW(pQwvusE7-kd45l(a&!^gWZw*--)!}SI17f?d!7e5u^!f%Z3bRc z%D9UVyPW`b77k1t;j~_V7x5QjXKi$IbMx06%Tj*J@^?LbmT|Ck+T#oT{i5?-ha3nxtrX@+D9qs?%@fz156p zJIl(HneZcqwddIsI(l|db@)^=KFK;T9%UDAi}#>^2QxhBMx{joIL(g5tLi$mni(}= zC~PODG8IjFeYjCVE@alTD*lxGkdTDM`~ar%d6Q~iA?EkJ*`Bx(1G~(u8jRSgJTIS{ zAZ1vpmAF~xILbnDaEdeCW-?E7*}J1?Z}IDObBfm~BW_yB8NB96Z60kYL zxE|A7lcNsBs;pGp1=Ep`>5L=OyEvwbx0|A_Yx@H<8K{u1BBuLtfLbDz`&S&j8VA*D z<*2(vST=_K1Nk2g0#o*0wltMbx5xBxKIh6N7hEN0LfMYgS9Kp&Yw6>$0DLWHf{{+r ziSN|dIpZ6t^oy)P=$yP<1mi^+g5%`MOVaqG?GaiOZlgJ%AyPGsD38RKl?n%taLn0l z^(&O;MbzZk>vI$XYm4mezT2T}{6Dn+oEkhDFatgwl{)<|+GIww2r+t^>DlmVH}oxz z2_~G`pKTpKgKU>O3yEKleQqNFYL16C8aH27-dD>Dkok*w8hk7iXwgp$28YB6=8DtH z`^(4Y_bJOlHfARCA$28Jlu4CoG?&DQ^6qixCaK^6WQ=#WM;}(N-olfdB7J6P>BiyX zQ?(0=O=F?<5Z-B5-}Vbl4DdQCZt$*7&}D=YX)=sDKYjSD`#`zzO5(!-P1PBfpsr;= z9aUQ&N0ZRzq7qrHb}XG{7Yae<-504e>iVeIw-^cwUK1PNwj>B}57arZ{+JTzV0b=iHEo3DL5~cKaiMUHhJI1qk$To6wGM=7qKp5TsUq8M0ZZV}X_$uOg-TTH%xs=9R@7efv<{`|8YuuS5(O2#c z8mewumQ#oIBwfpcZ3>)W$dunGtHOAHY?0>D8ba!<4O0(~G}|gqnh@P;ag^3OYe|3F z?sOHJ;%7m&`4>M`&Xp^UP*`1cLM%69>!U=feC6QfQpJiF4*%dle*b$R%>7P8j8M!=qeNepRW@n?xmE$M(9rQBfjf77|u%`M1V8 zJtV0}bbyWC(VFo3Yw0PYJoy>-j~EI5=EM2M>%M9w{4Wg0@}C%v_5Xz7TsPTMwmoyV zshvC+fTzRJf$$+1xklf~$?R1o98{jqi_nsW)TN2%WfUYjeH3T)vWB-_0pgN`Iy*`( zj11yI`1icGTz5UE_(<)E$r!#LeN#6R-5k6hVv)vO$$n-XxZGWwAC43+C2)_+CCSwA zmxM#gS^1Y#kbbdq=- z9xj7x+|k*{k-l3=^e0-M$(*r*2uDmKCoyIx@>LPz`}KTEe_maap_J%^9UK`fVEAVsCLcthisk>FXPIBwy(j)L}O4#+N^XC@e5K^6>wEX;U3hP(3ut+qh2M>x? zqK2%37v911`J{6-De(He-VkU~2KpHZWTs0P&y5t+1F$EB%IC^9{1)}xA5d@3@-*np8Dk_+N(#NhbX%QDBb#S~4DuuP2z zx}aVSF)qmioo9%1%)}#(g>jR(_XIQsTClprPg7)GVfX?!$?r{+tTH=vT;-HhNpm7j z4UkcN8Ux{unfEVCwnFMGdSY?hO7Q>X-+b>D?172Qz$r!9B!x}w28!7&2G`IfbFdxBOaYD}7?oSIfccGz$!faXImuF%A;V_fv zxd+uPL~_suRkE(Mezu%%xV#85Dg|hpjlA^sRJR@A(HCzIeq}SDj9VJg1Tqz~(&!IJ zVg}Y)F;bcvL`xw+roNaqkiv=w=~PXF3qT!g4y^q+Ov_qeUI?EV7#P$^KqEaeags-5 z<1)x8{^M7<9c#_htdQao8DH2i-yfEOElZuEVSCXnMg~JIwS?mZspa<6Qia-ivAVuH zf%BIJw7iAG%G@@BUefDpcSI-dH0RbYR%`70@o%fAm1Y;Ez2)fbNTqG*OcLTej6p1$ zZ^o!)q6ldw05MT{G?c{7F;ej%89Vmy@lL*UkWNwH>Ob9AS-(5TTFkD7zzr`!MzYZ# z@qw8F;T?pR%!e1M0?2lcEcE&fzGn*+2YNI_F#kUm+p-9D{cIrZKwIDiMrHI6V2kkG zUds9?r|HO4q6n786(a3b;y@VsG$^z~?I?=%G`xcX5?mu1*5fbPfy8m%iApF_f1{vu zv_kRia8FbHs@X!To#PO%-ChM=pb=?(j3#ck4sB$6+2Nu~ivFv{3P zNe<9{aG0Wjq*Hc&tSEart_`JOY(#7P?^L)lM0@HlWOKrLA1D(@zPJ7HKm-ME`}g%i z5w<_iTM=jfT{=zTA`wZ`V#f7)WM!WbCuo$ermA!TAKML)$2|Kcy(dvu;0Rjy!LZ<5z>sC!8SHylLk8~|A=WdxGVlmjxDk8JNT%glHIBP-9 zbbzu92BHom%fq&e6lt#HHvlQ@B7VVVsQ*%DJ+{%fc_&5J>U4>9yz6(3`R4x;QEF-S zH4p_c6I)@AaMC0Qt2BaeMot3YINiw_-$`T4NOHY5!B6SGHa#x+y4atqbti^eaU;Yy z90u(MFNF9aPa-u+grA;OvTEH)6*gktmK3z};s*Ob6kOFvjqS?Xh$^xW@m{Nr$a9!R z6}AWNrTKeR>7drKoM2P*m9kh~5LcVDroEaM@N1Z65}WhnP$7oMdJ9!6J&czzq2Krl zE1oxC_%3O4ft+;5TN^9{aeyh66Bowxcix}F8-_AU^L%8dG6aad9+cu~#f4OW} zhYJWAGE`6}Wt;#poqgC@U-WNGJk*#wxO+vxShLwpv6i9)pxU!LRHm1dBW|+&h`PhU zPb;HvQD<6=W|>POqVp#k3qdt@Z8t2;(n#i%|2*?ajN5QN$Me${pJ1tKpqI$Ik3o@I z2eY|nn7P3DDuBAMga1O^23nh(sfCvRQ{ie33cQl7&S(LZVHivd&3Iz?|X$DPafxwDzz`OGg$ zm}6W&Zem}spm8%W`HgCNtDYqa?c~dnxjJ_tdXFQ4O{kXYjmxA2PBhcm9omlzA-Gwx z)0R236>_F-C567w^U99~>t@6V0Nw%k$Z>tGFU_QZt&zzahod=>{RkgDtJmeW55&m0 zOT9We;>yzO1TnWa@-6{%w1r0CSs z)Aq#VpUCVrHjkuGQhWtJ0WmKT#%Rp=qfGfMUJ;ecUV~q%Bj12AUs*tc2E)XVn z3$H+N_%*)#bwGXarCU$;1Q7OzlCMEiy?CO5FV3H|5A-joGLv}zy(dLQLDuvHpR9sy zKsDF}xDodi`7@T3CSsd8xnR(ex{@&s#{GoVUUKQ3=wLaq=iki)_w+SD(#+V-xC82k zUVr%S+aOB8K19fyKW#RP1B}vwXJ7gbk$%Bv@vw#9!%<1UAB4xv`9X2Cjr&w0A~D#G zzD{G=udv*;FPkTN#^VXOFx;Ss@L1d@8-DWIBX+;jbtdz1q%MOX@`us+tq#*d%h~ zQo{6+aQZPB6w~2~59hk6uaMbrYn0`fTP)ISH1GDBv(@(JNTJ?$>4wnGj&wa73{%bt z3MBSUi4bLSlMi8u9yvdugj~y$4N;Gavr^b?)QxXcB?^RTuATdjgKH8{j+O0*=+iD&Bg=hBXS%8M8Ew0YS*v-~%M}~DiEZO$wUl~G$g*A3 zlhKve7i+_N-wlvXq>es2@_#T5eyBiQ*5ZmRW;z_o@nQ#N;FRg0Fft;ldIqa(v13Vv zn6+OB3HyQ_u!F8bpsGR4eUd6siI6rMC+$?+ea$8mx`iz3RpICZnI(~uUF~*P%PzDA z;9EAdr!M(v2tX+B+dn10_7oY4_$gSi;2XDs5k#jnta#d@ZIni)gnC%y9+fp?s=?x* z|4iPb6P(TlI45)-5ebk+--2M^Ftzm&Th`5v%61!t&~=6$a474wQ8BhQmbk=|lu&xV zv!AVC8Rs+#(ssVno3l=!Gg7D?yYUGL4whe^JyA4L1d!hww$kdGR^E&}ZtB$2MVfFN zTR~*%xz6^`RbDn%g2)FCWJ*;Sh*&b3&{(*}LV|uLOtCs2Xfh$!Y8&}a7$x9so?^?^ zQJA6)H+W8SHqj%>IFmrAbHVp!oPgATr7KSY%92gN`R;!*_&je8`?N`+j=Z(wM=t$Z z6AFwanjBv(0(zX=Ip0zuT>y`Dl;~KGU6RlS3#(e-Tx1+TVOJkN1RR5^H4dJ2Fl`-Q z&I~$@M8I#hyE^>o`a;1Aw5XyBp-PQpVXNKfOix$E9kTF%5mjF7MtaU>8A z8K5`$L!IXN_OMMAzBwUt#f=;|W!mBSU4Ct&fdStTas8{|W{$~>)|99!jAPw~dd+hm zwH6%J^b);_0CU|8;s&|Ed!oNFM*&)S;PpmFzo6%9#M7@@EG3PPI0?ss)%S|R1hrO*nA<5U*Zybl@o+!&xPYyT zvB~TL(ecep%B)Z8wd5!TnAUX76WOJ8ivFuX}jR%o{MNX$ue?FhEaorO)&?$bf2{F^}-c~=5`k{R!K{SBh+Dxqc=R?OE zrpBl3fG}%jV-&_F(zwnB&Z>XpJeBEQ*!lzZ$53~t>DXnTc{rSwrmsCj+|)fyGTYt4 zIX2~-GM7zl3t2HkW&;5LW&KIbI+KU5sol3{+GrBWstl>t7EjIRmXIUTn2&Bptq zL+%huPtmBWL|fX&Yo=Fgmr|)?$372j;F1{UMNE@QQ?fW@q6H7@dcz)RS;^L^b6(dC zlovLxfgHN+&Ixk_suZqa@187s=*CB^?8ve4Z z%L7DyQlk%Z^)4zL2g6*sClYe@+B0Ys*}UOeYSx#HOp7et-ZiA&8{>cXMPX9PTfd(j zlx`>M=#xBhag?AUAs^sOoD(Vlb&0$=63==rHNH*tt`>E4QG28@RsGR)lxo$LGbfgL zsj=4Sy*tS)f!QzbJ49pEbD*R$`2u_dg9<$;hdxW28Ok0?nsfZvMQ;|HU&Zw1rppsH zZp~SbjT8jwo?x!Q5&p(95c*2Wi5t}Qe#LrO1)gQ#UXytd8ydQl+U9e1AAV0HzjB?P zTArn3<@#23{$P!u>vimdda*FtFAFGA zo>1h@&?$EqMOP;*rT3{hSqV`x_2F12JEd`C-=$8yGjZC^v0TPoV~W`?*hX zy(NX^ivSg|$dA1HsmEou!SVF11*(Fptuw0^VEqNm`7ePD; zQ~Y8vE_b%#t(y+$-9yXx(Qhgr?svZ6KmOwX1>#u$3&hdW|DQmd28b3_68hfc`Ov$XtQrtw6WCZ`$2KVN|CpNa zdS`5Wy0ph0KTP|9@JS`wQim(dCENu4CR-(s6{SsD;oA+3L*@j*P6 z{sp-vlG@Wh&hLZ$b`nYAmA~y`iNWgapNf91eEzXXkcm-<_u?I&N6%;n6kLY~Pq{&FeJE&5-gp-Q)laWrWNFt3imluat ztf``$B(gRc4mcqR<#G3}gK#KRgPtDl?oY2oQAxPI{!e;12CeaEWzF%EF>GS2EgDG| zKhqy)2mMS5>Elcb_DLof+4FJ-WL6~at5!A)X4FXcv1Y7t%qOzVZp&8{1?In7 zRn(7nQH5_lDV*XANY2ePs9%6uii$_e4VUo|i$3i3J3J;gb&kj}Di`w&$?;1?N&%B9 z44I`PcCCpRFiA!Q){hMv$SD6tN`$%o`J~JfU{bg&7Ys|tv(a5AW@eVmpe?b-L?cq= zxYiU!KofJN+9$0~&91Jn^HG?09|ZWk;c(QaIm=agZKg!#ufN+wzt@;Jj%G>?(}AMe zaUC{>SViWacV@P1CFw<$cm@w!w+^QuQ%k9sRIRye5{|cS$$IX)@W@_H%j_g+Suu>+ zdR=XFh?{nQ;d>Vs7yFp@e0N#yzh)N+%$b!V5#^l9d^-Phkw-qa&`d-$u_mA8-YHxJ zA(JANRx;YwRW(p{qNbzRe&JUs-gdtmqP^F47SM+j89UnH2Vs=sTmhiQ{vVjASc@v%1<< zy+*9bY}~nFRXU?e3X7iMqD91kVQa_s{v4$|#+eGaKsMb`3&8ae9(4R!(zF4!r|7K1 ztaB?JaVpaILI!i`rox=rbub9L649w)(ard>sIiAejAVqeIVKLQVUSG;GeGDi0e3fS zXwK6z2RDH7B?2r6#vGN1t6GOJg~^?Adley(HTY zTJ8HDdNl#$E=ZzZwDqg@wpj-g6qiai?RINJ`K1Q?41|+Erk6fb;d}H6#xK}+_OVagLkfS*Lrg`nujP|0NG(}5L}cjF|_(jKsBL1y&r~S z2FQ0L<*vJB!Oz@+p#=2-pq-)yv;Ly~KN+nbYXwC5^t7Vvu+&y|LB&whIY%U>u8B4s z1*j0FEEz3k6iZr855V{R*CR(wwcI^BY8)&B{pK_cQbAF8QY|1s4r#GQgUops1M|lyXa~d&MMF9d5FOS;>KZLxnukmQ)9QOX#4y7(t~e*b z9DpI{Q6%-f%+^x095viB*CHL#QOCzT&8?^>8dDY{S5FxT$w_<>wU=-ggnzgOrD(1m z`jfSh4jNVd*b&Mx68p}6({B7LQT)quKvCc-;}@Xw<*>?&W0a*pWr9{bWhUTx6^t(- zz;z#JQPp2%H-VOSxy&q7nI`(jEB$ztQ8dSKb@p%{2i8M!plqy~qn$SxlHZ(qwWamw z6X;l*IAa(TW@0Dwx|1P!EHoPc%$x~o%y<29djWnjnZPeYHM5Omeh#K^lnsoVe8ex~ z>15QsLicSN3->KFDrEwTj#1ab# zj99EkQSs>tn4NAI?lcjH%3rqYK02PDm&m#82m}pERUFUJUfqWqQRGi(g zjW>r7+0+NzibaVx8V$3L7qD6sE0-Og4!6KuKyvI|AH=wRdT;%6J>`%#M(Wdzw45nRxcP9TPT)cNqX#!oZC%E0_4wbj0}NCg7f=G4o5obSIbMAB zV~pWh3Hx!SXnXht#60W+NCgN+(JCsf?XUMrRN{ri&ECYXJ>}t-@tG=f(=?_8X)c=m zs^CdK(AO9{C)TghW+I%k_xCt;QF9S*R8BjdRO%!EF3GdVfue*qQZcS=2+%HAIpGPe zv9eZW9C5TL$iTA{5G_IQGtL5bY+Gi6Acty~hDDK`o7E zZu)F^YoMZhNI?(a`ImXDdy1Kj02n{6NhfORTNdd_*+R+w0 zc%9&iDL9w;e+cZn0rDv@);Xs6V&)3fG4n&$Tx{u0zd#;&afM<5IZA{Lz{ilA&T~FN zn_xla#tjBDM35o!hHw+8+vNa|I2Jq{1|>?|B?v&IlJ-)0TRfV9&Ed>J@nkWos?2{& z-mY20A?ZKnd@(QYoCW0N>3O~yR_l@PR(P`s4AFU}oK3Gjk@l(2=`HC&LAf`>2N+nP z-uMN<9H1m)><>61O|ursC<&QtM%(S+4pr>&V; z2sj4V)8$PpO}7+6QD|g?e3z1mLX_+H-KBI|d|Fg$)n8Oqq|&`0QE|x4&I9QxC?7 zEv7P?m2z+`!52(({Z{U^Zd-Mcp)hw6wquvh>GX5_XokYyhzM-Pnd$CixYj^qhg@V`&RQG`SxB zlDh`VE)FaqW9YdripgQ1A}ov=gDXJG20W3882sfu$|F$x46+4ciB(Do4p+2zNn&ZS z8Cu#Mu)c_9d_pjh))Tr$8<^5#{T{G;~uo=aW?9K$tEVz1X z4g>Ui_7tTkSZ?}_nCSS>J%_%|IW%&Pi^yR?*YAC2Aap!j|7h2S*Yyro1*?$m z<|g+5qD>VVXAi@>%S!Vh(4P6{?2vQ1i_R4}aY)rl394=kpLRZiGjS+Vxnn-qF*&5ZlR zG^c#$;NYFlHR~U{2Ms}TW3?Uo+?$^N?7;jEs-!c;_)$mPV;|8^;XeHy-Kta|a-aPV z-}f#F!mjb~1)Av4D*uR9hmLW4k{1D`Sqy>8Ic1d;no#0~x#q5p{!Y1^Z;?T|6gAuy zz`DScx}+%yj4ApX={Kp2s(fqKt_$RKY8wdn48C|0Q6iH?v@iM-Iq zLW^pwKoc=%4k5PptedoIyqS8u5M|-l4#bpD;il2^z&?HFDfz)UwgXxu4V|1G$=Uguj%;;8ioRt?1UNxT!`<$JGSM8jHzk8 zjkIdc+@X!kX3c!WwP&%|OFIm$W_cEXszW*kTjm$pSnA%6v3gpdm7T2aPPy%yAj*o! z5P&oL!H=oYzl;s$CpZK=5-?PU(x!3QY4~G$Gu{(|< z6mz6h{#_nAb?3$Law`aKNJuk_6QQI4d05Ybk(AWW=`fqefnK6TtJ zbnfzaw(?_g)*7dJWBP-SLDG%PGUnVREb*LtjFf07TxO1k?IHVAw+)f%QU~W)ApN4i zAN6|0^k4b2g#5ax~R08^T;=bXivx?gB-F)eS z7wASRN)@p^x9-6*eBQZvSnTchGD>p9p42HRlU2F$QmLuo*|$zg*3H9%7DgK(xWv8% zz!j|&9LeAp1K4Tpx3VZCUV@_`=1G>%Ng64mJcrUMz z($~nl0Ozhzgc)VguXckO^6?LhW9?*d2!C#OVgN@mw`1zhK_YSW%wJyB$@x8|zRE>h z(*w0i^Pf`P7rE$Ch5i_8C{K30Ge-j=|Aj@q<^ky5SyI`z)w&B9P(iBY^u2UY`C$w} zM6(WLzv?g&u#9;c%vNfPw%4$KyNX2lww~wcng@|17~Fr0DfAuY-$w#Yce7VJ#sc$y z6ZlPCy(<3&5U2j6Zt->T@vk#(8$dm$Z`M2Pct7jX2X?m{5XK$!)Bd-16M0JhoL zs}XwbdEk5#`SX<2@^E!ud#izC1XViK(yU0d8I$1o!%%PuzaiD-dm~0rm3x2KVDa|d zziZL`f?4wmMB>joIN7E#(csu#yl=kNKsv|&!fkB- zf!o*^8UHW1P0DdoJfY|Ef4D8vFz8SIj!I8CiI%17Cqs&diRTrEJh61Y6ru${lIr8& zYgf943V^tPgJ(e@5p|t>nV5-Bx}J&3Ki+XB-tqh1H`jWO39`z<-B|W!*@N6+lZVCS z`EfD|GTukg-+o$d+r~)p&`s%iE1*nOrn>Q#mq;x(#1;KyLcQgBKMQ5`n&Dr zhp=yCPH~wY`dM(xP8be(-W&P7Fb-t=+m+4j|^ z6BQ4lVWYX9^zT!Diq*_Rb4D`W$V}4vBnpeu-%`&-I|Du z5P`yztKNuT)=}F(pSPC;s6ez8Wcq9|cU@4%=IEDY3a3ozL%7 z2bKO_2c+8Nq z@we4?4Kt5}da56bGgJKvkpkiT0Sbi`z#dbiuJw#FgV*v1!%rF}Dr5#A)^4kUXS3;` z^usk%*iNe8RaJTETyHMwEp)>~d)>#T4rSnn^#CKOfe9Fll*7SQd*Sf=e+~2`Hih3= zKZ8zEdAj7V?dz9=lD9s^1m%RUYo~!Vg^zcb>j=R>IDl-@NWGpe>37$EsmyE;Ekep! zOg&rMDA-BBKKT`N5fFG__1umF9Fd#N`GRo3Qq4p16WrF)?xVfV3gS+ar3xui)Qn?n zgMI?DLES{oK}S5vG67Zw*|^K{x?#_!0L%RP*yg<0S)$9=uf2lTVE8cJQvT8kqhjS9 zI{>`ljz&;LxWw8^zYVVfsP@s74^U16#VNe!$N*iqf|pfxDymGJ<%p4|q1A3#B(HY1GjE8Q|uhZ)fK?zZWF)iA1H)Sf-_uW zS!1ILt@v@rEvLuu!@u-SLOXzCZu|vnzn!C#kuC8}2{yIzO)Ykr(Ur?4xEDO4MRk1I z$4}Je$1Y*pf=ub)dO^i&CY$}lY{4RIWsU&&jNo)|=>12%>02vg)Za?KIJ*sr%tiM* zm+k(vZvZy#Bca5f*VXZdRo1_mdPRhBqU|_e!6k-Lf9giW=?rHZ+4K2k5FHw@V&ZC4 zWnnf9ywA<1=qzv_=^h&ihU_QFZD%o3CG2zvFIiy}(o4^bf0`F-4#AR;1E>H*(8Yg? z;6fzgB;-{pKFl?}fTK=|$Qh*Tk!SB6;I3DpOrR8+y!L=e-J!1$(%|0tDoFaCr2|LM zOJ^7xoCAvn6PM?+8%l!GoD&O10<5}S=x6ry)c5#?WDk3TAxmc)ziFJXKe+YGatuNx zOsOv&n#)tH`s!gC)C(`?^{>KP7?JPw*4fJ?zk$eJ9rlr)m47D?Oyol%tCGzwiZX~& zQk)5&ZKY=%kq=ajosPWvQ=AgDnwbas3Si}_G14D~v&6R2Mh?6B&HzP0$RE~z$m-PW z*-iI1|NcvzcA2qfqodv+cYJ!aJ(X5-(n+7~6l?V$rU;tFy*hbxeGO~fAykIAU7Kc0 zum-ck8-2_hf`C#veH9?XQ*u!7uT5<@ZA=FsXw1?@>L!0;pHSo5^Tc`t)vV1+`;Y<0{Nw;GWVab zRrTVv{#FJe{)RasLcwvI#a2rRD`jZ1@=cxWco`sdi-t)dko2U>B(m#aL|~h(09Q@r zt;;o}>=F~)Tw;)nZ0R#_`Vadl4FWlnAawnoSxSU`1V&zj+U)8s^T-78%Gg_yW?=Vp zHj1WWzLXXr*R&G_-eGio(Dc#L**i0~z-fR0dlaLDbu<2>mGOH7UsCfLzv2kc*KtaL z1Z@80ca;2BacyblHn_h|Yez)g8F}He7Xs1I=q2yzi+06*GmMA_IN!Pg*>e7G&;CHw0w9EK zmmu&;|DPENnHepaEJ+O^-hYOxU&ZQlV$Wa(3XZGB>T!a>Q-3Vk2%;1d%DC-ACQ%3Z zIC-zBo3&-;j|I@){Rfm171ewVT_cq*zU$*7goiY#0cy8EfKPZgC*ts zLFr;Mj8XwOAfpMI^3)#OM-c;ZpXrO^K5g)!mqZNBsjqKK8-UOx+}LueaYYxobDUqK zxWwssGHt^EIUxL2oiPk)6e+{ldu_mwGtn`BL1BZ7^IS_^QD!F2Wlf)Ag z=X9xy4bFhV^!XiFd#P5EcqNqKGuE*Kpc_Wk2+K`E4;Tp-5Ws(jcf$te(5+6sBk>i< zxbI4g`P|o8zvFyMt8n)iYA4T*Myn7Nab{l=a^Yv!W-oQhCI(?|BrF3&S^=^IjI?o* z5D1$A&^prSpr{vwlP%zHpF6BB4@7*^vj%~(TB$^Py?g*CTh|3U1PXN_vTF?bTXDEUd?%zqLbt$B_na6~Dd}2Q`@+T^7g8d$ zRs`fh|3L{C8745bR&Q0k0^~sYfT0?*ERP%7LGJ9Z+HjaJvQ_fe3;t>?UIO-IcX58d zr@*@$ZUmWl7L&-OfP)wdjIJ*j{^KB}8!GuYwpCsb60QE4Hb}_osCk4Q3;24pj2nBO z0*RQlNf(7TNw9ttG))!IjiPa?9pE^rY&I!dK3 z(!ZU+ebKtKne}Ce_&wizo*{VmUSB-XUlcFNYYbxz=13Pfvsw+AeAGz{ zMeJYHgV#w?QueO54D5{H*LHJf*|54t13`%n@+Kjonmk>GeZ{btV~V!P0xoLYr@U0Z z!VpYs1@@o_t!Nq*+c&}h+DVVJo}eAK%M1)RJ~_8sJPmjN#mHAG*Ib*+e$k1wmuRP# z>>;y*?;T2nf%sAVkd2icijbj^Rahs%s{{|N%cDonwL@78KNkZlGJDRFBinDcsGe<% z2q0azZy}tc^nqDY>8iLu-LvR#ZTv>86TjEuj3^D*Td86f@KNUVD)li5I0G9AXP{$4 z$@oXpSt@M?n)PqO3@DE1cuSmQBAgC7?q33x_f)gjd@h*+50Bl55%}J_8|U@9P!C@n zaK5!E_}do`mmO|w>=(ehI9$C+GyZijv5Q==@e#%l2mO+MMtG0Cne)_Co24UYcwl}X!>W!bP#Pj2xbz(42r9P)jXBj$JxVLKSAo!1Rg7do;C1i0Iw znkd4o4l*13J*~8X5$1bhWb=gts({a)H%_arp`G5_Y(}DpXLBZY4VXLxacKKN-$=8r z?_okQqXPA;!;gRS;p7|!8K6=yRlOr5?ePzuZ`JPvVLLEO8g-ms<;G6tEt-Cnqm@bl zHtW=t45!?1)+Mr%CNi7I_-k^ztigU z1Ev)*9p7~J4^r&$4KXa-YK`D+Qv)3J3kOp6wih?l#g z9|blfX(LiG{!uz(3kZBO3THQ%O}t3Jy5b$|G%EP6Ooym!wym~Mc!*_h=SB3rb059d z8=4eSbzw1uwVc^ZLG?0`i zrb?pH=6r|7?UH6t%wdHx_{FH5jgN<}$W*bgpv8T4Cg`x)N?YMf2o4X0w@~`iX@D-x zcR6t@XUAk`RHyzywd3A6tv!Qh;Y5Sbf41qn95*_PP7R#mZ5sMCP~|D{&sTV5 zhF=dSN~O|htOR!j#$8y-e~~15bYi9lmIP^8Sx!;zvTh~dl@w`$hnOf?k0wnR@g|EM zsscDOK})zfeJ(Nma8nE)2zOp~ZZBZ8Bo-}X(oB@-a{5qz=O)+t{CmmO8j#T%=m<6a z9(Ho>#oty;h_LVgMEATe=k&*El~Ofn ztPnt040@}WY=kK<)l3Xkf*b(xdTfMqnZEi;3%mxL_zpgL z(9|#aI}*yX=FHv;P;j{W1JD!VhUhX|jFVkxv$}Cd#|zR3N648e193oBn09omY)KEj zY%0(iV)DJdZs0~qaup+NstJS0Y^&D5~to-c45_5mi ztCc2)zC?cR;tUjEEuZ!-=VlIvU6w5+L?i!Dd7-!{uE$^FG@ndU3tH}8X{_YJkfp7C z1aKuZhIc`D*u`0CDnrEeP4cvDe>s39?@yx1N;g?yI}J5rtLnCMvA41(I6gB`h5+3 z@fpJ-laGK;K|Nm;3HOE_<}oo%6yOpCT&wu@B8&2{rBKzv;gtArWA0`B{+ z&9B}xy8Fi9Q2`XlYV5{62(HA>!^HdjI922AQX*$O5&PfpKeWrjucMcL#2;%Ro}rEsL*RCG{uTI<+cP~)a7 z4G#b6Nvs#bLoHm@7kck2t=b<2oWva7)U2leB&v%NSPAj@{jb%LDPjQ^!9>v<41XZpmzJ`g3zA z1s5%kHR-KML=X1(ECAruCv@gVYWW#=OvO}u+IS28!>$l2=Zp0t1nZ`GVZ z3&VP1(QG%ZCBOt(jX~hdoq?z?LvL~yeu8n4vy;Lw8d?b(+0=06E2QwaB;1g^L;*W9yr)e zN(ZGqnFxH-+6 z>paok$=FXUkcWzzVp=pVeNXz_K?s`gC1k&bQG=91lt_IMmEWeGf8zFyrl0{J>?T`i z4%Hs97vN+|K=>?&hM&ZT5-+p2&HmKj+>_&Zgfz!6ZBf8ZXB?Lu0TIpnm#t3NOc%Uw zi73AsAOZ-PW>{GOP|e&>p=YlrXa=9;l+zCE0o z4(2d(v=4+GPZaG=9;|i5w8d#*=@X&dgP;C`g_q{mg*^;a1skAo+CqJUjeew;EJ2~uPFs^KPcm|*xiah`2uR@ey<*oWWU^6(i+3!@Wdyk?-#GK`V_pYxAUfsdVI391x# zU%#4$OEF<2dJK!$MruktbgV6u;j9qxMw-*M`lb6(L_I*pYB#sFYVdE3@kA#~IM%G! z0x*~;V7Jo31VsW-N^uYT%XBaRN~2Zay#{UgSM<=5I>XI6t%Lg2%g+`A5;=y@|Hs%l zb!QeeSU0w9+qRvGZ95g)d16*<+qP}nwvCFry2p3(jkhoQ=KOI9WI>x{Ce=%P zRHkir8kztxQH+;3WT+DYd^nf{+<@cg2^sFsQp$EoyO~-lDsaPmI`FWL69sf|(zQ<4 zuU=w@ZW>MlYvaVe{@B5(bJetfA>u}g`mv~}$4Ab2M2YZ~^k-7JWDIYE9rf*R1*{va zGeH>?0Z3}|Q4D?)Umz?l|36F`Xn4CVLUY|DtU_z~LlcGzTojzf?`%GKtL3dCFS63o-G;rvT&|qgA~1Bs%laGuF%7i# z)huq&wTzx44c?>&+>|e}K1rb_!bc06ed?q~A=ip8-**uR(8{ zCoLg>AgKv_sOOAB!9bV}h2t_R#OpDoZc;*H$zm6&EQ`;~B??;O^Evf1Ons#z)7(s# zc7W1M$nK#NN3^c|BEZn%g5Qg+NGxbZJpA%OF~6aIr2Mvw zTqreI3#<>QBD7z?N$)A)S~uR^Gs-fKT56M4U4dAtEu)u@9bMaOj)_UszZ@Yf z3Be+24&R9uuQ{5XRaa$uYbzS?mkbYJ_vRl%e@EFiUbKCto>H$U(klvQ;4zf}9B3|= z5iS$Yr+;%Lv7I>eRYe4t-t_g@MXOARmTQ7x^g{)L0?iPd*mIpJ8p;WW{Z)3FGm`HP z&ktYv)ORk{Ud%L-N+WT!V!52+oRHmEcLBr9%}~Seg=QLAo;zPaDOgo_i<_8_;6n;W zsn38`COA40ssQjQNn^)vKY^KCj5=Bck}ms(%b`kD7LjyP@SLWdS>ITykHD@p9{|L4H~gn%0i z@i;)6fEIUTZ=QPVgc~djTAo$(YOkQEgl7?piL1;dX&_b^nx6YTu_-yBDtIpMB`0p6 zO86Q^z3eaVsdwJf2aHo=jVIkOxKAVmz5mB=+1 z?CXFg0D_wq#-1cG}xMiR+J(s5X~Os z-Kll$fc9k+kw{hp@oy_-5PfA$&vtajPF39#o47H+E)Eiud_sGy5G8x+kOk3R=OK(0 zCBqH-89n5HHfRVI{o8@x`jUdtqw>lA`1A6mGf}EmMabZm#Hl{Gpyq&vz*`qB8oEzc326+00JP zRtjOgn{u5+DqM^TSba6q{Tb=ZwICYk4k5&8xDM~8Sqbz#{;(DYk}v+P(;(kU=d?k*IDD3fu+XLc2uX}jd0Q{w9ma^KXG&BrJXKR(RD7ODR-rbp z;H$WZPX&X=RdkyghiXZo>9CZi(vPjoecF@qChNVq+ycJb8HHIZzPD)}nZKXe&NF0Y zCY$!I8RW|aHkb47EtX!hOJNa`qy46rM7$lGBc?ZZjby>d#AK!{uy?K6f4_kOTMF%c z%Fo#gwpV{H{g%($;8;qArK5|e(_4z-uyE_ypuf#nenF0ha>>6|B-5Tv?InBedoB}# zZdZ-q7Wb0wm{F1TTbqjlH{JV5*+_qK&9dOM;rQ5HqrDi3OaCN7+C}Wq`z;-7T>s}6 zxFvf+Q2=WE*S57!a8c9S;57o#Cf~sY8PHcBj-Do}hIPBlqU^!gN%9Rs6af`)Xfpxj`I^EfAU4t6VqNaB3S^mgf z)ysPLmwapA45)@mz)4A5@$h7&h2YRU@gXom?qs|4YZJ%nqsTAr*B=%+8sHoKchqwS z+jpTvoDVP*3XfASxkzlZve-KXD&;P@mmaJc?rJIW6*1VKSz2WW`cJqu?{8&Dtx22{ zs5^NSg-@vR?$_ju1_iIGfA@p0ud!AfQ3Z}g>)$ z{k`s->jheZ?n1!(Z|X6C6-mbSN_0%^7?3hdA(ahS>EU54^v`gt3^KK|5tz7WOcj0 zIG7&}I$!2JGxhc*o6Tayc-O0*SsR=ay3mJBw6b-^*6uQ2R$ZIoWn43UZ>`K#2yIM1 z)*c$Uf(ctpPZ_u$mo@jC=`LZ6UBslO=-xGGG$>`ICihm`k+oopH+k;iS(>JNk2}uj zHbTURD*0r}KE9*j#|%{)&esjW*x`K%z_4nvFl(&6*v>q68A0a^cBX;x=1)=b}!i-_77E;+t`hDVc`b_g|mW}^<)^x9j1fDz?hRMM8f!dN87Zak({teJwzAae%rcbD7<%_NJQW%}wse~? za-vHOiQ*y|c$VFc7=hEeRoZ1q+G27D)~r0A#0`_4G_-=pKgwF=pdFN!huP~(t6w29 z0m3{Dk0#5}$|ya+!b}>e5b7Yf)=@GIAtwoJYA#E7s#?oa%}<|i6jnz|0HC|2?Qe#5 ze9p-yq2RQjxci}cS7l$a+G#8AAPdskhyh3kblY`C)^ zY&A9+O(1I`n!t9FFX~$c*cP#^n+@eD`)tWZmpi}Cn97Y9ASXLAc>dsWR8hx-+177v zlN-f{S7qw*^O_vPRKsz6*;6@hCUR#LCU|=9q>8gHgNQ>@B9w^DsyE0*xDB(@KPGG( zack;!uO)d5=*Cl>KA(s-ca9J`NM>}y5B8UO;X>TWH(-KOO{*3fo#Z4zd<^Q}3~}h! zPfor?ahct$)%v$%tzsn~w}dbvrRK3JL!p#G`$ z3y^F#VxB$%vIoc2DE_>*Qea+{v@yJH)4sY3O(_NKk2`WM?Q!p|+;rHCxhNe3X4gj9 z`hW{JzKveowcBvyfj`kk$`h3wYe8Bw70ey!H!UQdOvZ#l|ooZBr%DFtC_- z_^ScIIbg9;lT8qkowL-K#*qcm2?A^VsRLmlt9-PS;|rkU!V3HYJQm6kBgDZM1=Ucd zlp+~7++#Uqn#Kf4OxNswafvi(8!EthF&(5=|JGD1WYDzChJ8lHU&w}d$>Ilu+ac^8 z>J40EyR!n1_0jNE1@SfKAxx;t5Nrq?Ar0E@oDP1#G}mWFm|>BZN5-56Vn7#RVt_H7 z;IJE{pg*df7jpTqrure;T$BUBKP7oLMnzk!<(hUvWkLy=UBmedvMN=mGuv`KHVW#I z=#@&bQZ-_GC|PZ|Y4AjVvL3_6f7{=D68?9O-yFjo_8X=jCwKmZu<0MW9)Zj$Ck{{7 zseQdK^W|S%6x`hi&KFpZ2?s~ckp-MrGyW_fA4ptytD6^sc?Xc23~ijlCm`t09mwpk zwP7_sh@}m#fK=FBz|Kv4%QbEVy%C4ANE!DOv?mWVG@Dy$K<_2&6e#7`7vG3^Ecq!l zH|r;wdH~&=`MNZQ1k?NX+Ad>3@-E#meld{Y1kph39|tONG9b}B93Rg|i$VB(V47|H z4pE36!HKPp6`r{qb?M;OZsLMAW9fCkQF;==R?7#1qR5%>(XAYgu(VWi`PL2ux#Pqq z9ttW!8lh&9J4HYO2#f??rNU<&bkML!5h6Q?2%F^-_!LZT|IH*Q3xSs;dL`-N;flHP zUA|n(Pe;{+o)C=5I&gP9i31H!OL=n9HiXnApiCIwVPKFHMD}l9ADx?t}&SE&58vAfVWm?(%WZA*F0$^KK|_weEgj z%dv0N`?dk3sGirf4lf6phr3*JZO|Pn59kIqDg!GnK%;c;fk@jX`oc4VTG;@!hdJrj=M&!+Du)FcQ^2ZBA9uglh6W^AUYYhy&= zUXKGDs^v?k6SjyL98sS}%&31IZ+?%l6F}b(_bsCIhRz;R6$XPwu8-6TE(Dj+o=P9M z8WgIa4sg+1$(E=sPCP zGq&Ct%9i9(DENMl&#pY;qI}%mW6pETkdK~+qM2S@?GvOEhLGzN${hbyC}dn5)AuAR z$wHo(p+11(2SYQ#ip~fNKE&izK8f!Cj9iXt;=Q*cT6qq2p+d-Aut@%>g<2k=LX6 zFb<)>6|_@RniCeo(ZXmW{u?o8`MAUA7vz!3XR|1tQJ-78t*6u}S`8>}&P1_<;^EK!ew=gkM0ttZwiSw{L3s&DzKtvI;GVxfPN zG3BGK;G8+gn%Z76q=B=N5>`=Zna^F0h=WM$iuj})fe@6fg)sBgycondekoZyh6;fa z*MTbV-57EoU-ljt*+R@`{th}j;+{9-MB}G=N^J9s@niXRflhNr-9rYBL*vIp0I=Ya zw~-{y)!lOaDKG)?s=Pr$DAX9){@iMBG5+^smoz`yzXoSWT*^EYraA6?8(utJoDT z*pJ8?5^IxGB$ho=EfiA-%v+GfsF%(hGL=Dv+8(x)w_bhYixb zCA7s@JD@e-tQYqB@&xJN{md)QZ;^EGqaYPnC^>Rgx>$_wXOYc7u`;b;37hG3-;?85 zTH-;Vp;f4{)-qr0irw(EC;Gw3$VgLdf}S@Jh0R-&%gKt(b|3q|@dE1z^Qfz;m3YC? z&~{X>IT$>&C1uRB{w%LkfMCHVuv*|clHQzapbE)b+zeO7NMtC5Ob})^k4m}7iQe1C z#u&)%QMN4+=<}3rn8O?kc>XqJj@_6W13~zM8#m;49W*@`;ShJ=uUHY;O}s|$L9$>z zi7{S@3bY5nkH16LjIw30mZ)Map?bqz+VmkR81&(YVW*DZlZQu#9$d{|GBn64H%~Q72k9T7Je~0z%P{03a>R7rz$zsBI6TLEtZs-e0urA#s?U zZ7!yr*9+uWCYqgzb+IQ>d*F3nRg|Hgj$6UGn;q;>!TdIvW5P@Cf)NkHMNN8ojIOtPvf>xJYi6PbB}Rn4e;&}R zpm=7!-A0CIBcHM)ivjkaA`j&8QUu41`qM*2JnOeXW?Th>ry6I+_jQ};!7~ZJU++90 zLz2+n4d#hCzl;nhO^+w-D*U6*&-()DlfZPc3S9f^PHqy9+;p;u@SexB+DWew&j;UdPQhPj?KS}|OouXO!oA49)v_3ditdw+3{jQKqGCyu<)hc5 zYA&Bhz00o_4&3?jhNs{RG0$-?Q8^5ed7agjxoNT&CxCsyWjs8bDB4!eZVR<)R!v5q zzxjxs@XD&YvM1Sw5u2q2-)Ot^024zZOGs@QpElsq0Qa# zA}tTs*6r(;(AtQEvC^Pt9&x`CN=bwaR@Syi5vBCm?8Z9#CKu= z&3AWVF*RRp4cdsk52@gY|s`R=%f*8zI z0zV~)vh;0EoRmVE?^6R!Qd#|Bmmtw8KWEmpJ5(>ZxhI%{*bA?zEG7x1S-o7 zl3q&3g-kd$(Bz536qg+x7OzjQkyeFu)GuEFn1?VK_I(aErWp!6Nz%3vDRhxF{KZF4 zRgTx{Gz{w+mVL&C# zhNyW_$JJzK+uTPpC|p|mLS~2tQ5I%PLdMqF7B<>xn{%}V66q?~h8Qb6 zR=s=ELEWPeg>SI)qw}k#gVV!XGX#jnUaaqcm=&pVF?F!v1tLiq<#coKM*o&n1e+BPw{R)j$ij*886_G z;xV0J1Gsf|q=$7sB4B8$9nW zC!gorT3|doqBy;#02Y0Vx%R@el+E-#_i=nhh9A3*w!t5(Ha>~-QLL&4AgC{zpp^ZT zhwH#f#0Vr!^3m28oDrvQ3=*152uM{#b4Cc6FkN{fKNp8^S%j%Us3lYH4{(L>lFHvjv3FEKjZLK zcnY^}ULBZzPPhw;+LOE+P3Kl?OpxoSny@d2s`^eHB}0G4T-um8Wo2ibt!#}Eky4;_ zbKUJr>flHd*rob=E>kK7FTGlNBk$h}>ae)pel!B=-6e0>Wy}Oi``|SQjlmJ;a2f3q z$Z@Ay_vC=&{To24{;E`TaQ6*5^)hB68iX@*?WHXTb-MXq{gXC-=O9eFNmM$hUk>h@ z`~U28|E>!@B+*&f^hT$S056?at*l7P&+UWfuInD|E&+>;4DI-70c|9uRZ6jlsoehs z*yn=C;`t)R6>__sUHkTV@n3GbIsIRfPU426I->umi~e#1d}jF0n9G-era)SpK-;Ac9}NIp?u^l(y|FZZss^E@E!4b3S;^+I6`VYA#C(!pUNMO_)kG1qc1)aDYAL|>= zvbn28(Cz6G%jO1_D7|$?u+B9hlE?q9^;K0}|C2Vu2l2 z(0&71WHsXU$!A;DDR)ZhFoK)Z(y?1R_2rvON}e9v;HNSj$MFX>BNijoMoGF_Edj@I z#b;KbuedvN5SJR^#5BL2l6&>cW)5+gS9@^^_cbhM=$CwoyK4|aYJK{Y z3m*7X!s*`<%>}YE`kh08I2SFt_ki2S3SE=dg0(X)rAlf)tJrKE6>r@kiQ{FtRV|Z0 zTN#@+Wv*E~wGTJ&3x{x+052jw+YGPub*DCV2_4y1R&v=~au3mxWpD^5D+vmf)A-L& z`hVV80mdi@pa*T|K|yy!=IYboF7H6}?;!<{&8PRCVCcX3MvJ-kqAxVAB%6G{DM?dg zSko7^;7g%Qf7@(3MbM@ewYfY5o2QX|f)#cOk(LFI<>Ov(hZ-Bs< zilL&pPh$xT!+Y+$TUk`QK$&7ux=CN2pQhFZ^B~aSc&0v``D{#c#s!%Nj(Ke;0QW6h zchZz;4W!Hm#09~abBV3WX0y=d-~q$JDxMgzX%hUEZ= z-KjM|FLRr|bU?GI?gt)hGWXsDf!D4&oKC>8GlT>g+My5fzH}OQ(-$?qvfmyCE|^R! zg#Tz2&()m9G^xaK>r73p+L)Q~{Gpp?O|TY9n}$$HWWTZ&YZ{l{S^@&9+eTr(rRQY& z!c^=nT^;`PK*PDPKObGCYYg+?6vi(R6OnlVST5NR%NLKT8|+JGU97KAv$ax5edZ9BGvNY(L%{{>!Z@}A2dcF&t3M>&B6KEobMl>qY9>*?@g zomg3eh@=PYDES(2gVv+GVrDGZzHKDWYc-_(AnS+Kv1nncnSLT@3x3Oq*@L{_08nac-GvY0i;L8nDeO=O#tQo$_PJ29kyB2E@$0*?T8G33(S3rTFo z0YXOwx~Y2hRhoJyw`2Ch=GEw-9Nv`-vpELe%teYi20SICBeH|xI($-Sfe9cQR|UYV zEZ&kC)VUp$1jePuN)9fgW7&}87riA|aClknE;1e=MQ0#7r>anM>YFaX?unz>G*Mc> z*Su=pkA3B#me57BE(_J~4QofQ!!6XSnNGq&P%)T;GeNM;8`p|Nz!ev7jC{x#?9mht zxdqoSGaE+L`qMIbV5(D8`QLRDbJY@GRLt?4ucXF(dHTCiFpYPorv>^@%zft_y6!cg z>?j98TpmhBu9?>*d3tqM31BYkZE#EjNNAuKY;wTDim9%+Oac1A3Si3X;s?+U$md&Y z#@?7@czv+y+Zx3~6>!u*Q$i9>ZH$F5<=@RTT2=rrxKa?@6?_PryK-O7_lYYyYWclh z?c?cAf}AvlMnOg~)T{>k(Ay^^EYArA!&!bT&r3~ynCQP9YHgo=aTH0r4#Ajj?~k1g ze-#-7VjJ3KkbrCvxv1QX^&1>|wps!>=S9M0xqSj`-&`{pjR?&3gsxHf?*sETYnCmk zYAaoTd{h<{VAxqYz?fsh%U+q4mzS?f^pW+s?GJr9@D?SRU`&dQ%BF-@ZAN_!9fFMw znVMv!Jk1hhTGlEu?_yA1J#YqWOL|>o4!E#b*))K^1 z7?!8K76EP-J-O|dlnlyBuV-dmFK1?S{ufL=w>qLvI3qI(xfPvnNeQ8AC}*r2ctAFN zy#{U@`FOa<6ljYk(`VM0cZ>7S{s|LoB)OFpJ57!{JlOna!F>tlop{f#DuQvv5#2yF z?6YsCz! zRZ2l7H9Vbs@vtz<;iuCU*+ z4NZ@#F>9VeQuLFBQPy-<4kadTjkyCF7l`fQ;GaGDrW;b$qO2^?v#r(!2>P$_0a81n zHurvM5a9$nv4V)3ay^S0vEq1Sa7rC2iu^Y$zXsj5$v^S+04|X zDKn*W)@;N$!a&KmUbgSl$kM3BW$iyy7PO`SW8$~tj;4?*gEc}cveY&ei&5D7UB?se zMx9sIC|hjgSkp7RPuIusCwj?ARL2dWJp8_{JcAb^A9$uNoz)oKp80vC*Nb1%8)%o) zS-*6^l%?h}>$8Pq&z;O6)Jo~3g>_`QO8tlmbt>q}D!rV0r&?PTjuF_eOdrx7pdChb ztOav&v%5*=#{6=CX%qDOZ67X>^ZFhb?!ATfU)|0-rf8McF7Af<^a|ZJh?-GP_^wgX z+AmWByd$48Mv)yqQ2?n%b-@0Gd2~_~95$0o(x%nht=E&i&uOj`S7@{?XO1gHI5Bf4 zxlsy)eb&JPSn)6B{4*J$7d6`(r^i7_ep2(T&kRq#Y4qxEa4lH%1oJi0uh`0A;i%c5 zflc@MIu3A%-DLTfH8&?s7IN+b*L0)r8oy1ED+`KbJ!O;r0?~%*xdwIxT2lEViHlI< z?7tVKr|!ufU^0XYBlli~=xtYNt~rsW;BnSf7g`fkYB zOJ`L;>fEekAKWtu!pruUjt0by!bOU6Dm$c@f-@`Z#NAyH`tjeTCZ<>P>+OCf!WK+pbmo)7bjxlMu|bP1vm42OfZ{;=~-bhkZ- ze%02^^F7a5|DLHSS^r{?@r;6C+&z_9DY-H^ATppO5ivI$Hjpmq{xN5o~}8 z-0fGA-TDr3bzG*NWfN3#2u}OAhh9A6Yqsa@MODB?oux%|$zIbP zrDRwHA~Y1*-wi={lSRr}6CWnnH-jPSrrML6TJm61dzTWRq*a{T)Il=IMVr8k?X%_P z^D|&{k`YAguvQ?1;8~G1RC6^aK&K#N3TkGvj}zfk+#cr8<8!|Ta1PqE9nOU=;Atro z0@GKo!M{#IOw5M6j8+Z?iNd-nfgOqKlbQlA^$m~qdD^L0sjMU*R9l^e>|&K50>pNv z2$q+8MLh*cWH=={X?7QTu}lPw4ialiYQ|!+9sm_9gcEX;3u8dsIKLo*pY~A)F0A=$ z^7G%a1j?Lm`|E!D^zi`OCC}za z?ArA1&Xm@wffGe|=Z;k%>NCHp=miaM<8JFe!8P1$3B)zHTr;TFP2fr$L{*(A0MJ8! zb9(XGIL`4*tlcwo;TvKHt(H%HE2w4b$;@8qUi2xFC@OjBl|&+h0eg#?V|kqR$+{>` z;#{;+9ywLI_&gNNL|ZH+K8}pR*(@xBx8lEnpl)U0*bGxTAQwk29sCuI;pI^!)g5Qu zg_rxh+Z;FEjzO)kw7eW$0jIz#eY^axP!yH;!jv=MQM;6>=aKHZi2V?`&zJO%xfijq zaT~IL=|GIeP3n+OLIANQDHT@8wLLnInfK`rTKw;euT?YPF62b4#Q1wGikwX01F7RM zjM|@Nzsj1Hb-OV-vf)Q5*~z}CK@!!4l*9$wAK6ZNu3pXBXU(>mFosW-KKz}KUr=znvvy+5|_N?`+g*)YFj?`E2(UZUQ za0WZu&d~JN_9fzS!;Jk(FDaZBTuOR3bu)7xAhZi#Me?T4u^OAJy7{4hLHZ-_(65JZ zO`+N+d4Y$f^lq)4Gl-IxlM@A$5!l=j-67)e z>%}%bPc!AVB~kg{IPEPrUrAFYWVb?agC2K za#`ABu29SHR?om#^X83rI6nOs?hqZDSW^F0#_y+Zp(!5!DpuUmh!YZN^X$mK65;-k zL0Co5S}Ab5Z&EB$L+`cdln5~Mwzfi~EYc$cq`I^oz`q0Lyn#=r<)_D6(+C^DdY0Wv zqqGrNRbE&oCCLnal!aB8?nSFVH{GUvu63+zU6MRnVPSs{h498rxDn7_EIU#!YjL@% z>?FNKCjl)3vnnoqwo%(fLFg*_Sa3;6njh4KVL!BYry~{9v`N0kH>maB>ffgp{(xTA zU(o+=VGHwrhAr&O%>M`8RqZdAT0%H55W19y!ZcKn_GS&JFSuUnIvIfZM&#XT}UN1p%2p`S8!Mi_; zAXTg#!nYUE`L)u!J&T~1r|aEk_KIYExO{tIB_iaO2tkNuS=my&d;Q!Bf!;LY(0XU% z?M#fMhJS!W&#J^E3D{}=_VSZZ*AdLdi*?(8wt;ocy*?Z1aTHbo%4(sYLHEw(45q$z zGEEfi)eMGuB~k212bByqeR9?bN_C9a?QWD{%!u~y` zjy}HK6}OUyD1H}I$*6M8<)yl$BSbe=2{RH(wJXz3_nVe%DgXHPF}Lc|bgtwKq4nzj-ys7NRDhSN;e~ICciF)&q?Gt?8;Y+Qpj+ zv}GTG5=Zu+CJHsoZ2i%4)hlbVWvj#7 zRf3hw{cu_a0TDuROwjh}()(V_>%L?`Q(o_~p|n@n=>f#j!l%zUdeH^7WvXQ&PA+$O zgZ3r>e`v0ywK7oMSc7DOTLPwWDgoZtd+0(r3_jU3zxN#=!nw^kXlNsugLLoZ%yC?4#}y-lkrTz3l3Aop@dz+1V!V3a@n z&fruLmj#kGV}44%97yx@77DD@Fa-Vr+rfTRR?Hv1WOc0Iw;m5n*{fo#?~7_ZyyRs; z4M}-tU_o2P!mIb2N}d}nKU!USa{eZgtbhw%al?opW48vHL)M_e{H41umcWOY{gOg~ z9ZuFh-0DO|#YXtbtqgiO&vQpwWjh)#8wEf=ez8K2W7&y+T2W`Ve#*2*Ck%ExT*ShsW zL)RD(Sg%`yCe548ABx4zq>$kJzDnG(3`>K%g1d7*GDIYMTyt6P0@%lyQRPl9Y|NBp zBj&j~cwkQ;b>Ff+aw=1`Llt%ec-r>?;e)9mMzG#YVvf&LCNgd)B?&)>zvGAF+FBh- zIY~)6_@l{pzgKCLHTJ=TNvL_lSZF4JCIdt1oOW8WKUpx<@kr~Z9f6Bd7&C;Kc zy`v+AR=TP%`L6|Q)d>j$(u2#!he{1k`kcI&czOBx#S-^63m07t0(u|SWy&bUM~A3xXkNoYEB?zFr=fPm%%s8)X<+cLN2%RlLJ6r@!( z?udKl69iifV=B@z?*kGom?QW<5`ko84*BSe-L$<$!2tFb;~CQop-ebL+xI+HHxB-e4`qVJoi!b>_GmYu`-_{6<1Ig%=&`7-n*B!+We*zH z5l3}ig~RwyN&aqKeBM6pW3q%Xw1l?qjrMpRZ@+A6Vs{w@mON8-@$GQZ=<8DijQi7a zbPvP9>P}+=69NaDTHIGN^nsdwrqsN2W+oB*(&LGs@&vlE!zYqguu`G}@R`vG{I0~> z+V2(i@IEhBQExgWZ{@fE-#~mh=2KWl`S-n^W@DC>w-gsci_%rwgD5tp+EYmwtv&b} zi8-bX$QYdj{TjQ=>k2LVjB&3uxl|#LPP>c}nwtvTp%Mo2`bf4HMTj5|DlPvXuFVIS zL1RA>8IbKT*eYfv;QKzu&MVZq7Ee}baq?n(s`WTQzp6YTtM1MX`;a zb<&yG&9i0TP#&Dn@nfkxpbG3f-LEng_?RbMTJ!9EymeffCR6-V))yh*5G%@wkDf+& z8O-j~h}eJ%;NMHFl#jl0zTVmoskC=-b9SpPPG-3%2z-fkiw|wtJ;kP+-H>s5pT258 z*YQ{QRG$nx*g`F&h-mL)FHTVF*h^z++;ERruQ%i+y&xxst$XPo<57wWIIL5)hSW+& z7WWnUcMFZzwHG;g>&DeuxHpyIM1LCr6*3%K5~aDCQ=;_G6TXY}bu9GY zAKBEGYrsC#qp!6~8+KM{z&iynhpHzofR|s{GCn(v!Wz;q0KF+PEmoLNylgL_8AsuV zZIC~!WFZ6$9C5&ds=v){pPep)ltCLa!}W&)SG8raVUa5N@A+Nq!gN7+pEi-{64!DxNEeOrmkCzpK`V{YHHF3M zEG2{bEa2Sf8F;o2q`l7OdpPfn)c!*K(onaJyfL42Dj}vRp=_q##%fHGv->oW99k17 zB+Y;`+GK!X%@Eg1I67ct%=XAGJZq_dwrXoLPE-@;7dGHsIaBBJc&tL`HNILni896A zh6O365O`r9Yq@aE?LOaJXNVMprPj3Ixv}d8qxbXZ8k}=b7R@HhIx3*z`?vbG>QkWf z-UGvFEzJQ=1(W~<{`$gM-H&oi@yxW(S%(>%E>jPHCjR$AlVRW|zE#R!XP6XTpO=)t zJVpX6vU{)|{FXFdLI}L>v3hxP*u|%bwVA2_CkRW1eoE zk|4bFehes)+H&Qyw=^2FBK)-70?#8Vtw3H^J%pbGP>YNww-g58vINz>dt)UH)(OzZ z%A|I6pAFv-zw8;`+Sct&Uj;bugr2AY_jlknV6ZCA%8Gxvu16jXcT2DC%nhN;OK1aA z3t+|I0c^2)lmXO$clIQ!8$d1#?$`q`mvlm0gw)BMT_xe7|& zMDxt^&iX{Ss!CYQrJ%jzynd>KH!UG-;U!~_=*u$-dVsprB)}ew}n^Tl_&yhP6a&rUSI!0r# z{^FB!iaLB;VEfaKxlrS*L!SGzj-#?IaZ4aJX~|eVVmF& z_J>(6x4h@dF!`;(*)_|6SKur4Y!i7n^CxHZseUFdj=-q zDstlVL&st+EknP2f7%AIQ7+@}@DQi&@e(`4c&uE{KYl!I@2jr|1s$Yfwc1*Z1nLEb zg>I(<5fY&?6Ft$ywBFTk|J3jDr`ptOr)|z_gCio;Z|npde|y5%1IoKkjo5p=rUtDA z5_~5HA5vYB{)@49>e4K1(=FeyZQHh$VP)8k4BIv`Y}>YN+qP}H^Q+!_bdOPWP`_cF ztTCVan)6Fwbn60>%C#n+E-}gz!lcOe`~Ksi-qocl6Yh>+bYnvi0}=LdMA~OcLP??u z|2~QU5r$OXa34o>jl3hnzxR-BJTB)v1Fg1~MfDuF@Yx3A(1K!Ot#VtQj=MvuWgqs$ zCpm)4!(Fsx2G!#&S?wpA1+@$5_X4VagXVUDPsfH`vGdZKR<~1HbmrEN096-eRNxU0 zR{TPlTUM}gk2B~2-^g^$QHtAWlFT!w{Ki%CLE7d}_Y4I^%~0e9%SI-Cg3C)H`oM)@ z8|@=MCQ-{ZG_|YVc zd`D65%bdjTBA0j*bIOd7qqSxp>XMxyYuxXh8M@engM5-1VX+di#CIe8k>U#pUs|lb z{Z5uu{O4CE~UxS~O)LxmHL>Pn5m7$dRLDDF~ux z!IbHK^_Rz;=aiF87o)Ljt5G3AhFc3ZZB*TpH+WlmUjQ~c3N@2A|H9R1_)0E1S7}`2 zj!t)mIdPhUsV1P^AjY87#|-K(JU4P*?BR(?ct3=PYU*10KLdC3?DL~LM^5+<7=D6- z4_9fZc(IdK{OkG&rt-tm%FIe;2|-G(pI4A1{q8WjPb|}L z#;Yewi=gy(Q>o@t#g}CWIXu2UVHFyK^CYH4g6u93#FgHwMs+x~>c~8;ac<1hsK0=N z;KveDAB4R6KRh)so31_p16da0r-~2Kp#h?+vQ3)ydF0i9kRP+3 zStTjdi?kDu>dEJ)~0yCUKz;GSg2pOsAvGFMDOtUS2n#|@=K1NbQ_?LqM7 z%;~vgVs5caDRviLF}5>+W)Eag30ggvi7gsu_+FE*zY$7t6cJ{j;CKatm#-BK1PUaS zC92|TSX;fu$s}A#O?zPLp-aW^K0@~0KzP31q@c-H?lki5}bW6+R+jIT-X-~Qx_P=5W8%PY!#MI|cN z+b!GWI(4jVR_XEj`hc)DcF~hhygmHYfzEYUlF6`ujr7}}1gp!^5zt%G7uDx|dcQEF z6ktT*)kc`VoCa;k2lkTwJ|Lybzxez5w(AR6jd!2=pAaJRe@BR{T^HzH!pRkZ_Hu9vketB%9ASXm7Hc0Tp1{@{%o_?ou$(X1Dff|oqrC!?w z$pW6zlHTr(P2O50({CliefN7Fri8dP@IPI8&#p+nS$1Ty_3(JT8IA@iwjo*C+PNV4 zUI*Vq+=$4RUqvkV+SYMv8(vHqrlX712=Wr}9^-iJr`1ngn>h7ngh9$6=w$MnWR1=4 zXlRQ2$+zA6s!M)aJ)ZPAGt{d^ttPAOIJ3;CaC0V&;)^#e-yP!OA2N8CJGJrT>~8%q)@0(bursy)-BaDRV2ZnyZ>lTzOIB zN=?_R#0?hgIxr?d;O#Ry@wPUw*90b_LcVtPdkg!Ltnq3cNc+oOGS8g9p)U4&maRey z^x9smp5*HUn{3l?YTg(`{at$VkezUXj}Abax|G$WRa3&&wwR6g$u8G*Zs-rWE-{VR zBIQt#y$H-Bvs82Uk^uAM7)ldTrmff@u2hfboNj2QlmxI0pj@GyVuyg46R`WdZLn7u zke;Dk-5&m|cUlFQ6(!{Is)Xq&IO}J(x9gr{{Lb|>rO*e+hv#le)}q<``$~q1`{gKa zG`y^T^w+|(jyKxP)(y8K;i|Yw1Dkckm8HleK7wC+9~)C{Le5xdE>SA1SB7!qwQT~R zCqXZ@2U1z!17R>k{lbSJ!k>sUQ(H7+4T~sY;F(>(dqz?UK}xM7@*|^YLwmHH2+(-K z;FPM@9=cJ?6rNKQvJR!l0y1oJ7{X0-Ppt2XC{tST5YcdNMOD!`mOSQmX=d%KS9Rjx zx-pFFh@5XV>#o3i)Xm6OCDT?WK3;hj)1X~uoz2AM1SK*cr+VC5&u;B@lZJkX7Bd3y7-jb@D#=8TM>qw}6a zh-f=1OEGXKJ4ZJ@6uN?a#@mpN<}+C%RBwTL5gQ{MO|t1reU9*F-|Ax2k_Nx6)p(6=?+1S&In77hm z5HT&K;ikA>iAZvCKHNSZGgJmpoadK0yBnFflizUItmlSk>W$1^oCb*WdC(9wUro>B z{?Ei7G*&=9<_K75FNZAwQZzPD6p}?}C?}Vc&!WL^Z8yNl5Z2cj8yg1b-sa%3Dk?piNLEkJavD=ExVrFlz)f-++6&U%@xS(Y^H zEXo$$*%e(r-sN;a8WOsF<01N{DTEI+*-P(NH-(o2d}PBIND$SjYRC#0p{Sn>4MC5}bdv|i9)=RY>A~Z0Z((=uWJrkL8-lSV z_dEMGS^@(?Eg_+pdNud;L}GVamrc0MKD7sdBWbiws(6W z8#xmrgSeIh5dds)o7JMh)pMhJU$eTA%HdZd^;im;h^GP zDXV2*<6OT_x`#ILgxU~p8%!Th$U{CWYa_}lPqBawPhpfGnKq&O50(t`NTl!xFBoVcSMyRkPj8Q25f8J^=)~U`r_(DK)6BC<_PelHsyw_zyX3%2UJ9lgI4#iK<) z$LN|J=UL2-AT3FL03?LpM4=v_nAtHr5{u#t5TO;$QkB1jXs5Z%asj!X#DuFuF%sK& z@ke-h;1utsb0%N|7mIRsJi`#^8+FHvMRI{nesnlLer`ai#i|v`_pG84b?b6GriDL% z8g#o*-LSCV)`mq`w-)pwhL52%wF8^N$e9GihQG%m3qg+?$>b!-msl_KCGBRTmrn0T z@r@6$$1qm_2E8}i$=&2wnm_bovF6x@&iTtamDzs>e(G9@1f$yG@dzj<=)evaZ7>%I z^qg$JDAxp$*AVoasnA!zoW2a6p69#;c8c%MWb2*ylFAA$Qj0ZCIUeYKkX5@?a z2*8?yvbG(xoG3+rgA6;NHNXIuHDHABczjozZ5(@Oxl)B3QNse=A-RAL^5$bif887@ z;NsaoExCB_z>NoGLHV}A=-EdrAb@UXlXr-KjpENCg9_KjoYlFq z?P{Lj`!W1d=gBX)wttxw#Vo;!$^jnf_)!)KZRwr}=M{8?xvQ4`A~Z%~jzzrdeOrDD zuaDg#!f;*S4?}8*+kLk!w{CM=w0Af)sC>^F^xqu2mFd3E^4sB_>P9ttn5TZ0&NaG+M1A;bwY=vx;2~jJS#(Bk=wh%|>0>$;)^oyKT}z3y4g` z#y*~-BuFh_EqnbN{wjj$_b9H-d`mmt)_sm!H&3(_&WXhzV1iPq(TDD;5t+}qIttM1 z@WD>#*vPQj?#Gu9;{+ufTUsPW3wu>Gx_6*L^(fQFejG$!_p>-y4wK6y8)sor5!3R*jBSuXmiiRhWYKicCpRE7lAS|5HX5J1+UzyhoEZto79w3?W56pYVcyQRKLIfnu(%ug=Rer-1<%5 zvw0J4>HF##FBzyk{(BTO3V>uX0Q?r}yQ5yT3O3jfFJ~sSSy3~NZnk!||AzfzwZC`R z&NvV%*ic2H2$a*k!$!1zNI8Y9j?yKr56m@Fo`KbwWzg+386{L|K5u){lho9#mb+s7 zlr~Qn1ar}8`Wj)(@0pD>g;192){i{)z)xqG@*NrXC+PGc z68d5U_m2qPy_ac9iY+%Xai5WMI2%}Z1kz9EBKPet1M<^r|NhOL9HS6K_KQM& zD}+X44OVJ@M{jsv8<6td&RhMGZv#2a^i%VQVlzL-jBZ?5nL={PQ>pbGq-w>9^<=I~ zTgxh)LPu*_u_mg^N>(SBpt#In6ea`E9Y;<}%%USWO685~h~GRTyh{BsMsnakEXv11 zT%~!&RmTFLANh-ncCO^|%bNMG2}tq(J@)H(!+x^iNH=K-%)q=b)O8yBBiJyQ1|gDg z92d^W8?zq4C@||AwT3Rmw8V(Ur<6vT_(#aaaK|D1&6|V#CX)gmQUA;hf0&RBZ;?uS z@0eBLB=C>b3m!7Qvk?^B|-bo$Sw~ua0d>B+Uv4Z~ zff0?n-e;7CjRB*vke+^%p7922*u#c1;uM`iUHt_ikmN~4{$VHdQ-TPa~)?eE&n<-)>{yC!+UuY$EPvaZ>^ zbq#jkLwYa>kZYyMxM)Z_CIbuv17G#qAEY%fk-C!5dQ@{is75_~Rg^|QJ3%F;xP}PB ztR$5mrgk7}Dk#B}yizhRw~w8ajVJ6(w%`f$Ix?q=T?5uUs@~$w>Ua2SR3$m&O{zgj z!f%&gJBxv%zyg*}>(Ok}juo>zBHR>2fsymbqI^DgaEoJthMm}6B)1ML&<%`oBHjcu85m=H)d!ZeB_l{I01KuD%^O zc32dh*&&5YrN7!)hjI&{JuY(*R#3mefbn*rXwn9R*1V;rM- z$^FmWbKl$IVY)%26@!%BiC8R5r20IIzbodE2HZVqBa9ol_sA_{~B;2sBJ-DJ`j>;}%3Yc3KK)T4+D>0`y>{tG=!^3D3qM_7dv2fO77ccv z3{$vamjEoS@MXgntR1x@L6672@&hPg$QGnQY7Pwgy%Dy%h>s|H!U_S@P{@)}t`B)V zXZoiLp(t3}PK4!@#*ui-5!c)e5s}QBB!oHz(lfP%1}Q5V-X9XtVx)>s_1*a$bK7x8 zda~ZFMlgsW6Lb83jnA}e-U}mX}1)Z#k_-#fNUaimI29|9!oc4_P z6!?kw2Um75G|=Yu%#&W9`ToGJS8p#nK9|>dMfDvab(GAHAE7|Ct5@&u=2;*W5X=Nc-yOVj(Un3#8TN#4U2L5h`GhI%F9Z~% zfmbg|pHI6x3kLlh{}Z!g`R~k*jfwStXPf&!nau?!tAET666IG{#brXJYRgr#QBzS? zHk79!2(bZ@HKKfA;o|kc=L*-A!0G}%J4?kaA=nz9K?nl4zsK>x1Ve*VkL-MqoZ=_O zPfiW|{Mb~VEkx$_Kik~H(&5`u-<;OJ*yi1TVw)j>vw?D%azPk|hEZ~b%Lx@rK%Q^U zM=uO9C2BptD0ogu-1jrnZpF$_u?_!R2-09q10@(zGFTcCK2QtsE%I72HH^fayhpz3v1R&dMR%fcs+&a$z=B2RjVc4 zZ`^ul5uY99tA|(8;5xSHu`tZ+bf*lGU~+I9GF-ol zSvOG8)&Vl3LMI$`Z7G+{ljUAp*DjPkB?f<&*Aw)wHsP^T2PubCXkr?y8Ya~meUaL{W5k`p&6Bu~CcGWZkHG^Xr$Oi@53$;XT=wn;&w~sD+NNi(AR2g`D>J z5^hT5f=@!?EiF(UC;LVN*-e4@I7Ro`203l|-OMCi^d$wrgrJPsgd1u(qkk1>?ig)Y zEHVqKCYN;U>@z`X`8Va4WPX$7*oJpdaYaKgSPGawJ2=CL#i>2Y`dQitdN?m}lnV4V z*cs^%YJcuYTGm8ZJS%N04^1&!ZUidm$nrsNa4s8VdG9UKG6a?li|Uvdck2qN#YZ$l zxoW~jK{t0;hqF<$Q2LZ2#&#>0y#|;=?v!0RZRC=|VK(a5oFeHA*K)KNLE{c)yUHrg zZ2x8xKLoX}N?^*Uh0Ya;Sp&r|Lxn%2N_B_UwHq~F_r~zkHpjjK+T}nhi+(2baI}lb z2t-TQerh-dXd2nUE=2*|qKeYUiMgZV3en`dLFhKP)X@>OfIliGiqsiULY*frFPE0#zk zXgQoQvpOIX5R$rhxxdVBX!is*i0B2%G;6jh+10)!_JS3N2%=wn7~N)9FZi7uCM*I! z{YAu4FAARuO*@3*2fd6PsTLN1_}r`ic>w+gw|x`i66|wGYDGNRngs$(BLuU`z_FIn z2_@S7vO?Lzndn6w-j8zti6FK2%a_ zff^0&^{;M^GcoRMoEG=a*=!F-a~CSnSR$OfXyv7*j;gsTJbu9-az6sxr)$M+hMcjm;IH=`J%!(6G;9?!@igz;#Q z@gtIM;IsN88yTHIvSvu@Sx!5!Jksm@mmy^^%;go%{4=E>?wC9!Vw4ci*u2a&-z7FZ zz`VLL_Oy^hR*WYa>=fK^xH)tUqIf0Z-?!~q$cyewMY%R)6FwWq8K07|N*yC1hrjc< z0rA}H_9RoFazV)Vb8fNe80KGGQ;&!mBu( zwMNG90;re!`B1YkF(L+cbm?(ew|c@>(cX}rg!Hon1LkQVd2r`f z`}_C2lP@hj(PF8K6w9GNJ{4xUu9V}b59#fj)pkyeGBdpgirZhnZn$wnGQb(C&B1lM zp_fl@V@HrdP)U9@A908dH~@lqoSFLVEn7cT2uwhfr}wD;ObCXVe@d*jN}pgEociJz zvlgGZ7GQT9BSYXG{Pis#P73!-I)ZoyAjpKVA8TvTOH%a5q^T4pR~8E5mpiC*j=usS zksq3*mslOhTkNJ>hal=s$?h}aD{GVR2>_b>d&3*7{TwxO7q4aicLOLtwrhQUU z&hk+DHX>LrL+HuO1zq)EVWz_;Mr}2IG!d%eDz4^N2l`mmI#H}+eiD48mj;@jN@`zt z;wQnMzrXAPMUL_T6 zL5kfV!6_vrC4}rJkW@9r=xmfi8i~?waQa$Rb0Aq{lU!XD6VDX2#l3_)<*|{b`MV5K za7?|$qy{$r;)dU4Uqn$vyeP@y`YFI+-K${F2U_E9qcv>rY)?vlM)w(9oE*^TmlT@q zgI7G9q?O_spEmGZ@-5bkBDu@DMdy~BNJi>nmy_6Cs(vd07ks4rhP#bB5FZchgQ`@l zMLZ1m7I7RG?Z4`j@U6y%3K=nbRwJtETCLe@Y9M5Fv~G}+UqA11uNVTEbeATe+oO1> zE{ji2)kfnBmr#Lr#Y@{YgsYK`^J7wnjT~>oczvR?v9pO#7*z3W*OQggVNj8Xvln=R z@o!(KXGUYhG1;vMmxO&#YkLs|;&mUL>1V>yzSWoU)Y=LhutEx{ZP(GQHNub?p8w4A zZxHgyosk>yh!_*^+uXxlTi|5G|Ufa-7DplTvXRjmucE42ZoAPThq=DI<0@5h*&^z+LCYPYkOs& zz2XJCavyJ1b8p^e&-+Cfpd*-Z3*v5$^9L?f?;+2I5QB)K=#-AKQ&LE!G#0rJ!XiIjBmER2n8IkacxF(i@0a6weRJ4f5*upC7u_+mLHn+#zkAkHWeF2a! zcFS@gf_O3H&_~M(VG~!@mNB8maL3oIBek_0JNj@l*tCzrf0Nl%N@w+j&EAN-f(i7? zoE!XTdm{Tl5r21VImjmLu)bl-c?}nY$qW&uQch4&_-I6tCm~|V>V5mTJ9zng=WG`+ zaN`H+=ZJydE=$mdr727~+>&C^ewpA&-_*g<+XLl>?}j&GPk)+(3Cy+`81J&HEMxe_ zsVj2L{RLi@)bSn_ZoY{9Qe*7WzcPz0Fc2bD{_rQ1i9=Gqz`9+#IUcgheelBh=Na{= z3KhBp=bDipJb63rw&AZH$y_7-{b*$<&lr#@>{(mZd+s`cXwZPJD8VFw*ak8Def6nD zuIfJ$$941OTK+lr`AqeDl|?l4K<{9$@OIA$PvRYF9#TRckra#&sKs;ft>hC0(j|O5 z`&WFuxkv&fFVWP(FA{ z(vEBs0w6wV#_!h43WxI9Kx-;cnAh?pXgR{OUyO9rN2p~DJc|l9Ip1<}0@mcm=Jx_W z4xB44sL68-jH~h*``T1MF{$}Xc{%*khk_1uinc?R~>^M}syV(l24Z znegLo^%Hv>KMWj+%@xY-=OpM{NddpeCg4=-bF8D+KUQ;=Uidbsm@%N?=lJ-vnq96U zbiDP(uu@AsgjsI|-Y9D>9|y7tPl~v zq*R}vi=39n3+#uq5%Nt#HAHs0rj*>&x|(C?cY&eEN_%d5?Ws#E>Lp@}OP4Ed%V7Yp zOxhdnCx~4G2cdxx?GPB=<8>BTW?S>H4LZM0aTjtV215(tz9N$d53p$zf%V+u>(!2H z&miP*#XvvJ5RnC;?tyU8)0kq;A6ACBu`9K$@BU7#VS&P-?_{V=WUOuf zKmimhA-xr5LHpzdkr3FSY}2WZjX(g1_O$DP3&HRoP&MWE79t`R(6T*ToSi_!J2O`8 zR7e8hx(sf2>RBtOM#dJK4p_N&LyjJlji_@&o??f@w|ddl zF=Q-(7FctG!<_(g3Iw*o$$yvdkQSr=41W}+bObAeLVmaO2v!MU>OXHhX0ZFiBS5|a zKc3?E=2JV>moR^@cC$G-mVR7XJLkm-p%4`^$UTg{`?->QL_5{SNm$`eB6p!Cnc}YZ z^2+gaZ}Rq|j|UwgcCmENo|)xdh1C2LDbfTOS}Me^?^AVIt-T1;#GuIhwkE(-g+YZV&yZ)7QCIwoe|Jc~8~v!SFbWYtXZfnFlAJ4c!UC;)8Kh+i6aFiDWB)D#0qIG0gUb7~ z+`8deJ|>;(`Jr~AIbSu;2xZ4(1*`^zCGifA1K==+mxPW}(XFjS`7rNc$egA@=v7Z^PO zC+oT+-xcOJi4U}l>T|csc^&wMWMLQN;CA1#6pGvkJm)z!30`eqw?zKVWfw8i zqHY0x$GvebPSdV0?@vSIQ&+?*IjSYf`ga9Wef)=00-!M^m) zpNG9-6>c;nD&GQDfCkDWl3X#7Y$(M-IbUjYis76vxB$ZiLdHj>z|i!_Os(?%AZDFn z7uEx!&FkK+&tnB$k{=@Q*)++@4&#hn81T%AUx#+oAy&U(S%q53cls=d$eKG>W+fb7 zEvztF6W@o+z=g|*6C~Q|+itYP!Y521usWuEm+Qvlyq!%u%psVp zzswu8N%r0vD4YT4RE+6c8<*iFpwu<{EEXOo+@7w^snPjy*XWT5t(}2Ry}AS=Y_hx-H8amHOFpyhk3J@K^Az1T<5tm>xr_v_o|b`K`qCgX`kI zMiT{ZEpG%Go$r8ED>%9n<72ZtawtE+yz^}>uy3bK(|gX5o>$*GuJIV#Xdmu)rx(s| zEvx7pJ=`p>#rM`x=#tI5p(mu5*pa+Z<-El@SJC1szV21rFf{_bud}7I%YE_FLvioM z$zOK(?>e+GgN`=2SEa!dl}~NQ=s6;MO8hjx@&=bKb1-TMU{6>Ws=}j&z~D_;DpX=e z9J7wvXleP}>6+E7l$JW)#`96UZ zi5lJG_te7^uw-H-Z^@#DYa4B46G3~rH~oJ4S>Z`)Ag<3mEpr_=-5`(JDxNb)Q5N`l zVr646tL<_3p9<=mUtMgW6r z>?hRm!O)3wSC@ZMuDK(~heq{@Dcq{&9!!z?zKcfG1I?i8QuFzRs0IDd*?Y11g-_F2 z_T0xwy5($Cdu-XY?IDpB{EvG5D&ziy;zLF{F082 zcUpQvB0G%BD($0d225Etm1&Nd2Cp%UCEV1n`dWlWV5nx<3d(4rLN9vr$yONKX!`G$ zgxBu&AK+u*lBp=PiB@w8f+XB0CTM?B*#mOxw&wInO?a)}EFvR-0%aq_c4$NY3<#fC zPqO4)T{$INsF-FJ6@7KLxs^wrnJTmnla;HTITd3V`t*Zun!5eCh+Pp^E&6gB%$W$d zCSbYEC*yy=+Q*rC?`s4T9u4#1;+W?#speJZXmPREa4`%Rg55!ko zm?6sBDu%zkEr&Ta1h4Feo;@Wc0GN_GGJZE0#85H1K|7Yv2VZ~bKF!;g@^@qIbu{kT zQR9=}_lxiMCnhk%p)6e^6qqr%va8T9#d<1r5v+!D>^dv*2#}}%3&2M>Hcq;6f+2Qo z$RBPXznRAR5P)D!9eZrfBnFgjVf3vbU$^Kcp^{*%TVz%c>nm(-2`8 z9+1rS$p$1=ErW`dZTS_g&hu#&WBlp+R=lp_X)Ye7AvBKHXkmPeJ_zm9DUm$Wr?5SMSJ@$W2X@Kn)*^%{OvsyswXNubYg&8OvS)`+G61Lu3p6~Aus#^T=aaKx;Kz+Q5f^7(oi#no}g(<&q{ejjOb;D(sQYtJ~mhpB)&ksl%i^8fWo{QIC`rOqT2ZaMf zT=Lt1l!kLQ+^ubxnvo7 zJL!4*vG+)mN+J5ZuS!Cg;i_De-`M74WjaM@IBB>#no5Jtew`plc9aa&&k&Y*+yx z(^ceyLP}a%F49S9xm-u^5D6ieTfU0R zD_k_o1H@Lg$v3=Lb*e7loqeV&h`J}})~LtZH}e3&s&Z+z3V(F2-WN*%MMxjL>CLhp zgkgNP<^16ci=)Myk2#*GErcuc@uDGxqgeJt-LOt7%_^F$!4civ^cY{P<-*O-V(ucs z>oP~<2utZ085`|GvBC;Kr~V?KmRT!4jd+I$cWR>8CY_SdyN<1q18wNFA~ z#d-<|i&7_>3vrNeB_X)(wPfl=Wj)#G#g15F5t+*b6iPb=A4e=Pii^*==5`gAf*2)n z^sqRWhTd`dv-C!I6xPr%7Xx)POFl@LEpFeVA2FIOL_epClDM=?gE7em;}jW%PN>S7 zF#8-jm~zsMEk4vAGy4PYes0PdGDv*w!U#~KUfCZ9 zRGfNBo3l~07<>{SLqDR!Sp^`4wx6sMsv+5mp-n3k*{>BT76xF@x$5j}%MFwP&wWST z=OvH(6p7{XscD5^`#*V! z=(|TK7FUV@nzL@=FXfQ}_D#!>&hx{yLdxbeR|Hc8jwDzn=ri5AN$LAtC`b=FC*jCN#+Fjo}LWFD#?g4uRu>uktK5c?}C2DvufPNhSVH#cn!K;9X?))G# zBkUDaq|3aI$>}Z&fsp$2#_aD8)bWUCxm3cCpD59&6zPrNza^3*w7w@>?>t<{orv0rC`N(C&f*_9*F>gKDp&+B*IqbC)C$j~^DIbLZird1J zS~@Z+o7XC{`OCHkU|wG6Gbk^Xt%W* zJ+W!AORK~?oMU!u0J)n(a2@I)G$j!ue=sjZOCk$QYn;H~REtWFMd}#ZYQBc}p#R{J z2{J%vbZk!Jp!BpfoX=So8fup)2h&yFUMAe(nvar?g5tCa6KvV_)NeR}PIuN;ocV|h zpFkf7pxs#l4eRl1cFr5F<8ovbL?_5sEB(a1jb5OJv*YjA&eMd)#;iGz!^oBsTn%Ag zqUu*2aLpbqMv0(tTj^)Ay3aM_)kR=HDm}0=ZS0@X+OPF>Y$ zD9!S*vPpknF96=u6e6$9^`rQU_iTia6Sne90o!qSG;h(nyY$SAMzyR$@<_$cMinJH zPrOLao63)y@us%T-Ow>-y^nY#nAiwSm+n(h>Ki*S;?w_?o!rz}riV+-bBv3=EZ;41 z9~VgV6(e+YefXwHSK2tbtTS3#pD^_WISmPy#?hFu018@6xT0tO~cLkw-9Dba)+W~i0 z*j-nd<`CkVQ^c4gn*guXFG7yvz| z9W{L86dFbv_Z6JCXu{8F66F zAmnf}i#-&-7nHtRdE?v1NB5_b!%+ZrPi?hUKg`wRo!WUo+0j!+G}m72QhU51$eq+jafQTk!>?y zmPY}<+WuEF!%Oh#*{3n6zMVe374@SmC+?dv-KLZOO zP>q2T^^CTw1t63xFw5#LPF2>Mai0-);^nU4yh&n61_zF>Cula-7(BX(Skq}a)Ycy< zEwfL9vGV94)Jm}?yk52v;&3N~IGzKBl#W0*980uRtjcqo0IEuVfKqH5ilN})Xn)B( zr>cu~svxA08k(6=B`UyAL$4@#gGHdEbFQXA9+RV?6#qjY?QyXrMotWZnt%0S@ZBit@o zy(9Ccz z1Yn4{R9q%kSJ$;+HNRAj|C1e<{ox&BagRUfIlDQ#rxZvd_hOj`;0EqL;7W=k0_WeV{+|m^ZrUmNgjuuQVNGZ5 z+{o@5;jU#YZCSIEVI_gZLjm!r*y7?CMrYK4hCei7W25+!m1|;8Uc4-`G`u`(^#EU{IcV~H5dF`|)-#`7dzAJg`$6PdQY9 z4U1f_eWbfcVD62Sl0wEH!RIE;nD*T5xb|SdQbmEiyi1Y_KB>jNWWmqnDAwmWDC%m> zUa<~s3yA4mJp*TabKbBjxoXZ+Le37$*Sxzq?Q(0UEM^sBRQ1~Adv3fQepmBgT-hHY z-QP??-dJh`JF#jzBh^$KQAIfwCrvdj zP~1wOTOT5A4^)u@E0SzV9Y6GizjvonGD?B{Ka8DIa3)N&wcpsbZQGt?V%y2Ywr$%^ zCbn(!jcwabCdvQZpHuapuIh`vZ|r*Z>a~~w)FkyIv)mSWhh6$p5NQSfB-9hRzP%u>X4(*=VF>HLY5(MyVI4$qo z*6PWoi2dN+U&*TV{($_R3) zH5|lAhE>5;NJ=%fRiwN1m-F|JxykB*L0AwFVJn1hgojHP`%)PlzE0%KAxizZ6&i6d)FQ(=yDpPJwf_Jp!#Hvv)Cn zI|6a$k4Fn})@>~7QK)x0j?IFdS;mDWx*3EHK^S!-cc}&i2R^_ESi%L3ounOb8`662I6N=c{Y;`fWOg$uM zIWh*wJTHi$te_^wLiS}y=4iksJc3ZNWV(N3agvX{@BelFd+2}^2BA7amTQ_xJ`$yX zipNb4B|rgI8uwOm*>Z^g?JB47xkW`HvCCyOIg4)VGlnN~_~A`tMLyQcDa9ksyO|`t zClN3jDb-D3hmVcXJxd@9+#%@_N#VFGwTa$G5YEZ3oN3im{UOW^pO21EwsC|K1}slKhEVU*F4J7naE1cZQX`$9h(rphBGumSrgGJ~G{ zAk89Zi(idvVZA$~A&TK1^N(rm*;<_Er<~L{El{OTrO~3{1&RlD#<9*wV&^mXToTPV zyCjiN1fGEM3Ly5oASQ3ZQJ+>o7kpu{h|eai7J7~Bee0APgP#B1Ko*8r-Ttl0L=Ura z9Z7i)!=vSDT{%r%z^*%&)c61pRgOc=v5Ok1Q{;q^Z3kYUA)QcoYk!Mx>Py!iFRRlr zsgXy1O1RNOooQ7kOkV_qo&s55yBvCTIL!r6?ir~if$YS*JzUUp_B*x!+>n^_f@E~s zm$C+rx%A$`h-4e8P)IiKji1oF=HplF3$`Qa9PSam?lqDj2TU^#C~^#B%zFKp6$_w> z2S+zgq4>^UWg~d*6H!oQ#{^ihf7yu)TOjEZRZ#AZf)0fe(kaGA?!~m@)<=R@Y5|&V zd~i$fmiS8oanx55+$$sYhV0T-3plxs=8;d;76munUJx01%b0#yX;jNR?3VCYhGBA! zs^HO?e;0x!vaUye;;Pc}IyJ-$`1HCZ9ANAx*IWlXUVSc}dXXZRgq{cjyOj-oMxMKc zNkBo?v6B|H6niIPl)FWm0c(7o109czYc6BH2v!Uj3lU35@H1crMUGEZEz=Wq69u(5 zqsb1RE{cBTnLm2({f&^7Okl-d66NdVqdCflp11vEPl>`23Pk}60u2;?HyhV6$C;P9YzOjReP?go^S4`tlL<2BAK*tW&%)Pc&D6r`DdW=! z7iKOK0!Lo>+LW@KC0kgU*Z%{wu^uS30%aTGcH+GT?#iip8@%dj6EqRG*NG&^1K6#q zo(g6CGP7k7-92%ThOCG>yMc@M+CiV9SdQ6nUT8O+(9$W5Db`Rl4iy1PMi#W~e=yx>rn{dH^h+J>M(xRXfU8 z)ds1i$&wHE{M2S{XLqh|>ha1FZ8tYL6e5&*xL7f(w2pL-Lht#72_q7w8Kp zZj!`@b=NWM=n9BAE&P)2=TB;fEuqw~e#4`}Xiz@{quU86=-wTH%ZHLNh_zIu9AAAwkAjKfa+GOi|@kIVSsl_)aIn^i)(k9?jP* zS0?UoI{MF6|6}6)y)kPc+FY%pO8#WZrmM2GsoH9`?NVHCn)CEhyaRMV%?@a={Colr z1|e`U|H3}#<@QRjZ$;eX@pshuYNl|6&+98a#Q_;_GA2YphhEUhM<_>kN#BR=mYUEaq&L)K}_gJRNJBnN&>$j$eG1L&2(=kCWmzo30hD= zm7Dg#3ZvNxqV&h~)(JP44@$49(i_{wzMG4`FRPl=!v z^Uslys0=Zrz~j;CrF&wsP?=xR5KCzY3exy_a961)sR`3JF_vHACW8plRLre! z5fW@t;>ltbI%Dh8INE08!CJRd&9A?WgT(#}y{A?G($j}rMVcfU4R(cBWS0g16x|yK zX9SwvNo&q!aF<^~;9tbduzhSk4>I~8mDgWGdkz{FXD3#J1)f}iC}#Sqm30t}^a zH3;Upo9LdHcf)`$$o&L_iGKTU_%lrg-z5v@a#QmzU%_NDV=B+`G>B68q7HTe=UgKfF7-){kGG!uk^{=;JH_@ut+(JlN(XfwW6V)WV06pxrKP!oWO#I39$01;N0UfTA z_hD=?Xe98-o#RQ1+^cd_>Q00_j{G1x^0(7~t4%g11=ba6wI-h+F=UsY4;3TL1Yg`% zUv+8MLYHN<)iNc?VvU7~eYIT2!yMRu_8k{xmLokJH*SCB^%V2-PwLB0*ki|*2U9^{d^+>Y}d+$OU*0+ktC4jFmIK>xx($ZNufO z|6)ikZYnARwL#pMJ7EAX`9|Iu2KM!L@D*yrMlJHb%bzSA+i=!m6JEEk)EdQ)tPn6n z4w~8IS}#Lh)zYFTu6<>0+bdUUhr;&gn!ozF)7semjJ7~PkwlU|ex>1@fP8@-=l_L5 zBMr(%)ZgIP`}nU`2JhN@Wr1b9 zTAaJhL4FbTw;ytUDoU|;zp}+UAN;%1x^|Qj?~RASRegcgdo;1RBt>O!?bduc9UphY z>l{3~nV;bRIu1{Ww&@3XHp&QD*XxBbF3rWdmMIO)E1Mfjbva}KhH`QIwTY}mWC+qK=Fs~%kVh5F{N{_kaBZ_TxYHwwjNy?tJV60m&L-q$HG`AEZD2%sJv< zK_Oq*0;`#D98{y?TU232^E=l>a$VDZ`uc->LI8y8qR<_JOHrPiTZFt!u1LC|>>Pd( zn9IU5VCi>8lQz{vt1i=|znbl?>su1zNsJCY%%>6jpeHWw9PemeQUb;y+1{btxI(s# zJD_~OO?qqidM*@)J~r3v*ixCdakgDYic118<)SnrJxUVYn0@KK9SG&xY}r8OnL#8% zRQ20k6W^_{X#Kh#9#(94H{fv!6Lo)J6?}&Tdmx_8RJJ`xn&YBp;XK$OK%yc>C`}RI zt_$M<^~*1t?$8fGPHQgCKOWh0!*wvnBe|p>}__2Vlb8FWhj981c&YWXxycPJ~MA zj|Xni$<*2gg;MSG8}C3bvYVRovdc?pUfxIEN>yN>7;7Pq{Q#K|wMkXdQZ`y8t*YTu znOLCecTZc6`wPjl8+ukRtK?c}$(1is0`n;Ay>ZJv(AADKopos|joEOMAVbM}-*C3y zbX$WNNq7d4t4QgWD^6bD=3fan4Iv&jdY>rEO_(6&av@QJUExWgV)+wlgO||_nq|+$ zbb}o8W7c;*eCo9zEOM`IgjojfveSQG>^n*CZj| zk^Kk-ee%!%j@>E-30l48kQH(S_Wk$gJ)prPlWh;Lcht3f^&By%U>#yG4r}n1?Jsz* zfdYN0fmpmg4>yg#BW&q_QpD8l%nwEo_LEEm@8|tzh!z;bEH0_z0K)cKoP#w3Er%OK z%-2>Vd+^Tbay_b)j7*9XuuTR*6{Q!177galO#(@8)n%f^)dC)Sck7vvI}X36I~7MA zkUD>!e=U$oKxEmZ7UN18jtLDq*@Q99*JOFA{pmoQTECuBq9u^?KG@`_ag#0LzloqX zs3Hd+Mg*+XIrIep_nY9zAxs?A%a5FFx`Z&jltaclJtYj{KqK2O^*jF!v}EK{k{sdE z3DK0`w2Za@LL#K$l&wv(aLG_?t1~iMOUaOKIim2Qh8;|IMbD+Zk?$~YBbkyEkGy+@ zNqC>O%=gHUs!{$W-SEMfH3wdE2MgG#!Z9kB>Y7I4R4``gs~?MRF0SAt)wq9^Y+1pa~_1`o6@aQ_>J;{2Z+ii4Sj{r@1K(vflB zY){(pW~-_$oyD)a5q72z13;sW108*mUGfM%<@=nlnuph}B#W|>lH~uBkcaR*LY-96 zCKTE5ZnI5H43gdxE5_IL(h~O(l|~B6o*Vq*(;+MlPlxIwvhef9{+Uh-N18>t>1%jl zCMV%~-+%7{!zo1am6T;ZsUH_J&gJ7bAS^Gl&lB|W;^+B62#y-Su0p$g1U&?vLR}QP z)Z5LGDdf;8(+T*gJrAQx-3^pSPYBKS^|9zBqHkobY2_}uFgmVy_;R;MWnJvGA2(Y! zoH*QlYdkNSSk7nC zYbBf)Qj45e6nSuxe41*tCk^xEeePe%b^7X?rXAW8N<15gLN^P;=iMsDLsn|OZCVPY znKKRRx+l~0k=x~-I`W##n&V^9PMi(Mq5;*rkI?a$Wjk`p-66<9R+2<>Sm;F?cJeJ9 zl_rzy3`=f$BZU>}GkkW6%g>$q>{aL8S}Rpf99162Dz@xqf;m>Q7oyEQ7%!Q4`qmWr zY$VQ#T(YHFHe8CX6KP4wN?I)`;^tOeYfFnK#;h|0c?~JMmFuP7bN2N`EI5AVl^r`v zQy;(h`cswa?I@|cK#^0yir}!??_Lp5fwQ9~2W9T$I zOK*=evBU;WNg|6P9vHi01oJY4;l}**4Nc5E^6NIOdP*bNrck}LH zL=-w!-hR1i3-+)62!(mg48bNrL@N(TB2nw3piAjT6xXq$!Bg-7CfkHEcx)7?f{g}O z1`tj36H@|z9?KXJ{vLHf&Q1W9ZVayF1h)Got;?2@zqWzB`cq;WHx4FdW`Z7YhJatn ziN{L>7;Y7xI0AmX@<;XRc78_l9@_$D71SoP*wA^@5 zLX?<_VyMloT63zJP7J1i<0$ztkrFqtvSa1uuw%Ijo`wUPUd=Vc=P!S`134;nr{g6r zX|j7fF1(i1S}1cI`Mh(u!3u}OvLeZBoaARH;~a*p3y=y5FNjW`FOu*%W^mqNsBJGa z*Wt^SgufnU)UHg&70#GKI<;-5#PLD6e7yb}L^V&~XhuROrJ#dDH|q4RIDt3|vbnz6 z`j6Vk>C|s_=_rYZc`wA7=y}J+kW>h_%0peJYCf^=gk;mrtVU@SPHgkUDAn%vAZ<2? z;#NUXDoTKK$rK&+hbkcvlME|mcmNX}Ygl&jSlwL$20qBQz;i=nk;(8N{iDqSuluV@ zqAS7OiRp73&)X zQ4A3>j_02?ls%!v#C*u!%o{hYoh-M_o_5F%9o4C4>W{#C4>ouru71cZ zBy#qdR{`g6=H7k1yIHdqBmK#x9dF%pFGVU$(f0;(|{l9(MOR747z`HQ*Wu_u-TEz zue{A!AgDpXR+>Z*@5*+N6PB)(JAY%C>c}pLpkdYTsWH@MG&!eUy7pjp&SagaRvkWR zIFxwxRG4qWcWgAPEmHv8&QKW1;u9e6p+QD+G}Rr+=w(ij-Pyq&*fiWFX(qEkVGT7M z_Bv}yryBJP5YjXmF&jgJ0=YX^E$;4kMTS3uygjkC=hov880sRqjEE2mr9=r!kZZdC z*xiNcgY^pgu1u6GSk6figegr)pMf0=L^L?~lJ5p=m-q=-ZSQNbC_i+?0q8u#M zho5+;b{Bzl+ya~OUQ}7L(t|nWB&YX7FrBIGTuxZkycc#}m z^*atcdfancr({>2(83#?4Y9J_qi^g@7$Y}>t5L!{r_Xk0q`!MYqU3CUIk{$u)8a7b zy8-v1=n}m2^Phj`22S!XQ|ToA3R9? zTCp3Oz;5x61yYUMN_AePRGJ681&luD=8j~Ix?Q+LSU~!N1BhDA41$mHJ7)Wo|&5kst&h9(>wgHASSUc0Ou1KK~t0tV@baDt1gG zU7GSTP}`NWML-%ZzEp|eDlR`(1{+D|RKBosqEUJ2w#ZOt#^BL@FP)lx!PAf0V^K7C zLXA)Nr^6R>K_}y$xfvK!?C~IZ38{w-DoliCLh5`->FNeEhi`j_bGX{#zSq4+AyvB8E$iu*Dh}%^MSSr;!v-_|H*aa;($0Z>en!K}jO|}}SoDiKQd57Heus}3Eoan$&cxULiz?g**om3L=(YNpn z6R`#++2JgdV1q~lP3>XJg4>BXa`ryB6j1B!E~Y4ke-8jzS9-Q z8^77_DgGtMBc|IaS8v?BSz8T{$7Vd}gWD($iGu6;Q)!3C6mtS*oYtZ(U^&H`K zFk1w?Kun~4DY=(q>w$m=4gMm%HeN|+X|oo@aL7?y=K8=?Zv@jL2^`)k5I)bnNA|Qo z&ueR64dToyUmUI8(@d|=o@5BRBl|jO2W|oS4HhgJ^)KM1G>-v1z2j)S1PVyB^zkVgh?_j9+ziX(J-QYkHI+_s zWN#hkNs~FE#<43M8BEpB`xC`hDt-{f^j5SoaTgOkS*$z`Ks)rRIEVRm)fiw$)%=kV zf^d(otHa7krNIPx@rO$D;BE6T*Y+*as>99-6*9tJarqj{-oiGaeBD zB1g$cZ@?C+K|~Ck)BSbf=(y70c2pmzUF_7V4VZ;NW?Q2#5Q~(_TjxGCQ$RzupgA}` zP8cGH=VMU!b}VnXN!z!2C?ob^yVQhUn>?PoZ*YUwL-~VB@*#WP0H1LVwVAWGAV_6~ zl2)zPQ1{C9i5DuX#2Bbjcs;y3cDS_vrri$pkrYdH)$wW_Mzf~h$&opY_c+zfa&x|J z8pQl$O5*j5Fz(J?KqA9Zd}rJQlIo5H8#q5fCDBy{H&}#t4P)Gik?UuvAPjDqX-R*J z6dTWE5GoW1u}AkZ&feCP^Q@?i^@Hm9bZXd3F$Dje*r5e$g;{DhcL6duU05hD?n{1L zqyD#DEiuAfbV!SDM&Gq@Z}RiMsiryn(yNLIs%Db5LcVwvnKvN%_&z>`PJI~VqVP3E zlV%__S!Xy<%a}2b!&jpt`kfqED8`bAZ=!-wzPs(=?%iqM7mbx_I{p3&q3`mO>@LTg ze3pvwU$+r-=BizlLP_j`cy7!_6!WW4Pd`i~a4>cge;tcH2?AC0yKDv7A;0HD;mET! z#wJ(DdKPq5dq4?212v-Et(yR$z+x#ZChL1$43EX<@+kW*sDe{o}ZtT6(ih9Whdz+Mc|UjdT-+! z;#_MwT{OkolB`~-K=Q+`QsIhNhNceIwC$r*_jW*kTQ#K6lY_{aip96b^B(H?=t5B8 z-C*;vF!VNOyZRM3ds7~}m-r`{eb^-#=nu0(^vZU0<;D-na&P(Bpzj*|4{NFml9k?O z?O59zOIS9+JHb6gz9KkLvGs!Y6s7euxBxHtIA!1-{Jy+{F9s0F=_SP4&=3nwUXVRt zL)NiVvq6E2PMUbx8Nvsf?!(f0=U?c6Gk8}-q@%|2pZ-t0*g<1o85ZBK+W{uw%zXSW zQltw#s+axmJ#P=+ASageaDnK_SY5F<%Q;GZO3o>ae1lFihhi#IG^iva80AwZ{4a^l zisVK@D9hnC5zF(8T8&9|rffnep(j_~PNx9GF5|U{D!UMI!w#DbFXxtG4Vgo7^X=F3g-S|jC@}0Y=BBrIebdOS5uLj-5X8j_tI$Qfw zTFBryAFR1nFTFVsvlq8X8lWBYU5@*14husAG4@OAn}XYI5lHb$+|mMHd1!ozD-meX zMs9KR>ON&;r#%$|57~(_?(SSf z$#Y08j@CW)x|Lq>M^pBn6U||avroK9akke=qW9-NisQ;F_JAXvWIL>PeitMoDW3Rt z!|LZ$FKt>=2=ur(<;7BQW#{sXdgL<>MrCMMyHtUuFr$4t2}<69DZr*xr|=Zy1beZ2 zuwH0>^%d*U*xj`TqXx)vO-MFzfHCNyD-ekMcZrHZ@Pk!~Z&-qW+oIj5674oP@%ernI(N>>7 z%+#hnIrHDo7W`I#<%XAjx(cXBka+QObo8e!M8VejV7~IGu?dCY1dQg~x3s>(*#XEl zQ_7Ep@wtubYb+7sH&!hD?aluN__+Qj;A7`zX8ZpGzU%*lW?K?TJ&(Xe37aR&zh0sU zg}qfSf2pO5WtCZ_HZM7-IQuDo880C`70fRw{)! z=X2*nGYt>p=1bO4X-o;YIXOO@92t5&Q(>Dx{UKM#VZ$8N+rDX7Eeu`?CZwF-C}n+m z3&l_|Lc8rY@HNrp$Wv*%vP`VPD+79rA_F5(zBsP;wwjOh2gJp0&}FXFZ6Y*;%fQ| zWo)(WC|8DRSkc@V?lEOdA4H@WQ7`j3TfWF%Yu%L z-i>QrJ_{1~^q|o!R!b7BayhRSe+|}{bYULH&Si&+Y1*z(+s=%L>Jz~U=`_RIt~}84 zWT+6jCMI>(f?r_cl$zpI%pIzs!S1;E`4}g=TvbvVh&1DKmDytUuUyqNOm}AXai+2H z3SH;!%RY@0spcy@@z5Xl(zf9 zY|yg4X!5y6hEk?MHfJ}D2osJ3HpUx#5K4z{@;I7PG66a|my~RX;YGfhgJYeYT1QhH zMbslBi>2TmA+J%UZ&dWV4O5=2;|1nd>99Vj!62|ZOCcx~$k#rOu&b|?03lFF?SFJ5 zz96LQ2hQ`&P#o-il8zwr0!e$eTZp^?`v~c|ZMpaa1p9mp10PmaMjRFB)p=Hk@y#JM zE=&wV^Tm_W&ArSX(^&)ps?p>&*dnf)3>w*#I64Gl%Z@v}_{1cDNNy;Fy3F3ZUgedW z%cVKKeNFHu5r~ZWA|y1A!xXk$Zm=CYZv($L(cjxDnyJ`pvwN<&9sj_qja1^k%<&U$ zQ#K%XK^Bk(Tj{GP!D76X8eFgGuRf~3jCr!hQj$hm7)v8+U1iv0EhK}4yTTsuL9(4iYx_gB)Guu!O7coiwP zWrRS>s;(}j!TtVm%)L2Od7j(~Rb(*2nW{!@OlO}UX0`tJs?Nz1l?N4yH33dXFD;c8 zFbY24#ipC$C8bHX4T>-@R{9r|6K`ssIVfis72J_EF~7j!6(x)GeHhC1X^Wvx zIMXJsR|8FN928zM<>VB@rVP2H+rp76GUyxN{&u0c6cx)!d5x&X8Y9GdmM@4-R{_qCpMNvi0)}u?lz(*U z<#XFn97$^^(>}PxkGPpN)lbK%xp7gNfN50_)7+C?`a$nDT`AxonWzy{t)Ful4xB4^ zeKl~!n9Tu0e}@`00<4@XsigR=RrL#qHP#6(%3r9?;M$Nrgj?ow$XqiC#MP=#wUUhj zKw6ux;1YSLIF<#yb9h4w)T=q@=Z4=u(@G58?H_pVyM$edlhb-#y2wDmpzk6A@@NL-ncT3c-R3*rMVD5Mj@Z>oLoa*WO|Fte33X(POKzqfb#0f&(7FjRYa zpom8SlNbrqz zOn*;l3dE`h#2xBD(Up4|N)S`Nl_cJ6#gq~By6eDs_erUEdtSj36v79gVeeH81L*hd zGo{F19}Ip)^rCKW?j*F5b*es_y-!&x3cp*=6ypa&(Joj>48PF&`HO(}>Tqp!+q|_5 zRLIyBA{Il^hWAo23&eY2^29$}+vC^-S&kZ*#A;CV$p3Xu^S>!KKz({ZB9 ze`HSsl5jVu4-NtciXPXH!VnOu7E{!rT{2#CS3T8}8NoDI3Z_P5psOv|iT1WpXX%E$ z*NDwer5x099=mvT(w7N-J?|;@+f%a)(x9H-lt0&`bf~EmJC7r#<+|aRui(|T?z$$%OJ57h&>yMZ|HM@8p38Fz z)`IRN=ic34M{H^xL>?aHF*`MiFcE_KhNP;xnBC^&=lgjVpIt&O_k2oISE9`g@cK|! zl5PN#7g;kL`_ZH)RvVA*E+toKJ&0$=u77cV4r2wkYE&XuCd0qbxxnW1)cwL!v{|27 zAn;^$I}#mK?zTfpmO(cjv|xRdf_XsUqV%vT^3A=IMsmK&Knf9wgT7JGf6XFz_Wu31 zA)R~>jWQxA85RB1DS+I#4WKZd-g4iS8lyL=o)bzBB7j^S3Hzs%0YlRG9+HAflKajK z02862ScU7#;Q)f_erB_dWZU_lr5l4m$_yEPNui?I+Qe8BQGF9xcEGlZ6*@Z9u2Svm zngEDWhsZACZK5nUzs!v={Pd1JX1t?Xy$MSO1QkN5)vHgZ`*XO#F2_j9jX6=Z8wK(T z=vKrdKXSY?H?v`dDIs~}{z?nITuCJgw31wl>&D_^a573h2Db>)N3{;bz)`aA-9VmFbncKq27_8Ti~L<$X(YoK}+pa^|H z=MdLLQqS&DaY9T?xEs+hD7Y`&qpZ+n^au5x|y&=bL$rKX>C?p zR+n@5nRi($fpB{M^x_$ofl&l1+`?2S0?S`c^wYb+kr(KImRcDC*Q3`38TjQiqe3?V zuLoe*$F^vAgDW@)b=b#R6K8wno7`Zfuw~VqC^)iym)x5JoBf9qiX{c-Fzw{9?YLT% zi4};KDDHRFWzg&|dghA+^Kj97oxV|~`fSVf(THa&bul(({(j&LeTYe}{y4mhuI9zp z=aXC)-Q^^0C_I8!883<)`ur`X5U>#L&YOSdI1Vl9i0xwF0SbF`##Ghol&}hc0e(13 z$;%$Vq2xAJ=np$0j24wwm&hNy8n~EP|F<+PIR&w0>CuOvkr8dO(Y`E2iDqpMfBDv8 zWI~jZVwAOIJrX$vIYeFaH16B=RPO}rwrnV^g&|Gp2@Rt-E#T4ruks)EQb{p zi2xl#iJQc-wXUcj7*1~5z}@9_CmeH*6yhzH`D2z3C~#jpH&`m0kvcf!xb-1r< zw_1yXKYLkk#TK2DgNN-x{qMXS)%1oi1M8Yv%!QK)&Hi}M7Z(2|Lrp|mWr|RK{dbNX z5;e)%2^6u-ivD&8?KI==c;c4inYN*cyXr;abi+`rh^xQE&oix;+QCf`S9vt5c< zNJ`pjN3CwyvW%)|NBJKb^$t`Tyw#Iwb}&8Y+6ikZb4r8;mfFKiP|l2ozd~f>K2K=3 zfOB?>7bt2le|?~wUBV0$09d~uxc$37qvqxD<`rn?^j~W8{M|7l$S?x|1dChP`PP#(iCp*+ z#b(_f;mMERuwJ?Lc2&Fv3{jrdofip8zf{n^#Rk%G+b>D2Y{;Nu|HPdv7j@D}YFyN# zC$zs8Wl#LDIXbryk8-pF*_3Q3_y5UtI{P}ZQ$afm6VcNe66}Fg!OgX1fbSusA%N?q zHryEUcfpN7$F-$aea@@>mnYYUTg84xm~@c^10B2tqRew^XU+qyUbX>GIDIv(>EqCK zKQLP7&FSi#o_-x%ZaIjQu-GJaQ8j)KY?R|6L?Cveo7Y40>8FEc%`Wl+gG>F7R&Fno zW}nuSFgy1@p#Cb8qZhBxJin;@a}J|8%(d=d!MtmN_cyan%7*U9FL>X*UF|@FMcC=K zN>0xW^LNE74JYqN9Hysm=#ovGF%*m~bbS0Wma_kw5#o5mk43n)n=0w;LJf+E#M?5i zuREBIPu%0LgX%3%1?p=_#~FMx){Ps!WViFyf}&YB40X*sNuh`3uqVE87**;B_5i)Z z-=><9@O>=dQQx&^qaRVcVNu(rkMSp$XuE-qU^0q&t_SS^06g5kKv;M{!&sP9#q*}@ z(l9eo!@$tUfiHGoBE0>LBP%GLqjH8!!!6f3gDjF;^dKlO{no?}%KE70=0M+4ltu4+ z^9JZQy~yC}#g&5al$Zk7C=yD`&O^;2+#MM?Yryp28U_-fI55t2e#wg{mu+e@;4DZm zeODT>-M?|$_#73HEB)!IH;=6VDG*^_uT0;WNY`R{45gF*kl)N;--{*#?oObyc%*d%f5`b zHbi%pa}QEg-#jI$E zlSp-NeHBIpJ|c8Qf-5nTSjBn_Y|>97F{++78`4k(TiUtD;jRmApbB(pdg(KybERs5;7F6Kv!b15Y6W%&;#N?{K3WWfow9*LM^F>>7SuF@#<*9>Z`-xrDc!`ug07ut*R-cn^)ZjyU40I-jvIJ!9qPCW)4c>5~$%DXJ zS#hgByk*1|%OuaplUA*aS$!{}Kh5nGkKdHfoc1NX#&wr71`i{o2JWJY^{ej(j1o4T zl$5PyAQjkK`~Nu2F%-U$09$h08RHfhInTY8>X-lh6WhU<`8uJ@)R;<#U>F;m2Z6v^ zhezWB&jCx>?yh#0BzXij_FWXZnNYySoD^^9!<5d`i%o|z-3Ah>%=aCB_WXbq3u*lP zZ#0bif6_1xX0HDO4Xe?aaKitc)N=@2d;)2U#3e&*VCu#YZVjL$Rt79>vsTEZ0g{9( ztph6?B^!jydz}#vWsa0X$;qSGvjRJQjUG%ro)Rkc-AsvI%lu3K>C3#G`0DC0ctsc3 zp8eULevPkvL1cLxHah6JY-t(lMR*EzJ2?U_EE^u&*|mQm9O;+j33;? z`x}Mr8*p1|_d1Ni#cVe>hy{eOY#)g)$~#jm&S-AW`*(R^&$k^)N?qps$1!xfV-Sa6 z`8$xNSao7PTjLt@-Kl)D&KK7qs@n>e9g1{op{{(CuACgS4=feudUcNn*I|_*8 zTUo`;Sgz};ZCj)z)H{phJu*ewHJ8T+WfMojTE{IJg4GvVH`3yTM~~kUGPag-b+s?O zo56Vp-%ttPT;BB?K@baP0t=8C}_96+w;>2OoNB*Hsy4WN zBCUgQGz>B`fWQq_k&-LV~VDu@Ox#u#{p58e+(Sd`} zR*s9fgJ0dcU*0n{#hBTFK`SQ2Z}{ZDC~GjXpGviWO<&Z_tRC!LI<*%kemkwo2WA~_ zQ`?ZXh8@^w6>+7CjPNZ|gouaMzcnwbHw}DoO;hpi4KHm9_vI2Q9axbkcerMTqKAf% zhy>@Ti6IR#&|48+%=Bz(?W3|E9Mtvekm{;RbM{1w8V>HRt=?r^7psxV77vG}&X+-~ zB7Cm1n(LBXsnycV9CxS91xy4mXe5-zJ?79d!W5{3$En$wI|4%kd3C&D0XUp7Ysu>k zRCft_I=CN-(4lq+mb~-5#EU@MeC{U=H1LZZl0lw3R%Lc?OkptVZtotS@YOT_;G2G3 z#m)stvT?gPa6R%NqrJkq@P08pY)9qVHLaT6Jn&9l1U=xxAKPJ_ytEMPrtUnoM|GtXCprNeiSQSI8?+|ljqvNm5 zqlWj5i(2pp7A~KC{b&TLjaxGpHK4awk7AISNsL!0Vt-kqFja*+Eff13IG~#B9LW!m zS_Y8A+-lM;SNg&kU6exd)i{|$qxk3dE2A+20^B?HDI`V5PswZ4LCw82e!w!c87DUz zhuA50wvY$4(aA|BuFdKmY-EfG(tJ)Ir^d)#z{cDueXl$0847Ow)sf2j%n zkV->4VpnIVMyi9?J8X%Th)re^u^!`x7-yKZp=Q`~YLodE_WKYHm2a|W8_u$W<1w5e z1_gMm_W(lESD$|tocDxCqR_m;<0>9oNcHmR@3R;zJ5J)5G1CR6nII({HeXo@o*uY5 zC8$DhVw1pQ!!kNk_AMe$!oEDPTS*c)HjtUWe3L+dI3wym$CPBe?Vv0VmvJsd6=0vi z6Y@O5$h&{#`5V=neGE~vAzpwcC=TQAxoa|Z#`Gw$>C1At9O9N@0C)pdS0!XoyUNX( z3$!Oi(TfFVc4K*2bpFQ0%1{&2jsuK(pY|bk=c#c29w>iWV^<%6@u|KT8`s( zQ@xQAH%4*zD&28_zO(544G@lMDxIuRrQ`F$fVq9&C}@r>*P;`YT{BKt)*_R#Q~*u* zU2p}k7*E8p$PpV?nubk$Jk0l*Y6m-HP93?Q+r}}5r;1oSe`@XnGwl|GFD$nR#|}a; zse?u+{1r+7x^Os98K|g=mPz(>7h=e!O8v@Tb=wIXbMlZ(`a4g?!#_CmE7JAz7xnQI z+uNe2ofXz{<4zlL$V&KWnMTc}mEY1kK6r!`i%#nfYypsh7n$-XV)pZNY|)MqhklI~ zqj=1zS`gVSGGZR=CFUQ9L6B~w2{yjzrPHch3)|6D9h7gb8WD&bVx0Q=WDqJ5=F-ZG z8pF+;z3*Gvo)QE@CrT{$ZZ+Xe*RlUm2w1myPN^c`T+t`0A-jO6=)xW;6GqWX*&ZWd zj=7w~ffbsC+XY$}fS}1__@)O5cHM#A=93-^EFQcG`0g%|U>@7Fa&wsJH>##E2|cRH z(ZFeCI5&EqTYB`8`y)*3uB048Sq4&sfzJH%A-`{8JXl%+Q4!Yaq0d~xiD<9!l zu6MM~3-T@@uxZst5G}b7G>x6s+H4KDb9ne_`<~&RgEvqHy!p5L0Vy7B;&Z`cmE`CX ziVlHHQ$trO(EufwPBPo2KSOJjD6Qa9ohlRAodGZi&OHrG(vRELga#lkJ50Ah?VB1J zVL`4qf`j0E{AIwtUrD}@KPx4z%x!_?^s^ZbXx|#X6tow{5Fkrpf>Agvz+|$#VN^0H zPXtiRkGkRt+)zRHrAQNdg1@U;YL<4u$G54II%F+YmIrWp{~KW9W-O{7$`U6Hu_-ga zaiBb)Rmo=+D;MaekC{XZ~3es;}`>35~=-Yz56@p(BOGVb&HeN znCeJbGcA;D=Ih)BUZ97b1BGlpjas;hfW8^VnZq0)n#C9NK)-pyQ5*cAxHBO5QLU-| z)K3&2Xr&0@V5z_xng?2z`wIn|6N52^V}(><4^w!5%5&BHLUkY8S7%*lr&4h+ut zMWD4{94wACOVwJy=A8k)C;w%LUf1kQMIqLG&A2DCC$uks+r8KPV?}oCC!B!M z8T9v39}>Qa30I@(onTqrMFed?SOCF?ToGF&8q0+Ap1 z4^z2PVDSUtvN@*@VXJm5`3ZtipbPwv`jgHyDPN4X48esZ2!0ludoT>FkjW>h%QK4R3X=D2TS|hHo9JD)&QsX< zw!O1LgnKXVreN%_)&mDwlG`{G&&<=M3m_KhWg$-Mp9mc? z*lLitYQ5WTCoVNJO2K_EZUmr-bEqgD=CN&CWZjSgCk;t<0*6b53nKo1Mtpp=g8VizdC~!aJ64D7`L7)Wd<@*wgM_J{(k&3Ox4Kq6$BZEd9lMyPHzJl98d&ti{Z;`+Ufc~h)UM7H|; zRNDv;#|fsbr5_WQg6+kS2Oi@=^Z_S4TvH8H_SMG$Z(rzH{xKpi6(SNw*3s#S(A@PW z@Xu8zK{7`HuNU7Vz{t6_8@^-X-|diiYbu7*ZDfr%e<<}_0_t_X+E5&?5=6CTocFR0 zJsgBjt%x5eIAI2O?i419d6z*6@Ps^#!~_tDG^sLk(;Kcl`+b3W+HEPD)!v#}^avzj z=JVAF6bmv3MHY0RUYCwUu@44dX5zd(UTMnE6b9721>Rj27Am)S$Pf;3 z6KJU{950YyvRQPk>m0A_@&xFlk$h%i`7|dA+!&j|E@aog5a$TcY3Ij-Fa=J?ov`k# zmlSiVz4UN#XxXE+OU!mioBmEg8NC)&%(E=-M?iGu3-omcm#_@nZ0NZb_5gfdei`5Y&6sW zcuvoqL_-lE66fN}HP8kCkWiY6A8Z#ZE1|%F4Fu*B!Q{5kEB75eM#l#XORcZ6yaChM zKA`8L-m?h#xno(6T*EUEiq7-fOUG(x)tiBP#sdYhk_v6YbZ_H}*yR~xzt!wO+U=@D zTZtMQlmZgiVa&^nW6R6q&ngkD&!;0mpn1Zl0EE2cqXJ7D9WpEx!K06PB12Q8RhX%O z%+jj*h|PKp*@SlDq(7iY^TP|z5ysO^dM{1q2ctC+M=1w5_KwYpaEvbbTKrj zbCZ@74loKoT^+$GqlJI&W_?`A!NEAAr@xMcW8+bs(p~@1K2Pw;6+qwxKwO}l#0gE_*PLy4pv@#0 znBI=FXb(tw?e!`ku|Gpb6E2r)XG&8{P|>G14x~UYT_1yjHqrCvGG3VVvfwa>7tU@0K7qNJNH~MUesQXf6r$*PLVO^s2TR2V|>Wt?bC7y56o3ycGh0f zCbiOxc@WJWN6pRq*J~IwZ#R5H!@?F^p_&9wyTF+GN``NKy`0wm1Mpsrnrb>X?E|s7p3MhL9+1b|O}fosI-YI-0}DunihgCebVQ26o4^k~;qV2->F>TD+$Y)Bx>hx5ouu(eZ%O>2e7w_3k= z^0v2haQG_&2h~Df*0->@3_z)`X1`w{%$UZBFh!3L7D|$K*9z%2Gf`e?%rSU*3OXL5e^!gIU+bEc75xMR2lR7EF?U2!?tgx3(w6>@3jTDFzu!=NXHJXF&Ar0$ zS@1$>Qm8O;{q~@Hk0-)sLP^EyATrrrpZTk9oao;3`^u zjKU`rJ={1lE@iOdNJxveZ`D|neTmsxs4dtNZyZW6Y06oBS6c1XV!^A~Lf zX?-772hH@SNbt!`5-nwEw95A6@WPS1<~+JC9?kv|(x`ZZ2peBtc|k`Sb%b;?7D50* z;b{OlX}iw+)z8`^NBOk)yWCj1&WIsrxxxJ3h^Kw&8r4`24d)Ki?aZtx3r9x)yuHPN z1%iD{#)|rKk0tp0!M3dO+{LbRmTX(Dks1fy3UQ98vuy0Uv`5Wx&Su(@8&(a6)-=Dm zblZ|KW%WBqSyHyA;(Z1fe+&n&=PTR#@qJie=@QbaRGYN8FZ7lp=rMvk8Hy33?%bZ? z9|b9j5a_`L`r*3nHy)1W>keLf)N=dyS)-us7rf+iqeET>K;%hePi4Kt-$i*XUEK`w zjj5O$jh7W>lje8TY3R4DbX8wXS^6_hPM$AA&wng@pT6q2Ca1aBaOifNRJcK`MK(wn zDurwFNMNmwrZ|z+t;Gi;yaadLu}c9H$WR4DN`mX8V@mWJ#+hh;8eBr~^A1_~-YT^0 zGKg$Ba){yuTpLBN41=+tKtf0#n3>CX0K#<%4$&}*hPRa1*e_2d zq)Us%S(&jmoV^_b1#tALSPC*2QpoIc*`Ya&=+}OrShZ@a5l)03@q#M)=G5!MO&CGa zhIEL6y(vs!Jqmeq8{olNVrb%q$>awfoyN6?B_uY)u0>fKWI}WXSGz-yy^*Qk`2ea2 zWuWJWV;*S^j9O0>odG3i^Y9TxhTprxMf!NXoICP2-VPM~B;af`T{2NQ)t`VAjq_7d z8$ug6#I3~dx2&}O&bmc-f%3FNH$^o>Qm0T4+r<`eDc){{zSMKdg@u@w%)E4BQ?Di{ zScA))UnD{=Wx2Xb&s0BPUF0ahK`#ceX5^(j4o)xUjW#-p`%FO#@suF^mn3FWRwAqc z@X4mVSV;U9QZB-l1F@{w0vNuVgbcDc$UvmoEOI#FQtpPm@9ZUb`Ar)%{VrC16T)Qh z;pXq9@TJ?-a#KU*p}Ek=2qQ@|iJQZ*9}&bEa;Zp zo}^Sn}9mMM`HBXS3=D40p~Hkl~B*Qml~b_;it)zrEepR zWLeq2=&XE|X_z=-C2JBI@lek>3q3v0QcHdbYuXm(bN_)adwUwV{V+#z;{NYj%`mOmI zF`D}PVgD)(zpIMBWDwWSeHAVigwuv@^zWWMAWpOsRK32-_r8C;A!Z^!srJusj(>ou zb;~-@M@bCv#N04hBzZFrk%jRc!-P26L+9hQ_phECe)%M+8$LR4Q`xvXfFioQj9jxZ z#u(B1ZKMU9d4dsn#xM40(hni50m__lHDJasbCSr9{Oz46Q0DFEZy}xRz8IR$u+<~$ z7yn}T)?mHxts7nqyY|zr5AcqFjLF7ipD(52lGhO%OoMsaYkO>71q@C<+jPXTSQK_cIS9!vY`H8XQD;VxfFM=ifQIL=)EED zZ4zFKcWNHNZofVeLnqw>jUxIr^Zev>&`@?&r=9{0aN2~+OlXC_WIBQ8a=N2n&9r97 z&&Oa2#iNn3`FX`4TI~$8t=D=^w5daisq(L-k~3)Fh|5hE5vZ?tViZa*(NMu`Q|9Hu zd%0ti@fymUjt5bxPnH_7&#nRx?^M7$akPsDMmw@p6UQ4X^?fXM6{meu3jv;7-E|t1 zSt@zRGRxB~eKI)CT7J%?UNnk<>tc9z;OpLPv4#IaezibdGE17fdQSpy)$Ih zHolA#Z2OFq@OUus@kUyDHE?uQWjr4JI$v6{S<8auX1nF`+;3oeiU4>f^NEgtnd)NB z+A!dS2OH`FYzshqgGKZwm}csZ(!_P}j+^w)RdL<;Fb7yC^;u6EWU)3zg0IV7Ug>(s z7+1QJNJs(Yx5K)Lw--oN;fke=!QCCagI)n4tT@*_)~BCd1Z^n?4Zbj&dGVx3elrRa z2KfgHEWL2%TZBxTr>5$CIYNGAY9a0P-otjfyTWp*3e0VHiuQMJFVqUk;1+P3VQ=rQ zsKGf$K^v9ituntxQClL2!iAn_p_r=+@SfyA(?2sVr4Ap6Zzb$!1o*@9s%IoZH|X!mmSzxKu4)?fAFMYe zB&iE?N64m2wBrG>+J5mpdkmprm3j?2L3W1Bb?83WyY0PtyaJ#xLpem3L;R;kGwmaN zZJ04wm^0b(yo-?zQzW)2JJCX(^iake_C*S4HS2Oq!c{0eClZbXmeKzPKJvM`h_qkR zaX9L`_*8%mWdS@WPpc;~q@kE1E=QZP-0ROrG}tKq*yIsr-E_})X--d)r6H4IFk}vF zVwfp+A7VpnDA&D$mK!fHuG|#-!IdPZ;%6vqVdU3oOHx~?JCVD}-LB2D`jer7qUf%l z9{CDdQA`;+XEkr?X{`4A#BOd_k?>O&9XSfaA^B#UBd9B}%a zV=eiGZ2?$bdm3y<#rP=4oAwu)0*qLe8{WMA*r^RN++P5^lW3_ce@c5t$p;B6;>Yr< zM&0!k_oFhYvh*nB z$9q#(tdAJ*Sa@#+RRo_)5wc5}>#VlYUr^@FqKqZO$m5k8l*Jmjhuo% zDFvhcVn{-X9eQP~gEiLep>YQ=oIYHj`#P{F-4){|!}xDSP$@EXon33LYa zk6-8@G(wc@2o`1O7c~QwXl11wg|M^xmxJ#E<7v8pZA>ov5|a$MDie<&k8a?c;4FL-Ti69P& zuIaRa2MmNv)7jT9HOqh-x^A25#@^vdbAciMQzi6zorpMrZ&b0AQjb(=4Y_sX^`?ND z%i#gWVzSZzmDtA+*7#luDx+PnG6zZqmeT$ zb+`Yk46nP-^~Df!6=xfdD5!S4Qh8AMxf!Lu=D(I3@XR$H-C(mMH;QT-EtZgl67*`fgpAVO;Gw}lx@GFi&%WtNs`J-E8Uu>+`gi^;n4>$Z zyH9|(^M-f!ESTMV*cbU0ygx)-$6@CBV3pu7!Za!>R%g02Ev>`N69PE-IR5@V3Ax{4 zxnFAdu%ZyN#{JDxfgbr-7TwE3s4HHT%F%O0vnljO5mGuP{dv9~3_0{X8p|wb<}5k1 zHEMuiopdE37M&**wF_1q&tJ`Nq*JXc2PIt-PI(j&>qUCAC+<3m64bM!gQKH8T{BLR zqPT+pl9*E$Osx@&5!!5v>l9V*MR~Gw%{1V^a*@H=p1gKpK?DLfwop=3!`56}*k2q2 zi9>hPMZd3rb!`#>TG|~PIWRRXjSybnCOjOzNyGwT^BCq_oQq8mj7Z}7qvi~BKUv3U z;Hi6~;Xtv;i2mY^KANncm?D{yI^+NrjHDPub*Sjl5aR)~4!O4KmLumB$`*nGS9QP6 zfgOtxT9H;dEGR)cY>E>@tuFb~bZXJZEm|}VJS@Y0tKQJPtpq%w$Y#+?s31pZljso1Wni6Vk-#vm<-Z zqnCZEPjR*Lh#42LS*HO78yMGhFmgQVzRCdke@*Sjf%vv$RO*F8fskMPv-XGV(xTA{y-B!KF-);j(I+ei?Arr~tRQKl_;dD>U>x^G&t=UVy8t?Tpj$i; zijRF*&$A#G7#**@D^#&bovVgB10Jt(7hkxKCE*AQs39yi1B7^kso8@5u-P=z)HDvi zFA0L$gd4<0hNe260rqY?$aLl`-y*mtEZ!o6ok;}D0>EPPkyN8sj6rY`2B>inlW0(J zu5_>at);^Zt4L?Tv!BfK99E9~~{JhmGX7Z=xf zAmOEji-%jiX_U2v2z^*=&+*AZ&aI#(HOnLo#XNGJp>e8dkLar@SEmrg9yjNeUI43s zHuBUh-96gvNLov);mDD#KgmW!kCq)?Z^0diD&Ac85dBeBYOvA9nI&VdCjC@prmwrs zQ4<$!+FEkKbXbePpyW2GNTm(3K8j&7X#y7GJI71-uGEW}G|g0uN!*C8NAY6t!))|) zO9wU5#ztykGwDUf=XVIiVSM?7cVtcBBnAw}c(HglI%f`J6zQBh@4Lm-L> zPKuEznI_~Vd_;hkGOSt0D;NaUAd3k1N}LtM94Z8cUn+E#QSkMVJ4KLLv(0`8uZRgwge3(V5p}ugBc^St|7_8IVQwAo>IIeA35CItH!G> z&xk08d;X*xciF_uo>B(!@JnBHTa9QPRcQjCcjOd+*|Jh*Jlx_7TY4fZRCz8+!8=Mk zo~6Fhx$ZdT;#<1j2RYML&a#PX`SS0f7jmX9|GeHCbZV24n@27=g2jVNR;b`f4#het z5niV8C2qkqt`lf-XdL?SWWcebg>dSy7lMR7lkPC;MMxN`?~W6gJOO_Jd(OU!vC z#nIfJNr_Kw!MGFSEbmF8MvfQ{6)%-T%qp8aE}kiaEl3MS@dvvXfx`C^J|mkOuATz7 zo4W0|`DVq#8g>Ko(O~zDA6;&hdw@6Hf8ie;Tlun5mqs^=bKX`5cf?5fbMag%N3ww$ z{iObh5wFxGfcH!n2ax?zaJ5~|E@4qC?6V>SlOrngwiE(4uEVI#`W~uwL)5p?^`7H; zvicsZc0$yb(Di2QZi9vN>)QwVAnV)cdp2;tpYMIs)ToX*<;#oyQ2!F3AGbeB0Fo<= zXh@3|%)!2`84~$p4vfJaa&G%NTonl3SqukL6xhSw;2PL^zwGTy^Wl>t9Uh#<$air0 z?jY7U3(APsu-;|l)Y?^ZvXe6vd-!*(Vu(g+a?;XIy@26#W258uWZC{};a&wfXX}ji z&Ul8}oGXE-275F0TOGAXR_SG$eecUk*u7mMJNkV55VNC3xWIL+P&iL#yYBd0CWvce z0P8Xq)TbNm^;^OiM1s3W-}y!Y`HfK^d`#MS5St0>XTZ`MzW9IgN%Fof5K~a(u$~8Z zT^>J}-P`&9R_W>0(?u=VH0e8!JYWExKAY;EU}&77l7ww6zZv=K`4++ti7g+P?h?Mk zNyCK69PRay<|FVc>Zx9@hG_+~WOMrAe4+DPrmOz9QwQUJo;q0AI2rzT-^7fzuJfi? zn(u1&?n_|&3@x&df6;zAt)?0hXWCWczwGUXvUUQ5bcAF8aMB{zuiNiLj(0a;5X;qi zq;hwqiYCZ>{_EwxSC6lfS>uvYzoY%_Q^hV$?oY486ZYiybJtD}Ums_0wJXUy6S+>w zm+<0+;W=+Sik(Wy)5no*Zce>CNT>F5`Z@VB_5On8HuK_BBi`wSyU$+|Jn*nBxGfHm zt(sNB(e6cARWuX6bt5xR_N(;pVN*#mTBz93)>yM-YSi@P%(*3%)3pAamwgtjm_7dF zKTNrnwWFI(thx6hdMuY#OD*|t!bw{9dO2HLEyu$yX$Oj@|B~FO*-m&`N1f>KJrd^5 zifxxwmAn5`6P6{)2PVsE(HtpHLS^#q@KmZ{uP$!BPvyihN96GdKd<6zwn!jWwg(UY z!bW9m@;x`ZG;wu(6~vIS28^>Y7ZT->bIywk-1l-=`MDw+{4OdjVQEo zb8^WPPMNS_Pkf4+S*}oCOLDbA5{>I-R+1fJQJa;iB$$*^J;nY9hdoudD;dM(9(j5m0sinVspvW=>o zV9J^AIaKkQW@6WRKUto^v=_t@7%{YZkkjek0i))ecl;K8mlWyui^B#80}C=S!P+Oi zY6gvIGvT8H^d6t!tJG6}@-I|Hd&~S0daItWpQ(N;yi$MZy;A*AdM9@aR(;CVSQRcQ zVM8Y@j~=_77OYj_-Ofys08hDU^=#0ytyZ>V%x6g}l@8J$+?07`Fs&HKTuPgDbXZti zRBSsT=<6ZY3lL6wDA51Li^S9aRrRsZ^11^VI;OLW_ck^j!eG8egQfvBQ>~=CX3J< zo&os#S#^UYmG7v0Ca*aLRos~J=T_%SA<0w{`qko+AcJO<TH0lV>nOIkPhknhx0o zjXpwk8B$jKOO|PCR|Ge?2*#sWqPpyaYkbjUg+~R=_mU^J^*7B$x(JMa^cq7EXbjy3 zLeo@owXIX$FtCmY*Q}-OXwsAs4~2(E4V!T6iS`FdP_{h-ok6Hk^JT1OtN`J$R=y zS0J+~u88(r@J~HBPP<0_F1a{-lj*X0c|r_IdXyhcTXeU#@`Q(P&wj z4_znB6P&+7*bSKM1+zC&>;gIW2D8$G7X~&T7|W*t-of{4OIS3cH<6c_keqnaA+8m# zhZ>WaV#>t=A{CcMvK#o%tB7I#1o@=<9auY!B{Cs?5{WR#di&qub+xJx(f|g_OGANIS zP-V#w9z;kHoj{JikW2bdb}GakG9YR@HtV{-sq4AF=2Fn~(Ibp*Hz>D^UpoQ;r3 z#^k)}@vy^ zR^{Yj9C&nY%qg2>;WrabSmLjeiaz_n_b>D+Criu+43^}22k#Ty#ocAu?CtOiVPRqd z0=>P9(-of@*ppTdI9HjB!f#P9gyL7rzL{{yn;Z=763kZHF_8M1SUeDv&d4H$i5k42 zWZvd>KP!3;EN*=Tn}y9e$h#hES&@R zk2cZi<>%JqFGLE6HKEg2Kn$13-60$97)QPgq+gch&19RZhA)8?$5>1m36fc;P|(pz z8+nOYp7^_Jwrs^oZkLHykoAFJ9?4W{e2r0nrb$qGGr8!+12m>chIMk^k4UNGRk@V2 z6edYn&I0)wu=C(%Lz#Fh+<0R7L`fkx#vtGYwy}99-Rf*mT4iCrd?#+;t_DU4IGL=V zf)iwxTyi$lQUidskdadpTE=6}WWPtzi4H6A6Lvt&3-#x!l!>Fp}Uji1OcYPGY!Naa85ebIDooBHjG> zO3*CrDewQgeI~`C7`*>X@9d~blidzc<{B6hcID2a#}Efs=Z%K@i~IfRq?X;3oKJI; z)wZ!ke@H!zJDC8|_y#9=UHUn#x<*_H%6PWy)(JE+hQ9V>+q@H_4%jq;@r-)oslag) zF*#+n$uz)-HE`nTg$@-32s{kLA*P|Y%uUQ+i}$9Iz)$I@wyWwop+SKgsLEQh>yC>7 zh$UYqHz($vXvJW{oMv6|I&`JM-~?N`jBL8tw{|jexhc!yDWpp=>fT9tkkyv2y>vI9 zu-+GgTxA@^u_d&J>;n>I}Vs2vrOql8mUvmt8Bgm+(W1BS(hhvG* zX>5j_$^Kj{LZBuDr0%isZ!_Bvg-rdkIL1k;6ny0l&ezb$nyQ1LJ$RqFSkgJJo?Lmf z)s55uJb$mMTb?jqj*~%g0e9L2Zyo-$8k!B3SeP}8QghOSv`%fw;wS^F8I*Tk+|$Da zQNn@&mUIUN`=F|L*)^NoyVTX1bLCryc2d4q9{#etJ7NkGnm*{Lle>HE-ZdL;4vSK_ zrQ!V1w6SD<$Pnk;EOZ}t6)17m8GBZ=LP^Mx1UR_6y3KZQ2FBlr(1?c0<$K;WkcU3H z+i2avl4DMp=v~+MuqPfeFU%=TMFf&8t$p`q-iHegiWPP?sVyY7`(z<1j+Rm;i8R$RFOxQ%J%TL<>GGGu+HX3Eq8Gb?bf z_2;rumleJCyA*}(Yi|}Bh6^Xl|JR70gNcxtEY!E$~H%hrE$ImIk+LLLY{jY zw9wA#+C%Q~O>WsraF*s$b5uAby1Ozx(H4cdP~x8$ENr0!v8(fyE_=D?YBRxkama~` zd@F}JUx{B{Bw*Mx*%=YxJuUP6WiT{Yv^G1p$iag`o8(QicUaHBwx`*fImu9@Iw@+Z z04E}mQhbd4YM?Wii9eSt%Uwb8ZZnYLZSI@%Lk)E$vG|dTVU0mcoji`2KQ!kJ=XU|A z@t;m~1^KI&HyMyWC1)9yI@MJ7J5S5|)(_l$-@<&&6c;y$eD4t$^STd-N6v{C=(OU< zlGGX@e^aCwq4@;n3j@5dYDKYvd%S0-{aK2n!Cy`v>(i&1szPsSGpqn0%=uL>X_eD6 zK9F@I6YaV&*=40rzi)ant@SvSP2Tk=-twpNkDB~x1&MpE;l5+!Qt$xbdQbfV!>tf)o*9{C|DRobJSd~uLvn?9ck=q8F?$WvX65KbZM_X76R1!ZtZT2|qFZ@M4_53|f1 zzhL)+W)DzeR0YEqEBSP;F32?Owg7b96F(j`?BFCRyUUAeLg5XBSCF=oA}RBU2*FrH zb=4oOdV*+`2U2fge@{RExjS_K1T*!iEtK}5lpS2MZ)S_O_MK)2={Xm8OHa^dmQN1q zds7A8IqbSF9duX4Jxk6=gSSWQNQPBsGPtHf9!N$%TLe`!s5`9G1GS~nT{}L{RO>=M zM=6y%Tpkp`lu9Agc)DsY7bkbJIGWWTjjrO;bePbnAtCb=2;YW^A5oMtA+}Z8y?5j7 zNDKL$lBZ0pZC#;{x{$^D@3Yti_)b`9&PuL1;6BnqM0CXa*mF3`ETY=7G!qlB@+|Eu zcMPl!TupK!TJUiQ^ms4$18350#zWKP|M3PRe4Y0}8&cVHz!<)@;-kOxT=K&(pD>8T zC|2Hov~q*uiZ)5nq{4T&dSvN&vPlKJ?|`&G$6s-c2eE{10X=^I`%cdzcnrRZc#)@Q z-=#yb$#|g%F~X5WbVN|U4=jUP2h&)AvO5Clj&QU`7I%`>bM#Pr!UP+lwZj8j{v<-9 zPud65sHt_C^cF27yjf;S(3AyfdV&Nl>0VkLC!+*R|po$G56enpoOqRJN$ z9aRo*Gi&bRT}Qz0S!tmrc{#z^|JW`Oo4eKLEcw)DkM1{$6o{agbGhw@>j z1VqLZ{CId*>B?Jg+pW)4>5(;kTC?$8NX|gHRsT5Hr>31aWD*wyA@wr@Yh~G%CO$6z zG*}+%`eW<51lg1BM@Rkl4}zIj&EehwNiGb$9bA}G`a(Aa3f_on&3!2P(Rz0im+rCW z6+?q1znj$=9o}MzLqdK>aQP!_Gw z>2ajurUm8-fi=}MK1Dw7uH=M^Xx?9}hbsCXeZoa~1UCZcW_goagynYjY3*tUslbNC zpkP#HsKXXH`nKro!?rJlHd9~h=7N9XqV;j9nBt+3b^Qr!6YQc6YHq0r1~wju&!q)^ z0HD+dKM3L=@xW}#~w<4E18@*>|&YyzP zFtTwaU6PZwC;EH&4bYrV;AMCR!cjJI&yP$Hp+E!*6rPvWRqt+-SrU^`en)>#tAkyf zeD6L3OI9Sm?EA8Kx_LOfs-HJcp#LA8>f}N&x1aOl;Lg+g z0So>zh)I)7j`v=AACay$GIxwz zS|71WotTgzx1yLHArn^vW$YvwciO4oXz)3Sp~wNF986x~Try1v53&9T z{Sw*l`lZZJN#*lVOCTW#N}GwOW-o2^eEzhpBq`#D=b~RnIP4H`%T|dEHB!BAdulNY zdcL{Yx}RtM(NhwSm()|1+y{YNR3CcZi<|;Hy312w1jd`v6Y{ zZo{-iu)Dgp(x}N&=jR4D)7?8LDDn}QNM|P=z?jYFf&S#r*Wtzl$>(mRJo@O4hV8u# zRG3qZP6^JXyYi{tof*Qul3ut%a4j`#P~ufV8O2#$;7)XC6!5ZyUsHY0QU+;dt>onA z=jD8jko7jBCnU_stlS(o=oTTizq<}@hL^D@YKd8X#(Ywl=2Kw63q5wQ(Z-ci6{vwY zI>ps?cqEuC3_zK)jV{3-xH|5}A>5WV99R8d=pVlK__>eQ54tU*m?t2W$)(D81&5fK zLSof6CVvt21vt89(&ryU`hiDE zKwH4>IuHT_|Ly6?o;hjr^m2LX)KHJ6s=T-?mLqXxtiKpdY_!Dcc?TP`fE>FyzMCRI zUs+-R?0Bdg`C)c|j%@N;JfaW1MK+1BjBBB!%7f`{F)m0dVJw%u(0N$HN4PwZmuHf2i zgAb4>9snZW27-<@EHelJkIj@kD{ty>bb((-zY%n=15{6zV!^wQ(W4AOaZIAWT@xGh zRQnrQOUH!(Y9I90hzoeFj?91+_cLx7y1PJI~dDTI?Of zzITZRDtprlcvaQOPq2Ja7GXRHn18gb#7`S7|0a~w;zyAU%QfCOtXgm$Wz)jFn3VAtKpWKw|-U*1%Dz*1jogUPr17FNW9?g;~+f~H(8l2Q&NUlH(gJ^l!bku4OfSkyHney8csp?=)7<%6}!x{>E@ zb*2wvy@If{3;pi*p~AP2#$~}YgWrA`)D8RwOdK5QM-N{xP#CD3%W;SVYTj@NwD=dF z73qigUUSVVBQOVNBMiL(=46-Z#77a)We zfFbK_8v!InSf&{S=ncy@R}kQt#2BGKjse^f;k5|_2&@OWY2O&{8v}p>07+r#C1>|3 zq;{wMx_l{RRx4)`o(=1ZkE*UH$0ozDlu}&8B0fuLEMS}Zis(6%cu*o# z#+qS0KL%=qV59Xq0UeZQju0x)`CUo;(V%;ZCRPHA)m2x=ZO}@KX);v}k z6&PHA2IY?@qTM6JyC5*HPT@k)=E`WCp{>O6c``g_Cir1A421;T)8M^1LLX|yU%(Lz zjbg&HUebdDhlz-THo8^ptKb5%hmZW@ZRNd$h69_8&>p@u4%mWA`i}OaujTsU z2rz=uFcM)-VIGOVoqa|M5+1#;ua{+aXd>PgN%qD&((Zr05(=Q0jE#bCihip&i{M0^ z!YSDGS;Z1eyAJ5HpZ)-{*bo6$85$8nG=1I>3#c98wdd=^ziGtS96GRqUqtw=n&6DC zUD%5t`2kcPh}o&Y*}j1#!jw(}<}vnQ^n22R=-#Se;-}*`V3bA2WjTZ#1n`h|fxA)fV9d(4c~IbullsCq{J;(Z^xLd- z#Hs>J%;fpGnn1B&0m+_n0T_>AeXw%hT~$*vGf|(0P2uiQ!2>~)vePpX9CEiqHYV%C>EtvTfV8 zZCj^o+qP|^>%NZYN#DVn47d z=Vr*%_ZKCBDde13`kYJDO#_|Hj*&w++_Z#EUwd&IoY|&F?fm|p2_Ii z?!#x9NIQkqE}xAM7hZPFB4D27<)kUM_91%l_ooo0005NNgZL|ZE)PCzY-bc8cTjG` zdfv(|#?RZH7$-L7vSdl>t?7xq-W6Pw?GemrD@KP0eI6|CxAaLrfIEOv&1slgaiO$- zVw0u@lC{~4=f4eg@@6BGp)7adp}0@7&hl?)$sup_raLZk*oFD;d2x${XL!O((? z*rD{c^ykLuHTT#?K5_e`cy2n(yKzH|BGqmZhwLMPGpdrOELdj?!zxOX;+QJ>v=<{} z5tLpe7yyWr1Ht^uFtUQg!_cyG5Zfb}`2g5-Dj zH;xY>2N>EAqm?bX8sc7g)KVy&(b#B^Y&a(5WzaTK?9Ku7vWRP@K12!=uu%#zm!;2X zGa$|G-g;sqkcvP6bNP?%dVJsKvbQ&isT#}}%!#z^sPW0HpAVS#xSa=1=?{njhT9~j zY2RW5WpL=#3@8N?R%B}s5Qe=&yuUT54@N{QDu&Bth;y`sx6X?a>lKV$H%M!}w7o~D z;oTt5m$Gk5d`hyQAR1GD9;0_ue97jA*%|fxw3qjjBBCP0xr8C%pXUQ!K7S!)ge`qzywqm5O-x zUuWP*Gb|~%uo#ot&iD{w1pR_}-XySQdre8wp5ku6A_}Ri+d}JisF-P_jm=QlMsVzD z2K@JgOMil{{q>UzfsupQtO?d#05nwRgh3|=>MX`0k{dAMZa8WqaX;x+m&dBMu@0I= zop9Yj>_-73lAOi)mM_GeNDJ-Tz&5TrumPmlhcLF`H0;VmFA-^~J zkFR;HQG|d@Gj_GJuT)WY_xHbM$=M2$(%n*)ldXqsEb4_%Tz{=ACG#Swn2A-*U$WhY zUnoSY6-9oZCbt~64H2K!U&5*awS!IukL%@|%5qr30~w4UGsN)_XlWBo5E>3h5njRa zRi5)*_<)ZS_Mu=>snaL}Gi;Kb$Wg{CnoQy(7&6_2^@SITe+PTFK@aF`4-K1<*`4tP zgJBBcjSxujGtQ6fTKHNGR7RjSgDwjIIcNrTm8}xZx@1;Ny7@kp6kbqHQ6>8wOI=Ij z6Oa&JZ#kpk(G)y8D5_p@aV7LP)6FmjzlP9Qg@ZerH9X~r2{!yVvQM^u@p^`GH~xW!yFIKq>)HS07#XF>4@hC98k#&4c{qUHV?TUM|0( z)3n?D$l=P}JIuDhs{K4JbBkBBf2djKh@SxOJYM_ctO10M$u22CtGGF65|L%TaR{q{ zGzO#a%hEptn!Q~?MO?1=r`$cByv|sPOkCnIa~jY&ql_!D5I`OGDTO>j2^eI$wdu54 z#~O9_iAUl-d}G~t0ovvX_pHMSju8K>rQRwC0?&WfOA=aqs#E6HU>r0yO}XjJ4H~(v zqz)}PEmG|lIR%U@N|$nzXrE?p=)9oueQ{I7PDWm);+T>s@(JlS1pZwoW^9d#SMOkQ z5|!#}#dMwX=+JZy``Wxp)E1cDnw{( zj7N=|H4$n3;K(5uESAXc5Ua#e_mu3JDIei#wQ_W#@?db{9o^CntC3~VWczZrmvTiG zGCIUuGB2`cK+Jn9B3nb)lARReBK4gSL`$b-HCnrXr*67=3qZ$QcrK4k#`HHE#ud7_ z+TUNbq`!kD6|r}o9Y|<()XerQw!p|gfm|%s`-?<#jy1+bhH9qM7>z7pdvEYM9q6Vr zzOS(g!{A7x&z2}1Yc_EO&z@~|+@!wnqIET_j5hciJi1-=fniw4{7tmM@zk`at9fP+ zSZ3w$H7Ll%<2Q$W#m|sd&LAIOypRJP?&o*}`m)H}-PlgQr}Y0dV0O7 zWz0-;cE_H>z1?_ipesT;W*6w}1vHrx!ety&c@(r@qyb!0jyGT#`u0PxP`b{MjAs{k zA=7PA$H{0oRuRv6yj65cR<*>x1lCE|605SQlA7g}2|2fRe*#m5CN+8e27@lMbx4** ziQhJLxjcH>-;+H}FTc0;Jo|+fuDQ)a`&2ivW;48TozX>JT{mV83s+#O_TppI9;vh4ZCKt zbeHc?cb%Kir892SJ$LV+*Iji}Wt#Emk^84Fb5SyzmAyT(5`m~@=lBwk>+q;Og);Ie zJR!U`?d^xe-v-kj>z?Mx?tIv20H3A|HnT`o*WxGCnJ$G|OjW=5k+L|pE7L$qp~JTx zZUY3myGy>MT`;qF(4YU3-C@4Y55YQl-I^Tl%7PEiB>daTAi~%rQx!Y+0`1Wn3}gu5 zbkL^$$F08kbQ7xKm{CBT9B};OUf$GEZ~mq0*9kmBR_yh{^C@~YzN&oyP1=I^71(y3 zT)SR_{)rCdA64;csC4U0Wkpzqm&73j*E~|2X%s#=$)NhI?USd@af=VdE%z=x0)L0u zwei3No_xa9aw+{rMXTe+zULR^_Dc5m*qYw$iRSEv`%|8dFXf^WHtoQ}s!!@= zvG3l4OI=L(J~vhxvAkJkDlrPQpWI|(t6{9s7t0N)wnL9(f$ThdyV24>>EjeAe8S!8 zAT6Jy*HB`#v8lh}r{yExb&^2kuFIFYKW$*%Te_rn-(sZXrM=o04hqN2lZFM&1hn0p z=knce?KHXg8sUKBf$8zt$pn74A`;(`WVA+{ZQ1X!fj_`YZizkrEgEG0FVP?)1N;9b z8kBL|5dRPTVM7)qb_fz#&|B?|Y{EjT;qW+MHIjfmgG0NkH zuki;?sZZ>`%kCmX@i0$IwsdWT@AOXc!e9~V>WO?5)JK#BjXBIE& znWE;bP-8QE0I9YK;nGkOA%YXZHN23z@qYzLzpb@aVvMoqZvV#}yk9{QfqXUcf!G9b zwnZ-cT9*cnOfEOEjTju7q4G}@du#ZCjgfH%`^{YxZ{vc)6KU;m#=zocgV!Hy%Y+z% zH@AXa?~DL$)z>UR{3+St{$3#q^(B-aM(T#4mYuhXu|4wW5)>_z7nZ!!Sm%LhS-EKe z?mY^Q8~UYi;o?pnH-w}F0`OY-PB>n0fz-$;q);KqU(VI(>I9~$kGa?tbZ>7e_tRBb zPZVvFhN;GC>k1OZtfWlOydFQx&dLD^EKt9 zOgxq3E3LV7=$D$73M2tk1$i#;>1Ch-$c9*-Km?07HS@Shk&fJBBgc)teLmn*h z#}kpoZ!`5L-bJW7_Zt8V$cr_fEvZvV=F~*CFl{avATE}H4KK8PPUBsc=?3Y7GQH+U{kNuaUTO8zuw>CwxnCOF3StwJs8ZQUY2sXaQ`=b530qmO z$4+GhX|TT!xWvG1VU14XMOpZKYBoqM=C}e(G}j+nR@Aw zgIWoP*kCKkU)%DGi{q|R5J|eXZEZ4ST~Cq0_y$od;Mwaz9fXJJPHX;nRyC@&IN}jv z!pquMs|VD(57R*-)B*cQ94@&tLX8t1S8kKq`b1W$#sHpb)Yp##**)9za8}|W9zy0s zg}C0{9ZRAvRpQPLZVpNjwi$c^o5*&c_8H8w7d0pRYBa%}R;ve^MeR)^ARWL^-q(x0kokgH@sIkHJ-Lo@7Jn;~O58iZQocOu zHE_rYAc+S+Hjow==?F1EQuE&S=73pn?=yZg3%ppsSeeNT8&F9?jOvTiJQp|9Fm2#$TkiEPB|U@1{5`f$IAkJ{RH)P|bCdoXhuuT*$olY4 zG+v;M(+?d}Ku#Xwa7)>o;{K&WK01lPVAB9v1uI!i zhsYUhbwd1t=K|J%z?9K1-XAFroW^6eX+X)dy4?(v<|0%OL_nh7<{5<9(rxm@iTy9y zR~2A(;RohW!}>s_1)~c^n*Xp<#>L~t%792$RgO{2{Q7KJ7`2Xl`E%LX@xqE&vf^Xn zwPwkI9hhWs*BzIXN|m@_A~}u1=p`-1$>|y}=TY@^0=%my^gCb~;V3N7Pjr4gZj9+b z(w?8+P%`eyi5*@Ft+bpJehFd5fIh{+?k$3Xctk*B+eR|a7Tk=K%>nRZ8@4Q3%M+}< zbzguc)!46`V0#0BhHKs(BoOQSgF#~ayK!ouk6MRG=Xq22u!bWch@Vkm$|C=OT0; z0G(2~A$5{9B>a3jy)85WYok+uV;I(*(NTo)PNJNP_Z!1m_#bdG+!uF~0kgy(kf4>_Yi6a|8SuDHifsSmu*p>C)=?y*_~XhAqV zVCVvSmJXW{!2756&Ys#0cmj~@=k@LRMl$kSKrO&+Hap&00TO&sZXZ#y2*4CY{hcF&Tz?oPp$J=udIGp>1(j3(2m^H)T@oL!e z85s8DX%gOxrpjP|=14>-w?#b5R5#k3q?4euMUgup-!c$hTCPRvZTSUT@61v=Jjkqv zNY@d0Q1lJ_&N#s7Gv3v-9UxrXH{ga*He)P>q%L~pPK7Ym@nGnYy=21?ewKNG6#y_& zTK_Q7h`B`GibNAES8j`;zgBVc^wh9ERpE$Y(1b`e!=93nt-`phvEdu_l7y$q4D2C3 zkThXP?D2YviJ#$e&H22;^h!YF_W=^7Y6R1O`1IRN9kKvqYQ^-5ipd3<8RArGdNSEVYkS>yL;nB;q&$|>gve?-gG7LKaPlW!kasbBmWSfr9-&dIo0{b1}Ijn?98gv1NuR zgn#JcK-}|aOmRB0(pT=9Q-U4#S>z7{V9xMIO|Y6{8jfBtYKp=Mj-A&IR6sGcREOn4 zENJq*FxFZ)&5Jx>m?4X{55tkPQ$}CMr_cLq<&jy#@fhP@jI4EP2a(nN3Jdb*{nj5n z7EG4+*wGLW;7#h-vssbZIqnwrg_9D%)L_Sx{d$N6*Nx??c;Ey>&Z1Aa)}y5Q zlHF)~lLD4{bAC)HPObO-awun^%Z>S36F95NaV?^JIhzACT)kkS9!5$a;IC06_ z`MS%m8>}s*3>aa-I6YIOw_Oa&eHz^CwYc|wRQgp;Y`3{<8c5fb&4t5ID(|}^k|1>* z=2KTRWQICdPbzK(T5IQUz=46FY}Zpt(ptWg+~|mH`6vCmt7Gi z{JN@51)%pqRdA*S;)9`Lc0qQ8V|vA}mMq3t0c zT0`(#2{SP12P%t^FTXr2r=bf=@Win?ZmU2A!$IPM6W*jjIS^7|RGJe3c%vo`^!U4v=InZjQS%g-X%*f>+~900Nt+(&OkdcHSP3_4~J;Wa#VgIDD(K z*foRS<4?4p8E3juON0V);|Et#ODLT0y7D@zP4ES)zgk(S}~mFBEOGIX{R4D_eOu$i3>%@1?>JmfM}m8o+4( zEpF;eU2$Kxb2lJTmo2`;8gTvz2@*jY?_8v3R+^F;n_^i6jXm%wY5hzSr(?YeVm#rq z3QmH}0@-T)@WMS0nJ#UWpS+4d(Pt}jx94GsM}37MoOt-vYArnkp~ILPhFbF!LFqOD zYB#m(lUj5{WfyQUs=S+{3r;NxU@eh#vQ@gw1TPuzv#G%|{iiE}70nCLLr&+QFM@R` z7`U19Ur_{OD8e15kMx?(+LOMLx>*a{L?}S%rC{qfyL$aL(VmN^RN*?^m^4M8bW98# z$!Jmbyna=*v*(u zfGb5jFK|HES^~1OKeLj>dPAjw%3?h%;zx_Mqf?-94sAoN4mit^zoZg8?_3XC29GIK$PyNcILIS_Q#i_J#+C`0o@nD4mEg099uut;r0hehHw4z!o&qNmbGdo|g=r6DGx(%w?q^~h){q+pdC+MM<;Kk&lKMoi!n z&$j~ak6=c8G8F?r8Zdue5*}emZi89%O=}azF2aE9o78~9IqO8I#yu-ab_iOr^jnN=ALk$8BmOJNH$Jrz?+1x9cE4BTx3tI)zP&eSf!OsA}_ieTW4Xc zhb$e`-JyIKG7)lW>9XVROf%U|wXKtI=bp2oXHK zRRIAZz^bhLqx&=D8ki%@j1FP|30Al5rrix6)`~eyWcojX#q>cIz+f~w6Fh}}nIqT$ zI%R}ur>OC5=9Czx*T2jrfIb8%-`~xnJ*bVp zFHxu4kg^M;OWOw(Pd3`so$vY8cHQEnRuirg;!duwVf_<>lt6=tmV$ko%?QmY*UQqq>En7V!85Vb{h68R5eoag=O? zirDS~5(-yjclXmKwi4$OCM~t|+mXE$m|B5A$quK3x{sP>4=W{#t=2EY`s8(Ejt_8$ zhD`py#fL2ae|*To%E9$CDKzF>J0)bKqDT}cOM;8Ws>x^}r=}Pa(N)G0xKBNE1rIj0R z+(`T@OVYzL@nbU4Xj>zf^Og6cK~tb~effSMaZWfR_{-clpX^XKvz)0(CEs6yuhs43 z%+>yhlJ)uBy%K?(%-wkS5OquZ&+aw2xqGpE$a7Bmf1n$i(%(K`?`G_O-ON>NXll&a zwbZKvxHHcSXpEaaj|x5tmezK^Jnve^lr3p}%5KP&ez6-*C2KVsV&C#fs$K{uY;4wF z4CyJa<%PP}XOu%#nrwp*B*BUujWumLT0jYG%esL6R2 zhZoc73pdaZxr}+zZG?RJw*%zMU6)7k&#@F4Zo8wm}=0S`#g;(c9q1k_>0+=$+pH5acDgc zui5J_=`m^H7FOn3p;>dG{2(KCLVn8Y(8ySIy47AB##aid!rG|m&t&Xt2uiv$s zYcwOtJwoQXVPIvic;5eRV{#u!;STX4kOrSvn@HBQ5Zsha<|SUuxKSP-!HEgM&&iIA z7S~=TuWk(7uE#V_uqBsKLq2S8nE)+qwlu=R*6>L!;qT zH<5(s7)sI0m9L)4Hs2PQ@55hrz+As5Q^DE;R4>*cPZkx4or<0h0ybE3_nPv#rf0k& zSaR-!zj>CpN%X>0eh}Q4gImLJW_<>n%v_9L4og#+-O*@xj$}5o7Q3pbmvgf*H}d_& zS*z@qi)KC6zA1FC6mN%Otm1dT4;u)X$W(&lb`~hfT3)cdQN-6E;YY=yAHtfhi)5l) z$??^T(bu?LMLoo3-Tcm?ut^l=u^3SN5De3mS)0Hiu(#1(op5xF&Y)nL@bec5tu~kC zzUEBC2b%gXy$j^8;!Z%%Peo)dNoyVtz4fAHZJUQlu9CepqLWE&Fowt7)7 z`LVmK0~{W~P;TXqrW^_fami=%+s#xdS*)-^O%8JNs@Kw$*J<83vg9L1qVSL7R4;ZE zB8fZ1rledQpx74c#PLhSVP-y4tQ`p%1-pN%GG?cDxdB_&faZji+Go@iLeR~##{n9e z^l`T_f?sqMCW)lA$e)f%F;q$(P+Fc|VDLL$X*W1=R3#dBpAke5zv|OeV;%oreg{Rb zW%fkL-SO2!C5&U}vyxxm;9e7p3IA!8P5u1m0_532c~twTl%GQl|B`B}^M?w#48I8` z{Lq{A*5t(y>>ep(@awiUdQg&^2J(>qQ}~Dkf#-b4K}I|`sh~0i*y(*~olM_Wlf5b2 zq!Qo(Tz&Q9%r1sf12+vY=hEZk$76p+f{%tOH9#B8n+s!-z^V zh>Y<>0u)v2`D}yg#gbQ-m1;QAh4?)KPcn`p14d%QdwARpl<6byDY;E)eSsq7zRO{F z!av-r&A^Rn)+|7FIR5mpJ2~-J39?csY`C&41scX;410!rK|EstpicNw-tih(Goy&t zv6EF{F)I>04NSmCZ{N%1{uytv{Qx%tUpxglNoT@7<#_ zv&(X>OXn_&8Ojo8XN;13V?fSEow~~tNo7UoCbql>w`aln$X;xatj5|RFY^%-gBtmuyFU0( zY>gHj$NCTrPl@?5Sp4Nz|Mptb9!w0v&nplnCuGAmVOX7c z32(_R0$9@Ju~+Yq4VGHs22S}WLeURgBSY^=(h%nLLXYYZ&xFT4Xrv#wN0)5UsQy9M|73JeteZ3Ug3Ktt6MK z-$FgQGxc$kpvr@?N5kFxujm%&S~@7r~m}~r4@4C&ra9LQFz32V{1fG^J6ZHT*QN8oQK7l-RVS5AK4ET4EMjRlIZYbtz zn!f?3ZiIcD!X5g9jc)yx~x&hl+hNesPD=e1YYLLPa zVkyVumizo5OUEz%TL%;J+?|U*;vzOy&n4;OaXcwB6c|F=1mt6S?zTNn=+!x#tcOn8J#iSVS#}Pk)x6=IU7xdoZsN_HdDla+Fjh} zkf`ADCJQZU4{!y3kRD4%f3Ir@La2=WRH62U%HyS>N#EYw$5g#6FB()lu(6e``MK(P zwfuUePMKg?_xo+GPJ1WeJocrz{CxnTOTNUlvJ8?+I!?`QaCBkQwDzysodq{#zOUt2V%oxN$8*jTR!|mMaubo6C&$6Muq;kcRvl-t{K0$ zJ=9uxgjD`eq5bxI)W4O5jz9F{k;lgpK#s+?hydUiKz5GSZ9;v}TiuQk7lWl-6s=5I zYo{eQR88Hee95n3K+g?ayI|iT8i`&V)}Pf^rD(3ZB#2{18kq4>C)12)-eg*f?|^o1R7|xsSzP4?a zrx9mP%oqsdi;a#9EikYv+qPoE?$-hvM~yfG0=D~)&y4go6x#@I#xva1aAMTIhytw# z#-72#A~%)&C!Wl8oJAGJU85D#we$Zfmd22#)p)>o9gdR(VWJpbf#*4l$_wAaCE*#S z6QZd?*kzbNeR$+BK>%Gi#h@0?v5D9r?Xh=5kWeXL5bn z8q`@D`f!TKJJB9kdS@&0QuQW$A@z>7hSTB`j4{Y_{+0kcczo1=itLeM*T{5M`%dZP zCF7H4?I3-(*{zvq+~);olADoKP^h3mOSUG z$(10X(L~dcpSr2brVR4}8@!g0cfF4DDUl9$S2+d#YbRptt7jr!Jb@=~RVZzg>VI$9 zuTT&8@W{+1HYSYv@M77@?M69BIf7op54-^;vdn_!V7PrcH{_KjRP9Aghw*|5r*hw+ zddfp?j}#dewH&g_>_DokZ9^Irp`ik#y=bWon9Ccu;|+^ES!noVz#MT1BPk2y%r{o% zVp%7_>0(q(o?E&D8v?TxuF5HAx;fOjc|^ae+-`mM z4~yQ#^hFp$bB869wrNU*r#9i^;q?r>oMBF)Kh|1n3S0DWs#|~#2MD+>&u~109)f>) zK#PgBBwD*{UmT2p-sw9)A!`u{+U4k|WTAMgXxuDUHk}w@ zPi#OsZSF5%Pgxs2YB>NZKKvr#pKC6zfbpDgiQ$vb{tCurFLm4g@z5t@bzk2n*UxBf z3grjgfP%*Kzi9>bw|c8Bfzrg7r(XJY6?q$O?w2bN2Z9NYdw&yuFo?s zYz`?U2`@cV?VdTZ+E&+cI`MhisGbz5bccdPMKyLPUm z$}R?-V3&a2x)FCs4e2(%vH~k(Q-WaT70|(GQke0(={9tl^qWjKaf@`df#x}$460f} zi&RP$t#XY??Z18bwY;CtJ77$al(*Ns25xOSx^}$2F4t%Y%w3fxmnTeR+AkrWgj^Y9 zso>+VAPjACn*Y!Mg~Zt^dLzz%jY^imAvSSnLoQ(m4<&y%*^wOH`W(EF|A6nJ(+vK% zOzF@6pDF!eWMKVYWJ+36j+>$h-4D||Q~sq}=deh^U6oF2d6XLPkquSHRjz^=z^kzJ z`Y(ZC?UWNgZr?*DCOm#N0ZrtWc0u7_^1&5V^gW+Um5&n=8n=l^Kaswsm4)|*&!?%x zvaWLbxohW#nHPl*@QOLF{n+}lHGKHNXrF^`f^LMezq8`S_d^foL0wU6PM!Yn(4D9T~ovs#8$!vB6{_ruylwu(55}L3u7;d zSK2>M4yA=kNn0;v00S2Sz+O^nIJ7OFOEkVLb$-zBQ&PIB*FE~zaAfxNw;AqaJhmD< zuxmjpQ$S_asn1Ju3XMC6=tx=z{57>8QdPh9uWi;Q$*@)=*;TFk%5a^GU^i#29x!YF zc-r@9T;FPlTo|LS&NfU;o!hh)vzCV&j-3575VS?F|EkHJP0}l7b#*)A2^9&VR;@Nq ztBa|r)6CAYPpf>^_e#xtI>B0ZI7Kk+j4?clSYpsGw(G|kNi_ko)yyR?2pUKM@P|yt z=%yF8f^hRtp+DIpy{sRHCGq^*to3;`26l^HCB2hwL#-aqx;>EN7|Q-tX|c`f*zO;P zYUPnSvRL&HiCOo^1K&b5Q-|LKGPTp5c+o1h^7F;$&5tD{>djYYR2RHYGl@+OMV9>9 z^Xj-=W!nJ~NT_SAp$-oa%iar5!O|dr5A2wvhpYae5)BV>py&s`NBStmis_#cjEOx& zEYa!h$^qNd_@bW=W$yLOkoVVL&CRw?kv0lVl?w(m>b_?Edi|Znd!_Cy)#O8Ki>~sC zdpnw1UKqw*3&R>vsD=SfPj-j@9}I|=5^yal z4WQ1RLrd!ur*o4Gtv{<-nJBhq1BbI*)@y6VwU)z9MfRMj6!BWYKk$oT)6F`Y4WQO- zpXiW-_Bj5hzPBt0rPogXA%;tkSc%@A|mtHF}HkhagRG{M5#JP&f1h$+_93s9oDe; z{@7+{0|3I%pTTtp=|Cbx(eZtDXHY{!y;@DlERg1WjC8y*IiOrb_EI)45|fm~G1(1A z=J3`E{n+`iZJB86q_fXysKArog{KO@X}F*SFpzz@cdo}I%ZE0BNme;dxmqH2rcXz_ zr)n7q)`^AH=Br~gocGIuiwwFjvRNVwe_d&&A-^j`ACn!zJCRc>vo%u&EdRy_f;E%) zfUJ3|jh7AN%`cFD08K=*^Hri>y9^be2~b^H&sT$qy^B_%Rd`>)=;IpQJ-v|@|EdU) zoQN+fxhduj9}A&j)HpMbq?T$Jjy16a|BRUp#|LkG^KVfUOIX6$vx-9Y77D@vmj%?j zmjVL*-qKscdp8j%nL4%(=j^5sv7eLEdF?pht{mAN(mOzSPYbo!ZF*x;E0Kis&-9CaGaShR< zItHX=%h$C|lW&k+J*k-5z}E$W)FED5A2r`EAVi3+0+7?G4pVzV5d5ZQ@hMhE=38h8 zW_TdHwa8Z}G)D<)__qjCu6-6u4v-;hW(ivoJSzP^5a9huqe=*c%y6=d!&zfYVKw}v z0ccx9*@DoZ_bwY%89PLP`?N+pc_W#mAo&ZGP=ZYFHG^@TF*b3>b!=b8I?a~51q+oP zTG`dCdlNWj7P|Bu3j}6;bZTdiFlae?SO9{PoGlf7tW%t0i{sh7w#zN z`H*Gj26G!j+YX95q)oA*2bisSgK7sT*QZseC&9VuVX0qiYHga@j7cj@*K~MkP zPD2tVl{iE=BYEGoKkk-8y~pQnA2fteYVY0ysB$HQ1#-6SIzsn{SYuAkh+>exyvhh- z&cI2;U$4or2fdl)cLs$Q=rZQu$!DhlUwMOHz>lBTD=r8ixs(7xG-fSiKaw9oCfTR4 zHP00R<|uOmF%C?Jo{s_urSIAihO3`=az?6tmD(X=V^fF)rz~VGTPPZEiK~sHl1gmIjY0>}leZg11+oJ^h7f}wlMLJG>tD1>3wZxA zYZ$&1^-=O2AcrGjkOBYYhjP?#e!dFfBh$)7R54FP%KAjSzlJ)^O&Z985JNe{`wb%E z68ri%QLA5tNq=M11nct}nL=pq?zSiEJ90@BO6 zj0-yrXh%= zbnX3Tj*h)%_c(tb-G4ww?_C?r^Bbeqo@d(%?E=nRJ~`nt5&+f5Zxg@sjY=nwzuY7f z+w9Nu5L}X`D%XzT3V_FrKMgvC<*3JHL+G(dw$eJnwWqG?D|!+|vv(DGiyC{yn6&iG2uo>j~uh0G-JZe#><@c-rk z^)3Mz#PfsxRJFoX;o2I4QX80}fcA>Z-f}DNw?Uzaa8=;~+P=z5d?57~26G?4Q0z^1p5+g85LhA#;@EX9d9|~xJMQ6Z(8sG@p~jZjWC>}VU|#Sw)%TE*soDb0Cq3<1m_-w02r z_A8>g5AgJ=H5)${Dp~$8`D@(cx=Na^&^T7bmWpI(J;Z%oe-)r<{6jBu#Zm>~9awe( z#x7bH%!iIz57lB7ZW!?IRzFvoXad48|SMhv0;Nem5D0#%^>H*v}O<1gISX!9lOg z7;pE6;Q2Kod4xm&WPcSJO9=mOeq7HN>r6ai6~5+=*u zw?d|?mGRq*9n>q73oxin)aoF6mco{=J`E*489rJNhJ3)x$2$pu!^(nDR;W1eL1-)3 z3{kqfiL4O*_5PKD6 zguq_K7Tq5QhyBs35e5YWC~M0nm>UW$%6r4Bq)vDKTFpW(AW$ar?*Owv&BN!@62B4x zv*1XGf6oY6K$q?oZM|fI3zPF0Wt<*v0!5^tO_g{Ip-jL-@_=L>8X($B;^B9A8Aac| z^an)Ws*|`Vz+HS7oL3Geo(ZUuz;y3EuQcc*D91bsceq=&RDna4GleXsD^d-+X<>lF zkcoP|G?264k_Uqvau8+-2lF|dW_mYp4onPwYEBA@;8{wGyn&zkFk;al!pmbnETl(u zzK(W81VfDd4lO)Qg``k*yc>CXZ-OQ;E@>Mt8mD!+xWA_mIOf|BR26gKD_$$7r+yAb zsEbg*3!zG{dTYLT@gloSAuFjGXxJw3gMdpSG)029TXw70C4c1)=oyudKqm5{**k+4 zt+zWbxIXX2v=YSJch*ljw~D`GV0h?Gq3JBx1RJc`(mdSW zG)5-@-Huw2=V&fPRsd%O&bu&Hl8A;nVV&G^U;)C~CoIIi9b2UDRT}$~Kjy_hm+)G@ zJsX=mTzYi=<~<11(4^K4FjAzW`JEoAbZl~1b7JCv=xSy7t|WZ7aw@6&f9CDSz~oPR zDAF`!$t=(g$~R-p^!tBU-dk~(_8=SgbibAMwi(PERK$?>%r@B)Mq=ExALb{5Wb&NVw+p=+dj)Wt1 zh0L*a4R2g8oMosO`G3r9{f8S|(F>ZK{tsj4)E!E+b=%msZQHiBW81d9V<$VdZQHhO z+s4gl?Y^D!@co8rRclqBWA?FYS!0XcRoX6ZLDvl;hF#d^EdHlIw0XEN{9cErVF^F1 zZ>86Ds`(=0^Jb8%y|gzHkC%;53M2}POLCOl2nf}XwbI`ioLrvZhHDsExvRyyWz%F4 zoRIDoK|NzO=2QCDFV9eM`~=lWl7sEp)1kq(m2cbOJv-KLlHt0&+|UWx1kp9mm@!)t zcC3<@dI#5N!Qs7lb#ongmx%LqM@$SCl;_ljLXC?G2IiX#)%*zJ#%Ku5#h$FC-$i>> z+&D!u5tk#Mh)#3^_9p!o3_wNcIu;l1vDlujWACW`%|u*DCp_U8HME3 zzn&O%1_qCg>N={5g$vZc@p9fFHz5kvfjE*8o%GaCTI?E`MrWi$JTYhKE1@H_iQ2KE zSN!?Qgo9SRj@k6*SCD2GD4DzIJ=%h45v+e%lqeqP{W>mjw{Zq zfqt(;8>S7?uny3TI{~=M04F}^Y7Q_GqCcs-W5G}D?;c1ow*doRo6ZN=83dA56OfA) zvemx6X_=U}t>`MDS|>?x9R}?yLmkU}i&-J5^8oH=`UaN{cl-q|Y>R;a(?QXlm0z;k z%#%g-JSFI^&5|X6k7Dd7g7eX1bZgiPAyr5hu^t+5%?*U?E|7o(!BrT7{#+#I3S*?% z#fiDV*>Xa7{072VZ%pirm%_G2tD;(4Awmelvgjtr*@8zr3hh_a{0fQ+2kZYVnObP-EvXL z4f+u1k9NWrII9&ww&6Lze%kf;&fy z^T@KhUT=@{IU7c=2axzlt8(hi2#AN{!(2Yu$%uA8*IT!jn2A<3;EKF@uJ&^E_1&;W z@5nH0*2r48tD6Uk+c4c$Hrpko_)GA9oy>k>^VMc+#fA75T5KUMmhX=fkn$o96c38B zZ5;33<8!$THUcFz9;Jk*qS+Z?XjcABx7zMM}i0Gy+D zZgE8sWh@lUAWNeNPXgrtcF;Ef!0J^<@VnaD>WMEuA2u2$Kpo-Z5+4NksvJ?%Q+b(b z;Mi}#`1#v2F(t^;!T3NBhBd)FD=5<@#Vhjb(`iAZ!a_{vrkm^ zbm{8UJ6$uQmxDiJ=NBkvx+s1>x;`k-v#|u|b%ZU!YZ))1nW6rUcB8?EPUg4_>B|1| zay7JVPsXGjIh(q*Z_hQaq|T=`N)gkx@OaetK{RCj_|5w<;!Rj&ys|dnEr_h4%IcxD zq<1EaV)Yz=yR)^lcHoUTAb-{>p`P8mHr`@CT`#U_oR2$kPF6Dx|L_(_oekmosc<5S z5ur7FfIa|^2rcTx)y4aI2Ow78FhlUw0K95*2wLuv)Vg*oFtRSH`fL{wltJz?C`eqJ zDk6I>F&ys#=kl1{b;33GlTFV8gXJ`!dw?dmEC!9%#|3M~{!M>5UmNN1)1;YFE4-}i z1UHXTS(r@#hyZv!tB#>im)PyrUwxLQS$DwIRWm+2yB6F93u`pJ#>flp+Qu_co{Ng` z=16C+K)cqSLFh_DtG0;cd&+<-m$x>R;B1&_WHp=B>k(zFf(gpJ??^(nw?FDk_A(%x zVRx@LkH*5OM}ZSEDG!CCQCIVdRMK)uw)0ri0*VP5(~S@w)~eV58oFeDGQvwYK|_0g z|9#}W*GdA3I{>qeh17qR0?q=ruBLjDQJI-olSRA|WTI2$jHIrT>@H3rDr>8oxAkk{!#0foP~9|U0}#_i z6APO;wM;q*b{o&*cGq-~C>*J%6^n&)rFB6)P~Zfskf`@Y!Vs2*q0a@|m3L(`lG2LF z5*6B9Wx)zz-jTNnV}keyX6F=U&>3yReRgnAM>eT4nX;c2*kC)m~o!e5&qSl|p16z{S$p&4CtUg$| zceve?HJ_zeY@9`U={j~FZdX<1!z2c({t{vCmfRwXmta1C$E-@5NY`?hywi(^38Ncw z05x*EmI4Ue2WWKh|y>OTKOjA#v^ zUtW7#vZ!kMjJsn&IhVV^rVBbl4QTvR>e}wuNCL>zAG> zs!}dK8M&d5fNdPYu8AeE50`MZwzXj|d_n|e#K>(;j175#!@^n@eZz@kk?wXC8E<3< zrx5n$jQH?q0FbI!z0n~j5a;`BFIP9$inRjD;6?TzK6|Z zlDw=~Yn{ku(4Tb>jQN59Ut+g}ku75F-nw7ggA>pUwG5YV1= z&P!HJPoykH@Pq!PlaD>sQmA;azOM0 znlWQm6Sx|uOPvXWE|z&6thx;?pU~~vLI|%t#eGYNC=^1iyJ2SNX$N3r!_x*-C#(haZixq5 z$Fd)32XwoHaHH$ui3iAK)ezKol$uG9c!hAyByzoF^A=6)x%3Qx)dSNOtf>O~w?xd8 zF$Mf_8JgHv?7(B6xw18>5As8U_dxcw$eh&ufLsW)PFlWlXQC^_@CkIU$~eZ2+i}JBEqJc|Em2be#pvGU)p}yiE(o!oW zm2h-WVOCJ${t`_cVjeG;WX2IN_9HoWhPoCke2MH0Z6X*jzHdLOfD~JER?>7>1Iin{`hkKmrNBkx z<3wO5KO2E9ek-MVEUizId!k>SoG?vhBZ|axkwVYfqI~((Ky!3pn^5DrV|*83-rg#TDRUI1ZF^q$lf@Ce+w{@ z_O*Cx1AAvV$(->-%S5h}dPEfE+#>KrJp0!wQb{GrB4I+2I{7MWZ$^5Fqk$-d=2VWk z@#%V~XY=eJe?A)BRXkk596$Df0gtcj7%+oQ_g0&=-!pxMX5cCzl;5KB1MGRlw4Ikh zWx&`kqsMPt<`a-G%eo=jtR(M&Gn|S6u9LugcSYFGFj$ zbmE4quEvbm|Lw<_yQU?f2D}o*5!6sljGN6bKm+8|wzCtDy3YZ`T?&KgvMRHZ<}NjM z#*rJ4UUUOpYRQug=iYX;&j0}qk|KAHh3{3>>-j!waRG&wxY-9cUt@OiiVbXw=U&gU(Iu; z*@1_P{GAC&J=D=XA}z$sSTw`ad<`_DUm#x#sdG%y=AfG5aP{!La0|inRxy*60PBr6 z8Vul-kYmTo?Z1+=Vw3P<#cr+IAOlt}HjuyCOkyB-k_@ZQnkmjfxK9&JVg|6=`Y|kpq`5 zkj>mRAEJN}g6i+s6gDFI6Ea-;l@zkSecrKh$t0a&lLecDAbOS-+Ky-{)CvNDQ1ie~ z+afBxq^kebYa(F{HzDpy519M-M_LXOMeI4=cijVhTHH#<&}mQlZU!4chIvs_aFk7% zkbzSroM2kWMb!J4k7kd#2qtbDCGLeV9I)X<^2B+XGB}CN`J}!JTEr1~C z&;D*wA;x@nJs+kJY-7@e2CI<2;dTkui?Au5ltvKNCx^NkA3rDecf$hy@j^GO7^xVd zr0{0w=07G?1xg zZYbO!0SSIW_-XZ^+Pl{=~Sxss&F}&j3+g zX9=O4Isqk5h6~Ri6(U3Te?pr_hBf-aw!&>3W_N@mX0lYD%Ej=qxi)YR3*>qp{ik6K zo~w*dsd|h$-^X#2rDje*#+J2-*#)*Ds%CQ=5jGfS02DSC2P{8I<%e>4gWD_xC4d{!+BmWQ$Z zhy`*2xP+q1u-(YZ56(re7bBYM7?!5*GnsQ&xs25WKN#8xiF&~vVMB{riHWRsE_A<9 z1c~Y%oA+G9FRbKt-mCPGC^Oy9XmUwr70ebue~}~HZm>C^#z@vgK2(n|rO_#Z|1R7f zK7=h8NasqjfzS=ta)}=GF$W2oP$`88nrqZZ9n51n&p%Ix)Fkr3xslY5pJoRvQXqKV z>NBW_##bfup}zAcC~225*N$ec_?}$ci?=u#nEIUgIEHSfGJ3;y%pzYngPbA6as3hI zsuG8ZMY2SG;VTe_e8P8F29=;`%OX@%NsZeToG%~OBT@mx=Xj_5bp}QwdRiZ@uxJk0 z@%AXi0b#|Ro3Fpy- za6qMzf4??#yw9>NSfhK(l=D%xy?JC5<=jRD5ju`-S1}hhyYEJY+zxabTt|usanQfWcUrje+xus!K1z?SQ2Sjb`ZhR)lf^LvE&v+EVcCF%}Qd} zxJ7(Ih9{2D3-=qHrdp!g7AA+zk7jp~NCeZ<%gWD4oFQ@p02p5(t zBpM?auqQ1i2hzeF<#gnUmEbN!^@O+66>t-)#M3)TB{uON6+9y69N5o_zCeXd&Rg;L zZGt~u54fP&E7p%~p!Bbsx1v9bXq*fj(=JRo+R5$0~QA~Qcwr8ejRP;PnHN7xZ;AyZCQ)$I zv*=1F?oMW=oGq#yFNr19<^$p827LE3CV8k-*j7RI-ln36WdvbtENtPY;wVgQ4!byfH?KKJcE_sZeKJ*On_U12@szYdGyk-3^`p zH>-^*$`8BG8h}1|ywCd=ySzN>mW%Rb=MXgE>B9P}^@k1W^%to-+U62ALC*0<^1YR} z$~VrUnn!=RUHSf;yd<`~_pc_2bQOEI3U}B~m&NK)G8*(d)$MO_%EMpOh)yf-P=uO( z6vF3)A;HJGEwqfvD!(P5qE#pLO7agczm?l=tQh!4Wt4{LAbz1&dG&^=Vk`MvI>H<$ zRs>qL=9by52)M!ZR5ovJcIP~iD0G#S#VA!$n{!#j&Yq<6$xQ% zsh#H-#8>?4*bI)Z~kU}&HetUCg}ahW4RbV;>JP*bb6)wWvD zc!4kYJ{2CfE9`}5k=W~htOze=0_h z{EeC=cRy4wxD&Z&RKlT!ZN{e$fXFtOSk^5Ok@N7n;`dU;fkLb31FIyQ`E87F%E=67 zzR%i?iR4L0;OYt?B0-Hl(e8X~Pb})iKFHvmypn%+sN)}vxQYAmCOkfKd^f1$`+4{{ zzsp@5#DSJ<-aTC*zQm1)jw&mhuXe;)SjEt0&LqtUAttq^eI30Y73_H*7pc_MwGQuDn>JA$-f5e8&;=F! zwr@nLjugD?ME;~t{Wgc|xbSBzi2ODG@Z4~PeMO;gUEK(Xp){ zZGp%`z&DSd22*tL*|PObojRVEC>dPyQDb+Y(11wGx%4a5##~<*dYa3L(J$P8C6@RZ zLcC-%3ee1Gv+v}qZ-dpy3uT|0TOxNM`2|N9W@Ijf50wyrE{|QWx|;akOc`1jEGfE~ zCJk*ZNZG>P(`R23I&A?i=&Ez=di&(a66$`LsI%-Z`J1e)> zF{UZg!5cNy@b6mkV4GVOeX9RWL0QC{FwF$QZr0SSR9LPoeKqSo&Is0xqqp8h<}@#v zYS>-tv+72;G|{S|Vo6fy|C1`Dd&De8++gE7mZGVsm+a|$Rku?s(;)P;E_Enkm@pKR z*9}(*AGkA)bmb+c4m`>5{1e!>%s^lpn0y_1aIqr3;`gfC__E=t^P_FsM`Px6fCEEz z-SHfvdftawWM1x)bLghe(>A89C~`Bts!@dBM8Q;*g_rrX@qxJp5!QGvAT5R&mxOp@ zS}MLNw-l?}18UcnQy0-xYI?h3m?2wLIGeV}SLh*U50g0x`LtPPraOm?!rM7kH1+HY zO8<{ExM3L2EoYd`g8Wl#U=r?&EY$9=VUI@hlM_Ni@SD8fwGLJfcfnA>R+H=_P9(o6l?@Qm?Lcsf14ifG%Ltd>0~q9i9z4~5dpwb4D7cA9NctbYY?>^^oj8hEsR!QWQ?|P$$Z9mBZtWx zRNo{M51ZvND9i{4qI|S)uG2zbm~ZlC(+K2x5O? zPa~oQX9$XNy_gcc)dfJ9ep`M&8N}Yn%L;tA9_bgos=B3Qy99y-F$3JvFqXfNTj+cXY@=na6}fa>K@2%U?9l2X z@n}GUjReRt$*By;M*;ZXanVD>5+uzsYxb_Kt&x83lS-Rs`-YpL#QYykSIx8fMhti^q;6g#* z;i67Wb;J~{veojdD}4#ln4N02oJF%o7h`$aV#ub_?=0(8Y@OUNgjxAZ$hw*0rV{IG zFh<|$Hp3%e?fvJJf*Dv2()FXc;{DrMOc3ez95zxM)WKtl*S=wqLJrZNyt|A|wtl5G z2H4@sr;cPynzHG1o0yR~;_~Oy`jzmo8}^jG&of>_^nkA)hO-~$(eH?3GiT08PW$+% z`m9B#Biq@>9A#rP)K|yKroE3M8yyqB?%)YgpJyhe9!@K6ehtpjTi|L2XVV; z%UTEjF|0sD;>Q6!?v`Hs3AhmR^JnjHH@cldCMb-0+~H=)GGId9!~?q74;oD&#~;Au zR~bxJM{jCY8zh_@_W&2TLVseuM5XLjgaPYg<9$42n>WzC7_8 z+jK0J(A#NW=xN8lLFUAgHNxrh|1w0k=+{>fcnvSF>@KdFd)8dsDiRkt0rd6FMwy4! zZdHjhH~hfBEkVE$I|;n=6*W)BTJ^qtoYVqb0{&q^ZP1?*4`Ap@mQDzf;1<5bFt>#7wvd)@m>Tu zm{9g_4pm%TQ3;>3WW(mS>)@Q7Qg`lC6s!3+8y;t{hbhg;_gL6 z#4FpShOmy|y571C za*r2&8jK6uRulXsPPuG@PM_tDIYL4v2=r%aCogoYki=m2INlnKT@`>fc)p|f5iG?w zoe7}h0Y=GZLphvdnTqotg-IAGDJ~N)ep88%0~IT&?!4OJA7tmd*TM*$e|@|cZM*nL zr&03|$6;Zrv__%~Xx57ZJjO znLxStXsa66TsK{VB?E<&RGXRu`*rVTh6Jny^HhRU+o zf_$fVPjIs)q%i-=HiRvvTx}B!WAUemh=}U)i05MH_C}={E)XfL=uSQW*yuYgU zB5yK4gtC+1m!DcR^A5FnRLze)T5Ym@zkZmGfXqCG_u5S|CqGY8o$Q%}H}#v&$2i1x z?darN=}1c~7Q9BB87>{=KUBR;k2O@~XOpogoUgHY5zDJSQN{h0>iHhkR>d!=ar#M$ zrx?nFBbvYg3E*1#BMjQe*qp^M{B$d%9YtgDsxjrx=6w_fpz%VvlqK+uVAZWr zTcP05*U0gFSFTZytd(UEO1vlogem?Qxa;}=fa0Uu>Rom6!w-wWv8sS z%d(&kvn2$@2jmFg4K>8tx>>NODGw(M$t%I_DcJT#zr6w#7`3+|#hQzH`l9Pj^RDVO z00;q3l;|cm;U@%IU;A*SsG0^EZRLHea+OLKL?Z9l4nP;7Q&7md ziOd$h;zRIsGI};{(uQ|rE8a2UO3;dj#+7gwP-npep%^jXF3nT*6)A2@vunQ1qvZhP z-n9oP`XVm;dgOb5n4m@zkl(K?hySsH$dpM;GhJ4Yg7wc7ICVzqOmYbvUGYA=<%)TL zk;bJigT9?1L6HUsu&n3Z8Q-z6O7qE-LBBKLBlH7R>AT4>%d9BU4*D_2gYi+8*mJ1% zYrRbdT@}BFAgAo;7R?A(CH3SnpNypyjXe%hhWRCtV+COY>*?B1?S?Lh%9z`RD7+MR z-WcIRbY@Bv7P4q;MsaN0HXH)O%DOL!vQ)NcSM=tCv}TO}Yi)}EAi#?Wep4q?hUeR? zcYx!BRG@o95WOY&vgkBXnVr1}Zz)L$R|pa7HfFd+gSQ}K&AGcKv@Vhu0mC<@O>E`e z>C}L2V6IKHW8KRd#@6Wr7`E%@ra*}ZS4g6U1`Q~44Q1D36HLy(LUurY+y|ImlLx2) z3_P%*BA~2`W)m53jU}PY6h8g|ro_XB7*@gugYIMZgdct6v=`0f6A zr{qBKH$z)Mpanz?QZ8s$_3Pp1s6o?`fyR03byaD?#Zt(7lj_Gjf=O0NLAt!mtNC~g zfznATCx+WDRwQ@<*S|4p@6(yHq$Ix6)~Wn>N#w!jST&oFTM%=q+g(?h7E`d{o`)(_ zEeIl_f=~L>pa~#|-XDN^i&%W8&7~yA|3@(Uat{~`z6U=8+^)%kV_5*RvPhadd{UU@ zwuDJP+s^OHf+OpV%8x;ihOBo{ZlmNF3oq;3lDy5=ze#xzhTOMOUa5|r!NDF!pKfR7 zl?ISlM>2X)3%*IUgbVJnbVgWFJ$ zuLb9TEinmL(Vjpy1sxq7qX)+6I`W~6vbQG73H(K-R5x*&7Ij4zC3w|roZF}W2`Koc z#ighBT)3RqhlD0b#MphnJ!G7BP0g&yE6LIC!Ev;Sm&joh$-AQ!{JWGh{U@fa1*oNQ zq_$@N=8O4D(?%||XQtq66 zjUb+K-a5}uvRQn?6$qb*HMCVLwSe#ClsEg9?7m330t(v7IG_1jQ>glX_TaOs%f@4%2iLvE z3)CnFc}dpK>eTDkcZTM(ZLZKNOz=aoGi+Y?qLGXoEmvEnLcegL*f(JlMuNN?v#Wk? zJ;`6{bGm?2Qjh7$3OeW_8^dm7?D_UXkQH)qmS#!dDdhou6`m4ZEEpirGc!k(SaPfs zV^Dk%kV>gU*35=c3;s+uZ0R)f7pm!B_3Tv1`&&Q*`PTvmrglDQd_j0(pUYv&%Bxv_ z+J%LKft&?m&2Zx)H;DnW#!{@&w{Ba6Htte8LV)DD)gZS^)Y{PDp&D;G>nQG-`J>w(BC6=V_-7^+=AA6iK2JnXn?Dw6;s!zo?jT) z*_AdnR_@0~t%yYF-N?JY-PWIQJA`FBGgI>Qtm*0JIub9~NsuBo7+Qv~lc#euD$4`@ zI1oX!7A*Dg9tYy`nsyHxuxP4D0*C>O$hceQNv334A`SG_cdPX?gj2SXkiBW$4I&v43tnMwWum*XgR9;3OjpvmY(6 zmFqZf4rFKyYgzv;3;GEExBKK(d4gMcNINPuXayisP>7Rd`gL}SuY=BK^uGWj$A1Gx z77h-E|3e&2>$hHEIC;yeX34m`(ONYP;k{!o%i8J_M=RXO;{I6nlS)*wHaL|810f;j zLGRdZ+7iv_`pOo;m}0H0uoEHdTfi^V>$U=gS-S`AqTE zrisJb!NK8M@}#JuEZ0^L2@hWPn>Sa3nr9;r@>yt(=Cg^mncG*N~R} zA48f++aXJexeJTuPvUFMe++5cOBG{D{3s$Fw(rd!lc1s&L%)qy9&1Z4i5B5G?Fr-I zrpqfUbz&-@4!=7s%}6dSw(Cb9|KE^S*>0NMo*+1SKXx2S^>2?eOBd(U-aos4j=b`F z0(i1H4BFCWue@L=fYwF@gts#dKa;;z9Em%ykE6jY)>$LCJfHVn@{`1-=$kW5(^*rB z-j{-GfuOEk{c5G*2zFQ#Q$`c(u0KA zpg~m@_V4C0n3;D&B~aazwRhsKqBYgm-;uzxQqX*Og$6uDp^~!A>sr1-r}?taiuQI9 zpV8ZcYhC;A)pn|Q;I|sKTetE7Du!LvN-t@c>=1n}t`L7I_p46xnDx84>mr$1e-1;x zi7EdJ0bK;!d$RE9mT=jZ`+_W)jObsgtOAmE8CAeB+A7R=n1^huD_oC^NPsH_5{xNx z2PBoz9W`T;YJtUd+CLWKEV9_Ns3D4wH>_x z1UzB2=_I2I({IbzdGd0@0I4j3EWlmvV_w)ER$d};|q=hqn}4EmA0!4 zOF~!vYj#VhMMi(xO5g?*)Ou!VI=o7$&|x6I~U%TjBOK8;<<5J?}Le zfRwY`a%D*mKaLco%{Lr`fFhPmRpWh#|7sDa(wJ^O&c@ZPHgUndUlS%fR^i^>*+<2Y z0f>fiC;Vt4o41ALgD5*8rM$;0 zC;dZ0SLE0Xp+zkjb>hDKM`T?EJ`AwSr+9abE}X<2}BU+&=PI{ba&JR+Ov0clWrB>G{A!$i9+r zq&rIw@FPn98!+M8?!veiF7tef?RvL0e6lOxNn{1?5vdP~0)R+b$D%pnVXae_-V{Xi z)!J~XMJF;ZVNW?k{2sDkbVya3C6bG^rt;ORq0<(e>#GsH9ZFNi->>*J3tPdl=-fx7AjT-W zJ1#0llg?;I!h37>kJ;xb>*yK`CQ=rz6R{ZH)r(&S{UEw1A)SGX0Y98@I{0W4mb`p|YcA8_-h4o3SY9I_c{+|9$@b4;#g@qyqX| zB}g=**1EH%U;|{TdVdfad=tPATeOY;Vi@X|A=e@$_s0MlAWB>VZkRU7=fxg_?kXE0c2iU!{I7bw z(h$N;yG^!XHd6L8hTsCqXb}XMSkYM;8b%02vH__<7Bo>8eP4KaLLDq7PuTKj<(GMD zED6*zSYAjzU*)HH@ERTf8|4&U{~dSm!6-iI4r^Rw1oA>$ePpZT zB=aV@bMM+LKhSovRua^d*6fhWC+voD{uSOXb^@1c?nO$Pap&WupxEc0Nnza%qC_Q( z{OC4mJtKzB87|<`(iD>KO~ez(M;d{oy!x)a?o9iUFjA2-XxSQz7}P2!p79$$D<9aN z2>-BE%$4p|9Ni3Qd}lQzl=Bi^s;V8#aEw)gw4*R$VXiK!E^fW~ulj^NrAZd%z zWV%B^LOSP|4gnvp?96Sw-jRr`e>R5DLQBD{gMCP~A^%q=Zjm&`_kOh9@6v?{E@^&h2NnS3KmwFRCBwyFjoX(_gHPRJI{j2Q!p3~xjO(;N>0 zgOqdWbeL*`S>uwD5=sG3L0uUh@J|SPO3mJwIkCz}+2UYr^KZ%TOdqp4{sa~O&Ygk- zL|R`|sJ(6Z=ND-N)*07JI6FHL5U;j*IsU6Xb&l zoYrciGu04WyxhkwYAn(Xfs;F zBz1*i6zmp%K7;99Gm3M)8!Z10t28^K_GiUz(Oc7`pTJM%0OsJ=3iTdq$(~l3T^kh} z#EWEFdxX!Dt{en>(0K_Xk-A6Uv7s90f5H%7l&g&kot!OJf8! zBrt(U57CJgf1sBNL2@GYmJRoUeX3OmrBZyFC;O`UqsxA47{a}qS$zj1DJ$F*R}L3Q zNhyGp0$v_f&p4ZK5>@>4=~5sUKoC`EK`owvii%oOeilS|M5_C{Klt%in?0-nk3j(& zG)nU3%wGSybTW9<;X+apWKv^DL^vtURX2!HQg_P5_ma;%$P@OHOl!;S8&~_vr4T#o z`Vj10I;${O+kq!xgKKvOIoxJn+ey9Z<>T%V){JLX()CfEj#{BKT_E^ zs?2Q8y$c2A+JC0v@qGMUsQGJyFFD5(211w6pm#|`nxja9j!E5mNfZrN02~CaT}UZ& zG@P%XhQloVplctf)*-Cyi_}5f9_;eZ?{|^o6vPw;rWq#0Ak&&Ugw?-deJNWhdFgKG zEOrKZnvEJnQ&?&V?8U4SDi0pFH+>#OX)RhM&(|e*=*ju}4p5&muefOkZ{_0sfxK0q z*dV|(qhFWSut!IItI_ba&LrIJ+yDUMYhRuxL z1%3>!dNRgPj7nH5LFbacDf+C(&I!FcP&GO`)3|q}^pz0aGWm7cAvHy*hq6Ya(5!Xi z;rr)(=lBWOgA~RsF&W#0MD|Q0kjISsAj+rX^Z1w!3mH;C;Z!RnlGzVaTu9j-5n|{ph5!YwgUfQ4iZ<-HPSS#;X<49$W=Yu)`yQ* zmB`0LTixyTIdJ~ugDE{VQrWf8H7cs%U{jT=)l;p|tah_OV@Eh)E~c>XQG;&in&;x0 z;gBtbNQ{f0efNEB*1Ekh8KR${Vn%pkD3D6%jB?s%& zKYoW}eS%?v@Tit2iMnn7-Va0)^V}Wd{pLL^7k*UT6wp>anX(}V=7J@8?7u3$V2iLC zKcw6mXzv5(jpEhL9f|LyDr8431V2^zEAD4niC>)#HshB`o)6Uz(-b7^p524)Tu|gj z9NA#a=EG8J=_A)%295#3E!NtEXX)W7@JBO6DEcwjkT+G%zG@y}9 zwJpvexl}u5HJJ3&_{A_ZSzJ(Szx?#^rJw;~gr``l1eBfuEm7izE9%Cj`O@-+8T0<3 z!t-3TfnFE0pX%d`YS5*YA#{Y4Ju*j%CZ$4WYAt{pDn{vh#Yv-Fq~zzZT7m$%F@7EF zGmrbDwAiDcVL@6oXeb;i#Eax&^<-wf?bvMixdl!_K>;&QWtxwVpTIafFR$NzCig9Z(y0SRLM}oF9IDfOt!i4 zR-+bYXza)o^XoU)HgGly}?pS?WBjuc0GbIB~`?Hra7)WQ_vJlzl_7hVs z@@}1LIefT0vsW%HJ;eatoPIT>E07+n}p=Jr5AH8HKDWAmF@CosCnHj^&;EM-jAiteM_k03s$#bTL_W{67?3A>l1FhkNof~} z66y@$d3xN6!9>W}sb$FUtMSQ!LlAzf|Wmn__+I7~f<`kIJQLs)w6CTv^pQiMaOi0~fK-LLm zzzT-$)D?Du5YQch7~fmRWI09-Ym0aA>qvs?CrAT(>H z)?RgBn6+x5$*3C;wzp+0P&OAC4#mmAIhk?yhO6O5k8wBIht`GibipW6#AMsc?rS(1u-;0Y?{S|!Z83#Pkt2hD0OPzhT*|~;QcnX(*GA)&L}f7tYF(vo)D5KG>+x>?y66Jw&gW6|?r`izaUt7wwO@EZTgD>mm@y=Yj6#tK zley@DyIg6!8i_jds;3>*TFRDD{~KmEr94;JEPSoo=|EMxFzm^sOr57DUpDzjU1n56 zYh-6HaY`fG_RRhHk)YrTSE|y zGJ!b>BtB_?7%YUh8z9>9?Tf7>SzR--THPyaZsV?P^NK>?Vr%MmD zT?vFpyVB4$=7`cEbt2WrFg!q~LsmHRRzfsI{k+O_ks>?8hkml7J&RzkJk@-BPN1{kR0asWqU`a`bOQYFHV-ojbZewQBHPUdhRH*yOGGWl58(Ews z>7Iy9&3Bn0&k{a>zAvh=RsEuKS(x8xl|+onkJM9ua87(1utN}z3x#e#kzr(@gLv8@g|b~^TsosMnWzJ2cV zsk--N)v8*5VAVIr{DxfUgewCEqJzx0a|#Qk7p0S&gxp)M?8+`4ei21;n|OHEui_sm z{|e1^@uObQJ`GL`_dH7PifdkKQdN;1WsnEDI`>*H_7DbErPvl_UJ&luAp*6xlhxa! z%gL>-*&?Aszs#8oOq+Kvi!;V04PzMAc0|k%vfy&_VL1iYQ&$R59jI_6QR&PR^ zY;zn$K0Slw$aL&z>85L}30MwHLaQGL1Ik4eO#dcNu|lrsgN1fY%93|G>fy-nGiMSi zR6Vx8wr0A4N9gzxzC|1+w{yZJe2}xvnSY<)Jnh_{9rH+Tl6$8`h&Je9tdy}1vYSXH zhy`J=jJ67ER4-}6uqrA0?R^`9v|vL%9f38WFy~%`O0dKm;lod+l%SeK3l^XadJ$vB z>wFUObT?C|b&@aKa!!OFn<_4od5t=X{)1-^-WYFP9 zKjGco$lb*;$w~GYO*niaeg$Z_bw1q_0!0(*`j#w_-nAugkWG;7_lmC_Nruy^M*N`K zE{{K{BV5BwNGS4Z0IfkUk@4vK^N|XRK%HVsc8>8O?y(#GY9th=MGt<_?EZEo+b%_o z;srrjqryrXXKu&Zf4e^E?HJc`^}z`U{&TdT@`X<4Lv(^af)EhxGC9H4MOTbpkNC#& z_#z4lyy|m5>;O#KN%Ut(otsBYADX5kWhqBEcu(9@Wwry{|7d$R+ahi+VGi04TPKR;nz1RZ2jY1c*|Ct5{6*rbRw+A$Zb*^{E@2Y*Jje)a zZ44=bWQtDR6z9Jmt{6BcO8JEF7me+!fpk!P+%_m{6t-wir#i@ilWNdhN!eRe3~!c; z=Mh4XFNx)7BRP?v3BRtY?yk^4`R7LFAhaW zmEbkx9b$l-b!E6{SEuTqe1;!}4Ib z(Cs2_0S7OGNJhxy^J+B0!^;o-RCP^c)G0ya5q#pv$Cn&lcBd)eX=!=Y)=h=D_HKLMSUR;JAu5LX_o94vs%_}U-Bg;pyLgys)6t|~|v4Dw@d|E|<3?`AzqL9R1)vlYRZ_LPf z_^v0i-dLNiLU*}?j{2&uFG_DxBVr7Mp0=EUf?tBaL`anO8=T2+&wyXAG67j4WO$&IeIO+@x1QR@%i!^UzMa8Q}4a5G6O@`LCvb?C_(O8nFGxPZPmvL(})X zbu!*2KOl*t=}ClPDkKJHPW&^y5-$Rb z)uoKV{kCW%q4ycYaFSy+m;ef^4geM=vV87II|WIZOLY3nsHMgV5nPm-QpN>_FuUq< zNlaWI6>`&dyPwA-wjb(CTJ@Sg8RmZ&( zNjzCTBMJthe-cUT%i{=qy5!~U6q9>q5|RDfwa`f#V__Eql)+-Pkh#ycFNiLIx zI*aavxkEg(?|UmNE1@tOk)A}ASD8Et5YAzd^peWNTS<5+OJ~=W;@}{9;fbVB&!VG< z5+ZF=g{|U4mS$1GBc~?HCQ8FZMm|Uw^0VL>AJ;Z0r(ULY zrZJcJ19_@m)8WOmYz)(e{iG zxGMvJ!Q7b#v3*t2rM@!kbt4=H$PVs$iv}-*QQHjVyMflelCYrf>Ru2TKWhluG0hlP z&yagJILvvVsyhu3F}AkPViL51bL$HmsB1}})jhC*m8`#rUx2mcAe1qRQXydMEMz%X zKxWQlf>-2WI>0V!Sj%^8FWa5H{q@6`a{BNLFR!?dMA7m2 z`S>?e$28AO%dcS79#}_{I06wrIF)O1mvI4JO(;&y;0dr!BjnAVYhK0^xte146i3$L znGc;663DEr5xo-v1_`r?E-;==`yqPm!R!Z4!+tzTY=l{77Rp+|v>g1B>*_JF(jB=O z*T25n(XcbT4rg7pISK#B6AS$~F^u{OsAb?{#LX(ZtN36@D%4u-w)(*l(H44A1}!yq zB;w-lC>~5cpp_EKhl-y7nK{xQXfmo(fs#;;L<(D(QX9OHsn0g0KXqp-C7{hU-X>)30?sTpGyM;a*Z?5!@MUT->Wa;>Ks!$_p0tHe_@GzW z8}qmiI&TEaauPGqMdglZ_%6YvJ*!7jR<=`;kNHXJ8)`Mj-ug`}&ymJ0u0BVRX8psb zRk0PrtBnLxKN{0vD$0AP8*|lVEqRGaUOP8gpK%B-uk|?fKp)K2&;3GB{#ofv^PvFQ zc+JdyT3*@4ddVAM7ue((sOg{is0R6fj*lTBWyy7yUMds@&8f0>;D#Dsh-Bt>_zQG& zNRO7Ly~n$)p1&o(3>HNwF{^w=T?`y1cYs)}5SXg?u02b{MVwg#Yc4rfLkf21pa=Kl zP8KX4YKhvo_52Xxcof)u(TBQ{x7KM20sKtwL}!^fa~FIgKLbG-UBIo>^hlTOOhlhRzSCfBjQ!lMYdSxARm*Das~YlrdWBjY7!{rQrM732erRb z&^-r1U?;aQnBY?3Er8-W>4ajI7*br2*}XdkG9TCx2j3j^pY<~Ajbo6nh-8a%NOp97 z3*qpiAZsd&asMZbyYAf+&v9)mKNCqgRJ^T^{7cfHeq)bVuh)`!t{7hFr=(OFY1`qZu>QEVThY-sNK(xC^2mTyYm{1n%L)3qc z-Zb!n?K!^Ad{#MO;DSb4%4Z(Z+7x8s-Pa}V+b`0I=I{0?n zUv93zx@4M~6q*td^Q!e*-9|@t*Iei!6mq;5hnaC4#)*B|r8ozX(UZ3V z7HqVJ%<#ERJK7Q2+#u7-GzedXd;9U(pW{IUu4$2jf(B%siEAEf)8Jc=Y}Q7mG5vyg zhBHIdIWCkRCLe~X;}L@B0d7-AzfIBb!#jxfTm5zH2bmK*=_*l=@p2(Owlgg7Kf-?@ zUiXNO8ka5_u7DFf;8^mR=+5$7(0vd|cOW9^qN)}HC{*Be+ z`A%}icLajQBF~=Iy$w_QAe05n^u;vY=7?NGja4+i&M6;=NG?p!LMf0HAgw)la~5ey!hz zj(cE($)d*{&y_x{kMv39uG&^=*_nli9#+chWUN3q#O%`2c(^kOy0j@B^vdB+WQXbK z)!i>WH`Y^OK_$z<*SSR;Rril6i)YUuMAzc^k*M8^i@VJW=abS|wK0r9+;JABP^-z) zbo0Jv1wz(Ls;zy)GhK!%A;x3)o;4_x3n6{@VmMbSB0gv(-k}UbPYAOtq;TY7W|sa} zA|rsB&6o78fie0u*aQcd=Ovm=*??yNbCxE-kJq*MGBR8o%ei@=d^i*EUO;>5G3u8O zgL_cwu4#K4?ZgUb-Knb6a#zDQ?54`U(NvoOCx0aA$OESbB7!XM{Xg-LAk7z6%Z8+rgZ` zy4_*2uV$=48EyqjDasi}uRCCXh)`@b1p>+VF*d}!NB z>={7#x?r)luW~_L&0lO6#9(8fRtI0|VAk%2=;Bdr5EtjLhAF~mpnKbngCsCPH8TlR zFz$Mpy5$~Z`5nI4I{&(pu5K~2Ncx2Lew{FSg-a}^d$LdWR^%7U_>;$*gWZp)L^gG{ z`hj@94KwoK*Jl-UDql>=(uD?8?rHJei+{#{{GAsu<+@F?qwodimCO$1FEI} z3xV5_Q2ax~K0oect3yz^qbB~*Q$GMNCO8Ii=vIwY$*m6R_x#TA8y?9b zk4{5d2e&>Sx7Rs=N^kDxA;8QvI>qwr4I-?%h+wQA@C^nc`bmCyPr;_K>ef*E`@LhG zSUYpj-~$W0sOKp|E)hVjGik8&{#on$b}+_h{kF@e!?jM0=_*^_)T*liB%o*WqC-+X zcuZ*x6N_kK>@>=lJwT`L7g7cJaPLAp*n|X$e4!T84cf;$dO08a2KA8*CipK2CO6A} zNibR2IQ~Cypq%TvbaK!6)%p;Gqd}~-=x@(7$7J384Erj^%W3%)bV3wZfOQH8G(}B% z;QU9Gz?HMvcb}!DMsOj7Pq&?vRGSFKSyq-&Kvvq@G0n^i&B*uOclpFGKR4fp8FEEO zsoTJX!;^=%lb5_y@&a%2$&yQprj=e-PU>izTWqLjuL^JBN4O%1WL5 zIFfYoUF#|i`CFt}&$+-x<>G)w;ypkH@i{7XE%5398S+vY>0DZxc@k9}PJYS}+iYlZHIYYT7q-&i5%F|LU7H z-x`PGIK6vfXLYa19Q&c#SS7ogZla@$_y;&OHR41uN!x|O0U-xn&X9*HHsIa8g-9uO z*GaPa^nnBV{wZd!-)H`+$X=Ub{%^aeE1MLyl}DJL)%W;VoPK8R@}b&ru(K% zD-NCbWUgUW;VRO!Ns%IXBg$|wQk+&%iZD*=zE0C}$UoM#0}+Tcd7ad2uY`(&i>Wch<+CJG&)?62=J7-V+dbZNA z>9`ZdUHUI~l+IFUnINh2scdzeR1qHHxUEb3e4%>pU!kA53r(a#8TNwB?Z)?#dM~?X z)N;0en_hvS^pO=|C-wK$I*wH3#%qCVvD8>nplZ+y$z&@-$5#~)fNlw8NQmINFe}a{ zn7;T{Q=0}X3@~WZ88Ab6CSM+rH%LX^rdRM!pMV5GND=t~bq8ovu`{F%NIwWAy`al= z0$r21`kUwhbm6Lr8{XVyphuI9t+IG20qq{8xAafcWHIFO6$x1Q8mm)w<=OrjCY8tT znK`W>tK9H=kax50e8=CRqzqo0e&1g|Ysn4tYTID>$Vjt#Q=Z}#6e3*&t^uvl_~Rr(l#nyFn@-$X-Z(VS`S*248Xkf1YiXOuGFC)L)< zeS-`3&mTTMOANx{osI@P9&jf=Bhk}=pXlED<26k$NWhyWM=qi|AKk;OBH z437ku54N7lLAk_Yi_=rGuG9U+gAp*}t~NX1G<&fJTV!J}qc^o7!O9Bg(akWvL*f!u zA($lM8ulI}o1Ff@$0#iQP}|hX(AsU|an62W!CvR9z`QU#JE7q${c+W-mvm~wmA`^i z;nN}I3^R_&;3>s9{>afL4qgkFu+2*bh7qf+WcucwgBQfsbBc%2N%u3R)einhR;{nk zxRy$%(v6old0Mpr{LNTZ-X7vy=+uo8z`Tp_9seu!kOUtM&cMFhb0%9@^xygl%05`X zC^REDYA4Ol6L>F&e!-YTe`VnS=5;VrJ%p4~y_Lg}G6g1b`9{xFN39%N9e-M~+v8-9 z4A_Cw;nyql@Ce%gocj#H$QOx(=H@K- z_aUL>rx(v5TjQ@G#6C6GJe_yKT%$o)+9IEOiSas;#66e3)pMewE~JN~bCNJ=+rKBRw6ci2r~WXp4t@K6mK_MM)B zmc@eeptg(R3M+-YLfq(dXf^`g-pJpBrz27(K2jnyl#H)ND{cnUSGU0F@37;jA=MPY z0*hOL+_hJQaRgNamF669QWOZ14)|cEu{8okFniAqVdK{1w!uY%ItN@B%8Xc7RPR5Vk5B`u5yc&XH9+>%CWSA-@VGn`-fa{LP-y~< znWwk9UdPT&_&}!FTy2@h(j&cSoxQJVr@;u(S;Oxj9vS^%zutt$2QIA`+(6cTLSeFD zaR}1z)EniMp2|;ZmkblOYJKW-kvV#^&0r(W?r&xaf5B1H(g)IiP0 z5IWi2`_V(1n1ieW%U4E3KS}k8wQL1Cv)Cc(_Ye8GHXYRV&{k4eWL$SyTh!mi%ny+e zbgVnyc)vjkoXZAB(I>s>RaHvf9olIZ@IBD8dReodHYOTo)5JgOCx#-$NlHluVQ}WC ze#<&8y(?oXW814-MI%(-wc1oo`SGF%(?cGwAkXjlM6${q1nEo@4KqCUmg{1w(Bj?= zvxV(9E7^j?g5EcsVkay~^@e$&u!Mhokr<dWBNVJl65JDDXu9eCU3h0+%MGo`hg5SWJ`Ct$T z_4K)@s8ActgOr^S`?+`ROjh4hmPP*Y{$PEYzGk&^w3V4259V_hCTFz zz6pknDu9X} zjN-beua zN_rF~U?lBm6ur1tNm}3Mgs2&~0z-(x(>&M60qB2~^@OI--3V)Y&F8dvB3EDTv#35X z&#@>_xd$2jJ+l3@Q7#g3C{@i9t>e0HWw3P9B2=YB_e1D8zq?!n^a2ZbsiX(X#Y{2y zTx*HnY-GiA#EOWW%-TB>z&AF(f}NyU*a%>^A~Fcp0EdnCfub}nsi~r8jRGKyDCTea zkB@*DD2UF1N`#o*W=gqMA^Av*f9~^1nAafm7;)Zhd(HE5k$k*DF@9t9k&zuO2dJ+n zui$d$E(*+r=53#lkl~XF=sj^_V(8tt1&Iy&yjA-i9j2aRIxSK2)XpRbupRj>;jPX^ z)UJszn!W>oX}8QDj{5mi3>wUEuG`u&*Qdj&t22qH6-fWCMB?GGiSyhTRw5gp@8XkW z8tmm=pLy>X1H*89(lEBuRsN%7kvvp!f5p`mLt5d@DwNDp6@*QZ(&O{^)fS*fXvwiK z1O#95U-&u{s#%Yd8zcz!@Uwf==AiKsaOzCF~pj&qvlrw zSWuo@l_P)&52-?%5MYD_ve%cjU)uPXbY)V3{UOhqgl^Izp#?@s+q`^%WLk0qf6Wiy zbwe_qgpWevk!dwi&kEyl`(`Bysq&k%?+=lt_$43bJFKa0 z^k8L9=pTD&1XJK3R9{DpC4#bL$q|~SCCGuz9U$wyQmg_4KvHs zG9?Q`8>SYSy#3%1cBkUm`-rFr^b=;NclZwuMoO0eS~h&I@CH_T6R6AWTUbrYl6S4zX-#Wt13sLT^$?dnWzHPpOhOE zu_fBHTx6S+QWmI!kgE5Q4`_<4kJ$R*yI{HVxHG$O|k#4J+oqXdM8Pf_M^ zsV&=oBzd?Uqdt-#`D|!R-j=7q0B>9YV&0uEh{>Ko0RCU;Ai3$+d6#$%J}$E_-Zh=uz_*g4(<=r4w{1mP?0QS}yDC?2qpF)XgMq zYYF_uLz>!EIWnMd6?p1@$d32iaKEUs7K;*-Lwdp%bM=GVNR(Aiza=AB7-L6B7`bI4 z7&IsPxh2}*C{A{F7>~+4&PXEPLJz3=fxk7ZYwQ+dQe4YJaC44|O~xRXvu2gvKh1d#;Tx`MfX|l!pWi& z(vu4bZ>^e6q@ph7N!nFlVBJ7>h#2Mvo44~l@wY}mH1WQS(7E8zg#*Ja*heO}1Rg7f zR7r(cmZ}euL<|oXW;zPvSv?nk@{4;ID|#+Cf_n&*6vs-92-d5s}L5tZlz9f!o%*ER@Dxo~Q4X zHl7zNrHweAunWnm`rE`v9>E?+xYq8mvZ=c^xEn~SdCjKEJ_wY}cRrYnpYfeS;W~lq zT$g>a7&L4{i0||wW{CYxFTUhsLG#x~FOMgIAOgbwyu3fnUEN$POdR0;-5gDA;n+C2 zS^rC%$il|?f5(ZJdK0eek}19a;zS}hf}dF|M{&1t>BLElKfWjqTb$`2RqCU{C@>OJ z$-n=+Q`=vEJc#SE4B~bCqQqA5n++4iByPPi9$Sk7TEo~ZTd|;ZdC*$t1lnG8|ZJvQ{Zi*Vb3t_%#)*2 z1Lw`NsFOcod^!Pmkf<8$e=h(gq*N^F=4vjd&?=}$S*khaJf#bow__3yUF;Rt{G?hS z2sk52Lw9elVQQCv32!Ts;KLVR{CQ^iX~LEzOP+_;blz$>Tx3pa{?+zfE&i<*8ZYlF4_B1o?Y zpchY%WeOhzvtgN}H_CP8W~F`|w9&*PWx&!3YrK^+<8jJ>2==A8-c{=gEr5^o(RgOO zU7jz4!@6~ytIEv~lJM5vd$IY^qhUa+jhKGA(t>k)U{#=$v|RQbAtGm-AD-(*+N$%( zQZzEdPNgW~U}g0t#6AA)mq<+EL6H+y8t7lyK@%u)r-zkwITeTb@n=Oj*$r3qkUi!H9zs~e3`^z6-1_Z%V5FGMq7 zPe0QbpB#=LtLtix84y86fxpMH!I1dEN(!trR;Q`Wd6uFeCrg}sb@_V@GOB!f`njur zPvI4~x}ok}{%o(&Ld+p3%xz=HsGtnAXQEPzRE@;$MPxujYKhqg`Ra1SR}Box1z5D; zacUfzNURAOR|=}_pk5N{ElTu9z68~Tuym<7pNy1b;!TepGL0``^Hbr~f$q0#3AuO+ z!*Iq3KUQUzk1LcN7Tr5VJZDqmmP2M}eZhq+964r|r<43pF6j@O6U2Pg!MQsz2&m0{#8XE1w|K*$Wi%np`Jw4{2rGar07je$zXBN(I z%GL)h=9nP#sX>E*e`SPannn$Amt5)WR8U_KZ~#A7&XejV)N7_$7RuwuKnvr@Ou7pe}X!rc_idm0hC`sDvnJ|ke1p|5_1b)`hv&YU--Mg*+XwlL| zqMAXDT&v3`JV^IZDH~VUsk`Nf}+{f!u)X z4+ehv2AXBBfk{e$e+@YkxbhD<`eD{8k9VGAIb~kUY{>IDm23$8F`XuT?)?WI5a z2m6M2Yg!AaWHZfRoJ`CfDFvxtj91x;EA2FgIJ+3%rOz_gk4eEMW-TpQ^5rz#hM@vX6es2OHDQEZ0|X$PB_-}|5`5irT%JR zO4%^7THa`|t288o=)0mR7bg2+9jm>hwdf-VS}SP?LJ1+g(+Z&IAY(5K^gRNIHQQyQ z!-iRI_KUr0x3FN$Q_YqHXxFHmldSoA21Hy6D1f~g#y0xl6y1oi)!v7!tRvn^QBVir z(+@@QQ#a`HY>nQVj654dOlF_o%$SD31w(Z+!6LnsV}*NjM5BU07}-z)iRhFcT#;#< zP#~w#ae(T`e4t^pUT#YqQTfuK zYypzbk08Rr`URk$Cl3S|*iRzEY<)qeO)jCV6&m@OBQ_e1oe9ROrG`7^v_0VSV46n$ zX9j`qzP4&GB>1<`eQWkn_|?Zmv4)og5LtM$@H{Xl6~B*XxAGOsdXzCnd#l%{u=H$n z&jCWLwTof6eOFqG9cHxah+HWg+G0P;X+Q{u__|y7cO>CKPy|COxQQot@%a^Ca=}Ti z4XJ7cW@r6ita@mo#6fcBj?@6%n+519EaF$wt>58p`>ef1)_G3$7{mP55$v0O)DSQ` zmX)xkNJIv!wb6$*_%K+Rh1&c>Nck|MWc&_hYT9?oUu^^#DSg4?%^ZH9WLHoNMkJW= z#ScWP{S;38tht~=18QoJgXt*;r%Lg?{v*kz;zgCS^Q1sBUgCQtr==^cWvgRO9MUX90 zmt{Vngnmu+R$8MJ28FO96oHWC>%|tH9{doohG?wA4#q`i%ZU_&KRe|p*H(jof)5-} zSD%;+|6z-S;2L8rBa$#IC8a-wyVx7)>O>kKHniLu(TmzcyV9mwugziz9Z-wBO>v~x z{EarJASnExGn|TcdSml^0kQx)2aoc7)#%sy?%$w;G7!}LgBDE*GQeftqOS4lFrjnA zhCu>0kz-_w-HHL!$bV9MFA_yzg+;|TbmDcW`wFzNgrW~g`5_$mUD6Ji;(auO zA#G7$W;Kj0$o`s_E=i;6A^E?t;Ov18xXE=R)wu+3F_uu!YQGiLA40=!YI_Kk!UWl< zQTtkG>5pi#4RLBP+Y(W-)$qNfs2ylI)itKd7jO(28Ml96?vpsiD=uP!okt8xgZ7Ot ztxaR(UcYnm21Q(@&KS$1%rsd&;Scpgw#!h#U0gV{7;^bnzvZ7A#b3QViy-)N$l0(Z8Y9m?Hz|;R6X$%@)klNjMNlWPoUu)Bk`%gvR*(!=Bta zYUk5wFO^|hAtHQ>nNfFBZ*mMgcqj)?n`G_>jJKoN7D*N!_q;w>QZB=1t-;d%Dp?TM zK41Ksi;l_Nke38z=Y*}Ak%{SGiWDZ|+_!?Pr!hTF?jn}j-zZFg5BSeoDlri{ADy2Q zf0v>*BNzbGDEESYWp8n-z%?`D6xWQsoR>p~<}t7oigy@&n5aE4K0;I?VQMtQTyi7Rpib>XBnwvJsy0KIrKQUWZInQH^)4?#!lgV1^5 zxX-((U_o({224oL9Mn+)DddG`uJ9%BJYiD9q&!yu@;5qHZR40&$s8;L66s zer-hB{FdaIZ`c{9NESJdB^={+=X_%pBYuWtoZH@%k2k4<|By~W)@ zpgp_#Kf+s`Z!$p(v5Ywfm!tre1l%mBmpw*%^6{K)^J~Q4|pZc`j;Vu4^ zYPD*+hF8OpO?LG@jqCb)w^ni7l$qBmFv3}e9aF)OSMDuc$6H`te^H3sVx6FR=p%am z(M}X}(+knLPG>{5d)t{pxXiJl{5CNHRwEBGkE;+n0!G)mU|!cCD~0{ncK#!!4Ra#W z7=tGSEGSe;BAh#-NPTM(i^2|H@vEe&YrC?=!)929;jD4ozIZ~ErN17!6=kmxEXsHG zxl1Wgh?a%t{#eqA-XJrVox3d7HpHI)2*ItG zzM}AiC)fQUgQ(ioR)R-}9GA&{*oRo|kWY~NO}~}+Vg`LspKYiaJ!lIO_DGqcz2c5y z-(fwtu=&`wDXNw>u;+W}v7=#BU3ca#(a%2eEx$h(Yf|A0!og)w%WffcmO|wR!u@1& zr|yGGBvYpWcMjqiG0f;6vjhSF9h7dNM)5VQr}>GvFvjn)fjEG6SzXj7ZacUu(edTcc6zq-HtwJ65@drw7X zNZBW~GLlul{ESpU&sn!tggU$5d!}P{ImT-7QlsS~mMcQ;Q5e1>*_AXvb|1uEfZ1~y z;Ip~uqVm2WfEb{9?@#wrz2qhn5MI}0*bsmd`{|!C!04$3U{ky^=5@`cXedC@86~}jn5ViV)(ptx5qfFi$O%For*Ef-QdHde@yNsaD-(F1nieaJQpwJ zO^q%*d=l}F>hxy=)~2;pI-q5wuFz!seoQ;Gt#{Ay z1TOXrt-9gHgPPVB0s0oBQVj`B`mW zm6hWL9{20@m|&e+xn{ZAdU!0e;}`^c@Zr*9wMOQHC5d+-O-Ja?(<<{VmeEJ#vepBC z8jhqv8E!9Z1h()uk)!0kC6BGcQx}L;ztyIO5o@+}?+iK^Tn(W3Jjxvye zasQ#SfRA6LIii|gIpi$%c5@f9uOKQ1>o4zF?=jei%}2-@zaEoTU#2SRtZ#-ilZ6W} z9z!N`?#D6bI=!V=)y%>^An-fk?4IYeh3L0Wd6Ctjlz9S=q!f(b&W~LV!hp9BUIk3q zxlNUY{5j1~Ixs6~_{3dP18~L-V@1vT+YvX$KC?j+^3g9fPk2IUPz%ZSEj15Vot_g# zZW#e1w-GA3Evw?AiZzjpT;M8KL>w}APRqJVVG~M$fLF$n#TL@pRA|>(L0@*++I{KZ zdFw>Fsk7(As%EO#>X-I5zXk|8%I~#}vZxiS%HB^g8ETEAzepTd`2h2i=aM_cmJ0$4B6$bx}LiG^H4rY+GFMU5*a-W{tZGuO0;{2oQ$No+)sM3&l?SISq*#2AA$MXMJ zey-8`hYm_+Y}#D|)u%!_Z500%3qo>qWhGk##+%W4P#ns!dDy~{nV_=D(16%Tef=YD z$}E2Xc%MD?MerK$oxDj=PLBQmH49!oIi~O=r(}MG_ufCM@VE26T}t1r%e=4~DG?bF z5*k?xj1@Dej2)a^CS8+KGXd1g{!%D3O6&b-*CmcV#-1tQ7s@jHj8Qx-Oh`|1m5B+^ z0+8RXaL; zzD_JW0W?$k-gE_xYNEU9jy!^rQ%bea**4-aQ2@8q!uRK)*Q zzoh=7yR~&lHBWg=igUo?{g?I;8f;aO!^Ro2M&m9f#gk`Pl>A!WYmtwAx8Q5S;YhD| z;mw{#i=0-B#y-T8Q&0mSiVE(^Ro9m;+8Wqn9UhhaZkHjun<~D{igEp!yoP1%0u z%NUc;biDliDFhs$u~t3nmrYY3&vEzZo!rL4$sMM-^9fx&!9GaF6(aQa@$O_?J{<9b z&$`cKFKwiW@Sid#u;98kGf-e>FwiBBc})!?rq*TPK6LCX_U6d5h&~E`o~o=U%PPSl zZtIpW9X*!_jgz@za`EAKNsgW{_}uyv0t4uL5+kepR|I7m#l1D58q;CpHahhI^KxPVa-fSW0M=8>l`n2Bj4OX~m+pa%w3X zqH6k%O4p2(PH4A8kg!wD!>sjM=7Up$XZfJ2kXMkoScupjgpok*Q?{Ds_3D_>xmf8Z z)B+n&4#!B`%2*T%`Konazp2Z|p(S6~k!PT=5Q|kN-fh=PTr=)^mM<@yvY3 zA44z`0qhO@>3<5CB}^&@I=oT*4~C1YJ}&Qsc`IpxiwV%uo+3DcdEXaQ}o@vVme|GDKt5rX;p6WQe&kk$n1P{dz~2V zk%!R4g&1|QqFt+89nx>P%6>EiSAe{CPAjfyHGEVeYJgQB+=PN;?H1m+j6&fN8_|&sh@3Xn zMd}A_xp9tpm-r#|;*ryLivRSE=Acs)v@p(K^Fn4@hshRBPM=0moU5K^$K#=EEP7HB56DfL^QT8EKVb>Cf^VXJwAd_ zg@*OrWU6LA;r*S)bqHxoh`&k1FW@hvp4{;7+o*ZIzR49-)POMW)$XnLmk5Fu?Pt9v z3@8|=VYHQ!l9Ev?ImN+{El2B^m=`-lUH$%g?(Pr^zvY0;eP4wQwISvlH=7HsA`WbH zd0u`b5y}|7wn?O_!k`$soXn6XFme{R_Wk~EQWG_r>UCoE6UfNS^;#iLdIxpgbwB{l z9!V}Dh~oL+L6Ga?BMW?fr8}EUb6jDw|77woksUp$0hk>8elWTX4L;4YCkkwx^XrrlwoO9HbV0WG%*nvFEN=71b* z-?oHGA%+boBr0x6KrlXY#;QWW6)ssHHQw(YW$|T$S#rXw2eGmAb0UTW{pp#UzKS3v+7%JEUgljFNu3HG%yZFOwh zwr$(C)9Ki@ZQnloe7w8r-v6*>ty#|;&lrq3E^>V+e_=ViKjKnbqzNc}{CGyj6c5t4 zVw7ySiHB0*DD)36#1E$uH7_IqWOCI}SB-_R`m~?ig$$SYe9IPd)@NR~aUCzHv|btP zxG6qOKK0mfNvMnSGng$=227gk47v%Z@qSBvB)_t-MDQfv@_%ee9U#1T{~N8~5m7En*yk;c=8o zB1K-qrV5@GY`Da65#+7Ck*ckyO6yeAVkk-XNShD{%)VKA7Mz>d) z8S|DPFtaSSQHAGDl#{H9!l+LSUZ|=*(?NdaK!rM{?%w?VnO=9l?!XyG<)`2IcOVt$ z|1<+tVlFZCf#rXmEzIvy0708b%?l#Y&Gm+w8POluBg0Kjm8_5XfCcY8t)0r|5httq z%e6O5nlNe|LI!DN7%K^YK9DHfRQIM7OSethdVz{<)XyJ=K%@*i%-p=UaL-M%$z zZyYne9a2(WWMYo_rlO&XQkQBpl|?n>HLa_tasBKfRr;iJqo7 zh}09cS+;eC#cx+pl@FTjF8QRyZE@dGeV$DWu`yHK4INDk<<{h!*t{WXr2cJuPD(bs zo~=s*WNj)CUp)rR7pH-w`||m%{4M!Wdge1L&EHYF36a+yRqi=5JS*rgH>{1272`<8 zcmFx^1e}$u9*W}daoyi4mo^tD(QqX?t!JL>53G)iidzPtcZmN?A>RpC ze^G$NZc*bFk#=J^G_Ij?nu7O3;BarByl`9e)YRHZH|D)4%SMwuz3*o39=WzrMtP9} zpdoGj_?4xI!C6l8cQ-86o4hPgetoL+$oIfGvQ-~~J`}ew0UAgKRNWET(p9>(wpo)| zVvc~Q1F1K|6I;$;0CvGSWaE%P>j?h11^IGkX~83{Xl4fLm!q)#n|45AIHdz55CQ}! zuKi=>d6loPvs_`v69uQJZsf#xK9g09&Dy|silIiO* zaVC8tM_5=hpQaM*wdP}`5QCU4DICLUjofh^N!R3%Bz1`9$~ug(PsSc>C6FMM15FBG zYChqmqn2n-!?$RJxPzMji3KmiGyWo^tFSgw2x8!GR>I^42v%2&N_DF;O|{^x6W6zuEN!!d?gi z+rMb#`3O4)kOK{M9|8$Y3O8e`vbA6OFUK8pNA3VX^Zkz#JoqGZHdDZvi9TEv23BYg zH6o{Ng^DH45A&DH6Awo1YRkld7B~t`#WW6gCQ1>3tW#mldl|AKi&$LzsIr*h)fPW6 zKlzG9m@!w6DZznL2P>k=4 zQtB=&yGY|Ed5*fGLChgo^3%i99sd-ys%_&=f)^o8#*6e>dO(Nc^cI-Gsq~o~$|8jQ6j5%y-HgFn<`6-|pDz zMnRJuyrD?aN8fD0X*x@bM+h3fwfD736yR&Ya%L{`1ScW zo$_M~WG@7gTs!e|tMTFj6a0J}F9m^!$~(K!%}x$R*{!3r?A@ z@mK@vO}p15R!<8rZD#UB6|gykw-n6t$5rJuhem`9c4{=`^uC1o?MiC2}(F-%c1QL3`Zf!sm>D|@CXEHotIBemN#UtDz9g6uQoN5s0sD4+n7j}&vG~50u)gpSIlR@!crd~sd!itNjbSa6 z0SuFHr^}5C6=eQf@M=CqdY_u8^PmU&m;e)@s@^LlKDmR>PHz~6boN1tZJy7h&(gK) z)ShOw#aYl5(*}BWk1Kt|N=Xc>H#4r>5t9UGFLef`35$obmN)@3+&qHU9hDdCzSo37 zB%O79vXGXcVe2b>NG5bSS-X?s*7 z6DdKSEc@S@KBdpm-dJX-c*$koyTEj0l{svnc!`0yPp+!XXe%OrKE@4Y$2~gz2lC06 zJvel1v{7%m7Pj|dKz}%gOs=Rbal!&7iNKWk&3ypJXz2_DdBs%2HZL3*V*O<_3VDHd zx3Py%sxr-exx*B}I#B}m-GP9fo|4t+>1RVP-o?wO4xQsR)9HU-FXocj@+X#<#}X^+ zewL+fwRx10Ybv4YQgzuXqLN~ zo1>Su`WE=a&eYSb5!sgOs(h0Ejk*<8x+2;bO3okVpYy5~6O4uDikf9d0@_2#wLceu z^rGi~icT>u+L)Nh-ze^Hr%otNd<^UPH(7lv=N2^gJv~|Ik}NechJV(0^0nwXiBjCxY8J2I z9*kNxHc_tzH?m4?ruH60kBT&@M?Tg_2m&4340JV$rff1BqxQcVjyRoKqkWmQe=JtbC)NCZhc$42T?w{Fo08v${sj@Fm zZ&^?ax^Xo&SfNmRkd(_7mRtIhjA0u82j)R-T-3beDySHrKJH=+p)lY4K=ySTPM^ZebX zW)5CycUn4pgyOnO*O#Z~J3p3iXEE;YUhqb7F+BIkL^OV?>UKq~>4zL84>OE4`&^$2 zHHAYHPrNzfF+dk!VUK_0xH|-2{1mRTn*WXM{rbPz9t$VS{}nDR5`^>$Q=teF$ z8u6qVx}NrHP(=rkk&zFP^HJY!zb|?1+-6{bln}3UyOhvCv-3{RzTYpd=lI1uM#VY5 zF@AFB;_m0B=lEkezPKWP{-hRRsEbn&6yv`A!7&t5&x;)d@Hz6K=!GY+@+E1-)uxJH zK83BA*U68m_YDzl;E;G6XE!ijTlx;`1;Z=7c>c!-o;^@ki~WxgtQ+^^`*g7F{69wU zYKHEfGuMI=2Vdqm?M%nYqGPUBG3Jx+Ps*1KPl8e-!iXnNK?tt)l4;XAkAiTj)pJ6Q zp1xgV*Bxn?`dJ5KbvD}w&*MKMn0vi=E*CAUQ%99x?0=2m&h6wV4m7vulbKG81Q^jB zXJ41c2c$oTVlRYW>f*O8nuz7g8FxiT34dE=vaS6zadr7T#Bk-n*o+r_YA{yZj57w}FaaNN}>?)TuIMyG&3d@_^5p|x?@7bI>9?w?g z4AqVd0aa~;Fz&O2F-Pq_nYQ@k3#A#UOO55DN+uoSa8~3`t`EPtXzV#;;K%-rbST{0 zGZT6)@4|MdX0n=!_IaHZ8?yiQH=Zi&&@D(_$=3u*k94Z#dLE^*dxkl?Pz)1eHg3j#-7X87A=bWX)De zL9=?*Y&)rWdYl7XVGsq`vQD4r=)#+e8H3|+`C_!3>}EWV!I4KC3(Q%)EZafLAJkbg zbeh+xz=Lu|xl72p(mZd4n86Y_?3}sxYv%UO6Gfyt#=)9PtitgxSr;5KM9ITfUK)o| zVx~-$NA|p?YeLBxnO7V(D8Ce&t{or$pK&4D@x8cv&3M~YllU4JlhP4f**afzgMf`~ z3dvb7!v*%9ooLo_8Tia*Y^%-4eL}*ch1n=G9h@~+3Q_+Gdtt>oMsqA zUAnF90t}=k>pyLAFhFGi{M-7>TfbHU6wO;>mY;%$a<0q><+*^BhI%!^peI6IRR<|J z>|x)V(xYUxup9V`!W1=s>uQvq;Jj30dCT{2WM_B9d>7e75^j-+Yx1LWk6v;UA<4kd z8fO-ag6(UrU=FabqCd0bpuaMN0BDODZx59LiI{z&|MEU2AvU+WfQT*7xMPls1cVio zET1S1?>f(wY&*UHwV}xBn0J zuvkQ1!PRwS{-h4aG<_T4BnT5U5~5;bMxhfUu{a_SAt)GhKdu&W^sm88a7GA8t5>^J1el=Tw^?j<=tk6TfUA4Yt0P`t?PM9b$B{fv(JN zF#&3W43q-Z(qNF3AjLt^9AP`+=bnf^K@>>2z9jM>ar^!xz%FW)OG7WfY=UrQSr6(r zxkEa7iCVd(;*cDofJ!v4_U&XRYXuYQg^_7XWwy4dMfJuH-zpCktd z%Yr_*{PN=`2EprSZGYop1_q$#^hr1}ZXo0~9rMeKOp+(S#?O&BbaQ_0nXNS8fr}P& zG@j@})2CzxOTD66$Bx8L)tq9#25*0+9f ziMT3AF>`T2OIll_d7}X}8^UlQ6K>?Ds0!v-Aek5}kSXLDJjPBB?7`u2fV8#vfM?t1 zXA4|`HoC77MQsYBQ;fGEb9p!!q&%ZI%n@rqgn;tJGYml(wtwVoYzB$yU5te~lLf%>hm3 zFqp&5?v^^ig~cwR2iC`x@Dg{>tDRNR1|fyRF2X)!h`1@{!&u0ovOhPvT@0|WKYKTe zO@)3h&eit1qHz&TigErt-I{wx!4`KG`k8+(Z zKSA~{UV7+-RQ2+t_pnznf2@u-ygu_dKOn|BN&Hmt>z`UGr4GFk6a-vpd$o6#C?8H- zd5ZW44q~FL?wo~!U~ynh*-oG_MFz{t!C&xb!NP(?LH(s{(PhISlx9cR9dfy*ybX`p zP2E9%`Hl2!q=gZI%#esJ+usy8{)s<#h1)$Rf*RO2f-yh5{5%^B!E1p@vBuq4`~Ld^ zgrAr@Fd8fYMnn|0%E+(FI0lP-kRPv6p!{|ueq?{!9iT}I7fBSA60E@6cG@$xqYA5B znT44E7>n$@4z2DaN@xXeD9{bpn#=_)03(-!PqFzZ^b^2whQwVSy-odn-@E-9yUV{S z8fq18?YkDwWN#kQ0RW|N6#`@Y=`yy*Vv&d^-A)sj&-^|0=cY*2xsNA02VSJ6ZFz&> z0JmO#-8yyMWS?WvV)u`LW6jp(`}!+Y*Lt`jJtcotr7GfleUnRuOAh(hOR4RwM`%)! z0?F|2JMnyF_wNC4k{h)21%=${GOI;X=;jGU1&nZeeIzJ&(7=&GZb%eZ;2v--hJlUd z>yOxKxHYMtmCBVjvsaF}olC~?2>zf`y&1^RIJ{NKY&7RwrysUO9RflgD%fW;JI{5O+g6^(F5X+N$`^ zA~TM9f>C@Fd7Y-|O6dBCGpfseD#xE4F1ax(d%FecUGQH-FNGn#-NOADtl$dcu#Rz~3b=i7e2B|K#Tjx%VH zKxgu>`hjW{N%g1EBGicuM98e*z2b|9zZ$C)4uBBiRZJ1bigX}K|C$ESR>uyXBpD~r z{u?)7-7DusjPL)PA2V6j0{JM)Gq8FX+bduZ#9|P0E%j(7SPynNgpgEwzc5SB0 z&cR1?qL-ouv=rxJ>@}Z;!D_~sy47MVxO*RI^b#F*3gX> z+eH{D?Xlu&@Lt|v@|k!E{lw)F<1o>RjnfMh{!Ac1A-(^OX`R)ESrhP0U96A zMCgFHfz9L(@{I(&x(o%-66U$b9uyr!6zEs-pCzydfH4x~zpQFTb4AK3LnyQ7#>x51`#F$?C_aAlNoJv2$1(TS-T?1z*>NRP3vq~lG` zF}fOP4009bgu*FG=HWk{i<;X)>RwDB%R^_ip%nH1yPjb;Gpa9_I!v8Q_HIiUTc+%E zX(;>K6_Jzh1pVdZrTd*&PkBabMYB+Y4_6j)8cb}Og~1zSSx)YvOc$z&~_iT{|mP6D3}dviMLc`*6XNU zar#;+eBZbOgQ>J&lfA@{$J!_0?|sB4A!Vs7yMa=sdJ)f?@omGK@A$B;v?kn3P;$z< z$YlXv;fk&wkr;BTYB@%g4eOY4eKHOH`1AU>Fu}mYw3Cb%2(zCJ5YJFdGo$UwXXL(c2icn_1gRVzEl8tVF-OVL}}ZI ztMYzG9+!=O(tQZD*=66u9+UZ(*Ej@s1Z&|vyj9OU5`-_6G8_xTFR@1y25JiVr*I_7 zgarz#iXoW%HI(B1T~qFrgusU=A~m4M#f4^b2PMWE324-9;(#0pKUfn&wkST5{GP!0 zf5$^qGspRm@ZFCXlE^M*r(#R45$sOGRJZVWW3mTm7@iNSrFM>eBEV-XuRwYr}e&LSwC9`5@m>o4Gc;=Utsh>L4TjqmI2T}W4-*H%4U464J zdLEPfTlj@So(0ZhVi_>!-UU%@Pz&4Kn&6){Kqta%S6aZ-s=^V4NGUwwq~D30PG)|d z>VOw)ld^|aql{8mlU7otA&$)xa99lgA`v~1hN#aE>`0Fy)&l9iW_>a| zrRr}K8qCGx70yF z^CXI=7J0~JWU?5<;bZynJNvh+`fvYR&i)cxaVX9KEX7u10*)8EW=ZP$QJDg?Ai=`s zMi}HX@bOK}dTl$Gaq(6Akv)af1ame!)x8tC;XtUT%=-o5aS|m{2J{d#JSy{(AJJEl zwrfTP5x2yHhc2KJgQRDNV~XvCA)uUVgEB$$BjlUPxoKu-9$0K4Ae8_N%T3=a_dZNf zQIinKUjhkrl#?`*6>AbMxxM!wwQLt;lnL#eoqxo`tHSY+Q!6znyxd|s#GVnA;%7N( z?51ivEaoZa z8?O0Q4xKNtC8WgvSuz?xjT*8-)b~mxowl~m>k;`w`$1Eh>$oo>+f$kAG7_jE3UIE| zp2e-V16BchEPm=3@#cnM%L$WC;;%9C9ruSBWlYBIuS!&uQ5HwaR>EiIPGcM?Oc(h@ zG59H>8);O_JLHph9R=bAo#&lXTUhut+VzETC(i{7K&ACSKEtA9h>OB&)aX8F5>9~a zJlTYZnhr`5$pl@OZM;Hm*o&hL-!JDuJrTxF#TJP1jifu>-4L;Qb%GVgGRZ`NeQ)+D zrJW-+dyfLw1NSZ@QXGE;<__fceu72|*;jBuWbFpRu~RmNG#Hxft(dQbVn>`;!VV6L z#O|+`*@C~t&-kblfHH?nm4nj9AJm=$F1s6(5IewQ?L#@09MpU`61dqZBHo|qQ#`N* z>je=S?a_W>7hr5#S+j28p&A;Ii(e?sPFsN5{uq_z3Ky=GXB#OqQdVgy{N{o;f^-eU zpzueyxd!{ug1Bvw@WX}BN$IkU$&332=jJ+-_-}lU<-hPbc1E`Uw`W_U^>5QxZ2ljg z+rW4Cw0cbi81_Q;PPFcGd{3@dUiLP$wh}th4iZ5@fL&4k9DQYZtq-}r`~e~ar)oJ$ zuA&7=pSdpU%Dc|cr|=l1;QWsAJGn^mcJ#WxfGTTCdNuFO<>}$#^kXEnA3!&IW9cxM8w#O>p!L6iZl9#8 zbkHn>2COUN``|pRXwTZL1HFN%Yw&+7`H86n?w`37dyB5cJ88p@8gaIRXQ z8;nLlSxg+`Zp^g1W+dA(2ce;N0_@ubL_xS#&6L|rVJf|m6)9fxn;bJ-W+{vw$4Vhy z`-45;SIgbN;8v~wvTDnp@tS|AYEPfztXcrijMmTiZma26?)=W=nA%I_D5NpUdS*NT zpE02~n46VpZ$FRQ)ObQbEg(<$jxTUVI?3cM+wrFuTdR`gMPuwT&{y|`lwiRns)NVo zt;&j#-w_}d=Yix$$Tn!OepcF}5TJY@C%=eET zjLlA&_Kt)|ke*H@jGLzOG5*l~l-$zndORCwT){u2bm)=#)z`J?eoehGNmE2po!yE~W_(?1HRK$j0^As?h)d*KSt zMn8Zs$58u`9K0R&7V8Mr&ryNB>xtVsh}&2Q?TzZwT$w%oCywy0kxufo{H#tfkL+u@ z#MDh$T&*;;spo5t7G4|mvLnxY4{A43PQdK9Tuy6Bsnw@#n&{dfEa5fM=^4o>RYWQs zKvl}pvGOIqY!=pFtj?IE-p942`g5`@&SMxKE=}51l1{M@{G-FtPoEd8UB|Th6x9Vd z3Y0(0%jRWwwaPKX?%z}|EQ2?A5=HP_*k++GAXzL>`bUNR!tvM3Z_=^JEqKzV=d6n1 zXa5Xz{$sGbcje@CaD5w7ks=gB-mcczRdC%uWsfi_&9rVdtl|vFGLEwP0+6Tdvo@G9 z+-`F=@e}^a=6o&z{Vg_CE>GOS`HFTRpc32yUI5{2-5zgNm^+5RN?U_pgP6kjQ+z$E=wh!okYc913rBv1%%OYQv1 zQ}(~kXz!9kpFHvKbZd^vb>|yht3d~kY3e&K3orw|-)_nYIAARS_-9chT0k&Xf_OBD z;+`;>S1WEwBWrYt0h?eP6q4P19q&7F{W85OMEUU{=;N%jj{f{$wMf$_5G#Yxq7Y6d zxryE$;jphbF3vMNB}}L-p)pD(`F1>(n%Ek(NtdcC3V~rASS>QwIuPI0P6b*14gfBj zq-YTIj@23A!w?0OPav?>tQ7sAuLs7bT0O&J0d*qBL~0I$*GfRr5xKj%V@1^W*HM<6 zaVk;!IP3}N!fbX@yAff!Bn>mmE@*NhZAsCoh>;TK)pswgX#jo(Z<%Vxpyhy){)#kO z7igxwi!c?c$qK*c(5y7P8S14EIyO;-cPnkxarMtJ#<-aw773E9g2*ZHZq}|P-$%u< zAj;|QS*_?u%XCZUNO)L|Q_#MAMPI;sKStu>9aLsvlqg?wEFh){4{&F*c0E*%OJ}6Q zi$3@$2s=%O#q}f#iXkd^4!up~UO1RA^bV9uMM=a?I&34!q=uRb0>uwLBso2hVm+7e zR+=?ZUuMw1bo}Y1PxndvUIIe8o3$}eZ_rDRGrMd@ydZ^^@i!_WRFvANUYJ}~L_vty z6xx?k=9VNu0`x99!ALkc%A32pt&ov=&Q6i>MgE)4nLF%IzF%pejB3QIllK5uK|1$q zm1xa449rUqQ~kVe=APR?ydetn0wRBeD1N#+K#V?9y)X%=r2sgg=ivh)7>)iOK3P4Fz4YwacA?R(z>(Svpo)&snhA)&DH? zj0veg`{@dn%}gdrkG&5>EB_Luy=(eQlqd@YNr~KC=;SeIaONnD>dg(5bJ9}&SB$cB zzS$da8_RG7G;IDaDCwa^5T~(Q?fNNGC+#-6$n->=dR|n6*av(Tutt$Y|Cp=KZW%m- z(>^ymFfU-}NcST zP-RGu@5klPYBP1x=vtA7vBK~j(%yVF2<$9zF({YJyaK8tED}MLU{h8)q-HD*RO}zENTQt*q=CC0jf6BJ@>wf7 zHUzyNDmrphKYF#L!bd(Imxw=eT4VwI)o9WQZBOkdKv_=hc#)v!FzrSCAM0O%-n z|2P-Yg%CAZXJP!T^}rl(b@X2p`Ewt902i0?jCi(R`9YvnuZzi6Kdhv=gm@Br2Iv$* z(UmyffQS*?LDUwd$SsvDX0okC%%ob-8POhpPaB*^FUGu<-9MPKQ;PZ|{wJ2aw2^oj zz_*CzUTifW%72~4NPSe9rk5GUU)J+p{Zd4Q> zV8~+ZBFs`RdbgLN^r<*Rt8x_ye^##HboF5I3lM{dN@RmM*Jvnz(MPI*!lNTm1P$p; zF$xUKQH(iotqG_J7F8if*omRd#7t=DY`b#Kx1kmM#4AQ6AhxC2mQUlebuL<{HB%&_2Z{l3ddD)W3CzUn7i2p%ugLjy3j zo2oHhb{3-b%{s739VAUpG5pv5H=}m3egwk+G;}oRN`iBj6#}CZbVmKr&h7+jtn?-} zLLdI?RG5Gh+qidA`za4%QK$n_9SP^=q_9z$6wbg#SnfPtt4j3w7-qFCLxvgSQSe7E zy(=M>fp zhQBKhBS2JGU^jLFtf}-ZxPMa(=!lk>y-CF)APwmNzP5;k-%E9<@x6oR+T~yChD7O2 zKV{pe>}yqxRB1et z$abb=RfC}{BHgW1qZhNwl7M0ryT72~p4UH(6%ad=Un6!P4@DAf2nZ_DMzw z01-yjhGA4G#qg59pNkdN`o_ED;aEON4T7Vunl5i&HzxNQNZRC62G zbV8V<$+2F=cD}fFpL&=qxj*)L12*DZ=cj;~ENWOo{mm8dCnd9%MdgEQKPp z<}9DM7Kr|ovaz)`ykh%jnp(_akVg9u)#%4fiH=e7LbV^sI# zVkmxhuCtim6N!BFmMGW;Ekxioz|UcKQCe&K#jHX^*Ow%#Z5b`1VP7Xpw)oK2sKOiW ziL|UrHAS63!6T|i!f-kcV*h~m`}B75)qGuoc&r$97~<3$U`44G@`W`iT~EUU8W9I| zw!g2qI8Y*sjfBU<{X33Tc9uV5gYZdrP$N1VD^38ufmUt2!A(xSSFXHXwxA=Hyf^n=1b4;K? z*MKf~->R|Y7wHjQs1O<}SX|%ED6S4>E9J(WBW-4=5A82|#Co_J(Js6Cb5ZtR?&bc0 zRlaRkcN;#C>yLc(B(!Q*Vm>U#0%jUPO3C&TqczaO`s2#KXWoNTvx@$O`j!GnJBCD7 zKY!O|2L0<>;3G<`h*bu&{mXCu?X>l{KILCiK-K@;8^R$;D5Ft~^XH>qm$>*`9PBxz z?*rPZuozsRc{OH-6xn@?u7Yq_Igr1^F`T9!d6{r7^X?Dpvf<1;9LZSkjSO**NLpdL}?+<{8f-AWStrSJx1`h5F;(1cw_#w7W{Ahjsb*Y&gpDVKorfg&i%d=6$4H zc*werfoKm78H1X)0wVOd)EF+hcfei+bb*E9L+R@I$kHWEuf3p+ySNq+*2kK>UI&9Y zW3BTOJPL83E?_1ABJ4FXq+4Jf-B*39@CAukhizk|kO04y(6_U4%Rf8xFS9Nyu=vuJ zB1s$|;!q+I%bbIDe@-D@Os^?VkFU?0T{ijIlS>OJV7bGCMPy3n@$=E~p~RBQT+NQK zxY?9_jJl%?BzrgkskdT))d&uilvYU85pAY|uHwi~DK)};HE+GnG`(!r@e>RVcB_3( zf_|7W(Wx3WuE8ZMpHP(pJ31@RvClf44b+ zO(k0Q*>Pxt)Hycu!M@F2fJ@z7Fr07IBIyz6`GM_;^4*G2_6XxbklrFr$$8-=!1e)&p56mmmWVeSOskYlC z{WNrF-(fG~v2kLL8U-Hiv){@bBb>9!U1bLy_O{s1O2L+wO|&@>Sky;KB#3))dq${{ zp4#;Nz|FF%Hv|SDi16=Tn<`4@;|q3Z@uqR#if6mQ|2lCZc~F-@DsCC2DfH2yXJd8( zEj~+~axv+HXy9bbS+GXD3c?Wk#GC(l|M#l2WEWt}QlgnE%k1vYe$F-=M*{IkBhwxS{BNUn$BLJJ5>1!mQmEswpRapTjAW)qHS#N{0ZH(6 zAwN%t{QvArr~d}4SpOTWVq*ONfK}Slj{n)1U~!t&9Wx4sTKb7|HD)*YS03?W^SjKg zh?hxFu|C)~(Ezk~^L%VoeU*JRY^^Q*BKx^MLsOMdBIta(eh3oSK(!$L=_pO#EX9v= zKZIldmEkMrFT(5G)c^xzU@m?M|H9kcLRHWKf)ri|vBvMXT0XpC{QdDTAqc9tY`75S z<5NrDetJF~o-PhN-7l~low_*Wke6k_zN#$HXBIYv(v?B-Q{t4cehX)VGq>29CB_k?V;W zv8o_fB<=>o{H-2IJ`^=>ez>r4Fq2`%YN6Yoy8-m|P0rIL26bY{Vvhz<*ww*l#Pyz> zd9>?`;oNMR{({j#EqAFsRP!#3m!@u4s|$`?lc1llUl^iVDg~ZNshb$f@2~+2tu?k7 z4~|Sq1{ig4+A*Wc`<7JCy%lxlN8Qu%`#kyV-bmV8DQ6zDmo71-VI5i{rAcVC$}6?s zA`6%>ujmCwa~l3gGU0XndCO6rda?y>v$I-F%Gm%+^7^yN&N?;Kwd0x74SY4l4%jDR zA>DR8P;HHxkiR<~e0W%x>itr?I8#d!^`(&r9n99ukH|$ZpjGK(0`9*hIa1Ar9hQod zFTrldPx$~usLEk~U1i}w-R!rdf=TMr4nV{SV2Yp_MOeZgehVUA?+RNtl&hOeBrR%2 z>X~1`oD?`^u8$DY_cT~95ypc=1JgS;(*v^hw3jupWVhVGQ zAR;h=87vXwmuLLKLO2Up+3nrQ>xB%%t;gda#FT$g2S@BI6`G0@}8=PVYy#p0G(3Xo64+-e11R; z43RF|p?;++Z-#0g87z%3xgcVh)xfs6I#hscrlV@yhU3zHUG{0^`4fm9376w8coS4)GV0&Bxn z4B}IGN&_s}RHX0>lepd;fV#zRN2j(<;jz%XJtqeAszxlRM>*Ry5KeA)EEZ1$J5_tMqH!7Z8JK<%YypC za5?zcQTo3#GfgwX93yFB%|r?@&CEu65KD%9xA@09dm5wj8(v5S9ixJV(3tbJ<#6$! z2$zmeeN{VVrk||91f3@W!qZ^fWzxb4`=2YKYp`DaZXwOY8+%_D3vKFY6yZq%9GtqN zDQp6#)|^$$ePcUxZ(`jLasC{E7bc|lV!m88=CedWSUdtSf_cp$TL6dtD$wah(;GC# zOem51YF*ZgGs~^ygA2*1O*0|q)+{rhtW02x&E#PV&?$PBNvJiaa08}CN@sw8rilo2 zqtHA}-jb|~w%s+6d$H9NdDFh6j@aS*J;aghhskV1GFA3PID|X?GRy&F;F^Ri2bZuk zZ?Ts^e-{HRX%7f!hz+PHkh>_2&zr~mp98v1Wmkej+N{m8lOO=mjm2|l_LkMXgyZu_ zv6DG2F6jUQN!V^MLtR$hMwh4L7q5%pX7P)oioqs)C!T#aLg|ewo6-|62*K=C0^JE} z4M;~4dT1H<4pb!G1D8E=jz?^dua_&QK&XMLY0#=CyJN$Si^qr?Eo}>Tvj23#v>S%bt$}XHx}e zD~uJOQr zC&H)icK^TCMs)C>Ue(NxQjqZqjsh}KaWa9pKO9`Z0Zx?3*WM$xqDr@VMMFTrhw#BN zgmj;xq%CU5bISDSv%95`L{_53&Qh%g0D327`HBgT>u1VfQt%$xeI|XicniWXA!;Nj zPc0Ic-Jt&5IQvFHnpQm-BxW1i)3vhB0PlKcOq3W z^~;gLq}*LJCro6*Cs$elM-1o&ksix<6;_{x&JU&6{g9~L2#@isp1FcQr1y#9vNmm3 zrJw@Hx2I2r^?0n1BF31DEg;iSu}2dl--6J>h3OV~OK4 zB$1TFV^*SSZ}e?p1NzA}+))FRBRf@}mi=;Z6g@JQ0cK>anF)HuqHb}(xU*+A82xE( z0TUZZ68XA`F7rJp;pL8D}y^23slgcn2}8>!9S@0j#Rha^@XE0a1h%LIuEt z@g^#{t;J)?tn-g<6m^{4WM zxx9JhXd?!3=2kX= z%meTiB^LDwaPgLYr0N7W!hDZf$a|cvhaj71V_4vh#j+Fbey?xJ3?iv37PO9g$Zw!C z9wW4rmuelA3s;UX?LNk0KiJ?{tUjd{&V6CRd06>c(p_=AEYK&+M{u(Kq z4L_S7is?y73sL5)_#?2xTWw{2ib{d3?f-H|6#Y(>`XJbauOCfiOXG}pSTk%vAm%42cV1CBsF?)1*Q~8w-aJe+ zLx4d}#yV>CP*g#DcFL(L^=Gi59o8;^g^Ju=>F}@R{=r|j)R)wh+Q29u7{kO$5>C%A zd72l(C-#|5(!);=%w`DQBDyX!1&#LZoNm!o}e0wNis zvb}!WAsqXJ-Ue*_JS9L38>PKwi~7$;CiXc?{ReF%eWn*N@On;p=v2(Ol1&){|SqI&w3c)KU1LC>3=mv zZX3E6t0%>2Op{2bQjDEQ=b%4IlVdb+U2TY*kM8yP8?5Y5gc|ULb0xj5gbR&P4_7jU zJgsbU1oJy^bJ8RMhjVRDTLRm|NSRa{XAA{xPk#jZiLh)Ox?n%8q4eZOGo4RL( zEKN8e68S;ebnm$%8$KD-hl*}CkvSDZPmvo9#$<^$F%_m=y|uh*5gm$z8UfG5wGQt{ z(0Fty0t%}~r)hx)$-dy{O}wf6xi}&}-!yEl35qyRh3D}33^|Z$SX2Si-R?}bp?wcA z^IzaL7jlHJg+d7IW5ImN!g-|VP3nPY?igG zsHRz`uHvznDdlqQGCRT$IHEkJH?6e2ob5Bl%nF>%fF?3JL)V4CH`RlTmS@|v$Jf)JLu%<7dVlhdDE!fnrL(iZDk16y+P005F6*!7w^13~v zevBXOA86|b=$jQM}(2D>aL3#2r5y=M64sZ z&b0=49WjEohB?#iNau^@v6=VMqC}bEt4m@xv2oxZtOHIavr{0a-oJ$Fpea7qts$Eh zZmbUKx^wslt&F(+ZR+BM5yZqXUWrHV9zM8Q!wBZlPFsz?U;S;Ec9K5lzFk%!>ro)d zJr*l02JPg;W>zmBb=NoIdx&r|5e>YypZb_uWYd|<^@h3l*AWuv?f)=#j?I}x3!9E@ z+qP}9W83IB9ox2TKCx}vwrzGaJySEE-cxnvAFMBX)!NssM=3*1k&0D7%MK|bS)YhY z(-XpCfogaWDoX5xMdsHz#Cm?ob*=Ss$~B>}v;XCv?WW^A;avB6?jljqad$3?QdvzW zv`wN96`dJ@m`!W_X$tt5m_eM44VNoV^%i=lAVI{mYcpa3{#Ovjz|Z2Sfi~tC8~jNc z4N}Pm*@jX8@0@TE-i0Ub`{FO`HBAE@n;HZ9i&-qO!t8Wgr-Z~xg7B@ZO_uI70GbIE+eu8<4f^0dQ5h8|Q5xKiJW+T-0Js;F~KS zg_lmvz($bP2gZc>F=BFrDI_ILP=n3K8u6BDFbjyp|Vi+>uFtEvhtXG)2dsTXTLNRGA!Ko>>15S z_o013fYv86=YQ3t zPmvyV14pcyQ$ciL|IR#PSs{IY)W~`-Ft?7~fxwfKGmqov@p%SU6|;q>RboP+l(EQ+ z$`6JJuUCBG)jEQQjhInf-SlUSYz87*iYw!^{WTsMD8qvJXB06b-$-$xn8&zhIg;?; z+bmL-NG^Dd8c)D;4D4Myq$a$aD%TF5f!VSB15%@3x7V2kKJ`y+BbPAD?>?QXXp7T> z2vZowaV@%Es*&E|87(A*jm4Lv+2BqV%hjJ|{GD;Qrxc}P_i!xuu7je0J(+oQsVKVE z?D5RDjHW3IO_SCmy>s)09Fd?nWT)47HMdG*_8mN{!vF-o;P6-JQ-0B8GuYwo&m2+A zZvK+jioYmLC0}<{>*nq+Bm~2KsHu;~p3Pz8_(NEvp}+K~dYpsTZOJ+~!;o-TD}|En z3?F}By`64hBFC2tO3R%FDMJQh~J29pK+J}30QIbPhiE) z%*pxRfmM#qq|=r}>b7&Onu46esunU0RK^%)QxVlhAz5UTROfY_7qEH&S*98|hd)Wy z-{|+m2(6i2xgi^iO@P*vk7=Qzc>K+|L$I2VKzy9n4gab6L+ytZ$*amI<0r4qP07cq zm)~~MldIq-tIkbt=bG0e`{c_bS?k1&f$p^n z(E;P!82pAQ%l!_Ir5p1u>TR2dIRv|nakMOJ%{`!;VuHMVHl;25^ZnrPmm7oQLb;-r zuAv(RbH;;{UnaQ~jk74ZWIdV6N6>3;WFF-r-Q>{fgB?o=qK8teng(Z_YJ9Q8Cj|ln zW7dN&i~K(GWJyCIO^4GA`@>e07fvyIJ5^RorADtjg-S(<+QD8;1x5kwd5D~WJ%MT^ z+w}RF@ogCY+4!ws--TZFiBHFn&P4CVz*dR^}Vu zDF0uzPa3_-Z@vtiGq#jp+46kfOfqClGH=F}LuC`q#)DuHxYr>gPi<9DxvSv(xt#D)ZQ`d^bgHGXLg{DY`StHm9ag-Jv^Pux`HifE* zB1he==wqpbEwnambf|5FC9oBj!2^I*I<<#igiHI*dig%qWevmnaXqV7r?e49ge~268%A zcG?bTfdI!Np!A9%d9Xu(h9i6!MFHE?$a-678!!wf5$dRq&p)v;%T(s&x9^Hvbl}C> z&k#|1G;(7^zuK#E@*TntCc)|QMF+U$F_)gDV@NN3wQN|HBq9@e$}fi{fkhIR`dGBE z=ZlX0S}MkgR15h>8xQ+r#|Vg{yB)6CpkneC3=Ju~?bRsu8c*T;me?L&O~%_5y)F34 zN}!7{$*{Mxir4|QX96O|g(cu}Np%$-O}OiWi~OA5;=V^UZ6m-IIIoIQhx*b0QSKLC z_2xQ9@B>=$51(`@k*$+y(gS3oqIR9D(MU$ZLry0w;%M3kv$_CDNy)5%#!{P2GCGw^ zije~7k0DR=RA_u54hJB}xikqvAPlUzv;Mb@8Vbfe!-$*B&4~KocMu3;Rhf_1TbWHnB zno7_WCyAmd6`_U7(tXivq}MgElW?I>T#eO(z){lJnhtBNSAehJg$K)MY!d18`TFqv zd0H!$nq@mHW_OTAKgezV;)M(zi=)1Rd8vW z%$0tfz6PW9ou0|sNJ3KYPhOIWKo>A0NYqNN;ph>+)0UgM^Sz>y&>VH4UX)X8LZc7=iFm1zrH;VUN73NmqQA6wJlQS@{*58~O;48$ zT=X|~pM0_9A#Op#XUc+AL;n`=bpsfti#v)TsI06|LF}9#?tO`OxID}KWbhQuRkxj; z7v9OiB;!{Zdgw!2ISUbjF00&x_>e77nH!Kr38gch*alFWYKa}D1BM>Rr!S!nNV+Sq zMK`&h@6Ibj*xNn>n=w?(AaZD1t`D&2mK=(WaV;{>yE=RWeoZ3kMfD#z<6zI&#^HO8 zLm^OFwFSuhuQd`VD06j?N2N`OC+HEW^08=l&E&`5aMTEtL_PmP*IB8A#Gc`Rj(*7K zKVyHw;o?UF0?q0(9c7!q5m+*2>4IyqLDRjM@%pk!6}$r-l!#}lF~xU#(aLnr${>J> zg2KODr|MWW`LjaX;B-Wi5N=y2Vl-D`4=J4L-a60xe)lOaCZl_tEul)1TU@{qAA3U= z&at4--bdcj@c=3&G%>r3ZYVg^>jmBj$*gaX1$Tdo8G zm`Etl_jjg<58k6-FHD#2+4RL^{dH1e&|zW$%qJ6p815c;Z*RT+>BfsG$q2sD}9y?jjnFdwD36eK;Yx2SxmY$+1wMs;Tg{tb%x|iTR!Ut#~8t*{f*gZ1x zydpJSwA{`mjvOMe-W$O)R~CrKlceh*#?qMOBLDVfAl&y_bd#6w;tJvV`d49wHJ76C zg+ox)>G`*;NTF>J3R87M`fn;G4d_}>+nOuUnlW^)aj4ila*CG3E#%wsbF(t7Q}k&X=nLrl&y)$i zHPv+r6VgRA!=u#t#`1)lYHu@7qaaTtxxmqVHZF<`!Dd^|I$JXGCaJsU7>mDNkma2t zduGy7Nz_j4&~&xCozFrV44iee>a;u7c1kGtyykobw8P? zew7_|d?mp3^Vs}SR!j2GE#WmY`G62h-#QwMINN2GyfU(7wWF7ew)7*@ECMpJb~ee( zNSvZU*u;Myd3XVj(!{OM-c3zK6m}9d{9jN~{nn*i&3T(D8>HVceEwEaGnxF{S&f{iZAxrq+>$b-)U%Tzp3 z2i#ZN z;MPmG>MkI?a&k;TU2ju^5yW;l_aXNv)^P^!EPi*Q%Z&T^t3-eL0=*V0|J8Q^P~02x zKAB_0oUBCfmNP^am!v61Fj8?yt9mWJ-qgJdk)g%;SekwRRshIavskt0;rcyE29eG5hDuUVfJxl1HdKtRtPIHA z&=RO?RMD1bJ$#2e$5WrJv-o+w{c(jX;8@ljaHc(cH zE=kY&D_?!}rZRzC^b}_#_?#ZV=*)C5MEHTpz9?G!{gs;Zxb1j|#RbtUZ ztY!ug%a1#B)Y4g%QlEGK^^pc*OSN+K9}WgF8hodF65VJI8}WksY6 zWj8jWxoS4(==%NQ=B3D5I9ymnjm8NQYb?y^9|L4V(>S?+I}l+OvYVat6+L&G20nKa zqk(GHvGz$Us{OEtHj7+hmbNr|8%(gfWy#9D8^isnn3xAQXDG7=~@!eZu`{_Crf@_|^!&=M61?ssN zwZ7_&7X0y6Q04Dp(2(B;+S<@wN?M%U@zFDs46nmkO+98!Y7tc|9dQq2p_Fzsih3hu*Qk)JL_Pc|chx#PSN!lJa$ zN;-?>?nqEFoTX8ZZ3*RGt`anGInybgrqxL~f0R99z+0(ipN(Zy`&^y!xMuvU)W0j< zO7%f=$ywkiETe?tx=I&aUaj>rQsMfffKQwhM02{bGceA=B**O5Yc;ASb0!leUbp@- z()qm$QV$&AYm_n&x})_4+|vpb@Jylcr^zl+!0OA9$p4qFG7%#O$Yv=q*Pap;8Vn@_sppWz`NPdMg6fx$@Qq%hLrrI6VZooF#)18Vkd--AE-T z@X|w{mEYI5{iRZzi&hX<`{hXplI}3a-%ik)qkrJckPWBb{sil>K7$;-je7R~{ln?6 z%NIvPa0N^oF|6K|b@xPY2(WqM234QQ#X_ey!*52^<8k6l*11&RFVa}GO3c8oja{d^ zf%RUwYTqL58vr`RgK!>vPfa?E5=v73Inq3+(|oB&|0mWPq}3y(KaBTZ!n1=Slo*>s z>1v)FsaTZdt^w5&miQrSzpr5LjUe=*7kiUgDa4)~^5L;1u)kDPq({dQjvyq~zL!iv z{wt6W$#rs1P42tR6OOwBj@iK@eL??`m^5T1;;aowQE6>T_xGPS3-sDnnS_omB)k%#W~6k(ps zDPp)JV^^Y^JlNL+q7(LA z=x_1XWh4+yQv`pFX27gN!S+?L^+}3F`LGY1V=KIsh;YkRCj->l>(sS02pWNkE+{Ch z9o_3anodkdAdDsnb&b|r_23l8h8NDY2S%#8;bCKp zc`M9AK%)P#4dy;dRd>g)#W_bx81L~9nc8e4zaLV+AC7=W_qm^vo*$o7-|HtG)%Rsl zfjbvs+VLKB4d?;)E6F8}tvkrL6+QK6&D9tJqd%>$?7xoM4yn8H`H6Un0qZN5i{n}& zUWJ2fzZLRR2XKl`NLR5a{+b?#1P5&Fmh@f4`&zqsj^5l6Kutxj;i+v>AX+Agd^`2R+qv@B!p)7*OpK4NtIi zG|3cn;W;>S2u7rTB@%P1?5~E>a0lGIz?vCtSd?ohd+39E2dp-RVkTs=Ww-Q>UkBce z+h3a7k8SR1`x3V`Jl3F*f6akr1{xmIQ-{K`?d<{%QK8Jptmy0yEKTocgPQ>KV zG04JRW&Hh6?OkrZt6wxB(yCwzMH`sPYkgAQW}NYuOx+M$i`z|WpIM22)bUp!&>=t{ zs>Ua%vg>-#n%nnF_!wgAio#24X2QW4IUK^$@01H(knxt&4GZI0&+lt)+=v|yk`$(D zcmuzKM-FCy3ErSnBrv$wF=f_|k6V1_m>T3)y}$fJ7tTdZ{QLl``WEvAu%9TRAV*rc zO$Gkh&f=ssa6oS8FO98{J!9tmTLVQ?A^MnI-9H=BJ6Mtsqyj6K6kSr#%Vj=0Xd8mu zAVw)CXz{(d;oF#ib&8KgyFheE1_3%gKbuacbvwnSPWCC*9Gf%C`I@ZU+SEF;?Y1Cr zr|NEkfRuG4M5G{<7$?Pd4VK5v(}9GqNj}|<4dBpi;81C9zF5#f%iKrlfl(pNzXe-7 z?`QysHXk|8g>95;5l~V1UzOtZl_vJSA1}zSFyY#CkphR_2F>^hHsFaG4rypF48~Dt zD?J~(H=BjU$~`_$%rm1=h(FW9$&^H*qM$Qnz1E*=OKw?gD&6`^#F;h?VxYLfj>*K7 z21~75(RF-X{ZjjSI0SfVr?j|)kuJEXlTvO$w@I0`uJ`*OdP-dTaq&1GvFgyXluUyx zwOIfQ5#CZi&uSO&r7-^E(` zqzKB6ocC%`JZbay&u6R0fom&&S*SW(8Q3^FZP8D4XsJY7SuxB>G9J`X%u^qM3o3IO zCVHn)IEu1p7e4^^bpmbUybapi&^#^m?fsAi`xq-8C;2Ypg}p9_=wAy2k>y4|4U#Vb zf-|k6ccQ1#%bSf>`yE}-b!`KRM$h+BV`2p4f0KY}%)=jMQl|W1rIVcU9wB}hOt78r zosL<@ zE615SpClLUPFi=u5tHu*3_DxcGiJZE(Q%_-@n+1`X&EhGbB^2LQhU0j44N`yc78~{ zoqOiAqY#&`y@gRd=4~!(ta|6e=vohwxVpM*0E~~~2yr}8YT+GMm9%!qftUuC2_f)> z)~zm$?`(OD#R(xOVXd~^k)qO!rr(Qq`Tu80l>y<~0f_W{T_gS*0J>&70@BPuJ7D8u z=#0x!`@snadW4*n8i?pKhGF%P;F~)9g6AabsJ%Sd9+-ijy)18yB-FChmzW=`K5b8R zzE-uLM*vZCfqNp;FnWD502I>?%M(TUEnyXESeMVB2yZ=EZig0r6Cw`9ar* znp4(05F6i;CLMAHhb(7NtJ~%jLrI+2-Fwq@*)u8+U7DXJdOUl5HEpmRd)vfnWshA&0*+h zxlf=d*RQabogGry;M@ZlkZ2{Z5M!HctX<+9;5_11q3%N4(z05#!ok4 z0ilaf3&Xq|$MlSUR{YW?9wO9Hvmp)j;86;smZr_v0# zfa38@mjFct^FI?}Z-EqpiDMt~aVMLg(TW3a~ z5fhfR2IuIuW^n9%f|I0i*VF3vA7=8mEC|A zv;!KpAutDjI~<@s(AJRuf-8;B?^ST@b#OzA%1?r*m|w+Ny3&#}@b$pLi-4hE2D!ztG97!Tk_DIJmiX-dyS!r%); zpx)$1G%y5d&)3>FxRT2^0^XMid5`!zo6#^^7 zCs_uJt696k;tR*jtm^~0_t;+2_x6Wy<{jG4onso5r-6{HSGmdo-Ay5#&+@W;Uh;UZ zgftzsyv!!HSR+vz;m>ykL)rxvvWu&+0MVnEj=ax8k|z za|po=__~t&Q%M7p(OCx&hzaDUQV4Ogg@hufOX{1pw@y{=IzIAxO>d9cd{A8LrC;>i zjP9Nvr@qkPtW6+J^X*Z;VB)4UeRicQyiJ|#4Lw5GkLdf3Hnwd&W6i}8^V@GvbMFEV zKz*qr3Y>p~Z?5?;axR)58?sJ3(PbI7SYw*`gv8DeZ-NjwXg6>KczRz(*fw5u7 zp1>Z<0z?Ra*{e7e3^ei^0kW>N!0Ot~tK+W1oMcFomVV00pDw${)Z7idt=Nm;sxEcC z*QMvdq!Vktz^<_W~O9kcvX=G%$H$=mlClIT-^tdAue!1IlzY z3GC8a!3b=VKHHGogZYW~0itQ&_b@Z?{Y#nI>fd-7KcV_kjsurYd!>M}#|r$r8P3k- zRwHO-deejzptc~{%s=)-mKwpx7s1(P&SMfsSQin7f#Brjtx_0)CgpJqxbZs@W2D8c zQm|yC{Gck?HdiYSMpJbJ(bCBsMRuW0mQ}V8qEBebk3k!&ElUS%-w@#EDf0DT!s}cS z#ULe!;Y&a;0>%DP7?QfO)-by9b4p7IA~VzGlER4fLfmK6txXX~;0I(uxic!-;OH2< zkf$$b9F;{~3iXK)q~6AJy%Bfd1SpRvXbbJ}P3m6%5VbYEuXy)h1y&p0L`p<%85Eu) z@)&A@gu?tgTMy$adlnjhMQd!>5`Ysj(^F>nT_v1m<}#(~;4v9bN^~SfZshi^do0S) zy;mbp@^jnrHx~kO z3G0>j(hsv@mvDTKAJ5N4FSbLRws;HX(PUQ4k?+;(S8UP0c0s3nzi9Lzha2X-iZ;y; zxeS5Z@0KgqsyGxY0fy)XtG?ovR+r7w5sCaLTRz9!!&0}^>Fc>aM?>73!`JW-{5Do2 z4l-%1s?6=d0pQ?be=x!UdF#9onKq21ee#B(3J2HY_wSutklA@4!oHabVU!rylU(JT zT{Vny%iO<6QIavInq3m8qGuv`Ss=985?mDfBp-6o$xlFZWt9BIP5-%7Mxn$%XO&EW z6K`p}qkp-m8N8N*6654{d#1F(mx0f})r1j6C6&-P;(LybyBnxocWlK^ zAag6^(y%+8kv!i0`y)$JPo$t9!IrXK&6otOFt1FknL5bPE3uZjt zHKaSvlWdK90I6l(jh)z~&dp&HiCnj+1@a0#MvY02KUPm%ci^FGSEk;xo%%DyfQ~+N zY}o+hhZG}YsB?a6X2gC7r*O5`s3#F;1?GlacR!92&-LOhBuv|h0%wDy*H=9ld?8b0l;`OU%abE5VQnh}$0 zF8UPA#3ZapAm#ZBQW;MmK?9=NL4imgPg3zmx6-7qI2BcJ<5Gk#{-@6CJubO8EG8IR zJ=G}!7x;G*wph-G<-POsH7!*(`H9K!@{86LtmEb!rg^>SK1k9{I_eXI&E39r*i2WF zQyjf?H2f7B(Hb&Qb`cu4612bKa_b@lOm4xQb-5;e)8i1sugMpjN$D9!tsrUckS@L1 zNC2{$B!D~y?f|shwDDwS%<(5UME)*UNJhPPn0bF~OsWDNXR4$~|BIGKp7rLCkd7x+fo(2aJ`wt-^1o#DU6);zV_;B?C z6R@aNEM>qwBp3@PppT60clw~j%DwlPD%H>%eOKimuW7~dr!~$^rY=H= zbrL*)(w(Qk;ARpbCUefk_~>{~3pdpTkrvjZ%P=S|1p@6O>uAwc=){7R5O?o7^Qa=07P3NAG1tenPj;(DoQk;^P-o2~VygbeEkmZmB~zPr zoDTR!`OOEg<4`+t5Hg@T98_A?vIdiRfBe~c2#Mk=W8x@5J}~`Y*1;$;ed02c&OpAF z*oX3o(x-87Jum#;fFw54DPv%IEQ(N2wBS$XRcYouhBRs3l#%`@7%bj39G^aA5ctYh z4Ldf`ioTB4IJr6HP&w)Q`=-O6yh-QpS;@-@<{!1&PCQqe%kAyY&uh3BMhZy|8X}^ChU{ z@`tO8y82C++8?>ZoJf>dZRk<#sEL4I9iu29vi*Qs+ukhvG@Vz4M6=Vpp81unbZLM4 z+0;{pVGcRc=?%KeV|uNqO=m*7zYrgOkl~>Whf}c>x-*v$$>&Qjt?tD)!3ffIQ^AcL zkKOuvuTrZUNqlOG*i9ZQ<9{JDjx~(umHdV~RN_j`L?00sVJ)#0dWbGeTYs6U(#T<}?C(9fUC%A~4b=lk7TB|o9X6e6 z){rKA-rb2_kLZ(Z3q=}q#CH#%xd%QZRapM`-djsQGMG7NWlZ`D>=Qs_aJRm9@DP$A zw`BTh=e*$)1yXykGK4s;(f1(`tac0<%qsUH(^on{R?A^tBxg)|A zzc;h(PH?!CjC(<3`*w}@6ksuv1K)y=Ddk6*h=WQm{n#HU3a2*65 z-7#IW_fT4k6D})oDFl)PW2KrKL0L1X(ggLf44C-yi@#&$X&^z8Aqs%mGtQScAY%Ng5tzM`T}*Ut#8Vub(_74(wq`7zN4 z(G6%T1VhfDkviOMh`FV)UN^rnQ*$+7OPnLB6@e?G!F?i}^4Y4i3`SPjUc6&$!HN6c zH@6;vu{b+KU+%r97pm`!+BkiELcSlO+nM|I`hFO4Wogxj+f3Cs za^+Z2;|cPlYUMiDo=;l+B^jPS`=Wf=Z8V`K|4_8E7-dT^$!fXanCAQ&f1K%KPRaR? zqzdxR?3X?LD{w>CYlouma=XgKuxhd}TU5{b`)K1jh0Yn6W0Pf#9(E9W|G9oAdL%NR z`>Phwm&ZS0YyKbHi-WA&q=QnuoO&yI4*1^-3s<{U2sH%!qzIHCQOwp4(WbfFxA)Nk z^$J@u->>WSp~B8~>sfHY`Fwn|FZF&HSKD@47D|70VUtZU7;EZcT7Tw+H|4{Cgx z@1vn_a%5$xwg38!d#l^01gU8XY_=8KW3AaX<)ZyA?r2cuu!e_WvL2R!(MH2H+Y)S{ z&h|9o$YC&IK2!I~8soK;XlssRM2^>4y-}0d+Ei$=&&}|DmM#i;ZN)Bl9a&17wcOfP zd7oJ_+2m-Y1cgoxm2|**7{;2t3=Fh?dKHoswiBBPRP4f4YyvuP52NZoI5+UlgqPcb zmGi=Fh(^FgM>vbqdO#p(J!earBqdm%)@j{bexY-A=BO2@$8lVk>9)+Dqk>Q5*GTqh z;36)WPErS-TgAAt&H ze5Z`SKQ+T&i~r)4v9p18v94|i;Hx_zyFCksD-!3X^@-D=y{0E(14&EG%qo9l7ePH!X#-Llw) zPm@w)*41s1gd>F+>Z?UJC7&S+XV=n+6rEpZOOjL8*&E(<)`^K_|M$&WAxKboZmAD5 zsLQPI8NqQy9~%4R941(c4my`8F6A>#IO$XwaCmE0DNLl8tC@E9%3CULg(fewgn#|y zHtNLz%Q$4C>%RuRfV&h$Q$+*Yl})`g`Isub(yj;6Y+7N8DDhDby6Um$wblGtIglIl z(=Zv!K!J07p}Z_8pT1m-4S6Ahn0y6?dQ;R^gGlm=R(8`%kn*!ISWiQG+Y>W24g)}` z&M%ROAZ6P1tQb^HTm~S%akW+*qQkv%_VC*5y`S7IGCmyWR9p1Tpr8(#DDxdrdue{7 z=|sKcg@OpqK^V4n zp^O@nhW+4l-nhY&;B>TpmMbUy^iuxn?Q7Lzm9q3!Lz8F2aZ0O-1-9-tanKQvY0X`fJWxVnpUVn$TyFPRx6K{W>YOyi@m>Yv5Gk3{I2ZN-wdnVd*b0 z2;74z*#o9R@a^)WS^n?`%SI#DbaeRUmS^3C*KzA3#RQsn-CK|K-KYKYSdS#sAi06t zm>tSt8x-(BK0gfbbENR@5uTNSmp-k4cv49KJRhcjv_cCT7J5q3GU2O7-tcQ(YaDAB z2mVZ9SG%++E~H2^M7ekjVFT7zNrCkjC5eGr5|mFk>aCl|RdVGqiXF2;tF#TH7gatO&i8c2*Dp zL17&8HNSC!^i-Sl|ALkIs;_2!6ymO#km*^@-A=C2rqu<@d8DDkn~V1YHUS{rUB4253YI1l%Iv#ZwlE7gwOae?%en&-b-m z2uI2rbsHS4HtPjcf*76!qU`H3NmVdUidEZ|eDex zv~u+Gsx^yk8&nfmXtvu%8F6Q5bcbzTUY_0|t%5hkE?*Q5-pH2dyoXjTM*@7I9=?>i zbDAcJ;(`X{&k9U>0(u+CD)2u2Q0v*{ob|~W#_f$Ih@?J141_R3?kEG7eK$(e8|@!)!S}WtN_}wTqBz!5Whbojt_Fju-Cw z9`gd^0P`J&0fk75W4Q8f3Z8^bUHNyB4bw_=O)~&R=k(t!A!xHat7vhec`lbPQl~48 zEu3xPN>fFn=fVPae9+~-ldGp^Bvc*Pw?&{TQqq2DI^WW)4!>P3mPaJcdYUR;e~vc8HRVDc2LM zNbc?(-v!LlBk?RONe~hPr{}-NUag0C&x=`e#1({M=cpZ7hlF)l*&iPlZg2`lmb#5k z=}Hr#KUW!a3#*l4;R$GT;w(?mE1~X+w`|a3-VVs~pOl#MAJY*o#C{l{egwkv7nloo zqD;%P5RZ^pP2+RGAt)0HSUj&lW|6c!P2nuV>^X|eU@OmKcqF&`EQ$|3jYfi^P;wD~ zWuN6{H?`;R{zs$$LE z?}32IL9la+Co@mikFf}k~U)eSvY z7{)9(H=)!iy^o6N_=l-5C@`0Y4)V6-nXOdKQ*^;fL)-62T&Q~h7Xb?3-}_3wQKB%k zH;P>*N2O1AM%1H$ryrkM#@JC|9>!-s__CB>FE;02&|otH2VbL|a#Frl6mN``MFsL; zjSV1P421?-;_oANfg$#6!$1XznVC6{bUfWy9$HruM<_UAMHe#mPNrNu-GuK|cck}g zOwFI5UgK2y(e!|NzTZ&wuglwgYlbL2u!B7bEi}X7fqnNXN>ifE-lR&q+z*j+j`y_7 z4_v9A#%}2qCfDrh`rW55XP!ou#b6=mMd3A8>lUQT7MG2q4%;3(rJ9bzTGEZtYZ}kG z?zp~&i&VA@A(x?lA|j5f{%4~$W`zsURsr!wzU7`ru_^P%`o5QRm)F*(g$y*eB1_#* zK4H=j8zy$^3$&~sU4K#LXFzTeOj0#x5f=~8PSyT}X_f{k(k#kQ5aqczgqX z6OHkRPx0HPOF3Wm4@aNmaR7Iy6sDiUzaO@yds!B(M&A$8-y$F|Qn0zR^IDjLmtyfb z(>5|6nrVqdqL$GaDMtqZKy%i*Q)pqQmP&H|*GTH6u9(LR<|D)y*MdO$3==oBny5|O zkqfH1dKi74xKphnqm5QSr>rHT=i(91luY8UpnOySSVEEFg7$=ZY%Y!Xo1uh_d`jVx z9fPipOsXgU*P!^BIJnACtJ)TLJVG-uQ#p@rYGMLCqn_Xir~Jw3YdVT8Wg;k!exBCa ze8JS|j&k5Zscq`^2FYznGk-}Fq2V8c-tXCqv>$@!#6779aW3J@^ocdYtle^Sm$75b zYNSWaR%*qPV%X;G#zw2M0oLjlVbz)Y0Hz z<49K)b$Zd|1nK1qH5rSUJ3)z(5r*Hoem?kvX5J7f)=}}*%mjz_OTPIT$A8vyEOg=e z1y+#xWH`ZMGDAGQA`Lp@31l5gY|s&W1TSTQd=`en9%MFw8>#I@r7%B6fRZ}?+3ZBB zh3pap`ugoUt@XaN`BGt!_hy`lU!92g%(gHD5hET#&?& z65hKY)MBJ3{DJ_cYQfh!K5$*ZCF1iERs`puK;$N;2|;iP6V4&ob15YpVPu>%S6iNi zWEcH2I3PVO!saiAxW}#=7;q!9=t5Tgr1bbEPF6$7&<>Sg5egv9kkyQZO4G&2x&Wt( z5>0>lJig50yItG@kt)TS{u=5o8|P)m8XVw}Xncm-AS`H{;AXZ2o&l--#V7ZsGqJq% zOhG28Kje(+y`;r`cMU4pZo{!@SQjj`ifpGK^tKvWiD~qA+dy!L1rw{sbhqvXWQTZ- zXEOB?B>PSE2GSYDw|2^9iuuuy!kCG@huR^vgoSlPPrg;>`tLxae27J5W45{I^oI$C z0jTgBF&W;y#DveXKqf1;TEL7OpoRycBG^WwhRm$fP>IF#^5r7dr-B8IC;*IWFSd3h zwjGWEiz~%EeJm=KE}By0OEi?^4)-o|-&j)4^rV;m(tE91Dz2-*@FS5yoI8jUvM;xW zxslZcYK%i~Ca)*p;7d`Ni}B*1MK_IoI|jyO1K=U<{ZHDbpcFZa#jdLZBiqYlamtcNwr zRo&Fawoh1h;dacqC(0_HroyJvw8%qy2V;VAfzdC`2U&;Fhy9e?=)%*yH%Rt&VWH$4 zP|a>8yKW%NSwN@u?lP7M9x7L)*+K$UqCS_`Aqz&x=4ALDR18j3CYY3R{~0$diiLhOyu zZ~hh;?4#1Wn1<%0h*y-?6;(8{`uhX$TV|+%t3RfjPMgW*2{vcR%139po+#E4!t+Pk z(h;iS@e7xixcO|#Vx+&&-D}Kw@ie}FOR%aU2UDJ7DW#2&Dndz%sw1hhU1T9C(QR+# z!#t2M$SToS+&3;{K!?~r?cSz4wsHzu09pME zvNg2{K7sih0RIpP+_V_M5_Ev3*$*yMQ|e>OFh&t=|B8d!boG?CxeO5q)vhDTqA()$;*bk@mmR@Bx!=al z)t1x7NNEPZ0dC@Bee(s(MMwuX{r!~nr+TWY>3bLL2kjOBT=wjPz6go;2tOtwQoy(b zG8;BI4_o<`P*ryzspW9soBFjnx8GUQ=VUsfu7K9{74OKx*jGUi*|UaE3;W_{f-|J@ z^ArG!@J9rEvF@tF_T%9FakEnA#;)Pr{HqJb1?SF66Ph7*-}}xzI%u%lV@K$2BU$zM zgQ3%iQaz9l*Jk!xh6!G>HT$raJ3bngd+L6%zZd9HEtUO0S>3q*0}kWnX8rFttXgN% z^&bw~cAj0H3hW%uBn=#+jNec~d1kSiC$oLU17I2=lNhZA!K+HE|2X~((w#_{f{Yak ztuH>@b7LrGj7jp@ZH3Ey9}>+T3IF;T_B);I{@ioDCG<-4;}8FtN`suOd5L9c61dB7 zo50ugS%Wng3tF(4pmkB?%Cq{7Sd;2?Fg=`wX&-B$(R)xt@-8u1MW+{VTVwY!)Ho-+ zSsXT5Mel^~`AL$QLBnF;NnT^&*X#LV*hWHItXALHR`_q$pqb{lzhLG;*K?Ad(kDtC za{RV8ymuO*dF7z6ClCKYtVGjo+I-vhB=n!xt(&W-$L=z&aS*R|L-trq#r7g=Ztz3d z`$(MUQheKW+Vk0$gEv`%hdlOn(;PW0u_XRI>0bPJ;x6}B&5Q5apww|crPR+dfom3c zfOZzQU)@$?`?FR>m;r`YTHJ+{2&rZ?YZA|bhq|W%&tLIR64=9U0p2DowzLqg!ODB0 zYK--(Irl3KX-926gCE#htN~1YRY2DLaocU!_2*I*<}^)|b|Q~nNPk;8|HYG5qd%84 z3^y6Xy`}|`B<;=Sv#kz&+Vlm-XZk6FoFzW}e32w;?Rg3qV~1xW_TRq?b3m(6p6kz1 zQsM735mXRbZRFiBuwovE4$H_dRzlxKF4TMwe}oD{+GFiYX)`fYCSd-^ueY{lu? zK~(4n-i~|c)@7@{MDiVATO~N#Fu>B};V#$^MW)+pV5XLoC9P;{z@z&Lg2Bj8mh`gf zy0zqip*qo%tTcI3;zp$l4SSU)p9<0TF%M0>1N-Lpta5G5AtX}R`&~3G20G}MY)Vx- zOwbq?9MFggQIZb~JwZoCHvsn0uE{A+Op$&vq)6$kni_6mLl&zfRB|UvLrw&Qug9-O zd~c$uA6O`<+|o8)#5^VyRF4>=qG^e)qHe~(I8H?26tu5Mc}hMi$0VTwc+iyG%1~;N=FvX+{n354Q5XZ=o5=$ zE0vQB1CBDHBmVxnwm}w{% zglZP*>dUm-DtgA|8WGP%5;H@$=tWf(L;R8+Cx6{DokoUEeT$FP!Nmh>S>o%PpD@KV z9lNo$|6pfPvTKG+i!!TGChHNKZ>++gv2>|Zm$OGitvRoZXb9Gho{KG4QK2Gn7)kC` zn_e%>aqyzu_M$oCkt@Y``q%Th&z+|#z2PwyCYar>llV=+-!<&K z0c|(>_mGnA>h@Xs+F0^Sn}x~32DDi8jsv=-eYY)+%ntIQqM)wNf%am|5rVe1N@D=r zs+@hJEOkyU-;Ln>K8vO`zy0Wu2h{3`>6UBfC{syP|9*DF$%W}>J)4?AU1pIr@%#yZ zN3_KcM80-x$G8#3v0H5N5dLL@Z>!_|Mt_ie(>)LwhF}WdzJin#jCmOP+LU9T5#XdG z-PdKIO|dKfF!L|doVMTX^_;6jj)F`ewMMZI`=&aY>i*##(c7&;v5sABYa-bN z+N0TTDES#$!L0)~t|127Fk={^;@*tGTNaUI0mjaV|9gqA^|<0_@qHs;bDY$$^x<%lS8`Thq& zhsBN1MoR!&WW#fT7>B4V1q}qcgSxYvY|=L;<@YNR0ZQWi*M=X)D2YQ7@n$zSH}@GX zz6wKKs$As|=JPAicqcl)|IfGl~H97cch|mJz0L1inEZ?HH1HEm8<5 zo2bJYdvtR(M)*tc53_$>cFkH~`v^YNI;~gupXeM0=dfA6>+7j=k!5E+(LmNr9E^8b zsFR*XwiK}6Pb%xa^xkHn@rba@{>0v@>7eY55%nH^Li)wL@9frU2TtwLSIWOYwR}NU z+0m{VWlHhEU}@l+4in#517~ZC@J0R9{lCFvS@R)Cc3k8OeNtq)Lzxm3#Q(C}hCWiu z)HVg`?$nU4qsTevThU>KD_cDhZdb4l%rHvbtoeaW2SU42qF34%VCp}ANHX#Ao|Jpe1!{BkGEq-jLa z<%+BW4<@6+6%<2>!0HrX=WOtc>Ii{nEEQI@mU?$%neFLyOAs|97Yj`d4B=S4-ZFFE z;pgXR2WPRVX?XMUWUG4g=>FDF7Y7`-@0Z#$8Gw^(O}4SRpZCAnD4^%ni<4(l7xol+ zs0rkRQ=|SpbL+?$UIr=bL-6+IY0GDs_E#*Ufg9L=xsa8$H}pXO3_=@%;3bI!Bb(dj zJV?V$f3l9ggV`1YT9$F@w){Qd+ojD?8>(>=Bd`W}H0<7ZaP6|qL~tOC!pxHi2>TRP z@MrWZ(1FLBi1%f<5SaMaEz!R1J9|e*ZX|8Yp!Chqh&s5sYTk&pT`l$w0d3`DRr~Sm z_2=#F@@&n96bYuJwg*tps^1LBzyLN!xjSyMaO=;4R&Zh@uqQ#|KS{2&Q?TZCAczWD zcV+>51!`qR33BxVH}$U_Xf?D5|CAYjkr&=t>eC6-zafa(Zoh!|iIA3VMvc$9@-i^F zT$iRr2&}QmTaA&&cwaz7%z4-FgHYjy$(hbE6CRJ{lw;~**Y-8#_!c1usQ(_usTZ^?%Fdjxq8qFhGc zk7f$1ju@Axu1JUEs*GjWVgp%+9g;(CY?mjZsI&7EseX-WcC{70M`Lgr3nG{<0>5|Z zFbX)uH2my=pSo(JCSwTRjBFXiS_C3gAiyDiuvh#ar47^W;2e9DPvT3q^B^ z-))h~URTp9N}};)F`CKPawSfbO^Wxs1V00A!2Qr~q1bo!U=vNB_6ij9FaGE;QY0)k zjPRl3qYQ^*f5kUlZZ6M$(vj06K&iEkov!+nq;s=u_+> z*Ik^`BOHm!4Ju*R=w+OEhqmup>;gBXKD%^OmJak>p;?g|^(O==_*`6L>Lsil}v+$ds-TECcE~a zPk*l9%LT7gS$Gm18D&v3hR^*X)~MOoSiBpq-*I1h9;nxP>Iulqwy(jIb%*@*X~JjP zf!C9;e^ljCwc4yTg^aZ#0uJ5mi3cBr0#c<8Uj&l1SPmUQU6M|wsR*Blfy$I|ZE8bK zty|q96k5c5KY@h4^PGiZxdCXmUeQ&iG;8)oQ6#a&33>KWbk%*xeZVsN^A9o=8p3C{ zZdI~2dz$#mA4-D}gEzXq62rBnO6`F04zG)m@uSM47ZPbTO{9+kqW(+(+>5Vox7BJ%jHltoj`dVER~PEH7-Uz3gN2ly;?NMNH`UaimOZLXY+gs{m zG+`<>uPn6zwHTz;y$O`lLw^XT)SIAX%s2nwUHC^74|R-T4|kv)tib1WVTgzjuc@QC+RbeNhtPenK+TdR5h|P9>3fh zP<76DfWg(t&;3XU5%PaM_Zia1rBedgMFCZv3nX4lmO-x2%XyCvdsnWj4zxK)NY<{N zn3sw;LV`*2pWTpfSCmv&w+wt0Zgk92Fc^~u=y){*uVW+E5v(5p}RI{B3q&*e0jO{>s$~)AH~A=ZIuSCJkQ8%hfjxs-(3m@W&q}X0G7BpKkLMq)5d*3XY2F8 zaPl^=(C_+ueuy69M80(RJ_Xr|UD@hYyRhrR_YInmaOFInJE?6@UB%HZm1DI|+gv|` z9|astoKP;{{w{M-p3|#1JR~tA9{T(>%^d14Z6-H zlDSo0{8_H80rj_@?h|u`xQXe#i%CUma>AE1928Wd4 zl-_)_H+t4+<$sVm3SR~r@=fAay87&PYNy%fV-L+--oqZ)B?y8AmLbpuZE>0!eR&Z^ zY_XWTe=}<+u5|h9joV0Ir`K8G=H|sV`!LPrZBPLl;Sah8ig(Rw^)Y+$TonXG+B^pi z$01=S3aOuU=0^BGZiTwQz`7eT=)w-$Lxl0P!x#tGe?DpH8)7 ze#W-##vRO|xt+P6>yF4U8iWK-EZqV?I1&5N7&BPMY*@OVvc@m^2<#M*+Y4p#ZC@E# zYG_1Pov{Wp%jd)gVMYAgVqfGCaWrqt)1BOfPuT?;rlakegP1iDC)u|Sp>&!C=RpKUll!pp=))Xa1^DF=cNO3af71LuKm2-n5}{7^Rn7X zaQ{yhKL^_`x5ILOF3fuXX7nM4d@T+Sirkt&krQ>sHl z+$v2n#^FS&e4=qa*mJIHZ3eGXCBeD>jrHM4K+r%GwG3DujV>*ce}eT(W(I`4z;_e%RpN zXYNy+vp&cJK(E8TkTv*QN~#=MB~d3?r17oA>gn9tKa@13XrWWQCXlVqvn*(1lWmE( zIJ?ntJhdin9VZWrF3RT?k-zef26UjH?%0Or*jw~DYm4j82_p_1?b}5f;qy0Q$_*X* zf5`HEGtznrObYnAchmt zsD|~NJ?@_}il0*ReuG{3`-jxX1_}R_)$!{Ts4t&Nl6i4g6FyjL5uNpVzl8RuS~lA%pXMxgiV)mql$m&KUaln7tE5!b|52C|RY)O=_D!LCQ0H{M5rldO*EB>U{1*jTWCHVaWqn(H(R4( zl;;%O&ALiC^;>nJX#n)DeY5FPjKNv4$|Imn&X6n#e}??n=wbHm;ay~*8bEKroXLB& z)^2~xB8!{`kC;-Q@ORiFv!c5o^_2E$|I@93ekJqOiXn74&XD{MK@XF4=vn;PxmUR5 zpqd_^?Kz_9AH$M6^iR=(B8b>c$&fXZffflZOUhrAv3@eLx`cANA+ZBhV=$jr0=(T^ zyzQ^K?gO7q4_K2(=WEb-QvoP4p> zdJXb7R!Oy5+f)z-JlJ;1zx^uKuNbP=0|%%``>3mf!MhC53gPgdgj6ki(eTJ*lNB}Y z;rvoU(?)A9M{Bkxp@fKNfb+CkNm4WMs_4H}bQh*aB?7_K85*!z{bs&n1#5hs8U6y9hP5swH@7S zwgZqdRq--=ttE&^haa`9f!{|~;(RG5Ku@K;8rV$o6E#kAZqnta+GDMH0MRsybO$J;hact&4{d(2E*%lJY2|1RMb3bT=q6XAo44 zlhxX?`4BiT#BL~{iN4`4GvA%Fo!MZeZW|59rw}U_s81Ov^y8#Hc_7#r;r1g3^^fY~ z>zwiAK*9J#Ak4|~$?vfJrMQ!a2@o{?aAb5k`9<4?WPVRiZA=^9%sb^ehLt0ZO^LO% zw58nf<|^}yN#mGRHGEYTlbEe=P$L_tnBEJTer4|=I|Y}rL+xl<9SM$Y5coGyMrN>T zYP#MHx2f2OaN`IIyIu4p^Dy)PTXJk|KmqPAz*K<-`ad|@dKh%DvE9K%?@RNhP3dl# za#FxUFor#}`PD1Adr)U>T%{jZwyHBrUzTQh5FeO89T0{0{IiZ-?T;<}zDo?&+LsSU0HEF6&45mQ<9aC6ND zkTZ?Q>=RLwJPM!nnj;(zfwG#7h7K@}D_gB?d26B8l-(aOWV)P#CGUJVx3Yh@4OALU zJtD;*ag%}3oGtV#|E}vt?jJ+YREYf8C!}yHTf}p zaG1S)4`?fsGlLe4c}7_O=Xf(co(jjE2NYo^-QiDR|L`I zibvdyLc-A@n4t~mtX)~}NW}m|KrQ(8twgz=#BtI^kell3X5vlUH_PKr5C>h0f+Q)V z7)}nXJ2xsR^zQ6@LZk8F!Kr^(=S?Ii4( zOndMijZmY`>_8H!Zp>6Vh14n3?(T#4T#Bfe)}%x|4QS#)F>C7RKsp^n=hLVkHNG6v z)CG43UoT@cC}~}upYuM$WI6^~;(uiTb<(NV5%bZW8hrMW)?JfN+co_?o+5XFLB{iW zOLMCGThLijA+{2HAOqy0lD4I5V1>BvJ>AbRFGrc|pjzWq9z_O;g_@>1Rs0voYTTC!4*&d1bgk(WBK(IR zuqvBJg41>%aC5NjH* zXq0hcL*78{J_C(N%Z@-iiL@K;hGeHnL^B2~j>L^wL1jr?=*`e)jEmC?76&q_PQU!8 zK*FFFuRI3}leV5D|P%;Z%F!NoIJ<9-lme+3ATo=pQWQDB) zxT5?iv9+)d=h9M4Vfjr=_YMK<=JdrRk?J3s5~BpEwoEP7AonSD_&bHytiHU@Leliz zmTtmD&Lyc&fQ6;_ey5%s4la|UmsiR|D3LgTOq3>!4hk_8-2wpIyP-7s= zRD{Tpd!!so;?M)fm3qc*Ci@4S1ya2?j=OSlFLzB?e4FD)|rC*VSnM_6q%KD+b=Xb zu}d-Zs)P{fy(eE4l7tjUXv!%)P3QTQGK$W$N4F1b1y$!GO{S<^)b?e(!(6gEKd}of zeR^Cw3ofA4jcanc1)9Es@;|LZUCrT1O7H^3(YQJ+m!910>+BJ@DqHgFcY*L1m9kYX zM$0sU68O|2p#;bq*7agIMLRF1_!YB{WDJtyPiM7TOv(3!WW^S2G`AyDP}fs8wqhZ% zw|}>;ah5<$HgzwKrc3*NsU&lU*HJL&(N;Kfj?Q^_MMK2gTs_EQ*S{Yh&W^ zH;BH8GHjI6j5n&uc&>;5DZ8LwrB-I%M(;p8%p1TM?Gq3$v&q7Isy(|q(H!=QaQXp1 zv|qUL>h;gfc@#C}PQ=!)qVe|a#f>%S9KjWU2ajv`K9eYXUxb_Nl~4G{a+2PD?oy8? zSx9O(4{zz<8C9t14Se-Qo(IM;Cdm6ceP#d#r}t@TvHsoBSyF$9VwGp z7?gbS<|Tj=81FoT9cjbZ36}MkVG!t9OYSgb?J43B!m5H!qo&@f-z{Mvi|W8>fZ2i3tnFJU@b z7o;x&aJFTJp;JJ&$APCo-B%Ypn#wW6(_II=^}3feo6n9=tG z#&WS4U#|fY)9{-gx_gvRkD&B0D4Dr9ik+4M8&v>x)Y4aV#g2u<2K9Uaza<)y=fCj@0m?B*&JdBQ)Su9bmIlaep_IVK9NS}7Vuslwt?X*J#b=OW)Cwa=t56y(sSdDWMA2C z_p6x5H5vIf(jo+0V4sV{o3*@ zXaSqjn|tUD<8X1rNp;EB_N$`uGq?qy+NO!f9U#-M_5@{rP;}umWa=3#P@0~c3R}cn zr}`c=MjP(`Z#g+b>2&AZL}T9p))(;;+n0E}9MTSi8iyv|_x%<5@3jVaTAcLL8b)4m zAWzBDi_xH2TB)^;S8V(4g*~e5u(PaT>vn%NpK6RGnNX;k!#vaAE!ltI@U&Y(q> zd=i4+?%Kc5>DxQw2?9qdu5lc22*s$~Z5uRIOt&!WT9(`a*ZdY$S9Y2yzkNAObI|A> zm>im}%)qSPQ7tR5?2cUms00;yTT~n^+~_M?^q9VH&<^=di!k4b&eY{QSktY3ge?Ae zNH~yx7Gxz$dCbWT@{DO)DD__)dK*fz9(@N&`t}?;k!5xd1)^M_CaoE-YYjXJY>iFw z|8l^QwURT6>ePWPRuUDItP--4loZ5zXsi7JrdDbPT_KRAaw-_kcF=lds*OWL{9Lrf z4@RBH6KTb0wrT%(i!bNAYK}IymSp&D+lg(?*BGfm`Fk5^=Aki{v^kPqJOn9@PH!JF z=k3ys{95AMSGTsSYqU^xsFPw zJrQd~a!$&MfN=TZwRCYCT@pLdHVpRF}WELS}lrgn4cd;O1=4Sg}R8L(~ zyls?Eu!rLg96r%C3?2C|OL*iMtBhZ9M)=QTUSL83}yQuOR;4Hy4eFAn$u=;4sTZPBoFP zKd5v$6qvJ0c##%7`-~**KOPDJ|7J2eHHHa6Ee+Ayy8Z{QTgn?{~VF-`O-|jf9xBF<_2lRbd9$Kx>5^n(^S?0 zjc^NjFBLC=oWFmCq^#-xCo4J@WvD!Lu$F}%n(;H?6!bpEH3$FGTX_pMjbT>dI$T;P1g_ZJ}z*SQ>WQyTPSeCA(bUn*{*K7syAcmOqN$T zqC_}4Kw{7H2SG+mFxub6mt#2>E*e-%W0s&cc*-_Surt;BAWJJ8@iuMd!~{>HRk$;v zvaOWN{EjxMUGg1m^cJJ{S7*Xdrh*Vy!?NZ;kyRa;BbHQNaiNrHR(TYS0d85dTgrFo zu%rVb1Z6Pm0H!<$&b_<6s|S4_1J7JsJ9S~RoNj7yU+FA8NN zw6T+)jobV3i&41SrE1S7#nGu_MlV>(!V_TYr2XeS_*bKO$1%KYihJkR&SehfC%!o} z>kOmB{uFHX!>u)!r@P16!Oa6GfP|tZ-ADqJi6MNuB z_f+UX(Vf*LntMBNZPvzu!jCd$$p3hWr_xFT*!Tn@s4{(IN6(NY+}(fc;)OFpaE1>z z;S{!_O?4(1xLwnj2|aSy5QaR)m_9kC%Gm-Z5BRFhm`mJKk}mioH+%5nX5V|ho0tPc zWRzCwtZ<#g+b}znX0K63*431IbkM)-EIp?!t20$25m82z!Vuyq^77`3OBcc7!+Hi| zyz98OgK!oK8`3^n7H%qru?L*$(wvUoD^!zuCB*`e4~5UyL})H64_<&uiz4_olEQJj zr`x%v(OaVsxMGjAGtFjlhyq_|4bDCv+c-lfGu7qlD_#*Yu{WBX)tF(#=7wIDPoa}( zmoDfPs&}m0#_QmjZaN@*DlU+~Yk_qEj6=88)wR_0u?_%IRUfh=TSMa=E`0wOq^29P zb~=b}M0CxsnXDe@sI@0xeaG8&5-?6*40cT%&?)4}N8k*K-T=vp->^8~Ag+YU0Fh;O zeIHG{2BGVFJZ*ESvYhjP@B^j1f!Do#@1lE6SVdy?v>5+IW8k2#44#{%mziGTJk=~g zb{1Dx==AAfHhl2#+DG{S$O6Tv0gsPVX}!>afn?F5%ic1)i02auBfdN?Yyt4nJ8J8G zQXj-hc}u?#R0jP7OC0pHlU7Y!YOG1dQ}-|qdijx7_LVnWw2gEd6f*#VXK0`R?!(ZFoBk* zZY&^7$W?bOTh3&j&Ieorbc%)!V17X*=KpqJdyi$SOKyiqn=SRGyq~CmqlXzEp$x&- z9sDTzbPq!MO2dcrZ#(emohaRh*Dir=et~~GFX`t1Y&@O^W?4V%AArMiI7PTz$na5l zWw~{s`WNUo@C*@5Btx298uz80gMxowS0YA71$suj)1rfsotqcK7RCZk_O9FgEG$Q+ zhde>@eZz$ZD_-NoWl_)DP)m_aLpd^z;-1q0JB`pboXi$3&rIPQqyU}F{T z_8$smMO67nUk?j5rgQmEMv+I6Ff>{?*Etz;`Cb+}MBky=!uH=K*7%RxtLE!*Z*nev;3Et|NlokI};Pv|Ga|xt1a!m(VDpJ?P{K|Mc0XI z5i5#nPt)3La2o!sBb+E{{PW3Iuk3 zx1_E*Y5(Cl{NntH`7@pB{q}veZSS4sw-x7aHt}^(*)`y;G`&cMzIR_K{fHbqD~j(GUZ?nLNN@hq$@ZO3`nCT{$|%?s`XFPI7EX_*h-Fuu4{`yQLTOc>qV zOX54uLfH(LDdLauId2!nZa2JawtuFLsSFfww_Flp+r+}QnqeM|l?q*)$vd=bDYhik=m8rj1WpP~F&&%`_NzE(vNa&|sYlUz z*8ClX#2IFoPki0HX;ptylNx2x{oHJHInXwyB%n=2Ma0=SlSHftI9pm zXf@g7ZqEV?6?E+kc(O9Ut{t$^Ud>T%R^Oa*G_8}q73`$q4031 z!tIldZ9b7;f+plbK;5druD+-gW_r6xrNuUSDyG(Uas<`xse1G?aJDRgMqzA2-s@wq zLC7XF^bD<)Md!{JOxFXZ(lTY8Gu0k0MZMrP&fX~K(h5Zux2!l$IWA!hIfcqYJJP8tZxQbUlm6^}j@Q@Sed@weZxWXg~L&I@rdh;aOmD*+6 z@Cp58tHqR%qElTtAUTC*CdNkA05}0QnnWP`ad~VJLiWM_CrI)nSQarian-ujD;Dl- zc%fczwf|U9Jkuu`&_6ybsxW;j{c~VM#5sA@oQIEGQbb7pR0Gd=7Tr~lkkKE(xdY9& zL%|DB1MQ&(lhfE_z{eW~Gi`;eHURtVO$z7>qnmTV9mHwjaYc)+ zkgPoW-WLkt90M;LBgQH84rzQ$;X#ntAYS7mkYBUHzke(A9r6E#VWcXEa{oe>ndRXG z0g&7nCS)G^j+>z1eJ8w(1-^7;b9F*>n;S5YY>9eq#LwZ{S)vW=fZD~>LNQ}*+hmo9 zq(YmYfF^sbL-wi$_Oxw7Cg!u*KHW6HsM^f7^V;q^8VGk4=9(1vj=T^!g1pb6#E7=;S>dFw^g{(3P0XZml8f0D2dN zZW0MJVSq*|^==+m%2uh=Y`O>EBKqJ#l`{;qFhs`bUW*p1E-@Z2r>b3eKJx~YI#OT;>h zvZ+hFXJNf2Mcdzx*M8WRJMn1rDjZU6Y7Na2cnT z3AB7p$XW+{_zr~3h_0SI5qOs95dsa^qR(o-Q2dd|p=o7kxbu zT~_4Zbsg8}V=E;)xK~KyW+;MO&9$+2)in%vq}d8r?WyHfu5)?`_!?yK#6^$*?YycD zU_b5XxKOZIntS@`l9YTh(st}WHVzHt)954!%Dv@#VuE`LwUCITNot)ooDWf}P>7Mh z6rdS0V7BI-J(j#e!^Bu=$fb82Bs&G|Td|*{4t>I%zaA2tH^UoYIb=W%x}3RxWCDuz z#|*5yIK>@7^b6n&0;oq!8_LcHwkMF8n5&=k5J!>bS-EsEQ((;OQo^n3;E@e zz;|;3YpVhsv5@GB#~xIl@S2DE+!2MTmH8NLz!0eJa0XagJK$U$z8F!Rg4rOnh52-J)Z8v5a>Z zDf(5L>5m9cfFxnZhm|*`n%uU3MstSoy_80ZYvp8W5J$?_Zr&5it^s$x8XE+jsSW-X zl%V$R~ z)JkQIy7VHRCCreiEV*U|fY~xXP&a(55ZD35EJ*yE$a_MAtn3nt!$~F$6IeE6!z){# zQCg@?yG4b-^$iHpsIb;cOn)e(2HP+D6S%Uj-@+9Ff6$C;fC%@@Xzi8O6MJLY&Cb2< z`4Z$_RW~g7IIO0j-rxdS4tyX60$=%cHv+4FG9s|^MpTy@S^)L!U9eD8LrFu#HLMKp zMytqkY*D@^vX1^MsB9*7@TO;m0cVn0MuseoY?D^2qH~+6XUITXbQUPlm(^MJ7>?Ak z?UakST!+d)r({?;Z($s^5RFrm5k76}Y@_Hp@GNGu8lb~EB3kBJiMsZO1vwZvnHZ|Xp&83P?h>m#2jfi7IfW254H9V3 zRE6e1`RAhzmhH>s&WH)BuW!QNriJy7iLqc%`8qBsoP&y8e03HHMabc$ArC@<49e3l zOoVJqDW>saE--2QpR!DL9gE z#tCRL*i)*0%@Hw0w$d;?{VP_0b1IH0BX|LmZg9->q8X)sjL0gspx^Jvo6>v+v@-Tr z_mZ83%vD!Yr2axwnlfc!h9awkh79Ag08A3RpQ9NPDrOlT6rDbN4-N3}<#kwX3hWIr z*^+(vx<9_N?dimZM)MU{mNA%$)3o`GpcTV<-buHY^k0`zzax>#Sz?v0WVj;_-iP(slHyA-pu`F5; z4^ZO{Vpa5VMehvl=NJ2MEr+I(FW2*EH58P=In%G&#a{S{cEXoy6UZYB{Q}s(Z{E5MVS%Cl4?8;Cv=Al5>;5PMQ0?*|ZvSGJO*- zJhjA1zp2JGGiAX=sABqA8!o9)8mz}CWrYq!r(hJ4Cr+%@NrR&>B}`Ju*W2I*CQ);` z9=^XzKB>2#j8$!E`s9N4b zy`!{`?cHDD)U}T#L&B+@&iD@`T7yFn7GAH|vfy~&Bo5jf$fVi8(Ty7@H(Hgo{dU%> z1E@3RtDDRae z-ts=!>NR33BN#hcbnVi%fE_?VHK=nz%b(HBK3;EUL8#*9!il5?MR5@Q**CRp!X(rP zUm%mY#u!pxd+7HJ=vv>x+_k@Mc(d>JNveXE0eU9oPi4KW<@QcOACwKAbEVFfIt*)P zW|~d$z*8KS76mZ#;W?;tdW<3oO<%ds7yHn>5X=>Nz7hZG{u+OrKpv}@m*9Fl-J(-P z;qwg_y6F~>m|6=LeFZ&}{9HDMq^$xIEk~FTDQw(j{+vP)!13RsR07i=^5-VS|Kg6( zw@~mhVICHRtu-;PGh(uu6>@A;9MHIQ9?z~Ti&S-pP>4WRqJ4iuQ9Fk)AP0=2M*VWw)M;U~Y+wv}K2v*DCmafdf|oW9#Rf&8OpOcWQd&2eFLN9cM&fVjjI zBRvFG)vWMr9ze=zT7_(2Zm|^N762;trBIw0Vxa#|vP(+@v8`Ho9HZ+--5t5IewKp* z2hclZ9rZ1r?cFX(C3x;+k_>JcknS71p9O`2-*aW}PPn~W#3;h42l~NdLU+9^GI~#E zqDVaj2)+8xoyM-BvbC56iSU0y+5MT39l zj5RXd-Dt{>*zl|AURF0WGE6LN6W)e<{#2abB?bD<7Gm$QHw^+V=)P%`=yn2aOwSgh zH@&=5wFKEuA^+EWj~@x_zzo{MUm9SkB-Wghl}SX2D!gG*^f2eREd&W!p51Rs3~KDN{T?j}Ga+slA%Dnu$GkTI?=@H_hvB3Bq4Sb#m}K8+h}TO7M=Rt0LO_2<#!TTc-(kI z_}ZKlN4DqXS&bfPM_I}wxRzL!E>#P9?jd8o=2Xzd5=rhesGvA>+y z9yJgm$G*Z7+YXs$9bFGnoqxn0U!%-0D&Eil-Z_{UTM6?@Ng4|w8{!e6lO_`*(y@WOd|MST=~`%g zVI&k=TeW(=JKuF90t@d}(+?wOp0yN~$Q%mF@N9k*L)&07gI5B~62g7WrjhY4H8RzEL`=6g0-Rl>2Hq2k(}oi!n&>OP46%32^)+W;R-3AT~vb2p=w z{$R;*9nD?-+e7%I?l6DC!>6_ee1z3h><}`Yy3S8uN`SjPJYa*yyVYL&ffB)HT@dKv zM1}+Od68uha5|u-e+XqTcHVoa|4iWi?VnJ9xi|?AZ!zf3%(Q9NB2W`<5rWFF(nD81 zoN2`!;Q-(w4ML7~od5pg|FY?3LOG7*;MmlKYI9Kwg6q2{7&=-V_50@4@x~vR(&oPE zKSg`2|2^7c?NW6g!DW-339dZzCQUcY)HN` z>&xru;^X)xI_@KthUDt_@!k@@4z`PM5R!>?h)}V%>EYKk%A3{B#Fnn~^b#;*5PKL^ zvoT!%FE#cJP6^iYRD?1~YUjXAO)f+-+a+*I=Ev2in>|MoMxDsnbcG{FDj6kJ-jqpF z`5NV4_AnYHS{)X3!0K%{N-%=QJxjI=9{pQI;;!z_Dfr3sN;c`**$^g%JGcB7MYK^u!RVb3~GNYmH1<7bYk3X!BQ)f*cmCCGzCxF%f^Z0 z8BQ{|L`;As_5h&0&2wa^3}NL@;|BD6nP>7pC>-sBe0Ry#1Ihp_9liz+*ur!J7)x+| zlQQznela(mLyS9s%Lp4b)vO|M98gdwJ`9-d62mmb0P?gOdH$L6eYAIW0@+@f09&XU zFovtnr+n0TyW2HVLUfM!RO_6T9eaR?C|1#r`=3H6TJVji1_T0NfjY=bR!K#f@WP4a zE^aQK&Ym9mW7&qT4DyFh`;;A)usTh+4!LTulSz2`i}N(2ZMLv9kDS_>ZGBwNQowRj zvI*B@aI#k2f3gkME0umiZdivz|-3>zT0CRYa29;qOd<<4=Sk zGKIL0PHQrt)?1{kqQVPUN(mg=yQwc6<;tRKmBzM(;#d9MKG>(F)O=!2;fR<#S3LQ* z{~fh|(!NOK0*aLe0`sY`R~l{M2*%H1I)q-35}^(=yK{_Xy@GH+LBA$=)9AeF(S=6M ztrFlMz>|a33rfpM73{_%WeOes3*W>Hxu$tSIrvbq;3+ZnxP%&hZC)j~6}Zw43R%H^R_J6w;);~kf(2#aPc~G z1vP$P%0`ixd=kW=phXmNI*1|+DnWw6t{fAN7{p6?aaUtE;L^|%b@(`XXl^;LK$}5P zC~9|4?itVag(uq^MdS@-3b>{4d7*}W$HBqAojn2^XtD2CGlc%q9T5GP{b&`y!Mqcr zJLz}YZ{Dqe@!$vcIv5(Y8Y2i#E-OAey&e_3h8K6F?L3o&j6Hcp{vDHz5b9T z5Y$%u1_X$Thpty{tXUDH6Id29HjZ*B7DKzgUj;dBzgahwoS~&rR(laUP#K8--ezo? z`re3N%+(s1m3%7RMhV;~%a8^>Sk>Qd(dO->4SZ=9K27)*zR5*(n=khQnwy)SlP~M9 zW44wQU2no=?-wbgpom|Z`4|Cr4bvUacTykVK`-u_j>;1)5#t;4?n8(7hYK%hVlS*A z%P=Kub`IAYf=gSMX3A89fD2gu?&*1hQz#*P)I=A5>vtS~e^0b3?$3Kc6(e@yIJx=I z&I4<{hdbaHVPL)8qBU#{peX>Af;Hg!mDf>MkwgmOHd{_|;Ct`%*oXc2 zOg(d}NgnLs;XpPpefjX2jHQ3zA6&5j5bMlD%jso*Zaf!)o40FaZCJ=HqDR2IQ>Goh zooa7tjt_?`!Ki_;wB(DzkdZTRcTgds(lbXbl9%+7t79v4gK&nCfy)C}C7kSZ=HeR0 zxKtc;27l|aKHOB6P=AAl><&O6A%lV2PzsJf`Cj_Q=nXLMYy4c<@Q}6-KFTLvK1qn?p zdHwLj_mGGL+xvGysVw35Ui>A5>t>t$bflvE0qd8;M-Y72Ig`x;l{Q#n-%>izPZ%`n3EYEefP6r5wW%2_tyrkf;FnmBGfgTrq>cTDBl-a+(x8k5;ClMz4 zSX@HgNr=7p%G@36%R!X*VwjepYzR^LV-Wmd1b>lo&c)k2AaXRm2o``9-MwTAX{pu) znVRVcj8iia7ba*l`DMRC?&ueYzlAC!jFn;ZbiRJ!J3x+_q?05;$wTkS-WvD;rV<;F zx&M@6VWn#PIPy{jEC}V#(xBE@Yy~Sv&YEQ3avjWKUxurLTlmHsa|qOT8g46uA)1g}61;qkrFGFgMRq6nLE;6Yns zaA9E7U|=7q3jRPu7`P2A;u-m2in12mAB+*OkzX%JRj!-s9OvDQ&Cn`xPc*K>kr#Tf z!|1lgkR`R3+iA(>_IdgF1nOmmEJ+%vS4n-HbnEeuT1f(U|E!#4=a80MCA2s@ZqV*|{$k60tv=h$1giDM64OTrW1t!tb!*TWky|o@E~4P!B5E-p9-Iu+ z{Y<%ucikmMDbQqWmRtxP%;ex>yt?kafZDb42HL<%JXTtr6H&Qp=sknM!1{Z>3D!lH z0_0`bl$AmC`O`yH_{W*E^aedO(T5?6*_}4(mM2H{_Ybf%IPtJaAtd@Xm zW5JgGJME8XtEh=GGBh1VmEz5h|mVP+KP;u9buLNnqTGMRy>Ek3U0cg!Adjkq`Vm zDqRK0HxIk=L z24?z-2c1%0Dv_J}XWMAnC}^tl31tV{;4Xi@f8`iAoTC&l72u-B+@CPW7xb_)^Gjf$ z_l`ZPFhco2cR1fn|6kdI4Ro;;aw$mk7sIJL-o;L{aoi%vR`QI^mZ$VK?q^ah!4JGLihS!)ux^`*p>WYo9)sSnO z_8m7DKyy$6)kz@c$A{~>KES6aFnLiN#(rLN4z%v>m+6fY8~5Yx@#PL<=C2nYW)an= z}Yr2QP!xw0pH8v|A>M~ z?d34>Rbh&~q_M+iG9n^a^IP`%u($E_hU~x z;xy7QmnV720)S9=1|AehpTDpA;DGeI;p8zmYWzul)D{um&>0{5y8$jIPyfy#Tt4Eh zDxLxHr1a8lQ+GqN2}4?S(?S0LBIt?*CUl<-Q$Dx1j;rr{(URy_R~SXClT7HDg-xE6 zi1YGlADDrRM5_+-EbY-xL?x5%1>3 zsesSNzm3Xw%g7p^e#!n5W06OiQ<+%?-_0_YNoXW#GU>37`}ph(0L$led56cnXoS9A zff>K+7V`Y&T^quF!l647Xqnr^`QvGseanrq>rAQ2>pY5;YGjfa@X%4H%$d_ZHbaI7 zTiO5y0k?psZnbp&js~?tY|RS2>-B|5YeZo0Zt=^!@eZoyjA#P8istR6{Hwg?DZ9@j z#O)Pz#q^4NZTlg;Ydh6C*(CU?UZ@={V~v0<#!`HY%JT5dm3PMY=XVz_j*R^q_gS_* z?#tN4Oq~)1qd~#n4t`H`ZfFBHV-FqtRc@e%5gq}aH%9Ij zWH`OQOS{ND$EUZqXy)I~WEtug1C(Zebm_*uL(dTJd1)ZGa9qnJ9^$r`(-U}22*^6A zH`WG@P=GIN5PR^!O%O`-4@QE1K=T?2AO97_iIbW0|89~^X-hj{izjwJOmo|5 zuM%kK-cae1nP)D~7Nvm>uj5%60h4Hncmc4K0xA%HZM;EZe}J>$t8T$oNP5}W4Ti#5 zJFBh`Q$I}nii&xH}Sp>B<44g9#2et?w6M48r`(=28pujP2BP`!HnFcQ#SEps90Y(^jbo=E}GzrZQ<+wvEzi<-lzJkocy{ z{GwTPUDz|`MFd$kJXf0X$_gl|-RY-mYB$b(nP~S)W0g1+XWW*4>1!B+ytMOc6cw)8 z>{%~KV=Dhka_F$SP{oo0Bb-?H4x>sb;MLW^^>G6s4oQk5_(B)2+Xw`;wAtGqdml|@ zLlbnfQxk6#|9=+pY)oEaT{J>5o}#=#d8Xt$yQK0ELFuz0Y7JH%iAqt%3@l(rIJM5I z1*ayXZp}k5a^GILUvm(#Z+35#hjKA)!Z-gJcN4wKR8>p=0e7dC%9RJzuA&(ql)QOw zrs}+s{wuaD(-rG;?8`{4J(kBISeYQo>?d+%$d1N@m9J)->)YG>qQ0Q*r6sgKC%Ym0 zM<$%t>UGN(V~d8^G-oI0rs_KjBuML(i~m^mtGfFB2+`B0{*q4HhS-E7Un|yWv9lFv znl)<%8ddF6x8G^hIbwEI(e@WX(q$#@(0Tq62(pNI>+0r_8TM|Xw}Hl9TX~8(SltyG zYKN0z0P_@f{y~X)@^%_$qL-$h?w~K^IB?24My=oU-VMn)AfRyMD7XA)sbEFpUD^sN zKg?;#IYg+*&}4lH7WFY<*=?J(m6!>Or()?!Q`)|MbuRml^GLU7e%g`y@WH zlg>)%cHW(K8ChjhiIVibN{Q^NS*|l@7Q0eJ4;*KydU;5pX$%GVZYsGM$qW-uGJZiX~9f;9rPnEj3OgtTa&Z*ylo^ zP`8S+oq1``f$su(@B^K*=l}T$} z8x2Zk4`@gEdT~&l1NKvac^`qfveqRP*DC-Aofe)Drri+OUk<^mTQFY-=zJTL2j9GK zWl}(Y_l_PtQWq$Hg$+^~!0G{iQ4ojj?AnD|B`ij4n3wu(pX)S*f7J7w5zSDZviOpT z^tm|J9Eov^`;N~`R@r)p@uK)8o?7RQR4FmOmXX^@y}RRy@-4%;O>tgM7UDdhg&nj9 zkWlS|WVgprCBAU`$pqPEl76JEgdP+oI>?3Sv8?+%(I3HX7N59$%2Q637G~{h^V?<@5hk`Z(IJ2kw$`qR1)%{gJ%cdg(q+>1mou;N$R8$!EzHaltgDDHChuy)>sb{5X{nS^fw7&TyU19y z;f0W*wI36eJI{qpv5Xi%>%<-fD>WMKqf>L|%1zW0u*6>(-542YJ!>9*RygSk)KE3< z<|f3G--9mBB_D3^C(-3gkgce%tFK!w8xXPIH;j!q#;1f;#sX;kpyjMh@R!{_Y{oj0 z(FH9z1^CRxnihx5jV>`Ma(T;1))O{UbqH3*6Q5bCi7V7H+3@4bxziWSl{oRWYT>22 zL&v=(oFnQURPcc(`|DR$xq-{uL((u5R$`$y*x{RI#xArkU9QepN)H$@ENsVIH$nv_2Y>#|IeN z988SgR1T{M<|waD|8RpX6R=j&^VA#hFkt0s@>ROFR?tNuTYtS83jRus(yyv&+J`8! zh6;JlxdH+%dnt{g7Ejzm*7CcdL;;5Nq?iANSKtHoo(~4wwqrWC%RlyTLywFqbm7t> zhSZYX3w8={v11co&Y-!9?cf+hL`Q4|_yFY$Fw%Z|`i#TjVHv?!JQtyV@3eH)Dbo98H;^J;V#)rmkE>7;%}J1q~vu#4gbckdFmz zStL>8o-PkxmBnJO4JK%ol;|TES6?aoM8f_wA~Ci!SS{S*YZ26%|->1l5D+ zVv{l4E()iT;%=%!i`>_gxPHP{;w}=8HlfNc{2BO}CCkQznSW}2ZwW6PoEYW^h;~;L z`q(?N9!6Zj`4fr}1>i{gJUkS*ls`{_B$y}u-Rj;>UK9z5qJfe*M$DGKzmT+*mi@Pr zzyyKYP4Li(MCvU-H?swk7?V)Gm#xck=QpwYuo=;3Mo38>m(rq%MG;HS>5T;C`+ZXJ z3k1JFAq1LMy}Gi9WlkXC-F>C6KqFF#r_blaq!ST0Nr)cby>&sKSQ9A)R6IL8LbykX=U=cT{h%V2n+50_F zvH==!@Db^jM9Ef;z5bMA9w~pX7)2Qe5!NjN^N-UQhSGroQ}qe*#4>4`CZeLEAU)`k zmzM^Mas&7OAxD{B&c(K5gaDb=wFKoM?8yEqp}jX(J&UQVA;9HgRTV>Dgo3OV$Z)L8 zn5jd`1tnoY3n2bjOu7Rt?*jrF$S~j{ghs)ax{wiFz+M;?p7idr8qAaVS`W1>F&;a3ZyjK7sJ5EjPY2? zs9_~?Vk&PstV>pt-tvWBR(D=&4+(F1owyPMem07JW0O66Iq|X1E|^6*mLYJ?!&EsI%%c`uLT!@>s|8rNmz(nO3ECkY#5e2%0&VYOs*oq z6EKyYLB24?DE=p+U0VU_qtoF!*~V;80m`L>GxdWefI=FHyQ(=i3@f*{FHdL>IeC;Z z4!aHr+>5pL+@weBuDwc}h@wmW&24(Hg&~$X>%Gp+UUemkE5XyHL2rylu<3l4QBzf& zVh#!*@En zCqV7EdcD>ysOBE@QP4jb?x+4~#g!6Qkasm_Nh+qc+=beDORV(Igf_EgeoXk2mCzI4 z8pb=;UHpEMtL!5(b(jv$Zs$#K0KDpI?t>t%1lA)ZYcRf%?7sHgl@T&j{~DP=fY>Yg z4WQDI=@gK@TqlEw=LIl}Uy9H1V)H<_)F6vftPx>u6oVXJy1p5BrjB9z1KDX%JUFOH z7l0Me9U*=HIUF^pG$y=FJnCU*a7Ov?2UjE*zuYa+&0wYBdSYd&qU~xJrRBp@kj0Y* z81mdo#(mzTYv=+yjI1()wQvltJ+oU$%VfiGC`jZQruf-m8`~oV2Seh($cfu>MkOQF zlcbdyVY2`^aeKn}%$Vm^(K3;s`Ewu={&bckAEyMSDb4#|Q6*pU-P+P+;i*-6GVjt~ z*9MrLqI*-R+nvAo)0xspz2@0C#Ljz)18NXVoQukorXDHS06u!(FS_j!yCP2>o*rMd zKNWn+`ZaZq!&B$iMo|0{z&ESR$Kv65`NAHo0|2W6G+~Xd)8L;e_{0GN9}3CgjTLU3 zIClcG0^1iDl^6ioSq|v3x3{Oh?E_hHz$><*)Ti!aXfF=Jfgyt$G zqfvC5otCEo-u>4lUzTOYI6p8iOvCCBi#>I~)$CYSls?V~n!SB)b4 z1r1N%BD5{}<4&y9gw&-PBm)N24W*NH@iDyt%8^fN8me4~v_a``z7^rN9aU+e+)$z82|r$w~tD_?Sum)>*xNrF)`4)y&tM zlI3s!Wq@D6ZXi^a7%TOJ5Z#v_|7GV><-CH7hz0Bk(fpR!du08LbtmJv$uHn&5!>bL zmrV``uAdAxNru9w z2Vc331Wc9I5P?N$!=$gu$Dn_3i=kYeML(#DqDDbv@}TOFNmQQxAUfoZT0*0OemH&p z^KgWi3A7?h4?NJ3*sNj%PC$6n5($_0ecqkT-D-MV>Mk$O#T z6E5Tdk&-nP7O5-90LHNeOD_u8GcnNHY=&+|zvf#+!_#}#H^rmS#1wtlB~T`a#ffCR zdQ>3O(m}y+CHqKgLNmY~$WuzD#`i zglfK1t-#r1iG&{W_JK%_YuMKHHHRWNx*dNCAaC7$)zb>!U{p@8CG!RCrr76^-T4bWPygSlc3ab7|@K-Zws@oSa(I&&zoSdr+JUavCJ9s+_XTwZnYk z?mD})?^0ZRqvX3_CYm{#?NsLt`uol!9 z)E-fvt>C#OJ{@GWNqtT*tv=<;_(i4f5@AA;kB7oOpvCW3;zbuA)UNsOok6xapwFQp z#(Rz&lTHDw8b5UiE@lWs4zaG@ya9P2P5SK{y!3(J#%;-56FNeMQ-r)d6p&cYR*W~4 zKKc5BDD9+n=BFBF=8X*@l-+J)+J^%B^}r zWgm=3zUNV3CCex3JGdKE*n3yreRTS7ezvY7rIgB-J0<{0S;Hc{t#D`V&2uNd z%YPh+>o5PnjwnxTUE6+qt~F}+K7YA5EL7z!O%4f$+H{qjC?9gUcMD#sugt7N$>G4Y zuYYr;9*e6P0r1o(yVfC1qC#=RefsG(#aqT(zh0mE0kD0eVf|Ms$NC?s92>{~M=JM= zSasXvaNV-H5tVyCG)Gs7jUwt6yfM){P~5QfqM0Jua-}arqIGDmC?nHsUix9+C;cIe z>V)?Fz G0q?<+@btv^(X4{Y z)5pQ-T6wP*CD3N1U@HT|-uj^`yx2k&gO=?@f`ATs!I{mOJOwHD5X`tslmitS%P ztA6eQ)UxUYtEY4vTDj`lXdCjw$R~mxz+WW#V|z`q*E+o=zo!Lsx3#TuA_#~(26y+^JYyPayig<=OufDPV_^nnoF9Sp5^4y%5)=J@&X=S z8!MIXEuKEG2N-c#F#Z(gMuJU8{511U#M2iSE)g3>I5uw{OE0%Z;&~a=)l8U?1DrHP zb)tiNSeMBM=4#osxORQ=4xP15{lfE)YXdkaLe}Zi8LX6$&ElK> zTnsY{2P}qf#q_vFv-=2~TrYu2=QRj*hL{EsoaNPlXTx!S+;(_zU#(;?inbSkRc$rHHW<|7Ted^1h2an)?&z`}LU^+lLP;CDoB9D9V zVe58z;BV*RF>O|bPN|O#^(hesbA5*V{ZRTMZpr~Yk3wF{*NV~X%OCboDSfj+gKo4| z*Dngh~w2)4{VVGSaSq@`BhY z9F+Y+(G&@-XQxsq(pF@(U)U>IqLK7>$=v|h(A1X-H3JR~X1fWds~|&Hu`v+MHU9BlF~$IAto<{6pWFF;DAKg|G#6xOCBFlV=q1WP*R}&PUtC02LAC zD}tI+1u?oN#~d7&yo>c5WigN=__d2}JOP`lWdyi$6mo_{pu4n~SK>8Gp+d+|elt<~ zgYk0Ufa}8%17sg7pC9gi9D-mtuXpRwR4ziVeohqIpymn3yLUhX_A-I|Dog7>%PlX&Cco69ZlMxt^5wbvJbADwnq1iE!sl z?;y@~j-wNRYxDe*Y&}782iy$|6)_|+u4s8*2IX%!dwY>8Z0YfYa{-G#QR2iDZx@+~ zm?&9AU==tJI0?GZQ)q#M_VE%0j2CE=uQAMvhBU=+KVtj@AbI90R z9nmcyAS{6Fu(o88UTu9Qu9#+1bVR1Y=b@-e{7aPi9+T>#Qb019g2q58g)KX?tEjoQ zXRaol`QsrE&!P2oQAn8ApC=f#q1eMZ(b26HFF=Rzaq+$)D{6-oN$sZx3~X$c;&s3K zy)(}NddWs{b4RVNnwf+&voW<|*@k1$l)0xrDwu$_z|>Op+^MCvl7$MG!~|F%_8AC? z#Lyz)1SQi0ErYJmNJO`mW3PzQc?a(PEt-lYVDbF3{nK3aoP<|?O8DTRSB;L-UWLd` zaN*mRJ`5?Y?Se4b{t6JtBdo+2l6>$`qqns`T)e|R5bl;n0b9@)ua+m~?M+ftRZIv8 zXz{xMB`3csB~U}=-ezMMZ-x~j2#AS2;taotvjW72kXA^!m59LTc`n-?G?sknA4QY) zuz@rWw)?^6s;&~M zY`*eZCx&BF6UxV^$t^H!;B0Vtk!~>|MNpJ}(FZh&5DaG&vctwO$x1Uy+R)jFsU(J* zzdtNpPch3{||xM&!nQdq3&?u`m=g z;za~aG-f->KCZpyS{k#GX+D1U0hLxjux+C+0{FxBN!jXGYkl`(Ny}f2>yqXfOMS_vDIGo#ZhCN5k`A3 zyWU<0KwhoTJoG_1o^WCPcb}ol>jY2GHX%#*Gl#(I+@*Hf%V<-%zV}{mBM0F=TZ>P-$*x zc@R0##uX<4&tpv}AJ-#9j2fbh6hvp*kn)SkXhHEYp;iXa3e)h3pCN9njQrSMqHZf6^{WrX4l&-Rf*|N=irn0%Fo%nW7Khz9mJ;;R) zqKQ}`uqY=?-^11G$%5%&N1Ly1R~CVyG%4l?7&aqzqQwdgL5(eKqP?S_f)Em3o|M#v zfd%j%vkpNs%LEM{7#YqJcZ3Z!pG;}AlZD?&XG$|>B$0BfOq;+1nDP3LC!uISJa7OD z?yXsKu*f{D7Zl|X9Cs1*!8s8(_7Kf85o>@Y#_Ae{TtPxkgbTpYh$a!z%q30+Cn7|D z?mN&?kM1M^eT=@~cAQa;hgo)vS9T7egM)E1V?>vV=WRH*{ZsM$X-gE<6tie5W2Fk2 ziPBAY*@5g%l>twF>=ABBJrmiw$}C9XMirve=Vnx+o88Wo{5H;v{LEbe5gZX-6g^Vj zJUl)an5-Rh5Yj%_-;DKX!Q3lqdH-k!CK=JxWSpIJAi{E49Ef~8iCHB;%P^1)z7kG{ zSz75!#{wmJ@D|JkgaOhgkr00jRx;#8^DvWsig6c;jPM)TR)&}^az%pUjU+so;IL`8 z0$SCoLguC?4RQ8Fk6rgu)0R2uE3U#cQW|JzEKq_+qAo~hUl><(H+4Vs2`EWrMO2z# z4s@zoah(N@#0j56phYY)W$1m-!RrgtT5xkx=Ce?edkj(kmO~@63(?=I&69%0#Tw-- z)heSPt?e(b>kN0#Yn}!h2OyZZQ&U0BgJ(<;NIlm`H8B_sgHshT2Q8T0vKLDp^&*D` zX+ECP4yu!q{Bx-Lm=Z`fGJVC1tJeJ^W)iq^sW~3G)Lx}`CEdecmUk)KcXWT}_ES~c z_2RfaLl!D~4%NMkHp^bkf`ipp+?R}3by11-_z8i#(_Od5uQCh>ru6_d*J{YTCkxnc zidr>pXksBLsinTP=^tOtnn8aq7Q!KfaoJz%032z^@=Ih$S9#l$x&|Q>u)ZtVOly@V zd2#)(f_ElQOPYj%!mQ`CKHtIfIr%#k0NhN57zoH7?sg* z5-jF@i>@sE2DkpyM7p?X3zo$GgKkKSCZNknoKg+kut`3wUvud(ACD*#KfdD0pm9VaFx#o8Wf%M zKul+iM--LH79D{DD5yX8;Tn9(B|BKoeX^WIuU5scI_%hwOjS`vjv^k-M}hp-cd9KE zil+*>(SxrnY~VbIiX<9Gi`s2-;w5qXu_H*2-Hzy&)$qQaWCEbI+2=zW{o0jV`+spU z;!15KmF*=VD|gSVkwS(Pu3apABkOz_I}r!pK;+|6xCDX0S#xiie7?2{OTu%o=RamD zBmHI_lpC}0fNHGJUYZ2!G2{V_r~^lY<->PnsHB`72EDnH`J?Ll>+%?7griYQ_3esE z1jH(D%CWfK6e+C6k96755YOqF|GvZL-1K(mRoiZa#_1XU$pgp30YP5I{`eZY*7ev$zDF6Tq;WSj_*s1v$9;=`9o7C$%TyGQpAKOH+sWX7&Zvm8z=R*&Y88IopDFTO@-|DA9LfS z-V54%*dM0-kQP~vo`o)dhk-{%q!^|fM8l}#MEPj@1`+Y;`c66B2Y-wpO)r97Te%oy zuP%S`zCLGizXqLf{k@d5w8iFPs6z`5Z5US{N2@T781Rn+LYC7*HNw zU~JC7lQ2b85U$wFL5T-i~JX zLT`Lrf$iJq5KQNk1C9;L(shl?8ZAhc^fCA-%0uX*X6%w<8-}sW(ILj}2B&++SUd43 zfnqP0jV+#L9}`;Lzja2x{LY~V&4_&Qs`17gQP(YtXZnE-pv@m>aq*B;r^C@*+lqi5 zR;>9xWu@nVP+sQ+Y67@&7j^kquQOtFXDlk>>Ti45MiUmrcEd)7nuR9tuaS9AVOL=_u(8`Sai zuKlUWcG&cC+ub`>dKhw9$$2}%fw{7 z+Lm2RQVGs>eQ2@`8v3~PbLAy96~hntGp-&pGO$xHR}U{T4I6|bmz3`t;8j-D=$+Sv zK0g2&&bHj4xva+OqrDC<@q@(77Xl?rb&x?`*2jzrKzR~ zl#F>fE77{LIyP}HrUwRw$b$Pyy6L*23>idD7WwK+9*;kAxZgXuwqLF1uwjQKXqID#0TYp zM+ocAuq_b0bpnOgJEOb9VlS?LTFW|Z!)3?oq|)PJU}A|*4X!xgZ;#O*B}4bG_DHJK zM$WC8T^ZfrwG^ZsnRe9FmdB}xg;hL()H~kI?~aIpr@kfD3rlN4hUIYuz94i<63HO8 z=PrgDa``uF*h%EXDKl)M7dJ{A&kO}-8egnX$v>6exOJkd7px=I!*{P9b_n@njK|M#Qp%#*-!5Ht_+e?LVQ#XR9x}o`pqAy#gTGmpfB9D=;{Q! z)BWhhsHOPi+XeZ3Bc>ZP0@G4j8~4KBIfHkujoHEvs=Qe~;{%36=t;@;>WbFr1QX6A z<00eix`2N)h3Z#m6k18p74rciC2$q$s*J0R`+(~9-6%B$18sO|-#tENrYa2UgPFx$ z?$8_rBa%!VhHezcPoB<8d1wiaJQ_9QCdzA@UVgtZN-k2~ZtPZB5k;_~fNqE0Cbyct ztB2$fSS&po{jPSy-B94*;{~Bgm=|wEFF$#eAHMZzTeTK_@jw!J_9wS#1@R?K`%gdl<&hNpVk|v!`numM ze(XK{M8I2Z<~4KPPo7*nrm&W0f!DXY2ae665mj@W9%hjL?i9#rm^Neuc|JLC+$VbK zXGqPFyvwT1-ULXp6Q1Js$qL$#!!UpL6jU}h_mO34QGrl&+Y{~i+@iSzp_4%rmW;A# zq&~t-Lj+QK10mxVX%5s^lVp^mOY$y#bASTybiZ4VL6xAHh=s%C=1bvB#0>}m5g7{6BNsQhIc8m(b2{1? z0dN5r{fpr;`#96azpn=EOB{C%7PVPxq!Q$G!F(C-vnIa<_*QWW# zRi0uXH%-L0=K$Ag@Qq3en;+35{l_^n;&^#@&B-vJmHs!_i!02RDC-)^OL^gN`sE;YJ9x;-SI(-u5T8y zu6zyQIZ2D)IQH=2++*s*xFV&C&Z*XIdt_LpWzs8fkpJvncS|v?dW*}#7T^7$~Ysw z;wU<;eFC>9*muZjcc2@BD?^3gRjxiH=ah}vnRNf~peBsFOPHPCo9K2;=2Iq_NzlFS zFSZ(}2t4Y318r z+6Lfig&+{NGc&f#S}PO(_A4lRmw=X=3BB>(=) zfPtB?U9L?pC(ru@#XPaG_|V=BKmqB!q8RTM-8s=y0z$Hux3fdMtZF@|U4qRW9mv>En|i>g|%?!Pas zR14IdT(`fju*%(F8Ttzsr>RYHQd&KF#%F&CV~|!%4}jUPhY~$qm+m_oTYx7(yl$;! z^(3bA)Gl{%Y9syB5{P|kZeT-7GJt13BcRb`5;lNfQ^m`K(OljutN_0)bOQT$@cl4f z?FQ$J@T+0YcId~GXk+6Qa9V+*Q^)CwdF1R&S5-T7A%T2BfI#A$i+QDJr64<3^XW2lnd$o820kJ>)`Q{_<<=tCO9Bx(DD&Nh1^h>1mqV9rt$vHR|@nF z+gW)J+V_vnn#?fsDfeiA0O6HLu=R&V{VBg7z#c1APXE&&o@Y78 zx)_Y$MY+gGr{$m=u4UgrcA5A`zceo+#|M2_bkH@zOaLqKA|}J2>-?&7Wl>wSu~^rh zc+@+4K_MV{R-n{ff-54iSJBMd+eCZBy%U=1;mZ&i<*HTKkX9p0f=u3(8*`lm_FF-H zZw{U*+4SK4Az!0Okd4ok+s<^s0@5dM3q^8vWkJq=*KH-C{$AR#{?+&9zzRxLr5aY( z=Ym1NH2nl}Mq(S3lGls0_PpCv%Ege=WiIP-TdYN~heHKS^4P#oSN>K+ezjb=&0+v+ zcrcdsb5+=*Z4O$kgIeCoh&=AusiilQ`DW24(4VWi)ex8ZMc)`F;(~(XM{=0o0NOL; z#PeIN_k=t{7eDAz^IyquZ%JcHsX3}+Ko1#&w#U8Lx+qjHFA`Da4}aIY7&w6-uYh;esr_f-CGBs=iW znF%Xy9kpi&nzpKm=r>ST1m|j#MONf@|HbA=;j~=^c}c6~1WzlKBluey&og+9VKO|% z$zqY_N^+p3N+zngvzetdw5#-Y2+Uro6gS-aE=>bIp#-?D9pd|FTuD`_8xI?&&|lPC z=&>rsaUs7%LlKgU?d_S_1`WVz2Gf)S7QuME1Lk%D6=4G`2VI(M?d(r6O-2539yfo1 z>A*wI{4(7xpi`v`;RlQWl%K;`q$DS4TDYmWUwVQjVt>lb^H#1pri?}8{H_4%tLnNV zhb`?pGYoJ*X+$6CaQv7Jz}&X5`H)x8##DndlGT@3I$VJC8X2${w>bZ*)V*Wum++9$ zpyUEjrpA*1oApVvPOH_d)N?RV)>^S%kP#pDl2l$8v~`1DXz($MTmA|Ky*fvzm^sbj zfG+3bPzPoBjKC8{EKEoO!ZEFJ17~0)1?IL55Mn_Rv}HttZ;5bFMBczn5DnQQcSL&d z2z%TId?vJ32yC)U@pS&je66%r5^nkX;K1cq#%)fN1MOox?l|gnvaepnAi__&ryAA%vottEi7!)dfaa_GI~)o~ zxQ4Y_?S0}cCmL(45G(5I^{dO31H>X<%O=XuNGi7IdaJp(Nv`j0zl=#xD*10DD@^2Em5Nmw#k~3Bl_Z-%56R;q&FoWLd=Z5x3xG(#H#YfB zz`cAG>uxTVAJ<(SqC`o!L?#q?56z5(=%x1+a^2al#3blh!`IC^^ZQaGtW}sPMTBJ@ z_T{_I4zJl)7|fUhD-N*QT@xNQKWK;8NOR^-m6m*bk01ZSb%-TrjtFJId9`tjS1Im_ zKC2}^Q75`h8RC#wkVd0O?t)1;PRZX9PQq?2qHF(fAF#g#OMQ~@SLVYv#1mFAqeGAJ zrk&^o3f*K_C%Gb{da?^LLY_J+u5KSN@}n!neAq^s%J}rjkHc_Ryzjnzp-vu;oR*-> zMMN_-n1~xruUj^xqG+qeMV6^%h#=&o;*guYO;KjsRq^RRaOa>ai*7HA=*Yy68*vpr zAcs=M{_dI>QbnjU1iq|#uZ74r^Nb!e2Z=LaD@$N1!dSJEkX@$^i8wGR7E*!J8?dzagiOhiPGOyD2{;3~ zM`0qQ3}H_g@Pu(S@|^}dFt_B2J1{G+4wd?JoCD*6tT07ldkWu#JRL<(58wkl#0OoO9Zpy7UXv(`KE_A2hjc(2Y{@ zxJfydSQUZIzT35fGENurXt8%?6?~?feQAjQ9Ev-<&-Tfk_*_H!RQ3H>8ueu}m_@_s zlsy1t4n}akXwUJ$8yoOg-dM~R%}wB16;*u_U89n7Z6UXfBeZG5OSc~#UZX$tZ_EpB zz-AZIU%6A>dgrL zIia>@O>e)SCy}%J4{c6fSaJF-IQC$Wpy(8}d`jRZF?VgB5P5t|$^yDHI#UO#WtQI63x_NeEp-#E4B6{Me{zo*p3_@m0=QKo>h( z*Elv4bST_(ynOU{!MwZF(H!YT{4DIAdzWo>?!38(E1m1TGw0m!^6z@Q>s@whh?H&Z zye-4G#(NQ>Q&Ign)un0v<8}Ig>u~>lbarIm_>F}9Cisgs3Nbs(^)!w9fKz?WRbdcq z@6yA2zEVUz1aO{5k>)y41W%RCs6?D4m zX*X+2Vkbcg;_lpopLBzuITb$$B`bw|{YY^Vpu2 zk)OdjBIzQJpJd^b-8h~WqE1y72a<;sxAg?NHIk*4m@A8J`-ZS=FBY=4_S6bg?kyBG|v|85(}Nk1fILM87`1$A|sCm!YQ z(IHkZOKmQMK%LEYe~%W5rD>UTGlaXdmP{2^4Gl z^Rr)lPp$SEy>?*hFa`X14k?0%@GM@dr1$IxIC`VT@*=c$-@xQTcx%v(K}P*AVD!=< zPXi2_?$;aA+_nY2NOG2^FCNVuLHv7k0GDz|lN+D~MXRf#^vDy5?Yy)@XW&-PQ)|ky z7QKM7vjugz?y{rmLo0RKzY?4@wmhQr_|tjOx#`^&nFhNGwF1so)F%R{!84J41A$Ge zUt;3W3SWA+-{NHKDzgZ!%SeVg@L35q$;Kaqj)DJmOhSH;PrG|`T=;F6u}3wHp5%5U>e5cC33)^mJ} zp!c+Pbp~|sU+UKRI2RPs#a1!fzwt|zEyeS-0yDrF{4|^t52%ALCQTn;W90))`D6J0 z5+nA>^aZl=1({P&Hb=h>CaMN(u75T~tEKSqoouVCgKxut#BmwhHvO?zJgM2nHD)SL#J~>bTe_tFQwmc)4 zDTmu6VL-Z-)t(h=&=rLnO{=-K;Iu82Oo1b|F=V}Pa#xofu6;;Q*Ex>)Y~P1 z64E}f#kaCp`!*y3?6X{fk#Bg>^42XB=#kg~6WSVc$#Lg6-t-4B7AhW)BT965f#jaF zH3yCe%)AlO^}Lw#;P<8J%+63;PysT>w%gw`Xd5dCH>JT<-wdbjF&-J%Mukh#$ioXE zOJk=HFedYO&Yq1-8$>NeN?9Vyh1>kC{J{dC{A-Lp?d?X-*(Vqy`rH@5G z@WTYAU$gr)g~=g$ehOVSib7<2bhEUI0n=59MgCeT@( zPZKWBLq!B2{tjq?zu5Z+mcj23-U$_le3XHd2^x4P?8|6x<<_Z6eej|)_9Q*PkwTA zWv3Q)yy+qxEsMo*V| zK=am?E?)t>FTR4H=Gpxx#f$>|O@A+y(EB^!A$PcV{jolU%rICeA;Vn>6qZFYS31+<^vVw8iUzUY4ga|i<#TM7l+%`A88pFXiGL?T78IlyEae_RxY0;}4 zG9*+5A+IdGYJ8jhL{{^edodX5Rj1hmcI0QA3b@x{TMmpW(oh~0NHPIAwbehCBcRB} zQ)_a#Ub$dw3eEjPBHvr7NWWSITUX2Qy)!?*x0Cc2Ww}FphV$_ym(+x{Q?EtVjro*F zMX$E2>qR@+M?8xmw*Hk?#v zM+sWs=FxeI36^rY=<^mGyZ<}~B+Lr1;lJ-84d^9;)~p2x7;-smD!DV^85Dvo2rs}( z`XX71N8Uhj%aTZgQ|`!A4KsDi0CV<;ba?WXzjo9xjQHClVAfQWVFDpqhK{~rnNW`G z>-W#9!M$OX9s119a=?&A|C26FjgONCim$aK53M3%{{t(jfBcejk9j@S?lM_&$&6VrV3t3gPCb6^3aicC~kmWTy|8-m*F z$qZ;&0{eU+Ng&Mt~I7hB1WG!>VS zLH4y^>B^U0iMOJE^lA`gA~+pn>~#SGmeJ{fqNVEo0S~0&q~%xim+R$D#-}F^l`JiB za~V4quEWj|e4q4>kbL^HS?5DiQacJ1hpaYzcLyDsOUdo=GeC6)8^+dDh<$?e+>p#F#dv05d9K4VfY)-E zY%&M9$n?C$s`zZo@YypREg`^+qE6%Hn9Y?t(hPXb(rB7Ezn-+xHPuET~eqy?m)KeSq@YHNe5;EdU6Hyk%Lw< z9^h5OqMJH;%4>fmn^TcwHiHHC3Z z1d_n$Z%7l0{XAbkA^b-r^Jb`OlI5s44){Tz4V=`f_RF!Z`lXYx7~%qwoAR7NAb!3DA0b+xD==pk7a?&V2Zv!@JeNG%TUc!x%v5T2 zXR2p!0YB`c-i4zXmu^rzQOIx?`OwGswV9>~sgCy)-5QbTvMAJpoW14hV0u*OF$$zn z5&V#$=IunwiJ=2PiW)co_FBMlS(vUh&fU}$gav}g8U*gpQ@}p9^tJ}{b{b65w{jf5 zA%cV=ou{ACW=@(llpj@r#Q^#bE7uRsZRdg^o!W9iN#OGrmGS%Tg^o)=H)o6Jh0uv0 z=LmlKU9u=)aQMIaWh0(>3!{L~csOLE=o|9jW&@&h<9yXr0d^4o*||Lj3!e^W6Z^2q zVg1_GE3%_)b%I~g6F#u8kufUw=kPHa_c1e?_&w25z`EtrCWgz6OUYW|Id*?Yi%Ca^ z%w?j0VBKC9pEwu3w>1;=D3q+B_!bTKjNxLVJ^Ayc;^QuT!ri;@CEB21Ksn}jtP+HW zM#6u8dO%%^zgQ(Vk`kILngG4TLmVRXxjWd~+b6oB749x64Vyb@*~vY4;nt4A>T*zy&w}<#^G>D8Um!EW z?kQ=vRAssTVJ+6zm16*Me&)7MDMDZckw`Fx%qEcP;5l1v^Wji)_Fr{*9>#9n2w%Th zSvX9E_qWs8;9lV1b5O8Sa&0Eoo&n}OyL`PDRh~T&V(g(F-kBO(P*ag~j&QU)?D%)r zgKZLI!_Of(pbR^6i3kg#Dqx7w`+||H z1Pj%gQztZND|auB2yIEu+RR-ogmKRZIt_i12N*JAW4>_upsLC3>jEj*;*@=I1f2|I zuw9<$TeUQkd@w@PW0DwHI>7gSLm~&fD2k~?Rz#+g36BbJ9PqSyFLlFrtTLW1oswuv zS5KvM8rq52k-?PD^6J3Zwh;9YVJ92O6U8dz>h}j8%K9r&f+q*hEp?WT&1fvQ{rRWF zkgy*aTJNuc=FnSO^n7X$W(^nS7!WYv%6hw>jSUN%)MU~W_3reIrmQy!zHn#Pk?wtB zkvJ%#%8sOc*zA`ljn%;;%dBf}eYP_@I7g@x1%SQigu^b-se{x5ali_)9Sp-FM^#1k zK>Mj`9%B1toz!o+-yB3xtvN@hN(xRl5(W@c3^Em@JYP(LXs3*UZE`{RcKW2@+p`b8 ze@D<4*&Z?r7rigvO0)q`Ej1sr+dV=_2o3MDtlj5G*aHNyB{gK=pwQk4dIussunXzz z4W3}dDLW$hvq!dn(KwHBRRIS1oMaC*=(__To>IiM&2T;zz{LD3bT&o6puG z7oY9}lC!Ct&sz^zhFn>4oy*gfZSZTc5;bRP6oo8fP0Vk^%U^FRe{EgaA@q|M+>Q6 zitTm2?vBr^3&t{txDnN}taX0^7j2GsP+D>xnW6Nru=W*RFZ_LnJJdfUqD%S?m21i& z|89cQhH>~~2kU*3J3apr?p zWeTcbY#+-Sm-+ZC*d7PMzRRu3EX7&% zF`Kk3i_)-v1OJbHQ=3$>F6It9X3OZ?W9E~@Bnb(V}Q?o_4 zdtjf)=O!UzKvU7h{+|64pniYd&`B2_5R(+=&-J$uIi!+0z?mm5BZEKB=#A$BS5KcG#W+gC1Sp_F0q3H%R2>I1v{||u z=g0M+e@ixGf8!D+3o+K8D@Og&bHYP|x**x9t!5t345(;|&Rwp~M`)zXNOCePT4g|%t?(wnct-2!Gf1V@ zwKg+m2(Y`nt^lfp*15M@r|Hm{;FycuKp?2w#5D^)oM3!+^csCg376u zq}d}6CYqo2!kwq&#^52=f+aMz1rYpGX?`{%)&^%BjB35srmCrUGPj#e9qdrJ$^)pD znEzSTda1sz7-U}!BYy^dqLu|c-MwyGLFMca2K}Oz_nWGM4qKJ*I)*F@5bAdDU+G^9 zTq5j+aImkCE0*?`FZ^?QdHsBSnL7Vj)tHLxNaU-R|K*T6I1OgK!7IdAoePsCOX}`o zQNUX{`5Z z;^|y(6T@?Z1kJC{@FrRsQK)zRsR>Z%!|rEp0iDD5M`G(95z6QcTM$d zVlu{uIh(;clWR)Kn3JRmHc&m?<(gc#?5?KXLNRY3eKy;kzWt5!Twb91_IsxvFf=cl zDr6opaj=)jVAH?;;MiTQ-k!&pj(vF@*<$TbW(_cA`rDNW1<_kFkP{pm25U5DyE z_kjWKwqiAHUwzHUwIL|_4!%Lub!B4+`RH||#$w}56>>BZnN6SZ9A@Xwb?CyWN0U|O z=SVBxLnrcuNDBx8dfNwTz`?U$%&Y<4jSe3-&y8iiqkumw}= zOPj%RZR{drmvy*`D($K%DIHjw&OYM(Z~Y~~#eU60{Z;jo28{m-ag-y|zMmK->{ z+T=UB`dy;JP_4@`AZipH$;=rkSG_;z5oh~G%PJwD)*M7FJ(x!@;|?L+ZdUd$OD`uc zwgZo6ZeOA*R>afm{mg|5;>clqnkta&bPMV}2PEy(Z`k6y-A9a)ale=2XfM};e^Ek6 z{f%6XA$jsH>fThEvnO<*=D3C8@2I8A06RX`>j)dHMZjkBk#lmS%*vHDLTl=mfrSIX z;dC8ohmF;9DP@ae|s$7zAk~$QGmo;zw zoZaZZhq9vEBppkVltrl6QL+*KWN9sh76F$(HDY1e(i)b#DSmA&JKR9&O>)!fTgT9H zwCsp|Lgdn`{X9e*A}o-Q@ghUpO)ob#_AgO@xOe~H-xoW5_-X(d5$c^ZxYyiY^HiNH z1KMyKvc)j1l0C=N^kd!Cn$5EC94DyaF(_``*Tch7i>NNmE!N=;r*=)Sz*$LAnE0b8 zlZXAbE>1fCJk@Q+DLav2JflADbu_NBeFiz`@1fehIs0&sYz~xlA9K_XVp1+3_dYrE zp@Zni?k@%1ak)V^39&p{_Nr}QG@YfjTwq@6`{mEz0d%FNMfSGoME_6C9pvs|acV3N zEvct9QwTgn7!)JP z*AEqaRk1nQgAQgvM2fs_jtEcT^lwZB3;n!LY++~3Fgis=OKH3~psTh`FVxqi{*`B~=82Ib$coPztVBgb> zrKUJVDB>aYT{44(@P{&H1fa@2NJjB{#wVaN8@b_|9O=a?R_OQSIO9YyQ<25rxv&OA z6CwCQn)CtF-fn*p6j8{9Lo!X2nozb?KpPw^WMRBP9=n_DL2@N_Au&L-s==j$>d(K5 zr5ooE9dSi$3$6yW9tk6F#ClQ(rb$etJfYu^M1z!8L3vx&Z#6jtk_ILuc2Sj4P{ayv&eW98 zsB@>6=^_+)Kl1V z98ZECC|(}lCoa~U#*XARV9E6h(HeY5L?-^A=*xRxZ6y^i_@=5y!rtyMBKD&jgV$#? zY_objeBE0Ka6+j5ZKI>c5^ynwiI%YCK2aH=Qa-t5TPVcBas&=+R+GgE?_21yaf5+A z%E|1aQ^frUKty)NMnr)2KAQyL0kx$jr8FM&CzN=OO;jLMn$0l+jqRml<#Z6tTFwy> zNOo}IS|en?!xNCsxr{z?QAOkx(azg+A&C_JiaG*MtRX#{D|eNEGf4qb$?}Gf{&nBN zG780wu>Ky7Tb^UiCCg9P`w<8aJpU_D#OQBGzNelM6(B)C{=zAN3J^=txFW@ygvKR{ z`LN@86Gk-;{n|^JAX7H#`EfcUHQDuBv1`B;+>}Gm6z&5|f z#}m!A{}1An`hd_Gjt)>hQNTvy5Cit9q9GMWTec`h**TK&RWsbx+6i-P} zN2pKKE|hgJZ=5WU^iS=*6Vl&B$0KBv_TsNAWo6f`kW@prf*hFo{UaJ}T* zE@~CT2ZU}D=;B(vN=7Fs{UtC<-2SDfQXgvbZVj$@8^wjo^ONe0S3N8_R|S;*)*0~tLj5-%NjBW0UQ zBt>-w1-XEW8z{+Xp!vO$)KtqcZ|l${1{2shW2N@f@XuD`{9Z?YW-wx zh!U0ua;H919+eV`eKbVS%1t)(FsHQq+#K*b!~s@+C-!g~1=j`V<2-cI8Om&e!!x7r zv0gp4r21|gAu+&np(%Yqc%|JmVrcwQw zs^yd(?{PF4^tkFiri}VE1d=GwZmywej-z0;^s?*I>{B!uV)m9$R?T~C>Z}w#EO~HH z2ya$c_zz{oC+1vmti~75jNGRp1QP;CqVCZqncA)71ulHWXTRx%C-OOh0=V)_xqI`( zOT75lXUvj*b+2O4i0Jbff;-$Hlh*(<&(bI~1+;!FSVzK)SYyZbDBhPx)@DB*bKSsW z)2xJ&yYn+ez|8OIUVmP>KP|9@3*o$_{0?`Je;rsVca7p){aj_(%wKXU-;GdkMBk;M z)gN<;+j+yc(jCWv+u3!yC(bQ|oUjn<-xse+p;mEkX*;+B+v*ZRCLQe%`%qnN-GI_h z0&j1@1liESS?msS5Ryz#8cf4suiz)e670yiOPWXAj~uH zP#p-Fy>H~jpN|rV=6PK7p~081_N4s-Um-n6!Z=V&-5~K#eso>^L`cFLxrtJ~4+}dd zSnz=bpOd`6=DYD1<;Hzz60drfGAU=vpZomj;Ths7=TSJ`zqG;N$5n1r)Q4mww3VxM zY97Fuhyr4zV{e`~prC}cchvURW9w9wj*W_Q{xn2pdiPvFskhl&0uKqoZuI+?14Au7 zgq8a{+*tu(q9`0!Q~(lKad3b@7dgBwJZPE^wEezl@e#?Fny1oWcMW$Xeh~4wWuEH$SDb`Cu|5Ro-1+@F@H7Z-D?gT4od$OA*QTNG z@?$rllop=q{F6OlC=&)1x8&}oia!~YEo2(Z>Zt>v>3c@dbAg9H7>?3`<>#I8z*Kmw z+4qaX{<)h%-0KgODrAyXZ3A^P(;p!_l?XHT<)g34h~oFaeI4H`^9Dqfn{B%OMsJj@ zC&K7sF?Q2}BMysB0YUa8LVg6`*9>cy3)U!2srg2;Yh!6IKy<@oV*^?ju2lex6hZGNU zqcMEdtUtFS`+J7`p`st!4m+)w6u|r2nef`@r}dXUts@kJJTbTc+R->+o#4X#s3_eQ zZ7n$*ASUW?7tYcP-|rRuTOjcdg17z7nF$!*qrZ_AGE8Gwzqq-7r!5HIf(M~_GKFM^ z@>DLw*c~g+VnN|<2I0J0I-%?j5J@w`LcxG6eF_D6>$PxChCu{Wfn&1Vm9tRVm~*y5 zz{5dUxdt)R$@jelqFRwzP06aw&F__sho%U#={KBcblN-8Wk0XTuv?)0yjd;QpnrIxVjDSTIG3GqUc=e{^VpJ|d%PUZ8?solP1B2c_(QqHdtCZ6K zVH_hz2EGlyUR3tFaiQPN&d$sLB7EzK4I*jr3|CdTAwOL{5!;S@za>_Ev~6!ET$)Q& zhYR`wA$tOBOry@g`|}{}?R>V8FQP$y{|0wHVn%MUM8Q3X!JJqyCrH2+&v#9^iOxhp z96$ai{!;am_P^B9D&4qRS)fUDbvU79$J;__$A_4!W#lC@kv#BITbZ(k{u{+A^MZ!qjAex!` zMIuc!JmL8)=#}8EIPU{ADl2s%idiEDVZlnVPo`)MzwRf@J4hQqUdazpmiF^`b0Lqt z!o!oPzpP(~I!J_o>Co`K`;@m>S>7spQL8zFsx;CyK9~i)DjIdU*EsFT_^XE8_Y>rR?gPoD2(eOyDO$7$)nVrNxDxmm1w zlN*P=6b9DfE`%0WB1RyhgFW69$wmgfr(0LvEQ#N6B@hm(%>z|@@p9O)u_vP@w?C@Q zP1Dn%p}jjtt;uSJELkq_E%*Jg3jpAUZm?AI0 z-S_q-o3lY#BdpPHYda5q?~-jK?9xZ@mabD$ds8Kot?Jd*S$G8U$?;_q73`*)-INNX<$DEldi~8A@FAKqTFXW<-#%9t<2Y zZlItT>J+6w0^|nfJ^={g&=%&AU?iPDs7%I0z+ZL-kg>2($X^=KrOF~nv`98mk&EkB zjtNAv+IJpIy|tqx2eAgj`iQAHdq061wqnuldfUAl zT|kDZf(9hT%wZyRk>DZF$tt_3?!oOOz`P(_5jD)RCGcx|G8O|JoGiyWX%)r&4H|S# zk=mZN4W+okAciZ<)w>O!y2RreR+$Q<7Tc3@*n{@F=v{P=L9`VL(r#G(0LT3uFxxqHaTAtj)Ki+cew8-6#-E4Y4gW5rj@guu!HiN6doZ~C^tr`llBRseEK z+T&DvmM^50-u*!l(6g#sFm;97U*aW1aI+f=kJ<1cOulY(R*FH@EOd?o6-kLHAuzU^ z|04GyOMP$Pn|l<5xO_w<6eev@2yXt%GNw_Fo4JGA$uKuGw5VI)$qm42Suy`AQ*}VqJ zAYZs#9-m)4Jh<@qycA$5lrVz{J45_tuQSma?qN!N+%E?&^Ns-PHx<3ynbqZFJTW;I zBd+$^lf#}889;{Wm1i-;LE*WYI$3e%ln?5=@edZHkG>_O_;V=$D!pD6QYoM?7rF{L zbE*)z{|^Pn<;0~GqetkeBvqVRcnUgrWkq!TM-I=3snU8y=BKfeR(!}G+s)FzNNo-^ z7ZT90;OF)h7s^!TeO|B10~K)|j6(a#Ts!K(OxiBmqcby!*)V1)U4) zvlUlpS>)gIm$-kAd)*AbpXg*3uJ*IQ%cyneINa`|>2YMNr567Aj8H12lpFD%dBr=B zC;1%Uo)KpB&^*LY$@4IjzqOEsVlkRFr#secA^>^mwoYN1y3kPH4tsffN4{qQJNMZN zO*6_*Ua%2=*MjEgyph-8F2|Vg$ZGrvT(d2-uYcu!jym-FP>_NW>3lzB$_wR!nR)_l zpCviXqw}$wwXVwF23|66es-f7M;b+UM?#t21+BrB%a;&bZUU%WC%EfgnW28u?C+ei(rl*)n0k1p(x0*zCYlQvNEYl8(n==$5wsz8^}GO^q>Cn|{gN4hxRk3Zzw zj1VzRF9v|+Hq1Pg?_x`Pkh4p>y-F*7Er7T*Oo5<{xZ7%4TmE~vMJ}9)6NE|~?s#^jEHn}yZ7Z@x z)QYQKuz}hjt+kA)XdfXkDBe+x9Q}A`a4K%R@czp=%0?nQAO3lpdFnxy{;)Iws}$|K zsd7mrYbYkcnRR2XJX1Q`&o3d8zf<@qi|jEuTxo|MGR=qsaD3REE+kk+E#l|<*% z9>S7Mb@f*7!;M|-ZMsBps?u7J5}$U=0OsDkVdw8dqz}Q}yam|(?guT;EtuXO&dMw5cEP7c! zl<*Xg=-Lw=Zq8%pc<7z9Y1SLIpK>37G1UC0ep#J)j|Pm(Ov$gcjIiwB_gqU!FK*JQ=lS)Gp_jGe>dl4x_BHiJJ-i170Er z)8#sqNUA6;O(10#+YNo3 z$!sWI?Z4@Q#=NjZ&ig)`@|qRd9J#RXw;ta_d*GdB(~6eQBFQ=uQMf*m-o2C0a?7eZ z0>g_Cg`?vWtK0U=)x+10d1;E;ZL5LZR$5q)EH31vng8IT4iiX)U%g8(;+B#D{zQ}? z_d5^h*VjKx|7BB-qb0a(?RR0co!|lcOBRw4VOf#E2mmyOuNgE+9R{G64Gq7ICu~&M z>~WBGh8^Zsi4WYvg9G`HN&A)A88&n9aqdvcSPRylweUwY8!B)?(<2=ykTIwFhIR{7 z8_PJV)}X#(@8PIE^e|rJqq|Ws&X8B#S@2px$L}t-0Or5J1O?b#m6mwbhEJK?xK`}+ zaGN2vFqW@6#t@6C>YAt3mFc9>tzMHSdJP+xMwPF7XC_j(Ri<9mK^Yx{0Y&b$a?c+G5a|K}pVWh}XcGXn$m zaV}bUaxtz>M+LbEPHa#fV3mzL*kTrhD>d4u$YhKNcv*$#&f&p+F@9C^B0OKnnThO( z<{vrP3ApFT(>2>sGk*uxn|q3jqfVA6Xe9tRTY?rHLTE_F#TXAWFec|+yC)H?0rQ?#>zyk?r0 z7}u8;|Hk%z>_J_`E4yR*Gh@4O>|ckRG+zxouoqTL$pVz;=I0b|T@rav5J^2!xEaN( zU#f<@<{`LEyi?<_n+OkGe1T}7`tSrzC;)4ENvyj7;Y@^n>6z-QeqiG0)6_FAoPdx> zKtlndp}WM7Pk)kENxI8DH)4xRSK<}0(Y89G-q!Jym7XgbU&H>;(Q<$or%_hR)VYqsyb;zo1C$jQ)^mr^0LZKHa{`*mtVGAi!i^|fr#6~K5p`h9&1 zb_%vG3)^yJWhibGPDyDnn-$z=q+;OnQJ&@=zg*&;2SO9kAgTR*|DhHWdi-!dMhbM3 zPWJ}+e1xDvG8u#uJw8DD()~l&Yeae@iSpzX8KTPecTe`KJH10$#xifFt5;J-5rNuv zS;?Dn%)y*C1p>daL}Ft+%yYhWj=~VTj^)Y!#Qy&=b`H_81?swuZQHhO+sTe?+uX5j zXUDc}+qP}z<=oqRr*Zq$tWj RK|0!||G^ z5#d7E@J?gwW*O?+XQ$6bu~*!svextdgk@KzK_#WO?;c5)$K)4Ea>mHn(HaW)lipt# zJF=zv3}s+}BBa!f(YA6)8)MpZMtoc14x&ST3q*SPnAn}`(>Y5|?7<*9{zaR`q~Nm~ zxpSH1;*s)U}5|DOelyhZql^6P^2T?LyHG6@jm{+`*@W@@0WPL zqbr`LSv@I9mFsEOZBq?iZk|rD<5g|6gP}|NgnP8VG(Q1BdtyQenP*3P_3t7EoYhv* zcGdGYwLuX8uT~pK2@FKRic(KP=hJ!=U#%~GRm2%u{{@_IGW`GGgq89CB0~Od;{Pu= zY4n=4f#*RZ3o~7DOSMHAsw<0^i5dM}lyl_B(2q2S1~~OHJ_EWUnkq;r) zw)McyefRshU?I>V*DmPbX4jWbPjQ?c9$!yy2Cg4e{I@=KO*GXbo^I}*O+k5+ zZG$R$7wbEzM&hGL1yE8tG5%csj)m<1#41_QP8PF&W0lF9xEH_lr1tZeqG!(1?e;VI z^~1cf35`B8b1n_Me9&;M8f^>QS20Pg6Ac_4UFX4;h0dWOP(`lssk|+UR;YB4JEyQg zkY%f0XHqX#%*dipd4IRq%B+l915)F<%1)&WrsnL_`=o~Zv~1l6{>-l>a0yK@@7IRB z)V*1q4=I@cN?e;>>+&M0BTkEA*7^@UK?S;+qrQ_0ng8a&(sI_6qSw`5ZjDom1lqSp zwM}yW?jgZ;*5I7;#=+P7e9)U3;q?sC03)7mv+-Bya>uF}viL)PA#GdOcH!xpt|>{L zgK}}PfaroNMR{(@q~u;Yn3NMAq$c~PP$$*1ma2%R_@E&sE2fbM7qZ~`e65l&AXOnj zDm*b$)oY9H0@an2Z%BgNZr;RV@a&2;rNo4Cnv8foZQg>>hR)7xDj^haVi)=#PobXX zFRa68?)s853)}VqFfH8ayaA?MmV8HXH%{)7UB)9`#g7Augm2LR$c881UsU~f!ODBL zh!bO;6x2ozuA*lX$N>CLd*>W1RU|09fO3$N3$!` z`(gP8$V$T+p7u3Us>#J~o- zJaf%O(@Y_A)naH|&qR!kN-?QF2DLo)ID4&4*QBTml@Wo=nkHow)n~AI18*GBo{X$I z9Oacug};3^=M%L9JpJ%rI#0?_V4p8mX zaJ~y^%#aPUD__JzTQIT;zPq7v6)}8lI7bUQMym-(HBMazd2R#QoOpyn+nnF61x4DG zM}2`(qy$)j1YZsZNpVl-RqK!`WFUb3LHj|GXAHFB$w8< z!($3Zpc-9pK;Q4WK(r7{tdnH)FJXWJ&7BX6qmc-cgX*5g^-$!325)Z598(d9=~NBV zqZ&fZItqP7G&O19Q>3S^9_z}Kcau0=fpHoy2RF}p40LfX&Heaout^7k(!k~9#hlSp zbBV;`cRX7Rck*R8AS65lXk1aV?TUt;#0_8Y-xUW$$Uxm#shTH@2U1=(DG=mMtDN}@ zL^Am+DZ5JdtV6IrrY@{WnFRWCc>dwIKU@s)VE+~h@BRpyEbALWt~Sc+BRyite0J#@ zb{c!N!^;X_u3^AAY3YTNH*8q>7QH=F_C$bOd9S;+*nb?dj$Y~cVyGH-72K3Z=x&c( z@!Zr5DYENTJGQ9qdk!Z8G1Cm0CcDph>ofet8RR7p!LpH+*UVvgdV2M*QZ|eUCC3;v zA5Mz&gan)JD9TUijgX}!hi;)kBTu2;mkuYABhnZ=u;|imN?fD~Gx70Z28PcAN2J0W zG$hy5Ua|8^Gf!b>eSt6Uy?tTq)BBtcA2Z~$SS6q)gT(dHLCrIKcwPM)9nO+rrfi%v zDV&>|UA@z3R*$I{5cjygE*4B6@3!^P84{czdgp!eDPJ?&_vwge29NgWOI>Rr0%0nF zZll47hRdu5_Iyz&vo5)uHULTs6$X~1yL86cOTfha4p8gORX%#yX^1BH3S_Y>ttkz; z~tF$zDvQ>3}tKR7Fb<5#90fB`jUQ6Yc+!kI#j}lp(3ZMPYP~PZ6f4Y2O+eWWPMLmq~N<0Y@F^kjyH<6ngN^ z?XjV(LvjYO1f|iV00_LHLKM8Spfm(OS@qBq6*S2n;%fh|1R$p=wXY3X#j_6lt?WA( zgeTMBk_`_Gw?!lVZZadA5V^2iI6&|yeE|Ej6PF!%d=Te$Dc>G2EWnD^5xf2l{fp7P zWiE{=4DFPIHI{=gOr1ru(By~RLyAF&Lwvnp#m^XhKGl`zN?{r#w)5KIAN#4TEB=hN z%02tkf3)sBYpN4)nb?A5x{HcEyKa2O+Oq@t)F_Jd>8RVflj}!G4=SI$&i3it@CCYV znFK|T2}sbD=e?yvT6jR~3EuOK-Qgt3p8oqiSQ%7n`Y+;FsEj=?wSVF?rEBzS7{+d& zP$D1IlU@IftC=Gx-BBNhRnPP%Tpu3529E7zXVC17FvE z0xJn{Pq7$?EJhJaEQs!CM>UXl)eZosKZ{LfPzwZV%1-45a0b?VA7U8{VD zP|_UfQ9z#KgA|d6O-DKauP<+jG54WBLza^lEZyBfGUJ~S4?0L(Co-&?47);;J4 z7>J($GQhs0)WQ4^JOA)8ex6ynJh=P$HMa(!%TOJ3nbV(pJTQtf8{H7JI~p1$lu(1U{kcY?WU z1pl5}Ne$g_%0X}=s1wZD9H~klmW$;cRv||@nGo_1jn1?-^!F$vB@h%n(hz+Yyhn-t zZ<=7WWgO)v9-%*ik6!Pqd|Glu+}%``s`A#n05m16%@eS_T7-Ml3O4U|7V?Av`opsV zxOXqBvj57l93Db`URJ+%w|bob9(gzg=j$o}DOCqu>{N#57$(91zWh4sPJBZHr2)JW zI&A>*T%<9|_~2;ou<;H_excNQSIaf%gCwy2h}wdyhV@2y-Agdo&c;6A;a4)sy?t^= zPtv}T$`egX5z?kjqiV>bXa^2|4S6K#fv5Gg;ZbnMv;v-9u@ZDoI1Hofd7r*u{~wnX z+Y#@y98ITqCnT#wc!9*~92i@>bJk>EZbuODG*KT`XAKgq>Ln}(7{v$7AoG2>1;H@c zA@{K_@n<`vT;iD68hG13kr!MG->e4mQ`Az|Q$H8y%^xzaHc(|t{tjN%%wybp?0fkq zx7E=eJms^Yh3p4EF>og;f_;*aoXdId{@Hv9FbRqoY~0&{tqp0d8U$m9=U&G z`Z(DKnu~Wg9c0KkIbwNBqjxv02F{yfrg`c8_w8S0O3UC#V?XZS0_XTBB+FpG)q+`k z2QXW7yw;Y7t}Y8oKS*v1Tg{&nqj-ea$<&20@5E2GFb)5jw!xczH2=8r_ZS;Aq=u?d z0oI|jTQRx39b+@OsEilAX4T58scIvf(9vWriCq(+rLEE1(~ayq}8h}*;Cp>NcnaN=OsjePr`f2NWw%*3~+o|EzDstoNEAk@=~z3fIvMaSf3m7VS|J_)k|VVa)ECZ%A?9`xLvg$mkFxh z6rNFF!SZ1}Ll)t1sGfc#0SEgycat;|VSG5CnB&Z&1+BkiCn(suO9_FESi0kbF!j~I znB6w_>X@kO@CS3Zi9ayO@OU{-RDA>y-%STuoXtvP?U+&#{mo*H|I|Ics10yn8^W`S zY6!40#@v@O!^7A|$&LeJ&}Uj7r59Xr2_s#%!pSa-QmNO`nT#Wo@JUOUya_yE#ojNo2BfydjuS-WRq81_UG_^#x0SiI8h6(JB9! zqW-UpNX*i(2b1`zhtUyR%=Cg_%8^hFx+Q=lG!G^r<~3}51Z&Amu%4z6|6ZgFg`AO+ z8CyPIW#0s`=I-_P^l=NX8b4QAvVml&YG+s&3Ty7BQCCij&qyeb_PhT|m4{!#pu|YygZm7 zWM^tyh)hUJQ4kGaE*;E#YInqQS4hy%CEIn&vaWUwaSgcvfT}g$0}fNTFF^@V7d@56)jXx~ z#aC*vbif3*swHM^h+s0;r(yTd9|7_+{||j7L#tN!;-9lNe|nggz1%Zn@9oyqwVqyU z6`tLR-Xh@}y~1Oh#4u7Zshb{b4JD58_@;(;!{x)7z|uc@4Wb0q7nsla0Ka~IPAm$* z-1XDv+};Sqyg~K8?oQ39(nKxqzGU^zS?KLMX-P}%zj}K{?A-cw`@GvsY! z0Eihcz3@2;zWV+$$cwc#o~hK^tFL^8l7kxbf{Q)n;1GGeqYLg(A!SMSw;C__U?K9X zWGmM8vkVjFn^+`KSZy&7GI3Ocg`xHzMQ1E1Z;&`q+KX%9>O?@FMcl38S~|4!U6a<< zwkSx4{en9Y$`bN&ZwhWTni(+CQk9iQs+_$OV6{(0`!f|c%%kCvZ&^(Eko(@l2yGmW z>Tu;dLb04X?vJrKCEJ4m@{=Cjo)>fT4z~4#&HvmYHR2PC zoh4r0k6oKsS4_t~%L%hi76h4Q+s^oeA3PsdOK(RxdRpL;@li5Rv&F%;75f^;*FIsU z=F7hR;ls=OVkT+(QA!?klb$FTkoIskl^;H5= z?|VpVQz85l$BEdqI>FyBL4I66{g^vGlbR@%dbpD)N*AR<>1Cx&rTPSRXRxIE5GfHz zw`}rls`+C1UQNn{gr)j`|5n0^i2fkjfYSJ z1?HSUPnwHhbj;K+H%t<}O)<>wHL7UCJ&?m9xg*$SI-={d0ri9X$Z(_(; zytTQc!uKYS9aQ7f)XT}q`KrZxyBQhrD#PYd@wYhXbx*cBbubXN^z6K9B&kf@yum_> ziop07vyl8JT|A6e_s16PnlRP=xLHL{fUd$unA{Q5u4#6epU7ES**_lUv}o@uA#fqy z$M=P*=UG*+TE38Z=`!G@O?C*N1P^AnG|Q+@sIVlLdRR{suFTAjE*>t*7DIxlDBf#( zM3PlKc;?7QKk`$w!%j*cptbRci2Bh3HpWi|>D;LYJ~_fXL4&!Il;W2VWNhoABo00K zV#Bi;d8JoGmthHhC_y6hepP!cI8S*e_@=ycOgw6t8s>@A#!saqFd`|meqKP+9?s5u zx}5v~jtds{{WtQ+`2Ukf7FLG;TaEL7$z$~fp8ZCD(Cp>=^>TC}E0c|4fv6`h^75bv zvU$8sfO#vR1oc<;mx)u>03O?Od#F#k>K7nF;fS;@n_)kAfX>pAYWkb5Z?Ye5Q8-S~ zqu!^Rr&nA|HM}ox+@~|jPx^LD3_U#Djs@>{;wB|~`=>p^cHBQ2d6nwL(GH%ulZCa=O-U1}w&$6^rNBk+k5M(~NXGa^%^AH!09ny=Z^y z2bQ!P@^Wx?SxU+sPDJrL*QO#bH`JN#kK&mbB;PJ~oaN_^`Z>f%b6g5guNp zw5$}#99!!aLd&C4BlqJ>pOtn0j(v9}JH*SzvGXKjIzfnJX3wTPyR^GFcq%h2)lkPw zeUnj)D2vBbJzD!A;p9F1;r`{NKsMSDfTr0}EeyDOC$ucwz4Aey^n4I$M7us3)@S*g z%DxeTtyyJD1)$MPau2B&_e(-hiZPNbVPKCHC1K$!OKy@r+Tr@sH4L7-_H~eCKEJFt zN=m9$$u5j+#>gh(N;~vBK^BYQ%&@mOK(Kw`#!57Q=C*6db#|riphb$ZD}rB?!0%Tb zJk#ddbh&sa@1Rw5am=PwXIneij}70jOdL37orn;on?r>N+QLdeh^Qe+AEmp{AwPID zwHBwM{Qd(>OQpVKl#urT6{_1H^W9$j!{7i!b(mzj?=-Zpj1;C7OPbzQw@)Ucf6M$R z5K)MKr~JF(W+dxpPRv)MD*!)^z!77fNm)y4I^;O@0ks1lI$d*7thuvUz0a(05L$e! zgD@-OrS108_%Oz0>)$3&dC{6a8PTwB1pe*V6a{V<`|&Q&IV5IVQ8R=U#rwUYv*Ao2 zC4)nNhWx4jSC+?Jt_g|@h+E}obsk-nEbJ`}G`h8vzj8otkca%~%5eZ`bpV;O*1Ut% zK!z$dMz{z4?c)wQyXZQ7x)3)CFop{=??y4JK+xwFiWozoKpQQOqoM+RwL1t7&QCHz zRkoNw6;5+6wuH@4i4*2fYmn0ZH0Q>bo>_v4>`ELisZjY@OAr%TA&{AD2l(gGH#H@p z=lP{yPkM;!op^>ds;R+{wF(L^De9uZ_`V)Dw1bm6Lkj~zUcU!n3vbPaU6%jyo(Cs> zhFf~QvJdLQGNOgnNX}ciy8Ht4D|=qXRnI^sg6KOa_Um~9Rf4O$-#EO^6MMJBUXjrG z?s9Ig?qXXQq(r0RVqtp7Y;udWUwO0mz(eL<;YNN@?^G^J;#P3 zYF_5f?jQ;!MtJ~6O1JNUF;*H37SqJZNBMq(QDi^3Pj^J}=fm1lX(sUGZnzF(0u!{j z;^c}j#F!<#WObg6+E}g6S5OujPK|~~qWsjlHV9N1pAy$M2@(hDU4hL=IwdALCDKCN z+{9Z$6R>^IZpf4=7*ySrK@226@R1qM`LeMR$>p`)nCC9sgLIWO(ewv)dxnZGNAc9h z)zg8KOzJ*v3^6zB4G!@ccQ|CZ@&)*QQLNrG3`!+tV@&=N!CpV78M`T-Gg6!z-pDW!E zodIaA9jXk}JP_3wei)Z}$};B!JwrU)Z6}6=!R^_${zP%s%?WuHqf?ds4MWIupqY{1 zIDCDU&1->nO~3u1a8{%+&I)#~(d}+2rj60Su_ zcLBNe#vuGsNo_`oaVss@jyJzYv`uC45L1?*=a?Eo-WB3arJW`WR$hL91ki6mQK{3W z2Y3R*H*B_kU7Y?9V1An>T#v*?O$V`!?tFkdsbq02elBF4Vb7U)iP zU9@G__wAaYa4WP{ia9`v3ue%}lT{l8mY=THsh$H~>|dqJ*Nf{cQM7RXhNjQ%{d3oW z>BG$tPP}o?v#v@S8__6I1xNC_!vfV8IaLk>ss}vOVCZD0n;_$L45kxEGI3HYn9R)S z@AcFTsHD16eS6w%U<* z6J5p7gSn+cJZ-*0x9Q#1L9x2Gv7ER!)hNy=VSIV4hnW49U*+*q{d??5A80(w-;RBQ zHM(J7^cIAHhvgdH(=3C_@YL+GEOgk8$xV(vg_0fJNRKf;0^5>FE8Mx3Nek*NowHPj1P3`xLXCK!aIoG1R78TxIJ$OtbSv1>mrUf$^N88z!` z%Z>+IG*SBtY z`bevK#cyPpGZY-I$&lQzNE6`jWZ0<|N~iS}6ISYzKa=#7$&cg5TrdM~>`7$EcM3tD zqzVCvpPY$KZ)Ik!ZRWLJM{y zY*i-PXy!DzV;D3zr z2a1lM@p-bPv^k`EwB=Gbe23*1yR`3&hQPdH+>JSH zE#j5K&bZcEkY5QlmZbsY#h|tSNxPAh>tTFNnyRqe5k`ahZzTe%D*@d+j{0k7ZECTU zWttQW+XzYH%ZE!Y68lR$ff>1WOM#U;Q*c_R5I)>(_qv|=XggysWTNEoqSPfrlTI~< zeeI96%JxxuHFZf+DZyJU`po+!Hv-a4n@-6=wzp~xUyF>JSfH!JlivZwKkXTrPai{K zco-wF{=EXDIav_7qU5G#C25PHxa2RpdQr@LSo2?wn()|5S&ou#gBocZ4pexWGgWxL zzF4GP!kl?q_CM=KZkclKWK%r-s9s@Kcb!8Wz+mKVI&28b^l^bC5#2uj4A10GfNF$T zRlgtpVylht@)SU1z6DKTc)$=UfFW<%&}y%VKHw@uW<33!uH<&FPpT(uX@i$%Yrqxg z->d|M_<{t!xHDK=GTG;@1~b9x^Z7tiY6Mrn;OqbWu0gHsjn~85T4{5s@DD8GD&N}y zKs#P+VeUoXVrdku_bK*b3USTgYa?Zp2AO#aAQK5=-z9mB8GbmZvVCFKu!Z@(v`{sO zNOMXIIa;bwR{x1hA62L!c>CzqHE6|*->LdDh}fW$V~>0bBN?T{eXIK?*WAqQDO~2* z#)pW6%E3=-KMWKa+rxhOd%EQul?}d4==TDQ6TL&2!5a{cz#ZtW`lX%SSeg_sY3za@^TB>vQu*tV#~ zRJ-4OXwW>M^_{%5DOnL)?j1IDhns2+n}&Ogm+l=kK;4$8`iA*k+UV>@#bs*hj;MB+ z0Oa}6nAbR`+dsAP^#O>bxDtfes5oWyax2ir66{{$%{`_>Hv2kk0Og6BsJ)E9>i8CJI`buW)r%9Rl*4ALWUQSrN{37|8CLl>Xe zocKR4O)viht>vJp3j`&6bj1aUNnVk_n(2H9I1=pj>S{O?(=J?-_!P3I3eHVcQ<`NcY?AcQh1e{E)K-jOiIe zL8=U^@(u;v1ggxuwj3QEd2fPMi!xImKSk(0Yp+6XncA&mWsQHR8pdX-1qgGT& zYfp}!Tvb)r|24>Nt?=1tQkEEmk-65o+@(Tkh9$aC$QCo#@FS6vUt_#m;sNp5c~;+8 zX(}lkhw11lt^u-Yrw>v#r_W7adc;#RgH<(SMLUNw|CPOzkKUPU}X-<6(W|MOt! z5}5jmobdtRFD(Y76!H;Nf<9t{iW^j<(oKp>HGdANYOGA39K4%*{;o%b+!~8m7YuW^ zmUrH$iYKPFoE%DuVNt;+v%wTdGW)5@Tg+FPubX?cu~ER%A3XE)Lc!e^S^>$V{xfJksT z=HN$w;Mr>#S{^gJ`8qaas8~bX1G-DW>HhoM))8sreH6ROU1C`eNR*CFGnBPC+u*zx zmr^VgIW{!5F%kKIA2V$Mao$f&-M%#(6$x{#Igz8`SYzdJD|wT&I$n}bRe!g)oMlC#Ou)aA4Pl##o5LvWc7ss+Z32Yo?kXQQ=@`4+u>)BM(r+Bk z0K9Q4zUDh+y6w-l3P(`RIWwEq*OBS`4otx$8pf1ogD$acvLh0&cIvEzTVpY#f)+8G z4^WHC{ADVr0dt(kO zvMWf`_%JLcz;`}tDRmwgUR|EoDWi8@ZpLZS+McUyg*;Kx=4ZdtSq8Ngzk|@gA5ix% z^zHw~d6@ox&cn*c$@G6y^EMq$u*+`&rv3 zz}3P3bR~GZqWEOjiptT&!RcP|R3Qmgwsm%J+4UM#)FmlbuuN(r`VZ$p1K&Tq8gnBT zKZtxeO@e(lPd{v%n7!G*7vOna_leXK;Z@U6It)fyxT`^@cMF+FkkLfN ziiF1c59iVTAI_tp{_Tj<^UHZUe?&ida()%PjdNF?y>Q35%5#?m-+lOB&cp4%lHI(t z<({ywh_1w`n0j$O-$D_gy?IBJO)IigeQx?@%ailJhMo;CyzrRz=%d3=o|1{X{0KU@ zA6Ku$+bMD3&ptqhY^Ok#tTv-2&gvqXWOWDoL*RqgjKsc3Er!{1PP|-VC&g)$FJjNA z&{0neIa**b6SA2)tFJ_bC{s?_q@j~$*7u*JXvC$Hv9U+H9}S;jw#p!$LPYV+<$T&^ z?qPtkvBuT7R;W=Bp|6zrE+g&a`iMx81NVW6nKwe_gP z<22M=KeYJCdia6ajIOfZ)P$bJap( z4x9P0#D|g^Aw7t$(fjbh7L`~=-)bnMp+OU}+}1*N25+U5{Anz1%NuY`ix-Tz;c)#? z#gt=~VD`L;V&A{Ck`|YQ&*~g#r9&2=nD62E0E=D2mP#>) zlnIFFFIWz`9DLUwQR>C(agJHT&AmnJes&czpAwc?pB4HUc>g?O6!13&%ylFX%}PsLASu{Vv7GM=a#LKP<~vRXBEXca!o1Wt(* z7k>{VJMlMw$ixw@wKpHf9Cezcb81LJ<4jhwkY+|O+l^~e^=H!F5}G(dOAwP={3QRr zNG}DYwH+1R6Og3e_`RlQ_$c<`xJrb%`BddXeSr~O4m z&8z*h3+fO_A@i5oI*~vp!%e6Qn}6<%xT^gH%e;j|eHjMg&{CPgs#xjnU}M291df#7Iv?BH8;QkF#5!*`{1x0JGM?Y^_)Se^G6N~_f*uMeAVTX%x2 zv2^yR%ToP<5}WpUpNF~IyE@;Y(iD3G7fp%H!aHFMz%r8jExiF*xVrGoU#p-5gj{p{)TX#wXJnzemzr-5sggVjV4M*&e1$hofCDeg>VjL-pc?^|S>y20&e1W0BI zn*BGM0ObfKCZ;NLmLEKMmoGIdUqv3Uj_IEge6^k>M%)&G3`6wJHG~NuL=9WuZ6Zb8 zSun;o`$VF=5$N(Dd9Qd4KL|Fg{c3Jj$s;zV?`sL`SB0cnKLb4IR~8exjcXTjFFMF? zGb!2fUHGms#Lu^=1PACFR2`k~FLZ;P5W;aYvQo$xKuE{Xk+~-#?JN|)tUx8AaNs!I zHl9@y2v$ReaqDPGD$K<=1WKA2BuZPKk`BN88W0nNa$>WB!VBx!5n1TP+*9W!DuClw zKuByxqWntve1dw2CM5FBQY**>XU$h%h7<|*!yQAG1>2z1^%l8qYQq#$z89-lBa+m# zo=1U)ye8@Y^(Pt>F%0M)thr{1opj!Q_rDPQ#(>aUT(o9ZxdAWtk_Iii+C;Q8t!zVV za*@84CTp8oHIlAEJw3YQBT-;9i%d&jdBXmBcsdIz%=blpl6QXg^hGj@Wzg>hQ+7DP zP@^A2sfYnPx9=8c6J6iaIt~D;cn6(lyH17S3-G7vY{+!tfpE6)7HOz%iP z$Ffg;A+o%vzsO;{dcH}FjqrZGxIb2Td*&mTM8<=*j7M({JVZOmJz*GkhnyxZ_dkOt zUWKAx(R>Qu$-BP|-2OgD&vl*aa0qu|>FQRTY%hqf%-B~a>(9DFX2U{&Cc|1B)XxSZ zJ^45*);-rQN05O7*1daib>G^@|K;cZaRgYXzG zN*)}S1-C=T7ECAiu(7`I#DFOM7GA8;eQl3Kf>>?Aw8VIT<4XBoFw?h`M|=el!O9?# zE#Nfd%u(RrnD6RCqe(iaFfUQTf7i-1=)b!QC02SE_8YA$Cdk5QChJJ+EoGGfzm%!R zJ?_(@{GF%?g`eE@LZB|O`l8QiYdPDyvWaPmBuG>FYK5H(7l``PJ%s?{_jjwUc+5|U zRb|frN9$)SUE0Xy^zKjS>$N)SgDw~d(r+Zm{> zsA9MGuIeUV{OZzt-i;pLb-I4l>ztE`J52cclz!YL29zx@C4{u@7jlje#agB3!JfGUMnGtPr+IcEUuBtBUm#bKKnT3VY|?n3Xl@!)9$cE`gfw>#=|VCWCgH5hhqGU5gLggCmNmRGz{_}& z{~(&3uZB9)J*o6DtX-c+U;ra|boX2)y%m++!9)1IuLs?pEVq^yWUcphp36EoR7{5! z9DC)>y(opwEmczu%=b@xQM+!h-2|~{0HRzn-7(TRUPy0dfw%jHkd*Cf>xS4kq>TZZ zbB%yOm`%EN52=$S)12&~cuW{y;~nMac(4-%z9Zz8{CXAd(fdU&dV=akhd%lt;Pr>x z9P~lRW^BjdZ8)EB4e$#rR=Od~(%0*jvFdWi31-K1Qd!)i6<142Fe_LTL*o`nV~Cl- z&^Weaa|hC@I?4wb9NJ*680SroG-4lLb-_YmMV2~H-pUQ_cG;!yj8&&BaI+p+x8rLn zG+N3tO`==v?A!|=NJ6p(&~#nGolZB+=+FFFGM&h_eX&M_DcfF!HGELp`gP0w6=n7e zabEsP3+uH@(<~p8;fqPJkcuE~nrH`1JQ9zDwA2cK09T&joG=d|1Yq3QQx*q^3QF;(Ao zzBQ&DnC|ez?`icx3kXQ#tQEEA7TB6#zoR4 zMN1%0zZm_#IO;4GX`-S@TsCK9h@^)N!GJ}7(`RT9Y@08FfOA{@%6P`qKj=UZ;-Ko7 z9`O1EV{E^ZCrM`}AndP_PizIN|9bJs(|Y?0_-Cbm++C3;JgRTx+Y++0U~(z0)+y-p zt^F=}y@BK#ZG`>Gr9pRLtu^T>0$#J>0+Y=f4)(ChALCC0lWVWX=cbFmXK-RwUF+0M z9pITF(;Fd$nf7)HKo5PULIruS;nr+YZJ(M2;I|v$bVWzqku`)6P>#Sy#JVtHK$3b| z`1Zl8b#8JXj>bVqVOq(cE-#~*@%7rAH?u?wg4T*QLn2V1u|wNAUcKB?r1r2C3Vhuo z-xaI6LJ;m1G1|XJ;(rZ;H34u0^^X*|t8*`^u(pVuFntP3y8JPuhy!bJHvRr#btKKg zbYYRbd`7(r2N0GKT?)hxc$Sv{mU0OgR}l-QF3b<=RmrVo;b?OOEa4g0bNOD*x2}E3 z@}iKAwK?byj!go$(mtbeTPjz+3#q5pCT;~Ah|G%Bam|nUPUG=T-fgHz4_dD5VOO^( z!==klKF*@13L%>|qUzhVC_#PX$xMqr6UM$Nr0EXC(W^X8Il5l1`135a0b2kcxAt7w z2Fj9Mm5Hv*vUI$d|A5S0URj&y+7%2ICAgCh%P^PQ5x}phUD|L>=CVk_#)aMq|JBDG z0}$M_ckq%41RZXlY~t~)Mr8Ib0>Do=BEtwV(?UQbI&N;b%>*zk+U-_&CdG3)Q`6;LEYh?hTSL8-fkXn{_*iCOH-B!yRYU=dOc4dN1;?Em4mie#@UAz zPds?&;sj;LPP=ud*fl;@xy1#a$)aGB$MtgjruDLvMs;*O?4Ii#;%<2&J)w!lUk;{; z&L2lz{vjX(9+J%s_Co#*p5#$|T(W|4XYEc5c4jP$#&fu!yR5>7(9cIanEH-*FR`=k zmq(}PhGbCYc6c34JgHD@HwHFxwR=Q-m;SsRPrX(6p6ANT;e|R7q_8^=Z&w7v!Xh(} zf1HxMrF%<|!w$K7@Hw;a0UHBJ85hUZX*k5nkEs|cKZW{8WpTfv{FDY$Z4wK0)q*ia zZkJ%7rQ*CM+2e=%MEqnOWkVB&q*R=WXQ9-=ejvVG%?qK7tTkd-2*{Z70Kqi~Z8PZ~ zSFhf=66zUi1RAzOOmAYcy4OHu52+wMihYZg=>6h*uDp2X5sHZx!g-SEQ+h_GjlJu1 zChGuxSl?W>yn?|4XxEbTdw?9Q`ZYveMPG5W zJu>e*f&Ijk2Av?}XGq#;a9r>vGx+L7up?GGs>*6ZS)?qV^N?h|@ZbIB6Ukt<;%(DC zY`E}m>OULj!dtLuykWjnpz5vD2s3O>3gKrT4qyf?1y2*S36zNjA^u@h}i;m&NTj{Jr9p)UI zwWjd;+wIMy$;VHxK9F##6Z@EhUC*Th@wd_VN`i}ZZ{#wkj_P9uSvK2;fMwz)w$vbP zt2jq}))#ZxBayh8ccyK|&(2wNJ9QLFIYCKDq*qetfxaKlvwRzj-~wp;0wN|6V5&Fj zO(%xi%nWh)6R22MY}(WxjmGRNX5pjq_&BWT$pu=LxjTzs(&7bc8Gxls=46pCJJt%$ zQkeJ<`mcoZ>m0%_HV?b!c@33$cdth>QxaXi;oc@l-s!qQONFZQtP<{7yY*w2EbWf6 zZ}hp_T-v59%4rT0X_{i0Y(B962HyDmJ;`*JmKy0mR6WnyxJ#MG?S_{@A}??~msQCf+tK?; zW*}NSlF?&Z7<-d-l76Y6eW-BC&6NpwVqZEkg*yT{Job<=e|qYo(;1h1nkF0LGBLkX=yWX~73^n@5`P&OqubJT2JSvvHV^^?tq=KqpA z)O6?fQLD0ZcXVjmp(#Zw*y8v7Hs?VgnIqh!wI zzy3KsxGb%}A-F#ug1+-+t^um)9dodt%xy1W$sO+RR1+$p8&*zd@L1>Oz1aV{s6=?C zSMn9j6UD^OJl{*;>Hn&`Di8felc(ZPv}(3GU&)iqv|zGA3vH5rRd}lEuAG2uqpEMv zwmx1MXl^g}=WXb%`B||s^-?qR?=gE&w`v&JH!72D}NRs)# zAV~&hhW`taly<^l|DTq?Fqk~{DWqH3Efqk!Vm*x_@qFW`hkFsEc%6n}Y6-qGKZ&xR zyZkq}QyM+8fQw7RSp1*$0kJ3_1(!WV0ixfQ0CS%n+=LKb%A?=A8w)Q5d=0!WXWZvA z!xwV~9$y!S`={8&fI^UzYAE^*{62)osc`;ZrO;}q;O6IBLr;R39UMO|Zdg;p9ew%~(UnOY?HY}VEXj{TDLdA_rAnCF3lEXM zcV`TV{GSkKJsiEZkrWD6kyJQx@YWW>doC}gj`qGcn37o#Xv*nbW`z^|p#o-)YMse_ zqdl-8Xo)HF;i;fGvZUL$J!o3c`%m6o=B7W)BkeY&v$Ekg%1*zVTz{kICkW1mECph9 zv2oTFA~e`V78|H@Foi2yclW>=v>dwMN6r@tDd*>;wzS%n6hKkyiEwH%M3X#4`$HD> zW+YS68s-gEq<+e#ILXa zrqO@nkwb0YwOg17W15@#a3rOTgDsc+%gmJIF8{-EW3qTaJEn!Tr>e_mgY_&3#F1mL zMR$uXcdPr6fO;d!+@Z8KBxIO??~}KzS!@`2Lc4l4zr%xCY%A?f zG7f8Xo`hAsaVo(t(pBtdUM%|cJA|Do|8DB9TaAA78A01iW|v1~Hb8!qQHv<{)#8 zhHMPjQ@{$`74}tTNy>A%(d=1QWj*!@61_42PoE|qYUd*Svt5Z)m~g)*0XRIiYZ@kafkK} zJtOh1@EI05s9*}Z@ftwim{ZlqSIFMNLUOTz55hv2-dJGdXX*YBBPLNWU&ZD1UAfQX zWuB4Vy~x+18o)XwPmjx>1K?~hE)c;Bw^H-yNKiB6g0B_bSl_mdc6F8qR?zT|b{OC6 zFh6`4avb?t?|!lU{@bOlR2bAuP~cyFLH*AUiWp=c6912}b8OOt36O2uwr$(C?P=S# zZClgEv^j0t_S3d)-if3cy*|Nbe5QQV(z0!8r+z!ZGNu*+0dK;SENJCPp zqN6pT+raVc4S7h4&k$38j$5_~8s)~|;?3!6@1Z00LX8YO|)Lf7A$iXrReE0ykVZAaVmLc**M3LN#_1gz2#_eiK z?d6gIv<=&f4eVl?eFm+Kk2hnTS&3cB5iSW%78w&MC8!YZW;5tnR_pLt!-OjFtr@5X zFSxE5??!6}4507mP@Sm!ocU;-uV;n@jF)u{Ch&T4aC>o(Yk`)w0p5Hcp?c8a3Ckfk z5`7db?`>Z&i2>biNL8|akD*k{==y-Gmf;k~5di!QR+iHUs5%gvPfLLv;wKM@7<1mf z&=`?|FXXk+HhdnHBf}Bxb?JIk0O##tM37yExWA3jt*r(5rCRF!MOKeXn8cD^8kc#o zyPUzb7{%9C{)=W(=-I|o^0~>(c|X> zJPV;RsLLK50qKYWQ-^wBp3|95!5qb6pAaj9mRs=$a6Gb->nM3{pyv@!ys9qbjHhHa z05&21R%{;rFESu()M|*sV<~cbRZgV}@GRXttVeefWFowN;m~<9UAD|4s$i80o~qq` zbT57t05NBXFmh*bhlHJcCtyW`iu#UboAPz*^{#dLPiZx%A1_L1jFTufoZ_Rjtjdgx zWRXh?(H33YV(?0I#*Ep+UwxJoNP(q!x8t}vO9)XxDJrWJV4Y7!NR?vqj=#CR2J=PC z43P=Hxz5E?(j`4(rU-iGZbfm>J2y3#zqnJ2kQm9aiJb{qsfIzV1E>VvxY;kz`(>}_1nv_A) zn&N$so{U3Zx~>Q~-=|`cG^l)D`TpY2uROQSyLZ5(xNLqcjX>Rn`AWSXwlh4EKYu#~ zDUM>X3~dE$+qvEEaMfUhRAv$_3dAC{S!ridHeMe5Ms7Q;kMt#p3sgdrx#B)ju<{}q z$W5%1hBPhL33X)H(5r+-hXclBK;(lJf{%8Ile)g#L3q%a?O6&1yHVP7-L&QO#Im2Y z3{MW2>+)skDGnLv+-o2t-Z8_k#*M;t3a&Eny86G5B*|*eG5r=*7OfD=Yhsw17s|S9 zioiE-zDr)SZ?*#6jLlG?W6K!y=DoyBx4)Hc-5nbfn#mk2%v`5n@~*4ASU~~JY5dLE zB(a$e*ST5M1V6gXf;((c0wN%NyET!>F+s{H!BpoDy$|*%$|XQlV9! z&)EKD4yvgzv%%MJv?RIfSQrg3xN-a)g5ZAS^!Uj*3$RLVfA%2on#kcK{hTxQoDm03 zf_y&L^xZXymmJ+;mD!cVe&r*nzr&Yo5R~MBw4-IBBUjK%nJtA-+VR-47m_`McHqu% zH?-d)?w(w5A%cj^1rDpjVpEHliP2CqF+M{arwtCsr?OFuMR_O{A|QUxP!lL-FdZ%K`c`mR@!XZ)xe4Y>+()c&)43H*_Bm@TV|dAQ zKB+^mb4->Ti!s)mSJ#bNROa0wpVee#Bf2t;AIvaP zPP;O0HHSTN@a?`LBVeQR90v60&tVT9H_T~%V%1pFZcghJbbJkIH#k5moxp_JE8bQf zN^C>S(|WnQNqyA9R_MaWMaY5MW-M{w);RO(lx87U#c7;X9*Vv)u%q1h04MI{Ixm+} zJDD>RYot}{WP8?|PQTX>E+;jP?VhR@`v6 z6qIt~1hGc$XIQsk^_o?A4l)}fTyP{3_;i#l#jPoFW8Mc(878Ezd61j%2d2VLo0=l~u(NJBt-9R~E(f z@PzZpc4Q2#ywQtuDD#C8U&0RFoPu_N$x95 zh{hr0;H|)gWohI+iBT)D#BL$0JOd4=WlF-LDv)rA)?;JiJw?R;u1PRZk-J~*&?ND2 z-X2v<*-$SbX95FV3gy?sUFWdO4x;5BnXOI&y=f zVfRTGoz|Sxv;A@p4|urKnzU{4FmqIHuSFuI)N)qu`ByM$j*2Q<#FeAoN6G`4%io&l^q^q%D8-%SQc;6AgLvbfF>jpdTJL_+tTDdbWx`j1sik<` zJMmdBl!3byS=CZba!~@c3qb-k4xO7s2Hi{zE#%*+Jo6G9-&T3$L84yW-%rlnZ8Ii+ zRF>oB_pM^v&-%=}-{+nnH~`#;aqE{Ed`m2Ldu(NKP+u9D7fY!qxl6?i?>HX$ah!zv zTGHsrSU0<`tGVYTzo-hjbN)sNzVx_l0FLtkM{TkMzEejVILg~?+c^RX?TY^X%cQp7T}mK%a%zn0YSburLE*ZI8Y@W5h_CUJRZt z-gSp^49$VGURg*Tv(aT4lS2=_XopwM;$zs9!^qvJ<)aoSRsy#A88;9wXVy0gQk1zh zMVv_&*5xQxx@LsfCvwN8LHs()*!F^MD`_G7$Y1d#>*1OCy?i|XLhpxo*sj&`N6=Z* ztv$Jin^?_Xr>}W9L3CR!ZZfczupYY?oLR%er25@#NfvY=iDX9Xlcqfxv$scwWjdpIisr=+$n@TmnAOV z^x5|sW>m#f^V?Q=GZ}`WABfvwMY6PkLjS7n13LDB5GW zKfhU4wD5*x-gZ)N4QJ&2iuDB&9Mbly#Bv3GZI1A`vv}f?sJS#xI>49VTaDy zqdifJWDt86THY#_l|q%nkyh(n{^2>J;5V%!-1_nEcBr^D&Q*cAe=&G9`s+U zJJ&ia(GA%HSg>fky=o8Us^A1IBuy|FpuA`RTZw{ICKOrcmA-e;c@5FhDbrDz!}biGlew1M@eEzM_A-{J5EQay z%W^K8xImAi-X#D|i<9!lNtUnoi#wrzSuJNHbgrh1(Zxg!d$ox2ytxZ>Q z&YxKBRzJEh7HfGTDBtX5kkB+h7cVF0HU}Bk^@w68vzAe|uU?b7tJL>@_W}6YO6FfX zbtYebM9l$P4fq=e1t%WP_9gN#6<&dcY6X17+Y$hb8q8l$nY2PZCTHY1Obda~}QcM@075t96GlvcQ*34b{-w|c`{0Z+f z33T^7S6yTVImyfv6s0A+4=m|_bCa0~ zP5j1DLcnfMr z&UymUGVT1OAAjWnSJ2qlNOG{(}P(=O)nch0WQ%%`a2pWsiT| z`8$tnClq+%QDSWgE2DyG-$W&@MAq-x%$f6j@+znJO!quA$;r`Kb}?&-Dl2BOL8eL0 zb5N5ZhgK%<+5ac=(~C_*hMaI>+HwkUBIT)NqvnFqmv~z4o|?P2vug5Vo@-F^D6Cmp z{c^$zYtJul^MR!1PFmZc>gjyZk_|&gD=XLG+#0zY#fzk|={jbrYW z&i>nj(i6ajwUoK^zIYTh9;@iG{N2da?&X{S<_;Fq;O&^8mVUgwNZ=$36# zxvS{%cw-s-3&mewtqQyMTFaE<@1ZRSCoZA>KYj1SRil z?gIz7!vK>0e!~so5YqXJUyQn=;OZszyNo(`Z2Qloz}`aI!mbd5vE~dZ`-Vl1htYz8 zkW?0r>(9nGtJk(%nPGcm*B;nymrawH-kW#Vl@HS(gdtB-gQIWa?o=T|AKLkj(Ku!+F?)igyW!81r0d; zS8Pt8%rMOf30sZ%YAm5IfDH7^En)16H~sP&GfuxXG-JJ(wim1!LuodK(MNv+Wo?a` zqy@@E7l+sXo9J1gCoN2t?X5gLfj%Wu$N7kpO<|+;PB#bEXUmE^9UIe?F3)_Bc8Vks zcFR}*mI)@Z)Eh*iH9hfOjeY2&CQD#lGEVw3)^$MOgZIZ}x)uY}jo?;J%3Ka86|Lz3 zNH>1{QTuggL}Y^4F$eUYS@a`<3hKZM7V#!aL)MVW9s;pH*Lnw?}3 zbWh*BIGtx~k!UCQyM23yzVa%!l&+n!Pv4fe_j4_wc1ZC_5XZPi0iR=pCRB%>Yh)k| zU;8{Cej^&)EMTOU%?S4w%@mO9rbFlqn?A}<0_N`UVY~?uAgos^I3}txK%Hd~D!LqT zdiI?Y1Tta#w+{rDNMFW7`d(Dcg}UrM*?CBTGg>72?GVs?YZKa5l9N3Q-f|z2c$*lJIw=oJ#5iPre3w_6ShY(N(SIb-Z}&DI1wINt zgtY}p_=btg2Jwj=_emZbMI`uhYpHci+Al{HY+ns*&h;Mmr|T12kk=e5BzIZtiM~sg z$bc#g6;0K7Y|Wro#8DN+&;?*1p|-gYjWz`kY8AEL7QH5i7BI^y-M;qdWFMzZf!ydg zHbt`zF)Li?Qg)cshQEm$j8uS(&`Usa>n`VOt)1l3Z%XNP2yf@}YSpKwUM%THaGCxo z_Y%Os6bg3D$UM|@!Ianv#dR)lCvc-+8>w7t0%7t(+EOMc6E&7RV(QU|U={Ry4Q4$%?LCtJg37oVGjCvRi) zZ9ZA<{=gJ6+Ro`$lM#+f7k~TLOF|Tz$+%(>9FJ-Hl_uE|@>6D2R~^(#b0^D~Ui30a zJM(JJ-j8POZ`kYmO_YlCDm7A8{t?`3(N>wKgkskpA?I25ms z?|lmAb70XkHSL|*e72K$d#MzE1C2#zCrnmSJEIxxK}8j^oiui{IJM{df>=gUcgimR z51wNs_cn=L>uQmsgu-hdrJtY`il)OWQS=1@h)wLb`jgNbQLMhXAFCbxNUlR#kua(m zt3bJL7yVlE!3Ny*PAw^Q+TT#tu5w$!{Qu%wOtK+yL zDTV;EF?yW8d&v*p3p0Wa){4YxsBut!k>^TlMMDdnWrpEHR(dxAb(rW{Bj)ne0QJdp zhuCYQaVQ_!od+mI=<0%82PW0!3S~VG5Bx=Ou09^qk`X0)oTR=eW5`%)-B@Fmd87ge z$q|cn^pngySK~1i3Fz+yi*gBI#(lH3kLm1T=Qj105_FdijzSwqG1^$_->fx2(Hi_R zXn)n^GUSV-BwF-})(&^rl$tnW4CCfrd?ZSc_?x?hwxwt4TnM~bD@6WkAlc}x9=T7VVk0;|-tNfT zSSleWBasUIVIGkzMHb^hibp5V*_O)co-!m3*~BD5%Cd}7_mz)QgpujHS50y1J41sU z`A~Jtm;{T90F-kVjqNqgL~Tpi*J)hrV%F^=NY8g-bGWx#3A-xSMUCu1%ndIp*|PT2 zYw(EJQvTMtqOxld-UHKp1F{ku`XU1hFk=^tv0>0)X*pT86%ZKH9iU30BkfkfrXi`M z{x>KgNB-~RUovM>A7dUvPG5eOCP+z?>ZpQb&r1OF_~y|8`h?0}*JuEXAZ(zVAh>=d z)HKq}e5pulvRU5Fz`_ZlFlJE524Z5&U*s zSq}XJBu2`*P^yPNHZl5(J|3;XIi7S1@U!LUm!jYN33x5i!cg|ofl~CwkhDzlo#i~L zZZd|IRCZ--d9c=w-4XV7+F~&efQ;tegv@F-2T=}EF%};V8@tofMbkc!)CWPCz_X!( zD6SL3QtijY1Q#LRIUvHO7t1tR|SvL6Bn0nOxW;K>j%C&`L+N zv-8p;)mstJ$`h@4#w$v`@^+dHTo!n}!{=uABQ~*nhxXuvQ>ICuY90KH`@J+<+VM(; zl#2ta``rPpO)XyUjahkw#&6$a>+zVVB`}MtT*y8;@Gxzz?9@l=KA+)zs_l`fV;7<7 zEayE(7mc*AuSz1v7!anDEcsS7{oBvzVUzP4zPPXa=_a$^ySCYi3JEi^YgGsZ1KYOU z?^JzK)!$d*r=m_;o0tvE|D3;7TKg`$9T(`&CGcao8O_Peoy)USPY-WQeo#ipPRwx8 z#!w!(S>z(kxo+(*cOK1F=Qlr}PDOo8fKeB5T87BW9fsPO{ZNrivX)*PF1~D(T9wqV zJzfAQcc#P^tU}hmYz~nCm_r&J;8fm_ywnuEb(!1nmTZR7W4I+&k2J_@S0F+G{lYMN z(vXoFuv?KU-_!ig2bu?&ZNNb0VaBL=!DGm5N8dJlBl4?KYifmkw;3vm zU1e=rRGCo{F11$D^v~vI>A~I}9>J$a`^bz6QJ)i9tuc5PbC!MKDclpr+a}Pb%;4KH&UplKX|Qypoy#9HOp}cg zO@WaZ7M8OG3TDPisrx9;ibysR5KHA=7tDn?sn%^q;)?lk#aP8MpYxP zyc=$=G}7H*`2yY5C}w1f7L5APq;F)*S@(fyQJ{c$3Yu1T)rJQa1S7i|^Vc_j7p>nb zvkVB#)RA?;sms>TNyA}y41!Vqdo9qB2b0(IU&e3ylm~4(JK6IN4%VAIs2rAIzgb66 z=uEk&?>qQ$X@hMC$g9fP=`dYBYlW7KPhmYC{LAFd)P3Is2C|d{1v>Hj5LLqN^v1I` zu}aJlnI2}Td(Po#jRG+6r(6M*?(z1+^D_3OX&(aNx-j52xzo_&k48}c5}zU@nzQyafvAJjWB+b?neJDGQ6P9@eLZkyuhKA61Z~P*mMZu0 zPQM}&8Ki?T;c3_>i^-=oy?ha`5)t8~(#oRA8%4v^O0hP&4V92|InWIqK|*E0bP~cm zabR?)o%xy0F{(1t-M;h)qBD8#kq<_CFB82*;Im9&jU6p6?Wltj-%)>&i_@*1FCP^duk&!Y@y9dC)*!-~>4wElk z{5tu2KHyA$_Kyf4MmX(6zVM7zpoIq>-GpXJ8 zOxBtFisXmePgLL}7^vAHfgnX}QI)RrPI^J~8BZ*IFk-YZ<|??&_DuV)8nL?AZ*6RhIQj{#PQUG61 ztP;J=%q;Fj8Zc!xpRcT|q7Rlm!9k20y71?*ZE1_Gx&NNlerZHuZG9?vRg9z>_2y5;^U*>0sGUs0|10s9zK`}iPWj|L9^sr=c?}~2 z(Yn=uhv6eYmFG~Y?b83SR<3|%8t_6De~n_mAz^tgjr{?4RLh-g@-tJ>0w>Fwkg$m<^=OL<9<9>O=HS_5~F@fh~i0n;;;I!@#{*jyem3ZDP4Lt=A$ zv1`blrk*xsBl@wwiJah`qT^Og;yCEL?GVc5uR5o6Rm$U;H{SC_$Ej0m{gW=_N`yf9 z5g{)fV1_MdvCW$)0_|KSDtbiZdwqH&+l3mwN#OBsfyZ5&H-Wq$i6WXIzHfJVe*Chf zZ!}HTYbbtC^T(Dy!TMH;JHO_Ho7N?u0CTd%NK73SU!kNIP`?z()C%V~*7RNUi+ima*Sx9C9(C*FN^oCht}N#;DOj52SnjdSQ|Y76ld# zv4|(M66_!~w&l@C@Rm$Jdy4i8Eym1Q@H`7K)o=DIn*Cfo*BSf5@IL0Mn4^1WYqf>a z5&T`yL-=&oXJ){Wq;2^hQs^=;fbuUrZv2_lu-=&y&8&Tb7YW&GX!87hb#l{FFPI;jFe#AaM0)UQg+UeQ}aqvV%4?fy*ViS zXhceGD;DNCT1f1vePFgfWroTkmKf*Zn5zPWOIQ=7Bp$PP;_>!gj@e|nFdEVEWG&nQ z*asM5okUy3oHhWnpMk!g|R<^3=K+-V`Z+AX_vq&HsyDvi}FYWa4D~-{XokzW~mg zT*-c`oi7!XA`T%NO7lZ;K%W%j8R?_R%jw2l-B(f2mC+1N9nqx}+U}G2O*<|54al8Y z(195!rQ3ODw;|C(VQsm5l&ru%&Y2XSOv3?tH^wfB1hokM+`Es@OrP4e2#U3cc>#&{ zbL0%D>w8ZppgX~YIvG%uLxtnIB!30Ew(;tl5G@#I<1JQy0wv7nB<_bcri3~#j(~aZ z&`SuG#!zbHY5d~ywYX6!oK~RSSw8?zfbMHK<2H;!md=rD*P_z;&wr9ywj%}A@hUBn z;l152#Y;hR7i#hkrOoN%N8-sA`E~mX9~vo@X@Q1=oScqVmYniiDtt_A+J_dccbpmR z=P-yYtI_BV>el7*b=w*ASuoz7s!6xg=wBm?#oxjuBJ;WW|N7vLK>X{2b@lHK2!<_D z21+$4Q~34>u9bSmW-sP`j!&lqPHV_ma%B`RQe{zY2rs9U&!`?*9{}6h!o^hFXFre- z)?41`HOTjx-B!aBFYP5RM_5oyrV8wc7Zscp_WEK7&7tYKU|r!M|K3~ z{g`by(@&%QyS34u?o5eYI^o{A5kVjH0U2{E{Y1@k&yF*TLslokZdZ~Og>EQ9lShWC zp;CsV;-Rlj>?t6Sby~bRptfsg?%iP-T%#^OvHk0yi^bWwLYUm^6YbSPb zh~XS3X7j04G#tpbAL}n~PRDfiY9m0MZkBMYzSI7w!7zvg_vwC9I5g7s=2Dru_JJwajjT3L$4fuuCwD%a4|eb4~) z2UUxjs;C&$SP5q>u((vhYW>4-681*x;4b-@pKLzF3y;`r1u{cyw-wkT%g`Ir*5_P4 z3WDnZJ@ujQKqQqjU{*|&Kht{)LaJHLKUPdyDF|@S@?gxL0#QUGjOplFw%GHPB(dnz zUuJ)zj-92IjwI@5Ao`wLmyLg=rK}yT6<j|UR_DK>jsJR+iArWXb*^Dz-bDw+@qCWdV4avA8d zBR2S2V-i2Hw=AR*!*~x}B*RCt=JrC3%Ju?2N+vO!ltq{Qr3>WB-lesNM{CbZ!koY1 z7h`ZcbXp`E?&2%C6Pys3lxvw2Ml|BOd3fgrKSPjhWB{BzrqwEIg6WkVGE5>sQ7~_q za`-z}hJ?Jp7e(EStZHr%XvA5jX_g80y*V6x2pp+8GS&=QUy5?h@)!M=GiiSi&AtE8 zV{VQXkd5*taMR9fIE@MaFbqk7Gz6Hd8Q(v>Wh00Xab2!|rZ~1h92K}qMSnVIywMY6 zGewwO+Z2dg&4m#S1d}-Ys^RjV?~_;DUZce3pryO{yf8Z z?zX5%TGc~eS(0{#S zaF7-2AiNrgL%z?ptaf&JehvcLeE|O`c~62Tdm_}$bGGHEWxT)7HZMcCk`N)HtJt<4 z7&PSU6WK}S=Gc@0xCghzhr9^*PAI9)7IQrhACyyWelmno1T##T>7MJ;w|`Xcy?4n= zir_VJa5i)}?SMVA+?yG6FkNv`(MNa%6)G5Wo+1B*X$c@1B*EX%Q8BcxQfW+IbkqNd zeS%6d5gc6QytJ!E?Dl^YXaW=IsNwWlW}oKE^y>{K0o~EB*L2!U?*{I1h_-sbD3BU) z2l46Bx4UTU-ZC7@Aif@F0Y4+L$hp^K5%ZO}Xs>EDB<)8JHv;ngiwh{yBx3Xd#f5

    )N@pHt7vp`8P_A#9YeI4&Dm;s6_k&n@g4a5d4!45OGa1Xr0M&E`q+yyN+25I%chVp0jgL^9P&0MdeV(K@!!`qwlkATK9l*xNsaQ(&@S zT$gVZbgYlC0x%13Yljc(o31LN88|I#nF|at+Z5_AMK)EmV8Y!~8JGdzNGXOlqa60u zW`LRdwYG^ro;}I-8r9ZHoz(y58-O1vEldFQ$WstwJ?Fl@N(mUgF-8K(J^hV&z9w#P zXE~vnJK7$bep7AjV_4TUOBQ&3;4fXbY)Z}I31_j_4PP4y*L?pCI>1Tp(vye1$xq$Y z+1eIY$zVnp^s3(%R<(hYUIdn*4nXI3CGXGw^_F2sOt)j}TP?F(V?l;(`Q+N!PF;nk zQ9@Q#es|$i9|ViOda%|1)SJ+~n%DPi#xAS&Z&SSncbs1j; zbC?Wf9M*U6CI&_BIDA2by6iOC8@(JMk1d>C?H$nAwyM%=V3-x*_|qFwlz4Z=4( zCC0Em^SF4oQWP)5VdeK|CM+q z@fEie9STG;L`;S^WUz6M8Jx5!;AFjb*F@?Fh4|1>tACUzD_ym)e7o39NJk{eqQFhX zSIcTGWQs?|vAe{2Ta};D-f2(;7|ZfO3;G#H6(8FJEvDz^VcEW$p?8WcoC!j2M~n5n zzv29)!2#$AJ+w!CUG4)G4M9v-e>RuY^U{|j*ruL}MPm>dBXETWa$t0qK)-_F{b3k` z?+ItXvo7#({Dw6^xz^r~i-@}u8pSZM`W!$71>!YWr7lxs9O(~d6W#UP5L(P+VXwBq z8cF=`@(WDG-?zr`Ry4`ET_lK9+0Zc-GzjV1T}dQ%ZJfj5i=H<2A+}g(9G>?sWYq>{ z0cA!(dhukn1eanGxWI{&V;C#89bEc#PmA7xU4u3p+^nQs{#w6|vxt(3L1rM<&qD{9 z0z@7qAai8^YsQv?fmx2aE{WYsOp|>KWTZxD9T4xpnv5wAMdR8aRr1{3a77UiLf(J_moESI-Gg-!KKB-8$?afSq4E_0 z<1dpWVn@pSZXs~8gu7pM9vL>+u69-17!v%h6sxR2e*oUwlD5!i=k_myJE}^njKX z2O^w&*t7>_GNcYjGPEqCcoD+IENnNRMr0dRw02M(YvvV^H?pb&U36B1{)RQI5^vaC z2CJ7t)exBFdEW4zAo=%FtVZMCYUms29YDtC^rR7URUlfCI#HwI zia%!NIe{qYAEyjf#RI>YKEB+4l2ht34m_W!ON8x0a5tju7P9Dq={2Gca=dn@XIWYz z3OqPu`hF~dOdk_u1~;M{(wpsbPbdL1`UAu!N22&25!+K+r`N;Z;Geqb5Xz zn|8lqh(xBC?)-egh5OanEC3^mc0p58Z6rW|C!lOcwHd&dGaxOwtky_!fp{Un3|sFs zz~`cZ*oaON2E|4IT|g+1R#X4M1s%T(M2S`ezNJY_Dj`%jm0be5{oa7eL9{LB>eySx zU&dFq%}cXVAqadoiJZsXxY98+6chlFX)B7zkSqQQ!cP)Cw|XFjoD}-QRRYxD(RFve zT-DZ#hfg}`OD7NA2f>Bo8}Hd;gZBVwYf2ta(PA*qJVefEtM_UyC4n+Prtv_X`jmx8LwL1Z0wGUdXL1EfrBJQ01b(|0eW>)o1vp0>3dqEu0TOpb6Ccijhz%aQ&0av@n(O z5Ad|mn&_FrM#B4vmSn_Z*gejxCYoqHFWmUeL3v*|!24It%ANelPVlc^bRS~cH#zY{ zt8#`28k~#HnTJK9WLN{*%Y;+&EXWjSKG5G+*0__mu^itJ$p(~?cVc>cSxNgM15A*k zk|0eJ^+KsY&i(J5t;DTTw03Ddx6__YKU=>(8#1+Yo)7-O{IXPJa@^fTbk)9QOJR^ zkAc2+K}=|g7~-&7h7E3XSdqi=7VRnHjq?^u!2Q^cW1p25^q#=Ya_htzfMo5Q$6+E} zfk$`}VsFsZK4gK@+wvot==!;icrQs{XnoWpgjD`E{Gor`YzG5<el$iI$_lV{0p}ggeFzc6i$~b#%Yo z%$lv|uW;H=K*gJVXFJ3sE;yPQjQHCGT!$}1&CQrFp~A(>$g3YCI9+Qqy?A|crWhg1 zP`9i4+fBFMMPaNF2(D5vb;t;~-Cj@YtRB7cx}H}T!75EWDk$VZJpsWaCI$GDS_PkZ)%(!rQat=2?i?r=abNr_zrR-N!;?iQ>*1g~1%kYzeAT6@^1 zs;bWWD#&nzEe--c+^*mKHn_kqi>_lI_#8fRVt$^#zU1{7CLK(3Cb$@P#d8wuIdSs+ z%1ve*0i%;-@T47eyHpQSDZ3-j1^}fgaiF2pMV=17b*ij$)Q;VLc-2{fI9~MEeG3Up z(^!iZ$S75w_b`yW{hL%4%mkn$N|H$myk7;e6x(*Y-)hTVD*RP_Z#zRKnw154v8&YD zCE*IZ?Pk_sCI^WAr!5t1+cN==OvC*L@+m~}7zyu^Vjg!UkS4=ojL!9By%pSu4&ziX zju4zC-#+hua2Iiu@N*yvlfL)Qq%TK!!(}5mFt=Yqv%g1Tdt+bFLfgTAiWQr5k3)C~ zzpTI(c<^bB)Vd_(AB<+d=N+kpQ318Rp6ViRpedht2+c-GoXTkUoteY8T;F@~ zLL?nIu^2;*of~${hZu{uE>Y;#!QuurIs3Bf!>DaoYvY0`VVDWub=ZA0n7Q^LN*LcQ zb~o9Wv-x#ILYrG;F;~Z4XQ1Nseut&L@n{ihM(g%|4_ypGHNqt^$3GcfoT#MzC2W^t zo@5z;(&?%o;_oI5wF}?4f_R^VfeUoS|HTG5{(}uNv-}^DjWz%Fdp9MLyB}&7uY<~e zpTo09zKf`zNhmjpWu;ms@n%EW7=vDhcQ9WLnxwRgS^N1m@f)Aq%o^n8&)9zh$4&i-JB<39739u2N&390ndq%XD%;Fa z@yPCPdL)-Qq)oYYb$P?-fY8AjwWkw0aS7467?nP7f6y+skjg7h&S(;aHd_25*w{OL ztRnN=)0RJdCYru%x4t}W@&Dmq<01f1_G-nUxTnm+z@l-i*YL=m+jL2ZV_Jq!o83}f zmC&@FOPdGn?!A+mGl$+9Tdo)ZnFJ&5;_c<~IEegcnm}myWPsmqi`rYDLz%_9O>m~x zIW+M-eSG>p&%a+!%9JT5f1MqbE>Mc zT7&6wY9EQcGqhdUt(N@3ZCAMdLa&ML-uhopKO_9=5BPJ{I&Nio;a|hq2fcZvtl((# zx=X(wztJArs?MyyC@oU$hN$jR?9O-te#@_}(-p6#OGz4SeR|OdrgqGRN|#G#r+0>Fs0GJ6yKTsEfWG(Q;p2GDKY~bh1@`UH+@m&=M9l>P?NkK0T#PrA((5 zakgriV%TJuYU`ru?u20ii-gd%U_PI3h|Yz%7X4eZp$B_)7A5{0^*q1Dj=%T}X9mOY z#cq}?+#xBm-)EpRtp?+uM)r5EqBkP8MV-vRL z$~rg?Vt43K+xMpX@Jg$zLIdV-OYKH&mX#}UCi~Nt!VGSgoDAkkkbk)&EQ`Bh6~r~n zcXpk58%70EFgh0w`FQSM|K!pJcweo)HXl-_Q6`)e=g%oVe~Vce1!N@0pvubUhX4*w zo+Z?5CQVwuO}_XRzLa7{an0j;4(-Z#RyN2!0p}yP)h$6@ZZA(ysr%vM!Z!JHbGKy8 zt&V4iw1FYcr}KyL<#BY)SC9B_pp+D*UsQx~H@IL-&)h8BTsKouZ+SQw)e zGBmwhP`ut?ymkE)=#G*!Kz7;S-hSCM!z;VJ_24jYYyBGe*XkD&FcYrX$>qJ}EV7ml zMiN56*Z;(H96~6X*cVU|hP9FUbfa+|xh~_CNy2m{lIeP(3cyU^0{;wTK=wqZ!f95U zOf>T&Vw1c8H+=3!4Iv_?_YZmf!W!U=-TS7-jSu*qxUwDes-mbkG$U-UY^7Vf>5y*i zJ;6r$016RjDXHuZVprt`!ZWm*89&v!Y~RjdWO+~ch+&X)xX^4*|Ls9YCINC?!A?xG zPD?(azI1dw+J|WjRaPI8jYxm_bM2C4@XH;h$O_t!W_r;Q2{G9 zvH>jOeCELI4ypUO5|M)aOvzz2-X1<-=0N3Bwx(x;U?%NX#UbY}7a6Dfv`_xX&rM$! z!4?`Fd$|BUMW`qCmO zKqVJN5(%=M+330A()i#H`Th3TqQJBH;CYF-Upx8X&5KfTd~_soiPmIYf!JUnxq3-i z`-F&O!QbY_Mtxzg{9*}ImfYc?d40jKI-bwvEV9gEd_~F0y5_#R0<7v?{R^1p39bCD z)(C<7xZiqS1b5i8K3kr`+H4TTo-34vfbo%Yn^{aQ)^tf}e1w58nyE2`XQfZT%_24y zN}E~B&6@7Es6&R-h7`znZK5OQpai_T3)E#x%`c4G0em9#r%v{Yku+D}>t5*dV~HK3 zJ7@`qRh6|9h4qiJgpddqctqvjQLt*j#CKr5 z5!q`T;4q-8F>>jD>+9z@l8#hM>!usI;-DeHS;YXLY-a3eLR=vqH5IO*VR9!{X(7<- zi$)t_S}j&l$%xcQSUJVFAx|NUccZ{I;}Svhd=t_Y=8eRvR>j-tI8K+S#)$`1JAQX?CpD}=tyP0M^_{)h8PD6uWIU5HrgAU5iUr`!#`hUb z683T#r#2HOfGH`&ANa?b!Plj2q1WHMm#3<`-;_1(P=mkNpJ9AHaX#A>>^|8<*2Q&a zLU^*qvTh>!*!vE(EiH#Yhd6CVoQZXa>xqhl%fS1iY-svR3g9E6_AXOWa~K? zor>6)h7uf>g(h<}7aPgcb*m}_E!fmgx0u5U3xtvg_Z9`{2J^N`R-}U=4Cah!ULTWL zQDL1{7a45ARK#T>_bMV(#*Qdv`AU=tp#71Ms-Jk}vP1G(oI41ltX4qUc(lAl1fuar$PxhjPX2PE~&R;Fr|=LaSVqXCHI ziD7CSWey)+-mX#iRK=d8q;#+BIlr>gt~&d#9t^-dJ7!JPG-Ge5)LS7~^;kiO*S$sVQb}=#b`5cL5dDDOq0Ap}e!;O*&7WydoooWpE&BkD zebj7q6bZm?-8aNz4x*5)cZ;Be>0c7n6t|ebL?E9*0K}|jeaKXxJu)*s!oAno+eINQ zER~3JP{kPy#gDV~j7IkNnSM09tqO-wVVUCszu^w|{&766qg;7b2eu~sL<}azF9R_} zSS~$IiXTPqC8#sbmhov5S6E7?S#(y9Sv55y7pJ2TJ%mgEH0wIo7`|$o+$UU#b?BgT z{pGQ;ENNPFE`gUDgyRZMvO&Sq>VYT9i073%eBrl)r6s zlzDNOagYQ)L$Od)5yZHABbyRx!p?xmMl+11I_7Q#aOektw606^rs0_oF+VR8r%Fj7 zC1{GUFnRJxv#<)LfgpF&hrJO_;fv1c(2W0n8s}lEDYf`Dnqr#Z#8JGK?XXX6y$pU> z#(4uHbx_49uuE}8Bs@Whtch8M5J>DDbUkYM3&IV3|A8zJQ?aGk7b`NwLIH7@W2`GM znnOy5OsiHc8U_yQrgz7O;($tnk|sKM=L^Vh6IeMInsT}yktoFsCDWv0t|_iUS~SAL zZTaJk)RKQ)!J7C>4_8JQ4S7YaLg-E-vcGRLR0yfz8CWEU32Z$B%V_|r7u{)uYzV5j z3>EP-$N?$1&e!?hRQ!BoJzW=FHskAX@uOT+@($Upl)Z2f*gOZn;+9Br2v4Luss=MTsXzZZFlYAUU@X_}ox`tSwp<|VelRO! z-*pdFs_xc_O|hsyYB&=5KG-l$J&fOeH($93SH0774nH>rBN@HLjwBK-z>5y=_V^8l zjN2JLHGII)-0u&2x56*MeI!<2 z@yC#PgHJ}x03j8c1`<{RgFOv!^9p6&pdRDUo*NKWA_|bL)fKYapk~&xjL&*fkgQnr zpTLRTRk`;02S+^ z_Jp)u{V9UYjtW~s2u0>GBby+Dy6iC>q(%8!xGC!6`(0oFb{+l6h7wbr z;BU@cnuY(aF$Q0S#$?{1mLf{>u~0eT`)=}cSiGvC*QI&(HJ7=i?(<43=-sg0*f2Lj zA@pW`67>dgPP=BmK!|-@%LWGym?IiCnA^Y$oUx}{zhy5GuDy>VF{<5jo8zuF-&be< zYkmg(M{_ksvGKgk+n1lp%kF!}mVmV_Q6&m16iW%su>wZV`*guu+ezmlMu)&aoP}a zgMwWUk^u@J3k~RWM4OAhErLC-u?Dild7dmpFPw;d_FeVV}wL{W#S9r>P&ZqGcZ*pob};l?RMxO zqpx0cgi4MAZ;VvyOIF*RyS%xE5s~mpxEw*JC#_bYluxI2tf6k6Kv-l~<3QM77fg7> zuPg;qW!>)8>x zLvRk0?tqXp$DHoWf1Q1h{M>szg6$~{7%M!4$=P1?Cb8xqC`257m4Sv*M|7ED<8S=) zkz}l=(n^EDs>8j^3@#8rlwLU!lfMWATXy94h6lGXHATUi<{oaNemOnI^1g{e@%I)T zgwxW@l8Lk`ol`qw|HD=)Y8y|-aFX7aQD4@@c*uf&N*cH@ioY-tBb~d)e8(EQR^2$z z68n)YHMUDnCuZdh2dcI2gKi&-Bs+D{_c)fbnzJxoogICEW~K)WK14S;nk6XXF{S*z zAn14j1yP9;Arb3FNIv+4IIWZn&}Rn{Rru;gUY@IRxt` zE69jZ3VB8lzq5sM=x5=*Kfb0T;VbGCybStSj+m=Na8gtO$!Y}A>&|?FYxc=KE{aA2 z)75u9HIp9Bzr%N6o!t$TSHhhPI69%3joCx`Dc)2aTYje|5)s8);fh6O_zB_TOv!MMwtN;~KY(#G8e0@K&i0*&1LhjmQWSNX6w7cyJbL{E4=**4*#3_8|r(@zg`` z=p`b!9OQ?Eq1yQI{c)AlD&|($af>JHzE{b&Z29tBfosSlxFJa0@)z&!hSweL!$$MmE~-vl)b950XU+c+uoA0u%bZRY)bzSPOQKkPgF(zsEM-Fz23$V z#(xt&uA`*=w>**a|L2KpT%7-dmA6)h_g|jaI-AA1X*ig#gM;)1k*b8YRwk$PJG|9C zt@#+lf(V6nIuL>>wKU+;5frkGqi29WVbp{warLPa%`wm)K)Cxh>mlFfpEZ({JMzl- zsb`09u<U(W+R%X?lir4TE4m)d12mOPNLjMrHR%{kCS6^Jb+$;oCaHz2gnYUs92fn2S?#RY;@A(4+cx6 zy%x{3YKggkp%IN|N=h&R7RCsT@}*R9VQqoev3tj0!CVa6yo@aR!FY)0GJrxP}Dj>Dmu~ zS-}a(zeRJSMxY9WG?7mo?X-BV!b1tX$3iU2YE#~NnWg|8q_EgzA2?LU(j~8A3!v$+ zazwQq$Sv#u^(!}%Q^&Wt#KiyxbkA$KDy1zEYCfWDfQutDLSgVu=pCZn7=j>BED$O5 zm=Lu9%QM{{?A~wO#Z{v=z#$^VyYd^+Y)trofw8znpwJNZyuAE;2$xX)zFeOzgi2BV zwsFMU4mb@2Ypy?2OI=le2?m&RRdY7=aODuhXf(A1jarp^QC+C%)~lQD=eajGt0Bvg zeXbEX)IrCWK8PZ4?x2pqdJO+Y{@q1JiKl+RrBP{CsDCid$*|LaIW+87ToT1|QM>bv zSY@6$+7Z{L*p7{tuoOYLgnBYhSs_HKW#*23IH;PP-$qF{vI3PK5~;wl(VnVdz;T7x zauS{m^dEBUoAC&R9E4rAiBa3LJ0vNn0$LOW@0{7pfyrysQC9TC%3mPEY-AQi%5wcUQ&ccI#jhE-+Hwuv#G8nes1=O0}0M}e0mZs<5 zQcXCb|5c6{1G9x4Kak;@^zjtkkq?6A;lpXj)p!7-WkScTfU}Xqi%2_M{{@a=scafz zP`xILIht&d2$PJ|zRjAMgEa3K9N183rVyJ_Di+Ta37(1>A>I@ z`@y?64#v&OQ{SirVUsl>605$9% zacA&XGs+Y)e{o{aUeNp=Jngty-H~R}6x$i`ErD4@e&V@5*+3{d6B9ROeW*HRFv3YlpM$gYHoRLjC{=HSn-Rh5u9*&eiU_wcjkfR zuOOZG?4A-+<(```{hnXEaiUmG?v`@G$jS_vJdxwFzcf)tkl&-qCE^!RA{@ww+avPf zAF$jK@BXNcIRdAKe@2e`9#gGOxG*ZQOm?WVP(Xn2EgFfu1 z2M=f0J!>Jm<4vhl>Uq!Ar`{vJ>iUO~-y>vUPDhfC6@>4BZN{pB&wWEU$yDb@6ts?@ zuiUqBDI$BadIK=Jvnpsh4GnIqLzf6nc^M+NF~QRne)3x8W;?blsYFPLd?Y@Gqm^$W z+qFITm-#F|Eh#(3ZnkzCj>NR1r7=9f4;`jAQdS_0uNK*im0-bLQ=+JG&I;upsN!0P)D)rOXAVDJ*s zI|3l^RmeKwqvLs;dPx?UT&O7mho3O?<=~l*7v67GI{b=F>N^_P)m7pDY9V~>qc|0T zM^4tgWSl?4g5AHjR zRyk`kB1Q!S;Y&kmD(ea^GHezw0WBD9n0GqQikWsU8AW6z|Ya zg`;I(WB-bR2rQum-rgl0r=mV+yAk=;XbW!1&tkuP|o|6=%!$YYM zV3fsKq+XCGo$$lr4x*be$X%$9T#Z3zeUExMH4mWCXY6TU8}~8*2GuVyCzzSAlYDYn zAVSg|HV}F2WNgkD+YMk?HOXFk8bCU2h)WgDNG6$MF45+)#TJ_{nQ=mzhUu1SvSlQO zbsTQau~IX(hNvIQRlqCfh3r+dr)aJ3Q@f-Ygug%tMhn&rd;-O#%j< zoRfI!s<@z-zfny#3jK1{o*Xt6!^1flOi3?rDfB96C!Tv@2}d5mB6?vS9CLhgR+hpc z5Bl9$l9IVGQfds}>aC+;1G}@5nuGY;F>Pa$Y1Op!Ox!7tZ-YN!OPP|)U4VQXp(-_R z)JM>Hv8NByyfV5}(M()dh(-gAT%rvQ-iyvvfJ`i_#&oPBpc8jkrbx)>KT_>=+)Ngy zA#>x>4!ppXYk^#!IG} z6p`K9+L0;sGbOocB$E=XxHyo%Of+tE9jxEIRaf5-Ut&>ZG*uq}7_fpG(*4Y61iLjb zJ#;m9OzRd~?A%fh$tABAz=Zb=kneB?C? zzIcSsvpibE&c?`VDcj>)d%7W&@Z@*x#dXW~fG(ON?Rd0gKv2vh`KkQ zS;V@gy(gfX;=X7>qt!RPvR@`^AU>IW`s&8TCFAyzkwOhA3i1!$am4!6L2#LE1>m;% zb!OPeOmBD3Y57u6=`xR?oTRA4CM8ppmx$-aAuarb@|ZVowLLoIt2Ret->MBrg&k!(~ zO4@pax1jBx`$Yz%J*Z5A`N;tm<>u6Kc-RM+X+`Gj!$9%WP1*o#U@=`b6PlNs#ISs>8zkR&&{pWRqfeuL$0HMa+$ zpqAgTg@`g&5&F?~uc-ABY69jeDj1LN?Zhd1=LQOrkyDsDqAz8&G;1Z4wFsDhz^La7 zh2u|fba?as?4?weec5X5;gGcYozl&N!tCr@6^G3x@ay>i@>?#qlFRylZu>C}0*ai} z&iilgaC6WB>us!mv|It}W*Pv@8qk&<&E@rHOBKnA-EM8Re{nGmazx@d0qoXu@1l$J z9SF4@&G&}f0AE1e>$AJv3n$k8h8&8$r_2|4}IE5Ub*+)iXO;1ueEcfYWl$ zklujWMWwDLf*X5YE$B?QmrBZs=UDLaVmp|HOxRp_{=!M18hl0$8I;Yvl|~z5(=Cl@ zKUpzO=^juy=0TMY7(CX=%0J=~hDUM}4@Tvh4YQ0mOl^HB^s&}+;(0(^?u1{nnVr+u z54}vfU6Cq9XkQ)5=ZEGfJ<1SuC@pFZj%YEi4aF0zT_+w@Ad>%t8R_-lj}qATJC&#@ z+3H|o5rD%pRV^qBPulInSf&2@R@4ak9oFB}4WqHfHu%};g3>>!)m% zBd11R*9`vnL`8uHRFBBI$O(62j;FNv z)Hn~DD1|=TM3lH^LOa4qj3dx0Y=*0XxD3LJ8(+JJzl{TCrm4|&*^44t-qrfYbYJGz znumS)M?0kW?)R0qhKB^6rc#uU_LT>s8Man z;&6vg$J>B|foz}umaK98|74AwiR*umj@0VNxcu`V_x|%B6L`|uXCht1-OJrqA17`R z$Xsq9_!NN_8A977D~160@B6kl|2mn};Wp+Z&;XTKy50O?<~M5E8K3;TpPIhEN#RRL z!TZ7d$z?>mpLiO$flSt>yi0gncfPv|v>2B{W{p>pE>+8n9Q zJN7*(RdZ@+??1LRZ=pFdEod0lXN2HvacIl;eU{WpzNY&3u@$iE z5e36vSf_t#9~?}ru~s56k0p+rOSTH6<1(9cvp`g$=>b2c}{T58l)aL)72q?K<2HE_Xfbz0bcrO=+z zL8KZq=Ven~U~ z=@^=vX%>{(#<{y{5Y}$-BPf^S3PS(zf?k391zn=ul1GGa^M@My&<}H^c?s3mk|q@u zJ=Q8N;oZnd(;FJ7Ph~yxu*!!>50K@_tWr1AHkoM1e+-oI=_8%3V|xU)iZ=;^vFMnK ziTDvUITNur%W=>OSL-!N4^7lK^3nVi`vWokO5C&v>jHe875Li=mcx(-axNG#9Qw((C_HQNf*%(oat;P0)g-^FnOTB_RMgg z8LFsiZ?)PrANtu~4V5qY{%vd!0!n#x2|RTWVvwQSezG6!Yazcg zm?toA^i<5!n!;Qbvt7Y}DIiqGqQTVZptZYhRjEdZ2rly|U31@=(8`WQoLjZgWskP{ zvP!NWQu|g#%|H!S6z`J&^gS8zC%>^jV?4ylnJQr%MamM#8K z3i_eiefQYNwX7SjAQ7LY16jzT&@s8#y{E2qG zWpL31A{QDCLWOtwO#TvFxA}Unp14V||4{!x&-CGI$%#=!$fy;npWp>G;>7S#mJ+Y+ zl~q$WeULP5^(<_&yRjnBp_OeQ=#GUSP<7OkRz!N? z2zA1wOL69t5~TNX(34h34I@3@eZk&K?_Q#AVP#vvC|~UhqXd)Mz8%gTS|;81U=tPB zK12_p8%X}=W_Do87|pZZ-K&B5=~hHK$o|+!F@v#q?A{xNI6|1lc;&-3` zXd*s8KOO*E$e|HChJ zstpPMSp6TR>-pWO8!S_ss>csH0PcX|A368==<}7G8z?BS+GvR>Byq?RX)YnyJ1oOw zW*9@LulBvqke`5(q!ZnfffdB_r}~?~cMKpNaaW1b3pzFtm+-0|3cS0G&|7UtMTR?g zgBO=i2Y=2^a&hj8fqyFq`PFrNIUO`>CwAOXX8Ewv|Gda>F>uRwnwes#2kxu@cFCsp zM-Xbx?qu5D)gQ0eVE@(1viRr`NHGT)Zr~jOaC|?B#x^62* z3{UAVux&n0KT|kU9^@sl?5RnGQ@^m>nO(Rhe)&GtBb%$-k^4u*m%!OrVm~hO9K8$7 z(M#~}iJFTILkuNjSZSf_KbvIUt{FK91O3c@zkFdI(_bpItzSy){~hZi*kFlUD<=Gd z?6D3XrxG9U8NioBv`vyB8dd+A0&rNpAsbevx4Z8F1<60(puUVnqEQ;b9jF=0Y+paQdl(_WR zc^48Q93ySUGt2sR`A9&J2=Gh1jp6`8`;vdm$R_%lbQv{#4J z%tb!;jhu1iwDbpHpX%H>4Xgo`9;9k%pP`I^S)F#_8QHEw`!Fg8U3&Zo&1(@7yVr<~ zo6V-52##&xooRE8F-`>y2N3HO{lq0+W{z~|d*eYpZ^ubQ-|-sECNhC2?4K!+G%1*@ zF7{HFBnF0Cnjpo11gS$+#cG+SCV#bXYAjJuOL-Ca-d)E82Utb}+%PUrOn%2=DIxQT zFPkQdbKhY<>?_l?gG)98&i_HElnrFe1GQp1SG)t@k7?LRinX4w6Y8{E zJX{cZdoT)mD{la3QjQMxRjpba)im7Hr(75|yw z6~U~oe>~7I(U723IiK)EkgppL?3&Iai(Eo*sJwNlhf7euKo0rzQx{N!|0O|J*|mAX zAXqmeyDHTKfuYva>Z-?`y)~F_Yp0dh&*zvF@Zod_TfuN>HbzME7U{c(c*()7EANH<^@|Wg`c9yH%mJ)>P9)zF(XaJY% zQI45si`aJiiSg9fGRY~_V3I*@u|G8N^2WSd@C6dbB`1P?$9<=P3s5kbYrze-=4wB^ zQrH>IU;GtK%_+a)=SP-N_X4y`#A80_boubvAs;NU5MoP2C_gnBOz1LDp^}5|5aC=h zGS~qcB2s?V;#zS_rbE+r^1BF-11MKW2m<14iKdOrGvG?1vMw9BDTw-EAbk;gc4$;r z|NM;8Q(%3F;X-0n_yXpLv8pZ_8AmGam3K-#22a&#TP(-%F7i^M4S2S7EVJgU<4Mj@iTi)n`E!g;{8eV zI0;g&jsl5In~GjcC^%E+WBg+SBIxVQ5?MkdO^TocNOnOXxa*&bLH}|cQOx(k(6Bj= zfb5Go*#+-s*bHywDvM<=CddCqatar7TJZBFNZ^6H572IYY=^~I+w%p{}>=>M|uVtg|_*Iy}+6adcQ z+%yYwUH$Ns@xly53(f#pR)xlH*jA#s!wmy32nzo+?0(`V7FSFlJ2SQgLgW_K%29!N z!sVtqGe;6n#h*vH!(g@7w+~@pEHlP7Qn%y?jO>i(w>oRYm}P*(-pKmeuc#(oSEl*2 zTpkJOLcXXnh`9mQV}jxdG?9NeJvbn*`R(51ZyPCTsif8?P&{}FmQe?MlBGm7hHa)w0(%W{&1zQ|E z2_-%?Bh2#8=LrUn##ANsFFPjC}c84U00LF1*f(rQb6& zj|~lj&SiNCS1(WPXe(*E!8q=$#(3%`Gdt0sO_g~0^grp=(i>wqB9K@f@jnL#m*9p- zt8@=JUg3a;JSQyQzkMk-Ybs-?OX@pPF~l#g<~+A_Bkxbrc(=pMHW*fvi{3WY_4&} z+l$QcwbB{Su9R+@$358W%k3fw4UAz@@tkaimRBE#!v|i|@BW0(UJSAA7GI;vF)yE= z-?*~mW}W4Z>)(Mfz2n@_TM9sqTC3K+Yi zT#lH=+=-bgg8Re+bxiAtF&KqTmjJ30CS?c-p~|pKAd7z->Jmo(nn0mm)6=lR7K)m0 zqVPmP6BjJtyiDK7Jkf+r7N_n`3`xmmM1i}L3!cNLVjem>7*U>35?HsE|nZ#;kF!?9v8zk{l#5w9b z)eF>tQ_j6NQR7t`eiG{T+(q1WC(Z^rO%sKgqpsUiZU+eN_B<>F!{Du(VS#DYHSo`( zBUDAeyD-KvsCNI8zFa!P-0~i&g1TQF=94Gqn~VY)lXAWP?(nAA$`#fjd!M_{#$%t* z#C~V}W+u@FKjLD%G1+&iM-LS4J!M7 zvd;vc_{05dN|@XdNR$@6s>bhIP9jjP0?;y(_Zx3L&^E>x<8C7ZznSyF8{V?J!&3us z2uFz3sAPJsH-7^3FRhNZ_k`B$o;Qh)0-K9dM-Cit4f}%l*wf~2Ew-F5JdAVwo!zH} zThDOPJMahCXk7<~&9YcVZlDDRV~y$eg+B(h{mU}Jed}ztt*%%yv z=<6%yN>)LJ;vGo?(HT9tN2JCKd0*_2e)EK1y^JH9a@3=?v%^MFVAuD9+4N~nuX^P= z4rzw;vF!Ox8taGkq`pi@mviwN@u5X;F_MxKyx66L_>%~dmKlr7Jk@6g)C;@K=O^aF zma52bteQ7(#ve4AO`Pv)4fYGPk!hk$152Q>fd0jxb^nWxk8U)0MdvQ%!KQ;lyrHnL zpwm?P-Y>j*v?#Au&M*Q~bF(?L(i6%wJo9ZIWho6{zU~0~)_1Oh6!h{fa`USXi*)8b zF{sws%)9;4Qkjfw1iZ7-g(ci0?Z@!Q2Uvwo;LNuL?Z2KI&)@D)Z@l^p~L+n0Ob zBTVtQ7~j!5&^4b9qFAdWDj}o;nC`FK#qjSm<_*-$zx+dcXC` zD5`j+@zIgMQ>Y#aL4I}ooy!u;iF^x++L$$-?a6`;W;;H&_HoD8&lIA#_qyRe-I4v` zkwoQ|JU4g^_{}8^xm0YBK-PyJ`a=;juyk3x#Ho!KT{5o9(eCea)R!32EvZ_#+;1jU zREw|67(aZRGZiz{wEx8TI{wtVz%ubLz29}&MS8caG2f~|DpTWD`ONN*@}$^NG7d{_ z9}&>!-@BG8liN(Aim|qwD_5i5ffwtVT=u&`_nt1@lKrgbN9xm!Yf(FfbLGo%82{L~ zrEXL0AqX~kIU3i=-Dgu3aY!jzLQgU+s?)Zb*Pxin`K~`?+LpM=T(3Tzb7)@@CyUz2 zeRug`K6N@wR-qM8*z@u zoGYBQ<;i!J7_~`u&BEI@DWY|w?`nGK*N(e*+=*fD+Um9<3EM=hr_8)!W=id4DQLcf zA*_aBzl9$F?i9azmd>h1s`T7M9E)^!?!pVVXmSw6Q7+A39i(>)+4bf#ylmm88Vr{l z#fUlNvcOQ&JKMz6$vekQ*VrEFf69b)Ml$GdYulgs$OpO`19gMnG9xB z+KNb~*w!JnoA0c@6&cI#z-vCPo?qUqnM3b(SEo=g3_j#1pmR01+n$QPHCr3Z2_C5s zmgY8&{hRA^R<63WYm7uI3%olX)k|Dzj`e&QN6en z*1%{lYmWRnbw}UpGz_FX#o9udF4^c3Q6^iPxXhed>tq=AYvF7x8>DmWN_D##D4z%^ zHcw-$^&U!`VNFn%%wAbWM8jdq_VHTV^ys~($cVw4d{1`N$KCvv(}thn&BkG zRWplbJ%@m~4%dodt-0m$VEm4(C}x4_Al0jZh{}{l&W8-~R}(l6dvn!oyq7`519G|3 zZc)q3hL4*p=Hx#~S4!1nT0KLJXs4zP&K+zK8^e)F7P{|FAh_;h1t0 z1xEL)L9=sCni{jjdXn9xO%BB)i<&2hlsH9igClJKZ+v)r38!BZn{2HpA2l#YvJ^@siZKIj?g^%12ug zFfuTF(`}>Lb}!owt+8SiOjva*Q^9|znj(RFG4?@F|Ars@N#*J*YK!;t_I;z*ECOw zym$(gfpL7i0MwGIcl6 z`poTlY`=~zvY`eG0Fq`gX5FsYP{7bVEp20R062h`n*gxS!U+ ze=oON5NAH?v#)ykS5=-H%?5#o?)LmBiMd{J!rJgDY7)J+xH0DCmPn6I zVf#BCn`s648IO*m`kI5F#4F#$cz}*`Q@%;JNtgOYI&pI-QB^bB>Nl6c91uf}v(iBL z5K5`#?BBZAA*QCIe1JO>#o^bu^w6cEg#B-hD^$LpSkm62Uqxn>LK2U-o+m<^0j0YT z#VyzK&A@S4&~TLTJ@z}76bqF*$%bbV*}}TX^_20SXUDB$pglr(lXkEcQ-$UsEJdYK zD0Jw?pGMp2F(T@)DHCzZev6~p3p@HKCO9%s5;PN;v2mBq@Yroxo?(b;oYf}SsKu;t})NvNS5_l3D7A zv;m|#A)^dAzH@jD{bcMg&%%Q(l@aVHm5zg&M_ALcNvwAQKAJO}aLH)q!#Ij2FF72U zXgy@-PNcI1c@CF~9ue$;EDZ!!J~;g}|DGg3;TvR_{{S;s2Xx0Y-8={Ejwt}9i=7uZ z+S{D#nFu5MubR$uBiT6x619(6K`or*;ZX_X5{_r0i|R^vA^M$jV56vk-@Zj04Uth2 zzYv)`9SRwd5uU?ueOf}pRd}x6dm#Ry7DTXafa#N78lK>)i=prgmo<*ZW)AAeRMq06 z$wNpnMGSK1xHu+*TyoB1X|dDM7(1nh4b(PQr>wDYqij!kC$P%Zkm_`D-MwL#BZ$|W z+R@}LSCWgUtw(-ghmcZtD~gTkb=ONtRu4)b|Fq_+%ISb$av!-VH@mgObEFLAJsVfF z!3!rRrG_?kcrFlO6s_KYs5aj-BVh0Au^WtegYVnF%v1{+gSRFO-zF{&!_E%Ghz_Y% zz!2F(iAWLf!I?~@Inairy#G4P>1}h~STT>sY(v9BA=JK;VUJ1`ejg?k&s~3!I8YPJ zeShfvWfnPMw=gs#yRZj4DkmqI_+g0t*w5#tw2eX}Jyr2Q2bK>;kEusb-3$L4x$&0) z#}54@d4L=S*(~ZV_^%6UKnQ9tTEr-X!72`L5fS*cIp}NEy9BsuMSJ~E4t^< zcmr3xjoR90D&9aT%{crnB6PbnhZE}-`>FF9eIyia=kp4hREwN7x?FOSatKh+@hY`>nzm}h;6RH$4XOPs zKawF3=meTjbiDJ7{|Go!4~-gRKB=i^`l_?(d!8>*$>GX1CiU`xb_wuIGdk?9GDV3%GmdCM(lM34!n@OxL1Ij~z@#3pa1gH2 zA?xCE+5RrW`(W^~RN+ZM;+YX+Xf?C}tJgA6h_ls(PCZ5=dti5}Mu@{GBLuq(<#;xB zZ?oq4lXb$(UYADh(P{g)c-)R?YOY6?EF=;RyU}jaWm>jiu9CcH_Xf{c4@5WTs^sb( z&p!O7$mvmK_kLJ7xDu?~3Fn*6P5Nu8KL?ydDQjpQHu6m7#=_q7Dh0)kCJGEam1QMS zSj^2&US{%-R=Cdj%xUFXoqr_VKB01NRo9TxEUwtcc~PrLAokFI)hJ<}F%d zz0m=IOP`WLEcMSRA^c^RD8U5{c%PI_ooA3OJ{2IvoZ1$Cp5&8zM>YhHPR>mChSo0p zB}REsP&!Q>(otu!J5VmHzD0Tb;lyLNT;zdj@0aAgdmz&0poc=}<_6K`7tP+8HU10% z@)0|-Dz(v!j#42#;mhY9YxH)^05d7_6^X~9e~r5PB*rSl8zW;lM55BJr<8`yfB2UX zW$jJ+>($AYr=$X&sU_@UU0Cl&zU&1KkshkEAWO7<@nOu@Q4+LVw>34{ZMyVkD35WOM8?Ld zy6P!N@iPaFP_GH>y$Q$$Dz@SFnl}|l&IWRO^DgS%mipayVgg{}pP7eUH`OmsJa8eA zt%83}^}a|DpgZercf1(N9DTRt(3He-fzgh{t^#~4b$4H84YOA!4o|7r=<>AnlL$F- z>6c==h+Q%ocN@;+v<7?61dtRsXL$QY&K?~>q{>*Btt@6u%z?z!W^tXn=OiUE-^;W$ z%vjHhLN;1=Hk!3}((brVi9|naL%1FyfPF>B>9L*k7!4QOOotVy>}ud4M7yHf zvTHJ1v9`IU=4_0!j?+Q8Kf#vgGjt#CsQA_OuM~5htOUD)|Nyuo?CcShJ=y-$ty9yj%IBrs4 z4QTyJ+yy*XA6147n?FNA=WN3_S;QnN@_fI!Fh^MqIq2-?8W`v zk9*Mz2y{dbQ0c|kYtjXw^V1!1vX5XAh*Zsbsjar=D5%XgJMG1e(CSqq*Z)49E{4lM zCr;_#{!1FKKqqD>vW-w-Oj2pQ{Uz!Ci`?Gl(@US-?&~?kL+WPv*tb)dja>(ik0x~4 znc*k5rU8tla4QBX6um|p{#N29B0D8}j+D#hc#d??=`M%33thL(yQ2Cf0z#zGrpnzl z@<_G42RWKfA7hLqh4(;Yl3`Ubl6{oXBpoh!gojny>=FE`H(z!EyQCZYY>WK<{}?;R z*jxgITi3R2+qP}HZ*A<_w(WM;wr$(yuDNUX?sIeheJ8nJGMV3#Ofr*sX07!kLy4M~ z^^qm7bXM>r`%A4KzjB>KgTVqGw$J>=;-S)F=*IjyjJPkS%T&o-7`{1j*0E*WFP8Jy zQsi6E;pN%urna)wK}TTImI`Kr&nGl~>Lk zx7^)-6{(N{dR(nzlN~DGfz4$-OLQltZzcN1h!PKyTkJPLI2DiaLc9g)@)(uiw5c&u zBfO5-DBSonA6^^Aq_tz=v`T)cxmtfzm_C=xGQBDHlglL}-UV!=X*EZ)C4}~XsY&PP z1HtP|Nlob@4;KiiCoE6b%d}YWXoH#m)thI^S%<#T%p2vFNNR92*EWF*H(!u9r+Q+B zz~v1|#$J?4eH&*baHkk2}dTCnz86 zgP3kASIgl>lmj^g?&nPwE@GH}`xidphDf*Lm>UKGI2VSO5gn`(9wGV}(%6?7?pPqg z8QIpm-U$s!po31K$Hs5dt?H9-b@Q|{wHpxQgL?Pz+5S>0q?_lG~$gnW-wB`v(d*qzPp;9BHhJaWxza|`Y>KhvD`_s zwS?B2)wC}AG2qfM+H0fF1I;T`DQrF8$zFQcjaZgS{|`RlmHnT*`VNcg)%ASWeyfsJ zE?=WAD2>Ku|GX(RkdLEZ&?=xQS~7#TDUDvYtL%f%9*%cPKfGfOz#gdne}hPp%d*{O z?^RUuv`1{P7YPM?mR26aQF?tHjav5kxZ9v!KemAn!PG*1V3XTrT3Mfu4{5c%3?Q z(uSTkWbJl67-K&sVHsjXGM@yjGW{&#{O2IlDSGWu8IqK~M-CL;Qr~IYu?rH(u_rCp zOm02J=kJ1KJ*VUCr*q5Kc2>lK%p`aJUPYP{c!3+7PWmYFG{=a8v1PPsyO^!)?3#|> z;ui_Ld3F$o_7UhR<vXJlkIl;>Om zUtESe+qAGpyId^D@?8Z@izje*S1(=7-A#Y6QWa2}U?fr}=uOR#E@WYe`J^rGQj_uj z`Tg_TMe}F#`d6c1Pi)OhZ zjQ8@}Rko=I<0B>=|KybNGNFi`GeTwOCMF1>k?-x8?=FaB?GlQ9IX?C^yQI4RPnyWV z#PnZiA`1)i|3(`k@4h9`xMOp>qwB~y<1Poa6Zhzi89H=vG-hPv1G^T_Gk3NSNvDz)$b)x^dMoJ?6jP`h{G*hk3c;9bKf}WFq(`N z&yzYCMPhTWeU?~iuzg2TBQ0*QBHr=NM4)I&QiDQyS*g`5D$Qv6>GDwkNWIH5(8K$3 z2}%4s!zlZ?hw6$Q48%d*aB1daX^e1W1Wjm8)16;b+CXQ9V~Krl)*TIY&|+O$uX& zsL6|}n_rtiIv|qc3>4JOLue z%@g#Vq0Ohl8_4s6k#aAj#9?6?Ej0&_ZUT|Q@K6t}+oP~n5=gjOArL{#GBgIrW*@o$ zH^g8yHNHY5i(=eln1{K|zXv8|`6ZW#BwDhE(0H%}rI~Gwafy%7|KUHeLMY5wh4@7| z#16fCD@LABq{f9qsEx%58+A8>cz5k7255lm9on;4O81PE1i_So%MfV z{~YBs^{>V`tUFpxJl9E77jMla)CavQN;egPAhmRrd2c7yYF+=5x@G>6DRE76YhnP+Es6mnYD_%zyM+Z0O z3}=O~EESMgR5O+suN$<=qMML_!7d`O0X>}xW(|h*wekz_1tjVv2 z3k886UMomyb~%c=D6?Li2htq3Wcu-M!1D*T7IFpka<*+L!$m862JpYvrt}3-k^|bMxj7zcXZe zdFiCaQ7x)m>F(i<%2`_0SutTvm5P7qP5DvAR~SkF6}^PjDfjU#upI|L!)qLAnPf?h$MP!PXG zhz|K(a(ln+(Zo+J^l~N2!3PL5T8^8VPq+gie}S z{Q{o|gzVi8V0Q}wLrB?ih@W>|0Iks=DEP%XeR^0)3tt+BhJ-GE-pblMx zYuRrMR%4{AGk47nl}4I=cEivDQwjn3u@5IzVgCS{&GxC3oUi#@NRr89O%Q6Wi z>u^jZt8*q{I@o*Zq-%1D`*|a|J7z^{fHq{2DOE<-8;{qxf|KpNSr=$W1Vqa4ynfm) z9K?6V3!+oXSc2hsX*($hh2;R@$58G9dCX`rHMj+z#XRFYkX;*G-UWfYFRid0{>H1? zWCiJ%3RG}P30XQ4=tlp0#T_n3Jn4sE!8_H7Og5rKQ;HSk$?5X4Q&ksgv>ALvEy%L3 zNs$OVa3bGS+s7@?+tU+k-nJocu=fO|A-Ux2xlPkMRP0UXlQ|c4)xO$i_}#kt)F8K6 z4&xDu#39`ze3#-YX8P^4I;z9m976GtF=Jn#Ijpn(DcvBj>N85%v@&*2aVFQCr|%1# zMsX(@(F6Bd0*%*33aw>l)Zo2Z4YBG&kP@pZJ&Y{HIwAJH2Q%v1S>71bNV!AXYL;3| zw?x?0JAan{%1M4TtZ{i?NE)QljO+o2CjvH0+zF`nWn(WkMX4y}{QXe@OLR+E%wUAi zOiR_k|1LlMn;7f>f+;%MQ<>d8`QfgFw9VKO#5@;@Sd3|!XiEDf7mbpVvTP@LYXY5I zdK_KZAo5A>P-?sPIxL_T9P=JMV=$_c`rBr*Thn_i+^^Hxxp08yY(I!U_qU!VG8CWH zlWPW&zR{m)Sycol;M%zPM3KNL>trVqbVTdJO$9RQSCBn_(h(b=%@7}U{G)8^M29Au zDzEul^tlX>?n!y&;3sp%E|>V%x6eOFR2N*KBj?0alL=yQV-=-mBR!N-&|H}S!Z2Hwk@YYT;HOPk*=;J@e!)z;2 z^BS>4HcSg5$1E3HLW8Cl(?VSZq4iBZv|6PHukGlb>;uynN7-uz&OL|=nEAnO#d(e= zPz2`DQB_(QLN3X2lZ0~A4OJuqalm6MO-6NQ8+YhE?uqIGlcC0i8`TcDoePKRA&$A| z&oQN%4tX|Nt@44^iNK1LPPq{$6&c#^^D!d3YB{{Jl@`Go!;Ww?eHM_2V%pjgM$%B| zr{{`xVeuwaK7tl!&1xRAI8;Ku{{9V+@S;*v+9C^sA8E4s z_v2@ztigEx3DmFIG{T^*8)6xl>#Ujwu2GsmH`Eft<|f zJ#yFIlzFF~=rbA-qP;j(c1Z319vGx2!i2QDfV=Bby=|n<<|E>D-l@_DfGZa?DgVk_ zSZTWDDjKn|qgPV@pJsMNpe?!eRLdn59J*}^FG8j@`DXa>{hEwaipi+gs?2h*n!!~G zwOgl;2}K&TWfnKk?Fi8Pl6~#r9aJkDnxkeLRf%oiNa`_Y$c$ii*V33@N%wvdw9Fr+ zwxYl&1l)$;g$Gowl;4}26po#~C&ktmR<|6d+dK99ela2u2&G^Pm{oIf5<^Cv?CjYP zW79iA0#U*lqroD{Su3Km>!XQV6tnw!NOVzC0QBahDsLIO$%H}5UTIs#Zp=#>$XKvU z(wDJY9}@e>5Xh!sN}7#$Es=Uz>H5+Pxhjy_>Hu%n>q~6`a9-Mt>rv?E`a+HwQ;TeL zcn6(&=zOn+0AtgY=MFv2k;d2UG}6}^#J9vdZg~uVL;TF2XlHyK%P~z)4NO~4V>U1- zuEEPLJ9yS>nT#bA$a;G!&1?#};RxV)hpa*a^XPx0tRE6wmH=RS5X1M4~)-Cg6H*=MHHqU}`$;_zX*0$0hdxeO~$PLLs< z_QD97MpGdD(4&LtttZ;*B0O$ybFvKq*gYpiota}MwvvYGxsq_Vp9fEK_~@fSQDxg3 zZCOMhlym_4_4$@fiTF*caVTrH@^%b$FtG-#;7PS3IuQMxA*xvK4TRy$y`AD-7|&b;0$GkMc(ZFKka)3Ny>uL- zf`4MCFy=b5P%a}f`0SkMaw6M9hLKY%S@%=5*MAe8AV#;cK)e&v0Gve1$!1b6zPs#4 z)(AgxhN36{gn!`n&=$MHa>e}H+0a<<4;Gc8>hl&!4tkZSBA&PlI8>Lkx0e{8!I-Zm zd`@{#4DlO@L6rKl%XV7G3*34O9*Ww-bKG`Z!Ek3QDfo~}OkWQ%0G+4sZ#OqpJ;Za3 z+UnDig?9{UJA=L*Av^foxzd6V#N)dOFPRPZix&rTlD2 zhGgmFDAlQ8bB+;}M5KmwE_u~=81YREt0wGtg{oX%R@*?wnJ$PQ7oq)uU`zz8- zr8WJ&TspumTlhp07=t@Bl1D~(3rP91u-0U-)Q8wab22A1hnqjux{G4{=#)nN!$^1a zI7>A`0LJ1fWS}%n?aywXMic$N;!Ob;&{aUKL9mTOLlMln=@zYcl(uI!D7l3`G9L}L zZUca0mXq8k-NG?saJE_7wl`4Ax6+fL(=>>xbk08ACXd40_0??J4|v_@{*bYPry89Z zz|DF0^2i(rt~T#Jp)Fcj^j!Qa?aTKb17d9#MSi4gg8H$}l1g^W4K}9ZU9+V{+fq%N zzF|NSUlB+t&qCW(`^#qYBTMo3%{W z^dZO9j-^c6b8^>MH1(Yz%XKt0<>gc1x`Cw``5_aN*~=JaKt>xt|5$yg6-bM)XO(Ee zo8l*|WcQJU7Bd?u%laB=>iOQNX;K%BG?g?3{Qu1>?j=KT)GZH4+v}9 z^1As4+ehPH+t9$)?zT28x(4`~stv8oS1&_C<{WgB{-t@#x^~=mOi(@|@IENV+S1D1 zk~^2$&%STp=MCV9O}w>uF++5ueJZMpmqv)F{^?||17Ay_fR{`syQ>t&+j$o@uOmYh z!a;xXR%$Ix7F};Ha^_Ml6P%=B#&O+qo|6TlR$uoAGUE^XG5LP5V6vfGPVsMC_@1#| z+qWFIMA@A@19wk+L=HZ?2eLSo?+oxILk0qVPmI2ufDr`0XuovKBeZ&U-KTS zCT(B2F7C7ot_CZ-8t9e<|K#M`Sy(n+GVg)zxcoJpAaXAbW@5*9w&W;gJ) zpuvgqmd)Bxz_RajeJE*)O~KGydB+qbe-Oo=)hquSuVnt8ypoNJmF<7SD`o%j%Cub@ zR#w<+uirapr~f+Ap(37?Q&A*sX0&NMOTP{U{;dyEBqkK#fE`o)aPv+!3A#{#iU8+Ss;w=%)}4V-V|IHNTz9B zs?6~Fp{$5?44)cAeoPqN?BdCtqEObrHJmB+1cYfJ~GW3RJ<;k6jN_L&hhrhkCxmt{daq_ z?OG}X6)04#OiPD1L{)1B)mSiDd>4L{Dp(%nG2JBv@do7HCM*)ZkCA=s20>alCO#+L z$4gf!J8pKXSXXhJQmP*^+ZMD@A9L-}8a{X!a zn6$Zg!zMcYStX(Vij&6Q+HLlcC^zm@4%y`iwbID8yqD+Ja?CCOaP+xa(pfF%9i1@D zXe75z)SP~0KY_;9YmI2jpM3}Jb&@T+L3^T|f^hj5d3}NsCl2L$>5>#nznXru)bum- z!{3=&N=gjk&}kZN>xI$`;t!|LSMn2R)9BO4@(Q?rUsX;}w+ZCac+L`KxmzVGd_b~d zc|8IVk_P(W;52!ZDm9{>@5&GzyjjC4QptrHM)TO_L?fF?%`1Q9np&2k_-xmwGHg8` z9G_bS66V=vjaUVKKstcBckSiLKoSmQsbtFEFN5g^-YK*cZit+UHmv zDe6UG)In`a8LB3DTjSF~z#Bo(k-JdyJ5sygUd|JxwqNJLcVfP-p@d{v1?!+3;~AGd zz`2)45}-H;ok9UDSIfWHRy3w5FlG?pfpz{09-MYpV(MyfN&e<`3&D$}^nm3nvXEI~ z7-9@4t?H!D&iVx1CW0Fu$%fCwif=gdQT3<6)uHB;NO_9kEOi!UrFEutX4d0h8;B|4 zEVQZ;oUWW5QwbnHeP(+>y#~=le5k7gr~fVX20~s9Wl_7Eldo?~DO%-;yM@MOw$oT$ z>~{E~K=vrTxOhkC*0+}o>v;~TlD$BeRLaMA0WcW05>D)4y}{_R+hVL;tiY*e^j(LK zmK{-Vd38Dz>e-E&v2J6KBo@I9z&9yc9B|_k$F0YA3@S7q7%}j^dmID2Sm`elx!a)) z?9JQ|%)l3-juH;72W~~@B~xl_0{I~n z8%4@%vx`S~?0O;~v4KkcEyM*4YqpF2TaaK}TvcDmH$?#^gTNgoeIM4q6<$xTVHAnO4iEepSsHMY$p!0BgKXykOJhLKA-}?~Rx}Q^4?Eyn zfK&7CAK|+OSVE$3*9iGgshI7$jfH@Nbs0qI7{KSu? zW`7;|SV)Bk@S79|CGkTR)n}-!KsQT?+`^_F9XKP805=y`WV{4O3g?g-&Lhv9NsGTi z89uwMMYq{O2Le~K-rv0X)jLY#1OSZ>BrM~P!>4MRLUsDzYO#G<*Pk2wZcE@2}g#_wLxNw zv*M}EvHwd9gK0X?@gN)X987AYTTYVzlLs$AhLI>{QtX(~( zECRfH(kPqOo9!W{StQO}rJ>F%96wELAFyUY4FfX(go>538mTP>RzVNnHK`(x-_{i# zzM{Ivb+pkxkS>m{q!q|P+_UR$HV;}HahY^L4ykDe+}^JUWw@S{9EH*hpkD;(bfgp5$g&TliA^n#c|E%my|@;p%Y*1v$Orfi?wz|S>~6pL#f zNO4q8<@>B$9-o{}kvb+(IUhnSR1%D~MEO&Bg@mOJVD$!3$WU^oI?28uAIB>Y;QivJ z_8$d<`Nawju^+<=22?K;LiwUt9K65HF{iMg~T=HL+!nqE*AyeLQXhn z7`J`Ws>829N0Jq$2Rv5cs=E>a!s%04(v4J{Sg{!onflEo7(>^*kUvoSqI*%&;xVWk~mdLa`; zAQPs!F_@VcY{CJ-nL;R&W*4+XLkAJHMVaYSU-#K}M}~xta3;o5ob;KKpFC>t;>er7 zpMB}cP%DbU8Zw`^$JZNERLTkvO)*JKQb82NJ6>YHxh(oQK_FYP_wUHe>40Cm(2#4= zN)t6~5o^-Y2I)9X!hW#PA>q8YXxnz18=Q$y-LWt*n2Gw=t|~kA_`$P5pY||yX=pub z%l2321xPhXbwmGHgYT7%oLW|8N#V2Ir`u#*AP?mMEd6od85AfLx*V_??is8P^cQYG zvInuGlK}jMp2<>yNQtQYsFKYOq@HOhzV)wEaA0UgKVu3d4Txvp)0(bo;3_yK66G=i zOna<{n{ln~vDJw%8Yscf)ft1?o^QH)!AnSz#SBVrl0=TH%5HKy!JfPYYqVfjQYSAu zfH2tm0Xh>rjRw3E9>!iD{vJR7z*V^G;(1n|<>p-X1l&9C>C?!w?(j#(Cd=rK*N&VP z(PmhYpy>gE#lh%#csi2eS#Us+q(bGNfct)0QGox*a|FMm(vCZYcQN6 z52PpQ?|2G#zIc5i@^`;^{2*;Pg)M!36jqI}!C?Y`>(N8cN^j{SRPG6EQRD|m|4Gms zD<4#j>dV#s;rmxexnE^WJ9T}11-xWT6|*tUH0+K+3RD+DF@%RG8T&er%5y&fpzRYI zAR1u*EONhgCVw(!xdKWyvNbM|0wg4m#vo4;3jEDV#((u~n3NYM$pEoK@+dcVGQ>ONrV!e6Ff{m2o1L#p1htu6aJcL{O~)>)?twGfb5wk{%YHKiYug` zRAozaTznYmS}V@tfVm}xc++VEvg=8n`Hxa!zvK8o>HZPkShpNf*Gtu}X#To217wjI z-nO;8=rHxu=TDkH6`%56b+eJcY?mOjea1!{Na{q)Zz@1oJ$E_bM{6#<+M4Y2pC!zI zX%_x5%J-p0sC8#8hnYa-cH@pz8gVuIku&Q)efyQn5sl@I-AX7EuY*rNa2b&qKav=H|z-2}>TM2eR8aTnPIIrgBL% z-z4nr3>H3z?S{5z@!Mg}6$}>xrgKQmce(apE1d68Iiqdyy`J;!4M=}@J0t-H1_l6W z=_yg2*64U3Vks%La2i54c7xy=ox$U}@Mhb&;3X#tm_hs4Z8;reJ=|~@65Mtu)OvQUs>C^~0~tVn#+gN8yXjKY83~4$4_Fiiak4^5 zWNzPK?{>4*R)Schi%+?c^@F!MD!?9Sr)&C-)dH+uXJNXnq4+A9PH-GQO211x*bcZ( zmtUUQ0L9XAqW)xw<<)M#FngvF?g>&zFof>{BD;cfg~J8x z@8N;9$98MO^GQVxMy1IVV2VLnE+86&O~`eGhsaP*Yu|YcSkyc6R4#1(01mLY)Cw~g zW(tF~4~%$3bnUr!T!*OFWh7w`T8Zz`hv4In}%=v_7+^bfD<5X zTYJw;bL>S^A`pewj@EzQ@G@O;b^^LK05A%dKOwbJ;^@Y>qsV+N5gVQjn^J$|UCyb# zNy7Jw|6A^RqXfyKz2*J4O1}jFTlCLUd+n<_GoW6tm-V0zEl{Oaj)H0%H#K?y0z|H> zVC(+kPo?=_=QHGn2|In+NT1g5hpkc_9HwE^N4!cE{*zS}G0x7uj)Bn+3_SV;FJtT^ z%<=#s>Asw!pa+#&NcRc|DZ1s(SB@u*<3cg zKmrd8-{+_q*LMH5=C?J`T7XY$R{%As5EQ=_<{^VhIG zI@Y-PyhC4MkX9y&K(vWZag5OR_I599UkZkxfqpC|*+BeK&!Nn$CObjLV3k*|WMd%N zvL#`s=M&$aZ1Qxbf506wf);}XApzh^)k5ovc^@X|kmKl5CeO=+s&uJj7|q$Lk~p=bHQSZP5H<9<)-c;fFug?fQqcre7vQ?zodvWbv{?3oO zh34&xn4q=3J!PUbpQ9W7H`?921v%0iN3AGL(aEg6L(XqR!&IlZ7A3YVE`jZfA&a}* zt&PU3T9O&>z(ObII0v8sOk-{=KE8Llu6SB4=L{=`ACqN2_T6@DK#i+UF&>RBM!yD) zCtQInz{dq zH)>7%tbf`V@jaJ%SgBted<++N>VC!3dD60YmkF?f3L@U#mxXtqE@-xRGPar->B zJqTg}-SwvH4Fg+)$BF_;Q1hL!GoJ0h7aiTb?#JO{Qk7kKq5ncepquGDrJjo&}L`f`XFr)m&ExJ+M5c zi-*4MWzn!I9ibQnk(W124vvG6Fc+PO2214eF{lM|=ms8t&~w}M>~5HsqIuha88@pr zU8@6vh!W>+HHFSem@lRf?NM#&lN9x?{&YEE(XtyZUYU>+UYfbsN`J+)he8M>nsaqC zKy0^GPW>HLBxlL1{Gz8^us}qts+8>VI#4d+KEi&=ozVDb^Xd5bQ)~LG`@g|5mj43FSULU| z;l68~|Fo-a<<@@%NJ)v|lC2X_x4o}9m}>xonJFzT z?To=oYJpwk#xwR!4>A_TL@S5=EbdKw65;aWeL54JT~mE!PD|D~AS%?R;(!xWHy3Sv z1apbR`0TStaXWK3noXa%`StG__W3`2yZC&t!a4Ua%Q)WF1Ut1Yfj;xNDV44bIh*K} z0^;YY@npw>5tDXee_THfHeBL(jZH76?VWgrVv`JF;?U<6F`QW!t=?lrG4fZtN2?~gTd%+kp6b1_Vr)!fsFI^TKY9Yk z&_)j;F34*-#kdJ%HBj8tXh~k3fnam!%8~@@Ce$7gRK$sjj(XjDy_5sP{VnWBkuMA& zI5X+0&Zmq^yJ|-`avv33-6EMNqanATrzv@zRohp(xJ|uK>$wsGZ;!pK5idN3`dJro zTRdi%Byapdo3Rt<5oGe@o00|6#k@(@z)eW0cQny5TpQIxk1^7jvkwc6x@if%3XLGMQs!V@imif5;)Z~!($Z_1`BH!{Q=@er;HPa9^J0b*SY94^3c>oxCzg7 zS4M}M)cuG-cB@Oh4>nkMvd-d9byu&gw*cviWfWH|+iTZRo*nL!2uCH~W+F`ZF!RM{ z*#DfG=CWTkqpx0T4qgdi0OLaEmfXf4WaN+Qy)9uR-L|t4&RE6yGeN5K3BWbvYU68Z%6FSOhu-3FCRR;-X4yh zPofe3HuzdtitV~$FF!wjC@6rm%>>htexlG55>lZ@%gLI2P1wPTPc9aJ^Z7DHu~^Z(qnL zdjxDv8y*0KCq0f2_gj@Jp>Vzijefqo-X!Vjb}fINjcRyE*DAu6IECZGDvb%0Snn)` zcD`9XPxcxBm0zm=w#fHy|ic*zBC#y_t!7Z8pxHYKu*Py{4w;$CnSaVD-?r)#do*ElACvVPhyCmwg^o{yX7f;3`Y#7!loZq8_eCJ5oM<;Ii>GZcifn zYut`3Q|_;s09F22er4T-8R&7pH(=2d{QWJR}3aPb-ZVjk~a==#>BrazE$&BbkDs7k!00L!f7^)W)y zY$(iV3@AH^CzXM z2#mN5(*DZr`T?~X)S!O28rMmL7>vdyVkMv9{=!Nf0e~-*3qHC$!~@@q>7)QHoAMXM zogS}aVvNF&-nxHwc^d-StKT%BmTl$nIOC(2Q*p4|mJaVd#vjkj@vkz!LND>fEN4!bE!#lN3S6wlHK?rB+kSD8Oq>jyiQBMp|T9hKPE1K41)&ox!+}W z{?bs${*cpx@PE%s`<-UJp}>#|z;)FhA#6d#+$Dp6rv441Vm-#+ z8^oBeLa3LnbC}#5E+HDQXF21=!BbJ@2t=jj7L;W&Bp{=9OV=c3sg}uP`JgG6!_F;PSVMhTW;bIBr4Mr&louX!@rx|W#oj}@gZtov# z^A)~fP%!}p5~9UrwU46mdA4zWvI}h;DAjOwv<=C=t>8Rvx?2}l!n$IU_`eD?iI`Zz zUx`#ACuLajwwSqLAiCpR1V;_$TME^nFKH{hp3#_v zTK9V!Mroh{`?(}*wF|IBwCw^U=>v)20QGq{ouDA6VFffS#X{xN|LfbAd839t)AG+T z2RrD+uzKk;X*{CXLVq4XmXE-mrZGxXsVB5{O!`(=5+WFC24kt{pmlJXDOfrOtgosF zz1K+}6Z8pbzfZB!^+$nMOJQF@(r>(RND+=zCt~{8Pj8*{Iw1p)jOPjJPZYbE$%EKv zFrWtiAdhT0Cwf=^2{a$z%nl`Ipoi)2eO#&l35Wi(*gRZ#b`R$X#LrjKGLC14vO)V3 zY9LB;PAv}9QKtj5DF?#KCc@gVa{Q}d`zjmBEqJ?V>V%Nl=w?nPNShM5o&Fn@=a20S zqKKmrZN0ChY!!vv%~@&-d!)d>DNu7|MsVJ` zW^C2rf-_R|?1q-lNnju-J^C23T=dV-vH$JRm--B+Ws^8~9T!~Vl}(wco>oY}7nwW3 z*?;xkK(|TXvVzjtGM9KrBzr5aIOJ^p&ef8+<)6B(EW2iLa5AxaYL+AuUP9wUdzpJ% zmalYEdVNa#8oj%O;{0~UOWIj#L39Pin#&F+;U8>lS5IGk?CKH!XK-QME0n(RW-seLubf1)A#td_Gf$wBR9E9_ zC~11S-R8Lm5}9q+Jml3DkbpQWyt^@Rl?CChKCLK+2k9NYB1vluylp-HB~&ofdRMc) zvNB_q1;x_f4>yT&d)U^D7HTYg07craa%Kbujg&VXfwXCFjTj>&)GGXifCddPIyalk z@ZON?E*+|ZtVa?$A1VM*{D9$IN1eJS=r5C#N>0)i$x@R>D44Sd+ZIu@j0VUwF&L94 zUH$B!=lt?EmKeqD17l61sYf;|7iWgS$oLcuduN^$vbK0L9`2- zuYPoI*6z^Cocpj!iMELOck&TH7baIn;q1x@hv?_HnQWcJ&i6m3QYaP!vx0BXoGwAZi-fF5cP(_h5sew1BR53gdkigsVfH2bJ-X7 zqSbBoi&U)>;&$(te=zTJ`gYRmNz8aD7Mb*iQ63G0l%2^lZOR)l(qCS(E z6x>Oz%TIn$$lN9sr$FLs>cj>nA#D4_E~>(47jxtyUA|*B9zdHGP`eeeg@II21}M`f z3;QgqCK&Sw#x}aJXDUCd)-^Btxx8U#&{iLAN!dxmQr)r#R~Yel>C7H~KJBBz#p1ov zf-wbt6T^g~`scUEzAP(r~eDimEDIG|AHu`j!_sXAktKGH~KN7`XciK2oT_tm~|C)vd_+>5Sp* z)d9UZ1a3UDsBsvCB_m4n4QBsK?FR+SRz`FFPO;~W8oi)pP7M!1SPHVKV4n+Cd5me! zgvlX{FVb1W!&TOW@sQrZZYqN(C^CctL~#acuV398yEV;n>S6{Ln+fB8!Y}}oC4{o3 z+qxv*R|=!uaV)^c+V<1kI@wQ4F#=Q_gK8*ROi!rR?PfY3kyZ;_PbKQ;_$!D`09DjxnC+ z+RnyX5@cNjtmqnji9ZL)%{woQC}G>4G0VbO?V;SB=owVHn`|wB#*8)H)}qErz|w=ozjZA%Q%~- z;Vpv$!Zb61FH-@1mx)$)xAj|;XzISaWbc(eZ@XfI896`&ku0gF{j|wqwQ4?^wL5xC z4cws6_#zr5rUCN7h-+S)juv{{-ScC$ffJ}5@rmG>f;*)0;I;(KuNM^O#-Bxetz~cL zg7zJS>Uv}t%V$aTlqW0)PL!Am?>BORds-cddNFbW8pm66vE#DgmwZDOS~SCZO?Y%6 zQ_0tt_=4&DqbK4>vAO>DT5&NP_EgcnO8(>rpuVuA<1a6^xGvt4E<|pWZYB?!3Z9+d z@JG=IeaDN9<$+{9SX)A+Tl!Shzpn;Fq#F)*AC)=9NiA0}>&28M7C{MrF&qTju1;R7 zsEPZD~3Ct##$AR4i^0P>{IXU}_!EydHIzcZLrt3e zY^fa+3zF{mA}|ya-`bn+CON0VgFv1xrq}h|g!yxyWOALnQhxFF@g9y%6&c2X?eHu7 zOy9=k@8IPc(7NI^8kO+(zB!`pA&m)`vPgMe*^Z|$k8!$7UfsXnza4Me`?Tr?YioUo_#?|Az*;z{Q=HpQ8?cjQUNChy_O ztBm=D49DMLnNQ`(*Zvdt#g|D_vhM06_%DQ%r@qx#aWVHubQIQOVbRv+`j;uw`L!Y_ zAuZKYmF_U&A73f>4_{Gsn6iHGW66Y}uQs)KaqUo0g`NWwJolGkV9z~2JUNS(WJ(tN z23z#^1S^{<6pEd&Uvl=M|bpJgR5Xe(IuI7s&b|=my7GUT6eYn@B_tB4l}V7&5A^ z@83ymD4|VeNS#M3S;ov7pn&_~#d;mhg8vM;u*K&oQjt~o1{asEL%Zn7M-jr(^bWAW zZwGsOrmXO%BFdA6D%)8%;%)(3Ews;0`vwp%nGc*M@LYJKm=;CfPntXG@W({X`74GeA&9~Km_`4IC$3*m52am5R z=V>mA9|PQoi^<`0E>rRx{S~_ZPzo~`1p)QkzRD+CpJaGjkP^01MX3iW6QJxDNo^Nl z5{*yJ>1U3gKeQ_Xe=rX zfM)|fh@cxh!3nOlVQLSCeKaK8RFaDTzu(#3vBSD{Dz{rHmkt~tNgo~|!7YeO`Y(O8 z_%Wu8_li2e8Lb#*cmE>H0m-#EqShf}$qv>EWxbnR8|$qcR290A8A&h$yi+uqw2MEb z=~nzpV^ROl59P9C7Q$<(DqCV+iFloypUJNyAMQ;e$G1#-^@Q}=Y~X?EV3u0v zF*^3E*L)-3o@O~GzsLGjG+EAnQh|O|5&l=AaE(HoD7S`6~iO7csc7aE6gcDwl(ge^O6)lv-$%syi<)EE@ z0S^YZ#|06fen*3~Ut2>@a{yofFekhN&6KX-He;|c=?|7*qFCJrZUg3GbF+CA3&knn zh{og`3qCm-8AqC7+qCV9Xb;F9^)+yj+c%HaxQbu05u~Ga2AxxEFzjKydm<>v!;NvPy5_tciI%zwux7pJa7yt=1#mzsE)Ey79EW6Yz2B zk5WawQ*DWHz+LX$_v61c`s5PAA)%JHukWr;t;3pmBOTJ{C$=Fcdm-s&Sy^o$-8An^ z)K3;F%JmwknklRe4Z|S}rTnJ9jW*$HGY!9T0k^OB69|!T`mp$#yQE|XdL=7WzW$nn zRCnl!ms*8u(EDyq@Yq}z)yOXiVfw*_UDfZ3p7QHUK-1vb;?@g%Vt|BEO#+_jxLg!N zA*a@knK>$wm3=Lx1j+(eo6N0!W<^~!ZwUrCo5Cr<`#Q>>vcHjsr@Q$k2p>!l%F7Bo zPM^%cyr%j7LbC&d0pEB*~>m23N0QnIU%VGn2EBPkc>&6&8P7OX{eMo2&i zHJ+Am0na>IK!$$JS&KnfNHv1Mes4?dS%rtPgxUfvr6WodEs$Qqls_pe@*BmWxmzjaB&Qn8J0Cxq;S#p%Q%C3&={^_4EF(3U0?7# z%EDeReef$XxbSNIjjZ`0FT9qkpoYs>EJ zbjXCFRJuVR?a#kHWw* z>K0T;@PZ-rvt|Mf&fWcF1CbhjBW|(QrmB9|O)D=RFxVlb8t}3pLp@+!%M?jaZ8lXP z)R6>CjY;6cCU)eBBHoCEW*bL_W-U7k*2^XZ!w~|Lj=>p-!crS5P|xZnQZpV!tJQ_K zFXqvK?Sibn7bQMjGj5{8^ow9QK*~Bd%>rvNz`=clL$MX=iIhVO`VS9ud+rMmTZap0 z$vF)gh$L7<2~$4w;DivZJ@*s*aVi9b0{!5EpAK~XqDDYM|hC22zS<(}3C zw+k6cs;hjS9$Y4HS!}=U^h6JB!@Zx9CYtpuyjQkSK?YcJytHqkAtw_dS%!s-rmGQnr>&e zPem?p`F}Z<&-As*UjfoQRkvb=uxp);HrZi}RB)XOYt0B*bHG?(ux$StBJfW*E5L7$&u7*NJ4nDe;3RiU~ACz~o%#`4;9t(7_2Oz-&Dlb(>xoe=P2& zn_JFI2)7~2W28r`Cj zOY%<23+u$;F%|h3&?WT*=eqLOG~oP+quee1UD6IuKX=5b)4_yCBbO0+~^9Y}JeQ@Bc zI%jqZ=7A@F6R_`OlfnC0RVFBZjchnWl^+3yq{|8=fMorW(N43KCw6u}*@t?&@yuOT zwvd@D{NJVv__Eb!%deJKvC`V-Ci?+x^6udH8yoJ3HS9q~nqJG3FbhPi#Deg;mPi9~ zMoBWKV}ysN6qexT*#f7VzbR!NbJXJr^2v)rOP=5k<~4*;DoD4G|G2KkGGOJ-hDGFGfFH(aHPzhngr{^r&{6s^?Edh2Kn6U8(6W>{egnvBXMxt4(|4Pbb zpVqYL#D&%`Q&C29j7VD11_UOk)b)cSyz=bOxxUEC$&*~`l#XZsxvl92Zt6C{_9A8F zn!5oNEOuQa+?!+8^~3sJPPjbTNnjN0ozsPPjaMQclO+$p@^3VFx%MpYO<(QWaBXm&%X&mC zYZbN(R)EGp7cM2=@r@s1h*&JH89Tb@Tu;72^Mk3=C^A}f|AlGFpCXF~`n!bbpqr#_ zpX>!zU;_dcFV0S!wBvGttq5hex~81(EX+bXY|(DH)Oyuj9#KFz=|=$~_Ni8NPVLgW zq?MQOhw-hyRw*^`)6PF%L#YhU9jbM=Jj>`?d=zT%vweMNSn*#!w(u~ic z`Ld)llD;8#e5&Bw<;x;dx3^zeL|0s~+SkBk$0Wl!zGW9TU27v;s(}?5ZXL9ntNkt2 zHNFO17LcYnC*JXHuwzh{&p|J?2j->YM%V36aXP1U>dkWo5^Z364Ji=c#3|C-2=jt) z|GT*`ms()`cbgN*ucYhW>icC-&*6tQH99n0Fp!bI3UEEq!c>zf@N z*RWYz?YS!oZORp|vNJR(uegvK*{(Uw)EFI4MmwW+kR=!)cuEt{nTC-p4t1Is62mWdT&r{1 zpIe_=F1%3TqfDW@ZO5KUx<~C`PW?fufCiIdXrkHqhp?NLybysv(G4V|K8^3 z|FaWsC{YwDIRTVAX&|~7Rs$?~Hd{)O6!dTZd5)sAs@_gbHJ(33obuv5-wFZs6f(LgtB&Ll}WiMC32BZc-={Hn|!nP`sy~;{2QTFEVo$=Ll zpR`6~3h&1PhRObZA;{RC(}8@;Pi4&5VE!LhuM3nxb?FW;03(8cRbwc@#X4aw2OR|7 zn}K9V-sv2KZA1^Negyqtv;orngYuJY> ztdkDYgl|X&%k{wDZ!>3EKZBL)?)gT#e{)o>RvX;n;#&@B@>keDQ8)Wv!=0MDk-{I~ zrc5pU|HkuJ{tM4z=VblAd0y4OoWiD9%9i6yHl~iL=wHG^)bw@5`pXjDB#Zc~9HGPnXYsriwx;I~2VEYIty#TyEDa(Rb$`me4{ z+#?Eh@g4$6AJ2quEgLyqT|8djil_fHSFaDxzZ7waJ%BaT<(P@vU#MEwla7a7|rX%8SY5qrL46iHnG26d ztsc^I_OvuNBc3*v>lN_W!{5JRs+hNIN|$97V`U?J?EGWu?ez9HhLj~>ii5d~B-@M+ zY<{k0Up*P}NefETlw)Kw&STg_!fiI^N5IqO!tb-^}|K5 z<}BOLLj7|?;lQ~Gpzm*4cbzdL@A_{X+5KoFu_cH5fN+eWJsr)JE^W0H(ptZGMcbIb z4Y#uY)SaWLq<}woZ4-hx97?IUPsW$|j&3ZV;r-EFX~a#cWRuDqX;Y;4OT+H9K%E~@ zl; zh5T!yEQajbRTu_K0Ew>q`*N(7W$5}VNe~D}C~hqE=W42AA`Ai2Q!#DRIopSfnU!H1 z3C12e-O6_SU9&|7$(O1^wXM74%P@1ekK4ADbN&3qq|0Ay%-MiF*DQy;UaW{N2t48u zy(mdbpnNXsWE*HepU3fx`D`9CfgYVi?>&B(7|aAlx?{1;_=2luyt%6R$Yq|ab7B`g z|D&C~;|kkb(>I~Bn*5e9vrpuMmFDvuv<0FDw$7>|*5ne1g3~S#x>!o7LzzU-QXi~a z@3VDVd@|U?J0t#7pRCcL`gKCCaCr}5VyJvmMcM`h^W(yKX%hr!?R8Ws+O6%+`c{hG z;r{2NCR$#A1yIX~E+wf`TzvmteY)y0#T|xaT#c_#Hv`=IKOyG1@S7oE#ryH7`PaHzhTLA2?V3!;uXRdvC_(S`_@;FBo98l$HVl7VzK z4uqku!;}*ps$HpaVyY+*lPIvzB`1}v1qY{OB1pvE^$6_ti03kC@@da7GgdjZ?(fh zJU+pE@_`2b7!%~-%m@!Zb(D3qF`Qv@)ZuLbqE*!MXdOR)EoKfMc1z!mnr=Jg#v82h zGI&C_@vqgoL2zK=Pn~cZJG*^bRZZWvkQvvwcWu@)3+k72#;($A>%wGErS%LQ3eF+R zo+$f_ua$~Z_1CFE5}{42Yl+X^M#Q58UMX=7nvUvspex-X1;h%37ve8B9Fr~$m+xP= zGp4-xDG(Nc2Q(?23fGOTv=mzqa74<%c!K7$;Gz1J%q?wuuOGr}pFW!O%{`)XMBhm2 zwIK{aPiBr&NYB>p+1UCfkhDfq!!86%kRB3uuhuPSyO;Mpu0ad0i-dftKMS8~(GX-> z&sNkBQynv0ahsHzap)Fn!8ivX+SB9USQ7;}Dsg#$O|PuJZ4$pt z@JpO_EDq%&ZD5APzfnzT56DlHdd6(4$h%nNfkSC6aCdl$w%hzx=80?|ECKNm@`sWO zi{Q10ekwnR*Yh3huU^*(!QKKU92`-9;jVN-bHwR7<~S%rIlaDjEY1G9Ajyx(G6k{Lqp!m_GPFJ zFJWw_r}e&{5^_j#eR@bzu!uvRt%7&rkP!|=m?W-SDiQ#6V&6Vi0u%L>8{ir!woSb+ zV%R1$m!9KG@IDo{EdFCvuAyKdf{CxN;v~p*(LTJAhEe#ASa@k4=Ba7Y9L+Jj^d~ZA zd32{RO4*r^@jAy2#-TxbqzWjv6OWm1K25sJKPM+>&Ca`T+5Zc z`>ThdspQ8=%_qM2A1Z0m6R44JOXl8)+{K>XM4o!^jAdrouWZ0$j62lfWoROHgQivkxfsaXCf?-$(a+` zta_OesC9AH!ra(VDNN3I#3zAxeOnZsN1z$`jXAb#qfAZ*nX3mXwPCu->#zS_zttVX zc4~Ya&@L{Y8rd;aZ}|c6Ex|%vtysgfR%C$gF@ll(U=X93C^66xRVPh3ZFS%Lji=$h zmwPLd6;4xW-KE&@UrW! ziys=W^M?qOav=A}5#kcA^{FQ$Kj&?QqmUjiGw@De{oP;M&9k*kfAv=uv)Rmnehg;zW+umJ7?T zqRWTlm|tv%e@EH`2b|MHO6WOdo-VaU^YwtpU5a#`S@?#2M$WaE*kAOQu}3N|adFWU zbG?XL`TCd{ktHp7w3HA=u}acmMK%M9>Tc*1VhNQb=##Vokq$=uoC<;;)5)y`q+4kM}E?SEGD-Y@#P#U zJ7@2v>VY9PQ2K3yw5S~8VM(-*C}(efce(Y|LrZcqHnsqkAX0iIpZH+kY~I-OrAt0| zeJoQhiZ|6ON2Hi48Rk!@8rVz~vUAK?4eUEpXe;Hgw=)mYQV_Uiw)})NR?zZi;Lx#y z!_o7kw)wKM^EAis7N|4$@>Ht-UdX$Ev$@4Ay(Mfvi-Kbm$6{z66DF#39*JS2Axg@e zn5UQpN8{q2Mt$3DUD5K9c2{zhhtTqBi)_iED3MHh+SXPW)$`UV;9#p{wO@V#haPHF zf#b5JRk`<^QcISBMH0Fz)3Z&%DwQ17Us$N`fe+}hm}KC~e(=9 z$_&w#F-am|%CqL|$pGI*DjvB6jyD}436F8myY@Fy+JuL5Q*T+UeAD(x2Pcq60|Iy%c7D^guZ7fG{w>KKA|pq_hl0-g1&bl%Q2YFCT~ZU{k@szgSC%4)Yj zY<@C&`?Zl_zXq?-VX0eGClxbnX!8d^od)$fQJ+O=5n!MGnRm_fGW)9-2{BOO?yolX z$Sdwvf{f>&YA)G8$n@G2+}+-3VaVEGexw{i)U+kB;I&zP0FTJV)M3wYH3JYQ8pd>v z-n79N8DXnk&Pipd@M1p^H2Df=j5# zAvu4BmoJk?0TZ0t{l@EX>_dBOqk4BV`b z&^(ynMKXQzQp>bpT<*3=}3~h(YJ8bV!>aO(r=)oCKRZj18 z7QQo=k5VJPhhbc4WYc1gjU5PPvhfl_C8f1%-Bvjvs$If9s!!h@5h4_(nNr~mC^6!x z2JA4uEp+5RhxuJ{UKAz@59lo7UzeWfnw-PEzdZQV5x?k^_{D_Tud^&oxhUIwXvuT4 z=sH23)r~%yiwSUGC?}$w?ddZ7CN5# z+TSelo&^)B%kAy=W4XN@PEe^A7KWNg2x#jZ=!yUCE=cTj8Y=l3`3#78!q98)4ui0M z$%K-4j_Q|~-n%Sw1W@NtZaJ1kw$~h9d3-lszM<|MVV$aU;;@wtuu?Jf!d=R4Jg$H>OuUXz5cE;?9&ojho#ymp zIT299AW&s zde4UnJ@*FTEbbyyre3tmmy_bN;XK{-YX4)RP3LOt8b9pWsiShM*gm?ucXLB%&B{xFfh-w=Im(4x-)T=McRc*Kip79lA4gF%@IbjS*X4?HhJc$W6=TDK1qaqfEtAKG{A2btOjERDLNTf^5?}Zu zdgu}Vg#apB0rb4u|rV3PIZOgkJAZH}@MQYOqCv)Os3RxyjqkPc#si~IaCqN?wjm`!Q8hIr4v0#x) zP1HA&e<<1UFtSJ+ImmKCpEaoOEFZK1mDt;EDS{wug1H>%?rZdjN`a6SgP|T}*+IRi zinjtR@fhy#duwnfX_j{YG#@ju&Wd}V2YwSpBdAC$@Vl++>Cv0U>uC&wPyqTb{QY_HGQcI94fddA1 zso@hyt`<-4lT=+b*TdaMu9bst|AY>0r zX?5WNfRJAi-Sig)IV)wP;f+!uReCh}xpgvN2-p>4dDFRF^&B#Dy3R#?&@YreT~+t> z)ASVqx?G)&1wocnInAL6mGAd4xy*k=7*A6ouPv6>UlWA*?tloc2v*81{9L>ZQGV5c~fkPUN+a3g^{FG}lIT zkJkq_4%4lN0ty?MYk#An#(SYaevL(iZdH_gLi~vXiYm>HiY`WtFPb<5wxG-RN7aDf z)l30Ifp_F1&v?}Lif=wu{+_LiH2+9seQV%VsFhDfaLG8U%FYPS=6kVB!8oar?FBwpLQ*L2bO(Jt zbn92=<$H7IW?Z&&J}Q>fgZ;+Q=d33_<)+`%H>J%90YervCg` zxqSe4hJHii;Z`O3wn6Rh0q%DS9^K?8);HFJ)4v)>k2s-lrP4jhy&$WII#l>5x+gdx zWt}hp;23e)uex3Ld=9^1><22vAp87hS>C8Wpvw}t&Ks#e&D4xY6Pf{3n)|-WG`wXW z1;n>6*VWWDBYh{h1RN=I%sw}4J*HDmj4VyN@XSU?!=@h)uROgH#0ZFp@hQet*vxm+54sKpSEUpiFXEvG9zj9#;FuROq^QT_FMvw# zRVUOxZq5uVDOn{XsRe_uHioFYpy`&aqIZ02|N!Alw1m+uzNVizkB1*FVHzJ6%pL-8CGZ%rTU>2bz7cT} zQ6h7=X*M?}A0NJ$Q(+r;p?>79&sncj7*J8zTS;0lt9P^-lpNC6ntS41$i8=%^dTR# z@gK^@5@_q9BgkwBwoHw-U)gfxTwgZn^|6oBlurVT=8dX8i-qP$NPntsN(9)D(A=5^ z*Sl$`ti~D*#0vu)-nwVQ?4gE45_W`xx8H6U9dxHIoUGEmAl5#`x2%O^1V(fkU7?z) zo1>nL{E`0B^4yVSUPMGd5u641f=6FUe0)-H{Ol##nPMavsh@Dy_EU>PF(xe2H?)@a z#TI!5-IyTv7fO=uJ%!^2lZu}x?KLqul6KMQfg_%E6v6EddrY~dzI-_x7K={Eaw&)o z-NuJ`2mv0BB-AVDw}LuTvP*l5ZoelhM^+cSoIzvwCu_v{k$5I|Azd23ecXlTboHMm02~(^0ylsZ6qXmqM1}p;ci_ zmVHBLs8_2Jxf{Uip&}!rGGNa)2f5_+9yr@rWOvEVfSCWKovx9)M>$3g&rLS0Y|O01 zhl79%ud;IS-z9K-;*P^O?Yb9Ut(ltf8K?;;<7b4P{kPnw%nn6IWEU35hz+^ zt;g*?wgYvDTar4iq-)^dyaO+RDk^USY^vGw zsc6Va!HK0^o}6prW*E7j^EN&NXnNb_d z7wLC9M>m!vsBKKgKsB{Z@nw%zu{0XUBrY~&A)%99Qg z!xq{*`!piz&2MqUHxGZ^zS(9~7Lx>|Qnh#Z))PyEkygc065WX;eYUNMMSk*v-2$9$ zK@m&6=N%nYb>Hup?t!g;{4ibfgdE~#?%tcNCeX#{#~1d@>UNl!>NzviB%uC9bd zkU8MjS&1UNLT-J~cH?mk2@c?i-+bbLjXf7P(Vlkg1f+}GPVUl=f_4v`oRid-Ol7No z+sk})rozRKyv<_1doaFvH?ua);7SKpC2es?3C$f2^1e91g*l*N39}OzDka^Qg2)VN zRwfQEp0QU2kjNM*5Br*=-)QB^k-P3J8)c2@jXc}$U+|ljqr9TgdWg4@RL#mhiWnup2ZhYMRL6eV(~gb$W|4VuU| zKfcS}x#Jjp1L+T*d1c&QvGYgukKym(9wyzlc^La?PS>!`xUdw?Et-63g);DN$5YfglBSPIXL(7` zq2|E#2(BN#{jgfF9=4J4YAH_2M?E6iPE3nfA35Qn54*BBY>;mZNj`0z23+Cl`InSHWm9;SBPDrIsi(u5hWBS(8{haA8%FI?KtSu@REfmB5{bPT3N zu3KhF*I<3Zvd~wvk0QjeZ6%suBfbXN?FB5ihe8)^8uJ-R< zA49)0gSDUx_e?d($%@O#gx06)2!I}Ev;Dj_o|(=J&uMt*OlZRtC#9Y@#9jIweC$VJ zA}}l^x80D)MQ$@=3aY{t(il-=mY73DI8I0IqmE>5!dTyhO(3e_{;anQk!sfGBqx(A zRgq7}oKjohm@#M9P<7+Fz{cl_Y`m0;WQ{r)uAW7U7pqud4dYU%ZZDhAV-v+(H_xY++n!S%<_O8jI#J!Yu84Q9t1-xwclCmgulMW@ z@dPf;>?rw7Khaz$8xKle@~~C%j`9vfc+`!vUaQjP48$?Vnf`}Hb-nCkm&8#S%TfT= z{8yd4)^AK7W!kfYN|sQ?(!Hw-cJUKtNg|CpIi|M{Ng~vd4+xC*C#{nfT9z0_TRW9J zbyZFNuV%`ci1T^%3#uiJ#+ATeDk%cak5pA#mTWW`oj?%EK}fxS7s^zT}(2=AhdVeD ztZ!G)JC)9DcJ~dP=2cf>Z^9V2I)d13(QZM1yO4a3ZQ&qF%r-K!&qxvF+3^|=Wi~9O z5;Y;O)zlh)u^MI%wim?o-+vkh>1}*=lV8Q-pYf~UIgWbe7MCCl@*~V!g6!!IN9RA7 zTW;ytX(XRhxIyL66i%(1NuTzkuzHqvf;u+;OmkGu%MZ!`=l!SUl)QLhW3M{wYILY62u#9T+rGF+jj$J+oF zDIOF=F6(OB`-%#$sbs4WtJk5`>%#TU@l07x7YXks{ZS(udTbr|UA?0cTW7 zaM!nYT_I)yv^P(xK)NoQ^@49}zY3>PyR|t0){G66aG5)N?{_m{s*p-5(u;co-9?sk z7U;3P5S3SI=t`0tSe}aT6lBa)>thAYkX_)Y*>9C|RF9L2qqx7FRMl)cVIcu7sXY9wA$Q+8AC+S=UO>J{MV>e4isnOrH!0ATo3D^vgwUA!Q#1t~%J%1a{UwF#FjbwC60fOB2^fjJi z9)*O`Yp^aiRl-A7(AF_4LYRyBTCS6a+>)^P^G1#N2q0#_AQ&`8?#?ziQD_YeZBCek z(!Qok(J57rG&_%XH~ekdgK}^M*FkysfXPqy_Lm6=Zw;xaBRkxi zPYa9_{t&EX#^-dM4Mwl{tcHSg%y_*iGnM^WCK?NMd+t0n)5O*3mM+!QKL8EhLK-#e zMt6Dz_6-XzeyQlLiJ;Hz6MAii_*_$Ezd`WdbGbm+p)VV*f=gz}nD8zAGqb(}PUQn{ z;R?A50>yLsvDe7jXplNQ0DUGi!!T$;`(vnG!Vgd8A_RxKrYZp-V#wa6{q+u4?9px7 zsQi65z}Y?^YeyPk$k#>YUz;~#RLuCJ;pQV3;{ADHz0~-o5Ol)L5;ft)nQR!-`E{6H zkbgcChth!)A_1&05FC>qHd8FsZ6BCZQE(CAo>7Bz+h$xqL}kl~W0NE>G}^b*vpQ;( zS(Ht-@2smA7^fS|Zqd5FOK>T-;%d@m;#T-F!Oq=s>(r{Yw?KzT`mwg1!0(aw{VgnZBzQ?0aJ&@{Z>nJzXJM+L5%JAxl0e;bCt0IGVyDAr;d{g z&*3BMR=Ht`4f9v=0;Cd&G*$Z*8NJa_chUPW8|04*Ahl)T!>Der?3+Fv@oD#Go@fEw z+%Rc%Bx0?=YNCRv6by3c4@aJ?^IIb=p?!PV($|d=o{cGd(rO6Gb+y0E1{BJi#$__w z_Vis#moG31NN4ST1AJ`%1@N(O{=WdD zvPQ=CnYp}-f@;#>HMP=|;TUbtG~X28Fz!dKZEfh33=nCBNSWC(?;Nr<26TV`@xw>A zi~zsve2~1-C;cb4-c@m8W>@Yr$5wWb?{UT0+|hPqtdgA0K2y9bUCqlUj+Y^TI133i zu}#F1w@v4&ZXvd`Mh>wwmA_Zo^fu-&yeZxPg$~p6a+E#JW0qLDk*(~Jw%hDTC2dfX zVs+`UCe_xpojH6@D`@x0RhTLL2yM%(`E*vef@9SnwuY{S3>_cNy*6fjJ z;@o*{c?y(QLTQSxZlI!~q&0V1>6Ho(^Vjc(W{p>jX-yaTz{y4<(SMJh^93iag9x(# zylvJaCZ{3VgEPS&;Ub~AEIGQD??y=8wK22W?@Zn_>wWlg=Y+N~OZn1uGBp?16i-a~ zYUEIK|2|_&r)c90jT5`*ZnGU&x=FoKX87!q&Ds`d9h@|1jgLy*6jg-6siJ^*=;vzB z(Y2O6?8{NQdrN9uo${zAcQ-On9J5Qi*-LnU*{tetH&w>@%~HWDljjrvTk0za=4Gp= z!N<`7w%Wfo1nKXTiFR92c^VzdpF!8*drRbhLlOk;@tXUidh>Hq4%AdoBSCR3bMEvz z7W;K~#J2RVa&`R(OWn@Fm|kpz&!cv8m`5lx=M+r{`BLVge`z6v6)ajpO@Nv*)og)& z_AD{$CAD7R4-K!H-ZY}C zGOa!-6^-BuDL9EN>qyvqr3QtOFtD~M-2vkwBTG^q)}c*AiEJ>^q?h6T!lQ}llg}|2 zYXOoe<`5tPD>aoZjd!ZFS9CJ+65H~k1DW!L0qJT~>K$3%F zJCYz_Kh$Gk(#Q=$N_;@Y!VlOeB=;(1RF?2nX8a39aL~?fj#iyl z1WSm$ro4Ts&!jxH?=k>USB;GkV;kC(-`3Je1E>Wpt5Iy^k-m%ur;{xI&qe!CyWq*p zyowh4bGXPo4SL<(<=0T=z#0{J9rJEUX<<(IcGcS~vzboCAdt;et0aq8cDwLqJ?N&; z-?-C&U4Jv@roT9Yz@4CTW4d$7x5%Zdnyqqe#x1?<=gH>NdlIq)X5=dRbgU*H+s})^kamV3rYfmKw@>*2`f~WkpyuH|q|nKKZJV zRcIYH_o{bmknMYwHabmq}O9!Yay$n%o#PTiw7>$gNZ zI(6{9b)n1=ZRE|zR^2q}y)OR}cmr8LQ`mhreRe~uQV~fM>k(!QbB!=9yES zTP(_r9x%tF@hr?ZSjX+=1jF5+0)<)Fph$(f4iR#{(Um`hTxM8+!2OC_3c%c4H56Qu~aJPuxQr znsZjJ>8CeZV?Ua@y2J~kyvGCohJwusPYRm-6~_@z+PYDo5i^egS&vlUh*gV!3-1%2 z9Wd3{sM{rRsNVOTSPjp4duOewI^-qA1dMbou7Lw@U?*Zt1D(N7Z15eYe&Gz5JBY#< zo(P*grg?@&+$Lv4`1;#iX&+C&L0^H(ea~T|StlUd#uvtn6B2+P9>ivE8ZR5a;>`oa zfE@f=3lkGr+BAo&J6rNi8`bgAhk*DT`b zf)1oj{db}N6rMbU!NycvA(+@nQ2z47RC#ZWV-v-xF8Drt(Ktq+ z=u&i+&gqkGlE$spKw{-zi~*vqoOz^X7jB_cZHNwhl$iR(uo5Uhl9{gvlP|n^d|v`3 z?>3f-&H=-~ARMRYGaon23bf<7zx?nk6TC$-nDG3ti8BTp!7(6^tdrUKM8HNDHuNZE z1*PNdMqGdtKr#%CR8nfnd^|<@^?d z?(t;owMonc^a8wNY%SBa1UOoQI_QkRe(cThcdy@Q{Y*pbLYpfv5T%Xv?Qp`VQg)JW zkwR$*=Tnb7tTghTP=0frOm=02!bLQ-<_r(?c{n;M17&)?4C1O^Sj&u(EWxPF&G#cl z#W?%BP~0kCvL(>K(GgPj;!^+7s9E~EP<+~TuO6mmk4|yxVHuYU^v#BLSujCZ(9uIx zUU25cOMz!YbYUXMpx@c6!I<<81v1*@rqMEV9zILRPq6TY7M&HCXqKA>o_ScGy!qIHUwKvaG@@LGIN1KxI(Gte(-ZI{PUYla<629r+avoMt1}9aWtO;3SF)b0B z*fPJfm8aT&ePb3DroZ783C?w|CkW=A?XUu5pL9n-e}tWqW48hgm%;~(^l96AJ!Ldh zOUT!sIe~F9$?wKP;t2#2c4EHq=;nm1oC0_;NXFjbe;7Nb=1iciTfebw+qP}nHafO# z+qP}n>ab(GlXUFlWLKTLy)VB1u;!|oV~#N%gn*b|?ucXvWR;uOR=<~AF3<>o^u;2g z@y=Bf@ww`%FNqssK+}m@OeVi>4&Z>dJf zSQ=$Srr^l_~fI{G|EdTf*zoHI@l4Z4wZLPewbd8ho}tvycYGk-U{nfqlHa zbKLJA55mN$FVewonDNf+09!%b79qzhDU~iZUU1K@Tiq9Fe98qtHJ>^x1WuZ8;(v?bSQi?Kp8-2=h?jmv+~B^qXE9fM_3;H1quNb5la(ll1Dd^1GYp$xRsI60T)8*#P=G;3rgYoVs=adNwA}cV`ySOiu>%XHG+}b)VYaka+r| zU|tgb@YlrKKAY)c{Bb(NW7Bs4T!N6CNMT$UsTBf|+ilgDCe|#B;Wmn-6n81g)Va11 z7v3$59AAC9>q_b{!4OEsbCzgj`if+qCNT{9vXuENNn%_{l$#c>>ML~70LrhZ*ILl< zUb@q1!Cp!ZYegPSrd!pTbv+dWx@gDJyz`^=D9i$4(yrX(ey~K5KZTGyJ^tQAet7mp zPez)uxM)Wm*{rHq)WwKqyuMJAJN%Y6lo+MbHB35_>|I4qq)#Oj zRe=;q;f7_#%{wUB8@nZ{F*&4cL?$_S;TVW40%`p?uCVZaBWmf7dh)p>$%1VW^svS? z_1Qrn^&(Oj(KdM?|AsBZqP6v;(W`nWSAFo%Xm0n)!n51-1Bp%>NoU@zHO?S-fdW3C zIkL?ao%0>(lVY;<_<>tWdV+eXSC8SS&C1WKDV~>P-)RL5?1Vj6In4d_)BCI`au#EF z=U!8<%X%yGlv6V0LVk-dNoZByhO#J^2I-N1VfVgJ{@n7Li-AYDMHQsL8e^)+`^Lt6 z&Ha~;R^qo;38-vmYGWQ#+5W;7{4~?h-b{T%E11Yl=}^Yl;DFttNP8LiH3G8`c({MJ zGGkNf6q`|K&)@f@ut^$|?1Q}1!$f7Ig!$||gh&wm&F5TZpZ11!>H;!!-B#PMwmw1o zdgo26=W=k{C)mS%IxICkJ;k^gW|$ECa`NKtf+HSqtifLWgKEoS!2|~Q6^&8bYda?d0D6L*Cw%JGmCGp*d z@4`qLw`!$9op7&p*Ule-70pgj3bS4PnqU@{Q&xU)OxT(bd3%ptL5|!Q9>`ekW7c<8 zoY3@U%6#}Xv;}A_IC*YcNVhk-V}k-XqyzO9Yw%hKlw(Jz_9+^m=#9u&Yhrix%S%G4 zALgm=-zgq!jVeJrI z*t;ox5h3x2;cgMN|3-*5ouM3bovoYo82^wn7eb^BZiy(zg$Yz7SeR5wj(8rK* z;tqx%%`I2TQ3A@pj`e|^99#l_%t0W=!Zzy)gab(9z3}{@PT%#^QwKl+JVB`;r})*M zTP;yYI1ingwZ;hu+V8$Ir|@~EBQ*5B=DHI+=)+ytf}n_muPov@w<2Mzsts7ziNJbL zqu4XN*1eFn(|*=bj1MWQ>)YK^u$b3^`tZ=b!P5#M-+v1t83HVVBS+;U3svOXl2|zfWwLs_QMTY?*j$HZ*yUp(;A@x_sf7*n?k) z+QhN;JLz5dfe38=DsAI}>|0Ag$XmLnwz~Qp2rjma_Q3TIntNJ;8}F}}w+e%OF0}LC zNS||EI}J;Jj8o!q*F3uQ_-T)R=rG0W`TZKo2b!=98TWlno#}GjU2HfPVvN4=pSFr- zzt#r8(MH7bBrrQa4htb5F8vek+C;0eM` zbtzbi?z;_O+^2;KL+pi$7u&kupG9&O{52ToBu7EDn^_t+pB`;MHO$*9+}(Y;>5Mv^ zsC(!lpOJ@Rg#O}^9e3;U>EFTcxb0@PFThr_x7$gn%X^l4w(CVZ+Bq;S6g>|T`kHP$ z-q7^gqt8a69B_U*6b0HdnxE_ni6j}sZmeI9v}Cck_0#a7rA1&`LJ>TPnB)5U1NRn+ z$5v?FdblUQHr**C zqV(nPC|7^xkN^BHRU**Q05Es_VwDV=ULVU(D}szhc&!0Lp4WLb&s|%nL&vM9^gnd? z7v0uTeKS(kC@=dJ>mRUHg__TA_Z6BuDcLKl`fO*pmXZ5`vsXdy-!}y(qhksvX}}V3 z-U-o?7fE`BbK*pAf>D9F$Fn_gJ@CM=E?aVZeqjKFG5enkt7Z_U(vscyY-;g6q5kSY z>k7~72{-%Qs_K3Z(F>35EVT4)S38*H+P7_gNUjk>&HGKjsLaI2Mi15DeFxPsj!o+$ z3;x>N=j+Ux%AF(`*0D;GlX*d8l(d2WPq`lVh9!6ZOMs3_CRq{}>;+N~!09o!Cut6A z_T%M)05#-3xdZROjNwjo>CMThM%};%U-Ce+6t$$-HyC5?%dZEeH&{Ue@xdemIo)Tp1q%+2w<9@ib8%i z<3x)tH@oci+eGL3y);4b)cC?0ivyKZrLo~3DnE0+-6+E?>3|8pM*F$XT+_*`vjNOJ zVa#NSDosfDJIVdk2r-ivx4*PMNfM2;Z@yZt@}c334mX#DLfAD-Ase3y zNZlvgLH?Am^^g%R_hUJ$@Pqd@1Q5l@Tp`s|=c?feYpfLpsKV4zh7nep`4pV;4`u;A z3G&F4V{q6-8}5tsC_lT`d~ zZptg{B)Snuj)&)CQK3Lo)k)#WTlEqv35k-Rkb2csqG4^yKGDX%a%Vw&uY~e1tDA2C zZYasd{}OOH{#(Fh<>2@~upDZ&Wi+-Vn|7KlI8O1^pG{{SC#T$}cw(?V0@zJ5&Ns(r zgi)Oe-BML+ssFZp-+n~n#LG&7fo=2gg@ciqQ4=5;i25L9_TwI9@=n|`eC5+U-vz(! zFnA~J;E(t>9}iAX_ns28@KZJoQMo(3TE+J#Kom}>7A^<6i-F_=Y9SeFzYbFhRnm2p97`n^1d87O7Pej=g6#Z~U9`F1oavaG0M#)dD zdDKs?0uy|)7_HGEmeATRZQ3vgtRD=BJ8MU=uoZcy?Lo=*?5+`%92{z1UdT-)`SftA zWXAAOKCZUbo6nwHmVBY~P-5qJ*|WSJj{Mo4sD7tJ+h2Ok%85A?eWLbL)o^~fvNclj zTL`W`Hd(V@vOcc%#m=LC#Hb9z=ddfYWbXqkcd!1E{*`Ndx-leqSd>S4*Td^>I*(X-H#0_n9x^XMwz zGGX2}(XXv=&lYIvd>D?t+O9O>3tC<^*;3oE91M?0iy~qUMe!U7MOwpIQ%elL3+t*z zSKaW+*orR~^T$9zZT?~(-jcW~vh8au;I&x?)JQm?-)LleY1Mo3-!bS+G*GKFR3I_O zhzWWAS*|XDqhW6kSbc1+U>q5?0k#Ip#W-jKvjA#PEYd4s;2mVOdX~Gdyh?MN-Wm+> zhPG5(jI<^;k_E#VXEJi3v%~qbnv;4WRKeOz+|p~K^uQa1`fYi^@Q}Kgwl!O3fZ=0X z5Fc$&?9j^DqhsIYq?B&O&`%q~oKl(}XOvG~AOat?f4kS~gz88kQ=(foyS3sj&& zkG^baB^N>lr%w-orEmU$Q~K3qZO0z%)O*cTEm5TcfPfl205XRInvA@}{-~p*;KEj2 zkT1Y!q!b%MbEh$UiyTYWkDUE?3yV4cpRkSiNL2&WxJcOX>K0OqK`W^BGIJ~zLI{Xn zbr zl7@A?$&ZQaPaqe+aeQG10^D9Oq$|yc&TeuABy-DW4GD`?4z8#cOh3i~@nVfymQSP! zy^-6dQPh#uqo^P1dn}i@0X?u?l5UURMD#MKV-^{vX@bx-4i(YD6O^6855tVh38F;r zj=N>I@dj2Q8q~}8&6VA%S2l4G#7+!xN}*I-{PK=^wT`8n%0{?RtEB~^a=tLP4PVSm zQp_#nc*^y(#a{Ks?S$k2vtq2|Dk1=%DMlE6kL|0Wgy=^|EL5PoySh6)|0sW$h;O6t z_&br4&fIK-i4Dv$d+kz= zPsW}VUIBD2%1NDp@XSb2D%RjvM5K*4VuP#0Cer<1yT;0L3BBGxO@kuUi9q0F0M9ts z(Vp-n+bh71<$&=vJhz6>6#s$g#Se}5>IN>Z0H!Gt;YA-J%B%i;H_SuPOW@`1?tUZ* zAslWetpF&X+qk1&h9JOBUv2Hro??6`r#u~hME_$i^z9*|&627fE;Jv4m40T9?#bv` za&lc(>wFTTr~jctglO>)?9_91h`Mc4>tJotMNNl7uGS9 z1&6VSnoSgMG57#rrKGUO%ph)%ufAgU?#g^$=H4;>%~DMhp@c1x0e8eRvTwW-)VM|e zdqi01zTr4baEPlKC&FNN-SjpWSkwK=Yc`&8*FT8|^2`SP5l#gY9r{lXuw2}r>y|6y z2mIr@tVtIWv6g5tELPTPnhkJhSpQ&=gh2epW)zQDo?tylCh_;`f$MgyJyoicDJb8B zJCoiii$klrV87b~1a{fn)tGhOl zat}s#nQGjcL)&egt-%2`fdZUv2RAqO07V&wXfXTAC4Nsi#oeUXLs{E1iKZJ{2;pjKCTd>aU0D=j@LG4iXg|0~@g~aoZ z>}f9{4pk1}Fn6OmEU43$HSw+sJ^ITM>&Vd*v@-CxhtXt`)T_+GpN3`k|*M`LJw}z*cry~mokm$gN9l&E!F(Vm!=xfQt?n! zYZf`=Erpu?TN=8cE|z-b(;8MW#QuR8$mq=-npn>Q=E;HM_@rY)om zjXJZ_l5lG}5LSap6Q~b5gC&&XTN836a~yl^2C{H8PSl-7)Jd9kL|nJh(bq<`q#zl( zX(uZIr9_qOWp60UqsT5u0f%}%L};gO@|R*=t{k^of) z7~@q_WVA5Ai8LOhU#>ZfiNS`4(EMflT36jWq2`+8==bF?l#{ij<0z6mDlAl)X(X{k ze{h+bH$f3ogHVod^XQ^ z_XiEQ^$&*x9bQ?gnr#(jhZAiGLh;F<>yezzbKx2rIeZ%{APfxsDKdo1)QOVHzlUHv z7Pc`eLz>fF&n9BUgjyZ^4gOOF_8}VEXM@}=c)|VFwzo2^JN*m<`A7+sJqbXa(us`E z{AV&tr3fukXrBiGhJQwVseb^*)(S$QLSt2S~bA%wG<7uG_P4))#7XS9H&bEE89U46TRXc`-`5hQf?E6~}dI z0Qf zAF7kT%NcuHKamD|Xfz~M637E^tg5=>)|-jymsNHDK9nIEP;j3fIBfdvSV;<*!PC0X zIuottFC;TY$%~QQ<+HldNaD?P> zsG~p|vPgtpQ#pmMswfa~q+bBVa1ZVdo%yA2SM^6zQkqBNafp-I@QW-{&~z5Ly|d>@k*utEMaz;jKhpDqca8rb2>4K%VxB~s<=V->%Xf3qi( zq1~k$YZFW!syO`U#`}Dz4AHGOP{Q;a=aHjw(LRAwRWPwaKy1JHh;YEfuc4VE2Z5>_ z0JD+4%Bu&B<)Odzy91oqJ8vzY?Cc$vAyu+~tw6*ZQQgLGJ|V}gacnXS^zbzYYg}P@ z#R$|ENpeDe=8QeS@7(b?*+%$>=Xe(<4U->x8(M zrI25CO69q%5=;mftLt6aqD_E4+8-=XH~Eu_RZn~9?B6PjudP(Yq~QI5sDt)a5UTQM z^({l)skRKY5Ho()Ic$ei?Eh^aHTB}hMNgk5Q!CR<0?QM&y;RqxFK#4$7Y5VzAr%P4 z*64}=IC7md9)O~)6<`4{ELeIokBz7o(Tp~ zjY~fh3x*Ujj*2{)k3SR7Q9-2pM2^z*%{kuK$#}t~arwtI2^kaL?}v%@{%wx!7K?!Z zTAIwc9D2YhBft}Q<|EYWr6Nl5DI^sbm)5ZXQfJF4Q)@pa{rplcZidgQ#+tSkr$a7r zg@JeJjj;`^m))~(x@v(3tk%=uYZcMCWP6wZ=V)|b)Z0<>i)PbL&-232R$sUkB6gT# zQY-Y6#CtmX_{5m+O~MjZ3vEzPRSOw;O<7NARk?HTTyxgGww)wp-g>Mi9uSkymngw} zq#GW-h`z@_vpu9p;141Qn)K^6lbLueCm!FhvXR|3rTPl?)hLRjmDGoX&`z%V-n`*! zF0QU8q(juYheY14K=V%}tCm18uEOQMI_e&RNXV-yO#R#gtyNS7_aN(_FC@)=iCM%^ z;F3GxdROqMM@!6wI78NbiX(*cAgE3^hA8x4{AQ$u)eBRMo5$Oz?FsA({y3N9@Plno zfYYzh$jM*0D!?K|dVIbhk3D(EL9DS^AqX!R{`btBCa1Aj+Z99?C#}=9hfRo<(J>mS zwJx`|!r;DAA^uYFCb6@Y#En5_6S>_{WPC4_OV!tUm{`IWu#M-pQ? z=o`0%*JoYJ*E;AsraMyop`tkR@*il14tK!Izn&p9r6HSqIj}ATze~N*4Fd(;5dNWV z{8GTcKcIstvj$j|UcaV!d>6f8F&G|Ayj#_x7gNQQJ?3`?2Fh>ke-qzpzG%aHZO|(W z^i&C`93rp#s~2$E{3#>+-ZB)frUI5gDIE3;CkLL~PxIvExd_DEmq3F{|M&dYr{h~B zI7kR_B7eV?g%*^G#I8EK8_chF+Kfo<$*6j@Fi;uT%T09pmMw*)C_wj@M`(X%P0(CL zu};R33G%NalZELAxr7ksVI2&M$1p7m0MnLw^mtF)%!dQT`ODf0@C(y!C!32HwV>os zcM+%Wpa%cM*y~`-^qQ)7DUUh=G5!vGZprxeu_4&T%3EdA)-~vnEzWdb*3Hc*2=}~2 zVFz;;s7N1odysdM?#OGA4fYt|F`kGz?mi65rX&g>KAd1#GfK9K?psg-T^)1wS_q=k z-l?dFOM`__bp1O^W?}HK*EBL6qQ%SOo_dX67Rjc9kgTPBUmEqq8W{yZt+efo#U+ZD z%^!s6_e!A)dSsgE1fe+vR(IPG4#>cC-gW<^fIl#ceA4!?7(`egQ;OZ`JxR6^;nZ_Y z(4EDvM)gkZWFq{|NLyCncp!7|wScQWcWc(FHQ@E<4-fEl&kg)JPExo8AcPcF+g~_E zm4#+N4S3Y&yQ3MR5MoXH_E#eIn~;G^;_Yy4XivU^#`m16bzvhY%~To1_1}C7prx2c z{4ceO^S{+DHWt?ZQ!Ag0^Oks0-+kSt4XC|N?_iiBgQ^e*yp*S9bdq?<-}QzQfSZU- zOc#zVKrZu3s@dN<{?fnkwE}nnm2$og4_?MnwUC-7-@rW0k6RwaGv~m6|7P@rlCO{F z?Obwk^YGuAvGd1icJb*=3;$d)ugd!E>^RAnF#bEQQm67b{WOa0)$Qzwq;6j~Z$EGD z_6G>R(~ssQ+cS^rtoXaQGxHS($=nsi>0`Q8o+VwUXaCaY?%cO&*}84?yUdG{YgpNtfPsTcEyVkNUn zS#)-BKmMTf*9(iocQy&9?GS7$`{do>Y@(cPp2v2-7J;4s6$unFkTEk06-h2R|Gbj8 zKomi0%hOGBXQ-IU`N(GGI7GE5b2d(}Dz>8xx;xf{PA6qmwi?v0NmNba z(fsVD7uMnAvd)TUyqFTM+vlpk>TFIzw2oaNSeY-H97f{w^c!N0)ot#q)cov^Vd-+I z9%2Ls7}~T^u3PfDYMa(7qS)iGidd0+i;xh`4jWrKj-AIMC;TWkZy++aO>e9!4FTzd zbL=ydxs+`(8-{hI@-A+?heWiBddk8uBxe3?B=rC(<*j}ks}hB9Q_sIYKH znF&ogp+pHYmqj4HV)Az$^m7+i<4uKTd}pGadjEEyCC9bWSo?PFSS+J|`u(;jIEv1z z1n8*s;nvazj_OF62Cc^;TwNt#s5O5{YwiFnnhobDEL)u6xHC%=x93>w-A3`I#Bp>r zp_2JQjQ_P0YvtzU>4zy1KG~|YUENntNCQ z4)i*=7$z7~NJb4OsmEn4s|^n7bI<_!O2QCJZI5nR_EmJVk z#t)vN6u!(Q$_N7;&4Qiga4TKob$nfKnP5$$d4}b9`ZAZk?XA;EcE!O3>4k1bZ-QA1C+GrCS@6}Cq0AW zBE5PgNWg|{(h?mF(#4w3#F~nbB%Yqk*VM8v(v2U0sxCN=+eTi%Ae5nThhTf_SkQ+q zb5ucnW`H>?esSvUuE?#_vfGkQB)ktp9aI5WAaE%93^|pV1zf_)<<^4Ok1H+;d^wtN z_3~Nv#vGUsu%9v&P8SB!I$-31rs}5T*3|^1)5^~Njg`noXmIEL*^24m@o?rI(rUb( z0?TrYD$*)Mm|73JyfWTpJx&8B<5Yr`ht3R+il|A{$7v)96H*Ot=(~w1=0!;N-&}T= zYa~K1o?&q;+hvv^mJ7$8fdsk2!nN9V`Tb}*R9)jPdJWU1P?RiUJLA8!{ZhX2JDL0= zj2qZ*nuNn!3vgB_g3(JW*rk8h34@Js6$LM*qQX!`S4knBP$x8w?H2$-A}-Gbm*8>o zDVe5UesoN!Y!O%Djz6CSTz_G%V{A$G*>pe+bw40F7r5BX4gV9M*_{F&N%>0j(I+S3 zENB&^BvcA)2ek~3Hrcb2A3_B{8AKnz6a5_+TRXOJ(Zu3foHm;j+CD9^9Dc`LN5fwU zgi^Thfv#Bl+0n?MMq}98IqR2?aGgd}5P?IvlgDcA2V!`2M;&p`Cw_^fK61iQQh5-@v!TVO zP#1H;VAQaIw4UStA+EO5Ej%bHOqqfQR!28 z$cRZ+Mn>sa=5cy)oRfAQl00{*8KzQxGYp9USpUNjy__oO`J)wecU3aqdC!z_wj${IuE}9l2$&u zs@*P~0Cl#nOZbK!Snfbe0K>lR6j7lnL``r%>VWrzJc2jrJ9SeV?JafJ*3ejvy?XRy zrVTE|G+KgJ*IcCrvGzJ%2_)p;HB&8tg=Qi`0f1uwFI^|W5nmXl@WS5t{6RV4>}4J8 z+ObpuT1bSkpr?sOLrlJh^yh&cgEzGk06*a&To5+LKnWM{`vLV&Jvb29Ys6LXx@uI2 z6)Ez+t}PB&6EbMxqYP;Z=}r;d$F9HpYqRA$S)3Gb0_&bNT4@RTe93#Z5UIU8M&OF_@@JET5We~D6v>_(q5w00gB?1&#*jNZ z!c=Fw$(=*9$Hv2s_=xr`W$-wH0&E1Vjw4{znSn&lOeEhS+_dBzADQ`=7nxaTc>$Rm z=5XkTR2v^V+7({m#NU#FTUSIq;TEaUD#1zE;a11!L~>y zXX~Ivzbiq^WvRwjv}VK@BgUwMTuq4+uLlw(pDSnsVR-x%dl*j|JOaD+jw^T(q2~E? z9FCT&i?hE0`nR~_ht4t{!%_Cnmk?fZ@|tGOxpCrTp)~bGtVhEZhHDz5O~%im9;kXN z2x{GDg@g`O7Z8JBef@bc$;@D&PW={ z32JBKyuN@1;x+b81YaY7rRaqh^OV~|bI+6M%;rW#g|$>`q)s!3h=PD|X$9~L0-_TZ zU7FMEQ`AT@NJz(iyl_m#oYM3S=sUmHA)$aXbt!>79XhgJAxxM{QKpYuuBLH(c4-#c ziUNAywDSa$PsY3wo<7a>5qJctN&yjyl0WtcBvbSWcV>A`DLye+p7VSQV*fK-3x2Bz zsVbL>sp{OnHkDbcY&?q4n2a``o0xuzLtrc;O*)4!-O>oKfKc&K=L2wIBmpLB)f7&e zLsoQpphF+Axe&6f^+a5zO z*zkm7UNxMCt_*h(yAs+_1SN_ymCmjN{JYql9w^9WG^gXmBzzy$& zr=nx8Ao+mDli8py1>5`3vatTo2uSWF3Dj3;u#TsN&nvUxJFra(Et z)8_plppbnL?J}K5>Oa{b^*0lmh>wbIxV9mVErk>gvsJa`(5zI8xfVzwAkV1%ud|#~ z!d3{)A#l0}!UpPU0Re4;1V^~4UE51a`m}vOx35OtjC2I1N=|pLx+F+A)XJnM^z(S>2 zCOung{WetklOs~k5b(h-j6Of@95h`0k*V~1@lIgvLLyTPwXVl>TEEq*a)I7#m0g!5F+KhAD%1Y2pG_;vt%bBdxw%dNOQI`EpIp#vl2H z8^ao_M%WejmV>yfkj!jE(phuk99!)Nl9lOj35P(^Dl13PEy^2dEiq|F8HBv*-{knp zX_g)Z!Y2k|?_kmHU}c%v&DIQoBs`@Sfxb7E#m*i)o2=205uhhu{!VyrJ3ExO0yI>k z%Z8q4#m*Zvd8o-vn#*iV5)q6eJY&e+!NO2;SumJB8p=HKrGx?>FjJz1BtIVC4ogptM@At&B2$c=|m`R{A zXdL)LgmT=(a5WM07g9}|Mn>?UYR5TJ*x3S>`Q{ZYwmH>bqs_LlLf+Vn6EjLZCCoWA zE^qX;in9IK?M*sS7;o7($m90jPEP~x-Ku=}7*sF>D4M7sMANIcKi43<$PGuoJf`^Y zzWvb_XBiK%Q;MLqrt;BsYG2RW~AF{+NfWXd~UUn2g_>Q?_)S^p{Vxi`+>xq0&ksA@@ZVYOr|Hr;vIV?Td{_uqCw*x4r&y+ zd}N^$Mc01#t2#%jX6w__(L9|~-5)l)OEn?;;?5nM>U^?7*tNv2SD3DwQeKSLUy=Sp zsh%XhG=K3uhN*sg!<%rTx)}=}iXb~$^U97did{8_Jp+D??-f6DF zY)=}15pm-7u@yNUs!F*W<^bH$vvzRMJijZ!gB!HJE+b3;SXlb^9aoKF9qvp})XZ=% zmKX4uxH*w;g#Vy3Tdp%*P8xWwOzvUcTju;KVGm_Y*EC1vSr%0k!9$8&<^gvMBgGxQ zCoj^g{)?(ILajaZtIMx3!Mfz-O5Ps(K?Vma=rUJA3QggBqXvv@P*d_4=!Ccy#4V7T zWwXOY{XegmJ46JUADYL4srKTg+Vx&XQYx`42POZ;;$^SG`2|sb_}suq%G|G)a<$Q; z&eZt+!%?NBOXm^pZqq>kbHRJN=D^mdRMcYxCq(sV^(b#htgE2AdNPZ=s^e9l0+S#E zw-$6XJlQ?jaNQKDfg@j;hGWaiu=TD$%{oMlD6C2VzFIpUK*JLMlG<7o}|1kMtvNjHw`*W)$LHu zoWM?dD%-LMhE~-x^m|KPTp;=fw`o>ANnSp}xLEnzyw1kmKq0woj9Kkd zP>V15mL>neRWnK|Os)>{MoD~i$RDd3lN`+D(-;*&Y`f<$(A z?qi2zvh6JF<-4cJo{WtgPM*z8Kqr9GaLGW4-I5Udx$NywFxei_Y zUs@^Ge`}>b&Opxp1HnXF&UsTZsrT}lZ&s*$`ydReun%b0bp1MQ)7qO!lat8T$igf>R zLIKQTGj-HClda>&Ko5LwOP;fR=3l;bBJufREK2+|(aMq4le_c=H*%*e47E|Yqh$ZV zm#7gTlygrzYiZkgX8w2saeKbCKIToUTG3A*KTW4#(2B0IT)VyY{(~>+Hz>yk!dEiVG=g=h!w_9822B`^>{V=v;l--QrAp;i~dX5}{F* zqT?Z`Jy2F1yIvSXOf#zM)_k{q@zytcv?(JzU;C8F)kCGbv(*NQBq-<5UstbZ&B1=k zP;B75!HrUm-kNXI54{$&bLQ=~Rh#>^N94kg$!{Fz+i_G|#}-Eo`%wMMW@k$6nt(n> zC`ZmWnr?W>!g%I9UW|fK4sH6^Nj+MV#WB6POrE#qz9M=U9K970FW0R#YStU?O-knc zV)xf!V@Hbzxwo)%?@85%3A&bfs zsZf%qR56@-Bt%TJAIHxDHWjWs@19-%c6uwiOGo1t)r!Mn3nKZ*ECu8t57l27MTGGF zb?+)jwQwR;ne&Ux9&hT76_Wdk9&!W2C|O9(vQ;-~ttG2Cb5tRD$WYt1!B6Jt>OYD5H zPZ*Fvj`l|(Wl&SmbBM5`Wc~I%ea-Kch!4Nb*aVCY%ew`m9=Rs zb`fQ2QM}`nr{3ZOd->fj&djW>h8$CVw~r08sf@1|t6!z$4b+oi0WSBmz}B@eqwJ)P zTV`HcOo!P_j$YAAG{g^mueV3U0k3tyS+dcN*Guq( zr5x%)8aJG zwo;j7k2HUYO%1O0AB0OV>09Mvox&6Sqb(=5;jmH7B2La@g)CM@a+m(cKoI^tBznex z7pT`TO8oa@^o06VhH9fMyRXvRw!Kcpy9ba!Mo;NSghQX-)tD(Ok? z4&T9a*vwd0PW(feULveaSewS|U6w@|Mn+6wVA_d=M3l2 zLrKNCVtYo!+Pj%RF)k28&@?6};=hbQCoAZ;4R+{kfE2anbmC2_$5prG0XKalC#qQp z*S5;jdw_0@{UnW9u{>2c*hJn@PcRLk|30MuI+7-h`Jl?6{2!H-Pm)UPz>3**j%M`i zpx- z-$1{hA&e9da*`C04`y;A^t>U5J_;s7GS~{WK$%_|xI6#h20@X$M=PdaM-a123k;Et z_B(vwTFZ3h+xC1-WD)^RAqJHzn*&?6esK>-0NCwQ?J<9n&-npO>3xrrc|(yQ!ES}d z5hl!tFgdqYMSPdSmoUzGex==9nDyG3J1PzAq-DehqCE_B8AGZYf=Bo|wO5mfn%ZjW z`$!ZOC&2Dp^gu)G!`wn?D$I?cPa;d46Xoz4s*-v3W-;21t6rIV>XFT`56;l$_wM?o zvKb+aqljH26Xf$V#29n+kJm);<)Pd6^N43~-kl@Y)iWLo=AUOXax0Vj1r=GI=UH}q zyns{tuGZOA9|!23dMnK!telt+it?UyBj2!V+?RD%M>5-^bxHFfw&f0%wgitQAMOl+ zrbv-zggm6siKlG-rE0&$ZTU!ZVtEeA(FBW4m#s7~tpr*o1P2<(==M~p=a=0JVgv>n zI`Df_i7+_p09LszdC}{|Ek_RU-djWuEu?66MUf>hKMR{rAZ@dRS_?KZSaasr6G0N$ z1Nzt$7m!EbOs`hcGgzxc(?jBi6`@cr7pE66FM{dh89fg?d9VpW6pvxP5oSF+WTCMg9Bx1els>`j&nVAAZ1gn(lMi{3tYn^mC z+*=YTM}Lo!#=J~)%}CriPHI0(#-&Dr4sn8N1^!7emL>HP>9NiGLhhgVM8b0Rwxgo` zP%(3ryP#0Bm}XP$+B$mj_XaKcAHEt_T@z;qtn(y8`mA|H&iP;IF^W4ft0QLHTRV@D zK{V0=iT*))HP;(2pxIIaGI!fMr#Ww0gdOk-N2fR*hl6LXt)()EobQ^5e>kxj*4&|D zD4Tu)4KdQWWyuM;y;Z9?+B1}+?8B>~2#lL{c(!r6GqgU@8V7Y3m|n3-J^h{+p|2U& z&LE1#)zvGse-nCEInSi$^z2y3(mHWo)0tMqtxvrSkhDaUn%;7Ics*mvJ#=8@qOKYs z-26gL2s3V)$&rm5>tJ1iDm`AeMt}E@S7`pk=Q(pNQ12WJY>c&+(jq1 z2+#?o3$KKcR|FLXOLj0I-7~%>27QY!r7sB*kgJ4IgMNlH!uk+b2Ta8T??F9c=vcHJ zibL~P+CLH(X$6kQd8jx5lGAw_5d3YVxyGzN1j7}YQYM@Z;J(XZg(In8W8^$dnYX}Y zF1ptddF~uQUo}=foWXpp@TDTtb+IJBe6;HAew590SQq;S&;=vFQV7j*mde8@kVv4R z`qoIhgWz~fc^~Q}dxgNf*~p~NQY}^eR)6K9?n~O~JvE_qVaLJb}6nam>v7GQ`7QtW0zT>JeL-!{(*Y|fvzLWd0? zbfO8EA_4C!(W4V~{XfReu{)G#QMR#d+qP|E$F}Vq+qP}nwzXs1wv(IlzTPv&egC1m z*XUZaYA!UI>bf1A!&6f_&w^3D7Er;RY`oVAAI3wIhB(1I$ysWZh~S9eK{`?={E zGXrhPy;VQ*NfVxhuAqr+3viVGU`6MFwc`Id&LW`kqtqk(cDwyHY=p7nRBOS7?KD?? z;8Nfe(b*h-XzGvV7%wtV519OCH7r)F6PDO34x&LzM%lK#$(JDO20@X41`|;t=0O&$ z!&_?xSo?;`hRV?a^OjwygWJ0h>} zT;Tm^D$--uvyJX==HkZAmLL@O-9{Q8g&G0S@l8+)C-Z|6=Tl7kJx9*XYd< zxHE!|Pr5hQ*ARC3a#-3-am66UElDJet3`=^n_w}j9rqaM(haAyKuV#a;v$288ZXox zdZ%(r4MpNy20jh;*+Klf5vJ;LGU1uyaAT@6`!J>v^af#m(UZ{sjfp;2;^2c!vO?)| zZx8a5>NNVDcTY0OD>OTqY$+%Uv0mu#brC@?xxN~twQdI^*<8Ul8e&z7_Yp)nR_$6E zPZ_4N+>RqOPra&k!A_(lbLmT`Z6OGGefie}1k#sS#2GEimnOA+$G*1;eflljk<$OA zQ}l;`aMW9zdImg@E@Rkxg$|!K@0qM0g{1!wtncPN>mr~2yr#W4N`=ndlh9}w9Y7%@L+)EXh$Ca)|JIfDaq~o2CXLQAX;dmJne@qoii0tsIZ6 z8fDaL4dobXZazj5) ztnN)kVpg(N$wLl2Ggc+0aWz}+p-VGhw`tR|MUI|D`mW{Xd{-Cz(+-o_AN~t8*K+%* z@zj~!k|kQ~<-SG5T%1b3*$iTR!L1V4%JJVXDwg)73EX{OT* zH#0UrV3#KM?IE$$U%+X&Qd4zB<;!FEYxkHXw^*Oj4>)GVe|dTWl^S?ds4CaYCwcxpnAsYF9kdc!dL z*2%7=S=XXvTNHA9npEkbOED4f5&7FNDxC09#z6Rrjc_W;AjX5s4+;;g%;-h;g>_U! z@PX6za*LYP+-k@cz7AWWySEy+A&@Vhd$$xkLS5y+Yhesa`lXcH@L-z})YGqU0faO$rJU+p z@CtudUW-F-sbe;bIJldi6ngq6erjoY-`A%-AsbHqxW?J)xy;7I*zS$7tF2Y(;0PY? zO`tj$mB8%EJyaH|z$TY*-aK`NHQ=uSAGPWCc6$B|^^@-#E}gvAl;Cc=_&y2r1x_R3 zo%%1K7bC-e3B5Qt*#Ga)Yg$XoX+tcQiwf0WKMaVP)j z*26Ovdky#PLvVJb_-WR|Xo3cDQnpZY+rwc4jY3vcLEIj=3yj%EB@};&kXe@>7KJWb0VS!&xI9dHl zAZKg4e!lRZRxI2f)PmoxChg0wxL3YuYyeELS!b(?Plim%!Z?3uTv&^KU*R8gI=&sg zat+MY+2#8#Jvc)G|2xb-Ul)MRY4u-muPBEm-s)z^)WzQPzW#I3|7*o6r78ZlVjl?e zVzwh+q&yy4#}^0k+MW0J50BoCXd)d`=rBjer$DbftF>E@D+OZe22BkW3&bMchNMtJ zQBZ$fOZx{l-5nQXN`<^KMxqb#zJ-U&FDhQQWHbEUZ^xheT&}al9D3E2t|~QxHAg^t z;92TfY?ZO7w8Vd}M4AK6lFND&-D}=)ixL#*z!4v@0a!7B-$A}~m$$mKQ;~xY1|)LH zClY;HYVpXo?^P3S=qhGeW|gI3*)C#Bm}UcDdHSvDf+Bh!7|<%VC>m;4GF67<+Y^9N zGoPY^IrzD`2g#GNP#P<~lxL0Xyc{&E?Oh1fsRgL2Dk!_^BhUb3Cq_~8Tx|-rTJ;le zg##?@g*4NV8~0jK(2sQ5E$z$u!&4U(`iO=zd*CT9E6zGHI#6zDCYEU3pm+1p2i3*8 zj%SBA8gx=^n$;Hrb*2Xf0mNL5eI39__7 z=VdN7KwTDC@M&Yf%hr!X3UvVx*sPcouF(a&&P2An9~ev%m=4pQs0!3Z^xFjvFo{qDD^F+PH#QK}H@M}!2w5Z$uOoAy@fLoMskhY=3>Jdf2i!pNDK+>H zp2QydBJtLD)=H8C2zDS8rjS6VVp4-EZf*N7XAL=IbDVWhQ9)mRn%-e#c5&GB=e2gc zvp83*pX@Ra&_DVipXX+fdlAix0quhBwGYbIWtMjLRH&k=Q$@-V^-*HWBC^Oy^Jr|c z+I08m75Fw1^c@nj^c4j~U%sIkY6&Pa70|keFBe7X)RFEo_|-K!D!NQME*(JmX>V-O z2!I&%QAEm|05>J>NG*a9(1Q%`j@{HUf5J*DXm-8cya7b_=G7m@5VUs;kGb=p@y~hQ zZ~vQ>*Aum@>qM{@!8bO{W#LpG5O!B{c~#Gwh?<6nUG=g{**!*>cisH7S=hdN=3MmA zv*n|k43DaT&Bro)q5`U3*>`kmEbH{_yF8#>otelNN1#WXP^`CU&k!CmHxzfyT}TAn z4hA{-DXqJF#t{S8PQt_i%SE)t&?%p1{^=k8b$=h-(iatqNB^iGI zU4Qc0dV3?{Mfi|pH0ldaH7yil@0Y#)ptoi$;slMR)Wx7!7Q!oMK0$$ED zCl>PQYv8MhY6vq06f?cETaEzKkSUR8xJW zZbE7SpLXh%9YH{Yy=8Eu(kvK6wq33&%CW=0$!G!G35gP4W%OwXs{$nc%M6o$13*R) z0R|XI7VMn!Ms0OyLsj}RTaTb}=QdEJ@xjIVbNIIBN=Vi{r4_cLF{4`eq=_q@LcD?y zBLO2iF2`yVzUu&5PYI1DKQJ!oH!3c`{1hRm7@?FZ;*=Qri(pec7x}VlB)JoyC~j5k z&xoOa^}dL0Hs>dGe3TEjFl-%JW=sI;14Vcqo|!UV$%pnVcXMed6UJ=Eocc->&g?)j zKJYj;F>ELZILNJnX+vnn`&I2Y)r!b6>sfcs@MMe$s2^SXkZCY+H)()m1`^`+CyoLg zrn^3rfUj*DkKU2~z*C`&OzXH@BwBA80V^z`Zp}&G%zEpk!A>HH*^7T8yriEh1o$nus z7_lLZiWFLlG@n5nIf<-krt3f2LLDuKU|6 z9LhROD>_C?%49?~(oG(?bt$Qxxvzl!{`X=+ApHsdxC-eITJk5pSsi`8^SYwTQ}t(a zig~>WbRp%BK|B%tH)`xKiy-P*QsD9c;{BPnq3^A>^pJx%-i}&qZvPmKiLvUY>ulj= zvP=)Cq6T-^w}&g|oxdy^s0Hl&cq8&3gX=zfU#V@i+&O>XFj1CB*kBN<6$NI5gzfnb zhw92H^Hfb@lL*yl0maooB4S(vnc0OXV{vkTeKGQ$X+vO!ERDv~#z*gu0iK8WA?o})QRw0XMS`C) z#y#d7N39w4$Fd8mmGrMLAC9_S8#EWFBos?+XGZTS_FD8#j%5!^7{1E86mKcevv{fO zA}3vdn>uxhs3P+AzSaQxixXFf!;#YshS(%b5U!kx>Z(Av*1XC8045+?>b7J;E(7mD z<_@@slH^mQ6{=^?mtbsFnx;Zd$e=%^D_=HVZegw0a2QFe3wdffP6$6L| zAHFI%Xkpn7(O2=OVJvlhaK-ihN)_6&O!@CX4-UDYgF|D zoT=N4YNtH-S8c9Tve;dYSoIkIY=wQP*V5b@t{79ULFX)r1j|$iho$~3)cE*+{zzN9 z2!~vc!`GBs5rB5t>f4ZLEA_2YTrueP=uNNoKdnNca;!z!UWahpw1;A z0OO|fsrCkZB%evoH0MfzA5f?cK^-*dF-6Y4sryi#ys)aYLq25x9h_w8I&IMDH;nxt z0FZn<+hDI}cDLk%(u>;?X8H;;5=1!*A&Qy>d@q*Njx4a$!}A2+rOkLwEn0TEyKZ}t z(G+N@hRwcp2dMgZw9(68&nkFA1W5v}w<8HC8`3b794Dge(@TmJ4KJ04nBf##m`;cQ zGFAo!Mg7gY_G?jG9ATplY*m`I%gG_H7xxnXeTg;;reea=3p7CU1})fFDz%W_g4HbA zw*$C(LdQs=+6s@iqtQ>m;WpFvz{?V3rKZXFsaKax{O-9h8-z8aPhHj2=I3h;%Q-!V zD^U(@hyo#|Z8IwP_q$rb*z0^kMI~b(v}#qEKZVs@gTv|1LaXDG2*=iwLpYUc&+vW_ z0X*rq?n&|-i$4x31q89>$^N40aWeRzGtqToRS*OfEs;G;hZ4O|m|Vss5QO1Vs#roI zKm|hal}H@~z<^CYiY*uP3+O63w0=5EmHnwK2P*x^HRu4Uy)bKZs4-BZ6BnNTW4g>@?f7>HI($st<@qrDN!_l)LChC3*SxZ`YQ0U zxQ}s$J8|j(GNcO4-T{y$18Wxgg5HOAgqGo8Ru}&J;2cx zmjStRu|8b<{Js$qsARm-i*cX6SRk+(kl8exl7v&}a>L|v@T(uIZ>@8+~E(;O#=>yipOvHQ^IIFyNEl$31T78C+!>>ORY=$A&TJeO_k5r9B3ab%L^|J z9cn6(XFX7VQ7l(q!TU(|4Zr)6(A9nXcu+t>ILGBCpxvjL^Q@Js7W4aph8$N@Kx{&LC74*UMI z?Yz>a4cPo%M?DKVF<9P*<42E%J0K?ZtsOt(y>6Oo1JRo5!m^x#_X()oZovc?hf0(J zoP9h%*{XC~`8OGtI5k@hAok&^w3Q6`5f=Txc(qvImolcEWfgwFaM%dZ8VP6ky3I2F z7)U*-odW#RabSlDaaoB4{o66Jlf~KN=k+cBkRhec2}p;luha&g~SK#prR~{{^G-ho7 zft!(pAqNPEqKt9X^Q{<#PJn8DY34dZ9+%=-bugAA&#D)S2(FM@2(IxII*@wD4Q4L+ zI6+DR9qCYMV_uK7*Y8ZN)8yK{FS|v3hv6gEWqvB~6x=5{b4z*h zZsMrqH0=iyq?b3Y>Qji|p{+)xI(e6A?y)<4%(*7C=M7mD=e^xE2?AEaZu_=qFSA>h zKtHpxc&CQNS+A`CO@kGZ&f9>ihY?04gAnVqcUdxu3t#}*Adc=0QXzF)m z5{wT`H0*;IHM0{5sC0@c5L@Us4yk}*5V?nh;qq$p7l_T^zAhT3JJEBNk56R=$)kv_ zTAT4O@ezCfEf%?on3IP=-MZpcw?oG`9N$n=<{$u2jG~Vl1rmj!Ng~<3_T`-Zj1o|C zSeN~YDo2lE%U2D;8&~Ks?*azXHR$hLlwEcqtvzHrUS?^N& z*vX4$`6Mvj1>6kYzt!Xp)^Rm4@u5d|xApb(wPba}vf12FSKf7@Q+yoa>@^AYhG)0{ zj-CW+I_?L^xP#Gr?WpvpmVmT7G5B(!3}?3}|D%rVPyF{6>&hO%ck@xjb=cp3-NuxZ z)dNgaT-O$yn`7@R^=0rh)4moRMf~`1ls&Uf$-_IGIRslNqZ^t37#Cx8= zGGZ@gNjIzsYW)>CL=87U#K_q$!0IYyj}jK{-FL8CBtgAzStMMcsDAIGkYzk@FqU0E zO{J*}^TjY%tG<$nr$Yy1p5|s{%s@b?lN<@9Bn>{ScIOt34^TGUr6x-P3_G`7d$A@Y zR9ZEvMggY*Q2%kGcbr=G%c+ovdDDGXB|39WRTK>!7bG#|%+B$Jp!VKsl;bcm+-|m+s-XhV0cl~X3WaH%hx9b{c|TdT zNdMtZ_Tu!8{-Znda{qPFt@ljQyA|t8ck&TC`-0aq@3CXHP1)-EZ9wlu_#&N4S-a@7 z=T%a@#h-fGB`6jPm6_XxSNUD^4j|^)vX*PI#|;$!`O=65_j8!g72O%@u`zmnWyBy~e+Y+z+}DYK@uC!rf8Tzd4?RB`s@+ruJ6TPYPKA^`ookw7MeYaAUt8 zd1dBtfT_yA2o_V+W!mCJb+X{^YnuzOtDMljyU>$Vn`5sHK-;q{(+z zjoR~Jkpqx!RzscYeE9A@(zTz*6wC2}PIzf78`i_Fciv5Es?7Bt;uMcaM9y>34Fx5M2} zH^Qc^w`4XV#}_qtPJ&;{M=lN7v3TW5MzrwCx^>tej-Ax4HhXDlMN&%`=4Bf7pBgYE z=IG6J8{QkA2eG!WPqh^z3SE67DUA-q-d(vGIZN<+gMcvj9)O;obboP(SmpdDV3k$F zIoFQ#{(s_SO(pHZFrF|H7Y6+exuq5D%QjU5a81(zd zW#i>ds7TMU2fHSw%#n{pIrhLg9H5N;k=hi2b5z-$N`m2%W*(5?Y+$1)MY;F}$46Rf z=xixRDk3L>byi0>NAc5Tz~`@O*d4ro~n?Ok^0uw7)nMQkJbMY>4y&o`on zPOF^vfoBsPO=p9-l3JCexFjj=UoQOa6}o$@LW+hU%ZJUxpELsv@DHd`jH(Z0{K!`= zB$C(d3wIWQPe$8~JT3ueXMIFawoM7rcV=_guuqDL@(rPOAzp#M%bHscVeU;@ZfVP5 zfB^z!lgxyxda2Qpr@&wdsKu_AB-F{W*Wufs1lFcImzC{6!%HW?7a0JqbBmf14;sj! zURK08s@W;$z(0m4ot1DscXJ8Bhmvj@Hf+# z{H3b0>Y@yOz3a$szHmw8^7qemB~IEO2TK!My#b80Y0TiOrPIptJcVW0sP;ho1oEmu zSF88!Wql|-_n4>4P>fz4F`aOUbCs|!l$?Z^(T9pEMjRO&mtoZ}o3q76omtMTls71D z!uR?FWn*FA|1{NDDIbq0t@X{~t5KbjQ3rx+(p6)BG98y{z~c7BK_uaJnKaJ;rCxz- zhyUk9IgCiFiOELwZK1IhtGO+e#@A%UW~CJQZWxLB#Pvz`P;H87(+NeUtM|LZfm>t& zS1q(ov>;Yu(L>fumOL8tIwWEVi~ zB8u$WBL@I{{MMv9BN&&2o=a8?TfziDu^antYpaDyMUCjjVcnw2#Bi9u6C#b>rk9P= zNDp$GkpqMtz{HJ=?^Nb_mrM2ore>p0xt_V#R8YCRZeV!McQ^rfh~&Q(2>_~D zt%`jXi7i+y$S_;FR4oCX%R}XPl>t3N-vB|5fe(CAzDL&1AMA1LanoMzo^8~kB3?2j zJxTP&t(3k(dgLTQ%o6RXe-&caLuv!=&_9i(z5W`}d;K4|XJ$@-#sg&ndHpaO&lf<~ zVZJ}P+rsbnmMsA)Yf|HD5=>AWf;dcOck6;QUwsvJH5%zbQ664vIQ|Fhi9vEd`!A>n zVUTbVMNOrmu~2LvFE_+F@bA&;MNdKzm$JC>dhvZ=6qERzFol=LWHo zdm-<};oU-pGlV*(_KMLm=NmOseN}RI|B4BAh|qwL{eqpzuc}t@55CoyYOfgaaHo}; zwT>E!7ZnjID4+g9go36E+oF_ZBAz~>iMS%^5;nkXawKon2-~K%l)O|tl>9Yrl{1W* zr>O|nSje)~YbA(i4`&bzNFoMaTug?ZS65I`by2N9k>EM6%gotCI05ZzbJ2iB?=NBG zlPnJQ1&>CR)tr0`x>-H;>Z!#B&9X%Z*o=?A0tbh29`qu!j7+>v$X!?)-5+o$guWpl zW!rY!Z|61Tc`YMp=<1eg|6_p|-e*3m8g~HJ8MFS5ta)N*)u$mi>&mDQ-9eqJz`(#yjN$p|u)_IjoLl<*#q>U;ToESg;>l8dg z;XFGC<}eoI&>^?`9FT_Jp^cH2bB}46V;_;^SB`6Rsx~--^$Yy6SC+32kWqlvG&8{? z@B(;TTrvS13!_id=yO?(x+AXUK6lTTmx!2ICSWdObD**v6}FvuxLfN}eY>sYS0>FN zK8g|jOjg{Lm81%(azfYxZnl0m1$lIzX!I0}{qB&O6nPeDseKNMBxgbsr|kDgP%}$p z*W3|db?rjwa65q&s{?lix$;nQ!IX&3JA?NXph@Hser))@)@GmP;Nel$C|UW2RQOD^ zr8v2r$~VD)31nOsOhzrOi1Qb~gJ-b+Tmp&)r(Kqf=m3g7eRwFzZI55-?a~W=$u{!w zO@1E!?+eA+DRmT(?jTAp=@B6-VAqn6La{bghQFCI%9Yz9)4a&6I8Jq}60VoPgD{){ z23oJxYniaqBg_tGq=;^U7C)UCJ=W(g2$XMyg|>6%8cadc#QBOz>o$@HfJHtM*+(## zDp=s7*fi=PMfVOF&G2E2w*3n7sjq1qo_Ft;OND)6eb>bs2E=(*GfCDVk?%28)F*5E zm1m+5>~E;9CVVjLXN?T*$98*-NqRwnpOyiPSdN#wtw>QIrp4@m1ZWfDWDQY;9a)zx+I~riIi$=2xTRK?ZokfD#r)39`x? zFN$8VskF@usP-PrjVQWiWY|Ut4IrSx%EGurb`tMRntypr0Q-*+S?#)PDhZTBiCkXc z92X55!F=NzMtB6R{GpdvPw`fFx?LO|_xoSQ&Q8_hr}-h6s@vrlz+&o9OdE)~PVi9K z5`?Y+&*?}oY{o!xGkCqmNk{lqrULv=HnXT0e>CA$Q91* zAXiyd;cm?t%p+D9suGU8buHZdsY=G)i&W-;QO$-w>1M+$>i5ur$M)^^gC!&k*8t=g zBUy2rL9+hTSe2T5mITq3$a@GL@K;N`MG;k+eTN4gH_x1(M`ZDr4g>U5dZ_A+Jkm+0 z8nTAnzPH(FR9{@^lYC|+DFD96NEVfL8UU&OEpZJB@CuN!{#TtwIe4__>EP^z zA_|RfI+y*yK_MBMeyEjqn2raO^lnEfYLd0bIVgS#u0QHd9@*Ef;U8eDW5FaFla7@2 zt-4PMcX$6t;EGK)i2yXUqVdDUAbN00CIBB8apPx|@M~{HO3a)(BJK;>0=Y2^HKhC- zs^iCl9wy%lxQeo6<<5%3Dmq0s`{=Fh#N8$)DK|rGLMckct?krqR?Q>psB1U=WPVIh z0z|6MVx(x_jGk@YW>Jy466t}H2f&vHCTX|5sPyCgmYS{EklokJ@H_vp*R1mMJb%a z1KoYABgg!~0ZFLV-wdD%mk@NC7|Sn#Er(XZ7^~L0J0r`=o2B#p-mZyP)Y;1LVn9{a zi`#}@4Croi);Z~4Iw~$+OO{1I>TGbhmjFmMsh1a5JaL-gl*W-~uJfLB4(DHw!oct) zAlH!?=gtRl0xvAqGq-;#8d0fI_d^Gmm;daxhyf+HEtf_yFuYFz4>V2A?f&*m@F<-I znO48^6`=`SmE@wOw)k(AURM!X#oW%k6iRJiFn+W5FR%^}&76=1SI6aLpD{Z%M2SWB zHd_XN9P(xP807}z6~#&@%6#1TomRw3m5iyhC}6S4Xf0;XqpS8Hxa68>)_mng+WUfi zb7YcBlHrnVO6rqfZpbE^^OT^MpUsL}KGd++d{lDi?1~4qaQWT`i(c(zNpS@37;O`* zGvhaztW=zq*!KfbQ~1db_KFZIR!Bb{s zQqZ!aptgOR;~fHUgUZNVnh-6U*W+!_t|DzXhIf19>!~Q{tl(pmaWxEAE9RLqXYsnBQ#E@OM2uK$8s=JM0dO$ z4U}X5BCdqR4&dU)1P_3-tAYbA3#35mXv73AHwXCL$sMq7)d`52p~>+FhPDEEAXWT@ zZX|A+t(lM#0+mM9H@-%dBqhlw5sz~yCAnTbz<8kX7SeBz4njJ&Q`}w5u}dBFR>OBq zYH;Asl(MCybmk-bsU*F^vF(m{5UqA#HW3)fPWc799rD$vN8HsQHqJVrMjH^S(G|W zq)um9q8}Kc7vK4rY+CyeWW~mRsSiMqo!r+hamXf)xtdLAnv!k4edSd>%ng~Uq+y4MC1o7ekjJ+HCug=DP;7sk-;#osrco=x(u(W#F6=oUi=5a#)u zBD6gAWn>_;`|W@cbZb)4ctI8KHm42;edV(&Tu10MCW&6X)@5{^Qb?4-=%pZj<)KsW zbQ!~;%wUVUQ_u4JDJm_A9AYr)%ora&@1zo@711C(XLkBiz`V~!QrE|~kvn+j+&9J& z47J+VeMaZ6JGT~`X+#9$eRg6#UO$<%poJ1d*M_?!gLqDzYFPpi2ub9;w~&&gSC&4o zX36FEf$9Mp8b+tR=CH>~uDN9U%)e-w;S0J+;;dl7UA1_SBwnwhF|6V+xGtmdT4hh% zy87q(6ee{y&XI0BtW8Y3~D<0cn$1J@_dQ#(Y)77 zp%4`hFXEJ!K)o`2dX& z!==7jJ)Rf3x*a}PInySGExCxwnu>pk$myv8bsGlroStY%WcMHCkd32(ay+&6IG&H1 zC;PA&M!;Tr!8Ds5EqL;+!Q6%=ov->A2b%y2Dk&yzci?|w;{uPcoWL{-DYm(neOODL z?ffm8Pt8ArB1>~{8#9&N8nt$6v3Pu4AuYCaOl5n6WY%3I>`qI{N)wQYpsD({=PF>7sFJUYlQIJUFz!|^ekmMhQjR8j#`scKfwUvSL8X4caIN|?E2noY zvgZbnK!fS%(b%y|A^sKYy7Q(7{5^GnCBBsM3)&YwGx5>c{eCk1u(fn=*3jwwX?lEt zUBmSp$t;<-+ue6xYiwbFbb`_)Lo3yIBx~DRZs;g;@+8%)F*A3b0>63O_zw!3cxPrO z@?CRnT)HA`CPrdudhMO+B6Bi%jlI<7%(u_OzMz|Xf^?yfiG%P~`jBx&TSle0BeG)*zx}VL_*=`5Js);&9}CxrurT?*?of}P2N+~#7bd?no8^Inds8Bo`rsaT8EL*4Uxi?$ zN{!wPsHvrSClfK6j&yH!q8!VH2OxraBq>D(jcGX1r(4ZZ2@jUfKw>Qip`&l6)pWg$1A3 zm~}`u<)2~{4+-cmTQ({cfB=3s0!=bBRi4otE)|X)@0xF=B~?J#x|^5? z8oRBfr8So-^I=T2@oWxXUS&#R>*sHs477sKbQQ=M517W*{km#)*yY{Mj)MiKjFso_ zIvoayem1|zr1VuKixIP%R)0K(vn?zdWxQvlP^xGd61l@k^Jz!qmy>s;HCrMYCx`Z{ z3J96X;S9gs1|r}m|MjscCfbxG*Zwft25n`fg)tkUqz?qmV147aqi*E;txFBloYes8 zIZSGyROr(`t3sJa$fy5^owFH9IX_zjR!Y;U%mHXCp14qc4|w%{p1Z5EU^|_wVZWbNMUAdodZw^OdO$9^@z z`;b{Q$%hHiXBF1vp4GgxoPCyTv$TR^VUrgZYV`_5w^t7=p_y1b)D@J2Sfl1n>wGn~ zMd`O$_6Lw~kBf#J7SS(RZn9r6A5_I`^30hfz#0DYewyH#OT+QL`4v+0G4J8tf)=tK z4)?v@%B!@!QEdG9>Q%D!;7)anQ8x~^p73;IB;lo~*upu|0!h>sDbSrQ?K*0C zSYs?LkKT>n7$1y=@KK3pXqrePsOjV(AR{0CQ1ulIk3E&>-=tAY^g~#7WoofdFoV_p zsBJrzy|KdP9>FG{3rCF+R($I<-L3krLK&NF6dN_)4+W+jC2D^bH!!Ca4CHnvUAn#I z_8<5vrs-DSP^^qx)n8X1kEy121xi+Daq4$I!PxR)+QG_NkM?p}7-gL(skpd<@mI8k z6K#alLSM?OAZCb8zx--E*gDz9Tx|Y?#4L6_lef+O^_mN!iFs zJK^u;MXPHX04rITk#RNqwsZZd>;^0=gippNmfVg0^k3q1iWg6p)X(MetI*Opz-CNbr#jOOnJyI?OcooE8ZLC$=!w(-e?fH~* z&{)Ukb=T*ddovB{)5Ei<`%SSd07R%zzQ-v>Dkpmu!n@~txU?GOW_SW4P|K#>*r<8WD118;MR%&oXJ3Ii_jhj+|AF8sYHIPjY2 z-Yti$hWiFusAF)2nK<}lFa(I^R0K5<7op9D9aba^H2v$ECv?Xyz{C>Md5~{-Gjnu1 zj`Ndd!4PT65XjfH55Gkz=I&MGjBiCST)+IiY@Xz{0U6~+jNns?>1W&1Jxyw#`&>~X z

    C>5kWI|R7wh$^jbCcNCXEAr{Om+Hb&tFnb>p- z%h=}zxxlSQJDGWlMxxIYy*7nC=Bf+2 zawu?{6c1N;0-rL~FTVls_e;U&fm0?&YED7xWmtkxiR`9%boh56*PAZms^tn)s|4|L z%)G;#Q3MSdUxNVZ1oGO5Zp}ThneTzK$U`jOp)NB`h~y1y_9)P3%8*Nmf|YF!SCmAv zfnIsUip(hYkSQZI(Wo-MJ-fs{PQ^^jG~ z@z(2F=uvOWH!pSM-QCdAqv6Pm=}HPTKXB!6sLL-DkJ+~K6B57=6~#koEI z5{g)zyZ@d4mV1j&fTYF5+nG5rKHyz@jT^tNuE~v$*HP!W;?$@cE+!x#fn9d0%Qr*$ z{uY=rYtGk;`(cz2J!TW7BCU;I(G=k|fqAnb6_0a|U0$q#KUd$I6yLb5afd}e>s)&C_uWS{ICFzAy}_*QJq4tBCM3Zme zy@EQl>#*rhayX`;q+HEwvG-*#E(pX{l&O2aIQ5es0e9C(TmBDm(1yqW=TibUg>CEC z6Txgc2hgY9)nFHAAZCOFKui`0y-tV!&q2Y?BYXN%PFDoQ*A|(a38V3e>K=soMQrSD z_3LC_Z_R{Fdlf>t+aT!i6Or*74dvEV!K{O92L3CcGy5L9D=`xd_*rw&uqQO?J>(Q+ z3&7ju7^|Zy3m0aGe1%0U(Wd8KYjCW{C6<@TjwBoVl z{<3m3{ZZ|wu7rMONyR{%IGf&2qL@O+5u08kmF|wpS_shv@+uEIDfwo~{e^e`Uuvcn zcD%JxouAh46XH(=5(qxlL;%j zomk`*Gv?Grl+HWWFo~DKybcZDEZDm{=YQaII;i53_!tHG(zV1T0Lj~d#xzxta{3z9+X z*z1Z|K8*B>sk&Hh>zmUKwFfu;zc_^OWp>=lyy^S}HM3Kjp=hFAJ$W13HeTgI6#c?; zs6v?d0fFOQ=B`=~*@4HmyET^+w}&ziW=zxnc55nJee9rD$*j*`IhG+RTULdFia5nX z1hdZN7&; zOf};0eeJL0_#mA=r9Z()4{D=h&^%+#;+7;-V-9MK4Hbsa#XL|@c$I(uRFP?V({%qF zW4iu?FN?z8N2P6l8L*Q<5kY^rOIgS(ogQeJ`?5Efe>S80TeYgkE9sE&Kyro3{P5R^ zH!Xz++eC@w_PEV7$wHO;u5>rbEslNX0Pj9@rrc-hy9Xl^1v198wqRnClg@IWCPbbr z6>NKYvSxt-5(L5zo`ovpNEMyarc;~#jXItrL>|bn9cdk{yZ^u(XJID*RrphjSxUP-rA6hD|HwiT_H8@cGluiUthcmkTLc}$M90WM#^M<<4Zz5 zm^IcVeN=%z(e5w$Kn3w$L<~(qGpO1dxM?F=(3;vqo=>Voif+r{t#I;IPub8N#bfi< z3r2rtp}_qY`Km?K2YXfl!-I#VE^oR?-D16Y|8%}`F3gKNqWQjA-uiDchM&PcM!Xar zI3x+YY`O&AynvI)(|6i9@P5~dtj`YaV_o^G0E-G{Jm=h1&+QUl_^Hg7Vz?cbuJRgU zR--Ff6)w%v>}!_5SlxR`$2XnTJrFt<(oMzd?cDM+c}C9+h39Y%N_eJw&6V}Rd_+*D z%|zNAsjxP)EopJZ(D~~-b3l84)f!Q$QBU@?D$PD&Rv8~%m=L~eo%s|vhqI^a75+_h zK6KcD-G?LTdjYl}C5FHGke+KQbx>VoP-?@dvoO|?>4*$v>J=C^Le&P*I_|QjKfy`0 z4paQO<@TT#j?ExI(jM2(X0v|{{N0Bk!JFOrh~F}DiSD&kejt|?-eu#DWhCDM8BIg> zUD<>lnGF->x;7ljdpK1aLBq=Z1a6k(P`dn|WwqE8UQxpbwBiHZ3jT7`0_S9OR{dN8P zukZNH9-^=6B@QbC)&>4j+E>->A>}M4KGYk#8_v}HX$oe=7Pq60wa_Xa6Iw%Dxnpze zhrqo;asJ2o>n_0|*xq50alE6fvw8B;pOJl6jnu8$<{vt$BRt?-%bZFxy=kq-hM(+k z{Lomi|6%MNn=}g+Edjq}8(p?-+qTtZ+qP}nc9(6tx@_Cl^qrZQFZV>u{S!NPtX#RC zzx5uuWHEwv@5$pLGlEF$8}`%WdU`#+taoQDWaWsI(JaY%vF|*0NGEFMpcr4(We>R7 znUQ70dOcjdxmemLV}Wp6Akw5R z0z^2`95b3fGj%3QY2SWeD3ws_cFQiEUrlB5X)di?n~ASpS|8jjn&B-yRg;`^6RGfa zT|GYQYBy&nC!%ejKeB}B@>3#%h5cWn%s>Nz@RL;EqoFbqNeD)t_V;%Gj>qHw6v{FG z_fU?NiGk^V(&+oYsK(6VjM#0nVs}BPf?XfGQX=U@p-7aD7tdduIoa?SMYkdd0p-%~ zneRW{m%XdrDnRDYOVMsP@X`WBc$#mYOihn3Nh}HEl;6L;^=xAok~WHZ@g?7%NIw5; zW@qc-^ZL|1N#-8PwI^M~iSr2O8haEvmdJJe7pXD+bS#nG%kSsw%hmafEpuPQsYRxF z%rn*5!-;IhBd{?KnWk}}JWcG2wk-2TGV4f+-qfqt>sy~K|BsP{MPF&lmaSIH;K{fk zku|YBn~ZuDtxB4|>5b%{{V!6Z=f6k|MY8jj^Yi0eK9UMF{Gl_$`DC|$q(&WA+C)U< z=>#@j!bwIOF;UL6=vuRC_ZP#zOzvNwZ&6pdW>ARYccAg2tQ808>}0H*MV@i(_m9-zrTX8bMs`MWf?N{KKT@MN@^^~Tb>06VHHPKK zt!z`t_+Zd?JN6fgDZJdj#d`v*+ zKaWJU{Q}uHq$s@Xkz$+nZB(B<-&itUDOJOSSp(gCq2ka)G384lg}x84CkTu$(W>y; z-ZpT;Q)08D)?3iHTgJ1f$htRtvfqrKhC9XCy%ps8g2jnm1{#w*BHrT7|8P5Prk6~D zw(t!v8e9h2O{ZC+&~ap^y9m>$0Zuv+*N*lR5SJKTz0Yj29xlt6w4eHS&TzEJAPfs% zK$_N4VD*o+HZo6L;qYijfyEG+3Tk?O--V5^~fbunC>Vkl)Mu(e8(lcXy2D81d-p^`95Jv_mI&m7u&e8frl+8J^|*M_Q%FoskJ{Y1sxP9GFs&3L;-V_z8^dtN|&+*)Tskj4u!j3 zIO~S?Jwmajxt6S`EkSnCI@8lGg8C#S5=>F3K{(VTtE~_ zOF}jQmG^JKYz;Mc0(2WPJ!`uv8%-jI`I+vUimTyFm}PoU?o@WP1yueM@S$8eE})d7 zpw)+j)!DH}INTQ-pHBL!(mrz%C#^V;c(T^-M*MTl`&Zc7yT4HBEN`X;fJ_ zHV!R1cSN<4MPIo10pEGlEJJOyjWhk@-60G5d1W5|qVlJ6-+2}GS2+ibY&VLEd$QQE z;Vi5>P~Rneql;Uh%Ztj(aXJMrpd5e9KoM3?$lJe)h3h2m1tPywK1waQ8`UETQny0k zeM_}dUI#8aD158fnk0sNnPG$SVb~G^xl+stK)v*ZejNwaWZT(II0%HS(8f(V;e|~O zSRi@_%sSAanEZ0H-xWO{#u|7?P6NL7%J3lu?i8Woh9hhqA)wfXz3u?{+nSW1M~Y`~ zgoyBm<#4zdI$%zi;PE>bCHcU?1hI3^dJdAF zRou?)U;#eHli!oo@eWUs(DzeDD>BBDQ?#D{$^--qXSV=NsXAh}%TaSn+}M*qEVC~^ z(4`kk36&rmF#E#yv!3koK9}5Y6|d_-0!h2g9zpKWQrYt9G;}`r%NgaUh4ur;f0zCpsUAfGMHI8Xpw_z6 zP_afD%PPXFEJ5ftV>T)-U_3bTot9aOwb}&rvC!~$d|=Zrocv^skbaIk7aUmc_grBX zjvi!SEGH_a)9Lvqzi%l){{<9{_oI$Y+W-hbHyyF}d!ASTP;YhC`_5@rnFg=-!+zp( zVlD$PNy$E`bw>;g7;fl}#k{(|FGY4SI0y6kRiWDkOQt}!`=6alHCd0Hk$}DA$h`%R z&dO^|d;<1=eBvN{`A32I-}1mQARcxmJ!yG?Q({Y_8X5)7~oR_mNXjImwv- zImsXa1iDBtk}1p#6MnTwE38DM#~#3LSBw`R8{5V~v0B|_1-8>2nPWOQ-;EL4xVU^h z%1ez(A;$CD>5uK;LL~5=XnHh>4$aPWr+v$yDs>;oMF9Sztfui5wx_c(YnnBgfI~wp zTbYyKSaOoi-EWr7dR=vFf5j)u4ZqoO$Tu%YBb-PpIu7^Ow^D2LAUDb^?aUY~F)oni zdWn59pmsvyiKk3nm@6_(`cs_x2Nwhd!tAEy@=XoMYGgPx^Bc>|OvP}&wY-hUJdMok zrnHTwjw^^nVj?Yp8hsq(BmK>^N=k$`3&(qm!(II^AbR1J_AS)BLKve~qd&!@P^m{6 z@JRedSD>6^qylX%WI>9C+=j%m#@XY5Xj9W-uQq;9P$}}r&=+ZgEl+J(Vn#}oEHlPb zkdDGt40x};Ci#eHE}~ynfc9WUWFI&qGudzsh+h$>ncBND*zUdOdvnk&nX}W(FW$dxtJQaZY7w3IxJ*i_YD?qSS%eqHJUYY zh;FS&#Sf%-2oi2c$>*f0_8DJTn5b+qrd0hU8~-OVJo95$@`;L~@5_e!b8{8uQeanaCw;~sLjvIcCy z@KGCvP)Nh~6Sy^YlAa_%AeHTFdMNB@epf%+Hu@IF1I)9wvLeAEqDh-Qxl<~dcD}k! z=-|MpkIpK-_5GYa`2$SGK@AH5Ty z39T;gxC){a3!CJm%`JaOt>krOa!ND*uHWU3C>FNIyUzW{oQ$R%{tGr+?G zUhu`$Le1AG^S6v*9d0n7ti_f<(406$5t*!lw}Ne5ZswvgLpi#-ftT~~(dMo-LEM{3 z(9)~D-mqM|7Z0{Ci(6hYVMFVbAYc9RMg*rN3M&5K0FlZKt|SVk$T*pgEmP;iq#q{w^Ov_}bq;I`^}y zEOiIX1So%LP*)q2KU?cI+dP$-T>4^2r+-?G&!}oQP_Yj)@dL{st&WCT0 zEWutwGm-^ggUmkx*3qfGYfaFa!T?4!lT%#X{gX=V5O({36%};g?|?teoeI5yF6X5e zv%8c>KJqp(CTqUzSwWLWmRPVX1Bzabng0*73@R#!7hwN91vc;%fPp`U z>IXOi+p|slL?MLUa^ZzZH?SH$MZU19UDaYAf1$t~{$32YHtgnLiE-D^?}1XZ*htaX zfDLfe-BQ{l0a%U|VLCwNTZ}$&u|wZY)^-#jbNoRkN$ZO#sGM7B)P02=J^H&tEXX;` zc1b~&v>Z)<0my_?66ef99>MfBD2J%E3lu(mH%^7aiDcL}1z@=i6Y4>V&+ln?54qD& zCB(*+AjE<~5`ggM$fG@si)%x3cuDz_x$he97w1fny2kG)CO_NXvwhPYG8U^hEL5}u z_1RAV>vZBGBU&5SQ%e-6ij`bQg%g|-7&-x=d`^BD>i(h5Zy*OK%2}iOXBQ`+d~Zd6 za<}VG8Pzi|TIw}D`4>^m8OI!~Z-B*&z-sAy$_FjieH&fcAE=L~5F51xJAIqP@28!W zvO)S~`}C|{R+jD;`cAl{PV$w}VA~(WztKV{US@XG&Hbp^6lRq~ zkewgs0HaHN1z{P6BYIQs%4#<2db{i|`$z;Z$4E|Y(V*W4Juh-B5Lcf^)mET;-6AeZ z)=4$DLilD(7?lI5ybOiZe&|CXwU~F%KQuK`t>iDErK!L5)g>8BV2WhOH}x!{MGX*i zHvvh)s~-KE(f?ER{W0E$i%Qu<9`7D*WnXy=>G&srBWyxohVi46KX|Ka*$4ZparfFM z7OYE9xeznxxlXzRC*Dx~jbM6DLHc*!X6^ZtuK8nGF&ci@H3Qs%-nzD}SW!#~XS?FQ zD_QITaekxp6xgZk@&4?|$w`twVq?M-&LMdw;hgU5(Q2r(BV6s{6BZya)bHy$^=_zMhKWd8rQli0s0KBHE!z z#Id*mE<6$gVD~x!85dEs&xR8a#0r%Sy8ez@iu1J94>i~L zuSOE^er}meIw4Ssh<6+Ois~TtCB~WZpu)^UtWSa*W3t}Yam*D)ril}GPw12lc^oOO z9buV9>{%kik>!7&|Z77 zXafc!jbz#s1Vo~x@4~%tL`lW?V!6fCMYVY<<8iJw!d*|4#zjgIZtDIV zy&rVh+D6$tksfyG46dSfj&E==F;=8Nw1sX!80kD%2V2p3-WJyTjr%NkW**sas_>tPRTDz@x%1 z(E1wZ;4{VXqOa;k#eQd{jYwrfJ-_VlZ>b!nLFr5L8W6Y%o5c%4=#i<&RS}xFlY9-N z2F?0H?7vc-oDqiJwsX$_Ok-_1W0AlY?^p-Ik1FsicI(4gghYC8xNYS_b)Z_l^iH^WM8!-JXpD* zttt?VG4WWArhVTuD0j*tW}gJjU~VMIF}GqpFm4+m=q4BizDuJVgFT%(NDjx|0t0JN z^i#_EFrKaxxIH}#nHE2uI(PA*yWp-NA-o(Ke!DGyVg}bX67q{`A3yJcBVCbmejA-d zkz-3uhmoXiMVG#R1Vus^%ivGi|J&IbW!qc|W~B%zfF{8dx7T&%of(yBmj0=)ftyv~7z;vt{s35gMXK7%BE_lV!A-e`q^iz4>$sjnSLBuQ?8@=*LgOELMc) z6S3S?PtB^ZQaHelagKvR1HXx2(n{K$Qg`X^eX#&S*!MygV}$*H+SoN&LoK>^ihCxw zH>SAKLG8pnNp2@io84;2{tUH*M*yboo8&+Ti|anY2fN z@k)NH#d@PH0mX_#vy(<>y#Hjs0Uj)$yu#Nn%rJQK9VR}B+h4VE6F1OD@*#Q8%O~C! zr)rZ2^IT)q0|x!^=Pa$l?rRWRglY!MHod09cE$JQ@r+jg;)_W0Z&n$aG>~&sg>9u_ zERCombHi`?TsBg7EPIB4te^UrV{!_{KKAG}h~(WZ%&v2!pJ5mQwJbQVPxnVZ0|YXH z{}fKL{D*LofrEcVK|j~Z*OqB2D4`vH+ur#`>&X^6B-U}6VP$PZC}{}^AwogU zncked`ag9a^(PB({>a5s*2wC^Bp=&fIXB)X&L3%tnPZBPegivAM|xZxTOTI!3w1@e zE7qKzKHYB4B~N4Ie9CL5?~{0~@x&)6M=4lG8S>44iY|_FIl8%goVYN5|A5Ba>tWNP zoHXLB>(0LXyQa;zsS``p$STXU`lP6Am_WJdLS{ek)A4g((#<_ws8rC@F>+&T(BkC$ z2fb%X*;O2pq?)sye|;CdZoDd3(bi6U+!4YZigZ_GdVa+~CoVmzf|4oCad)Np@GZQX}kyE%BT@8iN4I!mW21*Ibui5(x+)M5f&w#&3? zNPTY7u%=Q)wIr=fTdFcE(5vQ+@TA{60hWSosZ@QusaRReajojBTpfk|dY-REw-=r0 z4_@I_rHVC!(L%0o7U&fdMm5~LHq#qYw;lct;~P3t{Obh2(=tyVI7`zZuD4Gj1O4cj zUwC;bVg;<)Y*BUeY@Y>QHIuwzzDO9V(})-0r{cOJ11%hQu&S%guF?dlN1AqK(iDf< zw?rO!rS01Tl|#n>+=Ib^{seUVmrZmQSxxBLg4{`SvSTa- ze>litH&C|ZSm0{;b`B2y@A{l;8IV$o90vz12%ciM_BMLkNjTpS*@GcAO{V*hF5Qkd*4uKeNW&1#t9 zQkWEXTPYpXm5!lmX+AD&wetAY9L6EB(Hy#rULq|bpbz~VxU)4e-|6o^gZ``4!=vx_ zICTA(do#8}moB@Ds8thmoTfKqiTgXsHRx6UE^;>k_Jfkpo?Wl?+WT&Fzt#vx?fri1 zU{z){5PC+$TSv!MYVYv3K@Gn78vK~l5M`;J3sxtgY z_we1ErP6A$*&A54s{iMQogby24_J7D`Cra?S-OZ5TTjr&%$Pq!(rw!NiTO%Wmv<(* zG(ZPfWrk3zKzTe9@M+!^E2sH5cca$2c71(_F!~z%_V6;=HHMtg%C;-Zr4^>HNoDOe z6Nb|dx4S*p;2vPHTU^TrXh)=1$aj&7#_0Ho!kaALmO~%F4(7(vVUMJ;K$1Ly04rp_ zuxCzMajIHt5l zW-4XNM5lLYp_k3IS8ZOQ91CBj)uw6z@AUnx_TY%*O=Owq=;MX>4#4(MuT2|cTOUo+FXA+j? zyh(8>)uOM|Y}Y{D>q{U8aji$I0>zS!Xa%x1Z(lT7-h{&@TV-%xY#SeaLf$NI1G;PI z7{*Ox$+#IkEb#Y>askT^DWVz>q;Kl3SbU z7_b>eV?9*ZAzIiZl^nDQY}=KkKMSs6_H*Az|UQ*;o01 ztM)`(ZQ1O1%Brt=5lt{h&t>Wc4QVJh6yUVV@-_{Di9(VQmMXmje@(32%OU3@=%1d1& zCc4D8P(Et^I*ru@5yG55P|BQIh}A5|xkE?f4IyHFeR<=rbW+pw31HS_^mj}YfMrSb zM3QNHrSY$T8s24}ja+_~X0&R#I57n=&X8merWaVoz(DgSeelU?5k;qDh^Yr&Y;MQ% zPw|im_?wj5u8*HeWSS295S?$otfsVIxjJ>ON;91D*lAwDbQ6I}XDo7OW-9Or>EsKg zU}4zo0i{q~0tbjZv30~PCX0{;#YBZHm8rbQyG1E~NQTyk*Qe!}Jg&WT{@f6KvvL^} zJ{*Y@#e759&E(tfMPOZG!My2vtnC@KgN-L739W<=40$(IqzmgO0D>(oGh!Na#C3_5 zgqucwYL~RRmq53>-W7?Kk{s?vBsS2Rzz&42nP9rEe-?+J`U1>V_T{mn&MrXfrc9`jU=Jfcu7kv1rN=O zN-byyzLn@6n7z!6xEil;vBtmX?t8Y$ncaUJyfenx+B+)%t-|IaV5kCRah=tw+0SPY zi5)3BH0rY&^LhOpYIel)MuKzjXwH<89l9d`9WurU8nXE|(~_N1CAuY97P|@qOE!2z zphva`0}Yf9!HLRnJr{H}4}#=3;XvM=ff}c+v$<{E4q(sN5}U|8k*x+xe_i!evp&Lk zPB6n^aZkQbYWehx>6;&6R8EK^Zwtmpd#D+Q`h-(|MwOMRGO^m-5S6gTj3ptfQq`X$ccY^zyeXf?T%VsT-xWx%}nyl#15DCRRzBxoB9W=M7;gc63AeXfdJp1(hBt8!KQ zTyJk~wU13$=&=o_Z_D+LKu!k#(bscULkG>jJ{$e85$>-o>-mSKSCWQI-nk*ds~ay4 z)#-{=@-tcX>%8EioIy|OMoxvphYtQ z2QCse(%2MXAAq0k-IF-);6PVfWrEog5dB)9HGSbaR5$3sIkXo=5!8Odo1Bx|i>n++ zG&?-JyoMT@gqfjBofXl*OgUR zk~(B0@=)avA*_FLZLi+ipm36xuITo2Ej%^5VCsYvNz!QxBr5F*b;^hW&CNEUHSVW z`X*rG)$V{dl+d{w>rYVC2qcqSWfoijWNIr2L3TJ^ZwZetVD3VBV#q80l2yX8YrFvf ziMPm1OXMI{?v!LB2M!jXMxAVC6WHuPrM3jKzZF+ z83yM?@(Iz0ZN{Ev2l)-O;EMi-c9OalT+CaI%`MsW*Mk8un9;1YJT@1mGnkgfh=;SW zLQn!3#q~@5mAkM3kGLLr1MKI^hvAk25`<-%5&S?hq>I9wE-8(c!wX-^Z%P;(?Cjwl z0;bnT{ujc6tP9`Vm&fD0pXqUkL{nW!hVAa*n2bRS${6(?#W=YMi+e|xnbOa_Y(Q2co6C=FAYCl(5rs`S(uj)(ez#qhR2!9K3!C6*B;(8ODE{!e z?g+e1I30|gcK;ha$)yB%k{7TGua7KAHR$Q5uupkO7*p9&!OvnihFv?s3_d2o;nIdK z5%`ULurLUueWSoHEPk_Wh5j=FZX&gmzKFB!m&@L7M9reh*^eg6oEm(>7$QHvwIZD^ zCmXeuTS9M|ULP(X-p%t4+% z`1S;~CjzD2+-lJp3B3NGv!~2q5x=6&3%ZkNt5~(1Y||s&pGq5e<3B%WxBHYtXAor9 zh3IR~MbX{~&aFRg_uY2xJJxbf~%3`zov1hjaHPdR4)($E}~oYcws z(gWvYzt0CjV%ym&w*(P%nBVsHHK_x~b?Jpmym<>_lkVeA8e&2mAtxtqP;jE!cUuJS zF?;K*6l@619+@SWb-ls^wq50wH{kiOw=6tcz=uJ+;8&4-kXZoiOD}KVRRxW~mZb5$1fgYbHRHd%X%S$p}lr~A4 zKcXIWyc6eQJddn>m-nBY@jyVF=L+qV&I;mJIz@JVeS*H#dtriy?F|E#fzLgDD+OFz z2X1iRwYJcQ&`U5WE)ALrKA6wO8~sc%s6Y_PhY*}0?PBV-N8T5Bbu1PMl~YDSZp&I{ zB*;TS^GQd0k?ff{2aW%7XJ4)acuYfSj&mkURe@CUWHd(KPCv(mnvY_#!HHvLoEF=ucjM+doU>N>Bq%WTjB3-@}GBvbh$yOtkI zND%KZNnC+@^7kpsoT%Yayx3WjO2N_=Z8udumDk&G?KR-3#VjIXxzv~Cj>0lyJ|5Zu zHygAL;o3YXZ-{{>Cx;f@KDOMhk=LJymF?s^3_)OI9m7)x zL}Ctz0;H0liqOGNWW@VMI51cuTsvb1@tUFPA&$sGmrA>#GXM3}9y5#})XXQ_@}i_6 zZyPFzY~!CjIMR67Ccbf@!8(+1MFn)T^uEF{7&1_gIzoNj~uQ?2R4nCYjDUyV%4Yp(6IB=!w3P_G@L(LL+N3 zhv0f=!}4jNS2h9KIG^5X6rJSlnGC4LMujvI-z-Bq7{$;uM!NWK)5R5%m($a67<<{nh2Ma*>tpPpWGW`ijc#4s zD)o-Qxgt!GxsCFf_UFh%dws-Ywm$oVpH8pOQjRRm3K5Ix%0DQl85J{+);O_r)56mc zFHeFI`|BI~WA;r*YP>e`<1LUcxuJu$v%@1FLb7sA+||=*5%u?R6sm2>!9YFJP2?ZD zJ)M!BIf5A1YV-BRTE3YRnA$d^%f}u{_5$i$#@rqZTM;VM>9f!GQ#dxE8@>5bTt=iv%wx#we zOwt~vcFjPe>dX>zLM~8KE4A2%S5mhT>%~)j!hHtaH# zT|`i7aJvJ;SUsiYPrGn2Kf*Ed_yfN)rjj02I^Tom^Wg(Gg2bo6E{Y7BFgY(Y=tS$x zrfr{IfW;Cg)sZvGMm1uhT zUK@ba*R229xzYKW3kY(9=VGmWmQE~@Bfy|iwitTJ?Wb&%h>lcD-M z-C$J3(x&xu;EGo_!LR@6f>y@ULTfSq{%<{8gP~%+oyF^5gWeFPmxAp9&Tk+UJ_d9O zwi`Or{5EID?Z{MlSEZzhg?}A)M^U(QCrr9}k}-O{!Vv$`s~gAI@iWOk0(Wn|8fRq_S}%?z$_V%Ff#a$Gk}`FA)u zAQ=17;r0BT#u-fy{8hJ+)-M8z-(3ZIZ%HyS8FgW?IG6e`7cN%sU&AiodHb)|{)-1A zLYS8*R;GX&g9JuD${in(O*wY%A{nbt)cpC{kKgb45MD>onNvfJ6{==4KrkK8%+~@+ zK8JNlc4$@glhPOLF+BfDh@}1U?6}2t;@Wm&-!5?E7V0lQi^3(esgB(s?`wb#jXY2pP&IkHp#P6H^@#L0 zMpIuU?(F_$LXk80X+pryRc2#IMyoO}NF@}ZxUDUTCgj02@gn{SJ@a*JQ@a^jNqIow zGEiOe&NOv^Pan1Q)X??@>{6+}j=s#gKN>(lz&%va#rq9USNA`0|9?pwe(_^Tv*%9o}qdoe^Y z3L{qinropKDon4=T@J9DB?L`zx|;~5KSVusyE$Ym0qVDm@6ZtlpKzhb;wHv8s!S&G zFjgo~oo}d6kC0H?UT^{T2w=(oY}82a^@bXdx9yLow=~XiGz_!uH=&*?WkI$27sxP) zNnZ{24Dh-)d|)BuHZp&N-`Lvc$$6KkAiK7N?6cI_Z8A11Z)6g1HLlH!RA zBue+h^G~#)6EbrvI!tTUU`g^m};6KRz#1l+^~Ed+MnZ% zh%UvDN_EQ={sZAa4D*njA zt|R$X&<$WBIuuv3Y~gQHrNC_@pU~LApMirRCKes{^i_$7TwQmi=g@5>zA1)T_o*G| zJ?hegISBtsvHC8oJfik#mn~GC_tKxA?5u-9e5l~3rb?eueQ4^oARt5C zX5Rr*Rei{>3#`ls{D%`f&XeYf0v}LY$j=m#U$@z?v=+5{5YF}Y5{$ZDc2oE`zSX^1 zcFjDC;Wy$+JHd+5P>WBgkfM1k9sqzrEEEM9e*+cSoddK30NPsQr&vp;o?lNO^nV?} zmI46NY$@9k5Da;?QrS(wNsvaALLuHG>IFOKpVqpF!FcWPRhPpJyz6%VyLNp(a6)ya zl`+)!%$`XW2VhWsTvg{rAc}+9!PIhZvCn0oqRfy%aH$il$IRilvd^kdJj;bak}r55 zTz+yVa?|kmpUQADNS5FybTW|F(fcN0EI$A<+SG%^mFlJ;teOanDCJg|2j6FD#(d z@Yx)X@@hVXBf~~(;JyVRfV;@d^nHw#(zEG9?8p{P*S$oPGNo;GHgieN1!{A@M4 zmJ-H^C@KoW`-JD8JQgk_aSH7U6*(Ri({|~l^2}G~@7N+5k~!WxD3@k-ve*n(#VDzb zqZ91766<3OoGl^YMbjKEEASWaS(V!^Qkct>CW*g>Y0|<>kLZ)C{p2b(WivMOS;Rn} zm;0BmH+Dt9Zt0gB33*-QQ>KbJ)k2UdxM1m}$FZ=kjgW*e1FX=oOI|UJSc1|Ry8G)0 zV?35}#tD9NMFE7!XLyDv3R97?&LK^@_j9z%p2o1k?Af%Gq%%!3)vVG4iO#JLHYl?& ztlK>P=QlMA9ZOj?JKd%UqzUhwQn^pxfcH1KM*1(2pVW{NIwUn2tyRb%5V*2^jZ#}w zukNjaUrFwPn%S@ja&E1XDys0ZisH)2Yz^0xyx8BM57F`}qgJaskP}O4>HXeA-#>JwL5mx#v5$4w=ADrU^nJh-D3w_Mfv#DLe5Xh z1e|ci6~Jdw#M|td%m5k2Yc)_eXfD)!A+4Ji^t!>mIvMI22Tgy@ytF>yBEUy+z_BRK zgU$3eq=X0yW`IMHJG94v(vxWCL|y1g9j_%2{^7=WOxP&fMG8>&7xIM_)zHuz8NMy! z#$KI4O>OD%xyV4tlLhpxpj|2=u}$P_tZ`nZ=f_v2 zzP=CqH~5;F0Bw?a??uPQ^U?UY?Aq3)S{_U|wb#5Jdn-9>wRjVyot&YEG+^>7fV_yz zuW$%VIi%h>)C^yMCf1mjK)W)}G-bN2=Mh|O$=sRR(WQtfuiL7j4eC1zsc zVlSlW+eA1FCfd=MFO3f*c5c(<3h0nQl5H%VlBRy*Vg(6Z<472GpQZ}d=gpChQfG9L zXd}1N@Qs5-`Epcti~X}hV;kuAnc!TvzyRz?Ri3N?M#<|g zxppv^&Sv!1@ggP%hV*f1_O`FF(+q4|=ig6^!dE@WHAZ0p5~%Cz+)x)&9d!=4QAWTn z{ra_X1EFUUv6jS9{=YtMtm-~y&$gPj*jg&6tOtO3j3CE!Uv%Op+-ikoLQ@RPPY|63Z2@OO_dWEhly zdbh6zJ14Aj2i{o!kXMP^xRUHe8HFa*AO24fGv?;j0~3eVJ2WkbC;_I~S5Me+L7BY1 z4gTkq)b5!t2b}N+AwziUUh^~A2`C7){^R_^ku8PIm^Ex3vpECV$a|A?(Wje__ZcTj ztvwH}ncHeg?tPj(`iU^;AF^PaTZ>#?XilL*_>n6kwTZb_6AmooxA;^VP~svDcX|m- zrVe3W!TN`DX-p-5fTl}G<u!WBO4BPj>YXi24gHbk- z`E)?m#~4Pq{Arm!Z*5(^AmkZHjZ??SmTBGoEc|Ncpb;OXX*u)Y87wr}h5h`y+QB3~ zZ?_SA{pEe?`bL50zV@48)K4S3F>LNXzVN0}I1({flj+b}y-fAyX=|U?-S&aWojKK$ ziihMh4%1-J|3~)~?fC(AKRHqXd-N7DOWL(%e52)a(vpLyc~_%HDQ*Tg8y*ae>k|ZK zLvad48rbz`~H_4F#)H*DNlS9$xg!%)jb$i3tc zgC_az$ga^lGQTcQ$ZL%B&kA38G&dt*V|q zYXrwhmv^WdD8>C{^Wl}8h80|8BN#%YX=pk4pigYD9;62v7GJ*YmU9@e49goK*r!oD zkamGlhWi-2*bmhbmc22j2WPcb0(%j%s5vaiZ9uqNi=B!)PI~_k&M9FG1j(v;1s*2z zd@}B69G^7$gY$qOs11IOQ#VNy1WE|vLK+Vmnz^C>*|>i8AL@5gk1@C)%)0AR)8sp2 z=YK!FQ3J4)t!V>mMj6joC?e42Q`IZ!7QhCM^(Y76!sEF(unNT`-zRaZdWtE0q&m3# zADnvJ5hwEe%D4JCFD^DXoCXvZR_mSWdvL)KFVr-NS1kojho69m@#6A;eiLD-%ftmy z_sbZSutirDEuJR?+V7S*=S>*32z1IMeYst$KGS@Oa6qnYWRvAw81(K zMj{LQ6V+sYj%nee`}8jiwTTl!Zxi3KY*1NoaPcyDK_M4Hp+KM)UM^Q%+bH?+`pdjhc2|W(xI{RLq=|8hB5<$Hdbx?Ib{G$AKoBYQDRBT#TCkVEJR+qMk zs@=SlJ~%;K?K)|1z^K6UTHt%P(~`=4a)uUdUoA@cIT#Uvth2@I_)uc!u)|ib>Zpkh zN$@+b8}q$@?xo4u{}i3E{fFp`g@NV&Uy{4hmQC7ZOYDBpZQTm*pyenqJiylfSZXwh z<4LSJv%Q%KW@+fJE$pCJPLoV!sqt&`)1UR)8vG98v^E7M&?J>oi*|nGFx(XNCcp1a zgX-&@ai=rI`HA?^vz~dqyuMsv7W{2?_d9C(a(;U4Ek&t`h`zeHDNEg~ASC^bC@mzJ z*(^#2llkBFT?4tge2!U0$n^V$8t zuf#1=zs|S2V(v6<1LNu`+j{OKY|`%ZY1Nr_3fL4LIZ~+^v7O)by~a)@sGXJ1YuVuU z)+>LvYTL^Kv$3BOb+@u<#!E_Bd3!iN9xxvd&2IYM`Lp{iAM_QB9JQy|%tEt>UMhbY`8vN( z@t{_b(xaD?yo%kgqkvyd!Mq4dXungZ;hZNzjQ9`NHno6I5-xjp)TeD@dP$R^)NIqv z#QX8eBI}>|Q)Jgt?U;-n20Y5s9bV~z6<{-AQb%o2EO}_HoHV4@56L%VBrj#_Yb_^4 z)qAh_Rn|7=vspH9J4!cuDq7vnwEHK%de%~+WcIA9^{XmlLPCt{T;aYUleUBDVobwa z)af&^LMx>+w{$4-q*bSmfywM58LCYFlsmAWWHCioRF{zshPPTdhqcXEE|v;SZZB_H zzN*?i!?`N&l?5(qr!SOV=C2s1-q}iCIoA0&pYY@hxBZ(vq^9hW5B)%56S9!@{dnMG zZdQlyitpimC0#@F4b2F2YWXr7taZx(qHYhQy{~BJ$lw-`#(~^Yvpf}JRMvsCSOVz_ z@WYrYBasQ1H0L|9cu#Z^mrj??!?xf-6@H0~nj>}3J0QOC0^bf?WRKiq={GWH>O<)C zW3khBpLRgY)HO8dKfB4A(iUhkxVpYA{mx6RBy3Z;RR>Ue<$7H1j5*Z(=-K@W&;U%? zF=_7h<*F{uXsvFp8%LbTK2=&)UIDkLV9SyDwg^9f`su1&uzF4gOGXj7f#tJ5gFZ0c zMo32h`o@tL%;gVu)Qg6FfA8Q{c2F)7wHvQgVEEsHyspt*hc(x;kj6eAh%!Xb!0vsl zuJ2Xp$Ei3{%` zf6|VZujE5-bXIz#SE7caCyj1l4~n8&`bS~(o7`SZ?)K*M?Q9<^e;nG+cpvtQpyC`t zaR6wm!Dv>)McQ^zRDYbHm_?;hpsxZ--OO!j1J81jn-P>)m(l=1^ujKimE_tCJE@R$ zI-`2|BdwjZ{^29HW9t55ncok0Ms~tBhqEc2f}2n7JhqxNS>=4+|6^l^;Nk? z2YEnZ;CPV3;q6nm{^h5sl8a0iKcz~5QU+sC-i-|;WtQn8=ZPP+UH*ux60+GZBEv{X zdf2obOd!%EQ9s3ns-*(9mB%Eh4I8Tnu4?d9A1$@(Vo`eW5N&DpeC!j+F5^8DSA$oO zyvu}lsFxmLZrQ_(b0|ASV!O9kB7D78f-=!O=XfrKkLW9B{shKiOp4Lp3{WeOT#$SF zmwUc-TSM%_B9xj1zL*gTQhB~0zEPrNl6iT(zT!}2$o&}J#+vb9!&)J~pcpS$HB1I) z3~(<}tv<9?Am`1Qpf~$)=H!u08q-qOh1bgjU!>jVxHJ`czt*0I3pu4q+r)cK%>BK!V{l@k=o)A zMKyP??`;F$b2*~EY1Q&Q346;4L+U!L@32(|Xq@xM>RK$&h=+&2RCgfa5zEo2E&yTH zSPLy6XsGwKcdsj8M6H00&dlqE7K~#vDQY;c$fNWJls~_m%*g|Z%E08Yfb`QyJ+R|W z@-tJ0#q-p&++E!-x|zG=8gmJ&AsX>tvmTvMWbFbI7Z8iR*9^rLaLOovtV7i80FpM} zx+~BPez6Ekts}85^x&0`GKp{YSQC~B$xB|5%Upf2p;10HhQe~e)StrA zG)`x;@sR1@_dE9KDg4?NqzsD)QTWZ!LOx?|z)kBvd1ng7yISmcQ{d-(UHF!@C$Xcr zppJF~Ea(wFNCwR36i5)md^nQz3lK~~ctUbmY>8;K{UGCeIm4&PRzXV# zUJ6H5iDCMzLt4YNfg-?!mqt?1`8g5%sZAm+4Y>JB-z_~U=)=OYRD~pu2%i2HR@NCd zBc7sjpGRaY#Pxl_m6Nx|_2~P-G3gP6Llc@GxN{=?6K`FTJ;~=0XsE+cd&tKNCF0&8 zU{~VtN9a#4K>q$^?%G!OsWE#9D7R!8<21BF_q^60-HFHSd$lnPj0gn_)%`M?Ih%=c>kAZbIBSIGT z1AjzHL(*Q*q!d(>r_eAT;6-nef>KGwjVMC~BQd$1k?6a*g@v+jTw7|?UQlkd_RG`u zS8x@#$PeDs&W7mXfslXUQ0K{B9R!NAPAI~vEfMcbiUfu?sVR_8Lsb^iZ>&`G@hQs5 zcvz2FP#URxaEyZpT|p-c3F}BZ?S<^o=K0jD9Xf#8XOGEK0&1$LUjHHOF7KGTq~kBO zc%(ljDEn$XOWeep@Er(GsObTF z+XCs<5BqRD`x*D!U|T#+tZkGp)G=X!OIqo zhL?>n1g`B%NB!GWf_&-+F>VGd0wb#i28#OD_lZePTqQI)DBnIJZEjFZU+94(>&olU zN^)B%Qy~Xh1W%^?KdGPXP89aB&MO9nNC{|K; zf-{32SPdh?(KGK(e#b8rFmxMjDNfiU>JR;J9g<5?!4Li+;ZJ$-dfw>?*Re>XgYFn+dw9lE1eY? zq~tzI8=I)rY2YX;qu=$i<1RjpWa7gSmz$+#sYCBiFPicsYT^5nj$RSLBt zB_*$0Jpa35KxVK)9aLQ75i#Z_bt%52S*H_R7#57ci_GgoiN$}Ovp(1f=<5Nuw2Egc zMX1M<9o3GQ1VEeZGO~`bV<+U>6$iV|OO(7emkIq;(?QLk(BBCc@yt6DTDQ~y`aRPX z$F`Ag+Bbe!58qWxje!C7*Sc=KSx0F}K!!o+x|jo0TL|*KX;BH99|#(yv=lkW6VJjp zUa^|re6dbIf+vpI32-6KGfT$5)h*A1)E+vKj!d*nn8K|p$LsEASl;IUpcRQwC_OC) z_}j|Ybt|zaA;Q}a>8Ic9B4iUOCj@CqI>di&$8@LK$)t%tPP>tsW)FvBYjXnH-iv{B zSWvjn$W2rT^9$u)AyX+Tc6-e*~<*Rq$&i7o!1W?tK}oE z?2afyuD@1(uh;Xb$|$kqlg4J!T7tHnwjR0jPLCJA_t-r8CSnyq@ZoHM>W1D!cfQbO zwqGKMw!*X=b9NUB?TCKsF51n*y=oIdTO-TU?du={3M#HFQNd*<(e%_Ti)g*inka8~ zO&4KP;UY)lz-457fxFeWOBm6r$soNF;A-}xVp*SHlenx?RhXFb)lBbm?F16ygZwr1 z)OP#KwWV8ae_egWUh{ulF8B}#T-3b#zPo_qfmBfvZMhQN0s))`J_n_u_eiK1`*ck) zc91^rqxQ~$5%g*-q5y+oDR+?Wl4u_|q}`OBJGX%YsNDcZ%^j6Y>^Qb;2gC5|u2(W9 z*$^|~*`hqyEL$M>-;MvEq7Y(cz335*c7C~c+mVkujj^3{uVHe-^?3n2DVXmf?c_J@ z#V49)g>H`mI@vtT%}3^v3RBkHi2^TexoVzlx@4CTCN_(=Qr1QOa6W;P%F;o%l^l;` z>U0XHX3$Kq2)C9!d}V&oWGI%%@y;b3xDUW8E(ZKeROhXWwpo-$xXFEcpE+0HiIQ>=qYoIF|IRnTy^)r1pAxSdw%T%>6hdjV77h-kD}$B}cbR zw)gnHJgt{2L+dvzWcv`E!gTEvc1(*wwr6pYcUdaN+puK9`S3z;7U9^?k0Gfo^X2IFO8m}^-n2EwWRPJfAIfXQM=pIrC&vKnRU8sey;4( zOjNy5&D7Lv9z7VR?8{)v92HX6Es;D?LojkZ65D8m7hPQ%UfKU+MQuYiW3%as6q2Ab zesXV>Pyex^-cS6P4d@g3K7r4Bxce;4`tp|k-R(L17&2*1c3xdq%FT z$2qg!iwJl<2zbknpZ`1sJ;-z};GK5Xp97;v$-CF?GOh=d! z(qMM69D<=$;aQwN$Q6!Bw#NhnYQ3_W5jqp=qj2hn4W{>HmO%;hvzAO|T&DKfk)CML z9i(tz=$^Wdq}j`Z6m^-2dxmaI)Xjhc!n z;QJshL}ka>P1kkH&E3#3e*bB@t^2^g3wx`~)yUQ(6(1GUy>4Pp+Q2wVF>E?g)1|)i z*635Do8(z9w_4r$mgw6o!tDN;NV#UWth0Z;i%@kaU5WAvPqYLYJCVw+4+oDMkvQZS zX0Yv+U9fu%tvYPFHf<`x>zFrrGP}Cn-*7boOp8Sr#?0TxVE@S-E|hD_Fx4=25-0}d zYMA(dmgC!7in04hus^QiiMpu?Gr7aqOmj11-b!ua9&)4~?%V_Yy6dAgy5TwPmtn#l2TPibh+Pcw0;vD=4QwkTPJuTd7zuyx)jQ z3?yeW=+tvN1a_)C?vsU2SjXLIXj@%6e{=ikUlznYUOAJN_{5Z57Y*r}i_J0-LvaPnhY$1}w)(IRq|O&i zKzH~&)0%dL!7hb1DRg`F!@|xwPOH)P&8vX$CK&-g3#bl)oZ;|bwE~2;B4GO$(-ng6 z(8jy_RRh5mS<#NfnxH~(vD}m^;)evsS-g&(EPx?lIM;Y8S6~r13d6RB)tZu6AhMWl z!fpP+fy1;)5ehP{bZ|>Fel~(rs67#x$HEJE>ODuiTxPRr3Qnowdw#2gzAS>?R?9s< zgQ@oE8~n9}OZSgYjvxl+EUXl9Yz7_3Wvb)>J;LSpsLGE2gI5bK0sSBU#Hml-qFDbw zHzK<`Dk1g}rR>*4N92jjb+{;z1Qm)_b$pn_cbHGm)H3OQcwWZ6o2&-#@ZW$VOuT2q ze`Y&EKCn)vz1rvmVYB!WhxjG#(|Pj}0g^vViL1LzKS-yx)Joamw1Jn{V!9aj2_1Tb z;~|)@D)&q}ju0YJ%8qkOtrj#fJrY58kz%OZq5b&rTOj@qdeYmL)dvl1m}_^X)Xa0Q zBoeD(fa`}QRnGEeXI1lWnJ$A+FI_(sLJpAB2%9O%6$o)a^`ISedx6F!_z8swfWv<) zo{um@p)-h3$8oSQLHRXMF_-Iw^fvJZH+QvU(BBjfn@Vn>;xgQGwpFYntPPnpIuaD& zNOHxm0YRnclTd^>5X!>}FZtno%5d36t6l^3XShV(WS4qB<oVRlIkJYNF;2=OSC2+*lwiQ^UWq(s*n(7hlgT4V?t`Z~Y*ut=>dc=|a~Zyn1-^T5wrsxy9OGh7y3+Siy6P+Ldb zOd>jk`m5k-kztdtzQ?x0(jilHmS7Dp)w?3R^#DnqrQc){X;K>RY)OXc6InMpB$efk z!H5z5&-dIz-#5&rNj_--e;)77#ucWd&_03hego%OfIZ35A>HSTikr|bEREsTT;%mE z5VR-ZNHt$kdiMAR4^-_Kvsjs3hE9h>+EQ)$R%$vvyE52Vo&XIsT&+;lw8^9oNl88f zh>oyi>Yh#8st@JydK!ONhi$S30>V^kCivELP=G|WUEDCq9i$8+Bss_vnV1AYX0@rucD8G1)v}HgfJX~w>dEeJecvJBXnDT1qq@<2l740dET}_O(2NfeZ1izb;qrZvO9u^=aoXC0^{l6 z_w4#hv^H3^J@OAMM0OX*z&xindu_SsV(WHuCy8iK(eN{e`aTrCvN{(sZgQa@AI?x4 zC4rmgkVMWH8VQ1ieWQEqZqr4Vf(a!t3YBbeAi@n=yk8y_`9XpM9&PeD_kuzC<|~u~ z66Iptz$o+JI-;(VYG5G&=xwFvoD-7sv~7(jUTvA-)x-w@&{-u688zv`$Sc z#csk*t}{gJZi2e$Xj-8m#C<%&Q~57Sm3R{OnLS)LZzFHT=VTWDW9R+#2QDYCFQfv4 zVgZb0q3ZRN3C@{DuzDevB%_axJVNaS3?TCD$DXRz6&;XD?LyJ#OU4Wh#wP6tTr_i1 z8wec{r2Tlpah%;1~Gm0KnbzB$)sfPloyY; z2Ho*hUIZooz_8cHsnN4z9)-K${9*LE^lhtrGhyq)-ua|V>ESS=BlDqg{Ve%L^)p5c%b@d|6&-2?f6Vya3fexFc^uTJG%NAk-Iy zP`i>(ewjlqi<$NFn>)9_X@J<}(^wp;5D6i`brGnj^2bTYS z_7bD`NCy_rqy(mRoq7Q<4L8v8W^t%(zO>fZzeqyx%RvPqIj@*45&NKzyj|0uo_QNTb$g_0P6d6nX#C?y(im6Y{YxG6)L5{=dmJ4epE?rWEusqRurtTj$oh~i8Rz|_J4cMQ& z`!>#hnv)iKM4j=G-jvgS`23wS(m_BX&V4oH1X>p!ODZ2c=pfAw%c z{`=hrcG>(j(MgvyFHrsPc^iPWfG0yX9--osT^t{~!=ri^eP536-&2IHp{i&Qb|z^cfoTi3~fKf(&xnyQ3d)A%`NS^`j>Tn6eC;C ztGFTS++glI<--h&flUjrZ||YEI9SDp_LP!dxe-ep)yiG5^@PDogWQSw*Ro$IyrIB| zytbRto)61*lMV_LO5`d!L?1*p3_dT=xvNS03}j%1CmuMBw|=wVV6Of|s!HZse?t@{ z3+HXQ#W(-Ul_J$nGoGg3#zGLs$^EoQ8>WxFvWw7i9C2oCjIkXGFRW8S)4bX>5kH)8+UU- z>G)4>SSlD!rC&H@7A+lC2Mnjgj4S2zBi%COA3x14b=KzQd9qg^sMR&g_fRtwr;rd2 z-2|<1R^$YsUH@tL;t$E+*gkJph>y2V0juVaDCMhG&7{bgQ!yW&5XQ_BJo4TlvbV;e z?5U=oVR>EN!e0`pW3foi4)T-_!lmE!Fb+Q8iE?TQznIQBv(zZ7+Zz{F%4|^*KZ`8P z2`Z<9(VI2@wQ%th$NBfl_#(}WqcZmF>>2+Q@lBT&FFxc8{Dv;Kj720KFWtMz576X; z!3A-8!gbSTV#xI&qECu*hM9TW5 z%=2Rs+49QG=A!jhW$wc6JYTDj7TuX6^t^r5oKZ_nPU9lk8zVu)!_RQ#YE-%u@49I< z7s>)ppsDetuM>SVG}P_KI>4vN?_>4fLNoUN8Je*$u>7CEUQ$jtVvgPW{|bri@0Qrv zW2)WI>qQ!z2M_4%Oo?*okeOJZta^mbJA^a12w3;C(`YVWJD7W0u zz}WM_NOgN|kTW)b^DXtgMU8MbupXp`@Vkpg`g?;KAy*F+?h^d{crjs2!l z(s45eVUlXH9&Mi3|Em8hiFsrK$j)tV_xR;AD2|@q@?qE*b675?b;I72u>ql8%@`MFJ#ZdbZ= z@tk*<&;03PpBx0e68i__oSy?gFSCxnYD$}18*gP}nEK>t zYH#PRk15ffs20mibXxpaBOGb^q{+QXIl4!>7@c(LF9w~#@=adBD5(+(EF@*58VcC_ zkapzWa)+Z@F*;a}TEl}-Bo#K${DG+U)tB9~hl1&8XVKO5NUx}}wd}KdNCiZizyQ`E zL7+ATiEq&S3_;Kins}pTN(IzM@TPkm6*~K16q6^!Tq?oWQQ5{-`MM#57=N}tc%AP= z9DN=8`X82D92QXu-RfIzean&|xF6+oU*tr6shzv7)nHb)38hL-zO5{Y^`hukXv;Zv z_awQ-mF^5Pl#_6qeIx$eS{X}68>EMKNM<9?8}F%WWNUjlDmi;Rmu&#|Vq=?m<2 z5IWh=rDa$7NeH={3>j6VNN6OQLgEag5Zo584f^9J2)_(VHIM(4slui-btFHSK*4(d zkyCOu(QZ19sfs->{!$c|HQfyRNtN9*bRIUHR7Di5Yw z^>8q)Ope86=CxQ^dN{bWQ(!ywRSHS>;B=B81^y|Z-ymLEeK#+(xb2R)5?70sXu6MD32lipGNS|d5Y0gYNLSLSmB?6se`JO6wwXKVCgic0ppVcYAdp`;w zRRQA?dM2rnJ=`&Xy|+~W7;eC+jT-$BT@D;DiUizUzPude$Wh!3s1o&D9DuHm)kab+ z9X0bv+-)G5;wd@KntWM@5@^8C-#C;BD9JT~$--y#he5-d4(xy7JvaS8e-avoBTfWk zy$@XY?&>eBTvsLYHFpK%J;-=uYs!}wA$GQja}G-RpQouE7FxK=MO)MnwMtuBn~$ns14~6r;1Tc?%nCF; z`PR3t1K$ciI&g$x+Pz{qU_)*?H-pEU7)t8!BeXs4haz&F-Au)9U!T4Byk z*>CGLm6wf;dX;-L=sNNl@yi&1qnu_%2%hW>&7<0lq&Ca_*Ei?YgkQ}%r4G@HYB z0lO8o*f7TIFGSFsdqf+t&z32T!}n<%TNAKU9XV8D!&2SBzt9xN69*|OI}(OB%g%d{ z%jEm-o*)N9A2JD?=~#pZM1?USDE)qc!IMt*`0>)S#Y$X7)108)k2x9~lDAX}%=i6@ zEsM5R2x}%cx8t9h-|G+9$W=xT(!5)INrBME%JUoEvDUg7Z!VPTPMjS?lG@#uBP|mFfG$$KIjA3OJoUt0KZd z8D%Rv1dPy_^J#a)UCJAx6sVqiRT0DpSmxZg7hf0%-w=WjrarZqZWg50_|PUd zWd_;Cncs>ZaI}YV-Czn*KLU&F6^A9fe+uwVnz5dqz;-tufcH-$=ITufhwhfAmjfhd z96+TI{uPr47#J;)Q0U_rIKy*~<1HbOMHb;o2$DnP@U{v?K@>8R5t^Q#VO^*Lx8elg zX9XVq;`NA`K4I&VtQ?c@pR&qzqtXN?zI`5V-M(Ki4j zWU}l;Q)~bVfmz629%OCcnmo8&KR6bl9v9$O*uH_V*2xMwu-1iyv=Zyc$P09tv~V=) zx>)qt%{NbL@b0ga_t~4TjQ4z@*k0&{Wj`a`B#n7Pl=LFx7cxa5Z;b{~L7JHaFXK6R zG}%3c#2l-@F}^)$>Cp6XRTXz~y7&_~LB;dMWpLV-|A7a7#LGwPs>;L} zQ=90Hh5k0W5&boXfTzL}`*m3P2fM9Mj&A?^**z=6?~)NCR)RR|<$$DYJd3!B5 z>b|NpStE=ROMxlKVV#qayMpRChah#X99Uq7sBvlN#~sxsCpy{%yD-Uc8N^Ue0{n+w z5MX&cd@EMeFe!2|duM5%of+v^gYC2W?)LPQ#$nvA+=dJtoE;e~?t;kVGP%={rLLLr#UR^TO8D{}b+K|Ut zsJ#ngO(ga{)Dp!t9J#}yv+Z2@MkYltOrp>{ojl~<+d(aaC(F4!HHB12Q{EHWh_xR8 zBEJ+tw0XdhZ3GSJU$unHdr+KwMVO055&*O;3{geCDo>V?j4u8rOrFwsohMVGQAjEr zL_9Tb>~#9?=2aC-VKRmN#Hzo*xb923$halGH|}eo#h690gMUMD@gZ@Nl!#TUoNkJP ztZ-4Vq)s&HQbnX~2{oiN!-Jn1|xWl|)l()@dgYP})oT{lkIf{B4LXlKqGF zAqir?t0oqkwG|r9Kk_BT9-QVecspRE+s=tw#O*1Y--mfhGc6@jEB98NHwZ0sn@o)HV%8B0ls3TmgO|z(V;x>|AaH^m8mqL0iaIW6aSc?&X3Rqt zH?5*-h=I;Q1ac+>Z6B_o?%pW$!~7AoN=tG97jy4q`Oo>*+j&we_CSn2YoD%&@cm8ennL>+~;! zOeX~G(8tJfuoFMoY07GE52A$YDBR~hVI)hcvkPKCC`cxr?Mv9*yx5Kos#<~6&ro@! z+Srb|RR|D=S4;%@imudT(yJR!Y7xQ)>1+bQtIXJL8GXS|$bnJC&cNfBfqmn8d3boz z#V~2UKwtxttdW4bQ8H<}O{~yqH_V+#TzPoTxiqp=BX5<*o~VOZfx}h9yg%{1>U>b} z!Ohe0HT_M?7w7dmNvLykZzS%74Y?t|f<4#v+LU7L?{M1E5q_6GTUW1sDtG~xbOFdU zF_YZS(%XKP*g4ePqaA_SJ@SU4hl=-bz1Q?ZXiPZ+*`Q>k;gzO+12rvnP6$qmt?hCd zVg`+#{!}8*ACdnpG;UV#cXzpZ6}OR<);uy&`4I4;&Dafd z0yx|W#!>v>w)SmX6>U4UY(uGU;;JV@ltx!|ndR?X8@++5>weC%ThAXD`w3X`;zpwIv}@N)Gg#w)#+En@!vuH!P?ek+M`g0+LqO|dq#=m|-WEe~wP zK$Of)s|K_gdCd6eo;@q{I;2lv(y?TT_DF7qt2dGJS25@u&r$0)9Cu-1fhEfFh63{! z+IZckw&D#MPG0kywEy}p$jSf9Wh@PBH}dAYKEEcKMayO{EUkmEi5oq|Rm@<;*1adimi=KCh{Nwcc;hgcSJR`XD9!7t%i3k00oI^X z6)%n++eh55ZQG>FsyaP9>>urF@W~p>&`C_tM{wY4XFL7o4R7a!%auz|DePWA7Ae50 z*-m);8f5=a-gKFC5#brdkMzFWS7O->Mme&a!!%xyt2>Ap{a&<(lSDU1x`ClPtMC6X zjr#St-WXqW{pGcuYDY7lqqlLng?8Uuh>W4rV>dr!`1S&X?5(!Y0k$dhq~4^;QD^%` z4!uFFZx6-*88ST{pWHKA-glaI)jl#yLQhiv;S-q+9sEy~E8ep1+);gz{8Tr=iatx1 zU89DdH>(iTO{KaWy0%{HE{2_tOM(ihP5{%339Bm-#Ed`&XJb>-)lS%U^_Cs-=nppZ z-1*DkujfW8x0g0bwTgna8UVDtcgtpg1v3Zu3#y3hZbIbW+FS}@Lq@#teHl9B%=>1o z6R!F1rr2JaiKoWPB|};H74Rt_Wad(JY5hrTej_4t{zIa z)G75SZ0hDCL_zM1!8}*ui|L`ba|luO9}~@WKXgo!xx76ov!g z&xv)oCxnyh@-Ry8_(nS6f>}7>$v#Y&8}ew(MxPRvPz=K{8ddz zV-L^|d^KwCwO={VRURY*vX5bEq~LXTwZK}}Sp}Z+fln`%m`CH^oN9O5^h3?!49+QoUQO3Vzww!z`z{i^GuhDC5RO5M z;%*eaV3rpN2GgiE|0uEk!!=(74uY6ShoHP{nJf!i^h!Z&fBxBgSc$%qc$55@Ux4Fi zOOK;U3L*EDUlEkBA}ey44)jI#dH-pS$h_a?>6#h1OlO;)U}T#a+591XBT>iijU$_% zGA0pBaBq_LsS1Coux+8kia)k6<)kEI`=$AD(yZ#JGTA|?pD!Mez~R0U89)~dESnKS z_}`BFn2}E9#(xhg{KMvDm!7uH#l_tO{@d2^&vt2|FtUc;U_IF?y}*JHB|zWZ!ACh2 z@sQs>NibdvVJhxoP~TH8b5~Zr99XJle%sN4P^SiUuO#Rjg7f{h=)Yx+9RDS2q-Uf5 z-|r)FuXxTI6{6`47GP8kGc{9e+DO)E%1TO>b5{Jt)B4kh76IlhpLWM& z(vvO-nQ_4CMFcHwB?Aya{3`kJadbVC)Ss?tW_R=>ez)JNmjb@6>@S!6bDIM%+ZRta zVHH7S7@GK@)Trg`op1KNOb7W0>Clq6?L@|FYXcSQAP$}mZx;{dj*o6>2Z z{*sY>qEx{ry~$<-+Te_`P5i zeeVIfS&afTGtqW$o_dd1jdc;+QXoe-T`xYPeDNoGCKwRZo~;EgM%}A_&%HW#bcfDWt3G;J z>2-^^uk3l)yU-)IQlICdDX*EIKll4+c*_vEl+3=+wMaVbYU)JcYk9N;qMHHI)fahu z$Aw0zHn>(%+E%TIP_1p)5oQaZ8_TXM(7uCRnD4YaAEK-V=q*&WEFYF@`@U+eY~NE4Z!qb2c^lJjAQH-TX%SZ!%!p_s6Pcc5}Q6acIHyes+tTm0IScy z@$%1tt?U|ZE^JQO{LsJ#!(8U47xrA!nlxM=->`F5sM%lgej?QBFJ`RZTAiJ04Lv$4 zb;d{*7HG8IwCJ-T==KEtmDX1!#8?*RSK{)?ndQjivBw4$3pz2j(hQYF?pUIzsQS^Y z<0ggMNLRa|C`%T*az_x>f*DIkX4*JLd&!4=?r5uWY~xMu-X;U`3>iY-e54Z+$xdul zxV@p`cQ22m*iOu)f|w*EiraeYc9Iyud`?z5yrU$W{)h5z8np^`)GhfbB*e9mE^-|r zNjT1jEnxs>P5WK(8Z(Qg@H?u2*;?ER!{3&v!GgcQf|X?CZyVxuJDJE&I!!ct3BN0< z)ZF2Ml95Nw*E>uexwd*=RkeLBxRjY<5$&$p-(zZA2%eOMfnKJ-B& z0Y9$}+W}>ErKyi?sWhQUBa%q`cBe39Qvj#Gkf6hs+FNn1HBN@unW?N+WzrqRoPB2^ z7og~=2%blq{E#2+L;uT5%KRdG3jGJV3i86}u%Q1^OX8;Nk55s|*k1NVlobE>V&GKT z5$c}HW1FAtP>ueV;ik8A?)O`(UY$Q=LWlJ)BK_-e#oFd!JXVA-3kxwpe^%ZG{qPVs zKzXQ8ueS~#Sh~3TY|_pL>L;}%Z8TIDGDoS$ zHNl$Oyb+^^Uawlvwms~E0N_g=H33YWX1nAZ%To}@D#&5Pjw??g&r}iN-tOw(`WaPGt z$($Wsd^C;4Ds@ayw3moOE`~XM0df;>R6yAFHv}ai1YptK0IPhA25rAs1i475U}kdS zustS@z?6Pz4u>YkWd7UGz|dV5-Ago~?-2AH$+>~SJ^}s!{5^Wy8)`z2DUx!KPEZxD z{Uba9dTOT+XTb^OVq#9uHRVd|mYmvM+M*_p=?aMW3C#$P10l9P~lv7dG06Q7JjhdP>lP7_iPae_!yV;oN+rOp+RDfa0S5wgInG* zoC4w#xvuJCrzUX3UmI-W1Q<_aC|Et^DI1AI2wd{IGo8bU;0=u5b11UC4@4J;735Zf zjXD!_KRlhOpyF!~4vb7*!}c}=%T^^wbwnz+;V#VTCibgk;DZhTX+uU57%+rP?nj8> zH!$N^4NJ}V?{J=Rr~bQ0&|Sh)X*PG(PFQmWAE359{DEb?|t{tNVz& zt~}LZ$tDQOgiM~(ZycBmF@m9CwqhF1gxmANOWF@SZ(y1a11K{d8a3rlD5m4WD7Zxm z+)7DoJbHqh!Aqd0c8Dxt<~49HlLWbm!%QLHuXp6tv68;L0?ST_7ZvBBo5}b<#^_2U z`@+Ai_4!7(-SvdUp)?6`e99vd8N;%lse(`bIiR*NF(_rn61u4(c+`V~p?gGA>2D-6 z4|;JKZaG-T<{?fIa^YnCjqC$NqbSw^`{mbeg3!VPO&OFOe4p%v&Jm^7Yx&B#kG41{ zoeUhTQ$>;@?iGq@5r?8hmI8@tLP-QOJ_n`BaKB{52m}r`P#lGW!K)G0X;nFZ)gXEvO()eZH zeeaYHuM!d`p?}xG?kIoBP4p4~jalQtSFxs3k;VPRZQ zwhZ@R^mXMP!p5*fr~t0KqPwI>yCP zqXdoM{!v2^@)mV!Kx<>H1R1~jlM1_0+0VEY1u}j%f8745q8$OKIXPKby=(ru;$}7U zut%fwZRN=R9H|xP1GoejXL|?po)Ub7J;I|3$g6~0J0S0MypN@LgHg}@{#+I$_!d__5SiyNd&05ir+o3$pj zB35LODs{QaqCPOEX10n4D>J|7;ss+#SS#E~?bnPCZHr_K3th}u%g>qEH=J4rP>D?R1Eu;?w=jn_Q>w20tiJuCMTIoWIlax zi@1(Qd`BiS2MD~NwPnC>_`k5?De)~~I!5Hz{uRk}WD4vT5(J3( zIHljcrKqirLws|(!h4_-)j7l6prF4w9!Lz?LW`~T^zPfWWhX2^xRLBMpU7-gqdh@d zqhXE^_@_sG2hYpjJ8gNJiKSJA_NpIfbsfU%*C$u#ohiVidoKjkM&Euz?zP~Tl?Qj( zk3Q1fnwBPBW*Domndp(gc&XH@4dB2#btWG))>)hRU<`u{1vfh5%=p&@{P`{V7nlX} ziBr-|O1QQ=4s=*~R{S)4`HGzWCO66nGFN?s6) zTnkh0RSJvz|6}YNn=^~HwH@2+*tYGYW81cEbZpy6$F`k3v2EMzIC-yY57h9FAr*rZ(|7#EFxPxG_O=g|}5HVo7oMab#U{4}p1q zP|p%x5`J+>@5=smM3QXMlVR##mFQ9)-8es}*;awWifJH7+QabT@#l zU^^ZN!Yhz{i8cZQ)gJQ_i+!lca>tg%3`2u!Q0q$|rL93m?ryD}DudIuS1cFKJR|Hk zp78JV1Zs!g>z`NDJCql@q02j~G)n|KJ~+D<(>;xlgrFzP8|Az8@<6%162M?17fSkA zD5YEm^M}nW%OK2irx$E!PqwsMQzjrlNR(gV?8jp|^`@q_=W(#8>6Y4avv%?K1xRDn zG%$0?9;aq^K|}&CPAhq$uDv=bVSew<>UJc1#<%vn77SK00 zJS_~AD95-VjgF_*kzjb?j%Q){2y=|R1O)J%QYYMixx7_QP!vr{*Rnq(8aV&3&H4R7 zQ0QQ~r#e!`T5tkR?TK>MtChz$EzP5aVm`kU*`U9y1)s1*)k({-1>;;qwieo}S4t`P z%5*YDj2_qpHfY81+WV4N$YsYSN}}&_5(!+)#s$lfguY;sORWOLHd9UZus0%+kQf~- zCICq_UE4zuf^gS(aRG+N?qy`{bds0OaUK{WPGqE1^0)K9NxRuvOvFXqR}x2lA+m7- z?6aJ<9b!u3Bu%3N76>>K5Ug1_U$td^qky#C#L?{7Wje<@zRNzWg449)!r}~0dlJM? zhn|}(*-kZkmjQrYbGBW1jBd3ozVV&9#*{BMNiIwWc2yCWW0ch5@ufWD$jX~4C0N6> zGO6JOvW^P{>^VG=4d8)LLFOUA==DI)0ZDtk$Bk@fXbO(}{5_~vD8u1ed* zCgAy;kAqBa-u?d2+ThkWlbQy8np_KAUL&3nhnRz|-ee+xDfP$6 zI@yRzh3Px;g$2e&fs^^Zm_PiVTg_&jtePc7%au4^?H zFe&SjX7joWH?@-fVt)$&a_*+}g(I9Y8X_vmN;rGGNO{`Z&=?7N~UK>1gvxaxb?;AAM43g(bPt2+I4+qe(rk z5VBN+ze*S?dO5lH{a8{zK|Jr&d|St~QXPVa30(Hm4Aw!+crze)L4x5&CkXULQi}VGC7rBnYD&RJecn!1 z4$wX5!8&sSyIBAA2X+BXB+z0WGD_FXY?~u-R{2VaFll;CcIuRY0weG|g zs?!y7J+)!AcN$dn?x(e4VV?^HGqZ0+AZ%G7oGRHm}#95U2s zBl)H*vz)J}n10H9kX0SOf%u(sn!X@alL8RMV6aDGz}1cq;#aDoBjnsxDC!_Xd-F5kfLFKTTgPylMOyn6Cm_(M_e9xBLT zfW`|Rz+6X2$lMk*sSP6e&<^Zbu>qQB0z(oYr0oKtB(ldj>pY@CDbv6kJ(xyItzYb- zMu?NcBBdOE!gTbpGF!DwmGJWZ_0`(J<>Tjh``Eo4 z#($SFj=&&*=YZ&${x0fSjG%jr`jbOEn)<`(?f-B8(!=pp{q*p+ZJAq zbNfpQMusNCxSKZHQy9RA>PIok*<)KN`%CIrxI}yqF@;*ho4cF$We@20IAU+#SubzE z7P(LUGI@qT7gt^RlVr{Nk3T*N(v(z)xGsHia$KTR@|th_)cme-e*Dw4zKPL-I_HJ{ z55TmIjA8q({X{8QtW$DeO_~pz$2nqBnSji-ogPz~WY60UpxUT|p0#AxxUnydXnqZj zYt#%der&%Vj)XyEg5%X}noN{=1|6)RMS$gYP&QU-6FYMbT;8^5I{&5<2P$C(pQ3=0N)TzK>STb#E zZd-J}8YuKTl46Ce9D@0#dd(TXlZ>dW=owBqLeR6C^C@%`qA@8Hx=BNs9`VAj9b{4U zgs7YtI!bG}zD_M&dg{=AxKPn@oL|h7N>>*qK{C<|lPvfHNWzefnc>Vr7T8y}Im?&a zK!=T>0sV;=EK1sf$smtRi!HfN4c!U%OBW|wbe;>Ssn$nGvk<1$Z04umrXgpYzkM+j9`o^4oC3q!H-#@V9-`<^jf-U;qgLVM~`a z?i=(z0iV*1$y{GUdws)2naWHIsOGi|96AY#CSR?9ClIzpGb|`LaMM|x6h`?XQ2Bl}4>%XY=-)7cl5>U|Iv) zHs^Oby((pai<(w-E#I`7LCKhv2q{wQRq8Z8Hs_pJ>6XffG~HBcO-lAL8tSf_BJ}Ka zci1+jaz5+nV5Ddc0)`ySOfq#0g5#`N>kE?3&W)8xWxtJ!`ML74pcIL>7EmQY#up(A z1ZZjXZ;i!|6w2lE{?@ zk$dtp0G(Maqu)vs%*860DX}h1RV@eTO9A8y6S4q>$DQ`f z;-AJ>*a=O;+I+;u`~nN95Qz&&QFL{&I%WOkt~|VMtM0PYQmM8%khtG`^s@uci&tc9 zf)M?ZgrDsSp4Ai-M7aYp4sd0=A^JI%tuYgUbeLFpfzuTsClsug#GyEh9QYX=W;I&~ z`1y`?o7C^z9q>O$Fd2+QG~VA!v}RHCNxtP+AmT>iRS1NtuWNPIm8su2bJJ04pvRNM(+s;`THt}Ndn0@?$+m@ z9>w(7*_r^ztFthWG3y(0n=6&z2vDnm_%NU1GU+B7Te7?Oj*t}SD)l<#Y7U&+i41Wg zFQP$K#rCyJ$BY{;FOBESb(4szn~-mP0%`w)H_5cnS(*4Tz}yLp>X)#lfAzSD+FPAP z^=eZcp2W$-8-mOb$e%%Umq8@8#iO)Uxu~lO>JdFZi{Zaq zi!5|~3TLRh21Vlhe+MahiE=?W`AHiz=T2EX>lnqrs=5ZB$1k?5AU<(M-ssdzcRamjuLfGz~Tk&9u~vwvap*&(q_Pz zO=rPCOi09G464YC+XFO8dmx4k37{)p#7hQ4!3y)n;);O9rPvZabhR(Ci|+2(_skFg|%4fyJKjVwt1_E6gcoe&#dDD=rON$Dvd z@{)26&^08KwOn)*m|HU6g?NGa4^KJYRpiNlB!K%hQJLx11eyjaKrn#{Bfmr;!5<;1 zM(Vy>%LkA1*ofq&0RB**F>rg)s7*n5CGkMMZi4Rx zr9Ze>c}&@=Z>i{yAwx2=rM6F@-okt#ccdgJ!D)AOZ>flEIs^LA2HwOT%3&{Y2bgz_ zTRX3T6T`|_f>C9N)-iTiUqH4jgt+W>J!BC;Gx!0o&Osb4N5>A~4;<64wo=G0b&g?+|K z;T<~1nw@rb773@smY^{rhmX0AVXoy$kSf3#v^B@AWF2 zOlfBIz8Zcf@MJN?XKAb}un)wkTq`J-o@LcLE{`q`(j?U1aC4d9UxPI;gdqq$Z#?k! zO5-_Tj5E+Sa>f+5jTExdI;5(E+pdmc(#u09_(y-S2J&L=Jzbyct80X{zz~d<(dSyb z%>B^H5GL*OQ|tEtTP-y9l(2)E8D!@6yfo-D!GD|C$}ML&2G{=yx4v5<7zGBTiZkub z?l0Y~fm2J%VZ$OB#2V%A1H|g7>>|B(*8ndEwl`{(9sacd_s))J5|HiD;ze;LpFJ!_ z;!(c2kZv#}$n{F9teul5XW9uIgUwf{H#BDX6UHS8A?VX(Gxs3a#%%F%+1z`EqOhre z_`KsG#20HL4H~^(uJ&{xUsDvptMp1fVm)>oRnRc=V|r*fTf01s-Nb~66F!#HL30MZ zlqJ=6Xi{W8n>D>>8-B*tUlTI=jM|Kc~$&Z0`qU zYo5*LL7}egXQiR5EGWG_XnjvKCag7UWGmC!}$?kqAZ3>wUAY*mQg zv)Q%HaK5EG6tyfvoDhteY_N-7)78c&WVpz%FryN*?2D!M(Inru&}-xXBSHA)%FhKH zy%K44X{hE^+DX3FnO9ApI$WPj^~r+Sow8W_=C5!=S~jyO9xfWL=Kd*iln0AdpEjkglZC8S%WP` z-yfpADsx_9#JgkLHw8*)63&MN_tAXV)5c0CR^{q(=Y|_MIfes*XWAesC;xP*gq-2# zGBC~G$gu;$W=u|V~WW^5m@2aHh&q6-^Yd6u%dn!znf(qI| zwtGAE8<-(E{qNxnxK}ZQxk_>b+&26?qt)X>G%9YQ7ULm+S28dar_)kwq3Yk!Z341C zNV=uIv*7%2C4qjO z^mKwAQTKks-rN`Knf?BwDtVP=)^}<`2{FiHT25;p!)HT4KME?{qkjpWBFRjH z$hMrYw_3jHy`(Z@kb>f}Pel&`1iu2AwfG`qehfwrR}i~Mpk^_D%L3kC@|{!smyO>^ z-k=2{244eVt^;Mn&liLV4kL##ytMiyWC#r%3gLtOSu$(sxjZ7ERu7$x3m zcuj);GFJQOiLGK>;;J~Vc_M)~rz;iSxh;Xf@FVdNO#Eg@U?(w$=NpY6n3#u4EwM6@ z6W~x@Lg7?KNzbIjtiVH-Gj(EUVj(H1wf{|Uo&4dsDsIzzIOXg;1OsX2ASCs#0{&13 zkb?)kcv!M3{dSA5j(AQjS8cZA)7wbY!C#`ar!6~YOh{Bas-rTp8fW&$pw%r-wj_N% z1h4XXHLS#9uUZ7+Z3QaAEet8%tjLch+j4!ro;_1x#Nfdv_%C zan9c0L)%WfVbm~if&kR(nc^45SWYx3<0Sj}=7Dr34@_Oti7w(2}zFpH8Xq5)7u>Y_5A^?-@P?+d>x< zudcGhc0BvQNqJ1OX&~LYtj`99*&{0dID7aaYf(NLm3xmyS+bOGQT7lxGuX_|E3O(a z=Y;7S$Uhi@zNl6%*+*cLtGgA!F0o82k7qB}9HFj?f#E4wH&Vn%t}7v@Bi@0$QCz(@ zk!ji!Fz}`~$ZX+K%IZ}6BPCM~vpeUzpbf&ymor*%#fnCW$%Ka7|BmEJYq%(bGHoZH z_V}H|DP^+9!u0`>H9eS-n9PSr?1@tmwp@ZbDS0Yo7eC5nMPz1wHw>*;0iPE26)XVP z_Uq!9d>qLAP2*{7lNZAt>b2c-8tB+$f0m%9zcrEBkfo9EPpc)tHGUj3o&$dFLy^wP z3+QlWF7Y7p8-}f;a-eb^#V1V3za(qx2Q1f$gamb1SEfNFPvADEts#4ZcHP@Z4`suo zXtXEA*<^@9d*APMp)Gih1GzdkA572MoD!2dRZ@-3kA3O6ZBe%i{B}In{o~XN!Ap$e zb3wzLq`$A0d!wIbXj;`-+Jz*c3?V)3UAq^TsZlLJTqzg`@+15Q#-C8hxLo@)I-)j^T@W?lEW=s(lL7JB ze>}KMB;I@=wvf9n4uTZ-qs>7+HW(-N zG$CU_rwSju8oPUeTk-_h_FU)Vn(mtCQb1GoW zf1G?eDV-O5wH6}&17D!8(uzuI2*CG~>$2G2cFd&5Z z5!|z}co#R`mU0ggSY9?NCjH|2xEm2`?ZmbvT9l^Mhr}($=CmA5x!Q4$O0A%dQ#;h< z*G^26rzD%=W$jdt3oEO(;qbX2N(PY2q;7^|?f8ZL{rBDC!?xW7gy|RVUR164?b5@& zBsKnmiuSB8sfM#@aXxHqH1O~JUnTxW5LX6)6j{NXOz6MSP^SM$Ls{8b{|}YyDQ#KjO$n#&9~!z6%df_>N_gHW#UV*zZ5oxRcF5Vq zw4QFkPK46<*)R0BIN@vWE1K_mZm$y2FC#?wFI9Z5CKE0K-l`z_Im^yOnr3kxg((pQ z+2_Rr;cZ z6a>-4!DkS4${rV=fWX@zvKCB|yv8S{c)D!@#^U+_VC;M}l^sLq**;a=-2FOXDAs^+ zI{G?+J_?ECY7TK9vN+~r!NqLOf~wnt2mW!8T2aPa?!aA=0X;LS7;VW(QpC=dMZ@gYKwlO=ih&zOww& z=`LJD)SA%asl=x6^K;9xS65g^)U3yL-H%@!}sLyaS|Q*iq3!+dGSVqZOsf z0p|Dv(dTE(GG-`b>hwrdx|Df~dP+e9KePGQTnKF=+K}ym8OKaqx7poN?rrmEkKRm%}l#6bDjM(9H8% zZJPtZMYRa-3>RkY*X?Q6FQ_pMTxYhAuecc*LI6Jkxxmke{v_y#G;1H3V$0`OimJ0t zS!I3Oe>19Cm6lBU7Zf66`sYAS58!LXX=?-CkJ=ip&~@qk`b z)mDV2CGjimKdbs`(lPWoHK7gvDX(#=N8A@%sye=)g1dZR{dKMwse?P;f#F> zl{`*SxXRiZZX&fNDSJBb-|mB$v$F<<)(EX;#LX{HI}D3+4UcS1X=%I0ktr9mRd2Do zs`3plCjI#=)?KWc=dl>7qfJ{8T0vCJW~r5Y8!KoBFrA;5J*d=CIeui@Pt;h zjOQx!3j3W_Lht%cn22yHi0n+LVHssep`5ZvA@wv+PeF7k#Ss*)OJ?!ZbmwlcB}uH# z1igT6As_Hc3WPwBN=`7c=xJ`Q{q4z|{3gr(VPgnc-bnZ_%$hMlw9<#G!rT+#m%xV} zMz}zjoL}uhWwPsctj|!TJMEIp{{qZ@a@v9?aXg|ZA@!}19037T(Af=x`sGR0s@9^x zPa_>ypW7!o&Cbz}+C<^m3<_~NYAH1FhDi?b^x>}n`hGR4^w`cLq?w-z_vs%$SsvN$ zelojzd$G=KKGDHR=FTyGkInlSx@h(1R}n^gsGjGrEX05eL-(@1Ky{8zFa|KwaDe&i z$(vObPw7}Hf(vbJAS!boRs%=!Vs@g($tMY!@mMp;O@IDWkhf^m*e}l0on7RK`P8F7 z;)~vK;dM&zyoKICaTa_n<;nxb_>NHgJU}~^>qwSt;+IMfTD_mz%0zCNe=0mx^$(d( z+uU_Y|3$W2Q-c5ll#kKRU9mOAi+Ja3X1?ew^lzb}(q2?x*8XaGn$(9r?2^-Z7Y6+toD>J)+aE~~L{A*Taov^7$OybW%Jn7)a?1Cl z!?Zp&C**nTjGIMlq$g|UoRG!0S?X2I6A(g5$VeNGoVYr;)+U48DYUl!R8KW#`uSl@ z?EM^m6LZlxvZ*5W-Hjp>LXL%(-RwMaxwLf=#hD}Kv&Ng`#hIx#vLh!8yGfZ=_so3? zqp7S9!KT{O<<1So30tal?jCb{*Z0?y7O%CEeee9r)b3MzYtAb# zOy=Y5F+uZLSF*056^*~J#)KbQ^|=Oh`j|5a1li=>uFTjdG;MeL)VeD3`ByY8)j*X`D?X~3AUE)Oa zGk2`+bh~v~;WVLJL0jd-GR1wl!>Bc2>LBxey>X`c|ny z0i0yYk$Wn%DfiE`n9;0~ai!DHnV6rrtOC-FFS%kG2cb+TfyuY^{Kc>8=jHx3)SK)i zBgDqGC*RxdGep7H!w3+2v`J#gprB*3{mv)DJD zY_D9%Zd@&;>%)d1w#aYgUCkFE!_D&YFKhFAH&EfY_fJE4=i#Pv_J#u6I(wEpci&tH zQ%+HpDNnzWf4BWT&4%5T(fd?iF=tg&(KlmY_%eU8{(12#e%MM6~Ym=(*U0_&6<*&h*fL)ib&()02Vi)buaQRFSjF zY{)A6-6VM8_Q$Lj(`jb=GVG>Xjsmn zn;oQYPV)n~Z^2I-XpJEI9ywR^tcphNgt(l&8g1{~mbh`As>a(MDhuQ0|xzGwWPsBUXcXWNf z2%;8ky*L&kSVj*f^YsR$5ENaSuTyNL`-K(BJlE!3)7Mtevnv*%e^evIR*jIM3L*)3 zEsI;td+=@kKI2^c%OWWYi_kn+DF$h=+T}x;F%Cx{RnjnvCq>@({*Fja6yaNrfJly9 z%d)3zX1sWa}zmw(&)Tw*-W z%w$4b?pUlSzX4Kgy}QU6vGwK_DX~^s$8@o*c80K@dj#ue9chvXj9MV06-XEQTMIwk zr=XFCWC87+_!%iMs>T0~tqlS_FVcb8Q!%75{yB{GV^}8};ata~j?e{^0x81g0ZDAuYkyp(onFtz@Ia;+tDnfDj z@;uhh^xfye(Y$~S3@VP$r><=U44izDSc4ZZ2{h@o4WCDMO#42=*YrWzmmzce?b!Sj z1S4H8neHOwjM@xh_?aN8zIa_WcSz+v*|15@s#?lg4VXvrN{I;svA6`on-kJyk$y{Hyn43qR(GN4<1 z{*i2VcF0Yza(i%lNwYpV<@-cQvXXdB^4s5tx^N+P>GW}~mJ4_w`Yln`PbmC2ux0D{~-;)ZW zt?s=RWSXs{7p^-0bgZlXvd_PeUPw_U-&Gf9tobxhVdB3mtpiHux8^m$K!Y7@W*#1s zj8@5dj=U@Dd9XTGuWG?aMT-&-%@%GX)||w%%}*hrp zZ12*)E?I|w7;b;A!DrCkMgWGLWT56>A>6;aI8U;42u8K-;m;`hJd3Hq%9Q+p%h9@RCEgrv9t8laML1DQ2%jsUbk{_0k$MyzPm zm5GDytX4}qMrD~NUJ4Ly$t>G<-@hQEOg*D8hz0xx?y$-!Xk#YE=9T(^h30KSF zJi8X8sO+QNN#)4p8>nUYyf$~(`~6pzaaYUIPTI9|^5O;JV*DFU3YZGUAe`k9ONBm< z(#VCnXhHpJWW0HI5wfdCQg8(Pg0-`+EA6UB_PAcwpJrMFEP&u$x;^W7X5}%Qk*9?rVf&20$mIW+KN5#g)mRS92mE3vJ?Jirojv48_Mdj(~^L|k|KhWTXSo5dr z!XpU@Q@HAxKQC@{aBL2E73H@%G9A+V_4~`<+&55f*M0bZqfN~Jg*I_9{~x5K&HtfI zzY~9G(*~}`dReSOP?N{ z?!P*Q#)?@B>>@rI+tYjq1yxEF)1K#_1m4LYYeE>+C+s(a9m4$JQhy#I=?)*nzc?m_T*sxv(D)TU-A-){9y7B|BQ7kD)8Sao3-n z439sW2VvBr%vm@=rqk6PpRFx#!O~H+>idTbQ5DtPJ|SbcJzJ4!^>ltW5HBON0KA_H zXHx_Ca#wue8>>L)+r&<#yompu?EVP7(}Dkkb;?VLP%Cb&49g`{>j(;F1|`SDDm7}F zWOfDy=I8fg-m)snr0Dq^^Qj7jT!w|p2#!xv(H!7xrnQMAoed-3>LSaEWg)U8E5a|d zR+7b7ti1#Qymnr!SJ&?_t+0f}#V=_} zOw1N;V?)x@Te>Brmj1TFs}W`fCMJp#!={D~Q}v^*n{WB0TK9vKR#GkBTP?g1o=*O^ zHu1D)Iqg-Xht@yet(B$A>QxZ7Z5DY~!3o_UewD`{$Btq1DZe6+?&rPabn$#0*qt@d((6uz!!M;9BC zqRU!ik`d!396J$+0?(u>PD}Qiid6H&BxGBRbU0gqmW&bRt|5l$9&hy01IGJBHLG96 zlB%0h=kV{4UGZmF6q!%|N{Ic5clrs-n`7x&LWwlC7v}*&;%d}-UEJ-ryPeN|J_&9C zVO#GhsWxb>M-}G-WS1bLcM|<~rwCo(c7DJ`8&wslZ>t7tD@U?zMW5X-3eR~JZZWo4 z*C;UKv}^r@C1tdziERJsy7MB7ass$wsE!D!)Z?A45uGPW`~v0E9FG1HPn^_&taA!> zmh3bUIJnWlb`X=QNVN?VnYrYAQZ~+!66B9czBKm{fPP?2!TLo$k>AkK7(fdULEocr+(9!8>F0xS ztjvaRe8IqJ2vK)pah%oH;`8lw%d>uB%q3k&uW^h3iF|M&@SdR%Ri&C@lSe0{ z8aUQL>A^wAEu7^Eb8yEMKH{A^MWSA@>EH=d0B6cU>y;l+t~<4^?Q?9^vCWjvJ6}b> zaL3)~Z}KN77S^`oU&oyJ%if6O3%Ir7_d`#-P%3Y|IYxT(hwocVmhQcl~VTWLzC4!DyZ2W4{+S=*6&s z#mc|o?!(LOSWnhI_N9Zj4uK5O(W&Wf=I0}i9$m$jL}D!Z^)-HgMMb{7-Sf3oXht7BGir6R)VsW|hI}Bi~`IXon?;PgcN`j3--@!K%$fzdh3MxCwFiRE{wSVfrH` zP(ro1n7<_?Frn=4S?^k)@GAl|AT&zTw{J;c4RaJ6kwAtjpawp01s`Z|17zWKO29W4 zIAhDcj|pHVBNL8gdSb5?7ujd&Yv^~DNRHZtPo70Yv+Q&$o7 z#?KDNVLg;%**6W^Ff4S1z9%m{ErZZ2w})EgT?Fv>RU4$O+Hw5aBZ*vQ^dt;e#f*tv zo7G}(Q0}ZO3aBLlqC#kINm4iniC$a?%=G9#dz3kCWiF6#%dd0u5?P2$**JgbN|51y zIc71-0e-C6{IG8QX;$`f+qH8&oTOw4QW)TMX2MaQV(%?)F9rPwNZdxHXzdy#!?aL%rx-NYE!0#r)uw0WJ;v1^F zAgCpFtIlOrk;(jz8imRm_PaU=EA@ulpfcooB7K2Y8?(Kx?LLq>X6Gy#*6dc!HfPeE ziwKW7kG2;eOFo!j>T+7|Si3^?hH66)bi8PCM%eBqptoVA8?VM0Sx?-?040eE3aC;T z#<;mW%r!nB29d~pyd(;SrTC5dO9^m>M>xVN4O~j>9xu#pzzBRB5ky)X)6mvP%MI0T z1AaV-PqfuOI-k55&f{~AMvh>XvfsH}OC?kn_<6-eW)4=5k*I`AuG==9HPo#OcTn;6 z$=C;I4F!{pXr%&aQMt)ldkcC~B^MhF{{qCzmfu_Q?=YuebZ{+-1tlt;ZNFSM1h02C z8ON|S>mGUZYBO7|YwaDRi}!UGhRuJyb?kR`0jD(%j2wjr?|1GXQfkFCOj>_~T$>Mx zwbyVfMSUTpf;_Z=#D(EV`IJ#1TXBNd+~#_My=Fc&nkb{&Q*t0t0|`D+0`t2pm>vId zGBN8N)sZtiJEHx2CA>Q~d}-f)W@n!Is~>Bwse6UpdFmw(_NKHk{XHl6UKn`1vEC{! zx^fAXtB;)2n~Aqk*vOCs(zt-R-2aT29hUU$EbVHHx(ZR7|Vp#Z+m0 zj3-egY<@PJ35i5O1$?DI>=*>4nRx0kJE#bqubGPi*r#RSjn5l9OR2za4?t_D|NZ<% zLGYNgx78zE@<<@t{tHE}6nFShbTMPZRSxgyFxsg9hSQ*sRst_i5M8+%X>e_ zORLX)%B}G*Brb^Ah0fhxLp|Ckgw}{|qL$1dx&rb5L=LOFJWE2zV5DISAECHSri-^T z`W}@uVzoJ{6RKGJkBkIYi0tG~Cx$nxkCknGD2cQT+2S9^;0+VBy6={}IBL2@%{}Kr zpFWnL&xK(O<&L*uEPB^*q4zL=&5@0h0=j52=qXf`hpr>V&#CTddxO6@U&@ejAdx8e zFo6KYZ=Rzwf&Vg-h#I@&{Wy{YuTXCXeHiG-frW5=$txYz*_KWBQW_O#2Gn4e@E|#| zl&nXdUKkwg-DrUa`MvU7U3@tbtR$s+OPdJ7MMRr1|uXIfoz zQsSk6mp6fs5vKUkK%v3a_0I=wLc0E3s=*?oKFDuLY3YoA2G_4q3~G0OQOTBZQPk=} zVh7%jMr1ve89It0P8VxpzbbV|0=p{K|2C=%==nLn8Rz3gI0){9Y%SEz#K!gXzT7RH zWkr$V6xXRhy}}x$6ASF;Z*r%>niU^*^lJLcVE{!g_9-_}p;X_*Ws$bbf;9d!^Of-v z|6RAniho)PT8i6a4k_e4u4)EGirdz>O_VcAP*UF@wYjGElG!a$oU zE0S~fMNSt=*6PFGt5Bb;?2!(ADXvr(IWX#F^2X`W@1kM>(u^fE?;MeM&9+Kv+Y zc(lKcnKT4Zc=ZUax!7aeGEb8jHKLb+GMOMS7PwcM98A)|^>RlHXtQK1nEJR0S%KP_@Y&*DFO?^W=3mTgfLZv?n)ubqwNA0?+#7ivI9@ zQ?-8~WKZVMwlPcIV=oln?4`iDMsF_XA+^|*W1gPM9bpm)Y{0=bLVbjB zP(3j&2%!k1>5q5Tbzx2!@L*QDo9ac2R(hE_4t}hD$-bl5TRW^6Xwyiwg6CT$@P!nI zfdt@D-<0IL+E1eoK4N5K@Z}P*p_uUbfGd#W?L=pV$=kxtsn3Kn4hVh?-pkU(&F)1< z3C-IOhx*xsi$6dW>ltRdwZ%PVKdq?60|q#vNBU=SB&uFr3GgMtTFk}Wap}$kT%%{2_;!n*i))Ut}L3~EI zl$HmH56L#8txhU=fWcBQyhh6y83VI0R)gww!=G2Gcz!%QxSu2^$>GX3XK?K8!l* zE(>)b!6S$WyyR{Pr?BiaK=o0Vgcs+W_^6oIy{T>uZF!W%y@z&U5vyrRmTDPuiAS-s|@c$S) zhageHMNPJC+qP}nwsqRJZQIsqoVIP-wrx${dCR#G^R~6FimJ%^zfAn`J8y}7670r$8E3NNCyGG@{_8Frl%X6L3 zUmtL(MSZNWjkD56GvnX4ihG{2R*Wkz`o`0eqBfvIIoiq>TT{C1WN_tBvGcQ;kS&E2 zmEA|rdT(O8qUa+$_z$#}`BmS4!*I<1FNR}fU}OLPPhkJA8*eyu`%mrCq)=6cSqfx9 z-s!WMRBDdM9dTqymiSNwWq7dUAW1JkK8^=tkJ~Z)q!pUe^qDO{G2zbQ6fOO4y!MMD z?#BDX`QxR5WzU@Q8`=+jBk9rF<^BV*ti^n9#>ngW^7-)dZ?ZX^gIIdQ#tZp3w6B>i zQTDGV$RT3!{nFK|zh+yHx2HXQ>l+N;w8)M`zM2)&+pXhQA>$2RAQ8NUN0KiF!!OqD&NqSoY+O{ z+Q2DmqPdl-yTzj*oM`nNtGl_yI`n0o+h63YgQ0e;TmQ(L-Ef(UX4z;gx{&JfWXqQ2 zG!$ci-ch<4T+nLDWQF za|YUJD0-Tz1=dfVVVb=}APe`q<6Y|p);uYvIvs5mO{+lr9_hG zx)`^4qZ~;Q)xFdrMX(I zKzUrVh0aKOICsV`n-h8h3CI;u7MV6&k-m=l&+5xZ=Xp%!hFy?m0K3HBVKY!VMbx)a zi$KokCx~?k<+FW0g9|p7#&Bws62UpHd*PprUNunfYlA6|kFx?R{$%+;6y>2@W%1-< z6Z)*q_r|sF9Wb+DI)6k9eEz}u{Eyd{iQ|R#&&KJh%-Lzlsp}1RW+6~;BZO9(jC$~F ze$nK~4@KAywpfy3{=AhXDm$!m39P4Ap}v(lY?zX+$*r!54<* zDX>MEs)YPkxWC+@YXEZeh=&{GbB@D!r_TL3$cl$)Pp&ODqg?| zOGihre7&C)!i^_M;rY%+DVW?=3b$}LJ+RmfiIhH2f&{iG3IIaVO+ZDEM)nqnije-L zsydFhx(KaYLJb#RTcy@33^H->myS7h1In3Lt`GH>e5kMnABo$AoY3nsAfQ-+5Y z%=_t?pcuDXlm%{BxNomEvMwN7&@2nvFg==3@LcDrKs}f~xFGojVbLG4m#RnLx+l_} zVr6jDY&E)e77Z49O<*LKwcm;pfbqy(^ELaO|FF=iw415Z>ve|9Bc-x5ll@@XAOEY_ z4th!eH&_(0L*TO-peC?(LFJ~-Hv>3fbuDMN@xH>qP)|pr{WKyo*us3YZok(@H*M6d z0c2|9Zq(I5ttR8zE!7!9B}k_b2RuASA{PSy*VQW+MzN9X&8*8JEAmpceQ5s6ekaV3 zI}xVaVCRGWT-r_F?8Y3^Y}dPjjU$oBdh0wb`Z+z^MZq*)%TK zB;zyhBD?mfZ z_MVQXUEnJ}Yl5cZiXZWE#AOXNj;nI=odGAXQba$Hv74?ka5OWSSpU!6T6{$jz0{)$ z!??iT$XhY{I&>7G^0#}lSX(wycS2<2!__$^Sv=9$P%nqyc@efZ+#ii4y>@pT}crXkAf4Rk-A$R{h8BbuS%X=iulQ(>v0$o_wbRKkDbjfSd;_*mh}}m1@))@m^pBU4*F+i>q~?_?qKE*4gpfUg zv9`gZH|aIwmrq1}6#!F_`9mWlL1j|iVY|Zj1P1&Ft>S3GaB_mr?<<35^G~;$U_}z` z?TyBB*YB&Dcq=T|O4#pHUj>7`^1Jobq|HaMn#Bms{Q>jWF}}b6Zlj8L*590?B1+DNGpipvsY%W zPuQG)vFL+$8wM{4Tp5jW5dh=>%h#u(O%X8h!ioIaU;qt2nP1`pXt__~Zeq~K>#Spg z2nYedz)#|gAaSM483G4P=1)@>o8&=Q*tcs}oRH__wnv!*PN0vTPZgL)+kk?-gGS3j z++OE`0h8X(&0xL*`umjm&g!w}fa~|9026lESrPmQoKv{SoQ zVUqHDJScnopzJ5ew5C6yq9U!=3q*T@?%k=X^~;Iu&_F~(7A+4KmBIia z&R^^Pci+-dD+ckJHjBfSAmI-IDqvJ1E*y7!a>5x&R*|*zRyXol!wZicxA8S%uVIDa z%$CWKk51+k3Ui6kb(o$~#K9!YB#^&EMZ->n z#-pQ_@3lP}@t=IRH>;9jaM4^Hmd?$FxCQPxH}=vNakQ((N9v$B{e^6ds#GBHl~ewW zEAZ9Ly8@~bw8_Ut?#qW#i(TK4ibJx(a)G5mmg!0;n!%HBn;FQq>gW?}>cC=w7y~Rz zAR15|7t`yKZxZ|JF*UKT)ZPTz41T3|)L6+6(tKxB-Z48mv$!ChG}2-qAsE1`?L)XI z6rg1w5^>aFMStOFau6g9?mA$^Sew69boK=wej2Q{iGL~^o77ev%wKQNb4^I7dpd9< zofM6I0Y3y9_$Bz@yD4Dby>LMt)hoCK(Rm=Ekm0wWwladv1FrE8EF={DIVP%)M7P2} zeOzMhxdlg#c+4F3jq!b$M>bm=SSjNm3vbvavl(NafIH3~rV);e<6*^#%ckMfcuPi> zyNgdv`E^Gc&`=z#9(?_*w@Wg3cA)+=#s0ohw?0+!rj>y!(!GM9Yi0-qP=D~qLjq;a zHRr6y{R#Eu069INMRIB0%0l>}=cRsCw8(TZaXu66-cVJ?40D`y!|7fCyeY5Pcevil zRois7t|*J1QsZ^#|@-5yCHE1-Yg+BbZUv$q-%vr2L+ zLB~1{(oZ*qR9MbV-eNSLW*l3pJVlSOa&BjR|L|{RVy!HO%D0>7(6fkq0jfK<-MTkt zd=T%VR_SGp_UFmk+(O;+#9Z-GE~`M1)^-4w3#aQTKw2fZ%7AryYg>YsPb1?_aRfx; ztip+jXu-$Jwi%<=)hl>FZ&vO)#c*(r!IA;Rz!eA%gV!63o6Ok|X7#Qz_!sW2M!@kF z<;1cmvi-FXp!ltZ|ywzgd;8Mdqxw71HCM030*w((a zuedXSj4DE0H(cm* z6U^^66{N+KG@V0rov4Aqme&9VXZFL`EG=U?i&+g_Aydt+Om#qxC=-tvntCu9B8612 z%zhFtUOf063uo1EcNNo8*Guy>lz8!Pf`5`FlcSv2j#0Hfzj$Ic433U%!${Jru9!`q z;^A*(rRLqs1DDDuBNB`1s~N{*SnM~zv&~NeH^y#O=NVg)29aTsQ8};PJSlExp0I>n z8=(y4_b7UfpUC%)K1;8C?ig|THmJ!MOqE-_X}->h8>cEObIQIAF}#TN+B~)rWk`|C z`x8<@KHBYwG{bt$*hC~)d~D#{pU#LzN-^~~wt7b6+H<;S(BkjXPnB3{U25z5HD?F| z6T5#BK@|#U>9LE+_mXjxme``ZHx5;gYVDqYK{iSv zVE{r%qe}YnUFueQr4HcWpt1O;%i*B==;8SQm^1(-rtT`mCEG-xf*GJJ;+TQz=% zf-fbJbiAe&z6Y0SP7#YL=a%mcj}CQL)pIr9w$Poj*tMdNRr8a)v?%{&0~^A<3F`yj zduQ~ebf;m&(7JtLB;JVQn@)q3MYAf%XU@RI9ci{{tDME!Z^O*VaMIE`1o{R=Nc`@P zF(=ed(N+Bg)B##vphi~cZ?+P3k+TpT?fwQov#ARTp;jkW-u1(;`&_!!0Wv@`FD`yh z?2tG9ES+>|(xav4$3EU=f&Z#!Lg|s|NqFqNz#vDH3e@75zt=t?;q^^;_uEzHSZhs;T3s}D`pA_Hpp6JK`=(ms&Um2mQ=vCM*Kn+9^EbK8~a=IO~126U6{ z!P&P(`?&lsp%eCQ_xj8+Lg7;++Of%h5O>k_-+KsmKO^SOZ&7SPycf8o$9DW9P<9>nQt_f9=V+t(L7tjCjz`PL;vN)$th; z&GJZ@y14xIA0Hi)-vfEC=L3i@@goJJDg}%6Zla}sdAqgnBcfi=XXxb2)O!TU`Ok|? z(E=Z4Z_bY;vK@hK#A$WRenrkLucC|0M-4I`&^?w|~2o8icFVUV{x++_k zVX>ixq2enWm?l!tUb3S7Q1}dbRJ$m`hDGD|AC%*bY6LezI0Bl%^0GOu&Y}3i;T@_s z!_l8YRfmJ3QoE04b)pj^cqwb}94ZFOxghRm#3m5z2<$fiQb%uv?;B{Q$;j;1Ibtk6%89q4mPNCz;Q z1rZ8?n?fR0*N<$i^a-s0&g}@)9wq1nwL`BiOaJBtCO5$iCBeg5lMSc_^!N|-9@c&E zyx2Ad?g-{fUNn2a9bbtj{gXQuD?Fy9z~rhxXo_~N6?6T;3n{qM46JY2DAk)(3E&TU zUCsK^=pdZ&hTxO@JV&EZ(r%tPQ4IpC#O)QuD^%OVRH^IAk?h+Uyl`~TW=CCZ1>6K- z?X=(@N=1@&5o5b((%89bR%PeUpnMdo7_!|6q*Y{WKzsd3fzuY8w|=|!^v;Yzd^{T zUY$3am*9n=?2;k^PQac}aB8ES>vq5aMg}3*O93ed4Zl(LtQw*uh=PJ3*+nuDR0`za z1Huqdv?+{XZa)^S?byc%eP|mDnOu-^cjCJZl(UyYQCeu+KoWgmc~)~=+yI&7ksz@> za9OIWz@TqtJ%{NX2(()$ES?V$c+&=^b|!GK)&Yo&qWDe(g7u*j@oM?I@`Rqr)3pR* zwr-X9yiubwC4ayU2^K;`NdeYmpz~+}1(8y&bW^h6&#u@ZcPUoO-m+AjiDCd5yvX>k z4<`Jc%u}%doPM5|-gDSh#l1?QLu-#7ANQPClF|*pi+R>C7WZ0FdwljhPZ54LJCR@k z)mZ{R2(k99=+yzNS<#i@-#;YN5@xgf{w|dZ~ZOglwe+G!Q8R2WQ z0=PykqG#tB_?pk}vMfW7Wuu!Y8I8W-EsBG=@?BL`7EHz&6&uNz(@-C*Q;_pdxJ%m; zybuyye^z<-t`n%36dzopzgi5Lk&`%bUeiT~r>e_Y45FR5Srew-!nsE{=%EAY=i?7)h}qY`2>H! zEhrQQK?N);c?=84W(E^70>ei)bB`|md2-n#AMr%02hVj zL;p#ok9xwn{9J)UQcih*dGh(mU6+}4Gfmg6Sp{*!c{l3qfSkhd#ny#7XIeV-%a&QN zPcP{M1~fHx2|o6?vcT9-1GD?wKZ1vg9L40s-4+BAEX4TgKGMbA`#b0_TA=l~_oXF3=wNk2&tL6MO5>TL>=@ao%CdJ*`c)G%NI zn8Jz3p7%t@ND@#%B~p^>ma5i^kC@PHCG9W~T=r5iYJl2^gsOoOfH+fqf(Oe`Ga!xD zi#?QT6zZZGh{9uLU=Ag&;P8=Bz;B5eelFd1_N$u1xX=&K}^3&>y~Asm5nL|Y{1mxm^@(*pT;sg z;cpV!4xrlWT~!A|pCltbFscELF+_DHaZ$Tr3yBl-#ooN0bn13=JzU7^0Z~i){9L@Bbhv8<|6tiGgj~D7xZE%324T#oCmc5eA^9xS6JD z1^n%d6AlO`)Jv&S{WW7IP^jFu2xX^kxkQVOI6Q7V%$vSlnIyjMFDYLkejeBb6bLPbLGb6mv^#?>J-tau|8F zXeTgF)lSF`4PO|IJ->iqL`?W&9|5YQ*zhsb^ zw&WkJErlp3YSd9F7Htk&C=Z%;ReS~x{Ca)Zqu(tdm{iT3N0#0%&Aip)#x_^+>__d|?;>5mvly)vgr#EEJP*>0!^IRYl*{+)uoV1#K9w(hnI zL6pk<7qol%eWXTS0LWCBeXp?7&*9_7AN!8Zmp^t-enR|$jqB0o(wEEt1Q`s9W4B@X z`6!wf@?Q;E0H8EuhE1-E0fGS%&;dY<`Gw%1Ly&5fZuxDA?9@iW;RvI4ti&QaMc}}f zN+Jy5$gv&r{(yJ9UmgDWg#Dg7%=-@?UHi!egWg1QK%QG z%;lm$OU(O{*pQxt8P3oe7LQXqo>6&$fp`ZjP>?Y~D(kFvD00=tmo4~al41!Elf^8!Ae=-!%5FVZ@A##)`0~=fF%u^ zMa^;jG3{B`PIhe>=iK94igGOVHD&LwOEUFmcOVLYa7beQQ{09yZMS= zv!hufCs*Bh5eFl?bR55(XspGBvNlb;S<~^PO!Fn` z@!~MOI##cnPNy$YHO?<$x}Ewk&R0`Hd&)zrm`h>@af}-g#rRS&OCgJNE%dP;AF=RQK35 z>uNeh>T%FFM~G|P(Z0dGHm!ZV>qD) zMGi+|L|IYIjm;GglQK#H7@nX{nNTj73M?Mw^HzerikO~m9(Lp8A??x?6Xjz#Wn<#1 z%|f1Zvz1^~ioF6qJ%cg*tuX@wl_Qw<02kvrgFoa&L0N!YQ5l@w{cCtBRcTOBdhWaX zpiYe+khX7E85PqC<3<2x@o#VNYybXH*6JEWp}PN2c$?k^z}E7oe*!(-Y|R)1s5ci} z6K#@zq&hiB)$kTPK#6ep1m5)#51r%O_=npNB6DA_KocXlk3X$WGW;CP0-wny2>j{- zQ1xZ{TK1T*))C(MlRWmC@O_g}5@Y}nAggTP16KFypEI8Q^Yi10I(gCyrazc=o1X&R zU135N$0o%g;XkxwC4{=rK$$Ddo^PyjB?T~et-ly3(V-fQuoVWkf>mFq(daJt8S(R^+Gpd^FrT_y(qE&$}8Y^2Qszp9X|18V6sn_=nnOiJ0!5W}|qXWV$@p$`LV+em%>XqKxTNRD6Xn2qEJ_*8} zGNA?c-Jx?M3>>CNBUCN}d&b?DfxAKO2E|tZGPGh0X5c}iZW61G(s)3|XFWVTC~^Yf zuq&E|?*lR|KnkpYzksK-I1ky~T!-C<%7b&$ycl~^-U6N4s{{=a+PgjyHIwIN+E-D# zY4ccIZnSmddTy3$tsU&`OS>LtfimCnB45~y^7 zpY@sX|NAhGuxYOYBx))>Q8;k+^96JOvLL7wYgZZkFl)O2BZJ}$5d*}-(FI2HoLFP$ z#QXvveBQ)`c$~6Tb$vpI5z4JO=)J9#KR!u&-ADwzWISFxby3~-p}^CRIS%};a%`yR z>C-0SDdOk|;b@V}p^c*d;`mwFstlRh#XSrgajVV0GE>}!b-`Vg$cHMzq0^I1=V6e< z|6IKbU>N=YV({ps3kxfMgpu$Pn!zca5p#?7XKj2u>f0E>pv4uHrbFbTN;9eCh_xW8lYts!Wr>o#{ZYppWCWsU&u>|1f3 z+uh9+8q5mTb#nQrlQdCrsLrZ~$z_9xFjJ8O}ZB6VN=T)ho{C)1I0=gx%S;9zRiVn(+Y z7Y}f5WzF6)HDy2|95u~~Strr7?K9(2*yrgdVAMYPZhX!C<-@h3YlG*vrU5x;*G%+-3 zth!G_xbB$J=;z&a*!cfA7msxCy zbr4|6>5|xIJk(v}*j?y1DbL3@4Q--jZO;2BMZ+PDS|+W3aMZ>lTYQ#$&raO}#-tRR z)XxkFb4||zknDmgq*yYJ7FDEZ>FBd6Ytf^~qK=b1c{=-nnCLJ+T63nLK_61JvaQeg zsvQFuPuh^^-`l&}qk2lv%brTxD_VRSwpr3T(D_|`tdzs2WOKOc_Qwi~mdXN*4VeEz zXy|!ct^dK>)Hw&Lg1MB4-g##zPum)@)|2|zh-S{d73qt4eH+aUm|fjkQQHiCPQgZ7 zZ@V>As8?!vqd8RQmu8R!N*`})XUL?R*+@rUKTfCK3B_NGSzG4WOopukuHckQV-%u< zAN zopf>&Ss1*R2H>ErryfvPVRF-&f|SOvGZdkN{i4E1)8oX?^Qdmejd=mA$$SJp>r?#4 zBA%`?{q15AuwPmInPM(RxX_KvBKJUuGDytG6zIKkxPuhcl! z;JKv(d?Yq6BKpFp2NYU$<%qgh@t0s1TN1x%{~;Q4ps#)x<<9qd5Y>K`9eo3qg1iaM zrpr)4>O`F0^{dRqy0G|l)7n;vU94KrZOMffl$^%9G&0QpoTij2aSYmvD&c|r6P!3 z;y6E$sXcKll|6YUK3O$OU=lrWZt@B@59-Tpai@tB-@ye>Amj>xO(vdvD0aZCUa%NL z5r1G(0h!CrF_n3q&VFwbOO(B_W;?0jbHfzygR`BV1VP}*S zKmzlWQ~oIuj^#~C&l9HBewwWd!bS_}fYJiy^W zRQdUZThK~alk;1{gqrB1@7M^Oo+N6y@gS}=h>G0sXJGyq8GdJ6d?uz7Sr%OUR z$vTA|HLF;thD6|WsF==ILknvRt57wVI__LaAx<+V>|HPy4f;yom8I9ImEN&i)ZROn z)0|9R+{``Krff$B^vJn_8Oi)s=eiWjwZcn%tA3j@Z#C;7F^ zO-1P|E=%A>&8a&kLiK8q4Ym&0*%(!t;wpU**L;|r9^8KGwCYiwz^RN}V%lnI7dNx$ z&^SS;+1T@b7;918T3f}t`ZpSnH{*(O%zzyesE=`HM z6>l5ZFJGIuB0^|jR@*G`7nFm=gu1&S$vQv0yp;7U4^CX3Ntk@&m-z%5)r1T%-g_7x zjW;BH(?0wzudF2NSZZjbX&*B!Enu=B-)x~&9k7sHoJM2n?mEU9g)#&N(5Bi8j5 z)$o>2Xj!U)npB&k9r?uU$7CeU%`-a-Kqo)TbpI&@k~dX$fYcwA;Vkw!+})5By~}-) zC|G^aTf_-=ES}r?xBcj*2~+NsZ!HV~Hid}waSIhX52~Y0UDKRKLt5$&Za?}DBLc`6 zC#;>B9@z2Ct!jaycJMY5gMt58F7iv_4b^`VkLnVY?`;~w=}MF0Um4W{yaN6qHcX zkB1H0zC`$HJ(O7VOAa0~hR!Q+29>nr8}j~YC483iB`Ox>aF4Ctyrm;q>mZD12oxY* z5EsLnqj!2ViRUCKC=KV%MVizs`lS|smE(*%ZlSWJhE&l$42fvqID|)!GG5=%8O2x0 z3NW>!sTpD8M6zf21B%3bHuf;IX=N1GydnzG70Rh)DUGBulue2Pb(k!#$FV*bHjef!( z0X29jq=17(^LL)$Lm%Bg#b0}O^5MpItf4!&k;YrPx%04Vg<*WZHNf(dm`jVaGrW-tFv~V4N&>Ty+ zW@f(+zJI%#++SC$fy@(a4>)c+#QA&mY^b4NiY7D7aA=Dvo_W$zz<)o6XcCzn*(0#D z-+mGBW$I2IzHB~rQ8X#Ryhfjulh6|4JsjV1RIu6dz&15Bq&NxXO}=fN#yUKiFcnHP zEkbQ_di7bBWwx+<`KrRQ$kR2JnBkJo#q7#-ioW{%=gr8Gljx>FQAqv>gD#Ha_}X>} z!3Vvd_|4WGYOh6u=Hi(uWHIofJW~{oov`m-zcE7?%|&Zb|Ck*8=PvIlZ8g0R*mSIU zl+)oe7c#F{0l%kp1w9}5Xrq+=O#*C>k($gI#Z|DC>T;-xEjc>*wX?ho+JW1JUFi8i zU+D>by-t_CW#C&xwwbQAu8~~$mF_I0Hy=krM7}))zfRmA=h9 z?Y{r1v^F!9{~{m|_GpC!ML2}$xpwG97M=ixUq9|yrCpw9i{TDTynw&rm8E4GhN=a{QxldeNWdrA%iv8m&>;X_WQ;)&D+5p&Mw0+l=`; zL*!SxZrRFY0qIhQm4wDGHy9#D5LF~#^?Cf>A4e9W%3e#zFoPkiYe9M^sD>){&vW;u zhLw`Z{8ai;`UsMT@xvT5C6X>Z%v<)ma6qXR#D9>P7c&x+rL3cmszqeKNnUL_c1~iYIIp^=;8r^Gw_XRlg@RbVh17G$~pn6@lGIC;l)(yaqO75gCTONCK1h z?+aMOul-$s_Sc(omhuq_1>-yVGL$AQ$M~F7xz&DRY;HGF=fY_3y6h%peIQyGzY!I@ z(lO(3-+>RZLII2OpzaTOEV65*tBY+>@B zS!|IUUBA1=x}L+oGdCfk9&ihfsCKSC&4_pfr1Ux!^r)Xr+!N1%n_%}fB(WM@I z$R$|Z^O}pUC^|zblFsz9_(pmER{Ga%?CpX<0XZ~BYFe-h`%W!A2IHaFBNJzd&gH|V6Es>F|M{>{$)Nb$7 zaFN~{xkD$yOev|xHh*Ou);V2n;v(j#6f?z}s;a2j!qo=!_+7fCDhKLVt2@20CB z%o3>sWG{ntg6ZM9${!=1Bl|<(pZp(i&yR5=5*%x}Q#Qz*IxHP!U2MFDPj_=UEuDZg zHs#u{gJfGnnQUD#$uwTM#bFqO$hUSfkx&i|7+OQHzmgTUjFZ{}?L!&)c3PFIQd{2) zwPxpPp|xMf#YS^Ggs3Ny_f+8;3~3;D<;xJ#^q6B)6%jyK{imVyvY)p`KN`-EU#m1N zV&FMnsj-TBcnSQVw?q%s zTZWL^_^@yAA+$&2Mw74ANP_)H-|x2e>#oUOzL4gi)sGH&`s-VUG}LdKOY)lS^luc% z`a!|+KVJ3azm{rgx|vPe28Nvc2J1zEa}%*wfRv#Yp8Wi4_m_y4kP*^QU-uR{Qgpsx zfB==2feeruUw2ftjg_iMhK%lqGSO9sQVLw55O{Gu%TmBd4I|Y}{-gBB#~OsEE&KF} zDEcjV3TLo^O8vfc*PxvoMPEF)&|#w{EcKeG^ue-$8vP{&5+-SWhwISkSR~I#B|b;) zdZ0{zz(s?9K7m6W!<9QCb!Z8bFi_{JJ37Xrex*P<(^Ws6M>>j>)NDR2ktYpO4&pYB zWO`tb#wbWiI^f20qH-=!!3A4CT0Zneu7};M`CC0T)5vZ_Y*4$4RX50{_S|nAGps}_}{=F z+kXLrOspLLo8yGCP8(uL-51vz9ok%QEiw=fA%Na%iKasiMiY@G-YGd!f+RFVr4X1T ze^=gEKUSU_{fyv0oM&JF=HquvuB0k|U8rjdUylo0MK58NMDWV*^q<@voD#oKa2sF3 z_cybfv4z9e&BNhY@?@BR@}JD^4&E!?@Ao^BDuv7GKccy1JiTq4aP!(+UJqBEwol-k z&HT7j$y{@|FS--({d&9nt9;Q^jSHnK?M}2Ok+$MdXwn+8UkhCed$x4;ubUGsjU`*Q zTCI&I`^0RPhZ};oZ-GoVH6IyNM##l1>xM)KSyX;2-KS>3u-QeAls8^Q+Bar)yn3z4qxMc4O^O-$q^{A7M zdt^(|f=cHj?2u{T!Zc7aIIutY^wVtKT$O8+>_L>^ODX;h9$6y7UyhBadMMA3DHkzC zQ~741SY(xDl%qXaYtDg38OFFTS+hT#zayf>DT3B6rdj5wzKO4X=9 zp|>MPvm-T8X+oAB32Qk56TezvSgMd>n&-+O3!(j7ttC@Sl+slEOMRV+<^1RCVljNp z$Qpco9!y1t&@#FJE@SkN*zyrL1Jf5$5=CpYu$K|ZrNL?N>| z0=5{fY}0>`X6WFbO_eHZ7KKcrnHWu_tz7o2LpF07!%4QBo;i?k;*%kv1-TirB)JKV z46B@Qp!?>N3(M5=SM#~6v1Xct1eN@+3glWc0SCpp*uKqOi*6$IFsrmtNO1}h-g9e5 zkrY^_zEYixPqS?BQ<(D%tuxc($pImZ~jMaqNlKqL_%tbPvP*OyVN86^7 z)CLwd8gb&iT}U31Q-hW&J&utRKYfxdYF5^4zu1ryvn(V{uLf%L(q?vdQHa*VwDA7I zzr1T%+V49{^mqFe1{=c}TQ-2DhIMlAZ4iz(8s&nFyC}Q&&=cMNo;bqxRI3XM_vMi(LBv@S@=6 z#_F+Z89h{;uG)n4IUv*>Ts)A*D4L0@jGUBi9}s|jAIp>E2Z5sOS+ z8v^PZYf~Kpwh12&NL;p@aSNO3&_SbKu`ZmOD4-(4bo@?$k#U^k>T4kPl|GZcRIjI+ z<@~U0*g%7Bduttl1}@i9VAf^YXX>Z3V%<+bz9t367qc7TNaO5-z%P%(NCUok?5R@@ zHAi>S4cb<4u(9C>I?*0J6$AwwG$rD@d zpo5%*?EBje-Y2Dy}p47#3b97S6oQa@vK_ z?lx7BEsYEfxQ{QiW~l=J8o*tV2;+_sS^1@OQ`PGHAubNk zhY0!G2|~F>-Q)uT4U)7mzLARnP_y1kxm?DK-eo1Q2giPhw?2}`KVYzlL!F37>4_!q z0i3vR@r8_{ix{EK%5lDd*6vgVU{nv05>m&EFCqR7=3(y)deo9A23;1-27kYa9YS0q zybz%B){g;h`%kb4m^Lq>aEAEDq>}{s^AgP$TR)PHXiX* z=I;+dDs~0oBy)^wW78mAhU#+lb@^y3?6>GG-=&86>&akg{y0P!0ZA?NQ3lde0``G7 ze&<-JHbr(?z;8(ko|VT57U#-yW`CaCSzVufZ-@1s#oWFX;6xn>pPM!5;7d(0w>GqJ zaK;RUeu`+5#TS=@Y(L@EW?SJGyBF{fsYA@)_kLq#bnUm_pP_qRpk(ijvnPRS8^ap$ z`a7d<+`HEaAx$Gg*xB(2ECj{C)YFSnDi{)meH_p#dzB*ST^F=d0_DB4Eo0<8v()N} zsv(%T5k1N2QkxEoQE9mi5F(TwvVDB1(rjq@36*}$UJZ$p1a<}yjn(LYn@j^Ju{5abZX^*wIPdvOn7>cw;a#qH>~H@NLgMD zMux3`B)uu=TrJv*<7LhB*#XR@fwKSJ7f>@~Z#YDWxYkT{f?x28C^5Ms6VGTN91Hz9 zb5+tPcJo#WPe9q#aXQSk4oSL?09dtpJ6Vf#&@PBC@}8z803& z$!Q<} zO*^kF5WW4euV~hqDnCm|PE>wlNJkQ%uI&;KEuQn}JqH^eNX|vMILOx4)nlhc_z07T zy?5}evoL~iK;M=cUkZXi3RX=n%ai}0fCrsBE}r|V1Tb0BI8Sv|!0aU84tJnX>9#b_DqmyfI=bPkx(D`C8({8!P`- zGKYK#H7(};#0RUVH?aq^@x3j_qG_K6^5xIp9hA&3B=ou{B>4@iVNUts5h5~j{Afje zA=M+JL6`UaBaLgZ{*Z8$m6W(J9qNk|wrn;hGDlt2;~LX{`-=(% zz4c$wi1injfgsXwkdREcZ7C|7W6X{k)OhN&`*#I=TOycXIL0#UmTG#Sy3+e`!=1O; z=8fKN_8L)DD;PuP?Z&TRQ>s_%4fd%Gyuhj)oiNXnx_{-@+-Olt3Vo$D_(WlJ5;dwS zG%?BYe6B=W2!z`a4?va?^&lXWOY6k_>5G z+7kC53wDgFsp@7IdNPh&%T^WxpD6Ubg6QxooFd~Ej1SxjmV z2!c_*pA9xdjhF{kqCz0L>m~-2%dqkkr(2P`GO})0G&q`{FN~*7K6iPY`}Rt{Xy0hK zi@_kK-}`<>Z+X?u=!}!S1@&kY*Kqepmq&$$W#Zlx*Y1lpY$vWny_3O_2!q98qj=S_ zN!ta7le93(fwOcbFAot=r^ZPdqS`*?RbPKT5Zv%DAs~{5xlOw6MF8i|Rx}^;*d8B9 znxSX}Y!Pm)f82L%?$a*$K+;I4Fx$VIdI&q-R_b{dz{}c=yUR3LvPAOOl}`l;@v+6~ zOoBW1dH|+cZTXIeU&chG8t~BPWI66pEq+i+fpD-~k)iaao=E{SoPBoUThrqa*li8O zkdpVr7-*)y0;i9lDQqp3M-=^oN0vhZsBrf}fsnCGa|sUE5y;C<*^YJ4(HhjjR7nh= zWQIB7ueK5c7?r=n4S~%_oL>$f;ydX@bdGwsd)5v$pA`4kwgVTj&*4v;Oi7 zJ#G5(`tI3kEzlNQU(=?bn?w=7WF#rNe45i~v~n;+3X+Q!w-J4@$D&?n?Gr;34zF?s zJg9#8T3V`m$ptYI`+hJqusQ^%v8w)uj@5@=?nuGdf%;*~hE(P7j}C~cIiNbaoorpjG07sjOXUUCxKw?AgZf0ZgDoLwC!1Vj11?IZ*#@l`<5y z`p<2VdHlgsGwkbek96*uRd;q$;+C-lAZYHp;cy;Or@1huG8t4r_sCLtH=awk&SNNv z1ku0u>~M~+#~eqUZUU95A7_~)&fs`gHL>Rk<$dPDFAKXou%v(R--i6@WlKp#x0ZkZ4GwgG6~0e3j2LxgxCG*?azo?^!8qRQg8);wa<#cN+(euR)t*r zKVsjYgXe^UL?TcUR-hmTQMd-G=g?AOVC!WITdI9Pg@qT1w?Z4*3W{`x4s=TqvME>d zyG&#|d!D;_lPW1O-D;SyA@I?M3VmrCI`2=kY#AHk;^OY^?qfDtHZ^9F{B6M;km^3TWM(}BVz|n!pN`(+`|jT7RQsk^q`zU; zJI|)f?Z&(O7z#-m-t8CV+@{<&=h#QZH2`sE305)Y?Szg6;gaqk|6Dqp3IUWZy_LF= zFgq;_-#pi(~f=d60b|#_Qn>>IBPg$&yYo0Otp#j_o&fzV%K{Z zA1O#eDAfg;9=wpNapH`TE@V5hY+KJQc$l2KCgF<)KfcLU-4MhH(f)Yux=+M(1vgpU zteQJqulTu~)IhjgkV2jMzznIU=?w2m?W7#S4^M|%?f5>&5$SY0USc+}Wd6RDfjIo3 zmn5d9uL>dEM>Fws(lp`d=2?OT``aYOn`lxpaSy1}o;jK1+@tF0=zg62ku09hrl`8Bps=2Zj+v@zjz=md02=j*+mOA`!C`n822zJ! zBhr5%@dUSX3%nZCJMt{7&+2}4ZTddlRRd1#+7zNbJaJ?iT5Jqxt~Kbxi)*sPdRwC! ztXul$F$o~`J5%ykag@2h$5$Td?7*XZnGulmklb+}k$p+qeagNF@|e@fH7nY;g_gX} zw1vv!`KYJJzKo@0{ua%i8g*dxirA=xs}2_EF2V&tB_1{ zHXPLen~1cYj!)j+V}U~LAve38AS<8|Xz0_om(P)jm~j|&rFhYu_}i>tJN?}CuUl9E zXa^qtywf%)xc?g~pIOwU-MOL^yH10Rr#grI&f!tQ;$hglKj~55>s(x6mKhudi>ny4 zTy@!DV0LwRUl^O%;f|&Bmp_BrLRV}cVqAwK?;PO>j7M3 zSldx=dve!-nw;CTOL*S7G8W#17FksUCz`$qid!h`3T^vbKx1 z%E=BeZG7saD^VdP#!rsAx<4yV!YFmDp1PKz*hFzKy`xWQ1W-IZdEMLn z0g%24Q2O6OP4@o|HCfpIf9i6#w56Rlq}+bZiMnJ(s^w`&hL6Bo0q*sc7b6kqt#f_fz73g^9RM01*M4xBs4 z_R~9^KF%!dpMT}5SHvesW~zjHCNg;M`q^hSYeHq~CyUlwUMb4soFwDWrF3Ha+I%{g ze?!elmH#xCw9 z9B;NSg`=y_ev~rDF(j`mVJgb1nZn*Js(82$O+RlIN|B+eN?$EjP{%S)<=ALbu8ARa zEL0r61)oGU;LX*x`QZsqd<6ay`(45NA9>jNS^rcRex~wLh zSUJKAaneJz6}^Ny%IBY=!+HdEp5#3=dgC7+$A=yl^2xz-xBP>ci`}kbn zGVcy$EmE--jW!DMDVGp;(Kt*;by_<{7c$0lw4C$e55|{&xudXw;pH(6r&W zR>bEQbF1+Nxyv(0F1o&$b^8RROfaNX8b~y?$kxt+1Q89wRCTN`Dj-I?!7YqxjeCo@ ze@VH_L-oIXlxTf~A$)Y{$f77^4<)LlDvNmx$`f0u?b#Zfo^URVHBt9lhm#vyhXJ;n zKEJbu?1__Q-^y&HVA`OJf!1G^-sxBb!!B(O$5x)egRJ`JC-_WUg2S*cd)yc)6Jf%+(4 zNJb~-Q3**6dnu#1-fn3wW-;|7#ym4JZVS8;r8L;ovLKd-P8}(?g3h$<#U*D50U)~> zKJudzFBI2V5)IL0iFko!?Ez#MzEL=_;|M_Nmtxx`6`WYslKAx)*L1RFl#SLe-?t~c zWl@rNAYyiaV)WMdu5JVmI#udhWCt=2oDeaVK5U^K%qYB1T?lxAlftFiTtyp2dY=oF zadQBvd@+sX#%y}_i!rRnj-d3yd99;uzTPGZr>kp83@&7gl<{#+(RRf}&Fh}EK&1~l9SX*Nf}942iRux++^JZ*qhFpH{BC2 zR%zhs=(9mr|jM6sX zh{N_}-3+;Mvv|bQ&J;Nx+;Gs%TU8L3!Qo;ikKXlF;QKrWeZs3?R8+ARfK@l`|+J&<(Gpx>0}G$;rt@c(Vbb<)g!o889K2pXNjiKzwNo*DByLv`*H#!SI!kOEe43 zm#nw~aTPpd%VE#ak>!`y`&5s@bYf8Iu=#rYCl&&H*c)mHPgTMWduRhZ)6p+La#rA)~_L3Hjd#(5npLf%Yx6A3t_@4LEs zuY7YDr_bOXIT^gPogf||C$1b7m#JM?Nl7V({maw^_yMqx`}$6UGjO{dT--UpmReT3 zVpus;kC6|4Q=1qI`phs)Q{77waa=wIHC>R_TkiXG5uB$Vly7LCV!2j#e{zaVMcgHC zo5&$BtIbgqgT6tf($Ao2rsPPZBN}g%Y;*#q znDCx^*n{ zjUC>9rqNUL2J5RDSH9F%AiUCt6FP)wej6%2+1yEANR;UmBUUEt=cF% z2=!Ppb|_MHaNR&-(2p*g*Cz+CBkwI357{zrT$8FuVF*y^k0%L2LbIOL8Y8l1$(B7#y{DXUQ@*mXqF$)AqvF3(P>%C?OWDMgq&4K)T zB$J%h%Ye!&2QrMtgN7W)r~_n3bPGID5&;lmx@jyD8|P&BS6C#rC=-(j&PVul2iy&R zT$BN=K^QSuL_s0Q(pM|C7zgqb0@0yG+%uqCC|Uj?K#0IJ^ZX%zDKm&T7HcE1T;uY} z_I$Mmcf@>j*Jhzt*H%p;cw`xKSwI3@mCP$^TZ|vk>;Wk+9#r;yp)&KMa$E9Ogv!Z* zZA_vex7hd^5gt~z9Hq*cHTcZAf@3?ntw+dN`IfN9D%URHKgT?ju^NnyUMxW1astd^ zS=nk&4uHC?;OgnhfudbrDN*^)m8st3CcR~lCZRHg8FzRs>vf;H_7C*AT~Wif6#J!E zqk8eh`k_*oqSA4EU{YYW5-PI!-SyYq(CqwfyDegh|0%TxLgP_mN%$th6BgY+j5uQZ zeW6G(h74oeL`7#zF&&z?&dubhJo;8A%GS!*Q{4v22fK-5Hb1vEk5B;1`B0oFjpT2x zeH<`1RP12H2>gUxLywCn^qYZ2hKzjV*ohG=ASH%1*mF{$7)}tf_YzgNo4Ps{p8_t^cYU>QTvZ`agyprsz$?^x zby5;(@j;HkCA;tnX4J)W{hjqFEO;r z+;5luX;Y`>60$22*%KR)G<`vK6#py^1Pk1MVbGRXH)dmcwGCy8XpN{;ryq%~va18= zFzOsh7&B;zYMq$nB;`bZqzdYK0}xSjoHNOhByPWm(h!x((bPow?aza#IFu)=>2PT( z=rm5}eF9RcR5L7Vi5j`+m~alnj`4e$Cyu$AKCxm*XJFcFi(3e>C{R)Z_3Myy@$8Pg z*Mo&D7mj52-j|z+W+Z=nW<26=mV73g2W$F1&!3V3jIbwbp$$sc961H5 zD&8Cw_J&vH7<2RQ+CS9Qpl@n-<$5*ds-Ciqr+ilfAjtA+pUh~28 zafIaeeNDPR;_BR?9Ckjsq5}{IadOWN9fnNQhF%iSDaAA z(f8SIDKDP(ARg$*i=TvKR}FP7pb?prM{5KUzSq-RAH(=z1hXf9!ZDTre`uf9NoKLd zF_Y<7;Kk;E{c0e$lgfRK<8C*aSmwFT47_(nUa!WDN<7yHj?4(RSv}xo#url#O1M)J zTRtn)MC_|wUfF6LR03EDVm74}w3?yP7}>e8*SA@)>T9m?1kLN=d52kE)>Q^5rtq{` zo)7EehnZsIS^6)<)WizwpPw4420`$0P=JV;Gf#CRwu-mEInk`6_nt!aB9`Z4=cUsk z-V!Cx>#p|BNHY5|t8jmrvmh6+{+ZJ&SAVk>A?JQ6qR4b9RF#ylWkL2B%6rSz2y{!< z5W-_|eK9JJ1?r@q)>79u=fqi2j0^Whk=%TiKo`Wzy8s(zi*u*#9|g#951E-vg;VD@ z^Ygm4iVq^JELz9CfiLcJ1Xw4v)kt<{4X$jGquYr;UhJYb@unf~TL{1m0C`G~Vpa<9 zju0&dIaPd}!ISKKXItJZ^D&^nlEzO+2c3cr!;;YO@=x=rzvh1BI{yY@6iREWn=rs>}O*u2fM7h1#RTsHd-JS@1; zz9ut}2GkOCC*ymiuhdkSy)ZYTkgQOskc`#abADOgiga)FxM}tw*FqZ-BDP*0O~~wp zfnePfTtss;e~*y7KeGGTu$Fs4$n$3{ir1N{{HSmtijzfCdT5UL?nKEFM|#8(+fvkXRfVPf!&!)HmFj!jqh zy4QE~C9Tziu75KUIQzEr;|9LiM+^CMv>J11pS%C)!EMT}@4_F1F*4bh!_nhr!7+y_CQOo+;p@T_@W$S7NsM`mj2xvOk-{2tCa6 ze5;S-qfWOuMcU5{ZBTd6zj0Ez4azDkBPcq`-~MS#lSPSQ%&z=cr0spn#!+-7d}QJE z{1tOTxkdBEHiic4qrT@qZHL3}AddaTY_6_Aa{NV0hsIA>HeGW!4bBom2?3-a0nGT} zdhrK!E*M^2ui@X%H3HJ%h!x4pGfH6%nGC&CdaFgiM2_dLPRcnh=lIQDwNLeVpZw`a zkbE;R<$v5So}8XO*BZ2MSB|;TmLJTMo3bWeWX@A=E^|B*NPME=&bf!`lH-^~Hf+?MO-kGeX1d+J-DrX;6xijW{% z!H4sK=c}VblKT@OsYB9~tz!?ry0)}7wl??pZhmUwkr~~Hk}Pk=6lr|T;>&5H7Ams)X_;-;r)&ON<-&CkNgjl|rS?V!GemoEXZS*3lxHDdvI6eU8xVJSh@AQT0$#f@ieI{{xKFkL zjBAaqxz@dgGl}0t495->CAoNtJH}Bzy@%Und=nD?ezTAkCTMK8stOXmYmicBs(!|n z<@B0180}l>wAtk>ggNDQK|~8zR~Bp+Keb^O)-||;cHCm4my!WrWPcJKCOYVlOFv+5 z4(gpP7i#=+iw*;79lJLz|q3Kk2+2O}kbU9(puuoWjOSqtlCIZ;uXI>#)EL zV`H3vrj@d!0==1MaHK*^(`Xh|cI;7-Y5JX2`zl>)s#ode;E8kWnx2Gz+fw9Qv7+hm z2f2B|zicHh%kk2J3dsV*o0G4E6>!pnq$q8YBeT@akM^PAMn5Z4pe01Bp|iB1T*4Ya zIr3GKWpZ6xN7HANN{b-;V=NMsj?wykAS`kg65y+xg1?W{Sle3C=4`1u|M2AD|s$I|e&3P?iD=gwn z2+!k1FmJCj4!gx^R2U;b*ea_9fCJ^f5*r5#k<2&`bLO(!Rsbi}<_+9;^N<74XGr{8 z4b~4!D`Yh_{4;KH61`6ptb=7oM7#utD@hXCZT?Ep;Z6CN`qW30yFvUAW`~+qmOO)o z%a7ZIJ{y9(BRKp#V1t0`knTR(Ol58r$}rxUipJ^%Dtl9*XH=1VBD-~Ej`fuzG=r9a zZwjm7`UIPln#fz2lHgHwl6XjNI`)7P%yUT4k)~X1;<(Y9Vx$>Bw1m)Sfcld)u3u!< zx^qHME+dcQLqTl{CtWv*y_8%r7r^I;Zy|Dveac0kR#`C309qj`w;LU^8iyX$EsoU|=Ncu2JBZlM#l+Qkp2&V4%!@n% zsi@Uz=JnYv!F@ecWNaRD61@%Gw9{aye(numdHwczSSvQq)pMpe)A7ykU4~SvDTie} z9n2pnpr#K5;X5P&_geQ6w3XLC8z{Kq1y$a`y+eC^6wPgJq~=4NJnFq=AQI#_-A z#JPTj@3UvvmKb6TFylR}9zX72UMajKe?4c(8(K>?>j)JvxXv|L&BJz|}M zD4MMR*EkbUnvNT?`w<~{NDbA~HiEEzL3;<#(R8-`#;^t}lKf4V@+H>vrzboH2_!(G zMm+F55SXUU3&u(46^S8FD7N^=ak|DUcrY>AWh0Q;=-6RSi;b7Rz~#1^exO~Ae}dSe z75kv@Ty3DRx6Ft+cU5il$Fat#k z(%UctF$Hs@!|v_=L-L}~I5h!e>oVx-=E#nbK$!(d# zFn}LUL$v6q5zt1w5V8Qy&Onz`@TsrU>Z1%+%>-@^;90J4iaJD10r~I`0daI7(G~=N ztD6j5MruZl>)`~9#OtjC6iTYPw?WtIGtslvF2wzW7QB@W3)+?3Us)scuk{h^QnX(y z%I&m@dGzv$t7g=(&Oz9ule5zObO)r&gPZkqhjV|v!_{-+FOWGZNwh$u1yTWcaqn-D zgtpK}g5WMfhu&_d?+QTvVV@>c(x{o_F(M@i^yE(fLVSy<5dcNspG}C>51uELhn4=JjvLN-;9IYnR$zKM3D;dz6AA2Fq8^FScnkd7(+{RB-f&}!POM%9qzd{jtRegn zCz`RJQ*Rd07y_T`u9`Vs!Q6ao>D%ZrV*l)|n);~|AL$Mi4sMH@oOUqsvv0R7^v5~k zV<#&_MzP10TdduBE3p9V2O%Wt^Sk<}smJ!tE83=)cX1fe!jB~`+V#6`Dk%x0%;hXM zas?_>At99EHt8tQ55A4i0FX=zJOQ{y4kZzFZ-SoDXS{v5%)&-x&y3xBAXFaeUKhLI zk0WD6FP%(E1kj*M-29OIejC4@3TCIAKM|BGx~K43b%Ye$`)MWeDF{J2dx1W#O; z1y)(sjK~~OxQ296j|$KHr9E@?4v)!m@A2!gffchUyMPPet77exEv9wY=HwCRxCjs`tlG!9j1L*Uw_Qb06=lB6 zxmr1$FO{5kfpe$Or-*TXYCQ76kfEBpR#T3;?=uJ96r9!|SHh30OzE#v0NY0~?!k(p z?nE6L3yWiLFW!SU3prsrWAX!!<(<_~Z$WW8BKiZ0>koy%^i^+3JkKSG5zrRr zx-JZ?SB)_Hv=p!{ZJs}TH0(@GjCd6ZjF;Q9;V)2X)j7o&4uKu#`1S$kolf}))b!`U z1x=o8y9p``a+gMMS@TW3)AH3EQXcZ89%}>@^ZWmiVr1y@(iwq8hqs`6csYI^;|$M7 zKk{en$;$x}fR=rd);b$YuS%SfhHU7&u*&el=v_{FEy+s%x`5}g2mPC8;B4Y92 zE=q)(eNFuI^@4Z%JPgcr;!27lD8*2DZ_XuxU(wZrw+~lB8vmdZQ{KG+Ny=Ab`9}Pb zJ0&=#WW?a0bxWa#3!Z`$hMSHNI$UPvtzSq@+xJB%#9qY_lz8o{$tQ!UTBFW6yMGU$ zVQShZJQJKU$a-jw5;97)Gp||BZsKTu$?85=gv3yR{J<~Fzpc5&P*UW22En#h7|Hzn zjjiW^32s^Rs~FLb_}jtd)N^J1aK9m{UfDQB38bWc>&g=Ijk)PsS^(f#roL4ZVtbp> zkU>Rp!9cc!Ukemvdc7^gChUiAGW!CC8E!S}M9O`4x;%e7Xm_$*XL1)FTa!|B#DPVt zP_#M@pxh#f|N$hx{1t`3rgUdt^~hPb&;}5SZP_( zi!77(O4}#Z4P_KF+Z?S&#hb19W?iQoV4&mAA$m-lGWKBPJ3`=nd^taFZhIxQCTPZaO zD$eQ%soZI=ws)WD{OxnEF@Z=dkyrQZ+TIa~a{74eaEaSWOyq3XElP6lKF=Lh+-d(Ig;~+(irb3uo96pz3BMWwiM{Co3 zk?EgH>99KR;m_P_Cj?HliulULT~008D1Wg1sGRbMMAiY4;}y3v|B1Ck#{G<`0W1Qr z$YAO&aI`CHV3!P=AGZjqEMFVwoocV506-&Y<5iq)jD>gAD2HRYYR$7bNFI{&({5o8 zgS;D&%RsqhwJ*(1S7Flwo$Rm-!~eIWz$}WEGBCkLUzsS5eDOo{Web?$((bchRi!U&n}E4o0}TVnLRwVQ$C8;I34*vKzN= zHQs)t3K#ufQ?`Y-e-?_wEi>|(nH&9r9|t5SC|-+6#^3%ONWz92svR3Fv$el0m5h|^ z?k0JxF%hjkx0S*MU888O6vO1JLydFYmRqeiBK$tUv2)~8@@}4-JrDeJ+JO6FprA5rAo(lL&ev)TzZOOeSwaWl;Tbc$f8y(WB5oAeguD%H z*9RF>D9}suAyrA^%>WpRCXpIDONW949U5ZDSz~p9Y_(WpY2HY&6>0CtQhpd|Bps

    QN|g()BRr_$0&x<9b}ub1d}q&rC2J1z?hO|6 z#GC;GWAXqweV<=ByQHvQ0v@gC4%o7FD2k<4!Fn9H zdq3AtYbZ+6dX+{_{K<>kM1lw1g2GvnPrIqaMOYyfs{M*spu^wPHeO#Kv@{}^&f#Fg zf5FhZw&YPpm=f6|IW-kLJ|W`$aF_14fsj_6?1`AIE0LujEEC$Ch4egdqRR=3%0OrV z_xv4aLhePUTX~SOpe=f@OD~mn4MSmkUbVh-zP-vz$@z9YpDN$<+)?FxC$U2D|x?BQ76G1 za_Z3WsHg=}iYk{8_ZWG$QfBa!B=L%jSuqgUvJ4W(c?s^Es4vvdcf0IE3)XM&f>;1G z|8^2}3kx~)=GQW|5kqDI=OZugRGMc}<5$J!mzfWar?3K?L3F8xvVn0 zO!4I5l52f~C3>{N{Pz^*pGm-#X!43{I@*Rl|FLa%0cGvO+6}uR@^nDLl=n~@sz+t8 zhG1Ng?bFG(V3Mpvp-|4Dx6$9T(MNxaDx<3G}6%U1IQe#oNtQ3s^>!lHZ>P?c7EWgCb(_(Cj zk++#!9bM;7PA?O?sdStAOTkl4Z*kEXc{vgh(C0)tku7ZmAK6gj)8y$K4}I*_a`C%^ zo_6-0@VX>{$>EKO!PUmCb4sDMkik0-^ESPs!*w^I#cdT`(DmwlP>D&;ygs7k{m@0bK7Z*`~p_kY7|5_Yr2R zFc84}VqcsEcQ9MI{P9oU8~)xY#!iByzhl*s3$jl*>bV9excMmleUAFt7kMnyJvz$v zcY}j<(963KMJ;XYbcIXYKMpH~&d9AB`00q6uN9K$?^?C*Q(qT5gy4y@WmpN*B>D%J zFgb$#+4`^JW6LB!U~mU88NY|zsZl8kPB6Sa1JV?NTAHGt&^_yLO!oSWgBtK3a8xx* z*Z(cn!7wN_zFIJKLn=(pQ2uwpq+1igN*|(KSB+iewT+UOXoLg z@clFfE%J#?wBt$wupzc}Ox*Cx$n_|4eDv{pp%mKN7H zOAe{Kl4m7NMbr!@BrlJfL2|vqsZNRu*Cw4!m~djtLe+f>RrLmrBp0D#{>yKLGN#Uw zg^S=GJOeWJSD1?&T!R)#28(3UW%y%o&Gfgyv*)qP-SaW#cn(CG;q-0hdCUFI66SZ& z&-}l4KN!(WGU?+wopJ^sIfl};5_R1>^G3?SJw+ZXjYo2s_A3qLwl4V^GDZap`&@2? zkUO6h-910*O%&*oJUnF`Q4cO8^_=G~(32LwfX*MI4$bK(Sys51&KVKL4(;V7GcIyy zAfPt%gQIN4z3OCXOv9t+_qV6sn>@54C$EvH9cE99S_V+cfy7 z(v%I^rq*D^S*aV1rHKtqPI&lP-QQpMbX4p?ILGa&>E}s0*ZhT6<~WeliOvmT*#PvF z%daOVZF(pIQ1XAF(t_CgJ01t&hS>>_PV11g{EZUybm3_vH@83&c=OMwJZ2=2`s)ze z;O?m6!luaE{w%~V)3Rj8U);4+ab>p4;RDkESt?k0b|E-v8GFueGL(LFp0xaMznZar za63T2gRvQi!+CP<#DRG#z3}wKf1k)#+<;2ygJuJDD zbVNZc`xd#~t2_F%2$blY`J7}pdgShgYWrBd8ez%|GO*}HO)2kb52(}K*ExKiTqJ3q zMbWj4`@M&Hyn8A~kc6iP4=^dWVIz{(TGRe^jFdwc%(&=Qu93L?Ej)3*{%}ht+SaVh z`N)Qbg5s8mp`)|#yu|1%Xy85mZ4bx!=h*#$BT;$;hs)m%r$lj%#lZksvDDf5WAZ@R z`o;k=a>%KxJME~2+YAlUX&p)Wo$3|RHrrzi{z)PpfZH8dwtd+L>{YL%5I_+{e$@s| z)yWmwmcy7fQ*^OusLm>?Q&%}AbN$}WNP+x8r&L6Z2ARKWIipylXf4}M-+vY&{V5#v z{+1B1di5R*YNwBTul5YmD=tFk|E@l$Zyi~7lpEH0`K{u@ATJ8b{+kdXahbq7;YRqR zF!iC08|U0sS2lIc2}G6hT533OnezbPHqXP&Z}D@FcGGbc$(1zV%XSX0K z^h3P1qqV>~$rtsdZ4tWO?>-#;mbYg5_T=u)wXwxpr7|{H+!A+0GB<=Tcwq!OOo$Ew zA_xq*gYpZtK$k~`FWos|*$lu?bnfsq#S>RHNI%14-L&ko#CMSNy~HNSbs@}nh`hW? z+SRAIV_A$BWK<|P`(yQVXY{KwT89Xg zMzsg^sW0B_)1YXUHtuhkn@jg9;(~Ubxu)nykwaU}tx@^^a(e^mJ_+T}bIQHQ_t7-= z$wWYnx?{L0TMmsKCHPM@`n3yN*8}91MKGsO{)H7TCmi{pF519@B-TYdHvhJwpnUc+ zI^)`|WVBPI)8*oV{tzwJqZYkx9BuGR8(WDxQsA3FENLpFE=-v{ecs+aoXm`w7LYw; z63x|+)6@#i7f>6l^B^ba*S-k2-CLW2tNL3p*s}+ za_Rez_EA&_SqXVSb$hN9S%froMd|;pG^q~9BCGs#mS@@dMb?Ls3uk(;CmMdl_iqI9 zo0H|Qyd^|HZlGsrfmbFt%i889jxHx?3#e2TSP%_BLXUK@T?qx)Hl~i@7;NF!1nien zfe5iV9LXSJ)iyNy1MGz8J=mFfF=}HNdrMmYfittE;(@q_dIfta<`Egu_yME3q!1ww zvpD48pX12F={9P*yj)+6l&H=1GxL|j$Wss+I+C}W`UML4*k`FZ=nC~~6c3>u=*!MX>9lWHc%)5F;Oh=aYBeuo17tLXAF zEt?`2@65SA<3~r%jku>_|7QMN%xCL49xbC=Ibn}1u7Lhto`$RI{wL*tyahZB_w0n* zEYXv(J3WL4Oo!hqdMMBI8y|N#9V~Hdh3q#cI!nyFSY~!ljQ3NgRzGpowOh+vp3_Yk zI)}boMBJtj1DVOf#De{EkkiG8Kj(hNPB(AcnP-z^yt|K>2{PSD=PSHI+_8~=GQLze z=4PVfg@=aBRp=r-BuPYbf+tNR&+`oX;H3i#AEr_o!yMif=*uY z2C{y~LG(s2?rlg_hO!c`aUEG*`dVcHSj5)7EMp!(M$~J~Liitju+e4Eb!>Tme>(i; zUV=N-4pv=O9%~%YmM%YuFnu_fMkA75o%sh!U{{kt5BfL4{I?@5F--K7Y)-p2V+%VP z|A|tu6{x-JqPcZ!M1`U3IBkgsLH={ZIC*4k|72KkcmR3E%5c!$flEWxgsu*<=zwoI zZFE4uXda?KFzIL)N~As17gD+Qlj+}bxAS#}F5Qno6B$Ti^k#xLtoFcsokBJg4yc9S zB2khKiSpVp+|hHqwzjRj2_0{2X-cX1QjMMvU2EfN@#ytSCBqc%IRcL{He{_FBBBUi zRsGlXaGb5W3ltuG5(9;{ZT4zqPaAu%oMF@IyMiEo&&{wxU$tvn_l~P(YA-P^C5xT@5l+pQy#j8UmJGN`FJ{^RCS#( z)h1s1gd^JhD zH5fog0sJY-M$PCXl!bJmhz!sy*%`&p6&$!=nSgtk4Z!%ly$~4Z(bBnF%YeMTwG_gu zS1jmt(H=8o4V3v-|XaIsN>FYAv-)L&tR~=+qNWee{06tEbpbAfdgq?zoj|M za+I^+kdW9c-ih~imGe@qC71%yzxvZU-_~I;X3-cggHIb1hm6nqTE=U4C)wIw^8jsJ z{_~YA9x|&Xb~A|zng=lcL@fbaQ{n9Ps|FF(~_ z=-j>z=D^__jI$g~+OoASWu?z4Is$&G1hny(Pu&}FVeZ3=Fq8joccPg=*lPq&9_Gqo z>||Q_?x2wKjEGE{^gWe#;|CoYjf4C(!{4CmdK1VuJE%&GWqIi-bo)XReWk9YFza1$ z))xf(8080zS^iz)&M$x%K*I+gAi;~!v2WG(qBJ?qC@(Y!+HCD(FM_Q;tCDzwxyw5b z^#PX`lh>Do!N_shv?xbU*G79Q4~p|*jqD`RnN0F_Ag_Z2_cPX8>Uy`&$Im4dF#Q=>p&!3BK)=~EahC%8%DQB+i8)$-S6zunZb(p&(#!hV2{?W;Ji*A`zqq` zrZ&#{T2JUu>k_0WjO2g4Vw0dZer^$sGS&>IP9{UAmYLBE<3WY`uA%y_sN|N)OG}t8 zXvPkAX^M6h8x03&DOyo*G^eE&g55)lIWWIC;0k2fr8U_wN)oTNtyfVdzo&ERj&Kh$ z!@Zk0Y5zI0>Y}drlHcTdY>1J&O*Itk=n{p>!F>EY5!DIiy#Vj}gKJ;^48SE|zDJ(g z<=kn;tUDIXMQeYrh+M1848Gdu6QkP4Fdiy6FqADq$NBy4wln2ye$;`;JE-_xw)Op^ ziv?TP<_!1vPO;{`snW9lK6`jT8-rh2H0UyzR-=##5&`yy{#S&A<`D#&{z54kuK;0f zi;(cOVO1XNL^TW;8NxNO4JwK#Y+hq6yEkx@zG_4DRhP;!C>N~ZDtkd4^N(k8sQh9_ z16{<^&GF$1ca|D-Uz@eH%cw;&7r2Znmg|WzAmdVVSCAKM^eUFz7g`e_5Cp9&a%fuAU2PlMeLz`Yyo>r!vxeFUAxNe8;-VGZl-ZL&mdTwC^>iGE!p< z;~l3_7!H*-6|!4=3gXF@Pr+H*Ke+0rQ}RI0yrY4-s@1qC)?SyH`HEig z0=C*Fl#9cP$^;j>Yt*s$Cm{)D#Pq}2*{&@lQ$mIp}I@zBKY;?^dRIXYfG;JU9$Q;!@7sYX-m`Ml zDn=OiH>AT{!%--=`4AOIQDv8a9eNKS>TSKD=A`eoLCMpnDv;oP#7+o&TK@ao371hQn( zcUGE}IKohiEdo5eYH8S7>zu~id0$JK7!oKCo^S`ts#na|nw9T@kHvd`7RyU)o=9$q z>26cMZD&w83r4JYDtXW5X2)-s3R9TtC|HMQC4uZhcXkFYkrU_>tw^)Wgcp@9o~14h z@=l+zA`9WEh&>YMiHF1S0O$9Y2lAQ2@t3UC)od{URg$olRl#Z6{sWE-5#!AD?}7L0 z4gLkJHeElrR{G>8n$yk=kPyriKj%gM`{#M?%-^JJ0d7PEz?&L6KgP>`fO_!!lpvIhgKw+!I|)*OMS|9%>HJ_7|+c2@n9 zwh#e#TVC9dRp6P7hmBj@j5dr*LTdkHRTtqoK10Tl1X2@%vSYVki@n+evMfD+rj%bE0_!jjyNR;aDsDa#=i8r7%ZGeyOke$smL&ctnL*uA(DioS3ZTk?dI-&tXB1bVRiR-1 z8ph9qJQbQ|@ZYYsg~}}YrC85NPPBE0(~ApU2AA6>yT~c^;k=fcMXD3hhp1dAc|s2A z2%L}_5AaCV-T>kO#>HlaZ!yEcF%{GHS#C1ms8Ldg%>H~?%oQY$!fIa<^uLB8W-%HP z2~19c-Pp$3xYJb&b zJF@4BTp0m%dGaOyJ-#mU&%-QG95^*r2V~~gVt^kBmPop}RcscOnd`u+uUuC6t3tn} z&;U$5Z9bhjgP-L@{??$4T?Z4}3mKWX&bvGlW^lq?`x-`K=3Ed)0@WJ0pG-kNfmS%7 z`q*()D)i6-wkRmv9f9MI_jvr@;bl$(HZyDy)!-AZ`c9#uzA0h-8ZS-}fc1K!56-_x zACc?pm`wrM+^d^@ne!kM%mp-#?sa|U$W_GDE_<`CC9>We)WL|4I5kgDNHdHMk+^%; zTqq#H88eWhYzZ_Xcrct_BoDZxo~+UTOyG)4k%jBjFLZI${6s%d7P#2V6S-57^6f=_A5PieWyu3BD4DvGR&|`~{LZczv*Yn#=Z-U>##5Q*&ayoomge{0C@FY-m1lcX zUh^+Sg$}2A1&*>gv)*S+bqRp@Km$-sQ!%VV4iQR$ASlsdVUV^32jU#IHMqy^_kK>K z3KJBIe{0AqAFr=xI*Mnw;vm5gAD)|-%v&HUEJ`3WgqN<(&v4hq&*FG`OC1!@C`5E? zMOTYuZJ9ISclZlTZS&&_3lel_azl1#0o09u0T-Nd*xOYzjVYzFKdw@C6{l+9q>$v9 zz>e;pzQ`a*amxib0wkzA7YPC@LwX!y3WnmxX~XMg=^%;8GvH!TLGB!A7QxBGESWqE znW1PFWGL|Ac4Vm$&)-1YB~-Dvxp$;s(|0BGsb2G7upGn%yt0?mh%&2(jYs(4d`LX!p(MUZZPHz4n6ZZWt-wk_lQ~-ZaSX1%6-8bmfFM{WFH|%;i>LDMYM^wIZ*JKvyfR7o(InD5*%Wp%^`fmeTivV-H z@Q9J$*w>3``(C?Q6dvv|>df!#DhS50#{0Ii_HY=uv=HW%jjm1MT@rG*;E1hY%b{QL z8Q{F-%icS3?vV(SbL@uugwe+?5jvngDlRXoL96gC{5_ZC54PdX`ciFXna4vo3INq= z&bnd<-quVt5muzHG0Nd_3nW;#D{({4{LRD$jwYtP!@M7vR$TH`;UjVy6EgH?{3i=} zKjLkhO*grEK#$K?k(M}Ea+1evvmr3=A0LRt6r(c9b!4an`Cq9+H{nM+cf?q1FFe#U zo+LRbs%vbbSSuz#9EfUUC0k&bB)Cna_hF)D+Mu&7a{Me0gFvDD@+%1@W*xR_u6`m= zB3QLo)LZuMwI%W?gl3GtFS*$K#cfkEEdGFIqp*U*y;Wtpu7-wimSP{jwMEIED|o)e za2N-P6cZszZ`@w+I+#0EOUO?fnN$#FS}cxWI;6A?lzZzu=*>=|uQoT6f&fRya8?sL zW@G`6)Gpw=!Ko;`q$5s#$P2=DOF`sp7;LEp z4S1J)-r(ym@>$IJU5Q6_(IH}v7(k>i7?7Jb&k=uQifWTt3m)kC&W}XBiznI;5$xF8 z?naI%^9(Li`8$j|d=}BPbSgWzapRu=u<9$)D|avUt=Dg6+QckF&&ZYBhPQJ#O|#l6 zMP1FaKL5d2dZ^bjLv$+LQ7P%H2XdS&^dCmfuTBXl&bK0PJ$h){YyU`skl0t+W5JtJd zIuB8+Qn?ZPAW45JprVC~P-g)Dg$`0p?BF(!wo(>JNpD056?BIezy`fGa48flyS5{; ziFQtTLq8*LES%c~$`P- zfEE@?WFUBWhZj2faP4F0S+3L9!B02D7z}o4Y=X&eFNw)xO|f)@6sxepD7p|zHZhyk zpBe(SsK#%d5_RYS{F1F({&#N93gqGXCpbU9Ejebf)eDp?;Km|rOPNl~2RhUz@f(Y(X023FngC}zdLgNIn_ z>-?tS&QOP@n5yeG%57tClG`zzlOi_&N(lZTB;&Ms?3k(YuY4hr-!&)}8w_J^s-;^c&-5BLB*leP|y@D0PU zm@xiQ4P-GP9G#YDenROvOL^e2}!{U zEeyq^g6viZS!c3M?9pwGZ}QZdN+Yn} zfUbstd-Vu8$yY&Bg*$D22zLu{mwX^&KdbAj!v!?jE>07VQ^cV4GoQab_#ce<(%}RD zm4Pw;M+U~l{Qn?btF$H@HpS|<9A}b@Z0t1f`$7~{OMhN{5&4g`<^YLkqkozWKx;_Q z9fWM{8TAjm_a^tix|^*V9j%nCurU6*K27_ki^j2FM*Nro-Drz(TJJ*QUfQxi zy{+v9I`FO!t0E!)hI?3N@@WzFRDUZsR6cV;nU>#2<#F&(k}5zkC80;Rvx`$F`*1yN zRYPaiiNSsu?M>ezfo-YJenN7?aH?#n%yIw1?WY$}0 z^p67e%!Wnwb^*e=vwT99J%NB2?JOGeR7XLzp;~PsCSgp)Y#rCxX=Cx8D~Tad4)L9k zf*reOVQgrd@R>>3{sz0_+XB%xV>E;(<>iO2bS#6FH0a>H@46SaE;RUQhM5&`)B}`) zggD+5Kh5hqU(Susm;mLDj}(1P({+MrJQa)+>=@@>L68{hgr zYkOlMRT<2zM{>U=1No})qm4?p%s>)6EN~sSpsg74FinllE+~~H2E95K0)&bTT84-m1?11bS zIy@P63I}0=*U+V)%0!O}fg0OCLZLRwkH|UM)wRcf= zdx)C9`cl*0dFlkQcD!I#rB1Mj+&`a5X+GWX2F?sok zy1#Bg)cXP`gZ=zr?wxc~tH0>vfuE7E47cOibs(81(dVI>peEqqm(LD+s?RK-Nh{#Df_P74x=wRPwxU$8M}3e<)Q&Q7@IXI1X>hR;x5rG* zD6SAT$M?s0iFnV5EgXe0e<%9()(MZz zOgVy}8>N{Dus z2AM4y)2(Yx%B&d+rYWME!6+)0_vy;!Zryv072sW=(%RdxJ)LVCXYTR8nel`(vz;cs zOPvtZe#O!}W!V=y6YxP#vl7eqCRjOJ}7vQN1vrxXz6H^A+6vTmMRY%HG3dV>dHGwvER;^Z*j zvuYNJRxGpwcD)+9+dwUpU8e0R)%-lzYmiw?SfR3qH(b#22!3r{-5 zrO3AUT;JW}TDT;kDZ>J7B?naMa{e-r}pDLX(=KFlY)I1e4s(T4wF?#&1pjO~2hnQ_u#2@gFc3 zS*-8r=tK!T>_t>a^MA3_OhRWU!m0<~s1(n_gc&d$>ZM?rq`u7dY^iac+fwH7Q}Qee z0mmB2U-N%UA33UbAytwYzzJqR=nT;w)=M^QQ2oRsTJ(Esa70SU=+CUAW(ihM^-=-c z+V3ly&99=+1>EY(%9lPVikkaLRSno-8xrW#O2e?CjKQ)3tsw)`qR|at*&S{Jgo%ty zQjK;}Oc8{yjwGwO&Exz*S0 zxqlRKp(Bnw78;3rhkBpVp}3}4OUnJ(B1JGMMJqA)QWO3~DW*{h!MX2l z9Cx9f1AQ+M*)C+WR8as-kABP=<@<|b9^FM~Nz!A_A@&;L?Q3-BWD{hypuYITh!^~2 zh8CJ-_$xDjvLt*r$SF42v&%V+U4ESLz2TLm7OJ>J#ACKc;3v0nMRsff7(yL4`I<@KB%1k{Wkju5X!P^H1B#yak_5_)-l$lLjs@+CS+K6ynUhWA z2h+)`DLD%?;C4xtIyw-7BE*#q#I&Vwz+(0y=j03AG!RKYdLPw9WE_+}Z)#Qj5_tn%Mwx%K(S6H0l6vsAa3h%mzFZe@W z!=;hw`%X2?A5;#H4(0eTLuwnL**{@N^-^W17dzp}z-`L)xDJv)nAsam6vHa>lB*n< zvJ-VhNLy>F1=b^Vdd;52Hu43Q4+twpeF0MiJ$&-+>iytef`K;%qT~%JBxZT?;H&vNFYTb`6qy)|*{YGIqcufT!_hIuv4(iC!V<`yYY%6PXp$bJqzaND9Y7#Vy|iXY2pTE75UuJDUGi;Olgw1wne zaqe(lvcqAo*e3x0y*NtT-@$50MsnVGX{@Ft-u9P3?0Yy$a!!WgwD`ZQ)t|*+V(GNu zUU5U-T#}a8>dUa)*rzhg(d-^MJr!qmJ=<~*$z>|_RA%a?(}P=C9gfv`4~-W7+N5h^ z6po?LYB`m!`YMhkeG+>?giYdUoDsGqm#!-cy09|iKv_TK0Q<0uvSZQ<(M`Ye=q{hX z^IaRfV|jtGcYvlg1D|{u7nASgZd71iuW;wG^E;OTxd%;JCLR)Xp#%gxkxMe%@V)*n^6UKlKY!k!T zqlZczu}Eh)h&mZzksaNR*$aa59~~H5X;pHBJG-RD#03q&HqJrr?WFf#J)4lEa#}Ab z<8^l^KE9Q2CPL8`@;N-Fjb%BB_#(RfGp51yn18nAwT;b^T<56R&dkhg{Hx;Ai|{$3 z^!vbBsqj@8f-`THV-n68C)z)hGVL!AR#|+yD6$!z2KO6Vi z8<%!k);`tk$!VW$WV40~;LtCMSjiTqW(aj9w(1C(zDZP-)4O>HQLgF8SP_c);d zbc@6hHSH}_M+}nY34OQluMxBPW_7oHynn0c5pp}{{?0@QV7jdAM(IoC?&}v*f*Q=7 zLGdnoP_SM{__GwbaYyq-!LU{hdC+J6sWR~uv=?omxztIna0vWd;NT4TiU#Uqu~b}q zb5tq@F(z(CgvLp?t6i<8=%YXyv|{ZAJMGbhr$*1U?O(2+SqByPGK*m)zMxP!iM|%N zJMP{W0*0P4u=t8e1qsHyi3B@t5O%=gqws?;+`madP4-4L4&R1BcC-RBbCr*g*31j= zt+-jN#{qHS*8bH}OHKWqf4IJ#uaN8zx z<8}wo{OCF8u=6Kr8aQ@=fQ2O6R-UpCl7T^71z&PqN|}Ljj$R;y+`H$y>YfX%yD`ss zuZHs?ktoD3LWP0I7h48Fb6H&oNi6;|g|T)SX`!o@Q-~U_-_G4dx1xe;!@E2ouc}Op zFUUvlTZM^QE>{v3u_3wvc~)v_P2d1VDi*7h z);JI;Da@E~Yk+?^xQC6GJ^)J6XQmrzz$o}J`#w)4cEF{ZD?spAKt@`tZ+I?%pjf8tzXt!$3Sxi67mBB=r?+7o`)nZ&Zz%##? z9;3}_2agUf*?5IWFCjymVYj^9E5q7?zE!E@ci2~Ewt+&F8=565@hfa5vlYj3w?mk~ zA(t2Bzd_-U^~Rk6u|LN`+B-HVzU*ZDkQ9}d!|38+=o^2eU|?^`^!^YMhm9T8Rpr!K zEUo#cnwcd>rKYv#i*1x$zOC^>gz~S%992`U1^P*W#vq4lzkm5n;9gJ$i)`C{uQyp_ zO8EmI15$%c{q>ri@F4zCu&=ke@+?PqV`yz68mPAgl^*F?qKwy&*B}6Q*4kQjlv22K%-PwLJRs@geUyv9&7IBR_$Z{h9uR;7`-DfDP05D(MFx$en{fc*iB+abX*$@E=u*g>JwXUQ~^1a zCO%yI0DEWo{;m0I!;QjJd;eW(a3W#hd8RC#4EuEA4R`A*MATo$%)CZkEp8K*X=lxOVeepIu>-lL7(blFMq&5Nn%WwDm;Tx?n*o~ zf?Mb5w{tEGy(Y-z<9pSOe8NTrgJpC?3O610i5^C?r4=s0lZz7%El$}XP!K_#$$nCp zvBauGx$ll*6vu_>PsE6BCyGQDTLj*}NzlkgEP}=fE=4Gnqh&_HN>3D2g>rzNH*uOo z2K%j_GadPO^TOUiGQd%6dM(OqAfIkAo@*np zYRI1OS4$&Fw&Juy)ZjxysWn`DxyJp*WAod1&fhu&o~vQ5N09eP^8MbmX8>oY$BJm9 z>ddX)#L*d3w)6>yJ&Oa@oc#F`ZP3|{)R*7CmLsDvltu|X{qtbIDMyMeHQAvYqqz|( zGPw42Gc8?EyrwC!amzGmBS3$6vA*mmSa$Hsr^Yk$#4c6vL{N-C*}F0*RKuombVb+d(seb8B@F*N z3*NolSkiv!7)l<#eM@K|)#+uh0rYWp_~J$frd7Vmea}b$+`4$l>kvMD9Zf;s8ZCdI zEi4VVul)vk@MOp^BY}N4+$?`}M-u7zmKGOq8eOZ&y%&B%Aej*BCW88>Yy3cwbY1cD zQ?T?kxl#G&*S6wDjG;$cXSLdvq$r(OG1^J;<3lKpM1;y-=6Y|;nZ02DqPKAYocaXz zzu-X?&23q5)Zb^)AroKu-B~wzoZC&*n5(o74Vlm4Z%*;-$6mX_Qp!{7eCj(fQD^3D zl~P}eFT^-N-?8hy`wekfsFY9h2H00|GvZDq9WJaAg?~}2iV3hIWT(2PHvO`RH@=HV zGThI6S#xd4x*|oJGp7EzeEw{|S#}OTv9xK@XD>OpAqufQRx*@40Fb%G{dKy%mR0H& z8^64rTfB_Pu^r&<=ChK#QfIH`BpbYTViI|tKUu$8pT7f=;&J~g6l3|1P>h|0<^L{* zEouH(S|#GU&Z{?uA#8t?xdw}_Nw$r29a~>k()`)bo`#?#2GG_h3ZS`LZ!Ml%zIACD zszjwBhuJ3rlOd#ZNK|MrX^eW^ll?UfdFJ5R@yR65od%>w12drg}7lxjLx}Cna4?_+ z47hS<(+?}pLdQ#1%23vff_qQz!9LPjtO5#evinK*7=gD@0hpsB!DCqXgnyg$u8yIQ znzdHl3fz^W9?gYDq9fNC)@vL$FcE=>nb-{f0=ldsb=3CR;})z$yP{orhO&$NwQC8* zSQ&n{G%FiUG#SLDF*i-b5QN;6C&z5I8)}wR=hPn2WsKMY0v}gQJ`C_ZS2(;imTtj_ zjy0B6Ojh=w-O4#H(BDuH*xV{p5kuLtT-LEhmbc1Vh4uD2?EcKdU~c}-J#=Z_Wm0&r zt+!S6{d+R^P(0HIr=z-f8rAnW7_NUwt1hut^vrLEew_9PdO`yV-bS3%X;_s3BC^x+ z7^m+7%<`XRZ``64`wAX&xb!m0cckEoek}qf_>cKw%8YMVv`nA>1(@UN7|;8V_zJ`A zurOe3js_t%hmNdszRzy&0m2sZFr1u!R>Pk*u2MZ^W8;C0Foy;T;<%!`a=;-=>4KoZ zO+8rcGznN2grv`NjB7@a`4#Z>66NXcMd9Fs;zj1D51!#=tkZfAXaiZ542DQR@@Ra~ zhU#k_eNu!n3#JfQl$u$8Pm4Vdr#cWb{qveOg|A61vesTk;27}N%SjniM~q?_A_uOn zxbOPa92l|o+SnM@ukgV1V-lFJ79}h!^WX(s91(Z~PygA({k;kpPk8fqcft9Y%R~5P zu1kuDnvyb;W1~%o((gB8MLSX9AC_a+sEjaPnT(BJN&zdpkzzO6FY^$*F~x#W_{<8kp8 z2%xI{pu3CwiiGYL&Pj$-rN%QoQ-Tvp{3Hfr?gH0uo&uv<`8YYx`Tk=zq_LA|R3#<%Dsi zUBb)mnLaefO_op&YMXRfSir5meX$0Bw%k7f8uO{H_!bJzZR=mv7fkGEKI{ECO1S_Q(9kml9Up5jes0l3y?_Rz>n;Aud`O4j zKCl_?IVzcw;Q-4j4Yd^8SEwcuEe``FgIqtHa&FktMyb;lCH)MPC}WByB3nPbRxJv_ zfrWP=&cl|-ytA$Fw7)y6_FZ;C#;`pymWjbWG_@BVWvMdfr%)e#!<56+p6z8a4-N0^n zYJLIBL$Zi?67c~6Kd043u?s2_K%edS*_j8HE4GS+4UpKS^CDR(_PhZch{vkCMk6vr zZ|};$D{|~9^^F1ui#7#5GVus2XxRahgKFxGaQ18c?=EYSr#ZQY0paxRLeAe@QHKyF z$#!lTA(s8O=Ch{K?qa)HfYBBr6R;V0QuhQcEcYx3E-G9q(!9QM`mIswWdRjZe)XCR z<1gEUTR)`wD`E@LA^%{%GCz^1 zL_Xn3JXu1|W@5?~Oas?VmDvhF{Z}Ab1WW}}7D@L9ldHs7ELO(U@#!oxLqfnc{Dr-b zBT&I*Znw?TorX_o*Tvau`H2#>@ki^8i7s+8<+2(kM`lR9Ex1EUIF8&fJEZ@J9C^t9SBz)P|Ao zJ0(Dx%N8WL<)};qiLN)isM4izE&>~?+f}umOn3?m1ABj}x%ByIMBW!u8~yGon@2EbEZTV6R;eY z>Qu*&_Gk5+i=np$A$=k<24!j(ThQQ&w+R>p;18IJbbpKKsJT#Zs?PTUy>uPk`JN$) z4!{`-ut`y~Z$4ZSmcQx8u8JEscyLO_?9JEqasmu~fM=ft6XsDMp>L}w1EoJ)t|exm zSXvOY0v4T1n3P*~Q4}rE;ocgR)QmZPrrj4y6}0)@ST(#MVm4u73XuhzesQqn6W_KU z@^gLV%8+T`U*`Fa-7j_ytd4}nR?s(RchlT-`6D$lI;dijS>(|X^m1l=KX9-H}K z=LP_Th@R>KK6)){FGrCd>RsGm!Yh7zCv~+H89y5c*3T$)rN8G>@cFB(G|Lki{N07* z!lP_f7_3F&%7oJr=_2v3Z<^8xz8QxiQ=+-DEyq>MvC9H4;G19Az^5=|++Qp~?0bh^ zqXg@_;FTVIYOlWcf~cS{Ve$^lx=L^$WwojA_-5yPghN125=9T7540cQk`&=2fY4|< z9UyX7VfzHwnRe6XqPLtYvZ*}O>3Ydb;uM;m8f{{mGR<{uc6yaSEwq>UAe*%X79u!@ zv2q-CDxPr&HmTakbBEXhS0H2t2 zvN zDiWoQdaHxY*hi6WBnParQ;h{3DQGMLSL&pc#rIDfHU_DLDI4iAc_sTi`rKHbBW;Q; zhY`tA*9!&19uG$kfw)j1eCdPNuxUI_VgYd+C&KbZmKgl-yrm8VP-^nTjEP#*ec_G7 z%ca?pzP}l~f3J_pB?54^k!V=NAVMmW&dxm}AhYfjfjY-$7X#?(3lB0k{rZ3q z(;--*FsV;{jm@qDM8M~oy&PSTGstr@e6PlzdP2k^3}#$Q8A{425&)z7q@Mi&r^(N% z&*Tg+zDeg>G$+Jvur5Dm7-x;oJMLT}-Uf?Z8hQYJ8;$U0tQDG8A^PPI4 z+8cCe^MHi6T~yP7 z7NIag*w8-i{r`l_g!UJ~a|E0$NQHHH5)cKfpE+Q5?kDE0Kjj606PFIJ9*1p@QB+OFTT|(x2gz|&3fn|zoP$0a8P`N5w zK!+De#{f^-Q=?slw*!ly%#l0GLwFA7t^Q0fS!m=eqLL)`IhU}1k6uqT$}#G|k~{+3jq=+z(WBd?e{X1SzqY#HhY~}KignWr+K>?JeEW*aVIcYMvwV)RLj=f^x8NY;G z;w|1mEh6aX3V_9husntJvvHv~a%9}z%f*imE`Mw&bTM1?Up!vt`gDJmMHbaXTGST6 zJ7$CiU+r_$Q6+YhYkor`l#_P41xfl7a+sHE_zA|ZMz(+`2g^e5+X`I8vS*ZcAhgK_ z2enuE6u?KP+8SOW+%+=RjI%*^uE1aUAS#IUN71iavv1qRWS~Qg6H;{`oLw_?i*V-o z5czA491-pWyWRf7zJd09|9W(HXg3lBkrh$IAv5%o5(0ZHnztFhcEKU;4B~}4Mk?J5 z#i&R=!?I!(hH3K0mFsohR+X4Yo87}@&gvii?Ks23ndc~wOSX3W(!12YPfM>cdQe#tr}r$-r( za`Vq*g}e#u^-lKHI;O@K4yFOn6J>GF`HfXpRmUOjC0h)2e;wRizLs+N&6tbPdaGMV zKAjbk9?koWs!$-kI|sy0C)c$CZMm0cQ5s1qSbz*T+ANs-(Vvu*l_RRO7>>RTDr`0P7q&|3U=!U@aq z@>5h8VUJ1SEu(l_$UrT(dlKK-ih0Ri#u?+RtWl9b;G-ZcSDs2eG;D~1OmdkoF?Pd& z-D}7NURT}O36dEB9E2ZqAG%V#%JuY*7tK_POAOwO5Gp)CN%093SYrs|QzYLh4J`Rf zEjW@1~vWp6`;y+s^~?8WAT-8_-E$9Ad3fMIYUt|^ZnDbYk_pH zTzLCqd+ER$VO=#PUFPXG55zC8(Y|fdkeHx_YTJ{EOd{Db3#<#Ml782S2m8OUL#q^4 z_0n!NT^aMehZLPMN`^~~)(SZb8^f)nDNj0Aw~@&tDGNSrpQP1pr4tpelp)b^N_f&q zn*sT$fd}2Vu3yvH3AxF{#bL96+32C@&_7n~DTom>2zM9Wj&0}rE9oqVgrajT^4!(~ z{j{NSf^wl9&1Gq+%3I-ItG+EjEZ02S++Kf$l{K3g)Ko8Y|M{Clz-YWPBc)B8KBk=J zC0mEe?rs&=VB|!qVhemv{l2Scx8c;dkR;L`Jw)VkU8jEbqA?1QH(@%E^P&mdnEM~b z-myoNKwH;s+qP}nwry*+ZQHhO+njCNw(agS*UtTLlDm?#ztkTXl~j#qy!gE+V>*;4 zJVZ|sdow299w=kt`MqSTxHC2>zqMz^N?yyE;bo8W9iPxwkSUgd#q2j#06JOts<`o% z)AD&*+e`GjP8k-?>jx$`7puCHTJ`Z;G(e)8*LMpOH5(1-cI zKp##vw*NQu>Cu*T+Tuuz>A9#~)Vo+Iheeh32JRhiw65wh{X=8pHoAd~trDT;&>2~6 z)^OF}xAIfkbHl$jWJU&nRwZItT1L)BR;WuPfFNPg`^psXZj)&4op|_r(D%5k$k)OD zaUH(6DEVdHo7vsY>G`5J5i#41Xy*8MPSHI;K4U%z&DikQT zKU4pMeyljl%B;_$=gxmo*#mQGH?1(k4r1@WOkTsrLJC=Pc5%MXA;cV$ z=ncMC#p^d{FjOw5-o`H^PB+ZuUgT57SL1FF!*Bz}F<3jrnP#(J+r{zMzzaNnPS0hA z@;TeCRKb#lb~(*&+uBrfM3bC2F%}-Cq^}pja#j{ICU9Zx=TmIvRrAV}_HqL_m9j#o z@85(Q9Qj0nkHi{9b-!C>eH(6>9rClGN5;0Oz^T3dt;rjXuk)|dtyLGF zEX(^BpzYXKE}2v`^6xlif9Wokw&7xqnXa29w)N#gLT@!E<;n(4?uMG0ZVk*0Jpq zFz~8nn9tGb^(<+cG5&hA_4gTw$u*`Omv+|O+tr7LGrM#cZq{lJ=&`t5U9x_9NZ~hS zvR>C^i3(GXg&HP#H06WyDe8zZPIeU@C1Vnmey9%{p=fB+BE2-PBzBU2axCbk;NU)ZPBz@ zEKZjKi__m$HR3&UuTc?B_usaE} z^%C=B%GJY7Q;d)8vCJ38=CD>hqaDz|GoK1rDZud3T+FvlB6MFHVAyG$t!1XMp>GjNSMwsrkVJ5F@WALw`_)}NGkanWHdfpvJp=k;G28Xi@6Q@U4r!7;(`sGjQRr;A zQg2@q5-Z^!{L8``-a%;Kf(yqUU}c{*&e1OVrZ69z=r)!FHbJqk#zs}uc^ zte%lz1yD4AmucXf^n*eGcg1)b3WLV+=-old53D8Xk1%5bHO0Wd0G1mb3$=nS77b9r zW3Q6&r*M-TH)TjKa_7i=PiuHX04+Qe0RJ_d4nqwmqD;^OL=3=Z7(u`}mOznx;4@pL zIyYsqvWiXDLUenn=*)}foUHq0z6bw3q_6@|LjFHHYc;LfO=$=p^<)3 zBBo1R|KHsUue-dAd1%Ksj|7qi?tp+Yauiq)7fcA_`CI_Nus3Swo54{<7xFWPfyI(R3f7o^wORRG#Hq}0q#>gHa2JPINbc#Ks+!Dk?r zll8%hxvM`7ux15+BV{3IY(KwTo|h9$pZtr87(KS&H}t6(nW{Z$}7p6 z!mT0(r7W;+nkaV}%xKVs#_%bGS8`8!w!}eWWfiIoJs-dZ4>%lE(WCo{&4PO@2Cg4* zqpoa(L%^KivueJ`p_{|xPD_8TX}_k@bvlxzxb-U(-)hPL3&mzY_N&D~XD~e&`ub@p z4ZCkTfu`_|EJBq^{z6{k9X92;`3LUs6Y97)JO?~e1E=6XbdtFp_;;P4yfV*es_?MI zgd7Ezl_w$jVvkHwAOgp&pVOEIjsx2rPyh4{taK%@Gw>GgHM!~GWu(Awg;jy@QOmQn$d^|@6 z8%8ajB&(vWPhHvnf;>F+=Ds+)VfA(PtsRqRqNz=h-6Dr|!CU*Q7uM2|n(%jXb>rL< z6f%Z~P4CWmtObZ7bQaj8OX5mOSrEDXK|1w>Q~lzg$mP-kw@kY_A=Wf3Diti}fUXZ) zDAyXoT;SofRvg#^zNf5hg|-iCD{_Z+9PY-=>GigSBGcrPq#9=6dcQ>|t(&KaK+2HT z_M8}F_Qs;bVv7+^0P3n4O~N2F-u#)AAKa}#zBpuY1 zm??NXQBQHmgJUw#xN?`~whGvS)C40Km@vN42ny9rtpH5>bX}LPa7D5Tk~bRqlAWkP z3hIn`M3%s;oFZ?-W{mBtcsNhalUT;)7c$`yWIhf0#I1NV#5Xlb5{cmRVWljtumMaL zBsib!P9-P2Z;s8NXip_^A%`=0&XQ3;V*MdZ_+}GUVO+)~KVFd_fK!q{9xRT*p~mdd z#KlY?p~g-{aFgRl6QQ(T`zMIP>`3p6|B|flCd

    >qQGCh$CiXwz~|U|XfA=yqgq z31Q4`7;V->c9zC!C;&6X4CRX>HhB~^M03S({^50Z=O=C41F+m2ICN1jrU#Z(N}ekN z)ZHnuba%yVl!N9+|K0S#hDQ%zW;N!@uTF?)fe-K0M0b)|ytf-}sp{R+Y&Kn005(a1 zU8bK0(`BKp)|yn=AE<=zE!C*oEJa80t5KsoUOh|Q8z-T0j-?yr8p|WZB8Z_IWQIZX z&VwNuXu~w?%36a+>Y**y6AB72Cjt#=Ac0d`dfha=*WeWqpm9h)6~MRn*QUlrrG8iP zD$;>+enlfqgx+{))1}F>DVy=I%CV8FuZK9|L`^X;8HD$i>+AGzsa}t>2VoRmAKbKr z-0Oqs*{Ci^X@J#Cc?#+fPb`7`vAl;|x9PZ3JsFql5WUk{9Ex zB$Y2Q>q27c>icdR%}cnb<_fQSfC4_;coGpqq8)0%?JMWMyLyT|iOfamwbH|Or zdn?^X$qQh5nY^_tcNb)gQqm=L7k>b03*x1$=5~zKqSC+NRXRfcR)s%719gvy z125NGZ>hg~CfP$8@wZvF8YG>TwUrC%gv02#x!7SFb9< z57BRc9^qGs?SSBn5<9q{W{uhaCX2*gZxXhv=X&)IxI)zzTUYQ}Rn(Ix4aQRf0}0#| ziZ{hw&v{sagbZVc#4w8#6cWw;hqFH%f)Y{wBMhV?JwRkX=0?c}00_)ropW+qH)U&9 zSLwI9_kpCd0vjeMx(9f7bHwx>3E)==adk(HnTDdUIdl8t+-txblwF{10BSf?ZjhXC zDOl$-5)zne0BUJ8UjhH&p+$Sx2`n4~p3 z5p%sTE!3Fh?o5w0{Ed;6c&tMt@~fAoqNsI7>IEE2ffBnraVlt2GmJ~(WPEaYe@^3I z-N^6TK`=n}NKhMs#Y`57k;d@3`i_U7kUEX)%;W(I7k*UO0?kqej z*vAqrc|`TSm@y7{HccDLUN`rNS8Gz@lk9}P=p&90*o?N(r+dn9zVVNT5ClEV1v^JG>d@S?{7_q1Ed9@_AMYw zV*BdWb~(Ci=0(h%!*qm&Spz51U1Uihhr!o5wwZSt016P&=o_GNe~X(fNh>Mq~h z^RX7n&7h4?rIZbr*Y>IOt6w*Um*pi!u2_s34i<|`B@pq`h;Deh-vyqXjAKvIeTjk4 zrLIf1qx;wKqH^QlNN}_Bv|wyb?z7#vvf}D0NFl@d_;47C_+@`PB5gXtSWC}=*lH)W zc_ID@{=~z_hnF^#x|T9@wm{;?kT&E~!!_zw1`AP)jdS6cgT|y?38f&%Egar~+)Y;D zTKQL-$_8Tff(IbOQ5XeauwNv5CO*GQ9UIPVQZa7@43iYcI5i{%0HL@C-Ivr9cjnN| za{LQt_cBBzj?m+}>oXome9(!$41p%8V*I?jH_*2@K=?6YLFVcVd6T?_g93zKWe~A1 z8nZ}KI!^GUSGMp9k5(?*x*ks{hM_X+%YF~~6}rVJp$rHGz@ zfF^CacGYmw;E0_0eM=~INj;PxDbAdp%+ZkJ)Z+>XpuH!A#u@;}u=Ik^zPK_n%$h|G z8;uJNPc678cQ<2-oEkRC<63YuDrg~D;YS1G{Wp2~(obt0{}1zD**nK5)gQxUd}Za7fu*pB=8;52&1 zyc`*Npz%RdtOXZinj!+K)gws|nE4k@Q*=PYAzfIp6p23PQtAv8N8V()H{y^s$8<^I6IkBN*A<1&nCv@V0zWfYs$ie=`n8-+#!IlbGvfp?&*TtMuWi&tP!pVvQ(^4GJU>@ z18z;sg7M`lcXJFKkfQ6DM?aZyQU)*Z`}i`TQtqYD(vwG`l2ufJS^IHG-zdQUvH~2D z7({-5CNLo6qQ?NN+-+b&vU1HRkbt_Gqu=+Fm6;Oegn@O_uCZ;mhaMu;HDv*=o#h)< zv@fMbHSQSnKiV6q{dHYidv(vdf&Q`iH0mpLgtX-5vUEugwoY1b+`nNVlh`MAwBG_w zUeRBZP_Hr?Of-Mlqq?m!UF>@^ZORW)tFzd!o3PQgAs7_D)bj5M-P~0);@)kg^c^8+ zY#2kcMKUm2cLUV!yIrhvWX3utxQtdej(}(9n)fNMro4dL^eX?JjKT!; zGN!iX&K3mBO#h2?dt(y=Ln8wd6BCygSSCB}Ix^nsEgP zDOx$&f049`_LH;Jig(gevZ??bGVmfCr>AJisl{lNZzpBNXvd8IR&FdN1_(w5My5uP z@3Xy6%f+7W(dO&TQ)XZOm+bej1$n^v6x?l1N4lN9jIPNp&dP(+dfE%VMK@>2(LA`9 z>>DuAjxqOE;cxlQIH3Kc8{um&kK{J-Mrv1iZPHow&bzS3>|2jTGCGsb9F`m(?p0)4 zPCB&5cb@D&rT2F{8AC&p{Qv+0{0{H`Z9J_1*LeQFX|Qmx{x8ziM*hQ7#S(iSXBOSr z+&3>^S%W`>R6XQSM3s|8rOcdYxraIZxBodyvGeCFtSbDt`3>NZy|#zU%v1r-&?wgb zh+9(#2?0N6JTpC-nw|d>Z`=i;{094-R`opXncWWIm9>HI=e=L-tsn0!LQ}%)955x| z>h$hD^1cPQMA!@Aq)(A5T?S1q<8k+J`1pEqbbNtwWeDPrBmZFr{-#6sy2`L4c+MA1 zwXjh-*X}kx7D*r0Bwu?zx=o{Z3uTVj*9!i7g%q|JkvjMgR3WpF%q^=Qdx%a62X?#L zkVmFWN!mqmL#CWL*669*`p;^NpQzGA?dtBcjq;jg3B_M>NMvlfg9g1j$T@jqnULXH zZLwmit1C;UT9w|l*@PS7p_9Fv3P&H1MkI<$omGdz`q`}O{SNxMV`T} zaec1RH9T=M?P&LMhUvh8s3B7-HY-sY;Q}{%hU-P%tl*a?!wQq>I7F8?u44K&MaOnb z{uS`)L{o?MCOm*XLYH)xZDzglBmZfsNwROzjw46wGI*%iXB@@=A35pBMy>8P3FJqO zq4&1G56=A~&1g6I2?+^t0&8Pmy4Y&bJy(%7y(X!Q5*4;4)Hrl3yPbS}P1PM((p!i9 zRiNpmJ%?1rGo2^{$mO2Nas9QhBU9X8G`shv_H?xu&2F;i5fr_zx;Z`8R@KU9{2h zSw~vw1sm%AMYEuXC<$#U)Ivec;a|;eL9{S{so4SB=(J#;Q^Zcj`7^27@K+xR{;`gA zfN$Ca1l`BqWefD{4ZZ#*&i~;}x>5D#&uWv#?y>3>ChFvy(+9^nT6F1@=6Mh!>91-U$hJp?r4k>_e&TpAU<%=)yuYYJ{_HI|kiFR) zr+`ZZEujH0m7G-GQr3+ZDfPj~||@?*997 zkBU(?NQT+#KuBdS5-uYw`QoPD`WaV}BJl#%CX%=3hQ>n3rwdZ}EOAY=wcAHz8ruGr_v_96WnOs2?*A(hdo55dqO_pGLh>e zOFzKZ_T`o5vS0F48EX-MW~GLQvOqP)3s%8iB!i2pzcz9l^43pWc4^j(tSvyRE-X7V zz-kHGWdMpd74(js?Ce^M+AQ=(1;E%!wMnd3O3PmobJQB{5e5xpNYHaExhvJ(0y>ZS zc~TTf1{$rRNTELKsoxC7<51x~dQnh1<40CK91DLe^!8yGA#t}I#W5=L7=L*Xbptm^B1-}_k( zv zkgQe$76irdtPCnXlVKwKeD%by(Nclu zzeq)pkBf^dva;a7-g6dI4RkutS4Cx4K%Jd1Bw*^*psKKN-1RKDJ4q`pMSl_T=}$K{Nf5oDFG_<59E?~K%q}MpRCO> zG#6E-x#S>#dFcM6Slire)E(^3?R7#61Ej(yf|GyuyJ#g^pVhtOf!q8eILp)~<3QeT z==`3+L=ijDU-V-Y;TMvpM_jr=hyn4NDUoGhw=*?UR>sc@X=e~}3UIr*=6`Di2y5uM z20pnfta=EddUd;&wM8AKd={|J%ZELl3gT*2Y{8C_`mf$jCGgSz?-Wl9vCK+R>@n@_ ze2ufd?ZFO=7rrwh%4|~`MT@#u+V6dA&gu1E`M>jh{6c@*QHhIxauNRSw_hBask(#J zwI?+|Mz|h}D!=L7j@ho zH{?!(oTKY`Z@--kt;0hF8nzuA?-6sW5X7_oI=g^v`2h&s#zO(2aU3Cqe*))ljDm6&nt+kdgAMKc5W1IqR_Ty(EXRA znCG%9c4A334Ye?S2`GX_vLuihW26GO%6eMkiBiXItS-HOPucvFXA%q&4em4glx|`8 zE@3M2D99-fmIp#CGFfPz&q{{(7!IKyy?qx$e*wPBX2baPv9p8TPV5>_yt=Gb{yFAN zu3~(P?}ocC#%h>Ro#LMi{ zlLxGJ{AoKWY+%Z^8J&cj10Mw-oi7Z7wz3w}LSPk&2M2^kz7T#Hpc_{OSDjrNqHZ3j z7@`oeQ;Wen77zA=pTNh*ao73ns&(aCe6^+j^Fv3b0R%FMeq4kHgkRT@pS?z&zrd2k|nm z28j`2m(zI)cscpFJaJbnm?pITLZKjx?V^`~LW5HoAKOOHXMMS5sN4LIoS1<^RGLQz zga*EBC+lJ>ME_>vg2hFyPftG-O#V1B9%sbGrk{oQudg!ZfZ&+O#V|Ap*98lh$ITS} zN#PzqV6c)33?06mO^>Y1(g!!W`8u7UQa&;%(s#97DDe$*b=T12CVwRLv=W0<8Zmo! ziG`w}NGVyK)^OO6@`dqqk(9aPLPJ7Nz^&bnK*XJ>CAbATc%;-Ivq(NGY@0~h027^v< zCI5u?1aPXqL^8_gyuSi;>aoA7yX}Wg$?F`Xx9gz@*QPWuosRISHKkC5ju7y z*ff6J<(Gb#AX$|9*^M8E%s@8~LT`IvmR2nBvK1 z^&MAqyk83=zv-@oGoS#F!3rUdD4_?|6il)yNIC~BRvq=eWmBjpt!qn+mlXv5a&k*# zdI^3Cgnnl!EtcPox4S5sd2cvhz*G zhfDdBAW8GYq?HI|KTg0kioODrj-*x7^&Iy-N}di2*MMk8X$oeMz}+JWg6Ao!vkv55 z@|2n>tPoD=HL7y5NupWaADclhfS&J1s-?Z&m}T&E+o|duNHpa~=NR-78nBRN@kxW1 zyC~q(1N);N@-?dIEBtT_Rw=-KHVr8^s2c!edWixGObCU|9`qNV`#VHoG#yd`qlOf404&K~?y_P!H>?3Ripzb!|o`orM@&0SU zu~K2zBfA1#jd69+UfZ%J_^j@NC;yk{$n}$(nb*cagF+@>M-3G>U-*&kk}k|<9@rDp zn?H;QR7R{H^D31r+!#Blc)D;2c_G1}CJL%8@|m$wvqTy?`droDbpAfdrO|Sdo-YTyYd5y0%36A3WX0(NE2u+m>VX6@^O7D}OxvVh0O z?1XB;wu{&^3UcjJpCYnuEpY19T4vt`t$!ReSq|Nt4Wv8E@J^qv!qf?y0ZC}$c@1(~ z7kv?i-oD9lweC@Gsbg+D*iB*mVQ30n?p^KDE!{7R@@R}luqMqJDezBVcf9@ruDC9t zI?eI(B3=8BATpLJ8fY@}9+-uhidQO{^Hnhxnd7rgIo=+Zm)f+%*UR1+zU~ zOyxzfL3Xb>V7ML9%@N7WGzW(!iy8*nej9S;U;mvgiF4F z9NAIih9v+DtCtbG9v?3${kPp?#6`O&|7m#J>S^x`YJ+0wNwW+zxJ5!JEFqp!8b-XvBlJk z=%D|ys0=kFK4x7!c=QjDg9RZfM4mp1@KHBeV zJQ@z5>B+Kme!({5bu*9eX38COSUq07-&UK5p!0#|V94E>#*ResX)(&-F`J(Z0?} zv1&s@`^dhvS<8bX)1qeXq(Q|A(Xs_AX38(o&mXQut*A>Y?wolMJr>=bKO5kcV~*52 zQIB0+tsg8o$9Kx$6;O(4T~-6MLU#wqw2k$GDmLouS+D=<4zM&S9C~doRdJ*|2_MUb z;8UoDy*k@@Kks409+GwiKk8-;+AyJ)HmV_#o-?Lx#=38$)y&!Baub4)14d9zjxf)8 z9trm_zlnSmSC0I$W@unCrO9WS&ZjiC65Z$h@KAP>@yk|BJ0@Bv1jVQX#N!nsXkj2) z5su9gGPUo8ub`07`O6Huk=02hX^!-GoHcngDn_ASsi;*J0^2rqd*}q?)g*OBwK?c& z6kd=-^lk)BG{Q)wSn!p&seN<7;G4WOR%@*~WOK3aD{owFeA?uB9U`>4H|yGPYs|4ANJgnr<0>E zvKm+=*sT;yL=y+xrhB*ZVVn&u;_lFcAs@Np&=zL(^P_{>bG0Jy^OJF%bF|`P!)l$V z$j4nOTnKk+CZE*5^1|Yi-&}nlIQq4Ak z7EVNyOCG!N>x*6oyYu+enfwY-G(!Ux6%V?;F9Q7XKX{Ae9eVZ6Dii{*k2o)@5t9ba3eR9eLxU~-Zpxrc81T^F%LgconK-73Y1O9< z2&1e($!+F%x>4~n(fs}`@;AMataRaIni-1inv|WK<#p|KXrHJRV97-%4QnN!B97u= zouts2nDmlbU0{7qgUc?e*9Y|JZ?BZqbZKg0H6?f+2z{be11*6R^4c+A1sHj)P)1Y+ zD*oxx2n~R&0?$4n_QdxFu*B&m6+6k!wVV|5WuAP=%(naRlzL?^jT~p6k4xwfDE1cA z8ZO{v76!B$sDZ%q^O5W@;p+S!lfwXF1i41#Wo zV@^d&lVv{zDJ4QT<;zP5wC@V5zB`2&M7}}2O2i^je4#A@$As%OA+86Jho{ysXnr~` zDjRcNPo49yyd0V~=`W~^oV7*+!gpF!!-}wxg5M0=#*i8%X!Vk-K+i}v65X`!aFBQw zM$3TPznUx7d$1)nUU|nH=d1JnDmsyyKFJZ)xJ_?Ndy0ZxHyZHK6ChQUksu ziEWwOpADe@#T3-o?%lqf5Z@y;31K^*nw>XN95h(z@*2jXwRew0TAIJn%yd#{Is$RC zX^Ld^&)aOIE#Ssb!hvmW3PZQwV|ooUhbD>cND}=P`W8vg)wB#ZO#XK-Tj~0A2{ZLp za#N0}tcY>PfO^H%1pg5cv99t2F6R)(XP3?IhD-d}uvoh5CZN>Ji`#60Z2|?HbLCoO^WyE?ket1fmc(1txL$fmV9}v=8p)E!S$v zU?%BjRKF6uLypPmiAG^RidfTJmd5a5yZ&B2#+qV&XN8>DSCDQFY@orI`YzREjYix{ zO@$u$!0lK_T%jY?U_XS?YCxai z2}l{G+XJSfr=!b_g$$Mf5jy316cdMjF|0S>ZQP^If1*G{0E(`Ji3IN%&)n()2mQpy z(A%#Sh}lh$6|`ou{fhPBu&`>;QkBk6aJc|#IrWhy)%zax&`!$=t-POty2n-JR$99( zZMz$hK(gt{W_sOj#*4@21n%u=GJfVDYbe0wxfd zYyhJbPsmd-_^dMdCKv>-W-$4fUW06~;!Z<-k!XpiLP^c$Rz!URa{>(sxlyeK(%uBj zwN;X4I0Nekq|Js-KC4t^H4hsV)wA9_u~n^+zrZCZmDX@eObsG$ntA#sHHFMF1R$kk z-)=qKlI@Zfj7$O%G6-MsA(i6sFp)blN(l(a} zA)KpeOn!Zc31%rTQrEs5e+ik0XOD-|Lznr%X{|dbGV43-dW~Dr5t7ecH133wi%Yvt z?Gcm2*aTJ&DuFY*K`or_pTYsd#+QFF*cwTU_u%t<$8AAhC=SQab(SlBFVxNc1gfc( zwLtjq>QKcoFc7F?_Q0Sviun;h@C`Y|L_9TUAZj^|haXQhk4FNLpp7`J@v?z|PFXGh z?4xR|&eWZD+^JX7UO+ZQRz4_2+H3l-C!%f3x0)r6jj95czQ8>EiS;#9d*fvRHq(<-O~ub3{-Gz?Sz{DU*U{CU#)4jS{Fzyq<%++}WNb2*Ta^ z3iNU7#l&ydAE5FxH|_?j^d0yaJBPc3hk}(ruPwq#3Z^#KsBmD>P)C_qwD!QoWA$J0 zj#b9*&P$7UR5($4Z-8S!JWfG? z5KG8Xg?y{?UaFBX#;@!MWNh~w#{9o>`)2%x2vTuvpdRY&E0m_j*kjU`h9YVlk(_&P>Nend=z=o*p0q&*k(KY1Z--2x8 zz0}Z!LOEGu=W3w)2pecZy|&{B|6k?YcmiDRLA#@F7O-H)A^RUY2q$SBX)A_3KT@zu8&k9>4i4=o;32APA=H7L(*78!5?qROgmU@@}}GAlQS z=L1u#u-qm$lZ8?>fsdV*tL6e8IfHACjIgKf3qLvo^KIiJFgdTQIbP-1$LY}VJfP+R zks@Zv5r}5S0%kt)8G{fE|Na7%ia6YZc8=`;o%7)Ih%#iO%-Tc95)^7h*pX9#CH&Ze zG~cV)gR%_@-*``Fsl)gJ98qv73d$sP2>E0cD(u3~`$!UU6r?6VOMs4ga^v3y^wu~r zI4tDJhc()J*pe(PAyIakpwyiqZT&8yANbloSol!QM(qyY9J8VdN4{2fyZV+gd0TI4mSxymXY z23GwBGEDiD^2hF0f05u@b>FQ<9JE$hH`m;^g*J2B}*094Mfqo zE7;SIpL3TtfYt8`Z9`#(ie7o$xzixVUAsKPDP?+>DZ@aBM7)0aemMIA9d+(aE1q zOw|ns)f%IJY`BE&bX+T{&{zFIDY1RqsJi<4iEzvG2>|F96%fH2kU2)P5^1tP)`T(l*Vi~I7XWodsjMq1oL3h2HG5Tq z?f+ST+x^dIShF}fq&2uUI8+U29IS`+0jfch_ym43X#6Rwlrg>BVA9Hq4d}mYEF;3^ zIEDEd3!sIj-9*%Vw7D-%F=WlvVeg8`Skum5vU3Cg=Mg&z`qRI5EdO~q;z_E4uGrK&k<vH z`CVwtG8~X&RlKn=X@cF*U7v?FZT^C|F|#33tIGH!!Q?~;vFs$o87@|=@+Pd;CBkYo z)~4%~ydX&R=v;TXk4ljbt<9VuUbPpf@UHZpBB+@op>WOWjSI8?JKj-Nk_0 zFE&8{Kk^`oiDnHiS2xA3tG%P(h~lS)U?ouK=Gx3goi7X^Fp(HXlkR_A+{f06I>(&MGHVKCNF4(@Vc?i3Nsk=7A_WrLcMMG& zJZ4m1n%0I0HY#IC1xsZ-T2Pf3Kykr6uL{in>D{tVPp3jx1=k}0(<5UZ5zO{I)9Fjb zjQ7Qqc_UX$fw~!nCX0=GLq#s*tqt88$wrHOhq0Zax7LG=;%$Wy2olObc0+N5(2Pen z5-3UbEmXk(eL_$;1_xZu^GD%AXvw3fW4_$lj|oYB$(ZUltg^%xU03*mR!0S66vUT7 zPr(3o320WEQMDEv}Xw-PHc(RzlAOf`O+77d7;lN!Q zU__~%)!MmI+5i|Tw3H&)O3U6kZX-g-o}mt3x*$uz2!}2O?Qj)I@o{E2|CY$<9Wd1& zj)z}tDvFOAr4+N}1M^-md7wD`QVx^n&o)01M~{C3?w)pW0?)7PA6g%!iFIfhzi&TE5Vt&uC-OWw&A?vovTR6~r`?8Mt^oByI=N+( zUfCtzjS=My2^FkreYqcxQhXxq=&2F@@uCB{F0&PMXR&$9pxq=4Y--9u0AVz%QobF} z6SuE-`v{Q7osqtRC09GKPJ z4SVp7SI*n1%kqO7vl&dWW?|ny%~2flM#3mz=;i`b4|je9Cj3^tn<4I`n6U>2rJCU5 zV~iR{cuxW5pU^Dc$9G;K`O)zA{c%X4%_{W?hW@A9Soxp>h|hbY)O7Azu9E7n+lS5) zKOWc7i>xO#j12BB6gzEC-lGYTJNsA#x`VIFgJ9Jtw$>*r{V9_y`~))ou6KXYQ#N#G z#ws~7MmrH_H)Qc^zPFQYws$SdT69W<%Dfk<_q5SF923)aGDSelQ4@~!iTsm3om0P> zfmXQKQw6qL|MpIY>-t(T<8Cn=imXE7>JM}=y8N0JL0>!lV@j8M9_95BeoquW(v^CT zsKSg2`Z&>9;1-vsw&@%r%4h}#72Mm6>*6x^j5dZhpd2d%%$V@1u+xhdYTR|Ca97}k zg)*db0M%~YXKYUlg1Q;DWNc^KNLd2fBO3hfT0O6l7^rplRe~~8;pQM#M3AO0>5k22 zrB3nRbiFSXKhoF2*vGJ6{%MuCpIx#Ac@hzi5Z3C9*RJhWub-=Rq?gXV$&=|1&FR-q zjEa@0&g^|y&YJi-@6)V+zm*;E2k}-I*`5hNMC`m15>8XKc`KPU2B#liFWhU6)}tYkJ>{bC7~Z*e78k5+zoaaA6lpEpbA8#-qzKQ2{M z6AkXW_JsUE(IWpkUo=7U(h*my8+BTwV=7HA;&MDsogWwWl!?41GG+u7)||OYl~TTp zYtreQ)~`cCznKfS+mHH(GsCK8^;Oe8ZhF_+X)7Oe# zl`yli>a-x_wdAcoil!RfrORfeik~d0f~8U6zRwuwODPt#x?oxHWJkJ`o0^uf`?409p}B26hnk$LfB^;7Q5S&e&Db&8){ zDO%9ULPV1_6RXs|X4Iq=t0F0{=R9i_*d-g;DAoumtnnXxt9zrnufwni+OrA-)z#}%O&*I1aBM|Rxoz?q4&mFz{_>FYnzXD+ zo7PLzo*}UZq3Yd|+rrH03t9CVv*O=p))l?Gt6U)$|a{5h8S?UnAp3K~;0hHn+4 zp_Qt2YuPVJBAsQ}sLEutuI>(4T_{-D+|tqK+8*6Q za=311sbKMF>J!T~CF{$|9NF@Su#4>*ewxeL&KkEh@71ZqxzMt@=y4UxLJyOHjbiMJ zu&H8MrEkbEqjQDsH%8blT|RB_7zt$h6BF{+TZWA7`-!XBK;LGs0Z7y(>T*3roDHAuZ(=5Ij|pL(p}bBgXs*@V5gox)yIUA zlPvy4e{XgYjkerUxa#Fr%W&wi{uTPg)*TS**0TfrP^B^S+K!a!c@*5gwE(`B4|PMt zOjYh`anB%qGgP?!RdQ2+KTp#CFc+tK36#mHBq%VKpk(u2cBr2}NzPCSi0 zCSgfY;@*h`3iWSd@g;75rHjG1H|o*&T2wV^V>oGQ0PM1x=sywTX`Q5A(0vm!ZdQva zQ6DV^FHWcBT8G4Rmm8D)&>{|BvG!D;pECcb0X(QzkIr_}XM1a!1bWPM=1{)>d=V*N zjGBnkQ~CI!vg#qZGhH-d2vX!Z_YJK|A27mGyBt0HG>ecyU5Fm;L>)FdgDTw48M+TeZZk3ydy?s0nRu-(# zoMs$vFaoZ7i7SGx*0TRw&T=lV!1Q{!K;g+|Xnk!$U{4y96ZiG$DmtAR2Vbt1013LR zPfsAXm^;KaQbbY0pi=QJaRes8gk=D@Ee0NKC5<~Lc!bE$V*qb{dcgM5^>l-Q9(6Nf zC7`Y}w--TY@G78XhKEk=jx6;boS>d{J4r@MIv+6_Nn)QZXIxLC4>6h8duS%`tAT^O z$E)>00YKVm3!Uo=7nCk1z|#CllHO70ZVui|vFt~{9~VR;IJ<)JrG3C`eM6Z^qwFYt zIVbAc3WPZldWA$xLoh4d`cbBW@1SUDXzMj?s{r@IH%o^H zL~F!v!hZ>WRrj~)<9@yZ2iU^MrCFt4c59>c%h!0WA#}3cW6-9qMrFp6SFrVsr#+xm zY20u*XraKHQy(Zo_SzTVE%BwZXI*~p?HSAPpC7#yvlJ+}rm5*L^$mY@o=J~W4|V6u zGvLV`RyF0BPmyq)YZbObxNSc!^zJ3~X>U7$X^X$2%&yqGeebjicR|ioyNjOKX2R&X z`9rb>*b58K(*-}Htj&%Gy&Kl8_`DjxDLS<+RUbS-L3io>JSaqUn10OVFDPOu$0uXV zzS=Ap$|=D?v27mDmu*ZG1X}K6`n{oWl&R)=%!nWe4S9mC+O6w6UUzyeYPTor(^v*) z&Sf(Q0=#WHRknCm8_!;Bmo4xWWbtBtCk+|#_M!*rf`23RwiEGGZSjQ62_#zEc^bNv zv0!|P)Qit5{S@;DrRp*BOzU-e@j*)mpm&PTrr2o&?qqJa#+_`L=xLX-w@;C=@`A)@HS8s$(?+_HT8rRZND}`6 z6L|sp5c&%RCQO@iz+V^0_JxD9Q3*M!F9VCM1tvncpTu+3$P{_)EaeK=w?|_;3M&s- z`a1OionmDM!kdswPxXdE)VKw~x?AIliz)pUc5H`Aj40x03=m?kXg?}ywy9+lCTE?t3&d#5 zNfnWhfUG_G3+e1J!yDxcmBI{HjDd6{;1Z}UV*PpzkK*|Vn)kqSYUlNf@CZLK6r$?( zddDE=$`;baK?)}U(#lZZ8#5!>MUgOnSo+x zl-vO)u)}|utAx4m2uEwoW&!~VZn_}7_?5vN$w%#;#EJdvEhL({ROuK0T1-s!vCn6yPUU0~P*b9ErzT|qgGqqEmHi^g_|7AK8> zw{wm_SA<{dGAX^pajHM^BX?To5n^vWY|Xt3O1fet6`>8DLx?J?p-d*3f9s@`%4v3W zD!%~7f3^jufPrG9Zz-!vcqo=oI4HS3`smzw08?s`7Dg#O*bjKB)-EDv_fGT_s=lL= zS)#4ju^gHq(Dd_JxH5Ylo-?oGM6&Bw2<7Iv&-ePKpkgW?dk6NrqxB>pleH{EVH zX_rGytYI@&1{O#)JY4#hq$U6vwU?)}x3xubNPm6jz&vq0Mxp|+jLus>9{%dgH9(|x zP|efgnSq`3%xPJLO-W#dgS=rVHV}@k@)db<=8}Fhy+xI#a;>ozokC)8@KXDE&}`_Y z`&Y=U1FzQVxK8^#G?_-SWeyz+Bd%TRI^e0St-47KjQ5?+Le|tD0>Y*tCc6BfOMk>t zUqh&EpkCn&=%Sf-?JmdNK5Yq2m{O(CW-4j0*g?z`1CkDH^Ft7Mh7n$Jo5{RhjzSKk4eybM3Hh|Ii~#S4>){+trzFwUW{frz=+I*gb7j zX?fE@oCuJ=*$=1`AR#snre0C{9Kr@mC;d=r%4@X&Uw;8Pn){o*cC*F;=^9i^U=U9F zo&)u&^gDq4w-&-Z93$8Ps#17WwBR1(Jgamnv#fc=uuD9cYDPahbSI%vMmQ~0X$@k% zA@bM+Q9_)Q>;vaK&Ugt_)cg_^JrF=g8`@bCvzjkpp(ZC+>-%-n&WaYEvWAK$Md<-{ zJ1VajDuPf8Zvl_4LUy_#w$gQSNjGA*ElTC!sGnKMu_`S+bX4|oN zasYeaM+vBuM;OA%w+cHbE(x?C8lxg~w`kUmGu--JuwXb>l8X8GrqdWDDnBcqx7boX z!8GS4a#wc+$30T0RUpx@7*JQ&DK z9p{e>S3a*@Fzgamz%VrCb1g>Cr&A z(adU9J>{TbE`XG6d88a(VywbXwdtx0@;lms4HV8&MJ}$_n-=X`8ymrtn3OiDG7%jQ z37d>Qb_3dC86^smTzT|K=RlfM<(J;mdjn{ANO_x=cA2@#=^j7&K6g>nYlb~XzR@ih zzgPtVC2<_O0zv(vMc>q-&^x*<_d?T|?41|o3%x4J5|BVUn~X(x1wStw)p z#tOz)c!GJD%FD(IS|$cxLaeP=5yCz31*^q=w@$SFzvy1VyMM zRUIJEydm^9xszd2wT|`%pxyYV`p4-M?KauP3J%XTrT~--7in3BDY$P*6ZB$k@Nj|2 zPROwPVFt&Wj|L?74TG4D>PzhEL3SV41~7Y4Cv`piScC!MK)gmVP;=utl z(}=X(F?)-XOiOx6e^L=y(fpK3J|saYaYxC=T6qn1f)<5fNJzgQ;=5iUI-_mNo0=rN zw-Gx^(1L0b?GUTM*XIVwxqdmyD(U2<`mCPxaKhwNY z!hl}*uxoe7>Pw##Vef~XKr~cQz0FKB`M|ki0hu7eo^JLr?ySb7+AcDw9UkJAmQ+Kp zv3@x%d%>^(M|5nG!Gh7?&VbBGek+P_!R!sweP@Y$&00H;;=*NTR#AU4_st@1FCh*D ziM5ZtZ`{#u&^`3+AV$$I6K?C;Nv0!)RFtm+fw0C@n>BHvOE3kC1v z$Rf_Ccq$SZGJL@ZVK4EH&_f~lnR_>b!3xfJt{}to*;QHg&2t!v2S4&0D^<5>g)iZx zApcW(LFifC#N~omq(p-xY{*e+Mw1*DZ-Uq5H(hOU-+mSmZd)bo-k&s%5FW}Ix5>O0 ztLtcgx%8ZxJMkm9P)llJz0#bVlZXlin*YumV?Q_Z1}EIf1+#>v+S=q%cUtyXBiq*W z*=(BZC@DT{f&`LsJv;}t_^7r%Td3In`ffTA|6!nqpdTJg?h3e$jQ9~D{en$EYe@R! z@ZospHNx@&{%^S<`~RC8GIKDn|8H_bY1hqv8?lzP$_;ZD-6x!cQ9m~?GAbyVGFFz< zq)d)i>U<z)Cy==eR(3M%dkbCmdT0R+{Jdsj$3&N$NJW}~8f^%dM{{x>0HDbld z$NQ<3_-s7e%YWO~9i&ovAjBdY!?%HFq5LI0aXWK)e!0+eXkSQ=Q6xGiUJCh?J7*sJ zh2O03mnGfejQKdFTb4}G809L~P_;4K&j!b;k8#rW5sR#bD+zLDrVQ_<(7;rA(MTzz zm&FrzvMcfGbPn{jN<+D>_inH>YxGLt})pvc@W-sykw1ZSzM95?7}3uwti@y zv8G~KTb|ZHxpvL^Y{p)t+Wf8D4-=byn^G3J)*Iu+dgUsQY=#%q(r=J;j*#sN1D}TW z;Njtd=7+vX^ta6nor#Op5~(oB_Meoz5Ja6kEhAK?&#PkDCK{N6(aeO59(I-^RZ47? zRd4xv%>+@w@bxOS)l?NQ%+nEpDt^QjJK_|`1t@c@rmt)gG(AWu%jH2?XXL z?uDxs-G0UFg6`2HJ2?o+kG3FuSX*R}1$dA< zcw9Ua39S~Pls*ECd1kpWkei`pG^iQN*PQrjBls zv`#AK4Rkg`vq5cTLDVK?n^e@JJ!iG0bqUC?nBiYFRh^lEp*8~=9{p=$%kz+InY}vw zOpwaVgHLt;z+nq^&?l9(R0S(#>M~A-vvcB3H*XgYhG(>asWp9n;v^MVZ}1#6zdhQk z?ptqgQ49yNvIMZ^XeaI-ZeWx+2S*|PA(PFk&fEuW zttepZNL9Pp>*=)3GD;@uRmi=WT241Go5`J@o_MZ<-}rVA5o-?3Np!yV~#iFT*?65%Vtm<;M){%S+J^JMH(o;d%mP-Ba(xkVdxJc4$6UAIs%Fy2(McRW>1wr zm|yOACRu6R)tEIesJ5N%3wHb^NU*Fx4RX`?HyBdX;bv}x$7U1k7cljqWgd%Q6;RJ# zJJ3jHsJEi6R<(zCmBc@muJ}3>T(f`)pVqO>+(6G)$LGGB-w-l{P9+L3b`@R{Ur_pC z-xxCS+max8h59k>8D!RoASu-Bs{3oHSZIl_!l%y{WSiB2fuU|=GOe3bL#Ij_tb$OH z1WJ6c^OXEW%VvmrPw1+2J0rN^jyM`>P|i)6B?$E;doTgX@5uB>nGXPUa<8d{m%s0Q zJ~Owye13vQC3-)P*N<(O0q`4sN99#Mb1(9dch)1c9KB>r;5k4Mvr>m*N4%zL^8z;5-0LHWc(s``AYI1^Hga>GA~x zYXU2gKX~oEK-&I}h-iG$Ibk>IU{fQ_PDxtZOOH>UUXk45>V{dZtHMH{nTHfQ8|`bbtNp z27+&;zxIO4ZCTJ#J6EF7+8DH~fGw?tS>mgz?b#XHM5HS8u>H5bQGxZu(|epRGJJ;}1g2 z?dJuK3Os3kiZJ;EQr&ws2L?z^;5P4&M<4mtbUbjAE?WaVdm|La<;8XT`vQ1eE&Df=3Uec(ctJ_1nZIk zvcr$O-`}A$w*gr{P)Fl0VBY+2%`p{spr(PFbd^9qTFrc8WG6-lJ9VQYK27}(SNEj} z0>_T6o!OK!JOhrmjo~sjKs5G}WP8F*gztDtJ0giYL(*e{rU2;s>LrjbB=+kt*gJpN zyoP)$i9~F$C9fFqco|`QQ=I6#f{;VovM6K|!M%GDZyi}26OTA`Rp)gVM2OMM7Ob}Q zUf1OZ0v1gOBouh(3*=Paa%NYdWYclxJKa8q4`2-=KGHc&~U3Gdf-wRsKS4g7Nq`H^h zzGeWl4Sv%6-IFsAj3e3C<);;7gT~rZMaMKY-)|xEmGTFj#78{TmHi7_u94Dh`srCO z5C=Lh*Z3RJ1bO^f0aG4YB=*AkyPy$b^qWQCUCj%xb+qa0k|fC}1E?JgO)V7#D5I@C z=r-HeeS`CdTFHgugKx3X!H#OxiHhF>D@TGK3EWMn{{||LQyTT6Sl%#oVy;5K)5}D4 zI(HgBR#+yf6Mkm8Fx9vh#W*J)DYt~=E*^4k>VktGy)6N(6kDtV9@7M6I(Lw&Z-__eq+)vQ2T+6yay^>p8XxCrEV?#aefKV8!2FqAeDfh+kNifl0Lx5}e z$$B98{)2GQHK8SGY8mtajN{y4Av}$PeJel;jWw)sj)KI-w%`e{eLt(nstGkvfz?{8 z3N)6V@HaQED3@wg$Sm0F1}ga9H=7{tdYp!wCGAKB{^duGxwjqJ;c(XqmchX zSrJ>h=G|L$c8E2CeW$F_z%7FoNRl*_F-OaW9b<{k^I)N}IZx3accz7Y360RVIQ0-gjMSB!qLJzkm{yD<%%DjS zOB!$&rnr?JzD?x|G-zfB&V{fh^4MIZj4#Mlo>t_u-8WuWNue;*$}%{KZ;shhEErbR z1D_1_(|fxac-ZT%K2=#Hpwz%!8w|TzM@f=@MVI!~P2dS2A`xBb>Kg}$$gs9DZ3m_3 zHUKu^%#QQpJ>NP(^d=9j>IFF4Ppg|__&-}JB{W1NVE!LFE6^EUuUEYTXZ18dROlAm zr-QY~sz|!m?u>R=b+{r+P^t|jOs}OL6QcuMTw%P!Gf(i9tBD+w=0t+=fq1TATDp?_ zfIcHyH;6qvX>ENOPOw20IWL^I!;HgWS>znB(oMz0#&hGAa~tOekIx#R5q2yWp?}Rn z;)ozTAv6fSJ;)~Tg@SHWU=dn{oeQ)VI!7?1+U|Zxh)gn(Ghpw;Ow-(i1J=uhCiUik zweZ(3IJE+oT|fJ-x&Q)Joc82xf`1=1yhw`Igd>Z0q|?L%Y^+}1A_TH5>F=FbUoL{d z`j3M8etVU8Z%S9*UuAta+R1+Lw3l6GAN+luPmlA!d7<)7UwHh9>)?5Xr>!(|unegX zPxRLa4fejG{gHj8wSilSeW6R>F>RfZFmb9mi_G|ZI>Wb+9%G8jFW8vWs6jkki6PVK zXpzRBgM4+<{}{UMmqFCd?Ri({C2+ThZIeJMTMXfbYi_Fr4=opYJ?z1M9|sm=`Z`I4 zy6^WvnLR&)*sA`y((}Ql`1$*$ZC249MPhN*H@gpoV59@E(BrFEfxLS2hv2wn3x$J! zf1kfeyc|G_o-Pig%g}7JJL42+FK!e-TFS2@+T6>2nn}XRfKHH|3=vT_D-F6ejTv@V z$ijk{36b)(*a$yfK=Ld%T3mVy`yWUa(N;n?7rf_DY_9WbE8}Dbr|`QK;3#FKIFn~-(u85j`vx6GrobVp>pgQ+#1wK5*-#P|wHc9XenSRupz9{p!1U3! z-cEn52F{))7KWJt(ZTX;Ju+wTw?t{NMRytt90|Y?VkNrd;Ul2#!haU=0bt6TXY5+8 zNia0BKP7MLQP{|VxKLo4PESdTc4r^f!46DjH~ddJY~v3})Uy0>F<{mv?Dy>YhGXT} zUH5i`$Jk%DJ}+kX)Ie8|k8(iI2Qo;UEP!D4z2Z%^tH;s>(j%<{JbiatTF0~}gjsV! z4^@ZEPlacFZ66$B78>aHMOgoI#xiEUp>0tJO-4Wg=s`(=K#6T}67X1)3wf16kvthI z%V-J&;)jJ?9_<>gm@5HAx$aJ_4Bs-7Vvm_1mV)>fjT7R@P|OR&fN9qjR0t<4P9&l^ zx9AkM#~w?d^{Uy)%8_N;eG@36vwGz?q?Ipvu7%WEY@!k2yAHzB1wrmUN<7A^6G#Zy zdEzjU@zvi}ob|Gs52RJe$TdBwz&V!isWVqtfOA+K$u?Gi&42_^lG+}!guW9SzoFG* z&yliBg_0O8|2u11mF|~#xAubcP_>2|%vm@8Sq^C;OxipBS*mmAi|OSsMlTF>j1*z6 z(qK7SG`G11XMxT)8?S$@(_M_nCJ2O;KAHOAL)FoEaEUnl?8!eEeq5e99?3}KFw&_nE#)U27PP#*mp8vK$Llrw z$yn(p(#q$Kqj1&-CYP(8{;p$^^5dLf`an|Zdwy%+oRqtT_kAh$u$lN|TL1U`=Hl#J zvszXF4N5sntD851`{;jA;OsUcqm7Nk%H?==PtTWwD<_v{?7!{u_%uwn(4ju6Y@!F9 zYkZsXkwlFv$A4Sil$V%}8}96!q~xp=o9hP#*Uc18U!dHLuiy>N=*{<)Iud{-W2 zQd^bk>m(I7xV?_EE9|(~){v|F&CqAbHMUyT@;@J^*whnk9zR0qrNet&Ih*$%8lL6t zZJUJCWsk)scE*2mbP}oRg|a9Ap?|9g=_YETHrmK}iy1Xh#Csc-`l7K=)b*9@zDH_I zVpM__t4&lA9K`C|G}$UEEH8&k(5Ky?BrA}?=Hmh|hBL~Js7npyB>17HJJ*?rx9jIC zY>|uN0FiH+Ag+^*JzGX{nqoB@nvgXDqKyez5^E1l*F#}&o-h%5PU-2t;#8m&x`9d{t+y4o zK0-E+@I>eGVW}ZLP?gg*(GNUtB%`C;w0CVYMw8IBukA)kTbI9Tom)eY4EX6OH0Fk3 zYG6(LF*hnK7r@I-WM8$rD}3qx@rB`sscZ3y4fr^J9@+(EX=XxRY;=j8_my^kl$&ch zJp8@cBj2mr%uG|XaGHA3)k$p*8-r?+DwVr?qFM6<#;9XFjDY8MIL)dBOiAa{oo0`z zARIy_u<>$^Md90*xX@Wb8Iw9spf%vTRTcVS3@pp;c4ud17(I%JVLFrl$m%FXhq4|V zkD9cTrYGCGFx%^J*{Zy5-pH$01W3(9Nr`j(%@Np2ZAF!ljHgJ_>x5D;fTqGpv^|hL zK-%%})@&9Oby!gFVfrK*JuK#>923XENDC_TjNqQh|BK-42PgTV9fuRlz!=rTZn||Q z|GnrNOj!Yh1u=wMO|@R&By2KAzF{!OshJeJX;5cmnhKH%mK76@L z+v%~RR|y&p!~UuXy7|s-=mXHNQVwrbXkF8q8IuD%ng0%4TR^})oAH5RS1*gdH8yEI zIgk|93Uf+miv4?@W1YS_XwCI5Prk$0<1ZOzktj{m@7~ZAPot6knE%Ig7gE^FB*L)m zm|Q*!OCqJyTPfr@7zCQ9u|f{HE)NbPN_oD$7t80*3XLcpNJn$c1@^{xQsM0SIIO_? z>vuzWXBoJ^-F74@E-h{@%R!080{ARcn>JlV%b!bUE*{am!MMCCw>6eZR(AHfQE4yX zgs2HXZnB*Q%3alOtGWD8=XN6eak@P;^l7uN7Il&N`1B2KUD_1o^&0czTt}>H#_ME! z!kA?FNk4W+Z3>-bcah$4CiH`@_w4Kf69;3F-7NOV-eLHidz8c z9z7;_GX-btmiH`*XYFNKfEy=46t71ktlvI8Q9YdHIm$a>0Rs2{TAq|kVK3V#0(s<3 z2)wRr)-ym>n&Shecv;#;X4P^#?WaJieK~MfL{jGYTyo6#^CmmDO zCq9qBU_hiaENqs?sSve4Z-B(!xzD}NKer(W z_-){-OI9T`%ZX#)ug{MyTJScFmAkxkcif7cLZs6b?Vy`wqLx={ke-7h>ugp#rMnSU z`iuP0R33?qD0KO?=bRXz>efkEa%*h{y4*6Ut%6*>Gv=9a|!m zGpI~^bn4kob3OfMI-=a6U`q0@sARD>U^9pYcIQRSM2;70vSf~0E9?<9vf)IgJx{C-h{YggV#);(r++`I|f4%%{J{t3G++em|+yx%@HP!+3XsPI_yes`|(sY>i>UV&6b4pRwv$q@JRQ01ya)#84D?PJ+(qbQT#a-v+^aiird zCM>pi#Nj{+yp{VoYxEE4aUY67f)C43Z@xB#i}G~wwots8K1ZX}$w#8f)Cg!x$tSY} z1NknW(!l?G@fN>BO;{yOGp8_Vtnf7;@Er0vh+pe#8nJ=T?2lgTBX@z?YSJ3UZq{Wy*hAIq8i)JmTRHN_qwagMWpx<))zZ@HNpBGsBJ?my z6cOEdO6L`euS7d(`9S8;psEDWK9XE=x+-z zAXM=WP7FX*;5u(dfT@iGr{E2BRt{ znb)Q67%Bu-^I8^V)af!uIG&o#bY`MIPKs1;5_ol`1bQGTX{jke2ao@ID%zr4hm|Pm zOLkn}aE_Xu3B&m{I=1`Q(y%TXfWa#TEtn9!{WvQjdF@@5)DQqe0{w9O!w4n?F)IBo zYIv+ghv9k_6J@7lHx1LD!Qs``_!D9~n;#0dje+Z)78wZZZYO~}g zaXgmjal?nhlD7g|J10A*G0a)NEucAc<;q>37f^fwt?eI#v;;AWpr>^Db=uVbfo82!;+V;V~! zVU_3weQyM_3`f27O6xcKjA)v%C-i~w0y5mzr~U?e)Jq4#)L43mty@^Y_^In4Lc$Eg zi9Xlx=d|xLQ@^b5Zf#lvtnsJ6n$)t(H(!J7R%yHowBvyVU!Y0nx*Sjl=wZ_tS+mk0 zuR3=Y_p#=|I&!PSr|B$a-jMDI+MaYB$|Fk#8s3n36mdVX!bOa>UIAI{CyA6t;ZUgu zdUahN!jBqINE+hufpWO?k})stvE)V&wH$gq-yP32W%6SB!(mlS=|Th@UM(@b(yRNt zgFQG%!C!$r>SN1?4ZyCDt}d(r&cm-b25API12lI8rLQ&{9(e>*;0NQ1VE$uUoAxXW z;$ZY5crIu)vB&xXDPDbEQhf}9M~q21>zltS3}JZ!07d8t{K8#cK0klq#HsA=40gp%oWGFDX6`qA?1;O zV$fM|11Zg*nvMzh1!Ka#COo!>0v3!AOuSCMYCHfmyJX*yh4iJTg7^yz23G;8$zj)u z_Xri{6&EPX+``}jyb}r0Gcq#cQ^UN=);!%mGH4O+pefwj(JgT&fXU2{bmFm?vnwKe z0;t~*^W-#ER4I+Xd$s8ia&oLM=!%)RW zSs8P2TJH5BrYl0*iUcCfyYh;{u@z`@fqBbFFo3DzlKAin7CRrFWPSe7_X{CYvZB#yH!EqQmZ~J2{AG zy*el|c(CggB;msS@U^PE-sZ}=_;XO?%2}C^O)ziKqmhqb8r+R8Tl`3NciH3v3n6z+ zkB%Gj=KKt|VBvl}5dvEs)7HMUXfR((=VE{XTUR~|%5jG3H5l-M@xaVB#n|s`S~O}a zY_Rvt9oUxB<{1`OcUHRW?$})RSmsb&A03J<*5|btG@x37Q~`y%DtDToTTj^ohwde< zL}t{iNF{(){9Mve5b@!ru_s~sn6=QDDk4%g{H51UjiIqsN~Tj3j)?f$xOvyIw|}i8 z9N__sI3jG$T#({~Ixn!0z>O)!c5|Xt{2?^S=i0V_$VEP<@8Nz>t#D6I!uBPhH}J3k z_Uu3x^d$A6PF?QMfx%4Zq;fDng?U>K#q~YweBDY0hbR{@H+VTbfTQ{nV-_YEKHCTR z`FrS)KX2FS0eaETgZn2i1j6kL4TAdm!lw z-{Uv?X0TIHf%aL=uJ#dsH@l;hySS4#!0o3FBTmup>Gp|l8}vBF=C*|~i5lkWdIPQS zl!rIujHK^9cG$ z%7kg_dfm3n2C+gBjtbD7Bo5;H?nv55NSZHF3kx&#ERPfP&GO1aBj5-e%c!9Zr4G~8 zw1nIoNZ+&nm=Ghd(cR!O(?hn z?<4k%{eaHU6NiI4)M*h#!vD3FvMK#+_RV*DW18bIA>^d_)|D{{bgzRFDE8}_`jp@D zDmXAIF7q~L*#iRqRNOfme^9?;8`6hr5GXY5&j1VRHy{tFr#||nLuI0rwR5ri+#T(Odow&Ow?1@wszRVplb{%b^9o_U41P)U z#FMV^;z!|P$QpOtr%e07 zYmNU|`0b9&#C36UdAc(5b-2f7ZU|&eFrC4KzNS6lyfQEQw@48(-qg|#!?IIvL1cB3 zjeh0u5F4qgxxGH+Kx1I+;=g?^CAt5$qh#Vf(_Qk9^CBAF+56^vnKYs*HZncFHj==1 zs5-AOUi2vmpkBS=-t6tNn(Q2B4$WO~)K^ocp8P+3HnBqPKmUr}7q5xqNq=lWHhZlr zh5yrMdlbFICqj=nadNAFJcN5Xdg1B6=w0()H__*Bo{8G+T%T`$Nl*OF9G-da_8ig| z)uR`QzKNe=I_18Z=X{Ydqx)s}PoM2HCEKVD0?%Zen=@A$#OY8Eu;$nHy*LQzjQBJF z_1WwDkbuSCgKiQsG!?64d!Zfswk5=@h6K0#aAI24sj`c!anWJ9NM zu+yYoV}J*>X?m;wj+9e}CS|#){R=izkl1|PN_{(7DKLABwlec#QtxXRsWcOsex7n> zQ+yaJt7374J&kI{dLsCrUp3Q#V$R1?BC@iiI%SnjTO(|LhcX?r*i8dM%^VaP6uNJw zzv7~@LD*1%-a58rXwX8YdaTW#CTF0F>q^lq+OxXlMZ;pNt@ahGbD4`8N<}fJqMKHQ zO<3DCRzNb1+`3n3Zsku4zK0PwYz8=1fYGtRe3hjfYx^WL*ep(M5nHErci(UBf$@j|-u-|8cNMcKad+q>L`ZYs+SzO`{o ztCk54*p_#M6y#{I6&pwJPEr|30(y7W%(+CStB$jjWf+O9yq}H+GeJrxH{4xE4Ab80 z`U?FeeXk!tlUS3}u>y5fZ}DCLOBwR>PGQZLYCX|Thj8A35tPH;`0i%p4F2BUA`6+i!JAiQ&^{Sf(|(JSYRPd!R3B@U`x&9~$&nX_LYf%nRLk}O zYcL&d$r+F@O*9v=P4`-uB4%f1%IX}V<^TZfBAOybGD3|$_HK4|NZ+P|gGMFU4)7Sh zTJxI+5ogfzxYQw@ldk*t$)MzXV_WcoAY^~pWo$_~ttV?FYk;CYOTL15Nv7XcA^nsV zeY}z`9A84jr2PB8Gj;+D`xs9P;aH218>Pu-=K4m?hKLD+y6rP8VAqF;=>jK*W#T}< z%5=cVgZgPBTWayf!99Ti>vnl75nm;mM*|Q{R1qd?Q}wIl6E+QJV4&gmM#5m|9Ama z^y_*hwN$AsqOsQbh-v0I*BS!uso&@6#B-ea(&87v|68r{q`#=+JF=e_fdqaVKHe!7 zX)_f_1hhj}MfFm{Y;CW&#RR{h_gKx`!R?u?xn{MGdkvF6pl2lL)Ij#nr-=~qllI4f z7htQeA+f(}mKng>v>l}cP|>K=nax^S1G`rS^+H{VkHua7ZKpz%{xpzdk&8H+4Db@J zXoEAzE-QdJW`AX?%*?~ydq)~XGXheQTGGCvs@s7~!WMk{)C{j^w^ehI2PY1CpurSw zN<3{6(%;)z?9)ZN?gAr-wAJG{w|s)>ldcZXvpq=d7wGwLy29~dt8jgYdP2<$ZE4p^ zVlxyWCrDZTA*3u~Dt3|`6b%gxU8^Su_Fw>{gF*~n9V^mKfzod9_!y6G_}tsqLEo!0 z+qkn`1Y6~21iZTvsH-DPXa6lZ?xdMiHM>1b5YPW`0bCGd8TS$KpzN?fq;fl$%7dk7 zfeR4JcaO{q8}Q^gfK~#keDnSR(3~m864b*`iNncRgqV>h{Uf_3A`(2yo;9e6;l>0b zR~q`;oFdm)0VB*1bFHkTwW!r^W5qw`4#U(R#WF59T=AFdbSz}IWp_?WuupaBtdx|B zy{3cTYdcJ&#K^crOEQjFwh5D00P3QBs3KJ>!%y` zo!@~mhWO&PK@=Q4p&CP>-8QB$G9sbadFzS178?xz@k$5oGC~bFmLN21N|C;v`4`{- z&_}!aGM@%B)Lz}*hG7pei(3Qc0(dyi=+Ngr=y%XtY5EEKv)mRR@v<#tio9(?$c#v>W!sH@-c{5ur}g6;4Zw9x-!bM&s#o} z3rL~}SP>j#$9SC^@JM?wD4{ejw6MA2uIz9w3Rn{esoEcyie4TlcRygzAsKAOxB%k8 zws7|Vt!tnetG{(&MWgwQrx$+~nB0Gok2M)@9|2LP1gbz)RyhIT9li;g0d(EfTb}>5 z8Q+7gwi36C>;fv%s_lrDfzP~>rG1M$LbMLOlov(Cu)@}-;bYj zo706&w0pAKj62`3F6c-k-gi^0S_khhnyvQ9aIN|fyTmu`3h;{eLqANU(2tn$>Q1r+ zrd<6pQXQ-q;eg~Nq(Q`ED-8h#;JI?ODoD-K-Cf4`#}DiJG3xxq^zjEadpVY+Pdd(X8V$=38RiGvs(Z^XBN>93qM)%bHFn?TXm5v3_m;r@^t0S`8W zEF-+q@)%;H!ZBeEjTy+(gagq4e4IjNdjbp>i^1dP!A#h5*xFQJ)_2i4daKjx;lCi~ zhVI1!{C3O+%!d+4WHKi`0Bg1E{rD<#Zt3IVlA)hbA)KyX~cMO7`qACO{><=PE?H>vfc)^CitcI@8 zZK<@wMNKT#l6o&1b&%-Q6y#zx^Yz%STCJ)1L0>B1542;ds_yZV6BovB_2t;bJ9eY6 z9>A!ObnX zulk;^6@{5T+K9T%IFz|84cSCbq%+AJ?Pm?Vp9ev?xqV2w zRsmvF$c?&>U@E^-56%~xX#ol5*OPMiMxtb`_-o+3wC=nqT69nST!ATyyhqVf{yPG? zyIAq<_Fb!tCQBrU`a!`T(AeH*bwDWow=Vcm76qcr1u%>r!2)}ZIj6*jNL=6Hy$_cX zW#uC?Gp(W`Emtn&=|X)s)M)g#!ZIVNfnX) zM5mmFpKWJHtnbi3;)l)kw#CIq*0sN7Bp`H1LuZX8+5%0{SS4NnU8klX;%!-Ucg>Kl zWHKg_5WHq>Vo}yta4IRX#ujjn0RW`40*rS7wuTP~*}$Z+C#-gp*Lz-L3+d-0?(G9d zq__`$uqw}osm7o+&f@V!qNl(K#Tqe6m{E_llK_5gGuUTRkQLhM_sv%xA76+eC!axi zC=aq`R0M3=qZnSa9}Hyy)TN4H*2eT*1oTM&$ffx^z_S)9llXNEDly_12?1AJ%D}Ch zv|`Ni$RFhuqJao8sq#X(8SXs!DC{Da*P2u8_dELOst9O%Up3BbXIKEl;9o%4kV8|{8?HAsyn$?o<&BC3S zVph-gXlu(fAZTVUllHTWSYlEka%;5B%*~2wlc0+r#Dr{RRsNNOWwN>nhQL7@Bnv&e zoSPU7j!v+1nKF-{$V5YFmoAGr~AlZ3maTzy{q5mQ1N{8DS)MMaO|T;1t-u;r6C6c2162dp z<%NN(ruRhqWt-2z8Q)S-kW4^SqK6n-=Nj98f{Q9-;cm`UKw7Y0%4b++TkU?pD12w$$$M%;x(3 zJnhdOaXIW6GroBk8T-fQVFcPW=DQtQL-j2`B3jXE*NACJ@q&DMmW}2lryg!WI~gx? zFz|Rg{uuarH1*)xALKz2FgeCH&&^xUlnT3!L>zmIVZs4H)GKtFgIEaf%SqU6Jx?A(d}HrW9_^n#TiHE* z8hI*R0V+GJG_*`(Ty>M%UkNNK_f8x5EzJ`)d243FuUlcUeiq0Rc$(V$-wLnuz z%vw@DVzQPS3|&>`qJz9`Jud^V^6Wfc`sWY_{lOkhib!R*R&WMNY286N)oD_s%r2a0 z`%++gsnr!rcazOodG&h$b^jhx0&@#I*dM7v5~ZdQ zCt4@+8N<~@&~#XS_n9fh2^F=E($nrpX}UNwTNlUF&rKnz@prDr$n(DeEr$OMw3r$GVJ0M@rFqiPm1K? zvf@e(Qv5%N@+SpAV(r(yn}3iPPQ%}UA5)|diYVGUD+j>?$3W9>yKg-uOMG}Kq1}}Z zdY^uuzw)>`xZl3{7B&l?cAvk_){pxWUp4TC2|12qXl>!#g1DgS<>bqWZVmFQxjVA( z!Wv&qXJ=(ZQhWRHdC?1v{{%eJ-#9qn=h%Z=gL7-?SJN@Sz6moC9Y>-UAgUJjYWC4+H+{CRR?fkUKUUrH!VF7wwgIJx&21InKP|uM7}<6n@Qn88LTr`8h?gk zMl3|=4v#M@YwpI5kbEu4Q<7y^D9g5qM)7Yg5@tsi*mRn-zd18uN)lxwlDS$lN-B=y z8@YGA=Svuy$t5I}`I0)uEfd zT!^I#hy8JWihL7pU-&g8Jwlq!}aOMpV3H5K$0;505Yw1tyU|4L>A~U4|Rtr2Xx`A&(D63Jr%=AasM?;zx3~` zwH4d9VWEZ}hFzuM#Tfw@43(l-y<$C$S*OQtXalto)-ppxsIjEoSydxyCR9wMe_E1h z9LdWU0`*a?I?0|%(GB#TGi8%%-Iy`+hyXv4y}U|AzurYm#xko=RlvB9ef1DXO z`TGSrgeeYL$F$@Thl;EmtyfyS1OdKv6lG+RU2?3Y)R-|vC}f*_CP!~3P>gim9WfHo zGB~@4&o}1Y*_E;USk*u?$WlX3X_V|gBS>r%$(YvX$I*vAY}`NF_G$X-E!`5B5~E^@ zB32xAK=TaQ^!nJM)d_X0V^c+|ca(j(0=$_h#5jr&U@21i;E8>yvH7pCu#jy+UlZ z|F)l>7mC_8AZQiGl;P7qO_>7v*crDo0%5 z+j%IX3;kxwA#|}8ybEGb!SPO`9DnxYx`&QpVB8#g0`c&JNp2Wd*jl<#_`!7NVk9@7 za)}@FBsX4nrH$p@0B!FdQfv;vjzr>gXgh*PO#{6Pzz3demB*(>5O`;Xz0KI^3&tr? z0UOihfJ`TV$t>(raqjn=hA~vLh`;#G{SN-4Zoy|*1AVOcu2>HcCYPShj;`R3m|1Al zcL5y)@L`c)UL}NrcV}pyItc`#iT>#;jPz&9HqY_ZX=IQGn_L7}kINp-6`S2KHB1ob zZ>;02@pye8bC(d^rIW*~t(X7ihKyoPtnDafcdyRfd_a+Z)U2148m&}X)&Ni0#gf1M zMMm=Y0ZT1aStr#SuU_U+J|k%7BU!qV3uoWtiRSBN^gXtjvCg&Pev>1>0ZJ!~&ZGIhr6!=ntHIDsRB-~a4V+tvX#9*W58Zm>6eCI0zC z#g^opyoAF@Bi9L#pD9@F&CX}H(X0Op!V{telaKZVM4vh>7c@FOO2+@SdM{=(_L7$cLY@(dKh`^5e!R0 zfJc<^9MMsDRL^UJsXKeo%R&Vv7$+L)h%18 zm5*Egr{w3xW_c=ecB;g4kWno!1{eV&5+X*ILme0fyw2|qK9~7a&y$-b2ufl!y43sL zCASh3Ny%rjcd@E3|IDKE3#q4UDL(+8$yWtdyBe^WP$PGl2Bcv$J)6*|)I~Fn?h+M< zat#IDo38;Z1tSjZE|$t60l=beUs9H85_}(r+dV{d-`fB0r`TIT`m2u>x1>AeFzlH+ zn)9%|yN~dhk;Pxa=%76~dk@x1#l(Oz>OdB+@b#RD6j!VQZFz(Pf)ZO6@oBH5TrfIb z)um-V5AYMFy*1$+45e3yA?I=^jBS7ejyy=t05)+1uf9oYX}D-{0?a2!$jKHSI`*X2 z1IOrq0J(T1?x2i;p`s;Dj)GK)uJKv?@LYYwdY+L20eny%3HUzw;uqgB$iUXIF@ICR zCMlUTOuqRBEont0V(v4l&0)VW*kUT+)hMY%ZO4otN5LF!`qaeeyVh#9;3Gtb#3aO1 z?dv+YM$DtPRq53_%U2_P+e&x}qC4Ul-Qmr}Q`=vn30+Z73=#bz{ofnZdw+&y#pdN)xIozub(?eYd9a!4NE&@$M5tHPr|Eh?4%PAPM_Ql0L)g{RJ4E` z(o|K_w?esYu|{t*i% z&eijffjG_2SM(@i%H?24PTFYTIWdkyfTh%?*!923`g|3Wouzqj4f+XO>P;p$G)ikRrX4Ndpv%a)&tW|~@+#=ZG zWluWcjc9Z_TnM)TV4-XZB>n0sC#ryFo=_#;+z$opOo-;Po-DCUtl+}Kkxp_1h(@MD=+F@e2a2B_*gv24Sqvc2;vlCVV&4^g10>#Vks zx9hi7Q_J5-1WMA;hGe9eW>IiVk3tgM25ZB=yT|i>Qw*n8% zq@yvRTb@(JIFnefs7$%wB2zl;+^lA!Q}ltNnx~6q&LqSZ_gA5PBrt@B<(at{qpa}G z&F$=5ic0_ZIC63E_NACR?i36u#2mhQw8SnndU0eRINNX4@7utoCA>xUElya}{(~iy z-$Sl#Z(^C#ARPJQcWYE^@@F8!_15UVD#1m;u&;NVSn>)F7)Sr%Tp&h{KyM<4$RvX z8d|G|l@&y6oIJL`PC&)bBIL%5z9{k+li9JrVe4R5d) zuSLAYLIC#_*6{d+)WR$v__PM@Gw2wcYcT?(S9hXZ`lq^Jm}AY63)txj0za2yUPwl3 zVf|9c3Y$y}0Ez)|c*jP%zkt*f5O#!XB=`++c}cJ-MGC=8>voi|)%MO65WkD2)}})uY?Y*<}lyb^SI70u8>$VF(SCXRc5ez+9YD z1^W8Z9$>CoWWSI(z}T$H5O*v=iCOU0t6^)O1%*L-uag4^IfIJ-`fKbEwql`g;_c)P zhW~7a2AW1FcV- zSIv7IvT;Nn76uZP;$}lz_l4N^H+@~MSUOpOF+oCngf#uiZ@h6+^h%yT?L(SPrdJcJ z&(hZFnslb3@hfv6gO0ai+evCN!U%vCd5(w_Rr^axV$EJ_V>wxSDT~K{;k2etZFXpm z3GL&}h*}b4``q+=9af;DH=c^-sSoE>^JKHq+Kf&wBPM$AVjvUO!vG0(Xh?CZD-D3O zWjEv+$$`{tY^`2n;BWxG)er?TNdw#T)5`Vg*h1EbZYKezMO{zJbs*76Q6|w_WN!S9Dc}eb` z;^Cla>mz+%bG82EqmzVv5&1=c=donuEuFo!D{#KO$xKcE=>;z z+`6b#- zwP_j;yApAGQ?^1TS@nk69))(lg-rsSSJ~Np6yaQ@aNBrd#^`QfqcqPw#s=ZAUfYAfcyt*kl?a~oJBu~q8S4Sfa? zK*6LjfZAi6atwrr5={-D+So5s(aBRtA9CT&6e)&H8iGtIN zVc&4#2_9jyBGJKKL5{J1>97RvCF^_8qUH4Lyd1mt*v_??*+a?OmeP%jUETT&1|Qk~ z`Mdw)q)R^?UDW1nAb@9|CsrdRTsqX0K^@Bqer5ZC^4X7afAPvJcb$ioS_f44$sTRL zba&XKK7HYo3&fN)x+gp*dckqV`cwiY(W;CRYN*_Sf!bZdSuyVJ@QZV~oyDXz{ZR=y zR>(cZ=*-Ax*wcM+AI9|xT;E-(!cMH)8rTSh3!cF(^^e4!QCeomh3ERLWs16SkueQM zB%OM_OlmUc6hmc%b^-V}=zR8}tjGl|O!OW?S>9>lnfB~-ptaVs*2KHlyqJ@~-^1fA zkbiXs{4_{&bVh&%1#s5*({*r6@i5rrZC1NgJ16W)Fz;RZ3bSJ@D6X{((NDH#Yjr=# z7Qh=yv8}_KdzCy|(luS#b`AAg+C8WE7@h$S{QWK7!V|l!mpExn1^*jt7HGkWcFCy% zM&zmTY7~Yo>jTW51Iad+!)#P)B5VaK9;R9d;L{rWVXs0qQp zqB{KS@KdDk?msOcwxVR%Gvnv-57e$_LQGfk}|omRvxHI{qaR`i*7F>`+Lu{*F;Bh~;l+oACB)w4VVJ2J-l-2X#>H7TseE;nWnDSi*ho%uBOSJoOlKY+sWv)eaz3_`Aofxt$d4x#i zW*utY$586l15DSrU$K;0Y zmsRzJ$J?2t*W`LmGf%A261kulB08t)Veu^2*=;(twyLsla5WXr&nJYLv^H=f2_qHJ zZR=VpmjV-kgR7>A_d^_Au{wEE*7a2*=+aC4ob?)0s@qjJjHDFcBgM)&>>1{)wDBa2 z@%BN=FYQGSNL4zhSvbFOzaOse#k3iNAd&ARC@JK6L-r{?rwUwjJ;=RL z%0dL!!v<5s41Tfd+}7ouND3Rql&$~j2cnt_cE-(i>F8VDn~J=PqftUDmD@G8lO7qc zDpL)lcd12UVs>16>%FknzDAYcyi>QBTSiT1Mz62k8@05M(Uz+;N)I%k_F&XTY^iim zFDZ8ZH4XbH_YQ9{G~MjSjb0VDQQNF8l?zB4&zWyWAl23<>7}D*4-G{)t7?ZgI;cGC z9O14oXgAtx_BBaxitwU2Qp$oIebi>QrtRAcEBV1u1N?fo7YSFWI%uObnvlz96)DT( zM-5vwf-w*~9+BMZJN+2ke#zbrt)INOsIXXOdy=hyKFSu!3)urVs4EOU=h*a%1jzAY za}!Coc_OgK6`NUMvd~W!y5qCkQJ@Gv^I6bi#)za}%m2y6)bOT4PDD@ z>HC1Jmv;?Ib|qW^Z4Fkth0UBL@Fk)n$ce?Mp&uWf?#iH3(UM{$)?pbgxFOiIR5R~J z1hK!77gt;F?XdstG*#C=H-V5KB^5KQn*yN)p?tmUX|8Zry=eRB3DgtNkY;3FDAU5lEE4Cd6WKseLq1y}-<`s&By$`{LoRh}0KM{P z>hk{bj7&5C!3lxJ;iMztbhXZ+besn9aG&62bl@H8d5Vj#DlIbNWiUHNVO`AfA~ zkaj_09=Ghn2)!E`$gIv3t^4sfK0kj+a9hlX)q%5PD^fgUZLeIjKva&9J9^7C#K(x8 z#-15J6cww(0_?J9!xrQcCelU8yR7U$_rn5r1Q&oDc}%ATVF;tq^`Y6Vf@oslx+z7ML&7=>GYT_wiayhX1w-iD@nuod zPJL&6JIKnm&{cIbbp|wb#Lqa(06b3cc&6O?g7?s)8>FE8*zoK?k`7$t5B5`V=s7s3 zjJ5P&f*iDO)tKX)ji?(BGv)LJLCCE2zcB|Gt5S9OuI)?*?_B(#Cg0{NH6i^6MkbFd zi9Ixt>dhLeiV*_|v|r}i!AxnFrXL0BsbBOlQOZ;-j}+>s5j_+s;SeDiW+72+uvzF- zoKTl-AGzlyEhSt2`gV3oMeSw&KK~3=)LXL4O9zLxD-=%Yd`NbaG!KSqLsB1kmUn1s z%8eRSw4>H=&ww|%k^*a~)Q`VAjAQ`Jb4swWfXFcOGEJp)-+^VGD% zBx*|h6Bq-X%jb@(l_%dkN{;iG!b=Mcy-kJPII6DkFBBs09jbRP^*V`H`EXff_HEhR zu)~{WRnk7Zb=!GWmX{}G%mC{r6nJ=2y{g|CK1Sxljb+s@)QO&XRr4@{^U5avkf&tv ztJJ4t5xkg1F-7wJz}a*!&~C-q{iGKw&E; zU~I#i6ZC7PNro@1%Q0+0w^dE-#e2*((q;h>!LC9acb(?t32k|PN^Fard6~z-%@R^S zAGXYX9PI(mlu6F95&ROnY*QU1#PM~^bO(Ola$36m`8%nZo*0N|srJdm?FsE%0J0l(Djxj)ax?0O4D)M)p2WBy#6suzWNaE0`-aj5WFJJcn^=jgpgh=&L zZQhd5^h4Cr>og`95j)r;SK7NoxA?3kBSCW|F|x8UGIdN%un54XhSWeogo(8y6IEmI zY=cVn5B3a}xPbeG!_(I$*|20UR#{ zQSMx9yziItB@d`kXBt#v^p6PIHC{qqP0NWwrW8C^Y7Qgt?D;KA^%JPu-=EAM>hmm{YCbjOn%nAL6>il#ugTj7d^$dom0Mo?3PCc@|_I?=Ym z<}`3}if6CvdGl3eO@fDWypiLv-;;Fza+Dl_X!0TC3k76@Uo_}Hq_6Ku<;}rPDXR`xAc?lvmJ;@7%PwRC1nzvY)sKn z)az3qMTFy|N+mcsNUx(*KwEx6zU=hq)I?hJ{BV)G@SQD8_*7_c@FcrAdPl8z9hnpD zX)0$WGb?;KIq8hbxx>wNCko}^#~-RppqWhiM3KJsn!L$1<4&Pf8yf7t3(lL%3o9&iBC15;W zDqwmtveoJt5VKb+S&KT<8fEecQ^VbZn~Ac`7O&uHbdY&qc4CwH8klomjqDqbDir<_ zVkO$_QM#uv>fb6ORk&X%_epYhqxhGwEQLv^VqC?0nCW{SGH;riZFITIW77G^4J;qQ z4PSp8WWW5kMb}0EkVY_HeXao0#5kAb27i5{b2c=8YaJy|l|TIPLWXRx0{Zwi=1AN< zh45q5Pj<#T7z~AQL1py73=37m%|SrS4vjkg*g?uV35k!%FjsLR&0@8rJ!xhITIIBI zeU_F|)~$F>iEqKAhLTGpE|6V2vx_!@dG}X? zW^;2OXn$E_*FgLQP@t+65*74=!&qHN*srNWZZ3Y}`AI9%ZXH?*_vFXR)#t83+i5P2GTmO|E3Qb(&&vY)xqXonH zU^os(gfFKEScXnPv;`cmJmODl*0gix@UF^{WPX|eK-WXB3dzEoq%GTGn6?!dAV31v zEaN6uJbyrfoLao^93=WHRrq8`LZ=Yv9i z^Tu%XhpSgLkJ>J?i9%l@`GWKH3DnU#EX|vhB!Z=Q5pima9yV-{A6Zl-SPFEN4&7h| zr-M(h*EH6il2^6M7K&3BP1xBwY`fZ%e~T~*`b}giQ7rMiOg1mhJnF6f#-;&tZMWPf zRAXfS%`{#7a7aPyDol)Y$`0zCyrU^a4mq6Al_e>0I_F+GWBh_Bkq(;LeFUPEInR}@ zw0fJ0JMdUu<+IvVQNzm_yG^pv@#n{5*4`EJm1v(D^(hE6zpmMn^%|>qr##F*T2nQ~ z>(ie%1*BI*BrZ4hZ0uiilE16Tgq3ExKV!vWG67lb;?DCCn_9)xAZbPd+QgodhoAQc z!=1v~z3rX%SEHn!1WjCP z0&Xhv$d0J1D8bz2Hs5kHOx5SiQQ1S6*^u9ZpzO=r8{(_Q5ZU?BRf(mAOov)M@s$=t zRitbo$v9t{lpq^V9V;_n`V6*cdkwxOx$L&9onbLJ@pDYo)0`0_$)w0BJoYTup*^ov zf3@o_4PvOCR_vj%RV0KIvfD6|U0#skT(ifz&|Y`aNDx-Uf)(j{KBIFWO`B)YyM|p- zQlpONf;r37Dg;f5KMbmypnQ;lhr!-DsezODkzoLMhP2D4SzK^@Y@$L3HK|j?9BWsJhIvz>Xe1Ew&U0 zmBqPOm+<3*=Y->>Qy~!NH}@XKuPLE6HtQz)^~lNr`}Ua#anl5d#KAxkGMd}pLreM)&BrV z?e>^U&-dBy?ETgDs%wFG!t&o34C8-cFs$tVhuY3-O-aYik(3?F+imKmj>W9M1W%ot zoZX0%s|wiS;bitVK!o~;s>Ee}MOqWR*%j{ujx*;r07V2Ot_j1`C0fQve7v)VoV)DX zKhN~DZ~8I52fc~M{61`)Z_j> zbH%ISZBxOo4n7X94&0nRWzUQ}y@_Z41@M1t%fPRRo8ddx2LDG`q~WQzApVcAX!8`i z>FPQ7A7RnZC35gT!s5Q#e}qN4)ALB=(C*jVHQ$&K1ql~zo~RL~c*9EPW_4!SCxS`l z&tRP0ZRX=cv#i0wCoOapluQRq4Y%7hBANxW*$ETv>I+?wLWciI;QyLaiey7}jXJIE z1doRn_T=d0e)|CsaR}Ar``E?p`=8>@YQ{tORB(<}9>)gnMz%J06;W&^phyM_Inlqf zS+8Y z`r6BkXSVWl5!p2t&m|NU9v~f|fGC?qu>sy@Dvc#-w4JyCjRhj}8gc%KSKm;eVA0V= zi3Wc({hw?m3euy9n;tak-LLl9JE>V6FNR7k=e9HlSt@IWr!DT;2ub=%CU3rDiQqdO z+VzjS1e^zL2|zNbB{p8wI;1%5*-q}nKkK$Hmt6!W(RAj?>qcm=7l2d&<5B-@|F*&Kq1F7u<{G(xN14~lJ1BBP{#;pi_DsVx`;+~w{d;jewVp& z2c>M&QoJs#B(B*a1upBPqC;`F*#55mB9+K%#8j&QiM~G$UJ7iyHhO$MBYvC}9Yg&E9}uP z*LN4s+D&&`2Q6RXOJ+$R+A22d3$`312f>Etj=0=4Sd1tTL2R8H{14+#%BY4*vaz^P zpNqV8Ylu$zpb1{Sfl?S*)|NC~lx?ss!1BWhgciUr0Tg*^`w&?W;_pZzQn8^(_Qvw( znu|zdgCT+Ku%gW%F5ES25k}5HKLX`!0n!{u^q&Vn@~SSa(-`uY-Cj=MPxRQGQNy-FlK z47e3}iY1b)2chp(&*RA^^Y#Im8zz%P|B5%US805p?Qw2#23Vqrg^LRypv3V9UHQB} z6umlx@CKqFezVtOGU|GRM7?`O2?4qBV_qg00Fp9ixrBA*sJ$TM9)zNeES71hwJxvlh=ljsMmn$b*oLDkx?Q8o;)R(^!TN@FWZbINiDu{DAun7&RACH{Hl< zn3aW*SblxJwU5P&s&rFBe2V9upd3 z!)PQ`c~cIdFh@doPi`nr_n{pOdari*wFpT+iNPS{4p?f&Yj)@GHFhY#ok$SttV%!_ z6UqZ4E51TRi(U>+()N&|Rwh4xB85pe_ByhD$#71lRr8xua$M5Oo_<^wsR5@{mw1;t z{gDf7M@OoNUQ~Vtdm8Mgjg{j7P~j4Ee`+kI;z9nepmb8AasVqhf!J)f}pitl?d0-JO zjRn9l?IVf|QP4rh5chx*Oi){uF_EUT7K5>qIWimVZl*OL6_oN#GzrpsF&Kyxu()h1 ztvnudNgQK^!7szLV=SNGNa~C`x=D&bR>o)cOg`|N7V1Isqm}8$8%dy6AvCX3(u2J* zrT_}>Ye4!G{d`^Em(g{8tQn)q0tvi+yejbQ7HcfqJ%0OpL`G+{(TM?f%v??cfzmca zI&HE(92+mk)223WG^DdUz-GPZbR1c=!` zamo#*4tm9?SCHq9D>1%ApHfO>(rXFscy;P7KO52x%vXv;)818Vm25E`!>>f0X%JuF z*D7{~l@anA8N152UeMZJfROuMbU791U7mx@uRg zV*i^a*};nXc*xYREc3t-BXnof(xzX`bs!(#+;MrB_jM#wmCLpKFE9dR(_)bYjM*Cd zBXdDzSFW>B{+sAzrfOsXDa2k`z8s}P-Rhx|_P@}{lzLPCMj`hLl z4IeRlQrT`pwqN5dz6kz}p0v03?rO}!Z0 z?IEk}+Nm;TZUo1iL`Q$M}XWdj{XuzfzZ5?r4z{Up*wbo z`p(c&J@D!19{8Br=hIB+#As0@Ykg;H9RIY6bg(=DVy3a>XrMTvN|F%AI;PW~J28^Z zg>m8UbO{V4{=l^q*ceL2~F?Oi?BBu&;0fB3L8($H`VlH5RTHPnWlIRNQlU-v|4%X z`y8SiJ+MgV{=$BCT>W4}cvBLrBCWT4cUrigJlQ*^?y^U7PZ) zG_d~1$)|Hce^MA6K>}O{)L}nm?rN%{KS>S&&(;~&Mvd1v&BDDw!2rp-kCTtOtxbW> z0;@qBq#zQ=@clsELY(kCMX(_`GH~U4yg}oB;3{kRA0G8XRS8bct4fRN=?LlHa;mX~ zeS07{HMSe2rAo3%+WZL>unG6iC4){!W!mPD$7Bz*5KYRYT#%qD%z7r@L=U%i>kyOp z`UaC~_eBc|%R^*lC}}JcXygOdsJ?;f*kun86aoRX(W1TN-FUJ zJqS@093-t)PadTVeV4p_`rdpMS%P5J93c2oA8}=cumQ5F45#t4Vlas zuPXzluvN4Ed6AOq!c*)y`Mk8VL37^dE&>01!vK0c`k+qLEvyeca#-M^M_+ebDwI3( z<6k@6&+uyY*<_qR`3kxHz>_zjckIXZ`TI<8&Us&;#QuxFmOa*j>uyMt zCv6ZSDAf1fIO9*Hy`T@XfOrPM!{Z7l&JnISDZ8tYsw(Z4Jg^~${=%9ALl)@7Jy0yh;lP#b@U+&@T<_> z5@v6v0K~nK39%jXV|80My-9oQSDD5VAFi+yhV+*Ibjgx{OwE}8%R#O&1X@$lrw5xt{E%oO8>$ICmx_^ zRJd9Xes=HHQk_wP-Aj}C(R*S3u4yrHXbnjt=>7Wd8fl{WE=F;VZV%>A_`N{XeCC+D zwKY0*Ir2`~A)m$=9#DaV-zrFVMeciFRxVl-CSshrde_``vloBNdp>rZb}P{`oOSQB zrd@8rrwO_}ysjJ%i1O|^oKHs5mhvHfAEI@ohr0O8dpzE7(>u81)0x%=Byd}(Zoir*5B=9C3!v+s*v9kiydS-p$*k*7V$Zq5#Omm zLr5Vhu5-JDuIB&j@*4N*&Z@#QC6V;w=x*=))6^>?`7r{9f-Vg~SB&dj9?%1#Z7ubj zP6SoeI4T+b;v%-6(z*gJ|0PB9Eqc27JkNsuqEFe;hbwtd#ZMtVtm&VP z(vDjZoz~r%DZ%mgvhoS5$W1XA8a;o8fd(~lUjp*4?P7^0vLEEQGaCnP&2C6-R(A5G zg-WkKK#dek-Zsu5laSsZM-~PiFBkRM25=-vIR90s)%Iof+xfU4U8G5XPa=!E%Fb?R zWgWef_*VK%Vt+5GdIOIwW2dh|OlyXLF#jG2X~$>L=0E=v|C~_KPOZ-OzBj+gBA6i^UQM%F~N+0RFum1Yrri>zy{fFVP5tvFI ztN~c8)I}wrW(Pnv?Uwcw%HxqwrxA<*tTukwr$(CZDVrZ%(uB~&HRhA_Nl7vb<&Z&AL4c?Zcgyn zrKwzo;#5bf8E+o%yuVr^<0lvz$jq2ak1_tZbBufszLqj<9GvSZQ<#Xy2JpW9ucqty zdikb-E1?1Br>JM*NuH;JL$u_NtXGLotU4ANQ#voKe zYx2kW`-lE+Q6N8o6`7ycb%)vM;6MKw#P`xa#(j(2}M(h-dt+rj0ajRPdDeV zO7gedu67Dc4$u=wQOHB>mKhdnUQGLy)aw=-DihS*5sM=q?PqwMW;GMljYu=&7yHfr zGb#pF#NkV&WyZ?Z$&1RiB8eN;xd+P1^G5vT*12pOy}NE5pT(5Nt7y-UaertV0W&q2 zHG3ZmqO{JGn>6;~F#rPM_qE>;@ljw@!*Wjg!rI?%Q+4O=zop>gO>)ile$nHR@e2;8R&4Z$F8~BFKxV&GrG2dpl(Ux;EMpiO2eI|dao;Xq zp>K5MNLN%@=N7PxxHdLj{M6|cxP>^-`9$$MpY&m4&Cik(WR|CBPaQUOcaJ*i6}Xl- zQ&M{Q$$CJUIymYBLAW)6L}su19!TOT`R*{XTz3xN^$0Qn$Ne|j(4O^jHFzR)btwZxVsxm`7IB0cvI?u_~OTxfAsw}uLUXfUkrtiTZW zYUg&NnLnJsVz>av$U%se6trrK$6)li=Y6?Op1L~}bRmm>Wtl)Tk#i{FlQ{Thpx z2;HV1VhyWtzp-Cm7QZ_g9L`@EntbI@RzrR9}Y$WjJ_HH z7P{rlQ6>bK7{>a@mI9JdyC%*@5pRCN`Ls^SHZ)^Q;uaWtKtF+I?fzwD}3l61Z4YV8WP4wX@P{ThndjF5!U zjeO=Mjmpw9mW7i_^tNET(Pl01{2>iy3zj8kwB-kTeWIWN(#V%=fJ#3=v<8xI)5yJi zrYV=aQIlQJ!q)LXx#fF&@Y?p?tKZM53yB9uh08{_@Mf{~l`W#o2N`K9Eusk=uv(Qz z#GGM&koc&cJD=q-)|^$88KxT3fV(o|2?(7V=^p_AZxB``%MgE?^S}3KxMbW|?C-{f zP~^!+&YI2y0GS;7Q+ObJ(pF8?-yp<>Z}lx#XrF)cO%Abqr*sO)Ct&U1s1d)M_l38V zRGnU64p;YWhRy57M~(j&^bV58kbjX~a#93%@${kcY5kKnVQ}wg_t)9DcuH;*BT44$ zsb1{?(`%Hh4$20LGKLYGf;;Cu#>1qKPg&PlVO>-rIOg$o@g9yK2ZW$lbiKn!_1(~t zuEdri^fkh|gE+_>RJO+r6Wa}t8$PQR^9&W1i+(rq`>l?im?D1x0X&LXKg%qKWyF0D z1&1jBC20|s$HTh6Kzx4djbs_~;&nb$)VEWqqCbGTG4QFbX+>>*vl=(huypB($&8Qb z6Qrw9sdH4yC3j3mRm^6zPx0m*FeU7Empy!7T)NsDrg*I_C0hQhEBvi#DD=<|)3LS* z2g6z3cWwH-;e}3bX(kEdJ^E3C!^!n{dz5aY9K&idbQ=`{3`gG@n$TQz?#DN`m;9Sl znqr$4?ha1NdNaYl9A_xhRHMfhvjFs-O0NtT8;IEH`Tgl}xf9O_V$sBg8gc==*;@tT z_B%E?{l%1ll?=-m%3HI_yy}>?qTAca9*||lQNXSnHn2t>A2hhZoQz%f@ex*c# zqL4nkI3`)cfcynQb2QM}$Zp~5Q5&cRp7N!yq`-YkYtzoMm<|1F>uM(bLoc@lu_{X_ z@6Kp4oe<2rI;V0_toebi8I?Xp47jJ8FBQ&wOPB)O*b!q`X1~_;3{G$|?hb!A_NiU$ z$i;Vq@m$Kqxf(z^$8nhb%;dM}dZ}RWSoTq)(OK4v-UXa%29$yp;;F!TbzQE;Eu#sH zGNZfe^Z@awCYEnk5C0D*?caC?0wjwlj+ZcNb>N zw@Sd3;CxU$Y>gd~ln0(`i@p6kS6Q!nWjPN>*nDZ(g1;OnkX_lXBSV& z*(bpyB@T>Zp6(ewjdxd29_(|#^)i=_FnC4Zp7cp~q60WT8+drzn%ZMydc2xc5u z&W(Db+qnEF{y=`5dy-6i1&9B6wiwca3MK9JOGa?kH{{RYCeN`T0>p7)&P%AFoB}3P zQndUgxjUQ%1q>Z^Y)eJ25mC<87jRaT22a{$>b&`ygVu{Pfej+e)_vWKHKe&MA)C-# z&Y%vAA|1)@Qw!x@TVuN&KlTlt;I(GUPdts!u!wq^H`J6X4!FLiaGLYXm5=uf4I@oB>_t&-DMl!n4lB^BMeS+MwfesNd*H2wW@RUD zS;Ckl)a9Ic_K254EGPC)s_?rFXe1d)dB<*_ zDdd#TvvQo2Ub*n=Tj>^Bptn^&Fl#(6`T2(DNyeQqC>d*5ypjuPW(!4iFn+m>50exZ z@lXKnd{O*CjDwW%J9O?5kaJ&?eU^9%Y7ldQc$(+&@zh;7Aq0`T*x3*72)#g*k0@|c z-^7oAbs&$Ls&kL$_he-|e?S1N9(Kx_yPAzR}0WxW8mU5b?Cx z7$~~=a5@1$YZ(~ZiUNl!I8oHZ!;+@<+c%lDRq!JKHhQQx{*7N4E{_&IauQDOzE<#5 z_MUpLqBy6@YA|=#ApQH-5q}CO-2~u{|9g$r5z9OFzEs-yp>Qd1D8mls3u2eU}6& zLj|$c*&!djlLq68XGot>dLErO8%8GZhIPjS1<2aQ;)va^2`2^qTK4TUb!x}{RU9XHN-P8&{=%>!&-{BhG{9CQW+&{ZqqPDn3dRC6hHl%zL^?g9 z0cqQXQx>t~ywMzE1rvya5CSH}3)Ik6TF3E`mLX4}xKla)2nrOokF;?r;V)@+?&RLD zhyj6kMg{eJ)7(x)yKq~G9UP4Z+8^MP!}X7~^6vtYkLC>)$5?;Upm|Y2fv~D|4hVt! zU_ecI*MTn>D4|XJ9?m!>E~IY|2S51BnIH=JM#qs;*ZxdzgsMwI$jNDZ6)rg5C7MuV z4-_~HZjld4lZ0-qPwhF*0o@omVa?3+HsPw7OJF32fyUUgw6Gtz^%+N~z%Yw0$3Xra z$`k%X?cq8vw{pU01+)ycAGlEH?C!Wx;+R4hHa;Hs4scv~R3P{*#^*K`sYCf^R!KcCqK z=;Ycw0JN2UPYG`36X^VO{Uk8;H$W%^7&X_u9NPOs_^c@fiIPp))#p}e{!XBHi)p~H zPfz8a1?{E1SgrDE5_Az1>~g|EkoLi&=b#`YTj($lDt~tw_DS4V%;mfxa>*l>wwl+S zZ5r8h{ElV=^Iuz~VHb=u?KepEa*dN6H)C*ipj!XcER=}dH|!pYux7A|p8UK+Uk)r4 zzgwu|kU7HffrD9fCi?7c_Gv%)&E=grEMhQkBlb67cH0IEARzBka-73alQ-o~!6@Y* z0lpCc6;7^Z<(V{7Du?oj30s$&`V&i&S)plHBJXT{;sHog~r$S(+X95=a?z$p<&cjT^2H z>(E#uF8mj1TfKgA)e_r6jvhb4R40K}!zM3Y!Qf3u6Azg7=#QQniz_T16&Pxy>?P#X zXZvCjo2RwMV|38Fm=B7?{7t6gPpzh)GT=Dh_DhQ3DMU_LD~HZ(*-4Z=ytP#qeh-(k zUD-1u+kPnJV>Lv6S7F01eegN4wcR+dp8Eh!)kLAgxH(Zezmy4{$8y#uESnISC`g>1 zn?&|hv^E+#wiJ9J5Sse&=)1VYyoeD!P=f9HgcYg~FUPk%9*dQ&L4|Zi-$t)b&>@fT z`L|vtEx3hX?MCHlm4&<{uz*wJp$bB#{s8yLY#D_~Xj+mTk1Ev`fYlP$VGFUA`*n|K zgqbrb+{_|c0Z`Uk-q4-v?-a->&|}e+3ZfaZ8Cm)l&ZOe{+aoP zko*P9Kdxfi!n<@A@iq1o0HRpnA$om=Tr_KCd_P}JartL7|ES2pkx399wy#E_Q@0Rc z=q(S4yFaZ?&G7hbDjMy|2SC@c2M-QH&X;0M9q1i7;1@WpXUQPSy5hN za`|fVH8%$}F%;<0EIG0Ynh%@yAc^>h!5Q7v=f|JloaRKLB4p{419yIUn zIm4!Xvp5|o@4Bw1X9MVhzc-E-YWYe-z<`g8KSxv5<^mNc+wvD5&%EaDeN?1Oqv@(I z-p#P^6alRumoBLPQ9XX2mb z3%Cm(PCzgjo}L4+sd75QPF9!Nr#PU8*ZSkfu4MyP4eX`5?KSwb$v0s(bmXk@%r}v{ zJkRn_);m}wMVHET?YEbs`R}RhhmORrIg|Or{~n6U)FqCD6sEGh5=u)&%F%c3Mg?6q z{oZR0pnN{=NVmE!2Lo6_(s%J2<5mGOUH8Suu`!kq#_QAev4?n!we!C*J?8(L=`pdh z{(m_we^YH@@!c2K3)7ajEJLu!B3-@$J4WhP@*Tz=R8NoAtH{_0>dATu0N4OW@;`-J zy38ZoGx)2XN`3^`_njOlgn*?kD#t0hCQ07uMPZpmugKpttmJ2x&xa+*;<~1LGxm%g z-d^{YS<4MXlnhBz)x6oZ(CiMv`A|hBZYm_}>uXQPLRy_2u6A}DJ>MWv_wv}($S3tU z>)JIkgAUcXSEVA!Y8fTz|2ol^w2s90ph&HcecHdTOFOv6ij?!}TZgYL4Vvujn`gz- z$NLq=+^FU(Sl+*NK2JRJn$T`Oox5}WGR@k{ZPvm6h$;N#30=s@Sl*dIr|dizD3?+) zzERaGpvY`35jSZ@=)Nj1{-*m6*Goex*=W`XsV+oj@M`ywFJovdez-{B&E%7{eg-Z4 zxd&0nsO77Yt~Y3AuPdWQDhZ*zNT?8f5$licHTu0dL^-6j;c&{^~CCI8zi+T%e+($pnRO9=JS4W{KN5ooqC}w~&+uE*6mfHAMzKM?^WD zYOY}+z4V)lX8E_a=&jNxQoP>+sF zmf68N{C*^Af=JKnq$=v0mTIB)`Wm5Z?tyG6RE{{1=uo?XTUz`htpgf$+V^;JgiLwi z@##MJ*n2<~>ow`(La}7vw>SuUiFPE)3Yp9qSJfrBf}y!Hp{@knVB88(A#yZ_A(ZJNc^9S)f6H}QCbP?{ z@?LROEDA`_5xKQ);A1y}GYv8!rzl5Bj1wpv&eoxQWsRUmC`JwVRX+tr>^;Oh0HH!B zPOG-KL>dwpuHVZmioPp0|BYp}KtX4hU%FHGT-hAVLoCm@YGm#`Pg3r4Ax8%AeNCXm ze!1G;$7yGdHuB^at>C*dJ3(t9SRr_+*5a}7N|Ur_WZ^MXk8yhEFNJ@w@n=`Rd>PF& z(%_1l7gh#e^DFNVSEhZ`S(*}C@_}zJA045*`@<7Zz;E0$5Dh6!vS(K5GM98zUm(yX zs!KC9^fU{tH-J}-e>`0;A%iVgjZ~XcM{g!6_u_mjVtnyIhUOdB$ec#pI+5&mB~+(8|Nl05+r6XK@*9ZS^i>1Vs3ruw{BAln!#Jalba-YwYS_^ z%F=y3>RaKh)+bmCo_fv6yH8uO#o)`DTwb@Vlh3)luYU*D$SSA_RP}wB_x+^@!{eNR z|KYC(YO`@{a#f<;0Okrx`i>1{19t9V=Z%M|!&7|2585=11o?Q)$8hp8g$0PT?5}88 zDOqe1{O4vKctj`_-oly}0=OBl^ChflZI=P1R+Tb=+bhU!OkrUVi}N2oJYiv0Zpohj z%u$2nF)F}CQ}=d@avMEvq{y4Kr#Ji=q$D^ss2ybu-zu~MN>5VTG?F1QX2{m(yV6c- zXlg5>1+>lq+opSa)36o-gDJ`MIf?}M6DZ%SEkQ0;VD8QrEEC2FTepNhk9IcpnAXo z5X}O-E&8E;C0p$^*;eE3w^{HJWb1T~jQRuNhVdqr1Bz(S`Auox;X}t^Q^{b zD3;A^EUGktTc#IA7og$SiI~tfwJLb`dJ4IX9Pd+lm6xN5+rYQr`hAuWi(G!b%*%jF z01Z(n)B*DP_sA02QHh>7U!IW27)Rab0fYbnA_w$vXM!;BV9aS9PI9Pd%MD>BRxsNL@${V~?-Bs(=F-wH2?A%=tP5@B9de?#zFJ}a$ z-0A?WU>RStr9MAD*b`9-FAmkyFul#@8t;ZmHYMSN;L+j|J<`XHYU{EK1XI_1x^&AQ z&G0CD*L&8h=qVw_C&maOB`NclhrBdDf*QC0KU4+$GXE;9ip%F_Gl5T2TS#?iC#&xV>dX>YmP{+(OZcusIY zH4yqQ*U^h}X!CZ#x8@}#fdR*PuJH@-X$Z5yLEnG&;pu#!_Ul#nX3=E$i6Lu;QuGxZ z26i`;00VT`&Rzg9@gBo%0X0wjZh#L}^dajVR@h)}=8`&m>F$4K#cu0*`s1p1V`c8d&^u{>8(8VyV8|*E( zE^AGPU>FJcl=67w;8Iel&WRfd<~r};b$t_UIi|yvU!)N_;tPg%VWO0PfjN3icYqi3 z_7dgb`a)OdtWVkkZ@h#rK==)C@bLxJh$)E~$0rftY`?EF?KyqbpXpHI6aN9C1r$mb zMXOB`p`g&;1#$zY#y08I19Vytu$CGkfG1?lF4EbbaK+DqV!|-<)TXe}^0iB1*+4!# zSt-}o3)kULY>WrM>uD0V_%-l<2YV4-8jb_1TVlLcmwQpj?=}#qR2mvfB^%WC@_MKo)_X^KMAa%l+L02a}m9@DT2P;(JKzo72FTj;hQ!c>)ti~ zl>_b&w~ivczR>F|mAzUqA{^6>hCw$1%xy7i{=pxxpi16{ zmZI$RLvs2q!L>!|q1An(V`6aCtr5ALTlpwdxa=1||Du6|U;)X$jS-shoj<*w0BYQKEl#2V5;1rtW0L1gA7;>5P)JyKpfXVDrS)j@~-2VWeVT@c+ zeBX za8${Xii=M=gq|ivLf1`U!8onI?hi-tyr8jpU7myw>PJ|}cfsA`$=F6ofL^eaYK8PP zk9ZR221aqyJvJ@omMT}gD|>C#{mPX@DZAB z(QhXX*(Der{ghAg7zgDL(;TdF+ktf3onvnj*6P=lyk@7dV$M*`*GsKF%|B)I^R;vu z*qfjjqWnN36#`kdeL@a!_wA*QIEvJ5v7CnrVmlS9xOQU94IUPL870fp*7vGZuc7p0 zGci?Mh-{}aNO$Z9!?{>9RIL7S6s_i+{F@CYnlyB@Ra^oD4c@n1EVJUb8>6-7_4~Do zq$I+*1qvbe%!Jppg5@SEmnf~rN z%oir39Gg5+44*{uES$E8LVpTJi3eQ57JH0d!|z$o#pBw=?x0zRV@C-2qaRp3hrz7N z`6L+QM8_v!ub&!n$l1F`%KGPyV9eCqjxi-2vd$IhZFCK}jK;oP%DHRM*-y8H`gBf~Zegr*&krVS8>KhUqHo%Exp;n^ zhpMc^uGJ;%bhp+R>@>HL<^Z||yD)37W@VeNw#@!_CTxnEOL?$9J&#|wqzEAK#=i=d zTRvozvqg5`V1+ddI{$zn4=V6-gy^i(<3ZYma<=k$;(6Z}>e)Hn#e^|YI*rZ&+tgR# zEWk!KPTmX!kN{JFH5ARw=tSxSOR+tsRh zX!~507SWIuQ)npXE`oE!+heJTAllb6Ol$f0+zW+h1h*d;&XS?V{Eeb@BX9U83J+PO zirJ{8KbSi6qoJrzX0(x4Y@>yylfK^@j-I%o@G);3u|)c2Jn6;a>~?coc|-6n_|BQh zE=a^5YOwy}zs5Y8BP^Tv(n#)8+Z@wzlU}kbp0y`g(o5y)wIMN zCv42`nphSlbGXq`Q#KWAvS@pL%dqgnUbVJ{TNnnUwCfzznZ>0y&09$TK@+1!FXHXE6bXu@aZSO_C4sOXT z`}B3=1Q|x}lMiuBLWSFCBzWs#$8~uP+Z&j>Y@93ywY&M1Zt4}~mETGwoP@38#%Xpm zJ@dmrg_RhZ-l%_xYTZ?x2Q^(3w+K-dY9yPPVl=TYqX0#>a1cL|SIAq=ma^+};VE8r zvDdHcn(eR6MRB!AiqZ6Nk@QZ=M=H&B10X+lMPCctYa-=Y#ILx0!Ne z1WhoUFR4?ntKL(@!qYu1Z;I7!f!$n48k_{xK0A2MUw%5_?^|QR6fmQ{zKi`1HQ0ov z?x%l=e8sNXJ8(gIXWqWqz~Cqqe>MIti)?}Z+nFQ8j!A^X3jv_&$8+-Nx5y zVq(x~DqPEz6uOEb;_P*>{#6vs=Q-0b zbWTs%)KbzVg)JTG$+&Tay?;NgErB2g)q+VGaxuM>6=dwH0%kvkf!>7Tzb5WlYy>;P z5dVn;lXEod?0H!luR)BUV-#RrwjrKS()*#tW2Wfj6V51uA1%b%eP3yP2*6ZLm?HU8 zTrYYeY5rX~x&?$5vH&2x@EX|~p3efc^U8r$3#A>ng3@65W7Xhu^a<%zHa)UnkaHJ9 ztZ_yeeo-QxS-xE;%X<976H|L;Yxrr$*{ z98u>j!{1Od990AEslgFvLW8X@0f|3|7_8r2B-M2OoHmHWIe2k97Sm&kj{6z!t(TFC zDl1eX)qEaAUB$ipTkA`vgALiYof_JG?cmRR&%zMDo2$cZ46wKj`~vU7(B4EvzzKdM zPo%KpIIfNt$1d7E-JpPQ5nNY9Q1kPxg&n~(Pi7YfC!Y2fOq6F$RI)I{v|vZYnxE?! zYU9*K+=kZ~*6jgab5`2cmt8;N zpH}}74!|${Z{a{w^@j2TAE>bTkf@!f*Ot;tfgNNFhs2Q(w|NETN+^A(2UcN&ywX&= z4mEGIh@p9*;{I!O*-6=7%s{G`X83ZY;A2O3r`r?YQ%W4~;H95$029qye+|u6*H*UL zqA20nt&F$6qlEvH$Ah_!qlx%W9#3|k{HXKm!^|>oLBZWF?0@I+;zlTkG~ccC))HH_ z7OmX<-lG)C+%$cDy=<`kq_Yx|YXQ z+~0ZX+A|fPk0`a|wB8tgxcsd0Qt4y)78PVkps*qc#GdP6Q}sIAyB@$dRjZW?#3~N^ zM;|UFDI;zWU!;l}sN2I_AgR9eY>c$i$>t2VX{z`$^DDQb2ux0+!O=f!@Yg&L$!kWIG4mOaa37CbfUT97Jt+B5;M9ct9*2__u< z!cV(IEk9X3l&I9)%2Y*cJB?QsYp@W`)`4B(keZE`SMM+xvlxTIeLebVLq~})n_%_RKnx z_^r=pUl!fweRVp$BcsyPtqL9yJlN)E)5&HaNF9&nBK*RJZux}18&H=3-M|C1=1(#z z6SA`q-4NWasnaOl5m01>`5{dy8#nhMwTSRqlu2MZ{2?k4?5qwA?`Tjr$X7SszMyD4=QaAdLsyV?bIPM8WNs!nxJ6T!` z>(|l!DC^+ z%%UYDO^k4}cfsMPO|RV+XbGh1Xc)`DSU!=4o4V3Oi&aly{swof%i z(s(U8R4EnC{lWcFiI>a_dAIt%t<)$W>$qftpjSVF(nhr!+p#9*Lm_Eq^@Ibnh}i=n z56v2fI@BD|b|c2@N(bEAQss>DL|?fDXJAWTI)GZ%w3)F7|9~j(T@5}0my8Ryu5MBp z@vC#=@2b55E|TpbYbxJ}TR9>YO8TnauVmJ)K0kFwWm)h%nUwll8Y}OU@B=OA$^Arc(R#pRkbAR35R2XB&3;_$JN!0C zh1r^iAFMdxKKJdF9+(HQ+n{L8R(|+sKfkN5v%BA4%J7g$%+h|o1>!0nTGAWu-h)OkbU{s27_^Zv$sz;3#b$RGd9cZ~y2~GUDSi+H&O$VdY~*xJtV7Q) zE>E4^hUwA0$f(H(#y(*TW)$uvIV}li+7A~4w4x9(t-;3;j|BRZnU}6V;QmMTAfQcH z>PnV>Ac9Y>>-tBG?$J_MOb7zKKt!TDpr?5b5e-wRlM|#M);-n_Su5PviX{jT8F&d& z#|GiITJn+$s`{e_cZ@M}hxV-(p5L2{ZeQ68A(`M!2Rpmn@@r!DFQ>-F*3^;;i#qw! zf$7-~+iPPup3edRU{rXBnV;Z451I=Jb12d4EyQszspgLwyUzFfY>R>V*>w9HlgN^o^jCc z*IUHPiBc;}oditR6w)NZxLL{kJXm*^YsUK^nqm)D9Lg0fP^_ub#)_5JMw2}IDA#O^ z*{AOSXaGtvJvnCSn%zOB26*8p6?-(MXB4Y%@s-kfkJ%bRvP8Q2tSm>kbL*S)+iXkG+h)OV^Q-ebCeZk++9L%D{%BaJxx` zCoW4Tn?j$X?xeZyucR|T<)5yD%JI(*hhwFu9b)jwQtIJtRlkPmg}LTN;5ZyxKE&Qu zY7mCuKhcLK74M*yL4mhtZQ;fc3UL^4>!JQSt2(7`MG!u8?n=XYtJ^;&@z6qab1;9{<4ukyd;<=dx7=B*05G*XG?@X`i*1&E-}(IX$|B>7W6oX;Qc+p;isj|x=AJ@?7Sp2xyXOK{EAP%)5WKwsu%Y$@B?{p#p*?tLp{12= zd50O4RRZ;O1i^Vdu&$9*?&W&dl@RF_OJu$hON=1KsxinxCzDp}#||y(cZI zPd;VQg*S8lgc4}im8r)wop5?{W~pZ_2fRk?!GPHb>#~p#ZX9+@IoT+-mvGma&Lrvl z?^d|6R4XztdmS)j%0>ki#-nco96|sP-l#BOcG*-;L?yp1$#{r@# z4)j5se}Cr-6Rcwtz!CafJb-E(OBUM6DsZ1rBhQb(&6$COhK7%PsVTDi4gRs4(N`xX ziUS{~s<>R)zAEPvmUYua6!MBEQ%)}s&Z!R(y>)IE`Xj-#PbI-fH4x|@`-CXOr=DC7 z@a(blbsMmJQm?k8xA>z_TCo-Id~U~Nm0}DsjL1;IA(AC+1&x*XQ$|AD86=yf9@a8~ zeC4tslbUTFEA0ZKH0JO-R5>sX*e!IpXo0SvNPiqaBw+xcHoJ8ZQD0TMLid#VDz1{Z z_|miy?Ji=6-?SN$AFduT=<#1V#i|hW(q>&w$egB8k{UC&^-qUmP!2E>CsD>1hzJ_R zw+)6{eHM*|d%#KFrC^IKfV7>|a@3|h;h^bCe=Yg|8k`U!%T z4AmBBD!x*ec_F_`z#V6F&lq($%pP$a%avt=^ALx=2Ldd60mP?s6Ae8p6SJ2WQ_pSbi_N7d;) z_@wrfDt2Ox&v;pS)l!KD0!z&tU58AsI)(65cYH(N)=02od#-|#XX3(=0=LXENYl*q z91j?@yLxbD5`E$uSIK(1(d*lJwM*?wARYiPxW*aWbNQJ2LSbU$RdJwfFw;imcgmT` zc0CLL^|P<=EBlxc-W&+({Eh#)*xw;$KK5G!msFClo9EEbwV##2Zw{>V4crQ7-ExZV zShF~@r*%(YUp}}`3*TKqq%PF&JK@`Bb%?sZDqV~yu~%^@`-aQ29>IFlH51Wg>(KMn6skczfJYZ6~o$ZS_^6QZXuG1+407=gpUBBR=%^#IZ_m_t=FeP0oYMS9C8#;$KzHHF5uz=Qg~fC|g7FJ; z*Sa{!fNro#j^3a56fV1LV?!tH-99kuBy&lshOB~{xbgljM>%Lf&Bxu@Dxzt!#3LFl z2mp(_Jqu1eKCs33H%AWBG~1l(^;g!lEUpv4Mi+Rpe^N8*1?6kk0z{4FsUGLzrf+}| zd9&RqkRdZ>?rl=h-ReAK6VyL<0Vgqr@`vR^G7!N2l>hiVo}gGa)b>4bJ5arJQB)SRn{)YjFu%;aF{ z9o0deWs8lEgUm~Bk_KmoRxg%()ED+G$~9J_3wMmB*wK;x1_wpO6SF{$?(MtP5B*Vg zFoYG9Jx^6L8m+dLQvsPg70)8*nX`A3|L7L6dHh&YIQI&cSnYZx@={%kzI_X}&%_18 zeh8%_lZ0IB!0jB}mdj0hT`5Orj;;1g>}@i3B0v22q!fHQ6Gk)(`yzpA^!JZn22&OJ z!14cJc%nhVrm7#A>1nW7zT_&OV;ZXiY#h%PoO+bUuYoGIFJ7_7iU$-&`bXC1B8~1^5f^=cl$#Z3wr6HyKa}Ffbk*VJmPzv#?ZxZEAlWkLy}b2 z`+gS&B2dSjKy~-&dk?AW@CS0cYJXZaL^umXl3@?toCEw}9gpHg5LTJ&Dt-_gm0B&swY6!`?3v)pE% zYH%Lq-Tu8Au-W5$707_%%vsv)@^;ydJ<;0nqqJ!CAF{>W93`DrFfB@WIG~8#p@&ot zb@^F$)pjjpi=R9@lSAdh?4nO9ji3I(30uh8~*owc=FEY#-ExeTJ8w4PuSe2EW1- zpTrB>4{=(ly%f*CC(?aQh)lB9lmuy1D~`%@Nq+OLl=ueX(%OZWkky}daXl#AvGDR3 zur@#YdxB27H49FpC2vS~x;78`#!GT0fo|We9yjy^0H&)tLF>N(mjTGA~6e zbisGPOksV2M9GUd1dNgjLIpme?^1po+wv#Pj*h8h%C3IsMFH&NBcEWW2Bt%FutX77 z?|T=kNS_QKx2Tz8#8(`H$>mh8fw}|9sUXgrx~M*SBIjd000K85!uHF*k04mrrDLq( zMpvDOCK0;j;_RgadL49O5-0=49&fmJl**0>ARmWxZ#@#QLlUaBYq3}y&_S*@Sq;96 z5&MrQUi?qmmv__4lI_9xwIZ%Q zDjxP*hef|Au($VkTp<>>AwR1ba=+g{KRpz<5U0vlw6=G07-T*q`44_0+a40oGAO5=z53!^Tv-$9f6P;;;ynzv*tE*OU2-6+1FZV$e}`wo?A+3zGj zRn8TjQW?zD+0oKmjOeUFDp&7$4~PF6T4(Mu8jZ}dXmev7M7vJ|N4uwej8SH1Oa2R9le#q2k zcNT5&AGD%B0k45MdY9V~zdo_n^5q{$o5TQNhJ8mTP1cEP>#TD@H!a7~KSjt+3tNj7 z##;)o@U#bNQOz345pc<_a+goVSH~H`Nh3!?PxtT-&{>>Ol5Nqe|y%j(lfAsr}QFs?@fVa%;+RCacKrh)l`?OaKy4#9Fx(u{%ptfe>eNa z+PHDln3(QYm=%E|kNprSj?GtVt5@xkO!Ee1GmcELI_Hxk6Z;SWcw#yfZZgWPa!3(G z$z@iQy{Q0!N@g&aB-V&9nqty(Xk_b30YUzdYPp2Khx^KD)7nMk++|=?#m>DEF{43M zWMZw&vaQbE=aKE6Ub^DAB1clR-*v%PM4wP?})Oegc()P1~@H4zQ5*TOA<{%hB?j%S{bL)$)y zf(#$IEly|6I66lABWi?LOjPJl%SYKn5RECU`v^Lz0yld?x`*)!kS_Qs%=GBHx$jm5 zp(@i!X%7gkXx*;7Mp{XNR%Z&Mn8j7+G%BJqt6!sqA}EM+T-*)k8hHHhbI*O!NzxU1_T*Ip=pzD z2gsF&fo~#|xTgwB$a~B|ShFJ8>Fe0BWP1P!I z*awL{eU`yT&y}&d-q?nDk#a0I^A}TJ%>z!<=1D%;B0^XgkDYDgYP0A7pRn!no&clx zvFX#EU=<=cFRzhhQ0dI{*EXDpnfiuanl>hf+!s?gOB6jX`!Q?;y#Z(&u#|c7`~FoF zo!Ga@PosJ1u9D#56)#A0s1WZT6{4V?C_S~fyU4zQMFQ@v<4%Y$OC#)c3DM!hugb(O zwK66r_^Z!2>V!a=pp60m3O{Ol2h1=C#Kny+8R z-d$(qCIgZk0tT*;12`<^7`NQEdbQ%lfM~luqgS`07_-fZeu+cqYi*tRFO?TPKYv2Al=+qP}n&cw-?|99$Cor|yPoQvMoebcor zR_)#WtY@v?3C=}1D(mt^!vjZKgis+%VPn3J;UkbONuq{Iqc`ccLeFMpwdJYNO-qp^ zv&>YhfYy;7S(RoYhfpP*D}}yjND0CR^=4*A$3%C4;D!YS9(60k)bwRxC{uS7bZFFe zydRgxG6xaw4*@7~DA_b-@go{sFN?UNq~nUq^NhTluXasFfvv6lbvO-eZ($jhsu&e# zTkcL^5YKDn(wG6GK8*IW?2o4!giaR2Rlk}#h zMqMl*m3}XpNyPgqEdUk7OB6&Ip`9{fp+)`FAzMD*iOY(67z{nMcmOfRIuvwj+jSPg zDzD>`DO^eXG{O;gEi{tDZiaqA;aP1GeH}tzo`J{n?;fMISLIgkAN_3(Dg)-hF?-E~ zdNW~pr0WhXJ@tyBGBOLiS(P%lC>ecKY_0mgphl2>5yj_ZNMP?vje2W6V1e9x0X8l` z8~GyB9B_`5*~9gJg=l<>%i=N8Y_9sMi9El?(pA` z?$_>rd)bjb!!C}M=|OF4(;28Iaq8d@LKH(dC&}>JO_d)V4JnJpd>iDFAqI=yFYMK7 zgy`y|?PJ&vcNXM(N2#koc-#T!AC^r@?ftRqvCfB+#FL7|9Z#WWP72v0<1gks0}%En z23i4ROU#9M!r_ElZ-6avoPz3-Ih!Pk4XML%%e)6tzuNGp;X8z&GjHr&9;gbSB^oHx zsb)WG|C%?>BP62UD%!4y%%WT#Boe>)-74$qW!QF*O`wHutoM#nOf`Was{x_83;qkx zV%4Zxk}4pr408efeA|KcsxsPWps)Lnz5F$51*_z(M$%*g8%q=E_z`AQi4OuV@nBaR6R{pH8CK748NOgdY#Tj>Uq7f9G;@v!i}h71 zHjZ4g`-|Knt4C$vjRcm&!jG zS`g$olgRw2I*TO1rUFPo!ZSLhB5UTRw}fK$1C+yob5>YqS99>^8M?Msk}Nzkc(e~u zNy~)4&YtqpXxyy90pIG>h7kF>aSlfCA>&do`YRJB77c)e#GQ!q7Tn{QX?RwYeq?Lp zFYM~u4W=sgS=^GK?x)Az>&8QBW?CvDz)DA?!=t-~_#JlfBL-ds!Ku_s-m5tH!Ag!-Tzb*Q1p4eys}6k{gSFQ?)9|LaGjio7jOq7rhF`*7?gS zoPVqH)d6|TCZD_05a?z4v#q$8m%EY%P0!oj*YctkpwH-9naVjych%5<7{nmQa6D2Z zzbEAErs@et8oEhHEyyr-Wh$`pl<~yGMKSAHdk`vYv(Y%4g_Mx>yyU3sl7}PNKKE<2 zvIt7N3@s3WM<(c(F&#uuo)a>}N!K#{JP$Gj7h{q6n1B95&L$j02+IvwIvTD{4S<*; zt9UqPoYX%R{1@RZtXU0l(YgXsO6S_O4Zoylmgo3S4`k+axto*@+C`UXT&XbrE%gU+ zadCcw%>AfCtiD5VwYfE zUX#&y>q#NnOIe~pZmq5z4E<~1DLx{m_!mZjQ_Ltr*|@me6M4Dx2#%m?n}*C7COs8Y zvS8g52S+XBK_vxK9p5wGluC;? zI(vzO#hu640;H}DobGMSxzH1EN>XY_y1|u+=z=esVOcqZ#(fxk0;gQ7Qg(lp8lsXu%ME4gj%li172;z{Dz6JKqL#FQ8`S6 z1FG8*dC5G12`5K+m;DdDYqrNFyhhlnusNocQw=)F#attR%IIpU_p7RzdMWE4#4$EJ z=i{IAbC9xoo=J|cVe^u4`QwDyTpF0ixm;`TsxW~7Lun4$O_sGCy@?p{$v<;C$WluH zlC%Ro^8zxXP>T7o-G}#P1MQLe)MIF$%F8L?=3jPaN+UkE)eH(T%lr77UtiZ)SU~G7 zOM%;W0~AR|1v{rWO2{&-fS?pL=d@nL;MQtP6D5VvM3H;TG%jN@(uZceV#@4SXr&75 zKk`BbY-S-YkM#;Z%Hy4LHFT~IcwXgp#x+6 zND<@p#?Q`{ffJ7DFz?MsoCD|1Y(gDG-%!y-**^xdVhMT0f(*bnv;!{3$pUVcfQ<(A z+4&CmW~#43crMLIt2)ooHSnKbya?`hXA6LgGnM9<^$^k)Yc6CG;Ou^Jpk8}J!d5{_ zaMX``KX;BW(awGNEz>R|umYKwJv5;=$nDCrFE{1SPfdR2XG?^$G=DpSLmpN0#|>1$ z>D0(?ycrd6`&V20Yo>)$Q@~I41v1l&*J1*~rNrl3DHJbus1P$H_ax+B2z_bLAmGw8 zmxE$cYf{5mPr(aN=Kh)w_f4dXyHMi(3Bbgd>2bj|;|MsoJg%ZS&mYZG6g_|4H+~WT z)PeW5ntTrQze%j$we$utrh)ns6QUC`uPP$x$cOx6>2ZV`SU&ZbXhgT%WN(JXfD?fX z#5-gmPaaUi4m9t(HnyX>tt0KTdckQQU{s*NRMhOj?KNtF$lg$nBoOT8W3eqw-B*T` zL~2eEENIsmtl$#weKQwE<&2LYwJ8aH2d=p^2E=rQZX@%Rl%&!GWX@bzLRquc^fP(s zP9+6|a&G5Akiz6`c&ZJwhZdJYH*(HFyC@c@j}G%_VUepr@aQn+jtVz>agURl7svvQ zYO^;ER<+L${b2Z}gXdRSHPeVX&~j=KW{t!iC&i9$o43{azo#K9qY+GIqlr)jS$*Xj@r&ADMct+Qgl%}j(lgo*w2rGSq2&XuT zHbZ^);Hb)X?*q+fuERAWVL0$XMElf3^BtPnE;uF&bc!%Ib-bqkR4+%{Hr{o`dM2d+ zp+~M{)wrM&4N^EK`UNf>%giKa&@2`f7-+9ytpA4MJPP(?Ot!Tr)$=emUalu#c)?gj zI-1RZEy_QPq%`+I?8LYSWlp%XlO6^TOeRc$=NeC5LIU)!!xCyPfO!0Bk5f0lm?_aOKTvZ#ye!*P8 z?pNXm9*VRiK}ZNJUSH2nruzdUFx`v;VNf|IuS`J`aF_qK)gNncZA_rk_11p~fciZh z_7KGxl420!_s@TkO1^&_9%8Z^yw_BCjr4Qanfkv;V4VM#1g7j_3ARr)+vdY4;9RG7i_JRlUf)Dya z02Uefs{ASh5eW|#2@hWSqKo>hfbuMh^el}G00RNBb#fQ6v-fbcG`DajrZT1`X5!-F zpe1Hz{71+XvN3hEG&Zy)mNj&?Ftsss{`X)_tZZj&Y3l4jOvP*A>}=1?z~JWQMsH~2 zL~rM4&QDEC>}L5dbR|WpH0f(7;b&=0-(9cBglpXJ|gzwXk9$t|i zo;i_;gK9+{EZgwMtQc~h2WBv2_dv6JXo^|=tEq&B8EvL&)3 zxc7HsW5Bg_Bt@jOjRu)J^tO<=AeM#XIafhPC@UGvPb_`!q&kc4SFp67)+n zRAh2wabPy|GbE(1SEO5{H}pq5RHT2TeWC~S!*c~jLR4&2Vq`{Seso53hSM|PU3{Ap zng$ZlE%#L@2lCx06{-`q6OIrToFf+=raB!S{@;ZF0MNNkdj7X;;`+a`>Hj1T2NUc6 zTTs2y=JWd3i9C(51<`$o8ZzZ0iv78DC6mI|JbZyuQ`FI=`5P=ONGeJS2p8AP_U-E1 zEB?-lrCY1I6pf@5cWi6c-vAbPdUIahvRxkV%`Sz;FLh`4;nvM1MLCh!lWp(G6aGEP zk(kOsIa5=V^)iTONWnoQ-w)02z>lIA7S6^WuO)BGPI-BMe!q0I_INCTZ31OXCjSqy zF}y`9w>mK^bS{LHXknv_?X(_Wp0NSA@NWiJ=b#fI}q z+JCWD*?4!21MJAVx~j*?FV~A8!6{eqtdwR=R&z@b^`!^ug@`7mJNOg2#>#{!w>?jr z!{`DUh}y-_-6+Hh5vY4Vlmm&IGFdWV>&EMTYjU3XMH^R)KBR}*ot94wP>j*dl@gs<>hlWH?m zw69jNFQng7)4kS}??FMYc@5pyXm^@e0M3p~10x+)k+TG1MFgwZHN(AHb5-4H5fbps z1&$wdQCb26->;Wk)vY6PnK$)*mhOqc9rf0`HwKYmW3~hm&f3b*XesLH>FD%6O3zho z%T$eIqF3fv#pSf1(pN~W6UOLzS1(`5U$1#ItzEl~t*UXVhbv}7kE8OSX3s>W65V>0B- z#Lmxc#iB>M;<40C>FA_C%V=<6#jiQ@y|v=}DySGcUM)GZsh}zhi;^o+k(mGVjeLe3 zMzBJ9kZcqjX?G#BAhw|DC?L;TptW;6PobHvudV-uky$e=ccx;LeG>R{q(g)i!zg24 zk_3E{F%&Q4!e!APh)th*Q4}h4Dl;#QR`poqyNThz2kO+{_p&GtZ`GgzO|Wl&|p_9#;8N6o&`ett`v zcTE}-D4SE$@P^jcv|^xu;@sB_|16OgNvVeVnF#1PxKk)I+y$Rykbuw*E<Akf*?;IuAYeTz<2(fIt&8?etxM3VzwN^HITCE&pwzn_M}E%s_jz|P zRn-;1K@X4fnM){FENgV#M60$gYk;5(+o%=+CDo8@mJIy{f?zOlg|s>bojFl!geuI% z6P{2}^Pkptrab$=z+TbNoDYT994^Q=f{7Byb*RYw0-s>v)J`|;yRHj9!7oP(0$-tV z%WOz5^U1c4pY;tmEJmpnqQxFaVYP4j6aQ>u|((O3J;tuRp+4OJ&sEQ0gUwz<{R z1%HnpwKk-LH#(Lc12lnF>!O#A-6Qm$y<^4TLSUE7;`0@5qt~53DD4r@e+Bkk^(a9Z zna9H^A<0RWjFobEbyY(T0S`x59C~Xy2}kJ2@NPMY!=Qr=wts?CCFQU~N|8jY^vkO5 zRe@2=#2vSMq71+*ajtG01EHhFWl6`*DxsTaf+az~_tOFBHSH`Of9WgS`~;#y$dztW zB#Q3&6fJEbZqnoBUkkH#z%pGIS?wOhrhhRiQSDk2Pe+D_C zqlbbw<$VSC2p|>sXIGb3UK|DxDpI8H`qmxhtF$4AeIusCo2+_^%!EYE-N74 zHTys!J~$#(6Qx3i4&Eysr1S%0*}x9ehX0|!WPDUfVgcW^)hR)RCR@*op`i7ph|t5K z>$fye@?F%@L4)a-{I&kfyoKLgFVEPa`DloSnG_c+m!5trhk^{Oabc{;M4sTAo+8MF zv=Aw!0{sIa+_JF^Kkpw6qd-k`P6omE#dn3bH;pw=eyn!p@I{ND!oQTA_Jb{kv1j+w z|8)VnTP}#U{`e`x2I~*xio!bWA}?)&C8HPufD}!O{}5z!qIq?x}`*DR^jJMW(<;Y_2zm3FoBA)QUGAsQ&9h5WWy4e+ht8GIO=>HOkjW$c;CSm(=egIR>9==uc$VyPqgP?u zZPL}+m%HObD3U>dMaL9kS&|ZoWWJcx2pNy|b%SCLXpkNcll>a*fEeefDcA8~tGvw2 z)^vqqxDM|*Uv)MOssY*8!|1fOcOl&$A>Fq}X+k6S-Z`#8XCLcLo~ff9DNt4=)EqHl zzK!Zg4Y--dKfqoql2>27OT~Xg6!D850*) z_=2}V0rQ_j05wn{N6CSVI=BwewD1gtaaRCphC|n^eccv}2btDdtc@Ro!2IQ3N! z53OqlbH&|x)DNFAtXMQdMCg9g)GjZ5C>C~;*?6QvzIuNCnY$CxBGRCP!N@pqXe@WE zBl%Jge`8hfL{Jx<#;B0iKBhiC8GdZ}S@66krr0qaY4I7T4FGDd6i0hJf-jjhWy`K=abDbW%+`jlcakS;VJvU1N~FZmALZ4qe*+maoC8B(njQ*e}w2&zsX}BgWxKo z=PiQY*K^@UMpD@Gl((xI=Fq>meUZPf-E0Kx)n)7O8R&rYGz_`K>D+a`_JT&~EMpjA zI)gNA-f625|0t1(BOkXBAs(l9q+nUD3RBqbC@U3(t)};`Z^B-8llBo(xdkGeKEwGA55dOVnB9S)EgprjVMJy6w(!I-Bn>ef``4@nfW zhT_tLhc271HZ~FuA{85DT>y-*dn{jGc!T+yUj(`NtT57&E6-#UN2FtOK6Q?+^CG;a z1_D#~xg;F~J}IF*wq0Fgy%T5Y;gVuccd>aD$#2iVWB+!#1d(ON?evZRMUeu}7Nh`; zvIAQu>K|l(661s^d1LAgPNbO^md3ukko>8e`3T!ev)0^}TNAc8w7$$)c2HZd7JAK1 zhSG{CM)@=64%vBlk?I*)Um{Xsi^+CS9;xE1X8l8^9C1N-gH->AQ-qy$<}y^+mai>20O> z@qk}7%VHv$&T)n26$@!yC{=Ec^$Sv2AFA1Yy0O18Wv0ZSPF7xv(@cE?WAlWU;%>8* zmyb)kLTkwzSHB(rf53iHS2#j;Q{z#oSXLb4#(GsTOuZ1|p(P0dG-M%P<9SqNTacx(46U56q zC!#e<6TyhPORX!Qw1No@Hs6w0D=q?UFUSNnBZjme;V+QtSVh&4gh`>l>_F>5btyzSpn$Cx++T_c zx7{}Ck*ItG-xYKt!oiim@LXY4_`CL^eVlp%bCIXL-{Qx&9(s13g&EI+ z4p#7#@cnhHa(g}?8w+H|MevU7=s{eO3nsA)1^MTQy%Qqt58B^t-4S#FJn; z+*b&GEH*uq!b9mc`~NClaO9=U8?7zFcA<+WWfK`%OUiR5`sM7mUEJ>5zH{!w9 zp0g15a{+Suu-d2?{eWuM4x5Ml*9UTq{V)1-^~{rAg2;9abewj+x3mo){!Pxx>qzJH zghykYz6u^sJrq1OG9%Jo;LcsG^_uUGF(wk~H?}Y5dwDpLw5Iqc!92jquel zzAHoHuikDMAZ?0h+r~ySk{JqdH}k{Mgv?8-RK}>3wBJF`!;>O+8~@vpZe5b0bpro@S-c=xZj*Or;sl2pGD3ESJ~6sa8T z&>MOziG8Wlys5lU*(irHL_S~4C6%_p6TR??GkpgAP26d6Ia~~$$X~eAWa1S%a%S98 z%4Y~IyjNO{8SvZRv|mo#3tEvYM^nzjXh#LxYBpQmhaqHa`_Y`;?N;N@3z`wR>%UCJ zR2z>-A8+}wZam_b-Two}=uDS0XHHc64;*7C--O=e&Hh{D56uA|Ul;G&0*d%JC6WDG ziNvZ6D&k_=;M>^SXwfib+x1Rm4E=R-e2{1@c4qXrWQo)@-`J`7U4!HNmkT!wHfyq6 z7KKLEvs8Ie>x-#+m71yY!cT7O9}x&PYbU1_gB6x?L^$LL*MHFEU1i&}R9B5DFk;ps z9EC#LWD72f@T#X?MUChb=SMg=Dy;a~$cU073x4arQTKOqKQu@m)aGJC&&ruyq~dm4yLnO8fE89nq;^i z49v@&Pi@r0=kxB!{PPt{%G!q$Pcq7qBwgxD?rpRXNCs_X(@J!G%-7jsb?XiW56Vj zg*vJv$4G@LiToaX}Z|rmEPAB|4Ppi4OF4_aBF(c@uj9^`aY_ zE^q~XM zmXtMui9KZn@}a(vAT4clc#BgB{-tH*(;5cQeeQOtSYmDz-IEejzOxV-qemM&Nzq>o zqS^wwEz4kP@ zc?;X)20WPqUTKzwyrafcAYbPDJMR#QY3yF_u`A6a8%J0Pmk}dQ^xp@*NI(=nwI*{G z`y-p{0LWG#KPGPWH{|V|BQ!rT!#*hxN~Z(M3)#_r%{NS_vAJ@9_$1IZH)SeAVHmBq zcEK39XEv61)6Lzhae#VuxdMr8d_Rc<5cfj>`v0LCJ>v)7c_J~s0t`Y(r2=eT@Dr?H zm$ii*o%bP!hywoW0l3)2=_Wm?!7gdyT&IrnM)B7ISv7YO9r(FGR{rWKx;$h_tzX ze>`5MPhU(G=&5HStD9I~?t^;QuGi8W53lPCB1xK;Ph%Cq+NTiZ99gzeYYXy*c?_?+ zPRfY=rUVW+*FqwD>2j!c=-7>rU0b%qXFyG~{qejN)TDTSAg+-pv7oE)wEw3bkH$gn z_9ES;2FI`{c#*#Rn@GGcB0UC0GvU<*$RC;$bBP1;*xYxV=grSx*Ei#1>okf9rIn|@ zK@1$4Q|eJ}v4GEBW5N%Q$5FoX8uvy`SJ>*$FrKR3G@kFSG@?B5z_Ln#2P~5C3QMuu zbd`IR`{^bh7>yy{uKN?vMD!bk&HPvxOxZ|)v51j;0;bwZf@vb@m<^iQGY$d$-7+~G zpf25%sN|+tv7)-T`uDO zNC3d$uKK0kn13G9zT!mpLd4B!x^6yYJcA(GNHm^!qrmb+nK>a0AlGVG96`Z+0y3T< z1L9SE3RY?MN1HCl_J9=F$$)1V>U{&MOz80tg1;@9d0{^}!gxczv+1#=Zvy>QpbN{T zY((mJdpN4lhoWyrS4M~+yw+RK;h(Gq6N^2(+z_g^)-@;=ml~!Up}a`MOp=>R%&b}D zVNJEgg@W{Nd7*}QJ%bNNRq$iRV6Sg9brihCmL zlurC*T(%Uw(yn*G3U?)@86*iT37jkflzk)@Ft4@=NJ9ZGcyn-QyNKV-hR8PL&0Ai3 zn_!kMtaz|Ah?B)ol4w`MxVs0sIfaY8R?`6-ah%SuiOf?TVK*HoN%8fT;vfv6+H8E= zTYs+8(IdC;t$A4{O|Xud+qVaEO?dGC;LK2VhOvT2FxE8pDQE2B8fs!IEZ(Kr;UO%; z_z5uLn#21+ao59Sz<~L4(3;OgEJe5$WAB7i`~UDD&6dU3uJ3=cqh6c{;7<@hG&dev zm||;HSp%0R)c2j(xU7U>8mQvD>JYr`T%G$GqZdNIRG3*Y1$W}SZq^WkAP&7tA!K|1 z4s63=ec*fosE@?%8sY~-X(khS9&-o1#38z1o6P;DpxP>bGu!yaJYo1-alL~}MN)QF z18gBIB{^t>8TXq@v5Wk7G%sb*Noy{AZ@K^j`MbO6akUzEWpG^jX!M5($qFO8-NdQK z%CFxPl};cytPo>E9&(m5zvLT8{C1Q;OPVT>a_O_4xDYla;Z{W3S%_n1GTA6%eI;jW2TyR-V5k%thWOL^xq z4;W<-V#ONrdV4%P3}t#;2_JjDeYF?h$@g1jK6TD$-59fA_ymi!uMULx#f6(vj74cq z5IFyCPh0TDMa$<~FeFc+4d{WsS)8N);QqEWxDsR-2#S|nQkN!Gs7>P%7)!%HRR}w+ zhh}DpO%j=9zA{2j3wES%V*e&s1In0cR1i$Bjj;L(Gp0(PBcZGgmT6y}yx;%Oc}Y@5 z8}z92!T(czxT}qy&Xg^yj!pG`Q(h1(K1vLSQ;7qYyj;$tIffU@aNtYUAw2UWDZFmp1KN?6dh3nFdqP zYD&>QgxFfoft#88<({dIc@4sd?s;4@5yy=Y`|RiI=PelPzy56MGT~RgoJ}Y&#J;>_ z&M)l%FNjlUOBL^7`~cxchwJE97UM%`oQ^~c(ZD$BuglL%cj;M68eeY*IYt(yF6W1WWz(D<)7&DRY+b5fWj+MN@*joD)FouoKws>r@u~>S}POvd$?h+)C`WU z^fXu}q$0*ez4>Bw%17p5k86H(9M6NA5%=5WHQia25m(Kg@9`KkGn{)rrpKE#Qf}9^ zIhj{+F_gHZV-0asEkraRIh>GFY|3i)MfT_j0i%qY{rOs#>`O$_F?(so;cT&xHU&x3 z)~em+L@7MMvmu^b+k06nZ=oJC2e-eiUEdve^5SiX^iuuTsbf!`O236oLSu5rY9*Ck zDr1qRBiWasOr#e}BCU^lX(Wyvazyf(vdn!h@6F zSml6(&ySx|If-}xsAD)uYVejYa>e}pK|RR`Pnuw*T;(<0qU}~D0e8`!ya}SeZr$I{ z^$e}Gmc)u~Dsy{LTvnrY)G8&&e+=gsqnp5FMnJtV&P2t~z4sjRqiUs(#2+zrV=jxV zF3LIf7ue&&5%z_c($%0m4p`JMO!abpvGP%?9#=i;c!!q|pO4>pzDF;`Z55`E_ANEZj9R8QMcm zg3CU`Cze(geWa;!Wuz<{;z66d<3`CZ;inmHdP~Ba^$cd9lV>tZL1L^qYmE`eL-Xby zSn?87qhw(F>LTGeCX-j;z!Kv0TWz$|o%KrdWvUF}etW9aiemVU#|_tHjqQf%=hQ&l z(4C1|*3jG@qu%2V@u5&u%<`4@`q02cyJ36~&f-af)H+4`%DfK>CpK4|i7&VT?%~~!uWbrh7;0YoUTGE-nvnR6nrW9ovNcs!zBGDJ* z38O!c@MR)_lAj}Los1;2-;vqkyJ_&Xqh zE$4x50CvHch8IG@p>0!r>f-oEXxb!oNDGQZxnL(K?je85wZ&J ziB%;f@bia!8RnoV2)S#^q22x3C6&?MW#_!+?Y`54-uXagQ z{E#fW@sFHvCd_Zpo$#zO`lzR~5L$lj@ewH)=57yOKI~%VAwcX=hSgR~6@H50d)lbF z;V|A4!tmxrWd9;izogH}l1T3J$Ph`Hd^UWXG=qJ>y)uiby8hX!an=`)GD<8h*Qutz zZ&a6I={E=5iL7ny@^bNe=hNgw*kGCs?OPviuxAts3U3*w7*Qc8iu!A&?Dl**Y|kQJ zIkwqNqpG}?d?O9g2_{)?&*iG!)8uV1hlOMRCn@U#|-W9TJFKw&_Au@MPYmTXFq41IC<- z96!_^FiihHXqA__!rz139Tg^eHAJfZ0AWelrqT!;-OQqfl7Uvw#U#5x+Lj(!p{-GA zW*}48(w=l_IejTU-hhweLtvz(ANQL;D446`e^a@b{)@`R$;k2lwq`A9&p2%i|ARzS zFZ}A-chDdgPTOJWwjOJ~?UZW_*&=X2zV18QJx(2;I^_!2jCAv~r3ZcK*4a=fdU;pFZ==B(Vc z9yxO*Y1eeSVwQ~2IxB5f)lWg91eFFnS646X6QbeM4a5K$#iOs|WzbIeuIFYh%|QgdnX)ahS&ve5DT(oAk{rjn^UXuc&DCXn z1uzh|o%-tArPi}|9U0Lk^Dx{TcIYcMwwkFQu3Wj5b%gY1%H3Yze?Meo%aZeI%wR?q zQ1iZ!88XRk6@&U$sA{bQL_;VmJRhL9pQQRk|d#yQWcL)ShiDH6Mzo z#~{JD+#i~Vr215l@8@pJT#idHNZmx#S|xKSXqm7p*VQ#^ZHg&mZRNywqezyD%9?b~ zaK92ch}kOCmA%Zf;f;sci2=h~2_(_zPU3>H+nu^P~!=ffWJ8fS(<6Fz? zfRC#DoQl#Pp}v{*=>R0|f!+30i|nbA{zB21EfU|(rcFrz8!tDo%=~EAC59rDlhHaJ zDtK~v?Y%%JrixIIvxw7%hS&Gn*UDlINFqFb$eFntxzB*Uly(-$6Z+^d1xE2pCoZIjR=q zT$v^O-ehkZ~Wp10; z5zhyXIoeJ#Nfs5(!6GUOEM>dvK8EH9qaiIw%b$z>@5dIBA8vX+aksHgQ^E+ zzY73IUqTvYfW;4dtVbZM$s+RDd1d&km1w!5*tO8N3VUd5EQ?1=i}}p7gHfm}W7kFl zVIsI$NKjC3S}Gfgl}>rmn{MtNXu`;Te#GV-1p|OH?@zs+&%>_-A*Vln&7C0AhF2;OfZ zAXb^w!4bb#VrPhLzO)ll$)>TT4Zq?2GNwJQ+afYv4YQ%1L5d90qj@OpGkqnSRip{h z_Wg}XHwu>BBJ10(*X zI}2Y^RStecF42$k;I?Zr@ZcgeB8jmiO1*pTfQQLlK^JIwAiX8ME;6DcZZeX+Tf)~4 zmeixYr+ow=W)#&pz3F8#i46UV#u@9Dw|I+S^edY~j3AAom><*Ni{X?XzYR&!@wZVt zdIeU8hl6%y1O>R`YGSBjgBz-Fq(_1v?gB}8HbaWXh6Pbv_t=sz^7;m$@>J3sQ#0j2pjgx%A%;QrgCv=t$a zVTwS@sQx)7B|*Uh``S)On?O)>MsP;NLkB_@i#(PGt1cKd`8yXZ3c+njy4hRyx()!| zl})4!QvdD1qT*#k>yw}q^w=xzeKzP*cv0BY0I;0xXosKg z>M-VYZWWqJO^)jDXeoiW?oo*+5gbyLkaOdKL5XUfgy-<{ZMI5qK?UDExl>fZT`t%k z&MFzzrY+5{=Xb^n%3GI=@Dhk+|H>T>cHcv##!;TdBXJ1Yuujp`*`e2Sq<< zQcC%aTRcZ8eKm$BJJn5-i?W}rvxI!5OGHec=UB0?KMp&6S-wPwb8rD5njgI>|1jcR z#~iNuHTEG1B5=Oz15pQQL1xMiOSN#C$Iqm@cR_{>qmV5nTj60Z#hSKw9)e@go3<)A z>;Uf<#{f?7P+fc2*Ly7Wk_?G!(yfyNhCbzAB3e53NgE31B%R;1ap{;tsp)DC4EL1_ zF!tF83OqVB+hZ9_ub-i`W`eaE7bAwj*x%&mBpkK8c91OU9szeV_?z`)UwUu9yM9BM z}+cYgqHq^S}I z1k@W`EjRfZ)>|59{o&1(kIeBr=$8! zC(uYTwG&X~?60;z>R>yeHgr!99oowM>UCEhMophGIl){~IIhT2JYKuOpt$Qno@z9B z(VD*^)M6;YcX{yw{})d;x+$n@*XlGAuDba1wx~kNn4I4phTO$o)kXnIo@EnUjE@(7 z?OON}BnSM$hncf)MGzeFt8_Y}+al*VQRf&PYlF$J+^#_m6mfBTK}lydaJmGzu*y-d zUQzxK@qFNqQt{e+Afus#;osXS^hO36iZBAd_ezS-Rol|hY0C^BMp{{KM<((mvZOnq z_wAJJ7j;FP>rL1LlJBXx#D*=Vs=g&iV~s48PyDCi_bj$*r@vWQ^FD#Q4&j;`GT2@x z2g@qWJ**|R40BDhA!-99h)nIjJvIi)B&5?By2%pKH8;bAPOxa zNE-ZN*b&tqHP+t9hKClr>HS?)pL{|F5)o)6uh5IUbEN!`8hG4qZpvV#tS1A`y+#-U zNiUq;S|!Np)jxGsw_DE))#5a!i~+2=mWH$>zn@JL=^iA@Ks1pPW9AYYDP@6=6|x8< zw1;Smqay!{v3F|DtP9(`aqxaUiD{Om1R=E3i&VpY?XcFE zG%}LLuy@l$1+cK-*b!*D;5Tilg=AxPYv>}-fSw%sM5~?-zV$wq?Tzr;qqo{tp)18^ zh>4(WKsF}f1}^0iCx*@8(tAS=xZ z`b5Gg)^7^}Bf@eN15(a!*!Hf8yG3PZ5EQr0yEg1w_6ZqTVec|ggff?>usr0^>4L@| zO$JlMPWJu89a2;c*`VQM2dCJ4C2>g#u@<2qcTm7$moCz>AY8>R-dv~c$$MD)v7aiT z{-F({V}Fa4?AFgZ6*aH_yg~Qz>+%)`v$sYNS#3!y{?ZXw;{FbajnD6u>h@~)|2k5E zvN%)usQS^u(1>PI@Al3db;T1#-@w?>I;_un0<}vs9p3GgN==F;H6TqXA^xsw6vYds z;myD}6I5Jl2m?sjBn#+!5_M1OaB&TspK z&ypv-Uj_D5o97!sFysx#C(A}W@=i=$BRS7%a&V`Ou@FkAUnaVceQHjyo*Yru7g~F; zM8-`q`k-ucW_yaT7jXlpMa8TF%Mki|X%7bc3i8cjoH6(weECPb{J}B>+w$P~L000z zvc{{!&tNy(u`DA<#Io}AFC?4U4#wlV+_KOZi8A%-`|Vih!EB`fX7rMFH*C(ru!?7j z*qT1N+5Ew!!{}e!Bkek*y z$eITz1e(3l%{{dEzuSHMQRHhK;~9-rm_Dz+aQ�$Y8Gk`xn!fKfmsQF9Y61_Z5!J zQ+Mq^x#9Y=E7BanVT)F00rV`76xf9^LbdNlDRBF;*wY2_;(1nvEbYs)8hmYJdomoS z{mkTH91D@fAWX8;)!1SP$_vIWPu>*|WPu zkh~DnK-#Z^!-5|F;2Ue}#N(iy%HU&A2lZ4r4;$g2#Eu@AjoEjvvo25mG=D&p#A(7^ zr0nRGn~b5T;meizdXE@P((jISUQeO!%|Uv-k#$m`1hCgd4>xr(CUVKZr z@Jc*3?ns>S-W<&VP&u>NU&Ao3$RPp`4U~R7?$AfNHTw>cQ-V7qe}vO;-LvV_=O}HN zTS0=JV4RsSC@#a-kC^sJn_(3&iOulVe_pseLn_}3;jnpFevYB$&~uwx6!+GC+F{~UQumq)VLjXd|r}BknP>#_4I>FG|_RE`0pWz1xUdcOb$VU zPejW$A>{ueVl4kf#F+mVQP&_Z+@H^OnvQC zoM>9S84UF*qWIEIDj95ge(p(yW*(f)pZ0|f)5;bM!n2L>6n(TI??u~{OAT+u6m?%T z%6x&A$4^u30R{XL=faBC#~7U-{TsjB!Wz|?`(m{QA!TyK|LEVGH#nl%Qs6|=Mn4hN zXhe5h-Q8|Jfn(6({suhu@&?q=ul|qzjes7}*2)Cg)x*Nc@%02pN)tIFhSg_9GF?0Z zdF6!iHu>(?w;M|flO=UeJB^`pNw9)5!1-Np!vNDQF~;7G8*3?hARFlpr$>B|s=hJ` z09I$F7tF!#lDU{lgt_-#JX!7$FdK*)30Bg0Zus*7T#+ zTIp!Opv;zYY0_Tia}KB@i1D4T6{I&jVQbhMpt^*$P|d^E7T`HC1sVosY({-#58kn6 zHeg>_3WCV8Hi7x^v#9=$MGn#sZ16=1I7BBWBoJGgQhvrjc2}61yzkZ##IF#6$@?Mv zC?k7#cSyz-EjR|y<;ohAtAUE{#G-F|K3w?awecbAH%B1d{ee-}uP7|4QoqqS0bk#X zrXhF$0T5qe;M}loo&J=e{#*5DG)f;Y7{!3HACQW+)m`oB*Y=VH=_<1DQ4%s7G0)0# z+f>sC42J@PS!@zb)H#i{m1oB%@{C|w5cynk&E(?K^|D@VPoYwDE85MqvrPXkxD~12 ztyQE!>e<|976b`bGTCWe$2&Exf=D>zo{KXM#$0qe2*iT=8&-Kd$p>U-)2_e%Tutz1 zaySbZGWsFZK6>)?t<}FDb#4)e|I*aC8wB5>vj48uwJSqfvKb#qZiF~s={gvX4thP9 z2ou>Emc#ckJBTgyu{_5L@4v2zB^JQq87glFa%7nYLLxjH_)@~nUYTnj$ zh@z#R&vx+gap^HST(?e%xlRVNalh{!zAkYx2LM)&Pf^$utnNUFmhtaL`QOU|XLUB| zBmF_VvSwm|X4;nEr(#UxaU$t+ZV~f!2?oAp6D!Rc#u{luL>$Vt&x&|owV`e!470iJtCL^oYDlvpeM z&+v>{1|kCvs7bwb_Tv^rK9+O%RzmVdel4a@Py&UY+;fnpf0td_bX~Rdn4uvu^a*z0 zp=p5gl-}HK2zH7>T=iIg6k(~MIHELBU)`ixkDZbOFF;Inlv&Z&uFAgh**A2khv@fI zu%3Ffcd!%TH6L|kKtv22Xrb!_ijA;#kzrzBMDfesV77okaGh!zi-paF(!1vj-DRGy zth51`=N~sAt(@5DwvCPB**w43yOQ$S6%{aSmM~PqAZz`dmT3rj+h-3LioAw2(O-fHRjI!mT^vq4-?N<*AmVvY7tW=zwP4Zmo*EjmYW-aO>X#l&5eth5 zR$Y8C3@SFvZzJM1m9Q+z1>d5!5S2XXM?a)a`Ku;}`Yi)p`A3Js!vPz{q@?oMom&Epoh>- zj0i)d<)m*&wV=8QF{!C%Z(w)#PtU#Eieq|4S(Um%=qiS~Lh^%zyo53mv&K@$b5k_< zou?^u^h!1oc~PgeCCvB2TmKq7CIkbt5AT`!{IVbnw8D0@IHT^^cKk)OD()T5T*!kZ z!0bWP%J49_uRqe*c}^DYROQqN&7q5IBkCuQn*FFRk*2AMV9vP`;Kt<_kIphEer|My z4J})ZdgE0q>AYWGgvumNoZBZ*Fr>*S6kPTI62t`?Hk4UWztBOUxfHK^5B{4QqY4eI z+g*GQUKkRZ5<7I4AW7*?l&B;G%{?rAizoq&VOekfHLnVGE*~4WJ<#wyL{QY-;N;}a z!chc|k8*_-g6|*>?xn9wdYVR0A~Ts10&@vawXM<>V;!@=Vt9jBh*k!3f5akptgTHL zG%!hcHN0_tO{6iA|4yzk<+?YSxYJ{e5lyF!c6Vppdej+%BKT2~%!+(}z(v=Q02*(7 ztrfuknQv@X_paOZd3>%!h=|%GAL-va z?BlV8g+=5y+7foKWm8P0dNxTkKjqBT3hWzp9}8+_pcj38eY=68Ou-1!lVZ_368r)YcRxCUnhjfpqJ)SZ7K?*65${JA`<;+h>ku4 z5h7hrboZQp`%j3bGCPHeSFXd2*8cLdmOe&JH9$nT-9JKBtH(X=dy*}1wVTgz@vBsd zEPul$Q=rXyJg>F29sA_%e*U~LEL9h}?W8yAjx3Kx9JIHNh;#xL4!QA|d$!_YjRb>> zv42!j4LPKwN^0UxG-2SDbA{)Z?(+dc0l3{F7x@o^QP(h6p>gkjB%k|j2`kNi9T&79VWz_#qT1d4-@qzD zRCzwa9;he!yK0W*p3r7a@PX(e*u!DS(Px1dVW#R<#o7;@7t5Z|`g0kmC5>i=35K%h z%M`fb2j%SK2T!W{s}b9Zn1guMbixjDf42N+3LtEXS+Sek?se}1!z|_@>F)Q)fZTEF z@R_^W&1w-ve%s}!omxO6|NAbZ`4%dX@#1Sje&;UX!Hm+%xE5t0yl%)&NTW{ffF>W6 zJ3#i0N0w!i>k4KCi-TozR^>X4i|G?TA={5GG#_V8;J-T* zllXlx6mzv@{gUTv4xzixnUk3qpPRX7!gxA0l@|aCvFN#A>03V>kI=e{0LFCUn_zs{ z1Ww+u_4gi}TxM+k6~u#z8Z?<}c3wDbK`HHN>=on+;F(@=P&pjTub;Ph>_(CB8&;Zx z%MmiFi2q{<=gI!6v5OKMZxKBH?~w@s_1j!zNNog|?8%dZ>(k83E?ZF9W!5&PU+<_y zCa?EoRqda(o-cZVNQjA&Mq0-p{sL9kZM#`!hM?b>P;aySe^vbt+Oa|q7IUR=O?GfF zFkic|zT=AFJ7w+Du7a6-6jvHY5^f3qwKyZ9WNp$s(4b5iXj+X}xRCG!9LF;8^oiga z@mvRDJ0tO~g0dc9zEyzf0ep+Fxfny()uZwd>F=m%1n{3O{S7=u2ge8lP7>+9i^eWn zJU4VbYIzyn1x8dDnTR)e5ZjxfdiL($zal&`QH7C?xmq1TFw~ zt^-*LK>^}$$lcwbMmcp0yV9y_p)6}Y#@}$<`a5qoCaa49zr8T zwuLVsbX+h-^9sPi8YE}HpzL){4=eEv4kls>lx8i-RQ_~UJ79X`V@#Xtw`KQl??$`v z+GruGjPNUbwBh58cV->S0l#c|u!E($14lPw|Jxd2U-LV2^mUa=?j%Svl_Dutl~b{8 z12fKjqIl)`;#gocL(cj7p7%p^&ikf67yrgOeH{xuEK?>Z6Y;&n3&G9{@#DUI5hm~Hd$~^yZ?T!XxC@VXQu-QRWwgmSCneD ziK#;Y>~BxCbb`4mE2Q7EY4Cr~%Eq{e&zXBrV4`l84+Xn?Xz<{COgvq`eC-mp*BH3} z7ae2$4?4#BzY)P*>c}N;ayj=r>rwI!xUCCutJ(sR`!=<;y~t|VveeoJ^X6o2>#NG` zz0Zo$)wcGRd$;xpd>voTf(N?Y&{MKh8PL|($A{0vx;@_k z&x|vrnBPJ2OXcxtq0hAgUPaDu6j|w`EsplhBuDPdG-;#8rE97L+NieB+fw0yKcBQeL3;uN|ENGSa9lf$<`(5kKNtg zh^9{5WaR1PeB6RL+$9_AemW>Ht&{NlTOSA=KO0KpK+}KrNTWQ?&moP(4rHFe%O%yr z5CF?6@&M#kR)1UZGf{jah&$x9j>lAd@pZ+On$0`5KPu;ILwwqS3F( zywOtift2sFpu3_1-|Gc;P`kD1{;EaJalOcH{ybcJx^Qr76YSaYhq}Kn2jf1AVv_rV z1<+m5#of!QNm}a#_9H%{Ad)plNn1st_wVgr$r&E&j^rT4SdGP86v9kJ*4yemvxmV7 z3CNyy9^a3h9`@?7ZLRXoweMTn@yMPuYns^J$X=&GOXXbiaqJ-rPm}KMgS90e#{|4r zl#RDU;jBvUX-6GDdT$*3vJ$WpS2`V9+^h{5-H?$b(JG{!Z{C8KKetrA1(*vntqIJO zs+lbEU zTsB%Z^mum&9wS*om(`x21E&|eH2~A_;iP(QVy0FEv4h z8wO!&T)W~pA3X)fdOC0VV8fbIq9&CcS5ntPgx%9VOpX-8+zViUqbsoFn#D6(E#@k_ zHA1TE_+v4s|0!XvofU!DgqiOD`g}C#Zn z>l&)QXEJuOK%8h;E?pN(b@Hp^A12W+Y*DSzmuzTel$qv!-|jhXuLvw-#&$a)xUC!c ze=+uBBJ|6o)7`rm$Rf3-GjH@Vgd$Qp9$+GAp+bl_RfW>t`FV!;k~GlHNCxEhMf%cq zz!9_b0R`yf5(Rf9M?e>tLK0#O&rWw)C4#SGQ4ZOUg*Mh11#08SIG9XoEcCr}=7MOi znZN>Ga7ggkTY8gsV%Vh@kmMA_rfbm#Q0v3w!BwhbyQOIlRjtPNtn@$(M{X(LSd+<5 zv?*i|kU%l1WLlx~TaPqOP2GoRA>r`Ob*eb%#pr+ChpZ*?WdDW=663B^t0uUcg0yX4 z`EbyT7Y7ASw`vS!snWxv>k19bHCBQ~TDWCy4Af-A3uf z_3o}+G0awL+BPG$Px?BH+ZKu-fG21Nk=aK6%w3IcanXXdAtTiVza1k<9Iw>QAvMw; zT_t}oV2|9_oG}cgLT5bOt3pot(>*F{YX%!E9^Fnnuatx07$5@Z<|IcE<$_Pm2Uf+a z)VAz&=X;ERd_%V48C&RjkOj28CRhIrnQ$G>TIGoVDuJXCeP{YJ)s^y(pu>dWuvF@L`BP*i9Px%3K=7q-4k86odFNIybE2x!qC9n;*>Htdvz7$G;mlMF^>R+2;y=k=>)=DEBAL=;Av@$Am|jTL6V2!7exy} zc7_wI@UGm5T{*<&ndoL1Ru8Cs@HwsGhvKg>N8WcJYi8a(EZf)`EM7vj$Cp4=i7M}K zQU%?ht2FEt3iSYPbL>noAD{G02|K@%+ z2Hagg*)*&RreoZFl-}(7-lni{CI*#Z#3*Tte#fSTgwNG{W z%DBs$gJk0`$^5+l(%jW)^V&fOgES$ocgx2?@O5=b{m|p=iYalf6Q$fO0Fz%+SN_sh znW<_!`sjfUIP>|tQQrqYA&j%a>;MP{daQbZD%g6s$!N;aS1xKl$m0}^>n&^S08V8q zjiOwN5P+XF!Z9YdXOuN5O#i9OqIkuu$|jF?Y0W*u6CG|j;E#AgK2aC;p&|TrndAAy zDTj9JxaqsH&5kTttO&t{*}?C7cgFWgr0Rs_9lJ+t9A64VFrYHQeCD+wc|0`Up2L;N zvV0n|nD?9d*Zb`hbpMHJ8Z=qcI~~DIw5C*Ur7o&f_6(-K9!4N1rk|dHrsuk zj&YG8u(YPT>aS{T!lHXKbFlBBz_Lue=fjft`Xp`DEc$Hjr(h5YhGh)O;|?^^y&xQM zQX!EqxhxT(+V?v%vOqMnho(g`_XrCGoCO+1+Cs4rHG9g!l%mvrt1j~PioppR4Ga6E znM-J5YX>*bYOl7)_?G)e@d29i!$EZ$0zL=d1^y&fEA*Oca-I+6_SE@<_d;Rt;cvw0 zTT*Xq9@BKdj1ZqaY% z=^PVyr($GTmprXSx+k_pAi@QEGqKYn04i&Nv)qJ`DyxQH> zD{QB#-f$scT2DR-O%ua7n%hOO$@vnuX9s}mok(w9TazWNQt8CDebADvi=?Q^UcqCG zU&Z{khPyaTxE|3aUmy0>(2TD_Htj_zMFG&%usw3jcJfHKt* zN`&M$C?D1E*OEu(@RPORxSxf>5rzHjkNWh(r`6z%Cvh@_zf5Q4gai2?e=qA}z0;ig zg#xr@H|Uc=^IJV#I1h(XRv&6h^su}WtVb#mF)TPaWeSMYoIAuTo^)dNuva7olDLJ2 zu z2-Me5MDm26Eef_w9tjvnmwv8!{}wc`S0?TT=@&JK5Z6&boxI0yQX4%Z(U_WNRh!a^ zG!-mUcX)pk)R+o?s5)N^@Zv!&CbRp%w~Y{MK}I}?nLN!ASv3EyqbidzeKvfIXFci4 zKnnkjNHU3&x>2r_HGMXUCvwYUtkV4u>nqXpXU!bS!&(kp%czD7?3PR!v6wM z&v?CUB{nrb96%Pi`)0HdV6;f#=r(!+$_44DbVZ!NRugX(ps-LC7oM~(b!aU!F9(+L zs3&#cNKI`B@?uGYyY3i24u#i09Ss#Zob1+!GSXI;5Ii6r{5XEB)OVy9O~ql%yA>T> zfLXlxyCEhFki}T))2CTiA<%&Eh_PsQFCP{}JIw;p+0k$PDL2UI{FrjL9{-D-R+odn z^>Y`Wwsz@~ljuUiK9-wpc~nGriNHW?lQB%K7=gr%iJEjes$j7%4~{3Kwz`+35uA}V zCWLKSqTg`)3oJ0v0d+hYUY#Ec9(PyA$Wq1)X>-m0Q{&RM@W;I;%%l>aZ*3Mzb)qzzc z3TIml?F`1}3I&Z0w;>bjuV-}}j3F`$HqfTSzX9l?H@{5+sTCm1prSU|idVdU0b zRY>ez@DbVD7x<#&32DRZ9EU&TzAV|VoP(=uPTw^)7uUK z0h}n{V^OJ*5!PE@8fy0jSC;V-QK_(KX&wUEsAFlbYxyRw8o(dM+7lNGuCN=*`MQ6s z*g1Kxb!F;I)CcHN=!_NdId81;=hnVwjJKSJ-S0d~iaL@5M!2_a#`+|Gt-e*ghev6B z=Du4KE5pAm!b9X;Z?~?89_>AvK93$AV7NMZ=#H|NtR~0pLSN6dsjjijaE1=`HyIce zx{mk@dh*HwUN4xcNYk){Ad>8NP?Uv10yDm4LPys3!W=$s4|;)cwC6Mbi@dP?2YKP* z{$C{ZE_I|lHpdeG-q$Ak?M`dmcM=^$$!<|arcotsu|>D~C=vFN3IA zKVBw!j)OMi$2exTh}U`zDHP_FlfVj4IkCRJzHRlpimOXj0SujD7uTk&4-aouExZ_l zODM}uBh^QP-vuMDDOGA(+AFUpRkX*Vr(5kgr$PD0oM{CjlC!f*-}o|*Zd4#Ea>^%L zI}eeGJ#RBPPHmCYnrb-D5tGMNaZXdXIvmc-a-_Y9?;YL?Zi+|mlS{p6s zVgU3wGI`(^9_6RB-%1OLJvG&JR^5r=W0C;$aV$izR!?=DL?W6U0Zr>SuvGZe6OVRy zB}3wEA;_5DJ-X_081ZUn{jI5B{Nr}YLCJbuTB&yYjOj9586ueVg&1`l=Q}%>S>Ar( z&v>#l@@XE~rS_Orc7hlCrGMB<7tUqT_dIFt=~nw(W(}3I`bYNCQ5BVu0z(HwG&Byq z^-K5>#Bp^?t5dDwZ;bS{l*5}IO|9{C-$gzW_Sk1GhBlnlwwGTv0uj71;P}C)4NN}* zLIm>*$a>(61C`;aI<57tDj^Q6gThy0%k7nvgcqD#hHo~NLS-rY?ju7TVAP|njpu>T zs()Hmx?*QX(4>=hO?@UeY*!(Vcf$Z#AN!pDCvp?RkA&Uu}4+eS^kXHihA9 zkK$adgMTQ72_Ti(QlQkiP%g8z>ORfxu?Y?UX``exXwP=lsI!%k-17;(U+MTlUy8Rk z1S_F9Lb>w)e&TCc zt3TScbQ4E|mWc0m6peCqp*Ya|+lYh#7y*>1YmiUv;>SRgQO{tskXQ-_p0dn3zgwOkhx{K~Y$zK^)!7$%Uz=zvvA0d4Q?1Ckn zCmezN*r0#Wx>1fFcl#&pz0wIhAu{#TY{brh9N|2Dv%{jVyEwgY56$tsVEUjO7J&0= zF)hEjVkgHfh2`4peso5W>U66Z1nWZ1-mt%#tEDt=It=ybOiUUF-abXeRkO}^$9L;0 z&GfeaO?KucxYXyxTg+pGLF_jU z1n7|~<>lt(4Gf9Pw|nwa2LahgU}NC)S!BZj1>glZ7-*K|bIR}u4vhv_)OYT*tgW`v zok=l3LX^YzcPKsu&k<5E0v+cA+f&*AAA(BI?Vk~xNNi_{&h>DY2Xg0_T6)GIc(ry} z0j;dCuj1~2l~B^D8G1$h9DvQu0JF7Lv&EF6PRf3C*8`~TZJb%)c)c88*IyHXGen30 zJVfZdc`L4B^7tWibMy_G)!%DVQj96i(}DGl1E+v0lD8?Y$f!;!dx=PQrC~y% z%5s$gGPyNuAnW|{w1Z!3^FiT|4TJ@n(0m|D(9JVZljbf>)sEwamPd^2Jg{%zFkV`- zYw)zR0+&~9FJh=RMDuru0f=9#n*lKN$>;&Sm037_Pslq3IX=oZxk`U{Lp-D|4NtGH zlnmCUa5@d~bvm))K(X}xsEa{$Ke*;@1}8#f7DP)J&=z{*8a#8K?V0N)qpT-pXT&0O!Rdb2q% z;0AUwN5>Xd-4y>G*?@^4+dbeDX?4w9b_Q$gVU!dw#mLL4LS|L`G)pA_nWl$~1`1e2 zWy=%Fvsnb>5g`>`f6(rPaaX5!wybB9`F?M#!^)*jC< zi^zvvwHo_rl>OpoWpIe@d#V;NLl}@`%+`mzj=G=l^<2Bwjx|=a=?4miSYSj``1F>; z_VNsLamto83Bb``EXvFM)-`NZqr)7D4C7!aT_Ge_n?R93$eju3-~_fQq3qOMT&=>hty zQSuNb3huI=x7f9W{4_q{TAWCH%AcW15JB0E6%xX~FvUxi#pbX<%8Bs8!#&q8^H=O1 z-`!%Rn}#3J{UGZIBWfeg?n7D#3ri2_^ESziKCOWzjUuHhK`wFZtLFJ^Tf$uVX?J z6^@)Ha?bwPqvbec#g@eyY9DD_Gg#CLa#v$kcQbCN&5TTd{TXlPz~oGoTJa8m<7B3w z1MUBb9+w7`|C(=JDZZkLYzAt7`ALdwZkDJyGTa`!$tmHN;4{51Ecn7~?hBs`(mAR( zHw>v98sO-@^2pL$+-yi|-Co8}@P3P|~ z?!Syu@Jh48Z`0%4{$#=|UN|vza4@~G3PjuQ?rnfYcH$_I6bRSCNExK0BMys1ZtZW- z>gK;Y>{b{Na#e1jwYly*vdATNdFfWpDt!>r(m>u1?M#G}nx&_U&v7p?@nul*IeMci z!JoGf0pPKi$Y?VqXJoQL*0u}wS$Ix`;DEVP2g zaC>B#Pp3QA=~N~IOmJ2GPl8W|Ek>Z^J~qmmHP6bjr!8|X8PuA-6kCDEexSUY%BJSs z6e-|0$0rz8Q#mg@@^!EwOY>|7d4Y_$rLQ>_5v6Q|;}J=CuF%y0Gj0pXU{amkOCi(G zTQKmy5zg41OLSXLfWyXtCJvk;@4({v;A)&Wp3a@oy8xlO;462*iG-sZyoaQu;FsPP z)QqF8G>u=QVXDy{$ZK_fa;h-2mCh1=zk(i(Sb&}tq-+UhkKXxZ)Jp=VAp5Sy=^l&w zYsPGd*Cx~<_bx8I^oQB4|{$4WsHPDp$>slO1U z_ooOgKHEQdJvAw175!?+Ngvssh}bOVC=r*Qb=$V1YEnKw8lg8 zXd}PDbv;*1I*X1?(TB0b%p$yoy@+xM@!CxGvnX_W`5yABEG(g)IPPum**ieAb3J&fqJP><&j>1D@azk z`<{4E8YOJ;leIcY3^_ZDd&M(nuupq*7k(XoNr=ebOE@m;cOobv%i5u0pAyjCk5e*2 za#6GeVjJ%DM_4cDo7#>MWUZ&(y_<Te%(9V(#$;6lps^U z#{Ftfhb(VLnQMy4mUdrE8ZvpH6#i{CU`8i_k)5Yiv#&Q{!7|B`M*8v@13d(G`e2mGPG@g^JcY!s2 z!2UG(+S@8^9rg^fBdi?%IC;i+1F&MvQKlHYsObcJSfnHFY1wsv&5l?E!q zfxaA;%QLkW0Fn9UEkR~3HhTZ0Z!|iu0?t}ZydKQL=!F(+rVkUkWK@IM1LxJDx*iLj zMJe5d#A^|{L3=69z(e&ZT#lHGiImI-+I|XR7d25Vhk@in_XX)$0`|RIA~-_VwiM@e zcv<=zSc24yAkQHG>ribZQj}Ks%JO$5l(H8;nP_if3ss*g`cszjbH!)8bb=yuw<(_lhhS#)hnSQc`r-+&Hg&ctuMX&ZSZmkvVY2h=vA zA!Noi9%W7-#8fd02FTZZS_V5{ryj1J-;&6 zR556UY*Zhp@?Z(?D>)JH+?a3mJcra4leIa&J)NVyKGZ_$qVxil|0uYNJ*p_#(JC5~ zX)BDPK&ZxaH?i8L)^#FhyKT6N+lf)+u`82&#egt7wAp3z%turG`NW8Q#P5{Mil)7p zCnYEqjJ_z?odzEK3=l(fHk+R}LLPx7-C`y6d$9Qj+Pqfk-41p`ry(@ks+TDjr9qjC zN|tHiw@nU$K1?O$oT9-|UD#B6Z?CU$Nwx$*@Ya6R2yBfL?HLOCPCyHI{U`v*?*z8! z^7%5q*|x6>w4#dY1plrttnCfL@XIfO5NQ+Cd*r?I_=gXb@ltwD`47;ETL#Jg+W6Wm zhgycJuodA`!tYwFzg{uGy!Z7RtM%Qj63wttwt)70Zo+-BLvuNb_C*jP=` z6{fqbI{Nf;*;u#G6Md<^L#YdMnAvXpGtWGaSXGwbE>axsi;E%%K85Mhex(%|g$wvBJRj`X_Nz29>`dVw~T2}SwWJB}l81K6^?#~rlPb+26%97t&a}|xcOG_a= z&enLM(R#GT<1x>YEh?r0m3QY=rF{Omc~j|W`2&Ow6|Zy8#ULRELt{npt2l8jx2pM# zyqOUVn8(itMew0Oa1#LSAM6BZ)lPe?pJMrUXd2mf@LBih@M1K(i*9Ds9`3rVhdDn- z@lRb2ULa?EFVFibVz+jh^OPL?qQdI^X=xA-3`m}5OSn#^=T`z50>aKuaU|`2#lnRV^>wPtsDH;< z7O;KR+f(fuMQ-4%>$kS6V$1~?KdWNrXRfX>H5SPIL!Es6 zV)x0)%i_Esa5ydD=6|s^_WxjQoLn6LZ!1=fw#<(eD{b4x&En5eacOhcG&Sbkv--}iQy|bI~q1Kf; z6(XzfPDrB`wtWhCJ9#HvrRIHwTJU!UbI7h2s*%&>(jr{@7e=5p^q;$2ZcA1 zsf2&ZHZCHB--GOsGW)PYsREl_x)8kr4Mp6iQ<|CKwxU1fzG0`=@32aq`s-lHju%RB zCdnNLgFiAl#WRI-gU3pZ;`2sQonm3Ne0K#E)&ZAntX2#;ynyvm?6KC~l5A0bin1O8 zj9$Efp^`pA(fsCyDDOUCsW782zq?|>`qq(!JuK!|j$BIcA{BY1yAsZjSxZUH-jMLW zdW70)^tP^73|GwfZc1^BWZ7{~Q%{*oLpClaH?QeB#8Hg8pN0@&@G>eK`kDJjvHtEU z8>hvTwrD_#dpNk3@zl8L!1;P#wtnuObdjY_D9Z9LfVR?RnrFTJYS3X{)ldyel`Vu% zfAB6*H@5XM9NRlS!5Yu%OwvH}qw9dhEzX5+>-z&iJghT&VL1nv@i?jEPPb`$DypC? zORzUINq)gmaR@){GKr=&QroaZk`U1AQI*;}b_jtqXL2OSsBOu6_ZVk^F)2CTWW*X{ z7v4b>r*bm{*YOrU4|0q(?V0%oLMYHNG}YC=NMi(nO?wcF&WU4GxkgrFEp_ot{m6T@ zziTNw0XuQj`g?^KdMFT8{5pAOa39QW@0Gqv7dRZU6`CxW3TjBSt+>8b8t@x>rE?(j z7^z~ckfLGxs<1iZjt@DXp!!x_J;y05YLREttT-@q2tC@{FrV&sir{{5M{8?IXeeOm z4@_Y9sU-o1{Sw*lqJhn%Ohqb)H+J%PK>=!TtZi;E$gDikyXY`l^PGm(-m9D9w z2gI3H-4_y^?RvXRhYOdUn@Y#puIPkx#P{SxhN9DNPA$)$OP|x)U5Cg}sOmIQ}f0)IPVK&jl7g5(J-1G8I|OW%b#ve2EeZhiJZgGn={ zHUwwgbrleHx)1pK1hTO&t*ag1xk&*)vZ4{e=2hIVb-H5fRG8SAv5k|%@c>oStm2)q z`GN8<_hbZNb^*>Lx?7XII1jL}+b1#lF>ONifDyT&G%{et>ON2uBBjmf1($^q5`zhc z2kAImsi&mnI)J+0*S(o`ap$@{*!zTAn6bsU$$)V+r-s0YvD=CwaQl;As<&)&r9)2b zB?erA|2FS*=Hgs`Kr1yE6>B-i7&acJp^`&^F3+&nFSvYU2%fKuJ_>xs(ilW!#Yqd) zAGzPp%^hR} z;w$8&n*fH4=VcJ|NU`uKc$x-7Oc`G*wqopEtJYR?D`-lc6&Cxqt9oC0zWP4K6#u!O z3Mot>hI3FBBWfd=Q__CI!Y^>n13Gb+e-!EolHO~IkvHc)t9Jf%I3XEf2hR&bmx~_3 zEjkG5pzX>7XyENC`(}MN66=wm)`?_eMQ<9lXZah3#hJaCe85|>AeI|*>64xa-sGuR z|2_b1mTd;AIb5lb_fv@O&VtESV;qjibJy+0YmiXjY(mJA8sM2d-or{Kkd_8u}*Ony;)UShwogWgkGig8)#m zkD#n!n3(OsL@R1{9Wz+8|3Uhlszzyq#m!_~%gh9tel_fZ5?M;ie84 zfrdP)1NNwqIaNeb@{4u~lDI*cBl-w%y@HZQS%hOQk)1%i{C;IRnG9V2OcH?3kcra5 z>FE2%5x8rsf!a8e(P+`>&JkOOq-wzqSxhPt0>xxANf6~S(8RqvM+sz2INMM$0O^?f zAt`deJeu>U@m!s4F|z2_Q(Lw9EEQ#OsEVbE1Jlu6T>QIkv+2E^zU`?!X#%i#4f3+T z%HTymR-95Q1T4S+sNGcEFfr-=Vi53B;t()g#+UP!2W-GWv{oMqF)2~AemNYFFIZR%;5K{3V2<^~&{OJiOP)+ut>f86JP|fw2I2?)e3C{y* zK$=I;V~G&|*JgH!%PFf@2GwhUU}H$VBLmG|x&?Y<#Fb(|JXN4*T8TCH@X6205X@aN zklU{#%-y6|R;r2|gM9x={2~M-G%x!JlEC$d%$)?~1_sCPXLKPBmjfLXFdW?|hGhFM zoghsGQG#eVGg_2MG_wB9K)iU{HC~2I_tBQUVQ0z<5%fb@b&-fB=xNMmWH^~yQ3NeG ztcu&dJd{a5_Vr?W(PYqHkb5J!r0+|=riBe)| z0DY|Y>_i3`3ak@!#fv^G1@&E8_hqFOTv3UFtLRe>6rzB?m%qBkjVJpo3ea$vrT!3S zEO$jeHDZ&`_RrKLAJUnY&<@3MKZ7E`R0|-HWtFHB$>Kzv~`% zl#SO9rAdTjx6k)+IkC|@&H<>3Yq(h|w7VEBvHCT%5F_dn;Pz4HIE=~b9bu+Tx;h{M zLiJK)f>hovCknJ~j59ej)d|_Hh82A2eT55?`ROocJf+_7hmREXDzeXgt6xS1q;+VY29KJU)h~5$*@|@@$c4?MZ-ZvW-T?g37N8kR996YxDm>!BVNW z;qQ9AI^BB6(O7e;%+T6#r8>n)M<~A*yfW}ElDedfyP_I_R0mhV?vDluZui(ZrE<3{ zziy#`M{0FGWRO%gHQe_PJqko0*jrbDWjlt>H$9t#Xx}~)5r(Lu)P5YQubvV%U6IGO8AsF;a%4d+5v#=B?n8jZ>KHI)$lhpVZ0jrku z0Y@NxoVJh}#!)XZp1WLMSwgrMuq&A4 z`l*3z-)lk;T-c)Zw4LdxjL9Q^z}S{L-V{qDWbWHT32Q12vFhaFo2)Jx0w7w4egE?E z@)r$O3m%cu6R0!x?6Ny?6kTDcrq)Re;3GyP3QqL~de=XI^O78em;5FB zb|ksTMmElV-c|u$*@dM5Bhk{W+Kb01;^LMz5bmWFcQ-8s?sxL;__cHnhTEpC%m@rJ z$ks`+mb&)wm)!dh4daYoX#k8d#9h;n%ir_>;{fQOJ@H3)c@`mCmP&?AJti915Cnsi z?!J)e+tysfH-dd6(^E%ogqysiF(91mh*EKHK2k)+ziK^C^S-UhtF|GWhi^~|nUQ};M{gH5Zx`f(-m4+x z6wM~-(feidVy%i~dRH(oebtu6WQ1!Kr#6%sM-<(sLgUaceUU)oDRgx&&YYW3_gYeC z4TUYa#AdgO3}3~N8j@p~c2wEWnDm@aSx=+wh##6gXZuj?$6X#?`@S=52J9Jtz*Cp>9Vkr6*uEO3 zrC#Sjc#=cA`?x}|aC`K}BBAHx~$B-drOc5GcgOV}|;yI`I(@t(aRw*9@pX+L$=?T|qN;O(|K zZ#{;>|3i|sdhlW3SD`QjF!HAHEs|e59fmh^vV+r#T~m7qc@~-vdL%<4`r(qOten{B z-+NZk+n;SrtLxABn|pL&Z>|;h=4bFzRr3SQ?cx6W`;k_a8WktodOdJO5Syh}@ig%T zbx9f&;SNWm?FTs^WB76UMV;SnENo4$U!To=E&pqcl&Qet90!v-wLPY3SAG`<#B4bv;=^b>X4RYN=iV4*epfqE z{IdOKn!xi}UP?tpN-=cmFP{g8zqYBRa7I-rV-n6!jQQBezs>0Cp{92z6U$5(#l!h2 zuwJ%|N*qGjmBP#u4#a3zP@Da<5VC4xkhk7R9~t-lR3p6U#1RJ<_;TTp+}p&<38tD! zP#opp;cYvOTtCuo1H)k27G0U`$4rAWPC5vUn z?oC-mh+~M(^Q(xT<_1_jXq7X#z&8+iQVhknPDIDuhR6DnQ_zxMH~Y6d`%Y5EVnwJ0 zOgo|uIHoIOy_bBtE1p-bMvL0FQ?;??%lwViG{2B88xgxUQMw&8Mtl&?g2Os?%#1fS zKWwP96ZgwP4sQM!H0DaLWAUwzSFF*zs9)20WDP@dJ%*>;_8e{=R|-5HH{b9GGbIkPD_o8}r64dC8wFV2+?o;H(WuWqay&L-!Qw>t{$xjtB+Vq~W*S0}9 z2I?&we}of(t9fvXI7+w0b{1lYx!*N^Ob)P-4AnWf9>-L>=x;NxLMHj7ykP=qvL8j| zV+}i6Lfp$KfS@uhKie5JHu+~lPj8J)njYT8w8hFU|aw90tG@C5_{uP+456T(kEw&DG;Q=E-=%+yA>FMIkwcIG{u zr7a#%V&aZ6j;D1CU-wD1%$xo8+u!yyUoLUH^yx+Vt@}+LzRGWkdLQM5B&97kx9?rL zdN%p}fJVv=>0R0X8R6yn|7P8_PT%q&u|s+#Y@HvAsCExEBmK(QSq-}-vt$l>MFD%7 z4MLPnO4dbrcmDhM z5xRILF1z<^Mw!fhWS}}_zDYV_2K*#@dvl{KH6y;ea+4%>sjiMKtNtD>lID0Do^f=n zdcSZ>8&o{&V6Gm|@?iIDVfhZQrAz(R$M%bqjP}Z+tC=IRTv>WIBWWhlnO5r z(BIB^cL8JTm=KWqT86#kgaupXWeI2$od5k8oxGL3Jn=Bkw(mq*k5eo;D^(VKPc(Z< z_C{Up`IRxlE|=vp^~n`%`QCbR`NQWJXeV4d6Ew#@^?j;TVM+7DGDaH8NRZ}Q&t4Vj zIW|>RO#N}Q$dOVZ(R$$#be-8|_fF*)S^bUS)eYdWJ0PFA;IBGf+HEnJ>b{QjXRKPA z)n1{$Zyjke^jHV8%z4V4z36P;tiG3jj#{WLOr5Qk%aI;28(kc>RiwS^rFGPN;k7!< zLnu{D0nPkwnVTTPt>BfmH8ie}=>3;Zr@MiC5-lY8(kobD})K|%$NTcr{f@L7<6($QcV1l1ZK0E zY??ovMko(i+&s2%jn6UE(BPA@2|3jTC6}Y3!{@~TFxfOMaft6bp{a+F-Q^?qkqTw& zPPTzue=$D?+N9a#9*_9TZAFfc$J%xSG)jon1vhLm_sm@`bj{7els;4tH7Ws76+ctlcLa!O>+}yT+Q63s9 zW+k*7N{V*v5;GUDs<3aP6a5dWvy(y6c%wAh+Vd?bcSwo2`PAZ^&<`4f>#g{2%EfPT zZJP**u=amha=?cndv*E^&XI4ick8q#$V<>b@yR#?_chjN{C>rLrT&&;I~#~l z&gVn2YA6z{<9lBGuCdgyAPXX}0#j&NQoz<7q6LdUBe4&nBeYBiOe+2ECZ$hS1XZ<* zDeyT1n^~ivviOD02r1TRXm?CNY#J||9*KRzYu=$8v*jV?{UARUFQ+}t1`uD5a=T`x zcRub<`x_WZQA)&x0j!*?nt`d!-YOE_u~R04pyPB7UrBu0;Rs6#az>RCndsSoRu!Mb zB!$guFsa(F9qAzxHcBDDDMX=m*v2KfMBhRq82;)P2^y$jW#<6_6Q#k8tnUl(+C*p4 zik}k?<}mVi|2`H3M5{|625l&zyi)5qftBWqcKP|aTi>>GcdhkizI)ot^ou4xS53P+ zg3XyCFX5+Zw730CwNc3BATItW4b2azi zHBZECXH7WY!;+L^LWKxN}T{^Icrnw~Fz46d2^8f!Xl82Np$#3JEhKV)bx$Z23cwS5)e8_HM zHAg4FoOi{7Vg09D+5yuKUasrV@0&eoZ)6P*A32_>1<*Oa+jtnP*`j|#xW;d6gZyPy z)4OGcvmU%aLLm~wy4}s+PJO2VkL3X*qC^dsR@M^w2479jWc`ZB%X%kD<>MPtY6r zM2MKWEa3+?zH<(2G@u;8v>af+48EU!sQU@Vo~S9OS~t~MsA_N) zSy70`wU1IY@N;HjgFga}XvYfLR=>0rWIMu3*!J z7=nr5Kz`c1XV+YC&F_fgY0exqUzxrV`S;jC5>+_MI<-Z%F(E`pDFg7>F z;C@o20|{;Z$i=+^J4+^mGSXB54y+*SFEH?B8buJn$ejzMCB?b1Qi8Woy2NOgfE~qF~xbhAM3E*?L$~_q2QH zvTlc~IL#E+i{=F2=$PyK(w=XJdQ{ii@PJXT}KkFd>gp6!=1%{=YROq8s%%7Y%P z9@fkkDc5o!~#EQ~@&;x<1zQ&1&bw*}R zKu%YlExfVE86ChUiAG$SZ9L+w>iyRfp9exzVB?BZvoDqsItt9tUI!oH`Xwj;qY$!f z%AfdaTNT;vlr17#BmycoZW!F`WXml?Utms91#rC1%Vz?k)?#z?1opIz$?kM>xK}uu}n$DEPcU?#?fy8 zdJ*BRaXrq_Ga#`W?At|5FB3oa`z)ve90qtYa39+U#Zou`v#@}2CV(%G)vsm`8Qas@ z-QJE@1vMegca#^^zyhkkoxEl2LyuEuBsr-HZdPmVjbx9C3HR@EUmrhMcv$bqeTyQ7 z;tdy96{l#uE@PFpib;PMEc>HH3ZX?)ZVuo!%@UyoBDe;qb;oT5$s%6-h>sLICH;-JBOO!s1*f4Lo*#wBrFs*cx)Yg6{{!I8 z4-BgkpewN^?8-XJ>aud_i*>5 zoJC>#J-n7Mp6o(KwEMaHm)X@yLSo_AyJGzHrJh<2NQEQw8zIN%T)Dse8^#YJgW1(_ z%L2OchhVKqU_R=R*x9Sz^Qx-P9bWgCL%iY;7b?0_kQBPAV?au#F9MK*XXcGGxQ;yd z5Jr@(-p-DdKBswKX^q8FdQ0jJnjhAb5C+R!R-0?WZV#~r^3g|4DdAu2)xSKm!<|lr zH^Pw~H@#6RF|s`vgXCph!x6So^OH=V3<#V|5xxhwn1+6Fsi))?Zy^jYVYI2naf?L) ziB?ZSB7LzX8}Dh~T*6?7;Qj%uVI*tRqI{UX#*gLSqfghUj0foeApS~f5z}usFL6FPbL7aw0BR^Z7X^d?k@0qV&Mm;^AgY0vHT6Bt&KE=X59FW0 z(Ixv`S77K$IB|2x>{JRCA|}hE^rg%kPj}a1HpbLnTm{}ZMGl4tC5~i*F zQ`{I>0LoAkJKj_K*Y(SS4~_fxQzbX3x3dQ)sv*1xrrL6=T1Rb=_1nPM{oG3L zRy$X-nGddt4zDfe0U^dp+bQxLjRW`eCD|vG)W}1);TBj)xl6pfxgtON^pEc%TQN@V zrA5C6EBwfAzvaA&!ecJ)R-!f%>7Lg|1e5F@h9WG0t%gwxl8`O>tWVjjM!Hw2O`EqT zC)DyVDm)$D)s86<)kc3@)omYC(D4`12aK|L;IhKwpr?=u7Hk{- zyyXoLS}6wCJzHQ?gQQ?C*4M`Ol;-GF~Org-I)aX!Y{i0L#G8_B%g zNVaE@_0OAT|5w_MVS)xnQAMR@`OoH_C5-LYD^ zU(guSX7ZF{vS8%_2*}n0KgPxQ1F;}|qN+SPow-pk(}Anpxx~K8&XK(Gks8JWjsM-M znTLqK*x8kRQWCeY?@&!!DKr7xDUSyhfD%^YMh1c4g-BofNAd6h zYe+PB^`1Wgf9RZ0{7BwNh*XQ}ZMG5L9a(mKGd{j1_76a~xPw4jmM?``yF0P*Dw)i? z<@yol#$}kL0Dow9Ikev$mXp_|uoek}#;BxOq+*OP2pRUa+y>vkX?5W6D4seFrNjwB zA|$!48%`a8Q8$aM!lqRd7ypP8#C$SKz#%z;9lGriJ=sdZpbUkCwlc7k95HBzs8sdM zkzQKr?u39kJyQKDoxGjrjbS2iD4X0~-$H8^@)_IjiwLp|5GFeC`6YvOTBcG%l@pbG zDU0>>*I-1o?T&|_aK5rHQ%y*SuM50teQQCT9F9a4{%;*EmIUWeiV!VUm`&AY6YW=y z=J@2rP1W>TQYuJ2B~q!N-ozVCni2~?(&`>6D)e%`p~5F}d#XA+9Kk47Tz-LOjm;-JP0$IMMw|yRAa=AfDSVSou=g&t0J9T$1 z5S7GV-kHxm86chbKhIx#Un@I5X4wp)*x~~hcA6ewg>-Y?@5Se~n*OcY`9JO-AFlhB z&WNH`H(PQME1DoXAB&NMrdZBIh~m0>Gzl8lZTG!CE{_}@qr16;3d&$?>q37x98+CN z*cD6{gfqlB>y2Rb)%%d@qxvOYdfX*s6URrmhNDZCjb1`UvV|)z#5Vn=<;EVEORjns zpJ*ERFkJhRE0dvl?Cs2!JCn%c*z{R#HarsSWc62i=yGNMSBSJSrs!(7y->%K@uGb84kYPV`X{f+MJ1HqFZFpzRQ+`a(c2mY z#ZgwjXXkIy4jZod9nP9~D7Gaxk}&!f$-?4LhP=nsaKK)!b*Mu0OSsKKctYoBHbNaSxa_+Ynw=V{EQ5;p z(33Ot4E1?rWURuj{#$}|WRdsk3q5-)F({n|9ggML1-kYSpzJ`oFl4rC(?-uF;euQU zm0!2ZU=VpJhn@TzbM6O3YXzseZY>Z>j+|Rd44#Y+r^fE$&O?3POBJ{O_rOqD2=+i2 z_e@zO_EhvQD`L=io)OHWnZ`1G&NCUAv;eqT9KWn(EAydN_4SzwR}e*=eejv8YdOOHrW?_dMg*c<-WW9%#k zIFIsdg)iD8*KJeBjEeO24AU|V<{Ue9=YK*3^946vnzD%%*Yr1o)K`-Gh~6^p!chv@ zSm}T564Xo^4EQ24Hp5>+G7>Rn%w2MrzzQ1j4Rvmwq7za56jWx~aFOmk=;uZw_F>H@ zaP!Izbn(1IK@aZXUdJ|sY=@m#UI<#%w;bwaX!{4j@H?4ofo@5pshCmrJXE8lI*ZJ@ zpLzmzYfccX9^MaA`un%GTRj>7p=#dGzCrZ5vYc+wGDbnSSqB6=JAk)RStu&kdMwic zf3@%e|A^J+yx==c#=6P?LvTB12{Oe~7GIG$4&kaDNF`r4WM&E-qubR+Hh6cwEC!*M zjEXOF?B_!45Qtk*07BLuvrCiBcY$c2$45bfnOD8jP|Ek%+{TmJwC61^TS$s|)o7>Y zadR~Xn=+yHb9(`?R_s!D;aWMNvCQjQH~m|FCLQ-Ps}!FgTo$?I4Cr@MI5h%cf7##r zQ$GA&fHZ&;UEpwfqf{vatu$1^29-2^=8-vAkQxTT`Q>_q<98e}7u78m8p5!B<*51K z0UPVN-3QChbcg$_8>geXTLbWSK{Ipspmvpo8HQ=PxpM>*;143F?*(-hd%9l|@I{z< zq!BTgBTH-ZD)c^M6^^#34rRmsD z64wAqmzx3#p$>v-h)UG6h|J9=OwVvgrUYhYV1fj%S)Jo6OK<#D)WtPw5|oS1n8t5s zh{8bPE@$;iIar==duzFQ{5i!J>O7bfRF1NcPf!dR5TAg}>(Y&u^uZS&sfpG{wM`kI zcw2n%#Q#?;AOT$?;m-FN`;-ui2ka~e91n$@oMTIn{Y5 zRJMknJj2F$bDPg^3AALeVo?Q6)`sp;b2-$hJ#mRawngq9NvQSm%)z_nH9}#qFO82O z0Fhn-rnOl0#WzPap+xBVS*e08QO0^6Dgr;~tKXtqYW18j1iZs_!i>5L*L6}&K)>W))nKO(buTd7-ulrKC|qoo ze{@Ut(4$k#do~!9^}vLF2xicYk%a!?2Z-~$A;a>-Znp$jV#d9pXi;DTvne$28PxI! zoM#UNL_g?_m8D{d$L7=~dm|>Xd-bioS>hED-76^&nkz9-xWrvIzLNo_e^4Yo^DLmx zc7;Nkq=rx;-LcBPryp}E?;Yz<_8l?n#su9P9i=^1z{z8G9m3Zxho2x%+oko8&H6Ik zzN2i_yn{qoEJ%S@!zt-4(yy^VY(p6>WFEMpM-Q@e$j+?IfqKh zB;|@IG^a*${uOlPH~%d6}xT~S2K=@1rINSl53oMIKFO8+MyV0Mt zqi&uXZ_-+Z>F-_H>?w;jx>?^lFC8K=nMz*D&W+?sa6J>Sa5i>fw#y@s)0?8(R1!oV zz+->$QPdK69n=6jmQ=4BEk(4ESAWhDFBfbAtP#COI|T3J!RFM3bM$Yhk+^$uB7yLG zK_=Ys9Ge2J4uMn{rA`s}_!w8^mXjX8_uk5r<#ViM%u(ub@ePUipxs+05xQX~VOwZa zsRLaJAxsI*DP=SVXVM2#Jf|P*1@;Bv#G3la=l7xK!ecOos#BNL8>B!Qt$F{GXz+de zkzU79Z9N_dfA;;W5m1Vi zxSsWtB;g?fh+>QM*wax?H6wvL$I>*lu>8{1)-Xz4J{$cyh0KUS1B%Obn&ovO&SC!b zNW;Cq18Ry*Mp(Ly^RqUsRa{|IjcQ6(TTTCxiiqJb=E2AX5h&q*DmAJzWI%)fC=H}b zBdKK#AA|{@F1-W!v{h-P-aNFKTF660ikhna6m1VS-bc}kt8POoIoS#<-8TuwG41CC@n8qxvmHr++bLgq){3ZZH%*# zVJfx%Y-#pZ6^eWVGkU#L1*N$pf5Vc}LJEc|P9m1ui>AG>f>bmXSAll?Nr#`BK{QZJ zY=q?;BO7aRF6OEJ6;t@g&r2k8Ok9ymy5D726c6zgy0$Mx!a|#xm>puf<5n!n+x^zJ zz4~U9$filG)oxjKO_e zH0v&Qey#7zOIQ?lurEF#QYH+kp)}JThH!m7z&}g%x1W9^E-HOhfmxJggV~L9K?4(l z&R6%<>yvSSX~i*?5_2_5r|>aZOR;ssga|tg&eW$ydUm6)N}(28e0Q}TW(A$!zw>f8 zNC+`6>+0^k3PVFtJ3W9jvt?(LZz9pA)Z1hgNbz03Q0>U_JtjrlmwVOD^;MCQXWR%^C^?;{%!Pa-pJt0a5U3}#DYeW z3q~jx37JV@c#7_oYvbe?CDLOZwpHT>$0{u~Qozj1n#xgYyq z3#Pe&|M092AV1TXIay$5p02-?@evJvz5M~%n;ku*8R|&R^aHY)!E()U^Ppzi%x+@H zFBgQQ(ARLc#hLEHpWYn0+67eV2&Ad$s2ys5h`=+E9p)=o8bx3sS`~NLxK!2*5t&#N zzhF!_=^cg|!!>A2GRQ>g8C3zn9{D%bQV?vK#%me=B4Ut z0mj0*UE&Z<9G?fWX?-cXCgF%tYzxN{2g`G#-fd}-=p%G)70b{muH9b_QWS_PE1ue> zg-G^zjbJ4ac^Z6PuzvrY#G)h4pVib31Yn?0-e?fgXhKLK&dWLG+?@>9HO>AYV1xY6 zhx5#}aMq)h4L({wCA4efxYZR9(3XwUev9wr5Z)>|UlwjB>ENqYYy7AM5I>}R{7NXA ze*tH0=UPbFA}V02UMJf#99}I`3_Yq^)UJ>f2{&yGWai0n8nG=gmT#x_x<;NZCB)i@F_iF z5D|$p!7UXJkgA%`hP_epYyQbi@aB7{ z__uz;7}2LY0+ihfDV{X30KPd1mo7tLq|&f&HLR^(6711F+hv6UVQL9akS@af+=eZr zDYzWW1Vu=&XO2%*_Y?AhAnmW&uLOq64U2G3>3xG$_$Uiq8;NEphPDTOmdFMc<{n2P zef8PXc{q@jOs0$wfA z;nTu2+X+V&x6WXsow-7l8!beDb-;ANfywgUw6NUqvDiu`7S0QsRTb^|(T{@!`;J6i5@ z&I?SpM&>L|>5~{3`F>r+z|}=8#E+kxafzQ)TRXnlF%Z**cHQH73d{34G*x23w=0xd zGpFM_jt2jA9+W%=o>6Byy2IwuPHS*D1k3nH@Xg5$r_^0b2y^<%PL!vte;~ zoHwi;q6Ch&G^PrmK9Q+#AC6>hphTBJa}-e;(1L+dry7Fb59M~6#++%kdpSr7UzuP{ z{l$_2&kK4+X$5Z&83x38xdjKgl;y#$X?(?BW1zk?a^LJ7TfwTJIMqk<4;IVzz z@d_n>(G6U89;RG!c_W76Lp==`zbU)0kvHS`C1@mUzIrpfOnvE&R~gkgMHMh~oh#Jk za<^V|2lFZZLf%I41P(oU@6vu|8t6vR8?dvjQW}Zole?2jv*1gp8Hn5vuFdgg=Y0^@z_37cJj))k6E9#}ZSW0s-5zP|gN`hvOOE47h()m1B< z5N?)kWP?-085*|qW;wV;`)^?9t4~6?7%XzcpSD?5Vj3uaAUFJ+&SwFm&>$OI#S3qn zKwoCShU56jY_PVUSsDZZ{{x(0d_x*iBVIZhz7Lr7%HxHHW=#XxN@JMddDO|N)Q?@B zPUWDW>s08-z)z0jbj1%{IMA~u4qnzTm$t4=!N5m5P~x6G28%zh+VbU$glTja-uep4 zT}4#hFavGhR}wNJ$(+*3_nT30E>?6zR?Gx*f@z4^Q|5n_{}OP3K#-fA+xv+x@c!vp z)Bi=wzMGi(c%1H2xRsy51zha6Q|PfGy0Uex{tj zleM3yNjki`vQ~5xCMP$ML&HN!k^G$X-MAp)BvuEJT67tcizKrHtN5WI*f&l8Whx|p z7!VUU*q1zF&>7hGlprAO_%tMxJn}jJXL}%Z<<0vS-_fJv^`lqB-zV=Pm(oYJbgGSw zkL;1W{{&*b9^BtH1Gz~k*JinHPu%IQ%RAFpZa3Hh<>)L>{t5X{In)xI#*v_ z`NQV_-DVVw9@^Y2^CS4F>b2CNrl2OFtKco3=X)z9&YK9IV5yo-W#H*}?7O;?k^~VA z!ZLH2lB9s1e?nbsI8rxB_tZAOHdyp~df1v3&l2wCASj*4pbRcSt=eL_2*(iP&li1g z023SU!_hcjPPsT26Sn4qjf5$`0M-;ufGV?s-mtOE`|3Ep9yrwW)KR)S57VgTk;`k} z6uFf4`2dYt9u@0UkwOlCd{nqM67iB)XddDP6h{#ipgmcUXokMD?6~ zDMElb#niNq^gtuVu(Ur0#2oQV8rIh7W~j6lIY8jYcb3{A$h#qjC4k*z`@u^5thW`S z4w59eSR_8LYq=*qe}`2HC!V`-lOpA%$!=&v1mh-!@^AORx`%#T$>?A#2@;H7jtLL% zT4^2X;@FMg6>;iaE6(UAJsb0bE93^+@nMH4W@zw!<)-ZZz%EyXIF#sjrN5eNJJuxz z>N_HFn22~-kaW@W)L|!y^&3>S_6JsYWp_qDpRz{+g`k2+EgeRd8L`Wl-_I#~D;+*M zWeKpcN%M6L^cC(D0>a2%uzW?L&=Bm?LT*Fg4tQet*DKDb9?Pd;CpONB->jO-63bf0%~sT%beb)$j?iK?gik_Co zPY_Uprt!LV{O}{ByU2#|VZeW{dvb3(}_a|}#;$)4v< z=kd0Ze(G(BYjBy6U&`>-)!a1JaIGC`0SRfM&C9Q|K_k0Qwk8uZ-tGdy7Vwc&7X zFGG_+pME5%O+t5Aalc4M6rq?qDf#7*%0R)w9VRSl?Rv@q1LW^hoi*(3?JLk-Zv&CC z?q;=$^Ac=#C{6&0&8`6M(mdVS_^%r)em{vNz4TEc0h5VSN~ko0vh#XM70TS0^EDK* zP>Lc6YY^W@_&innOXmy0fgGdY`yDS7DOrHtZ2keZtXlwA+M(E1ZW7;9JuE!$)C9cy zfZ3KB|DYAjD+ru%6xH%)M*uu07;%7io%ZA(F z_DonX){TJRj3G-Bq%_`bL_I4N9wB?Xh^%SO0Uf*;3}8mmF!z!^z!))*<%M?167l<* zKBLJnTEBGTfh&k^rnF)8#mPiX;wp|itWv_4f-E7U`0`HKt1Ku8^A!-*PvGs(Kqk;g zn=rEI#%V-`>5*t#0AGG(3zS@4fxZ!*gR&Kafrw5=Mx6EJsSV>q*0PZ?Gdqo4O)gY& zJTaG})B{re3#<24X($$xg)VR7y8~y+#a-+8#=^6PQ|9s$(9{hP!dUX?Aa+rOW?Ad~ z>|a4Rl@tah$OyS41)6C^P@!Ya^)iPhKt7J27oY`J&(P$?0^q*E=11o#kAZbJVx zs_j2FKq=egJK{pp0*c~~8jtmM<&Bzug-&3V1_0=}ui0CS9m zeo|{kCc=SIFr(;+esyX|wAu)tBWu}_4kryj3{C@;|2BroO)Ob^CL;63fpp4~Zeg8& zK7XDgtgz9c>dfo#WU3P1Bn}q4fZUj=TPA3 z=x?JuP$3VGkJ#2 zdj`ODzmw}Tj-pX@-IfavW+s?WSYxEy*JC>EVRf*Ef}NQN8BMj+dx4a3XgZc_LgC5V zp;k5%t%If2M7?rwEm!G8Jsr508P=4JOyyh{Or#mD$X}?Z*Uj`Re6A{T&j2HFM>kXz z75e#L_Q_lO2GQbi`p^k$*l0no*4v_Ixwxr*EEHPxs?Vnq1HY=)G@Eh1JM0Bn>2lR1 zqUyRJc07JkFqpWpus{9CocCw70>T4?jdin7Y}JUNqJb5B7*99=^0IQ2X-;|EeH<7I z`ldMX&)f6&1UQ>SYl_KEfAEQjn=J;f&GPbcvyjLuF+74dSmfv%WDRJu`ujWK&A(^= zK*%|&D*mN_%IKu^5f+vN_~JwS!_PX)~`{6Mi0!H(0d#iE`VkGPb z{59EKFDf~_=GiJ_5SOyGz19T1y_o# z4@-qq#D;SgPLsO4P|l9&-%N%%d?#vAh)HA=Jf@ZVR=h8_*l0p5kDg~=Siuj}50JSO zp}lP4l*K7>^ys|m>mpCJ%*S0C}334F&6s zczRJeI0H?tOP5Qc|As4vxH58C?xS9 zFt};(D7*`{4I|iK`uBQlRWB=#(bc6PEjFCAhn>ueUor#7ILuCkjyEL`xg(g1h4^@a z?UWs;Hf)IkA4_{^Y^ylRygDl@5r7z*=Z#YnOLId$8fTOZTbI9?F?*!~cLei!sq%zK z;Sm|TpZinS(7N8zqv;M{dK)`R&+B#_py1w6JAe8;dsM~u*o#etLZZnL)$)Zo1G(^ft$Sr(A!NbhN*VU)bF{Rql&%JiqPF-%5@Rs3q9wx`4_6w zGL;{+#HGnrJ_3AALhX|DkLZ&(BKh1#8!=2f#c`^6rvuuZ0Pspu?w__*Q$Yr*Dr3B&P5y^LDmk0h3C1ZH9Yf-v2n$X z)=@{`cv9r5#~Re)Kc1=ICQwh8e7C#49F$q(rG`#CE62ZMJHR+reZr$=tx;LsjfQ{l zsqZ>>)@&Ak>?Lm9MohPQbkI{g?dsaYsjYDCi%i%cijHL7m^Jo zvjJD^(f1vv#dh%NS*9|${WGipI@Btnl+Y+?F}zpNZs0Ki%#eC7r)cDWWqAwL-Wr)i zsOCM-O%0&EZIZrvx;j=4;p+PO4xnwraa%V}>}9r2h*^x35ItQT`PA!gOdvJk2cMXL z9S}qZ?-CI7Ml?-O}5SiP(#r>AthGFrswMkl^wFR49v!+b3e8Jt!b-7X;4Z5!|$FE zeG$oWdKerN+svmcn0n5weJN$hYww=&#okb+ zaU>gg8-avuh+XSEhJ1!wqoKy|ySs$V-M{W%sF7ZpRQT}s@C@0M0D6}OqDwK!K{;jV z((*s%!od#!c-W_&qZrgByv)1Dih@O(;srZ6FAX?;%>X#}&HG&-zxO-bB>DDpa8K=m zOLp{;EAwBHb#1;DCbHpU!|iG{?@f4dwXF0AhmA>eCDJy5PM5eobBTJz6L+QYv*fma zU#AfxPv|Kys?Ex73KenI_Y+94@ma?(Y+Vf~Y@~5D;~KWR@>CX@vAFr4bKTKLc9k2k zQIc1TLC|}fQ?8eW0&Yr_9cr8!UFfh9;X4tj73YbtgFgP12sfXCFrdB<<&y46-482d z^ZOm{pJSq}Q+>5=lUEVoX)y;n#ILxuThM|RYi6#)_A&!ow6U3jckpk-O>I$ z&QuH?eO}cNV9Tg7QawkVF=#$z$}JfH5og7h&WXVN%KTSJ=3W$6vAX ztB2*W97w5OYLcfHCBL~ufGzBuF1vepTNfXxnK}ivpd&lVitQ^_;qU`X`B z9Ojdu!VgR$K5nFeh=x^uJK=q6qIzyu)DT~k_;b`dd%|^mW_W7HmAD)Gpl9|tznGLW zDyA7{h?wJ8l=R9~5sDL{US7U5)~Rl8B}V>;>tOGE_oZ*={eVsNE{jVOg^(@Qo%D!% z&x|_PqBMZI@}XFU)Entp?Tsj;eBk`GnuKr5atclNcOqmMlEnrN>Tlus? zI9kB_NPrVtq*-%ya=7!}ohZgx@=E9Lg@&yp7~!PYCVr)I=xLK*!{aG#6N+eF-3Og3 z{U#Ajd2>iq(R(b74RxUH?v!in?KW`)Zohsu=GH$`B%$cWf09S4v$xgzp~87ZPX}qG z=yY&0Mg44sgJv}Hv@c#^yA*XDI(krI25tP|xOR0INgTOav*|MPVdnr)Mw4+RA5H7L zxLNmR@92YvUsL=MsP1Hi$m8YDQcr|s4x53FF)vylGg-+{pd41T2UC_7C>j0cG$W}_ ztH12nS`u2sA|Y;1yZ?NcIkg5Xgy z&6?6|Ylr969yUXb#|J*m2fTp{g__wV0Tj}~X7mYb{(H~nMD>ZRr3rNC>O7}&c__7& zmrDRwQu3P{?_0;8$CfiXCutZ|(;1u)+-8~GH`5;7Bg{^RfY@G*JIasGMb}ws;C8WD z&DZPDtV$oI78a}e*bJTy?)HHjp&d|Xcs%k}q15XGiKE5@C5n>CoHiT625CCZ>5Q3b zgtM5eW)rc~-6oR+vyBJExeX4KmMRg%Kf;G82kR!yZF_c+JU3}CS!lmL9M|592c26a zn42xR7Ss#%X}#CXcS7CfurhA2uz#W&rY)}wfTNA9hHJ7{&6Lv5j?SHBr~1PBZ}Sud3~#9Sf5AhOH&3pRi_Wq2iTWe!+cPRDh7X28?9Iw2M8}=wAOd74BIQ& z(wsa+bPb2}EE{BL<&h#?a4^6ixvDc8bh`=|`l9CDyZ25ZNH&jv0M>YXAKhsdT`siU|JZ+R0@SSr~YW z|8c78?alkEC$ZNBAJ~oOG~7}XHByJWV$X4yxAsZ;Zg0_r_E>c(!(e`^i0axo9G4iA z6hOgvPC*JF-;I2-{TA`(utuq}{!?+_#$ZzmLJZU-L;7!h(ZA`B>KZ`UD!o?4_lR4X zIum66bn*jIRd9llmv{a5T_b?jLsjxk&;vOqItP%Oe>-cGgwvAY8s9*+MIt z+TXy(YnzxLtUYqrFazd9Lf43tD@BIFyN*T|zssGMLq@j9hVk(Hf_fgH>nSliL?2Zt zxB&Jl1AJn=1~$M-!0t*0ghj!wK-BdgL@(VxbzG?lz1G9aWiGa22GJMLAA_3NJMHeQ z?j31(_C3M$Gdhp7)q`2bRC$=uwFl(B;V;0C9CRCDF3j}Vx+8Bb+oNSdJSg!_v7(*oW^Mm&U6N>%RDD(6Po|z&n8NxAUaiS3k}H@%ubwYRN#> zfj)r1dH`XLZ{VNWIwsv%oG*1iP_5TIad_WE1OG0S(W3AI8fXl;o=3R_5Posp)zj%# zgke3}%jxP4Oz^BH30l$ASzG;ezhrj>=2?4rexPfF=r>0W+!r7U$vBnitdBxX<*M`Q zAL+RR^=Og$)}D;ns_9DJBDacJ8sx-tvce*}I_i7tX6@RvE1HtUK#cYPLh8sscBpS1 zPjcO>?)MOs-updb@t4^MWReeolPulbm2bS*g#MJ>r{neG%w6#>n)zpG9Ae5+<(H26MzBl`+ zBjJHsIw@G@=d#jBG`rFM_-BEiYF_S%l)KC2&D1AU#E$W&jq)`B#TKF1{M)!}>e{kE zId94RKZ5b+m7-!RTwpJ3N6_&zm0DjWP|^O$lPEEY(aBrxp z2hS^_Q64|6LoC=cIJ^HfA+sBaMReF#caP3hw@cHBq2&gpqF)l5m;*sxNd8=57+k~* z%z&sJgYh@(?O{W{Q?*h{@{+*@p}(ic#(MZm;0C)nC%Fc`Bn~Ds(j%72yUZ~lJc;~S zzIBDU>0n-GD1R+}Q{H;sFqvE;urWKrl^DU^VGotm_8Dc^ID1ht`U>S9YMv=jRQ_mr zMmZR?BX>pp+Dfqu?g;57c8&3&b_q9WqS;4%614aPACn!F$-Z^X7?%}+e`h*6Jn_YM zOMMxGduv+_B@Rc|o?z17eo>$m`Y?CcK{3QHQH>C~c~rn8G)}(yC=X1MG#SGf?*D?4 z9jzzJ=Cu%<-9bS{F2UE-R=}cC&~82e0UnJ!Z#-;*-~)~-*L~0o2bIiz-fjs1LiUpH zlgy%tE;3rUN=FMCDDDpwPOLorP)6Y2I-K*s{_7ia>ou(ZL3+AMqHsNExaiE3@EH~5 zso>BO(y2LrK_*;ihSfb?QrxEG)oUfW?jSXnJVk*(5YU7f z?rlBl9PwmB&0)&(K5IN0UyvW->Coemyo~g6s#3PM@5pf4M;%{6NS(R| z+v1#Pt)+yu2P2h?HNr#(tTENfa>_j^Q4)$|S6oN`H4ci@a z(cfli*fchSzCEh2NW*l1j+ zo-yf}uuy`~P@wgX;HGD%&nlzxwvyK(`MCJu=&^h>kgV>8(ZboATcRGFg>-bNMQu6Q zwh2}P;5N^#zlQ;wte17Y*{h4V>@jM856qVY`tUCcG2eU|d%> zlQ!L0T&$>{8nQn%%e(8n<|~0AQxDzpO0P!9=FP$LtNlCsA>;$E{5fE|Jp6cJ%`say ziP^&nB>?U)m9PzMRZV^7;A;cD6ZvxYuIVbaY?S!aBVz=SF4ir z@z^o`?EtxZ@V)GLY_fI3Y2FMv7rN3D(O zeHUldIWyedSHb>80EHI#x)d1|dw_#b@pknleSnUZwT$%EktNbF7(6MX4*7ez(uq+! zXYHZ~PIrD|z(pWY?x#IOpLWZ6Bq0uv>^-7`?m3u5UvH+s2O56qD?~yp61{=QtpQ&- z%UR$_*N4Ky_p>bb#+IHtQS*Ar3b*CVwlY|b9ZDb~pG106=#l8#oEWZn-e@nXnf<-! z2>O=C`29UT7gAkM2e*ITC-gt z1h&QvF10!m_~IzT7{~5XDR1S|3fJbNXvt#2IH_OAw7sU!b`$;(!#cMgll7zbiN0pm zX@5!pqi_=w?Sh0l9ESGHrU%U5ogVEbx^b~Kz-5PoDTnF8=atXtO3PWqW*Kn6f5vmq zfC=>JfxRI2q!o}vW6fILXmh zRp%8drErJSuXNtI9#?n|i9dgGX7x<4b_4`xAW-S^;gQvoK-~7LW>Rr&!Y|n_8vqXWUr#D6IMgTsPHJa?a^y$~sukpL9Ue$M85c3(a?MzXxLPdnKueku zyASYz;JyY+_y+jt(NXoAr#x3mfj)x7??E3rPeO%D2lQum?&j2WU)VUg-sr)(C{3o#H8(5jmvo@D7C?A~1bI`R@P>Ls~=e zH4P=w)VsbxYC9?S(`n9uvuQ`tWgQ<#O!9%lekPiIcLn$+?PMxIANn7G10?HnGDZqH zjK}gxT0!;E?J2L8NU+B?L=KeYaj8x?z#IYTPDAh~AQZVLB-%i3k(%o$Qhlj!0~OJ9 z13|3-6fw2hQh&gsUPU@^hSVr$E)3amRo%}&Bghr62QuG;KkA~EZ*{H}9yW7+l&n#< z#ryUC6~GUFtDBTa01j2YA5E+mu_m+-Ndhe=XkLDF&yoN|{plpL20A|zW{n~hez_Vp z7p$jP^?0vF1xmbPfZLGOvCF;e%Mwm)DtM=alWzCw%Ib2hAC11&8w6N?Rv>1s>|`LD zq!z40bdzqGlP>;_s>q>PwRC#^^Sy#L0)AKHEj2GF%8oqKv4Oq`DZA}Yy51`I>M>_5 z)ueW3&eYJEu3pavOj?@W>Ha0@ZtsS6bJS>VRbTWmf1szHdsO+Zx1=p1>g%xis%^sJwqJ*Rf-bo^6J4K*Ry@$zP*M3LvShnb0YkDMmq`~uu7 z7cek+!KWTo`jLxC>H=Z0sa&|knFQfiFiAda%*7ZSn6}J=t`u-*;3e})Zzl2P>%=md zww64JZe7}#0k>P(ZX8zTZal#%?=zgrp zVgq7BXAL)9HB+NBgvWsk6Ouq-0WllV9p5(xlF26pfMo-iSeo7NJ37R}^?Ej7YOL(Q z)?duMrw~Q9Tyj$}?eWe~`Isb-c_SG99oS7eR^Uv>{XPy?Iw$yI(SYk|=izKmaNI*O z3&z~q!GFi|*4H7*MhKc}6Sm}O)4{8s%$l^LlM8FbFHl-JC-yM>JD$IL{S@wVh%LTt zVSx0vQTYSyR)YgYZP=5&FoL>@^|+-Hzk=u=(X3DQ_U7WifL(IG%vl>l^<>lVLW|vO9hhdp zd^FNiNwo@oa6kM%L^GY$V=1IA8X_gsVNK?4M!8E7p7PN*BW| zeUDSP%N_bZk-|=CA*==RTfu7>zFQVO|6S>H0)eKBLjlE zBmzU)t}O&@5_@w+f?G4t7=0ymXrna~5;2GUFZ?j*!@=!~RqrC7EoS0@FJlk`gqFWonnoJlD=_*{kv6st*+Wm@6 zeS=O{C}IBBoUGHUo8Hit2!Av%`ILrd@V{W_)tc7~8gWu&773wTo3;4Yx9yGVJWSU| zfl%Rvfig$Qgw`;`G;E|W!h}4)B*V5*Bd$HWY0{3-`Fz-?YIw(#$u>IN(+NtZC=-WC zs};^+dP35QUn1>Ab_wX{LvBOS`1FsfWEjTdzVC+^zVD4!;J;%x9;Ek3B&isa)t+R` z9<~eBQ>)hN@weNfb4Ek!N`7fu_0U9QnklHqb(Lr|BTRSE-0DNC@`Du{*$e>K+2oTN z)KPMNhEb+Hop9txFDugnjcnx4zb-Zo7k0)WbInaMlelczKv&{8VayN_`(+thd-&C5!9VliFFdP8*5LD?q~OC2y@N<_<#W!Qe!x0y<_! z{Kcy1Q{-9ni3hZH10G?+!qT~7bqn5kcyjA`0p0|XW~3^C+OX{_AVStv8kjskJ^ zh8GMG7}~okNKZ=wMcvD=k8c-k1`w|S$W?8WH(K9V<tpz2He#m&?eh5pn|*a z3;wyvP0W4Srqt_^(SI`kr_{dToe>L%RuWs_=G>+1*IhS(bx;H9rQTXo@%NSTwN(kU@tE$L8ck9BaXZo<{}F8ipICM)ebSG`@hdFcwJJCmttWzt5$OSTx7 z0{ZQ5)CL!A9LDey18a*;D^8UB$B#KELNfJc+D58(j{`}So#ZXUJzL=we5beO>CUa$ z-qF7Q&hm2gOq$e!t_igODawwA8ddhf6(%g2K2_#rEEM({3*O#w<*G|DWF0*AW^d)h zG8&|;D9BJg=qP^LANFF@cGD#*5`l6Pv*n8(fJMP!)kxH;ad~1fR5aX9q#YknH@J^j z7jR9Uni~Vf42Q&Fjb?4|$&?HN5<;1LPg)CHXk{?I{laY2sySR(hp)DEv`JXB;$koS zwYMwB= z{Dd6nAsv8m@N_CAkfySwc?A7l8a`O!`5Lt2vbd!wq{A}Sx#w#2VJJKG-H;3iR0Gp| zz~MvIU6~4v6$D|vP%u4anui?xZ1Gtavp};^`m2_ViTX#c{5q8&Ie5Got+DGf9*4zf zFcZ5UI~fBoI5lrVY2ex^(X{?EKy;8#tx1EQn|Vkes#Tk~yKyB!f!RMyPmv=^2me$T zc0&9?b2>{dbgF`HX~v}ynzU(F5Xxh&0p+~$4wlFBFE;a`AR@4REZ^T`OM#KD{YY}q zSTO2XXxkgi$mTXz0K=$RGHq9^!fA1?dGc>TE zA`z;5**2GFL(zev6$Z?rawE$SdBhU`2x2U8jin!jKX zeM%S!J;pwGTb}tbL4b6l*VC16JQv?TJyO}3MkYi($9}E6o~4i1Firj*3B}z!J3PG} z%F_3WICjnN2mi<9fX()=9iD@nkqGUBY+_H@0+-T`{Ux?=*Z zL-#}LC3GIh2YU3*5eb=#051G<_P9is8LKDkg~oa9N&XboNfTV;&k~reo=8HCPFqZ`ZdGIj_@G5I=;)nBXR(NLd?2^Bx_ovJC>8 z8ef{4b1R`>_t-*^FO1`?l9Dy*|0q}}0TR?WP+O+%VMavQ^ykvDc?iguEVPlvKpgDK zZPCoo|C3u$gDwhKiy=ZGlI_m7C({5*S_ubgs`(|G$-c@jCGwy`I{2nQz0zgj3-}AO z2jwB8Ne23bN^dwA41j3Jgm%QlA`@e0$V8)xV?E$RT#}%&EHkJbRUISR3paBgN6r67 z=C3&_Gz++qg9N1bjO=B~teVxhYj;5#0Dig1AT{}9*IL(5Gl$?);{{<@9r`()%%Ks$ z%s2qy$lu9c(LBp|hNMT;WXGqwsqQrP+aXxwfz&S2W;1kr+uoK3(G(3ex;9qjm$ zTu?qa5SUNjYiuw`hoAbDD7(-n@H3+?>Gp7ia~K4s4E1X1cUh2xvSJzHMG*PEO<2qY@x9V2G;YuHtdxj%l`YtpQP5IK8=;v4 zi$UoJ|6aj0BqVx&KUj@&U?&3osZCT{DZLq%H8-L{%IEB zc`<@8VTBJaFeHtv6p}@D5SUfKXr>8ZPR0|~KN{$r7Kl_EQw7AR0!AxDq_huC$k|7Y zD)-W7(3M1tn}q)gGY8wq7P#Wig#+&JVrnpaqzs`Ib^uUf0b~$VbtxE*hU!<|l}QW3 zx)DsS3bwtw+s%WY$6OPQfqcOEDM<7!1~M780Ky@g-1vb;(Tcxf5KE{#V6ia6nmBaO zly5s3FMNrKaxhS3xZQ$S1Z zV7CWDq3>NGy7>1&Gn?3ghG#C=~wrx?pqjXWa z5&39tUqDzW+dhHZqkne_T33E}?_akRh=l|!qdzE67^f=HWk)UPPaK)XTRoBnKDeLM z@IA<%cf0>UGxhqoT%A4w5~%BO`G`{;cn_SS^lCbn6Rz74Jp zADZ<6C4z)oLs@T3YTe1|^~wLmUdeG|)N=2$f3xby16t=)UEch(6bL0i2`{XU%__=M ze)epQaCErp=>rj~=ZF;vwc+6vr-5Wk<<)=~NDYUNVx)NAH(<{wPv3zEpf^>%qdkDk z&uRibODAUH%~=F>h9H`2ie>|cj{utrU1;7g(#c{P2FK77O>^`6^tvy>U;6Hw-i9A; zv1eZlBizQ8%pT|!2e8`p{2X(>P!fK9lbIz3TJLzF!_f!v-ls2Eg;@|o2W>e{Mtc0R zLVi;#<#C`@WAL*@O2OrxbU*P3>60QT?r{efPK(C!hp>azdh&>H6Uj#7-dTo!$|VjF zq7nQRR{%9V3|JLt!UqE*=h(R^oz;&);2yvF`CEQ(amz}~YjH60UYY&oB_{!XvP%|X zOnzos4%p%9<>ctGxDYXEr*+i9&4W4O6A)90_=VqCWWe#s9kyeI$R}jcfUjo47HF-M z&60od8*198wUg}TK5GUYP#l zUU47R?_6J1!hQ4`+32mnFuAi5=__1qzJVO0X4-bde#D~6+8 z=`@*Q87|y+Y0S{b&oJ&uevO*Q?w$K@(t~ZKc7(lJmaqQH;GGGeLcT&pLFJob*k$bu zCB0Qe%=#e;E51f1Yl;1M5ApFCfw+g zhlYbb*Fff-E&Yi`;5mP|$Q?Rk^H0Gpdjx81_g}Q$OE&oUV)1ETQ5G`AklrxT6v3mJ3Jo9gI6pn`Sp z8vipHKYb+XusSbgg&rKE^XBed(_zz) z+fG?_9q>IhNPgG>2;LiBK?~N3c6Hu4K+U1%4LYFV831EMC|8K$$>p%T&+X2U=e!ROg3Js#NFir> z!Da-$QgASX#ZL`HKlUVv%!s~90%oL&E@N~LM(+gFY z!;Y*6jD!`uT+4k``o!eU@M1lzc}{9b_@nWIqVa_H*;|cV0snl5i)!b&qkZ&GCYgQ~ z+cb8!;bK)-22hKsdtRCKa=(V3Q50sq4@-1fQ}4{*aI@cj#37V_{gtkb*FQ}0H#f9^ zQhb!(E^Ok7hK-0ly8S^d02*%O)7{^6T{n}Ss)c*ofO* z^S{9CM$Z4Ooy*GbU$RODrvG74u32l^byM`8ar9PYRJ0W`^8To5+iF?ltE?W{EuJ^gvhs|pIhLR=z_hn#$rK)(0&v~#yx<>S;K_kIxP zH|lpr75shwYcCHxdk1gC&$MbV=`!f_wwU`VgKYpu*N+{7=dIu|<4zDq^N^|HdH?ys zKc|<=&%>9y{RblFDnITh!)ZjY_i_{16$wk6$J~JNuMMqGRGtP4BJ1M}v@6s5=<{Xw zzO`vPT0KMO$i8!FiT#&7RU`MQ?qXWwlSt%X|7YI%E}0S~sh8T0Oeu4Mk&Bki#Yq6K zal7g5`u((w(vDmy$5(PL^kAi-(oQJa4wtK36!S)7gGH6+w=ZLc<8X}rsGCf4w2;H7 zyYJF3Lc4rZS>t68KAb*@+Z$x@?*)Ks7HEKF4yIol=Un?s#y|POasP4==RrQAdbC0j zXQIbRml*E%lrP#$x}S6@7T9#ViCUWo{HK46@jr(8o0PwL4q%4GiY?o9%S0hw8_v|B z^$0GTgw^Qf_>yLFlabk_NPTWW+d|gidl%*DFc=A{p5rVnx`|Tlg-eDmrR(#Y?F8x% zng;D!SO?@R#e1`f>xYje;hIQ0y#wt0HMkqc7o6}1f zNU$2v*z&c8%(TmO;l)y%KMweC-;zw-8hh{}o5Qx#nWnU<4))~ha>!9qkD<%cbk~JF zML)3}IKACHk&h>5-0hf)f%I4~+T;oPenn%o{SpoYbvd5xIt8WuWph2Ks~%s+>{Wcx zkXYJ;YMg-6S*>Mi|E7#n)ePo6?ZI}Z#D(p?#z>FamL%Nk~nq@+cPe6hSAKtIdWEnlv^5vvOY(fW(t~rj3KD1`X89h`TETppH9T zDoqMmAhH0%=|mc1O>^)lM@%(FYAC`%UO3EwKMxr^V-PXy;8r*y@X&!u3#9$iv9-E; ziH?~-bs#+RR`zpuXv@K>u=pfi;WYheuiV;&p6g|iSFs(Gt{pnHRnOkuPp(V(dyP(p zdn*ft#<1-yHW&@<91Z{8XeCH*@ zBQFuEL!PUF$a4>(+DrLA>mXFir=BEk#XrflOw$=(5w-S%BC3O!)i$16S)Q8T{=0j$ zmruA;PbRMkjd^IxrjSzhSof)oYcNX~_WhCxWI$oG zYYV#$x44k}w5(fT4bZ`@h=K{!jb`r1e2@?dH~Af|Bh$# zw~Qh+LCzaQ@Xt2jrm71A^t;-GgvJ9<@NBru4e&Svg1urkwT%{tc!~h)9#Ruf{i5L3 zqwO`V_R|Ja;Cj145Wc|;^+pjusR1d0K>)sqwOtVgj{si2`9SP5Ch>yCQD4}W>m$@Y zd5GoJkV7!EqfmDQgM~KQMq$=KJD0mdSm@L2W2TB6k$NLsxj2ZC%?iX0cUVsbGkTB> z>1MZATEJPsW~m&5-J@hg#|2AaVZS3_QCHTlEB`W7QCJ_c2@dcB&50<>0Rgm&Mj)k! zK3kfXV|42C2|kSgA^J0Y-~xsDk;$wo8|61>%;pwk%q5SSm%wL@GGWw+t<{5Ine=PJ zthF-*;Owp}Ln9y;qrbeL7-=bI^tn&sf=B}zd(yumLoQNasvR+S)9aX_{2(MSpRVEP zqn@-ns_E>ez~}MP8D}CG+HRsNjO4HO(1}AXl#>`bD;}u`DnW<{wWLl5ei!8hS{3c| z2VMq%lu@OH6C;)p_+0o;!244P8#^lv?+lxzmyru>!e=pJo1~ZTMiO!(SYE$KXZdfmE;8 zsmMLO!Y(?EZGF?bjFXByslgs`EyEmwVG&;4CF*kwC6!)%H`~O8bJs13tkYJ8x{DoA zmaU3e0Q^Hdhxv=@7q8rAupxz-dq(o3^5h^dhQur`G?hBc)XMIJV#K*7y^1qDhf2}C zDa(kEzZ4jiyzk&>|Gxj8$hQucil^4B^Hf-#F(w1bL=9=+F>8-B%hL+L-I0vNH?YeVw!{1#*N;HG*qUE;(~)|SX|7f~?tsH*nB@9vhDPG@pOwqb02P1NvuxQWJ=9}bWW)KfwWUz9!!dZxiAKuv(j>OnWW`rHBu$;Rl4Z=AXi zKoFK5|8HIT3hKg|r*_4J01e?p>L|)+QjxXMle+7bmuiP;W>-%42+A}R6i-ABtQxI< zn*-x`HR=wCQic|u#h*NtB@~gD?d#pk)!Su*?)K{Y=!p7Icbsx9nefhlUsyqgztVFl z=|;-kKTujAv%e5;9}wXZ#Z7Ml7(iN5T(HX1Q>p3rS{w>6A2=gwvmYhaBLpSS2nt_+ z0u+<_R5Ml;um!rK>XBB0$1qU1nMzE`KwCX@8LlP`B=i9TjI?G@ffUF=>pBlqk36K?@QyZmj zXeMqD`Rbs@?!F;g#YRox5YGAZOIPH&LBKmQpYhKo+P)Yex8TS>2pN!6cwb#TYq228 z@Sh{~f(mS0vnJ+orqONs9xsXZNsnmhHk*kbN;b#)^*T1vVTWsqSfzWULAtYhD58-; zl79Dkq8!UgwJ#Pakl99j@L9W--W_Wpu!;(NuGx=sJsW(%kQtzknz+KZg9$fjz7=8` zQp%yWY{J;5O7j(cUG{+2UM;`9_Oaw8pXcNU^l-;kpXdFaK4G#GmYhWGa_q9x*5gzv z82-3A_bB{uT5zCL2>%J2B@Mkch!90sYnCGq0yTNVFQlIi&Ti$@!sPnB zT>4W^zh56$ql?ra&=Wrb$>P)f{a5JuG-$lZI~lGxwxHrW<#eK2Aem8VX~gt@T*#hi zoq`CG1=6J{syx$C;|8uyQ4t3;ghRVV*Y(L#?Ntj`gJwaphhGih&FWb`Wb7G+RJm)% z7akY+p%V<@O)IJcDHjk-g}z0?9)Oz1?r{p}3X@?~Id6c40-d>FLGE7)(_OC7#6=Bx zAfu|esDO`v==tS7w0{7WXj_HYo9e>@?aPg4dP)|t<$hcea{JxH)YcwtGumj8t07_$~#`X~?W3 zOu2Le)~YD3Dga-oiX!6pLb~gatWfa#cRmZ~HxNt+u3}~Zy_DA7G5OSg z#UG&a7)X!(FhVS_IQR{&;01lM5Vn+ zn+1Cw(s~|*+}};P7oRbCZ;Cs3XR8|!>Dw;>mEZu7VNpQDyQ-&YYA zB;;rxbk4?Ve#$6UZQ5h==6ZoaUG|fYWHJG^9t`ikuD?P40@2v*sbx*G8oNS$xh|U_ z9@<0QA5{m!qqGj=?l+MbQLK_Os(fuQkX)JY*7)vW0G#|(gLU(u;9ru04zdzgM=UI4UVPE zp|lHq4NgDKg1LBx%M1w?s$cZ<2GF9%Af;N;f}wCcAob|4hQ{&74r!# zQF6CU{2l=`1f)iXf>j?0D(;btu@QyESa3r4whfIfL*IIwWpZZWAFz^4qIbat*zb)r z{d$8*q^;m07Y|_-0k5Vq^p6x~;Bj*8%3g&fsjZqrIYwm1%NZ1GTB9X|2$GM6L}%bl zZ48Dw5@TNaA;fXHX)w&3XiQtqnU z-fkb;z|2W-)C3_gVn*Z?Ng1m!r>joyKj|PE6}n2KvUaz}i}cs(Dn^0dT9n|#tqQYt zNGud$5`PCzAX1?+=ZBCh;Pyt5b1>-Wt*Nq~7XgqyuxXMZh};SjiI5Glkh9~pwM)gX zc)%{kFlIPf*sr5ePG_Q-5@!*=UC(4G@a#3RK6hDT<0VGrDmV##RHv@Fsrbd>!_iwSe5>7J;ooWS1&zJ2sXmvn{v?cTpB*aLH z-y9|`#FYZy+NyURA79XElb+Y$qSSEbZJG>l)lvwAgyrVe9f4}2@1hon z`NHpZH=(qNa3gKnsEBhj+lc$jSbuDk=VTcq0^rbe&$}qW&>Nc_DDvKz?JIWKDcFy@ zcm9(b<~~ugH0jO=nOxV_39Oly_^elTGxW#TZc+w&vWSQPo`(pSj4ycTuM_&zp%}0b z`=a#d;$EwOtb4-HwUI9jDlY{*+G^_k$NCwwu~{6rC!&p9G1HbbWNlUx_I;IotPQOWA90LlN{u=w z^uCLew3|28x>mp2|6}Z&8Z%)QW*b`*CllMYZQGjI#uMANZQHhOPHdZJzU%W|?4Qtm z*Q-}md4VOWe|tjMzQAWKu(#t&FaQe=4_QySEyu*AhwnDqUoR1(;~_^j;zjt-nr<8B zZu72kBedWzA~$^5Gx7wny*2QY@Mv$V%A(bf?wQSrxsAFvQdT@sK`>9@vSoYx?wI?kqxolCL zd%%=qwWtdLGRLeR2YIGrDpmH^D||IpwB!I{sgCAj&H1~bmnt8gu0$FY{(#QIx{56t zu)Ajz>GHuDhzKe4aYvez@K&c_J@zxu=)3VFQ26Dr6gzl<;_Z>v5M?jC-|!GwK*{dX z0XSI?Es5b9yt=j2hm)*FjT2JZgx7F+5U>w-fQZcoQKrcVmADpQ7rVg)A1W)(fz$Fo zut1@4b1y#!SCByY&e+4$4r8hQO|HAwW0ADbUX>B`=_+>vlyvH?T zFv53Q?52SWLz<+q@O7ymw5Wf>0}ZX^nt%Fvc}bH-x!tx4i*W}*Tt%X0hNfpQwZAO{ zg>3fn#PoHkR5%X^FC1fa%=Y`hD-g~Xf%=l6*cQd_kjd-SD;^p&ism`Q*1Y-ysjjPo zs(-R2rd{ucV>RtR5)zvwA$nT?S;T%>6tivzWhM7_Pe67 zqdJ1Jfm)>yl)TMjCkdYcC3D#5`NnoJrb@)05Gdoi=Fl-9KjfVH&N5~U(wFIJu>S`{ z>Ug~9zeGr;|A31OY%Kpv3U}ObOUiZIX*T<+EL#x)EhsB4{XQZi=TR%^Rit7&rW65E zLWraYdyQ2>);pl}8|o+Rr%#*S>$)2nwF0Ukv^0(=2*j~#u>E95=F39m&{AbC_wi=< zG7nQ1_w!8PZd397#E1<5aJ_x)W-Lq_RZ1;vl!sgpjux77SppnRru{wS0*eU#-Hj${j^=apn6ySfBMAeA}Wf zRX(_k3UO*@<|>;XIuy+>IH0HBpcjh;-1ts^Gvr|YX?S^>b{{|qwyw8ai9Dk99fM8v zo{E{7O?_IuE9#R+Nv3No(OOMiAR3ADJ`zo#GktPgra8tRJqT~8SX!T-Zb&c5R~b{e zGyf3A*tpNp%fkJ1^ot)j{ay642<(~}DUdxm$&rEEHZaVSc0K*L_o#z!JU1iQA%6iiEIQXNp-ehcvM4!kehJ4(+DqV>Xn!uw)i_)RD98OnhFIq2DrX^xWAZZNGlgc}?cI zx%4~5eW+=7giC1_;6r-@NBiiRP41*vLXI#7^f0p0CO0Cs7$Rwxi$1Bz+FQmb8|2@Q zz&HLUzu}2hZxiHwm5kfR4KVaa-r2$twps>t=(r%tL#MVDqv$Zo6?t4hC8YlhlC65t zoJ4#L&Hd5Be*q6UGQr=t`&a)>CtGCh39-!`pF zW~-6M1&OSLP0Hf|6?-hM#2d++(OS|h4u4)^N&lHUnDA2FQ*aPzY ziQpA3W^GnYmoO}C(tcpo#|~9wXz})xPppw1C^& z%X0vN;1pzY83Apv5c&M{(Q*zmhn@T9^k;)!3qQP%-qNJYavp8>t$lONDlG{zh@LP3 zucx^FFvv+U7ZmZp{7j~$Dk4{)>0Tq^+Tc*1$YjBmXz*fWNsjOZOM0^=FNf=tbRp!Y zj13wl4r+B)Xh`zHGt!NU<@g;bAam(Vs77bbZMpnqSyN&6p3nuI#TUd3HZ-qXBl|H5 z&sQ*8h|V3@4xJimI|Xh~lQbegpr@fg;untKI>2Pq$+BiGAmgot8uisg%Pp`xO*gFO zm&Mvu{!PQ7B7^Lzxu$PL>YNB~dr{o(h9tbO5iE5Qait-&P*eMZt9m^1fU8zT}m5i83J>U1{sp?+=EeuG4UsXOo6uFUC&DakMB5j<334G*X(V zsY-@B1+?OG_DBbr4e==T@u?;y&SxsME-N*lkfTmOI-pVt6mhEti&uNf)aa8A zs5Ju?TA)sOuv2!5L|}apVy#{(x6bAm*?5*Q?Q+?urLE%R$taP>fA7MbS(iN#HVCEO zV(-)DLNpKKQ*JOOr+f2#asCm8NJoDbf_79Rlllga!9pWtJz(|XjXVI7HWBpmzF$$c zX3W%hNu1>z4*=5Z&Vv>ao4&vxZDZt6W@D(mL>*>PHpG%9SyhqJLfzI}XKe|JvSD)isbpsN%2`MK^}!J* zADm``SIj8TN>3@Rd)6d4O3-ZC1T+ICG@*lN^-0#y8v~y zJ$;1|#E;aX@L9)AdJQY^7mUdhDu6CSmH5Ep_+>aB>l{5Pb@99+*s4g=?;>T3QXv36aBHP1T0o54)28p zRie&)tZvHV6jf0opRsQK$2(^TwA5LxUc3}erCuX}{7!2CSAHLl@uQXTU&Q3so%x%I zOLv<(>aH4lSH>)@Pi?%|@;(RBZM?`K9q19jTrR=n_g5U}3z;!<2* z#4)|iVe_Y%%Tw>}M#y^R^qUym>P5izhcV+Y$)`|#kMGWumE|k~n4(M?w12n3I9qe! z*F}K$l;HgYTq0j`CZ=Z&EcG6$??rn9D?AqLUFgzr%fD{WX9f1IC1QHk%MC=_U9w=~f z7@Xg0I8{l%Ue`_VOU6db|E1UZIxrT*K@wu8Og_A0`<&HvJwsg+Z8_YpmCY-1YT5;j z^LCzH+$+3E1aa?yv>hs4U+D`sYx(&Q5q9TqUVUW*-Ca={tt! zmD;o@_omgFm*zs*#1n>!COp{v^fu3GBS z$c11AD zmGp2JMNq#Yd`Lhm|8)1{w4!0jQ(+XN7cd1iEB`X{>#HTrdbIS)Juhz}8V-9Z@o@b- z;yo$JIh(%wnL4pfQRB`b4^9>rHUw8>%DAeKr-INC*-WTp{e-hhUtp?A8440eAAnyL~SVOaZ^6PZxK*v$3FisCIZjE~?3zaP1O}nk4wjpV7&^{2%{RnW2XN+D5amRPo-s!)yv!!8fQ(wTg|O8IUG6e5AF;* z^9=J?SrJF|bAS%Tuo<+IB%ElVXL5IFUkAAX<9pKyX5+c%!Xdrl)DNvbc6qmUuUFFX_|Ex?PF@YPie50gpDB9poccvlGXY|*H#eE$7j z|J?><+PLe|ChD>!`CF5&m;K4?8rfY7no>uRIjiLkn_>22`p5|2lgPVjyJ3mPbnkGZfVJXE-1w*$QSDBweNrkeqBoG%R1=w1iV?kGHF$->7Sa zu9l&y{QisACNxs(Hj6<7ZcO$7ys$N;@e?1}Rm?{C1v08(4Xd6T0)*SnjP{Gdf0EVKfO>Jt4R8#-l#Gv{|r0fuY-u6K9=>3A%` z?%t}J5mfQRRH$(stA8wkEUKq|7gJX@gv)Buc-R`J z*C_W?<|@kp%;~nOaC0;j6NSE>knHi(Iju6X;BTvQC{+E}_6CbEs=>of!pSJPwDCn~ z0$Vm^y|W0Z8XHJBz|>vG<^L#Ac2Hqs{l;81ZFFWKESFGk1$D89$kpfx@`gP znm;y21Z?-7s7fo#z$%{%`$HxCR|&G%gxVQ#j<4-zaLbah6z4Z?jQq^WPFr?IE4z?w zt2pa!$3dp(I~;0T`!R8P&6~C7)H_&}J_ekKPJD8MrP>clcDr-Dk*t1^IiOL;TKWg$ ziF9A-T&%P)60A=lg|0(nY9c`$XvjyUfvV1SCKEc8o+}%S?RhX&@hb!)Qq%3!LGC1m z;Ckxrjaxyrz)>4PE&j!kn7nsTT*;YqFF#Q~*Q%HaEXqyi=MWeD%X6+VjHT<3vA@`wmlDlhvNj|FKh=ozGMSQjHC-V!?%@ zGR&$2Y&xuL1s$-bLUfJQt~gLPJZ8aCp=nibUHH$=tclprwVp$sb45`iHOs8s$Z_{s zCu8r9M9@N@4u({~FHpZyN;>CM>ab(<#e7eBSK7fU z;{4X15>?CF2X89QNS^ZLM7f!d=bb8UKun3cR)W{Yt3=2_NlZJ?VxFp%f1P3%Y+Y6x zAL@#k9brnFxt2<{SRc}+aQGxbnLaEeF7+!oOhP8W^DYPPyE0bwAz&-`jze@xXv-2P zl$y7Mn9IWuGNd6ZS0p3u4ScL$PbIJ7q$%~W$;FeOU=*X6p-~L3qnO1kXR1HUY#!wl0LjIAEjnf5g8b7C->odZLQFBlM}K z8}w|ToY~IHQm?7%CBd$Z&4A`qugpA+FZ?n&2T~K~Vf|e}o) zdalzAi^mrY`7WR$n+sPqtviAewSG9e^O6dog}yo37zFE7<}5@RxVsH52k7eZyK~FV|Bcz{UL+tZ<=ek`4W~rN-cFfr zeM2Q;YzPN_UHS+SF%uSu$sxQWG-fJNLLk)=E|8xf6Lzfc95u*N>09It@jPb-rZnkH zUA#C>tfQxbuCHHVp9Cb^R?H==;L5WojkYNhsyQN4Jr51~<;N}SwBWan#8s z=e0_2h}IxQ^G6|Cr0QIi6EJmlBfrT((tWMtee|l=Y&f_ROuc0Qgq-edTf@Rl9%dy~ zb*Ju$ta7LaV!-nxVI9mYQ}ys08Y1xBNg6AK!i-;{Q@Uk*_0eS6;-Qbkux$I@nx2LT zetG!Aw?IbnW-hP6=%Spa2E$y=P$6ILLLNj=S=obiyydyo;1RsN6f?T%7WuY|M_hpq zhAuIKH^*$6-R?Q7UBzoW7`hn3y&g%?p)_KRIO~bv3Q*`DZ69s=Ts87C70_*i3tOzx zmvFUn$=sB?M_8;(T4H9L^=-S}IxOXYedH@+rkn1el9hbKdDSTq`2qivQ}#&tMLl)` zF{Ivh^irFo5mHJIh!Srwvb1t3WbUEG<|X*@izA$3<$(~g-GoV0>~#Bpg`q|MU?Jv@ zUc1t(KMQklzd9UI-hvcAN%+fSp4@PaSQx^j!l)=X{Wd&V)i13ONX%b*vu5 z1^HXBRay%G8xW;jgt|D=SK%u^R3A_2sPg|;d?KnR3XOC>IOak4<2O5<;lX@y>CG%{ z2qIZb|9s!?(%eOIEz3i8>;3(^hZZ_PV}QrZ@dvxC=L|$W?&9mG?KB3XXZ#5S4=BXd z>8IEYZJcgDum^Q-GJ9`l`kcM%0@G?n0RB4Qd&2EH9LtN%^7M_5)r?D+Zf zNIV2u8l&0>+6Pg-7@O?Uo?VKh5X|x;`!|_$?tR6Ci{Pk-m-;=@hP!{~PM z&a(c=qmKcpvm|u9s~f2^BV{p#(wG2&>n5iU=c_OIx>C)5AwlN<9|^LsGqC*+NKi}q z|BCr~9%?o(A?&nq^C1-F^PCSPnnc2rOyU_k5iAf4P>2ZK*f-2s&=z{L5AdGrUU)B4 zHG{>K32Z`fBar5U0mldb#vg7k3^T?ClfH+0BbNmM?oPKC{)a1ruV(dJpC3p4^Pbg+ z-uu~(p{cnD15h4mA0i&b2+SuaHdlX-&IPi1_`MwfJe{9Wg6rl{DUvyza0hg?h(o)= zkT&AHIz~1r=W_>vWu{cpUf0QO8D9_gPu*t?B#COnGy4f8jp55wsi98UMdgnvyIIb5 zOz-pv7l1Pz89MzZ`N^eV0p384D zKdW75SZ=CC+N>Twn`tH6_{-G^e@;cA>Ap}**Tf<>JR#(z9vHrK=7ZQq4QK)agZquou5*enWEv>MOAwqo>CE-4 zBbT0gE+Ft@O)Hu4!KxjGP;YO!!eOeZqy6$oZDxzizwgZ{XNwE{ECD+&sFq z6K27LXJqyddDF#(ckb&wf|tjZHKf`xD)&=;iXA&CSCk`CAlM$uqT7ekWK6PY*jRY_v9xo z`g0rN`Li4QNCiL!kL3UnRO9NY$L3;UXifnQcDKdI<%aG!J)UiiqX@-SLfSR6azA8`%-6lK%^(Q8jT9lRtI2 zzeWFp(QBZr*^0Q`^w(bX+nDzFSo1s%*XX>RHXV9*lP$I0WL+%7=YD$3labr9gNc2D zB{x4c`cn&Su4^K;Lpt0n#??>rZpvSqx~a(dO^o3PG<>p`!eEw9yXi84piWw-XwoY@ z!W-WRTRVm#q-YC1_a9Am_Nmt{bL=28J_R{!Wnc;0%C}4-#hMag$60)_39j}7Ao9Pz zo*c!is`9}4Ek}_#Lvd-tY_x1RA9HFVGMPR}WV`9@1bQ+gPTEkQ2IxfQ=f)9iWn@FD z-#z(pI`n6F0FX+O2PeRAALS$eP##JW*^NTwa()d(y;^cTH(dYk|DD5+yM5#?4d8S z?4&R#mFZ=HWGc}XgR{{OYz6M~d_EuAT+LShAV@FrEFC@@AD4VVhL{%+QWkNWUH%$d zUdIO+A&_nkN)h-}OlXPOLRjb|`&iH%f%4KL7}gE#UsD?OaLRcyLFN`pu8k9?^a+Z& zcmh4kkd)a7T;s2}s?ku}UHgW|Vo~`FFb>rZF{I>z)QfQSxqCKFP&&`5yZa@TPP+1Y zw^n-3Xdy}S&7;1pSh+7UOg{(wI_=EwP_z;63HR5e1s8NlE^#hypfjE$6Fxk|K*Pk% z?aBDb;=yPd8s(R< ze$|4Os81B$y%j6qGy%opwNS#l!Ct)2CJwfnM%{T~BeH9=M>g}Wy?tC64HvY-k>|S@ zohICHb{7P&U{FU(U*~<`uIR4iM@`sq^5$u;g_6C@T#bYE7bXjuQah{ZCGL#eIXQd7 zj!?E_Nl{TN@=?n2klmmL+xjW})Fc4TAR`=eXw|H;mW|@uh4R9CTSdf7DR&12)KB3V zKI{EhvMU`O{W2t!A@%GDUlrsHsF9-38DR#!qx*bRF|g6&cm788a*!ARAh1q2ENchE;% z756F^EuLpLrAvrq9S>SkpSv!VNJ^!a*p4~{kOzOz$u$Ga>~i%*tWTs0x<+^D*;EKA zGjpP-p}IThQq^@hgnT=8&R#$bdub@w27iaGI!%vUlAf;Z+cj`J#BmDvi%>jQ`$r&) z9N{;H)Tmck-yvdNYc6z;frjV5g+r+uTEo0_(WWk%g6N%!E&t@HENI~U>$27%ZiqE{ z?8w%K&)7=oF1GiN%nf%&f$c9T4TD((IhQ02WY9A;zrHhc;68gqs(10s&xNjqk;9MS zjrg>fA{7^L8g@S+eCYpDMxpu){17IYxo+o2wJqK9t+M;b>q!URUS(%D?Jc%r?~!u+ z+hn$#wU=CF7gK&M=zViyg>GD={yMzsQYK>TrQA{tg|{HUxwdI22FnJ+@k)cbznH}3&-5i?;(@G<c+^1*&vSkdZkcSmtn{3sh34rDB zUTq90_?vctmim+9mNsW|g-2%GD+;|OoaWbMNI%GNllxC5nu#cAum14fdCJjqER9+; zVH;%-a>C;4QqU3IGm{$lC^mo>U9=?!4Gidxx(=qVShm${@-(nJt>9ZGM$xF9?AcJ7 zgFj^Mh|2H4bQ+fX_=JauzP@R>6UMP)o3v#3MHah=2X}QCyF(-Xrw7&ZS5CKSX$h`;1IEWFxqVWoo&`Wr9uZ19Xi-#eIqD?({C?@E5lSEj=KEDjS~ zpj>?z&0SRWbvM&y0Dmuqfv4|P@7t?&-g~g?Jo)rNV|-ND_kL>s zS-yk+P1VAGYcQ#HZ@JoX25*>dG*NC+#vBanI4blq&rshV~V^T3r+~D!VUXO0~W4oE(v9J zwo{FiPENna@e7&FWbI3fib98b5);1;~|u9vVr~j zB9rO40N~K9`w0H;~r-2fI_y zG?_z@-A>X{pYJ>jo1kXyBD^=dpA6?6CM+kj^2zKyq7WK~&;Rc6@vvx%MXh)9y2DGC z$8cEt!67n166#JdGk|fEoD#*m0Fy6Xr~4`CKl5(&=I6*Gy;EhzUr4qxk^&FP%Dm{+wK$A$bxc zSN?g<-k^P2$MXz6qQ<>gM*j$WB`-`2@^L6UW_?~m1{FhT!xm@wgM1Yx%p;Tc4;gz$ zu@lXrLM;9;Up{!ZW3Xm7ivVHSoi`);Luq38FH*d7%AzhD*1f59r||E0+9AFSHsYS{72U7k|$fZmS4 z+?OQa-gI!x8gg>9o-MRxg4Hr2Axk(FR(yW_+<7Bx#orT#xk~3|POi`(@rRf@{eHiI z@Zpt1@yOcmeRXT$lZwR<@5Pe@c;fs_t&Enki`UjP@gYdsPvGu)@$Tc@uZ4<$7soJT zAlqEsO}|U%=;(H}bEI#11# zWq)2BfBCol-)wDteVd>ITfLTt2hXx(#)QGyq{oM7zcF^81uk6A)K?FcsEt5WRTm}0qiKD%8 zQWcf7JML`qTd;UyE@Q?n#)lI`^7|FI>E{|sEwfIjT2_Zc4}X0rWa#30`uy~LoHwD4 zm;uX7>_wt991-~CwDGknK<0-nT?;8=LPqBR8MF!%J)w^Wm*?+(RbX#9#v6pb)E0t+ zs?X-0R#q@x`YWXce|WD<7buz){Mexwn*BEwsp)IaxKK^8pr z;FNH8YSX`}jWC=_GMVFR)Vcr!Av63&vI(^uGD87f%~q?Ct2cL!E{LFF*}Gj+rbNv2 zRfq)z+L7tcaKhu?Y9|I*Yapwq^fk68+m%a8K(Mq+!8`UxX$M+$pG?Ja4X{(*nM|@$ zc*?Omm0%gXaq6lpd`c5VB(3Mb&JGqV3R`|Q6U_>x^hB>fF}OXb-H`s@ZpnU@Vq~+) zf)8Yu;oL3`ES65&``Xybb2__;E9tB@vhraZc>EoaD-&Vv37ifs?$kz0?^Y{E9i1K_ zXa*v=O3J*3Wu?D70QX`H?fVVJs!h!T(uLcd7zYww=ig_ zWvJ8202{EVY>!N-D@bK`h%RyeeoTAVR(2{4?B9PlAXFxm!mmoT2(b~WUm9N(c=%MF z!2S}0`=;CJi?dR5p2AFkNZ2)J-DFte{$0dYI`3LbfxUP10=`DwOloY2Mr*Wu@U&Ob z=#&CZa@%7kwTWL;i;U=zfV^FtoZr8^HwO5~(zC4hcY3vMrv!02lzBj3>+C%H0P@I> z=8;UI7>ckGj^O}IP0D;!pz|sSbC2+a;$wZlNM$4&QDovv3Dsw_K*>>lw~WR)`lyv} z2_RAof|nD!5wc9@2~tFSO@QvLY6JMzU6^O+w7x3J8(28@urh*Yz5&IV{%9RGecFl+ z>lCZCj`DhrGBmSFoTkSQEJ>I=J;z5{EI!-jV9t}uQJel+B#_3iDV7xU%zgha4)Z>q z-cXMtQ|0Z(A;vkUaIxg)JTn}kP}JNMJNu?6vN@W5Q+IcHhS!p|$qQ5X*HA-&04KIB z?5K=AtdnrT125GngHf|E?N2&6T;D)3%#DZ)QDJk=U7875V84NRI8lhZnwb;sy8Txo z*nNOl{-iLNW&r0tjlmvTWxnom&dY5TrsNE}^(L}RLLQ_wbX0LrYu;1<4X?B+k_#t8 zB?s{qse{tWuM(i&yN+-ndup>0K1q?&EtHwK@|o^)J0!Mbc&+LI)kDL^S~Tp-6s3{&h5ezf@0siE+kXX&>7{5T}Kj-GQwMQGJ?WS(5V4XS{uku4!@FcV;{r0Wck14*L+Oq zMDB-&f_Q74uQPb4P5(Mn_sz^-i1HaKm{*Y8)SAUq9`Fhz@GO6g|Vn`TVJgd-2~)lBDR*pjI7!Hjxdd1 zK#Bgu4`Guq$>{T3#VFkeD~TA=oCB5b9b2iQe1qSW$}YkC&w4`^r!BKISzz5Zi@Q=7 z5Cr>wRsba&G#+BCAm|rIhtf0PCsU7I_{K;X`{rotT5V3TNbH~ zZNyY|TMc+cPXdLX7hK=mogesr@N9tjsn26>nLMF}@a|WD5!t}I6wg^M86lx76da>C zPfG=-0pgm+k}sRjQ~iM50-Z=R)N5{@LWl*%5RF3t5;Edvr-@!|9AZWJeN2UWmPJUf zKLH;g6}S7^{y~BL@X_c!#%VQer3ABra{hd-OnU^E_?i=Sdl0PS zUbG$&ck+Pmuk*>nW|ik)7A$Xx@J*mn*&nNLu*0Zu@WI9z5)B+dsokSra@|GDwU>84OSzypzidv1P+d!UY*H(M2`#;y0QFDb)LLOJDK^w2C^WGLZaLMD$8TsCZ^ zU775zpt8(uGLe&6rXEr9?CZN-&4=#H1tDo~2uLj-%80g|i3-P|)%^-H^JIMji$h zziDl!V3(OA@Y!Oj6f03xK#p>DS6mt+V3I4#p*ED+EWRwi9&NKJN^|MTPrIBC`P7zFPOlzKZkHF zFS>`B0N&ONIQDD0BksNI%ATJ$5#H|dTF z!5nU6t9SG(bPgAH$ZS$Y33Nf`FM5)T2J*uN0%8*PV~D|f%;X&O*^bkexa*CroQLCV z8v<;uUqe^WA?5ebU`s4$`LKqG0KN6Y->BTRmBigmAN)bH~Ew8C%+H#@f=@dH-NDZo$BgCMbc$;0U2O!nXF-atc zV-~3~Sq#jVGOAfZN@>OsCWAapr9EgOEU!@Y?B9G_O$zgGavD_Wq4|9PUDcBw*3i!z zX8MmpFQ|^FP=X#(?Pg^;@;WK55^p^l%Q`D7Soxgm$;Y41bFAg?K>^eQFMh}E!QYAK zMP|B8YofPse#|&^oAQ-jjrm7T<66z+xHMl(yFli(k+LX=DikU@x!JKH*6@-cyOoxW z8w9xFU`xgH1lT>7uwtnm*-gboa9{l*T`3HhB8Blx|FlY^vV1{Br{9X818>L+Of-|& z(s(pG^3RrR2F<=jy)DD;6&v6+t9dGQA?`cIG$#qATY_0B4mc8^LypS`0cne&P18Uo za}iC1A75*-T1ID;q$y7Y*lzfb{djyYTV_sgcdR4Cl8GOG<_Ts&NvZDo$`RXKRDDk} zr1}a|);dJFE57{$|1`(&gDJK^iW?ZMyn6?>Pnk`TMr(lV894;YkMZ3bF)GsE0~XP7 z5iZyTrUJc2ND1EV)H?WFjWWO~^Ef)A(4j7RdfZ?B!mF8Nle6>ud~8=lup-udX&gpR zgPtB4c2Z@a(;!Tv$8spQnze?&={{>)tozBj|6OPSKm5LN%hXsDSq1Je{PtW&Rg!k9 z^5-6>`Aw#LN7Dz=h7%`68SPz+QzS4K`Qu7>!_QfznC6A;QpHR|{3@tNN~p!bjok9t zC(W4lRm060*y^^dNJ6G78Zl zdjH=yS44PS(I7jtrtVs*u+S)(92~28pTOt}>ao|6K&<1dXRJ~9@P26+-E|{%l+0BX zeex{VEGQ!F;q<>Ht~P#NdJdODNgFXM>;%6R`_B!H;w@kp{8aRVAbx_-x^I=-q9)0? zf`vh|=vBMLJP{ev z_Z3|BfUDH%u4QnM!tu8TFy&UqA)^daX2vTFJG%|>QvGZ0UA*9}-r7gnma;*dvvNdE z%?m+4qIKto@r)+y#t7{uy49diRvbWNoRsH`m}~21?ijQJ6_7T(o^`4bI4AZN6tuGf zyX~76J<~N8ayTvdQznMcFR{~ z)J-ifP1I4DMi4@&t ztowi7nlUMffh8Tt%Id*TaHuTc`@<(AGu;;uB^9f1M=}TqP=+#|V}MOxA`2Q32x7?R z*?>BX*#@kZZ;^rY%Lx}_R_ZhgSogfUk>q<$2@iyNQ?jVKEqnvJdx1eWP`#SY!cC$C z3y1eRe+v~J_;G;_QTcA?55((9t{?xaoY$|pIj@K;26b&aM#?M@!|s`0)72oHuZ~CZ z54^lMFqpFJF0pYD>pR5g^qObVX7~yu4%m0|o73n@Y}%sr$L95ywY?kYZ=p2P_A1@T z$Qd|=4Ah(15p1$B`mO*pl&ygQmbJ znOyEg*lcOn>9_Lnr)U1b;vpADVHDo>ruhz{RbIBk=n@?CX2)xKX67<Life>qP5J@FSPuEFuOi>O|@1N?Uv2r>ztTpoH8mcqs|DGL>iCFi|>W# zwTthd=(Q=ZS49&oW^UJ419WIVAf|_-N%7NJ&5T>kh@btR$}2^#PR{p-u*Gf7PrK%9 z-gYi759*^S@*ODVPJk~e;5Pjs{8nhH>Mnc<;M&Y-??Xo~>RdaxG2wuRYhi5(l657>R8A z>f>RTC(*FA^N-te>zk0&_|CE019PzD%*{z8s*lucu%YP6 z@x|PL=jgRXf|zH@s;px2&G7@c`WBQcKpi8K1Jy06baW6g9!AuPtC#Ej3`)$gF}Ci5 zAFtmsu`XYWd}OvooNJlSp`o{tQ_E*r42K3Jk|A)U7zgANSlHe}jiM zqlK=_)R08YKdDmsx%SWya-kmm~om3CW z9wm+PtcbnlbhXc?g`w5`XMg*%zV18T9z9URfhh-w88TPX55vL`%sI$yJSjBhbjb2> zNIFW{UIKnTEa#M?{bVgS)5uA`X*#svG$-?rKjv0`Pv|_ z+V#o%X3(IuT=#?tn8vFw*%<_ick1?>4U*oE5Zpj7+j-<~xa^m|!;JY#9pi!lgJp{j z+V~Pl-)5AHk1?PeY&bBhP?h5(ff8~=PlBN}={$a^#OjzcjXQg^ZqY%?fFRxyZ;Gx9 zXQ~FVP8u|-4CC$mjJ zVuVYeCI;j~(-wy2eAKbcn6E=;2mx09l4yWp;D?qbI3DM^K)kR}P~~mbcj4llCQrK&AER3(bFx~*WaPlo zEKlOK)8Z~yhS123RZZFM>~%R2=;xqPi|&q;8)UJRat_)xv^h70MEnST~=5svz+QaJ*$UlX*cQiTo` zfN>rw3{g{x8I~ha=3hO7tM}UISQqKEQ3G7GQDgOO`tGO3amS$?t!Aana>$8aZNtV+INsni#1tI7L{k49*P{8^ZAJ-fB?9{tVA6uB>;%o+ z5uI{?iQY~)vAk&es+p54Yy4jW*!f-V7+U*y*Wf>6exPo!QhR=*!|&ZX{6+@J#8 zA*FFFh}`d7!*2KF(LR7pY#}M9C>gYrSZqPpQ0mMq!gT5<>fITVMInB5tq|}zd8g_< zm60Y*NiAx_rF5OI%GiUPpgT8k7N9UbZov01rU;>~d znY~}DreXxSe{3m15T#O5sxN=r{XMn8akt?;0W`kWo$;1Qb zXE#bbOE)ZcPch(_AOOsE5p}b91$##`)(c|wD*CE}9*Ee8?jxRSN|N-=gwHnWV}VX# z5?$-$aSyFStvWiO<0EJ~wSto%jL)olcB}>F1>r$a*I4UGM;^lgB(@p?E&*)30!=FHq7s5HF7$zy=BrMk~7^3YH$q3(s zutBh@W0w$s7Vbs+Bwix&dQOzb3bL}ue7?v)eEO3@-=OjDa^1CmZnM&m@K1vdfpQzw z!$f3QCQ)$P(?UxL&WACkY#v(Ih~Cyryw(Y`k;0kGNID$-5WjU@60aTG03!Cy+Nde{v}d!{y|jL>zVm9&pr!X807c` zRd;}yPlydTUYl>;1DqP{(NUq2U)GorTv zsY)!+)?5E1xpKM~gBKbbuJpc4WOSnO8=wioZ}$^1fRYMv0)wp)>hYtd^69|S?Q0#_ zpnjIxYj|!VTxLzQ?V^Yr}+*$p8+Q zMt^}4cs98OCE5nr@_sSjZpUNOa@n3G)5HZo6;yDMuZg^Jvzt0qZ=VDr14m#)SQ-`N z+MAK>AQWn7c7TsCcFNbe5F!^Z54$Zjk~1a$_ zcPO}ouUES%p$E%rTvVg>L0ukJQx(M2w%v@c^dISMQ;EcpdAgPhiYZ zW(7WN$4+);uhxYW1R*CjixJ?0JraLD*ZFv20Zvi#5%vc6o%dFvdUUbRX66^BHwA=^UgJBE-x1mh`@#%mBLnF# z%LH=9c_1yC&_~D&pl#}WE+rQQ;ziN;ke86lK{n9UqY>-}u8|i@DSFa)DVWJ<=)gay z8FbBxyEqHo`88N=c3X!9pn<#mi5g&E0HMkQreYpVY%{VpSg6VQy8q|Zc^1+L++LAs zr>9-rZPFO~&>XuC0mj2yunxwl0OfswV#6}`PG z!WVP`44JOpp%yfD_yfEBnS{2_o^YC74nQ>{^9o(!|#_~v$Tw(R-V7yN^S8!*fN zaeU6tm8+~-MPDP|4k{tKoDfNTpC7lx+cv4fz88GTU>BfDi-xJ@E0mfCqbFo0p_$-K z1y&uF;9!xj(*X-x*T7Y2Q74Pgc19gHuImc@fpkj(4SpmHFb1#0hJdbBtR4BNjlCtY z!Gw6jH-|$+xcFzsua+N?C1nap(xwiKCZ+LwHUa=MKsI^Nz>Z1ZkefO`N2R-HXjn|) z17|-Lj&S%h&2zDmpPjC+rS?jE+M!4+_%E8T-|vob z-)RSzcauY1N`%=4BHUs%N7nwKCprPdv3eq`}W$hj6QgLYq(;>7*ioGGms;7Y4z)NE_j;{^Q-neCh_)7> zvc|}Ow0E2Sf;&Cg3)8XFri7#BL$!Z4>pQbAf01|52AYhE8fmkGHmnGzX~5O)kCfXi zKRp~gICvj`KNvue8=n{nzFg`LITQ$FEpS8ef8I-D`O3nY0KRH6N9VNyVn2X5TPL2} zo)>8}SNQ>0j&!b09h8%$HBE)X?X2614KBrxf}Swr9InvZ9QqJB-IbFu&r+A|*|-{A z*5trKR*yQJo86F5?EsY;H;&fv(&wpO3-x`A`bOC}e;re;BU0GVl{jn%|%BQ)5yRqk}v%$UA&$z8L z1Mj^;2Y+7#QF)YLAL>DPH{g zI!~GL=%$b(@s3D%=P`Sias=~nkBLC>eM5R5qD^VG?v-_o1s@MkKe-Azefb57(cl62 z-+IeW@gZ7wOL@K|@`0CO_ZFwg%L!u6h}K@E!7oWP@TM`C@SFg^C{e}2_w%*ev7CqY zL>ri@XT#!N{A!5wBT_Ho0xS?|Qc^=o+?@W^Pm>v4V3V^t5=Pfm=!;oOcOdes>ZA{l z44u>snfya^tHvd$;c{x$z2mvhRK=GkvHKYORxcfyx8#(JuuSfz0ImE+a5-bWe>zE_ z!if+zgJTq?nCF)Jmq?5Jz^kf3K|KDMsCW89(cl#{haL9%E@ z-MWDTCd3oV-k5=z@BH@V0G@Ib+aK(4x~<+0j{D_aw0)b9(er4P!S|7q`|)w< z`JJ1ZMCE2~@`0HE!#l@HK19I&QI~^jzL(!f_O|Qe^-@ z#NA1!7>}w?&Z|Y3D%@k0*{B^9Wmea9UXS@+%Pi|SY;E}J^}YhcS*1%bq+7L5SRnC2 zdfj^NdiKeNx|6^=e*F2JQT;dN@Z2v8p0JBE!uz<5=6tKv^tb-_TZcBW2Q5akL*l^Pt5sNL*SjR5>{Ri957w@~@n5 zzp^_My=~8*f_bm=kFIq7IlLkpEp$;(pI_4Kg<#IA91_hbVwQQcS*+T8 zZCsD1DpEdpYq3nJr&Nbb=f%NMs)70`S)2^)#mXzaxI)N7>?a`$b|~;KN`$uZ z?C#KZ@SOn6twFd%dz^HgeX8&e0qemiS`(*1^d^hZsg54UCbupTl;+ZCtQVe)BTbNfI?$1P@d+-k5o`F+oq3s4^9i_Z68k4MA&97${8o ziHwD=dJPewbr&~dHeab(&@ZHl%aA10`B<^de3$#n>z9G5_BQXq$hJ`QGH2hQvwhQI z4=k;3CdvC2L#o*C1Odb43fgpd8J&1={12#VRqEu+4hR#FwoYlnlU$_*Vy=N2*|GM% zz7H1BY@wmMDXhqfx2FsxPoqj<2QWhGQHe}s^sr0=ndR(Vw+mh4ARfm24TqB48ND@4 zhNNu2@wpOhKA3)uHSApj&v0Uy!yHiw!G>MgtMgnN7GvbsRY*}Mt`yYN2U8MyYF1;F ztzS?;I|*k|AWwbW3n5fcKHo9}>OEK?Ay?FY-me@n1PkQzb`8s>aTI@!B?25!`P}@; z{kd8E+bK0R8bQkN5xnMdHVS)JM?+RC7wGAkK`+H`sKUMa24|zg>OdaAhjc&?uB#R#MSMb>vVHKgmvlv}#mW?6$gOEghSHf7-gw>R+y7(tXGjj@}?{ z&}RYwV`z}KLyZiap8jraFh`Qci5SRo+_$F%@W}CfGrmq<=5LlssW*;7bx4jCQY?DE zrF(CO1XnJDT3(O!suTg=c}-*IUziSq-DG5^Z|BA1E1X4#!)mU!%}K%*3kB8^+W&z@ zLtYm-)Hfc!7NIS9$H??oNM^C+GPdk)e$;ghlL`K?V<5(?IcHZY5I};8u^n?ho(82c zoZ!M0rzN~ zJE9jf*KERz-*mxPp!(gaBe!5++#<#6-`rc0U=Uf&Xn>Rc6FhrgCuLhTln%N=XkFL4 z$0i>zpByNMkYyX>J3ngn?{uR;UJn0Y7T)WO6jbyiqM{>bEs`pW{hRA6(wDQ1_;5@6 zdc#{9=(a^m1m5e%Be>)>_sq)r^c_ar%yofO5^rpF7i;Q9Xe>}PadP4g>qmS^ws<)p zP59_Zh~{9m5(45Kus4K!m=%s3xnx0g6b~qde(ifVS!hd~bS+2wJOH&!X*P$y3vF)+ zW)WZbSjk6a|ES`k`2EkST(e3WFWt)a#q1>ru%DRV@(L@|W~zY`ph zghHsjFWNX1s0s*R4+09d6AHW6$Pcs>EF*pKEW}lDli=;enlR6k%Vj8nYPd8=zqQSL z0?Nq;Rs`VXE$Bpdow*C=dsLe$V`!*&GQ&obiPs4my0i z{0Ol=S+^^4qA^6I9TRgufv!v6%s0)3@E%LF=ZAYaG_Dz7I8oD<+%I~0nnWkr8W}C7 z%59{a_SLe&hW?6;mDiZO6FGwjL*jz9E{Q??iD3FGaT)~kmQ|{}K~*;YG``Yez;gVY z47I<^8h_!+m2lWtY5*pk8EHJyeGZ<*Raw`L)#5+cE+r|E<|1cSJC(8w{N!)~cy5rb zBG-BKNSeTf;p=L@Yb^Qf9UTIQhXp10s^%iV7}+-aPhu{vTP$l_G;yKEsCK5uT=KzHwTGjp8J^vb9||ea!lxSPN+FC)p0HibDTGul9Z! zqc!;|kRZ&X&`=9#nFQ&eaUiYZbG{4h4et;Pg#?ubZ8QbEG^hqLRG_!ZTl1UXs~MsY zD_9|#ab88cWQc_(Jiax?OFf$r}+Q1rU=mOfI;F5u)b+|wn20jrb#q| zkl3dx2NP2J;}~;I!S_jW(pQ&h-Tk=DL{Puqa*@r?H`1!wNN5RRo%|M>`zqjg@=1XYZ;X&S+?thU=9w(m&N=}s|A6pN8okg~A;u!xAtH>gTrN^muf%?7D4~ps zWuze^kD9Uk54?+v#s8GY-ym|(1J?#In8yoLpHLSdUD3j)6IH8b+uO0=Hq0Fv=e}>v zHk}=X!`V$|`Y@nkZ@pCC3IkWrrXw`o8ce4N`rC^!$3Pm=m2dKI*O;tZRRAM~qzquv zy8~KK%))N-lJ%zDMiIO>XY`^|C-(2tc@tYOt)U}RHyo{aE7+8MfWXak6z|Wm`&w`t zm!b0HdjoATRa9@nFiLKhkE6rQW=+z3~1QL z>v!7hUo0&S_W4BL8AK`TgwfiBK~|}Y!w|w#-}GWQiu)v!qRJ z{Ows!1%Cmxh((*!Kr0C~vhpy6E2h6S-x~`m(8q;XpV*5g2G5~T@l*hZh`HH)IP}4$ zpy#YSxg|oZS+=s+gDw(iwTStmBLxfpt0jl1rySF_zB0#z@MR(mqhC(R7Gva%jV$${Q!UnS(0X;pBSvS zJ!t9Kuob)`u zt2yLD`AB%*PM1|RTQ)W&B)m7PdWYIwg5aW-?agAjiOY!v2IqHBH#TlXvb??=Ij0bw2s(mtL6!ip*68fA%RNm1SsN%_GZwXqB$cK zMUMjtiR~xy^>0SrSa7cn3;m)0MCP^4Jk0s-9cSMi!&MBT_93h0Zf=oX1`({bR0uZH z+&z`VU5hq&H8Zx6lDax<{lzd14!J7m!RU=4^33Rg_{*|XuwoQv4BL>S4BYVG3FAiL zN~xh0sm`jZ>frkmZQ)q4Hq7X;r+1f*YtMHVj!xFn+Y_;hRRZ&P@x?a;XjqcPK~O^( z4u?9$1eZy4hxK*?C#|5G=;*p|IUBPwZD}l;Be*d$)ovGiZ=oih6lLBqmP`wQH=I>Z zn*9PqXOHC{=6rUXAFFWhwrDJy-zt_tPM4Q`e(=xBOcvegaDvD+_8Y*u3rscHC5k`9v*Zn9agLuixc znrT=h1w|R+GHDy`GNb6kwu$Fy5y6WKz7Cy|rC_<-XvJUpoZ*e#ggxzDAFekSp)(xV ztP8O0+z6VbB{k6COL9(}tWFPo@3o`M+CwRb<`j0T5~a<9Z){DAhVYyqg5y1NZgOsj z6?Ex* zbdClp7CNciSqgf*wTD%V)JNQ0+UD0g3GBU6nzGq zN&C=_6kXwl*SJc)5KnWlJ;jlIg-{-8?Hf#&Ps$akYbqdus&y+oY%iawSfy5U>bX=n zbs3FTK9|r!wV%^LqjnwHkR^M<=CV%WX0k}dt+r2w1 zoH2#83KBT4eTTKMVCyuVCO_;PCpV_i2}qr`dPQgs1wC9qJsKzAL+f)Y79u~yGad>A zQTrNKDaFLjv9MllF2T)XZsSr?cgC>#tafwsh8ZSjS&SFD919(+kK0?AhYm#Gg5Sj` zY+Y`T+O5bWlgX;LJ{S0K;VeHp}y#+bOAlx zK3s_VHXA>FYN@l+CjvwY@1%NZVvr1Jba-1)N3*#3`;1FZj@ficsu z{crSQn*a1-WAXp=Vv9Hatg4)A2p8H2vP7W^b|Ohk1g=+*u#;+f05XAs?+g zH@yo|@I%9!k+(E5B&)pxqCQX2DdG?_ zgRZNO!1fc^P1w)>xJ{nrt;1|G(g{-av81;2uj7xkD!_*_U)kWO?_ zpKK9M2DFt5b7Y1W*i1ZMd>b(&i*OLj+*~w72#(_!c`$v+l`}MzK3)dq!s!*+y+O+S z4g+4X>;pHm*st$D^p&tk2Hfl_fvd(&hy@7g(a&5Z$IB$f8Iu*d9%N5ceN7r!=+7y6 z+ws)7+{nFX}4*pz8+t+;NA#k zp`EBLtduBR3|(H?Q7jTQV%VN5@7Uy#MzBxY8E@>p@n@kqE7#m*MTL$?b&*{d_ju&7 z!ljVZM0B>O?L0+>L<=BCVMobP#DGjNUV|wMwA=8lbxx6x@j-izV@1sE@i|z=HZ?41 z&>x!Xae{UY_8!c#TXmcpv96GU2D({ttnUvYP0_10`2!RNtgS4YgnOod!K#YYf&744 zYWuF0BGM3vc%V+d9alC|f`wd{OBZO9be1EdNQf|M<(r%6`Re)U*;Z~A@EDw%0gKoc z*m>>I_R>zc%WwPjo4P5OHMNS~vneNy2^k;wH{(d{ljRuN+4Z_4dS7*-mC>qja%tIbM1$>S zpFHojO%K=-J}h@(E)AJuSHCyW@Y+UD+*ZrS#-bjPcHhQ4gN!_u8Hg}te)tr~oA-_3 zbi86zW9QzVrLOs-I4d?EJHeJ-*fHOY*A7=v>jN|av4sNh5~nW@h*Ov5vQAA@c*%caa9)Y6XwRzmZ-C^ zvLesPtHZBoT@8me!SvdJ{`Ps8D*Xv>ty$fI*@8{m!9WpcGrF|w{2A~d~& z_rjEO)i*2}E)T4p?VM5Z`ZLWGOx=5AdS5DDM+OEiCM2^um~w}xjQeD%WBUWWVvRyG z=)Iln_b9*POxomq0yYYE+@|4c#_R0krw5mTP?^P6`A~x4(L-em)|;?uT(3*DLbS7n zh`>I{b2n-+=mcu=E(BHu(RU8gS0DpH7|%C_P^^e9eDr7l8PWF;G|0p!-E#D9;;5KB z4U{GfLLT(E=)72?3T&aB^Art7lwHxC0Jn~3&Y1AT%mHh|m*%@?M?^xYBVS1f z{^H$Z0TI_$uW0s#hIOvKyBHC7RyPjx}-Wx33uwg?17cv?F}IJsnx+<1NEcYenj0dz4)+ ze4A9i_1o1)eRQ*Ov%7Z=S@#jF{G4OuK}cHG}+GP;(E!`s{r92T9JWFJrsB6h2jZpAsp8@|zT8_GHTCJ(Cu1*yJi9 z910w4PA;XjS>7}fW#-~FhM?v49*Zpfjs<+Ngu>Y%p!P<6_+asYUBsG^rF%NIYis=m zCSL4*4S=rzOB{u8n%&w(dIK0O=>@^}_R?x1x^n7iGzrnPZ~G{niS7XEr<)EftYg8a z=hD4}mMlCVH5O_P0RW=Q)~;9(6OwjU<}bxX78na=NHrLe`F$Ze)DbI}F>=S0bx@}h zl9(EwPGCL9y6?c92FLDD!OxLQqprjc-b+NX9dP_3P}Dy?`-qF&W<{=V`BIs|7_dv zYs!8o24%~R#PIUW6ijNNP+H9785W(Bzj5WACyFI`2Gr{PvINy1z=ypTn+)&=%^_3R znLJ{`g+vY2g&mN)eT3OW0m1>LyipAs&!0pB<(E>MewG9F7wjlVkz8W0I3$qfL2>8V%4KP z%fqoYUKX3)nbfLYir#G!BLuiGv`_HL*-e1leYJL4XulF|C(hS}reMk4KDenkC_s;w z*n>TFS3f#{f?>?Bk;>f6wN2jc;5A_i0N(S$0@g8Rn0Uq8W|Rm4fE^jlDvKtjn4#}E z;T{&K7X)Hp(3at8aAMRpyO|*16^=GyWS@!u!f2Vf$oCJ>pDz4ktRi+nM@TOflsa=e zX7`GImj;#PZ8RGumQIG=W z&&K5s8lQvV2%?jT&$V9(A`N5>B!gLG-`e;GLr6H#lN<6y4cca(IHIBNok%aV^9M>% z5eWh@MC;pTs6p6Vj3)Nwt?QDY7ZZj$n)Q7F*8xk=&8!|Xs)YvS-I?c{8awcw6G~J| zq>B+Na%&8zD$IH=W_pp^B*_ssh%;yXqEiMx^f+^vO*`LBB2y)Iz+h^h{wj2mqohQd;S`+ zxBa{5yZo<)4Admt--m1xKHByAM~WmWkK@FZI1S!FSi(G<-fXvKZq*4k2n@;l!11C+ z&WrDZl@CqzbDkyejHbEJ32BZK`fnPdmtfyo(-Yh~D}z?lk6da0o)2Ye87?3mVTtgX zAPDS?s|O~z?CUjqsSAqGqOMUqmVHJx)xvRHfdh!-AxM0R(T7Y!yDMaAcPV^-@Pgr# za}h!g-D**mO}oS!M);nx`nDq#1i{$7x5P~(qYF94J>;`$km4WlE8tG68^8YK_ot$a z?`y}HGK(VOvY)>d6k!NPxx*i;f`U5r;d(>jkV4*#htpU?5iy3sgO)Mvu-W!iV& z>26ETKdV)>w}4*K%dZ7_dvO6V<|3oo@J~_;0j)kz{y_Z=@uZ!ZnKe+&y{ACXsB+}* z%DsBUu4`1XzK188GkX|pof3jME-R9qMn(v``V7#d%v9ewnY#1mcJ`N;!N)fn045Er z!jQhd2rXozqHTdb^IinR^lhgC!t@gQ5(A;ef@ z&jid5F#}7YvNRe7z73xH0oDi1l4H}V362VQk|t2UDLgyB0Z$JBhT|{p{nVE7Evqy9 z@!-v?iKq$q>3BrKZp$@THDsR}3DCtCB1>xCrA8HxJm9LeZ4uGJP6Ej^>b(z@$^P#2 z2fUw30XDL7;+Y;SP&w-_JjNYoj)G>Nm^BPl=XDQP+dm^}_4ED;+UP2BRD2+KFX%j3 zl>1kRCVi)TtJyY>9|v?sGdqsj3{&1S=v-|9xvlC~=cF=tDS5=ydLhs~N@bUa*~AGt ztg!wVs0#`4Y}@#jw#Xv^^c0hC2H zG8SNrINrZERQC1UP}J3Z(o)tvm2b#YTsbPF%f_|FlDSM_AP=W7;VD{eK~MdDkep_C zrzT8g3WrZ?Irn8!fHfkHeVvs=%qRJLdU%5lL2nUIjJ_Z9z*Q5q5ws5H6Q`=)d|Ke{ zhJ{)E4GhH|wS^JtkN^1)y%94z7}?)+B@!tXl4_nfG zUM2sSIJEO5u&L?f&JREJKKuaj3yx*N(E#R~#G%C86({J->i|&0;N~PCCXCc{;x_om z_BERzba<8ea)0t_A;SsRzP}i7p-+e6NOARPxM+U_W%TzOV?4l#Ds!{5suz%qOH>? z9!23W!^L2-qmAL*DR6gp_?xzNw)hiCQ@q8nP<$T&fwADjDba6O>Y%rGY@xEFEW{u$ z{vnshpq7VO-4jI<5ulV|qP$htj>lgzR0Ie{0|qtK!Zi>?h)!4gP`S)(5)q_;6sxM%eQ4JVZ8&I$bHfi<5Jex;_Hi&Sd7$RG$K`1oTB12w&SiKhn>u6-pApWxUg8jr8ds05 z>OJX=9wl_#RV_D#-H&FEt|LAn2`GeW8wp&_4Wycct0Yn!KSr-2aMI!5yv@;&=ap^! zClG%w6w<*uITn{i76JNG+lkNlBA6t)(^k;4Lar5)Lb#qp_G`U> zYL^O8oB=m&nf>ICJyh)iOwe882Im)xB|UDA%hGffhn_n66w+28U7s~)#iI%v?T6Ce z0AC`=UtIS*CK;yQ4EEjarrR@Vyg518*GFJO>OB*u??>>uMxXtK1;`%{D^U{QesSx! z_rn+Szvw_;q5@GzRrx-qW2uHRy3F{v3O-GZmc0tbUw2$N`(DL{|b;p^d z&VUp^8HksNyG}zsvl1|JzIk4ol>J8#LjuCDybMip4A>|C+-|w?18)WYQ(2Dfzk@<1 zHm3hA4%^bO-5vb0u{}&huZ$D%2G*2lOq^oK8B^RrOMXdHhDc1Gj_*8<(dp!S`;A}Q z2^g0}PfSWnqq;7L1mHh%zF3O+)NATKV#@v#{!6n8{(g0{w+o(iGjqT@`FwT!Z-hXU z+?3~0ByIipn*XfdbjEZ5n!a(wRN->{_`@&Tm&@1LhoScin`2!bcbxGwD%dBL75vKB z4DY!tc)Y2l9fq2x-h#&JB>T^m$4~7ftLNbQlmoSap-cGSxs>Gon-zW|*Qw58V&jug z#QNC}>C2@Tez6hmWaxW<&XL-pa%PEl5x9EIIx1&-kIm%lqcB|Rf}_5gveiW9e@IxK zMWX1JDyzDoOLYRtGTB; zL(^b(_Bc~_WM55Um|0oAd= zqxs$6gTsdF&3o#_*U`@10_z8BV&QL1Uc>oyG-M0KCZ;#+l&I@6Rn97VgB!AK@V~TF zOco4VZdktLQ?{~B9~cML_V+Vesu=rY8uZWY00CUayyWsR^c7QA&_|dJWnsSLTi2)_ zcVO~K2h*ZHmD8%@mHtSGnm9VdR6o!=wFROJDd)+@Aq|VjHCg06F6n7 z(WmVX`ipZSP?y_+x_}gQ<`}b-km3Kk#9EL2TF7%{{dKE$&otX{h!b%P=BBN2bI=VkI=q{AP=xxGcrrc^`8j{ zCl>;M#7eX4AOjnY*EdZQoLWzQx@9}*O~OMq`cu%*f3b&jH82|gs00K2QX^HYyoY*o zj~77effx!%NCpZW#)r|#GFg)8U|b7*3!r~P`Z?A(_!BNxFPG&QYB!p*c&(!2aumsC zgSMRk0r_~1mb)~a*Dn|O3Zy-e?nh27d6=_3)$n`qK9D=)>WmU;5$$Q(&1q-@z1=Py#*oJf(b~TfGKrTB+=e6ACO5b2+F8k2$A0_ zQ@r0nz}7uE68hj2{$zPh$K05aA{{{pcVKEER28Uzd_ zWX{>&Dm57-AdpHaPF|qEgnngyJ~ypkK%UJS-~==zQ}b;aRXalJ1lCfYJr!iMV)Oek zPTSpFouRfy#y`p8iGRHdxC6O|xjO;I6?PFNgI3*MI;w0YP#wgwjMsOD5dxCl4cBY@ zgXDqzpK%Z&h9U~dhC<(9tPS$~^NzD}pqO0X4bN`+`aAI!vDkI517X%-jlc{J40j{4 zc|+a-O`|i`go^Mzx^6ceUu4lyT|Q#lLqswQHg#G-tC7UWL+QTITVPUjmlFG>&F*Lk z6M|*qj`YvTFy-dM0H&nH(N}jCLVQqdn_Q26KkrL$FSXm-OJ%M|LTuGa6Fu=sgqR2E zWeO?0)s!dC#N3eoz zk<2E&5Z2vm8C*f-%TjEJZk)Y+YA^PCA#L^S0cQzq(O4Z3aZV3-!yMsO8sl+=g}%n` zFIbMktb%vd7NbX19MMg}TZcJ_5Q)ftW!}rUi}DS-h4$LZ0>>{RO9AGw>%61f4gkLh zI%5W4pA~AIk$%L5%n%!jA)`#2X<-&%JfF~K+1Fi=Weu-?@d{)F5kDsN zce`mU7Xg$Ca3!eV6u1p=;{S#N%3Oo73a8)%VA@Ar6=IgF;1ps`q@!;ict_df%f_M6 zm~Uj;v6TE}E%-;g@|!E32}I9^E8mcEZ&}3EHpFVxPJ~W>1F0c(FZrKYwIGlo)DqDX zc`7?kw!oZ+gQ;<&JU<}fPV#l_rX+2(NkxB;Rb+WVm_|E#i&EMJRW~;MecFx_G4#`@ z#se0Zod9Gv<1`lVf~}#a2JIjBpUrWB1&|Ed^TTeF(#aR>SkgnW+dv<+F^m$7zFMYn zV0kT?gdrEL8Ao=Qpo6Hj=57yP3DOb2%>a^AD7KOHRCj7b3Sz&XxiRjaD(s@t+XCK)jG8DjXxJ&K@qwMGr&)Yk;LGIHol>8`MVl;O+o1W6y;YWB_FJvH2d-mxZB)`=07EI0YdyU^Y=3)CYmIL;boh z#O)@act+s*06x|02~q?WR&vL{DcoXYS{2>Ct@d<|;kIwGkA?5mMHPS#vSuV6snc(z zJM;mN_kDmLt`=8OPN>zrUV~#uNmmW1ABW>X#beJiPKU#9>IwzPr3U9=BtCy$7b(ge znph0JkkS;)Z4QosAujexF5czqkd}gpt^Ocok9~};5W4g%`)ye zjNoHgAo>C8>AS(QjVw%U)~{_F=xXeW`%v0fBy^kS2DPRy1c!HvE?8E8v>0bDKX0b2 zKn{b%v6rtb@ryT0%STHs*gUAXf-d#q?i!-U$NKxd@Nt=RpQ92=tnY&om^*|ae-b1qgW@(f83{|H^J2M4LTd`07P<3M21CKZ=5#A zX#6yeME$Q{EijFOj+_p3ir`?M!%1b~J3XJ`J3K!o7GXytCh@5~@KZ=I!g4fUYXA?h z%lkPKlfCmt8zU3HxY%R|sB5ur?wg=Z6AtUAUE}3YKwyPc;cB!6f*g734)LLsi6*Ef z@+W%_?ig?Px6dee1Js9b6rvK5$ii$UZs%Lb%b?y3tK#*oQLB1loQ%j{h#B|aAP-VE zYc8~bHVQw>MGD_peuRm=j-U8g$Od{kXvhF}H}r+gqRQ%K92(@UA%Wmwa^>EIyUYd4 z8;5q7{j}?ZXw&4I%ngz8oIq`4!zK#Ta$LiR65Yu@D>kU2_@En~-0k-E;GBi#jk zdpjlwsF1P!d{VRE7=ip!0C>ql(TF3gw!A5Y<7uOie+e!Es^?U-*xqeY$OWfDK6;t~ zGJhjiH+-7kmJ6o>>(ZLSaU;5!L6xt~0y+XnBdEE7FbAOH+H_OeX&ss2hs5@!!{$JU zfJ!*?>UknC)rf(O!=@fsF6Tqehne@&puAV$ib3d&gWUgN?3|he0U9h@wr$(CZQE9t zZQHhOcXipeZQJJ7e7mtPGZDK_`4ckZ$ETy%Q7X}5VjrQ2{IJDR_=_$=Z`a`PRKMd>&_#oGBj1Xka z(uZkq=K{s&+=a@Y!rIlQlkBsHF+?O(=O^C1j_z}d07QHe9mxptOy)p=n^1!IcXI?< zkI^BFbASSJC2mk3KP}-fRG&g3q4>|ll|HH)`vc_(jPYlRa&h>d5|mdq-559?8q_uU zBgvdRNmY&x7@MQgrffb?f^I#K%u+cI?9nk z(5x$TsRm*2M|jYt?%WJ9<}!hSl@($%LTWT?jqv(FhmHl0Zh%;Z*JO1kre~)lPY?GpLe{5Da z3f_X-u$=|rAj_JQ!@-Z3Zcvx@VhbMQSq&wzAE);?Y>_7+%uR|2u2k~IYhkZC;upgu z_3$hL*=P`YEeX80L(XqYRnu4HM^toSP7hSWg4XSb(Yq`2D&|a~UOLTi*HtxMRnhkY zSd_m*8RpGL+itW>;X?R}&cPP81OWtSh&U+t;_dz;f>mgaLsU{T7dz*b2vMgDmFzFn zPMz%z+_gC?N1jyt`w!Na%iydSNzYZE2d1QisRhLEn`~A>J=5V%)sn_WCX&Q}n=r5T z;WGkdQ$zMMh?0X0tbCmUbOgMd!KgnD?T~Iu%zxSKN`ddZy+B+CBc-KlUAhqQq^s|s zpe@2aj`VsLI8e7?p=n`JjL>n*c?wK4tx0+041Q=!VG!zR3L_bA>r~=Uc!@S@iomA_ z!8^bcjFHDh=KgzX*ZzRn8-oekiYozfACM2S1@FK&lqSzfXNZW_ctnpsB~IV}s+dz>QEm<(gWv0UP6^@G&kI{KkssmHq%cwvmfs*(bPeK5T zb3YQ2FYy>MeT$2CVEji+E(BFxSHuv!e1s@QQj`AvmZCp&+~;%^Wv@+HMxe1=d03}m zK`Z4RVB32!4^k zCvVhcV8t}VWuy))a<~FHK>6UU2$LU#d<8&4yZ6TgDj@Qa>m*iarS3lDK7xmtzuym< z(<~q@0v~SfEMRry-(7b{Xy4tihh4~i30SZG_+0=kAMyU%ofrlJpa|jD3GylEzFo}q zaL|vO)3oCI5B4H>Py5837#cZ-p<$V-+AM=?%NF|#7vaSAnt{#`qqA8In(SENVl`*6 z(7V)B&?$*x!~6lM`{aPWy_{6W^=s@g0+P_S|Kfzx1U%+O_Ga=F)8pK;)QP11bUm**(08%bHG4&iU09)$cl_o6;gHjQ&!yiDC4SN3|L zhQ)z#eC-7v>j*3FBQW5nvMUp=k$@`1Mcxpf$-HoZi%LUq+X$u8dtq$O(oz1OBF~*V z^L}OWkvqVicVN-}*5eVObYVi9_6te~?;c~C*t_Ho%$3~pUVT(XzZrY+bG{!sW~Hd? z{{Ec^h*w=G4VTA#gNvRPin~oe`aJ}|=-8LU|C1g z?TjO@m#^>bWl=J78VB*@<>!5GPoH^)bTG1!S&Ve?a?`^TAorX5{r;i7?>BU(iTI5~ zu1QA2?e^QbKkv)`)yQ=Zk*iII$m$$}bfo`0eNJ^L?$0$hSKbFR?yW zCw9cLk)(;I_WfY`gEPDkm@EP1n(QXky~SlwNN>Bdx}MSJJwE-g-t@x&*YernDOzMryE+&3wm zhO*rk*C>q4_FIDG4HLs=RHcSEbga6TL6VFpbl{S#l_jX$?!!E#)_$5(EdJd^)Qi$s zZ0&?;cg2~Q2DkWvu^qDCRkoR3mN(t56Q(%u8=~X1!*#*Lk+NMg2a+9279vm-YGI|% zIpWNf#i0_ZmOZ+_=lSdD>Bg)T4}X2@w9E|jsVzj3Oar!k6x(MrK-DB4q8zz9cunal zAskLNMCm(U_wlNwax+72NNTnZg{`sEv@ejKxy!>4e~@;p&+FOQDR9N^7=TWC@vKMz z7Eg&sI!{|JK<2K$pK-!Wi9iOu}MX@jq(_g;ptK;sK!TO=E zwyNLKb12v|lUjYwvsiOrH*nKG7szq}?(0^m>KhC|mc8X1KaQ|n^7UCEyRan497Gug zh)EED(U>c@SEXPQfA+$3P@Q6G0myW8QKL@km7ET@Wwe@WRd)T|Y}BjC`KSAMoA-#% z6E+K^83?US_yMYqV)r}z86XOF<&PlKwTuY-v2Bo#3OH{;st23&6X~w(}wssDDEbbGpjqCR>RaKKnfOEq4Vdh|bFUR^{ONbk9cYe`$qdRuk z0IOHdN)iv@Zd2h)kn+9LV+D?&0nh-esH$x@0`n{_1{PqHum9odJEqcLHM`AADy!^*26&~}V&s@ihcI(Oj5kRXKFnUw5ef}MwW_~6bS|IAs@*U*c{6IJu z$@%+a50ogZ-l&vzXo<3urmxqBNpOO|aEuVP9Ud*0Ku_sGywAuCt1x{PJZkc}kLV^S z#485EpdfxBC?x5{tJ&T;9zo4J+zpR*^F%l<7#07~=64ZS=XNR^$~BG#7kLI#VuEv)H{Wk|`P%f8|BAQfC}4 zMqDvtGHfSZ3+&xku@!JKv9C2lkt=#|+Beq=xGV2dVkiEj4v%j}_h~BheSI!ho)xRv% zqbQr-%60NPQXOs0=2$7UK^H|k6+!KN<*IW5rOZ_WhqI<3XXUkR?S|c~F>~V~<)nq?wu*`xGhum!e^3^g zKF|qTlBmQjjqPJAM8)qvdsI9N=II`Vrq0493+Wf&n^f`OIe#=>*Oe)Ud~yf6wcbyS zrYdNLZtVFq)XGld07-3z*^ZsWp%(cnEo#muwr9CiXxrs0p@=8R2SeJ*><8!(f5oVV z>^dByiXSFW;)|HySD*||&W+!MYnFP$KGRX)Eh{jV>pJKL2R+#LV{pgL6Td=jpQ8qQ zI@M-I!cRg?tLW4|z0a2Up|h^EDV8S!Yb?1H)F_T_kd*Brtn(Y2on7Y6b+?TA2kjun z+{j}Rti@PXLQCJ}aul5Yvo`8z42~uKX>~hxZZy@UGj!(#(kAjUiXDNg#@U6UMA6@# z27r!Kv2cq`MTvA)!~)8p8CMB?VyZ7pycc7ESI<~GdP;39@{$B0eWd0vjArope_

    )eZd}-ir9VI-Vfltjifnyrxp&+UL zv#cH$h~<#N2s|BT9hWVzRHU-2$SRc=g?JwdOj$N@G{yfW5y(FT{>q8et zwCS0e&q&D{H1%6_(-H%NvO`Uju)_5HL~qT;F@Ufa0T#_)(c}}I6s==vqCGK~^Q2Vq zUR?dhE~@%lusg$`tMq^H#ZWO6X;m=*++%HV`vZUFUL`N(>*(RuZUTsp`I ziaiR73GCPXy2>$olX!vfD1BhP7x9asM~ zfPlA>QRE9_1`3HOQ1)~4=AMF|#-KL?5$Q_6qj}63B>5%00X|6`y@1zk7yyIr= z3YT+xm*%`7{(8t-iJ>(;kI6mQK#(K&E@)#wX+7veC=S5r51;MMqlPr~4W9E0#24~C zdU2I$e;nL+vZ0d{n6{jcdgvy$e_Asa52a`276?JJh0T>@N-lwGsLpT`&95VNThyIe z_Y%iX6YA$RD__<&dH3Tpjcrc4FAI2q1zG$gVwj8u5A0?CW`ldnV9#6oRs}%GX6i(5 zBL{`;A;9=?7`g-%`zpT50vj01&rYK|@-OmsZKY25)?G_S$a>c-!4JG4ov*EzIpq7c zY?+nd6c^7w|~)T0KL% z*UxC|!I@TZ$nlxIV;tjEY#Hg&OO;5H4?mEFZgkDpC`D?VZJUMB+b2r&kShrBqKkbC z@S>wONS1YK!Rf)uUG-LkAl^qXNQwSK6KQT>GwByd2FpM7QN0=7xI2L?y>gwb+o-iz zmt+rOQH7mPEW`N5k-X&Q+D%L<9L@<*dFKHSb%NQl8LxdIG2z%ZosbLM1y$$va9mUP z9DyUG`xb@Cshg<0SKO&e&bm?<{k&qQ!Fz{GLBHu=G3kmkthCQjS8Zs!zrz;?&u(gJ zlRcg{FkAFGSQmcGiAW;T>~A|bAW5~Cle`ou3{TNdgB^px3gL?tDYt(T4Ehl;jh>tv zOWfVA7vpf*5os*42}>VV(eTvt$fW-gBVq+Cd)%l0#IPYzN#7FJl$V9URprbx){RzL z$l_zB2;0AQK2pKpO3z~ zR%-qG@HOJ*r{<7?om0Lz9h?Fgj;62Mb={uzoMk>!*+}`o_QnJNYx45^r$Le~(tJ|` zR_NKU(oy%v?H#J6lw>kKAYdOyWmFbJ&j%nP zWK_l>5SaKE=0S0Mbzh$6InWeGG%GM%CxjHU^|`w0)37_ubiuVD!+?BbWMq%9m8KXp zO;m!_fg4)bU4lYfH=h3BWA}PQf8<2j2wJ`|IHt_-!~I6vIA4XZDjv#G{n8T!=Df0? zt);bX{#{m-3W_LCLtN1AQF5MiS$r?d1ro_l%tig#M%nOu|3-li$zdV6D%;j+oT!_h z+Z*d0EWUGXep}|1^irsG6|EgmOeF~r+S&HlwJ`Yke81T`O+%(I8ZH!-xqWY%L-7#| zI_C=Gav+Q1BA{|7*cgJ+>X1k!>GL#^;#*Y%OJ8mNdr*h|4Vm`wdpHaEr`UMGaEt$V zEi{tmseZ#1^Tup$tl^RXMDOg>G@VJ~s;jyOKby+8-sWp!`!B5Qd9L{0fw+n-fBcLE zY^Ms))Si6SyAvGH;bpP3z-GTNx85L+PkHpWF1mZ zjzrJ1qxsblf6McM>`IbGm+P8D0E#=1+dBfDt=C0!T#@-vyL$5mRd!m5CmY`528u^~ zph|x4uYj-hLDAnj&IVT-T~}CBL4nvsHMERFiYxQaXFr`_@gNZ{UjLau(>#DKpF6TGQReM%niu_}RU;Zmogg#B3rQv9gf0guzfwguTJ)EAT&l z4|Sje+b_KKyOe~H$}WATHhgT&h}E2 zZiB0r!5c`Hv9F5sqW1d}P|#vJfv1M&w+pmjS$A$HzLs6;+cm(w=A3_39da<&(g*B)7ve}OBut|{4z_PYD2wnhP_ev7)+IOm*lDt2)@2O`cB7t2XXf=o z+$Dv9ILf&2?Rp{ zwrjYM_W(uw6}Nr~L>aqAufuNOKj1xq$PoWSMvk4~zq39T&fje_dNE6DXH&=D)7sG4 zRK(QS&cu{}j}OYp+0oR{2FiWCMti~$hdp)6iM45Vylb~s5!sCDPNtVzs*qGOStjbS zmAe{FyfNLVL{(LCC5Y+k-vcD_1F+LIXncw&>5=Qd0YoD-q&kpk-j_}2*R+Ag_%PD1 zv~M|d@S|U$96flVH2gx|+5YX=P~X9Coi>z+;|Pv^8-Eb(el|=9v;>YC9m?ivZ)jhj z)BEFoZQ@{TXvRVXcHKhL{f4T!qm>Im-U!qNoK?rrM)h3dKzvS4MfJaCP_sV|_OraU zR7FW^q*8`JUCg1YQqsc2+bN_mX?i%v?3%f7l7A?^**NF5qW(=y9fpJ77i#~@vUM4u zkDQNm+kd@xT-ub2mHJ#1q#|GUT+kW9B6N3{NOEQiH~VYbanvz(DDhuIsNI@>9~M&{ zg;Mpeean?Fze|&%etF+)68~rTlAo`C9qoRgmUgRsD@T0^Y{-hw*yrwNA8+C+qReo{ z9Phm8o^U?zyW(fL5W`P8j;6ELxD*d9Lz}{W(jYvton=n8MeWJd)*$y{Z#JWkveCO# z)xw-qOvsr` z*(R;#30IGp0Bg$WmMCn`LcEk$*T9}czJt)z^wN?B+$XS)2Dv4Mg`1UzF2Hba(rGd+ z#LmES^(yY?c!x| zDgdYyIcep$)k|yEhDbDx2wNk|rx=t4NHpAMLeU_mck-`dz*FwPk{hHgFVEW9(Wt3v zN%}z49p@`YVSZd+tGj0HZaeqZ=ri3``YTXV@yJ2eMY_i4yI9!BfIGUJ3c5G-l2-r@#}bfk^AXQAjOd6j8!DvhB2&?FQ$Zp%T ziBhsy;$T3;txnySo0Qk@PmZt1L68aQ(0tv!ntoejjH4t{1cy~X2a$QM4z=#ql6Y8C1`1KL_#S32vCx>RK&t8bvGKI5r$A`r32Z9AA$Z(8?3Hi&Pt( zyLxRNWK4~G#)gpnuAU0@C-3dUT5p7f*ij*842y^^?Ks*hBBuB?bJRKWt2G2o1|I46 zV7vMyhwj8wL8NXE#=j2s@t%3A7`L50yL|@{j)P`WH2(iAgQ+Yk^@h1jKu@q4raY6j z6U~bojNnr^^y)hBniie~l@h@?Ts=XkeTDD@q-ctL$fJ{Q$n*>5h}AV~QP=cCdxNV2 zPaBjBsR0whQLQcGP}zIOF_ZW@*$@j{Nt0d>D0Z_H^Ho#{ZE|lm%(fP*&Op4_i$d6V zT-JB0o;@34MOkKCC1)

    er&OXt61JLIP^bXf=v4bqQK!t}gxm+DOX$3I99dzKEVe zFI@X`fA%(}PyK?*N}Qk&qfmSw0r{RM!k&TqS6wD>Fn0aVusBH6$l-((1%q(0$wg%4 zdKhYW6Cf1(m~*-;KngA?IbBIWpEfsLj6kFgmUhI$_1Cd6ZUqxV!pq;xT#}4cBFWeq zG7QDvblyK~1T3E#!XOl_3URG8UQEurD`#br)dYX26lr_qae?u4M61ec#U#hnm@iFV zIM;9v0TL@2;1~~-FFJ=AR$!fZ@Hk>6p{rh(!#J)h*M8&iC_OH{;w;rPi5MTkU@$)i z%M1n&zcb9dpUQ1tz{1ZzzA+KqobT~|;i75n4PrX1PW}__fhf*qST!t0O}fiz4;`Bl zQA^M7nM+phTfP0YH~U2WCVwA4W&@zxwd5}*2TmOq1){%Th}-)yF*Nlun=mml5!nta z1E{-gvf>{5nGNycV!|qx$?!bwyGdl zf4>`Rn43!FLcCHK|7ntfxO)w)x=LQ_scSvEXEs=(J*n5xaG?RB%f0fp={PH~1&VYA z)Aod?w9E2^85r+B`0$_uyE4KA3VB~~6&w;ML)rlCLHOZ?EY2VpyHM3ehWuS;1AX&- z>y9FFp2P+|?#m#ge*jXiE_FfGJV@s#qFO7@b!{Rw>`P>7K=QnNbx7V0^R*i8M<(irYC`#v6Bt|y>7 zkZL5LqmCtUT?YbrM3)HaX0P516;U+1^YGew8Np{d$7}ef23I|Y^g_) zjPs_^91;PKu{RB1GmKu@13^B{i;QqD%W&2UQ(wdM3qm)TP!|D47o~USj@fH%#vN0u zh{^$kY`55$@x#UTa_Fu47=3PT*nL2hic@v~@u3H+**%;G#Z_-iFDtwM(l~xkW>8Ev zPR~32GJ7!*q3;G!hW31_@+-xcS#R5|YI8Y)bR*&4wBZDE#h|Vb$kqMX3JH2cYI6?ue#AG^L z@!mb22DsSfss2s-<73X&Qq#$?d7@jZ)FSPx5yfi0j<`Wk|JzZ1TvWwMn*d~6e{DuN zu*F1Oaaj{x8SKef?3}i1Vc(hH?p9s|G9&B=0=)3~$+GrpRBMuhRpWk9SbXXbe-PXC zo~oYr8@DB(_d?hg{^sTQ=CkOq7P?ldS#%@xjK>ETTUoZ{$taC?fQe$lLk3#6(UT)6 z0%3pQ62mlzrny5jUm+!yIbB`%1656(NWoJE2>c;Y-~sPM;E#b3*5A{*1{Hd_6K2lE zxlbHKETJ0VsGC}uA!Y+z7Zjl5-hpdTRV>DLUZ8ol_dCKf5g`(c-bi{43GipW?5?70 z4EtNF6O%V0eD?&I@k&*jPEpv#^A*%o6zfPOK8?5k6sW_aOQz$?vxKE%2CsGM8NmK`P0ULCF>4IWYg=2ey9HQ<# zoqVM_q5*GP^B9LMbY^QuuJ)+pV|5Vp27a2zG~UvqCj=>cjtt4J>TNPN+ZkMEJnypV zu~3t(#Ezl^k5E(KauF2V>arr_FW@${359kqlEgJ8Rh~| z`e(6A%O5Q`oZj!hD{h4GccMMmO^BbZ7arz4!PScXqOeKe&S#kvWjb~?2#es%JKtq3 z5tsSZpK_UzzIJVI!sIg^WD=Ro3}60--V6ay=u)~pR#gC^XFYi#e9MlMhL_#mto|E*MW z!BAGpJkYG++spUiPpnA2dtt6pW;TSNlkzg>XAx1trY%9HQz3uq$!pgr7n-Dp~ic3$*xG+Qrp*-)1&Zc+uqDQM| zN3(z5kkqz^29Nyb&W3cNpaZxt>DjeH5sFs9c~1jsB@*Z!?1$zpmGoy82?kW#_Oug# zuSsBYh&}T((0M0r0HYgMs_Z=(ciID`BjK@iMcUzUcIQ_!TVYSF$Y9HTY)anyObDGLExGX6BZcd}E2RTP@ zLVNS%I}u1{)jVxPIwn?DvLbN4$)ts)@2>Q~YGPf_bz}46GQ=*gh6`0(cY(!s!V?Uvc<+}#Q_`d|^4Kj`ZXfz)qJbDS6%Q*RoZ3u~zvZ$G<< z$`n*9Z)aFmloC9bi|g^_OaVsApn=*)*{-DnA*#B#a2L{HQ_060A1|&FIViTG}p_XvUB~lbh`B-@Zj!cxV15GxZ+ch%FXY_R9 z?&H+L10rL-Oo8sIbit^5t~)b9swX)NhdB~?fk??`5fnbYG(|0MPr)U(@if=8UL?A# z*V3f2oRSfI){={_uE$*`;lxlG@%J;%qGW=;uo4wl>35ScO(b12j29ys!M#EBQ&F!> zT)6X^BMYw>k%QRPc*)anu^UO;rF_b_n@&)0)$(&ZN|!6}3<2tq;7IbvE!j|0Ix?gx z&#@3df2CFaziWyKK3@VYG_5ni%R`NOtUI>G!h#R-SrlAcr+e;kH%FPIn5$R@WmT|S zoskUp_UO4|Fw^MPY&Pq7nufP96#Da}!VS{o8zbn*m6?M8?PcvV(Dp0TX?5qOfF_)M z(Uq0Y^p9m81gG32QBVawry#w#m!W!Q^SN1Na{Fgkm~{ z_EXk@-iGQEXELL7K4&>?ZRX2NxC|GP%O3+G|CgqZx7t-0ZpRMkoT;BAN2yfRr3*9W zDDg>xJ-VEBZ=x5Li`7Fd6(cv9D03(M>_h=2Nbyap@H-XJcVR_Qufma}p>tOWXzd6o^Bu-i0F`wYh4qqZo~rB%TfM z+DXUR^||lhhry@)U`RL~kU9bOE?IKs9RAGc zZ0OXUs{~~GzVMak^?GznLYcCQq1tG2aH}+{R%mEA0O^Vr^G`(e21bFB|5l}1G#jlt zi?P#%tB)?<$~5L@4S;i;8nd@$PjrP;Qs$4;X?;{s;i=XxsiuW|i3F4bxwiKsTztdw z5NK{HXby}`a^VZCWtzaD3C4&t@kvN8Dsl!|w*_Na(OiUm?>-*rn@pKr5A*0Uk_8Q_ ztNc`%Qn<&8eOM{P2~NNg$3Bar-~O>Af&ba1A2buCYz?r03Sl#L<>;&wo*czCCwjBw)mT}#oTa*8J-ncRpy%sm0hi~7P9N+rEN=aNYrgGuP#KhK^>WtlvE6`JCK{pgWGl6SI~piGtk7{5&6$! zzAH-H*S_wyT34PjC6}Ji_b?y2V=ve)WDyp3O0wIkTam~4WZM2 z1ZxBF@y+XW!5K9se#C=4#w)bAjZs#H zAXR8{H<3$$CJ_bZ9`DXV6U6~?lphp8bz){l4gR{RZ<*bmF8`v5pN#-C2pczJZ@1f@ zGTgMk{}Y8{{2wSBI}`i=o5D@~D(yBT61y*IH*7)dw4)ga7dxcbHC;QluEMmG;@k#u zSN$a<`qN1)5Ukg}yFQqY88S3g0VsqW@x%F`m zld^*3f2*o;`+09x_mB6xr#j;iGo6TLjvhV;X15U!8Mi_*)pr?-o{nu?2J|!fd|n<* zU7z4GjX|7==F^za4|M3@mzEd=PX@vNsj3ohH(62rr>bi8{xYF|O>53P@i#nZjg+p8 znKtAgDP!_J+e=nHw~Sm#2JB_0=_hx(IQg=3Q^veG!T71g>f#uhchqr;uRf`&s`S&C zYhthLQ3pc})us)CxYO+to~w8+H)X1&as|J10rS6BTN^DaO?4uMr-s$0XZJG!aYM2kmtAKO<#H0~CDyGfT2;qVO$UbjgvEqkYtAwpKFq_A`!hY|2ZG-i=nGO@Negc6ef` zw4zKY--wk0xOR6lJEh7IHc3Nzpi#$OP5))ZT-Ij+rOIaATi$ls#DKc5w$Y$;6Ao-u zhKr?%VgKmi#Kh=U3+s@YoEa{0kq_s~(TX>Bs==(1bF4wbPVzTT1AL=KzaWpWY1y@Y zxo&-_12;&=bt0fiEvPo}gyND0P1tL+6n55d{H`FKHtY@uHsYb-GR)~7bZ;-Gq221P zzyOg853|oCTn8~Pl^ND|8_ioH9vSP$hFTvP)-g0yOUYlV=+O*l$Qs>rSw^E%l)7!p zHsnoZPa8jBvlhJ$04}evT=K$OvvT1W+a8v-Pr*Ho@&Uejgr+8`E1fBw;#OOwDP3-m!>Eh=Cv5@qn)R%FGt}xvB`DObrwOFDfi`6 za((wayKB%Xh&G?WS1g3v83LInx)#-4(cy0`luDKpOQQ|}0k5PsY)!mV#+Xi7&2(?~ zEHr@od4*X1LgrM^Q~`f4@2CD}8^E4*AFd+{4yq0@dO5=KcIsWgiv9U-uokp!lPTt| zJo+M$4Fp@ET4z8kaIVwffK{`YZMcC|*4SLs9zcK~RC-QwS7yR-?M4@-l(vv88O+T| z5!nF#v1O*xWcBfk%Do81JaN4h#pQ^Uoe0p@-N(bG%H{98_?gU48Eb zTZMZ`J4&lic88E77n27=hd<7U$t2I-1RDPg_YmqWv6swY6(2O&dV5f54sKThg-y^W`jkdU`xQ$Z4S3NBNHC@A(NqZ=;%#@E^lW z(eV3-#)yX^T@9!C(d2!eeZp&?uuUdoC;ppfh@B_&mZ0HqGu_v<4B-Ud0kSNHI&uqv z)bnbQU+Xm^xg&4krgc%zEmSQBokAHhql9$~&T_%UxE47E0lU5YB3vXM9NZo%dpJ7a z*4Kz|jI}=o^5u!8HQ&a1gbIb&M)?91v8`t)7&&UsozNBs1q>%4g$@7^e%$9Sq;i8_R7S+V$(y$Y^xs0w$X^|qU0|F zZI`n9X!&GLSHA=l0NTLt8}z9!sR~<3%EKR!Ci1olX^kI0Ic8`Ww~~&boo4p#KU_Mm z%unXO*LND^xegUlj@FQ7<63NKsOEv z;NyWE6#wCvg~B0<9>R5__9mi;T36E-=f(ACw0AVX($VmIo5+{3)Zok`69Dx@##xq( z%zPCZm)bW0V}&3oD@f*$E6+lEw8bf8_MdI3(LzbzOTD~#&y&rhfb^O`mAEoydUJ2} zN!H%KFK2-SsL#!r-J_8(no}9Nd||zHf&v>>P}=&|$A*PBxLqNwFMfK8l|z+5j+v7S zG6T!K`*^G38w$M&DK!dH^PK0=Io|ao2}#9i!sVQsHxnz57uspLI{pn7SH3q>9cwCH zHU6-M&!oL9r6;=tfoW!_2b`m#A#|2p^h`$7bhJIP2A`0H(e3Px}b*A z!0tPGxd!DbhENNbB^ExF<_8v>5W2za@)b7AnQ!_1MoUxh(mi-GJ%JbNxU91+AgXMA zhae65oB!Y%N=pE;U^`Rb$Z@VN#gVX%Ks%Tj@@ANAVgrkQU1fUtpBX;20}n0{*3q5B zJsl%wQNluBznyt_={9TL-kb|}Y-t(NL0O12%S;2ifFI05K?Jp*fK%KIjQJ27Eu99d z95ll^H|gG24%#r$8~KZs&`Up68{u2Ll5>i)8)g@olN6;#Bl<>X+7AFISGkOuei_P4 z-YZ5qZiwrmwosNiN0kQg@rj66`j!)gPB zDM~1+exTdVIF$TkNX)MzTEKEuEKZa`5~}(+CMrBUzwi&p+wvfwCVYh_LPtpHD+ck1|cLni=d^G^yi|xxjIz z$(muDk$tPepq^3_!nTc;0yw6?&}Tg6{As8b^mrlEFv9GuVen5`=%&DfmE8AwEyBh; z_ZH}o^ZGz?d<(sUsH^{+RQgS$+PT0C9XY-g5g2Ec4{qk91@QMNDwFDlDa}@|ugTjYE5j;u7!%y7ReExFJ(`>j+)#;uK>H{~?o%5WS8*%xLh& zA8zlO7xd!*l$OZAjtCNTShz*ef6ZL9pZNtJ2C@AOy2S|CFquF1GO_Hav0biG@la7Y zSX7&(oX&i=(ySfMMu4bQ6IMHIBMTK7J!<%!e*fbPm|r$Skk(bP;3N5lwFy5e`j(lR zNipZ5wG>gvI}NNaBoj{S^qX7uKtttSjEtK);ZJ7L7k;aBVR!B zFv1nsJRnql3VftY4+;rx zp+I!mhH|3^i|2(!Vh_f0`uYtOg!ZCUN#JI)i`dT(N$znQ0#x)&DS1HdpR%YPR>T8g z$FLPNlL&DsslE3ehMlpPwgD8xr#+F8eHP5>=0FnBg)5Q(_uA&+&Hd;FA$d^}UAl;o zL}~9=X^l=pK=44+5WD*mJ|ok*)ZT6nT~)&EX>&Xu1)L*8=wfxv{4uq#qmcqJ{lwJ+ zV|J$<=+2V8e1rsupIM?Z)bjq!X(#Md*Z&k)9F}SVwNPwr8N0x)r3*H=-SVSX2!6_>1KVi*HRBZn$Dkp9FzQ2E4B6 zr^7dU1YEweA(-S7~GT7u+yusfpdz_NY-?UX4!f&3UxuJqY;7)7>dqc4(C z6gfFi$&9@edDmk&{mUCh+rgDYFWiv%0tFVKj?jWH+NwZd=VFPiAfU*XSBK#tE&tenf;)aSl8 zZzq08pB1x1{Iscxy#Vh|K}%9q&Ur=C^(~P;W9&z`-w=6*Upm)V<1q8IxR|_qJ&=~( z>6U9p$yg%2^^wz%;QEEM;HVYx>;`=m05<}ZA*x_)-mM8lwyVMo6RUp(0J2r1&my_y2=`=y%iAF`RNW5AYf+Xw>s@es;XO-vV6nUh_k+y-o*j+g_gOo9y;_ z)l#zkugnT2ma^Dqz7miSI!~b!05s%+WPE#T?-nS}CJlN9_wSPU`R|wfFsw472B@Zwr$(CZQFL< z>~lY!-P*Z-VSZS%tv<)-!&8&ac^sllQVl}oZaVm%O~{8PG*s?u5+;AxO$IU8VaU22 z`zR2^KfK#`K~_n{+S3$GuaW%f0s8fm_+g)I)q#NT7Qc2V3?63pna6E>0=&(iZoKMh zhYp@{u2T@9JZ+gl(BYOG%LZ+ARNBJ!#uq1h9SKElSG3YQMK4PhkjJ0zG6AMu2%?kb z03RSPmelJ1MzEOvH^E|MVBq+_3D%~jw8Oes)8^mnMaveCNh}m0ANG!TV~Scs&Zxpv z4!gP^e&sKW3@tEb_jUgcVRJjx3Fk`)(rO}><*r<=S&Y8h0@I$7T8Bq$!n(-{5MlJRMEb(_hQ z?Pw^oqM<9{PFJh+&zNwGD`KkuIB=kpE{C9I{wbsSURk`{0&Yt zKip1FtsJ!c@UeIDPV#_CPEh-OIgaChGGxqi z=E!vsQzxCS)ryDSsZ%su(uEk)W}z#8Z4xDlnB(VyTyM+XBND$D;NahjGfQgpF)2nO z`1fk3nQBX#eR`22!+5_suZO~#509ely|foT9VqS|^UgIp4x0RA#64a99LMIuJ(a3a zT+aUCRZEL|iEyxlJG7cbb>q#p5JgGc1MhrIa7U~zm%cUPa@f)*{;jov`XFUMO15YqyYB1kbkx)|QoSpqcy|mfb)xf9C$m>}cIigbshE^-mM@=-VaXOM_EgA%d zUm|*$X}hSJoK`Kz+}_nB?Yh$0gSt+a+FT-BEi6rsB?BWL0?=1}U3Q!~?+>~wE}VOU z7*6Q>49&=!-`4d_>>*%fQ&dkGSw@=83L(2W-azGl0z0saroiOXp(e9NCnADARm~VQ?K!trk97> zvSQHY1+`kUtcLcAgB|Vs5a4#{sb$5X)VjU-A}ut-W8eagOcUYzo)#(-3kIMD#oSCF z!7T(rAMiZl2sP=DW74$<{+zAxch(yYUkorj;Xh=5c6=+ zb-r;aQs0%SknpR_odJYRM}MT_i9b84h?cEwR%ITLK)6t#PB<<>nIMfm#5=#Tw9sB@ zefxK%86f%s(Fwv{i#pPoei{A=djVz~)LylOqJ=jCsyXnTqUaOjX~MeTW~9Ezw$rxk zLFh&nWw=l<9}L18{`ub!olRICfE+Ru*MjMl0K2zvYNzqG!7)4!dvt203rp*ezL*R> zQmm&^$p~J} zw6vF-R7vMK`?80A`S5WvL8~}!Hu^N-J)}y7SvnkLZ&2y)^yh8#UU6NE4!TqdHHy)%l1I?hVte8BeAP$xsZVlX8Q}8pRkT z9A}kF2TT5be|N%D-%-`_D)D7!ET)FCaFh85Lv26pd_!>lUaU);P4))X8q4V{%&87F z_RIhyMn8gk=`GgVS-U^>dx9BXmo}FeSZk97Dmi7%`6mFE)EF=HP$mcSZa)e|OI&B6 z&^MQwS*c+VwB{R(0Wk&CnYeRlPoe#zi%kUt*}JL^4<(IBCp6l-KfcQaP*M(i1%DbW8qX$8 z9M*X_`KHWW;>x;Agd1a4hF^^xXk5-ceI0_+#t|c6!LRhv3{)fV?TS{v8Ku6n_2Yr5 zQ{MR4+_3BV$=Pd>(rMD9!=Zs zI0vKg?xV_ZH^K5W-k(|AGe-ZM40`^XG<9RT9H=$}EZ24Zuf*sAA6%jgCe%-iL@gGbu|nFgUJ3Fs{ZyP|oZsVyM5o@0S@Yt9Ol z3blPjroX;q#Bn+nvrWqn=SEMw&F~e>E+@Eb1|D+{_SnW$?7_I6;F0X`{A!_hx!hA` z&1=K>l>VevCA@HHq0xs&u)|vbGfS9)Ms&7<~^nnxJX(LqT3VZ(WQlku`pV>vP zXBtCwbB$`U(O6Dy11)SFf6G=f2ilOL_Ck7Kr=Wp!BZ+|-*`s;l{w^2x3sJoH)E3W6XDBi(g&(sCp%HH6AJOC;C-$6 zP+Ip64~&%^_#Ez_f`CAN!G+A{dDsmGGa1^)j!&;e@5 z{^d#!Bir~=vpR~REMuF0J#1P)SFuxTi^;$phZ?51!vh#3;_|CiLzQ)&$Ge>*i;zy} z@)$uQ8i8TODbZ<|(}VKCrA<*K_j=Wx)A97gPcK9b<6q+1-4`4N^VxkJsZn+`ENx(R zw`z2paKIVvZ2kcx*#Tv#-06vB085eAAjGZr7HRSGi1b*oD=%ncaU|JBo8Oue*E>SF zbi6(}QAGKqw(~jH*yJWP0UbL8N%YvHpi;@AZ>bDqlI@nQ#dLf0!S|iHql(x@qG+cF zVk#zaR?l7Z=6LlB$*=wZmusYCw%R)Wx_b*Z3_R>(MXkVKXU0e2E$c;AKMrjlAP_O@jgPvNnp1UCd_pbxs8pGP zPAj)=HIl#ryVhzagjSb7=6%3uc}U5d?!-x_<}0s*E<`u$&Hs9E+=LTAi&jn-EK)5i znZ>iMYi$*ONBVnOqx!wW5@Kg9GIa_h$!AsmJ21!ufG_SuS~+A64!50 z*@aI1s&t2S^Q>sM$Ooyq7T`pmV7<;3%;-;V=bP~RU`QoK<1A8cMBP>{@tO*ycf zg-#XlB|B($s3?hZ#0Qd8H|td>dz0ltv=G<#lUF4>v81(ytzRcRgSr{$&-=XP3>8Ne zt^!NLBr;1awwP6CBkIU*Oe8w5%*M)6Y#d}^pGJ85w)ytAX@jK!o-cch&AktWd?T{s zEZlsYyk6ekSr38z`M)Svaux%|Rr~e%RXY{F$u~!8{t)v0Ix_CRlII+if1>_otpj>W zCW$?EA?0Gars=pz(#GeMI7AH|P(oQ1Q(fsc8cfIZaGT2nYt?Y1FY{|uSSQ5Vn&ylo z&%|(HgaW|e<3I_K9b)EyM3Aw#&jSS&#xmjQwIyCz&Gn-4l&hZrPmNg2B&a#?BOj~; zu$e#@o>GO^yxnWn^|HWgUW;WqYjNc)av;HXseFRx=vtIZgeNwSH!WFCa&0_C(~6;p=2&YX`IX9sn+g$cElAIsLWPPBq-M5d7%^uUIU zU8qxOhtFq2E^~pN&$BqtO&3En89v#;lKJHzx8i|jYQ4vG=Z;qkFF7*rY3KeFMokfiM@(k=lDq4HEgOC(yv(B$%QdVy~GqqkrM9-Ix%fHI;M!5JSPCHl5-!lYbH!Hm4mW4 z=7#3*Uni%>BO8&VOw+XA35Gp%&~LFCEx6C{NmJ({fd|P1?we{Ef|ld`eS~sj>BAU9 z%1SqR8s}FIri}gdkEW9JeeTLAf<3k#13_by-{l6)!BFdn$mO4qYZo^L^2n)u#ndRw zDw=C`cAe`@+05%?T(~xG<3>aN65LJc1jR}>ZS>m~wic;4iK64iZksAhlqG91Ml&GQ zs9=F;#~Dw%#jt2)Dm&o@%sli-s`c?jekm)z!Xq@(B@>#nJIhF+~k9x3Q_qD^Ba_D3H;0uERu_?LS62xhiA%A%xpk zhX~KZ?;5{yI<3Xhaqt4?VK*IdPgnHFeDs*T1_9cu`aR$Jw15W<7^W(n?@DKTUn0+? zlIaL8+!5Z4Dv)L^thfxW7mZF8)~eY z5h+&w<^@t?_S!!*9+IOgx};{s(&n_>RJfZ`H38zHE3W9-vtLM2$p643^SzLwk9-4)`-c+U7)Swtj3m1E}# zh&G^Hm~j?C9$m9^GZM@?2=T1E3JZS!+;~y7;HpI+Ng?$ntAfeVQos3-TI(T(wDmg`%k_6bE46@hWRR&Bk+_<^L*8p#(2v26a0Z}x>+igX9L0-ezt(&F%UN5I`@*cNWkf~YnHHLv zMW25^`nh|ju^&KZn1}Nv%<9dHl&CK{tyd$QG*?-Bh`qAV37#a$McY*1(shc@e{M@JH>Qotb!K8-T9q!+CdqS0mpbWH5UdjudBzrO0`|oOa zHk!36hk_lh(yFK2K_{okSBtm@kh&N+X8XFWzk6h4J6zq~-khEzuM7*UE!lz4Fr+Yd zE*qOlE77m{FL#L zb6;X+zK<{?u(1wwHDlYJyZ??d#Mpr@+bc(^>!u{|jJiXMvH*Mnbw0E+7P&x09 z0WIFxU3>@m62dnjk{2u$z81=>z zT-QNJRy$Xmc#!@HGL?5Q7-2C@L+0WFr)5t1ykrpKobF0 zSc@q5aEm79MpO?qec2e)h6z6?uSL+z;9V`1x)2-rrmq>U(4K=4@iL}Vhld+azZ zY;ax0c~iFQE?=6baAK8z5DRN}n#1j0=sctfa1haM<;emygTZGZ%6Sm`P^U|NHB=*} z1Aw-qm}T&Q$}hG*FVV$I|TF-W8RNZyJay$|n9siBIh>4V+?Q zaX=q8XbND4xDD`|u5i^|`~X<6#gwWGe0wsbf`T0PL1)zPN^GC+BAYfe+eOmbW@i|_ zl8=#*I(GirRtyN1EacF6Wc~u?*&uMh%sgWg=O8c9U8M%C)r8|2A*?1M{f=&KQF< z?UC*H>Pu&-9vX6>ofh!D(bAvG)PX9IUP^yr+2XL`#v4QJNZPRRk#XqB@m>nx?2sg~ zo&m^>=ei_!m;L}C01Qo}|JqAeg-A0cKd?M88PP?L7cs*J0#q$&;#fayF#k}anAf>B z6hGgmNr5b4f%9%%mnnDsGJUmooBka%^wOa5y>FtMe4c)kd3Iqi9z@jn8<+C7hhYS| zVn9%drG3K5@?@_bMR*SdLu+08{7FnJOvQO^(9R< zN2hR67aKk`$|OWiV(Tti(Tf#b)DjY<#^ zMEHl*M=t#)$(3&nzVA1Bo$+DYgGc8wbR12uZ2;?31@ANoItLAC&s#RbHfS%G(1=+N zAftLH2hfFk!jvpN+@-21&@+-)4FUNLT83wMc>*-Z&rU|Qd{X2SgXZN;cCvwPtoF2* z2z_So?#6>hvP<#rNXlW-B$bNHpSh9$Mf!3VPpZz+?f_~EhG+r!r9-r!HTD1sXBe#e z#N^J{?;y>Zk_(F+^rB-^jA$fuV@DV;B60@36~3WweABhYq0XIVbVu=q21$yV`F;nV zHEtIHhXOkRB6HTQ9lsX7`mpMe;+p^t`ftjk+j-w8fHv&aZdD7F0e_AmhLe-xaXFfM zO)SdN4jS#tEqp`46ZQ%=C*q|DqhYmPqgutEIcig2xJiAGFpwC?qkfnu5_ojaM`!3S zRT;(5sR4!#A6?5k_h#y)ivoc`mzk29+FOtZoXi`G`t9Aah`pG7dlg`O3_Ava34&!$ z(unM4#|va70^zt#Qj2L`gx^s`!kTGgLOkSPn3k(UK^BYN! zTl1co)35)?8G0J)ng*p^od~@>-!1O=w}EHKDD#%SW(rSiF{d%wH%;Dua#vSZbg5SH zpRBp9VAtG#OkQWJ_=Hk-u2HQ}qbKqb;#IWC*f>Zp@jIqf1-gSWn-Tq-=U97k1uTui z2o%L0k!%Yaa-{_kfiwE zgW@of6~=q5dVjhQk_l8~HnBbS^Ff@8X!#!s7TC)gA?IN;5g&1+y>2vVC641d#=Lt# zL61`kF;34EjR+N(wM=1s@#1E#~KK3Q*}AX`P^Q{{nLMBVE3 z2oNaTBh6VH|JVjT=2>Jl4E9L>TC3-bj{f7R?6jGRzBhoJKh=?So3ljdHp@p=afl25 z2pu6unx1eYxqnm`P`~s>RJ|*)Y@i?)@FBeuQJO2Db>OaEA%;R%m0Wj>5od))~mbb^HjOFZInYWKfWWQ{q#C&AfUG>e2dc#r1I{5Ro-rkvksz zK96aS^H-%LW|( zl7_5d&Rg6fY+6v@F@wo`!oP+j0}LoH>y#w_{H$uS<^S1bxUt=yObL2mse^#OwI0OwSV^~O#DNKY^UlU`F-h6#k9j6! z_w+$b9lPD$v8n0ry>X^@{DFrH(s@?48jJvoD5MX&W-Apwf1AZ1$Sap*q~ks znSeJlM9I68@sdh{GD+1o;Jf^ z##g!@F~=(@$S?LCGlfUOdMISP-YtE#aA?p!3(_Gl-kRe*ehgzqRwjFf&@7@pU`Es_ z0${5cR6@wFnYA~|kAzT*8}P2Z*v%`a3B=5OSk)uYRcYHPzfZgQ)hm5nCe`YFMGROj zIL1dq#I`}hg3PINoVGFEH?wPXmq?dNNEiajRf?o|UVZR67Ii#dL>|IpRy4j|y<2}3 z+f*R|KfaQg0OIczmJUn-BQYwzbZi69;%6M$u*c z3=l3&D?l4D!gUuVLpRvpRaS$X%!SvE2%BL{#GZewyUehZq`ysS7XZ8RI@KHi2lM!J zFU%PxIQ*S8uLDtQU!D9^;CM`PakUu6XyHFx5nia4-LPv(xnbu8sz(w?CSB@w!esEy z!1KIw?+}QDo3l2fPgO0IYkT0#BxYbh-E|BWehJST$aPLkJGsO==C>LHD_j6f$y?<@G@H4d*eCW&99 zL4{BvRj$=Cy#TRs%BnGbEQL?&SsL-G zef*t$5$@C&nP0Y5LZ)h8tEFUl3L}nM&u)`$+K;Na#pHiNyNJ}#wgUnZckAb>wlh|x5(=82E8&cB#uNsTBxwW#VSbD7``t7M=4 zD5);%PS^et2pHnGH*#gNFarDlpwj_^O5(Qcpv9Nr#nht4v7>s-YaKZNT-fA>cozU1 zee#Vy_-3*Jv0wGzYJ)=BlQ7*|wRafsa#dhAWaOIZ(6a*swBMo%7^EjH3|7k}T4Pv4 zkloFg+uX~_)@N(Zl-Di5V- zn_HgqoNK#rqVI2#=2#Ca3-V^#<)H({<{p{th)qDmFE)wsJFEzQxDNDUce_aRJYeM< z>N&-hGy0s=BHWtvePsndPbHUK;1D*7n^$xUhF zTqab$(c{Q2{rGbO+q9UO!pf0FDYp37w#M38#<-l~7aumNByZVtSf?RtBhV@?L?{3Z zxqQGNm~(R%K)`r@L-?ndW*Nq#se+*vddy*!Ej-yt-a2VvD7kCIBJt3D$L#{MrAPFd zkw~K|+>JGh?#XeVnp`Z*OWPpRC%hNKL;m;N97u@0PStE!$o5_P7Hdx!^}^2^jxSl$ zU)%Nta19MP1J0ykL=0D8ylxjY79dkdjx{yj6{I2q2yA6GRlw{jzb3j*!g7YG?asd> zQA@=cl}6HUIvvymIA0S!WwqPY2*r3*#VH?N@sGT<5R}@_b$fnAS{f^sO6)=|847I6 zX6XQg(~jeGL|(4wWz3q2AXQ0p$L%osXEq}8VeqUtx&77H+O&8PgS4MK24`*g7T|z5emNbh6@qwz+E%L=d)wu4ANcs73gd4{H!ps zC@Vhx!3=Ks$L?r$wdh>5*oRF}>XNI8FJhVU^P_O-nU%Nm91b_5(}F>Kw&cK|>@A8; zE`m{y{1rcTwzI>!uc%%JxV_D>l0KWu<5gcRVJ;d8Efq=4o(_H~4X)Got_*W!r$OP9 zF7GBZdb3KCMKE_7JXC*`tg0@ClKH!E!Q4aaGtTv^ny)#e50M|X(f5UI9Vlp&jiY(3@oOTH(N7sh##L$z)v1ZcTQ}nP&tame1nkig`)%BZe9oM z7kE0C9W62T)23UH^)8R*6rR1$8PkCn2GiKtl9kU@=-d|vdk2f7y$5F$FC*|>u{;h_ zYA;u3Cc$_?@Cpd-DN=va$7l9I%E|jow%mHFD85_VpUpVZ)vH7=CaP>XGR>)Qx6{XH zrdyX6>@ze9R&YALd%oCGe^afBy0QN-xv0t|<#r4F5~6X|_YwH2>hwErrpaA-cRJYe zu%-?u-k-W?7P@(Zsd0Uk%6N4AXu>aRQ0|t$WnA!sc>Ay3--PzLX0pSo+&-Iru^U?$ zF<BWlkm^TL$9CYo;>JHfNdRHd#o)D1L*OtnOk=(;3gMVgxG?sef;yN0t6( z+LSO|3N5&`)8w1gE(lG!F$E1qtsa@sk$%;98SJUa8phI}$boLf&TRYzu-!?T}exOB}%|% zsaD~pUkr`Zb*3>z3yziK$}vi?<{~gR<>g9mV8Bd5qGf2;jgf;v7`zuGFp4~0^P&b41l z!NhY(&$pIj?Z1u=v5mpgCuUm)>87>O3n8)nVOFB?T5=bSI|us{gixL=3ZK(#^68dI z(53rSJlKjikxsTM%907?2HL0qcH{#j(nmbc^m1lcw1fVMx!$%EX!AS-G?!Qe3}N&_ zwte0KMZmrh_#e~quOoy!jRJrN4b~2#+oPcf;9`=7r#7G~A6f|ms?qJ$85&GKv z$}qq@`A7#<-~_lVgWul`VOA$d@QNh9Tk7hQ?2Go&lRx{d*X=g69=c7R$z0g79|;T! zxz=^{%8E+c0Fi2DYp=N}kOqJ_wo8M?22e;f5~Nd9(WFYY0y9O1$PO%8cWw5LJZ@K6r>mGBbQ(9yBL!rSi z;5a-l69E56AAJS~5MajZS=XwgC?iWV|s4Zoxj};k@5sN z2{~Wl3@_$qD7#H|s^|kKpT9^Q)JnJ&mTh~$Er0>z_z^5~Lud}*0sPc~Q1{Kb*bwMB zt&JmVOlgGaiz7&i*8_Md`?5`xz2FPxz4{q;{oh+N9 zGHt5PExcfK3*B~VqO3!~B#%dXwY^K%s*M%r)pMLI47IxHK;5&6taBs-2-9~vR)t9o12Bm?Py!eU2 zV9F_DLBVTUX~M~82oRLc*jRlVzdLdjQ*L4F6+OY z8p?p=eW&E*G`8MY#&!WLY9hLU(%0APOl!r5#cz8>^ZYhr zQt|TcoVhg#gCqNEmr?S_e*?Q^?9z$viUH^+7ZI%+h*t!h@bdKpjw2nW*Z#5;Kd9~P z!Ztoe*?Pkopkck?jFfhyQ8u<@y(Z^4bKdF@nTy zWR9tZlmJ8~e%)ZsofQ>9H^LF+>mW?YN=|Rh)En$6vg**XDI=n%`FLd&Eg z5)C~A#TssDx%AB{U$OEnhESHXF3Z1ZL<`5 zdCw8y`l993<_XB{BMw?llR7a&n^NLRv41UoCI9D)+=)&DNfo(}LFtmml8dCmr^f)q zxf8fxXyA!gp^9XmgDZ%M_(Zg}j`)Ps*u&ft5~xY+h6+${sYwek4hA6f=peFlT&*MY zb=UdHq7T}$ogKDJGSdyjwDU1}z;tXr#@Z+m!?FRItH?o*6c~g+1Z-eCmJeWsBf?Zt z@b#;EKk8-+1+=DKXLhlL7|AEZ!P#DsZ@6do@FA-pDdfbfQXT*N2>hnQ? zqOi7<5v^QbyO0C3gY*o|2W!VuUs(*`DFrv0^+SaIB&XBTHB6yEl66e?93hB)?hk?k zWy9%{fUFVM1Q9ZlZ(P*rAY2$YgO&EhOXL-gWyZ&c0$&XEIi|s`U)VdG3ycLY!1AcQ za9M?nK+wS?T68@uLjll%)>d)w9T3;N+C&Z%J{P29a5C(w1bX#WQ5Xl1ABvZRye?KrEG zj%#EKE1TaHeev!7oKQRwto_Qksf!{<#1$v4c-}#bIMNh-zj8mh2cM04d2KJ|CuU9B4%SjC zIa%^6kmsbHXeCeG2bM|;Gv~yilhX%VsH1GjN>SSexfhZ%HNhR1)0qRQGSYL%JRnkH z7K3H0l1&2{|dNOy;{nHMz>8;qLwwu)n|cT*egfx?Np3(MYC zjZ;;Vj}IYhU^H6FKN^Cry&Tu=g$I|8Qcp^6&Qs_d3Cs}$tRViF)11C`N}3HkaAe$g z>mNS`ieE2;=7usog(40Lc?y#mA6h-&x0t-O1R|lY*EdJ51crc7kN4G;1maIF%w1oe z{IS?59VOcTde(Bq zMwBoSI>RKmAia}!8SKjLMH5`NarVGntp7J|r^}^&)k@Ii8 zGP94J>I7c^;@3WK)Xwt^)h}ieJ--vGNO)4ZAjg;U9>1}}#0Ee6ml282xkQ~RzGu@S^$_Rm^iDG{DhUrVI z7d0nkit+`Wcq@Xt4zL~Yn7jVl1*8c6>afHL!PT2UA{@qbD7VSW{!(%;MokQ*ExUxb> zg|Y}CIrAUt)#ntkB$hr6jdbq4|M_qf18PxOhlF~4-udk2X%&kXj8-sAPwpJ%tz?{2 zM?brWO=z~!_txwD4cmiqa1oNn#_uI)04-aH%b@u^o}i-Z)H6|tQT_6_Q%g$8N{Ed6 zW(hsmjWblSHxPi>1b~_Syispz)nJ)!RLI$+BY;Fiu8eE%GlR-zB|Dym>Z&E=IYu#bNH+ zW<}MxJ!96cE!Niuo4UX5=f5ZfMo~Xj?2tvds!|?wWOa)1bhtZKXh}4|U)(xVwTQCb z1SH8vnoiCfA|#s(|4L%a7_z$;ES6I~QT3_h(kdZgr2inyJ(x)Oy$JZhheD<+X@V9~ zA>vC@=7N2MBbaKdPH&+)X;0XUn`k-cf%H9}x=0n|>Etj(ObPqvq=3m(2;Y`Dpd9*g z(wAqmLn@z=C{PGzi0UnX$$OoJhkA*5X3SY-;eiO6nN8(LAhN36hbR(k%L*x@Hfh)> ztWL{i5hjqpQN(jbIe}D6Bg_p7RGU*pGr2`iqeN`DbB;i89KVO*q1p4iYodGS|2(uJ z(fvis)8F9T49(fT{T+zz8ZzCEx5GY0TJ7rX{pwx76sZc?frWh45Fd$MGI1=ai`h*;9k|6J$RHS={m^Guz-4o-7TuIJ)Zq20LIO;bJ6q@L4WK($s@keq7W3vJ$Mc(zDE>@F=B{9RV#zRg&lk>U zj%qbE_Ps+6+ajp>29F6eTIFK|`!~T+dH>FHG|4DjDImuJF305xXy#<6=*E5=p6<}t zJ#}eZo4>%)dE8SLmBxv`&*!B@z()L6D;adN!Q_vZruz~>_~she#3Z2o#(=z9?lsHRN{#xf;WHCwLBAJ0ha*#=GAYgXP*mNUu>VDks}L!n@$XBE);(+qWg#I5cI@V}efZ zv>vTCjZF)zu|S^P*b-vo&0EY{j`GZF(W?|308S)Ju+b8eb9$_JQ3DKV5nl8*^vQ3_ zCi7_YkFM7@`<&Vgr^5}0^V!*gQ|D9Id3DzL@!?9h<_!}k5aRQ}py8%h3~}Mj#!U5L z$4!+H^yKfKwv>Ma@<6=O7{k3|bMb%I%&GW9;Pv|`*x^vb;Kt*PBab31TkPlW#i3+v1H<%E;|K1Jc_XfGaqy2kZa z{*PNuniJ-KU?KT`U}5JEEOgnw4g3!*>?kjH!~Kt2?(<6QKxj5ihVHp*J0N!B1gPNK z9zi7?%+Dedi(lj7RO@|W^l8-ddQTqDZU$0ns6-HNl*eH!KgwIbdt~FpCxQYSe6rQp zMt{lHFL~89Pat#4?&*F7tEqRCWG!dxsI0}6B zJYR=9UlNDr%bUff)Y4%`2Adw#4V(C+h01Irq@hjPVxC!l(0McXrB;pNONGC(J3(MEETn9G{;NqWwp0Q>8`S8J zsL3CE8va;S@HWF?G|bMUHG1x-Aj>}Jt8PnnzV6f9Jl1?X->2)RTIh|U<^A_=2Vsp0 z!{gF0I@B>Ibwh+m69e5@bt6Yq5{+*Xh&s#8N-1a>wDJ(~0zMh?+$%Bbx3f<2`J;Es zwEkE1I7CM*DQw-1@~1cd(=g1R)Tv2%Mj5awI#%F(*+ zV(V4|dbWk1@hJteMq5GeM0e3{Y)243bXtvx&W_4)_Qa2o{OjGdpkK~CzW z5bp(e2hNP`inFp-?H{|Mu!|NL$-lxjDYy3gLDfZqLrZ&pHZ$9ub|>YBNy{i|=dpPD zZY&CkHkhg0NV^K6M>Q~OG=@JQOLRl0WzRQ;vlewk^uXVKy?B!!9XkIIxOep{+@Sto zI?|=u6}xW@Mv6G)y1=RNN@%GKD}LX*S3uDS(iuh$<*}wmu@0~FszXiRbh~nwC$ZOf zgIY8&0`S_zxUH=JbnJq5=!GX;A#z!tG1SB_OTGIN5in*&r4`hdF~= z`r=?(*)^Mo0d(dt(+0VnJ?%5zDBsuyT8QNRzA~N6Pb;rE95?OCyLSfSB9tJ@v!EbX z-)Cz(4UQMxHCP5QJ;}v4`C}9VN@r~7Xx|Y%g{hNtKn&eJH@zW1IYwz3C7s>v?dMpL z@r_0$tcvW>a1{YH%~9qaQLlIr#u-E`P(-x@9A=e-1sypie_~IKYr|-%gn}?Q-?=+N z?jc4x)st8s<1)7u*i+8gZ_y$Ws>dhu)aZggRSeZsKW~hADOjjY2DKG8qhtRb^!&*g zbklf)1;cssdsvq4Ol3>Q=6=yLS^j^Gym@VGFo4f?7^+8ZHvy!Y? zHf1B@lAwWdW`wH#nzp?yI?yn70)?P)(rDi0+fJpX^jK{(P(rHGN|30UVB3Y}9`EA- zq*E7o@e@^-zMv0i0?;8I4BEt|^E%j#*G9cT{xx_(=oJKcd*)*HfDG%;(+rX55w&$2 z(&#`R^VHe%)xAWNs|m*LppS7Nt=^k2E>VvOoN%Q-VCxn^q49KRbC|GMZ{e}dN4!=4 z(69VSymhqhs@ZRbBq)mo2m6TBU7UI{)oDoW7ZhH^2S;*y^u%UjInrBP*1atDfww z(GAu3`yDo}B3VlsgeUP!^}=Dr+KpdoLNO7o0%b%EBxAw)7|*!};c)A+lcHr8DOu}n zCQp(-4d=tnHJ0-$#rKV$)U%7&R`CV^5EL1rVh2%FrbrG(Dr_R4_f}mr7isYWha7PN zaWptfLBk=Ta7$DDPXjE(Ep9{C`cQ> z@!S<_q&rSGt&C!Vs;HI2$@ts7c9N!|R?ut)Hc1lYNkRW(N60D$U<+IHMlO#V4EL+=af~Cg+_8 zIHL&iI{k}b13DveZ+ehHxP%E3}tc>ep!*(Yax=eO76pztxT47z_xR$%Xn8E z+d|8u6on)20J#%LdH<^JlVHl)i8W(924=kHWAs8dCxN`q$Ahw4-6;HsYb0+0N=AlP zU=rhdq6HG&~C*G-;P`fDCQyD~jGuBe%U|CvF82>uV(Ng9^Q%T*92G>BK+KytH4 zs;B&~Y(sWx`kUudCGB6NZsY?;i3SdTyv-Yexs=Dkp4f+2y}M)t)d&jV3Hb5J+TR0J ziMpzjAQ(F68nb|DZr8iK#Y!DRnnDo;5-+abx+QXy0sLBUh18<*yKm!F4+dK4-4p*5 zTfj_UR`@(-s2q`)DPc^vO?01&IyHI;+k8cO>^fr?_kxsAxws}giXQ|~ z)zsrNc0hQk9u3+4AkM`g;&mYu5!pU^_R#w;&m-~P2;g8X-O}p9xcd6ua(4$ry`Mfj zIBN({FK#_C(It+)tMx(4NwG(A8Ht^l5(X(}o&Nj6wddBrz6lBewgA@6nNmt^2&pYt zEIPq?!6>wXtn^LS1TL+t3#~v=l7^(ip$us_nl+?`RD#60=p*Bw?Wgjvm4$6X3L2_g z@CELro;nh`>%oTb4*Bv89tNql)Cr0N3UB!p0ZPntcH%dIhzKMXjDpl?dVV}l3wGTD zSGe5C*1DgRW*2UtzXs|}2tT7;Bt$qAPMZGINu7#p(?_nJCxko0#1q%M8VV_ua-I&n z&d$5~F@FQFxyjV?j-P3%C~S`9G>b$okcEttu!-~uVh^F$R6ooFOoX6381V2rgRINT zAYHztrt@PRt1{X6T(cv@pBN~Yn%5Nl4X%@oT? z&8Ud3(iyLmvxT(By=Bx#4;YL3(U|r0m8pYr)nHJ+$BtZbTDVrO7m#8(6IxyrQ=ki_ zq42SG!nhMYZwyvZxg-6Y(6~DwdTJw^RXB|PIekZ$h#hZqn)MB>=yoXo{J?`~7v}|d z2OV-R$ZBy0WE+&DDf~$5Ag!F0hg=mJZOSk6;&nDo;4r8iOG2n(eR3Aq#C0w!>%QC-+5d{VX+r{M=gnLBEXe9PEr#@@|k zz-+euHXz5P(wU=F?mRSEx%J4TRk5RCuB&4p(g|--2O6(WXENf}m(+SIFZ~PqvB^CG z9j7UFvd_ODSqlkQEizJ~{cNA(_yqzX*ej8@ zML7MnHG6iw)exTnSqGrMqbsSazUf%EuiS@)^;SrAHahavP?;s$=PJa^*};$|d*MGm zysHaFxGoI89Vp|o2f9jLkbSB9^~_*lPp!ejrr@Oh;l%a~3<^PD)Uykk{~<+2WCMC@ znz@Q4vq6~1GS8fQg(z{&??R@N-WR^N)HoJ_5Md}e}|2(gUq7j ztgS6${%P^74V}+#NLEN{LjdFRmfl(dXN=Y}EAU)sT_0UA%F<#G8HY#{*?77ws?b-y zBYID01vAVj&KF3^*|U5VhdH;_U8e~DEGLQ5#OgBQ0`Rh+ltnJHDFMV?>b&UX(6O^WTVW^q1)ycX!xu z1u+oRvXg0o(qZ#=V+@e!Kl2hC4m3u32N}{8Q7{K17>FOgb&VWgP4j+!wKGE>gi?_t znK_}2MSN#{o&-hln-b85?rypUHoUC9=f&92rG)vd&1o7Oq7(vQ4l!IE11wiyoAWJ< zm+i>bUKiCz4ydXl5srt$)xv)N#kBJy2W^HGaDQ3@R7dW(I z?NMcXc~oZ_)h{^1Bj3?QF!KwiWfvMRy=OixtSd{wB`<9Vj{YtUn}I;e)xM3^G#IuQ zLk|`mc}TnlHD^ymUnro}WDb{kFbQWapZ7wo<}A7_kHT||0*lFsg zP=NQ(-di7&-Q=37-iU2rPEDky(%Mj#3?sR3YXWF) zv8$^cdMR&gPY+#S&M)C}{mEEr(p|75<>36$OEyum6L1H!$1#5N)JAVyj9wmZhdVq0 z_<9L?W4`mt5yuiFQRlvH&5-4wlZd_+T;iyL`R1&e)Us(cYZf#@Lhk+2Po)?liC~$+ zw#7i`*0nVa+Hb&6UYQcuHM?vIVpgeSTqsn+M7!SvmHp_`<{i!KrHtB@-ip1p<(O+iNPD4>%oE zk%h!>e{~#-APn2ABiql99erIk|4xpoX784+8L>|?x z3V1hViP3nb$12{yBAMJYtN(szqLS>@shbpNG|aTwf_wqFForxYZSUoO31@e`QP8WY zmBL-KX?k&jSwoSkLMrNm_3>k-Gkipc!h$N0L9j;9x`)PG96D2m>i4182}p=B$bu_V z_2A_9VD~;tcFLpbZ__>Ex+BAl4cjwW-SY{Nus9)wcjR-wIhD$M_$%cdX-4Q70%syZ zd5WJ3W%uG7%kXdFsXl#_WWe*-`#cIo=kN!Hb$){6Jxx$b0nSo6vp727hex4~b2;^m z59cu=k}Ur>oT3z;uiaNV)D2-2G-chJ@@KS0Yz8!^QHQrFG11iueXMq+%xSF-8!vdg z`rF&OKB$3cDLw~N?VTYU?#Ib)H5n+&-Oc9osz zUfVE5KuG=6*%LbC?xak+G1VMU5b6LfBi6}I(ji%aIyd1c-MygXO1=iZ!ze|{n;%q& z`YTJw-PAw!ekUW~9Q`U_&j2(^6j?qP)W6P);_h#h>vvvf`&BkCBvbp!J+DE-Lyz@ z|5*)=LzD$%Lq9vl2pSB;5f6BHSQummU6w3Fgkx(j7tq&IsluUj+{f*? z3mbzJOh+8gp#Kz_+}kXz1~31|{q2KdAn4HuQT37V^NTDm?)EaUwjs}uGb!fpq|>RS zl^~-zE=hlf*{e|%T@WMMN)HD%2A=KjOlQc@WIWbra8o1hPL@7b1ny4M)tT#nE5Ks^ z7|t{*>2rwI9Rk3|O3rV%a9G*zWS9t>L(68d3)g;p8XftJI$2+k#D%RY5;|3iX~or= zM`l;|`Z`%B+@YQvP-W)Xyw*&AjJ1A4go{3*=usdKjPmNhqEu@qe zU;Xj<9@YJq0W7Epb9rR(Zj2l?gbC?1la0OeywAHo%JmiX&7dandV6>BOd=;Ax9=CV ze5;-&A-^r{alv92%+>p0kLGz}bj-3F!rA&CKDa-*jLhTb_4VP+-tmFXy(UCB&Tjm~ zrl*v{>@)^#H0bzxO+Kk2IkAWCnp`<^tie^;-P+Zshh9Lgl^ z_#Q^~w!wh=k4$)bCix$k@Jarx=rzb|$Us7$Nj&B%VT$FL@9KZNC;omiXINsf9;f=` z1t9CAI`yx0Z4_z_7={cE<2;e{IO+2Y_SUrS!qa4Cy`c`@3@$}P^EA7Y?AqzC0tIpG&)KD0;5bkYv7!F_EWT%na46O=F4=F`!&Lx|ubV&F$yHMcoH)6@^W(eDL6&mVe%N+Ll&^r>0N|IS=)glHxsD(^ZE;-4#B1~VTtq~ zSHRU`P&mEhjrgeWnyg7o<6fVTe`N7YCFe?(TYxaLgM`jOr5+UlB zsaqQBom@S;b3d-u{+3t}kezf|jZy3)OqK*s4~8h#^cn(%#&&N^cv*I7Vj`))c7~gD z@Kl`nm3_c`$2-l_`k)J1eO|tJ)nxjNGUuQOr|25axuv0}^i}fazSlrR!=8A_N*!=p z6cX+|HC>fGq5ZeNf=&OVM%BJ?68#O7X9(&q8Dpuq=!fueTT!h#n=K?Mw}`1GpZ~rM zvH=tH>;?tm6Yc`J+4s`_KsXh$dOX>=z}cyzB9=qW zVvW#?f*DfUFmI~Wu;_~3#kCXLSY>`o;-9e7OU}WxZj7Z|db&nH@gb=f|7&pl(N47q zuKspwL@+3q!GH0uE=Y2~b(VZe6BNWGwBfzjD);L428JiN<$$>%DZDtGDeEhh?@%kv z$}RL$?L=QbCtEv+4}*(KZq9Oj#d&K5Mn#aZ*>ofIq>I>Cyx_r3fS{eKy}094T)&g5 z{$lO#ZJ#&c*eKyeGAC~@=hr{iYip*;Uk)eDlU&8qz*ygKJS)t!TJ^R#b_02vS{1M{-T|$y_!i^K;E`<#CDsO zhLG*dA`RPDnj!m?Ex(Nsb%zbcfcnq}XT)b44HeAviGVG#AGvuPEo7re)lV57mb)8( z8aijt)hlyzAQy(gJyBg@jU4dJ@NkTzwb_TutKmTDes_!0@tT&SEt#Hd~%S zVV1tU0>1wD9x2~W;Ul1B)Ueta8*_emHL7WG@mgFBny@AZBdA__5d`Qul1xcgBi~d9 z7mb0Dcu~H+mAD4x#pdt+#e<7h8}uzbJ;v>KsSK#M^dR?>6QOZ{mt0|qP%*_*geqjd+s`_P{vnCvnW|@R{6JaQ10elCxkrJDh#(*rg(XF5g7d3lcuJeZ9NLfX& z1Yl5CQOKs!`iy`3ETA|CMkJ5L#4`hn@$D`gYkET zrH*wh$b;8sIdU1SOARv{v^3@-V%$ht@@GC3SoG0^IfdrCAejdcS&zVPx2;s&eW9K( ziP|F&?VGX?cpNJ5)vhDCirw5+B>FobC1jzXLyBG}_0f&}u8Jq>BS(!?aI!oQFhYsk z;REC-6c>|OT0_DljlgjG;4_Ncd%(&hfJkK`wmK<)_0nYlw9w4(Perr<#jkEw*#ag> z!)(Qur;srCI+)&zKAHW4(e=SQHuGR2aBk(!0&ZSAYWuW=XJLailCn=BkK0K?W`52l zPN+ljKrRaQ!f(3#GL4e;U6(GpHnv_RV6F7*4J6@F&mtT{i0(t*oA0z0xm%&hh4 z0Ykt($brFvxF02Maj{Sq1jhBQc0<3#qv@Gnpn4hB!NQggY(<+Rg38H=+!4gG&;qD` zm$}77vQq`Vg6tbQh0im*)e1F-9KUDA1o!PHK;J#*%)Jy!qA5QPN}?q?SP@A@Qp)y^ z8*5l>7=^^4L!i`2s>470E6}>v-}WgY1cSlC6|6zD#26V#gq9+r!a_n~ZW70}ozo!C ze|`A!EKHb0A8U0c1jHf#1*=;}Kn|Fv*sY<4m>V)TITO9d8vs<{{9?)-#a?VGNrp z{!8wHoNT085!*D|eZj*91v|l>NqN>(x%%PoV@p8ol;*C*isXj4r9Hk4eT3 zBINc1aPPMUo`r8z+N;CfanXVvgvy9}c*X{g>ao#@+N_(`Tj?rsRHJOsPOkQsfo?HIsj60So+>93ovP`4>5 zZut<=A=aUX)O`(q5ie~cs8;!&s53-AI%^*WjY_~4Tq-1E-VD(v)yj{)z#zIy45+l= zZy-%9x%z3tKx%v+5aH8S-?6po@nGgSf6MO19(o|Wl4Et14pA`*VA+_SYDbq4P6>IdP zq_mWVwL%+<2N;1oEtOqwV_?C##2_hy7NZSyT`!Z3X-;Lcl-b)Y&TWkO&OjWiA`LKg z2?`b0J76j*QVdi~BCPsdaY97Rk<^pSoMdMg(=>Dh(R9mtjW8T3C!Dcofw=i8H%hclxg z$M5@ETfLnEY4z+#M8}*#BgEEj>iw-0#*YQ%X03#>y+EDGA=HPWht{wibmO#4&SZb*Nk5BD0RGTLy-=U=<^ZgDxjuE z+xA5rYuN1-7^~Ap2{1KB!@&un zdLxp;6!9YAwW%YCkB^uNI&CeH*hkEy{ZBn8iZ_@gQE-$)8vq9S2k^hTP>d_!GmSZCpSn-$>~ z`U+4=Bv(`(LHn-qs8XW6Lfa0R*Hp5z90pq>EeISI%dORg1{zmi?ymH*H8uh^(%{20 z%e#1PHA9W#P=NTn7dr{YopU!1b+7#Iyx)5yvWkXvP2r;40~wHvcweA929!hO$l5S zUtkJCpPixyliax@`f>T|Q}Y^~3CDy3wb^&o{hs~xh`=DHL#x-&lbp5QcYAHG_33)d zM%Th3{QSu8Jpf^L5K1mx%qTGbhR7M)$OVKFD=L;i($%9`#SRd6$MkqrwrW?seLa2I zxTJhei-(4O|CRWZo#wNQU)KHk(9E`q8y&UX6pYnQuHxckYkhe@7-aONd#hh*l|z#pKtNZtY1(T9ExN&1-0hq@1?pS!l3wwaB+^EkW7P43FAIg20o@Q zSXpmjCBg9p5cddW1>9Wm7GV_U&c@}7;1TwVhh&cm?i69fUjyM z(Y6?KY$(aw$%})L036H!S<%REXL#mp%Mx744IEe%-MBsKe}sc$Zt94hyqT$k7wu#w z4RZKkXDx@F8a3hI;Ht!{l^3=mQfK69$o0OSlyxF4unZshISV!TqZlp?YsFQ9wSP2& z`~5clC7-|X>lwkA5z1~9vkjk%$NybXT`2=R6Ah0v@LOkTZ23?^-`>qDgBWyd4CfUB zkJSkh7?3Ke@WQF*kVmwOC!vJZHn<@Hgt5^1Nv|^BDYRaaoV#RA9KS0miiweM26pFj zp|J1>UUy>%yG@C0?v}6Qk!p*iQAt~;+$=CsNy{ab=E#ieq<|CUK$)yp0ZgRpK~N}G z>6v{5*Vms~Ps~WVl`gJYeyzONuDNP4{ zKC)MPPaY^bDkJMESxK8U^-Ys@N)gBx&wl6BdSS|;n@_{&;q-ZZ%{%$^Q@t#vF@9AG z#01^j2vf8DM$u{kuXaBb!uJ=ai)7ZrPu$wgRb{F`+_^R?9I8xWijf2nqZ=#oD`b@S zjeoA*({}q}tHGj7k+h!F17T9p-LX*UQ-6`SZ%^NT&VPdv&=P6>Z&(t?|AHlPu>D`~ z%`UYkUH^$EZaUR1TU&By`se`eB4nl`6U{aME)s7nl8l=nND9+xJUSLowmwL){9)T> ztkRo`83Th+1+nNPyFN1S2Z^5@3&-UkR{9fjCll-VmiCo9i+DGG*l!1!xE;HYcmI0Q zf7-umRU<(JE7vi9`X00@66-ytqF|}g9&ceKQFae{d+?YfAs;(Z_NP#J{=CTKsE20m z`e9FyHypnCvc*25MXc5+s;!WqiWruk(t-W!@L{xPH?lfjtz={`IkYKbd17*2)XalE z=#XLb60=6U=h<#}djz$n#a*{%1T}Xo%0$^N^<4ggX!1YJxS5&1%cHlXqeTcxT#9K~ zXT+^g{ZVnXCKj4N^kL9ven$EmI`&;Uw%U|0DwUXU^*#<|OZil$MgS!1n{t}MTN>q1A zZ<1oDrJUb4s5LW_lwKCT5i8;pacCOxKox7e4`$hMZ5Gp1AB5){# zA2OddC!QY&s(lhYGrS=$U5m+>br*7cT9UL%5r?Cu5AHQ;wKet=`-LwkB1y_~VlH3M zY8EfoS5CG~IR{m7OC`evu7xDmNid@4e+T;DVb*hIm=PzR+Zp|FJn3tYg137NY&Y`99 z?D?z-vn=XM^7r}<++tDN9;XX@09tHjYH%(ugqk;Nmrm{ZwDVT+;Sp8Bw)>E+=vBCE7@W4*_kO|(b~Lo1G4 zcDmTc7oK2JKjr#ff|cTs+;tIHOt7tPD+;*{wHq14ujA}yrn1HN`1LuIJhM{{Q_fxT zzIZ_CuW0N514X?CCUXpTZe7r#G4eVnXUWa3h>IBjFG;X7pPW%oooNklG2gQ&ZDmw2 zhob5`JK`Ef$aVgKDl}+(IwFih?5l#K21R!|Sw9ttDS8L1#4q+ebgjjRzBRg8)8%Y- z5n|>&W3KnHOq49{TYh#`paGM*%_v8w5QbtA!I{X7D6a-wd?*rq9MQ-;NO>;KK7;un z(wgzii!}4{7K4@)@|V_(Kfv%p=8krQ_&alcTyBtImn3*MeDKS1^rAu+p0fy^FXb_Y zhi?Uys8|ag=ka|CxUz{CeT+y5>nBg^TlN<9(s?#g5T#0~r@3sd&Ak~XQapmhXtbk2 znOytK;6h9h?HYfd8^EDe`MhsFHFLE`{jXY(dlh;*>~>Rvvc(4NEEe-*;fP*DHS;v> zv+$(|Vlt4y8Z>A*;PeL-u9^yY3{Jj0p^3|(J0zttip{Mwl1huEg4y!atp6~A`yXba z!)fetgBLw19C-X_W1L$m=$ZuNWZ)kIg2uC5#<{?HTo62GxpwydGL`p)E1P>LAHHnv!Lm{-FDj59y1_EmNa;Z zD*LfEf^k8fwdcM@Bsp5$kfml1G~;31RF{sPW3kl+scr;H0LU!g?A)?RMpo!HPOa*( zk@r~kP!92492%h}3O}wF5)+ECwIBZE>h_k2O4<3aY(KCk5CvMV-)cfk^>0^EFga=!jog{Cln3{_gDMq{~k{UG7=a zyo-mDQaB^wk()(L%tDSMYx#3<@cjJTTG?-`lDdk0adz$X0cj>=D06Tot!i+2c`bU6 z=Hg6q)4I3LOx5YdLu6Ffv?}WPH<1gd{_Wi7Rws%z_MGMUCQupMdwmHBk<6UfggJ>g zlO9t$dHbGS;{IH?6|(K{qAQj{lEm^FV`Tuv>RgWQP-Z6)NL9AU1c(`-AF9N=&yw;C zqccSsGjBAl!Rmvy`c+wbNUsjy9l-4o|t`39sjXO)pXZSdQ28GMH+*TBo$#}pjGMMvH z+}hU4@4$^ZwHyzB9hjw#v}u;&Z3}E1UC=ZfXfCeY9W)s#EvHc&q~!O2ZyI~t6>T~n z3?`i3<+fC%qj3!>0N}JPvOMr(F(d@b$*ivPsFP+AsW9 z1hqSbS}b(>6jnei)1_Pb6!XsEw?onWy+cZ&E}a@Psc9CL>q|Ti-ftqAS}ePs_lSUB zXbldA>ze2E14F=X5Dr!Yk9njAL(W+i2fLZm&DINc?L}=q*K(%5uvM!WrA51px>k|M zWt=spn27ImP``ZBSmENr0Z>B*W<$S2^Pz7_w9PGhS_+oOfCZ_(2o*w6k;U}y+gXtF z%=4P8w-{Dz+&s#np`JG;lM&H#vI8Nm+wrgfW%*pW&{Jj2k>|dvpY%ifIA*H8`oG|l zjeD}`>Nv--?t|hwTO$(}=D|zWT#ZvMwjxWjQyB>n8d;l~3A=z~gkKsEU^sefC#(LH zamL(mh5pu=+_NwemGD4!jZE&-GdR$khCMi;zkskW(*;sTavcVHzH0=drBUPT$Qp@8 zE?hUQgVgC|rcV+}8{z&sKX2HJ@w~mey^ShZ6hw@3&(CAuCC#-`ZtWH)3+$nw(vTVe zZu7qT7^}S+&0~^B*^0J8^%DGE z8fYRC;XhfYmYJ0nN6W;8szl%n!y}s8ZGS#<1Y_CeTL9`cRCQKPlhPXW>kUYAM2W9U zOR{HsJhuvdj16%vE_}tb+ziZRcGL2bBu*`9nkB;g?Fc4n?^4SSkQY(ufFKZ)iW^Y> zi1C;ojIGI#L-;xo|M#2e!^(PPfPINO1)5CK0)AtA*h#06we|GdhN`~vSbakRy9y5W zSZgl9_XiS_YMSQ2KfR-n$KR@92^mb40;jSeSUFrKt}xer5NAWI0=z@7jx*3mLnnLH zdd1u>SyT~PwXbNtMq&yFUq3^G-H0Hg8y-=mjYB!p&kB}PtGbhn(&RZ+p(YcDQQ(vt zf!n8^hWhkj9FJkFWGSk&rmY_9NtuEDuZX8{>?C@^l;ST(dLL#ABYYxLacj6``No$l zaUaI2TE+!<>lst}Yb7~}1#4B6c_k*)vU8LD?^FJgF4{7g$H71Ln5s^P2be0SuENb zv+OI`51Tu+R-jI2PCaGA*{qH#KjPhV9#<@8);zoQvr>8f;nqLcCR4g>Q?}|79}-$nD|uUd3aoFJi?yWNo_cJtLXAIdbHSMYGxqDQM1>K!8g~3#PJ0DqD9{2rD3fojQsgdok(5+ExiZl#WQ|4%>J>7qt z@sJ{mbNs0>RlSDkPO^0qQn|9dA|zm+RIPSvTkf*i?3#Am5m~L8(V?xwY%=k1SZ&T) zb2i4x;_4>f51R&s0`{vu4EsG8gO0GYkudO^;XTA*n4{(qQ`CUG9ykRDuh`JG zd)QA6zAk*Ahk!;}6r!$b1hs{h=ZrC5OpOw<@}&E6qtAPHEqj(H^0lw@pB_6=-QP|{ z>3BS^b!27zFnEJ5pWD-6+>h;SZr}rEFl}z&=PMdK_amezm_e#g%h+8sO}<+RfsVaK zKcWB$8TxZjP+HXGKklP^SlG&rJYw>!5yt5=#LQA_sW0~AOwCU!Zd`e4K@>p>R89+z zYgy57E@h={WVrV>$feTie!Hjry(-EpU4(7Bj^8x4?>Q{kHzU2a?gFOHUcj~ zr3}5HuUG#hW+W$W1|J(^a5z@$VcKkV zk3d?Ku3@@*`fUfN8K>?l{Yb6R+_BSkp>Od#p(OpVqvookI^PmWS!&ynFCIFmCiv0Z zqRp;dF^N&Bb5FhbUtIzvZQt;hzd8FY7Hjjh%0}eXQFAQvxI223GkTwi7eN=c2V;bBVdgZnAvn*99i$6j~$qqc4=XKboz-RDzk|X0uH^ zTJY$3-z%a9X6{VP7kVvpVv`z~5C0+1_4dehizt(tl|hdp_S09annP4h%OTB7rtjF8 zp9&kq7_G>b?Cq#ajbkeSl6maVP?NYv3)g+|(#{|fnF(rrG~Vhg zm|Hp^IXNN7$VxB#fm1bq1P4@eJt?>h;#P6f=JKkf61NF|P39H;t7pHu)P$YnslKXv za|c*)92(;dvV2F8Fw)A^$H$9p2zd5QA(TRQl%kvz4~b-iD(A+|J-5+?|$=N9kCD46;T zf2e7SIwy~pY<8hl&;DZWmh>ym&bq+>`x@!=}MoP?Tbbt?uvdf<#vH)f<(-?5hW2s=i)1srp1sT?*9tI5bN4peb% zSvPIeL!7}VHP3sZ4LGeAsLa%%g^XC>gNBU$lDFL(FF?M<6G~AFmn2?)+~r2juF$WO zyP+dwRF`dd`*zPdmSWK^IxCp=DUaclD`Krvw1)c;dR&W1idWK{wPR%j{gA)h>Ceey z2UpoAG3F4Hp~KxSdds;%p3k{tPp>#errTuoFl;X1bGt`kG!^Nh#N<5P;~!Cu)*)C1G1B*lPW`OK-xPnu9SBSgw?2t*79vVwc7aNKH_8VIcU%e1 zgHScx9Tzr2_4Wi4MIp&T4&YE%J!EshzFL2ZNimeoF*cAtmgLn8!=7(L zk)Z^ZTd>v+4}|dIKsa+ZDbV6pA>(9d3e60z4YuZv8oNLf%-V8Lmqs{SEec8f86rbg zf;a`snjHnOzLP6Arw!^s)#AnNR1Xg`olAL0ftBuG;5;O%26s7ykF(o{(F%;<>;#)^ z3l;B-kN*Py(d=!Jiz*9xf&@ykErt@bS3#!tP~V}Xc|Qyh#-lP1ZoZTB+&qWU5-rkt zwDqq9G@ln-e;~Yhf_*A0IcA9U`d+Qew{iRiR(@(M4eROI=E}ANn>)OEG>0)h-kgm( zJbHdKPzUHzzdntr@ABx~s~GD$Dpv1H=Hsd~tA2U09$_D`OvJu(516n4uqibXt74Gj z-JRbedC9CFDn(#ISH`s6&o})%zlcKKYFY$$`Aln9uLTu{=MX}B;br-;(9y^soGOC` zshj1KD}wd#&wORHZssXKdoeA57AOXRM`(kXdVu5Foy!==}z58L&z&RXBWMo0Ta-ubQt<~>D%Upc+mh!8mJ zO8aL7Lu`GFMO2M?#ifNr38U(LSA4RNcKv2JRJS(k=gWNup8t{8MScLnfVeG&fwdpA zpKY@lqwNAoE?a*$Mbe^g3p>k;hJ}CUfzErrYkFlc6rPwxCfNrI!Jxm_4|^u_`X8cb z5Dqc1LhcIJ&c;G+d>?@i$9*}1;+2X&o8DB|f4xzoT{f=nUm*H}(%Vj+u#K zC$dK$+>`(O|?F|so@Q7~|{aRTu20vJT>tX*tuod9eA z1_ehGV+$i^J4XN$Gk~8Tz@QG$1~36wehys#gOshQ-A~L2Py#S0+1WV*{xf#}iIpup zO#sa7oQyvyq5y4PP7_lTRt^>}uAjw)kufKefe|~0=}(i9u`w452NN4VFPkYRlM#yv zs|k~-p^*_M2P+#Vi-930tEq`03!90t5x?%w(_x&P9Zd{uVBFU)vn;}84Y2&LJEqvg znVfpB;fU8*G#jsljbT}^){H&?Yl$WC3Rp^GIGVP%L4~10(Iw%Oz8a>)R+Prm7!$7B z$L2Gim+xh6zPvp>$9&t~+wXkGtWMWlx9J~4JsQ;*A6QYp29&A46aE;vJdrCq6skmf zo>eJIs~2gZQz+A?Y^$dw*V4QA5z*^vR?<`H>f+p}=}>Ez|4-N=qQmVdtE<%6snRT^ zBfX{cU+#1-3B8SG-G2#MokBWV%C-N6)VUP&bYJ=Zh3lSzI+-qd|J|0HO+}CHqx#=% z__pNq=uiLCLk>M^8U%sLf33PtLnXOlXNe2TCE8I#CG2)fWF=L<;=00r#b{Ull2%XK z0^3W(Fa59Ex4+st6V`?FRQS4g%;cpyZ=WOu|Fus@h5x#y_@88)C8V_67{&iOZmA}x zSCvutFLbe?q1UzfpBFyMBcsp3t)l<55q#IaN8X2^*|h3cFtyyrp2^vgHp8Vld)V@Jipjyo=2Y&fK^uHH9k6o|VqE z#3l6)C)O-mYsfpz3>W1$5kCxk(KH3ZtPj=H>u|Lt3&`WzX89}J&&Xpo+yaG%h~JgP z9;}SvUETqIuke6)^`_+ye<@b8v2)u`}Gwf#p%`1HhR{3Rs6C#Hg6 zFtwpVK*5sLo}M`}lotNlJW42cMCIVLSKo`wHmPyVxrbu8YVw}TDUv>E4ZpwPGUT+& zWtpxWugZlw`xJF#TBhKUS!6ZNOBB@5%6|(_3$H$^x9Vx(r9PI<8?e*1>vmg{w`J;NFpTVEc+UwNO&^`+m-AUFvWAGjqDeoJU)lo6yq*$T-0yr$K^mF zo;5t*sj}d~LSZA%#CwvhrrZz3+e7%TfZ4f`dUIcH&o83(`Gr6Bi74o@l}1O(7phyz zl=-+3le}Aa68m07i%(rSn?3`TvD?alkj3EpwA(=Uh=Lc4X^7r<%4A4KqQ7|Jec?CF z$N9WVf2SckLuLoGu9k>*0zs(l4O-`AU%D)Im{)a2nAp9M9}sjXBtFdlv2b#7HgI%y YcQi4DVPfWB=HldnAtMu$7l-+O0F_*k6aWAK literal 0 HcmV?d00001

    9=hp=q5HX40ktN9)AbNFYpN1Pqya%7k6+V zc9Wp06KKHQ1^(NaKL=wn&Wv zw_d8&$;7dFgHK)jyOu_z(=PV(%T65LUSZF)59jYWFsH-Ej)uO+@5PUn+TY;^3$bI1 zug<%WJg5!yJ>t9S(#Hh@JPZ2CIl-Iu+tp7e&L-vjXPki63*#12{g44Y#z4(*eV3xi zV_^X>0AaYNgueCW5Z)aq7TwN9$5t$Upt>c{9)b{gG>)ip-KP37;Y566K2-P>lM{#f z^R`|(#N07Zqt)e@h4wr8(I~m=O$J8KpgO@U{9hxz!uiu%L&MvG4YN7L zxU6-mKn!g>EXBs?&xd&5D&j_v3rn4W$z-lZETcR2tQY(xi1VXm4|V<;qw|!;nCh-a zBr(I@x^|{#KAEw6P=04z@jtawE#o$>GY{N|^Z!+oN8u;zY}PDtKj!BAl$aT{Y{4*D zJH9+{FN2?aj+gFqaeZ)-;bGg+n&g{S!nJo$vCZs+_B&_Sha{^xt{1i{I!A-`~!h? zJR^-LC_$V|z6nT}IEhGzKA{0GIzGsbUng!yz$6II(LLFT!7mja?3<<6ytEzG?+j;SLhILVh+)$0UOl`}L{rw0 zZcp6C6KstAr}`*wA|)*c=w!)=G{_NoyaN9vaRMKt5?ZF@6YWy7HBWZz*b%d$soL2o zs=o!Z!9+=L!6y#U8VQo+j+L%v7k_429hx$`;Q)OUYI=)eQVw?zxjH&O-{2V2@T(Xz; z_K-E$+cns5WgL_9Wc`Z@@>!<)E)?(*)x7-T<_9m{TJ)lltp=4rItN`}3p%7NO1axk zW8?Foje)@Op7o~&QiU?ICGR*nl99g#^q75>4=@5Ij*wM&6wK-NZ2&PtW8tp#l>}NF z40J)=r6w0%TpvCRM#vU2J~l-B#2=$O9*?FHhF@*4LRMV#F{%|M^cqNDAqCuPR9aoo z7G4O6*1Wxh*qGWV6e}<*!4HhT$ti7rXfmN4`{~nbH=Z{}fQjh@1M_FFIk?avJG=?x z&9e3KR`>bkPCvIZh7rkdMIVDBawAsu#dc$3Vl5jkrf`c@Jt1OOKjYvivbqF{Jler- z>Z)&6JP)Xj#iJFa$tccjl!u`x=sP8k<7gp&P7``hdLf*M1%?%n&AKg`MJ~&aSJ=XS zwUsJY(>ly`uRCxHRrBGge22Waqmy6v{esOA^ddd^Om5RtG^##|9I__v{>Vf+ejSrF zIAS*j>N??R`XMCw?frXI`4Qi)vw3K8C6UJC%ZVA@lj!@|f1v4HimvJ$4^>W2hW-IYo6~U4n|IPI2AX1h;#W8u$S@FgxqOWEYgArwtRp~x<6{#Ea zz&z}~qY1P&e-q;Kk|Spc1SByG3XqDr{`Tka2P|%CBo42e9;Dbl1JdDBj*;j=%OD9i z3?AX|itBGdVWZWjahO%|iF44X8|>sbEwXJ2Lq?fl`)BnBudARZ1%U)GxRvO(QeTgQ z-7Rj7Z+Tp2ffO-br7_7;OO4+$n`-q*LU;D6UUJmnb`Aex67~EnPZd?~oPZ%N zf^oEx6D_0l<>Tw%5!kt7wG>7--FIf=bbx@(j_dJFB0NZTdA>kKqFgi<^WZ|Y9kPp+ z%Xm@b3EGV9-d+`g zC;liJ4*WDmR0@Y$Cka%nn$~UHNd$Qg5SjF`%lzTgXl+{xXDUJwfZNTpDwiZ*fgyQHR^P38YJ(@2{e{oXuvlT z^OH^l?f3eB7?uTqQF^kT%tVCsr^!yvLcs>rb|#p`($$p_@p&fdr8HIR)ZT`*peHkZ zIwX3hA(HJL(K1?>#z_8i?=#@LLkb@nt*-Ij`*YT||M+cuH7QvGBOq0*44C$7OCOH& zMSQ+nH0i-q1Xj|Pnz=In<5=|XLSW^|0q8QM?;0CNBK7LTx1BnH{xf;TJHKCuZOemr zwuP7)WI-xr!mWs8j_?tap_uz`S$VDs9+qtW-UerKabMW&k0mxv8&6a_drQo7v4Q<2 z2z1iMR!+ZZkUb~7)*$O0oJ78ZY6{kDq+riPHIFguQ*>d)CJ?KQ*I`UY&B6t()DyFj zHL3I%8vtfSqbPt7M*(Fmf;quCx#szzpX{4_u=IFB?o2m$xgB;iY&-5^(j&YVa-}+R zSP7%!*-%8J?6#ni9Mi@~lc+Q1d3#mKVKSl&yr!Ip+I(c13iX(nf0~zgzQ}#p>qI`C zT=5g8hg-1oy+E=0yE)D-yKMCjgZKSqrhAjtvWLl`2}NoOly&hp9URDwYUKQRz`+O7 z$4AV}EX5u^FDh>3zJo*Xu1M?faz0~`a6}NO;tW}sF%|z!h@#e1p^8;IG7Vc zW#~T&X5`0JeI$Z^GhG__fb)E{>t00(U*rssKZXpGWL!2gqv?nFT`lw1~4}1gN$Juyr*Sh?3GS?l>18{ka-GD{(|T zjQ!vpZ2HZ0)Q|W;0{15#O5M`c$J66|cb3xw`%SCh0l0xS3N;K*6fW<&-*+%0$vtuQ z1$(aZD;Tx-G-lJe+!eTl9|IV)DSqP=_}C$5{LkWSgbX8I<@dxI{A=r?qnC(H*b0)j zWCCc{SMy*CMndD=Nz!&~RE0$yxXR)r|CV!{pQ;uU^j=<9Rv8ERIcOr?PFVj&XhsqE z!cv!8Ba%%JGBiv%s7ep_3HW^)H(;_Sou9`dp0i7N8}PMV?W;%0BE++sa4YP{KKZJ+ z;fJk4k?*WTGF+Wgjj}>$bWB^!-$F?>}SEC z9lH=L1ZRUdu7oDe{h(7Xh3T=)$K z0mXd&M|7v?2Z!v(kpFPYZITx;yY`7p zs{E@FAiCVJx9v-|6xs7qXXt)!wR@p}<3dmOhvk*ReJExe^wrXERt7;{)$iienI;8e z6^*u-A{2pq$aLUcbKDI2YIaZf!+W&&Z|esQ!zxok8LfmYqFwgoVu{18yk*3&__?_| z1Vnic!Ozr2^D8QtJB<)J5|S0-&^c8~mH;;+zF;82jMdnhG7^o8I#0;*XU#{Gbd{(@ z*Z<8JT}>W#AGHhwT=%gJs|%{-F1sS?%PlW~k?k|Pj?}YXOU~=x{#e}bXfU#?QwWCQ z$)@2}V?9(AhqOhi>e{YgrR5xrti`x)`KFaBLe&v3u^*wPqgv6L945@M?nJta!7fM0 zyrsGn33tC3G-MmhW{fA)Vo&LPcMoSf24t#P5y?fM@w3ln3K!A8g2$A`HUODh*odAAcb&Tk=E|Zl@ALFvquV;lSEn_17@7KHCEE5&@Jpdi9>7FlLY`h)iOdo`3^EjdU zx@VR6O~fexpX6LNgh35FZUDMS8(RGDW)6~qU>9^z1Sjzv7kIi%UUC6>j5kXK??7#z z2G@_`4+`Y|)~j9tfLecDZBqfII+S>4PL^0mV{KGkFfgsQ5dFeQzpfGwQ=&+gSFdoP z>yCl;xcsm@m%h5;FBn=tmp#rLPoa_g^Ce#_Or;N|z2#dy4x1jlyE{YP@}|XhyDW=) z^07whNC?WMN3z4Q@Y+J&2h`=UI??ndTTZ*43|VE$_9bmPKztVh0{$b@xzhGV7(uDmzQb$-V`!1=(bh$YzjcWTIyV&CE)B1MKxqhxPoC5u2EB`#-o%zWAC7YJb|DHexa!5C}>ChzK z)QpJmC_Kv9gaZ3$bl_UDE7D_s`wBe++s}r09pJaCo>vQm5`G>i^L@`x(^6^>+1yS` z9HVAfpBk`XqQP&R9w7K4taBR>{-1($EdRHn$i~gd{{K>(AnURv<=p?MU}z_K*Q#t^ zMZDZ2!&x1Xy&p?u{U4&&O$ry+2$iJi!od-^l`S{ml5j6zFNC$F8pyRp+WMaWmn#vD zdZna5{})Tm(`yRfAtm=uOhCqE$=A`z*{x*mPX4GM;CZTMY%xZbTG96;j)^ixQF0&F z|G<}`7m4!Uo1Bebvld~~BjAy6PLc9I15DEWDefrCWqjBp=92V<#x>DUUz<8cnwo&6nfaN4(TXVHbx#HzdJ$g`xv<0fR1z=Hh!_tTkSoU-pK%%SpP^6%a9>HasCv&C z%Ljieg}rif+G()4+1)3j=Ugt&HpM(eab87qS2 z*ZaVRq-*xTg#_j*+offL z_2ll!&JP`ITrC9a;gDdR;gNLM(_o>2JBMWb8OhQ}H+&04KJa2IdrRavZmAw0fwAP#|Ezgbapj za_;eOU*i}qeoUpFLjn=oUf65eI~>3QjoYMn)3nz&RI6w4;A$z`0r-2y#wPOUmV9&V zi;F_)2Syg0m;m3X;q|Q4b(i$FcAJHD!F*sfw-7C1lneB9Bi29;u%9sY7{+Dv=|lJ- zUwmu5?aKcD!h(iQLm<84LAfC6gGKHzASD;M5I*G;_ioQlP~`>Yqm94%@JZBU+77vd z)3cjScxN#@;LQRj#$tg*L80Xy%{P~8%q7A>3A@q%EM zg&3X~z*d^rauxbYo?datHg}&*PhB9ID&G(D{mT^|9<*DW-hzss`8MLwfn_ehlC`1|_m)?E6mm1El3m8l=nUvqY31+}ckhylvKQID;bQtDO2THkx z8SgCmxdx<4gpEkpu;KX)dvn6#N{1RubgNA)g~;)1uQlkN>datPx&_PWy(?A1==zkX!{MH0>W zHtK<)(4q`kX_emMi({Sd`N5t(mWXLVqGa}97XiDyd9x)Yw?e?KEd2>bMoS7sjdV~o zL1h{8156|SG&PsbZec(1)y#vb4*8U?Ct1xtcWa8A5@>Cz?@$!K&LuzD{J$$n&pqazYBuaUKCLQAIzB2x2 zow%2ig}9Q?`RLc*Io{f}`TQ}#c#wEZl8L#p!C)Rw6qJV`k*uSWKm$VVf$p7|18D#U z80O-wr=)PHg@1z?$lvvLY6WN_A^DhFAoVu<{$NcJpaEI=qy)jW}QV)|OgMvlyjB!BS*i zsp}U4abhZgcTcHnsnDR@pMnXD1#f7Jf!}rFckKKq{bxmRGc^qXSDT*GWfdZIih^{I z^HzBQd}`wNRNR5QIM~DB8KD46G7OuZf)>|^5Zo>60(ZXB9`5H;{ zBUu*UT$1kuJPE%7$GGmI>@^!;Wz5ddm9Vg!MX1QxE**y6H{R?Cz8LvL2N=AFs3^Gb z1hz5g5ik!~%%cBZDtc>a%uQ{QA)IH>j9C3bO4C;GwMgw7tF)+CVF@w}-o4rv< z|2*Cd^z26N(8_zwh+}*TfBCDki?5=~r)BTsQ_BUR1zbob^~}2k{sGRrOz(%=h$K6I zFS?^JDnTOkF;DJ7PQVw25=e!rD+A?u&O`ML%U;VgxWwItwD#eiqw$S!AF@kl+(2b+XWEa#S6$m z7~mjSo@#lwQlA~6#*$C(yc z*yTXaHan`0)PCfAoZJk01f5wI4F|tO5h0W^AiT?IWS&}1Tn<{PKqM&QRI7X#h}n+) z=cu`bl<^k=%W56L8mQwX6I2pWKZ*uUOiE(dFB2+^dC&>vR#+>15Pd3PAM`zmtj(^{ zp38@b*5i{a>fCi^^d5Wl6Nt(Qf$-}hxPOS;R7Zu)h-9VQo8W>p^)h)@DcJ!e*W}M& zBXcgnTsPr}7BCBDv4`xSY8e;yNAIRV0!z4$-Bbvuy$GY4+-i<(CYO!X8Q(x%jzf2* zCk?fY)+Emy@XbSi)t{Ogn8%XIL?B%A8;`u-V!fC32m7X~4R#7u21=K4IlaxmTqam( z43v)=`?KIuGlIZ$NC_XGOohPdF#MwG2!7Jy3%*GL0NuWJlcxOp^JC~9%7eFzHroCT zpQ}krH+A3yeU-#HaI!;9=2lWiMbTiy4&+!8J(zh!k=W7j@~kkY;A>6?L zuNgJn8m6Nc`JJUt`z6`dSL4DEEHtyyA=Bx=@aE*@PizBkf`zb`j02cbQV)rq%B1kv zq}(Pt8=O9rRE1q5H=Bd?OZ3AaCgyrf+;)FCN&=hu*4y*$nq;6gnB@});C}0AS2cwr z&5%dbzTTDK#^mwlydGm?cEctV2zYbEx4>Qn12`4qj3mrA1r(*^TgrWtL&R$$A}?N% zR={wf+vM|5Sp=mhn-Zl3MY+`?y5IS@N}iLPR2XH?u&48u=aJoD=1EV$&}^yf3=)uD z?u!!x_0*1koG?G!(S!d5J?pp@HR=Nx<}61{h{;bw2&{@m%|a@$Zz-hXT3kF zt(tfWL!sQ>dwR+l(*4f&F!aN5ytvqc2#b#Hse@*a!pv}L z(y<(fm?~3O1tf~}%T?08KB?K@qZ0ixv33e6DqUenIyUfi@3K1S)V?cgBf(h)1B2ux zr;*p#*g>-eDoG$nYS&X+?zVqBjPpD)FJ^*R_9vk;qX8V z)YoJbLL&ITl2}1z78IErPih;Lse)xFu?$O!b_FV%cs2|Q*x8AiPXAWVTAeH6Z4mMy zl|K|LQ=*Jix3zAldE#bdmG1~u+r35(-!k$wtuB>orAO3?iP)j=juD;$ht&B&4f_&? zy2z4c3URH1DMy$z9-$qByqE|QHgS4VQ(RPG=EEbyoQ5B47ypV01`YoMhL~19A(<-U z|2+SxuBspTqqj{B%nH`F-ecMF`XrzchJ?H42ev)>MYV)Lj4eUzA zo10Cgs5k|3)qH!newj$Ox9_G|I_Fi}yh+pv@K!IQ?p1)~+~<2Dc@D|sW&C%46Fz?r z*LY2-srWCDY%NbgTSEo4i%Ns^9f5t_k^*t83K&6NfC1PRUu6{%9hMJS_^&bl>l*;I zo!=Rugf(aLmMa2#JJTy4f_*X;TKSyva0efjj{}-WJSa zL(4>P6iRJ~jCki}m1zF7VF$qt_fLciHD{XYC?Axup5pT_HVf&885EjsO_*VzLA zE1ML^G;k6UY&_)%A8On!>x~sM2QA?E7Z+;Whas}I#2i1|2)HIm=leI%h&gy~m7g&L zt!da&+Wl1^)i5+7Og#t!H!vl$(PUvtP6KC+P66d@0==azn;K z2Ndv;DL2y?*kehNI)lFfuKU~8i1sN1$BL-aMb-8TB)8c*mAke$I|5&~Wo(067= zx2KktmZ;TPB}f>qG6k~CCw;gyN~(xFjTN-=ENU|r&mRt(_`S1WS4=J==dfco)yyn& zf-FBy4=ft&XXUd5+WRkB=c(6T{0s$u*PD2ajt_mDP9rl(3b|3>58}JljOz;^{riH@ zq6S$fS_t_t3KJe(-+@@E3G11{AOJt4iK-FmEAq(IUYfKRR_W2%?C27hLVNoZshnlQ#kQ-o3Z;snO-5BNRATdOltD60bQM zk}b)34`nUXWxPm-gU;I{JePuMC#Y&`2$U-fcl0y@3qC~4xTWfz`&DqOfpAKp@aaEb z1QYfgx;y}PlZIOMp3J#37O@3i}%ABo4lifl4wnbbAQiAOmR z{@q1WJ>eBlO%1p@@>nJ>W!=7Vrc?L z(GCjufed=4H4?O*17Au?{sGi8vc%T>S8&C?VhT6v+v1T|K9+(AWq0pSD|L>5hSJdU zvL+TGh(#}?S!vyC3|)ln?$PF58aP6H1@%}L)A0Ajykk$gsGi)ten=qbckr3G4+S!M zpk%Y=hg1~_nkzKJl!5+Mam6ek4@78_G9<(*zUeY(==>Xx)c`nqF%1~*KfF^H5i0_Z zbL_iDEmLBAF~ zkVGH5X*8*RB;%Ol{SCz^BV`cXHkJ1xfC3P9WS}1#U52WdJs(vKRL-`e+Ll@v6NqX8II_X^N(%_qAtH)D3QMi z6eRbvB!$G6m^sjPPd{`Ep6qcx;>37`2U}0_xt8P?X;-4>g`n+Tn>;8t__0WbzQ=06 z@BIM;T4UdI`tyf4`kczp4KZVVG_?WrQ0odXdY|_q>)y-f1EG%*IZU$G#LA7#9aOUL zzwDokP_qq0)vx=r03bx9!T&J}$ohXMcdBkCu3nDjM2u?xz09ISjB@4Uz{>IBZc4fD&mQE~|@^or1o*6Z80XbDa19KV*CiZ=;{ij^cd8so(GX zqmQ^44FyRORf%g%yH<-vNftGRiwtvz5=gIw_m5f=Tr*iB`L!+;AV1>WMVTM+R8V@9 zMENJ&M0C4}GusB|(H+N?J?_~H7v5Wc=j5b3+gm@5p`r63lewYUAVX6}vubrd{nwT# z)C`LQmsVCm?J@(kIfstaXG)?uSS?ZBjB<)3ON2`Vl~Z_;{2s?C#8U+&-9oA@L7Y%P zw3#qQFo$R^W~*Cyk?Z8ZZzgSTKG$$amY&bAGkM;L!zpcLLv@?lNQvd=s`)wIjWMkTx9<5GLJM@W zVbkCEaA6zSvYAkpgN~%184qh@FB2-pk=j8)R-k}JEtS;k314x#LSr%Gu2xRA?b@2L zirdj~KIV$U=*^AGF|FG@O~2Sd{;;tJkt`o zICPRj_0A%u9qgl%s(BL^!EQ3|_LNW)v{!Pzj~)lI-$}!~6pNFQ#nJr-8T((a{HgG8 z0-BsI$>FS)5y8?B2`W?Zi@_Kkuv*4JZ7`x$G}FNLdvIE&C(O11*vZbRsU+62-6`Ic zkeV$eU3TX`W5=>K9Y@)Q9qS&@p$&T?V+C5wQ<^hZx3?s~u|1y+=|bv!vTmA%2LyjX zTh!+8uW=3dOZoK<=z%TRJuuTumjjw4JJU!>@FcTJ3tV6q8NE{IG#-ty2O@78Ia`|j z(y(ONp{Vf}ZcO!3w+|Dloq*x*wT{$|Qle-L{rMIJ*APVqOW)|#U0XBSs(9_jDRR^s zYL-?A?mM>}5>i%;v>h}Z7bs;HhnEO_`9lo&4$v2vIVA;Vq`~j><6%6qiWpSyvGeT3jYr1Sz6gG360Cyh1td+zpJ#6IQ+^sab zsNeZ-p;>d4G@Nuk;51+X9%yYRk2FTDYBl)Y!Tp0X69-mzJ|Rd(#rr}T)hQPbZo-&M zA=E?S5AwwOX%k(!k9zUaqbvCtj%{lddll3s<3jHaZy&vgpYgeL`|zcgO|C%c2~ z{}`J8-;De(K>hzz7U=V%LCx&%fl^IR8kYYXtBLX&lNo^B{ICgYTU|xv?<OW!%iLt}lGWKJY&v8;!z({9Q8 zhX{QN_oNl9-n2Uw)8A}4x-(hj3b_lIBFj58Nk%zYV=0)FzjOji^)8GC^DGf{pLs8Y z7Fj`Pz35}k+6Jre9`1CRQ^1n)T~wULGIVyME)K2-@^PY*DQZ!dHr&}|T80ygU0k;m z!=EwNg4qu)D9<7niypKUD01u8i;LbIoZ}BRlC1FeA9{EL#TgKvE(ca#s+{URJ<;E# zR+nqh*LuqUBHT;7JPx*UJe?T!Tcdi{bgZ)%z2chLk?D$nKY(X;+L#KjGOnF=`?mgb zgQJ7s^kV*fqtRfG3}KT2CB)=fl!lxS8|`rVQW@%|iw&0heuo25im|a`>YMQPM;raR z>UnY4yeppLVx5UL)kycX@eXlR!p1?bHa&M|{L7T_cdQ&Y=P~2|{x|A5D~;nf)c3?Q z&Y4p%=f0|t(C$TH5dqkgRC@2tLfkQyqL2%iILzx%jugjZfh4 zOjT#H=l#PQiO}|(d#ZFR;A_s^N`>?$I>#BcRD@RM8_KUFwXh3QDD7E(-L-hQ#?stP zpP?ZbSo$zfjC1={nmj*!7RI|;N+1iRAb6ICx<9HrXId<)>7p44b%Py^cgh?=Doy)< znaU3HY+v&1KG!YCbv0_~h7;kwlty4F<;1U%S`{_wrkAjIEG7UMWT{E?=J@n|eMRu+ zmS8h|j>4w#y5+gp;iedT$l;cyO8-ZnoqCR)kM-{6bI^okwg;JrhB6`X4&^%hXB z$0*}}KDj)lK%&Bgp`2Q?JF_?(r-*os^n3%iC!`N;2|=Q!=FSmODK&+;=0j@@BL7~f{jNNP7P2S$MIR) zQZy;eb4-MB(WXjk5g76SgY?qZRBXhDM!bJJE&b}s-?u6aGyH>6iy#1$*Z6LQi0D`$ zJy5$}(W}`55J^o$$eA2_xpC00m*Ln>-Dw4PLLJ#)D6}zGMF&tC%EFC@sASa`RP;M@nNNXLvvF@?!v&kjRU`W z=>q4Fg*b9JYrYCi9NuMxwYX>3( zY$}YP5&b)E=Pa`7ZG?HlV$l$gqC#zg=X$#?FPli^_n(#6_Hwu+`_rg{$bW-l0?$^G zap2vvmk*q|F!ISS+pUXn$!>H%WtXoDk%_R&;z(kpJj4q0gO^>8rLYf%?q&HyUH8Uy z`H!|J@eIqCqrB*PzqNz*Ch3F^_}dxik-{Au17&$=a)c74n>&2h&ym#mOlpPbO}?p{ z^kzxEsY>7xbv*)M0)@c)1`K-l-%9aS`h}k>WdbH-#_tbDw6DGc*28(~{vE)i#eSN4 zX4>3jkcMzjj;b(Odp<7ya%H=noPUuU5W1J#uO!*Y#?yJj8CuNEcjf1pWqdwUjjYH- zVg7OQNIBglIX%x)kmOdEY@PUP15`y!5>XgBH#Ed52SoPeC=v^ulz6=2)XfkD1{x2S z-9Ew;SV%T57!H-3s_-eAGcwalNCf=-DB3)urV{XWE(Y!TkX;vQO~o-9VdsI%zi97(z=>RtE9!@>#DQ)03V^cU1J9TgpLuY%|dE^
    mHOKFk<3vKKy8pp>*e7D|xz-#b;x;Fb)utn}xad}0;~5@Lq&W@C!Vs>s zXiCgGQeIyUQd%zOAjS=uWjUdcb5R3Gl&*CiVgan~@B>*>s)MF%a%KgUEwG<*=BO(E>Tlj=YABMkj+ z`<0q#HZsK-U{?oOlZv@!m;oLr<+I>{Lztfaitdj(637pa_eFsiaG)vt7y*%!Vyl|n>O)lFsY{h%XlHBk1I@yU8M+O$eYPBX48_v7%Um!SPlf0+Xf;Ip7kh;PxDc!_?5TFEH zcqN!Kc7Ksef)C4E;1$M0NWj|isc83p@+ACq`Rx`jDTN#7mGg44oF_2Wh7s}^27Z)+ zYnqOQU70eNQqrG4b=qoh-4u!OVmaF1S#!n?n+ajoMDI?4AVJ=_LemFh!{u+xSPB)_ zB5E$PyX&=s2x4S-ymgfyK1R=hz=g8s-jx{(e(omfH*~x&xE{g^W&EI-2?c`jL9iE; z?5#xALO6_z;A921)E`I&Rrz7~21U&6z?<`IZMTe=SF#mVe1Ei}9|yK!wBhWiKGlO^ z1Zn901`lzb-(zp@%3KW7?&QG5u8rsaLcJH)BgsX~(sUxb*-9=HQnd+>Mof!o@rxErtm{i6L}m?z2bKwEq}}mH*jUaL~jBobptPHQxbJ zu-kY9mn(2B9Tt(w>V88;8)2Nvb0?q|oB;Z(663Bh%ja4#SZGORIDM1q5=W26*7l_1 z58FEoo=mf>x|zDVVFExIDuIUdlawuin3{3P22gS$&0f>hWuoE;o>^&6NPFP?CGyo9 zlMIoGw7=qoEeR{*EELTJgr5Uzd<>yL$yR^4#Orw^=JQv^duhKoRwmZr#{)YA*iZFYNPEsV%0~CusKi}CM#&qXZx$3oQ5x;@%TsKF z=!PyrxtJJiJ`|Ykpjm>4E>nc>vh{o_I8;8m9s)J#EZeY>w%AMr8gg>sa|xzifoxGZ z881{Mbd9+*$UCuvrxAsmGeROl?tvC1$?%mOt%zi_l9j!6D?9lFyAB6j?GpSOuOhCa5hd7)5@zv^ z3nM5&#rq63e`%YX#Q>QX3nPT#JHerp3D{mPUEuqPaSb$3`ax(qLY8mjJxZoW{C$(U zE*DWn?nfGwYpCtewc8qP#>|7z>7~TtBEcf*MFHca6_}!UM;~|9gSof7 z>%tQ$F(gtSkMbR7c6UAPYi@S4W`D$Eeiu3nfm2>%+q0L+^EO!j-Fp#L^e|!36p*v; zgdo;uaD2>DDz$N6_!Qn$22ndYwJ2)JrQ!3DgB@_jnmyd`&-?V>!g1mJswz5dn_u}?5&hhGPZMi3(B{67V2`tg-;IeFwpe z+CO+NAM*}&1PXrGiC*-Q)ScQZL$z@q|6n}%WN;&B-^Oso*h;d;3iPL*(<(^AZ;LmT z*@P#osizArfU_DdxXh-Z(@J6o`>q?2L!V1n7swSaS9xHFh}CaLfMjx`qT-55RoY4; zSF7?SO&7k%rAFiqJi*y%5aCya)af)3&UR1P=EvmW1R{V@;KMPZ)vD2FA=TB=8m^AF zOa3yarI{83*q(egF>B9x%Bs9ZAk8k|4-j3KXf&2xWI1{)Cr8QK{u%0n83&uc54Wdq z#a=l^sH=J^v^+OKSvkD=nQKu}kP;#rY-(s6wjHL|Sav_^&SC=Lr%Qym_J+=&X0Mzg z-nyLfD^Hy?e50FLnNvY^ek_ngfDrHpcN25IONd)NCJ5!!Y7_pK9*jx&=m~1@-h`nlci#0=K8f(k@QrgfUK5|UZ084(x(~!AD zGa6#s=Oe|d&{Np5Q>E)bEf8ggp_`?xC?uXCjn?=o-}l4?sd4WeoW@PT%ZS<-%+mkS zBnGbK&U8SCniq;Zb7#{~IKYwiqm4fmv_(@u`^gBWq<^9p!-SC?{l`qvS8NW-)YQ*H@D! zt;#paVj~fl^DqS)mLuZRa_75=DaU!3fFlV4{+`KY5-3{!J-`Q=jRtMT0(nG6r!>Cx zUhxgdNZgjbL`q;mCZcq?IV(ZeIL81jiMhtfc66xss+gRA%B{n28PM{Zx1f_@~xwF{DAV)P&$YH?j7E}pdb_JjS{1I1@0=WA`jNl0O2nk6;gp$$+z(1dIVzpo=#y3yc zmXt-d6JkDEhJ)Gw>d<%Y-(9-du9)^u;>5OP%w#%ZPE#BKztp}|r6D=hKJ6kyY@;&g z)&`LYU^RqW)o9&}US4Q-4v;ZwK_+pKV&a#EP=1hm85WRfOSv1R0T-Dk8zqOuWC3}} z)yh76S)J=Ic~8DGs)N^kpLGeEHd5wA`0>GqT$(5Yb?8C? zpJgI6u)|SC>eK|6G9;o&@Z2Wolyrq7v1N$0tIJTg?Z>|Ye^SurGN^?rapgkk?Whxn zzfk=zJ|rmg-xjr9>|3NUk!5HD=~~fpmFlj|LgZf(-jF@=wSi8xZY&(vAI81yRAssq zk`OALJ|!J?V#i2$5J3dAHXS{rpiK6!vqCzx>ecWHw8d!LfnUbxu{Z~PnVX=3N?~(c zae_`>_O!%Q<1KgB^M;SlW+IQC^c&oz=Yu{rDRMxMc^V{#K~^(vhBsIPs$!wh4G`u_-0jCFW$ zm)(D8;jYMB4dF-37A~(rjM46UD7t?B6@_4#DF1lg{zQs6mi?zZk>x++iS+dB|NG3P zMoTJb<2O&dD!AdivIeo!il!G>>dIei#$7S$qKWAo>N2u|HAgfYB8*@am!$f6{|yoW z0+G9})W;#i`15l$GW2{iG~RuZ%^Z^#^B!NUT~gyv&HC~re7Pk3rd5l`R?QaOG~*(x z?I&Vt`LaiF9g!oYQ6yu!sTggpPCZ>nb$h&?E|=e4-`Sk+gLwaoLK^40`oMHY%sS^f zH)yi*u04v{7wK8;tr{#Ck#*NstA_`3!bDdi88gBHW5!ga5?>TCbFYn>+Q)s>CtL|2!_O;f*2V@I4o#Bs+K1nl_3m1m8!Z zTUZxVzACj{xf%U$2et3&$|UoDIjH$A{mQj5wih2Rn*6d2is|2>&-&PhY@5jCt(erJ zRn1u2L{Oc7Og&s(ZZIUW!BG#U#+ep8P$L6pPSx(Jpl3d^b6-=4-h&%ceb>Y}}e` zzAUOyo|p>*{T-_b+SzSB3Fcixm;%3%1vIb`iKP{nM3=O#jpp^D(w1v$*fu;{@s2As zu3OntWL_r1ZIV1e=vv!0aj+5*pq6Pqcf>J}SE!aFbULRr%B&YA z%2+6NguURVVT&TBng&rFva8n9lL=N;H&q|D8g=Q+<2ZhG;H}fsjhZfw*tk^4!bhs5 z7>(wl3vinJ>q5Ck`%uo+(1RNT;m>o2uj_UtS3(_0pOE$t$Mvq0X4R!#sz?py$HR z04i-k%e?-PWHhpx8Gyy!xZ*|_$Zy%Cu$hPAaJPViu}n!6{{kp<6Mhv0$oi0Ykwj6% zxACC^2O6o(9r62(+TSbDDgsfhfidu*n8GX1^5|jfgA9#5p4(j z&(|qqq^lZ?kW-(t>6*q+F{c+^#{CwYm{3#l6z>1hX_U`TIZEW)2 zzV&;JHzooTUI{VV0z;qwUV;Aol50#S%rtDaOkx^GpnHXCSpb{;!3Nt^-`i7P-Q>Q# zFb`WIl?#_gUx`GU_H_6M=oLQklgD=RET~Ik*p)zhT2-rP@%%uwSZ6_OW-f06B*! z-z}n(qwGec4KowL$6J;BLP0VG*vywV(XcG4xBEN0vuHKAhBo%%L&P==BS;BTyr-2B<0HXa|?4? zL*B_~3PP5>H+NtfW*gP5N>C6ABtEYpR4Z`43}FQhZ1tKjSTNWGqX*CGC~hJAAhfR= zUc&B8cncQs(crgIpR4jFgf`kAaHH0MJD2Zq4v_W^<<@%1MvHLcTz%83t(teSj(6!z zZ6dq$wacJdS+TN?$nine zvRpgCB6gDRq!<3&xOHzfK2K1H_)`|I1`Yd$=;%f&po=uy?*EgMV%_$bplX;nWKfgn zy|fgi44{DLa<9HX0|T#)QTb%S?u;4SaiVrZPszPHoR!`4EyAN57boq^({0dE-XB6T zh->(4a*jU|=BPr~clp7Fx6icrzFytAfhdyOiqhq@tHNwJ-MugWdh`t~z)g7RXZ~+k zeAgIl@;G!CrdpWU4IkwxbnmekRMM!1$oR#IbR=iMiY zpEMH>+(W=M@Ab%jr{p+|^0><7&_gW9)P~n6ObQGuPh}u2r2m54&`?Ffei?0`+4pt$ zyVOq)(v6meriMn*%d^d-9;5Kxp9|1GS5=o$ISbb5n@bN`%BB3|k)vBQF2s)_w9GE< zecj`yRhwy^4)_Qr>BgxI!WE?V??4eRQ-wy^My1hhnR8g(TwA^#rkrY{lh}*_d=>T; z;3;2)USHt`0Q_sml9zyPN;!bjnVaZu0fH17v_I$lO`A=V`HdQjzC>tvw-$E-zT8ZJ zw~w)jt{LyC(3VGr3nd4>)U-~=CQWJ=?kT{@B^C_4CvQ9#e(uhYU+Z0qAOI8#pMm2* z5Y?mb1?F8HiPO2r-+mcTFuRB98a@?bnu$*ehX)d#z`j1spHbMEzS!q8!j6tEt;DlH z*U_~nT*9Oe`$5E;{Pir@!Vqr|&aE#XfseD%(+ht`yKNm^7j@Wpz>+zh)GX;-$2DEt zv$h$S$p`g;gAdxV38T^%k#l~td#QEBmGo2Ac(}El&7fUOvo*U17>Ie~=7y+B>s5%p zqG7Z%EZl?UhP@=z$nB(=jQ~!;2Y?6GIE5#U0Rib9ovVKW)Xzu7pFXci+J9;`CTd5X zvf3)92gvbZ1z_M3L_&t@<5aHkD6SNb$^W3|w3P@=#_&1_p1%e=4}2fq2ET3^@jg&q zST77XG#`d#I4bnXCR!hHijG@<34ChPWqK7Fj|2MtTcR(+hFrNp-bR*w;*XlpvPF^u z)i)EdoFPO#!Vye%v=6S{W=2Jqy3g={4qdEQf3eoE?KIVgpAPf!mPY;oV3q|Ne(+SO zIVki7?{B6}Rfr667AbBC|~_^DTd?_b3i6{xsTYp@wv{PS-|!WXY0nJFaoHK@hPtPA+W%?qTuu2JlyP) z!xt$-5=6%!>=ea0E>yk+FU!;*J!+S#G$9>=>?6r_XjbIOX6SHNG;&!Oy)c4Gni7*! zPme_SRIwZeao0DHWWWI)y)L@nhLx?IP{!?fEJ6C4%+A; zB)eLkVN83~G+;?9G_I=$iE*Aej5p*1dYsM^Dl$wY#|svP?|aONqH6e?WJXzD)Xi+h zr(G?g&=TUBa-nRqC#X5EqUBhbEs(r#+z6&TM@b>aF5N+9Ao}+`it#>S|2<090E`gl z>C3V9usa7&0ar=$Lk)^v9RpVUjP?@D$YCe{12ljPeZn3^r@g#@csu8o_E`L-$0gbk zjR6^WpLh$7)-2UcIqEJu1=t`u!a-uatLN2gF@HN)_maXrme3Q=Zk_$wg_qJT!^H>i z^F8v3`^b<_QQk~Y))y!?QRlC%&QDC`e1K)Lfw=Z$R=czAdaUMQ_p(fiVhK)OxDFJ< z0e5D)+8s)a>~;km+O9O*m#ich)(j|KMun*(UJmpZeyHQ1sf)A^->Gr<)Sn0ajwbnz z)Q|4gec}j2tLl*K@(GR)@x5AA1=$=LL`Z1D(OTuHnkfOCWC+J`6husYS1-qgeqqC0 zAyH%{B=DhaC=Pe1*Cx`*&K@5J~hJS$0Y<<7!&f?_O8s+$QqNGVG*d`bQ<* zkq<})k|CqvKnd#)03>R7qN8tt6JdztF`lBDtg*LRyN$|B=&`S+0+(Uc5O=|qy_gfs zDJ@uD;JZn4Quj#J!RdxFv2j$KFYbv^?Eql~_Dh@5EJH=@fBmReTlhyqr^d1KN6v%9 z#>ZY|SqLTOT!q#8qY0hP52f3x82DGE>lX$+Ijn9fSTandIR;ab&Gh30&oj1V%bpW< zXC0F|7M+OBX&>v>{jGlw18XKwu}!ruGN`U>hW*JchN?uIW$dDqoq&D2|<5 z{H1tYt7YU}^QKXJrG7sj-EZBcAa9RCJ=LYFJhF=f-cyE6c6Xut=AV3piG2>E88%+= z>!ivN`M{|(Llk0XvDUMP3~0|OMa)xo*bh8H>R@P28g!E`LwG!+m<@z6b;&aBlW6e% ziG(f$h#p%zlIE>#U-jpE-7F@{AH{oBPpy&Zwo?ME^eCek3q^p#@*~Z+ zS*IK>bFj4jM5KY1zmK=Sp^z&74l zk=LH&^Ymw+&gR6N@#~ixC{9t%z=TuB({f6z2HlH9>08Qbo9hgL1f_K&M$7eu{7Wg*O&Yb=FhhVq~x?s`aa! ze2l22g3Uf0b@?igyicr#1*1NH1AW)EWmsAzvo7b&s2YHn+&rd`fr3_J$V(kDW9(b? z%A&kLe#5LYdeUi#^BJ8kQq{g*nJ)b??LbEE!*FP+xwf5r$|u7^^b3vEfeU>r`A3t_ z+h7Ay5gbjH)RhdtQYy#m2g6V=DLbd?MO2dKEKTg-l-u$)a=Qq@YP5jH94D_y^}@5VG6vI&geCJCyy z@6sfqRloe}5{yZE>Eey2MEB2gUz~y7bv{;-M)z}eV%Ku(_A^=*wcfAk*kj6+;?I6F zwsvu8%k38c*^H8g)M5#0@bHW9`W$EpZg`^{+TII!ecyZ`$4a;~ARmXmtLU&54RF7~ zHGfukGN}md>}r%}kC~%NR&5xnv~I3Smgj)!aJLXl(U;?!iIBxByPsOO#K4S6$F>i- z*@ljn9An0{FcnX?%!mh6`EPI+vb|HD53ai*HsNlIM3+LO47WqmCyLe;2u%dBqr>i! zt@p4fgB!_hM$Rke!8}Zk3%la3MN$V9h z1-6m&d;)zMBDS_Hj0)fpP<{X8e2XKgedJvM$ZV(;*M%xGiou)?lSbL0ysg7>`s)s^ zhRA*^Kvj;_7{%neB+n4~1QD0@6h}krc|H3)|7R5%Ka<+U8g{7fnnyfTq zZM`rTv|lU&feV}ay!rV3&8A+~^k;uHso}A)>fXFee5+V&1O!*0dA@x41b2o9W7A{Y z5OYEp#TVKStt!~r-t+(;Eb$Icv>$5K&`BDY*oh|IvBY&USGTtx{C0epMsB%wh0x8t zu$Gs52Pb^lNg17NoLL&)Z6aPzkx8P-%t6oO1@Ij=^qed6Aj!(Rau@<%l;v*)&a4Ab zt*KA1cm2P&1;PHmZNbk&lNaHtq4}@8ksVW~nih7t8&jumdD3&4<&|kP-}vI+k>9A= zZw=9cRqjsdM3F-Q#l{0i^w5Kp_l_`UT4de2((}ufHA{*p2XWL*rX^ZwTzc&B>{jSk zxSgx3i~D&5z+@7ZNAR@@wq}C}@+;iFA6gEOvB@>=>}%p~cb6AK#2hfp#<(C~l?a{2 z@yyvLB*Xcl_MLHlep%l%oMCp^-vD8>e|4rEq3TT~_P8tj>k>*L74lUxc}y`L8Ei7F zg)8=vuP#<(bIEe7{)!RWv`i7o0*GKl&62Crt7ANrVI zWxO+LtK1A{b7-kp+J9>OlHUwXSo5Nt5d$98dMh>mXM&w7sZd>aKwJ^2Rd50AUamUg zx#fciL{SHUl0_US|DF(!;!NhmFKcMW3OD9>MWcMh9d&H8j@owAI*?mA5_;W5{L;a$ z(-v<&55zRWxDC1_ch17f+th4+Ch|K*qaIK-DJ)iKb_dh%<>86gk>^*k4Crd_W&ll| zkZ8>jK>&&|^!RR}T)iON_UDdm%WAFY#A>$5OlgtKEa*6y*pF~svra`Hqm+guKk|}i z^w+EeO764@MXV8plV?O#oXmH`3=88Ab02XRcWR(Xn{13YaF4P3FyN3EbhP@AAg0RA zGxaHC2|aSxwcD`tL9!F4n#p{Rlp$B@XfV{MJ;iu~rQ=nD z=I~XCx;fwH?6q7M2cR@>D_>!f>1+{g&XQ<1@4VJ_M$~?U-PYNUerAn*?X~7xogS6J zX3v%EB32T;!jGfB3Lf8qc(8qj?s5kjzel8im2zi z5ftSiY9HzoCpUwv1HF|uipKzLFQ}KE5bLf@+d!F`5MRlgV>Y2BsV2m`Ub!jD=K3Qn zykq;0ckTX@xaqjY2ST?NWJjnCBh!WkGZVx#OD(Jds)5;!>f>!Wy#E|-JOcFq4yXkZ zy1U^7*FZxF3v}3;^E<~SjDkQo^l1P`!r7K$ z`s-M~j1BtT!6ZlJ8)$|~|8n<`_a2Wp)fR=^F4@HWuLc`h?HYqtd2;U*MX$_SH7S9x zgvGs;78h*1Z_f$^T-;8VTQS7Gzv;x&?3ow4Um-X8uXewHEnC*N7VR3IA3&%F-Tyr+VAk% zJ{riE9M$^HY_vfFm9HjPyzgP54lBPLN<0L&h!oR=o5 z+0FET*SF6Cj}jN@z|KWGQjm+^DG_Nw!f)Pp04>sGGx2uj>7>RoR-C`ykvgXEi7VDA zmMs)aTn}0pgj@_pq4bv`-WB>Ry#~@(VlKQAMjHHXZ<7AFn1Z#<7z0k82&7~uUsN=c zR}=r7e!spt%g|9WubetZaD`mF!|m1YS=um{48Tq<4c}=uDv2s3s`{ZTUg7wVC)DbR4(p~6tA2h3KwiC!JGb#BzrAl!J)xM6d9p9&4+ z;1AIz1}Ri(2J`{|2(_C2a71R$Iuq4_Y=kX#kkm&TrSq(3g5?hJ(UhR2CJL^sk3UGo|&)<)7WAtK~*Rw|T4B2~G>fB^P>4joNw*Be<+EE2{V<=iOgpPA$;_? zY+T9*sgwjJp`&WtDzda`EqTK!QVYUflb5;j3KP>>jAel@(n6Nn7WwbO*`Rtf1GciB zvs?tGU7ej(Enu))kr-qiJ$BxpusF4xMoQ9`Q!8Gr0hOcg=amc!b((ShFF;|E7p@;wZ7xspIXzGus^q`}g zOuTpEWzEPByd@o0RK$f^82bI z4+MgQha28%hSMfZmB@f}5bFU7156S2(0*sdHd2Wi(UkNXMZVtoXz~ft)xzG;DAUMu z%&DUi8~%HhWjK8j2=J6@r?;5cwgb#mhI~*2MCf#RW4G4O<9o!Uj~>3We7Z8u^l22V z*UI(FQ*lQ-i@{YU30VfKe)8)x--cch9gH)6y_+YV{#UK`4z-Z!#PT1cl(9uHo^yG2 zPvJPo`v)P;AMc2z47C9HM+aXN5AYS`m&04M%_;WFe#4Gzg%*^ub&LSE;F{=g5+1nj zMJWdr0)Kun9{>rJ#H(RA$;{eRW!MaQtCUmaA3j#AIO|dWn9dwyFzj9?!z}#sO{RTo zJZumBD6U<7Q79!u8S!Hw8h7!fWsp>8lCh30nKiwUbAZk3$*S5RB9V^lR)@nn-(AYO zcHS&38qD$(sZnLs-g5%r);9;-SCA;;d@%L`N{uwZp9PiLEq~m<;3;#x&~QRJ_}Y4D zS?8wHXB`lnC{~n%!Fb{%6D|?n@&z%r%foL~yZzb%9%t>|qoBlLl2+eeFwx!&WO#pw?W5xqBv6F+62 zQIyBsLyFbn_%JIHpEVMXkd!HQn&7Ge>94^!CQWe|?e`6bId#>J{G7s@l-Z}mW|^Dv zu3KpBnkd`VKaV(XKWWD#Qz|i{IK*Ejk+^3Dma$2@)9I9 zP`p-UzWU3+=Yay z-Wdd-qya#|zSz-KNTEuyI`)Q&7$7n`WP4#LaMywLSwXoUmu;0FWvK+qR$EHp&yoNH zZ}w9HWyzWFM$2|{wq^W5<#PSAJ^vmNs+Brnb{}#5MMV7)u`}HEHsyxncU$k$LXP zPw#0I3SOB`OFfQX!U}ZUrJ>FloxPh1oM2u_+=C zZRd+j83>^~04x~j^*HQGA-PTh(4qE~cBv z%}B&<2$3~6x_CqxLk^*1n}zQC1iAs959$T7UsPihn9^V`|A_yKOFO8Y&NB8zO-Bm; z`=m6Td?BKjo@#dCex8 zfnO$B0MZQqY;s9?-d3HIUQZa^v<=*bw4V&Fhy+*;HI7ehlWhLx+DF;^W&7$AqiYHI zd9;@X-ML8vF3@gy5=tW>?kbDQnUDUN&DQ5Avyn);ych4OFLMj0Ge#NLOV+y&*M%fF zuxeE4CDyT!Q(mNU(TeAK|Jd(q4IwT&!MJ(m`9dcACrMd_3_j@dOZ&-Yemk|Q5a)zu zvUVl;=QyCiu*EZfl(Iqht)*}pteGienDqEOTsms$f|3^9@l1mRq=Q^F#kihw0 zo1NIxpjh(WI;TNc`;f1%f|DP__vGR_@%=<5X#wjVum=f&lN?1?m!Kd(U42)T(fwC( z$WtdQ#pPowBlvKPHs2GAdyT%ptx((HK{-M>i?x@=jyFcDqt@*jz;Ts$HUi+<%>sTU z0d@SkYBjSJ;Hk(H)hPV6gaY`R9y<}W+gg9TS3DB?Ay5@$?`f;e4eM-d>~2zGX9ovi zA8`y1Vcsy=q@ujD`B_yZ7kvJM?c)~j6F|)EH?9dfn~B>D;yLw6^R+k95=s|OD@N(b#f z!q^h~bykyxOk1<~n2fsyMkyP{wU0B-%f!HQk3J*$6hjqn5P~ipwCtziJ@IuBHjq=# zMHJe8lIn~5Sn9~9!kV<3i?W6xoH*YL_rka?u!Z^;cU=(C!?R{Ff2vJ+C!X%APSJJ~ zDE2Gezd=nJYMiv0GV?tGCn16tA7{oF8%k9K8zrhsViGpoIJU;Sem>U@_patG+S$F| z&m&QLFW(vtNudvvMNRw8pLRFZ?w`8WNwZxnSDphdUm1pw8J%d8`VG#;C5ILibJ{z} zcT=Rx=T$Q7QR!G@=W-+}d(!#p4B_|uAZq~FzQ_zO%!82s_O?R8I5@7I6NM9wEW4aK znVJEvBe5v&(J_F!(uAn)BvMK5AWDC4Ui>wtcbJsQ`Zi`V7_ux5A7jBvv3n?zyEd0V zcM_ww5|d;&AB31Sme|@BVAel=QzaFn-x4t(cMdC;@6A~9{#PoED)}zJ?^A(R@N{I~ouq6ovTW!G4p4zsqdlejKlOgq86A z|0(fe`wxj1Gwc7Q9IGkiv@y}JZB@6WW6#lTBAEUX?RI~#HqqdNtVNbW$(0f)t}X~G z^6Oj~`5t|P($+Q2&=^yQ_>uaNy^)=mpB~qfBj&*= z@hf#JLQ^9%!)w7EA>=#~RnwDm55W#mFE3vj>rywj5>s;{+uuJJsfv2ITh`N0z^k5b z1RwBFcVpo~pmPsw4Z^6cUrEa}^G2OkeW z(n&6fP8HQXL8uW8I>{Ts8%3&Vc~z$yMe7x8>Swy%nQ-Td|7Kh#leHOTHOBCb-mSk;yUD3+U7f729}ri_ zJh1rBG{k1Bet@P{J9{TrLrILt+*i`ucv%r=Vm<~m=GX<|cX)Zx?7ma| zv)n;M2+!m~#DgI>Lq2|d>nre=&ft!2w7V%GB43;|An-UsKEyd)7X>rP+WgLf!;aye=f zGmGV2Y+2-s2M6}9j0>A8plVQOZI!L+?1@SOKJgR3FlRq?6=P^-7v=!t{0uVUEpI}7 zUJP^^#(wb9Nl7_clP80ziDpA6pwZqk;x0gZrCrup_i>nsh48BF%Je=c9=m$MNTXHV z$rBDQef?Bqn5j<|8$;oQR7*RSRlG%Tl=bmNXztM0nZd2)>#+rw-C{yC!p;IJW8`uh zN;$cnUpTzZpkh*;UcHDUUpnBZO35tfo9ALcvI}GC&QYHLR3G(d{)eI@oGJgTK)s~* zaG5s=xP z7peG1WH$1w$${~ps=RM<{4{D6u!?oq1BB`VY%_X-!p3CPM9!4A0+UT~vLG-J>eU?6 z@hYE6i^A0dodif0nEhAY)D5InB#rB9VI`sA=O&?Q0}%tGA|^M~+R^URg>Nb~_Bcw) zn7S=f^_2m_|71}Uc27IPH<5PWT1CoVpWQxn>(+I-&P!9F@{B(q!tjc^qtVy9@PzE8 zriA(DvS=hg+H=>T+Vm|238+2T9r&-%c%q$xXiFFDv`jhEqBj4Twyhk2%QoV!v?L?J zv5;H#Tn(xM1fNCgfdXVZlnRQ4i}n{sK4Zs^2y_l6KZ3^qLjg7WJAByQ5dZ|7F6aAG zc4os>*J*5&UqRDsD&umB0@~feCPd|k+XB5VTqp`3bVl(2pn`ZoTA$LpR*H42tReYX z-B5JxK0$UB;5^vRVT~ci+W*dvu~Ti63!EO!co$7XBtDk_)P^lvy%1y7t`}f7QsJFX zwlEV!5fj}DJ`qh_oCDiP!VNNxty_NP>NV$Zq;Pu5h^U|E53BBS+f(%UIQOwvNu*{= z(Z~{!N_0_ql6@R+mc_>QWT?p(hJ?Qdiw;Xc4zTsgdEHcyNfZ+auZPPn=zrZmF)&-b zU`GL%9IVwZ(KSvNXQ-g07^(1Q3bH0wbl0ShMsag(tMx6jOnpB^hM@>-Pn~{fY2vV2CP1e-D31w| z3n&;u;=dC~IrhQmjg%z9L2LsZ#4&s-y-$&xsoeq&yck=#T$Ixw4>T^9EAX%N%TKC9E`O&ue_-ShBJuz+ z_W2CXCo8!w*Gd}jXW5ZgK_Zwymw9Z+9%MlTl1d#&-9PkPS!&wJ?|OCBlG>h1tyJ~P z!4@^(^Iw2d#r~1gG!XWZ68pZ1;_7YFoSQ{N4DCHHdQ=BcAd%DQI~b&%aOWf*nE-vg zMwlkDA=%+K;Aes{U*|G*BPy%9HLoDZgvIyRb}csvA!}OFZ_v=Z$@AUrDQKcG{mm18 z&?-6EmCO?P-uO({0CbLC01lgCL!3M>)h)3Ca7nsXrs_m3k-90f>x{H)nx}yF_I4rJ zW_|GJ#a!Fza%B)|oZS^Ry|}DZ_WR;k)VzJ}*Be*D6o>JPpXBmEUI?v}s~l|>;JPDF zVYQ%FVU5r&J13Qi)ZQvSkM-^BggC?5mBMw^L$!eGK0-KXQMcvPWCVG>{#r5IRo`|5 zwHGmIP!xLSJ{siNd_qNv^ypNwP;9^%2*tI?`ZS>Mzs6Mh*ro22u<&Z}ADwrcJ-8j# zWt08vGzag)Cl*b6%sDZtlt&gu(*zfj353rBnUEM4DxdW#kw z>0%(m;FIs-sC`~Q*g0E{?K7&Rz0WsmNrRSE^co-E>ML*scHR>&G{W{0Uh)h&Ho`_D zs-P9*+vQhA8z=JN0J@~Oah?r#h;N;ufGPDs;6NF>tJS=9OujzyhbS}LBhUj^mgo2q zHs@0Q9d&1>Hgrl-xgwGprrn)21|2sBt(FNqGs$oKbVW-5GCeECNJD%iM~s$aO9blA z@}_hyA@RecjU_XEM}>b+q9^NwGQ+nZKKB)1H8CEyw!eL#lGJ73i*3rYT>^85q{=<^AE^arw_ulDUBy zAu2UAvMy4QB1mNmK5>0LUFsx&d$o4GWRor{=csXE?s7ughTGB#(R%unZ1S}6Bp$6! zuP@ZbM7%|!k^=yht5yvFpECcOohY%$+h=FyL9Vi#ewiD2orHxKEF-np3-ayHHx^&p z`%*{lHNWQZlPtKn>cnUq(OjOGHj#<`d>>cG`_JtD$@>N{I~x$AxoxXwE^N`EKFnHA zi?2UPsQGR)8tl2xUqndsL_K5w{G@}1g({#;P3#E|LKl3!dONFY_qq7g!+&5fpEM?r zn28&~YRm~<5qT>=>VA2vVKSh@YiSjTdn~;1UOQ08+ekcE|KNCvCT`ZB{>HLMZAORs8iaTDs^Ktz+5ySM$LyK%AYy*-#v+cv(vMmhft zdIGV@Z{25mMV`tfk{PhiuN4MF!H(h7?hA+Qz1b!fFH2vRD5&94iLY<=_YY!*T{tsBa?OV!kV*4W(5OWT@cPU>bE{Ae`$sR%HK~mKS^bf@kD7kK zHRuBtJ46PqC&noA-DpN4b@mc)zEcK{<9t+=b}N|R%5s&8To!@g{VHOa%;h! zIB9BGp*#L9?it_Uh+7YKhmD*rZOi=pWQ90&ff)#^Cm~2wGf-9z7Nr~MD!6Y&rr%I> zuihpsI&GEtBq+D8!u5NbNg79Jz!yIT*oRvuI3u*P1{><_Bm=i#f`+zd`t3^WmHY39Sc2{w5KQ+%r2F-Vmx6R<*!>yNB*%l01!jIG6j-1`+5n1V!&3!2 z&}YNUUfiM#hH~w>h5u@*FF&lX_(Qos!3L0?K|&RbhYsC?yYTVI(=J&$XurmzE4P~A zc25(a4F&(DbI2zc#5i>GKC@q3Jf45k0C#-$g>Aj~B=z7gG1tW!A;a3ey_bhp(2zj_ zWXkwqU1QHaB`j(5=LZ$&{E%b{=xD@m&2{Tm+eP!f=@-k7^jz|6VW@{m!pwV^2+~P& zZHi#baQ^Ju;WY*7+e5?-tj=d55?zu&w`e=~?ox;GH?kR&!YwNu}l46foJR&cp@%%yQg9jNQym>Bv>XWwGIR&O8TAeT}Or@AcTwiM&xW zgBvGy2Un_bYxO4mX}sEh+Xg=UQPt+JrdgGhOGt)Q!rLJ(IrFkDlWklEoBoMVIP*wu z+F9E($EsDq%>eXq!rcWtNp*0&?CRpT+rUn~^2!|FBUT`p$tNPkV9Nx+T)~RA$4`Cs zN=x;sd!3@mr7Pv+EzW|e417++_VOLFQC@J|4YMfN?rY#uL^1!$U}MYMlM&@o_E1Cz zEVzYUdaOYre7;P*M|Cbh@VD>p62eGPy9^kw-|XFAQ@BXCC)wF)@Gq#r~?&F6-Ry7AH6zU|)aU3vFb!*EeM=zsKVnD8{PM7e!3t%=~@$*q!%p=~|}#yD8Je zv#mh~oICrhn9jUK%_Up51qXhoA2|?9vE}c9yFY9GypV9s-x-UQfYOlSmV;U^7w3h4 zo5K_o0{7I(=<m{lCM}w z+07GDG=2p3_<4h_Y1@NuZF6#N;rj0oXFKO@;%$&M*@=V++cLru%^lgunY3Q^6&Vn> z6U8}LEto)w@lGd884x5rJW{$+jHMVitStkWsj{!rV6qQwjxouek|CYt{Aqn-tN)F( ztEEiO*tjO7N}K;n)~YDfV~e7-^U775C%mKzpSw#%)Xm@o!-%9h#kFQODzRXk9Z>PHzK7+e&Cx*s&;^r9F_AhlWh5{8|joc3i4cPMO?5J;#etpIfsunve^o zFb+$v$WmY0F_p0P!yEE7V{H#yLdgc%zSOlDvX|lJ$KdXdzVBCK@CUJKW(~bM!iq5& z1RT*_pd-HUr$}ZkgWqVGQr$sC^c&rnGcGkGJ&?L<%MaeJ>Wr3FiW~RH8cGCCf@JyV zt@j$!8m?qd~Snf(JlE_d{f0!Sv3CJD9i# z`3J(sK~)aK8RB$4h5Calijn84!E2R$=ozY>QPQ=KEo9)UbO55R-r9_en%BHlM(2O9 zRJ1^!W9KcMDcvP@1E$3^U(fEJ1^a7Ptu7nnz)sG|Rq%vP0bU2a#p)JXz`SHyNkueN z2m-h^O3UjfJU9cy4-+aftP4buGSF_0rBc_x`4sYOe?gCj4gOT-GX zhXkNtZ0=jEt7pn>a_YAB4;l*=4;_vJ;F5rT(j}#ZBtYLQAEcsW-6GPRCgO^bGQNny zlyVUUQ==pthS9KwtreHN79Fw!O(1i5aYX|2cz|}*Yd>hvA7UQKky3X!{`Gp<5S}jZ zYXNlfd`E{DlyZ>c==@eG&SJ&l`t!AiKI@-;-9Jvj!M$mcYvgEuvlg!#U>lF}I+Np^ zm^3pO-G~97SCFq~^>zVvK!>Q~G7s>X zPM>&?A};Tu^#_5%U<;0}zFipIKc=FB0o5dMeMnhB625rs`V8#Eo1q0w4u<^YBaY0Ksp2o z!cI8KyG2i-e5oEdyql!WFr{?^bb3g4-F9%Wkt{SZP7f>au8Z(G>KK>M+0hr z^et!iBxZ;^iDd#UtvgU6JR8eON|_IZ6)j$kMbugH zro&@nTyh$6Rw5g913TMAI`h2Az2Pyn)|Df1L8T4h09rd4`cXSt)MWdJZ0}nQh{kYx zkUFV!6fKBnDCq;@0l0<}?)M;Zlo6p}otBIc;jQx&(wWmygT>ENL%6U9wj_nl^aI?W zIFdO+cxEz^r8Tw$gjf02n3^>#=^?&c z`0zKu4DgoG?QJh{yNp9loQ&g#uUwpX0!v3@=@IT~>%qGFW)0_)i$3FTj`{$t5BF`l z>bhw%OZjkxPqs=fd`PrGBBVivjHz;4(qBII(|6subT@L!$~u!<(8wwD&k5XQQ>pTcZh>6{}3mzC9gfdnCaan|sC7(&!-(?4eynI#-v5u7Z0Qnz4PY7?wFU z-;KTrNZs}tn0l+UkLTNNYPnY2mA~(S5)S_0TQWeTnERcKhX(M;e%E6|5GA^hAu)k% zxK3t*EV=OQYyE@i>4o+#uYe%^MsJhxlAYwY!836Fg z96Fk^5X@?M&z87lE(fQrwa|)j4%*kKnmqg_#C(u;a-i-i9P`)E^OZSu&6MpICY`-> zPrEdhslpW#!H$-s6A^r(#|TfRc(DiXZk!IM@n&@5_?idwn@4z~N~4 zVOhc0^c2&cAJ;P2Z$>lPWN9Yj_ef(j0z1>qx&d|U?5_1Vr%8^7dy)p!DciUMx#dW- zb!Gk$aSP?@HwkBClJTX5pC_=LTW|K~^7QG-XT9vriPthB1^Q8f-yZYT*S+z9tSlrM zI|^+367R^+x*>rjC1ihrl@LYJk0P>8GGa2R3sC0;Oz>G?mZm0JOd`?>7I-Nck!napkXl34|x|>AvNrUr)>)E^o+qMxV)ir5xcdHe40Ck!<8w z=6{8{ETz1)z_>Z3#12?3CobgQ;rY~OV7Cl@6|~`nvz_1=ZQcAeu1O>}>cqT+*KJ-l zP-~X&LI#`dIA=MPjOSbT?2;mDgyD%0#577>)HhCiEP@my&L%ucdfd}RCtv|7s>kS* zO|zgSmeO@lfqY9S{rxbHBDf&c(l>d#RcP<2|AzopYLd74!CtzbacTzV7RUd6F>1d? zhF1xK_IvHmf{LeVcET-2fZ+2PK@WQMv*P9DipDc0Z4u*OJ!VJnX0~u0@f$sxTk(gS z)opwj1TxDs)w|T2f=FFy{%7bPM5^c4C0qtj!ePT1ez{|&tUz53Iq`XBf8DmXUw)2} zp!Ucwu@F7RBtStc=6S}BOq3s|J_t13DabUG7Qc2uBZhY)t!G|3@p<4r%ta#46OF0r zdLVJ~HI#SQfx)uzGl{}6-+%~M#GD(KF3UZ&Gdk0HNz3t@C^x{m%JXnlFrRQr^&mOk_b_=Z1FTL2%2RcQZGybMX_QUf(C4JE1IK~uM z8sDynAsLEMpcb^VJBR_@GG!E;S)dbv&DzA&NU6s`S z?ZvU;rXz1-w=%D*-m&6k+aH<$S?|gqhuj(O#g95blAv%whx}`mb>k;I zcU`|EEj0*emyk!bHGJtgp|)$cvi7~7UpMVovq8P0<&C#hUHdzmE}rMPpt&?Tyg8zo zZ8c|R<{eNXsfi@HZGhIia-ck1;ZTD-W3DQHSK`$IMeT5k`D+7b7PY@sKhVHs>^3VJ z&;+&kZ$&EU6kps`_3%@+IHXK5<3;Z#`1e-f8#jit4COy{Qn>$TuEEa4!SufamR5A9 zJT@h=cWiI3@vh6)B*T<7)jBns{H)2_4z(ubSUebsP)HVtp~1UIFZV=>iU|YFaTIt8x}HDejB4S98k0g#HV(&_OS>00er!>yV&q(2m+?Sp5jLyfQBbW8?)v=vc;Q8oDOoT8aWOCe ztb0bQ@UH~}9^)@N2hkJv7Yk=M=lTRPE3ki@B_&B=O9Bp0y)!%lkc+cguSB8Dj zruWT=tQH%coDMyUR#}*G4~AlPTR!_M?g&9mCC6Qj5Lze7G0E^3RjLSa6*Y`Qo;d%6 zNe9AlQ=$Cg#pp+qzKEi!^wDXxJh)0zgZ*us)^f8AJGS<6>V5GLx1B=ra}KwMYlDLP z_J(>VaCAm2mk!fGjkdn#^&c55`JpBRgyWI(>Zw>Z3>wj@mb)ULH1VvCmOF=Tl(d_D zR4viwH3r%XxX>KBRU>~CipF4G^fBpNXiT-JllR1+0cN9lt3QkedHM~uS9POi8)XuX z%F<&xC|}^2>oyv#l-Aiq-)bvMHqPJ{_(o}C!Pc{9&Bb2=^&94$B;D};dN$s8e{lXG ztu>aRz{$y_aM~tp@IUxe#pfG@wKDFVBaNiL3#OBYZmpzGUYf{m^#F4Y(kHU%XgBSp zKqRxU4W-v8pd4;=M67I=uhKCt;l|u(uYT#Hk`47gRwllPmyYkpcFa|w2=pU2>io67 zIZL8Bt~{>dQU^ZU8b%mK@aU89(+jk~`cvyQwH9V#=e?)k*s#?~CcVMGF6H@!aZ5ZK zCdJj;Xpy{Cf{|C{r%;l4%)(#^x6f#xjDAL*UM!H6f%E!oEw zd$ycAo*{=Dk9cyD#!7_tEwfLAdJQu`oTxs*U@;UiB|hciy(_TP;-#ItmZd`7h6Ql? zl)8KgNsGo)t(VA7TDb?Qu~bjvSHr#yT;v~Y4E%@;RI-zFLxOhZrO~?huE`^1G%DxY z3K02IH-VUk<{g1| z{#f8%M+DF+*S|X+R|O)@ha*W6GxsUQiF%p3D9)-eSRiF`KA2w^Lt;c-uy}An$!8xo zMProFYkfyxTKB#p%^-}ewH({TM@0U;AjdIqgKy!2W1pERH~35Vb}igJob-r!rGVMi z+`Va$R0B`)somVO?trj?`f4z+=DM2iFsa?D(5UzavM<&RVTMZ4Lah*cz1xBe(n73U zabh->^19;0NC4DG^YQ%3P3$-9mWR%c8%T-U(O^6G)E*C}d4Y3!0mpIAMjMuH!1|a3 zx6mrJV|eIESZ4GU{h+Aw~?Pm;fH;9&8okQd4ah6aK`u3fuVNbr9%?mOv`tM zbgvYs35#J_Sh{yZg}!bVJ+Y5Iy%n=U@-oDwp8ds2o6f%>6GI}5f};b5dqMcp7K5TH zfc47y*_yINlP$QT%J_lokSH`$+j-2`d-U9w%h@xy^<% z5ur}kP9~nHXAFXS*^WC(Lbe1l_Dk;Ar-6oL&yNN%^vT_r0?#z+4>?DL6;SwYT}ywh zKu;u>bf3e(>VVw3R4U$MI2>9R5IAs;>oRmU>_!G=V3_3Gxi82G+&P3AJfOGS(#edr zZN?)W-z84BI|+3sq#uoUcp75{Js1wg)M7pu?3^f*Ew1J+X&mdhJ_8AYS~~G*?Hd;) z9J{6;UoPL%e<6nWNhk@r#H!KQ)ExWjI1&674W%8K4Y{jjeL~_@X0(z2cpV0;{mKGD zlmV~U$R-BaG={bs85s>ch{IW&IjErh_vU^y5gr^R31j%R)od2nzrzO-Z?aa7O#9coN5sx$#kZ9BPPorrqg;RDSeho3Bh>x6wJ%(ZVecJX0sEErl{Q4LHhdCko} z+>1BZlz=YsUbI`Xw+#zwUND|-AAaOW(6ig^U`3Hq83qW`C^_*ZcxZ* zWBHd5p$g82(gum`4lVwrJGWa;nm@u@&nckwYIidAH|CN z^saCSSnZB`!IDojT8B9bL?V}9$+cfQs3M)Yw~MzA7;ixI)@YM~u+&%v_{39Od)PYN zD3Am{Q;in~%oY20M}X(g@93bg)>p)0G@1N9th9G^u>&KyE^O2kAsy;J z=>nHr*aSk2Aso&hdQ~@((fL~q#6LzP z4gHB|N>nJ*DZ8!9$ZyGW*NHIoj!c3u70=kyhG7LsMVBUERK2sDs3;ny)JjMc7s#;M z-KS)fDB$bnx~%U#f%(9CHh#np)oJ{5Te zfhslq@d^yTRs%H&g;kYW=H)7|J;kW=9JqG|IsX#HLs&1F0a0woVZSe7<(Wr|z2QD%sA|My3#CkClX_pDk$}$=*!F=$M zAESJD9nSsjzzKQ~1M$f<=k7S!RT?~&qb8lLoC5%_>S(BRqSuL;AdiCSKu9k99^n8} zN#V2hpqWAAV!pqH#^#^;Yc)j*PEGZHh6F=0m;r11de97_WI(?CIXhbhcg9x4)V)P| z4+BUXu|N{hnC|W~Ne~!t;74`_Wf4S2vCBwm(rOkf8nHLLhiy=X3>2xkuwJJAfEc}FY{Im()a7N#fmPIvp4Tx4F)vbhPEH`EmH7~fxQIw%K&H^_)VoMmH3@`M#h zM8X0NZSCxgWC-E^7C`4kLqyL?d(URS54hz~`Aj~BEt8~}K{^dS&FNuj{r6#W7dz+O~%s$pAXB^uM zo!M6*O!rrN-TPo*u(20+_Q9=OSL~JWQ&22iIW(XX0Y>A}GP5CMoyK>8aal0ZFCvk{ z=?>1Q`6JQ4+)mwIp9m@`{?O)WK18*%-KM@Y4o)u|D(G$|^<9d#fOnyJb_z6`0d=b&u_~v2U`^)Chdbf0u*wYHpC6tmC2TNO z&pWD+Op>V@0K$X(1=y`7pco9b^d*-h@rP2xeeZ|zi~FssbBD>E9$98V%hk|CB$DZ84aoA zHJxkaph%;caSE1BK+9_5)}@=^0x5ay6|f9^h|b9}di0Zjl&z@)2nmGq#O;KWmfCjQuFZLd%0!=W zHupN&b-1ckQ}plBu!dLV@CrJ5_iZV-=X(6PqH~6mp9!Y zkG<02A(plch6_|H2Z|{gPDW6?DeI7=6-&L4CqyUOyp^{#XuAX`)k(6?HlpRlR28*V zukpCet9po&28Y_Jt=BUK1GLkd2r@xUDx{O2#jWHPR z&4r0>b9GRz@JF)eMzJ!ex4Bml&)cvgP`c`lTm@cBq+^6)blC75K-8BZbr3K7*wWoU zL39sag9M4Hh6>@m0@5NN7V4`Ud;Uo)K29?X`hGE-MRd+; zkhj9pNRbaWK%!;CiA5~862EGi5Nnvb;nJp1;^< zV7Nghypf9aQEI;neRA$}_uM8^n5AqHc5t_6{VH(0q((MYF1L_GmwnO-3ac)IX&P(CBJ8~5&LXhm8lAxQF;Wd}Wl6OHRiZ6Y| z3dWWrpC#~-`Pb{tu`@2*F6O1S;~?^_g7ET!aXW~(OuTLJ)LQs%f5u~j&H)10PoXzT z@ym}zjH9Ms@ukbIySvQXu&I>x!0XE+sjeazJO@E1r+eZt#(@~QxJGJU=Aad!99L-{ zbLJpjOCHjEhW$lN1u&S*c;`danoc&YX&>l88% zD@^-nm?d5s`n?;C=j5NFDuNrEJa;iad)re4^jGJ*y5}FrC5i9ZeR~z948U`6@ZH7| z@mI85kDN5zblv$a!NPpOUpGm%D&C@^CG26yQvD6Qo#jt=^rH8zT~6sr7W~u;-qkgW zy|B${xXHXo_Hoce4_(DempErw&BIX-=ck5l*1qp9kw&ft6i|qRvUPoEAqA)`w@nFF z`CEuUjmNKrA7IVZLGJ$$2eL5z&p42clj;93QM=Zia@mwh-?qJ7adUIhoUld}=}XMo zc~ftrA=9O|?(!X~I|Gu=6hkMZK}yvzm>!?rowj^z!f4Jsvu8XJKp~PEY_h#N786gE zA~EdnzUe&Ok^5nhLgkhM^uKyG@yS3oml!6HeepyEQcMjtSC<^Z)j-Yk!_YRVm1S@D zqMR&pk^Ywn0`A7E2@xV(qf<0S#|~g3o1<4Irv{}ct!YBbC>e7ob>MmW`%we26aJN* ziaQ7rtVkID<}?{-dOi|K*LpGLCyc8*v*MNAsYCu|Q9f~JKE@#Sau$u9wnGwcrCzOm z8Qzz|$)W~4cF|xuQ=Ny)5#IJys$u(7WN1nT@`WUQ?qSsYTtMw`7>DZQm>DOdQl%D?@P6}sUVnelTu+qwG(IIP) zIo*pzu>`R8oW$DjPHDcbh(gldudoFqWqAo-_jF`uh^yCdNN+Lx?w%T44SE`%>8)9v zm)`$|V8&jgtQg_jT|PGBV!^>EA9-uAPSyl}X4(i^Z`ES!EkjN5L*^aG<`g#oI019m zCh%hR%&;)#!%1q#Fj5Gz*Jz<1+s)3!fiK59buY@rKDA|YvJR!N3nnqWrS7^7*W?txGf`WQ&J7zW>Rp{_fGV*q_z-n7}`;JACfV_|n=v@mb9rt4}S zG@z7`j2-Gr5|v)HP9o1@US#9=0e+#Ly2Bs2G|^47y4)$M1?z9CJCFLq9-}ArPOO8b zQ8B=O^|6myu|=9;E}ugLysg3vAwd(m3gPyP46e?v1+GxhWBvv2uX%ha0N85k*oXak z6lsmA^tx)m74w*kFeA3<1RpXbw=xEMyG?`5aJtD|MZ5M2Kk1ztq$0kUR~`jJ91}DLS?T-{_vtz)0p!l&E*s1y>4!M z7Am&J@IPUGpRRn1r@!FKd4w~-_{|^W`d2*(^-8}gM62b-X!OA^P0{r|I3!kSjX{dK z707AO0SAa15zn0p@AM$Rc3(#`{8 ztcQ(-x+BIl2{OdNTgcrD=xu^ZFBXNd4Zxbr=2EWwD6l1UkRKbUD4{9+FcP(7(w+S@ zPJl0@^Ji>)c2d2QP7vh??v;15=B zjGf!3;1bu6lJPd2FQ3h0a^P9xGz6tYo;B*og8;(!!O5bF5a`2f70V%$8w9p5Nckt)nCW1F`+CQ~MP5egKlhPTea z@vde{(LGu7$>{jPrj0y3tIb-0KF0Y>$c5_lY_%2w*ndIC0R>!>AAXl%J|j1+)L zql)#&Cm$^YN@~0S2CE1&kVidQLHF|uS{R%ejpFyDUk?!=jVFo^+y+fC(PzmDInyv1 zLpX#}>aHZ+1V>%O{-zoQoUN*NLvkY4gq2zY?~0ElfKc87de4oK3OgKYsE}+ z_47(vF?TTilQVN}olS`)r10+)aVnH);nw}?c%Y>e2JgY)4ag;HZu$biw6ac-rny+*#Kj{p@dd1Abiua5 z7DTHf;#sfk@EV!@5%p08l{9e-xMnYXNOX zB8Hc-3N&7JuQ%3660C-@AhZU=AxYgim!?YF7=sOn>^GMTsyoL4V6x!`$P-z6uWo!4 zNiJ;f5nn0W(K9@hdXStH!Mr-p--za#d~nm76q*<&xPQ!YTwq?=W~chAtpJ73Nb&34qQW`g;rVJ{5R=MSKNP#Q|4( z;i^t=176($$>S5mH(O|%4C%KA6|X-Jppx-ZWMLWb(ks~`&QB|8hvJ+Zvbs`_i$?{5ExaXg zQDY@}s;jY=)hb*lB5&P4OB-?|ak0)T%tR$q<6M(d-NNOt9Mb4C4Q^~J4z zSN3G-1NA*kZj#9hfjx&_a>M7FAlD2;5P23rEgvWA$&Gq2Thnqo@d!Wef6ehEMRjYvrh(DSDuEA~nYq>7B zmY_tnXhHr3|6z2yyp#sNnO9PlMyy()+rGCPMr3JzC|jdH$G&rbN|X-sv5P~DxEE8i zlnhB|+JCT&d##M4Y91$8QU}ZBTH$lmthca>V9dDvG$gjW3t)w$^`K6P$`H7Yp)AFI zxyC9l*d|T&JI~>Oi_@Mxl{-pa&V75Y6sfm-R2PfLP_#->IkZBMzw2Ey1eY?cdZp{; z>Ie{?oTorUmAkv5!qEQtO~j&fvz>~|!p1R4giXR)XYF+5|^RN?Ll% z(D;Y((xr{AYJJ-(H{}`tEYs&PLi6)Q=0U!Ws+g$6{FxqBGt)J3|I;)=9jqW8x>hqPkq8w`l zI1WpX$N98J=NIN^lVA|!pRjJ^Eh|Oj;-4On9l=ui{98gbt&jCk82e5?=`uN);Lb?-k|NHt7Pxof=GCnshZ!XG%T-N$x-5Yd}o0s$SvDX&2D#`%Ik=di@`2E|O_vvS~br_??g` zP#(i6`4M83MdxTw{1YeDT#zmRQH5F{UKh>r+l74&Q{KKlTupL_j<_X5j1%Qf1s(WJ zDA$|xZPb%ov%)Nd(fi)V`O=8brs`$+Dk;z;v^~a?V7J6aZp73sh5Sv%aa94kYiUSt z>Tg-{P7Ri%?v%I8=_P6$Zh096(G1`-*5>xKQmuSpt7+3cdx(t}5B3gxLTuf&9@UC8&fmTs$JZ&vz@`EazwkGQ6FW29sHl%R*+Wf&tW zj2w?|h~VY!@N+q(g_0zt2fgm!B)SF|l( zE^K?Br4vyhvl=Smzrb_t*eqQjEQ~$zPm70~NA=~!{$7d$loFh$@)eHxhR@3_`}A%Y zX6=>oN%P~c!zlf#@f692;#9;-k4a*AV*N;5rop(7LtKbNUizs^Cd?Ltl<+4-%_q!f z$(nnRrP_(03*}j8Gp4tdUQaVJOIn;QNNp`}(abS4gWDTyxA5~&*=CN8CM(&oOL@ia z7HSn%_C#~hb4ZOh_20HJfA0=t22E7gBsLoyiNRWam4m{|CLObO@NLpcp>3sm1-23S$a6eWm_4E$cl#hl%`c4rc8q&A*}%L~+*A-T7Kgo`q=rEW z8B1T#Z!B!yliC+5U4rttg&Dy#0aa**Nsff{f^s{%KhhlA5rAAX($RVE9Z)}eHxO)% z3LWR(!YaGrM)q3h`Zk*%W)ypsHl5lj} zYYy&5xJWQjOiC0!g+wjB^e+s7*T|BtiIAU-!T#I{9^l*gZ=pF=a>O0JS~2!!Fi3)# zSJiINJ52X?t7JRW1~!zo*E?w`*Y-q4CW&G%_7lm`-jE4reZLHfZu{WLjFbbI9^>1*d^rg1ur{rqqi? zXy*nHBI+TE+D`Xxg6yNEu4wAeVXC_+?lx7b+yMV-B@Ev1CI|38*Hrmj2OW{v7Ln!J z4tF|C|2Sle<%U5*e#*NW@^D;+I8DKBsk`o=j*ln_{fVp8o=PE`;07*uj|uvp{7$HQ zSWYQw0r5FOVt-Qc{=g}!Ut@MoI5sExob5VD4d__#wo7pum0%lL^D0nW9y?Evi${%^ zwIlB^iuj)`D2%HEfNwt)573|HbOxSkM)rk}uvMMfimjAuF)~YM-iYuAi`_SS$Z_fczu`MEH<>9Y+E@5q!!$IyW=3`1p z76}TbYqm)=04!u(8hCk#M|si}>7S+k1&kDCo`lN-C8W7(c{26VlpKhg8jhFx$oM1J zxj_qw=^MD2@_9-0!`X?$$Gg25Sn@=Ku|Unq<=>yNlNgEwi=Kw#g$Zkaz4vINh}Xyc z`LI8D_lVhZxO$&-E&Pn*2U zHpspCGDqYOAC$cMZuix)PvThK7#ba1$)~rc3vmd{y2;z}mO!Blnc!LyEUAnQ8ins+ z01z)?_o6$NEDOR*V)&<*q3&X4>Yg)m(ti87M!*XTHl~^wFiJ}PGe2}zS-me7Mq@B; zo=D(*Rj&(Hi$8h0c|C6mR?pd+Qjax1Y~g_B-RB_W`8qef@2;F$s~-)gb#Z!^`5V{R z$n6_dHZlstJOG}mzL}2f*(B7?W)_990FIE&kFIeCOI3YU-~L=8h*<0cJSL_7hzF)N zhmB1OfMc%-e53p_6iHVe!t64P6hNvayNXK6Ih;fjU2(iM2jz+v^FFq??T9(+&-aOG zJw8HRmSFA15?U@-u5owqXtxd@<5{z5iWf(3=j85WkM$PafY_!5azJQgwR&%rb*${g zc+5ramB(9YR`hL7aBziH#@Er=-Bb0Ex9yU8my$~BlK)~H{Y0PjPUapSYE5=0?!HWY zH}&kUBktzg5IA~vqanZGI!7CIORS5l9pOK$R0!Mcx zhT^^R;2LwIpPoP|Lx1(fgN=-aysWWiZ{ap;0|I$pEr#YHb~?2Hd}xs?b4Yt78-P3aB~OEv+z|TY+e{cX6n1fsF90$Aczr)OOgdmfG6u2bK=EvBO;nj(`oRb4uZkXoPMCGbqwE;dD7D z5@djLZ$D4EzrnBAoO5kvi$rLlU}{BtwRmlZr8hm>DV=(|0F-?K(L4s5gs@YvF@!FjfY4W^_(fh-U{RmLY#LW zF!}qysUiRR^-IA8S1CO7eAG|5o}u1_)$8`-jl?v)DXI&%=R})7+w?7=m=WPK!)9@) zEFCOE1NBt(Am==n8Qo3A4=+z|w*DNO8JJcfiq5h^ij3Ne+$f9FDG%1>)axaoQ$H2O z&xKhqM?-VU9tuhaq<9x}pb%vj6i7vKqZR1PJvJ!9ncobo^iT~OcnP&PBB%+^xE}{5 z2|{l(69WZQSyL>eHZhZ60|IqPb1oArs&c=VfFgl}Fy%tn=x7)Y9tMMTd5yn<&u;J{v;LnveRvP3Xe97x}1dz!Wy$mP=YUcLDjTEcHrZM+^BDIz=bg zN)$iX4ikYc15?b`vbU)Ba=$nunj4A1=X^aXf3CjHQG;vXq!HsMOK?cckn)gP`dK{| z0CVq< zoi4AlR%|Fs!#|tpQ(rIG#;qFoD9S=p>@+=oOR95Q;LYh%(5%j@2e1ZBz^$Aect zQKUE@Bx<9xM$F8Dh|O0`JBf)3EfYYQ1jwULM0qSZVg%Tce*F}f{QSL-wVjRIxhaJ}F*Z4IoRq%l&k$DKSXP=kh7FsnrvzGl`3WW_*l z;$W!*%EW6oTe|+?3ZiR;JW$XCJ;Gqs{wEUsV=J8DG^#bKcC^ zu_d!|c(lsp8X1H5eGX0GCFyz*88jBiv;)J&s&Dqnl++}I<%nhyRgI2MxCrfEYR%G$ z9$R?7;ylC~s`?G~s=vK`IZ%Tp>pe2>)Fm}N`@>S-XJhqla}YLy8hO2UJ~Jxc;%)J{ z*u9EFOn4{OI}*_Zy1l4JA``k%7I21>;!#=r*1;mjwAzAn12cbFp-UGSH8$XpQqeih zJ4!)2_&8eX9h6DJ3Pon3GCPAYgOhO-ngzfz%jiR-nFym9;C0q2S(=zKO%`y=mj2}Q z69eNUU*dX1KsBCyj3dq44rNa7u3D8p1^?vinAt}tQ(TMkk3*e=4ab6TrE?130G3{q z0%KKX9CLiBmb>>=EB}yX1oZR&oEq)L;haGXe;c+<@B^n)a>%q|;ZDKaBx;2!NS!EB9=IJS$_?n~YALtS{b= z^Xi}W_)QGl8hTN4+~Em`f^4=Gi`ze#i^%=9GjnT!aesPi@RlyI_Eu%LHblegLdK6& zg;A?8AvagDq0yW5QVh3ah3Q5c65# ztePGemV)|+REU+LOFR8%d=TEmLY-AAzO1;@8U@^(x1k8g(veNarNqdHDVbb~%mH72 zu@OQ|&GF?pgAWc<6)-1L!;3G?*cP`qf7yOCq+FJd9X4m&p#q5s8p<13=a}_Ql6rBx zD?7(h+j)l7`x6f3@TNfgCr3r$VXRzuAw>M9D`FROSp)%N6SAXr@Z{U+P0RvvOXbh< zg4{m5`G@8++}0vMLE_G!^y=v;@Q-hwr=rD3(dtz=)R}_P8UiAd;7oqM;7&MV1K;Wd zPR@5Om!VA<+&N0iGj}-WlOzqv)W?ucFKc?AQ}9{*2BdNMNkTfKV@4r^zSN<1LAXx| zIVQtzCf|TlE&UxQzI4Ba21>$$DcCjAyrB5979 z{s$@us+9&7F`kD-JpJJ+iqc>AqrX>p+SPS79BBnagkFj*+gjZEAn(e6R~PVkN7%)SfuLoJkPz<61!OEbd_h%jL!T5)jqJ8rz)Q4BPm_elX7)ss6siv_^B;-$@K3mv=;KAB&J;I(^ zkc_9?m#x3tTvgQmOqJt%zU_md6`j=g*$@b8gP}RWCiC z!l(9pb>>7o_p4tEw|ORPbM0VLQZt4OQrm1Twg&NQDbC9t|xm)0Rb=s{FW_i-e%Tva{TJe2Vo{nA}UTmnxO3@N&~%!|f-pj_xnNi;InYYuD~I6?dkCxrcXkSTJSC zkk>R&7z$s$bxOAbb7P2mnW^V|q>^6(wxnb1(i{4E<<&%xjrHk?HAx%U%`vmSDn zmypF|3Vxal65sle`M>6&HMLvCY7=ml;(BGaUU@b4j9Fjaof-1JT|IaXdS^JOx;H;A zE(wfhz<8p+g9)a}x!{4$|Jpa+5n5#a{DdDkAzpCbJ#*jKQ)v6gcKa#T=Nzb$ujIaHrzVmZ;()Xy?nFEqtUwCn6V5rS|tXM7#7IS-GA`2%> z_4Q<52h(P;j@D}K6rp_HdVQ70sN}?c%Sv zj3^xpOD^YwcI8Of6mJ=N*G6%R*|`}$eGniV2ao1S33wGUPo6g8voxrunc!S+0Vq+a z-{rE>^oNLFGc0*MBGI)OAcmrbH-1nVbxR>ix$exo`a40XLn(h!eCv^1eU-X36Z>Jj z!Gm=uX4^IcFEV)2*Kjx`xO@rrmi>ZFe!!c#1*|>$9)Kjxp+|Wt;!u&yb5ZXg1IrTv zt=c^Hj^FOBcFZY!Zj2RcpL+RGvg{Q_Z!dM^vgiY4B+wM1i;HKBgZ~;s?4?ilqM`3t zTj~A4Fxg#{Q0XqBtG{S^nA)PPmkBU_q~NI3D39KDxc7kw9Ul(Ii)Au@n9pfw?J}*) zyDU?q;@|A5BnU_e$-QKR?7qwzLtc84Q+-_4pN|bz6mNk3iSUYMuuNA~Q9TEUB!ufe z*TRme;w={%ORgMt^2n;7sSjrRn1JP)WzmW85_(J92sCoMijs>6#p#f7qj*cBAXl!s zJj|caOsRnUkXB#n%(z7QCv;VhkN>!Dyv_#wyh)FuE4b z5?tm5UUue$67J;)j~^A^Z%q9YiMH`SAB69hiXZbQKhM^IgLh#ok|gk9wDEO zNyu4nG6bEZrW+f@)o(%E-P}G-?yOz!m`t66d^+Yk*hn|}^2j62Ya*K>38c*{$Gp$4 z;>%>mO?x3|tHl9&y?tys(>X1)ZdjV^IkPny283~L>8Gq(_B>ZQExfq=-{PNs>?=Ak zG1fe|4-q|<8c$tT{mx?Px_2TudcWF%rwiQU`Mgr2?CLV$w{ee@#CkrqcE~at#Rn4eyb?6*H}UX+mq+VR0>Q)WAqSfCDxY zy|#t22=oL!|35p!F!z-B*ZYj&bLJgS5Vm-Xd8MHKNLW}ip5bR2xkTvaz(Nc5Z0sBh zH_SC#@S;}B6@DS z!mHCy-*x@7RrI;l%OPHL`leOXycjw35{IHta!b+8ZosiCCUS$@1I%0m!T4g-t4RMU z!uM77Z1%jgI2Rk!sG7gnr$+ORuk5y+cz; z6IH1m6Y&R@uQ|;0dd%4Z6tnKSX>Hqm4$fx8-u5C%9Ja^=d4PX6@k+Zom$8rxF~;nh z2govKh*^!~qd>$cBBdCJFOSW8Mt_Yx!0d&GX{UYmsV3V4zz-3@36gRh%Mvq2ym&ydDoN0#AS2MwW5Jed?>-Qd$l}4TSKv`1(xd?+3hHj07`*BMZ3v`Zo(=XsW^`K zh(QT9HeKE{fo!JiiX0MlER|s=53Q<{E7tGQoh|QZ>U+j~hOo=XIzX=J%p4-g?NToQRz2`?E!k+KGSmT5F+J-XC030^CSI_r zP$OGJ-c-34mpKfg19Ekq-wKO{cTp-t3xS8@5DBXqgy9lyf&rIrK~m)%F48&)bY8Ite%<9h9EL zk*PC^yZwOKk`>2(Zfe1qq7UqN*o@&9Y~U$i9k*IX{vQW3xLv_xc7%~&@3u8O_w_Xn z$yqm4Ne^A|SO7RZ?%I}HY0)rdEp10X zwUA4=$rKJwRf##aoThm4t*4pNGm{7lqQR|PyJ(3cugsl-drh|d2sOBP70E;{F5g%e zy8`#(AzKf`Y&%tKR)4*shbA%tHDvS}*K+ab33ecQnr_Y8le&h}i#d74ZsbbL_OP3Z zz5%C20#5Gdycfeu64PhMULcc)vv+PJGj6>rp&f|* z8hr%t==BgWScoC2`gadjo=l%GN@NelG$2C>PB-WYe=-1rGl*po8L3FXf*eh z!J~17`G*Qx=Et&cOZ3`vUIimJIrdy3B$2*no?UtEzB6Xde{9y%{n^~Pu^}`QCfI0T z48Arym`qbCfF&8k=;KDHx*;V;NuofN5JVCEVif$;_mvtJb%iE+)Oz45zAh2mk=Uj| z)ZI7^jT+=)oJ^Lu#4ul9i2*F*;6seN*EQntmk)(5;!Yf=I5O0nc+~^!4H+0Oy^t4X>7`vR zplf@z)Q)5OS`zStGrbZM1~B8x8(s#ib5GbP8WtURU-TO&agMXtoI98VJP*s=DTjCT zuLtcXgd&qnXQo66Wh^?Cn-{>_yn-2=3YQ6Lt$>G|TZ6R=K@2npD{SO~N%;(Cp@U_9 z`Jy=)z^fW-Rl342x78|dV(_K*n5w)o|LlF9&&384-T4fn?mKE@61M$#M|`1MS#}Dz z`KdSGPgQ|#`vr^MRhgEsSoQNr_r^&PFkisjjc`hFZKD|};(@U$Q&&GfD%qWVD>CEY zK+0O)5sRu~QiaDseMV+0m3NUaJ7g`6*`Tu3Z!FFzBuL4&QS`GSD&E-l%!P6Z;ft)2 z0d=dL)>XjzYF(6R?1hlk$h5Zv^UaP44Gv!)payWHB;BrQ_i`@!7;-w}LDKe{B?F$5 z%{Ie`dXDoob@n`rOKvT(~qhRI7<5a zezcB4cm<8>cXMU+ZJV$P&CQi&i&H0rDKmbD$m%X}E(VhV&7gylT&@^J@NTb<2lwB| z+yltP@t>2?J+c}lz82!VL|O|j!IdWccfofinNYmU-Q3q^yRuRJakm9;tNv9Hn+Tn| zDguqQEIoBEPc_=*LZDR-)_&8h26BU^2HHrCifshM6QDhKa6kka27<#c?wsw5_+_Ls z#|v0+jZ0pc)a#|q8VK27;u~gu(h=`P=^m6)v;}816v3$<)_Yd!Pr(Hh(!$e%G>IGk z7BVUk_1>!(GTM+@h*pq6Z*P+&f zpt@4*D66^*XeSW%X9VVe&VlLhdj%5&%@&{XCoomF`0&3C&$E8Cfotb~v|u^GvQV#` zxE$H$_@KLl$IsN~YJ|QFC6%1qrbGn!_-eG~A552_Om2^({8^f4`U_u_!}Ml0TmIUX zABr}=UjYzbg*g=b!TxtG9iR^MYa0%*b>kCQGDf{XAWYVkdZ1EUY6|SO=B#!gl$vsD z;~Ch0FmD7q1Fff7Aah9e)ObFxs3!w^siKU25WE$+ws!X4DBV9mX#m)9#yVT4Xajq% zM~^d-C~Jg$(uxf{S|5P!V76}Cr;O7HA?-ffoD3|8meyby!>A9*-KD)e9q)<;`0ZjH z(QcbF=5nX;=2CFEP_8HfG0+(C&=KG*31bhO+{e6Z+&(;_>_a=yK_tH5uqUU7FUQU! zL(@7_%>BAOgYGw91ye09_BcJ4IJfKkQ}#4|c|ZFBug9DQe1i33b2#JR%G8=4e^Il7{-b)Co_B>jt%xPGu%E@ z_8*P!aDsR&^@-&MR|!A{jn(|YNopC{y*T+@o~Q}|kpwge1?8&jkd%Kzers{Tu&_?# zKU6aNOspFFDP@cA^o@+HqEmzm+1hbn*9lxM$7IbLPk1JV_Usf*L(LEc-{3x3@nw@v zBf~%?lN4KZp1|J?d@FZM-`JP@|1?0!Wq#m}SY7o@lXrBh#+}3Ab^^@7%U3OU_YP4f zTn{w11lK9~824_4!BQBbri`Aw0rQ#rqY*P$5*betZ^DdJZg0b#IYsr3l{uD7b<3>a zQl!QNN#L}O2VU(W5})kGZLUge4C2MxW1#q4Fv>6E$6oPT?o`lZ(#&2yyG~U>ho4y~ z72%Wjqvi&S+IVS_PBV7AP7g zPC~x_nAUew7Dr<8$2R(CN*LZs?8FQ#B0{xRMwBY<$arJI4IEoz2k_JZV>y%E|D~XP zS#kXb?V#n{oTUy?7_gBsWS}}nz1pkuY^ceOhf>^!(M#YjJm{LbKz$p_mg1A~#$)4C z##c7NOl_L)@^~WpMX82nNfLgJ7S@9xR%(Szsj=~av8kC1W({o*2-`6!(u}Qh_v(|9 zN3Hw)SQztrh*Byx<5rPq`;?!g$ZQ&=5++Cx1oszNV_{rbZ8vV2ItOHY~I**P(p+t zva!CyMGh#SA&^FCy7IJsa5|gQ;HA&>u4=B1V!R%FM==U8*R9HIs6B&OpJ*h2b6CRb zo5ufnE%a5JQ_~*u^xR6+TsBcd4^_x-2c472t!rz8u(&;muOuUOnuO2v>Va_1H08}J zEAK_ieYwR_*>k^erWNQCba~IK1@$RK zfplH89pW5*)8;P@C>1IvhK$Z>-H!}xuGw%V_I>a4`g`s7(q6(~~^AR__sV13S97=F4< z0DTRj<*#2`{ne28({M81WW}~3&)#b5Zf`nSo2~2}gR&l(f?p});GvuRhbmBSHttH} zbH>n8=!u+mbThx*%UsEL4Io8`GI~3;$p^PM~mH{%I4h+jp^n8cZDDSt#_cces}p^W_6?fB)QA#po&BwRep<#hp2bdvKhJ zb6kRFp?{FwK)gS5sxE72{&%71k0@0YK~juG@;O*=-+%;#iU$`%?@;iMv$06&uY!yS zuv9gM?q3egqiq_odn8JXYgNNT2y=Q+5L!-HAg66w{3z?Axh5<0guWP|Q{`}5;1QRv z#yLF!9VLCkd$#Ao#iSm$F!&!<*wFpd*!jtLLwFegX!`Mwnd))X3yHT{=1yFmx~7Vt zzML(_Y_C>(qKawN^iG&zIsqM7g(qwlU>(9TJe`Rp;f6H`} zb>Hm4>C5bffEmWT%adgwCfe3=89I>@K|8PaK1 z9wp8Qwm_v;?_~LXLD{$NmwK_Y>QByBtq9|#LRv*sm$#5HO%$hZ)%e{NVAl<(UT#St zFI#LcCZ5_K3#}Z1n-9!Nz0`6wmuP9~ShMb9EXK#{HrIdj_}6pDk2~{b>W{aQE(}Gm z6uztX1$CTgfA40-fbpRhtb)v(@tcVq7(1m$F`p%jSQA2)jU83Bi?-ls8NW?_!SY#! zeQV%Yweb0LPw*>X2WsG?UZnqzL)%n=u0p8k}FY7M3kTLeN^K>>;lF4 z-7EMb3Y^iK&i-uj7MkSHcszK&y2sqlI*zbQ_EH~V4VMkgzX2|Axf4MsC;IL@3#uN~ zqB2Io0xjgmOTg#@{~h6-NxXqKga;^RE(b$ncW))b1k$WxsHwvoeL-cOuqyK1p7ohmXFogQ7Z^RmnbHziWWm@D_GbUQ zchm}M*o7&1Jq5@zbO^8&!i>u{^P$t~!6KcyMbAgj?Dyk;;<>>aef&3e$o7BPAv+KA z|6s$f)z$qEJM>@e&Ud$tNtmTYmH4$V0l=qgw4Hv|+Hh-{T*mr2h(Sp{5LiHP{4$5U z?s5(#4A;a4P^+Q`^AmaL^gZP(=j)$a{5hxc$@Hmjuduj&akPpq?O?G#@66`o?sxy# zyVA@*nd{tqnLx!aI_H~Lp;xJWa{p78uX`saGUf$)o@wr6^*dB#-7+>k>V+}HX|9id+|D<-eut3wMzsH`p^6dEX|!TKy#0c2AFwk=7WpFYyZHn5 z=vWo4QPUOxbifAxT~|P}V5hg5BpN$djFWjyu^ts6C)czqoRIVKg5p8B-syKH6a7q9 znL>};Ui(mM<*n|Z!<{d(t`=8=<0Y$SQcUyj%+vLbL%@C6u|3jr(TGPQV1D${Pk-dc zq8`Q*Lu=GSUjl~3C4dxGCf&OQ)=clf&H>Hg@@=RTjgqyfo9>Nde)KIMbs8yO#nZXy z4M)e2rVx)avbv7m{svo<;4HDPO{oLK&Kw?=z;^CeC+#XT0W~@%r4_p7G*1;HP-80% z;A~k@ky2Evf4!>T>gfZ%+bo?IuUV3k$%wS9Ap zGY*|9PnKy}26&`Gam64Pqao6K8HSZvL?^WVccuFj*HesM>+cL zY;W-s*HfKdaGe(Gehhuq40QP*lg$}p?JJ?0_IAzmmxJkg9;JU4y@4`OVqp7$_(INo zB!7F25AGaY!HXbxP(Ze>;v7vhGl6#r`Z0s|(KY)NIo>q+EZXNJjHCZsI?1t;J!>OW z59P!ad!3-z_$X6wa`samhqV>UZ=p*~*mU+gGhx2U`$&QMU0)xLC#w|8(`M-WHe=nI zsqI_pUVqYu*sFlTLDFV+p(d$;WN-C0?kMGw7i=>(-8u)QIOZ6MPi^I9O4ZjsTP|?{ z&CZT|pO{51()ec>VJkvG82kKbd?nSPjX}ZJ7@=cI0#JyJ_(}5Uqc@9Gy}Xc3t}CXq z^~xf_n@VKj(&Pi6ziLaPd~%ZrV#_U;I9rQl8>}@rx*wqsIHKkr_exRBMAjF>{LsiL%oROZuBISHvC1K2x;NCx1!OkNZ{G@MOj+Lv z#k-mwW^uSP1PtG^9hD$Vg>A6~NxtxMZ+^RyF{>{hYv48c0t;-RBiE;ee1>VS`F9Jcg%NlKBGEvmm0>aMiZF8|e7k{hH3x@fb5Dm4*^?*DpA;)^3O>ADC8 zlA+dos)xrPI;2;QhLBCfZDg|-$kZU^2!@f-vsOaFi& z)jDdp^3C#V%3f;h$r}O!l`O^cue}Wwmk|tuw3qzBDj4!O$eCA4OTEbjEfhMG&5B#i z0V3-a(-X*K`M?vb0@$8CS2pN$C>LoFz2HJl@Mc>bSFIRMYX9vFJN0V?-61v{g%!4; z<|Sr*%h2C{oqnMgujgL3jhZjg530zy8eoCOLfMPLL==L886Y6uNAg+cqbUU>^f%{# zoW2S4m$$NP5#M?rByR1ERJX0&$OLDw8_TkWF)W@#5T7~OHexl3H?@)t>MGEWt`n=)h$w@S6pj=IG1zGx_b`S1V*X?@|i)sILUajHkp6z$< zF1jukedg;>7(M{!o$jsF9dmkdkp_UP5Zh1lwgn0sT(7zH;Wa>^w*7q-gdV3nA{ z(tE^ua8l&FE&mH%7($&zE22uT+h!XLrtsUMo(d+OiEV;-yl2~4ZWL2&5?do;;irwq z1qO&{@(a?&02b5r4p+Qla;mp_{K6h0_@S)_c z$4N|nMJgY+Gop7M>hQ;PIx+{nF&(jq$(QIo!>)# zhOli}{QEMxC_nTi8z134!J}1U4|IkYhOj@Gph-YbO+ZLVhMb+rNQR&3BW4opj{@Q_ z2R_$XEGs zhrF!|2HfGzwDtOP50MM~)B~%i_K=r_9V2bNIGzOC+)3 z_Rq=DZq508#RD=eCrWjP3NZ{Bg0R_DeDkJ8nGdDZCcMU;#aL;L?SddWM$y`71Ksv& z*5}}@?u|wim(UJH3ZRH8;}#0@pm`p)zLQuPWgO~g`~cmf_N`*hg|AT4_!SB!4dsec zniyl)G)s`o=m@hZodU&=+gY*P9CVs!d^AXDMbTLwk@6Awmh$9!&()S3VxKw3Td#iF z6+dj|L+8yj0KfnY2ILcxsVOYAv3lf_T28_F67Ckh)R#}xVZ zxQa65Sk*DqW6RMH5*r@R3>NR|mS1qbHb+=h0x#8~mQ$ZaVRo?@LgmY?j*n~nGcg{vB)VgQURQmBjpq_^*`}U(x@+%N7kHc+! zpDZI=I&XO%9a!>$g8oa{{t^^BnDW*_TXTXo^%!?Aqe?zdIO;uMC3}r&z+OF%U@9eq zB!AJ7*)Cal{r{v?%BKZ2bR0c*~CE$v0# za(&`IUuLC*iF(P|StQAom4DFS1PbpB(Fc;mwxu#o6l01ik`-y|$yacFnm#P7g=L2H zBxYmByz_rp3xj+R(-#wortyVWIo>{TYwY_cT{&7f$p@?t{?dF>0`@UH8wHlkuq65+ zUGBySj!JPM^6nXXb3b`%x)H@Z;j&R+elAcm@@^G)J7G60>SOQtG}^^bDCx@7VNhMB z>d{bsh|3j%F^h0>`7x|0J!)}2?P`3aPhTPz!!R?zV0i1<@d}zqqk+f6Fn-60%hhTJt=jtZpnVO>LTX)zfn_ZYN;({3k z=ZbE`=#vQLbaF?sKLwpV@@B1nJtciCs-)I=uIa6A_rBKn>tVm8&;^8Hj+lX#*h_b? zi*1y>XTZEcNU+>-WaP(EHV20?p%fM2Yha8oo+yG-raN_{GFx6j!k)P%))jT8d>KEM zMsgO@S)PK@y&w%3Z&cABm@RicXLthRuKj7!%JU=GdAz%1EDOdD;iiPMU%?}KRg{ko z5*L59&+<=qE93-~4VpGsNMJB;A#>^9VADR5rle=B#vGm6vzA$KYp>Kl`?lIbvh82~ z&7bOM5|kH?86#%RZ3kNW*#Z#Lk3JorXDw`yLPT#at{AH@QrUTn=g2WxhfUR}!WQ;~ z256$iL9P3D)t8``QNCFD1Pk0LsjMB9BUE!zCvk4Ji?E?jHtfPb9rP?V>|)Fm(XNSy zSpO+yf|dv_JY1tIBF%mi8m>{Crg%nkM29AJE|;?mj;9d*F)^fRCsX|k_7v#!FE0%M?uLJ|ff&?z;h5{)l+ys;l5h=)}NH49q zz~%G7A4&2Fu7!3I4-qjmOHCe!DNQB|5!nw!sJ(a?nTaezE0UF4N=L3VqumB@C_EGP z$k&Q{BGo4!==bXdPt^z-i`6CZeGldW!7yIJ{ig|0^ivRo&KCCp@6Nf63 zTq$`S;i1NfoxNL7`GfhEzrR#TLWszMHuIFISRSSJ9CWDTWlBHh+#uACC)}(*dmOZS zA@-LtH~H{_F#>zy)qVQQo@-_pTM%QM>jHAbuAP2G9IrOoz~uI3$}GU=vaZg5o6LD5 zaqzws#rRzZ0t<=-qSjqj-Dos^doPYzvhCwm6ekWE9>PBk9mLS>n0AaeRNci>0Fk&$ zTO_%E5O)xw5o&k!ewTO!JgbhR=YeGS3}Xi0xkhl3g6ByRd;JxID#s5|r1~mhmXHSu zoPRY7uxGGny3p>Ag;s%$Qb)Dd{3Spg!^q(BSYJVXABBgnh>8M zN?t3F(A*{z(3x;c;I_~`vbpQqa{!h_@{U_9^B4z@Gy|Rl*8s=xN*NI^33|A(Go`+W z@M{G6#1QLSYHpwiknk`$4!^%jihT&iQL=UK+zCVW-PJNFB}8jT)%KEdY`wEtz2AHm zG6W5;GEsmDG(f&@NknD6em^dwxkPy{q0y=1Gh7$|REe>mxp4t&1N@SOG35gKDo%}H z8T{E}x#0YuZmS)LgwXcv_+Z1d5ZS}_Wa5AZy zvsw?Q*mvhf6m|dtEU*Ax@UcHuuKSNf)`S*cF`TfJXcGk5qL>Aq2#qN8TRJtDuQqup z7(Rj3gO^t$;s`-~>|=zxCk~WAhmgg7-uz^C{Q@BfW`4Y4T_XZBZ>RYfk6JdTn0|5W zs!(WZ+4tAk@#^11L)UObn-g4AMlPUb{H7ll20VZB^dXu&wbL`jMrE9~>n;d-xt!xI zoXAByBh|Q1E2Selq!sU;^$5iG>$&@{L6!uxenI#kwdeXGCmr)tPHCBrr+q)J5}Q4X zpoU|^GTy~j{+OdU{GU-XArbNU->d(c-R3#+C3EA`#vtD>Z8d$50kKj&VU~yvZ*`Y# z-41zz`s&-;e<`&&0p*zEWJ&!-)(m z0C1F*bM%P)J)a@@c64&)AenrlaM1tee1Cs%_GVE_Wv0PTt)=(dnX_33)0uTiJ8ydNU!CspXZ7UX-}wW@rK}Yl-@2PJHh*Q3 z6@YoSeg9nuefvQScW007<;OYwxT3zy@lRE|%WCE#dU&UKfNL9uTesCpHPh0$9PW84 zcc%?NTu%l|^kn{>h*mRv-qqXbF@mf{O^%TGYgLll4hzEC?qq86e&VEU25x&>8$Va5 ziVQZ+kZlgZP3@v)V4u?9w1Zk&QYa>0*5)s7u69kKg6u< z(0(3Xq2cP7a<~!gq1`c8ih6?uY|?jdXs)2DjQC&{KS@^jRjXO;jRaeFF$)F1_fLN} z`kq}!L4GlE9xwy&<_NQH`j4Mtw+7g4b7DGJ zQl?Gf{|F!1vCC>}D#_5}tcH_upgo(y-DxSc`MqtbO-LSKt!->^8Lsni5}6Z|zmAwwt$l)AZplYeZ|zprHz zB$Ze^<44>Ul7s#*YHxoWPGsrRHbT>jN7B6z1y>H4fVUgz!XAvqPhL$6#a(rqHc(D% zzJ;J5xd!rgpinNm9_Xu_TWE|s-HAKv)z*FqvY{`JteO=fgNa})p(8~ z*Er?*pNHy@p*?w-Phn*21bDg;?i<|zDVOSQ#Xtt9gKj!Ep(|Y=l7RBtizf&{!DvFj zS@qtIpTB8g_^5BUZcS`uXzse~Gk|c=D7$5lA1SC-?#d0!!6Y+#NCP-OK1GbCCyBNW zlSGEn{a^#m)PL{DmBP#wrkQUBOp6KdOijMwi8WIyJy3#rDB{>NPb1#AM#c<47i2IU#FCXA@ls1WrBJ zLd!iUMdg`X0K>b?_z4%*(dtHhvFCTYC0mJM@j|xq!@mWwlC6l{bW(xhj-e_tQkIQm z@2GxZv8X{|ccQpqM9lv}2OSJL#?pQ`GJTRreK1AgVxa6BWt4Xyi_-G+gux!6Z@oC; zgptEw0V?$IY=R#5m0TQ(6POZ^Z^H&B?N0O9 ztK`*_k>v33cxowrNlgb9hN;^VOlUZy=MP{%r*0A!R zaqcSDXR7Tm`Oz1>_SQiB+&!M1G<31PBHDYQkqsbU)wR9T6!A_Y419wdxV{C8(mxOy z1R9A0C>#F%TpsAlo4@obbKW6mjZyjCC#nU~vxng$lcEHPB$oZLs+H9e*{UEopCH>#54V8q!i8N3J;u6o%DP;`hk- zfh^~fJ}`VM2pUBcg*r89DFXBMZ^YgMb6J;y$>jG&l{V<;Vl=MT*;7{>cq;0qzPBE; z#L}j=n>2{e5PY421HHFqdA>muvE09H5sn`BKpMP*owgm@H+kO>0-xqu4VP!EX;()3 zxN`90lS>MuU0W1_J~>I3jPam0ZjcQS(Cnm@vtSQb+ zd=KIBZb=nY{Anx#@ScuZo)5{X*{#qf6+rhu;SIQ;Hw_fVj|9Ys-_}kc_|87vTRoX` zJx1mdYxwD+$?FDRnu;!5PNbG%U$wgJ*Ec&{Ps5gb#RxP;YnBH*3iOVw@!J@#;N#Cc zIHVN08<2QnyuR(5_lQ%D^pFR4N2+GwRrKKa@>@+2{OY9GxiqH5?;U?8s!P<^7Ze@G zb$FD#oJ*t``6!n)w zgvd66K*YCs6K9$XKcf3Guba6%(J!PYbp1V=Z9YJ)BKQj)4??vL>mT_eAyn==w^z5V zTKH}MH5ya5Q9!Q{98$ex05mpv4cNR5DS`xVpM};Lrr>R2Se!7?F+P(<9q?Jv~aub8ZpGNDXznyGVE5VPBg2$NF zPYU~fS@ZyWv-L7_r2hs?VPu09~vQ%p7 zVjgt%9z$>V?a2BzdSw|*Au^uU6vAJGXa$PN3)W;Qd&e#z5X}^r&0t)+-%*ZS6CtT& ztOzz+h%cg%iodU9u3*vi2=0DnI~jEnc#T)L@7z`oO}3NM|B2#-So+bqLuUpvQoFVH zM>w3S8H5Tsv_9u5IK?@;_n2=#0@^IG&pEtuV3XcJ7TW@E=M{Ysx^~6jN;corpy81R z`iImnrv>u=9LK*sFtLfjPny9}J@sBuN^S@;!G03=9f-AVx2DqFvP?%pAh_?>DC9&e zrwR0!?-o`pl{B5mc|mv6{-6;Y-Dp*kl-S2^<6T0XH=dBHpJ+)J29Kbyj@yh?O1cNNQcecDjkh%}kT8$;tQ}RVYX5B3 zG{`c97iiL8HwldX-)yKQq=~G~`-*m4e7s#rq!^Nxn6#~tT_0YMl?jU=MV@_%8PKT= z8oW+TYUk`*5nFK4E=&D#r?&%v?gyB^oasPT636Kc(u$^E>N~Bdy{1epE?2iql&Xt zE(%(xmE`f{IOaDeg^8EH7TFO5Y)*11?W@JVe$NCDlvco@vb8}ew1#wKNl4W9bMD_+7rd;38h2=6a47a38!*~(57JYnS zftt;shZdSvzxhGq5#dZ*xpUtTtbCfV-6hd;2{xW>N4_eUiqwJLlKK69Ws{wi-9m{) z(A{47yCIq`9x`#;Q>KJj{7K}%WxFQvVCGVdd4}M{ih*DlJNV*EzoBL$5${gyEYZ0* z@~>AJ(n3J54tq7P4pkc4Y9qN-3^-m{&z9BXo-(pb{mUN&2neyJ!-|zM<8VR=7sE}~ zhr3J%{w!P!**tTbzGgC_ z=e*Pgxb_eNrBU>qgJ4SKd&FSL`QMrGwQ?Ej#t($ksa8%^Sy-!Ii-%|d8hryQBs-Y(2| z8TnMW$9Lse>&CugudI?QS#wzcwj6E8i;G`N)>i{v4Q#DYm|{i;6IT(pOPzphIDRG!MRMTyePlOU&4Ka@N`w&mb&gePC7zOUT8BV^66=W(N$%9{x zQeMxyDVr*hOC^NZE*IALlBR5i=2PD~M^`bpU zGg$DkM)w*rK^O2$))+*Qdm4}gQDi8)-rgJynj!m1PGHD-pyEujx zKuud^>E-izb+SZlT(^#r>OlN{#t%QNmKeK_l^eS%*=_b&obg4!fbKY^e(gCbw>SHy z_*=@_VZ%WE^S9Q>+M14BoI~)iNo@J0ih`W?H5!T6PtB*NM4$j7`lS)%WjMlx0ral; z9JLqOP1cZA>hl{XX+_6;o}!?{_DW%+i3Tgg4a}$_cGd{l&xBnc-yYv`Fx$^NC{Uj@ zcwJ&wT@yQf5%92B41Wbw@H$Sl#jA9j=vlTfUrMQ&ot1xh2zl!Phb|QM`MzB;|$b&)z?G@qXzAQLk59Ms7}+eC9N;POFOC;-?Lb z^yI$XMc)ZfmW;lDkX;cUBUc#4o3;;Koc_dRse4{KQ=M`A055faqY;98M>wTk&IZdB z0o`M%!&FJXntbB!!o*bxCSN7xOL!|zgj2J9_)Xs)Ttd}r9 zs$%p!`L?{bV5TJILZKMYO6(WT4__>)*YU-bD?9U7V2GdP4z6T@bV0Fc;Ypu?WuKDd zqqBUR(-gGB%3+0m*Ix3mLDlsE#%rlW7tB;?py+~N z0+y;6zAT~!K3pDe>T1Vp015}t`Fk_@sLj|nsXPJRKzf)9`GRXMTjWBYIuw;2zbtNL z9RH2tRh#6){(~#_Nj-!5bZWla_(y_)&ir5utZnmZ`*`HxguOpt+mub8K9Aj&qEg6t zMpb#Hl)1JBgVHT}>crMpSGZ$Y_gmH%mA?@gI8Ilf-kbZ4rR4LdR8C*ZVsG*r)t}=X zJU#x6<-d_6*8fS8SeaPa{&#g+jkdJ&mV{I9Lv0qk@1|8amPn5R=r%=S(H_hbMpBh# zc=MhBJ0Vh)pTmMhk6(@N!qQamkaLZ=sTx`(hHfuE1nc2=dInW0R!8AQ;VSTE9O+J!S+$PZyHe!YYh8NZmgTIe zoE9o{w6xW%nHn{{IC9QW2Mw)1r=)%}7ObB?^*2+lMa_t(FL$nkxL);N**2@b+wfnS zl*5rJxA}hHk#wM%u7p}x(fLl5LLp7hcTWJV9o4DJUUk%U=Rp)plfuzyo8@UE<;Q!Si|rK>M~pUj6+5Jk16TZDp&LQZy1cW1Hc*{lmoH?%@zm;sz2-IqeWH zmE$R6cjyuqp7i|N{M1G}(jP;PW_##Hpl`Yxrr?=U%AQ`Wg2x~I8wn-7M_X!H+`k)3 z*{BnlH;QzmLaBsXGAyd!htRj9B26bN%k7xWcchoC`e;+Ho$DvCr@Hx9XDvTVNyo`B z*e#|#UzNV0na$0#{3VYaTF((BceYIxT9LL+iICJU5>+bT|}RsL%g-)T+Wxq{!#P*m| zkE>enb-%>Jb(1yv*m|39;19G6|JeJ^vI@YclGI(b)@lVIwc1S^FjuCX5u<*@L6+5q zjXwSPdZ}$^sosN&XVPpu-5ea<&r=P@8_$J##8q$Y(gqx0MC=4$PIK#8(&8=6(l@a1 zlT9^`yGN6aS@Sg;Jp@*vX;-nm+|r_P6)k{-*?Z-?z!8o7O?QXTi-+sJDJ$8qMIH3Q z0GhCK9HO6^YzJsJWk*@)GS3dZKHm=9!5q4HyfQVG8?`P!Jn6TOUu%nzOAXCK1*_XK z-*oZfb%Ro5SkGe-P{O#NwM_}Tu6viClgI1XQkhN5b>^kl$mj`cfD?ypslY)2i`0o2 z9N0Z9+P^7;N$T4o+~?SBAX_<-nEZKAfjF&6@svJVT|R<%c1%RSF07C0j=?^lkonuw32H=oqhHAH_Ip7 zUDnR&Tf=$F;o2}@rB|eS?Q^`Ydpb90t&)K5+dm1?e{LZ=)-;1;1?bY5KPr)XKAHQU zGHBm>yg}qGjVl2jCh1iL)=DJwv&goL}ui!|5(s5_H zI(30$@Vzep*2la`EWAXpf%astDoQYczX8JEJ&PKcbjgu*V?z1GgakjqoT& zcCC)Bs0QnPCMleqnts=v{~9KNRYuD))!?{HYp+b&D*W!UlB?Fi!gHxt?M~5)=6b%G z-}4!I4=CY#bUb4S@9fe{yg^j|SlF{9T*gmmq0-grPFHOM)T)@mfr=Lj9Bi^o)Fwg- z$sXazhkS84sXy6BveS=trXfp}=Zy-r=LVXG$H25-NMGA3#<$(!qG)B{&BU zESFW--|z_I4^-I9J_`{TpQoPqmmboiKvTn(Dt!7_U1tyW#LAY)OFeGYQELvLKyXjGnTP zYi^&IB1}VQDjWQT28e>Sk@;q-K9)c;3dPd=&O^Dz9;`H227d|2{R zK<DBNwi4w1bPmOyZ9jHhB!ZqLh*>5IO+40eE13vtsiO65i_U}ch)+qj@Q1hQbE5w~h z>F<(U{gLZ>CIx_%AW{HZDi;A>{{U50GnqSB(*D5ki(*yE{^@r5k@1%U_h8j z zyNPo$=bzU&OJQJKz1(ikFoc%)9{^H#YwAaLjoL6iT7)fFF4?r`V_cgzy-#!eW*uZu zR|#RSxH`S!f6_gu{KW?@6Jk~}6Gw6H?@8MftKss-4KsF>3ui#K6h3V z_tonk!fjzPG(0)u0!b9vIBo8ouAbQNyXJ~>CaSM{*4jdr3Ug(uST$1!Wo44y9&8!F z*$y!^wHeH&OBA{UY|+|m=Vf*%rHl)`6pjmn*{C>R681-|I?mOm%*2eun7Mr+Bc-)K zSP;k5oo=V!#C=vS-G?^rp!xi2&2(ivh16%)7=TnS8R0?+`q70+RZjB<o%R zydmGYLUK4$4x>nYYrA(H8n5-Oy+j@SWQo-TtDL)Q23aVn;N9Z-_0lK{e?|SSf_iCc zxvs2{%NOH_%?Kjid&vjwQz?R)$lNxbOUIrO^rP8f{fA)DNPSqtFmrObwpm4f|6*Ti zGbWgcuL2aRlMJ&oLS$g7Sm}4`wunh<_K3RW_F0XdDyxE&gxK?f%#kSg$2rb$F&L-8;#xeHf6Ou|G-=q}O|*?52O z*n)@LHekK4aAi1w@Hr58CgoMGGpyayP9i9R0qR3yVGGZOoJqk`5mPyDuhBMboJrmb zsMTZuHn76>CQed9(igtZlnoF?!9gb(qbZL2iOviw*`m+Kt%EgJ# zLPWHq2F|*#4o}$2;PVq3pMLH1y6r0RM(@2G)Oc~WXKP_epJ(f1uPpzNH~Hii6sDz} zJ^v|pmdXmC+(-3MIwC5Etp#TNT=o8qZwp+p=DSNUh#0__J-1dO@nswfOgu)IDM6qH z5stj>s*D}^u#vZj72`Mq|7}=dYIC3_BGIgSBQz&p9q`I2mgm06?ZlvtIZ1VMVN1Vi zRM&X?R*AO{NSrbm6K-B^AucV%2A9}ymMHB1FqLZCUGRy*f%;jYSqt*@!A&INI8IQZ zynyi;Y)ofePaQO4vypHp^Ve;Z>bZ%G9QJ3oM_n08>)Cm@BO(BG?pegF()-X~J5aKp zCheY+T5vr2I(v1hV*oveJ^&A@u-Kd!4NHuT4G0;1qD76dG8d>RmSc2PfD=hzbUc@) zpKTvU^^i!5R#2+~#VoeBd@Hb8TXAT6w8&?*@YcX!mLHw6-JX5pk}ESmk1+U3~1 z8ck1f-&>b$k)T2@?|ohF$J+25<{B%xqo8&0l{=BY?#1j1#JtI-p{=We?SWhJ;Zi`u zW;P87thuemEbn7~s{XF$|7Swsx_XnvRqCAMK;Wx7hb46*gSG-bD{u#1egEMgHB81f z?N5Xeg?UTWB9u7X*%5al{sNTcs}pW8!Tk~o5u8PJZ?w8g*XhQ5(b&FVP=n5rOWnTz z-6qrx1xTKkwBv+xeoBHyyn(GYu47h8^Grh2vQ9>n7r>7%xc1l(TCz6y)!z(g!nf5R zCQ)6l`mzXSeHr}3NSJ`Gw_9i13P8Hgewh^d_?r7bCSs`&7M%D9E{8SS)eUd#4?cPZ zLN@pZ-b?A5#RnQP*ukpkpy?pP2YR<m;#7x#Mz#cUR-r&;Q=8(%?Z5PIopKLjAq)urlxzOXo;BZvB^mQ|zZH9mnuEVbdx;R^ve~ro=L;mDC2jjW4grh^2X{J9sDLaj3cDT>6z~Sw<9S+rE zj}Fas6bpt$Q!$Q?7D` z^3R{?4kLEa62?HgR4Xeh6Yp7(0h6G95ut&8)+;0G1rDs+Jp))DKDo%;&XOv&zuazv4yny)T9+sq*T==(W1};F*bWU-u91&7maQ31VJZ#O7Yzu)65n>`=ioRgtWU(EGVu?pkp3XWk@A} zMwZSRC8B=2$W5KI7%cpZI2 zG!ZmB7cs;)IML$;&zz859xF)^Po*X{Jw2D23&Kcd(lZGQmgthdAzveg)-;hqHrE~u z4nY=*!wlf|%o32r$I?w0qU3oA^!5RBx7tEWW_WZ(a71RgOnii3NRFOQ1LoAF6%8ho zb#Cyvxau?HkYuuZc(}IRodiBwysZXo(R}Ci!qx`R;-fgiiK5p5V~q2U2_`(IH;sev zC|18xialoEDETTZ(NQWD_^K4R?|d~p>)8e?bATTZ7XI-w{42fi`DXahofd1q=7SBD3_x>N7y(Z%RonyxG;HM ztG6Dmgg;!7q<+LYkC2IW)eitZ;rBi(!MGWg{~G%pRYw_4 zKC?Ru<7m{WfeaM&r~W|6RBG+)PgRkADoh7g?Zx$QdO<4pT?uCKi4in>*$GR#gf?%p zi&WyPi35ltZmq&ic5iBnmy zz4+6oH`2ab(KzK*`avH&jC_)>p<2O9{`Or{1&^L(y@%Khk5k z-BhVNg6zp>3}N|@FPAs)lU-@JhR9xU>B)~VdfIs1*<1cqDp4%l z<&=-8yynV!je9U=g%MYCCXw0exKp(o%_G%a%`ONhq!fW?QgD@H%B?H3PQUo1H=D4= z?}C2ekjPK6g+S=1>AuJ0AXC?K@GY;eACfmt_mmd$+{C%9L-*Os0^PtVH17Hi%4n*| z)5zu?=^uF_xY&bvosCi)=HD+lp!m`8hSBu&haHF%DH`xMsfqli`jLKoknA6J2+e!i zXYQ$N)=MGgg3hY$_TuCtS~RJ4^<>DZi!bzhnFfzqGB5d3ucY#H?V3s+f7Cl4;VIx_1TUkLyvJa`MXLGzXB-s8AFS~Sw33*EL zuy`34f}xM_y6>;OI0!r2>c{=c?>d!D``?fz+y6wGESzkd|2L$$DUsBBQM=@3bILXX zfh^XepxV2TblJM8W=$je^1OkJi~uQZp8DFLg{yRvDQ)MlP*STOA!OsF-`w1(&2P)rwXX9lo^f_mL^rw0J0Q%a%S8 z{&q`$k!4;v;Z%(WxtY4UtSf)U@oi3*z&?TvNk#6sm=ddT`G-(C zlUzbTDF1PZ51QqnxgI~m!TZgYYTvPWJe^D|4`zGnMJ&OX1fL*ObJ%Wpg*KwqTTqU~ zY(RPON$1K}IuY$n3~H&D7XVlR@a`%gaH{XSvR9?=tKNTl&0xWEmUmZl0(v4GSuObX zN8)$JA9-CM9hYm*O zkpq&Y5b72p)-(=!tyr44 zxwnPPv=pOIS}3yd~4sQuh6btw6n7*-wCKbqdh2p`_8eyer534Z;?3P$lkC?NNu_WX3Xt$+ruP#%ci}?d2c3O>8C|o zCpdnMe!D}bg_;YZ+Bx=I(HL(8?7`q;HOOWn)&MoFqM5frs0kg`J#btc=|;GBvDjgO zUZ0br`^UZ>`s%VRBXpodtZs+_bR**7l>ki_{K55t0b<5(i_2&;vw1&IJHifR4QL&* zq7t)N42`+G5JwD&2{djVzIvPi?otDhPUA^UM0g^KX2vplFnS;tCg7EQGg<0#B4#7` z3(UK*=s9IKM)vZo3~R1WD{<_z6Hwl9iv)(g6(y>tz<;&6baeJiPxcl0)Dc$1aL+6Q27( zf3<2QY=eX2Go_iAwk`n3i$&&9CPtydZNE47C0TS*khp(1{$l}aa5`Mp7kX&^w@Y<> z(f2D-*%egivmWE2`Xs7yc0HT zrrAscHjiO1Z1%h*Md@AMECY1D<75C6<4ln_$Z;Q zX&9;gKZPRj$}1kh`RbdikvzKB*z`xIuHlpw;w2D@Y^g#MyqZ!3rwAm5DT-mRtap$S z^uV*N&W=)!undAf%kW3fyq9h-(2pGd^k_Lrj^G*r8N$Pcg2)ouPsI+9@)!+37E-D_ zdot>{@A3Qkvh-%W%5PkUxENztgrAu%{8ub?o~_-ko-ehGF1;t7z`}-#TU0zq}oj!l^YJuedu(daf5@eux4W^yt;4_4t;{ zY@#o+g_CaXCF{lyGfkyP8JnVUTdV&N(10q>!iQ0 zpn6n^>l7R8{K~b)m{tSW(OYaXUHoIvd&jtsuA;(8_Zx z51u4-c2iClQ_ML4w{BlQ%qU^{2(+)Wv&@JG$g1tqcLT1LN!F%;;2qz9B8zUsII`GV zia16k7bYyY<$62w>N*>T%TWDAft8*`!2-6x{|cZtiCX9q-T&Uh=tN?F8jYa?`Y>Xc zpy?C-`qkClcoOM~e>Yz-0Lrni2q*$1Z(HSx)U=pSwim8T1XdAEt>S@EFSMU}Y6oR4 zrSVl{9@D&=;%>@ZH+5#i`7>sJDm426WF&&zH6>B@Mk*(~DeMXNbLseW+FMD0Nt}#F z>H_8Dk@FH6?=s$#axi&y<~E+bn*DBs>CYgzoaAJKl(Q!yd?A4mg$M(3Sm#01tc5Fuy^k$&f9*m+McO8_IKs zbB}hlp3jzb&&|R>^@sWnx`k5-hQn4?_H$CNDC0)w@oQMJ=! zIgo8;UMJ)7r5~I>Lj8hpV;WJR0-az%6qIg;WEM+sXQX#Ckl1}h2qi~Xfcbb!L!>`b zusVDkhUrkZMc*kAsIv~4YE6y{?VoC_0HZK(QDvRI0!BS4`76jY3J2=ZE?5HCpi~Q? z39UrMQ&b+~mb1QfJV5jJpa|;E(?x_ z6Wxl%W&>~zsue%+;|-a7Kh97AmC3@%n~KPb{)9!4%=9PhxE982xc(9|dZPSW{vvZ& zHR8YQ!{F|8-8O+Seip3O6HQ%O^en5E{MxwsxbDC*67zoX7+Q4KK_&H7_B_sH|3ra0K z5)lO8nAeufwkKcvjXMsJf;ji7)4OShA=J)r35aTaa&jaI2g=IlZihaSpv}F(0C?Rm z560q4QlItupWebg2Z9vSKKN87CqW=tgetFdH%;I`tZMX zjeGexqmXw?nkGbi-a0f+GLgMm&19R49^rD=BrJ`J^b0l$0LGsa<4i8#twAfc=S zW>weh+rO}^{HXPy=@2R)2+{u(F!$)GNEZF}*hWqzdH)`jz*PkBx`h(UAUQ>T0YVBS zThSln6$Bf*HE7km4c|*{X`XVzu<<0f`KQOAQ{jUa}3=+57p+Bq%vtD|yp70e12jFupOd&Ufz zG0{B@*`Qwat78w*#?~FpIm)1ti zO(JZTz10Ck_*VQa3c{mpl+xkp-VaEws2G8k9WIoSGEcP2Mg1~6Y6FM+SaQNYIL_x0Z59T5e*;X;oYigQ` zwgWN(wc>%F&$x`^A1Ru{e;w8aT!e1}Vo}J+@dlh;#|6M6Va2_3%bP$YxN_7u{kCdU z2&qp_9ZOCuZr6h&B<-*#=;jAu;xaiGb~!`7CM+Px)qjKpMSbndRdg9Jqk0qBGZahQ zQ74Fp59|Hx_67sxEn0NB?*DuI)0K$_%&MoXDsvJR4YItRr2WaRmW6VcBUHU{~<-}`TMWo*GIt2sk*YA%#0-RPe%b-jb0H3UXv{AaZq%pp=u~0z;i5R3I*f14^x({^Dy{KCIW3sLdJXFw$ zgpIs?PJVJ`KWX0gX{C;z2I9U3SK70WB!Ps|4k<4+RdElNR7jO*N33yDe>^UtOYT!} z%&T9tdFMe#@-TIE54rx9qaTcf+nv{_9O&8n9ddCs=*qa+rkn4SyDIJWBE;@-)lF@i zY`2j>U*#i)q}j+f>sT$yHG*^?-xiaEMpm+sRD^yW+Y^{3^9knZ^m5dv@hBx8*pJ`1 zSx-0l9x2y`2Pd?APtqZd%HhG6NN4#j&rwa?fh5ZH-1VN)S?nkOU)tmBn<}~vMj2dwWI<`h?669_B^%ud~yFg2|Q+9m4r_5eR)yzjS zK?+p2nk#U+{Y-^UI?DkFner4u24J_pieRQl^aOu!Rte~59?Nw5MQpu~#Ak$4Ni`7b(owpVpW%3R(7;Gjf0++?H zRh)$FPf@?$2YeuTu{OF;g1VnHMUF%% zh7(7N_%yv9&(Gc=$xab)t;c&?*HNZ5n}iux0!5Su>*Il$rr|QilR3@$sZofX^n+Zv zdoua`JQ?`JF5vsm0pPjx4~qy;G)1)`e7ogl*G%$Qd>x_khMs9ZH~y;5G_Zx`xu~uj zBEO;-ed|=!nCw$;kZkT)(^H*R4|XC5)7Z3h6ZHfqKp%j`*|%&MnjBo->%qVepx@MV z(0^lg?Ei(?v9bIg3a&L;n@+#X?l(b^XNLc(vBf~R+%t)YjGUxpH)XqyC%x3u^LH3b zWB_S{WB^Q{>uPRIe+@ooX1WqcAS%s$6HZ(M^AZt<5ozf`j;ztId$+DnA^|#ad2nqeS-yG zTErY?IFAmtPh|nS;c!Xto*OdN*wP8dw^e6CWqF=}cIAFQ+`rAbur=vKuVv`&yKpTj zaqwkK(9CtLE<0sv75|CZ=`;Od%#oyA&vN3;SrCd_v0Yi2A-5WLnyFn|cyd+t`!nn0 zk&+8NR9Q%EF9d&s`yr6UveDFNS>@%*nlszJFIHdVs!Y4Dvc;IokB28hhmv#lc>g=9 zWJu2P0Xh5e04SRY5@?x?$)kaHsr5ZJ`7v{JxD&;7oQJR(E$`t<@;vSq$L%q-_oTbx zFJp#HKHh3P!L*SZxY21)ey`c>F_N!xa$HKQ%6gQlC*s+?G2ce_!Qc<^r8q|lxb=zXYU+%aqPs2EX&P4lb*kH0J^}F_?{!{I%`W@p z5Vxcg%H$v-4makPUytQRm)BC70nebt`G?oJf~8jda-y`7LFVOIp-vEiWkd)2k>|N3 zt~ZzK+5mHcGAI2akzPZzUx*n$+fnzm`H~5sVAP$BVAS7G;(axh+Cdf)_FH&|gf1g0 z9vd{XSBNG`K`i4VST5ZU4#>H<5OdKq>Z79Myxcx0MPuoHhFG=Mc&zw?{W#`EVGdl5 zXQL*XrCOGzIA@9_fc^2h^ugy>6b4wYJvD;ya(Yjf;hW=uOH`w+@7$RN7}q zt%VKVMTg<~)NE%rKt^L%Z8k}st$K4)S7o}|1)F(d*w9xh(plu(dD4B+dLqZ^Ir62q2EXULw~g5Vwf|3OGk(gd7Kd2Z_kO1^o7g`+!0@R8%6l3|N@`2O`B1DP^1IffegVb=8WJus{|q#B)TZX;POq#!l1(9 zG6E>&#K71f3lCCZ1XL17Y+j|LPJhCaQT>stC>TMEKrC6mq$_5%YXCS|rlHjtm4j&m z_$L+UG?0g&*-OHt62bY1 zHg5#I*^;d(&WlV4<0OUcK7C<_!wc=NQS z-3?^xeR1HFYqrDqcI&ozIK)xbMuX#_WWP5U$PG4D&z7uq!!_z8i_3{nYorPajGjwj zeGe`K4U4w;VdDUkfm|PhLrfmsbArKO^>9E!%frlbX2V^F0_ht=mwjk~#m*{Yt*uDYR4ta;r9_Scc0C^C>2;^*V z{IQoQB4j`wmIXu78EZXIB|_(odUUScFEu{R#j8jCj(#l^sJ$&roqF?%{>oAD!167I z`1BpRBP7HkB43dM{*5tu=dIF*Ul^3E!0#|0?bmIH9>DJI?F_WVJ}1hQ>mOnx@R7Zk zG-Cxv(_$TE6$<&sjjw;d@O<(Sf^s0c+}n<%7y-W&1bLTCqRWDai;g6Kk%@V{_+rifS-qN(e%n(Q z&OoHJN5s_xQn2p{HmsA#c8d{zq&Ks@@^%!{s5t_(owebOyXb1b^B)QFC1ExGasTDa zeD(@NyCh}IO$N&@uYmGFrdWU{WfHIKovunVh{?jT~WW3=Q$!G9%N6^PQMSe+jbA2J!|Oi{sM zx?#Z01z`r(PF%8%jy*#65;YYY@t$sef;xIg=$XeMt~6Ou1%TXI7P zo_eX-@CtqCShWl1-0(OlrNJ zHFI+QrJ}6{b2YY}0Iok>k!DI5v!hr_M*!wW2mE&s2#YDv78MCaLW)=LY30f&l9KO3 zL}UsxC?WuOIr0nPqbS0^0xjBU{hA&waHoGM*-1Vc?N0IwZCc>%Xw*u+8sIdE)4T6`eS#&LyR?+f)%~7VdnsI4k6UMNVmrl|o&6!iEq_6FF z>qjU}6FAVYfLi^|;qOS*>-zwbt=cfg^8Lwn7i0rc9`KQ*$L0mxM8I;eJZ+T6x{f^% z(%?U2r%VN(jA3*#HTy6Cx&Y|6nY!v#K!xI}G^ZA40`$M*_%B*-)-d+@JOaxeyK`%m zUA^LohU0O30G>Qf;IJH)&+Z)-6%}O+Lcqk^lu4{8c9-mRghC;FVZtczqcVgN1`HrO zMUFc-ATt<6Z+FQ0^VVAh#Sq{%W%Zaj5P?GE`0gS!sqzLVrdU+*kL>_N7a(j$%RT8j_c{p7}X5c01QzrI78L9ZEm*}ST?@J zw49rJ>-i*T<3R!nUFD|VJuh1o3x&PTMT&}oQw)4FKmvO-Cj4-8QBe`|Ihs?=&_mNy z3S0%@mr@Xmi(W07YxeR3V{HfsO)o!Zl!aAbJ$P%YHM14CEG#_RZL^ljS15FYMQ}+A z2+D2!%2%0DuN|Mi*T1pLy5#`|XPk)HyhdbUPbrEJ%t>dRdgLMD7C=_81*#12C-4*x zJ@qO&I7fI5nLlAqEer-sR_WNWli)5-f;5u6Ce*%p)Naq8hlem6=yP?Y-UoQw#J9X+ zB=-dhu$z)%Ak+DsBz!}Jf747@761Z3fy&xmq&cPHVE{D43XJk0!4$saeE?$xZp81%w^ez{-RogbJ{0_P0_iqfS8S>>|3Zeam)h*TLl|2zrtkvI8p^AD zHC|qqM?t;9C^uJ#mc=HA^a~iYyn@JHJQZdp-yx1Ki7qFLVAJSf*-(G>mQa1GO0ReT zr-ny_^)Yq|LJ-$2iZIR8Hdv8U>iAZyNtKlQG3@ci0lM5b&s(pI z-w9npV*^cr?}+?;jAy6s)n0^+8K+`S1b9gZL)`i2x{XzxxM+LD=wOe1u(#)`+h^aj z+1aa`vhM0Bv0KP_nWFWrw*ATaGlx!VFtsqTu<&4CVFb7(wT4Hs0;Dlf9uaD|(#|_? zB$jwYAq|NvGn7sMJ(y@l?IVqKbV!OlMNKIn3!2g5z}FAQVtRC-a=>1e6cu`rubLV*ol8 z16>)=hEjd^t`XU9Ntc8H5blRpSqe=4sy_KT?cg2FYFy=})QSHN?K{bE$LCK`xC+>w zc(W&eXzP%lW%iIhW|`uB%ZGkyLsl$bwvLAImyhD2?%%J4ui&rhD49gB9Gd7R9Iw90 zDxx9jv8QTB-ifdWzlP7zxsKK(mQ zN1YYIuxeJ%fl}>%*HgF;fPEtuQlcx{!Iw8whDb0A?c*N9;_N)nd#$IU) z1a0QB!E&js(B0*bQfVX7GNu5B*Y#zyff+DkhL7Ce>(w=6+#dtN3UY&QVyy|cqo%7Y z$PDzIIk%+1?vjExzOWaB8aF`Mvui9A4;%Noi^IJpE*n*ANaGta64zpY*NCx)QG^M% z?&dl_MU2HO3u#&?yuOrUcJefd$%9m%Ezim}IXu_vv755_zy#>~I1(iP6IpSDV%BR^ zPtGHtpXPi9r~WIDudEcfY7>&G?Bq44R>hUVmivimh{358dtpWO!kFBkI=1-pc(Df zZdy%5k6o)K(rv46pLX#W2sYeC^H17B1T4-7|JI})zMwB?K8Om-3B3@RYiW!DF-Npo z;L^{+t?srYU}wa2AcqnVOq=yh?k!fI@eaa%c)%7fav*Xcp1}AA*^IjP^@-;e1TON= zf>r$n+Vpel`8C_>dbb{)+TNyFKC!H;GAeU-oPds#O;bCc&7`}ja_uqW+gs(=1~0c_ z02o}oy*0VU7pOATlKDl_dSdkF*RDU{jv$rYe}hdN{{=R2GO_&c1F@0+{CO2i`u+2| z#Kq>Zc@B#x^g9siw3$z;4j0uVX0oxDa0bXi2z7K17GoEk@^_c_Uc1PCuGcqHEm)9E zxaT2a!y#k~@G|qwcr!J7TODdl2q*nY|NcczM|(#Db>K<=@Jjg~j`rHj!Q*!;e9Do6wc>W3+Dgg<!y^zr1kMk={IxX z^!U^J<>Fe_jC37axeX(GF4SzwuKMr6i&~C=JNo$Tu;-bJlj6B1R3h2-Skf6qBXqw` zM_1b*$oAQFt%R>%N?>VHI{K`%MYE(l3Z=of!B?t;y*yrC?#hE>OyKVbywAbcZxTT* zZxlj>PXTLhD0uL8YvN?@c>X7PI-+9cBPNKBm&NVUIsBjKsa`jH!TbMX?3{u_X`^i$ z+qUf;+qP{dd&jnI+qP}nwrx8(IS=>kue#^0s~@|oyH<^F%`qP?M-3rlonxY`?jP9zkw&NilD^u0{0wu%k_}v3>ld?gH zFR709&1lbONUO4xe#nKhdyM%aT#K=)=$dM@D^Js;QZ|fh!W%%bu!I%g#KG`?p%FeC z1(;(pVRX^|3R554{Zsr+k*985{e~*d#6%_0oL=I*jhJl)!KfqfYttL#tngm~Ln1zP zxzAMbGp<5RS@2nz9Gmw=yw)o5yna?xJleV~)G~@k>M&5}o);$F6bmJpFJ!WHxNY7| zy59S*%-LjdW>92x!HBErYmB7QcHIl5t;N;h=nk3igk(MV8!O{w(F)zf-gFXk9~1&@ zs#7Z23HjS%5=-6)+wiCi&F4hL*$PaOp+X(bSO({+PuXxQk!p6hxj1M~(*lWQ^K2`m zwmDiL6g$l|_1UNR*Eb$h;1~2I$q*-%5%f&A5pttkQ%5RlW%NX8@CD)p!212@scPow zaz(5QMdUDBc-eg+CK#hxPlutFXwa^5aotHc%%JGI#?J4G+d!cBW{Bh=6;lJ91Rtlc z+JX5sfQ?RbQ@p;<{hFM-Ks|;>$r~5FltCxd-&b8GR+|$wX}CBKUL}-PIGF2oST6Fs zlo8UwMP0XbYD~Eq;H(x}Qj@R*TrVVO%#~W_qv~qS5TA4RP+PdOTBa-)XtAaL=#^L> z_|bZgGa4%#Ed)b>3rfc#bx`s_d>K5ig1-?Y9`B*?K~d8+7o+1Qnvk1E@m$_G`^?Lx zs;a#8Zw!q0ePCkJUWlyl4ttXTe3?fzfj9}@llZzSNTy}yHWWdU@d|XJ5$sf-RwfnROFC^C~kpO;{(2c<~1?Th9RmOSW6t45F`-jJ`DsQmbCHw>ub3x-*$W> zt{jE*i6pY)!FbflAYDeUW_iLHS0}10e^+(N^wd=- zU{9MgbCQ0{@^-CV=f-ZXa$^D&TWOV7*Vpf>kBJK=6o_wb8`N%%ZEENivK>C?GWqI8 zm_L6mxTU?croPPAJe~bkrRyzm_|BLqcV&+9I8UGKy*hj3tqL)R%n763Wq7+_pjy5< zo-X3~sKIO|_wV+@{fy}=d3!ZL3IyP?!+q79E-@IXKO&R4*Z?n?AW zXE0vpwtHTv`+j4-!HFdjfyQ1<7Mwk+c#|tNiDdiAHR<}q2w{Jd zVDP)+*Nzk$xtMN`ulp&DZ=70;e_lfyLlzIufZIHV4OUK2JfKzB9hqQ z0Z#6V+w#E2XuePl6UbPBCotGNoVqOUIn~LP(zn|d>YI7n$<2&w^7#F8;lGN=tw+8H6$z_U)q!*C5v-G zmpcUGk!V`b7;ya8+7uZh>QoFew|I7ph*|=xm&!L7n(@dxd$+AVG@*V`$gTIv^W>mRH7ZruFZTG#zsV6C~UNDX|1}Z#TGnx2W$^A*Jh$xb%q*Z zQj!+%M-UMk3kh+!hi28K?D-*LUt5TuJ}Bqmm`l+Ce$qX}(bzV;Eky~p7dAnS zV%l@aoPBfF=c#|J`Kf&r%T^O369mVeCGy#qH&9U+=wYf zuU$TwqlOYk0Ls)m$@YV{)YnEBY$nH8IY7A;LS>z86TyRs;~3%K1_=J@q?;$QQx=Xq ze{h}~7MPGkMxM~)E0GpO>%9nQ@bTfR)15fJUAagfA^b7$5`4sCpw`g0^@7hU4<9#Y zJCx9MTH*O1m7z>dnbGd?=#x3;EFQ&Aik;zL+hsfYIEJ{P!6|b*{ zFWjgnF_g3251SN(*$&y@Bo*U)B8FTM8GelP!4@45?|HwFil%8)lb+R$Z0|fWo&Td0 zPS0a5+4Gj8^5Ag|jExj0x?|TpVZy*C^5+H`yhxXIR!p*igC+r?#m&Cxak z!3=_**$cJbGWJmKx$DF2z1^N&zFR5rgT2TLNFGXktfY>$7>7if@fyb5eeGk-$Cuok zS>&5Nlbb1hCq7P;>-aO6tAml4B<~)O=l7 zP;+(5(w{vN^t3UT^cs&6iiyc>*`Usx!hUk*nE)i>k2yBt+KC0!C*$aaLNwJO(h|Nr zZ5vviv=>y>Q~9FQNV-3mwd$5O>_@j5&9YE}5q(?Akeg$6=!bKKTsKOw(l~A9$?Rj{ zb{LUg0+jUf5nWzG*-Y3&PVwGF(wZXucjd`{lx+``6OvPK;$`>q5<~+umlkJbHwYF7 zUbuWfZdc(p<5Enj6I?EQT1d-|$-*7=G@#Y_LC%cS|?5=|+< z9ARNVwiL~Zt>|J)7b}Cw>cJ}1K@1BXZ{obHcYUZ=a|3+&5^Qi4$jzrn#?x0O0u=qEM`Xl30Hu0&GdkxvpT=#c{6 z`q1#IzmuAW!IU7ssdcQhVvxIoUo9BoZOU$PcCe*{KbO-g@t9fUXt*&b=sRV;;&39X zd>Fk1(I&`9OI7b6o95?g^~Myli?9Ef#{<72Axy$Inr~G^8V($7=D{X}*yd%Ve_6Ae z*>OW%SHMh&^Nx_Iwj-IAvZ}TWLj5NJ3V#}-G)jg=6E&p@;p#JHT_7&OvPryy+OGaH zBK9I)UE55%(mqm!Q7;&s>c6}3ozLEEBUKx6lR=%;{SnS1wY%^uP4W>)5}`JfG$>hV zOxU29i0YjhrorSSj6X@oN<6a%GqindaC`QZx2L$Dx3XXn-tgw^ge?R@D*fX)4zw&w zPTs}fzMGazO|a@i>04ydjpwrauyUURd|UzwxP;%B8@LWWrJmhA0UrMqbg#<0~bfuicNN=?4o}!lw(uM zbJRa(v(;4%NjWVo+_E@^QWMMiESA<&HjpD(Z#tIQ;$SUREMx>6@O=HejVR)mViJQHVbAp zPOq7@7u!-_&AP0l+=fO&Cn)0N&^JGt!lUS-(?zlm#>|A@T(atRc&IL;zLPa2lnxB4j(dGR_l3 zy`BLoh3~iBli_7pzFFRN2qYLIk zCJ5uT-l(RQP`VgfJ>7WHQ6e_7l40%8Wr z+{ujVgaW$IIdE z;r__FeZpuPP*apC`y8*7pf@o;hV#@bj%{S~HxJ)}Fiy?HF#RmmCuMRUEb5=Qoph!7 zuyj(GnTui}eCnfH+z$Bi1-oRDO`vp!MZ{KQyoz#>)15#6PL{)WHqFxWlYZx^%I~b= z^Z;B%hhA%-)eb>=8p?VO-9$Lj@rq>ji&L$BoVs>KKq~?RFdPMd5I7!mGguR|7vq3X^(u z-nLFtIPB{fL;H1pKA#)oN`M$M{4|OjUhKJpJ2%QevVP%;K!7folGRJ_4qqjC>>dKB zwfmuiiC-$~B*lBOn7qXI<5za{Uvb$CsPdGl56KE+j;}mqNg55M+q~B+q0j8#=5%LO zp2^$GUAL|gua4(Cy8nnmGE-ezB<_8HRmu;KWr9S2grZP&OM8>>WgbQY+&*j~%eyhc zlwPY*oce8Sue)VeV`>b5oNS#YE4VvxOBFTEnGoNDJ3h>sM+?YiwCBK6mdI_u$68nV zme%`+bcZ%#Oqb+Lvr}0?6JPCUzpmD&NOJ2ro`|sKQIQEpj14S+Vyp*dRq}=s(o6Ry zn(zaZfukMxPBCZ`5OvvJCx0a#ZCi{;M~5fv61cosCoqW^OUs65!i2Y+?%zx%9#9l$ zIhq{ya&Z(cc1C zcDByJC3K{d?D1J^k0=a&W!-hyVq0S^&yapqG_tCPD-+J-M5r6d`PglqK zY+Tc>!sl~$gpAe8L4UA*3w5x#646?)(rW*>Jo|0CDab4A_%=!sWUB*X?7}PQX#k_5 z34O00ODSozwd5Mgf%}FxNPNKmFAT@=9~h3Ek?H@B;i|MI>^DV|e@hIO2BmG&Sn`14 zZltd%!jsDoO;{q5ON7o-0W{s3sP7rb=8{{zT|HPQp)mk;s=KiBDLCpwR$? zO%6^u%<#OJp4-X^6g6pienwti!psE2dvzcI~%BJz_f@*mz{_ooT#Sd z{qj!yD2rSIRhcH;CHO~ziK~k z#_U$sTi@XvOMER61Ev|+iuj2}DB{X#mn*#k+Z!ep1ZVWVie<7C={5>6YHrqVfaqd(i90mET5S<05~;{{qaT_m3YBeU#*Cabj4bJ zlb8?g-1c+@$YbP}=zI5&ak9(rJA06UO>SoF54n%uyMbVRz4kVq(_9}|-}g#7NNUB> z2)Y23+BOz)@LhT#9Crp$H~5?Tu?dMe0G`JiDZ3hu#aT~9LjB%*rBe;3gEt2VlfL5z zd>AK&1sqD_N!$r^(06}F4I#e-d>VKO5e=4|?BseW^pt{D9v70@IMqA*!y|{_9g5Kb^rYs^9LrNH02RmcbxbMbBPCK14L{OW=lm@xQ zC}S-)3f9)VU!7@v45L>BF$JNQq_jd0o!aTZUK}J)F@wqgqbt&9Bd!^`vQ#gvhJ>k< zfRAw(TUkj<6S|?5x)3P}k=oh+2=Zu!SEL%5FNnvYK|NwEy;V}3;hxouvDc|^tDS72=5|XErZT<5= zfP`rVK|=sJBt_}aR=t$D2Ukpi0e7YE^>K=&S5i!DqPX{32;YQzHtx(QCq(dE44p}~ z@;E~zjqET0V%n|K{!qIjMo8(-?YK~~D}2Rfnry7jMSI71J&4tqcL0rnJVDDd z4>k59oi;zIzHG?6J9FcnFO#2sV`kD*L$d;C1uwx%7%)X-LDSy* zE=)#T;6OvNA_FKT1sh6fK3N0V{VM?|hrZi#1t!nS5Q<6ALwJ!sY!4w6Jv4;6bo(y{ zH{iNfs&!L(HdjE~HIlWTN9KT+)$Sy;%Rua)*p{k{_NK1qa0h=SMANrP*3= zR?jHe36boXgej`PUx40g)rkL68In(oQ-67HeX;JVG`;Saym4$nh<1Y6N-G4MZ7N|D zutQKmSTU9%g%j=8<%(9LPnJSc%WixJwC<)1xdBT#XrUe~yj{`fL~wS611A@tL>2RR z0)w}-;O`lRS5`x#(a+3GcMrUpgbB`^K5b%CqrH^M_WH@o@*=tVkCRx2S{+I}@R3G^ zw|`wS3c#aKh)?QB1vvjrH>1_Tx-0%%WXwbI0fR@acppiA0|>YI-bAFV&;1%um$v|x zl)}r~@JqO?&t-mAcf*yza0*sUyPsyq_6D!ZlkkW965z_hjFIOnp!$Fy>Ia}d(^>cu zd%K0~gyNtJSxRRmx9}rqwbBFa5U@@m&|3mU=u6p?b>CfSJwO+4OYLN62yQh|ePD0U zoC#zTuhJw$Bs%Zf9}X_v+Y0^d1qWe)MBeAItX~Rwgh)N%h6M$NpQmGSxYXqS4a;4i zo+Ya)@d_4|l6pp*GU?Z!H>axoMTi=;pSHV^(D6NJom6kj>1KX8RZ9IsS*OV=?mCz! ztE~pzjkFYhBPg269#^kS&k8#l{s?0T`iD?rnSyS@$xhAMx5_|MyjHxIx44`636yfV zSYri#Kq3k_=IX7&Ir}w|9bOeTX`kiYF4XwLn)aeRb9u??WW|MK3^?jr<|c^k2n=ZY zUL(GTdpp|6+#u2*d-S%1T){o9sl|_*?cxcF6BApL`KPfHgh+n+Mj^eg5vvB76Oxm@HO|*_p`VH$!@7mW!-*h~)I&!}dqe?gCaJ*}T;CSJbG*U~b<;bKs~;zC~CIQOVu zYwXrdwr+WZ$Wx47@<5p?+*s$C>OL1hQeM9=xG#&zRNVSu=06=Z(jsM!Rq;rh^ZM5~ z5`ZqC`BO_K@`;nnU)4+R!v^j5n)>d6?o}HBHQWeRbk~GN`luEXZJ!3isDBb{uUmuc z6QK7Tf>ck^d*6@;IJXi2^!fm_nF+%`){?OboyrI_ zH^Xcq0VEB*HgOBq!&-%pVx3>6_pChr-OJS9 zu%+#l$+1gH|HSp&dBXj?E0>2ly~r;G?V|3CO2H7~K&BEx?T1GJg*jFT!}Ap)C{D&M zd%{a&&8XD6;i=-yWkKACviLwEoQR_JkF>b`AhVuG~5wiKvJ4(~i|N#VL4L4Y|nNss2wurVIFujDhPmB0t;EnKd(8qcX9K z9I!HJvV(==rwxA9DZ@0Eac5yXA;eTikvBXR?8h+UAW=a-YdfZ}H7CGFYbZ$I{kBhI zeA9UPM!85)h0OZG83oGU_rOQ`6$@@ z_``h&vqxe7#1as02~9%B^k_TZZ?UU@8{Zx~lcsBgqS>ziDG!4_7V}*t1`xpk+@7LZ z3Er!Cztg?))ItbkT}CG6H1cxkDFU`{(1s zYEcMK{60|`VGt#qE*@leGfdgG8O|XT1!3?_ju?qinRA_67YR(cS>t;}7*sl6V%1G=b=h9wU@SQ&dHbP9qD7m zV{Do14cr9}^-OrKJ|yLy?a9X1VXHJ)7aoch>g1w9AA>5_=6XY*i>LHTmfjIb3;Uv z2_}N#0vtRQ0}d^iGS_E~#*+hu+l|)0lHc{U{bPyS6@g7->TVZ+=1pN;+Ia$c*d7nv zV)nSPP_0cLy=bDM1gj`feORn^+d;Rn@{tD)jnT4&-7SMbx}qi5u@c~_q2Fci*{re6 zQ;!~^F}@7N6RkI;)&M?TS4*CZLKBhkoBOVQ!%{J6NVi#?E(=4(zfeJvf+{v#+V*Cg zz}<|BDj%Xr&N{d6!m6=#iJC29w`4VGS!h|Z>wOOPugsSbJf8^DD$hbsWU~z_mi_{0 zW4V}eY=r{0V0@M6fe~73*=LM|*M64t$S{w-=zw}@&3KM3&SJN=SdhiBiT#!=15xCv zQ2I`xBGp3!*;Aq+$>FJ-?*ppEKo@HFx3+eK;2-bx84 z4s+Sqr%wS^yi-FwN2wgQwthEP5q^T#)-J+gNoiiE5TNseXYb2#?h&eStSK_TIe-MI z-qf1^NoZEf9l5!{n&p7;IU|f0Sy570KScSbn=GaO$>Bo_7VBa|_K6}egh@dSjS3j0 zrTr}2Z{;q~^g(N-gE#`PNyZu!S0*4-k=U2Xg(Er!A?VTlaJi#DmnAvuz5&fslF}bEYppZr-kc!s=lPtXZcE zF?D}Y4tiGfMf#H~+{*@^l<1kj3&zg=Gr)=kLx(`XA7C#$7Qymk-J;Td&{OP18Dt^7 zN(mra0ai*?aCWFH(_8$rxeR=G(F-i5-q^Iv-TeTNeM%VlywosFs1AyEb<@C=iHrut@YNS$uP4T zbll%~#9@l60P^q0MZ@Fbk*kBk7xKS#Zvqs&F8cgovXn}zs2*WeJSMUDP`do_Zlc7S z3*;#5%+p((PyTx}Q$wdq%y*l;f8ETVFI1`6mPeH|;2ztKQICQ*Z?W zZ+`)Yt+?c=))l4EJ=d_=B0RgBm3cWX-8)m~8sY&mXRe^^P`iq8{7O6n2;8?;t^t1! zhzF*~$nEt$ZYB40i4HuxY|C&{n8D&H9CAE@2s3c|J(fvwL}8U8cglgxl#r44IGlZDOgm^dYyv1bvBa-VIRPY~{#P zcWT@u}?orq23Has>`(y87M>7#S_mijYr9FM?4kmL1uqwKd z4qyq>Sie{GfZ(lC@F&%f#flm<{bk{l$F_PbpMd|2Ke_cXC#${8^>1v!EO3u-$*%Q; ziGo^%fo*_-Nee9(gtc+x)vtj72_)r%K6Hs+BdsjnpTkrn{*xCr31^-J%Uw_2;N}6dE@ritmEmLMp&ZVnwHv`IoX?X@oZW!|g zJ3&d~v(Dtyf4$>J{udS;Kt4LVV?IS1#0Cnj)!}br_|DykQjNI*&SZ`)x;t#-BO^?+ zKp7b@^{-cB*#{FV6&|S_z+y)*gY&A>N39M`hTY+DxoLSL2A5iie|j+SvB z>2X0>$zL!wttNJKczzl~EYTx7pYvvUfA}?b*S4IXE?lyC_!UC@EhsBUqpWPXuS3(= zl<-nPZ0GF#{CH{n{_XNQ8q+;j57FoLdi}L8&b9_`-Ruc>EweOF;#JdXZHn3S&M z$MwhFhF#CqLgk!h`@pujRuhyX?JS?hq;c_{rp25E^Xt3f)s1sOGveXhl(Qhr-@K#P zhApQokD}^TCR1D2`n=<;GEnV*xmek|dmoP>%ziZ5**}!0;=k)pHvg~WbB%d|IT1!M zdB;XixeVs&Z0)0<2e(h$;2yH@XCJ~@tCqitRy(sY`k#dn;^Xba{pr3guEZ<^&2U-( z=e!A`p)l?jm3LB6{ZHyN4UDEFeKz`SB`mkpoxNuMYL|+?chJ!rMYdc3@ZBCS8|=R! zuv647SFs)`K1aGFSC~V{Z_A*wLolTiDt-r!NSsYg&bxQX`oNGjKO>gqPS1$*R_`#taWJ(sFacmm*Jh3cC2^7ar@*8 z&~&;gPk2&Awuhqqd+j%!(QBQ6tx*C{%>H8ZgJ=#n0Ojz6Kl$U~N}h;xU7o8XwfCut zM{Z&aT@bMi-ig(eI|W(JkqIWCJpoO1eelvx$DpInY6|LZc9h>-z|u`nzb=PAl!+)v z)q=Ys{a&0X?h3uSH(3Yk%s65~WfYzW-Cwb+PKm`w@-ubGSxu3!wWMB0pRhQH9*uC^rza7UXj5+j{Mg8R4mCWwvvkZM!@=G0J} zB#yD>XZ?<~iM8YZ$8b%1>96*^Ytz4l#UX2vl$Q=-P8<~G=zHS1Jj2WA&*v(#SkMIN*@=O?;+89$&F?M+At^L6KPm0QSAPd@HO!pi;Ozd0)N)&;Vu%&oOqu$CskP zZv>HX!4aOlY<*u}mGGzbE9%AbgF;nOG5mm*cYW{%4^5W`83)KX$V!mJa>x?ENlSpT zCuQPoU&D{$egq@TX@ELj8PxaYru=mR0`*2vyUd)-*2?2mP?&{?m{pIGg8^4mX+$SY zA}RgiOsd47SZHTd6>*Q@(|yr7P(u)2TmC=)z6yNjpHS7aAmcfE@_041^%sLkwIZuZ zoh=AH`<;u_f$sO<(fAE{&YOn~JPV`Z;jz|}i&_mKj3ZyUCZVggGKpG3bo5#>(s_$w zNto!+LUsiJQ7-09c*g2`OgBZZ2w49u7aA`{*cme*W!e_D&N zY2&kpDuklE&sDixU(f*e4ZNS9R^;cps5m`DVU(Tx2(-v8A94P>*tNMVG4 zb`Jczp4iw&U#oo~u9?K(^*bDc!X31ML#C|N4PLaAppQAZkkdCt*x`iJE+j$g+g((+ zCotzn(H16jP;qost75WBy&ftp{Ak#KF{l`0)2ORb?;^9y+NYPc$R3EvK7i5SIzdze zTmb%CEim9`)>mo3kUIDF&mtfIqUzLKhFxchWhb%UM`hX5*W~W-cU>YKj~LdosdmM= zgmFAG$%e^ZI9{m)V9)^}@q8>Dt}t+xkd0_KP?O`AV%8_1jU!gXuI)N;OZx5hw*JF_ zN9Gz3C~D~ayz>B7Cl`UIt{I84Ztf^-isQAQyX`DgfWBn4Q=M9~Ah!!b-PbYm0N z_$e`)w(OaO8iz1V0Vn^OoWIyDYmcDHP>hOQQ2rnA!O02>ko00?a3TJfwS%3R?q9^(-Rb>a+s3f`^)p?mWX>0Xeq$jmH?4Wg#PfXNm7w|0!12R9F4BP{kyop z=!^~qI?Xsoa$#o9w4M)jei6?n+18+_5{5bhV%0@=kDI5br_vSx9Qws^aR%I8L+VTn z{l@cw9TKIW=q9p76aUZ>u03@?`D`vju`YYKdcbhGiFMHI@;{a&?e8h*?~e=>&wFUm&Uv$I!wjXaI}gh+F829firW&R+Am~5~-Xwcwokm7fn*iV8A)^W$%=w-xQNj^e^5&mCDGSkAz?IrBfddYiFn3t z;1tF<0!uxyV;}s2o^WORP96*;cL?~(XE)F*-NKUg3v4dh=SWwuX&|bmf1_)W9<}@| z?2n*x%L)nsszCMQnIz4jK=|=n@s*m3MvD5fFr=tjVOVqCiaB$@J8;=S#kYkyC zidlEv4+h49qk!qn)X1`2v+|7DJ8|cxQ{@26XQn? zhe)&{>uBWO8j|1JO$WawQS+-qk*~_Ic481n>SfSjKDZz8KC6wD1pguhsWI3vH$82< z4JS3PN~7;A>vfF7n{oh8I}dsFe-gq1lb#5Z+4)?_q~=}stio`sbv%V8onjvemGo#n9yr+Y%@CpOXyLwqTLc6*O;L^Ies5;+c*l& z!$S$nt)78ch)bnjPP19%5qW^%C6$w45DoLV`qzLx9{} z8X{@JA2#dOqe}(-hVkM-CW8R3-BuDa_yR%8FU^=P1{I%=!gSQyn9M11C$vR#encup zwJ9vAQyPpm7=gt&#)1jGI(|YKgrECW2JQdWmCL*@u6gAb4=sHME`J{#^##jfzea}b zW)Y%Jm!Jd@!Ki}`SAe0z~2YcRj7sDZx+ zuya;;3vjCsozLng$BAB!Nmb<_0Am|i4BFq09t~sjP%QB8-=-yJg|BO8i!*jKV9 z_nPuWj51U&H8hJIyIc-8NMJ)}wT`xzv->-7^8Hvn!|I=-mr`xCNGjdxD%881V?>ELJ;g;2|h`_Z3JhNZGBoh%Q=k9=64f`tMnM69Zx+| zBL=1bn|h-3mKzq(7`ISb-10wj?m0uJ{Ea>nhPA56kEV?!2wyC{xi-f0L|%aj-kzoo@x10i&!`63{9qUZ|bAC z1DFQGDFVVHM5zqbCEyYwEcQSLC`U^q&?B=xtB9>E5wW_hG{Zurmk<=ci&}#~sA1=Q z6YvOOCe=kG;Ow=KI_j%K><|Z)zR+!p0yUbhCl3;xFUn80leuzW=ptNZ9<>2k3zB}% z1z9CP@syQY&<#EMbER9F?l?3$$G~0owRv}TPx8C^G@n0NN!TyUqy+wcFXd9QRW$9s5f&W{W? z@D1k$Tq3Gxb>RYk@-<*lM>b5#MU(&akY70I&@R*gU1WxFb;ga|8wB`~G})*Q(o1zS z-*>)|!w+8VF2t$RgD^ysCq7xf2G8Z??)wFr7gI0IH4efE*_P6vX%*!`DWEX(W$f&nNQFRVjnhA)el$ve^H% zfpEz=GP6(*(dp8~QO*J(W#?FIE~QCK@S=i_BMYb#SEK>81IH4OIS&UbuMn;t>HCZzj!(s|O1e5urKW4Cs;##gZfyIoM2yUT+h~MwvGPIaQd6qwqv+p z`j$fYVd(`VT?I7GQS|~HLzhMR^U|J0?CtW3tT&x&$=BTR#I}nyA_^a#Sq_lP-tTjMvwN~ZOAf$(_C!iXFP%aj)L2;sWoGshom!wp$1+L z8;H8GFUD62l8kXBcIKcWnFP}%S9CRbhs@fV)tkK z>Slxml$WH%*y4xNABrgH?*sK#v5xPO^ILx$mGb5S%p16hfbw`3CBniV;cj&Bn6emG zg_^&eYFwTJFp(~F*;_@QZTT5D%t$3}IE6M1fOFsoo+B%3x4GUt_*LJr;k6~(gpj0W zZNUG;fQ*2(`}niRduEG@as4Zab||or`r_den?X3H9`TWkRTL3Z!zK1F_s6u(DQ`RF z47+3ldP*4@cHoxp>O&NeC0K3XciW$N&uOcFNV(xn^0r*7J#KO8QN;&(Ky!DhDOah6 zRuCvYfN(df(S^Ky)t1?``^BE!q7F%QEO{RQA2?-Pi__<|p{`<8+qU=ami&HrL}fBu8vvHxGfrcGK>E}PBq+kb18vVv3{W*s(} z??`sDJ|l6)oN1WGIU_^W67jIb6fHset+#Pb9{_nqCI2kz_x;o>u^U zZYZSR7NGnD`>1UMJeoV)ZUC0Fi0$Pp{dN!x^&Pw@;xf@Wzubr^CGKRKl++?eL#VKqr`I70tK2_KpXQBAI3B zt3~wab#^mpCmU-=%%+H%e*s(UvkN~GJDoJ~Z$ii8)8eZXJ#pi5DbdgyJ^Bs0gE z$f!Nbcu3*LD>q0vdpa$l<)BZXcuS518JKLN|J~{5nBZL?igKGplp!|k<3rkAfmK8S7x=Uk$zvv*a(&D^OV zhQsn3!eFT&P9t09SeyfY_AAl+^r~BAFdt`f2CE1LuO^k!Ps8#q4DX`pSQ1pyhm~>+?kWtXX|kkL=$#0o7X zR7YG3u8U4vS&K*)0E|Db#j*#Y-CzD!Ln;k<69*O44Sggf6$x-rV7SNFISkUx+{qTU zZGhgQe&-f&G(9jZd%$ zuV{J5j4|d_8Qn#ZEdwdVz~^NFnLabw5JINs_O+%9&ux;{14saVgpak2BMl97-BX7jSF{O1bGl_voej{?)MI!eCu#;@M&TuefgI-^)rC=lO^qy9W% zim`Q1El+{OzARSn6104vzkI*TRfOYOWR<{6qtnhRdUU@E_GGXX$l#B^b>|Z^(BZ%Y zD@3nUu3>Xz82PA;k7lip`69co%NR*AtxR<1_&hRz$94m1XaVJM^Sj!YZ>8{* zO!ngoB&lnN*1KoFRQ zNt+|rff2k$dS&=f(<u#okMc1KPw#z)LDa2d)IOD0 z^340ufo^a^73cMqF5$Lt(YJJj<3M7jvgyh}%qQkLo~P7F@JsKI7@btaYe(@?ss_@_ zj>J)@0`1Ho7!>hC2r&4SsjfIFxxEi5jj*;fY*Wk~z3h`Af$!GtxE4&}geg;bHlvgt zdm%DN!B6f2^N&26uE2-xvjHGzKCZ|_o?p=Iqo-DO5R0R5%mIg349cjWW$XHL5Kv+5 z8@P}zO)!wVxa{xhsX}=T(Fo#rn;Y!5 z0DYrog)?~S*mo}On1rdXNSNGPyw_Q;D;^>d8=GgL5Zikb|^N?H7B&<=-(41WU zo%R|zS~GBBGebGaIfA)K6}(swQ84w;Ia{p(E?t_+gNbtmE{E+#-TOTUO7AR~M9xr= zl9Or5)DYfisz2HjU^=N)+8?e6$}8SRci%=5^uH`#KFzWS?HQFLcJPyUA!-`-{^0Xh zg*#sNXg(W#Leb`+t33wAKw`r}^T{4K*0x7n)R%^jVZaA{o5T(%JJU~ZPY+Sq5}CX} zROyYjM3;dei-uzeZ+cmWi9mjc2^LP7hB60_f|}`g4luifVErR5)FP;}$tUbVh7DSa z2=)vOCT#JKEuBSrZ8Zp!$c*%c*orOoz6e&5wK!{GUZ*Kl~3@t&jf7ta__MC%vJ&^csqnUk$0agBP2|?biVCHVLY!@yD>3E=8oxVo{vXE9sk;(rOQW%!oTy@3m5Obf6{n(#ZJpS*ZQHhO zI~ChWXWZAmV|2glzp(cB)?DBGzRqUsJ_62xKqw$Mh*=Dq=e*#d6Rk|y;ls8_-f!Zy zm$1i}yFzs(P!GB_F)*bLZ1e4G5kU?u5XN^P`M^%wQTzaU0qRG)r9h2~PtM^*G6aMa z{7&X+1Kw3o6R)RnL~vv(Y#mct%hGBW>vLcSjNgS?af=t_pr3Qi&c1{f^E!N|`*3ey zL)O$SMfirJ*xT!@;BmpM_zr9_{2&bA>Cx%*dy@29ye*Ww32pb*CvXBxNPqJMAlKtu zY1pNKKyWB?xgUaq#IEx94?_iC5Vcc_3mdcQ;1%q9; zQD^t``BIG!mz%snS@U%os#VvavkSH|BZw0fk#*r^#&0W zGkm>U-GY9+C-3uhh{5aR%Kn8Ei=BIgJ-lVu!_o?P!{r|8V;h=pv*szz)(e+q*)XH~ zScDNK?jrF>JYXJLrr8BI;%#T^f+_<7Nkyy5nIi&_PdJy|2*uFCfc~<$Z1|ZX0h)WC z2BpGq;C(L?>+ibE5Ix=waMXDlUJ9I?eghC?u9Bp)cfIWy~btOfPhT_|t4jM`hE zR^#iLxh--8lUYM2k)h*1Zh7NqFIQcYbe`LCj0U>HWwcId)1x$^$nyC3vC8=xc^t)j z@PdW1SgPa?&7M-F6wq-+Cfzm=W!@iR#>2ex9tR8KLCjz-^3DxFzC6#xmap?Q^*`UZ z$~3~^V^*9HD}jlIbyUSne<6c#t!$jSWCr=`*{1W3D8Wq}yCqDL<8tOiCBKjLjG=`2 zX(ienzE7RU%fdQ-HuFQjxBoZRKYfFyblYo;qJ@gm;Y2rKgm@PL2WSy(lG&m93F`UF zFz>4C&c0qN!!m7d)V z_2SFSc+J;ujf1PROc#5JLbY%S8}+9QEAZb!AW`dHQP`^0F-f@7*jYyyM~TFvwNGvz zp)#>u{tMG-F%DoNf+Wn0$|&tpll}u5R2I;s(M%|1f+z4&Bl;1<)eN49A^%`?>%V#X z^_~=9af!%CJ2VXT6qg}nuVn~8!$;qT0Inbv)D7@)5%23=>pga28+M5*f$pUwV7lJqi8)iyxGbO0BF zK?c{-L-76AiS~hfKUzh*z;rQ8-3=0Vg^0?W-5rDhX=wX|+Y6P5{}T1`A&eeQ$HLK| zoEZ`^-t#6v-m7D4?>FoU9?Cw{7>}WTW2w<^vI09*HQU{d3K(a#=={k8N5?U+f>-Sy z6N=d2femVQ*C_`!u>jXN92F37b&AIw=OQow{;h5sr&$`Pag~gAJgLu;uaAW;oX_^Oc$~`r zzDTitc~gXx@4e3mvEo=7d)8C`-OFtm1BMUEK{=G_wULCg!$xDJfg!SRa8za)S5x7W z+mi>hqXH=(odB(?aUCp7u><#7Mr2ZN@~{?Ib#VMng3vkjMwDH5Ef8VHIta1pUXWyt zT$kBDhc=Yv$(bbP_0x7x91Z&B;l2{lg@@L$BjgD2TRi1s-3H!(OR#f7I75$< zxBjYjuwTx?-0&4nZg}FL65O=bUdG}$KB~&IgM>a?`>CQDt9-zIVk0s0*VYQIwx=8f z%6_*KDije&3KEktk=?0+?-gjq(VvYJyvskYi-Er#a2v$faHPz;U?TaYaIyS~j4ZO- z^5mpDE|rxNx8R_Sj5sojqpVpRTAuwaz_+?xZcHA>^I3<$@8d z)_a~=$mqkUTeWUseEXU%&s`;6OVn~MsjC7HKgS1uIlJ3;c|JZ)j+4tZK>()7R#N32 z=Pc3)NG=dC)44S&POGLc!W-{{PUZCqgUz|@6CK0DK1|;lOKv^tq*!^Uu8zq#bsd7q zz}))qBmQ0c7x^bMlIy1a+)yt{bJyJ69G{V|t197RNOz83?k^SVEb13lLMw&M~v5O+Fjuo%I z^k_D$w+~>2qiO5bL7JME0&N5?ona#EgmfT9UaxY0?B#x0_YzaH|8{G$+k9f*{4W1z z&ne$jtut?mmi7@ry2suqnn-p9;6j?OR<*frVEWpCSnLkXT-`=nnvm1@Gr|Au-ICg( zk2$mnbB&Ge~HN0n3?}?%a$e74iw6zXBl%Z`+6=zEPkDLTRX1S7g~fmf95#TDIEBH>m%`Vw-cg=oZ!)hrd9-A zZ{tOgl?iUfm5T#L_G+Y}`qabU15@Wv#q+62eK+x?6So;2vNlumKd9xPF1qV{vSkVr z&Wr7Trp|p!0M^SY(>>f{57RnLi<_HeU+h>WcZ=_}vDE-poXET#F6X3Hb7s?HCR&SU zPw}G(wAK*Zo2-j0a)bE$FY%wjBY}A~T-}_n^QaRCa(uvhJ%Zk46TzI{rbF2kQU9?3 zEZ6b0c`JzHv4W*Do2!VkOlLfm+P=AS&vy^KJ*;w<5T{$2$28H9oS1iF^Ra&I=ES8R*v(lHd^%sHXDN`agk6-yf0adYVKmERw6pR0#(K)AYXY+rX`~oR6g$@!$iB&bnoK;Co)ayqUOP9>~6w>(A~rB z93OG_Su?`Jl&kg?y@niKIser*#)*>MCd&E~aS(3Y@22jd&04VN;Z+ll{8>N2KGmsB zc}P{+-^qVZ&>U(aolZU9{MY(cwVLsIjI%%o(T$TkEN<)O(zuS+BmD{aj#pNGVZQ3Y zm=qiBEPDrdw|}KnH6i`f<%rK%CuD=}QTz}u1Dc&%O|qavjI+*Utq`5}n7_Q1G0GQ(tz-PJ8zs4c(>1>EoR%Mk z_?3%(i5!4R03tGXUP8em*+lYqS`QKmwgECOXGy)tB5nf{phWG4}WV)01-qGrvxGDjd8uQ90dGcPi`?K))%+ z3a10o6>oaX*4;F_cdu=UL*t?>JS(gpf|oUxoGeRfl$LIhurpb7Q-X+PY|{ zVMt^~XJtA^bxnmf!)YR!O*hYlfVklEs3_*LQ!3^f+^Ha1`O#)&RfUXS>Ei-fiGUN9dSd^I8DSLB3-y-6x(hAXo}znFs0m1 z$rB<|g6$HBS<`H%#h$JH6#nW&dUbSt2tn^ZPf7KPG9ApgXnur1(KL3vg4-M_r zsjF#z!)HG#O_WUGd*>=F$f;6)S2@P2?0rUA#ogpk(eituzN6-gysd&Pedgux+vr1k zIr=2mk!(GF$@sIVt6dpC$+#@Gq1sX@VUQCLE=Q0mYLJ>`vSuJn^*H%8vZFH0N_|ob z%nH{KkrIQc`XCf;taTLi!u13UZ{9RPtsqNzoGufJ?LQzB2Z?$Vhz5w z^}x(?&x|)r#6W+O9c`bvU+VF_ObRPgt39W=tQkZQzmaUK7EReA93ag^H194byg4@OG2C=&_4vp zkkTCme-VsPtJ_TBu?4hzC6wu!!xi69IH5Ebg8>FCXcF^kJ}IgYgCoTQf25>ztTB6K zd|_(cU7AA}dlsw97p*mkU~^FbA0 z7D?wA3kYjA>+8c6?SMvfgyDYbi_U5dqA=2P&JC4B<^7C5AS;`z+q~AWY1+wRH|V<# zdfU;uKw|@ZNaYHk@(OnTLV%7KA1r$0G<`cAGG1FG#%j^AX8;d-Y-YW-=AMg_VCUAF zhDgkEMMTL=;!fh{#&m4;1+Xis7O9ezRkXkg&s`@wrX6T*DxYEJY%X)RMr#7_zt?zvTo=F zwSEiKuAs!;>te;vf(P&O5=&eAVSyUPwMd4 zE^dVIxgnpz6=*pn_lcZK@W&E@Wa@5P3HwD7N(%}@yOO2RIE64VmLXq|6#igwXkAaH zFa%snz&V}(%~qxiw_b7!I(;1lrI~+XwqQweENSZE7P#-{B>P72iXl}5xcgsqn_`(V z%6~16LzjW=aFtE#WtVGc3$Po`rBiGn zyZ$K!%%H52f5n0R^-Q;g)%9~|bnljF`qN&Dqez`n6&51&1v6l?% zNShmwMu#_UV{SpuOVufK&kEW(&ZeD{&A~w z>B-dSz)Ws+Ms9&Hcyd#jRhi~Gmhxy-gdH_^xwiiOr&ZIbxGLOHYu>FUIiT?U6=a5% z%cr`fvd1+R4y8ypc^585Dc*CAv=aB5nW$Gl(r%f~0qBI2!p8H3H zor_NMgXf}caEuyF&9=?d$njebeJX=<|hHYY2Q}nS%RZ#07rz_54e#Q zdf1beffu|N#M_`0#``Atz44#Ji-L{0y50k0Gb*Uwwa7qwm>G@EKDF=a_t{Gl98eh{ zW)cX$f#}m%9UYzS$Vj52@a*AENW8Id(2SChAUzcd48SPy!!b=S|CQG88HSqC8u`$d zS#IBIEUySbZivwDqD-zO;+;R?PUr6WhhB=psFPq?RON215r^huqUr4#ho~CuTpSQ< zv~CDv+RU_I1FU+JK;`r{IYpkQ&bH{p3wfcK{m4o51TR3+9-x0F{ z3Iq9r`pW=6;l2?-D(+IU1|7B;Pw^V&cazVL&5XN=XMZ%g6%45eoo$>2SGdA-wc$xf;jv>-k{FWJpzvE%gpt>k|f;1@bW?)f@ z9EZbb3TukhA(IxEBn5v$A+}cVGXx{&!%;#Zdd=~88UxQ__icIDkB~n7NQ`+9vzHft z#$X+{`p_iWrOra|Aic9d@8DH@j39>2&bwqvGs*nYU|@mVF~UhUEU@1D+^+B5tPeyc z{?*cS?bz`cV^KUMye(&Sl`kGlO@|K{9Okbs=O@ZW(771_({m;2RB`HI1VbGz>u+gK zm*KoCX{QA)M)D?OxeD!+7RVI;2^&1I8UENJRD1fAf9_-&p8LFd=>&makn#@Sclg4` z^QqY?IY$1|#((YE4}Q@RF+Ec7q2~*OsN`s}ck(;}_(6s=X_ zNRL>v;|~EOGIEFV)@FlXbzOueXAK}^Wyi8FIs4Q_22`1Ixv*7nOXDPT^{JH6dgVqF z4H=yBscvOKKo4H=b$bQ}OLL-oVM5h%(IX{XLT1q(kwhP(?dsP=?>VvKkZC*3UtWfh z|C)ttacu~(U#}*B)i}VR_?1v6=;vdF#N ztA$Iv;#vV6)lxK$Jnw=P5n0ElG>;|t0{g9f(SH8oY0zO`!+grNf4tZVOx^cQCg+vQ7l`ADj7R?$=!tJ(_ks)! zL(v@=KqGqQrIv|Cx_JH4!(j5$j%cUh(_2I=Z~8P|;o#^^1S;WkL5m$ST66^@I_5dT z8bqM1G)u?@_fD|>U~1MTfZbR4A$#KkeF9og>d@nd%oZ5EF9GTG81jy)z7vMRD+^Eb zPF_v}i7k>KS&$Dfa3Rvr}TM(BmVlvDe z59@5UjQqJ-33^qA3&_pQ|3eFe(Rkz|-3$husQsWGLMP0w>Vo-LY=~i#*-O4+&6Wpw z>Sf2xiwvA&Ywf#Ry}vU}z5UQj1Bn?6VhgZ;4fA|G!J2h6td;ul_tWLt5lIEXfh)?a zY9W~#K7<9HhXRAbpm>9$c!6KJHs(5U2oVc$>n++V#`W9>Ygd0=Y#8}>!-dIrci22P z)9t1eTe-S?If!yjcTt);q<}Iqk>{11WbeDR@QD)dj>(f32+( zFWw-t#T)sj<2^sI>#@OqOOt;7?=l9O?ezFB z5uy0sK5^<#tk0~fsJYqqlYP|U_WbK*$6xNw4)+I{38cv!^!F!6mm^O*q-SWm0oXVf zXr&7$CQbuW+nN0Q`~=(YV3{)2(FdW8YXaTrtEe7FKO`0{a?uNAXuQI+H90XU?3Q4i zxj)Y)%c1*I2W-iO{y6*hmx@RlyqJ;1vmVOL#?L>CgehKqPM?1bqkltJ>Xt+3?y<&a zMH^LT9_}&Uw3F#@bFzBP|ZoRm!viwJ3m-a#i`${!M_XHoS93 z>A3NKsH@>kJO)CVeZL%9N=Bds@d~+)>9re<2Xhl{#L-HddD!2m z>>UiA*>jGjI={|BwGEi1EtPXi^#v!zrL!4oSt}c}Nrq1x#A-5H@AQoA&kw>un+0q= zY7O1VJ+mYo1$oJM_cQ!{`y`(Yz&Z)+Wv}toopBj@>dlFa`l4%`vA(u#@kL$_VO-av zQRmuu(+lPjLzDg)H!iA)Dcbm=S>OBs^|APEC;gOfW)+*Uj@+Gb2e>niGQPPMYybwp4@#GlX>iOT4@}a(LhG0?8%PfA z&a@XT$GO^m1|*~%5)O^^e3v11jiigdVDn>Z**$VS`WAZYZty)m_Fi#5$gh0ktqsq1 z5mj;a*!5*$Tusj2lOv9Hms^fQuktN&wxhF?E1Ehov%esoP2pBu{wPuO!XD^N?uLMwnjOE8@Ru&~ErCF(cY47{iYhL3 z+l$BqQOQ}<&n46Kmxi5kq__?yDc;}NXL*;&G`zHFV(;G`rSmT#Z;)cbx~G5y1sTe= z&5uWTmEpJt^LH&L%^4-g=@jZ#12RDgv*P-F!RrUP zkmm9fr3gQ{qF_>%;A>tC@s`i#u`MB(^iv5_95glrbM`QI1XoY^*dZ9>b`EiEso(vm z{et%jr!}aN7(lwsGf2g;FfE(>5B$SH(NmI|rT>+~@_%O}vCH$!MjfKq+qHFY{{3~| zjQKy5u9`I*qX~lB+Oxl!D)WX!U_|Ab%aDZF!-dqGkif)!h;{6DoCEgjZTOz`tDfvc z+XigDXXrnKEWxZDkqp*NOg@Pu8#~ub`(Zslm^BJR^l`fi+x!_7%`mSaQyw&k?P15^ zw&7wCLARe`m}WNHbOY%G*jxP*f0;HVl&b{z% z;-D?#h^PB=?+5h<)`o)!`;f z^%6BnbexErS?=i?BP=M%gtr!5@*y6J3ZB7__8#+1_H{$iKMHbWU&R*te)T&gl{VAd z>^78?q+qWuKZ;sB3P2eBd;8cgaOez4Tz`@O;Wx-8eZ*YnKRUm-bk-DND@P$B$DKpz zmHJ}x#pvNC+xb(34%+pS_&h?aUd$p|uTj`&?3itgR)ndDzPJ1?>jN)mf_9R!`!VX4 z5%6dYacQ9xbpc9h*_XJ%{&=?cd561!PRy)qxx|CmMd@12h>aWK=Gb`*!vW(j1gf$J zfR26^uplG5-C-WyMd0*P;aizf{~Y-L_#qISKC}D%)U-qQIkptkmD%c z4CG?hO39d_J7kKE8ATDYkJ=wRbkjoTD2xGHB51^@LMK5L z7RYNPw&Ut8xXGLLY3sG|_F%3K*knYCVm|;Q352w53!Wqbp{~?No^UVCUO#NhI#zu~ zjMh7DZwDxxxDJalfzyXt>J^>8zpkbeiS=o{Qm6X=)XemoZVSEcf}Sc1zy!aSc<@AC z)GER_JiN1wAoc)iTcj2eAWBs;O#~^k&h)RO9G7qEa?SV zIvH&V31U{_lul9a1hy*sJRwJCEGCNy>baRr7zA5lp%aU6Brs<>0dqSG zUT^sj59NNEI8=a(f80e@`A9XnYio(vpr~eJr%Bcl5yaNr`QfZcIdmu+Ez9TqrY3a>PQUhf!6{iPkvp-0hk;IpdY3@DkemdM|F$A>W z0pDP-{;||#dMqa`C%zaAkEczeX;Re8nkv0>;%qevCf&yzJa5Jm2(iSWYL=KZT^{@RtqqVVlO5?5!pe)Ae5&Z46`3G1Ru8W{)0bzc5xRjh-51{a4lX zcRtqOWOt6gTpXz|Nb=|4t|ROMyd@eEwFn}n_*I!kF5JiwH3VUSxG0rulcuc$MRBrs z&%Wjo1R@F9gOq7MIS2AG@>cRGcLGuxgj}t($Ue@y2LzI;`4QR?QuM;J!s4eCzytu4 z%EAh-cvr<~CJyyf@cLFT_N63@)J?hM6C`>RFwA)1`!3_$cstzT2tk>dSsB`h;WEbc zIq{%NJhQk0hdG=)|6nJ2-%;NU;!ALTPeM?i9L;L_$j!IC)ayqn?fxfGDE(JZH4KdMN_;0MRIy*-9?r5JQ3)s$VXHC zd0!DkH_P|yZE;nmn5$Wvt)yJcoZf%sq9_%0D@^bsC zpZ01Bf<>`T0vP!LQ0#XqGrwgt|BMQP$0B1)K)c_Eh9BFc`lPD*R2HKU)ugHx})-6R#OQ&RZG?5EOja2Aq{M!RMcYxUUkp&n|%pK zIY}k*l|Vf%^G9$FV_{_x;Kd|{A#)crF9+ErG-A%Feu71zpoeh?;WtuZvyzW{TUG;? z7gaffcYa_@gBj%?(&ut_d3xMn6y_%X8FoFD?n>6U|6W*17Q$}})WheUPU#O?b%FCE zMf+BOGYmNAyx)HRHD3+YLm>vwbrAdnI6k+j!#x}O+z@skS=D$hmJK_nexAQvF0bR~ z;0=OXC?^~HNv#u?8stH7cy5qOk64oAz*9A>SC*Zsk8zkKcZH4$D7|UR@`Jx#XAZ_Y{zWz=Kffz9EgR@Cf;g3PFjX9V zKLV4YSLDfVA9m_&ozyY=z5AUEomRDnbtPk)wzop4C2q(w)Z`@yXxjS;AqDX35QK z39=;FN%YaDUjqi9lb#Y$V*ezo3fEh50y~Z0^-$ihVcu5JmqYhm|Aa>dq(HVkeC3t) z%0;c~ilB7RZi25QGFQO>-H9C6C;qys=H@~5|>8uIsN&!ZA9=?~YFm&i<5=ZYR%_N}d zzuyqj+QtMoE_F5-o1x=3A3@!*E%qQ`b=s9rC1+_oBOVOK-H2^_L&{a9cu1 zi{p?gc63GI*A2F3T>EfL4E}z(5Ij;CLTI&j6K^Sp>J2e1HBMY5CIBIf<{F%SuRZ19 zsH%3|JN?lEEx1o?JcUsk4 zO7M5S9RH0ssz%fRjUMVG7ads;uGj)hBJlnF!P^)eMc0b2D{8#5&bbvp;9>jei3a*J zi&FR>(!o}oF34l8uD3>&MTG^`{!{lQ(K_o5d$Qg9I!eIgSR03ox=50s(3Sdh1yQd( zOb>p*%pwz32Ews|&_ z$b&mav!A)x{H`PF9H2i?Li=UR??dYV6++~GOL@osEr1BE9s@@%wLJ%eY5QSsA&BE3 z*PDH$Z7z@vzlilY&3TYDQ&H2Qnpy<&mU~$qwJ%$*ZyJTD<*I!1whI8BA>R^aGeCxn zj(u-(QjzduMet6AkE1zkJ04cgFRUv})u9|pfyu*FU=mje=2N30TSZHz^iK69y4xe^ zG&<>(A+0lXa`)A;`KHL8jX?3P_PxrAY*iFApt2;$Maow)RmJ5mj6^;Olc%~TO>JCI zw+Mqor`?q!JC*8#*COj zzyRrmUQaTn6Jxmxv@`7N3VctS>%`+-fo28!Clh1Wr@Pc|_QbFkAt#XhRfYxo8hnt! zW3>E{CouY0AMEDD4U>5FT4GamT^VK0Utxc7`yd8CAv2Rh=ema>Weiz10ahiqS?>I+ z(r=%96mCKi;$ex_I@Rk4J`(+K$*o@<9(lI$*mpcwsh$wERe4|B2&=3=HV#fm>)bWq zShngJ*UxmTt~&4%=`&R#CM!asqyw?KHFw|{hpzaYWP7z-4sH5l&Lz5K*7NAmr`kShyAv)( z1wqI@K@{Hlaw~b+c$pA!GM%&a$7Dr1&B&0maOIiN((^FLJn<@$p`(y|OV0&PZ5cW5 zszyE^(*N+h=}8go?eKGfF6>a@s2)iZgb)gfyQYk*Ry2r ze9~7{LngVt70`?S(o@G{2LCyVfEiu?kqY7uQZXP^^R-M6m%0VoeHHzQNGWz*{@)Tl z*8iRGvHoKJfAM19wP0XP&~An&KqMAO{AqiZX+QXL<%Pgb}kW zqb$b;8xI$;uA}%3`1iJU$=yPiI%iMR%deYJpRGxW)CMi!|}^ zsd95FM@heRhcPQD{Ga&4zp|zpmiFp4WK0_rDE?V3E=`J?ODNHLqf)Z7%ewJr7G9S) zlxCEKHFX>kxMhWUtQFL^Ql(p|R{NyNna)WR4~!O-CK=GA!lv?F5vr6SWS%}e%p!|3 zDYLnOoOW~i%~Q}RSfD4%tBJEUy2YJdPTrpU8|99(C9T0R6}yrsmP|T!73UAnHco5r zmN&(ukS)ZrLEFdxMuuD!rGj%F=20)**>f|RZDYz=42Ukvh9%3^rqO6H-u>QC87FtS zXUayU%>6@CyI-|ARcjFjrY)nIRUqY=VZu@REAK~gI3e_Yl^f|=Xc5hlMF&I+Sqr4% zER0!#8phc=#{lAE@!)tF^;(lz8JSu10{E#}DkWwCn&XAh@)FiV`7LCKNH3QggPeZZ zM2v1nWpPfYTNUC4`-x0z0ktX3aFsPn{PpsYLD^(cY4{69HPOo;DCfM&8}=y0 zJe`E+oD-Cm_ivbu^GWoyo2D61f;I(u+o=^Cn~L@j%*5oh*Py!}v{t11h+p*&`9kFn zbp8Oim54iaSP{Z}f1poEdJSXC*!U~dVbD)=#b&$s-`c!8Jo^5e5xQ3SM8W1ol~NV; zy-4V67N^j>br1`i;?kY&zNt0KGW{f5#)9sAe{M1>wNS1m_NF3G$NyDnPZz6&+FLq- zvl(y=2sAQ%c*P!07pUkUGYS5U+k(n%t_uiST9LeCMUvt+{4>@f z;&1ngaUob6xw*6m&ygmh?9-7rN6RVfFHCI;riLtU`(e8smM!qJ0$ZT}7{zU2P0_fK zxj*w;;ak^ds0M7w6)lbv4HDQv_9dzhQZPq=y+v?}K!|KBwZqG?p1C4>l>X`KNu9RY z69EgbXSL%Of$UEbtIzqgu#>Ad6AEF6Omph2YEvCBZVmx_Q#6+4&ICnTj-BUQvs*6Q zQ1J!XOb=FWdLM1xa&y5hYOBJB1N$0(Vf7fs%_GGdeWzVRmun}F^?~EJ`(-(@O5B!* zBu&S<+JN!5Ql}Yq;r5DBEir7nHG1N~H*? zo7JxvBR}J;f-K>gt3eyJq@4i~o(k&hSnqsEu#0(uelGUoXzw)l(;LT4jD^;f_%a(*oG1`g`!I4SR$C) z`~?JOExiyJKTVtlkyRKkjV9K^M>IJ* zDSWT1%mr~T;unMt2%ZHl^mEx@N9B@>!B>BJ+WZh<*{d%4^7>_L|Qxkc}h*5Z84?TC=cqU2jWixHiE&HM*t6t9BMS zNcaycolRwls)oXa!VmEb%y%m%;SyZ6)L8d-O#aH-SsWt-<-o0ek^6YCZ zmZ*5GcUpXh-EJ5($O6KQ=lg}>sh^szoFrZSl2xt8rqIW>aMu6O<*VBnjoktaREy}r$1>ss{HbQOXxyr z!4gRL=_zjU*GaDI0A+A=ht+8Cbsjv2|BUdsT5zZ?=WvF%fhm(CPnblcI7kIlS`7D%!r1(-Ql#Mda zkQoK`?#Q4LI50QLiyk1ydlPM*!ryD*{n=V)Jf*huv5l!=kSl)DyxFfB}Bld=usZ{gc0!I#Qn z1b9Lr^I39DBLs;$zSW$F@`M4dejwre^nI5L(YlSs8G8V2uM7SPt=9oLX+yqF-)^BD zJ-mHQGSF+__ z6P3(+CI)Wg1!~NZ0}MrQ%o=mn@7!Ad#%knv?_pTx1<3r0y3NbO2^m65uBmrd^%D9A z*ailB95@L_>=Bto847^~bdbLp40@6VsOGy7o>*t&?ELy6Zw6#5?8x7)M*ozZ)mzXU zVA+`U5aR!vn4_DNaL~~K&b)2d%tLP1L3tx=NZw}@Lj)y8BF}teC3Pu3Pjl0BRx~Ku zsk^ILQxCn|_AGurqG)zL>c)EkT5Ti8+6eu^m~t(ER-ld5UKhIG`BgDV3~8xSaHW@B zlPp3yI?>hXujX$1NioBazzt#|pth*yld68O+gT9b*;Sp~{s@bnm;=o_#Cvt+53VlZ zPc*h!DP`RQ50qI*eLhjV5JLh}qaih>n%d6N4Ky=B)*l*bAfn>-tom%0@&mKJ51dkTW$BjMAC;08O}Rr7=MGSgO?h{ z!-wemXTFHDZ#@bLFc*G;>9wg1xXcl%F*Uq_vt%shX^ z2(-8bF^B3)R>48S#IkC4v-3oNowQOVYd@4#nkZHxV?VsXl(Y8~Lv*E@iNuzWs-YTeq$_o^X9zFK6*L3xo2v?cx zrTJ>It&mg}pOY&-ktqn)qUk)d!59IK!`&2jZITkH)Tx5#2?HYDflxoh1520jG)fu6!n#5ISusu6c!JwKmt%t21^~GEEsupvl+?dLt zqiiUx*V6D5m)MhNKb)~zX%~VYU2LIeO3@0fJ(dKtWRqzY@G57M1EQmoNKC$PkSZC2 z@;}wsYxFdLN?9877t+*xK!y8X@8BjYv$^K7r_L^nj;Q=hD_Hj8>} zQC9tdI(G)7BMC3Z-* zKGqklTk*WtocJ4`nX?9&4mf{8x{gdMqQk*cvR%cd8;Vtm)UP)oqwF&V5|d@*Y!DLjU;aSY$ZgOWk!DmrO9Mqg?3<4V~_ z?>~vWPiPC)aVdqA^Ezj(xqndgRu!2MiQee| zSW>68n3YnM$^&TA{1MEJERI2(K@%NdoHI3Cx%7GaZfbNjv@0TBCYWnU~%w%Z61jRif%j2PNpTmh%C`S26e>{Y;izYSdzTc zU8HRU2-3{T?a(AlR{9B{U_~xKr?1>5oU3Dcw??ubbj6JSxf;-WvkrBPUVmS@uashrknd zDb{xH-uucMG1GCrx5Sw`nGG0jCfdo$=Dnc7ryN%{s>c@ z;i3BZ=PE8K#8RT9*zWFqi=Q0*`S^D^VMC;ly9F4^6FR9QxZc4Xc;n|U29aGVG9FAY zd%bzWx}&lXpZ1}7zyFPop9wrzPzj9EZx64*jh8Kh8IPYU8%*f1^aK z|Ai89aB}?LR^f`av@^D(Q}1JKGC7-rv!SqsTTj(7vZSJxl}Mzt(_k0cD*zB}18F|? z5^RF-&lc=wA^(0dd{&0)H^B0-hdIiD1BH;)a#tf$-xpKW(<6oK8^s9!{oaNXJ+ANO zr)RwK1;r1ub{ri2>&sjT&igR7p;#LJs}MY&6Ca{pL?SC+qEu7JL~BbW3=Lm{4~^}97TS&XXX_*@=fK9O1FeCvYxuyW zu*AWKCGlTKI~OWuNLY?4bqZG~5rq{2-7YX75i(rzsI zUwmsW)bQOUXGR$M7PqTJ6zf`3ZN*blCw2_EwjJ@xK=mY=1I0b2Y_1ER2sP@(#i!%- zj2B}f_D0~N?$w}85}|xG^1kRKSFS}q+eRAA8$CW3F&z0n@huNACfV#)4>A08@LA>G zwtJDGe9E>fg1}prP+H9wt1B3Y$tEi;p5qaZLY{~nN;cZl`|#H*Cd*#VE{)&055q*# z`}FoJdf`ATi9MXZfYZi?as#_^&?(g8xFA!f&o-SxfCEkjPc}}xHHx#FR7sD5jG%^2 z8fmJTJMjfYEe!ls-oeMN++Od9{IIq+sz1^gQ2tKys2Xh9LB<=68H-LFeHUKJrn=nQ zbaoblpb~IwtOpFqFNVbE?@lxa_1$VVO%YNDhI?>vy%dcXMD%>DKo@z@?2)Q;e{GSb z1Ge;l%hzJt_GKSo+3x*VSKFN|4Np{NEt}4aXBMmk!d#hH_#u|PR`4LK zN|+dm4VhHcWprZoCF=4j^6Q~!hk^zRhjpn!41~yXZiFi4Lx{946rVZ#9C-7y7KhLP zfN8)*l>}MOOn-ntW|Zw-_{lfSv)7$xNvW$qV8lZ<2FzTE1RM@js);>Ip+fQ%$_-uR z0XFj+91uh+XU0oil0Yn6#E@afj7a3bS(#1*y$r;8b{dEy_YC;B-9i?Z;h;dgMQcbl z7Dik!9pt3@J0knh@?%L^%N`8em5x?PxNXe#Gkd0mIkCnOV}h~vi-C@_9a=R~g25=2 zGVbJtYNNe$ZOtyx+aa?DnOS81XepgIe+{!zwr0+1#r`s}VbLN~A1jB+@?=bpHCkB-1`|A)|iwk<%@Z> zj+d$iwz4Ibi^3V?TUgO*@3F}Cm{d_!H5sM4u0S9cWSLBL*A8(kXU~*fUL*iQ8X^sS zBcoL#XSK$~;GUZ51(>8Rx$;#RKmG&&+;O$knkNn(DMFE*pgJ2Q=$t=cevxi+u}In# zQtk(pKvq(AU#TSDmY1HDC;AAma&}^tdu}xYJ#&3>*%h=*;?)cNI}-lwn;J+V%)XqQ zUufteT|3|`n*NYtjR3Z`CA8CL(!?4Y5hh3uyazSVEo`G<58&g_vG2?uw>8Ivf~=r2 zRf`j%ht|{CiJ5wY0w{?R0l#k!y|-tZs`+Pb{!%0Et$ck$gQyv+UII(sIi6a}>@r3N zd2Rot3X59R|oz`rJ@LGq8Iz&5!})z_0>8 zXTGPggAnsOw~TBdtHpx891AOf77R!wYfIxXxZap)LGXuTBC0-` zrZzRv8cXhXMcaq4$TJZ>iwcaT>epjueXKIJP~U4H^w$a!a<{$av@ozDHjm}Ak_G#;NzVCTX$(#5?k(!W5GOYt@BhaE z?p$R9T03t)9Mq-K-2yWOaEgte*_$)mXKk>FLq>${oh%Lec-mLhPHG)Wvss;FK)7Zi zEnA-pXn-6(vlajZcHsn(TnsD*ZdBPnx*ZzD^m!hcI1IfK~uq0?|x5b%wbi9@RNUXWv*FJdkieR)yvO;<^MR zm+B4RwCUG$(dM1z4+5sXYc^$#M6*_pFi#PylBYoZvaX0vjMN4ir{x^gsfKWxFJHhB zZuvM}sv>O(fd$7|2d8z9n-SR`1B7o0Igqf7%UBVZp@Y8UlS=b&0R!Y4@hahjO<*k9 zu-->q#Tmx^><3eZV*@hnBF%WC7Ljs}6<9Vc#W7~o)2B&0dC?6wbTq4`$~ncf+?~3# z+?PCT&~Ma#V7Xs9nn`PN(mxU|5%X7&Iw)Ui0gM7qErQAp+SFPm3I+o&AW6d2dl)_I zk_#FWIl{qz7vuy!Dh9k3q|z@$loSQ?WyB1%(_y+NK_TUjELju=T&68EPOHk+P{Gcj z#kf435PP6`Jzkq3h-hI=5%$vw4#UkAXY+b7 zY=N8F_m%^QFOz_Y|7ryBc7ii@KjP97CkT82&-NMpC2i9}ukNuv0$TI}W6X?~*l z5khNHXPD5GT+?4e4rn@8B2#(xQ0R`Nwf$Ji5CpnowcS)g(pXEzsJ#l(yVmG*69kW& zrUWiv;ND2x0|9T&Pz4g=c&9*Gos?+pAk+gI(&MHU`a>SleMUQbbNk zb|fy5EX(mfxIb$)+u7^rA>6f=sqyd~vt{7Gg$ZkcnV zvO+Ilss3qmctnS+mzTtP(r>-2u6KJCfBP@Ad%;=w+oJ%uFos(Ai7%ANa1aTZkp%j| zr0kVI5f}w@mcEd#j^EzeEh!?4RY3z$sZdd!x19P$vz{Zh*&Q@lcu}}}Y0_d)w5i~LzB&heBX5%l8cUt{GU&)coINC*GxUliBXJ|2;z>Tv*Tq$$=Rv?kPNeJ z^DqaI7f_M`4whCBmrr6M<)6vYe$8HDlDTTXib4VJ_RvNOqPItpje=_3-^ey$C>WzS z$#fC*yX%zPu3fPk=983p?fOO6-15Ctmo=L}LGHwzokf1}KYQQ4FC^4xp9c5{@&rPKLvE+iPt?mOKe4G(_Rj?@rU=ZKI6&PGm| zF4;k)c+-vCY8_xuG%lJT5PjW3+HBCipuLz_a*M7}fE%_e4+pFc7jTp0VJLBc3wi1$ zj9h3_)F#|m@P)}rXjMDT2NTqy;89*;*CaoOGFv&SUAh5_4=N|qHDvKOYra?vXc%Qx zvms`*sm3rqO`+pEc0Z#<%0^iT@tce+^ZPw~qJV76Biw^`1`dDY%?5?YU~T0$9?}#5 zOj!ZW9=Ucvixh{L6cz|@&pqg5S^r*gLViW)Pa7upuIw39&9lEbP z^@vBqo(>J6W|HlF=xlNok&t)rTwOf3?5;|b!$P}VB`^`AbAmF*GP0t9lq_|?k68P4 zKYCHan^0dmCJk~=C)XJp_MGgt$Qb)Tlx2-%`;lbI<(lg1cSyAdwGR&Vf|x9lH=h%R z!<+3d=6Hd&KAI%X;7MmV`ml@F*ONXI11k! zC{IvVr#3%2$t;D9J2HZNF(;fw*}>+Om2W3euLVSX6z9O!!y+INQslJrkaGv8CA)hDM8F`pyKR&pj!E2M{TpUT zaL2y@Wutf1CY8U|kxw>84Om7aih6a&4}b}v5ThzD5V5(20WP7{6B9rXfiL|ttT5`Q z3Kb@?`TT57K(Y$s%i&@JmW-R${)o~BNBOwI?6=1cf<{GK(C#87GvqTpy43TJ@4w5M zM+xSOL^2);EPLEN?3b>bHZ(x^tru*?T=aDVr_nGYwn{shkbXmixGv2uv7uCY9rTj| zsKk<>uq{U5y>0Yc=Y=3R7+6~ELk<1mk6sFlo>T(&ULuN;QK6J!qz7Pt3$n<)spIHv z#<9VLf?-ZDIzi~@2e@#cC`8jT?~GXE4M$iUoaHOW28^GdL(PS+Va1+1x|iK!Q@QOq zcU+DO8ta!wMMbZ`_gMV_Zj-j8se``CGO6s=$1j);zvb~BeMauN*fuSDroqyEr?j=w?CVscH^keugEOOWDfW~bS~H7O%msf0q19}l$+zwxy)m|Q)}=&@b)@OTxvpJ^M#kCgTm|< zI0;(w?bQ4(lR@_axkiSL>S*(19MD(p!WvKFl%DH0;gD1ufNQe`$)a8}$s2g)mx40q{~R7)xgGsSdU-B~fQB+0}`M{%vw} zU;$=1`h~mkq8SE}F8;}thRgusLZJFEurIB_m2fUN@@AS@aS2=Ln~vJ5wt@n1ir-xl zb!`X#xj4-45Cu=X<-~vma)|A6yUSfgOlyjW{c536fU{; zzTQjX!JGF3g|Pa7bVJ3f_^#xJ>rheoVH4X~boKJvJ?~s!c=n%#jV}9Hf&}1RIIfYt zmw*2-bIzHbJn_S;tr~E+P}~K4DA0Ou&qY>pL8ja)N{$=Jzrxi+(s!T z!B7eHe3GS=YT808Z9BQMw3Q1iGDu>RqyU`XeZ$}VYcQLf@vn5xQtWso;}Rm8&zJb; z{lV1i`9I2KywVr_S56K5V$sL|K6v&nUMPRl^Fn+Z_}s3f=P^=QC|~Ed$G-U8|2UT% zOq--i6-8VB2BJRh?}raAPM?rq83Py-$kr`^j%|;?FGaTmEBWDaDNGelynf~@H42yI` zSvf)%&`hqF+!XsU))BhrpAN?gc>`PIc5T_;7Ly#xY+=ot->^mpI{d1weiKUPVa{kl zsnD`D8N*|j$mH4)BkOO@L&_qRXX$_NxXecrI7)3%#k6M3lHaT|*lBMfI|m-@>(Q+; zQ*Bb~>H$YA*P8)W0_X-Cy2I|*#i?3jnASOY0KrDMl>s2FY%jFw4y2iL7bInozi1^< z*hw@QAjRmJ*+S{HN8N7XL%;^#VZilHp3Rkv62w^pT@BU4cvG&Y`#=*92FwlmX0gDK zfUzy(j%sHd*6Kl%AKEXoybAA#YXTw-jhVm7?ruXy^n(h%u(NZx8vhdih)r0n(n0$( zB^{-i9+p+!t#?fBDL#0%3T!RG!IT+QRgZz>SX(Yf~j|I730-GOf@`YCS`1+k3 z{8L)QaSevyb|2pbkQ5Lw66J-tSe&*D%{PVl^4Pllm0@pWJDLJpc#bz(fWBe=DcxiAc7l;Kw6>#_x zdqOr2PaeH_F<{9NnWoG*>5H>|B-a~{PR54r0+HK~30=7n=`XGzd zI-h#^Am7)3*n?P4W%1j;2zY-1FuzC|!0+S40V7WC*{O@d3Z*H{HOy?YL`j*1-bh+~ zE!VNhGD1^^i23ZroqjFbQWHJ%t2(BL;m6HQ;UUZ%7O5+v)35OcNCC8YH`ZzqU7H2>S8&Pn3N-fK+ve7HQg1T;Yjn7A(#Hz+gTKKp1P1*OGXPsA@Kh$1I3C;EU8dX#vBsNM|lf)9+SQ z+Bs=Dt6LA{08!0LaH~}ySD?sj#8(9OZGIj4?qIB&l0CM~7liOBXBkZ2RZ7DW$2t&Z zhU@B{Bs^O8G0YgSFh5j%70!Zrc!SM=rUhv96{AW}I{^V<)G3AOmcebBe^ByX4~#%M zmkNY-kW-Y(XoGsVF9!KEhx~xo{mP|*_OdgAKFIo>7uFkBe8A1_Gx?o|kR3#aC-3YM z5>TQPgQ&ns&oVD4Il29$pg}J?FGn{%9S$eH(P~YPlXB}@06ctBNyL~50iMH?LQoPc zrvK9fwD*K+`Xp>x076@(q^N@70-7e$9*?c6Z=iBy<;6pA4REYzqF0`;; z!Srvr9}fvg+>dR0AeKca_o5EDQiy}G-k&+a2poi?3>3;~= ztYp{FCF)H&Xvg}800dfC4~^y{o#P_hn@iE=#!-){gC$S*%q zTP<=;l|S1wRi8%)IEqgM5#7#izK&S>+)qb09C9%C7AtJdw?CG0(D-U~k27nENLM(a z3_Z>QA^pxA&olmM90~_#=y6X=mp>ZP@;dT`7hqX)3Jd(szp- zgCK#U{H@(R&HWX$2$@cWo1&QN^->Xiu_Z`47eEsLcsa)hCWW}IyHriIk6!3Fh2xF`rZc&QAl6Jf!7qYY{K0K14@=46^ohUv+@b*6@7h=?bIRQsKQn1 z!JPkh5$af3ABYiX_M7MjEa$yK%P!N3*{TC)R?S4$=-v%diw?(;?4{6Ra>PssKPaP~ zr<%HExw-aHaR1~3uTaD#nu(Ac|DxU@>I+wh^Be8zWNJ(Qq#bcfBocS6&AgQ=QH+~) zy10PVs&T-;UcPt<-6SSC}ldH>&oKYV@p(4c{6V)hDK#1>2JOu)OnMBGp1>T9B1B}>9WnVLM%DvB{AzE|2iqFM@!;2oLKL{&IT*> zB?3uiDV00*OjQ_QqbPoq$GTas&rsO|2Cf074XS$rYAkHfaADRN8k_iz9{{FN{sh^- zkjr~7Y&QNzL2~lw@|wZe8Ni_YRX8^b`#F0=vQ;Mj22K*qVE_lgLoK}Ep3_wlxYR@^ zpsBjnWhe3Bjdi>#J3gkO5q?Qqvu(hFS8U))aKn2;;uV%lukVsIbS_C1|Jr76uj$66Eq3^L59U4*Ae!RndV*kfe0d|#)QV;}AR>l+ z+TgNW+lxM#JcScQ4EHsVjG^L5|Dr|)-0PEnJWEMZuW@PXx?iAPqo4QQPu!RJA^Z%X zh6R+B%j(G|%M2-nX99rz?%l9@!4t_vIE z3T=yLmmPG)gkc&?knS^#C+UXcUqyUYep2C86EMVHHN-tY+O@^@vs0-wS?0Rp(^qyw zj6`4-9ET5*IA?C2ePs+S?`{gvjAG`jd&ZZN{jstN&lgj^HNQ=L1DUH9x!!PBw3(lj z*LhEIx0w9xmWTeRyyhvhkbCx%eE_q7>mtR zJ&=j!vXCRg*sLV7Am;9*>5ussl%yWt>${;KA7J-UspVz&wo7g+uKB6MJP%k|Jc4vG ztk2eaDh{1&s^yWDP*zk5yc<%;$g^BIEf(}b-qypG@aXG4x7`63+`<;G{+S3*(p_OQWg`U}? z3Gk$QMgkdOv|W}mx+NvPl|yx{PUYFcL{=3x8!UVCHXHU(}WKP}~RsuL~@^ z^gXHqf!itNJ^DUyVab}}$rnTQau1*5xd#vua^e31V6FYJ zHli*-heghZSh12;^f&m*k0u-77UMp>S>S!+0s-OaOV1G>Nj(t5_MxZhlJMPSUZ-EJ ztN%$dOO}f*$0lF&0nX)-f!=p{*N_r?d6p%y8vVFqdkAyb5FucfV+M0k7}bm4;S|(q z2=gDyN!E=!PZQ&@97MJ-H#;#D(-6`|2>v7^^aK>kz1JP+rJD3?fI`GClkSX8|3oMt zU`2%U01t$rc*}Q)2f{TI6pi)KgjB2^%7|v#CPbCpc*?pOa&uNhlXmT zMAsUa)v$sL;D*24Yh(&oM&>IB|8!^Zcrx&5zBJ$+uChLtTetiR*7*-m%qBu(tQRg= zVnN{xyZn0KY)T_t_djz{5nzlz{Wu7tLv>vnIHqtHyyQ9Y?x~;FPN1_adPU)Mg?-)~ z*qdPS^~tfY_td8-sjA>Jk7Uc?Mx8<&b*j3K?VdNtSRAtYv#%#)ll1N;evC5c;PxP0 z>93mD$v10nx3@i63>wVOd~n~#9xoa}zlf``BpC#US)OsCrt5}MX*d$?BkpQ|f_f;4 z`I>G`J(rl{l{TrOJ9~PgziWiDzx<^{xgYp)0H#PaO!ib$tv&X+{NAkqT75mzW-A|> z7ZcDl`e~0$`~wu=&j@_huCLW}{Y^o_w3)X}_{@#SFrk2DZtg*8x|(u>_iZ5=xasF} z;J+aE+GkV$4G^;bZ$QYx$;k45Z{z(35N?Yl^*sKYCIz-t>?#PC+o#x}gr`zOC($qs z8fznC8?cKQg@Bp3w6Q$h?o9oJb6!nV17XRsv%YK=0AmaIo{mkWX_n;0Kg!^pJ<)&b zR>wUYWfkScll1Y*`J+`GnX8M-?^^oeH14Nx_kK8_c#j_unNU%9eZzSlALqzgWwHy=W;i6g)DNAk~_X<0qgeYRvfi-cOJxs+YoyAP*n zJM7`;=)I|$Yx0Z}$1Qd$q+?%UYK163$n#8`1yg*!?6lsSGjp~qRzCdqY0Yj)p$U<} zzsp;$gSor3v9jPUj#0t!19ZX9KWNuPp7yWygf(E4%~?*I zV=f9&T$DE~uk`ouwy9$zzvJo9G;D;Ey7ygu&+lsL1>LI{mrJpA>Incc$B{djPpSJ& zeM#lJva%|-^_0Ptp-^j;q*kHapn84UB$YaYd*S(#rb6*!f=Lyw>n?b9?}fQ4pLD$L zL*ZVS0!Pdi{f(AtTG=D_8`U3L@2KQYF*N05Z_QSgVQK|K6pylBGmXvQWdBr8Na|)* zJ8)jIZv=YtDNgdrRC}@xWhRz=2~^=hngXYbAM6t4{JM1cZ&-=IP}JEr*K2}6AC-11 zB1Ua{7AhOvGh>}oZS^e9F=HCWQdm1rgosh(C|2{z;n%~4b&SopR|`aM`eg@m)U?b- zN7jOsOF_=~V!Ei=A6f>om%4vF8%v(;rS=)%NcYO;w@HzjCE%7Gqrj79|NBO2bY_`A zF}dy4s%K*?Wk>FFzsLGq)GvKp1sC?FW>)JFqh%5oF4Zd}E-hm`QLM(oPXzyDQ9C**7-s^cGbO9>~_BOK*H%*)anB)sH4m#xyRIAoulUVvo4 z0zyfJMpBEQC4$-3iOnbi2A*Ik1?7E<@YYZc+0uLK;{veVM9cj)MD+Tg4G2&H-c~_d z5E^1GBB5N#RN$#I*PbZnOpAk}hV^Wkt!Q)Az2)mWAR=Eq#DMLCSV$0BDIK*uww1qJ zgZMrO);ZYncQ_&04ax*ckw&%4^Hq2C5!)kfZ}-^U_J6|{KH)TKoT!#Ub2MrUU0PAl zCJy(LbenWlLCog!Sq3R87Liln#rg83Xco0YhA}qvd-!@{Obq?)-b$dIxZ0kp1vagi ztgj!YrJ8PgUhh&=DFWky5?3;P>~!pQwtu*JSe*{B=O7zOjAOpS404w{K*W&TRMOp>+F}}&A=nM6T?_qrBPn|lhvK#i{Zjo4bP>8&K z7)Ls7nYU!?TSOCXxD^{UHu%1O7Z!40FI{4#`jXak2!rmMP~{+Y>8~xUuiJ z)LT}W>uR&ha@v^)C@U4TcV4S?OCwDZM5~Ld|2`*kh*FyxUwym1!8&u6?S*lqz4h9W zY1#cP7Y!rY>~o@dg6P`DWpie90q5&`hptM$m~6@*VNLWl%5s-w4XVivqlT4>%A0tn z`x~awY!)x%p=U4A9rJVeesQt%7UB!UWS;3Th2;>=wn%-cBluQH;CXeb`tN@g2WZu0 zHFXU(@#c*P04NLoamJ0r`_YgWnL7xy8bK}~u5T0$L;!;PKspucicom^T&wOq{@eFQPU{h7$~T1(uv9)(M#$-dt22rpg2*H z+EVl6SxCGC#xM{TO$0g$8VV#+g!|W4!$scHiK}*6S3C{RO~`XEwl`-xCZLT}Q)@{6 zs6smC`~HzI3D8{{pyrQ4uc3^K%Y(Hsk?WyPFn|AnAnVng3p^x-Te0)^GG&W^BM~{!YTN0lzRkntM`A`O= zZ1}1JTPwt=z9NrhZ$;09=R6dT5u8QBf^-Hchikv|^RP6se4uzDAs4dn$m4_AbY7-c zQl~rj?wZST?>JUOMvBOslVzynvBMtl~Q?d4&QNr#y#~svgz?AK8#}AJ) z2~ZISi3gK0!w2&E)J`X#L_^7FzM7nl3{rMe$9*Mdvwo~}iijFp-*lcTYs6A%V_?Ji};O zl#S7p7@|0v{QF&o)^xQQ65@Tc`F^igs9gR@IoDuf|-Awba-xfO0;103bKC_S1mK= zcETee8eh^r%o7<^*rl4{yvZgzxAmzCwk8aH{5eipN|78{^6)lNwoe`1AHFUXk+)>Y zAV`UusKd7T5JJ<^DD)0a@>DKH5Cx6Oz}Q>oP*eCTKyIkt>t6t41<&!6K`UsZDDMzJ zsD&Q@lnn3!2$pb3zWnkEZlJ6K*shtof9pIYAf`0A`)Ajcdb@EOPpMRKl;M>h!3-fF z*l~Pbb8&LgB^e>~x0_C?^lzZdH%{RyL)1o(D0G1;x_$5Sh6mx)03F+^x^POC4!T(# zU%AlNIK`n|HIW?YjLRQ9W^?GIJT9PXTGdO^8Boa$&j5AG6e~rGx=zu&vmvlo8ISUVOjTT3%D3fP*;v~a$x*(H( zPUH<1!=_@8#{@ndWYDst)wq;&KB;59Pd#n7Y%%Ea$apC@FCx$PT`_{s!)=$gON?Oa zZCQ3S&SZ)iQ=JeNx5(5?k!`cqS$l|*wbW%>A~a(4jFd;T+#IyFTUiIO))a_s-v?U| zj#|n)QZY33wqU0eF|Lt~PD`;?g8@Gq<917t^}k%SYpG58!#J=x#FEPo>KAwkX(<5c zk0RJsVJCG#2DpceC_brp0l+?Or{D|Y{5^4->83>}jv1_1=6f-6$hJC9`Y(EhZ>;5D z4R->gHY#Y%q@V+9>Xc?k2%k$4JJAyZ39c27TJ08OX?xY1f)}h=V}6 zu`%GlHe9)fJ$ra>*<8jRS#z36?7{^u^ng7jsKL(2O&KQdax*Z3f4&kqBG?2*t+0#7 zf!pY%zUV1cgvL!2VS(!O>JdQ12dpsU37^w>`T!8`h|FHlJP070tE|JS^DJd5(tUz8 z7;W=r+s_hqq@c}kk)0i}jl)_smNMX|C18 zoy?u*XX@VFVCz@{YlFClyh|oh$rVGnN3__<&&roDwcMndMe3G<8-C~4Y--fGbN0-`y%{l z4Zxz}SUqObG@s6pM(jH3noBNq7B0{(UAF3zwRMHH1LgFLkC<(b+7nM=yk`qs3$yC) zEoceAX1F-l8=?xS3~u(@9FwaoD{O*L4thr9*GE}S1NHYYNZ&(4J}Ok`GvLn*K7#>O zsnnGxlR1+CQDEpAIE;svYlO7Oh0*B9+g5*=`cX^pWh1z$&E$bym<%d$@4p7!PW2nL zirZs==n`NejB%i?(nYR+__=uno6vLf8}fsYM1^+xdC>bsz){vy4jMl#9&F^p{QEI- zLbp!LcZ63+zoY#U>=mRQ{*CGik|9%fUKOT%sAv!x+-2;6`95!^wOE>|a!myu^1p0q zNf`%|=Kfy?kNm>#W+4p{r+!rriM+Xl$ZynXqlyPLWrqhVaN2u8ojBQ%u4}9fr6G4j zn{I^X7N@Qq-R5PrS>eCk~ZO$(GPq)eQ_o@K-UryDA85(ronZpDKFQ$3Ra* zJbAr@_gf}>F#6uW)ur`bifoi*X~CE}uiMq|mnCRH&+fI+6W(6Y{L z!}8tA^^@2bIS&lVm?xSyi`;FO#VQQBlefmH;`BXvdmfp|LU@5F_=pALgzl@%xoVan z@;51+-5NDir&-XRNq6?6!(bHhOUqyks#?0e+9PD!8!wlSF5ToF;Z6{eq@#l$NVlax zP#i9gCBXasu*fCu+P9DMCmkT>#J;mI=4SKXCX6PfPByc*!mVam#*VfJ$1a+1G0#q5 zE2o*$M`N~W0=oKJ!hs%8Ki>*;7yRk*J8Y{jJRiy5*~boC0}48%yNT>5@%(!eIm||B z_U8J-ud6E>OiGP_`&YRmKi5`zV4{Rv&j&>J{)@7Fbb>Q@$l1MIA|L`d}x)Y!svrGc=*7; zq#-~?;?G*>PXAaYk4_$4oI;r6*n0vWFAlj0nvvkCuIDf+`0X${Zf4(F|BChZqn3P) z*a%t)n|t7wm(C9){D|%Dzl$>%3=ty^3^F&hClXq+LHNhz$y`u@J~#=QtpkmPvEVs& zKuX`nZN#OW4bp_W0xk^+A~TkO`N%4S`O8lZl=zmb9tA%L#AEbF4~Qb!!KOgE>+~rlva)DNe?nX zTJ5u~9e#^TJ#Lj+HU8M!du8F#@5kpPg4X^t}^;J0^-xnblKv!j#C6dwPIuaE{k0&NDHzYwEarsj=`BFT+axWyPq%y1D`%#ffyvDb z{f#RblQd_2OC9rU5Zj7Vl2+;l_U3ocFZYVNWL?vfd!sB6r)tYemHl-&!EuPMzG!Ny zv^(p{`K_WmfwL+~l@t0uCpP2AGO*_+OKbj~gio00TO>-mxooLCzEK^d@5_JSCrQih zxi~dErr~-dZ+(KS`u76sXMqM|C7<}&HyF3yXDILFywAQDS(6xu8ZdR>=@O4HpYhf% z&|jew#=LEr?Db1`)L+3)W6>3xV`>O`@8z4`o(HNB$q;|^M92YTn?hDyUhdv@m5Rc?7=VjISodMsMB4rX!A zytL=8&F8Q4!uHa0&e5#<>EBIb48mJ4o>?dbtK6dl;pkw3n-0Bz z*5afgqc=(;rBkbP)AfBut=+hMcTqrq0%Wd~y)umG`wzO{-L7O|soRUOlfLh~=85_{ zHCbvBGLrFRpqEs2LNJZ$Ck^Wt?=n-Ncfp9a#txA?2YRw|N^0wg$_&sl>&*AZM?~sR z_Pf$nl9yRUEHDHUj^-~(J}aXCKKQ}qj2fe+N=DhC#$P^P^N0knA0J=@TL_CXO1QHd zBivYOr6D-kb^|&WC_thmLn$cQavr@`R{RrX9#IT@U`X4iQM1q<*8!?L<@9f8z>le; z&7=s;c)KXK9M=#vKS&@L3hQClSTkB?qn{SYFTrZaNHjlDb*X{++}n_Yr|M{iMgIb$ zxUeyv^UZhJG&N9a{E;;12z-@dftuEo`U$LnPZlm4u3d`Cgb~dOv7hl843tj|E8nwd zTH}%X7>q%PdSGxQxY-Ws^TdTPRZ|bI2QkS1+Ba778{WSc>StP$U4@62vF@k7LQujC z_fUhY!8ChCMP-_F5E?Sd^q~hgK%y!~(8~5rHfK}+f0Uh5cO^`qZG#=VW7{@5wsvgW zwr$(CI!Pz#*tYGCZS(XU_wjxY=l+NqRkg-ib8-wO12|9jq!9K_&Fckx;Xy8)9gblG zJU_7~&7GwRKip>jHu+?rKeiW{i806+3SSzW6$|Z{6U*^hf4w8l7U0FB+OkBfHx=}1 zEjvb?syem}>nHV>UVfmyUP0o5&2Dncb%dhZXiYBzv3EcV8>`;qZ25Y!6(n9xScv8F z4%h9JHc8-%vFFWvVF*oT2k_oCAh?{Pnr-3@tfJC|x{Zyl zj{(3i)UziS42ku6kZz45(8Ou=!>m@gCUV}ho@rg2UCD=NsW}_+Y#>ig8*SQx17S?) zVvu4~5{ML?E)p8twvblG2UZ1=ZOp{%A989j0XA>FXL#xYZ^9{F74>5^VU1*<@dIPJJiVQ~ z2tZxAIJk-|$qP8$r0?Qy0a~_N^Q)T2Ej|)c9R}@2#hZ`_WAw)eR zIbih_IaU%sC6m*picVVZ{Dj+o9H>hq3~q4{XM>YsWBwj~C9r2g@?kvL0_4|DNC zK23C5@b!~keF)h0Pt3}6ugZ>JZtb9RcvpAB9+Z!fYmF+BJioY$$2D$T7xog``vaGR zrBxJ>3_j~@;6REIhd6O}3>a{klSpc3)Vw*UK(X8MF$hulWiXWbS1YIl=LS6Uz;Q4+PRrPi(}nrIcbdq_rXLv3Y3tPMUd5DWuIF*Y12yN7^# zUwJbjC@(f@AXrSWnsd7a{BZ7A39(~qJ+rgtbZL;h+7FR3|e}hrLSTH%aa3n z)-vsCg5T-fR9nvSQdE+1M>!f~yg(w@O191$k`Vf-=qhQxDS0-U>AG&|%rFiFlMair z)v#{`5D_wE}3;TKN(_IAPQUmN42?B^88v2|M`3egT3?r5${o-s}MLpoKAR zxQY9TQuRgwHYgcn@O2rJsj{fb@rsc%F)Q80N&Czci9E{SR&o1oO(@Hf)w;$DBlp1h zhF7wC$Vd-u0R|!78pIQ2( z4qJ3a0@%I8VLf9xt3@7waNl3xLAmMGUnAZ>Y7i$i0g)F(CH~|lg$&arUK3nE4UOi) zWi~{ZI1+4r4zl$oH_w{rhn@*lHt2Qtyuv@eqx3xlls9bGLAxiH5GF-g$IROMUN(8_ zW_Hoek{hQ$d~pMK+D_^sNk#E8wj9- z=PzKg_+Ks5pu6d(mV|BGyzBtwMwPe-xaQ3I4%ke*G7J@Jsg%MB>osUOz&+SPBy-{x7sUf~S&C)VO}qf}u4V|| zz_q@FVB%|!$?9&#Y|lTuC%Lp3R8U#^Q3&p!1#HVPy}xl=C*X?A<(Zy4E2M&Ol?DdV zT2$Bmkl9{Q6X=)g^>Ww`L$oqs(d3f7|F&8g>J(8Ut49C=AbV~*jMQqp-r zO1>p23!L<(>SaTsO)gkRt!6cxBD;dU)qOEkR*zVJR{cq@CI)SXwlOG};6AU9sAM8? zS%R|N?e$|j{wgkQeBOcE+tbM@sD~``))##EZNVxH=OO2>pS;ZGYH^mSdIQ=fA>a*+ zJs8$eBO9=9#wKshm;M#6LQNbDQTn?)5$B^`8U{2;$cUw1=IcA~=ousy&fDSdJ)b#c z1s2_?NFTH@*aVNml;xc1$h`r9XuN$tqMgE6T?)F+(SyPDz6FVtPar$Lh!M<qsLyL<5hNMn?bM%qFa%PN{i8k}^MKbQA?`49@UF+zUr zZe7R&Uha$8pLRZYP(V#?^{9eD^f>s-(-C~NdkAb>(Uqy1EFX;2<6!grH z8Yg7cF!Sj(#?zn2!4ust@lQ1Gsi;+f0TS( zYo5d-{JZ^3l(hxl=6-8zc9$GLOqgT+JwIwFfF5K4G0_IzGA=BRA1yrs*ITrPIyAI1 z!!F>9Zp(6h9apx>gn}}Dq)E#!p-v%)-LxZ|qDJaX*+rQthPT6w$TrKUb+4xNVj2vT zBbzB1XoeR}TA}hBq_`Hb=(+sKnDAWdVN0kc-ppDl+&(e*6U4PEv~L^FtJVg|$u332 zuLSbYia#%QcKHua`?Gu4Wi6{7N>ksk%zR)I99=+(F}y_9rQyN4GSMx3{6*K)biPur zzQvVD6~R(>c|twn^h3T_MD~oCEG6gwv?a4=v!8a|i|5oWlocoMAnqG{u8Qc4P4cX- zb*t)y&j3D}vrfT#gN}1!sQT^;wS*a~4Nu2%nLzwpWN%IYZq-V*rC`pUd>G)tzDx#R z>Nw}kR3O;A8S<{ODr*`^$vl|?GbA0wI*!Aoh`rSk28N&yKKL8OVuQm0`N+!n-h=RY zG!~B<@_1l?TP0PE7%_jK=38-aN3i3Q|J(u9KW#cO!ht~~aqO+ic#ZDcnwyqzSQVr` zV(h*z%%loKw(mUq4id~=g8%^nKV*!9nHkYB--;}1snP;)WE)-GhYHlBCuYx0E3Cjh zlgwK6MucmE{{s(*8+j^&7p-5`k98o;m4q77GiAmjb{Jen9JFNa8dDmFBcYb`$NvGE zA=|>8P;Tp05IOSikw3Z|F#4V}2}Qv#ix{frrfxC9$pHsAwTmO)4Oqq7BKuY%IvOx0 zK5jLalEm%;aGuRd7$)WobC|?0;OMZiS~W2P&IS%RHq;{^6EgRTJ`99*{Iv@@rYQ|{}bm8raj%hN4@$QXrs4;Sw1L$LOvtrF~k@I z@PaSrR?)U^-F8&zaT2*8s9O?}tFD(r_*rIzN~cMKE5aKp zk8>(V;;^J#la*vW^qsQqT;U_-xhwC+z-4aYl6E!beu+lby)f}qH`dUzs8{Vm|vFNDu0NB@U z`+qU1@CF_T6UT0ToF6*yNKqKGunG%W!D+fMSXfv5i`Z+RH>Kc;nGTzNp<5M18W;(< ztgD}rCv|D2%kXGR!lXwNp^D!CSYCtyLC`FyeFWUbW-O&|EM7@{jU7^U!ncJQaueNE z((grhiN?@)VLEc6y9ETgT`Iz7N6nc=`vLH|^OD(C3qogDc3PNnlvV;U+6j!N+K;Fm zdWgcGuyypQ?&C8pw+={oYnC*7W22qbY=Dy>w!x-4lsRrrtFmHpY7weqA`|Hb>D9-i zEzJzzLx!O*rA?>E17j6=z=5H@E0+{2DW9TiM}tO3S>fr8Uhhh?)dEI0Vm_w)`__|% zX1EIk(eIrC`fNpwzZv@e*4l(2-gnXI6V-7Ypl@YMRHg%>bnZ?uExI!scoEs)ev-&1 z9ZAP;QvRY!cis6X93RA{^-A#x&o?@w%wudkz1Si)-%S>3-x^Brp%^pUriQFY#z(Do zX5XlYDR*?aFVL}4kDf)F_Cv^At(Z;vHLFo5C64v zhf3PqyWvN|UamG_-bQrmZ^4B)L@OWC(IyzNHy#3X6l2Qy3RIN$`l}M40@Fh4Pwdf` z?(1$(VW7n6%7XV!@&Wi~02N~C&gy%^dcLrIL>S3d)5YG$XGO~)i}ESq&ln!&GPW(v zd54sglIp{R4qMRq>OhX6;!+m1fvgg9!_Vo8)q}P|5p!_+RRcK)Y2}z)YZTwAzaA=e zBoepnp0h(=1c<5eU*|t|6^J3of^Q3`)w0(If`hF!QXfqrJ=_&S&%iuHl&7kU2)Uu? z|MogFAZ6WE*&tu9?+oX=nNX|;6(D;#Q(O~co47QSe>5W2N1uhiOiX`+5=LxEh&=wG%5 zY`94Lx-^*a#hP?Q!i_&D9&c*K-C7+2i-y!PISd|~WbA$+3Hq`W03P^7eGm0SFX^@W zcRVlTmFPe3=MPzZA1-a*!~J@0By%3W)c4)oM{yD`i5yQ^e`b(r`tz<{&FARkcz1T- z;Pwa(e^<<`MM2Dkv|nrYG$?aba9I&9Q}bAPM6eU>ROXWuDjRvereC*xC3}p@`0JlC z3|-C~wOTdd-8r|Ua+2;3o;$tPWxU_F-iIm2idJ;=o*zd!6olgc8NVVeCX}f8AcwE4 z{1c9H$p2L{SK?MYvUMEbYoLpeZ;~`;Ra^vW{t}q{zvZ@@Ziav8N9+(;q%ho68j))k zpFhu+1e0h_0s~h)Yr!o*50PI$o#vkP#>%Orf_MEi?{s}WQW#3GXh*ZV7?-RNx&hWV zJs-x>M*rL?df6;##<(A?YLI$k(VU)LFWmY0hpzAU?H(bf1C_ngmbdNKr9&GYxzdyj zG^_n|QmL(#{@q(QSONc#=iS|Xem5zL(T#pF80AaG-L**`4NCE9wK+gGBGSFx`fi2| zo)Kq9>Z?>#Le6tBxQ?w)lB)M+om8Nlz+Q@TsqT+?8-IocvPNwA$2Oa5GiT$!<-0@w z>hm2&%IQFDhAd2`%rG&NOb>NKN~!Co-waV60F5xp`5mqzr&EKs9YLYp@mw}ME(%S@0lr6^c zg0LV^KMmyx>R|@OpAKp$C%ZJ&SVM2%V5P`=6}aY$%C7SC>^DDc-7g!Z?)RBZ9H${| zjDyefYq1*_sZv`{FmqtEPD@O!>`}300Vm`drrMinaO;@1I$({LyGOo_ROvF!6Kzk} z8!D>%zyhGfb<~x(Psf$to@RkBH6|fdvPjU6 z9Wjh3Ff1f4;x!pGEcBbY=!M9+0j5qq_)tNVrj$4AP=XdpbKsIP;XneAy1tp5C;7_J zs!~A9O^7nn?x6QkN0w>V#AxhsLzJ2 zxU1+%+xu()_)(Ol)bqWvX(=N&R-S)8m>*Fw<#+1+r7(|(k84>sYNNA<>!oK;v}Bg( zhe)^}M5cM!76hHWje`W6`?EEp#0HJ-sJ>~FMg&qsn@-?*iP#z$N@GaX)q_!-Hkfu+#pUC@LY}FO+m_2)h(b3lw5vx2Ky{m*ojjp;!<3O$YWnszGuM z$*3kuxM=c8jmAE&vVgQAF(!Z^8QPKSOo*^gq|w)9xdPLmV~?!dTmAc~>!r-&1wK3$ zS}qoyb1Irkd~6P@x2Lg7p36&vNLyooA5HYoHn;8u3O2LoA4kZ=VY(j-jkews8_Mgh zbjsXrAuG@lC@!5pk{n3`F6lj)d2mujZ(ic1=w4S$3TZ-+03N6eDc0X(gd~b;WgdbJ zbNPZSs3pqL1sEN3+G-lzn(8~upWc2}yLOW@{|CUV(o}gKEWpszV>^t}8yvXfAmSYL zXoWExNUsfM-2y98TXJoq+sG8Zhnl%zf@}44fQC|U4A8DFjW{w#_?0;s-wglO;z(mZ z_T_-#5kLK}&KG7_Ot3Mg!wu2APZ2@CLChhd1-$-|4HFk}nb7Wm2ywFhTaCD1Sm-pe zzMVMNewY~<6P0=1L!m%2W?>LgPI7%)PZFx|R`bwXfreuIo**SEhb|ZJKspTi9m}{c zys3W%&gdxp5EiF|yQ&Xg!J8w23WMO&9xUhwhvZ*YAh*8(={Q{i>u~W~j=q4io5*lN z^gA?dgPfDAfbE2mowVY+Zz30ukkGg$-u`SC$alrZLVJf_11NP*9wgWejT2CnX+QKe z9^<~0Bc8uEZkVSU6qJ?z7UCkRE_i4xXsm^Zdr|#Ogp{8+0wO)%qdy9{%{N+akDe3d zzCoNV+jy)#vV%`fL*5MSkauHKs&n@16Cp7ZOhyz&z{{$M>@-1`l5VX0XIcqs)LC~u z8EgrkX3%s7vHW-BEAa*5vD##jk#7#+A#H;BK^z^I9zm_-?%qWPoyP_+_(#W{CJR$j zcGU_}1*X)fn^R~xu`~BY85lQL-kGkWuDWi|WjXwH*-$kQAyIt?1&vyQ)WqG)deu4% zP=hq7a2fSC+XrEeb`{ceEWm#~aTV7&7##Cd2We)&Py|g$64A|5NR;H4a*@eQ zNP4~kGrFT0$iVBNcXyu(`dFP^SGAQUUGJ~ON#QN;JAOl_-LQc8ANc6qDA9SLejib! zvnQ)A*E4^ai1tW1xcV@gMEWNSj)g?RG7i#S^%kna#Od|Y>XmKv(u<#Z6n?(;F5J=X zH_v@?3ABB^3s8Bn))Vgx#_ggEy>}3D&bcal;PAx|1HJTWF->*p!^q={c)|(!+4*8_bq>*l|SZ_y)I;qoE@FHdQdslV{ zTnuZ6awebrdZ^FPD~JtPXZgbTuBcW(YWqLq*7%Zly>X!KMlDuwb*&o+#QY3FxSCB^E8gP-k4s2ap?5bQ@AZTJvI^8zbDvQJm0zN{3G4Z{) zTp|&nUdP(XHFfLt}D!uLaYOOb-R$x0O}AM z6_!(RQR565mTKm`!H#=w1`o}qp>~Tb80Q=60e!3*uNM?$q05_X&_*7hIB= zMh2B0-{REe#jGEx!RHgnoFh9^XcDv6x9;-T6<%8_@Mv9af%F9~IW3NV$bjdBbF#=_ zl_c@ESp>;e8JAOrcr{F@S!TM79)s^*FdceB`W0MFA(-uB@BB(1r#~uVF<`Yn-Bt^I zx(cmmdPI=dbfmY#E(cYfCoQ;-*;g0@B~>RsA!`D=V0jTgLBANcw*$E4k6;7$eUABxmww{wSfo8r!=C zi-ZVEgRZsHF(sPl=#Q6K)9F3ncfC;PGejSf9ZRcVHB*-F3($=lb)6LRuZ|K+zhe1^ zF7OX&kf>#1s<=l9ZJEwt0yPYkmkP@~z?;X6fjAVWuv?bnQT|knv%H39pcn02g{+zF zL8L8Lr)&}KWinpDrg(XdhR8-NA6zCA&D(0(S+rJh3uJY@jI6!8@Iu6c7zOFf#q!LI z(p0DUW_q~iX3~7g;so%$pbld@z(cE#K`;}JFZPV)E$Il0?++FP(h@DyfM}Lg+)_Uo z@xfkkOXSev5pnEj@`Xj&vDMoV!V-aAT9DoIqJK+XqZ(xQ)1lgx>dv;U&6b5zp-3?S z0WM70TIc%>E5l(E|j$f8MHxjpozdj%W{a1~CrSJ;De|T1dei@dZkeR|3 z&5?=BS*E}mq*qcXf$Hx4PzLKV|FmZXu{>F{$j8}yE$fZ%@$!reVzfxpa`{D-9X!}i zt|2n>xnFVFV1-lLcZWwu(YK`gko&M~vNOfh_WiO@=8;oIbYs$ofPN+4WvDJA$7nUg zox-bTz^CqncHSxWr)JrA>ywznAo8)$G8!Ereo1pk4KLdEu>E=J0a--)lAu#vJvLQv z!S5J6)0?QoC6=h@Bz?y8q&Q|puJsi6F#@MU@+h4_@A#ak}glII3 zaUd>TCw-{ZW^UP+Naa@_s4@q4v2WC$Z#vwGwe!9dZZKAAf8{|P@gww7He?O}xb2^y zynNUzDK^aq)1)EY?kn@Y=IN)~#T*oQ47XS*pL7EVmE@m*FBcUeENm6oCd;TCmPsbQ zV2ytS5@w70a#hrxXMhR;!#(<}mB?k%B9N(017!Zy%r(E8&Ytk-GZsFtuc``qxcFt&;3P57Wi2 zvV6gyY^*hiUzV&T&l?R%GOk)+;*hEoh;Ovyx|wOs_)?$9niu1e_?nB?9s2^13-2WG zjgl?vm$vc)Fk36ecXJPnAH+3m)0p==;VE9RnrY~m7F2NTM|g0mP1)24@n4#do5RJ= zY>!l)Xyb2iTFR2u-n|Lz#WiB`*PXdJXEPR;i-B|P55ih-hG(K{G4imdx}sX^bwm@Y zAQ@Jj?y0OnRc5Irew1ZA5{VNahvgo-{uHI6TtL^X*`A2Bf9$3r2v|YK2CxNxe@6NB z+Mfo$slFMaVcCy?%7w|PL%!9RQ-UK`%DG~buEz7H3F~D);^4Y-9aRJvFI3);e0-U8 z641D#sOT_!`?G%aSPg$CH@~Crm;U-dAI|TqGtF^t(#+=FDNxR%S;?Y#8J|_1c2A@;W-A;MM)58or2wJ79%U^*IMCM_$Lq^dEU0`~Rv4r#0rbDe1Q5 zJd?_)vzB;FMLZY*eZ^Lzvq9`Zyy1YK&^20 zRDVAwEK4+u`Ml-6Bl2-bH*=sF_r3SsxFqHGbV{>&6l+Kc!e^AjnRR>Zr#_2=sVOw7SdchKWbZm%6HN5!8^ zIP+@M?2QD^Ek4b>O@43jA94VU{mB#TQ{H>xU2I=8!cw~adX*3T5?0IlK8v()4Ps1IC*5eNWE#-p5Lvak;_i zdn-=g^)%UG+mJZdwYVD*8`znDpoQqhKYcmFFjlsRh@X0oe z-z)7qJWF0M^{QZt^cbpv7>*Q=y>aVnlVqEIXmtfG;VpfU?1+E_C#|OJEs>5IZ5)ui zb~-GOB0%G6FMiAA-9*SV;f1a(g|e|jU#N%00@AA8noW;fQ~xcJwr#bp75R|Y0>9lj z`s&-Q4~y%1zp`UvWg0s#4NBkUd?YYc4cj5nGFYU5D&2g2wMmF69s#27;YXZOcn;5w zY^O*C6&Q2?4S;q#W#X=TvqX(YD{I?}pYQ`W?c40;*%t0tE=*dvu@F|CKznKqu7xp! z*o2uEE>%-QSnF(QPQ!p+DrT0CmR3lf!Drw1(h#<{LOG^huyX*V49X#^wY(wR`b z4$pAkccW&#Z0ipYPsyj?v^QK7LL|ZouCCxcL+Tz6A!9jGVG@pI7Bhr;6?N3COekBlwE_>~y&fY2BIoUMR zB0zy}&%TR-CPAraF;e`*$IpgMgibZb3H^q7Y&pltum&DuAyZ!ZtzA`y6BoE3a)U{o z>KyT0wKAATw}Yx!j~%rOw*=!`xL`l15*ajKb33IJigbE7`wr>$uJ7{LO&*&vwhWg1qdu~FXwDAzA|CC^hu%@l@>HXj=CpBirg-ZB{O#( z&gH327Ezd;!Oz&z&~TgGyL)W*BJ?Qhe?1_8&xdgKB_^Mo0Trq%?LpLte zfE*z1G;&(o;1h^(fmHovAin>vyzZcp!j!{JQ#$QL+~I(y6A?( zfu)L-zNepEk`sZJ_caC1qI)pylmxFe91Yw)*{OA8pQh`bEyg8crqW*uHc)NjoGTFs(DSXJG_ zw+3YpJluakk3qsD4_YdY-pLgSkah(fC?UPi2(K{E8{=6s6(!2u22j1Au5>T8;Ge>< zY}IXxjGihohEzAI%QV|$XN1r}Zl~KlF&zOZK1+5&xgftLo^CBu@B;)=n+F#7+@2 z)~53-U3U`F470;YA}=HeVF0ZSPDTLIdeN1`&KO-tCDB%M$*-dMlAx(K#RqShS%$eY zt31)zeRgx#yVY0HX7)KrrQ|?_loo60{l^t4{r4H5#KIeCBL8E6S%XN7nIYtw1XGT} zWwdM6UO`6WCW@{Ezji*-Uk9lmGi&#sIJ^Dd_E`%{l5PuBFIf_IRK#7c!a)mbtxFwx zssyJL1{rt>&GrbAmEjYJT-dXtT7g6z6MS040fVC&?XmG^z^n<{A{;CXsVbmnB(Mf( z7JLo{uBccd3|SVAW~ON8UUJ4P1Yq^6NIUF7$#HPzC=b6rdw;hb_Z`~6Io%P*Au_G{&AWTNeU8h-dC6Xc4+=34S(KjGn{jkDTl9R4A7YGosHUF}I24TD&8l&CKr;zHdk7u^BByT< z^UiXnW!J5s^aO?Ha2*baWSrRqiK28%1h#cHDyii)Fh=!-C(wRzc6zB9fK+PwrQg6r z*rCe4As$K5!w%d&GHmL+F3U4qy}R-cAK7E3muKhE{zEF@v7asnICxC0xNtXLNamlo z6R9=%ez1&Am~yFxf7h}bBz0>L@v4&EoLK-3LbV(iHuSB%kXgK-nXnu!M8Xs?;QB92 zsv3#rd^{NI(5ZbT`*L8#4~$tucl|sOf&$_3+C?u;n7~Q(?E$i7&H1^JKA%BR54|>m zvIc%G=~(8}!zJl$ICTFH2gW8n&eiDaW?_8hA#C{_nIwFufbjj8fRU1WXMGxp-ej&cfBao%NeCpc-v>1B zoUU@)k!FL+zbIQ~mP}pLDpF$bK=Z~K6M&lb@vtd#pfQFyh2U5NF1dI_D`TJ)Z`lXH z3+BbKnZmpc96{d|TWBXQ$VRe+kx^C#BO+-FcDfEsCFk4Pk=oyQpwJ!2ey1^h`(88} znZGyKg)eZ1@~xmd5;$ig@ZB^I2Q!O5WuxR@S;iw&>!AtM+%DjMDJ>DWUq!)FukNZfstHR0G%q%PxWxofvwmnV?v; zY$wg{ICf~`N)4wq8$-q%+C>0e=H2Ob{vPM4>_$BDV*fksqY1{#2V9DXpTF&n=#R12 zBkK0`rQjtOH$MtW%0(z!x>LDD(ne?qSDzvd;_-}Sqa_nuwqwxlV}7pi2!!S{zRh&M z8ZN_(k!w8s8}P%!6Bt?sk6@G@O*v0?$ z#z8tcx8>w00I_yzgB+21~%A>E$>V4X1Np_n|yl@c_-K&fp$QRJI_69>&?;BzTBPBM4)Etx4Hz=~1c zbyWXB7z<=Yk&d8oZpUvM`jvU;Y&3qbvJF%m?@m3)kS6^zrIe@|lB-0h;$ZTx-g`=% zl9VsJ@XVen8S#{LI@JA)Jy6A|W5`b5*g7^wqpMRzOp>STl1nu&&32&_hEm21K@+7o7bCKSyNy!Vp^| z=NozNH#AuAfdq`9i9Xl0oDps&S#vn_#9qkxE2j$Y9*rCEb`IqV^PZ8|yaaDU`N&fj z($h|sd32fX-6gUlRp9_Dgi3-X$XjBFa#lCk^p0H#y_~5bWF)!&YOE!b9B+|T|4sBi z=+A2-MHBvJbzivTp@ypHRAm0En}&EVEqreYUBbvCGZpQYKVc8(C42RXOf8It7A0|r z-%D@@v&Kp8?;tme&RTGF0gd;Y!?5Wj(^%m|y*Rxl&^m9tQDv4Q345(yXP}LurFaia`re48!ZCck==RP-FA<+^&CU(>gU2WUc1WU$#B{a@oZ-NB~{_2 zZMSHWrh3PB_N``pmZH;-2ZlI830OM3{jRv_3*BkWBPVZFg=*h{T(pD>!^Ai zf`^7WkcjkHDwiU-6eMIz`TWKBS_6g%b%c$o6bh!j`_|C+UYo6pxk;l5Ew*i?%bM|o z^>>6KquEgXZW+zff$DEtu0~Im$_QR~TapKlf;bJa?rOJLxJwBVSz8R_Qo=JUm>*|Ar7~VE7>yQ9RfxY31wE!Do=pnO?^rPtnrTD>1fB%K)Nu^&IngDui%l)o_HuWP7FwuFzLcQn+<<5-(t zahPtXW0Ne#pQz$tjFwLxOQE#`{A{bf16uqu8tnfw8Vm~0Cv-+`sJM?y={FaA3Mb-7nGWl=%Benl4_+fTddyFEk<;c+61pXr_6 z7;YQ>Y)pd!Gw(|XZKU)thfRAZ!6 z-Y;OBUej#L?M<7xCt-Au&s`eUXbxidA()nIv#in~)>zHrVan}dVZpV!q2jR0>&|v~ zQ|Jb$#igQa5mQ?sm`36d+=##(+PLi%{jsX6by_Ej)@}M;9OaN9dk^9)m;z6ouP5MSaKgJdl}ls z0j-t1pt2!*ST&_FKFrqoGjr{=3K*oUO+N|l=g@lLD_Kc59cR}EInmD(B3CNzh;{j&|xVQ z{4L7~Ry=`oLLT2e}tUH7O zU_zS=|Cs4GaKr&CZEtLzFanQ-$n#AaD0mr}Bdx*ANoYc0GB$qjEM6d)*talTU^jsJ zmUSu*X%Mm(?S&UcgUNGSEkS#6hqFswwBSP*_P}zz^re-~*DG4DT|vc)Ad94nMRmde zN?I_95d)ojmlphR1+aTId?_%}?aQp`tJt_m$m$ykdx{_F@|JE$yyM0?a8PPju~LY!VFj>U4bgN!ztg)TnG3x%?+;*MoH&{1iyiIhi z&`7gXLCJy)c=A<9^dx)eQ2z;rGrJyj@4ghzs;|E2rj9fgOm=l$21OfX3X|C|{)sBEag<#l4ld4%WA$ zrW36N2g{%P;+!8CC9e z$2}aX%~KqLOwyDl2YS0lJIprfIHf^m3YI#c4hn=3b>-+F@K7ctSN)SNzss ztuSdyRZ(IeA$}74yzeF9)&ya_L(O(^>k|S8Jc!6rUy9sJh>*u=FGEPnR9CQrLkY%B zK5Bc)#fbir9<6;t#8J@D^jFT2JNBtup!P)CRO?Z}%t218l%Lapgc>>>KF zS$L-4To2_6X$DWy$pMyFPHxGcfNf|vc#+R}U{85cKTDTr3y2kTCM zZpkmgbNr|!+_*dHvTc46tVzl4@fl|@c~k2$PBAiW`)7D|3w?imO_%eLDM*Q7&{|B? zPb{|;D6@>^bzF-q)fJmff_kwW88e%+#BfWEcZiU8EQT1=HHg(OXSncPgb~7Wbz*V+ zbDV90AXLDb_K*G5tpSZJnd44CA}6fEi{9tQNQ#WNkuQE~F}j{VWVB3;cczUIzMrI2Aj?H$6TB%>TY|~D z*byGsua{(}z=Agisr#qecOQtY`(cE5-rTReQqYg@2NJshY232ZRMjD82Cn4XtDD|- zik#E&JvF77IxWpXJUXSLmjs;v2>7v9Cs3eC-ZW6vFZRa`hH$SF{gz;queO7UAa`!d$MME!S$N zDc$_soL((i@>oHQ!ZiE}j3Oq*B4}@BOJZs3K`;bi@i&7!NUAcLfAO!z5mLaE83l5G zTrkh+vT+}Q`KBRnD)Eo9!~r)}Sht$y>SYBk{qTK<9PotH@QnR%6H7pmEKZ1T!OMgP zInK@?AekfaMSqV+leoS5Eh`BS1e>^StjYu=txTV^$cqNi zV~V|nvIQvz)Ctu<5c8T&`N-X=t^}1VVD1Frn&6zf93k_s<63Sw3cm~hdp}4mW&?GT)T^S>qrj)>)>1Z_?v{j z`~8zKH!)D+TgV%;k>+idKT|KX&$3@|-4Zzr#v9xcnu^E8>LB^1?YzfRGG6d!jHiP^k*wGEM)(*f}+4 z0(DzAww-kBbZpzUZQHhO+cuw=9lK-OwoZSi>brX1i}M>+)vmop%{2!bdAYlr=hroE zg6GfTi5-Kl)!(u@^>&8t+z zSYc`$zQ3b__c)M1R$SRDDr8px+L4R9PI^Lp`R30YEBZ*>rXvj$RrALc-oTn+Tjz}X zh5`g%^eDdHk4*1e(fb{9?7W_!6Fm9(3N%c=*9(&w%Tw2jQ}A6$_6HZilW9f(hF3;H zX$fe8QiOCV_=|(vcAm;`1o4ei|HxMOJA(Y$zC!@884w6-TYS&@t=AyH@H9(Zvz6zW zc@*i13yql0e!;VMAmz~!P~?q?vmWs#%qL&nbe6^Fja{p7g+m<5v(dV&Lk(GDoje=g z*GU!LOG%2EFEJrpc^dK^4`?|h#zCX8?I`&yq`NbnQ)@>h7PzHVO=H`b3O7a)U^QHE zMUgCj$ywJYv=UOq^cJ!NbxIBheB4rOe6(0x#p9;!k&Ep?SpFv52m6Q-+-)H1SV4*f zzA(9Rw_VT1Ck1TVX5r?lNu`QX+XL2{MY}xrnSdBdm}bq6Iusq#JzgD_0VSF8MF)Sj zF^i`UMvzCYf<OOQ7wla%bOGUs4j@8 zriCKX=0VW%7?he>50h{bC4IXHz^Zs^Bl{#+#oP^%jHhyEkj6 zVRyIwjV^+1zoFKj*|$}qUC`(H(r9=CNid5sybk>*^67_lr&+M|qD_lX?_yTlZ>CxF znd}h)|L*21VdqfvgU`-1q*NPRFMlo}V5QmzFdg?oYUO|Zl>b4FC~Zrmf>7e%_vLrF z3(W%ZXU*_8cs;}Fn7Q5~J`hZYw8_Vxe6GilZQy&2x?E~n6PM;{Bbs%A1hD(-7@Y|f=6zuAlR}xyYyV7GK(eCfyOhLCS^y+RO zVrH1+^Be515KP!}2%LL&-4gO6LBUvZ*l-KRpLlSN{+WRtZQ4R~ z$pzsn#wSVa=+gv zoB0K}FT@|6#ZLKzX1j5Sp=@JcBe8;X8NuPwmlAXy9?i$$&m15aFGkAmQ!%qgh1$(=?&k!j2sCd}t8e5FEGIq%Z-u|rRlBy{SN|3C-O9UO zCnF95FFukH;<%&x0LgXUK(Td`bHBs8(-?BYr_YwZgr3$<_Dpp7OcK`vkfcF!HorSB zzn2@{q4WM(A4=b^r0XOnuvgWc|HPizu?NtA+l1+|&hfcl-_?f8sqq_E-)F1X@t14u7>?l5=oFMabZ?kO3z zzj#FiUl}qI-P6@Q?@AF&4e&GW?nTpb;U0SUYCSoDg#8;GY^3`)TjczH$kf@prLbGq z%w-|tpyoB+LC2o)FJcjI0->6`9tGw{2_KG5-{Lp<6~HJQD^f2S27rG(@4#>lJO2y> z2PZJ?dQ%;0uutoL-H$}8Zr8X!SSZ@*ACh9Abj#(?Ke$9eb8Vtex@Esb{|5aYd8+f* zM@ou`0kcokuEUQ5V;+M$RQceiCT$M3j^xMxg9}6q<{{i>4{(psO z-3G45I;B_!!#QAreeFu^L)f;xBy&x}8-Wm7`ji}+RbsvHZPt5h5uc48223Kb39#{t zF9HZAv})piai2fIQ6}!uEB#k4t^ED`^!N=$(skTI{+iR{=lkJV>EGfE(elmH_eJPF zB5@+K-@jS!>CD#G=b!FHy1RUyhN_$Ip6YLtZhDGMF@hd$x1T;mcupYJpZ`}-U~P9N zsA7E7j8%lPc-L2_PvCOGOkXP%JHqN_&iwC$vp90*UPnw?Uaw8Q1?TD7AL)lP&!Tq3 z)RR4LE~seEv3F+HJ^uxal4YI77^ZSbF#_trd9 zreaM9+mIOpfu4B&{fA%=B?Bx(zfIqG_<)94g8?yAX!=SdPhKR;J0Q#TfM1m360oP@ zAuT{05o3vE%ts~sK`eOi0pk23S#hU&$cbXMK(UY#2N)0uY(y_1AL4@L#a>`8bk1U; z{qJ&-Hm;?( zfF|w|tS9036Gkt0+oFH`-P|nOHpoaPYhp{t0JqaqsosgJsaR+@-)=dWN%{cc(l3;y zF%s^cA-A_MgyWgtFE1KpZHw!EGfi6`xM|;*F4N=PLsu6Jg#E4Ih!r6V8&?}^RIOjK z!)WCw*5^p0sPh*njuR#9&SbN$Fw~T|7z}&F7QPVnuUe#&s?B7XFDS&IbjTMtwO_%C z<5fVEI$-6-o38!3M|*}5t}ef2ayh&Mq(k$7i+PB`sWiocZF$EG9l3%LOt|)n@sZI4ThizN z(@x7knJq-sLAOmBxCS^pio;nziMKH@Cc+M_h6NULfW;x| z-9ub`wTpf=GtWf1X5!*@$7@6pNIJG!1`OI&X?jmax?gCv0vU0GxI%9l$at{WN!;}K zovKj(v;XBhgMs(+e2U%7?QdCw%)LrsQv)=??Eb+6p%p|D=kl)Nl?w!~Q?fgxIT?_t$_AIAi5DdwF)gt% zMhe_EH~cbXgVx-`QLa9}g~m*|wX#1QcbGuDIEyVjX6>k5ho!I{Ma2qqDg8K0GA}!b zqEV^^+hb8B%PQQ|Ojl4Da5-A*5zza_CIfFout+y)cw%`*Fky&k0Ta^FHW__oq&!`1 z+BrB3!Y3yEo9Y}F>tMR7LmQJBKL~!G{S@4kM&rO&`QOU!#A{-sK?5qOLjMfhKG)9p zyv9QT*E6xcBlam`jZ1~pqfKIaMA-ARej94!rL!fCa?n z&AyWoJ;KWu;6RxOUyDFW&hRp*SK}-wkYQDMF42)RAh^&ReT{ik)<|OBOZMhy$09= zP4QEQUH{J)8Wbag*u}QNR&+DP)+CTedJdkD*OFEL$`h!*x#~kk{YTzhG>y;ditoL% z*+Jj#3E1zn1Ib&H;2`n9fnXc>Csa5~si|!0KgesxbKbFZ;dc~cg_jQB(7F^#GHYYo zgfCLJ;y3Ttl)M8E(?NoQ+pV(4SJ`Vw(;x zt=(mN^DpVB!j%|)1e4;sn7cg_Ue;!Ahwe5F4yE^SwJ89LUpbt4H$SQy)(7Jn@$2|b zkp;``tN0z#Q~e{dcYh5-KIR3%PdYCTl9mNPo?yiV{*lBK54Hf{b~T z1Ur>MBJ7l8MMAe1tkoGvtY$Tn3DpV`tXd87zK;$`1xeZ`f7@ph%}Gja;<+>upsx_j zF6$@)k5dqIdv#PBZKS%HLMYE(83)eRlcjoO)uya)H0Mq@rpmM<58y;)D@*3q;a*

D#^+hk^v9UzP}O8}Y}5Gc*V zd*bOAe6AGlqqZydvnf9Jt3`9}f_g|9R{ z0ssVL~lquMW z8JvVCymc6;*umR5=l=qn78zepEhi?2h^j<6#{#e`g4z-c9#fku&97vwDodlvn@{;I zfZr`K2K16D2db%+?_sTaHw{@!kR?Bge%BpWFr+nw_r-oZ$*|{SJv=vEXK4^-k)j+C zg2+gE6s&^rPIS7z$qCjr9^H5{7Lo~68K^sGyx9pky_MTf_YVG* z&>P_FmcS*y?B2rO=Q69Q9e{;R_^NCc?A-rU(jsNNt}M~)ag3Ai?+G1;8AdfmWK|z# z!aqcq)kPVkJ=ZmmNh+vGZ(U?7WO**VWcK={ruI~=1Yr#N+<^dO&^$*na~{O>Nt_~m zuDzS8K0&KGOR*_zQDLh>zZAvApos@1OVjwU0pF;VH66?nlS^F>k^64y<3#?h{-Hf$`5cov|B9_>o{eWQsb68%ebS4gG&-co>mPTyFPXU*e)J&ss znh(1_oQnxeG>}RS4=N<$`(|VxXW+&69^VhYdp#Z`Wik``%es?aBA^Q&zLht{Y)bcc zjV?J*@W!upKtmNeC}G6!h@++Q6cnAQNr>EMFjx^0zN@a5K8=3;D06zN>nrpq+`0V1xxC=2gK#`pY5n@8WBrZFW} zS@$0nY;*rVo5Oa27Om?z0Cgv@LX}qcDfaR?oq}1}cvy50-go2ro~znmWCb{ORuFs4-osBwe^)-u)*2Fw^DF zfKU30^o5lr$x?pUSW`sRR z6a9t%yVUgU+x_x+B3{T$P5$?pp2b2rSJ_wA*kA4aID)zu$hFO~mc4Aq;s0|#(g~jg zFM03n@AdGR6pBRUVc74GVE9LeC3-pYHf1Skx_Kt&;+i3WiKvMkx^h4okJH*B(_-4E z>E7o_ALr;>=$F+g$4S~Ra~583W}l~`@A6g2=B{7X;JuU3)Th%W#e6vpCyMONx`Rlr?IBfKcY}Oh35q3VL{dxvEyK`Z_kVn8WXzAzyn4k^l zcFEh+tS}S$a%GrZp0Bzre~FBBIOHxH-^OqnoLOAeLG{ z_s__5LqOgSSCmSft&dHnb3n#MtWPE{h;Ln%o0F;oGjj=hh=B)g{ zXv{K7&X(dWG_>Q=#AU@$jiFq-g&1y>R`3$>pV$LWDD>;h15Us4mE@mlFFRcu;SSGr zPvu%HgBu<4d9m@PWl?5JeE#4WCr$cxEx&)Jz!@C#sIfSreCGBnQPUb4O0@wkv|2ArU?c`lx8G9Nod; zARv|-K}{ioVdOw}U2Zh!VA`;-b>{I6LHeQzO7^le8S0$#F9a)t-eWO6A;lKngC2w; zKlfSZVybPB3sbUU+qavSX*RHWpct}P_#nTx=%q_Qt^4dZ^0bjp3iLu+kbg@bms0gn8SSBBzW^2S z4_5hDq1J&{_+L%>0W=;k@`{0lKUu68!A$($KOPe(0qwKB%}O(yKUsm&@7Ly2|X~`q|P2pQb=~sdG0% z+eTIgaMjf-Xcd0ar?b>7)Q5Yy?oRy{V*Y7q`W2NjZ-+XFTMZtXdimYt%86j!Yde9I zu4OHW^`4{VOwD#C*??EWQd3!o`G@teA8w+bZ2|6DjJuBHg9}5fLlT!LP<_3fZ!_|~ zlp0!Jk(<<*a27u-Nalux*J6^`*?DXZPJp+4GKnZjGqk9b4X44n_KoS>%{SOkxH8+` zM~%TMt|%qEIARyPUqdZVPaH;098T0Ew!K0b3nB%#ts)iFTS=^{K3PvAu%JKVrygcI z=4E^h95Nwf2r;-79?_OsTfV@kVAyb`ER{PEl*_m~9I8D$z+0DI(27WNT`z$~WO`4$ zEgQ$4jaCM<;co|D(9xdWnMpaJc66aWY@wgjc!Hcl(uRUeBew|FQ9>Tb(Rz-_JFcb5 zFfz+KL^0nFx9*xKjumPrP6*H(!EYxlncJGT6rJ<6>=0|Q(%IWgec#u8CDK5BVSvq$WFu@>|Ir%ffojH zsPt|W%kM%=dIlKWJRpM@bfl(zVAF>vSVeqI28iXar-opUJEfQ$Dn$kLM;^dyU_ZN1 z4&j%UW0fKyG_2-G9zj|jhb$Ja)8wOokEqjBoNYo`CY>Cb3#SbxaeTIXXR{U0aG3@) zso<709jvAq1lo-jo7z7?;S6&jq|-isLtJHMlXh9;Bj05b>T1rg+oi-Rfk5B6IX2Sx!u6K_AD5C0lE5rmdBYY zB{xk|hW3_JreMQ9O)XkQI{@mxw5y>B(eumINaCkP1gP8vlQAIhG=%#lXNh{wf80k|6SwsjWh+Bo8iAn@$9KPD&Z{ew6h zJIaoSI~3jiMJJDjbayz3uQw71tr(M`iR>{Y-mnmSwSL}@$QoqM9AA-nWHva{d}ln+zB4^BIcEP*p^@i?eu9 zD1IAZ`5rY>C?R`yA)bdd{uU@=5GJ!9n;rV<=W=55@yAF(5k0FME%n2cyQwX`LqOq- zWI4K@45Qo8Cs9u1N&d02%MTWs#g^RZ%kn{tgh+GZ+>XyWUo+}7$-1u5o8rzYEGqj6 zaw&%~;~T4jUJ%}g7)zIA>?JqJ@*OmXR#<{@>Jm%GFvFr(Ot#)qsU=y4uG&6FX7_bu zzCi(jLN6x&JwEtJ7S_a$ z5D~#k8C&%x@ll14Pp%pe_R57-=bcZ|06OBhG26vtTn#6J2oa;f6Fu`kKt%ne86z?2jzwNxqF3C;a zu|l%Ww02F;9EtqN<_bQPOK_Y%J|Y%8zG=SZ(lJ4GRPH}{k0o8fcEiD?jx|^)>ZW1t z*MnwyD>j_L7b%NKu>{GD`m6gez+lmEF%s**EKMkpN1^qeyL#X##mrVe`FapZypr|7 z278(EmYtq)TGLvHPu=7<%0km;JA=cQpt~B1404N^M2NZ>)9t#6prht8x1F9T$Nkh< zGRjlG)u%gMW#1&U(KdH$U1pi@-+>z%TNIEQ0+wb5jT9FR206J=Ni7y zZ07(qXqbWMEzK2DW=h!<^IZ@2VH!xU)!w)xBa(E{sRz(qjc&W;jH?O(({m>mr?mCU zwIc*FNDy}&H|-A9_)sI3V7RpdNX)yvzlhmdi?wh@7ggPP%;#pF2#2mN!qE;KRYX#d z#UY0eUk;NX6@sk^T^I@CM(~-A+IHb%`}!P&Af8@Mw`p93R>qt1!UxG*9%6Y!BD7*i zl&cEXH9fhegFgzh3>`%j320hKmO)YzIWI96cp|?Kz)A+1$G?3uV|3Y`CBR5DFk)1G zdnJClhf;0A0Cq{^ly{0<#nAycbYx?o0PT4PDzR++;9|eDVkC)Bck8TDwn(I?6N79z z&ZSkxhab#DBw&;Kz`Pb;aqlsD$6ho1K0{WN>1l76U0JL?Uu4bF(t#U_d2!kJb+Q^xb6jRkkW!Lf90}~aNqf6hBb&&iA5j?{^8nbDz!XF4cL>@KWg!`?kbm`YgEHpV zDpTKC>(M3D@sa-QYAf-UxS*%bT~8DC>h_rKh3ey|Er=gfwU-#&xfA(5@EmqdhzjB5 zsW$Zv6rqUym$}Z%DmMN)wTCe$9KRk5YD$>t)rX>TVbCa#>ML_lh z@Dwqyh7d%Lq3d_=&ndcov4$yIl|+#6NK0ajm_+s$WW;E>a zIK*&y78EF%RElUh1}c^@nixkWHdY=H2wCZ}S^-9EN#w(g3j0}1()4_m?-C@{Y`LT( zwyrQ1=_yq2#0voe61*<=jxC z2c0Z9U>%q$SdY|Zp{Va)x4uMZ0vZ0@HCju)O^2~>xLSv5skp>NEzb!&9uI*}1ElK% z!Nt+NAV0QkYZZI22*Scvp)A$;8aJMu0{M33m^oC&Kb4Q9hyH}g^Cu~r8 zl}CMK(qc{2Uon~(M?t3m#F>HYMTfpjGegUHx zg)OtQ(w?0*)pu=&^d?}g35hjOlT}+X+z(Acwb@)x&c!~gVojhWg>L^ z*+4$O%&+uTOgg|g{M^0h+*Vk%#OCW+z;84&N>exOuR`>0D)l?3^ah)jdm}|SdB4Eq z&I;XK|HNYFJ%RDj4gZ-21Y}v`ouY09oE94;#~n0sXu2s!2zeH&kNK zIx8h^_a{)Xy(w@bJs9nIR|HQC`_#}IB5-UE$SBD9tTd!;OTw_i%jm`Us`>eeF;TXxh!!8aqFqJy2)|&p@VXHk|xuNNCy4W>r=z zULTZsRwa-8l@>o)UEp6c1N1@dQ*o=Gnx7E^jeR7~I}6;x7xOrm+%zWZgyMvDSKYnb zHdBemp>pW2%D+p>FGm$HB)~~VKQnhh|B1`rxKYu36|euie4T!S5a!io{TG16$?~57 zi<9+#0a(?#@*bOioVQ(SHLj-irUiNNU_QkPn+s|5|4Cqr>E~C*>aW0Zf_PXg0$Br- z#J-CTwY)N$TXl+vws0n4l5==QPBOsr^OiIW{4|!{UO51VR?**Ky)jd=-rgQ}_fbWy zrF%2RE^imdhhw_cihld-KYwj^3GWfzaXd%e$^Gh>q+VZN$exO7^YwZCxU+S?M|X8D z6#7BsmCJkT(I|jA(-y>4h^1p(EzjBdqN2`zgqG(+X>ak}=Hr&#Zn0}FahIiY zAzc`GhyF@#EiJO|MS!nF>#@x^OQ-@bRkyc%*$h#yxt0emt!-_UPc^hu(etEy^;~K$eD#yZmRc^z+?;{4 zW^Ing*y$Pl|Cg{ z8-?Z<9t8adGx_V`yk$S#2TTP5Z=8DIzrb#A3vV<_%)p-%2#v0)^v)XT`#6Uh;*Oq8SEYU_ z=O{dinygGojJ~k+e8?9PcJgIk5V^atee*NWBFPT*Pqc*OZ7>$w6_Q*YEWaEfROmVj zTfO%>BSxXb{Ngc*MksViNmXNZ{P^x6@f8YR%+QKBx!Ai1y4RMA@dDtDWt~+WsqKkN zU88?%)i{#Dtk7_F%LU(8e8SGt#G%uN^>Cwcw=EdXkGf3U9M-Ot{aifRhNSDi70l^Q z@p@fgV~p8T2Q1dsVGrBVj3AZjToI5V66!>nv3Id*a?9Z3IYX2x(Bzj@z~^ow@N? zzlZ3+w9wAMoBI~XKkACON%~sKuuXTj@dRMrK%hk{i;3i7ET)i%WQEu!(3E=z!MiiP zHe=We2tbJw;EB|$&d*N>b1NnoGpFB6hM_skVH?U+fU0?q=f(OryS=%{^B>{ zb2NOMr&1hqq}#ts7I@X;52<;OF-BsASM2?zGsdIiF}&ZX2^;5Xzcfl$$oJRPA?P~P zo7;y@?HdMzz6I0}0&*rKxDY)AO0>Uwo4yT~_QhZVahGu&^Pp!XSN&|sUafD=3p`iK zd;S)465P1|df!&<5j4IT9eGtbc}w42vi{UqUfFJ*nMf*_1Jsl zc#D`#tHSxQ&ck@Vnxl~;+80CbdiXu7}B z!Wa|1l-tgwlc^NQXQ@@d_6o^=8#VEmKaq}&jUDJUjfMv?F-!LY%J^cNBi#LPy{};w zf=MWF7urefBa0g)A(a(1u?+la(`QGg|fr2ej*8#NHJ zvVb?*4rgG)(tb-YtA*jwTn49z<<#YxksRO+ssBiCA)fRQW*c9@88Wz+ntGg_hV7e2 z_tl8x14mw>ZdDgR;L9$uX}_hV#042a%Ck1u1-lv~_6Fx3X^ud!cL8k*bOgUuWr2#!h;Db3wKMH0WBFtb7uIPXRnPZZ%d7hu3T_ZT3_wyN~}B zJvrOB6)BH`FYHMO&x!rz5?Mn@A*-Nx$x^nCGol9WLI6T|O#P)P6y zSt9p^Tse8J_x9qoyk-W0MhntyBE^1%hTLJQtCaKjo~=|7*mT=vr7$C1{A~M{*_N6} zcp)yr=#`~NWQa1PjDPzJe7JvkunUdD24Q9Hq2=bJAm>BTTt2=1sYM7COA-~pP%cQG zyur}R^cuuewr`uCBS>eKQAvuC5zbYwHZ*^MmB|zw-iw~R18dtoJ5@7K^*!{&GN?w>?q-+L<3aQ z)+6FnwoAi#$(u}`vp_*hmOC9wY%*)dOYR;0p*$OG&~{Q2InrSCAC{`Q9G*GI3Hu;M zri_+bdTi$(xhB2QtdOrJf`R@9Es)4%1%622_8Z}X2@q}ueW!L~XKWAnZCUH4uY!^te9w1Q0u)6~jJQ1&ph6&<~jymlE zkuWFYB~c}&kE~xOCLbq>#5gzLp?_$o;!$cU&07zp$FCMAUI4+RHrbpGv1sY~T{Ktc z-&u-az^QOb~5#ER?|d0B>&%!Bj>`M?^f|- zL7HCdFG>ss45J)@C1}N~EtDv>`Td7epb|+e5N0}Qf08zv_#D7V zj$sE9ZAF24q)<*GpXH9T2|v^#=H+JtlWitK+Ns9m@jzU&A$>y?vf`A^ZH&K<2ijh$ z`#N)~^N78{dj9!oSm5YR%{&g_EuL5o^zrmAFWp&4ZW~Ab%Z-nCA_1~)sUxw*k%i3P zo!$6Y(3*GV`CG%*fLD^?)Q8~?QHQ9%s5@URM{p#$+OHEIU8iq%rQPM9nl4!m%sELM zz2K!FK9dBHalq93qa0F7Qh`zZu z4hj<86(+m4PC$1K1okv;akMHuGR)G)zzo#f=gU~nX<8Rp{!eAb!9M*M3%kZbTLRA5 zTlnkosFNEfpzAyKO!7hMH2tdPz|ebL>aV~oq>PW5m3uX{iv6Lz5j?;kQ;721%|U~=$t z*_%T*5OgAt#Z35(ImL(iok**uk4O%*L{w`(F6G40A(Q^^#=WRJCb+xKf}Q&JBpm6H zQwsExO6QPbJxN`nE*w0&rK;?H@@+u{Cf&@@7Dc-oEFWs(+8>okS_HSAP+;U zE&;3zKfRe6d8qEe1c9e@ zQIBfSHBZn;ekqRQD0LC0^lwOsAV==ux@c5M*5|V$*giEAoMu1z#M*5$ZX7!H>&oA5 zD3K?(lChy$0#;lFe`XG>#Qe@{E3=md(O23jmn9sU79W4fk%d6Rn55pv40oceTTK-_ zQRT!{EyD!Y+FJUJ9>J|;)>|+cq3ym`-Ac}PZE3Q%$O=FmSX9w)Ou&f0y}XENH7jnRZhA{vBIMHU2jP=AJuVmcxd&OzZpr9d{SlNt}FT85&ODN0OSVan{fBx+FWgbjl`e*XTc7ZYbk?AxmY~WJ+mqZsSm`( z-tCwE82*itHX@#k<4xazu{(oMf~vqyH$lT^u4cxNgLfV2Q@07ebAGXo5b7JBpqY3) za=0xF3CSulYQ*k9b&Bs#tGc|KpBE#nDkk7?+-5>HA|e4$Dnj()huKaUsBAZ;35vdh zTBE9nHFZE{ae9`U;oWwvVRLwk55Q>=Ql)Z;ejVIwTt^{<3M*3vz9Bp0SHABrPoLZ8 z%*Es<9VGU?@EUOj-vbnFq_ph0Dg<;`5=guTYrl@#YF-S!IT z%0dJa8qi71%(MsmvH0DnPl)78%_OE+BW0Mb-z+pqK4iqOlZS;ujUwCJyuT{A=-`qm3R8tP=hhDY>O#YDzKoFoy&=m! zPoP{x4co64bIQh{HCxkZtXc-TrsL%EZ}q8^gu55mlGm-{>@gz=OQ~ZUf*pJDc1kDT z;o}Ko1bRH|uMpnv0FAF8(BeCFQ3Ocyd#f{I)6t*bBsLj;j&&*;gw?-2V273tWDg(% z1(Ya&+I->E?4U}%Qz?66MZ5-;l#vCc0vLp+{Rt#={@pU32tzMB2M5KiC|5|YPqY#7 zX@M{TdS7r~0zC;O2lq2xRtG%%z0dPX&nr7!1UR{Sl3g@^8yJ@{)5rkRsVCyn9y)6C z)qm|qLr545F;TM>-~0dt#H4EDl;C1i!y3~rIL3gW>ixeqGpGqPk-@9^0~N#uo_+7B zNACspVg{xy!@#Y@rGrn*xcdU&Lbv2Fa5}q(c^@UCrO!s`w=m2_ixkZ=J)!OrE?bAxP==qG=XK6GAfdO4V%!B6J_edw)^EVy#Ucqbz99cf?H-h ze0u)=+<=4LI%)uu*+Wt=pj}t1clkGX-A4oC3Su*BDCYaP>Bc@L79F)%s?MKg5T;bS zN6&qxMGdzZltNm91w95ZhE=CawF9(W(?k#}X#}HygQ9l&l_jPo{F*)|eptH2CJ)S9 z@rzd=TItRH4o)|XX>H6qF3<@Ih4G4>2*vI=21%OtSH@@c_Sr_CNDGIVJrBOz^Et9W z|96jO@dNf3YPdSvSd~$q?&D{R&)eY`;p)Z}SCICb>~N-O7o`rPDCbS6GM9EryB{0H zQ^yW;h`li-)CDfeqa$U8GTN@)es7M*4omQQ@2?MJrcqRx{{|FU{~J)`{Gar#0GAD! z>4wQoCDP6_YGPiC5+ z5VSJ6XYMbCU^EyN4Qx9;nHq2ZkF#@(%`AHRb)VX{ZCg`&YFks=c01Km+jcv(&8cnM znA$k~zsX77RLelJu{B&Ex8N(3rE-5~J5h?l??Q*L3TH@i7GIuq;r`aFD8?w<}Zd5j`$&9cJv zd3L(92qII}1#$*sGBzw7^EN+AFQ^)2*BwVI+uLW{<+nz(Tj630kOdz#j z(Q1SayLvfZAA?0BQW_Zc5g{_~(c=!SO<*Kuf@Q9g!nOA_aC7@FOwB2PQl3uCOew$8 zk1*fxPKzkjW*cyFwPd=sRcD$u7AFOB{h5H>&O{w(7LQrQbL3J32*OUd)wPvh_YOkE z91UoQ8;Rf1tooSqZ<&f;i7sERmk+7l5Gcu=28+iTH*YoGi174n?mz4db(rb>nWGTU zXF7EkCX;3>U%M(tJ#J{%o@&iBgm0Jrg)^mC@CU}sB(mjD?HyfCT(dAoi*M*(99?n9 z#7R8*vF7@r+9ILTu>&C2(ws^XmBAK;@1w5bcySLEF)&_6C%mGI6pNX5Jo!1_%6{xn z6=Tm27rwhVdZ(1#1`0}8LkCH?Wz4Bl>lfAKemv!6I25%3?og*T zkKo@b&_$%0j(#dN^J@9xZCUPE`uGMMl^n(e{$3e3Z;i&K52mngaJhPB*#@V(8?{Li z5Bbi`T2<$TNVd*xWp*Pdus22(mcM-i^OGI(DwtC`yS^r`s$g$@!CYuZ{Igc*cbzt1 zBq)c>k{QdjNaVMA_e zP^FbbXGb!XaVWPsg>9*NHT3ZHxXmy~LxwyNhM~E^RB6{M+__m*=MYSC+kUH3wTbf} z)qHyN_PDdR?FXuQ4bPvohwEIdAOm%vm4X*}NhUgb5?po^a#=aRq8Us9b|b2NyHB$k zs9U4A{KZ@O)1zn>0_x$Ez%Y^dY3P1rlytWLmA5J2t4rJXGBs-aopiuDU??TGB0LIE zJ$+pwCQjCvBV6;3w>zXYoQ#H_pCJZ3zOT7#hGk1%6{fQt=Hj{kl!ez_-B|&IjOR$} zIqo4JJFm?AcnK+@AE0MZ-Zg644!RP9CLV{wt%?vHgx$xgbOMDh96XVH!A%;IPDAP^ zj!4~)a%kxBg)E-;XRZ3&&?$m6fM2AXVWG@Cq`@aw^aLPj$gb|E6QlS876Z-cKzD9t zv?;a}AN#i^W2V$Sld=j0lw4o%^`{@q7dryFzIG(!(HuKk@lZuR#X@Ju^?8Nvkkrqs zy6$R?&0logvOE=VWZVU9(U{m`R-jXib)mC(2%v4c^dsSED7AY6%|ZJcc8pjLXYCpG z+lAuMJsbkKVdPug-k#kAaJ86N=p6bHTEBbs^bsJT!p<{uVhW0a z#TkFjJ7_9xK_{w}0>aRYFi)%-vXnHd{_>z@XgwJTS1=vmyT~TsNB#E5w)Oui*I)2u zB|ZfX)-mfHkXZdDC+W`LzfEmee_Vc>MwkA*JC(baUT7!Fr~WpR5H;moD#k`ZK9MhP z-j!!IxN&xljv(<@s>C91TZ|ZALBiOAILa3{L88;{>@dSmTVg#X9$TR$D(O`MOrHkF zmTiPn*?l9VyI)t#nafzoz>$j`2BW=O$Z*g6=ImiSwI&9{BMMzvhT%51wD2Y~Ne+Hz zV0ckpf2Ft=%c4Aft6>DCI1&X>{pmEX4H7zGdRVuG^XCy%*Y=RaKn-UUx z7Yh5PJrtImT#=ZT`k*Ggp91V*{JdFWW?LUk1&3fh-Z@X`;HAhSD$)xn(oD`{1i@(H zu0rnW*yzTp4m;L0*rxfoasI1YkNqLTwTOm(`t=Q+vn_bPkxm;z=;aApt1U}Sk5Bv? zx!o$Gl8S!r4frUx{FNa=`{*;@JgL@J9V3tGr)dUrlarqi<&1|>>lzTr0WbddL2*E- zRSFKvU?_^?@OKrFv1Gu+(`BcXC+)i&xaQpEJ|Lzz#SeR@&r6-U2HO>;9=Zb*s7;kI zo^b+?6i!AkucV{ljM+E1`4>i8mJWxq9K}>xUkJ`$UP2}}N&=SF??tZ(FpL)Iji-^2 ziX*XODxg2Leg>!KEivCf|6_nO*FAx&`-R*~>Oghwar8-aB_Qnu*M|Gp*lXGak~(2e z1z$^HPC|5i%yX@f*XR`_ZhCp!zIsTS_ zA$KuH@2KVg(`LTpW5XWm~KCfQC3?zcPL(0F-SX zZ|OZ|Lci~o_}?y9Vp0N#$hc5laJ*4KJQj;n9y|3>HBFH!RUkbJR8VAoA|po_Js%cV zFz;8lB_6+Om6)dDUn;wnFsrVzuW!y%e%}J8>K1(_boj}|ctM7Nnl#+3As>tCVE;0_ zlksLs@1eYBRV)aKbhrojQo?4EsPUqY_LJhXp&c-D9BDqm)thUOm!; zMRS*8(OBu@qY+00wM2*1cLkbf2(KgS{i2?$AS@(n7{VU{2ZA@svau0cr4`MYgbJ`kH=k@p9PSTZmw(t9FKVVN@@iCAlLe=x778uw4EE6CDW zjzo4Nsew@Q6=kLptxh6OkEAaR*y2?EppPuP%)ls@8YuOUDbf=;Rm$e3PywG3vN6g+^X=t; z>VbP03Vr+5m`mes;#l1$=%R?&Vsf^qgrspcv<-QSW7)3wxB5o)K&3kx;c-SvLUp8N4dk;$>H&@x)-THTk( zCk`(&v@>x*X7adAnNE&5LkhwXr`Rsc+o;hj>6GBvk_z5aeibOCvgIs3F`=<<>LWf( z#M_XIjO~P>Z|yiqeb6OVwPE-{{aSR_Xi7C|?_5!fY+-Q#0@Gl8Uiq*2!F$c4us*q= zE7ScA;EF}FZEJll*m!v9+I}x)2Q(A)dhci~kINJWaj4uGiUiI37lJBW5jYrkG4}Z} z{MY(vWJC2P&!eVzZJFUL-_8iVQB+!nt4$_h42%9|XuaZ8`{zw_+oA73E5%EA6+rd3 ztzwqip%q#pS`CqY>GM;Y!!g?TQ})u~|rEZ=$(RLX8`=T+%ow#1BRDs%j4=SW&;6~4Z#O=$ zzR4xZCSDlraJf%(>Bpm#!A~J?rJK~E)JrXG5buD;ubZ{^rdWl=k zWG)qs)Oj21PJT!Xk>)gwl%d0kb28z>TFbmSu`2K?pV~`$s+zE%A#5Iq@b4uyc;$eL?ichJmZ1>r2nN-Z6@iy!~pe=+>Os#KA+3HvLEiT0)>TXPLf>K?9u>T{y zRATAWTeKFS7oj31#sX?$o(*z@3)*!^)@nN85Yg5kFB^_dI|K|cJc8iHH|a4pyIbn` zAXBUjMK>tOZH!H0@Vm)Dfq1wJ0c4gdlh{C(ref091D5voE^kJSmIHy}eXaRBVLGRJ zO0^AtV@O)?E_*w!O2y6Aw-y`DY75Uuk8+l!{HzAAZ5A(gMXwN1_w8yr6{3(@#JV#( zV=gN#1{ArA41c%LSWC`9+#!?bp(KA{lAPY@D$)<0JFU8C#vY_9v0N=|$T2WrT<9{+ z>y`Z`s?4l7OqAHFpNhpE_omT*|2>W7U>-wzry4md$*BkT8~xR+Qw;v85X^Oy>dTiW ztfLR*7={ZT7z(^YS_gW5gsqyTw%u4zm4@pU0QS7o|E+gx8{R?#bJvb2^-;1_%eyf& zvxW$1h;mD@L~a`Gz8!RNqDW_OlsH-L9SC9aYq?dc-IQy`CUtG6xJ5q*y8})IIa5!C zFbL)?PsN0muZnbQ&Rr+%k_EN=6$8FoDT#0>tp0Vy5vY#+2#|gQ&1Yhpv2&&cxwE_D z{h^L}8}y)1b$XJgNB=lzTa6BJP*Ef!yWz8gCjaJrsGeZi-L5tSMc?J|cJ|S+fRWUQ z6HRTBD39puv-Uy$*?I0x*fHPgc|e%zO;9t(__boG3)EL*@f)EyW74y_RqXz}42_zUbRx60 zPH+2oHR7$pi#>2NcV&6mENtgJ@N|#C0$c7W++%_hf#Om^B{>Si@Q-lRYU;_vx&Tw< z;*arF<coBQz82jxrAdZxSg^GD9TWvs61m@ot1 z=yAz*=?b0M@NMzMmb|Q9GMgph9!32@PR!yq?Np7rZ4tS0qJ;-j&vWCT( zNer#ePzZ-nZaU9iX+MjCp<~r6PkL)VWSV5G*T4zO)yq8)&tgZh(@f&u^uGSx_UgtS z2P)MS-(q0bF&W^jCf9&&81m!6v$A#gzs?Xevp@b;olZJs6H>mO^XZu6H}Psr52GH5 zO57K(Y_5S)6pqJ^W%QS{!83;Mx?6Poh@l4<%=bw5^7Rfs&dl~AWsfB)%E{dE@s1nM z@pMe>XPtFM`>}?(lv~?ut7YQb(?V8r78@tFINCoCTRH4BUfz096(AmkHU^LI_l}AZ zAu73+(iE-Yxm~Vf&K=fe3tY0DPROIbJIfc`qxrM`h9`FBRO75!K3`u19#7X8E=*E% z9vpcRASv`=+Wx2+C*y4vS4S{zX!SvqA~(4l+|l?k92+%AJQc**K}G@w8|U)mF5hUm z)3W`1*Y^cB$1>>iKZT8~|Nmj*w5dSe?ElQ$wP&?O(@FuYgbBxL}h3>}a=4pJ{JNGUk zRsDD09;%=@TBRB#QYx>sO1h5MEvrBvMAX$6xt}h5)6!mNryjj!fh_xah4q?uD`z=a zSNQ&Vv-W%|d2)K(>HMxp@mQQL6Vrb*sn>x6d^0M38uNRaXyZ#C4i1$3Yn5tw1XWS~*gDe#9kBwZ(r1)ZV9eQJOvgpT+MnSmqUw7-k$ zZ0%ETP}0fUty!(BiFYwrDj=uf_UC(UI#rm1gq&!( z*YV&lgB{H;wTxWrU)uD-f^E;%Y9Z$7c*sOo0b_P>-68)Wk5#E=AN_n0e zh7kR0lui0>MCp}oXxnr(?HJ3QdAHo6)9{SzW*r?1s=Z1TlSsyH}NQ^brjSg>qQ$`cwrLPx1EbV^1QX#(H$ zVzR?2_=-Tk5?V)_DtR*Ee^ny4WzHzM6yQ9sSe?Vmfj2ygwryk7fX!(~>6!U8Fk^PAVg+9}GLuy+iw5!M4ZGhY`^}z}I(s9NCPC;sxs@sODHy932|1fvF zjgAovzkbcz?VUS(cp53OGp87L6_%Stqg0JZ|KOBZk<4@5LpV@E?mC{e87! z+FM$1BMnOA3)NXz@m2h8%t32+*Qry8Pe(o!*>QHHwz>p zs04tlU4?nGET^;ZBlvZ+LV2bwDxm$u@Iz>kuGu!0HqgE!Yrzi^KJ4UPJBa`W7NEpv zm*IE6NFVDNwJrP4hDM++WOY8&Fuo7#pQfg!wI(()d!z~?+}BS%-=iB4cEh$vp=4Uy zAzwurk4$&lHVAw4@XdkS$M6pi(XGf;Dd7CC_PibC_rro7Wf^YRA5o-jW0kOVvnt=- zAH>|)iFA;CGd4mFB9dL#It29+a!?d>5{mRm@lEi%`G8=B6N+Ho?$tWHrF6GFFK{Fu z9A7EMEp*!fVFyHYp0x+3F2+-p9DEu+T^bnW*<$5Xyq?h^AnK`aS(LMcBRsBqFNopcwYhbHT`w99 zC%-y*;gj|}=UpcmV-f*8YkV^J-G7!EE#K8Bx)t^sbl-)pr2!SDYR;h0%KnD(G9(++ z!a*tBQA-(#Y8KWoA-Il1_>w%j2}Ue=P(d!5#ziOWj!Z3fp?u z98(pFkMv+`7>|J$;00>M<_U@GLD<9%Q>LBTCbTQW$sPAF@5vN_DpoV4k=YMStHfqlwbA?U1sEA&yKS6?nYLg1c;_0O;l6P`_H1cL~}|3b0ew<*Zg z*b@lZ3T{ZWDV>anTZ7DsfHYWha@gJ*^GCtDQgt)6{-qq?hrN^;t$4B_OHgf=vs`Xo1p{}J1YH%f^ZdyZf1I)Nuy83-eN7ocVaeEC$qbR4toob4^`j2$ z>~gaklSw+5@uW{v+|h)V@7ML|ZQYscr@}3rhRsS!Y)dGcs)bQRss3=7TZT#HjB{x3 z8)qRh8H7)okxK*Q>>2Fd!eXwPo5eIk`WLeOl});1e@;qKWm6I>${R%_fV!OF?Tqd& zEEh~6fdG4>jfYS|hv-D^h7&-}M-~K+NgyDf20NHesEgXwb4$0NnnKx7P?*|+uL;=9 ze9>i+t4}KvIK&0JNcj0wcF%@~bfMC)<;Y%6La>Qk{@sg04L15LYk`3crH92#beCQ5 zTl#ns?Ua>cufyHkeySAT)UPLml`G=mb0+BAZ3_V-%Un`Y(0<7ll;m^PHdnEEvVz{S zC)N(AlUG?dt^J`>jGs$^nlLi^UYC++DqZpp|1}0>#dC?3Fi%Z&)ndR~5I%u?s4vFo z#`rh>`n)8qm}evv?Qn4?Cn5(qb9;9%#siZlx=2VJFhlC>Ds+w;2*)#By=v>}LS(La zj#IlRI9_>ZfHk-R5?JsDR_fOP$cQ&(5|^ZcJVp8wnb%k+1-p`JswHgPRHWM?GmZrU zOy}CTkb%&I$0l);=j}gJEI+U=%yaN>>OOg|0OmLd6R13tUcM6Pmi0Y6P1pg|WPAC) z0S@p)6`A921LjRKu%lq}+O#5^E}TqJk!)%4Ap|R^%wh?W4KK98&E``=e2d4Mt*b%T!BDX1c%2l-!+>wL3Nik_&4A5+I60pJoFKF~t#k{l(B!Sz;jp^iK8uuDgJbBXNK`S9C3_i=Z{QEay zMb9p#)D8ZlM~*Ux;TWP>&_541kZS=N7|<}r4Tu|c3s&8v-?nYB10;j#D>1hT_d4+6 zRnHKl)Lw{fYqWmw2@yHPNv0{VFX}lS4&;dpI0}#O8yN}cl7-LuWJr;Fk8TL3%snlnlxho-A4;Dq4>i&b-}a;^?~5Q(>t_M@DPsSkxN_?TCk%ebTL-}T@@?)!yEl>#DvAx8) zh!$PzpF9E0LCUFi7GGs0NC$tZNq}xOW}P}8coFU)Ca1FXmj@m~V5#0ncA2#ldv<7GwBsZKIzb#&>ztE3xaFW zR`iouq#>;e2<~^x*Of*Z->6dNn#o9g^{Z(w3| zg@pD`@iZ_svo=Zk!Pl^vH~W*WT?=JEj@J3;Nau9+X5`ku?WyFY!(TaB*3Y zD6vD1@LEwX@?efVanx?GlrP&& z2Y1yC2-X%=5qNy7woIp#!IrL8YR(o>^I~1Wg*Rs8=i#By1{F??d5JF3sK0mYRVL7s zM%?eRqOx~n`q}l`N zB#yu0cAZDzIir3~C7^R|y_XBfBR}r|^mx1;2Z)X2Siw;MJ_k_UPKJSDXEMf0xmDqv z{>iKb)P7MZ)}NK(7bHL_5nr;&HhRmpHY{I?V;A2Hq87q=AFt5oJaz-95P_7VpU}jU zemP7=2x-SEKPmAx2W?ECE$EsZOqLjPAo}$e>ua9F8Wn?8INH|6r^%4dYU&PMHt@@*HUlfZs#<$8-9XH*x zD2o31#^>C5RInNO^EKFnBowI>N7^r z%#UjEw+oCgS0(+ms!T8l_Q(=#f_VN4P|)WFXBHtko$?QW#I}kp&s^?eC&FgAd+Q$lA`-AYOrQJw8x1s%rVESD1+BYpiqeoPE58pSPi=iavSPBl11)yuF;R zGoJ>5q2qa-tke-eIN&~3<>b^HTer?S#vE-YqECbr>Zsq-$XZ;M_KI&QA9z1L-3#s8 zFj|7cMf+o+rzZMm=6~blS!Sd|Tj_pISC* z`pDYLBkOhS$s0ZBZ4{NqD`j)&6Lz$xPaeBWa7j|hTO_#^K{HztMK*dN8+R*Ko!LGK zXuryi0n6&M&6(DTdi_n8rJG0JusBnOAo@G>4}o0IK3UL_gTv!plcCq{n8G9v7OU_p z8ED&3`6)evHtHn}3o3y_PLTD{_i|ePr1(?EOb4%k-ZcXoG+f#OkE@;jOj*(xJ}?*N z(bl91e{_=tKH5LrP8_8JN`a}e>`4OMKAvLsy-sY7PMFiimEb{BG`vruVR0_{8}{6p zQ-HV0<~7?rHfe<6=eH0BEbJ6I`V585?tV1iA0ohAX7E&`J^Bk}b|O$DwQ}Ef{4TD{ zKW&5SqX1Q$Gu)c)V7MsGzV=VD$(&nDak1(3TSOw3y3gn?JBi}OXFq<}c&9(!Qhlk8 z{T~}l3#Y=%eP;b6Oc_YJA0@Ty6}z$giwOF!6n`tU^Q@xOop3Fu(pb)6_aPKKJLwH*~%HP)by#xATnuJsWbZRhh3`9UGw(Djh84I`@uMlT0N?@TF)Gmavm zcOE=2VMh0u%IG^ESAGJr%W)6Uz*E{|?hbDht2ktXuTDuBnB3`FrKZe8phB2itupRizT9%wcB5-LEdn04?r~vL9z%(Y?q93 za*rK(bQcLO^kuaS1ZN-r>n}$CG@bt`Y-IayVWWzxv5TjJDKV4k_nt+Rm`T>u&fLX< zn4O2`|IYpcoB*&s0U+R@AOHXWQcgugj_W^XZ%`IGrsejQBv+(e6(G#;WGxHD&+nPFA{xY&7 zmNRm(Fts&u`M$6rR~8rzbY)X# zQzti56JiN_I~QVkBU@AA{~8+Of9=IxZA_g&KtKS>ARufYpx;9W0S7=lha)lW**H`QQIc~4UqHIPrVZ$8Q1g(fLJI0^nSzycuj6Cq@F1~5DEcS2A! zky)blq;q9Rveu;36Uag4ZXNIm4*H4cjjH1fmhf!~($gF0>6HUa9E=OmB6^R2(h3bx z4)y?gd1eB$37<;->q%K&zyRg2#7?3owWmUi(%P6_U<|fC49qwo5D!U29^Sopr50mfXTp;;B4q8NJu|#pgYh9 z`aK>B7zlJo^n|{9s>Dc$ij7JHW&rb}Gomw`pFrLuwz#2bArakkUxafY-;7hCI$=BE z2w}mwa^YdB)8XO2Hv|w62Db^X|7kX{|F_xnzb6mJ_ZwuAu(WY8b^1Qq7`d2=nf|gj zF(nodfOU3pGBvV=^;o;qmP_1_aO#>K6_~*?IMVFugH+m9>zs^|c%M<;bNWkXJ^I72 zz0WSrA`k0?lH;|7_cbq(%;(mKt)XJ(gv6`}fev5bJHVC)`~5oN^>S;f{&C4Z{cIWY zHRO+#nfUJFdfS2~ub{9m2%J29c^FOPQqFpvz%!Ds+0G4NdM>|9zZS{T-e)U+IDI$@ z>GHVStHd&II7hmUC-5cnJj$xO-h8+X$qs{5j@C|-way%$*<+|Q>o?o-?6WiT@AC63 z{VtNN7PpzGuH(zLq`?==9;csfTfRSH?fTBIb9zj_1TufptY$j#WzGx5H2F4Jcj!F- zaRPHQGGDK*DA!AI9`^Wrz7{`$D!*i~7K5h7Lq`RTQ9EwQ z5l4sBd5LYkg*4-E;Q?twKSf6_2OG^b-jLgeO|=u$rnQ?A`z};eK$mYuJV}!AVVTnU zeMfbfgkDiBeFv?Z+E=&cMFpo@@^NI>)mmoz#=WK4{sSBUuX=An6O(!1vg~(Y5B-zn z6%!*9lhNc7fRx^*!adTJE!w#*0Yi>R0x5Ws;ut6F>Ina$#PiM^)N4DXYvaPcVldJ?Qzp zF;fxVG)xy--K=FDqr3Oz79GF3#m6aiXP(xTRoOIpERkBXu=C*0&Y*`yws*Lhvp)Qm zo9JKLw(g51uQXNs0{4cChtE3Ty|QHNP97Qv1bi!&8xj0mK}`#!%Nc+AMG^l2Crhl< z=ec*%)p+Ij!0xFbh@>BV)lys7Bq?u9tSv|RNrySOwefBK^{MQZ&kTCwdR<(Z^H)>x zD`};*@HX*;DRHB0)&I%ag9eem=FIL3|0ULU(&n-EXFrRNnA?_|6T#Utz=0|PL<*|x zV&R#8yeV&Ypf_vWn|VMo&66W@zO#uO&uK{ixBs_PxO{!G3m-5sfYe*AVXJYughvhy z#MHR(T`wlex?H08F!ayyy#L&&b<$%vW91Co-D=#?uHC)~gb{)z0MVikH*o5825;y5 zTj$hPvoCp4uqQ+`kT}g;vm>Q$(yvA6J`lnDbZ!+KvlowrA|%Hgd0m@mBuln>EGlEZ41_JBy4&O3&<~o#z(h+48NI_b#|ec9_xnBA<0vdj0G$K9~MmT zr<1ro!_8~)28#S>8@m;npv&g=W{H+ z4VIZQelH>l568B6;SBAS(%`1=0_TGWiE5)~#2q3f^@uTMvauOKX9}M-v+bzZ$H+~W zAh}Bx!wrM@Tacf0f0&_~!MOa?Q3ypC&3o4dj0LyE@teYH<_w~dGaxINgnYNR^K6R` zaB}$b{L|QFJdU|ym%h-acM9fXp(+`s`AcJ zq>PHyWIa(xCPjFi*xlE_O$0fo(F)2}up%hh)M38tGL*H=VhP-%G`v1*NIO-2$3n0V z4X4CF2$1Q>C=U|$H?K2#e7 zYh3K#PsxYC5LddiBcs+bq>yDCZU6B?>MEWlK#wQ)4KIdi&Wg2>A< zhXd9TRDQIfx!(NJ_zbB7cNbv!t-_+B{_l3eL%5W>ykWg zObFsL$sB0xIzGt&rcmJ`D2s{ng?2A^<`OlEc8?Gs`igPz(e*yX@V$o;HXwP^6Py;pBW)eXpd* z?HCs-7(o^wrxmea5lTtV$I|Ihr^T~Kt(=tne>e753jEo(YKK43;}Erp z)`R)?kZTuiN_)aAq_x zaCAUZYw^5n7@&6fy>vDB5=_&MgkwTiQd2l(=}_NugRBz$Mi9!h0k8GIwmy@9)bvIt zcqd|xydSot9iE8Mw@YnLH@GVsw+@Au^bhwQl@!qZ;JV_Jp`yY@0a1sT#{=> zb8dO5(P)MIYt(Y%e!>_y(S*}m=>OpIPKf6WDM6WCLMWI=uCO*F!xDFXAqwvJWZvx&8FYhHuCZHC0H;94F-8t6`}ZKAz!FC#2@Ox@1>@2kfkX7{*SrFTc)3G;wpic_%w7w;;(7w?0yin z#4S-SQ$5B|eny6ySgK`z5U9=}7f2EsVMW)wFi!+J!HA_RhH+>?@aAz}dU25?E9pNo zJfb-DIOO>x2oNS$9c4Nq>9DET>``82-E1VTpM9W6T$&$8Xb^-(x20Q=l7`v)gF}e& zPz|@n{ihcrr&DL9*9=Z&m_IFI%GeAE+Ly9m@h=X$t51AqI*!GAYJjE|!gP2j9&HzE zio*lFw-^TiL9PYntjs~$+*1BsAIQs0q@kl!%BYxvm(CMEyQ|rEWUS;imERkUWwNf+ zLI#k%utp;3hx3Ya)^8?w={Sy51(((<4Vvr#>sqA4YZ?!#SL_H6{@oSBh{F;q!w+At ztH7xa(8@?2t!Bc{N6|$er4Ootv9P`V`Y488l)rxr+%GHWILkR6red{o_u;{Jk0_Hu z99&JkEtF#TVJma`imPiYnrLd3Y0+8e9xJ3~gR54_uqRKNon6BrInP}`UYSi6}OXRJ7n=DtE6 zosXb#gh`5y=q{l&zV)R=($?x|7FLJV0^Zy3UC%^g4k*l)e*A%B@w@IzPQia7xrXxL z5JiTvJEy5k)yG&&{VqKBZYgELiv3^DLe%CT-F<)jeTZY1D`vt>Vs7bl~A zTQ49)m28$9_y0-VL&)&J*6T>kzCCnk?V-YW#^llxw1%~c&E!Y%IKW&p+0t9FPjuX< z-pYjRcR{eu%^z4&nmKaZ*RtGi)h~ z{!^U0tCL@>zp%aty0~>VLaG**pkiy*1^XkZfOhEo!we`$uO|Mtlpu@@(Fewn9a+lX ze^e#$0Pl@BJ_-geYLAObz4%2!>4Tq7YF04zvzsvT{?ZL^>`9W9|MPjg-sZFLU23!8 z&Z-5>n)sxw--u9k?T&q$Y~II{yR0`IX@Pkl$@Vz{3Qp>^tspIkMC&*qNug|O2g&gK z+!f;-rZ#>28UJ>?BRQ6zyhup$-}nHOX#;l9!%R78*_m#l8dD}lPRE=lm51+Y{Aih= zB444lD>mgqC!+a06iclp)gI~2AIdJ>{;f1s`j|F*FE^7(;F*frwBxhtI|x{+FOetU z;dHV3>@5EADZJlJV}P?%y$z4(XCgL-N`|oy-0lS>I=oY`u3pWO5G9DIeRy2Wny(tx z`Kb)|Amm}#!F|OVd@8h%%}gM?m94%Tw0DD!?V)3I_ctq!Q-(FHv zs4Yn>Tvgeg$A|00Qs=%je#{@)c)OP5E6HUD>wz)e;_=YIRJCGQ@|Xx(%*qP~957KO zaquN-f^?avzr3XuJPSmkKl~hJl8Levh3SZ@t)t2y{!K)cA37#r5&|KH13cZniJJs2 ziSvfUHxSjr2daplu5-oj(r1ZF13S)~Hyy|k2F5ct!UJg`cvviLQ5gIVu;7&#Zbx5HxUpv8D6A@<|b&vWyg z{%J~ST=;JHWE`X;4fg)wp|XY1_C`AsdrBj|{C{7*$ZoiZmF5p|gOGO^ca*|t2_dq6 zsmYl_!2|ej^*oSYJs}#m)WnoCP-{;4M0#Q@l+n-N#LkI(Vo~JCUcfC0U^oqNi&=UZ z0yF3(%3>t@6LCrp?CjQy^+LX^0S>SErZZ^NyQJ_{s(l8#@_nOW5tt9$D%mjhV9qy1eK{IZbUn^wDWE{%l zM|W9#c0X7%odBc)izZ7%Y~W!t)O*W=?Yv#M=76N;`*u=dUW%t#lqbld$C<^ZkbN+^ z8tsDe3A0!saYW0f3-u0oxN|4^tJN*BhM5G>=7JLIh{R-Y9yr8@RJ!Pg$lq&n2tq&! z_*^^o4ocy`F&7-Y{=E;QD8cM^0vH!H zg#Rh$;rKsu9u8LS|EGTFpRpTKPTfy>rFz)Dto861Lq2S-)|TvEQAM+EE#_Tx4DfCk zi$h8ns{_W;WFIeKhVqpO4|*Bm1N#wA8K_dLxZZ!^fFqH~06a z?g|QK$AY3%f~SS1>gZFVJ$Hxk9=!;S>5UHQO#;gsdrf8LW9IqLGP)C=Zl4!x4gz;` z6)T!q`wpB9+8DPU)eZcqeRHXjr*Jlgx1UYVEvvfL4Dywgt67w^VjX|pn%|(ODcy;& z9Gw;yt*ufI>ycS|_yo#yakQ5U?+OJ8EU|ZdG9RosGsGf}r3a^zA8#*jUs0#M0fX5LBVKdu`mJA^CIVSkAP`DPn@Qou%pAaN znWCp=H}6c+eO_U(ti{h4%Ta@EB(a5a{FM6%1*;=g{t^^szbztpY>3U^7qW|Sfk>`? zQv@4TaB<#Dvs|8zRoWTeA8Dm?vlfMJDp_On1YFFI4rXoNDAmiK#StIYJ7|RC=BC(! zelc?sOygtK5Je*qHmILv%x|g>aavm0zcTYld<@t;#edEGm0O5G&Dmk9vkg@~qTBA> z_UIRb3B3;bOIhZXi}Zx6L9z`NvL{a7n zrV9EE+}YO-$qTZ1^aQGX7kmZq>9%H7N9;2a4JDVISb~LE<=J-`%$gT!GyRK5LK&$7 zmQo%wp1hGtx8Cz7sM73FJ{Recc>-u=4_;5us$3t^W0Lo8<)d!1lg!v_mtHCIG(`ZS zOAw;7)ktrSQosaS)!q>;Tdx_Y6#Vi*C$Gog#Sxxf9^ER`jGHO5;2XC#&BvOTsA zEP+W9ez3IYhw%8$2zC=k1b7!@O+dTLqQ%0AjKxWwy)(s>*&O_t7ic2y?=X1HC+mpe z9aZ^fMM=R6%m&1dGi@9NLwIW4Ah9cXa_w0?4^9L;>ORMN3aQif+w-VM(w_c zxbD-#VtXWla15TRZ^Rp$@&97%9G^Q08>}7Mwr$&)*tTukwmGpqv8`Y1i8HZnYxBOf z-}b56{SUgUy1MQ@ea>}pbBAZoU{%0szD^Se1y6Wplilu9ze%ArM*Ni}(&;(x+7w1K z3337F2iDeThx4EsS1d`@4&?+vx)eYv=cQR$ga9u?vi{@@(KZn=zjwUot`;TE#H zx=!-8NzB>c6uI5`Va9|OSgalphG;_jcTvUX0yi~WnTLj5b63m@hStNZ*(E=Kn~-Or z5^cEEbn_05g(o3FAzLrtwM1MrIrh7cm`7AfiDc7?vohvlY}x`zHT#BoUJuVU6#^3h zS&=KA5}_1$y<>k|fv(F*p2W<~!Cf99tpR6GUJB&={%4`=Gs22xILyj9&@}2MQ-f~z z@;OCvD+IE}#*X7nl;`s_S*u@1%i@93W&X>V2eWSPEij|`d==;;(6j!$>anvo#yxHX zHoUn;i)&}|&&dXM?+Cf;25@=VNE8& zV=?lwtTXWS?y5M6hH-PSACfjfYj(AsLh)7t7euNs2B5chB2dvxF~ex*7dCmIW>M@x z9Gx^(s|pH)!xqWBXmV*!g^*v6DhVT}(4$bC)3{X#9IsUXUmB+M1Io^X%?&IXc zKtD!1RK55@rtwBb0(+p911(>+@I-Q@OqE+y5x4b?xQi1=dFD|~a*d&=! zx2+G1?-)UF#ceb7mWxsv6;tf8b8E%vl#(DEm?>mW8gqeth|bJhi}yML6KovU=?C+jLG+D73EP zC`LCiQi9(W;tZ4_eUi^=&d7F2hrVT0W};7EME$e&p*afmfdMk2WhtwpW(q*8Syf1( z-yN}D@Edo2AvA~R-Bx5ze1|Q_(>3~A3bp^ehY~Q5hh#19vy6KXkgDSZmFwq&Dmyq& zjoAy7!qEl>4wb^OGVFv!!F=?KsHDyC+DC}%84Bh`{g59ryFc0QC~j+o5k%xyU~rJ4 z)0I`Tmx!odVviu%K?FDu4GRnI!VWP~-X7^0G@Li9fVDy+DEgDA4^YEz5{9y2^9>Eq z5Bf=X_C_$GBs*UoK*vmU=li*+xdSlKfb(?4i3A z(ww`6I8wCQ4dq$_qoQv@m7Ipzr}JB2A@wGl;td9Y!g1KNMCZs5r^q?;!+$v{l`?^h zz3(@GkY{XFFW{Jm$`xsWix`kjDLpfTYoH0%Wo15nZs!7g_!c=_u5^G%Zro5Q0xSB! ziQ=nyLnD1Y8ZwFY(3;$@iM7F~X{P;c?kZ_5lF4LgXo$o6sM`GE>I>3<*>OmY5la9C zEtln;g#@-cU^!fDr9f|I)@n``)*rSd5?a<}fP4#ZGpF2>vq^Lb(21BYro`ZevmlEP z&!v^@H`IZj<)|qV-q)Q2EytlTc;Vm=z6_pwsL{rr=nN%sn1)lyx|ohR5>sK;zB{ z9Ft0Y6V3;o)35PY@J1$F6%<*)VZ_6?eLAJ2#r4?V9wJi!&d zirJmJ+y}hVYJyyXur(QI23`V$o~4H<4E4(t80(`jL%5g%t~edkaeouL=v5x4?r`Ze z_7z_l?1G)*yM4Cl?k~PgZi>Sx_Tqqbpe=GkKOLNwe?Ne|K> zCdkkv1SQ z@PYZY^%Drq!^LR>&N3fTE2ctIC&u%T`iRb&Z-2mUTg1G=5)&34+e$!SS;O?k@$hDm zpNZuodZcbYzH9fKjMm@(@*Qyr0jj>~nTvrl!yYc!CIW+d@2q1-*0p$^pZmWjhkpqf zO||%fAQiFV~|IC~i0I*_A)YP; zJA?#{_Q>Qi+hPkFWqInyi~8rQL;m^JcJph~`DJP_UrMqSf!IWi2@mGN_qD0FoK!mpc< z0inwe!A5c+PoN2V_8{ZZxA&qylgD7wPcT~7*#PCLq;S6v9Hu` zg6N)%Hb7!No2=!H^%Fva;WgYob_#DX?kKKSfw_TcsA+vP8!`;>iKUwQXQ>7M;G^UK zZ6a%b0(G;;0ZwJd7d+dYGxb35qwM}%48kPh4YUi{s5w%Yc;vZ69M*PYAIHz%{YYWD zc`iOefl^x{zXC&R?w1bpOL>iVck>wTytV*O0&lP(>aE7gB+s-rsEt59n#(}ITpe>Q ztH7l}-~*o5*aXwh9+rua+mxFK-k;Z^b)4LD=CfJl6!%4TlQ5qYLsg|>dJ8+;bfQLF zc_eI&0sy0Q`*K{Y1mVP=(yOj93U%aF96ZM&1EuUV24WBVR(T2e>h#v(Sn}Rf-U_O} zr9`5ml2a>`-|Z12mS?eZo--MfZe$>vxYF&tF0y|J`d)<4$;$S`xjC| zn%A=N9iP8@5JCE3tCz5G)Hm_yGt+tCxs*~CM~3DhXzyP{$ta!aqVyLXzg&l)IM28j z>*0;~Q{5D^e5utseaYyX+|67%hS~eOZ=VDPYON~ameg`^ z-e&Gtpl>egD1fo9Pf9tNM6gJ9QH&H3gF3l3t%EhcUU& zIp$%o!%}R2rWtPEjvYzQsBv)Z=E!r!d_uMRPXl~OZ4Ei0M@ktm1y6*7J%PjAgij{6X8=u5oP^wB}IJkElMH_M(COFY+x z|BN-P4oFLGehudLZ6#SjhOJQ>H(S)ybcBx4%abc&x^Tl3bn+OkPHpQzPJr377H} zj08?cG>tht+4oN_g*w*1&qz3i^LHbl0RpP&O`Ajl$49{{5<6zz>>#s(b<0l<81@(M z@1NHfMNeTMcuvLi=Zbz^bSG#xcYkGnU8e6fi=RFk68`ue5PVkpGCz>40dmVIEx`>d z9wkbtdt0KSXpWk3t^G;7SGr(R-JJcIMD^ruG*}a}S3AE20Jn#Bi>=@%gr1uuSL&cFkh zI&kAl0*p~z6BG-=voBNw1X*XtK>DDCCT$;PGox0n{=qW#UkX)jV)oKfy$1Odstbpb zZXH!T%9eXI!rc{^c%Z@E>PUM=0p=vBp9I-e6+O`%S518$84qN6AIDW4#IhO%Wk_U* z7!3?<8*i5qtM030b?_4YHI9@Dbd-A9e0jA>Be*wpnp3hEd(j~w4&(~EG8ncWNd6y zhm;U3xH3-qJ8vi)M2iFn1|^8Lu@h17>sK2t%4%5S*c9A5=mVD!8P?awY-2i;H)Da(?hhi~M+d|DO zfoGsM{PZB6?PnB~Zohsp?3Q#=NH4KiQGj2%XZq0&cMVjk=X6m~iFh@yrM?_QTG|pJ z;0ZuP)`!bZynFx7vbwFY6;NWWW$;qB=Sm3!i0{4fq*Gba3NZY&B&JyQj3oM6zhW4h z#DRKLKdhqVG*cmD1Azx}_#+)a*I0!mLsT-6wxKbQqPJ~4=pE!l>_Q1AIcKq>wmskj zXK8-HhT!TNHtAHb`?f-CH=j1$rI~dAduzJ#eDO?WPzkti+$ftMXZ>`E?KNl#C>UIm7MhWRKH zd6&W32_huvm2`#vDkg*w_Ow$TbU@D;i=v=;32xkdN{qChlk9qjFszLh;FN4E5|?Vm z5>0lz0j?--2Hvg<{k=y&bKMR-`{;<8r(mE%A9Ii=RP(e=y1~@`!Jke)sFB(ZlGeD> zQE!dsuxYugMMDP6lx=Xya23!L3=Df>|M6v8-P_k0aIGH|-87?J;7T3X>oMPY-8(|h z`?1f$pBta_^=v10s`E7>2lCf#IaX(FP@e(#(jD_B>-<;!>I#hC<*k+BQWGVz5}MY9 z`P_MC2VT||=`@Aec8XX7CWF zcd~JFu>9|}wxu5-GOk2&&(9*8i^6f8N)~ClwyxG%kAo^Ey_DH3ZVdanhK+e1Oxtu4 zD9awGi2d6={+;lNiLt3l5Qvkzv2@6uOT^xF#X|jPGDB-TpTdfSlKK<#ThIW%H08WF zKc1{Fcldi+BZV?Ug0ik9>p&{6($@L)is~T|HJ_43!+NigZ+=;xK|^Bi_54^NKVP14 zyEp~+VV>nyz~#iFg9}AL1k?_KzHwMP<7{p(uHxtE>6lGx&-%K&0}MDaIQ_r4Q+L6U zqe0t?`}ZlCoSFUWq}Dq)E>?E{>6Z=9U+tKc!%No@R4*lmD!WbJVTTm+#2Bs~yDIWF z8RgS3*@SmO8ny5hH+Qdxa?&_p3XjB{Cej)%Xpshv@w@S}u`&*Y?yIdIA@WsPQ>bdg z_*CqKR(aH{OY%MLtvL7LH%G1(T$VJYKBXoX#RPFf(~B%@&uV&Cqf`-GDQ@hwT2U9P z7wL@0q@7zA-8`p41M#h!irr)x&0q|h+ur+zT}#!Ml5Ms3r{|Bgn5;5+@2)II8>KE7 zEjxj%jBV&b+*hy8-b2S;p;*#is|NEWNbgWzc&u>A)QbRNX08l$#3j^>CA?ft$Xv z2}4kI@^0U-PSdVqm-wY1=P~8?u)Zxp8cg+^`0a3A)>~D3Rr->9RYlh{VDyE3#VBrF z^5-hLSGQ7WYaI0HHYiY%0om~S_Q|Vgj0TIGe2}y{T>@uG()3@&mpL&p-aXv*eV(_C z-7!5S=mEKJaLymM?fK`21_>L7mc%akZjf0D1w9ODHB8NF$CMe{_xoG>PTWcy*~!05oE|DMJXEO2 zrzRpBJ9~|!$}~USS=|#+w2fS0L7>-;WO*_)!%Uo`5M_v_XAs#os20j|*t7FhNEyml zFE}IA4vlr>rObN5YurPg4^!3T&5Eobopx{b?mDpPxP+VaPm`LYL12TKCgK+>;5)>y%0|WIO~c1`;9Wl z6wcdSxP0tO0BvSXG|Rv)=9i&9dtT~WY??&a`NejgN(^H%{q7P#OHvh_ovp{du$n{b zOsMJNK$VW#zQbIEa`7Z>dyk9BTKgpva)`x=Xg%&7ncJ~Zqm#o{YhE(?hDhUfPI9Al zdJ{u3u#0QIuv)HOrLS>FA+u=zJu)^yz+^t(WR)+f-H$SW4ZaKTIP#z_6l9UO>8VIqQoay9_vTsv)-j5nn`CAkF=5pvMsno zbJus^Fx?)pQe<8#P_uqxM6ru-vh%aLa|u@yYZ`&VOu`7NjE#S8K7V}>)3BWtwJ=A8 zF~F&#-&_I}S}-X&2KpLI%9q=sc4*d}+qu@7NM52A#M;i{LAc7~rK3T7k@Pnfv8MN5 zY^TX}^_I>&p;*IRcM zI*QzriU@9!3sU~nJRNk?v^7&aX%Yp*C{cpz%uigsQtdf#JYJLpiuB|!oc1f|CC4$N z&hLQ>qUs+G_iIi8{YuO+=RG6o53fn;<=q-Og3 zy#;3G_!_PY=MV~+rFzcRK4ZLB>hu5VH--!xf!MK{ZoQ9=%vdZpDdi@$NZvNX&7G!Z zYT$0c1Sp#-Y7P3sr1<_*-M0m!r1v@ci(DhS+BVpoE3jPYTk`j#{n$mQ-|2|m2698^ z#evwc+L!b99jZKjCP03uX`Zbz8zp{Bk%s&Bq*y!6pClD0lQM>>lR)~ANUsSxnvI}V1 z5Ipt`v4A(T?J{b&DN8{~dI8;umuPU8+>v%Tp9L)K{>@vjW$GWK_8+OKf+Y>~MSoOY z62%=Pjc_-HiJfvM%Wrk2WK}}K5>|uyPO`xFb-xo4XGo9<~?9z zzM`%AL;0RK+ocT?@e>c*SeJ^B0by;mj!&+t7Zf-_HJ|7@ENI(h$RLL>StAh39R6V7 z^&sR5HoLff3eeUUtx)w4`SRacnnKijC!iuWo8xv3-B%{_WiN^p*T_L+G{x=jE}GTA zJfi+Xm7`j;;DXC9->|Kly}(KzAolWvo!sMrIREfAES_Lt1gw zi?(w5z}U7OeYSO4j3&cq={dXPHnd_^)OxVN5y9hae9K#Xkl)!~1!iEmh8kXIe18cE zl~`>#QS5^_T>!@T$-u`zquEl>($<$7I;g4w#d!`>R;MQ2=2hL^#iKXKi!A9vl!H5r zzCSxZ6rJR0-C;;80)HRQZaDXJFm7assA~?03LOkI$zht;O#o5uAgutn_~rT)?}K(} zNd|y0NI{n-b-$?Y^ZG);f(L*5GUIYWJP1rvj=T?ZjfOA(Zl99Z6l1TDPrnu}^~_(a zzU%xc{#+yu6$r$CwH(zdj|p3#iY~`$?(4fqFo`sbS?IjK_98Uy9+9T&@(nLEL{8x! zz)`;^;jcGhuihC{y|yBxWbv8ux8lW-7}BilOLd}U*)+Oe3jRYY$e|#I1Wv1)aKm-B ztFAkG5Nj&Zh-r-ML)x zGdVV7YY@v#ZxJ>Uy4g?IWCBEqHPbzJ)M>|?0B+d0mr>dA0&!&^GDd(8@-^~%CIh-s z&_6+JAR(PolcLL}NB&BY& z1<1hCE(@vI8{R$C3aMcUpLm?VgN^(1UC+BHTR@wDZC>Om@nO%PQ3(yZ_oygPHz-V% zfKa1dgWB~OxYuCy+kYBsPdOnE%s*f6%^R(}2-;`rsd4*8t`yC&||Oy07jaaFzSQOh0wLZA@S zCSnPRAKmdNO5ihKKr^}GG3AiPz2>JxS%{iLT>p|KaJz<*hryn=aR!l~ZDAK)iTKwUZW=N`684wYSk&=^;qi z79_d6<;zUxeH-2STe>>jPrl27Ja1(f8v}hiczgh0Xt^`!Em^WUF4sY~{uo$syGk~1 z{Ct&1iwVv&ZR+9Y*xF0Ie2-*=-CKO7N>6fwHC6P-0&;1yuzBc8%HH48E#>ZVP#uK6yX?yYs&v0x~hpW#cb%NNAtRMpXCO`^KSXvgoYL1AhTwG=Llh>-PqAVl#DS zNY=?zu!5; z;Vo;Qv9FJ3_=vMg;6ONgf^$L-TUFZ(p+YCUl><*6Z=ZD*+8oosJ4d@|)9PdGhB&w= zN?gnLcZP~`%=+$)yvW5#>o4X!_?h9OHzH}O0M7Q05UDigfa%}Ke9A;`z~=(Z*ctv+ zO7YbEGgIEL1>~o?kK|x=#zhDWEeKKIG8$-hdGX2cATk&3(ZR2qc)LJZf=ThUW9zx@=s*`98@V8B{BmWnI+-Z8%W*MT37 zkiW)%EY->-42-I5cDKRDnWF*BT^e~8EAnja7QX(8D~@jm?s~Q;r#xK>JrhAsPcE4_ z^$_N)IjmWndT3tch|?SJQ}0&H-A8pGL}pCbGfQC$u{$%TT1k(Leby2~c&2@)+dQ>yh%h~buU&Y`(Ti**i3^I;Qp2lDV&3#( z|5-{D*6#KGQi8xz@xd0RkpWRtI$|!Gl3;23TiaRuUSs@n4`>Aa1=bif=$!K<=Dg!q zJrLc9_5S6kGK;D&Ga-hazl-GpKj~Djoy|pgo#2dzM|pd?or#y&f%gx*MOd`wAG0gR zRL9u;!Z1_&Q(yoH{ap6dikzTiqrb{3)J zj^A<(^c$Qg73lW9qSu0)8h}o4^jy<5;ELe|N{hWUER&a)Uae((nyt7zAfpU-XGI=U zKcd!IFPHd`Nle|B!<+%G?C+>+;w6`4$H@8NYP~gE*hU}4bWAr(jFW#>8&f9L-@0kv zJL=e@hU8p}YBgCv4VGeok)KwZY;)P2m0d@XUl>Gs4ak=_QuAy``cMJNOwHMiP(d3t zo9q7Zz{9aF-NY7iBN8P`kKXzYmJB8;?zakAo1OOA)>+n&?7I-r{2Fxeu%)oaO9{jj zCA{Tk+0QdS>P#)Rf7B&ogf!905YCt5Rvz@7FN#Qgg<7C}yPBaQYPDQD$8=i4*-J zas=|9qDYWwIdmShd;YGu@cCJ|r6(H8>DIAn5y-L@xoxiuNK-=WEtw7;yXG;;&uop8 zzsY-R+UW4|?_e~!@<8uSHrg+yN7ri=ppePKeAnYI7}DdhYwN~BulAAt5Fpov;iGJv z#*77*G-&!Lw$5v$0PfZ`<;3w+fe9*}5NPB*Wq}|dxOVDX*$m>NQ4}~F>xoQQO11@C z%w;FL>#$Wuj{Fr_)4%iroXSzMtpNzDMtWU` zA*H(v+a9eripQhTEl<(WG1hY7#mxz3cW$EOXFRDGVJL3+H?Ikihm)6b?@J`UHAjc- ztl&ajU(v5tR72;C7g&AuXnCgFJ>kC?G?*N?caMnjX~3-0LDaPKU^^NlK8eb2>Bajh z)C+y{14h9>+d$QGJ_Y_6Syq)m=elH`)b5tpJ`8D=yFy}LAhpWyFx}c}8T<^S6oq&{ zd;(uE+Lb&!57X;yXuR#@{IN|ubvk&N=+V51ChcH+`af$Bgvwc(U@;^S4aE@d(OX&PUGIz= zroL8~<=5VH3e|x-zfcy%9z({t#up2ioCob@cdSC1RU;r~t_=e)lKoR{xqxui#Wn*E zd1=8LPE}>8;WqQ{r*oY4t}f}L$FmZcM4%GuDlxUGF{>1CrL>DBDV1K>!`^HKUIETV zl)DT4?9a6BEM?&}x3#a`6b$JJ|9CC>>MGhc4b1J7-8|IhY`XbS?2oJ^EMzHf+HMSnzgz1nSCGa6S~*OPr%7doXEqv1ai%ZQAeHmgz#;+zV!9Nw z&OCy9V)6&@_8X>N>k=U62W`XAE;LCAZCu2$=OK;0zy<*OnH!_MPCA=torzYo9N8w^uYxWO`P zmT(k|*)d-C(@->DZtO%UwgArU-=jXk+dGlWNV7a1tlP$`Yvl+ zVTWgz)|MDC4SA~IVsfbYu4}Ki^grf1l3f!%SHk|@-$;mhaTc{@7Bk`aM2HctoZmXo zq63t}tK7YY8)g8twdp?lUL}rrl;4|pnlT#s(8$4r#>j<`nkRi0n4%G5lfJS|To=kk4a^z{s zfIj^}Y|>V=by%iCy_O1{eg|Jt!1UC>OgAL*kn(${-3zKni+#Acsr~EP-J+8|m)frY z9Ksc$B1w~f%l{Tr2LDL+a~KccIH_@>j}u>>ZRTmJo2~2wf4dTfty?%^oI!VYO<`X` zD%^WRoT2eqQtEitui5=wQEQW5R{0CDhEWw|3|Hn?-#DI3da-Fa4r$OJYEuHy=P}D8 z%_8|$%mf%+E0BPJ7Rif)10F7GT0H+vtTB z&+2Je*cJtufd!=;8{)-`!{ils@LEMs0~M~zQ{vB1;Q^ueGOSSmvw-5^igKTRJ}+(B z!H$~FOX@gv_#=x|9h9(2!de<6pG?c{cgBH{6yB7Cptn6B@cFdp7W?jo0s&X_9V{E3 zymnhxAiv!Rus7_vs6FA{NG-;2&yWihWIgZ4n7M^;l#1B!H^c6M?)2!DZfpD68|ARE z^F zg=_@-andb9s(KLn|4d9G&>i?*_;%~B#F(qyj)7nJu-s{__-hZ& zb@{eYvlK5rY1w<7mtbQy4JO4@A6>u>*mFGM#ianJk1#_!tJI4crrpx7^U%gC7&M@Y zthW$;2;51d-jEt1M%>*zgog{;0pW6zCWEn@(1J-R1%80bf(5700V!WPCT`$H>8VwX#%0YKiK?9t+O2x85h!>N#cz zRCs~&1&{8h(cv>iJ?ii_#Fu&Y#r@deWAZRO90aO3KR?9x(uk-+$h&EV%-__=co|6) z^j$H=8#9fUj}r()FY5v<^mvJeJ>I=F$ z^rD?((6W-`iT*p;J4O~x(JZ9L%uU36u3(fN^l!TI?#ad&-B^>JoUJ5SrIKr#FmDXj`3 ze8iVvN@Y9L3fZ~PU}$=Cf5AX%TMNK~QScNM+Oa-ltw0lK2|`#SzqxF7l5VYb1ApHs z9q))S;j{6drO3&u_2(Jc(D}1R4JNJND*H~5(RDJ%G>DH%m!DDj8P*UJ=O!bA-(Upq z$B2MC<4~Vbq+_><&^0CG8ZE~BrmRs|VS=>y!3S!Wr)l02R)PwZe&%!zMPyu_dH+r6 z*wYx3mG}4e^S&(ya&@IVOdKU zGhJ|PNAz)t_M?5y)~NT7RoQDXh%b2?EKjdUQF-akLLQ=h`a+)F1FPuQ7RjL>6)$UUg&3G$7P1q{IgU1RqE>DVy&GX`{X2@YKVb=XAtZc@n`CRIx6@1*Bp zS*h8ALLAr@pfg7@8nzgWJPZ`=6(XBkx_IHh)8O9#H;w3R+Ar_doTQTR&I|O2>2!i+ zSF%U4#%n|p*1dbH1(;P0v5AAkK|k9*-GaTTYwu-`riT6%Lax=S7l3?*hZ4;)CwHNJ zaQ~Qk-P^UTzWsdNv{qB9bv$to(h;rQP7Cqr*`;&^%~>uElPoYqlCMqTcl0RDze~mm zc0MmV1>L6}MnbZ!4A5%q_vEs2F|3)*IjsMzx~?<8!v_qfObfwmi1haKOiy_EL?!}; z7{DS+yQ?_#dcserY`OsRFM7Z?N0(-gkjbzxR5rlbY!D~U`<39QR2y$Zn^M2M=Go|3 z4>7-EFS>tEr7n{X=yRv$wH7!fv(hX17AI1nLjr%Bh&W^E@w`ifg&s+33+lmMDstI=be5tK$J0!2v-r&&Nv|n-~gPE{1Op|h;B4Z*^5?^OL(5uSYFGd zB;`K`Nr}DQUeM?PuH7_wV~oKxLMN=IWDx$58d_6`A#k!nla?jTV!{|3iM|{5o#tZ< z;wIz03L%9ioQUDo#sO9R4M>h1uw7$_`{c7(7I;JVW1_cL$JQ3+D3&SI?uYlKgt=L- zR1ZJr6)Q8K_%c!FfStqvkPhr^N3v1#mvR5i+);cUPrX3P#w4#8e< zOf}B}N&UkZ9gIA^;Tgs-KUDpWtw`rv41O7=i={cEzN0pfj2N5cUP2a^@b#+i$bu|&Fb06qNS~dQ6zelmq z$yUR*>rjz|pJ*(9ooR);e$m9!u6S zeVD!%vV;R?3@w>FzjIBFeF4|URE#UP3a1)?Qv|CjCDDgt+Ek9D;a*xd&8D&XA19xz z>1_l)2HzhGk&z#Sp^H_%WDr8zNhViviOi_nZ-1)W&d4SpnTS4{&{QKn4!JF5zWb#cTNPI0J~+eyju$t&$hp;#yzO?JJgzdrEn3|3TxTKgt;jic zhze_mkJRT+J$~PkP@R3jfE@cWi_r5tNGyQh@Wc8!tQD-RWT_Un>0C;`0dL7%I6c%i zgUNrp%3&nap(qQ9+%#+nQB6gFm$T-9?Q9SeFCYc(hBj~CXD)<;9itBDJ;vPw2iwu= zp$7=wZy}){!evHQo)kMHan0@?%7Dq#$QUc_UZ(n-3zO9z^~6GI4PbxK+Q!94BQu z5#08$Ey(yx7DWB-{JtQ~WP8`0^UF-}ms))WQZ>=2o*x=g>+b_H|1bX+X*Z0%)6B&t z2PjxUIqNvalSSEq4n)2^W~c2Dbs>*DRj`z}+IK9MTQ6yfdvxGKgQS{5o$O!B`1<`C zCa^%vU6iZbYfLzmk}F=RiNiLAdHpdNlg@G(0?e-*r_5@j6>N^N0QD>9syxI;2)nH( zvCDXAaC8V64{v8@lxJC?(%_Eo}s(-LU-Yue~tiT{qEM3>Mr;O z{k=TZIHO1tv9=V3c=sUeInWtj4Zv(kKcq*q@S^$s05TPxh({wC;UE|>Pq;kV6Lwo$`!Sza# z<}QOZ%X*fSkiv6Vs#MwHIzYdH=Z)_KGk@MvL;YC=Y= zJ^8qY3k!N?gp!@IP*8AsGh+Z>jNm|SJlxcdjQI7}i)Do)_08pelYK}%CAyjXnRIEN zfyTr>E9-G7x9)~Bt4vFE$a>i<9!}|54cBDJDtjqz(N61}JL7)cz~FZoYYL3W`5!po zrS(;OVA4$}$--ULla>7e<0PHe3&SL>IV5^#x>*FR(wJCcZBvqmf0pwVkhp(Cbkto% zLlImJ;Ng@(H$l&Q>)1?i3~=-Zu?&Ir9+NogEdR8 z-ZjI)(VE8#{b3f1DT6~$-)FUIz+66H_JVDQVVS8)JqI6u`Br@;q(%b;J~u3zlpe@; z$1X!9SY**^8q!fTx04=kfyQfg#4jUj_g`VaT+%#Uo>mr@_mzN7$J2KFjE(qfvyxDc zrdU1iy(gwI=|PT!yEUwdbLrnnso~q^e?5Y)^3||@0zR>+XzG!}BlNP8@rXk0;CpLC zRSj5=a(x~Y6W}5osgNt8;<*=n@oFpS##a z+s+krr~tFi_7&Ud4rZa+oq>;fs;*9LS?tEX$H9_13nG1~)V z8`=M#565(E+&LOQ1XSrwDqGevZH9H5r$%bTdfjc9QB267e?@1(I@uB#vIfpj&PpG) z|8Da8ko!5SH>ZTrF;M^8=V@bJQ+j-6;(R`n!?{LZ?Y#39&Mpmi2bQWTXzOp0HDXoFmgpr7r0vFTrQ<{ru*x?~V^} zj)q65xxEnM@1x0HT)q%ne%Avv2xL2NQ#UZXb#A{vy>bkcZeyh#aEWCG zvmn~#iI(dgReUHpNbd7z6^HA9OKSM__@`Jh+!~n!*ygsVvpyjEU`v*RNu}OHvN$%I zmt9{nw+c5SJP+sM#ZWf$EMShnUWbHP?q6uQgDB%p^uJaDL$irS*3EcUZpcQK^x&>b z*Wj8X5jzHS>(WP?<(%^QtSn}@6mS!G$vibr#U7oG5kaeiNpS(dC27!f?d0xCcfk6U z-9JXhBQweA-;tZe251~vo(MCRmKtdU0ZER3>Z-X=GMPXJ1=Lb>vI-0f$N0q}<}cdN z?V?aJ3 zPNu*uQj)}J@GQuE0pEZlG|w0bPMpZOny<({c=?2t5W+z3s~$P}0eJ;^`2`{9ZrzOO z#viaBN@;MXa7O;pWI{n?drF9TM-JFvs7X)Qxmi}>&o6BL?Ku`GL!;0TH;k2Rc6!1W z)Y**`@~5C&0bZ{s#w~v=F^ETi5HyhXS}+R;kt?yYE~WE5P7=uO+t=$rPMY^=*qTQH5Y{zd;@+E16fY$S7Ny|gL=hzOSj2!eSl)EVrAn`=} zFc$Wqci{%b+=Oe`5&+VNN?^Pds(Gd)(IpZUZS<-%zkHA&LMLwIC@QAFE1Wh#4mhZs zGRt#5Aa*i_pMh`gKWzFDA7(U1BL)<*5ngsUq1l;Xa6Oeg5);G+@c#)5*!9D`6Wr|H z_7kpx2qaqm;DumP%j{`_7j*xob@%9N#GG+R@;;Kpi%06f8|Q@Kr=A(%WdTB+1`Fyv zkwWiGgzdznAFqJ}fdtOKx%;z^$c~PeKFmvylXe{ejq#{L9Qv5kojJ2MI5{dSr)DZs zRbtwzi63*}{e5jv3(#{UQ980vWn#Pff#@aHLEsjT=E|ya{DP9(ei8mjzjauZ|jb%2#4uf0~`t3W3H#bL%oG9;knVoNcSVZsp8T=eGnU8=U)wF+#F) zUy-)#BzWdfq(ykohVD%DQjM)T{hrs8-llA}Gm0Vd3rCDiEn+A8{NHG>dQNi9LURPp zCTY?F%z_1p3NWQ+WwDZnMp5Z^%Q_YhUVSrrsnZay1IfzMZfR>sC5M3Lo{9GYl#+3( zPxu1mmv6I*&|*!KD<{;X;vG|oX1qXBBtn2588hyUy6i~Wj=-O92hv4(Snob92OMWj zE^k;Ulj!crV^^;q7=3?`nUslQE{gaB5bl5+73!6?uI1!z?5{cqHZVQ5@>Vo5O-p+L zre~vvMegrmm!%Ztq+_pCVW=iH-JQS=_XKniVRKP;kxU**s0LtNt1!jHxI1Kr{cB;L z9{f2?Sbwou2#ot#+Pgm?WD*&OIfOjBYv$Y#y43p?PzdppTEYu~U;pIP9I`6t19wii zDZ;qIY^atG+5VJ<0E-!@LbDVAPoPO^*S1X$Zfb($iUvVp9SFJf$5A|iDne$BYcxNL zBa`s--6P4&IWE3(WJ>Z|S`G0-*Rgd`)&j-k1u31c=>k$1;cT2_Aoe!Lk|^oMs7xr| zGy%?P3k7-kLGWmtEFmwzbWd7t*!E~)#%yx`#tej9I13F;N>>zSU z8l$JMoUnqYg>V^A{>tI)*ot+=t2h zFv-hfPe67zZp;?EJpxO@afr-T>isHWj`w4Rqne|Qqgvbk-$}ubj&A&4sKC-@o#z7U zHN?^Sq=Y$5t`&$Emx5}OC?bt>;7Z`bq2Hb@RwS+}b#9UPyC)qJJ7_uuer$apm|pz2 z!*&npp%;~ft-{Gt@On@i2KLddW^2`QVaXs{jCZ?cEiYDDTO+e4rb5C--gStLX@OVV^nZv#; zX-wfh2b_jcj|X-G{O>N-HkzfGwg#S3N=5)wE{m10? zm+c)Rx0D>1TFT}tDn}M|DFo-dN?*9LpP|$^tX$zs(+PH9h%;x;e|k~pwKBB9twYOx zwLnnTcL<72KMI|x!YH!WKCzGf3Wu%yYo*5@rs7=tP*_hCVZYcZwR59m*SyaULogPD zcx00Xl7LQfv{g2c4=S2$ff7i~v)L*}l_H2y$TqA-&tpUyx{<2h1uDO|8H@@N^+m!J z)WW`g5s2zd79KH4W;s|!_I%pqtfoibAh!}hgCbBV00OVY@4%9#8#G)XO#MrCf#G0z zt7W;AjJI4*vIgJmlrvK2zZkp69!e#MiVx$fgZA1R~^c*?W$uu3Ak1|Pz70SMp*-@TlG6n_&dna!Hx90fje<%9eN zV*}J{l=5w8O}>0a0W5juFNhM{^mH zkU(5!&|{A;L}-vx@V7|pi$MIWbs#u?1a9>7yp6lCVuW=QxMdB%V9$uS+x85B=asz z`yq!?{E9#r56I5-VQ}yRKn<7Njr23Qq?KL1|zi^zJ;o6^Wqrn*!~g z;&zDzeylK*gtl0^tzZCQ!d2f$D#-l^`2#{#)4aq>fTR)ZgRhcbl8UaFfZ}<~RfwBh zx{eH(j$h11enDv3B=1E$p}?!MlG`Nn<~lv7!cfn z5K2|Xn0|)0Pyta8p8E;`CPHuL~*;Qc+o;Cf`qGmzoM+0ihaxW2}dPF8h1-B zf!vzLyVN8mLa3COl?-?Nvt^@#7*`+ig`?z9;k)SpU*u?H)3E!8j!`PYl+|)6kYe1j ze#+6FmH-(mZcpfM``F3M1eZTHt&hCRY6cR+im<3`+6ALDLm(OO@U1MLH6@dMt^Tf5WB78x+gkT= zOTY>rT8}L4?vUOxA59{R;wR!}giIHaPd|99dG1JX3ea5~COB|K{Xx6CwAcXAZ|gju zv519Td<3nGydl*A|FJ43Z*EP2H~^{%0CA}ht`|y2PdF&>kx(SjZq?q8<+2U17TpW|o@)1{*=N0CrqDAd?#uP{H4vqoZvHwDPH(?2WwQwU22 zc#T_KmBV-oMi;DEk^oAeCAXJbc1f+uXiYbz`iqMQ$GpWEwy3W-tFoKI1mt&pf#*9) zu0>hgTZY^`7Ee%qaQRZjTBe6IrLmw0Ro4^bZ`)dKOjCKJw;aw$l5mZet0;J~ z$!ayKr3js*<)$pdHw3PHzdG`MtoNbWNt1YmBGIB!;}};49!B4#I~v)WiCeUiu`fU3 zi(Z<0234b51l#Egavv_wV9J+G{(}1&b+E_^6m z7!w#+P!bG+YxZ2d73`61cn7DG23R(f*_X>Z!@WJDeS|k);D|h+=r^c+}A71k;~s0Wg4{7H-x+PqbG`hw5G-Y=6pi< zqw2tKo~|)nwK99�A^xSD}^amo-B1F^KQ62$-Dox>rwj&DvU3SpIfzcAEiH|><^QJC7){SukdD1ZJeX1p!E^^xxhocQ3+CZ z>sNOd-Mzr@0t@H=E$DFl&!EG`$o&5y!Tl%b{NK-NH*wuIs(0)H%P0Wb;^`!cCHN!` ztIrcLaCJBahCx1<+jMYWwl5S`m(K7X#wPjxTR(a6M>gwsbKIzKuZ=BMP z_L4uty@_w~9$uc05AlCnNAG5he7sz|o$gmM4vDD(phj13Vvk_FvoOM`qPS)RITZNncjgm~-a3Svq+0wn${04>(S`uLgj@?({0|#*VCLJhf;`1 zL=rZ)Gyk5RH+rnhQKMsR{pQ0< zqttsVre?OU+@7XzBLF=Nd1XXJkZ91O)ADr19PU|irN>u7iF_j1I#`N{XbWw1*LZB8 z5c$a;q*Yxx&nc@IWxIS4OS0gOO*onN=cB%c|A~Zl_aO=qhi~dm?J$<%K)6y_QFcPQ zTT1|=#_n}s9^;@Ov=gf#d1l`YVWPyHulRi?3W9B zmH_{ix-kY!l)kx$Xm_6{MDfG!>ot~=8T5ukC3)fXY84&ET;~J!W@@E{B>KBDhb?=S zYNzl=SLR!1N4dB~s(zkeHUqh9qDecpyAX(_*+Qa>bWMpfRYoI~3j|5nd4m=>G2MFJ zU9w`W`d9F@wHysrpMIH&D4m0{jVN{uPsw$=vq6Y^ia|BZI=3Qc?TY=_Qe`WNh6cQ` z>V*Y|(fY*2Qu%mQ=C82z^|<@ze!Iy{3mrQ4M|j|(^p1?BWdq_}?obi@tvp(LKu#H@ z%NcG>-m;iHkCofH)d?C~+%KaEI|@;f=7A#8-rzgS!h2#~|6H;=I8dMZEV_5uL;gL> zCx&g!!WzA>hR8#iEdPFBO*3Hth9S&)qiNCe3_)w|sM!^QUZ@LrUnW6l0J=FCPs!vv z)!#)%Ka@4)^M)n;lR6^5GmG_^RVv=j(AH7n!QVp-<|#o-2&T$i1@-gp2L8~7pLA0A zd`W$@KMIU2{#yJve(B@m`Y&~`oQU?crO|7nOepr3(L-st=&e?g-(l6rH+RZFc9j_= z9o82jc&0$G<*5dHY31m$Y!)gbT=V`%Yb}G{z;VRv1(L{~27C-a#H7Fj85?oh_JQ^m z8g~%qPerIg{sKEeo_(Sy`bK|l?(=5OTehqwQi7T_E|tAR8~d_J7kb@bE1D-0tLCy( zyY$_JFa0f39xOm|2{dzA_Hq?8E$ZgX+czK?3*t*upA6<~5e%zM!!`%iz9Ahyu}84m zG-|wIzYSE`#jKdWj;f;wd<$~+8(@)#f7hu?eZB`>B@>Hog#MIg%V--v1)oijAZ z+x8SsMOugkVk!c9Z7bddJy>hoQzIur|J1>PQ!2>?6rtj1f2w{%+|Ct~&y(X!tgO?q z#%U7V)z$Igql55z0hN+CjVZbmS4IO+AH7!};pTCrQyzM}8m(-`<)uaY3VS3BP+Imf zivfnBQ|8{7Kmnn8>ZZK4o94?ZL1dzh^{vOCA#=ZKH;zN<$U0t)*5%SzL(&I7jByC| z!pWjM;7@Xrq#Y?h5ILn&B6}hG^%mQ<6Y_@%!}nldH(Q4q+QEFRO&q5Z-=dy|9*lH| z@32wo!nU9bO98myMv*yni{ttcapp1de36SSby4L0p$q5hRn@DR9hW-)p5Zuf59qnN zXw>v!;17$BxVwjzh;9J`nas=tw`S)lbo{_Ddy3^8}er8$`mPr04#+11l?yhZVp5M<%E>rcFIeI$2pft&t>?=Ky2G*c z7o_ZoJ~3zK2j`c{vUgCvto?0Z7ceeaKjw2Mh`WxoOC7tiJc$Pa zqbV5Eyzj$nX8!VheO9ZLB9$^Tq9G)Jk-^o$v+jT$)grX%rtlAO><2g2gm#~}k2u6wzQ{xz|-Zvd^owk;6KJ|*Wy&HIC5Rm+#v0qoxw;|kK=D@v0e&93L%Zn9Klv`f#WH$a_{H}yjQ@AwWm z?wGmCBtL{h0xJ~6c}`)(E084l2<0xyxwf<$OPph6x>*`;kw%gxIVe5kGv9=h05xJO zHx~m~CTOvNs74FSeBP|9C%F$?n$yV6i5t$rj`qyzQfmS-xBi+C{W$KBBQ+K~%k^Mx zD~t_2&Nx_plaL!)aitsTD5tyo-MexYbOa1UJ9RyBcdAW8yC>kjk3M*>zl=Rc*yA}qEM{SEWqR1^e5o_z!+DkmvnH3 zF=16`!+ia_%&hN@&n-4?p;2JGEFBN{Hy@>0{20te1w<4|jn497MSD29*%*gS`+7sV z3vSXI#XR3dh(a3P#gKvin)-8`x3Pzx_z&0NZ8R&Ey|x9m)J*6k5-w9!p)~}jUHi%A z@rN}86TDDzT+sZkbTGg$XOr>W1ZUowl{fZnAi?ca|s!5a!^`E4gP zw^dIibQfu`AyEmxiGV;wBF%$(2AG9m7!@SgN)4&ovS+UYy^l)w@`k+ z9k=+44GIV%VtjbAtuHBvFCEQkujbW=DHYBUexpHZb8rk_v@UzJTa#?LShn$Yd~@vi zE;?ypD7`d>Q;$gr6!1tX-Fz@D@GD-bJrBRVxt~~K7Z|r21V97`Z8pXd8Adn5hQ=vKQ3*;3QM|vhhSP0T8&aVBpQ9o zg>is{a%&^ngA+vO2TQtcW}^F>^(Sq#ClN7+5Ysg$G*nO4(5aegotsS#?sHGa**?8J z1{79y0Fsh%*prkfZ;u;fp~+DJUf4y#s#lae!F@Hew9kGLM0GHUsnD=Y zmOwb9x-a5XQL&SR@MlY5)DWP`MkPva5x?v7@N|>)emSbZ&;6C|>@u%#w_4HguAs`3 zFstwOE)A3%bSz$_XF%Rc8^OIzG3~z}6efNO`$;+h2<2D<1FUUR<*vs|+YTY+G zXBAsC&eiI@(eW7PeD$a1S6HlDugQWvxh`;Ms!Bpge|~%kK0fRd`K~1EQ~)m(L*3lM z)1c{i*NXs&LDY&3I)D^E&uweM93spAp{2*eI11+9gM7{*VYHG!#N_N(O(LCO7a4qU z;s$9ugYCTlir7Q|PW-#X>M|PBD|y7BE}?s)Wy5|kn>Bx01Oe=j)rVwp%1e}b@CP?;kY+?^W26}C~$FZD(at4)Ojyh^s-N(YWJ|xIknXMLO?+^!eV)n{)JE4;j z_=X5$05cTxdR=Ek?lw|+^QtVF)zOoEg{wz$9NkBc%E!`|7RN?HcA3Fjucy=eWT7J6 zD&38lLRSzaH#;NIiLfpqfq^PmQ<2-6r?BiMDHeBGjSlr;pkZ$o z9zXjOxA{w_erfD)_0!;I4)x_NoUM-%SXa%TyZIX(He60+UZ!*wyw$+QK9X&ER-&7O zhw2ZlSvX#YJ>tk*0naZg{=Tfxq(@`;T^JGeT+>_gY5Vv+n@q2il{5)uiq1J zglCf_M*%fREf6b>HSaxlZgO&_uOAKQ*QZ&>&EA^CFGf&!+0oOfcZ>%z!pK3rXCq*M zZ$`vtI274`Mms|`>Y-^jt`~Jj+4~@bZ!NM{2b#jQR(}k6{;)W-+i)SYYj@4B=nh3p zf%V^OKrZXVmv*hOwqRF<87jW&O9r~yP=T@z#T}F^iy*kI@0~}W>?`?g^JP#6 zPGqzJRK|#P{X$Wq%$to5rrUd1w3JUs3x!dh8fs!Ebd)gZOSn(s5QO=bl9rH|WzhBL zTv8_o7Zh%fC_TU3-#@5`>9*L}&6YqP zRoW4E29=Zu3cmN%-1ZKrzjX(6nRNJI|6=yIjW66sQR9b4f7a2g9&h%m1QuFAM$yVfm`M6A5OUiL+KNi^C4J~<@h`{&wDUU*k8y8QmcIX(=a3@(%c!+%c20MAI3Kqu0}!?DvXBtro!0FIXc% zSJ3>hfF*2+!ZIg|08_CIG>Z)frt&xRfV!?m1$v;OF>RB2UscWcP%UrmYlELrj}2FD zOX?oAxcYfM4UT1hNw6>fPY`%fvS0j4XEe!=lw$k?=>Qoe%QZzh6rP^+xKWlo!n1U? zu8+Hg0q-14X!e}htP7_?HFy{PoeF>$hQ&-^LhiLB&KFWLsxMXRJlLfD-Q}?{!{}ZX z`Z=$gpfGwe+#jspe}Hf8#ZnCVmm#Ql)}uU)0}JomNF+lx6fuvy3%E>%0*Du`Qd4bR z{YOzwidL?xy0L3pIAnjC6m7&_f~V@CuvKtwfM?hSP&~Drsc+S%~tp716 z)rqm}JnFWVkQ|#+nZFz|;6_r>hUG#A>&GXI!@R-1LjU@obP1%13amFDAiRo>`8xJDaf@UdBQ{ED=DYT1+A>y7P6Y1XNZ`eajYv$kN z2+s#h8#$=dslW4#m|z?^NQF+- zclyu!R;((M#wO`C4s<6v&a#xt9AhOfm2(e^<1lDxY?O}zHZ`GsfY){;^$Q@Z%vW_` z*SNA>&cz(G%)p?!-QBWFz98Jl{(a0VSM4%xCSnTPWzMCkus8?h5Xk69hK~T+ z_(iy*`}86T1n@mrw@D#OhZ%LUOR?mVA@dF0w2Q}k%wz)JNuFl@iB^t=h7Ced!{eNrKf>ocUOhm+9ZT)~+m|M)V!O1V*O>qhd8z@YPn zCpLzUo4fp(zmgT<=qbhTK=vYnTqw+>3DWjv>X;|{+2`tsFNAKNI#ugjHjwWVMEqi3*=Pe2EpV`vN)Gb^aME4#as4664H zsP(Cd{*0Yu*9+QWb1SO}ZAlF6^eHts;L;D)=@5j}Q5U>}uX@j6`&UZ6Q+h-0p{>{t zgpI2C@PA2t?2P|2^|5lYasKazmH#jGv3kyp^FLZwa4Zp?YZv24&$E&(EQH~@n^uKG z{fm8r=ip|n8V#R9C;Y$kYCW^Zr>J>5pcE|sq(-mV1qFkjrzgK1r=}0*QkV}=P<>1h`Iw(NRf-7*YS6O(hUk37h4sG z)R&3$KqdV|QMlu&c4=H!6q>~$cUgy;9dINEo=bq>--uRXFLl{wKWUu)it>>wnwh_f z(rrg2wTKk}nm)!;!>t5Jfr^It%Kj{+5?bl;JSWh zXE0T%$CTG#X$_00A+o`U zQ|)3C9*IM>bFat4ZQf{~&N| z26WP@vtKOhl_OxDmEDVT!kd$)1m00+kQPXObCm_eEXl z=WzGfv~*G$~H0TR*C!0Ca=d!bGf_`zz~7} z>D8+e*%WT*rDs{v(MO}E{`eE$bWRnj0Cus`)mYcKSPFpBnujz!E=V&45H()FqN2?z z_60?@1e;Ral!6j4x;AgHizqKdkge9?4#uc(C~LD$I=dx1%eb z!ytm^t&OTX^JQa!(6nZKCCz3M&ydbfnJVlVum&1Ztqj*<%|to6Aumm0dloW5A%u2K ztY*qjg0jhW8p??7iBYfzCpR~6adG)e{KNZH6)zBd!76z!$0nrNN1LtQU7G5a3VZz9 zT#AC|vh+un-(5C?_@d|*LDX6>Azy#9p8}K=8Z83_d-M;cAE8?BkzCR1wTU!5gkf^M zsGJbXB}sEJ50S%1TSCe)l>v#H z+A1-d@eScF9m~m)R~1wyj&k2!^vr5u!mIr`e;*8lIx5K88RTAF`^SU8ca3nG80w+PQeUnF-1~A^=L)z7?)%gGf$T14M|P>J zh=Z>gNI?`!BwfM+*@czfLa%tPlwZ!-1c58x=AN}ApKZ4Yx;0j*7fh-E=;z>BX`-j~ zCQZr2J4i_;R9-g{0I`+kp4Vispoz;(!2fmw4Ek^Za3Q4jl`F*hc&F8T6pXwz6Xa?g zEOiyin>3U#66i@{mh>)UFoYf3ta+f~T~S1j{8YAa-ew9^V3vYtm`FUg^1Fw!tY)}` z3J~Su=BOO8sIJxhAgP?Sr!+{y0{d2Q!gddwE&w@h{&R2=F7hk21`#eaFPS({NILjw z`rm&*DdY4Wh;2(=(+2zqHtBI?a6Y{k&NKcvonK>?{(%^A4`Llv(|A38eok@P0B#3| zX1`kE>3SC6;-kS9m}8sI2Q_fx z$+QaS$s6V+3uQM2IBWBHuE3qG-$8e(9zqiRo|1Hdv=2&&-lp9|y`aM4iYvaNwgd*w zn*iyio#fSQeg}%d-zfGR$%NaY1%P#^xy?2GWeWfwfB;73QjRuSzT#B=NAy+bEI3r; zE8;ne6A`xU+GY$k0jTe=QM9n3e4A*WOJX`wc8`0Az{X70zud;feOMJ>) zJ$j4I^1l`s>#Po|7Fv)BXJFkjfVW=M$zWg{Aw+)H(>`7vpwPueh59wY?vp zo1@YS9fk{ogH4awAUX*N5hl+k?Jp4tZZCJGZa!IaO6iaF(PySaLP^{ISs98&EH=8b z9P~-^SfdUtrOo?eb#Gm7alVgOw~$Ih8XP7g$bv0{#BR-TwRFkK@KBQ0>TNa4E0#_$ zGt!Oo%2JM(ed(&KzKlb<6o5$^=AZ~51SN~ub`id1?VUb1$xaOl)`Hd`7PL(=1jvlW zM82SU(K!5)MglZaFdNB01I z9q!xF=?^Em)_^2PQIEclo6igTDKm3_%qa*ZWG(m|fR`M}{WJHya+!21z-_L^5dt*SFk%IQ;C+KH3yAD!sXdLA@5;&FTQ);~=G*%lf=1aYKs3?RT zI-D4p*;>%&(Ea1`_ViZ|Ngop6o@-=k$zO~DtAi7hmXx(S$jxlA*LS;EAdkxo3WySt zUbFh%lebi+=#<(1#J$D8k$nOsPy4%T+JbahG#@uU+@K4wi`0yIh~C*MhS;6&X$=+^ z^OllSixpnrTZBEg#V6SzYTh$rsMz73)q&2Dd0;8hwt=D~W&4Dvv* zZvQk-7u;G3Peiv31Mw)!pk}(n(13E(zlQ^k-ZCjVTo|%PXq2-a{s-J_?J@lJK3-~Z z)cX_2W`EL{pvd0`@tS5Wd*6oRD-PP-4OF(vS!M>(RKYg+HEJsmxFw_(Trs8;xA-nx z(NjHrH{PDNH%q&x2Xm6+9@M2V)dn^Ej$}v$;R@-Lw8JY>ypKMxnsW)I!2t*L7<8)4 zMic8fR`=wtk<0-xZr$O)ywtg6igAarA2;XZ7#P-e>soKUV%E?1g^5`f;s5}wm@LaU zAppB8VDX?q~<=X*iXV_c;uXr3avzI(KaMMBGgn_Lz&{)&$YX zQ-@h|QSQZ2bOLx8P65*JL=SZgCk#WVf0uVw=@5}qodgU@ZkPs_LqsHTZ7fY#RF2{;xW100_N?f?xNf0>`LkvR;en93^mENK!lBCszjYlfm3)#eM4A6%l zVb+LHa~p2c7IJchO_Y6gjmQj$U`f-zw^8r@%_lv$onjnCrXhPf>{+w-c|P#GL{00$ z7r_Sma26_(S5bU;@h=IGQK_^JGv{pwUF_sOB<)(I;?Aq^{c@3NP$8Eb%Qw1d`*Lr* zdK1KNiJZ9gUY3I!jdO2|qp~8KFOk`U>&(I+g-L=L={w|E32$&^$8F^6f3wVmtaObX zuzMIAiyMM^%bnDOg}v{1*wMi`*@qd&$SHDnQ(1K7A73D}Zor7<25^w@D!Scl z6cJ`wLqe3oR%o_Ew?Iz|88PfbnrGD18w@#p)}V!9c}JhzSbj?bE11o1M@FeYh{Hx8w3K}LkKO~A-_t`>8%@-k@rX4kdbS$` zqqPr-h=GC|)%&@_SnCN*4${^0$Y`I{BN!-#noIM+t3G_;op+tYpNYZDg@yOoy|bLM zW^?yjR3g?eJd8u;gO-G^F_y=`ntR}Nns<_8QUqYO6E5b<5HYRr3#mQlvm|9*N*I1C ziPzP=i0~*!uq8irP-9pJvj15r5z95v);RE<@5TTkDe?VpnIG4C52?|?-| z?9&+QfJMVb{Scapk#7boM$UI|fxX6GY|KvuIA|w59JnO^j5?V>$%0x|yy6wNSywoZ zonq5s^sNOJl=Y)#=4}kDa%{fqc$!aozu=s?iP+WdU?v{qR&bPrK%cl)!W-BL5s_+@ zFP$Fu`FhPO6dqxo2V5JD*w!wX4e5XT=zQ^A+#TH55li~QFGBDsCcHd|Xbzr}q<+`b zw<9IsAz!-kYU|BVW8)6L_U{#iQ%%YTMrM7PCI1MD`a|Tw?MdJC9A@byQ%ICC<-cV& z&+f_gmRHs&JmhN~CqdK^{lTXRju%duzM)e3e=BxCWR)bdv zJqE`j!yhZd!lzqx%IpJufA&hpaQ@ALO+2xy&=01PY)zzOn+PD7q*_Zs(3ZUI`LE1J#vE8ETB{T`n5`5V+(jChJnBob0XtKf= zP88_n-KI1A0r>Ixh^F&Qwhi|W{OYza(<`K*TfogJ7-Y|&XYg(Gf<^vEtn0ZkGtl{p z@eTbw`GxpQNp1U={{hpdK1o<8!zFK}Eg?dk zz=@wGk)&S;raW}frB!^dE~(4Ya`Fu+4B>7ce&|J1+FVbf72%$oT@(*XHpZ??YwRlT z-?19r13@M~rXSE;%i%BNe2#v7jfkQ*0@lcFV? znR071`WMVc$8vZP9NYb_^O9GOm{do>418iTikJL)>D=KvDsPvcH=t<_$r#cxF1Ns< z)tAD;LY8;53ZBsXy~2U!I!OBn^hOKg6w*!Ob z^JGD{{7TZp&SIA*_lM*uN_Cuy3=N)hf5;ko7w=CR0F)XKDnaJ^haDSM)DwFp;3%$i zX~%&d>CT%uir=)ibWCMuqp=99kL{{^cvCWD(Pb?8DtRmW3Biy+*Zd4bb%II+DYU)N zT&!61`o943EdJxZ-M+a`FZyWuR({kDZ1ea5H2u`R^QLMT-I42~IdaeLjAS?(^8_%k zpDGL3xPMHzBaHLj2QIaCh~y`(!Exd}^e|$*BF*Y2ub+l0 zdVVJ?Do;zW0#Y0$t^4@7f(% z&b#^YMIH3_$EpTa9sC8c#>>Uq1?0V^L-AnY%?b8U9Vrs8e-QaC*CBnZ<~z}Qcx?Q)pwBiUx7INk%jmH-|uDE|8EhE>AyrYHb$2JJ)$*h zNIPzECUjq1FUnY*wv<_;iFkqZj5k}^});Y~Ey zHQj-1^tJEC$l3dtF&%)Zw~d}HTKT|;Yd#m|Z>_ewBo95u9R$u0yXPW}98XEQX^(4A&e}X;he-lBF zI*%*ijBCQ|5Phw%St*|!-jfafOgnB-n<1cSUBgb6c#59qV43cy~P?+ zmwS_}+q3+BtSbuOM``+w_xM6bKf<%Uvg3OD3#7ZZm8Kn-!A3**ET0K1K~c4+1MbIL z)w{J-_krBrn70IVy&eA{H{fBZSF3)CoFHk5@S|L3T$*{~?IEsoht$B7dSDbnijvbk z%8>?bP&hL2^nLUM<_5QNDorBVsm@Q{rktj$(tm9K+?Z{D4|jG)`u+^<*j`ApWwX)I z(&80lcw5cK-EEs5du#*hM!P6Z()YI0P3}z-w ze%54IYrsvnq*#%X4Ow2{VtBPv+}y!hM8sA?#){iVdg{&nz=urul3B;u0o3>oGsaLgR5_@>oq4NPzIBe0@?qkMj95^=yn-IofO+APazIaB;!e~a(?zNDH~D6AK4R+Sm3#$k%8mUKbbIlFnA+fbgwZLs zh}aoC4HQ8PE-2_y_5$mo_kuyFD{at+k=IuYUJ)<(+G`n43-KxX7@dVcc;8k&3bMa8 z=UeEY+Br$EFJb2(iWa0%4KtyHpKAog7tifne$Z4Px;UPmIF0?6t1fQe8faBlhpW8c zSDdqUF57H?=f|?T;%^}^BE(Ni)QTMuu=5FhNIzmhXX6P%ctm8-YsZY1_iWBAT zkjM;V!Reu~L7pY$cW$i0X90qsDvQQ$0JLY3p!*H4oMDVu&*b+1Xq^-urBc6>h3aK~ zAa&85J6ZYI}RInCVrIyRBH3)8MXj3we-APGM?e8}mp$fR%E8M(Xdd+Snn zhRIF!wE|+4Fp;4{Y}kN8z*J_qa6n@FJIA;%2;=?}1kIK%Kj3=f)ey}H24MxslvMSY zM+wP_iOTt5*=U_Al1+%d4P^Vz7lm{?#5B+d-~n|K)xpJwia)0XCFb;V*cO{49-;`l zYkZ^Jle?zw0pXNkQeK~p3S47*PgjNtt5Ze9bwBVaZjhI@%;`&;60-5f6q*sf#|iNt zl?T&bZRBbe4hfZcP43=O1~ZSE8mQ(HAX^4OK@;{ zA8z_&7xD)oUTowF18RshP&MwRTHu6a)uq5LB)e=ZYn9g@uLIj{z99haZ{l z$j_!s5a$YT>Wp+Et!vXBYF+v50_cGg2`$~5c1#`5Vb#^M`-Ao1CEuj$2MHJ9ij+P# zCVx1f!Oc`=zyoPv%Un@N18AjRV61N+HLU1Z`&WRP+EhBma z0}!AdvFtz6zm22nwNH_*k_@c!4S^o$f{KU5#r$9%-))&_IeP%SJN(4M!`PXmFeh%- zsa~P8=2z)qQ&>S$`4A*%QS=u)?~f|f3E*+yuEwF}j_G*12B@KQ;{iZBMzO7hqMyuv|%l^W$*n#4Zo9DmUZ>a!osR6o*{I~ke{qJb% z{E>UX4{T-QCwytLsg!d(?~H1@?~7a8X^iXR9|-mo0SP)C#pS1W`rRWxJ6t8DHMx>U+V&Cbm_-D>2?^06!ZdXQypO0jMJuZS(LGZ}ya+FIu5Cs?%&rNEP_dp~2;jDIquLX2 zXN|2@tH!HY2-9c(W*=*ICHVe7eiUF%vTxQ&yhT_C+XVYS@cdU>G4z>|gr)$wAqtfP z%2*&6zjL!rKSDyuRzy97-6XWKxKmSn>75{9E`q=JXXKi64|ij%-u{j^@A0DE^gH>*r|b=F znTAlaQ^O0bPt)I-4d?z|@s5ko1BV5lFq+ci8|=*biRy=?Q09vI&|^A6=@MNQcH){y z!Buw7#UK^Jl!;CuU<`_-yhaFBK|AeL;yV9az0IRaR@w(n|3rFOP9JkFEt=ujtc*Yf zLnsrRWhFjRK$tAk*53X7sq{ih=)gW9(v+-FRbvx?5zA8ttgGxG*V1rwQ}CeLhsaf+ z_Y-a`Y_|^Ft*Ac^?&;iJ?PuE9V-4{ukaL*4ECjD5w)NL! zj$JF2Sp)=ITiZtGwe6ydo~HFGb=;K6ObhiCw;Mb37`m)g26?-_NiOf3K5E`NJE>y^ z#CqR!X>Q5^T<5pyjV(MK_UTk1$Zf?Tx)Aa>?pKZxZgE@z>tPYo*0|e414oVEd`X|Fq%0rmt;|NL z64sg_`AEb9Qa`aAC8(E9A^s|4b>uE-7wD*7rgv2m{-Wk!4N!ck#L){sh|TP{z8OTy zMXdk51h8Coon*zZ8|b_;S)~0aH$b2q!?L0p`i3tR=5A)KN2dbTOqw+KgG^XO=A(0h zFAZ95-oGLJl&Y>g6bo4vZvj`CtvWpKDT4evuz}+%Ki-lq>HK>!@=X)SJdaKKWk7 z`9!-xnNYd!-1r>kEPi~xo$7^tY09Oot1-ngh%wgeS4r1=HuhsC2myJuU4c^9en>Mqg*{! zCv9ufa%N;Zk6LNMG6VF9@mMFwtg$3H*;I-9HJZ5DtZk}9QZ#9TLrxyxD*a#QeCVWJ zZrMa36*rjt&SejZXw#uDzRO?H$|F$Tet>omO?Z9_s@3g5VigvJ$zOCQ^>WPK>P@?& zj+8)ZyT2>#^Vb$qbZM+|a4;+givU__U(P~LA#P9g);1dk^^nVIUbv7UE{1hy;2*W@ z7ntk6YdSr3kp}E@n+2br%z%50-CB`o-uY&%{ z1Vn=rM&2UVKLYKAfI4(k-nw8m`F35lWk*Ix2t~Xex8%_Xm_9}5^dFXUwUaoYW0y7! z)Bfxo=ry1QuWF`W0O|8U0aQGJ2SR5t#uVR56E{P) zS`&z`Q%waG{9svUmQguaNITeYkEY%+mP~&5%M2<)l430R_`#n2b?w(Q5rvDGDEONf zPquu*=1paFg>V!w0beq+-EI0zDCdBNLR&C`j^RvJ|hvH#Fb>PGS&>7QY3CQ%;>o><%D7{jJ z^tNgRSW6rvOyY5g8&|-3Xf&6Ifvn3?E1MQ9I=3NGqxZRGZ3IMTkKIR=ugM@MmtE~U zTdO>=UcFGA_a;sI2cM2R$#jm1nyY9lb-*O|cjzRc%E5{+7LGq^u}FEn(8Q8KCQQ~4 zR$>ICAu$FPpfoNn3E1HgWLID&V&)i5X>idUK1+oydv%)0@7Mk4zr?fz`Jac`TtDKh z7Ae}b$MwHHZ*Ree=ARQRVA0SFXx$t&wUviaBdyZVe>uc&9zN9ufd>K}mh`AK> ziNg#2osr>}xs3$1*qpcpOW7;YJ#6j%0n4#?CH-%OhxvaN9yTU6uK&I8RR6n({P$U8 z&%?}O2#H$;r4?j#jcf9=usp7MgYh`AsVVs-7i&NI9I{^sjvwhc!Hs*Z=bD!i8#bXh zPZv%M_-Xz@%}WN^#Nm4^e7Zny9A!Segr2eI=*AfF~0AqV_-jh+0ElWpE~UNCjlbMXmU zZe0G7x(Q`rF@5Sgw6*@-5Bh!Vyt)kP>Q@Oe1-Bq^t`+G!((^|;7pGnh+a-# z1T>f!dagDv#s8q^os?%fXoR9DPW6W9sgi#53hRRfrO$^yJL945 zq{Zh4wA6;otKQc75)4G!jxH7K34o|8Ok99FgTjG$me+&|%Jv6oSItgiXP)U+Eme@&L8)AYf}^B!30IDhO8A_ax;#dS=i7{!gagcJ>7dw#5oY} z^RE|#%ZCF#Dal(NWts7!>t*Lc&_{KO)Rq5@J9f9*g0nfcQc@3$c-L8S#hrx+;xkq_ z!NO5y(i-b&BF%~E>vU`~L?5d+Sgw)4D7x2bEW&GG7?6KlND{6SMyQF)_S1YgX*aGP zT6$!o1+<@6Ww&W7Iip38Dmx%w7geYBV2a+RKB1VZFZZgF73^CGq3itU7*MT6x&oHf z6&3E`VmZXzGzNpdu;8D%`gQ?}34EwrZ*yC%5KzpkpEZ|S%XZ`+j(iy2Ic8^=9{kbo zYp^wx0ZIjmX>LX&7%}Gw7x=11ev3!7;_z`QSpDIezx{f7vtLbBOGEdDCOtIlG9-U8 znqoRA9p3-(5bys`v54Fl4vTiG+prhvBEK!2RmR@AyaN{bUqZND<$0jZfVASWW{*rx zMn<-}b2RCk$7cJt(*uN9x7uua+I7L| zqH4e8Jgbn&euj%!#^(zdqEG-~WN6nPBW*m49h2)0u+${Iuf`zxGxUiJczhFt#JZ;F&^9F_NB#ft`VtBbIgfi3BTidrE9y~qEu%FL$cCLREh^b=fVELHzeYD;cZR%Hm>C{?5 zH-G#xf4eqzrlzl5gBE;c@B%Z)n~WH<78s^+K38gPEDc*k{9`lMokzHljuco4>q*}v z%+8N-(xj6RkXJf88ONEe{;uZ~8M!a6LU@U7A^t_}ZE*o8rO0~$ydLydiD@CfM?8rO zAG;Yg!;kiLGA0Vo<%Mv|k>e-zq(u%nP|A7Pk;@GSP68T`@{nsTper2wQh=Eq;PjHm zPqcb`p*SC=|6uPLR)~+yCpaNtKo@3jdPW7U&5hgWk8^!_RQWWq~7uMQi6gmX8=n@MQttmVU ze>Ba5(8t~q;PVGJJ56>O_D&bLP5O*jx{l4AgpH&ZfS4#aD1x4g00V-@ULyIm+iElW zqes7u;8m5+5@Oex%J(Lkwn#Qc5zsx+>g;Eke?P4!!|sufU){55r<4)Uo-TXj2gh2# z>4Fq5P|Mhc^fE9)BE}zk<9dIJV?88KKNF7fm*FQ@(Ie>q9m011V5+^lAlEaUEaFa& zSpwM7e|FBCLX~jXWHA}jQ8f4n!-7chx{WL4#xIpIJ@MHl#?p-nnLh%)05vWLm`Eww z3G$+m^*yZ^8gMixqapBszP_TMipF)q!s#(%{KQcR_V&=!Kl93ysGB~R_iH8R*c5%c z;$@|0H2J^43X+iV7CR@@aXz9O#S9YAhA}^?j5b*W-lwmERaQiIlnCkj;M36C>2Tcv z?RFKc@ZqsQ*{sn(Xz+DnbjVjh0~0JG|ITF3Zm`HNh)^J--RU61l{zc1SfF$glYAd{ ze!kmPvhq^lPsq<8(_{q#u$xWMDQ!eS^}84MMcK26UkU+VCK!oR8RK0u_*rKFRexy( z?<$S0N8^?{2ZHayNF^Xkh&(OtEG74m=1bipv;}r`TQ}*t^UT4&3=bNNY2oV=H)ml= z!-zz_5reN77zCXJ2mFlfJ^|zGH|KZWXoXVAZ_B{4$r??#Ps_vuisf!&uC=k&+M&g# zBt*59{D2jKrXa2*s1D^w;~uOfp=ir-NSz(m3%Y2HSzf*4l7J%z!sI<;Y;_myw%G1n zSRjM(URr50GF}pr;BH88%?2lXj@+6>-!cL|HKYxcKX$%)K zi)aR@+n`f>ML0Lm0?~tVgXw_HR5L92K@f~O=fPl*(z`=|+>w2Yxoj~WMF8n};4_|R z+i^j#hmqhb$?hfF*Tj~Ugr*6N<3W1jF3me)a9?EWD2m&}iQ<i2iJ*XDFvA4YbpJ1jGpkritpsDbX6b+IT8OnpI5ew=&cOw?KKEMPxvR#0H=Pv zcfn3R$1*Sd(xTgHS?*ZJSb+3m_ji=E@XtSYXoY z>aJTy8A(feT~`-x&s1f|uQJ6KfORmzBB69ZIfN`~=af4El;|D+2HxD(&MuTr8W2&q zkaF;{KD~TIZ+~1LjoCRIl>!^JrP3kQ#t*cm7 z82gLdCHX*~SQ-T~%<%*%2b2U-QJ(OvUtktV zOPN9(W}aSVo?n6<#WAb}_{e<)T$?FDP zGY@NqboO?ulgNur2P^x0hKAC<(=-L0BC#oVmJ%eqLMH(}DD-RM6^r=d8^`6;%_MTTbsezisCc~#LR8{* zDmjyDZQ2t6!B9qcA^iQbxjf3xlkq$i*Pn2v6UGoU-l+x$k|Lp)ATO`60YO2RCt+xM z4Z-49Np@U}#ANWRFSY#nbupcsfIhF6GX3Vh^Mq?93kWVs>0maiD~(tCa9E<(&tiPg z%K3+>KS6oz4^EqGv--3MvhAh#kspUMrkgc>>;C)$QX$9Xqg0;a)N&i@FbJWKCn(`A z45N%5mUu9NtrWT<=C^PX(LlDg*Y@JiUnuWk1(381lMwkE zk83CzVm|>m3^w4hw?1;crtpdp29QMwCDad6U=44LmVMf{%AqocrMb!sxk+8~0IAi} zWMQsMXPp)?txPjyhA*AC$NL0+zU8lxT8Bm~$aF?Tpo(EjIdu_-Qu7W=Y-uGBVf+dE zbY^%3`R=4r#6nBCFVI{>m+k4Etv`^e`AA+9M32@n4V-VYZ{^c%m)_h}DyyX5qjvz9 z9rrU(h}P9QU?|Uvu!0nTMbl1ZULh$+UhIAS3IyxMOCUN-3-U58?3?rM5rsEzBEJM| z#acivspQAmgv`mf^?&nd8N6#5wbvx%0xniWO4hmMxfOoJiDwId_1|YxMjCSVXKbgc z!U?$n=d^Ksj!6x7nW(uHdr5x)tzEgizAqwx&?ctXA&;uGi~yYF>P4K$8hAgMgGn-2 zCL|SDYmy5;hg&W;73t3sDxPs^=*EFkQxq_3Wq&l~m3A=6u!yS>J}rrfSMZb&gCzUT z9lQc0V1BrrbA^xldm>H{1*1fYumi}Tb%zMB)5?pbG|n40+cStzy7nH!AGsQGEDC_j z%QOWuFWC6O>fHr?l?3BCagWF2Bb~t>cc34V94Wh}#O!N$nk1-KI<2yBZfF-^Xzy*;qV*W_R{|Dekd&}t z^%;1#%;G}7X~YJSM|-cwyjYe_sDZ%eBH7#kH#l{bqYy{^-nk=L1BjYRs^rwL4h=ro(j&!<5@&}~5&H39$6TS1@BQz{zQqHz>WU@729lUhq ziP8bg%D}k?6E#|M%B04nEYfzpEq;$D1g*(Du%JP%mMgU+o8utTQ?qcD2zHy$yhXlo zhr%6dLO4U)4yWMmIU80dzQM3!%Ztgpcj4D-+ix4;d88RI!V7p@@^J>xpW<@zb{AiJ zHwCkG+tni~j&<4{$b{U2CjE8o|U_y}i(U=|TkDr83~nuf!almz9spi-_VIW_AQ1>B(`7Gn;N zM2h%0uIXThiyC3yRLbf@Tq7AwbN;_@T z87t>6o9h1C>Yo2>zN^zVytA(`XEkzU3U0wLUBvM*gETB^Pxl-?w-?Dzo)9ywg(Y(};%B@dUsHe> z#nrs3nySmD_VD^~?^3X&!C~93a%S;(kI5Kb$|x6(N&bV5`s~+hERM8GXoC2Cto4|& ztJ4e#)6#=hqe*&3r}}RL6gm*YX;4KDw?Y)~K#x9ZzBXZP{R9U_p3J61gSjeX6r;DC zgj=7Z4@L+J7BAqVCrUXEYb+auR#mFn*cK|o%7UH3`Q;^tChs4+3suyC83-DiQz zi&!&i{_H!bA##%K!S^1$M3_In4Qd1x0GwTetCpTJh*QjXOtHbjrlTmE0kQJYlRLe3Kc3;mieCX;UnZcg>(TTr zlN!$gf>bvUtZ{YAmT*nv+j+@2%|}w5sz_eH{x{v&#`~AD(mr}$993NN(~zU(t(7`C zkhdlep*=rUS%f?%&>41#5@V9dfCn&Ts&=*g_ z$K&j`aSeyBmye@k>4E`vK|(ZiwH3T3lt*SkpjZlbL;=g(3ci*d{NU>0YIyzr>ch-A z;-B#nCV?bk{?@~tAgA*`#!C_ZW`{DJ$ku|iat?_^v_TZSR#CB-LRu z;>Oh(E|>%sQtn<Z4NU?9J`74MWQGoV#N-2Q;ijqnGt6Buon=cvBVuY z!|EDk$>ydhuKKJpTc*&KYagy_f(p^zJ)M~vq6`tDF}lVG1eSlzt*;WmLE1#&w6yd~ z^L;Fch7Gszf1Ty2hBB7x3?F0)*P=}x3ezO>eX@9!>b(jS$LY_94BM)c;N+a%C4?E zUY=}GsI2EY5qkkce1fsd2_izvJbP$64S%+$t?@LtCPluo4&8RS|KO6L-H@FhQ}pxbTKvqwM13Dt{< z;ra-JrstY>3o@lLp~&(%%dojDjmy2yd4hMvLl4Kfh(_ciIj0yIS=^4k?Qe#0$-BZ+ z#~z%6PF6e+&-b^rIY!2v?2)B{+@2`Z8exQBcR5on$D619JHPD;w^yEA$F-k+wgp-s zf0s2hql24smztEm>wz@-Ig)e4`74G6OGekVQrW!pMSk`rUz&fv0k|Fud~9A`?ih?%CJ*7Kz;ed zcJk;jB$p20o{4P^Cl^TJ95HNGIfG~3U`t%x0y1#Kh(Q&iE<98}AqGJi8Nx@$`PvnD z22MBx?2nlRynX=YI7A`Qea)5}gA|moX35r}N(;ct5Er#M2vZA^+NP}P35`fy5}?9O zvs*l8-H2uv{m1A+*e&h`qCi7UknD?))H*}h;{4eep5)7As(bR^;Q?L(bj&-ESZ#?x zOjcI3t;*15yJ$ao38$(ek3!kH^T3cqNXnE%bs>ErF!qA+pDv8t+&1-1R+n>pBwrNj zIodefV7i|I6)hTX>ju2Z)Mnv9zK-Yw0Dw0nJd-?6GtbNXWcJ6y^ahbbO?|GA(j~_J zMaL|)gCaM1qD-Hct`iu*i^%C3+?fikQXV9V@~YrW+O#+wf^`oSBt+?UzeuNT?3h=u z9g05&9q@OSSni~vy6;10fR9pLGZlT!;P(4~1~^gefFF(m@W57A29NJJ(i^-Rdc>TN z^SeOVBNCg<9I(^zu1KKM`=`lrR1B^2*p zdw?a*L4T{)uGB=&-^fV!o^a>J-nr!s$drIzyzNMP#w(U-oej5X$c>*cYggBvhR$tJ z?j>s{K0=rQkfiJ0pxQ3w3wDllgaQ4r1F`HftrO)RB2IL6738+@5vBP8yj9e~B*y3@ zelS}$8`h3gCW$)E)SXG=O5v9ToZc)Rj$Q9H;{D5zgXihZv2fwJjv^9@gPnLC3%Uu3 zV2Nm-`M{;b!J+=G#6t+fooE-yr7Gv+gRqcC5d16HAqmh&JONDc0HZ`Gr`hx4Hh4)t zxrb$j@`oyIHczrj=IrWL9Oiy;!w)kvt88A01QQ&?#3-z}&#>S=GX{G=L8vC6A95~T zirij4TI`6?SSO8Ts5xa)-4{PX@&kc<5Jb%Z9p<2KGZ`4Et5ar z!nZI^@53Z6Op{eaJ1uhK@y$o`&t_vvhl7QY?vh#E#TWlrfPiH5;yY;Ba=I4^LB1*( z9xC&PWi*Cvn88yzFOaVXVk3o2crX>!NIbrnxigN((eEvR0Uu}e>}q5zO|{2_^$@h~ zXDQr5%4hR{c#Lto(vYjrcU&WL>@V_dm7?5CnA%5J{b)Hbo=@Ms##B}2Ccrru zTzvLaCLM(Ka0-25*Fclb?KK;L$~f^qEzYFtj5@a-8qXxUC|tvz0pObFiO<+FVa2mD z;JVZNM(eo98Hr}F(fk}y0YO~dS)DSC>E4dkyv})x9&aWp-@8zVR>tBHT5WC_tKb94 z@~hx;PZ~C#PzqS-{H`g62cHiA6RO+v+|7?bBWd<=j8v?ad|z76m3(^5{L|$LLMZ6V zXp^2;+vnr%qk~ImhQ6R{xB1Tt_|xImt8!fAF-ib0^nT5fE@6Q3qcVb=9Ogz0HviUZ z#W}VVAz)tWXbG+CoahoMwh_)8P>5%NW?O0a#T8p))wh%1`=@8Up;jc373DVFkIhlS z7!h_BY!NkN2p_uhlId&(%gE@M3X2xlR}jpLakSfvBEi){0=`zPSxf7J#mYWLBsMzn zyja1OpU@0;j~4`kNSCx{XlYa3G^p%+fXuOPRId@h`Un*R;m`E>n3a~63QG#!%YPh0 zoD__2y0WiBgTqV1Bd!}!6>ZB=QLp~Kt9Fg!-w0ZzboWP)o{c^$I4*!QYjP(kgak<~ z>gEA~&+&EV{B~7i+{+`XQplN~Hxk&u5_|QN!NWH27;4pEJV_$93hvqBZ_w?|Aq+|M z2`x^d-0AUSl#lQ)4??_N>!!|2<{imJo7W7o%;j$oYZXYIc{S@GPKS7_jWfgRh2t!o$8B-m?o8s)3%{yb3QtaAoHH*z4643xs@a30<= z7rZmo@*Oy^^-fhD_r8crAH51rS`Jf=w?txv~oCs$`XGir80O<5k`I+%JMJe2Q($PmQ!mpQMULg6z&zHT3hkftGehyulFg* zj_w(YDDLzMSN98;&ZMS%C*l8Nw$#s(=XcnTgFRn%?VCBY&F@d0#B zPU@!it6CRG_VJNQ(heA@lm^h+y9wZBS)b^G2&6=h#Pu)7@@r~}=v*xkYDL7NR%zlm5Mlk$HB6YC_>Ce}ICXaoZJ;Um?)4-8&hJGb;^RP!jU zLiC@zp2m?$e{N4^F9}-|5g5jEtk8_!#Jm-4>a^cHl_CC0(w-)V2i`(+HR}c|zJ-S4 zzmzHbM27_7^kYGkqRxKMETBo@=PkU&&WTA#$eS3e%pmAxOW*W}x!FOnhjBk27 zN1mBzoPZFf^!!pOP$93`pU#MK!#AIZBNCniKW366k*xLt=Mw@YJpK`oKGRhg=cj>W z5wSaodzqON=Q?rf+y4Mc`k>$+28wAA#Ws!v-K2vs>ps&4Xi)r(0Z}kWRbR?y%d_Gg zhbPq;Aa{(-BtvG;_Au?4(xO_~9AZLq>hbxz&qocsP!mG!@z19Ea)J|vL6G$`w$tTy zpDdG$=>1u#-eYnKl!rE<$Yi`Y=k3T)m=!=WKJNaj9A4R&iedcqeFl~lLR67)f<6+K zou!{Omx^_wNR-}setrWk+C7^tlO(NyvrCaW z*)`Ttoy6hum=ucO+sWm#0go%oWv8~;^4KVV zlkP4w9pxl@1n`*-nZcQGFQP4)q%puZN>^kk#07+`sT({?`(=}S*=f{ncb+g^pNPcI z0g9|eRrKPV6k+F!aSjXk5{lde;ux1?R0?K4ag5wkkx40K%myA%}CH?#{6 zxV23m%!TGy<6mkT!)II3q3tNF z`mwvTSup+0g5aSKSi62R*>w&M($z?&^*Ts#-_T~!jVIG4<+{(;QgF*zX0BsgL)+~E zDtcmjP-pSW7*zNMY9|3%J0VsmHvJUzX4-Fd5CI1VL0S4oU2$w7wfi2 zz&<(AaHa*&IBN0(I5&&Uj+5@O)c0J*li}DfVsH!(N57H24j+!OilWqvO)sua^q6)%3H-vmQc;}i4u1iyaE1lPQnH2|OW(OpT z4wMx`fz_1xqS%5knCe^-X;K_3`YRblN@ba@dQD5A{tW1vDS~!ZMPj< z6~8jKvG7`IOj*3g)HrJ)SV-T$qj4c9f>`TyDI?MjheTvRV}ZG06eiiPmRleX4lZwY zc6766k;)zB^TLS6g0Y4rZ;>T=4H&DUZdI?#AOBJTkIL-pUMfj#C18kI69%ui|Kgt=n_;??*4yK6wo)BtR_Mo4#s6v zzlN8u6o)aF^@|&h4RA7bS(JY?-ZqYH8RwH_o1qcav`& zoiD2%?Hm-TcW`Cymnapf)~}*w$;DXS_Hyg##=haL_uHl_dN_A^9J6OatzXlYNY)BA zT@!tNXtsp@9nW}r-j66g6uu1okH$WLuy6-kRgng934h;?4x4AGae2ST<`?-*l)TKQ zPi>O)2kcSB+5+iCie;!Cb#rJQtvi2flr_tC36FBk677s4sd9IXY2Slck&I}dUi+8+ zNMzVG{GAC2p(Px{KDRrF677E9Qp2&pNt5tB=SGDApMQB1Mtu;U4vhWMWIuM?B^mN8 z&pL`^OM=<7yA=HFepO7Yaf#kg5d9%u^5^nFkC%KaOo8$O+SFtoQ3+c9{QT74e_{6_ zUEEr1IeS;|9NN)55#K0@hl{lrb7s*d3y((S=PhOaI1lE}$(?zZ@p<5e22RY-apke{ z<#I54+zgNB>jn>&P#eHKY;Y+av}U)mjb}ZQ?wCF4h(KF-(E##JJ*Y;uKuDD-p_Py0l`Wg7eAxu8HHFAZp3odKw-WR6GYVK#0t*2?=RQVk{L=c(CEF=@} zfpj@0-lxxC*=a;lw3*xNq|VO-kplkwEd;b^kQw3p@woeKq*$x}Z|NY*|4s*4I9VD0 zFX^DP)22j1&%ed#jbJOSZ5qP`WRhJ&QOk16CZ5tamM0tnXry_fA2`VMpWn@QSnT>z z-hU(qfVhyNoHJYTCc(U%|8HkV{yBBX7&}Dy4eCwJEPQ%9Im;DK*ue|!P02hoOBI*j zv2eLUzD?rp`EncLJ#x4~K2GME$&zns9!R)eaCj55c6)5dDt?UyM6eO z;5h`_h|+8u+Mrq9I_NGlrIPl#Np8#dK79BL+;1RBR1F?EKniUPU#3bOb;>R(e@f}e z8l|uEpZ|P4NL|atvWn)(z$VhI}RypnnicX)3 zOciaWJz%HTRWn&tanX1*+j!kf4*>m8j(AusO?TliCM!z>dZ)Tiz&b)&nXk!M!_@0X zy0~QMNIm%6?)Xr{^))+iX{@;*f()g7mO#+7;27Yk7=gUf^tgCxq2*7p(V(AvZ#2}> zRl03|pg?xvNv$)vY+rp(#mvqQO|Jf{v%Ca~*D}uud0_0l*rd|B5u~;g5i2{%udh}6 z_p&CVw>@JzTh{Bu>0Yb-s)B(mEHCPND05b)x!i}|KB6u2zPGa8CO{Wu3dIw3U%_B&xLqPkC;Hx6KNWw8s(-a@Xx53Mj#QT_e z=!eZ62JhM5i$uZx{KN(j^p&;47u&oz*MH4wOC;LZ#w3yH56^97c>`KSuIHk!2I~aO zidvV5(iDvXWc3wu$k)EbqIGj4^dTki5o{i<|)=jUq~Y7unujM+|B-1V;r7MB{pzjKCk9SyWxK8JhJ zXth`%SZ>amG?6ZE1jXt}qMcb5EEqOJ7nEgzqpUzNHk~a%!BrOXczGj{*jMJ`Nu4T~ zp{)m@4ILWw2c}4S&AHIDXujbL{1+|^l;p_A55UimlY-^y>TqjfP#eU!QIho0ZPB+w z7Ms?su+vZz^gZ^GIQB)^3kJEcR9|>8HgEyt+}(b-Ko5g)V0V_@k@mnXHzRLsXait+ zrvUy?uE93<=5~V^<;Z5os6T&L^w)>`4Z2)_Ui5Y3UjFD7I<4`=Sm9zTWu%p19QO9D&c0vi;x-T02Xj8PJW?I zCXz-1DYdJ+cDjz#Od<0uaxvcvFD7fD;m!q!c(l`VHxXXXHHsF2zV~IqeD@h@#E03d z14A)4A8QA8zB)FUh^X*v&knu z=?fdq6T0qf!P$*jE!rO~uZEdW@W-tpDfjk-C;At{VUmE^&5OgH>Gqf6rtUYJ`dZ?*d^^Ih?#gVcKi0{Xjf@HYyVm;ehD_Mj@kF14U zItAI7o4*)XGgDd-7-mY|Pq7+h3Oa|kq4lB!1|#Umq}$&TVjb$LZ{+yB64=%mj&Nv) z!2EL&0Qtvc9BY7ZMaR{Kg|10@fc7A?u-NPn=+;Nu1vPE(C#iA-|{-lr65_ zSBnsJZi^YEoE7$p(IE6TaB_2&yj++17~T)oemA08nge5pp~ad)QcR)ce)bIpOi zUP|xhCh)GK%x0ViCkm4DN*%)0%+rploC7<9|6o7!n>kCua?g10oPVseRN{e)fJn)) zY^)7Kxir6o%=qZV#bUbTGZE-xU={j}382U(4escO;> zy@8!elU*0p79?xSOS>ThfDP+7=tE=y^bgq|#YKR~w$em9qf3(lBgZH#mb6Z?kKk%F zf_8|AgQPXaE&>>sZt&0M1@eo~#e|{ZzMI&669$<<6B6^8_v9Y}dGJeVHZ#ugOoaf%@=nQbeg{q=L^S>hGcaH9WG;WrsKSo|19~hj-ZVM*7+*RSuBcy z4OAV#i-0lfD5_M)>5rS%pGBk_s&W&4Q?*4DwZv{vIo8geyLwzwwO+KY1_qdsq(Lkn zfTk+k_VhW2V1NF_KaoRd#8w!!EpT&OQ*A%EX>30bsTO%hTrAKhZwWUiGS^rC41RN8 z48H<&Xa+?R4BBaE7^v7AcwZ`M8`Ady&)~Bn=%-%{XO0mH(0|V_@doafWb^yp zefE5~ZFDeJ@R~vWs>O zTJ40E`r;QR2Lp-;)S*{(Df3DmJA=1Gf6i5h>CfZ)WMXMnq{<#zG?zq)zZ5Vrg4$lp zpKlka9wYlV27D-gS>i7@2kbQ=I7G*FAG(d4t9po6rFjmr#HiSPOo_qsjy2oLUv0EQ z_oF!2+!$|RCikR%J_!bJV4;dUreTA7m-U~JeC~M@gij*KSpaV{>M*tQHkBKQ7J+Ag zYgI335Z9JYx&igNu6t79_1A|3oqkTzyO!2v7o45Q)5?MxA!Zr@m>n(4L$oig=^J^LQ!rjXscFXyMdM?FZvsGNu0W|e$8L1wrUQvVB%0(R2vVA z5%=U!*UR(AzP;$XzzI#A5sE{5KzNT1`7OOxFJ6M^jbnXc2>?9+!?TR59h6|UyfXwm zF`;&aYz|l-I)E}YYf)K}08RkKw!^iAYrD;za}g_b@=b$nulMvx@+BWrH>9K`oSzg$ zqtMhg&taH|s*K>sk&&nLvol0?rw(vWuuHvUI%FM}#{WQeoN~5)o~OouY-seCC)8vb z^LxzoC2@cQ`0QXGnRBT3k5wCVo-u1)l;hHzqR!#RmmHris?MTm3PNw_ZdtBLA?R)N zH>GnX(!3Ubq5hcg*N3$#JuW`8=>C z>?XCfr}mD@lF%(QosA60(DVtAWj2k&4sBhifk(=-7V9VC{l#xwPg} zX2L$DC_!gI>TKC|IyT~(togLLb*2?f+b1o&LoPgg95Crho$Qi~`?)_wfQesrW#_g0 zz=;}7vq9NahO`zoQG@V{;*u3g;43ryB=xsQ?la#*WtnJ*0|VOm3tqFlg3`zRn+FwdE;esAT~DtumXp zR)_+I4}RzhtbiSk;;$W!wP^yQX6XQ*RnPXyo6y6m@^36t`2sR)bgO(bWY! zeWPm7R^Je_{zN=;ud{=Y5Z@@S!bW~s#I+|B&g5aPmm%#Pyld$38~l?e({4IfM5*T^ z=Gz@@Hia`!3GiR(9udI`orr*~MO{|XRX$vP!45NrNW73DbM&ArAGn0!3C#LPg^V4t zg!4KI^pJr-Ks=Tf-kqB4Mnv)xsSm%j0R*sS0H!?{vRu8rhh}b*_)$iQhjLenHvSCo ziD_!~#j@&D3?QfVm0YmxO&fwru(J$m&AYUR)1jYCUHz2*jKTB~>6&o{m7xSlKwFuz zPrApo6i{R?1_(#X%)2f-M)w3~fnGgY#s0ewhU+X1vZ_+<2yB5c-`k5BBqLskwM+XNb)`Bi}OI@qsHiswh!4H2&h+NLhap0c~;y_qV80XStFL zBSN-{2STN{I&>kcYJ!vXO^}D*{c|!DvuAv4GzfYjizePG&{3}ChBoAZTv3~DM-D5* z8Nl_Ey-)`J-E;|DQ|v>IMjA;L4W0ei!yKW(n6gA{=B)xl?zohs3&hrEVP*P*YbP#5 zLmUYs0c3ZBY^Peoe6t2Gt=KB#L{*aLXs$2ra|aDSKk zS;bP+WC)rUA>gAMgbu}{2}h=fhBvOGUZGToCCW+Rpi|Z`R5(*qm_&1-(ie)@5s!Y@ z88+SF$SC9BZlu6LpR#YDqjC!U;pdX0gHBC3p%G}Owsp~xV7)=Nh0y%WbzN5DgC*7< zuMp9E@3*^;_whvmkUeLHjIdc}TI>qO@JT;Q@SZ{OzI~k~s6k-c3euhOj;%Pk9g%I< zcCt5`5CZ?rpbgVe6GOvAa6skbtsp8hc>k4Fl^GVGlPNqJrc2Dy=lajyfOnMX4_TYbckL<{9W3#J8)5sfcERi-6Ia=9(b&+VX(QgZCe5OkBs+Fs6+{>J!qhyyEO9z3qDw4PKr=$M#RiVIt=y!ZwmGe zWVMp7LYdBaog+!RLE+?t90_O6&4g5K@* z@h=%vUkA?FY*j{z(=cC>^X^;v%RhO#kU`SSp`l=hSbI?%=s&8KD)d>j>civ z?r*3x*fPmck){2-^A=SY#d8BaG6FI6c+Y=LhF)TAm^0NOJNW)=ekbp#gi%m{&%8tB zj~E>7#{A&_cB>`-w}g)Meep|K7BpVutiu+4CyOusrkd$s%yNyi(ES^O7%5jvFW%cFz6Qea4693LP^K3+dmGeHVXvl09n zc;_1L6XRD?7ki)5EKUhjHjAaQER!70TZI|BHSGPQk4fVSgSjj%SHAR<^jC&ZvtbZ? z&eH@NvVr-4R+rA8YED>b)T@FIl%q;g>I^70vM&GJEMD46Q-zi_6kWIJp4>j+oWZh3 zM^RW>m@i=TP7ErmHh#ZCzwuaQ)Dc0_m(QY0cPsbR8g&&ddR6v?v@%;Szeu_ z%<-f_=t>sq)@7O~!2(U%GvrZ~71?H;nnQ7QimK=kL#wTn*MU0RGEB2gNvvw9?q)=5 zO+C31v3@24(%Z@Plob$sgCiz7NU-YKmgSBQZR<*p#3G(0Qb=$+M`Tzs=$Fy!ykQnr zNW})auyffE>#y5qI0)cDpjZ}rFou(+sy609 zb>)uI*v?37Oc+5`WHbOWM9~mj&s}dW^|5O-Ei|W{sFZsD zc2^)FKVx)i#~#X$qtVt~L5N&Ch26fQ_POC0a8 z=ilnL8M5_KWQGuCMgQCY`gHwqjzbh!@&IBB?m=`RBJUl!O`?m?et6F1`F42ro-H`d z9DQQPZ(-vOqzN@WCO|ZDXJ3F6eL2I&L>gIQ6ph4aW?*Q4(flF?9RYq3QUI;i8{Rid?^!c z(?QWTjpyk;B-^Wt5`$2d?D`Mg6-JD4E9@kk6;=&ab^ipS3%C#49Yfiz;t_n2@+aPi z8E^FeW9%GwG!5Hq`)Ly- zpr?b%I~i)9l?~ajHoOAM{$3@44p>C$D%E_;`*u%QuU($!I666BzG?ALA_-50`?r~n~8vcOD0@mYR?xV~7Qp`Yvx4=gK zEVoIVM*jAYzaJJaZU0JblqO0mgTZ%J1+*O&fZ=+GxInyDg4oTkZ-XljC05@Y@||91 zED45=4lAwZAeVa7v-Erg)XrsPOG##?S3BQ3c$1!?`E`j zFjkPqRP!SJJc9Pw-l{w(iaVDF;iG!#jv}!OQua`|z8$G9LRQIG)$?KS}F7w)u#0TjtCHiNUs*vhZ7|7q4#lS;!vX>8$C_I2$B4 zX!Od#$AfkvEgP$Hz2-S?s7gtm$R-8Fk<+%|{$Y6MW}u;l_3$?iQb-(tRX&}-k;XFi zX~oc%eXiOr^(r?hgFK_i_M>bg7bxtg}Z^fg|@g`nN8Srp%Z80WXhVo(a!b^5G_9rcuq@!&ke=^z$rOK31Bof6z32p6g=Bn&x)R(z^9tNZLs<%MNhwIN}a zC?m0Cn}w1@iXuV274;TuZh&q9xtc@19c=sH#(=EgB_G@!Yt~TT9N#m8o{*2rYP)Rx{LOJ$=*Fz;&*Y>fEK!JfOqJPCvu*J<*cav+g)0j|BwhKIv1J9;&ZAaVyS+M=uuXUaXc%Y2y+$e`ZbFYbfcz+H0Kf*|DaML`s*N(Lz zCuaRD>3~>^_hx9 zt%a>|&LhP2FiU+ezTtD3Qo6avNDC&Ba70sZJHQX-n!{}q*YSP+HR&d}HHzpkls|89 z1;@i^T5|%CJ-6WaOG)wcw?;P%g;SPYiRdMi$!KdyqBio2LzjoTn0ZLFA@ez~5sE)- zSYK`~4t~#|UM@1}4mbtMugN%O_8UAk*zBwu!YO;zK^hW5TIPO!20-Y`aKaL92nw;p z=?hJB&}Fxf%D~|P=clituaqvwJF}UMM7GO4;S|`l{zrzdL0qE&EPtOz;640(UasFW zkrWM`2S%a$d9io7GhGSz$L%5#i4RN-jBkweO;CPz`oLBkWwI|QecNTHtu3=BP{P-B zI0XQ;&F$SOkitE-9x#=ycS(ouN3BQLM>G@zN7u4m@pSt~{9aXU@}K7BeYV-Hg&Nq@ zq!Ks`uvm{DPJ`RIU)9mKP~oVJA4-oZvPhVdPxzu9aPgKTKkWQ30&zci>mue$=VSP4 za6I*jEf~JR(uhPMf$}8nA17;7;P&Yq#N|M|wy-P&bS`im;}?*VLc_XG)q316(H-?t z2=ZVAfY3`0G<|yr+0xi;7&!v2*Fv3L1EeR2H93IN9TFo9o(gs!BCD0UXE$Vx91dVE zB|e?+J%jnN>;NvPFjmhWw#4-2 zUhb!}lq~2H6s;W0mS*X&KS}{ss%qVROwEu(t&H6()y zQMB|N>_kBO=Z? zub1&}+72jC^DBE)rvL|H7ckyHkN<^nf_jrB^W33SpHUyDN!z4f7&@YXa{u)dSF8_? zP;?gP-g&(sJ3@5RzRT-elFoh5JC1DXwNfoS4;5r%(!2^1j$3og(iD}$&4~N4I|i$5ikhf zbn)jn;3)Obj0YYJ4@RV0+G`L62Mqg=N&%iyXtR6;rQ*D~-$d;y@HjtGR_=7B8Osbm zY)RCenx{wCEe(b9z3a{NPLD}!32_8;dxzzH5nTz32HOw?eygagPpb?Ej$WasI+Lpi z*@3aaleL<82YAva1U?C~s8y<-^IL`GoG-zxebuNpgJ-y0>jF&Jm|08s!u2owSLzs) zxgB{NMALqN67-iqY5Zc+UrWTigNQdzNciB^oV|ZW`Ni^C$T{A)dv=SjdV=ekb*Fsp zPgX7EM=sFU#&wx|{Zer-8s-gHi(BBGoiRC}=Hi-X!~kt54oWF%^$M0082l6DunX!F zXDB1pxZf31UimN#%&aB@@U4rm)O}(hgNpdTNu!~si3Dsz%q34L$adNO!|^C|WVxx8 z8n}yCPs`K>bCZnl;>`YK`nGB+n`~|5SZ*S`fcB#yRp|UvZMm_V>TTSNREFpw8EFLw zq1ibvLXp2W!AEcimmu>4Kc6e_EshJ}#ljKiT0nBeJvN#5c&UVT-oiCGSJ8|C`J z;3ubi(=AvO09}Qv!&iu?ALb#PnZ;QW8-Cgmt_3&kN@sHpZumw;$B(X>ET7j=rm9NF zkpGC0pRG5HKT1jClh z#b=WYtAW7x(I*Ju;{{CB%5vBMAq+}zj~h4REMU^cdIWM6QN_qIjhZ9OUd(oGd@z%^;v*H3f4rVDRw4Atb*(z-*j)8v6Xgig zz&I+0K!Eb|a0(2M)!n}-DUt?N8}P@WHx5k)Ie~P_GzTpyaMkuiSrpi2X{8kU=^!c) z?NwG=q>BL0i=k4iM{4=;>*9`)Ks=Gl!aIrk^yRvcWHHxciAD4xg*;B*Lw@o=;sakO zpqoLgqM6WNtox49#p40bo~YH7Mxljb-$*(=GU&Ae)gZ~Mbaeq&^jebm@rROig1?5) zL+#&=trVZJ@yMGgfJ~oxV)u5t$44ABr-U>rAkE*kIU^D&nAi+R2P9 zn%#TGmHy4E-6YFE#Vtz;U8Q&X54Ei1>{Sa5a-)>50{|aFX2U_*T zq{>W6#dAABu%Z-X9slYPEHv~c2ob7-SG%YaqH_;XtBWX;iV;7SOE_^n{kZ23I`I;p70up>-205TBC7>q@>QKS8gjBr*&$s_7Ve_bAIP!IA8!Le_4w<|d z!(=7PeNNPK7>4qk5_JBR8rFd?A2k6kVSMXW>qL0{`kRu@-E97GR^?Qu_x6&mL9^?} zT5c1h%DU7^Rs2%IV1bbS-0i)17`9$ReCTRI3N3^c-u$E<$YdPeeJh_0kQmDMy;$_(}U~a z$qn^A$)1$9EIyK^B%`wz&5(4yPP`o?&y)s9{v&!#`w_h^XA!%)di!|F7tEzU3t$+MV~`qSyM?ylIQVXjR!&wMUQ-wKAZ#+NAe(ouAa;A?E(+{a2Vo zNTEwtH{aWR!hefi1q1D34Td7nBmYG>BuypAHAyqAc4^}4@=}?@D}~5nu_iMgvyHj4 z)iKF39V<-p*FWEv+1DnkG%Pu%l1=yB#HuQLA7o3X4mi3*PAt)E(-kwD?rIkc=wtwC zTBscm)4P4=EHD0gO3+lJQJ|W-w};$-N2gFNj1CufO;}c*GW0amh&>-=I^_UCvci$p zYNp!>8)uOrSp{nxW*!5ru>Chg=DnM$t9p)b=!%(z`&I-_0$6kACS!z(tneVOb8DsL;MCxTqT9EY(}q?Wq=n)^>+7ysD}>_T zwV$Bn`>nHQ+0?e(BpO3c)97+xrH8i88IyxK6j_JsRFt@Qwi_Fwe);Y#$-FBjaNb~^ zn4~l&gOxgnkf|Hs(M!cam5yY+`dd$N!QS_GNrm^qXY$3F@#>e-;_unfE82`C2~V%U zgkUn~c`5m`5r>HLk_*xhyNsKs`qFW9dmSFx%esbD*kojs!GL~gZ}2kcX{E8-FtV)O zaYt0&VmanpeWTib@?OWq@Q^vE{&fRI0vUl4%oZ}mS#xl%6O-{DKRE@H%~xcDm+p5v z#WfF!XBQ$`Zq@l^^R(L~*|zNf`LWFwzthu`-|x4;am!VYx;3<*2I<1k(cSTax;jns zO&164#eDUl3o<)s4DCpiw}z-P5zzt6^TMhfUbj-tB_D0uas9jI%H~^Bu}JK^W6{l{ zPZ&vQGMIfTX~0fUNFR{})Oxu!Ag^e_1i2PO$ariutiHM#+ncZxvuP@QK4^rA4DRT7 z$j>0Z)|Aw%!M#g2bv>{b0u~%4%Rmy#&ip!z8>fL$K%bF?u?*Bk=dropFYvpVKG>VA zp&HC9%kI}V2s@Pllb;APVLi!IXEh|bmu{W>2{ewwpnGEWc`}$83j<2|`@4&8lq~7l zw1*;l7G=q*WXXr|@`n56T@~{a)iZ-7Y4#wI%*hJW2@b8WbhGsEjY2Pg%}G(@URIBV zzPMr1jAUowOjs|?V~|NMi_qZT1&`hN-Sro{#jYfTTNF+#SnSw@^cg7GYS`H_K49{8 z*3tRdIkWap#Xk@`aEP3ikTA0wf{8FiRp38h{0xq*M28fd@J2`aunVBAS|v+^6a>0u zx9QDR>+k3zBllR1MX}*zU>!9Tc~Kr&Dvx(Be0keI`QIZO_$7hFz?k1a1j?nJR|cdf zXacW0yd5Hv_DtsK9Zq>}$}zSeI*^rQ6b->~Y^Z`+%*(%>o`Tt^V<7kfy$_n;X!oXS z*5!!1eg`iVzPD}AmOp+*X7<;?JjrfVkH)>CU1Yg*msR|0>BS8&Ozs6%Af%-*z<}|s ztpK{tbX%zM1^fO-E9#&7Pa0J0B5ZbMI?0!FJ7uV|_5J|5cvi^#z5ETtI`I%t4qQZ6 z&w+Gu7>Nc3?-jQ1k@wiV;reqv`zM~)n`)3H3p0$OFZonV$Bx~jCPxpvip{DDzmvhW zMYx{}NBJw-vBh>`)lxTb7V0Sg4>@Q(NXlJ0H_SbgFv+-D7wiP#{vu-68V1C+%uNnO zX&x$DP5VZ2L4uz5)UcJ9K+|NgW@g*iNv;u*X2Ervg!|iNy>U>LHt zlpShDg|prgae%CHFrmD`E#_jrd2P4Di~$G^f4_k#FadU2V*dKBh--KToh6SmiBRB| zI}A7nkjb@B(wa5{qxL*qzeEMgI=nUab=M=;XbiP$U594c6f~L8#SCmt8+pgpy+1fm z(H4J(&l7dZ-vR~Y;@jlHKi9K6h}fF|YN?5y|3>UBYiUhFOVP9#y^{}i ze`KtJFdS8(CAUkC>PcNm(~z#(ONKPT&-I_4ZVPdn_hLa@4gfrL+84r11eR}jbwP)P zcZ5;2IncE=HtR>;3Sxg2&uwiXdJcE%gc?x%+lX~4;jkw7$9fKvn}Ld@gh{mTb4Pp~xz*)OSl zRl{nfYCM`DDlquLHDdlS%pBLzjuvjXnnVRrtqSj|w_J{-W8|v_foWbx@yZww?R7J| zd!kzl;2o8K%N4#bBl3%)g#4o%r#^j19)0C$4W0tXh*Op=*lv% zk1V*joP!14rr}nJvWpChGnjKmwl{$>`26bZbARvdEp^FD8C#Z~jdpE<0tVkTeuC!( zf7vRxwX0wiY*MgCF@5a(_1Rf|4K;nDgVRCF;H1Jbvrj#+X0WG)9B)T-n20RvjqCEE zfFUP@QH!oH9jq7;%KQ%6Cd(yo)su`u#VuLP5I*T&U^8(;3ZvNR9C`7`^9+Q#vIDPd z*_~i|{Y75PIM?r+k+b=+uRlZ|N5G00S~*bkD>7YBCkjFNPnY8W333&ZbvIla!Q~uC zf~X}vA@YJ30zvd9%?d&&PH!?LNf;cmvU+e@?gwka#~c6)Q=r!xrmmYCcJ4zUoDC$G zQbtUvt@CP<4k76rt>>cTZ4r06m|7Ic>x|jdz5@ zYe{%@b?lbq3MN`@YUevaKlsqS&QL5Ooat~|um)FZ)$D>_F{8cC%M238clGPnI%Q$? z^>|kM4ZkFAKXYP9+T+!=0k8?3KP^d9+K?ya9S8%Td57#rI`X}$KJ1CYh;kX(=pW`b z^p{kZHi~OaMvWG#*^;yGlwsUrt%p=vJk_@ zjE|pC)H**R|HL8*mY(5cNbL$aFxVhWe{AqTs2?y+css_xgEid_xQ_>p$@UBxje4 zy}|@g#-=L}iBZJ0g)dBMxC&`x#dauZ=>9mTArx;388WTK*2Yxz6)*PzhV9Hm%~4(4 zE6(g~ywTh-PVhVgMex7tg@m7fWS&sYMPwnII8KS|XxMb-BP0P?(aFr%VvGh*if4kF zx9Sp4b-uE=rx#yttOp8)x8$g<#;!OOkGEeFzM0(=0<_Uma6Ws%^7&J5QNIullX$!a zg|)sYZeJ2UbuRs_tCQZXv`HGJ=0&G=ag^@G9nUcws2LR*ep4+nw7~Z>8dqz08!3(# ziCFO0QG`H6u$b_iLgc@yB6|TqD*x%nyld^4No9SSIF1q;pP^WgAv#OV0#CrNL1^JO z(z6w6@EoaNH>yV862LBBxUPk`Gu!P0XUsBfr}jK7_(ve9v}hbIynGZ?o{^D-L(pON zBQ1wikJI`@^Im31gtBL?BG53nq8zx)|A8X9<}op4hVSHArBTXdtnJlpCQk$blOR zbP?|6;Z7rXk2IACGQ~yTMLoknKTm$;$RjK$3OXhiFOiO6|HNV+dlLQ6i@jRgGBZ3k zrnBtc3-&^3MnbN_1=R}mBvJ6ge7$(8<-U;qEqVa`g#ev!GwNIV#U2(dBufHxWbsL^ z6NtWhY~5tlA3yJTymOUw$jHt6?Q8RWVxNhI7gV14Z*AbUIBSJ#!xu#m$=N#dyk`Y-iIbA5YbYB(f z3pz_WZy|7>#w+pNu$+<@KDK+TWhSU1Z09?x-MY55rMHqHBiBg%wRvKI|K$aZfL&ji zeM@!A!cwETO*6&iP!jOg|Nb&m%U)5g8FNN##%_I`E{CXs$jdGlni?8WTBm{Z16syYUmQVZR*KUE*DYL1F^?vwqmPHb>%2YIF5FW-r$DkJ_(0x5V zIHkOQ84Xc>SbuT&$>9c{sYVyUj0xRmqrw^!@I&Z@l#Tai^g1DEQ22PdEkU9O0w)Oe$ zVshcT^Y>gW8i8_9S&d3Td^8-e9*u_h`4t!L0;B@q>*8d(&ABZ=P;w7&Eo2_wI6M`4Swf ziPD=F$6Q&b#OD=X97pd?aN!%pJ|E~NUvO?NEua+E(cO^F@ms+zc~}JQuQ++=p_s56 zfUjl^QvY$qH1|FGCq1!-p-@eh`UuK=%3S+mj|-fNJ&_24@aCdO7x@ThS{IRvdV4ln z%fi{JwLDcJT$hoaF<_|UJ`gBe6Pm4kJW)F>KrAK-wKz9d~Sm$E|`H@SHgKN{q>rI-K{6$1P*ple# zUUvVv8`S=B(nNR#&=c$#P+LH1a96%|5VOr)e;=jVMt2>QTg+4UwnMje2p_JtpfcJ; zsvE(0@teS=fgNmcyAC6*E^SCyrr|ZIg!cE7XWeOa^aaI;d>g}9x8&Q~#)PiVatK0> z6;B-c5sO#HL`y@<&iZ%VrB@EROI`JzPCiS`9oAOmKq0`kM?2-RePtN~nG4}DQ~~mX zj|7w`w2Qm_OWBS5(1V~@6D>8A9TE-1Gt@GrbZS2zx_V8b%`ZUXrBE8L8jd?2K^OZ; zmAgqD>v1`>h_96m2gVy>In=nr-iCuAwYLXn6RVg!zM_XUIz*5svEh2;H$S|`EF#ih z`l9R29)XLbK2*x9o6wLOs3LJ&4U=A}$z3=Y>QB~$FIO-G>17jOq)Q8O&9qTLEQ)B( z(Sr{W1lh){0Wo@%?rZ}dRN?b_si=XFN*sV`Ur(1py=U%Dp+xR2nGg;)5Rps zL+8;k#>~NtA!5R!z47n40ULK$LbO|aP_L}#4eo8C8hFwa7Not}!Rv@-R@jG9pA+6H z1efR;0p8qFugIflaSBVL9IPlUFL_Em$Y+^MI-f8wmT<%#4Nd6UrE=?j-p!RBlj;y& zTR&Szg(`TsIpj_JHaE#r48Cmp&aGHQB=?kUUAyb=eB#XMyKPi<6^InXp;Qww| zH02TL0()QBCVDh+(tV<`-q5O)4^D)kh?-cpbaP%^o%kKW>2ud3gAK#r{PrV7SB@@C z6a9dpsHN|F=lt@1@N*O$xcE=mN%sEzRi-Paw6TasqWAQ-Em1SI#TQ_tJ+z4L1?fFNE?@*LAKZP8z^#baX2_ zUYUOELNI>Gi~dW;3SQAj#UK4xjxJt!KeYAz<7IL3>%Sb$5mO?r-k-ap&pRW-Wc|<% z+2(8sTKM{QaVxJK_i4u^4d-#Ql=zPcJd9)$1>28z;k?g3vB{=UH#Gh_N3y#? zEb5i13;gBnO>8=O$7)fF$!Z6V91BWV{@e)?*|w#JLz*8_8NKyA_~piyrd01{%bv9$ z61rrwx-^B$Q^5k#R=c39Y+I_E=s4j1=IoAGrqEd3aQ+8BmIT)R1B&G33{oxIy}%;J zZA;&=9}KNS#%lY-^7V=}nO4KlxsX~$p5pOP3dK1Zyp4QG>&FRVGU?nzL2 zrdCQaYG0J=3ImphaLNW&)udQ>!;zh~P>!qboUPZwZ|&k+thFY`P0NGk9r zT=lJQG-og9CtO=7ue?EUdcF|(iX%U3>C3AxBC9)HZfVVrm%`dt`zP8X8|2yST#O7G z3|%!m&77XfDP7MU+2nVz!x!p@V2`AoiZrp=$5&1F*PbZlz#u@n^eS~b=&+xY$pY`y zA|~tiIl~MaAAfbW4lk(gI#-bL?dUljrIOMm@(rwN!G+~HCKsvQP^`Qz-Ln_aUr^_) zN6**?WCs!aMfI*eDnzg21$UWHz7mCn8J)U)JpXgv^EKG*=9YJ67D}GkADepFFe`bZ zNUi~@A%!woe*&GZBdxVcn~%vQnCRR%4u%cm0-$Kwa?5>(vRc8 zX*Wr+Qa8C;t5w*Xv{hv2X*fN9Y$`-iGFLjjPBK~1w<^EEgKK_uA?96YuPp<1C`2lw z?;q=W&A4nEZ8>miaR`6tf@_yUsbPNWKa-0@77PV|3y3&DlOFq!*}^k&WpC1hKJybi zNfEOM>Qi0mUT3P0vh7(_#Tf^O7Z86)-yEq^mdhr7F3zQvqXO~RzF=pmAXi81q!5Vj zU!B^tYFc$wckltwiW@A*g>)5p)*E25lu`rD>C|`Gxg9K6x?>MzI`!%ybLstD^{6Rd zO*Lxol!dV#I4g9N>eAdDP|K$Gpg4$o~B0=`B3vU?G1&J@Hz zXbd_T${?8=%-R~Ql$0#XjUJ^*R*DP|m_Xgg#;|0I*;+fsl?wM64nKacRH}GWD^(G& zzP<|&Ir|{0Wn=yal+1J$YM%|B$pRhxUe}Bxf6+aT3D}Ee;rDvFyuM~2zt=G@SI@^i zSo1@!Ul*SjubTZ;iB63~^)Ld0pent*8^HldXfl~|Z+HcdLC!&H34H~y(q}Wa(Ygd4 zN$EU!cyoU6h5(vss?7S_ZImkQ(2QXi5CHOcI#joPz%Y6DAUa=v8cd-`teOri`GyNT zO-u<@n4K|^H5xO6=kMLiH5c4K1oXkQ3+y-%6pp}X)e4Y^7!Pz=E%1C?$VtHEluW}2j+2J{+S;Dfcr&Gg{ zsi;0~loDpy1Im@Vf7$}@Us6ps3YILywjRnH*dr6gj|AL^I$T4pb?K^M6lQH4=7mh0IcUicdm+gxrrqn?JnvQ}V6n_QFDCUa%a^%1*%y&H zzt{d!xeclUN^I&iUY5ysA$Q{YuyZr5z<=f@LvvXSv|$*~OX1MQHvlemNbbI-s9F9E z_XQWk0DIk$oECrUN1GD}k)c$eQ6e|0F<05`jwb;&a&#o;RpM0(n+kv_$m13Gu#M0& zze6h4F4|yup9^@}KGGf--Cq80rw50&*MY&PY{%QXJHS3L2^6o(2_qJ$+o^OvYEP0q zY1mSK0IytY;xXvbGbH*2S3IpR8gYHhSu^(TwMF(EQ@LVPkv@WFLkVOjFFTtphq8%P zdb&`Uc)HesCDA>lqIA0PAilM7$&(}5R50Gb=aE_DjCQKzBV48JX##H^e3u*zj3m>gM)5B6y3L0LG{vnpF20)Sb^9MynP?fJi}lB*@uyF ztfyi_Qn|=^4&{;$T-$XP9`#xdt}vPBm6^-!5%$O0Y%gg)IPCaS)Z=k*Cke*j?^*|l zz?kF28jGwo=+WE3nAmv70l`y{fr=-43;YAmi@O0zFMas8!{9-o+rzU!^9;7-8Hlx8M0y! zTVqytCP1t{k|)5LGUi|5Ml2TJ>h_CvTI63{2js8#0NUVNM( zA3}~>yoW=EwRq|MjTw&_jvUJX?kgg_&oUj3;FjyXW8Xi4qci$1Z#sPXZ@7;@BSkg; zFJeap1ml1P3OS$j3eXqmy>8jB1fFGX@htt;s)|`W!CglgDG7clfo`kEZNeTfHt_;D zwx*3jqyaX^*;;KYM(E#0iP1z!E6}8CaiqNJ-F(>}{g5KE4f<@djwybUsg5G4*8~?s z8#!Yr<8Q}C~ZCINtX!+)zC&;CGu4pFgq zl!=4R3J)Jqrgg+`$#}NBMKCNC&~i-CN5a`4yi8;LJl4uA)bbPy8s(0MR+q%4W2-R; z{tG2vgKWjIXV8X=Q>+tU_%f-%*j}NWIz>vKn}2OpGr&#Z*&fhH@|H? ze_RgwuK_PCYJic|;HTG3gb%QY;3P^kXdf7@eJO-H_bJeP7K953N_!vT;y3IO5_sun zmv$hm`O@3;mAKCL&iZ-rY~ zp!tJW3cedtVGu!qj$RZTRdtR+Ck33{hC81XH5vmq-qno|x*gWU#^4*()Jt20fH=PZ z5s?E?{VF#tbH;egNcBMfRNYrcl&LDqgC13l)_#_xMSpdh)z?Lu7GGcf^1N;O@5hnP zsUzZ)jLyN2<`T|AmN?WZ5g429MaIwkSW6`3$$%BVfx=iZ&DGw;OkhGtp0lMyBFHz2 zS`yW(@o(3RN1f6(W zIQ>QbKOf?VPAT961C>$31(^qtIYN{?iuS)_O@$(RA~Oh~a7=S%mFdIcG$^f9GIc`k z_94*Hl==f(TPOgjd;m%zkry0Pfa|4ORl{U0@;k+NYdc+Rh@wiMpqSL-!w7Q?;=IJV zxPy6%;D)HMsPy>|gTDWT4l}3`C(_}f#E4NZ6WbA*py4##mXT~0N;JxSez+21!npWG zIZBbUoflGXf3rq7f@i4;)j-nRTtG()r$&(&4b-3RCqY6o_qlZqXTmFy<4~y&-#)uD z_yM+_gY2Jq=BbfpeD(vlHEwv#NxzM49N27X?o}LB2{%HOzkmL6S~i42!6g_?8(XA; zUZc2gqjI1>-voW#>HnWQ%(aMjP zzN@144>&@X}CGNbt&)?Y@UzzfX@~NV~5umMz%AL;OZNLvWB8d^Y<_0J-^gf>UYKjUQdu0axRS$Crx3`}Zyb%V805a97B1---q+HaN(#fc$K_&Jm^l+D!o+xnTWYsxiv-b zR|yd9jDc&oG8jfua4#DgVM@(h-2!;5!G7T;r0W+GR@y0Y@vpkZzELp=D+RxyGQJv`6L1eZP&z&YX{JJL$s=lW#FP@`O$(TRvt% z?{!FzwY>B~_y++$lM4g9Zu{(B}WnkQP< zGrw%FjqdR;8*7i3W=$e@I=p}|SgiczX`kp+Y_w_~0-R=1K;bNaq42I} zu7gPFVnJxLbVm(=la=0O(L-cqG@W+K~l&C~nP&#`S+(7R6?VW9)w569G>I6D3YUBE;KFYMLd$8_>bg3+M=CGcTl z@ZA`aD@fjF*`?Yl=y&MtjN8y&8t*GcdOcR58v%a{FjqlwE-Bq|IDAL^nG(vI#IAIQ z@NMnshAaD|LE*B6ByIP#deKVUi7iSO0`#=l=~IISGhAAGUzOJL^^;oXMb6U>PEG3krPD@4Po7+vU-+Go z)q_DDPT(JBEP3<43abqS%{rt~n*#=hkH1d8hy3qs@kD}7@fjr)Ucs@kpbHB=W$Q~@$afuKAsZ;uOc}*@3(4Wj8`?N6Hz33x@3%c{Ncn4#!TN{|R-8%G zm>-%bcrTK8WFQ$QZwKaj0H*_wf&YoDar{rR#>&9V{Qut#Z2IvSZw$L`InC%M?oqXn z7X{4H=ND9#mXIc)(JGC}O13ENAnpld0$IR{buP+}vhfr7(PULNbVYd|JMS}+NKBZP zTrfch@XzdXZ*y*b9TLPI2+I7d2Sz>WdE2)=Pvjo7K-Bhjemw48TrhJL{o&IRnoGOMXjnaYdePT6Bd!R^>stP)02ouhnxFRJ9!U=b%Zj#2&W6q}!T>-TMSYvz;(5&4c1(tT-j^^y1konO7#u8$f93NzAQYY8kl;^Q4L*r&n(~^ zt)Za$mxe;e!=O@n$kKBuHA%m+h?9 z)1iYE$1`#jDFhF1G5sEd>q^E@k1|rNo}jXuXg>?xSQW1PBLdDVs&~nRJA-MtZ?r!3 zTD--k3SWj@>JO^x*KY^OFoEol-7;Ym(}u=e0oqbXPlrUFl{Hkax8yGW{ws|=DT=Wc zJIlq1w@~wjRaWFlF+Hw&e&@A#9oMJ_jsV{3a&sjY#H6gAq*IC|Q)7vrm)4p-juQnXDy1gMEn<2j4JA~TZOR&}}1(P>J zc=gC)x0$Q7ZC5w*$Mi)ahh^rH)cBx>)GIZoIxE*pGL#RD0HIkAo1&8AsRe*H_+XOf z$nG**Q)K}OlTHeOug^0MvmaI4oF>@tR;d52Cuxoz=PlM@E|!pJ)I(?gQ0qr!G)|lk zMER>!a}Vbts{vhdxxHpr5vVQU>y*tNwH6l>5^?OJtoQGWyZf{0Jaez$Z82GOIVC=d zCLKLR(|e0;lf&Vs&?39l?4c~DRU4t2Y$oWoS8L?HWY3Y(s@uR7JZ;NH3KDZQ0HTmi zgB_?@)8+89_U5?zN6CU(&Qm%R0EC@YU4rz3U@`8(JazTaR2A*S0UPY0IL#H!+4P7u zzzjk-?7IyMd+{X!9{gSt#*qU{q=PGT%SmR71#5O--OH$GUGUrwYS!2Jy9d$J0&CNs z(0If6Au6>xMGfxw(W4KkSx~G@_LY`9g6>r`4N|`^1fBTvl3fFjEy{>9R)Rzbzch|L zrbTu8y@)U;4sMT2E7|5M)gud6jm47Yu<-Khk{-qZroS4PtCm3qv4U34I<73lu3o5! zy#Gj$`=0m;y?1H*QkP33>rPg!O1Hwv91e0@mWwRUQ92Cv;AXA$Oy$$0yOK4>^p2rw@0s=z9bgc3JX*m`{Bv6{_;0>34vtLAkrkAszH1$McoDN`^_%8w|Qoz8e9rY zRMEK6^lEPZr;zd= z<3WaGsW2p)&1DI}v@@ZQ&;(N86^Dy+g?cI;sm8)-8|YsV9-CxUIO!8vo9{c$TLYyB zEwH{!&nNiPfc#+Nz?MB-_fenoUpa1pEb4KUdhLILf^%GBt`0I8fVzm+b@2EVG~e(= zq@-rKI&(gHwnJ0Y+#bF5>IadDd$emGzNn*}{JYZIxa|u}(@`wAf5{{*NG#LeRS{RPi(dC9-wY7Jd85ZKU zGtD{guJ*nHAE9(tlrtukRS=b;@jK|VIo+0q5nNgr+bn&xf*dAazgKP3G~)FM4h5Ay z0e?kd=1uXRc(2JMyjBGg^FK7T_Y|YhnD_(G7SKo(2NZ1{sI~#6BqC9LqIi@RAL6`j z@5vs_rGeh~4zmI{jv^gLq6!OE)1`{{hT3&p6ylm3tYH=W7UO`tfY`SmHeuJMwnc=4 z-JzGzD_j>100o7n3h*#Eo|FzaXzwm&VopjX(_9MjA9q*_5Yaldgv!`b~72~Q-P$9{d7>~dmdCGkvQG#gcS}T#L$u3I3 zUflIvG#!7VO@u4&<={wEUR&}s-hfLxu0#oYBRMS_sV!L@KsB>*MjKHAA; zzIm%wlKVid3zu`iXIb|D_LIGh1cpydCT?_0}Kpz_-kCLG!LYvf z1GQeB_EL8~HFn2=;{4zorFNmr@t4TY+DIEe4P)V>&z7j<`FJH)O?~|19f2*AtHCCkAuvCY2QGU;g6-9NBY4pwS)rvMUHTfr7ZfP!MnsLYCeGUNl1jHAGKj-4ty62q&V$TP zxYCo`YV+LIUG$iN5r!b#_8Ch8mJ4FM-|bx;6FW8RL3KTr?<;LKGASoJ1BE-w^6oRc zEmb#SaSvND6!}50cTidW!C_7W4h`E^dmv*>c;(Fx)yqLV)Te?cs)Fym*U+{M+ zKm<#kTmiH7Ak_bnQS;KIIm-Y&V_3Uk@I2eoUC#f-Gt&~>MU$S8>n7^bGFuHc0Md;0 z*H)Vz%pntipX{BVn`zD8Ysk#o)~sp1==a6H&LO{QH_1Gg4TE&VMAy{QxN_yO`v9fm z8m+#q2&wu}ZW$`pYI;ykjoi~NY+nc4-$(ksp)imt7t&&?8vuCtR)}DheNlQv>tf)_ zY-Z9eEa*+rxr2L^od+Vhy3J~c*DFc()rJX+0c@0jg{~h?@_S+Z{wwA9kDLbBV(~F! zl`rWZx`p#lTGo$UH!YJjthSbwp8^}8Az1qHsTGz8z<;v zdASyCGB>28s>3428JM^n+zJF;6mNLVrC_(6bL{nwn=Yg z`vO=|-i$9SPP|)F;n+o5j8I)G79)b16~I4;T&E_TJ4v?91v`${{L@Hk?iA#VicDFE zF96K4>lJ`BZ1CUMCVl)e)(o=fh`Q+U@R6Lpfuxv#t6>e+&^Fp6GYj3tmxY#&&Dk0n z3b#q|Hs{V^q+atO!FTX^KI@>SHf`k7V;#o&Ir|!iOzkjzEo|RaJ0DubS%ZmQkLEq3 z+K!h_qg3(B{{EQNbrN3=Ph~r92-Ai>5x zI~(&OWQ=0D_fz!GyZ~2BiZUa(6FO6hr6%l3Uo9QCg2JS&5)ADkr)*A7tEa;Z?Bxu> zFiuK`vnr)>@bb_d6^RmflIYCR!V0%iH7{gku~e{6uDXsBKGxYg=y*|6n|*4I4&w*T z2iZ&O?0xUcNXD)#OWq|Z+YJu3CW55ciRrXG?d*q#NxZ_&$0o~M{sq*F3YZrej=M(< z1{_+#O}?A=);p(-jAokhDV`{0f7=&=r`M}90u;@!y8h}4Vrt29%O{w1g>3Db_Lp>~ zom^w9;j#|eU5W5J2W?M1ru>u4lz*$pCx^5UznNG6yU&XhG&jYsT%_LOWCb`6|3s7C z+&x3#OQ{D^m!PXX_Hw%DP<*(rwZo38FPk}=bmKANjV71ax@kNIN|Kjk+4ezt6Qmh1 zBss=ooQ~4$Xu7)0cMFZ;EZN-R?O?Ym?E$-eZ^vx4q0})=Cepe!yU#k(IqG*PYaYrK zcQ7%~OfqGcW94tPu-eD7&+D{t`OhNF+2M^eCoAU4qS*_aFX^@O>^Ch+3Wvu(aE<*! zC4WskGU-x`hY+OR75kheiVVJK#ZHL7GEATNN4@#@6i-l;*`?}nWC^pgI(I+{HoAfl zFiwpGM|{aA3SQIG9{C;pDQi28H*|U2KVRS9IW2bQQio0pYbe)8bTgclxW9RXWhr#; zrY8#qvR$`g%KYk2sRH#(QV^LXpXeg{x?QUs-C`MWG5b_cyz5X#^GIV58tz003hxS_ z?oC-$h|^OnfJo_EUX-KSjM!bWCp$N?f8f?&5tz+Bx0*kfeXqjgcO&TvZ14-Qk5-Bc zc>s0H!Rv~NF0UmOd4J^i$4cINR5#4wK82CY9YZ(mR>Z;ZaAOhYbD+2vz{rGt>W5KuDDJ{0db0)Vr)h zx7V!bx*~Pm)>Ifz_QdMcPR-}OU)+i00@}S_PQf~K3a05=LL~Kt-3>+Tt=f*w`$~oL z2tW*bjC@U%HrlTB%vw*E>hRCtW8V()L({!sUv6x`L`3#+H&_hq>G5Q!^jJafX)bLB zuI*d)jjbD|ZYOTuOP?X3<@ve1r&9TK4q0_4qYN)VcpL#}8r=T=YOI^GOI`K@+>ddi zjcV*HBxtS3f1a>uV`SahG@k)Qx>)8h76Nk>Dxnv6HY$7`Ko5?BcSVNzqMq272gJn6 zNAA1^yA~yzEuHnS%t)e4$+gSIODl3B634&?37s3`)cz;q8p+SjkJpbCJe!wS;9=!) z804$;rs5r#83{`d4V~$c&FfkOPV(~~o0ooCG1x%WYZrTel6B&a(6V)5XgwyA^H|x$ z7BCg5pc1&MRlF0>kh3m`B){*rELt1>$l~`iSFVmk^o^MY>NVf|Ora}$N|Fiq9-@F9 z{Eoku&^v#BH$rvUVK!0!*Nr{Ab>5ee*F9o?W4)hk{|)unGXwvtFt{E5;B{=r@9_j; zgEOucN>b>CebBt}FIKs6m}nh~{<+4xJu0w@lbUEjq;k&ypOtjxT$CEZXQ1SsqTK7q z>-%D4O?HSM2T7FjHil}&XTIl}UIyd0g8_tu4mgy=f(kY{SC^EMQ;LTRb3VlwAURP0|11oU6wx|DetBe+;=-NTtH=W<+$?=;)XnEs*~l)7N?)( z!*M_YI}$7KKwA5=+!9Jx^iiEmX$!0?Nnj4`tzxUJE3J>ad0bKGN_0 z*})bNiOWo2p_eT@dpAP>>@9B$t_kPyy@p(GAUXS?i7Y1&3ej#+~MPLD&CJw4ZeA(sg`mhq&&3At9oi0mhCn2B9jjcE z?Oj7D8KoSQ8o52w+s4yd;NCrP>VNh^(nEgtLdM4i%4Fx2QB!s@huKzpci(3AeMyz7 zn%tDC$ds~2!%dW2Wj-K%GnfN)2JhOd&^Bo$eu1BN`-f1`_-(sv4V(}6Ec)&(5%udA z;IB_tpq^ngw0GZI+$^4XokI;5qG|1h&}dIa^+GPY?@LjHY9`!M zcPzM>A6WRU>800X?74A?tG!$Sl}Nl7SBPuHZ|t5wPmN2FzC7WfRC7&Vz`&e%nqIB_ z%4g4>Kd_5#$CM_?*uZLtq+6`HNR@3=-YCWB%ATyAOY^z9)tN5sm|ICc=+I7Wij;Yc zEi_aj`svA(Hz37C!)tRqmA+j{tRh*tacrVMlVhN&G21+7e+q`wyyUIp4IN#Ce009t zQ%Iunk{HEmsNXTve0gqyqHDN%tb;8Q;ofBZtJTl2cEp<#ufyLExrVNwfg5T#ALU;8 z@(zMfN3e_od=7rZPlnPYOrCCv7Vz1oTv)3${Zli=NI4(O8IVhj#9Kz1*ccaClBJ$3 z0I=BxchyHjwk<`Ft$<%D?zdHF{sj3*pE~HfAwVNDQt`b#WE;NG*l6kR%;HX5Fa9n6 zWa&}5^rIws<@hHXzDTlH3&T%GBa1a~q5O^j`L%}rUZw1{`f=K_d^-*NUS*UCa{GW@ zb_YK31$g188?oSL7RYnJ@#alRn>dSO@&@aeYf7;0pz(9ys|Kw}vLGGh8AyRh8 zl3O2xEjJ11`KeHNnZzc{gj>1<#XlV3(5R7e6*>zX^I2^!1Xt7()XbI@aHfrm9Sk7) zU}Mrv6UAL))AGkfOzQg7LvA&Q?6$~%M)M%Li~4sWQzgT>4KvVB(s&-}5nZ;99I`*= zn0?FRQ(_eftJY{=8?smCz=%}o09zN+e%oq8g(KcjUYr0??gSBMmBhF z36xtZ>l!G+qe6-hBeeG_FK*Nxj2;*yXe@=1%RggSmnL5EP1Yx-dZswfN8Yk+HfzT4 z!H6}1irSFnKo&0%;HY8JR)!F@0rM|&ITg{Ei!W6akCT?SURyezIVsLVoy#>?*+P-Z zx(FbCQCXSXNg7WVwZw+^)`D2Skx18F z&g`K4R(6?Bufx9(F8DtxI2lxT`5K!lG%fIS3G{fP?*4%E2)jV*JC7IRz-4w7*p4b0 zFxjs>8esb>;ZKVPL4CyK&wG#FO@a{Ds{}YQkRk6jinDORS2pdht6_;I8p{=Da2K$& zpKbZ%F{CwUM;1eOz3~G2X;{Vo`Z4b)oJ5V-w_Q$S@IjPDM(jdPrq1PDA>=yF4EA}$ zP>=_zf2bQtm=7`H(Ue|zU4-%I^fSGX5o`P?DYrN9u8L27FH57 z;SF{v(G+8F{}kpXIMA(!hIMbCTWyIsWgW_bn%E7+U~Kwf!~_B?$(rsrS4le1b6Vm2 z{U}bKVwWIS1TPCUSJ`_FoA$N#)mc`#-23fz9{w-|)U!UHdd$yc$*A7vC_@2uM5YT>Cu#U>t_8U2TyD6Nzm7dSPsa0ZM>o zINwHlI&UUEG@j&7HQpdxzn|X|VKQM#GhV%UH0G#CDH!xemt=a%d&(YQDXcO>)rtv9 zI#$1OyGrd{jxsD*@_93zxUatVRaCq!ZV$w?4s1OV)nB!q1zDFG%%%;{3PQQOx{`wJ z6-_qOapp0&I25%~W4?u$4Exx#0lo=+a|1W6U}(X2Luvc=w%OwbuGOMvY?Nbp1)cOG zFF(+WSiP%0@@LJ^!D7sCnM^Ku89XzCx83doKL7b_(AG!epwaX{lrtg%Ti*JXB&_J!8r3 z7Hqj8UvA)A)3*62|NAm4#C$A3q@D2-R5lzpIA`neeCf;$L^(VLBn(-d-Fb`l?QQOl0%m4InFI=FvnGl9(!3L$ z79SU+Z|7JY_k!-Ofe9ls5I z(860GfbjY~JE}Q`?Ovy`$|coUeEmUpPS6C=)@ zQ$+9p)R(9V`@#b3w*JMr?0H`giipHL3Fm7;6J|FSTrSOEdpjgbM2n!h z7G%ajSD~I5r0qH79uowZxY`2$!(*SB%+ErLiq4uuGe)eJ7%z#yXB0ak<@OR)&c|hX zVki%t%$lS?p{%)%Ivo%flL?HFNKe*n@)xxP{=NDhd%u*jDgAAY zxK7@zh`kWAnat=+n8A%{&gCh`gr($7yD~6H$JvM$cdB>qoC`PZnLd4Cz{aAK9$GEJ zO1t0@>^My1&0n?vVsw^$!VBzfBmt484=`((qMQ(pPCZ!jspRryHc`Xr0C31ndu-83kK-yiHK*pjW&oBu%~GO;cGTqsX4ls+3v!qk`!tEm^tirLqo(Yd*NoSpmN$t7Hc zxp^YlFg7L~){B%TqmtOtSEYpVx)8e3!R+0QYjR`hy!eWApgC|5i*o3)2}r*Mz~0LD zF;u6X4SUVJcrJk{t|0qNT7n+F+Yqc>@AE<>Tq<6|TqhUS!> z5p?r#s<)tOA6)!s<;S@9!I_ShJzMD~UFs#~x?Ujl_|jHNLLWmW{%9Nn*qQn<kAZbzmF-P=iGK%u<09PW!F{V7IV0fBnTjvM1(Y{$m!3Vf z;#boGgSMcX8>eU+KfUg`(jI;`ESWVsP6R-qTwDoJ^?N&v*~)wC$a|CdNG!cnqIQQ) zX_Z=oN0+`PLkfyYDaRb%fV}oJ`N(nDDYBSw;sN=`1XOttmfj;jvB>&1+R}u_(pxIP zh?!jv*5U*PxHizkg!?yVH7OCETESV`^3RdakAGOH)_D5#;e7 zNzkLlP8MO{9{ z_n4E1U(;2h2oT!TWhDfc8w3%JGHLtr>DtC+N#=|LnO@alVMSK5oA~iGzvw*a%6G3u zd-sLwJniaUl$L*wysJIw-`PePG0A{KFTV(RDLwGqbr~i)`sg3|OrG#E9`@&?prn`L zBIID!BW=>5E!smBhH-;Y9@_Pe=Ej}D&@YS9-3_CU;Ptx9+o(OFkcd^vKV?G4H!DQE zt)pYF`qrQrb_uVNbns27deFCZso?}Ojm;w?OC`xQ#{^HO$4eo!deFd&?-%StFPw#x zWiIL;Ako)p9cMCQML(-vzhj6y_F#g~YCu?rIt||zgojr3w}{?1024zvPFiewAs`Xd zuhqfj4-^O~3^g~$3q!a0bY?VmSu-VG>rK#>c^Cqy849a=W{hbt`Xeug!F!o<3UAZl zY*V;%PT!ytH9FkeZd}*icz??;<7m2 z=uF(U{Q*2w4)svu8z1ks)WquKav1Lp9`71sN+7BVYm(+|S<;t=O#W)%sE4l|#!puJ zF)IrxNU3<2+gB0Tdadb|YV#?yNd;60{$K?@E%c=uZ&(ixz+cHdgFCCxwM5o$M3QZ5 zHHtwr&~9N5uPV-i!_1Jgm7F7096&Jt*rRzf5)@R_)!^2wM0&u7f3DEQif>DMe~kN@ zk{{}Ie;2Zh8p}Jug+d>mW}ZVY?`Y}0QFc)|v~KbDk(2AD`Z6^i>iPRMCpj|$HNvz= zeQqSc;DTUVL`EI;k9ah>ieNARN{{nM&q_8VRpLe8P0O0nJ z+-vZ;+4vQn&aPa$Yttf|wY|fl`xrN>JqvxJUb9E@)oAE=tGvBjN9W63x;jD^lQw#> zW>V@vQ6e;HpJz8FeR~;xm1MwMjg4CrJvRY1j*N0FV;GU6h^;3tL|eqH8ibN>;FO$m zINn{Ng6JhhY>t4Vz#G11<=^bR!Ra4RNk5N?q^A%TN7nWCu2}anEd;t4T zfAZZvKMPvD+viNTHjC4Wwp;&|{PUjub{LTebL}&q=(ob{pbFi8Ep0WexT>$1`KDRB z&EJwSdj2k_kxL^yt5@?R4$+L-cLwBd4q2$-P3cQ3`jc-?R=H*U~7qm$8p_Z&cQW@A^Tdb zvW?*$o$uo8iDC3!cn#q45b-LLp1 z?~iqWv#&yOR5olg{b`gOr;&cp?dLeoxUyrAhA`bvipPMK>b^rPl}+~SOrHcdP6*X8 z*~7)xIT%g4@2GaSohSJW0f2rU0e7lzCoy_wHC|hE@x{jCpYM^OQw;?LTt-?O8Ai+0 zu@=jt9L*t%ZL}qJs+tbR@QAiC>rlA5_Cud+I6l1fi0|mb<_$fFh_eAbN_Q#j1-^x)ySJ;kGpwo@A=FHVk(Qg-S6KH;q2d6=k>EdR&Cjnc&cxN*+B_Yc z8Jj+S84+_lnkL&FK-V2vIN^@Cxl?9A6O}X$Q$*Vl7Wu#5>)gSu!_*Bl&jl0>f-1?t zAsSgzri$d$5i-sRM+|D6w#(|?KT*>?C_kB*W>rHOZhRbh;O+{x<;^y@d#i`AmxR;T z85R-;_1Yc;4m2Xgv-Ct|hxQPGCh$Nj(A_+wTh-51joF*d!% z29Sz=DNe`PY0eoh2(y!x{VFK;S7Qf5gL!S#eG`WSQ-WTm0CKWdCCRcO4<`@QjInTk ziTlmS&~WEki&vHEH!Fx>HzjX%)3I&E*p-9~X^s{5lXftr-bgUIA*&8*`~BU)p{KR6 zBbXnbourCabHQiI!qSXbw{4R=2)ioXA|kunjlQ_o_BZ>7BmPYtWyUxckzxlX);M{A z0&U0U_XOo~~-sRMQ=k z6rp%>p?G)DkQ4xs0aXF{!}NSHKt@J(YV6?EHDO8M?Ue)lij<=*MxfCE0{r9&nP{nJVG67{Uv)%W~(bDP2jkQapk;3Qz*IWJbcbC;BR_;^NW~B>}q!r zgFb(N6pBpRU3)u`lLUhgRE0Y(aUqW2YydVwm2&<}0W!zqsFKnA?10@+#IG~r?c0Yz zEh8Opw#kKm+%5auAWYVN}hL=!!+$&LD~^=mu_^1FRj{|=)4!hRQ4UE zYQ}7qdqZzOaui_>I$^Qk?lMwY8aof`N5BXv0jK4ua7!LBebJ6wbF_rGxds8NalxI= z(~+oM*=WNNCR`k6K5`q@NQoD75H=$Zwjy7+5VQLeSRUnqp&)MO%~cjgMiqlH8%frg zYW9v{W=6V3{0inreq52+r$DpKbb~{L&jV1+ggA$ynN_AT z!c6xn+G20t_8L2>)=hv5LLNa20)rT&((H+-YEh^qNrR~})NTqj*DcoJwyLygrM)x> z`we8}Ew#6-g8jx7=yjFH_B>guLFOWRPqLev+Sryy_1sVlcl_ZRiLi6sv!i;_BZ5L;$cXvu#(s3aU025XZI*eTj1)lzl)@VYu zB5l@yTh@c(P13gwt!2lfU`{~%>BGtkHqZcdxPFR#kh31U91UpzHw*OHy4n zNIK`YLPu-;pQJYk$|+gVj$6SGg<_8p?g;Kw!{5_n!63`Lb=mM`Rd=_LA`1Zz9qOq2 z;(XKQF9EsvEBzV+(*N)%1Br3l5avK78OCLkClSo%#n5;3HG)L|cRMACmp`@p>~g@l z4*`m3mWftspi2|WcIx_VMUR~UjQR{{pUk%Ea!0-xFI4+6F+d&xn&T-u;;$m5Vo}hh z#yw?NKw+=w5t#%>4$@BsQV#&ryg2aGEF*-4-51!W7+$z5O)JtW91Lz2N6mNrVl$iK zLreZNE$EsgKS5gYLhCE-fHeb+Lp_4!SJ}k`IHFlQ zsm9DO9x8T|jNXw~Vvr-hqCy4*$e+<>nV#A29;zKdZKtRL4O_sCqat(X$$Qp|gy{0m zUT$C5pC2Ldq$IJ%DQw!n>-5rRLT%#PDI#QWbO$1%11?L|3u=|@#ED3Kh9g)EStcKHuDw{G)@ozJFB~ zcrKRau6BHK(gBr+ij@W>ltB!Fo?Wfx zd`u>p%1c2L^Bub+LS-`a_~wZ_$;QINv;%Ui!Ikf4$={IfNEqjcD?Ji1FVe|heVac? z0oq!|*k1$bB2EM=cM%&~F%@x&1Xe+7S0~A4jc=Np;`DC2n}482;<>{74D5?YKnNAY z;T!Bzc~Q*< zXhQZ~;T7EiUM+(lfBk1`-$L^deTmlC?6-xaXVwO0qfA63ph`itnxnRIo4nS7o2G32yQUEJ(?`Muj`gw#V z!N3zhUxI|BIMaSJz2LCmk`DevMTxD^u=Hw~Svo~TKO)Y34FELxVs4nFw4}XfLFvVy zzAFFT=ehOHzOe7{Rr!UcB%8GkO#8cTFQ z3+sUDk$3NCv8EkzK6a2XT!hi-#$zwl)WY6vI?gl~smdOcS4M{gYK8?NnvV1^3(W zhA*&*a5W5w-YnIGR`<NHd?{w>CMAmx68&3t+!5+hUX;4{51h4bQ3wME61!0ts$drUy>s znu8xhDP)7vc zxsK;hw3&2=X|#vxNRb2mF1Dk>za60#@667I_sYx<l zu+bMZVHy=Ezz)GeiCMV6FmBU*sD@!tRaT3&BRXl<7IIM^a{g%S5X;=s*2YDi)-{oh zMwDKuPw-f^oJ@3bc|+2eJpl~qX7nRt@%HJ1iP&^{1DX<-N&WIth(b_H?XjV9vX5ym zTu9Ea53R|wrjK$G-uct7GOh6V66L!f`gG0fyG|(S=>k@U8t~4|3ew>80DGIQ_WeE*!DFI?SfKXXOdSU#k)>9-N5v|J@h?KkZ1ir27P!7ga9CNGW? zfdGSYzJd?<#~~P6$m05NW{9PSfcNtF{B6SYUpzJfzTMkbz3dduUzc|I#U)u_lz+zy zCKwyA{2!5yluWcygsZc#RY>y=tLe9cM^3!6UUbbDi^=nCGhWxv^8pGLwUHHbMACqZ!;e5&u9Cq!_2c|A8IWtVypB$uL5Q#Mv|2SVdDb@|@ zqwTi_?vYw*38o2~WEgM?j7&DoRQvda=%uX=7^jNk=-1uQDWRxd1?F=0D)aEiHVu9H zzvzL*YHkxVL6EooA^b1?sk1J0^J+b|z&YhitEU9eW1I<6+lcC^*wKLqf-Y|34g=s$ zxCv&gj*{Bjp#>E*?<$p@;&~d58LxwaE)sHes?9I zV)%OG$@s=cr#0fJ%u5q3r z;pB1=^DtQtmEnD=lSr=Pt`BHsHSrHX&_9lC$L2Q!!7^xJi!JHnc;dQxsX-`# zC%gkm#J*z*L9CPohb)*7o(?kzn};D8JclHus)%I6#` zYLpM6Cv%TGC9~T!A;|V698vOsYv27<8k$3uirS(jOeOEcUH}d)0p0M<*9GH2$bV|q zyiR@2NJPRD4xZQ>#Vqc*mqjh^PTz>{GwoQxBpt=TL` zXx;lDCJEzFTA+EkyHYUR`k)rO3OP7^03L@AydALQSxu|T2OIRea{RtGW+fyQ_8yW9 zUI?>kvc{~J2BG;|$EC9%qq8(6FDnX(J=ZIb<_)DcY+E=nEi`wq zaf=le(IQ%xIYU>TpI1Xrl-sna!&BaCJY`C~13@$h-Gv^{HBY(9T7vQ|%HtpjZ1ud0 zuD+B}F89kkL;x+VAy&T+T9!i0xg?>*V4=(@p8J8K#k9iKvEj~?#HrVUuiTO;@li(< z2=oW&EyugVJwMM8JCQ%(evmTqxdg(ctYARI-_QPWmtmrHH3dOG+JLc~jOXcJKm&3~)8!srPC?&sLlh6mN8v^Ik>5JXos)#Gh*5 zin740b)!`TxbRnJNyr_wsp6~QHQOk_jF+wTMdKn=%NDgPlaXUj%&8|RlZ$Sh4cj)# zt&zOEFmIKvSnzf`M8V~tz~D#zcE4q6lo^2Ii;9g!%iu>YC!@E(;Xk7;g``k4wK)v2 z-fJ-9@0Tx^Y7}3|Kb!b@Hcl=#i93&DEIt}K_OW31NYurYa`JpcB%b<%9SfuZZFK-! zURpqp{|&k^{1@oP&h)<-q4+h? zHbot~AF3BJB=3^aWy}R8(M)o&M8@ikBvV8*-F2Ql0bB`aE4M&l^ozn)KIgF=Z{@^Q zvad1p>yX9C>ZDZmr}cR~+jD%{O?|jzk=!%qdtP1ZI3#0hVtu$0v^}wZs98tk=wfBJ z&7CvJ+>7fECg8;KLb;#a^Bsy}bW+IH*AFJ9;@#e^j~Y@-8YcAS#-iWN(%f=59y_#g z!%Oo16Ft>1tdepz`!&%_#!1>{l3UZi4(^w-_OyTH+Pb>d!aXa)ru%!h>O|(`wj;Wk z54zRf(r?qdmo;rmn(bt^1`fGHku6o7EiZw!LX1N~S=)4rsjCEK4}yh}j{wQ!J!hUC z?)N?MzkUZZal4U3mF$p$b*y9e2UGDf*4b`6ilfv92fbIDlI*YQ)uw4YKt+o1kHN&X?b%6D$SBXyy7C>sp{fdFkmfp7OFjfpB)pW%_SRT7DIU!+4Jf-X0W+E7WkO=}yQ=-}S#E^Rx8wY+4Q zA4qThWL__aY#3{igZgQvkGnNUAyZQY_D)q)hlg}Wx+NTzc!L|zbm$fNN2GkuWcH{s zZ-Qq0@3cip-aIqTB0XP7QM>Z}?j6WXEbY&e>7CBo5tB7&a~vjB%2nd3*q8!d6lBzy zG5U;-mf{R`wx{0tc(W&Vq++;IS!u3D)T6aU&-P6yTqt`gy|-7URHGs*_HIE6v601uaCo;iZnOz+G?HyYWXyDFb{FUU`~dewK-DVL zobsVYZ!Wmf=dPsgp0jcp=ZW^9I)JCytXz8vEr_HHwZLE(W2B2@ON~?+sYi6|!Qu^#xmz3naX;hNjl>x()Q~jhLa+&br*`69#I3lVK4{X= zQWRQ`IU46saqC(ZYtL0xZhtP&D+pr`onH8!*fth9IaUhoK4>=d@FM|4dFO zdd|?2x}uhEt6s>dtoT&Eb{s7n;5Z&&b5*MjrQ?$;liwk2jJjxfWoiUeK-+!PS&%hp zP-N!fOy9im(Uq6dTDrYMw@Qly-QGzXs9 z`VIQJChU9RZ0emu_ub;B(Kx$y{!`P&c^z@FLcXGpLMcH3W}~^4LHzD6Wjnh1`gF#pbrBk7K$$h`OoDQ`llT+(E5G_bf?ut5;M6M08og1D$eSC);gY@zj={ zab0usPE6!pm@hI<+k4k@)nq}UUKma9Z|jS8c-g0)JoH}8MT-x^z^__SuXD9!E%Kfr zC?w0?-;^31m z(JI*w-=*(nM=d$@Gy28zP&kb7QudyXJ46{r^VnR+Ru^XJe??RdmHqvr2Fndm>AU32 zCMPQkwQj7)wnD~TLjuxiAtySqL$~i=YQwmW-wx}9K7%y|%^-HM7=&X=#N~rpiM|^n zf`dM+52ylJR67|C=LOj4lg;q4GJC(HWswA4!fs~i4^TiA8X}sB0)yOYn;yO*fu8ZX z&o&vzPjq3eMGJLZ)VAM|uO&e6UQD0uTxWi6v((ZjyXw{N3i-Rz{3-!xUqNW)N6NfA zzVzgwGz+D4{#=yyEUeTV_1Zw8^(@|J?i6pvvHt*%b@%54T0nqaGSkizzZ<=NIl+h<+fgZh4>Le;(QCM4cp%Jcc;cTJ9SvGYr9I+2 z+jwn>I(3=;i1_e{5mGCq86jiL;Tl{9!wqVo#+!z3CRD&bEwfa%+)4?lXlS;z&?6C4 ziac(e?iO>0H30wxM)?px(9m>jfIunE!R!D2s?gpb0u6-qFfjNqrk6LJzl z%s8_@aOlLXJx|xG{m#bgF@Au3^qDTafY3G|(6+9Z3@E`+f7}QYnPkIkT}9|=K%@*s zwmH>4n1^YSNthyeAd>KXojA3@zP9Si6uI^{&N`XXr7$SPku zi=~yq5>WA7@AOGbQ(z}g^QqHC#EaC4!k8G+1{yDMK!bP$ssX|`w3!R?)n$~7w|eDEo@m*P2IP;4{aEp%*1`vf~cuFn8# zZ^N1HFYG4F41Iz^1dvIk@GpRI@K`rxUu^>E=gWatXf&P~W9{WA@yh&V_QL zf^2Y|knZF0DL$xW%1>~NASO8U`FH-zXth>vymfF6-{3b-9KTpD2Ez)ihj{|;>mjSb zXdBPP=D{si(iU$)rvr_J+Pa;q*vU%%8M1jF|_Sas( z`!0ylC-Ls=23ooUGX2KXcidH0_|s%qvNjj=bIV%bM7_X@ms(Dfvjd&sy)T~froi(| z3CLkg#|>_UbeV+1_VoYq_HRhEU}X5OEb_RDuC98gJ(o~2n6nCm3kZbnlBuK19>-6^ z5e*SE0USq^{ubRI6D@f3(54ht7&hR2y7E54?|h4`S^an(L#kNwW_2B(@;((@HZi2E zq3mCp2%V!Wxx5)`S(Sqpz_IvewEdo-3>gkc1Zfn*&yA{&$D2Mg2f z1Pc~jRZ!nm3l=lr!%=1mWCXqZH*D-vl6oc5CoBJQr_^`T>7jt zWn^SHH=!q(#e>QosqPYd{oBHr@(!w-vt44L5xrimW0nD(u@CFrWh$`xJI6x+MWrDTphqMq_f1SL%rV{?l@3BT2x{M8O6a_RqIUQm7$4~j z>M*q38bGZ_N%kK|bz=-S@!(LVDAm zh?tL^xPBXqU;52(tlaOZ0Ghgc-nXtbNCy#tk<`kHwL5%&-HZN_zq;^7weK>#*Pb`v zY<}&cdl?@$kiQui*O6tgE4S_9DzqN*PaqboIcMKNXa+4RvTN=BlMT#Y!wZdoAz2L| zWm@O!Sfn1dAr`WnG*}7(nH0VRFq8QKCsEJv!_UhnW)D2gQHZW4Y>m;2UPfY8&p0&w z>Z>7IdYo}9feW$dM0KizEAeo<9v^I4knq4HsrD9`Vkf`P#C71s2{qFaI$*Kf&D?&8 zId)fI~dPv=>bj}gp#Z%o=)zCn9~<@k=m$pIBf4+LHTB{*b1%VgA=X9H6iiSoSx zjl~EzAvQCST8c8H)Fancz>K;JxENYcg8UFtyKPfNrzJknf|E1Uz4<|QU3QvSL@Wlr z&gyD+QN&=NyMqsUlJ2kis7paaL%l0H_|!ZRaUh$#bZyx*p-#%?4hYojEFIR=@}F!= z-l#Jw=s(gee=lr7J|Gv|0UJEy|SO% zXIf&bW@Zf^u5@u$U$=d}*6iIMSSWKTdV^4r?3CH-^FAZVwE0Y?PVOV0*AvmjV{C7` ztN4Vq_;661KSjidv}J18+L3c-qnTRX-Zs;`u*|4}wq87EXCThSCy(!81+p3%Nk1 z3y0Ir<68CyUy8J7$tulDlN}bm*z}Lc-;f~Tjj6sBpGG*Vyr5`$tsq}DcOOHn-8uUl z62_F(yJ;m)bvtz*fQVv4YqeFeYZze^0pT~rmcW~D*WB=<>j}li`$F&}ndkq(3&5Qq zy#ac(1H{4wO1@=@g_Joj9(<6y(`Alesf~$C#+oOIuewu9#p^}Lat$I{_n5{f(B!D-V;~cRt4WHT;Wr`oIsdFIkffw=B$!*ru z3&Ix)`!)JkQ|C!Lmi~d?3LGh)_W27Pb@mNmORiQ%h98S;^lZ(lqy4n?kr&}dD27)f zgP5Rhd;?*pC>h~E%A0`dAX5%b&pbK!yb?58*vlsQtz~K;NZ8RXKgCO=d;{;^l-kif zLT~VfQ502@;pYYIA86Htf$2T#tUa{=*K4SNIV$=60deW1o&dhT3}JKXoWA=p06qZ9 zyXyY#?Os7Fj7#BaE&;sJJ29Cyl>fuoIdx|Owb?dK#kOtRX2rH$;TzkwS#ia-ZQHhO zo}9kzZ;bBIzhU3(F`qTpYU@%;z}gE#yxfkrK7r4q)Va=kU#f=j5FN-7e3}}ce(v}c z|1dBKTYI-oB(Euq*G9y4AKW52W$RxI1r7?ih)B7(8xFq=GSCKKn#o;E!55mtYk^0| zpyWcuF1sZe$*M2&$h~JoYoE5v#LN-%Irt?Zu(0ywBBYpS(uT-P%gh_~VYnnABf$Te zT?Rb-=pj0tajNbygJX*C5eMSF8Xv5r$GL_yNUFOkp>o7H%JL~JS030*rWwXFdKt>d zP#oGs^|USFhFh|$60ZNRV4cJEA}el*s;y@9i}`mz4I;#gu~fQfL2=b)iYko#nWa+d z4Rn9ocThRGgel?M>_ESN)HeTJeM>dw5Vl&bH%azzD-!Z z({3B>UVFDsa%k=4uHy8pQAuFF>3?p`D&JCpG^ODpD75CZWk4%FWhq$XU2U{ivIMof8hV zz6ScDW(l;dg;#j*HH-!rqaRUGZnwn5pSz5*2ZqwVecknkLLP38H@kv&>-{&=*6wZ| z&h{VT<^9A|%*l9qodP#}2Znm2WeR6AregV(d>vZ&VaImazTY1vfJY7|z)xEnxqJ2x zGRq(cQ~r@R6@TQ-rGIP9ZZv1%0I7HkX&w15kH`7W|3YSYwLeW6KgewCa94%Gk{V`Q ze`+7Y+Th`<^5x#7tP$O=Y|>;5<xq0*37Ir%Q@pfObFmZ zvsEU%-Yr*z9#}LNeoZl-bZ}UK?95i3;lZv5Ubz^qUu!Fs>9m(Bl^tHDsiv!5X(15t zWIPain1Ov57mx0nagY*}91d_s!THz9Nt5}Q4;zTq`hxF`>fC!>X8D$a{(0on0y3|= zS(R$a1MB;y?_*gyG?!rCrWET^rJO#(8{2l{#WovZmXu#^n^9M8@j%;8g=R2H(K^ni3*4RZjK!%c(Hi zsQ7DtU%jD}cnqlBDa94mQYLzqRx4}7X!Grub#g#TxEpRMe-P`hoX~j*dZe=j$zkWS zgpi;QGXp_t{jG{4a~O$T`P^j|LdEL9zfN0nwb=*h3W}6qszGa`e*$&T*C*9CoS{LI zQ>FjNDz@+CXE)Uz`fsYTYw0n#muD7L7lo3&-%B*R>QRqf&j*{zm&+3_*j)s)893If z3ZH8UoLoCS^w~NhHZWm%?tiN)=u%cKI*Cg2{&ZM+qJ>hAOw44>~~|h#K)bZ2c2r~ zk4nPZ8RulxUzV>mq>AQST5Bby2jR|l);rLOIr1GCya?vMAZ8kiD)!P?%g?ZFT9!Y( z=UAy^HL+Q-xk^alpS87Xx1HzbFJuHZHm>SMpzSkAqEV9dK#SiWo13teL5 zHjsO@D{r&2*qQoA=%?Nw)6ku%amTIz;+N$MW8HC4wuXSvd$gv_jF7?YP^XoAI~t~+ z@S+_qz=&i&^ZW?yOCBKutQINdhU<6gLjV%#67SG~CeR!;fo&z_*?y66S+UMOD`Y{~ z#0*$=lmD}p{1`#l#`fuQfhebTxTmcBiqRk)s~^(ghg`g9bJ~k%Zm(Y5EQ>4V(Zix= zi*eu>Xq;(>VTv*o$OgGwJ58tYWJ}Wh#<3`aFj>z`I5gzx?*1{>b$Mkj3n{S^xO8Y zZ9Cy`%kyb^%6DT~bzHoHCLL%xVcwP(6MsB*^@n-8;D%gFzz4?BIuAu8a}=mhk*$P_ z4hzYN#(x`(247zOqJK$@f=b>^Z)qPe8bgy$JKVAX-dOAcO~tG(9*k>n47oH{-cn)g zND()1rDrjBo0R5_*Q`Q*+jHBdFk!bB$=F2f4n{1%$wwNcTtPEEa3yso7D-?T-~q_p zS6sw?0CgD25JwLqJxS;xcN{RW_VY;HNd=G1yl4xtrj*F~q)~x}=v+8Rs{PBrO9)Qs zE&fw3TXwlD$n4i=z*fLTLMno(bP6;Yj8C}A_iMC+os)yFPP1_y_7b9d<0ZDUmX|@fQaem);K?tWe?ErZMc&@pjMG2u$?o$^Gr)jo@3?yC?>W ziFiKI=L)Fsi+@IfisbZJ8BFjwaxwcsf_^6Cc?BWSH$dPO@pV|e+}bBr0%^zR32|z^ z)!^yFVDSF1D9ibF)y*pxPtaDNdgpZw(5e~rRxnEak-}r#`!`;97<11B>SR5A85Ek2 zl6bZJ+IB}A#Kk20#Mos`W1}ZBu-w-d=3-Y;B!L1aFb?2_ND=wdrQ+H=?jrEZ56Z+W z=#9GojZ)k@*%OgKLH{sh^oz?=iCLFR+}r2y?j#T{M<2``pH9_NFk6RX@)dK0=7M7W z%`gel-eU#VJ_(<%$J5*z)q4EW3g~Z%6dDRFKYc%~ThPlbjonxwRlynk-aEf7#%*g3 z?Hi&$E{E_u@aAm(jeod=qd>e5Vy%=+&mNLl;4cp@~n>CdW%Eqj449aR$MpNf5_`7IweC z01Fra^-bonIQDuw^bMe>l8)s6_BFmKJAn37UGW=O@Gg1S+4C>TpC}Gtajs%lX9N+kC(vB}yM6j3}X(nHyW7l#hsXEozawyS}$~ehi#Kqz+ zj#{AxE6;_VvUe!xR*A=|V)*voppV8wlCL=eg_|&-EDT^SV3KkU4iNSZJ@S;~-iyBy zqVayezU+0}S>jCNotyx**~hlr-0SFQl6) z(NP|QS_s{RZ;Y+l32`We#4ETV`uC2H^qAg5eHeZ;&!-gKHa^!z(HsGeat# z-FtKq$m@Y%VUbo>^>felWb35YmL@YoRw8f+-xbdm@KG=t?=1*5wgI+&UKxGe4%bzg zXGwGXG#o$cFT|bsKtjLLWWyRfe1gFDlXLpkY{51BjA`^SLn{dx}|A? zS1XMANF>(pyV43TD(}u*?9z-v<1zBTIAMd`TNOENt!OVbp}T|NpNgt_Y*t( z6Z5%7*aKkOts-eO%j&35rV@G*lL2IXbT}jrZ;h%;(7j}kZ8AG%)Zd%s|JN*3C5&Uy zx>F|rKShtP^C~(Pop-i%IEjUXC$aP2UtV(+;-po>oxth(J{` z3U9X>RHFdJtm#ufgFS zcukRt@J?v2p4 zi14*Kq6%-sYBk}q_lgw;HM;qoshJSoUR@zfU5GPan>bh?8dLt880pyI7iTWGy?UwI z#6TvyHGz7J>BFBi2q$q0u-I(uL&Q^Qk^D5(nHheP<{1(Kf5xyg0H~YkNEpt)FlRrO zILO=x*97d!0G_BOX-Bhc)KA0a&9ZR0bn1DK0gcNmpFV-4z*it(YZr|Akv1yYt}HFX(_tP3(Mm*?R zE?AnJ9Iq=THDmEYyu)wqLw13 z8)`ou7W?~MJ}ZH%B5d%dVdd(Ky9y&u@haBY0Y}Bp8j!;4p7DvUooq9vY3N5lpv;f7 zL-daz!VVSjD{g7SqBU{4uWG~YaXgJ|y2*4fTYRYH5 zBAt(x^lS#QH_RQuCJ9uZU6;ATo|)#)d&+&+#U{y|VSIxoEBWIE=Z2jZEUS$zv34tw zbgR%$VQ)!y{V}=tQUbN*Z|SCXDO7^2cWyZvBY>V7i!cZZ>AV{GoZ}u z7K^ny4W3h4c<>lz?*DDijWPmTVHVFcNR$YzMv=N2 zLc%#()m@goXB*8;u~Yb6v$60T)vMW$b}Tj)O3yE?3O|KsIwpfFH-A%Hwe~vf>q|#;X(m%9k>CU0UROM={sH;eiw9B5 zK3HuD66>Ea>?|pP_?@gQr--q4*tY+hvlZ$z`h{xijF9KBT16py5ta3Ee>6=mx_~N^ zPw^FzeI%&s^}waDuh!r&_2gdyP4`?XS>=_*j*@xvVM1tPno*Qj>aP)g^+}MmW>9Cu zFouywM8+xkt0l;<)J8I2iy3f?MG92a&8S|_Tcg03j>^y(tkh`}W*+;IX=DiC7 zmVpsESNo{_>}t*~OTVq8-JEJ;U%IIT!@Ba>>v?F%MP7SzpcNpj$Er75;pO}^T)HMA zy0lpuFcgY@zIu7x&I+R@J+js|$mwdn$lbG_w3k>?kMwwR67)2rezuOB_bZjb6&*x^ zr#y&xc@loF^=`)!sO59AsZ0RlX*uE~;9uE|I;IL`hGz2MS!0yE%CX*N)Y*XzjsTiZ z`&coCbt?)s7LhAKRt`?YH`sOhBHCL=vFT4Z_X@9>S|44*F$;F{^Mss7Rj=j~MYg&mjAT{xJ{5KG6$=}4@+ zXzW1hz$x?X$Hzxjwu-@4+ztoLDyA~w5}ns6`Tm}ievMS5u|Cm^$^5gakB=Ep4)^XL z`Y*B%(`wvz=|%Mg$E3mmMW0{Za)5x=t|Iz8$~8038dI;y)BGT6TfTHd6I(ty5-I+5 zHAjBQwtd10nOX)TkufYRX#@wD#lko?B8f%9PK=@IM2ZwG<$>)Ft=cIWkV? zoWHY7!qxXK>O?kh^Yg2jVhD2@fI_tsE{Es9kg9knlmMZj{TxD^)EadvQn5%?7VMtYNH z?4ox3jOUEHI3o3%jC2duHYKQ!{}{gD9W6fb)_qtp2Zx4r*!tNMa1LY~|fqZ<7m`Z5C>x+e>Y(;Mh%+PPdt&_pj_gLsZ z0W+5W4$Rn?m|6c%|A*-zo%={RM@{>|iI-gieG8@BJb3pE0oY$wC4@aI8 z7{+BrjhgIrIx}8*I1}pGhswx%tPEhLnr-ZRD$*zY4nUCJp#5_BeCu_hcbF?#(bU;@ zB45zO8y72$HCYZRR@zb;f{FN`_D!haw;8%Cs{_Y;*Gl5038Q4qgdB!(j(nd=wC<} zv-Ur=`ZoY(Lj9^pI+uhQsfemmNRyiY0G4?V$5rP|5jFd|4}$VOW!ebj<~5Nft|gUw z+vB(exEP8|{wOOH`At%}k%NhHWG1Tm zcwI> zXm$sETNuVa5I#Zl4)r|}y^#njMT6t)rvzH|zBO*TGbnda_a8?g6oHgT-2>SM%G}gn zyHCfQc5C@xtynm6fEpq2rc6iOQxpcz8)?M+9UJ!eF+WIQfx`kAqqujeIQ_eLVS_^+ zZVbp_a=HPJ^+PsJxQ5)@21@QP9Q6$Kz}0%H4#`x^b4VEa@!yQGQ?U32ekh(XTL*+j z0TQC4166v5CmC3{tuXl_Nx5G^0?o48E`P$7@-A)Wp9EM#XBZnk=8*|3>&~U>Ti2bz zfInyIjz+*rq}dF%Ntp)^Lv1$~vV@;tVdno0$zHwzz0kJVSiRGAj}lS-`a_D#}5{H(>g~<$C^SaFYjb9ki3?&sJYouXaAC=g-2XRc)!Ku03{YxTUFtDA+rq~YU zysCYjzxY=_i~q>leWCgE+9Duz19jR~>H8A<{j=$>^fk(eKw z2$zwe9{9u_AH@y!@6gdqVYof4G|BYkNoAQoVU&BQ2_{}wgQl8xncVzAqFC`zTfr8L zg)RA&r$X00vI0WnUH<$)uvzAW$|92JtOP-bFA}pP>`~&yz-P!{^j18-DOG^X@YS3-l<#==b_4q)&s+aJVY3FfNF4)*iIKk8NZZ9f~>u z;XmuJag8a7IBL2sC72zM%bNjrhpB2OgK_{XfT`%4Xw#?O{sb_EbbKzMItZCDiCp;t z^XL#Ky}}KwY{(kJ;ih6U>9kEfr-ZSf<+R=D+wc&9R zPFqrQAbthzCGcl(y5b_ zHn1a)DZ()E&<10@H#OYOLm1&8vn4`s$!$a4;GdPwJDYH`9LriY1T_1Ho8QU;Z22h? z&U2>NfOuLUd=Ew`(4*aZ;2-AC+1&6zA%^zB_>hirRl4ZgAvlHx!s9+@r>Y8Vy z?{E;=RUcjxoVb@1wy};7^KIjT9)3z_bv|pstM0!TTa58fE@yU{(iGmY)G#wEkwt4Z z_SJ<4g1;r{_8Dq)Cbketxh#kup$5VR{Ai`dShNz58{OueR5T4(;KqoQCE*Xg9BSrv zFh#N=`?8C{Yxy$AvVJCCtmu=4B8C>;ofZaiFb>N~Gz5g_dX?LwnZgwYie;GLU2cIB z!K{@d4Iyx1=|GTigEPkieu6#}HN=??xiV9^M`~dK_rCN}JHG{&Ttz}bI;a1gL_Eg6 z9{HAb6YsdkH-GOGd!Tl)YDUa8Rc6?1W3~Lz{_3D9^HMXR?vbcAvEva`lK2Nf)|6G~ zjb6CY_*@7(i#!N<6tHtVZRB;1R9d>n9 z)dyLqGp}Wd1j9LVqV1H?S$-i9ed^z9{k@4|elcGgCW$Pl^dYf5G>~2z>l{ByUd3l> ztXqqTdt@`!5bi8d< zTekU$p;#H!@mcFsL^$5Y$oc+7gc)Tj`gmmX%ufwexF+TWmUUielrO#>eJh0a>wy>1 zHSMmZLDC-F_gKR!lZOZjt0%-dKqK425>a2JJ&*P!J1X$7E-Ve(3{#Q_{TG*sCkp+& zYAq3(bg=eFN_%G6MpeN>uJ}&=WYeFMlE(GpNfe1zP9JXcDAkN7fbt_P_-m;uOHVPy zmP2PY;wO~m_JOt>hg%oQ&>ce=D5RB-a+k3TmHs=KHaGCz<`gKuA`6I1`4+ z;3&3yP?=11BnRR<7*v`M$${aK?&Bc$3;~{=l!75b%~RP& z`k0aof9(6+Ds}hy+lq@%MdK`~KPh7-6ZpfTZS|4E?cwa`;4RY0eJqScF8-wk5?C{n z3Iuykw>>}%H~jqH3l1EAsZ>CNM>At~m8K|vRNSo33#dllIa9)!FwjhX+Y1QC1mIgM&Z93^L3YIbjt1fr6P*4R^!T+y=DpAdqkLrTh z@L^ii1Pcg_24b!)frk9`^Bi=_cSi;>2RlSr+f*_ks*PTMoO&?@h5i$Z!N!O@B?yp3 zHTg?LC52z>xLrK8<>V_6<9#^16F#*?;Gw(QP{mVP0qDrxD0nk*=#d-C4rxB!5HuB) zn++HC7V7mrt3moT=UYYAkoh}3bd}G)_oUZ$q5cy-IFflP?zHrb3?^kQRx!w^RkH_P zY9&Xk6qXU&#fx0|ZQOauAwT;zk8PHw4!kw8G*IPG)(B3+g+YDNPf{}gRo&$Ix_cN< z#+gLvhZW?5Lok()GHRCbWQ-dIBTfR{OHUr>$oipR#?q}&m`){uUVaICY2L^_*fjp| z12QgNuXDH0($25nqcI&@g)s@IY+Il`wv86|yMrV4V0uoutjtucQ7vQTMv)a^!vV#* z%*W}Isz4-LyyEVDb(%V6th1-03jSl)$*ig9lqqRY9U&;DjTX3cF#DP~N;?`Y%JJ#4 zw#<;;Qk}GFS00_Uc)^nVq&tTmJ`$S5n|d#bHh4}7Pk{qcnmaJkQAR&MQa4EYHWZJ8DO zpJ1n;9#4VLYc%37GVy#dP*9wi-o4H)VSP}II(IaJy%muqP==Nq3;v>D781y*B^*2174gZS;hLMU^gIz?M|c4&o$L^z5hFf|POoh%NR66Pj?$$H9x|KUTUAe*!A)mi zCnXeagIL-?<@`e2s|p&=xH6l{6O>CbAUxX~x3tGz^9>Gq9d8w7LT)G7mUZeN^B>M3 zte<7$}`v#5HQlH zjvqzEpKal916(sb&1)E}8}iQ$pU4*zGqGvrOl4;Wzv$`p^b&%Ow9C3}TeHyrebMCZ zXd5khkDKBk0Rbt{Mqaw;ozHA4lRrzA$B2m>M16ElEw)jXdzqwSWzGibsuTP0rhr*KFyQa_pYIVJ`A0oy;6BU9@`L>aSj7K~c^8($US3HN+R;IzkJGD*Tr!PxZX2;4nU-K8#4^$ zXdIJn85YgNZtL!XMm+qQskIP5hDoZfUEOI!LEQ7m{)Hs4T^^5JbiH59@!HgH_Js}C zU4wJ}OCT9Xc&}i6B=>E$SjPt}dvTd9bdLWHH3evje2s6U7Q;9=xmzm}orOgz4)F!A z-~8t=4y0jN|Jm0{#w;}>Lyh{P6QiE$-JqAe_!;2g|Jcg6gkV%9P<8IrvCn-Dih;R* z`UIb5m`;GFKJNRG)tq>@Z%i3UybzPB@VzSmRSPIWCJcI4EA&dE286`eHIdr*P!c@I zrvr+q{&hkwf?)6CtH3C|&Fv`rMK938WvLnu`spPZ4)#nOD`aXE7o+KZh_jXe>-$&a z9&2A`6W)f=QBp%*xd3L5<+H7Gnl6Q^L)-|nSPTxI3)U#E<_+8l@te=efd_)@U%REZ z`wDt1N|TPrEi(eJU^)^G)fO&#lI+H^#9MBeyXhj)e8NOLUD7q?1*Nq2ln>rt0mCGg zFw5s=Ea;kUatu@B9?91cF4-Oa2`)<0sPIB?w|s@O{u-*mH>0LPDN?oGJS2%d)^XND z2)fQ6OBq2(%c=J-e}Qk7Wi?Y1wNquikX2MZZPY(dA%%l>-ip{JUECUM%0|UTR;E%Z z4=l83%d>huBtxB0DY9FM2gdWw1FkU&v7-t-zclTJ`U(LkD-NiPVchk%Ietx0-ArG` zy|i=sljczaE$(_Yqp~d%(PN z?-9v&IDHW*iYXo_OyR=_pZkrvEG$sSo+CmAEGkP{Wlsllnhts1x?=vDOux#K`w;+9W&a*$KaeXMkd%1o4DsaoCg-ke&0d~kI!Gt2-c(p5qx>ftOk}fPrYJjvCCT@!Q+wFVs z$5r)GD9TcKN-4JnNsGF>jvo*rzyiK_&3x;whkHa;WdY!@WPNvByteO&KR z9I&u0Sghj7mfBQ8(YwQ}@^$tcs|P+0a$Ktk;Es_gDkGu?@!-r+%2ny*i@BwgPw9N! z7rlEy#IJpkzSCqHm-a+j|H`=M>nzwbmS;lp{;VeG4GcDws{O1cn9`egu=AFvV5OWd zZjNrt|IDO*!eM`Wfgw`}Q7w_D3+U#oEWD8}e`YR@UM6~|-J$lisKh+d{`(C`&*A<{ z-uhg4tUq9_+F-NVW#KBTNSA^Fe)ysixmn!yYLgGpNsIvQwcwe^ehQj)1(=jjCz;c9{}jG=9K1$_UkkpR(aQ2L|oQ+ zD^{R8D_^Gkd6S7yGQxozWl`inv{hYXShK3v7%cUs7i&y6uw22brD2+@%Br_A&2Wo6 z?BZ&DpVo&zDELRGzaa~FyUg-~X2-lAe7{+g%>EPbH8m;|>A{>FM;_QhXqf&Fc~|8WH!LjMM-GmqYe5vE`f*f=_o7c zHbfHeVof{~_DfZftE){G5zdWP1td}}^t@CoCMMEm;i;Mrg+U+S@CGQx3#Tgif60wW zQ(whiX+MumtmrEHdd~sB)egkG_0i~3%twd^rW1MlJ^zLq)l%Vb!W{`CP;IHzd#Eai zkkOBSwq10mmsnu~^Eg7GIf3@^Z=t{IpHJJ&jaCoe@#Lh@UK{LxiYCemMdeeuKBBTn zQk^EZXwYT=de?CEJhRh={!+vIy(??S%LA_FSlI*VVeF*U!aPzjT;x~ZE83v@0|Lfm zcS@I%x02Kf{y=OCz0!z|FzlJMq?l-MR6;&oO!a7_xP=Lsr8PgbeJa;w``cSG9-=G81&HVepYJ%BjN4nBXH*_35dYoN*n5JwiqVQ1Kl@;J(kpFu?=vPGI_hk25|g4_I3P`d|+ z84$n~TBDmgz-l}^o7RUA86Sj7$zc}Dre%swBK}gwz;(2`7fsk^w5F_vT!|xD0gC&C zW;NomhYb}-9Bqrz^#;Qqe?cc7*vZ9x4M+J_b)IZ|k-!YC5Yz!zfrU`EjmZxlNZLx1 z3$^wKA9PEvx5v2Az*>6olPB|bpj}bz+*N2Nf}NSGx3w*2+mCJ*zfi|*5r(w0q(mP! z)1+rNtvv(srG_K?eH04eS0f9|^e-HwEs#NqX(HUtT4THV%o901+mOch7tUPiK&uxR zj%EgP0Ft$!x-}I+nXHif%ido%&8ajWe_N;`^fcFDabu;y2uRMLLn^z9l||G5Qp*~y z2n^=O@MOMLPXv-I)pfMhvUrnp_6sI~LVP�RkmSO(2F{%m@6rB$3dQ4n}Qod}@S6 ztR@<#Ku7dIUlo6tgOxqF8U#PV-d*o`dBxCDA!791e*@1(okjN2R?pVY>er>!CKJI3 zg%YX=2rN@kyi1tSZF$$M_V%}vj~r1{DvaERbPlXysd!do{oMB5y7|+!NG3- z2R{B_G@&y17|D_gD9!|p-$jq=*&|HToOq3GKQguevZKEZ=xZDo z20_JsdcLsomduFGzw&5{`Ic7Zk4}Rm)^4Rqkd`&TO2kvvMB4@XkDVWJ?Z^~vbmaX3 zv;oLVg%3$gWd_aUe5jc61e9NXf<@nzhmvh}CI+Epu^?Cm4IIYd#DOH}X{)rt4p)$r zY9y3CSUM{_FdXOUeW}nf7Rvn=_uS##@%K?R?vGrUjMgzv6B!281Z93W@OciSt)a4P zBxAAyF`%-)ST_}~Q;Y>%0tJ|U1+_ZfVv8Cz4;;bsC2|t?^|-3GW;7p~`JkRap$bOTc!pM&hlWvr1)M2*Z7pQF!~LRkG5m$&=8Xe|Q0cdj>cG9g zXZgfz3lpa6>Qr-elA9=TlFP=m?)xd>gxF9^LS?wz-M1lXxm~{VwpgxiuQ=ps?>eUE zu|vsj)9Vfk9E5N|I|c2mF{alo^uUTTMB6tq?@uWfVfO&+3nfRrDUru7mpwkgpV)dj z0vRW@%P?h|6h`zd_qAle(4P& z20cirv-Cu_J?2w-qn@?+gQ2)XWPq}Gy7JDHCVtbU4HvZefM;D6;$UE>A9H13qr)Fx zK^(+EGz6KUMB|5){jNDiFPB}$81!a@%PO5=do(eDnM}25E^VhQQ90CjS?g7^dUJ;n zT>~XT+?=sA%s3J|aYtnumv+K3bF|Q4B%t&bM%#k-U_8vkq9sU`pg%@T(zJAYN5QR0R3jdUXW3#AFV1 z?`gA}P>MY36(ySmF>VTMcO zpIcy7JCF{Mmc7eJ?-# zi}I)1Wy&p-v^&Z$nQ^*VqS*60x(Q+(m;Oq5 z)en`#fzeclWc2RCZmv1xrU!v>@=w+N&AJR3Bj>8ii=b%Q+1uOW<|m?-w8jz$H{Z>< z=0xW*!F=pm9ZVAk|F_AkWtns?_|I;YsFhv>7&Yb%l_+j^VK7)317Nd1Ct@{wQPw?K zX2=+cx1Q{Fad)fQ$VrMh><2XF>T;12LW%aGFF9Qcp5k1<{4K+kIKYt`peTZinEMPc zu(9#r(D^>g4@KzRLDGTKo+lEKw;)JG zua_7x`1L_X1ln021sg*1DNoYBat06c7bQu0+*TL3R#D5h??O- zq)(vSbGku2`1OqNkQs`zXcM!_a?RfihMT$}s1(kxfhu`n{hB51J`rWeV+h^h3v~vc zVc(gay58j#%$g3o#Yn0d^ZTzUi!#00_&%9=B5d+ay)BWPSGtpFxvO%HE0*_y07KBFGjXnZ>sjw78D z^_+U1>ewbT&mH>#Qqg$QgQaD(EwFdQPOhQncWqZ2qDyF?dp9$6;0^?3EBqC7R z|4MSFc3JVvZm8%NqhX&&{Q8|Y_-$L-BhO3vR>+A9N>8YqDI}pyE&3I>>s$1iM@Wr= z&3;TqjC|YNCjSynZqRfsTZ*T`22bQ#R6@o|gaR$SGCWXT>{VS2@44!IhEVFvZiX!^e}?#%!=1|5ey753ayX3lrd7kU{t) zMX*BgBxdd-c&4(}WOXx{%QItmP!OM21?uvFB|lR4l{VG`crQp2&xpfUMwXxh)Z!Gg z2Vg352E@b{5L33kl!Y+lljprkqcd{qw7I}^z*=xY3n9kv7ZtD z3x^83U=7LkFeyX1%Tkhd;jBtAEYPEQNrR|dNh3~9qj{YGtFhtea)enkZ3w6@wLDF} z@WEH_E@f8;vT;RKB6S`LJaDga3_%*zWaxmr2Xrd&Dlw#TpRs)cVl>>}$X1a=qUyX{ zIq=iXx&ufex>n)m4A_)vC~)7MW$BJ88e+s9y$gon6}ol_i=9=zzOdnRo8>CUSj-Hb zXHLxXf)*cdX9vE=tNnwLZpp-%=~*_QR9A|9FUr^xe^l0D)CF~oUDJ$wWE z*&<;ASD;p>6HY-}>lr+$1N_opmDn-U;fg5Rkr|m~nHeFy`oKO%N`Fmmb#k1j)723O z%Ri>}kw4^Vq5!e6FQO$LUM$~c83MQC9Cou~OF5ojG7u=4DEy*}dJEY-39={Tp(SY* z7|vO8PZy;Oa~?hivPNB#Gv{iiv%ethD(^S`>|Z%l%uQpX@%d=u*F~`oq^BgN_S@*1 za!Cv-W?T?=PTeFeq|Y!|rvX;FoT(eTvP*X7cTrOWI+LwG2SwJNd*%pLK9|&WY%6Du z5gp`(%Nq#{W{Ng8s(`_tC%wIzm;EK&G;S1Zk5nc!z_w2(cvXHfbh6E^LWhIqD%=%` zHwur?HfZ0oT2r#?~ z-zxaL>L8BV>C;_4L|?;+|5c;C6+??q%G)2KL=gPJ>7IRR?ml+SC-Tt!n6-&%0$nPg zYHxv2HnhprzHr5hNPHr|)A2}TONJb;BUFUji{o2zs9&!0d&jM6qBOzjjY=ZoOJm1 zc+rmFzdL(rHzq4j?I0Cc(&yi7`*K1hw-Z_T^^!Xy+ZCP`z_DY|yG*ppyNJHFI`LICHntP75ZWSLnS{hRMVnEc7Qa<sRmnCt%0+AAlV@E8G8^|7-l2b-Dt6QCSMifBQW>zSor=U%icYFoZM>vkyAB1Mx&bL?^eH-jX1JPER`s3 zjXAUCPd{ofoW)UsD2}H8Ym!Hj)LG2Hr%c zqlRmQk5H@irE^(6kz**FjGFId8|5kThC}xnE4|9niPBivaXdL$M#gd{=0%Nkd9Q== z;ql>Z0d&y&(P)SUrh0DqPrq+*ZR7inu&bU6VRSFX{);KAxYl8|Yq5vF6?8S{0Qz0s|^!i@9q<%V&mwk@N5 zO9if=)X65@WSBQWoMV;yTABr+PH~aIZUd5JCv$`D!y?-@@))?cN~c(x>LoPwak#R^ zpMqAlOadJxAZz6d3fO)||EP&_RDQ~L!#mOyd)zEfQb@DtapWgWcv|!h6z+c`?Y$bp zz;>S5?#J?L?_5wtjuT*n0l;ic`wq6o8@mS3K;WiPF$5}i1mv8m z7^fsf=@p8Wp|_zYw}jTRfOL|L>muS%el{+6ZvQe?Y3nM}`(LvzK_>*5+FG^V;E^m7 z*8yxtF{9@cJ3~ES_W>13S{5L)<89V0+2)*poVj%6Zp5B{*{a=g&{|IMDyz6dG+Jn{ zeqr}g($)t0%IbS9e*K-7s%Ej_~{f+>=wRz zd);2Fvdmy(LB)CUhlj(DQ3~l9l~j*oMK^CP_V2v#+%oaMZ_qYHnY%HpqEw7s&;LKh z&apeQa7}};ZQEwYw(YzzyJOq7)3I&awr$(?^!Yqzt(kwZcim6bwXfCCxou;z&Qfw$ ziGAn-W_$s+kt`3E0`g(L2zdl#VY^x9`tO^7%`5NOi1{zdGDL*&#IMzmzWOci6nOCg zntl93G|98pP5qSTeYqMyowx&0lX=$(%gB(Lzkvx3Mnx{ zICkh)x*wbZKMk)#`4KPE_B1b9z z^pK5McxWseKS%8unNoa+><5ztWu))TVq$~*m9O8dmA!8ByyruTzE*o7hgzoR`-rHDz|Z`j$NZ+6j}iDJ@a>(c5^`Y(;{ar=@ek}ZL%xlGv?;E3*=HHq2A%3LV=q?7FwKVGWzT@3 zre_AM0+Z$3c@Q$*l#8*W29{J?U~XG$+bu82tw<3J21_#wNt{Ag;2TRuba{wPFFpGt zua9KuH)RoBm5cavexoAPA6^C!+ABK6oP81?%xk@_aE-dJv+ondSA}&7+y5E)n^XyF zl>h|D*Uyf90eX`h>OC~ZUt6QUKBK*~M|74%+lh~Tym4#v)Ep5LURq7Woxt6kiy=1= zgE7eB8tzF@1%80iTV1FzPu!M(w+dSLK>d{v&(@%HvSh;IxgqX2V?mw1M|&?(7<6-1 zsesEx)pn3vUb>v1-Ys$G)l&-QsbrO((4)C&H_Vij-3=;zD9!aXGxa3AMgNrm zq;dBux?3LH8|YR5f9MX*eq40S3*MdnR@aWzoxWTQ{ekLruz!$P;@_)kpzw?P?9TGi zupajil1P$esA~d)Z$zWYaM7wZBbzjoN6XDtPW{7jS13jrxm#jHcQ*uqQnm1ia_nMS z+s=2rSJSwV8HZJ*K*eFtR4p<37yTpVviXpa#lUi8IV1*_mDh9W4)&EfeVbLKk_s7P z1mv&u2)PX-njn%;Otqmu0(I)@R$+h&NQ*Ok2h$U39Z|kDOtyy$%?A2~UFs-CJ5)Tb zVA2mZEiSHLl%>KUS$5f5E?YkMhbZ9&B?UZrVP@2rxc!sH(8qMqX_D@*gIH)gt`wwOhRp-pIbpI9@vOKEA`y0^VZ)K^Q54TL=0Ch!_b8 z#r5Ik1U|cQ+Fho{gb}ME$!Wkp<2hsB)e>5VM46rLdMU%YBb#Hxnf+1G3F#LKOJ|zh zB`Br_FsAoFp`MxW!J$TvzN(Do@ZKxrvPfHWUza*zMKcIN38VVOMZ2k4v{8fkpc{Z_ zCzK6sVJAk*66QN4Lf~q^ZIve`)zZ%pzbohWlT}iD0om;NY^Z^K0sySbJEI><0fjoE)HmJ>+y^Pq1NAaB2s_d5K zjKK9@4C+4BkpAF*BDlJ&8#inbWqu)961>MlD$`~f^k*UYkQWbRvJK5jU25+pQPrM7 zu560Izwbm`wq;m+;cSClX1$fr(7fhqVEvfVBfD@GSyo^hW z31Z>aJXFh#xRIG&jMIQ{f_Av_$frLadmxl;=pNVeyTR;(SDQYOe75Jqy7O5!WC`^u;Y=wqirA9ek) zd`I;cs+4kiuBAOwk2#B)gE6g%y*0dVgr>$X9Db$bcMy-mpG=qK1VEm(Lzi3cs=af~ zhXW>-7KQ8Ws;oyPpt~*gjcsW|>YF1bs9*OTx&67X?ccB67E|@}xnY{8pBLj=>bQ1mGT&iO2iMlPL`=U&Ey58LzdFg~Sv%Oi zWokH4QUY_4?CIw&Ve|RtXji&GZdJiJ%72zK9{4N`nsQi%WT1{vU`@CqDjHt}o zKb1EUxWek^yuu0#t=ma(iITsj*;d0O4^Tl-RwkaqcP1R1NqsUKXd);4l)feu{3xij z_^{+rM)`9VcKqSNY->quB4@;fxnpWw;)T|q*NQWJ5p?>#h*ihD#*p8man zHtWix@IE?!n1+l-by2@2WiUBG`i1uUBzRW`VUtAJ)w6{@OPDr<1Ac(Mm~7hkx4mUBa^uMc9CwD?!j%Jyn+~j#x(`EQm8qOGCr z32OBAgV1Nj{^ogFs&+C4Kn2-R77~%fIQx+QP!FuNhs7#ZAoC1-k z>AlMT<1BtW@Z>OaYGpSmdq9kXpn1eHfAPgXN!svwq>R`Q{iawI?#SW>L@gZ*xG=BL z4#(25pHx~an$J(-w}hp@Fe#u+UwOVMr-2^UaFwZ-XyW^ zG>{DIF3LUe0^YbcJl!7IUc_(#LT!-PupLm!vyUb>#3bN_x^2V_DkY}=CTZ&PgLKSv z+rQY!qox8+k289*jL@OAirFEWBRBbD@!F@61Of3)a6@6@`V7VGdqM0)6pElk3ppxb zlbUyD*~?J6XRHfo2rm~cVE_BU5KRT=I;9%3_a60E<8O?dA-Mx-E%hgHNv@_i4SAo7|L+3myJ4=@N%K=^;- zO~3v(Z~Et({y%tA(Ldg_Iqb6KSd*NHhFR81NiKY9!-i8;t{RbV5m`?)M)LKkr)3L_ z(1-n(wHJ%fsB|a8`dj2$g2OLQXx*J!O$s!(=VQ}xrfci{iYeyMFv|C{XW~l8tCjop zN^pKdadrL*-D zon}o6pT2-jhJLZ5lOL(3)lZ&+e0~%Q_YcEeBo*ekID@0KF8#O9+q014ulji^eN9~> zM+z2g)B9UV346M%JjRkeSXqPnZ_K-uq4;HOFB+Ff(sL?QXB`)3N5mPa%>MSU1AvN6 zsV>@t?C9&;Ypzl9Y^yvYq7U!vZ0{a^WJ} zN*Y-fIkOe55v60i>IcSs%%l3ln@Cd7=;8T@9{PGRhiRO6jh2Lb zpI}rwPJh+Urb_h}*yKKr)Nl~mT}V>B72`HeWWqe%SE`}0_Rkynv8|m$ zU0{9byXs1k9$#gJ<6`Agt_H)FnfCm#4iIe_6e&5|-xfjZwMJ>Sw<2%~`SkZ@ZM}@s zWJXszJb$;%)HqCAwwo7s;^!GysCR|>T@PAwGTQygBaKfqVs6tggcq}rG*k==pXQysXVIt`Q zv2H;@NovD^4)(6tre;4Vy7*GG7d9!pm_w06U*S*#4+t*E+BMx7ZD2NN9m)Z;nGZP) zpMMk^8T)x7$pE$01sfiEnd^iSs8G}T+V$f3Xn$0jO`MXyKPvUeMx?cXW?S%02sq^4 zRD&V1ZAcx5TaMDM7DFNr3BgAgTT=i9;%_0Uw{v6l2RmblI?H_88b^~L`h3vPEw8Gr zjrt^VU4j2k9EW5BGuu_B+r!S(Q*my386_4CWen~4R)EaudBhfI<(D^A?x2Ekct--y z2!AwHb^Q9qiQ}lc0j)g3s1GPWj~;G479^*ZDMGHkTkL-181Fzw$7pb+k@}My*_V2Q zeX!zhSgA2|3$;DvC&f zl1QN2s2ziL{GwVsMnAw;1?(Mn&e5Ob3|TnF34!L+{0frf0dI-I+bMn5H!1!UT#*5!i#0j z1!e9_)@8HMiJava>(C1w)6wZHNqR}b|FFQ@SY*S<;S>!Uqg5s@E)Yb(kJ4tqd--U4 zQ)ht3Gsw18Hw(7EkvIi-(&D>MIf`=9kt(wxgi@tPspOHO{tSy5lMjh`FKiucmVr8I45T=g8u=cKEtd4V4x#edOIbz%C7!ey#U#fnKv(q|A9yQDmLST za=yWU;Q*jDsF$l)Ipx+RrVlA5wg6_y$dXWB{Dv6Xi}hW+1BDgVf<}`<#t@~w0!ln0 zXVL{9;-es$b=fw+rI{fqwp`sqRWwitr1cOJLHJY`6J8MpjZ0l!orGv2br$!H9v7U{5r5(2v#nnobD^t*9SI@1CU)#k7lwd{VGG%ACi*l)?Q8qAn*DCV9Z09Te{Aw3zqIc57T&-; zKBJa@thYGOD{NDMoZ~==;N{P+stH}Z#Jh^dlAur$<5CmG%=7pe_1Z7fR9mBKZ{FP% zQIOWIXrUi}M@-$@m!y38Ytcw-aE- z2vM(-n{6?` zteNoD3Wzh&H#s`3RYvfA6(Ui#51*Gk)acTHC*nid-SaN+wg6s7UdyE{+ByB+S=QK| zZDbu(a}4=)!Q?~+xChhhdm>*)cmSsWO|9rf6zOBiJm<`>|3I zPmy)|F@u|#mX1&Ismx!4qt+oP!>-PfpG7s;`T9wVaJrS2MtCBTI=={STPm@jzlAeQ_r3hFn-2O#|Q=DdKg{fe5kf5cNr))YH$k;8h z>$di9ri6afiw@p%f1fN$Z^M=}ms5xd1|2?`*exP2sGn1x{qdL4mdikft%~{&-Vc+|DFiD%*za4-JgC+E zBM|kT?c|BIR>>mndLa`a10}M__4$_}GIt~*Sg_af)yi?73B?>y(xj!}_i@FTO}`Fn zf|;0q^?@ceW?&G0EgI5QE7g~KKN=BaQVlK zTwB!-X*S$R_~)n9HfJ9jsi=4&^6M)^AZ^8F;Ad(gYQ|?8K#f9PembA)uyUcrl2h7>B?%-_YYvs8}-)JEnBLTRQcltZb)_O$Ne*Bv)X2pSqSiU-b# zMI_j&bo+Jdiv=p8R8F0J$t?wL@~?oYKsPVl9^R8A?(4}=vp^-3o;*#;)|}$Z!|EDL zHUW4ea>Cz~`z4VFC*2~s1ufocQ3~S$C`I*j>g2U6l@g#uX>EK*CY-q;rp%QjDn$2r&b`Z2-A=Ugx{Djp&|=9`e%WX>WDW;yt3eas zA2Kq-r+_bPqhdIAP;r%FI|mKdE`VbLXXX_w@~_Uje*HA^W5N`p z)4TkQ66)^j{QRr(!vNDGRPIzuhSk8kQ5{*A9wB^>02a;$v%lRRV zsuV?@>&}u+Ex_ro-fFioJ!Ia$3$F+Fl%GeYh~ZplG*vnixEh-wx%wyFoSbz_A%Xpz zo%OVQ%#L>ua{@Cd=CBz;`7GmS0;;VEi(UvLo-9yu;K7d3XQh@}Ku5$IwK`6!y;%<3 zG5#>>ylY}7$7isXtPf~~#x0Q&bx#csx33H*uUI170pgwRD-A;mIde7TuEhNrHdFXk zhOVsK4MjB*W%)hP@U^qBVb$1X16zuc+W|$1xNO9(qU?Kb#17h#b6s11BO14}EJro( zK|h3#zQI3Vn&|>>6x-D@#hx-8nWL-Ku0H!vigH-Woh6V@9(%={_H-VkXo;=r%KeA^ z2V!5RwEuzkYU3N1rWZMCro(>lS>z6!L!3+QI&r}@MEw%LUv0v!Wf+zl`f|{D@<5IS ze>s&HM-=t_9g3wSg_;&B5y(#;)u*HQi|X|X`kdNpE5+S~Z*X}^C`Ax}?p;oNTh)JrM7b^cy&XC= zn9=^=zCeb4v5^pDS3D4nyqtP+xhB z#n_dYrep`v1J>k{2Q-%H4DOdl4{J9t2puAEgoIamOt|g%LO`Iv4DBtF8x@!DIyPv* z%u(L744Mob5mmY%Fq@YKaVs%{cr*-f?Tm=fCizz8YsU5)XIJr6t}!`b8xY3Bae7<8 zeCxLb5Qq1;j9_@|=B;17gQQ+h8)C^CG@XefpdY7^7>ev}*Vg^^34~XDhU104jCZxE z`-#pde8j~)P^vJR5n-`WG!KRVUK=7-pErT6F1~&&F?3A69cc&NT+m8!myzX*NXi+t zd5uz;3qMExu^;31I6fgi7e*2|qCdCr>b7G4!^c>7C*r_`_sZ&m*jtN#T9+c)>KAC% zq!-2*mX*|-`yI+TuS^y8y%6?rJg7xm<=~SUqUCgn>4sZ~$aR$h zg-vahnZ}r$SXU2zx(*8%aoqb;7?{WyK$O#nYB=T9DnZ zIcuu(!Da$<6pMvap*XBvEE_OWNE}5c8%VeHZ z#$Q=vIl)o!n~jc%pfU!(xR`W)rG8~)=l%G-^iToGst{#lGN>Cs?n^lNrNH~}%K5?w zc6YMwm526VXF9?Y1$RK$Zr|g>w--L9MDONmkWr~rrmRXCiv0}SAS{)?yk_g z1r7B021*qnRP$mQ$c%(U!;UtFzcM&8e{2)J;@+bTWSGf#r%;M+9)%=}gZCwN~P zkB~4YI+^;_2bc()UR+7-ev!9ELzzdJ+ znprw#!S|3#?rh|{09saFd8@rALJ+KrN{UBQbFR3P;enKwd5GC8e@C~= z0;U%TMegHddhv2n5@i;esO(uHpoJTZ+x4eMq)|~j#{1Llwr97DmFU0mM2`Q)6WRY? z!s&k&QWLr!8a;O{y?3oENF|}3f@w~N8D)|rqh;qW(&8xa3^-QdVc`{lv{{~PIgEaK zhxJZs)3nrp+QENc)(s$=fl{HB*XwnArK!EUAxV4^4gY-Zym>J4YhHVA9yq=9{-o6) z%G5+s%> zAhtUImvyKsg5^q)Wb^BT^G$9e5ot_Hzg7CK&3t>jn-;Rd;a2)N%vx5nVj9s>2=T_K zV%*Em#@yXNP_edrN4~jv5|ryNJ`G(&3F|9uBCBl(_OyqfSA+2d*u^kxYxW#XPe>5DV!`+Q}2MSbZmDq8|+pBh;uALssTviuel#t?VS zdVY0%{po3-x+MgW(rw8YW`;W?;(%(idhEqN-ct;cxT2nKx(Xhpz8;iiCYQFSAlEWO z`oa^V;xfO;G|d(Pu9NgeN{_=4Kn$WHB6kno|HS0A9wk7pI%tZ*|J9sk34B8qu*4c| zlgpkl!}hWq==mz(+Xm#>!qkIRm09~v;5$GGZoWn1ifBg;X3Mh!2j{KiTnnQmk#heG z)+lM*$rtS$$W>MG(V17H+1H4wiDbEi}g%r>o6Ut!f9pawHT*5M@L9GEgm z3!)iv)oR};? zVlfGy^xPO>fUC=a(RiXoZo7F-5Y+%OFWjkV9&&2IY2pbCo+??KTOLssZl;o`9k4j$ z$i#ReEg(`JXm7uZ#Sg)Zgb4@N8u(2UeYI)VE%cfwI#?Br#w(@#m#&sEhA4lgm^#Ye zYeby)PUcH~*!~kt%#`(WjEl|=^t-&Uk6g?{7;;xq!5Y>@9U2D!I)rN57dfHke>JyB z-4W4mJEd21?$T6w(?R#=R4(RPWjl_kXup<)XF~n=;HLBsI z;lxW{DNGfMq>smK9}Yr9%5oI}UmWgsm((qj$e*oMn-e2UI<>VmTzniK)HuXnDm}#!Y2}~*wP;TPo$}u7sm{hY@b{U3kp3k-rAfPOy(hingLh>}GF@|@t`(oG#nvstuog{X{WAHjesM<5a3C--CxL(Y&8 zk>Mbs1gPm2%-24|k=}WJE9rehIP?g?PfG}OQGE-T%s-I|WJ9x+LPm1-z$ndS_p!J$ z+a0~2F&K4U5Vr|bz&Ue5yXa5cp|c;v4{{?})Zi9OsER0Om9En-vivGQG-U_gX<-B! z0e7m}gFv!WF!2zofr|wC#_!0aUR#{+!00LUMgY_(wn0%gG&txrpP>3?asr@kGSryY za-63IyMcc2_VLVY5IZZyfY}T|rPLoxG79lZMhQ&UpRb;`R7;w0C7$P$wjk3=;a-K+ z2?<@(8o_(B?lxS2QA1GwfmTed&6g+B&5T$Jj0Q*&xL_|eWz{MTQa(gO9r-`e(#hwK zLD4>bY1h~W!-)HU2pag#jB+Nso$fD4p#Z{P3Bruw^zBU4P(k^$f}-BgtAiiAW`m-u z=&uKN#T7zz=nru6|UI`NyGCD-dpTe5MlTt4HK&leSND7<(wj=>I^@PctM~X+S z094Ob1c~xa2uxk%cwU+d1jbVFzrWusfwe0B+B-n>kR$3%Wjo~2Z|MHC)+G{oOA${*WGfVHv>&^8-78oXS3}cSZ45;Q<1En&C-<_`i|sX5R@VFIdKVqFj&>+ zO>KRoc8tO0S5h6gsgQ%V1z93%g}4pYH~aE9is2E3fdg|#ioj0^wWWF$*Qg8}P~f6r zp~NhJlY-eN8IJ=qnQARud9~LT30Ar8ocrw15${Mt1;srzOpcciqo2IlCU3+^4w`8H zomr`BQnZjxVL}9Y!9D(m`OXiqr@WzjTv!C*9-UPOHc2nlxfI6^4!BFG8|_i8&%fFi z+xbY@8zKb0+SF2K^v}Ys_-o0q8 zn}Zk3?2!wi+0*3?+OzA7*5}*x7X+?XD_U~-P6XsWWD@!tT(o-{@|PqS75Gk+u5tgb z-L$f6=GH`=Y0JXRHp6RkAbl?t*V25plR>;wMm7hm09drc$46BV8rX4%xn^jyHkEg2 zV925aD>kOp0gGS>5|gwRAG6&8@l|h7x?@0K;JP}GjQzuGuXa3XYb)vMM5pFq5#yI# zn$a2MC>@jrqr)Kv6TN4qQZUG2KWBa4`Io!Wgr8qvT(oLt>!M4Tc5T8;Z%Hf?B^4W6 z;wvsVyQrTibmx2$2na~>eJXQd;=4KNkD!%`utaX&$iRIaJUK}0P|&_=!KzCLNH?Md zngW>p6lWn{N-x4fnGBq=faLFSI`nas+u{olRi&;zgsS|(H>~-Nk!C2@$_bV-zSczA zO-+vSRILUGVT8*cPA3P@vJ*+%5J)pc!x?jl5E}kV(=Coqym$FfI`9Oe>Un&h?rEJD zaH%YnX~tROWd`hdG)(5Ke5H6uGOL&-Sy#iBdxjnGO~6Zi9v0ko%3D9>{VtR1mz@N_ zeBi;i;&*C>C2-Ju=Blih$-&WkOO4Fe_r%i*_OjJ&h)j^K(F-VhFyWduE7B_I{*>r772n%+@7UJ>=ZerVY$d!2Pl|&^h;o9^{ zvoT4krr=-lKGfMIW;rA6GwEn* z%Ni>}=e&LWQYQPzuqw0Ym(TaF$6S_}^Elqu>B;2F@2bK|i?lv8)pVQInY%?Bu69#E z;exeqfW$d?B=wiZC7?dhsB=-xiZr84QE^`U3X2N5JybQeDIC#Jtz()+zg>kP5(^6g zsfNVu9_J4ZL4q}-z*BvS53o!bXz&K5bnL{Zya!Ns@M?G^56M`7?}#25@RQ^ph-kDD3aJ3UfIm;H zue*duQmv*V{^#hs5Fi_pWcJZM+{kc~qjF`VGKH{@Jv8H+`uE*V3aaf(*M(yk)+>kZ z99MPRF;)t+Sa^g0xf~SW-2lbEfav+ExLos`Vvgltgc?1^A7lNlCiHsrJ1t>25? z1}M<9-?Pm`=eX;1CES7e&iPEzh`sH*FeJK4RW@si64pG8_Q6mm+&W7w-9xRL%N*V_ z@-qsd-FGrK=83`JP)J`|fg&gDDjOt-X^x%|+yQsq8<2QJdHGofW>Ci|b!>uzqPaxG zA)ZLw77EJIfh!pr-U;BxSX2-pxWSJR4UrEIjz0)-`xYfg3(WM5Sx~kvBerBIY65A6&dE%pWKu4T2WU4E`uZNpo;Nblx&WIZ*;0;!4f zTIqhp(6jdNhi|-fXJPXsNU3E^jVO=3(*Id+UH5Ms{0%LkL9`9*Vh-9|B8&+YqN-3v zlHH0@FZaBV1yT>S3Vd5iJ&sR^&J~&IZ||=N&vii)q%5JxVpFlpi9_eRIY}Iy8=CaZ z^@ydh=XS)i>3!BXnU1J%=`r2>!KXuP+_zDn34|#+m92K_?5iD32Z`yZ5I$K!`uyz_>ImFrzSI! z#KCTX7M5LGgp6J+!A8H#aD_GXtUS@KVeB88`JYzeQu4or?!dEnjm;l&ku-X-j+;w110k z;w}>k+0!PShGCumecE!@T2YK2I@&!-pxd`@yYh}n^w3d$qZr&O2)~X~5vW7!)9J*V z1>(#R&JeI}!^`?0J=$Uw?fNcpp` zmE%T*J@gmh#oXHnMPDDC8Z3C!E`?I{SJ}9fG|lta^S?!<)26F+R%uUto2U8g7ar#6 zZ)wm|SLrl!@WJ|hn}xTCP^r@Wyl~H+?4JFDw@_<<+!J!4vo6NoL{~-mYb&u-ZWfvo zTM1xZ+!Q0-D2KAcX1k%q+%}JK9TL3kHKz>TEeC_DRi?!;U6_VO1YBx=UI0R`%k}`R zj~Kwd|9JfV$%!XkFVBhRG4tGIllA3q8G6-PSBW(Q!OtILpTaJu*RM?`-GR!r`)8#!f(ujzXt22B4@Ek=wjR0gUwd}}N-)*Q{#FIc3U zX$(I1rDT)pw%dtZq10KtZTLQQnETU zHHPaEXs@ZCG3bZnj<)Xy3`iKGCpRsgqOmD_T&gflveKV-4N znvb7K+Vz5$m1DHq&(L@($05k}UTd9#GT;r*NXYxc`*ai#zJchIwGa-`9;DXqn8?N* z6Ux-C7SC6wfJ+`29Yr(4+)DPNRqar*kJ1)+@q;b;l-q@=|6 z4jY6HEX>HanmSw6&z^;U=>s>6^sQ>pp0!g%qT8|}XHSY)GT?#$DPD&S%xWX>wwJAWFE2!fbYvJXZjni`SFn0&HPl2u0v7ai6pwI*SkyGIS>e zfZ-agL-%60^Ao_eS0+1O1eDfJ7Z~95!pqx>ps`jXQIat&u$%3w>zd%bX%ab%uK<%1 z727J13`wEHM@^-#ZznaHUOM!81-40dTzT~=#XBJ8OOlul#8}-&9|E5+#od@tN#8Ad zLkMaiQH0+M;cA7AFugB-;Ds^j%uW-aZZkr*m-5+==Zn0zu>`LUhY}g<=eb^eI?D~s z5!CVF%XTf+h@fm6gMN#s7Zb#64LNq-mv!8GFQtb+*JsJ?^{*m!c{&0k%C)40!h1P~ zvGwCbV25O&mzo<>{sA9dE?|ZOS!Md&`DG!?jpE(g7uaGplO2Dr1&nkt~% zc+7}VT7KIwwa4V6yZ#x?Zop_PVtsFf#p$utMzYKPhUCG^*UevIs=_#K3VLm7i#&dM zUyEwUr^cuya61IX7ck7Yzt&3J=`gOnvdzd%!vZC^Zc?_61zwOGC3?8M`9p@P47c-Y zc=H3y9tyDeZ@7{3f8$0r&R?wmw=R42pXZDJ!Hv@kgW`{tEgmCaO8TV%HjQKx_tGhJ z^A&5!l)ri@MElPn)`6(P->*DZQ?B1@JNytdG7SHidy2$>@IWLv6hzRR@Mdt#xGX%Ht#+ke(wTvX;Lh?&!R?X8IT{D z*?Q_d))8vTXf0B63tP%4l8wl2oz)D989oi$c}I>hzhHsDfr`qsIH6I~YQKZT3Rb;n zijJPkl%)$9Do?|MRJhBS)E-N`GHLTDIuO}4u1fAT0tBK{B2#4)7`4C;LP9NoRvU8| z>Jr8e^iK8x?1;qRA8m6Pj5gAbd-yfCGl0%Mj__l8KtTgrAV)X-M!g>V4M;H~Lnar{gMHXm$9qV6J*$u3DtWlt z49SuCCBX&X{iND`#o6Hc)&F!)G_PB_R@d~azynJ;gI}*Ul-XnhO&B}$%#iggczL=D zjuRANK0^ALk&z~?Df?1&!wDI5ZmRBLx=Y@ylT&hXvBYGr4C&cdb%4}G&3l}GOTFDz z6(sypMQ*ALRfSwMb>;2lc~a&Hd|zx6h4PGNY(iT6aAC5xzZI0M*(xkoT38G<0YL-w)_ z-)uJ=+zbX|VY6fZz`Q(mcr>;42iJ!xsPoQ|8oYz8*@OK|CPxp6r!fqyD{K6sP zuyUab8EY&IW}^Fs(v2ICy6%j_9W=_nx~H1Hz(g(*%+9O zp(HIJV#Z=6l{%2%vN5m9^ERTQZ%ZvJF7~z~>j@bQ-<*XN%`|d#pwy(s93yBgHr}*nm`zMA0A&+0 zP7mOVAqW3V)iXT8qFowbFE?Npe19dpfuUM1x8NGdSg)RoUk?V$9XO`*@yjVjC!=s#;2WS$w(7zmU{4JWmPc8CcQUoSlfP<{lbRzdr%t+ zDYA)1KH41dgh$VwI2(}wZlc#@8rWl*sTanbs-^_^$ok;Ep);*1IaL}zEv*5MML#<|}nsJoY_G-+1Vc&-xrt;tE$o$K4MEG#;{zJeme!W~e`UBn{8FH8% z;+e3EsI~;ZoaKI3mN9F;S`ZDJzx8SN&{W{SYa9vMm5qV()R=ldSJJJghh!V$Ll2yH zb?zb+hJlEuD-oerWRx?IrWovbFGhTXI*g*%v*idS{fM0wq4m%a)}ldO7r5-{9gU2e zflB9U8pxs><{Hptq+w$nqGIvF;Ge{(V({3O?+v3@pd&z&?XIhqt;h1(`>@tkmEggiXp3EkU3wv{ELDBzNBvYz#+$@13S)QmiH5cCNzP)^>YQiN29ErbjLn=-~^CcuejbQO6@)?GZ3Dzl* z%6f_LHEDebPR{_FT5bX3Vrx2tgiKLysx`(dmz=R=D#>m{dm{diR`2@V!aX?Bsj}2o zmgK2yGC^1xD+qNFz4O*%R1sudT|(mUFy-{dT@91(Y=IKB74*Od{fE7AZ8huyvaS1Xr7UI(K9f(X_i?e*}$pEAGR;^UpxL!|vw zdEIiv{gpQi2x{X_VS&)>klFUGG;FjpKx;STH|uilLPPn!nn4~$5H zBr?1$%{IhwA+N6gVeA~cGl8}>8{4*R`;BcU6LvlG(%Ml`Z~QvQyEWXV(OsFCJKXiiEoNO9=fCQBdBd0| zPxAi_iUJw($y^G}hTe5s4I>LovXCx=N9KHk?s}h6H3iy9Tqrj$+eiR$*a#d*>fpYr zBq4<{L5uOA4m~nBrt1fZenrTLILd_vm zRX*+`sIOok$qbqeb8$yn5C%8LT)0YW66LS>)Y%jVdC0R(Xfv*%Fy(`Oy?azd?6UM{ znF?=OmjHnEmI<|ItsY{*b_UmVgWVtj5X#M2c3OW!m!(eaczd+^G1+FC`d~A3<&0n3 zNAw?(soZ*UtQ{P?gm_c1C3p;*ba?wJB_@xWj}7JpQ&{rF8UNfpYy}<- zl)e@Dqaqu08+plb0TSy(fBs98)-4w=`VoeM;9f*C^$wV^ZCJWK(FORvwr=oIW!p9! zYBwc-9Gw}w(lFcLjWThV4YAG6?4$b<=+KxwnancF#dN&}35#rb#?(dVZn4-(fC(1gEzeb}k*BG;iFIRSvt8mh)iuB3WjOv^M1^hl zILzg5SGDDQgd`=?AzcBg^KJYF|KT-SUSl0i$?1!j(o>4B`+N!c8-E%QV$7W^mUeaD zWW1>9R8V1H&d?T29RuuUz!rSg?^~~v%+J!!z2z8R7~)vpf{#2Q24*rv*A9f2N&hJA z=Owds_WOfAPv+?>muffH7#^b@P9whO1M?UFHX#k*y5fq1;{3$|H1{Z!_utT4Oef?= zpbk))>RKRuzf|tzALwY6>dDYr?u*kKI|l{I4+?d78+M;UPyS9&z0T^hdh0R|jQ6vC z=yYzdAQ6t7iqL*CiuJfr_wB6^*_=Jv(a2O%nI-#Gnm3MRG_?BfH816}@-y`%7hO3p z^2Ke!Qv+#3XAW)5tjt?zY;?XS%`q>=f_@1vb{nY?$9e0#+}sAf_s_aa1Q9I(yw8(bSAY{PX5DM`^>Sa?#vw{<*2~ zJ1Y=E&PGl!p#$kIxB~@a^y(>CYpn%+Ox!S7omP*<7TF@u8|@d8#FSPZ@B!Wy`N@XG zWfA{6f71*t%x?rt<_W(IcZc-K74P!a_mT22qRBtQZ%-H|{nhe-qGjlbmkHS7`g&JP z@_1S48Rmy<2mvOFJxUM;L$2Y+eF4OOS$_v1lnp+SQboGPom`B$wB!g4G4W&K;kSKi zKYsZiM3WeowvXFnqT2Cbakzh7T;GT1$eUCP{5gxQl0>lVSq?V}r}n(Yg!%MDRj^x# z+q>#JwG<-EWV?K{WM0;1>{u8J_3kU%#hX0YblTY^!b<|&r3Kp4eEVL%A9vbOjn==< zy&6T}^nLc{)p%FWV0Qkd89A z<IMRzh@D%xNB!DF%ClQ~RO-rG5cv&U4Ty1$>GkK51eY#C8Y>lcd9)L^?*6dNg9 z2CcTM85P!d(R`kd`Ddpl{rOPsS93tAEtPaq2K47F@*WCH9~XXtzzl}~^Z$*casN+} z#?H#Z^8Y4jRoZetO*!%1=T{5U{w%da=G28K-y?n__JbtJZ{14ozb*NAw3w!r#e{+MZO*HMKIs{~o#^Yc;5DB;Pkzps>C% zMfFfhX4c9NypY_DQ*P#th#L1vWkyk;ur#LWj}F}*&o~@}M;ejmnKGb1uFhkiUbKvr ztk-LCE|P|`q=Z}H`BdI50BqI^mMG}U)zAMUS>Vsa77#C=%%<=?j@j3 z22KxicJ=~R(}ggg%)HBsUAnE=)R}R-c~FO{ z@tWA2GN%}s&u~@tJDECkpJ+FK%4o;iyJK7+i8kK4AJs=c<=V9|RCL~78-?9O663rB z%OP7s*OSDD3Q}(SwHPZG8ri33(iNUPzPiZAJ;tQ1g>Mc4!X0Cs3FCrWpP)q>QvzV* z3klrtsZCn`TK~d&BqKYSuZ!8iqd(7j+JHejx&5RYo9!8VzkfMqqCGI5%>4@|q(zzS znd@x{m5u?zs*fyZ`OtbADbHr|xLz0S%U|9o+fEp^_aGT5vr=y1>!qZ5kMuiUJ|bT; zbXqY@tYvVM=DDFD1x9^>w0y>O-Wj+0BP_}41uMxLI3ahbYr>0mdKkuzEa;anjewg! zvZh7KrHnObRJ^7(+~R7KJAGQ|#AEXiKK$lU-c^r3;E{Y|~nUmE5vS>31^Wdns5 zxRNKAf1<}JJ$S*I@RQ7*u_C`Ep%mgvpA)fu{W`*5URC;J+K+t-qa>5aGh~Ox3eBgu z3;I!fE%0#zrlc5@)MgY$@TCu0rl@UD^9!>cCSE{RxX%KkL&#T6xEci5tQW-=(-a+n z;;ew;%%hx?K|d!iMXrIMKyLS^ zW9wKdY|mgbKy{u2(}sxhQvq9Y5@K06L|E3oC;QnM_}Hmvv`iQ+^|W46w3Vd916Ohm zzekUxEdm#wE$8ogryR1uft>$V2aZw36Kv8Aq6ReA@qLa;TiYWYy+Hch z27(@`XO95(iGBqOu5%RI@)hnukNJN>JqsW7K~bdFp(1g}QMR&P=F0`zt-)N+jmE51 zKxq9G-eumqTfa4dJI1`C(HoAWxKPO@44k@i%*5ZzwPwbBL9sPGMTy@T|2)9x?pWQTJ9z}h9bJME4X;>R)w#;| zEyi&3S@!imDi#TKB#e@c5xboMyYWYl#;iS2o8-ACnXWjF(Wle%)B>jXl_OGg6S%EC zlsLG5`kXMQ@;f)d^C8AtsYC=w`nT_q;-5eBD=TcaI0EK9U<9k|vp!+^h$l833a4^$ zq>hOg`Fdl?_g<)CUuJi4$JUYAO!ur>f^0zCP~yQY9y{_cVFv%?AIhPzo`ABvK2J64 zB=s<#_|X@I8=?sWnxHaHaf0K4icv3$Qn{}*gYSf%w?bH`&#nTR^#8elQ4;%uhV5@9-bPGCjbJNwHz*lI_xHv(<<@iN!|>pg za{#Y?L$#PRG8f)I(VPg_xiG3>q2d1icx44pdUHWD^wji*VD z^;7%bO1G(}ICUmq1S! zN2Z29*$auX0Zjqu)!sxfcqi6L7)e?zy;AY+zu&y**S1)8tVHbkw|&%?6Tv*R+#%!z zK83kRv|m>-lV$dm%~Lbba1nyD5)4Que98i`i}~=(;jFHJ+ymB--pAMVq$k@&Y_vu`@V@421E1A#nF$K6jX^zXVh}8H7J4+jaulUpPN(2K7_7*4J>uQlH zLpFNHQMA;*C7BDn3>W%7k!}ng`oje0GrVZ!%dUDINDKZ9_Ifb<2*&}z;V!Z{{YEPOv-xBy zp|9a(hz?aOc=Gyh?6^IFoqC>b2B9EAdYSsV%~{x?@SIi@4+@cYx-G>|yxUGOzwVx} zBb8Aw2qDJpQFwbGCN2U)0B1?L4fAFq5`AgN)ch~|SDQ*igs14DX>yZ-vswP z-R)lYQ-CmF(%Jb(W;)?j*9y}qWV$eQSC@r+gJ`bJgN%Gzc>xROH}q^#5q4*mO#alA zc*Qv~K>Sq$D6FS?Fdu8F4#|Y>=OnsD_KF6rw z7J{4fRew;WwjMt&7O{oui-CIhrYcm9J(`SbtW_0=tUCysK9P{(o5UK~W-=%c3jghQ zOBy!zG=_@zcxuHnLAqID7xgusC>ODWV8%P0)_9P4Nf5uP5#7Ug!z~MvL1!k`xycT+T)c}Yc54ZfL{TE8yZTKO#p}Dw64fwrsi^hQK9O>m ztWCzY%ijnj=4~_z+`}KUenMAcei>^Z1#juKWy#low?CNNQ5covw9-Gl*jibjWVqmY zCrrl-8%G)>AGcqRD0EI@SMEGIXZ6&HD~MMlKhTdM!Ok&#vTUCNcEScSAV3B@29&R- zdj9$uQ#J5X8+V0I6)uJIxsZQ=nnxOnddyz89k+MqTSiEpXS}dAWq~kgo8P_iy7PQL zIwO4KekGPvoX+<Ox|p$c%~PfY)T#pSZ@~wy5d!Y>cY8h)8>Pe{K}4xHy4#&Lu`SA zYI^Lo44<4cOagOPE~)@$;f}Wdl|)w?M|E zUojt2A`S`C(Zt%vAl6!*-8oYGde0Ux(MdJ~$RaLjK<(e6zPe?E>s*{#PTlIPbvO>SXI46nK2xhx$x<0u zmCnGxH;wncWF-8FwmV7RJbMJ`S}!01B<>Z3IZ@ZYf2F7iY-iYl*+pDl`!wPaYqUTS zEYz9RioIW9Bjmo;e3jvbt?GSlpHmx)LV-jS3`=w4v~JiR)ReQwKsRGvfvK9*2d*Fd z5gq6wqD%av3kH!;i;ToguCT>ppUii9xzI@rwez&P&HPSr^Il#(KxJaQ-u^F&`w_;w z{U;N)hlZ?rTwQybXp3cFjd@;R7!_j@4_AOSpOsjEbvU>4oR1ae@0{^c>`DdRW3s~) zrrnAt|6wZDqQ>7O7EC&iFYV~95v0v9oLr`#%Eq5KrfE-Cmg=4Ergq`=EekOe#jWBO zLcj6x9s(%O`P@5vt@v4?i~7YEbcg2;cgeU^&@Z&e@iV&ZcUOuQtMm7Nx4(PhzD9sM zvLwQI;=P%L=S=D=d3DgWjgefK78)MV$cI=u%4@m@2d5@jBi?u6k!UFw%Ek{Rdwem| zP}KFPrXk+~n}!I54svT&#em_T61<^(*=}1T=R@KTzz|iph=rYyQJTP+M=9UVP<@09 zh=`u{%znZ%wDhTcPmO$HYmQfD5uzY&ff}+SacQo6Q8~Po` zV+^=-Oxd}5i!8WXcZ(pv7Z%^cF!vgv^AB>Y&OuN{Q(f(VRP5n6P__ST1Bb|sf}N{6ksIr zCOQA!vL^%gwC9i|-#zGpW!tPtY5#i-a>N(jr`~h6wlgN7R3e1-eoNnAu2j%s*ObG( z9nm~8Phni^y*V5pxf&u+%<-f|^Z>@O)2?XGp6Wn`- zylA)S9H~n0dF_t=zLsG{Sh*@U{Uv9+_8D@A_L4uK-O=EmTGBXZq+B7tcS?$LAw*g@-4{+I8;j$e0gt1`GGyhDNkfPC z-Bh_*rif$x*r4_ z21@LxVtY>kaN61I=wPUY30}t=lB=Zt@1^`G2WVcb0POwDZ*M5pWRxr#i`u{OAxdwR z_~K2g#+f7a#7{x1JHQu8Oi(Ee-oG6XBM-7j`B!5*R6@d=R=gy zW18sAcE#=9DL5p(N0L`#ICl9$alvbRhigGD^?#U32iJ7-ODiSGK6gi+0MX_-Eg&;7 zQd1u#GSWc1GX4Dv&Y0k`=~@5wK{GtM4R6%=i2Qp=V|X*TCCzpM9zSyT;^HFTd%RsT z7&dovr+ditMkwOQK#ef?qXr|a(V+mtD-{`a6tV=*T7hh&LB0+!>IkU0Ms zGD%8qtq)lN;_^iJ7F4}Gxb;**;C9E9T=U30sH^-BP`Wjx-(hV#u<}Ce&38E-ioq>L z>is$!@A*1e)I%sQ@=(Ah$1_b}0AlrBZ5rUl<2V*K??ftKTvK+e#cvX|@9hM0!TyN8 z1Mr+;;kbBz^-YaY7ZX^%{PDJWl z(ew4#ZX{$OE(oqt>D_c0V0^J=&o-sd@<>e&{O!m#0Ws=WJc+aGk6!nY?)keEOKGq# ze~>WuTiSP*ajh_bvmVm<@K5T@6-T;yg$kb9uapLs61ljhg>7RxJB1FCl%ww!Ay*T4 zCpqZ5t{#Z;BoyxNvDl~bf`?q;$Iu@e2loK(>Hh*nxtRVZ6#dcX{(rP1{uhdJ#dp7F zFHHAISY(G0cSEuKsb{?!l*(*1wzIpY_>)VnWfY3CRA}<|J@krkb%Njt@xt;Pl!!Xm z<=N#qAy5nm%wPCFP?T7Ja3qc|<~HY>R~vt5B(fmqFIf}8yFkOz?djuOq=FW{xl(fJ z!OPoI*mmrQB}SQSnQ@9qT=Ub5hcnrW6KR9u+ zxF&{WzP$!2%5kw@>C_IauOpA^fu}kLg=#*vt;46*CQZ_Zx2pf4qA~#UJV*{Ei*Gug z%>O8KRqE1YPesZ~*B!Mv5pRB|=mCj5hup8Z*8ft``~OhU+4kw`zuf}!OntWi$(u*y zN-m@SQqkM9(P55A<;^Bl35~zwzXDVZ6T-3oFBMgs&KWYt$9?=ew%A`-=d|}}LbDq> zF4~qYE4Q~J6iMh>YnITLS%ylfbRs!ea@B%-O|if=&SIL0J`p~iuNl6{C$Q9OeahwxgonLU%QEUVZL%TU|!gbh(H^ z|E&ihNsYR&0R16v6JMBLM#iT%Me|blyYfBG*lS?u@(K=9-I;Y`}{Sd20%Nsh2r$RF&ukvxm}>VajBvr@`G! ztM%fVusFYyzj-usd!Tg5;-Hjx-rf$6%H^P@x?0AUY)Cfdbf^ zKeB>@6k<4lx3)nPLhX}#`KHxqssZUb7d{=n{nFV?S=Wy!1_AUoFs1uR0gsD*Rk796 zDBY;zaM$AazzZoN3f=?=DF*ntth=;8z=B5y13>aWdCMO1!9c~><&jXMTf^)HWu;%L zty2y=ubnHCH{j(%eq0PSA%;w~C)<2IV;LX6xnbnTO=|au22dMwqiTOC^i+(?xnU4S zIM$gg`ro$-X@KU)G?r$VMl0kWafq_(U;rZn32@Lh0+XASUM$9ez#L3j=PQXd0_&JF z(?IugzW4W3Auqzqry%p)fnJ=KuRbBxoP~>LG*VM>Oq~nb;nW)&uZrn#Jj$t6iYG~# zE9HZLhIJCdqZ9av1uzAaPAd*L7~fo`ZwQsMQgdvSULwiA**Y;Q$G5m^U*oX&S_zIg zMfQx<0Y$~W%D}sgZ^wn50W-U4((p68jln5mc9#IPm4fs0XvY|LMpZ)ETtu`wtAP?vFXgPKT`QDBGp+(kj*29{(R;y%%rHn$- z?LIsP?Lb$C3Ha&Z`}#lSYNx)x3U+UfLg3AXn8~O?h!R6843z4fNX01-tVh?kr%^rAL-=;Y3baOv|QmJpL$@8rz zwgl#{^D(swhbZ;yV704e$!o`*4Eh03jq(kcEg4USLgUt`FAUYd+@Z;V=zPdpQObI# zfjkHTYxoi_Q9)$MPIxe>DRN6+4xO|FSHN?0ry1z?PeqtD@qw;FMo4x}XFhkQ_&{sh zf{BCWenCZpO8)FBGpo?tnAkQ;iu%PUz4POX>(pMk4Jy2~3ki7+22pWtTmAAdqquL= zWTer6U)2%Z2tv3Rc6Tt6#J$6S2BTg&ymKRmIJ*`;IBoqUbWa1!1V-}x9qvXq-DO!8 zf3FQWx>8ddQYoF3P1ouBHETTTMzASaH6;90Z`=#yPsrCmuu6Qjj83U|?)n*4Var_7 zUdT};)XQ%y`y11n2M&2x)~+ zkY6}N&SuCNa%vu5R)sgnkU&us2ciO@Jq~vb9eKTMJ}mM8LKA-g3Hr-FBrmqO2s^B* z<(OEBfcMD{(zkOZ@pGp{nvxtCxYU*SV`4pAjEgq<>v1=(CDC0ETU-xc)E00sxul2T z_sl3)v`xSg`5-trkc@Y-Uj+gkJN~>b0hqWJ5TMd;z>taFfpV&!l&kx;-J*Apx)CJO zUC>C%#|r4qFaJs*(7vj#FgKJ$`~pXiD};ox*If=cYxahiZT5v%NcaG(Hl82l`I7m? z8#YfWWCqH@VD{~=Ul%KKP}0DmJKNHr_65|r6okba1EU>m1O++FeBensS@kNeo?lb! zkr7#jv0Jz?vvm7#LLzkCO62GaFrodCQ)?#o^!&Cj5WEaRNn8XGk2NqZ{@C^&W&GvEtv&=0oWzR`mTPB48#UrL(mMM zG!JYC0W9a)vAI^YHqgd;Am`r6YlXj^0!?P&tqA)l1H_+>{UNl9_?l7rojrlDS=?_` z7Pnksd^A`u1Ldy=fhnDuB__ctF%Cc8nS}Qota_oKid88HOM*%#P#``GrQIUvVBb(+ zAq1tAvYgco;PK&qd*e0|XcJ#`GE$=lTRyUfL?|)oYB>Z_Wl{FAzzxcyQ(7Pu?xGYku$QoRyKFI?icv_W>Gzl!EwQL_`k_J2KH*i0w<1 zz*I{5l|&qM`OZx>L?8OxvT=~fP)2F~1SNyUK}AR=6VrZ>5YA5LP$AF!LJ?mF$9~G1 z4<$?+Uda<(wDa;S%X#t@H!8vpXTco3tSGZj*=22Ep4y-af6t+KFyb2gd7wyGaHei5 z-y^4_{jdXXY#x7!uhp)KT;t=Pk4dK=NaB&>O#kyQxnj(15aB5Fq)Y5=LkMeqOa-i`-Sm!yihi|+XB3@_Jcm2WX@gIQeZ@B&F$w9*AC#Me`GTGQ zCZABa6m$ujA$JV!Bjv6q=XigbMLU;P(0!8KICJxH&;m1WL87B#cc)c zusEMz%yyOXn!C;s8Nz6o@9r|Sk#1$Elc32syw5|!*Infzy?!&f0i%EiVV=rC9x;gi zHydR?rtP29ep%=j!FYdKwY+t|^CDY8fBxscTE8}O1#+nb zD(3bY9Oqwd(!s;wGxB6$b{Y+)b3>>1x{s2@^kOlIG2|kTL9zIzqhpV3o)r6yxc<}+ zF^K_2>rP~x=KKD`pJBqPR1b=*2h0s4MT zRo~ICvoyl=xjHFcEa|a=HoHh~VE5PIm?zV)#?duy`g(fasbBU!K3{xJlzVXYDSHqQ zdv;gwFEEQLu8{pE)XdmEB05JtY&*0++{i=i#R5!EiGkaIbRcU>Rd*{lcM)mTlOz^G zI_IFWg138>q>RUPqUl;SL$I|sO2f+=tmTb(8v&*q&6@~(=%m0}v87COoBiv&!=V~n zUb}*LB_i#NRakOu+c~$s1Rcm8CWj08S<(nAky>hAzvJ9*p3Y)mQx_5?gzCMk#+y{^ z&sHo}{SK)g0P;0+JC-`coBT5wx}A&AIrZgJ<*plG*1zjHVyY6R^K%YXgTDndaITA| zhydH?u2OC0RUfl}s-Z;amiOR;t3uACp2j@l z8&iKHHsVXc`a_Yg=gG7Ofmy%{;S(2~?lqo3U}7qRIZN z6$?GrUxdvYUJ~h`SKqt=>)crzd#_+jIR4tFUTT)Z%Jr@sy9UMJujJSii*BcPt@!Ga z2oNt^d0$~0U3b;Us0E1*cgKEy^$1>Mx%1A2{TX*E#%5Z4Jw?(>72Ln-BvC}DcB*02 z!+}2J$inX;RQaH}pmJOU$>}SocFeA!NE#wl{pNpDHP8Ywoe zA0EK2D@?55P(v~M^$z8N!`X$IPNAi>Rq$y=f|Qobu%ca#2P+HSEW;z5t@F`B{eKfn3!~`bqjVkN< z-=o680aa>~-Ou_l4;6#+t`w%V?{QvTL~6<)i9)!aE8&i|)+fiEXY3 zi(WKeZ>YrewC5j%X>5wdzTk9rMZ_ik%+YfYu~fEWN;6{O=G@`#;9C0*;DpnJ2T*Y%3sPfj zl#ccc=z%j(A2kBRpqVKRN0g8FJX^ZP<6YzgW}Aa+RUD~I>_#CFg+_@^3oQNX}S+aSo%bQ1EC&)H-X52 z3&j78*D?Q3UdPVD#`gbn#3*C8AsNrlklTR}iBIRWUKt!=#Gl4i3n`Us<4)Ehj^a^V z0?<@cQ$-(96JxI!*L&S&s7(P}2%HQIlO~vqCGtgMzvPmn&5VXIc89@zJH}Gn5e@n7 zUYoR~5KKP#tR*}=5q+f5{il8~rn-q|G+~3<)!wsb1aRO#gPKI^9NWe>w@k3X@I3hX z_yP3YU4q#V`P(TWqaFDDo!Kgm_zHp`8IL+PB~yI zMNtJ4Jw%USj#`nF0%@F8ST;%T%^j(40NV8|5GfEnto7SLmp^KlEe`Xs{ompgDPP8- zzs}&ThzfnF?te1D9;<*-Rcx)5waYESCpI|y8*s807ew{UTE$8^9iQ5bhW})O*Eu7* zGRzO@gvx8B%pnf{n+YC&HNS0;uX>%Zur{1kHQT$aDX0s=PnV4}V+qKar=pBlvaoLS zN7VXYyQ3JppNudnbo5FJpINV|r8UO)QrvU?R3N3zYeH#gE@|9zJIHd&{9^5t&dW?vMG$+Cz=`yl)lk_y4-`kn|Apy(F}(+jU0!~mdT&ydgY{=N{_>V(p>XKYj8sjbcl@hE4D!)u!T8beP z_0Z{>C1+{?=?=|EkkCiw^dWn3vhshjMVkr49@cEd(X_x<9oWd24^{L3pp*L*XBrdI6GUJ$$an?;(+VE5 z$(0OtCuT?ShdQt@R+m2(uKL}&@y*G45wj}id){;1S)k-%gbKk9H1cY_>;e$1(AFYx znkc{{6>dPSd>E!@rl)VD;>UH_4?VKLGVb?2jt)Shi46>mxtlGsV|Mu-SIG^{ax_f= zlW>t5{Ib!thMu*z5SW?8X{#Le;x`1nNZ_xiT3*l-jo7=zT=eN4N(P`3R^ROe(UvQ< zsr8nn&{q8_?NgU$CVf_okCaSw~;36IyX(Hail6V_nkPpV;baT^2aF1t~TnP z_iLMm_h=d;rXQz8Hi8nU(LM}5^{&yyr5XR_sjASZm7`wxIC;7lSpcILS$O!jv!V5m z@9~!%G~-Q?kbo@^tVuH~E#a1Cpqg7Qly!yQc~w<=c2zb5%^eYePAkfHe>3##+@Fib zE#X7n_qQzdQ~zx;$bv;hL9SBMiog4 z`SCek8i2vbx-h>=4@0hr<_5bbd?CmeWH0BZ^14N63mT=J!dE<}^98^z11+gQ$SF5P zj|^wC^DFPr!5SK77KSf@&YoqfbsC>LD>00gjmm>YNj~ne8ZlH;gaUOvF3L_SFisMO zOs0$MdyAjt4^&3uBQA( z=D6Ap?G3)`RjM?VLOw1T7po7okmQh~6wuU!%3i>V4KlpBHW3(IACVtEa5uwBE*bJO zA0Z5&7dnfP@aGjP7E-?uFkcn4w!3+MIXGcu9C4X~?KoRN!M-{q64#BfBp+955)Lbo zY4)2Nj06$@Bd|`m%q1bCt-=eHI!~j18hPY_?fmNob2~1kdN7_RE_UEB6GoWAp^M@ZKGi?9@$2!75GLwRm3#9WEJSgi%X<>rP(O5qO z<;VWFJr!h>v}nl;S@^5=c%(RY3q2ljfNtk+LFOoPCf6jPXsYzh;#YWt+PR#{rcEFM zwoHA8yN=!Xp#c{VrBT%aBN7Oap;o(#%JsQTgRtac)B}-$DoPEo?A2`*WjopvK~ma3 zn%v*IiGskaUQlL8;=Dd?nm}Yj;zJD%Q9(&Er0*Uw#DU0qzGyg_{th17ZhK(VgbY-< z&@f^5)W0FoW}newp{OsSh`LMs$B>Tv%PdnFt4}-%!|WId%QXrmoZ^pZxq9D?`YjIN z`eF}&E=V)~B;Rv$;#czV#d48BwOSL2TCUq!z0ot&Ob0gjZTh067&!Zh>{l3dRWn(T zw32kXcD>HM;-r4J6h)(h-j99m=3H0#s<$9MO zVn9Xv4jnZXGMHhKI{QULlrU%2mA;!Yt>fF%@BwHE3Pi!*tq`x@_>(eeYz|2E{YOBv z@b_JXG6p&l91(LwBLwqYx0WVjoc<7j5(Z#k3J+M`r+X`T;y@4&x^l7L+5m2Nzm~`#U(y=Ei}Cryqt)QqNmllgp*fiPaYU@KnxIu+ruU)al<7qquk<+OH3f#*0BSVkm!6(Ba?a zqAAQ(D(Zov1W>UmtdpG+Ycro9frWRlAY!8>6q$@k(kV}x%)-tVWkBZk<1wtN#$j~P z(sHeI%yIqGT2u5ON)ol$4$y4xl&rH_vyrUz(`pQ)24jU@aG!$dZ$sWsbu{o14MbEo zwr49};KOAl|CR&-djzK^0gBosqkBJb2O3-=#+mJR4h?DDdIB(PgxH~YRlO^`Kl>58 zX%^jq2G#-G+8N-IP(+Ix>;@E5o`}vz%5xyH8%yC0QBPYybB6aMaNQDuwsY4KRsV$? z3$8^K9PCbYrvJqrgwm|2UJ+};0mFkTLahzu7y)L8QxQ}@J|N%bFy#jKFp?9jb6f{% z=!Zk>x$AY->ZrmF4+CN6lf*GXCt&&slGWoW+>ZP0ox@fA9-kw@OMsgrIl&$*Q3i}9 z>ah!wj3nU5`qc04kW%uZjZA?NhEWT$wMOfcBd=E!cL$~##R7yXcSy=`3%ze-aJJ|9 zr|0}ugg^2&)cmMKCeOUsH%aabHP^U!E*q9AwV1aPuw2+@{J?V!_u3Z(`BU@qlSv_v!WC37t()}3FV+`gJM z=_1xTtwYpch*q0DTl=UG9)l12r~}T8`bR`|W5e~}sWchkG}U^nC!ANp?|ef@>fwwB ziz%|CqEX!Dh{r8ItOv0bvqFwcOy~XnE%QzFM&li(`MWYU$7TPKmCtK#j*2qhB>+$4 zJh)hOCN`uzk_WCkE#ZNDq=QVp@~l=qj;S9$p>ZCE?iEmbFre0z8xiiz=oH|}h5s_k zqQQW^b8olsSlgCbZK5@72Oa&v;E8|KczOHP~SJVyF>14T7%HbJVJYlYv0`Rom< zovj;e?S|z^3nY?zX2K!ft=0+_1t~FPn2=-U=heH%B}-%@zX6*AlXMLE`_~S)GzWFz zOrLUnh{9)TXwBh4`PbgU1$Dyj2vez%Dwgi~4QqGe2e|%Cv~z_4s<=H<+Z;vOehZ34 z>3ewmJ7AN#utiLELgL(f4ED^}dc3L3VR8FqUi@-Xp((yP(Rd&@x-Tz{DRs41@@6&# z<=MzP8TMBkO2L2+lo1G_zVV(wZn~nKc^#6RVPek1kk3|^YId5ptO`3@MX$-8VAYeK zDJQlteOE`HJO5nGhqN&M9mU6myW1`bnz5J}6GJa|A?zp+TMEkI%{AT7Ad&;E!4Jh< zi^D*;-CWo=+w~B7^s+FL9QvROu}#l!Z8rWk^crIOgNIh*+@27sBJ0oBOynh>3rNme zwN72~Ua#BN_1U4f)FK-z2L6wrzy)Oz5%YK0ITgUpGAO}FiH{9D=p!{dcnag(B|uLJ zw27mx)F0;tLJ3m|4?q7QF>gsccB6C0*-#600*Pvy*46{rL{hLf|Hv6CcE|Plk|@ex zPul!Wo2L?;)V**(i#p!=+XFQY^Unzw{)d6!x`;kw^@I`d1D*_OXI#E5#$~TT)nC-f zr`vnmT>aSPeK(PCWp!x2+>^+Cinj|U6a%vsUNt0l5Fz0$5xl?9db#ds{pS37gCF|H z1i(@Mhp}^P&MZ*3ZEV}N)v;~cwrxA<*jC3K+j=`rI=0@}wr=lxKHgn*&u>_@)~XuM zoMY}#-*Q!cmFSZfX^)ff|6<0aK2REfR)de+TPV?4JiofP>0Jv)y)n(aF9%6^ ztgberZ)Ugh+c)H_`HsBVvm+pU*bC_WGP0$eDUn0a+w85B44Rj>>%xP=o`PGX{%Yb5TfVaHayVsS6+etOR3ihZ&vfH0DY{DawId*_(JvIxkLcV~C1jyM`J}`~WhaJNgneUv)~c>PZ%zjvdfR`mbDWC33n_+`#29jR zo8BBhD}}oqk44~FMS54+`#$a3s?CZE&c3Smr6o8J{8@2E&o|vaM|JRuK1qQ4hU?UzI?3D6syO^78g* zXZ{XUY(U#3-YMtrK?ylvObhoZnvF){gt2E{?Trup@EFUa85+Dl4faTucqcW(nYRW5 z(F*2LhAUleHCzjMESfrQhGsq2zK4trXEp3!e=eieehmg3v1K%&Ym9irtzUz;JUi6* z!83Uvg>xBnY76KdE6{wXX&e+1Ek(0f$6$|>yFV1Sf6-~5e%M)ne5 zGl=Z|h*B&I_PhLdN|I7FMeR%QAkF&;%HiU4MM2`%joP_`P4A^e*=u>r!m4_r#g)5X zt926Pv6xO=VCO325|MIx6Ks*%cBVRUhKfh%sR3*Tec3+KB3*xZjf7Y+G9UJ+wL9On z#rSzbD(VUMXlbARt%7QNlrx`G!v`;XO%Kja4_{yamSM29Vr2mze|3eUd9AQ>d00#m z3mLK;5;0i)H(iCke~FJHEm+5Pt}m<=`6*R^~upB>H70HQrMyO zRB462&$DH_thBfBAV(-N<$Nbls`3H;TTjw7Wg!v|qBcdlpO-f$MAhj!MW&a(n%j<4 zMXExB^-nH|eh2mjU`qv@YLld0_uGLhcx^xyA&+qP>Pq(JNGOfx?@!PXkaZ8YkN;Q= zyZBYL{u{<*`!5)ih2?*NF_-kFJvTdCx9zSqTwQ(Bi`*DuU&ukO&2?_a8Z8oko%4D9 zd7dinqWv+tRTgQk$9~W2)72w}RYeE#VbjjA)7U&VC#p~a0~S$T-$xhRKQkXbtNH=| zjh!3j?&)@O4qDOOeB;!ZEzrx)?Mn}sqGUo}ySRHB5E`^vrW}H0WzUu?TW+452Icef zd%C#r4t#~@HC^>l_J z(hfCzjncCMoiXCDSuDGZ)+(d^$Sv->yo)9{C~?D#fT`@NWNvc-Q!neX}ZYdFAr z8}>KK^;7jzTqnKXygN{;lsW1aCd00gDco7Q!4U<8f>EgrX?3Zya~+V}s|`#of~*3Z zUFr*XtTCywt9pGiY)tdT1inWCQ*1211ces=+Dst&sk8$w;h>!MJ`+A29NF6#>2R8y zS@^A10Us<8#4K>5XB!&XetR6#tM@%rB9K~-?5bTzji0d@H>9fZ+`g6Kb zjS9K=?b$lQ!vzn69c5-+#G`a5!KIwGt> zTk%`XX7$=F&NAz^Pw8s)6}|^#bi`sngmRE`n1-s&$J5srZa&r?Y#ICk4d74`(MIB! zQl$Oh%XjxiJn>pQVSJ(wnOQ%C$fA9xxMRBpXq&E5XwHQFg!WUvcxbzfb?x`Mlc#7} zS(_wtqGJK`?i~5TDvVGWA$6SUGQUY((9m?8SbIf|Lal)K*sBJ~;Pi zB+5|?qQ^1^@zYWzH;)effCchpfrHrPT%M#`!Y|WyXesnMYyf#T!&m$>o$e@4mBclQ zn_WlxC3;*hOhYEG89wxvU;?Z`B+{K=VbjIrEdXd{Qt#e{qp*p5y+m_+1G=d59HM21 z47Qrt2p(vWW(e`nG z#o#noFuJTy9*dhgD;Tt;R*?z+UqDQ*n*iT>91P*?jBu?7r0{|{p9w6INcsIoA~GfB8+%hwpR2QFz+yr z;7jimumbc@riUT5s56qpkR5LI2b=>cEy`e62yh5=46};w(3voWEs_s^10xC89wM=o zK*umkGO10;=&9XQm@CEmO73eb5nr|n0|hdlUrY>sw?c>{wDSRR27;99c#TyggYg-+ zHUZ63GJm~G^y(%z?)pu&P>lGI%L9vcT+x}F*SJtfqyhpsm5mv2!LD-oNmyCsi9rao zNdU4TY0$vutbT*v@czvTYZAR*(V&2Oj}y;{=ubBk0-W02Vp)MO-dR|VMyYcc4j$?_ z|8OC-SShD%{k(U2f&pYJD=WVcR>-9jRyNu5XrMnDNZCChI3uRY)OOT6OCEy<^1dLC zCTD@@NB~~PL&q$v#Arw2P(umo`y6SLU!xIKPB-gI1<@pVRaq4b6v;hI5$|7KTJIDY zc7w#(5~20<@I!0IN*y$pe2|_eSOCm7==4H-D^3p9$ME4GWi7@Pwe zDibQ@$n3m7Vl8~Wbj6?Hy9T5?N2nRwVEZtBbH>P&vY3hQ=9f>To1{M_9C~d^LuBxR zzF3P@(CmUd0be;C0x1k2*c=wnF#@%qsABrR+Qb`@bXCZ-go0fNfGm|>LsmSia9Ha8 z%C05IvQ(c#MdmabM57e+s$}grsqsa7qe%z3t2JdoNUE>#`c1TPSj8x-%skUBt@d#7 z>WdMeGGj@Vs{0@znxq0<6EW<=5A1Dcg-`?z6U+VJYGe2Q?{c5TgiyjXvcL6zCI^g# zy)A8^lYTZOf}0N|4(5$c zdvaew*X0_*#X;Iv7{LPJv_Kae^anr(GH=|_yj_8`Z%obG#hg1BqgHNCJ02A1>w1S5 z(@jZ9>U@Q4jg{iXD(P1NV++3i9u78?wU=1iC^&98D!48Z&g+9|9zJ_l1)Ugo3^15d zvX|BzXKy6@z5n5qXqH0O$Jk_c_lvZ^)}3!Z+Xcc?`7D`wJgM5safl=fg#SmjLb#`< zc8F^b(nu#{uV~znu+<>gHKYqTyZhZYY~(dkftZ%&MH*!yDo^7dWsO#=wf86|bTf7y zZ_|Ldfew{Q=C71D}d%NG-r9W{dD$m6b7|R& z;6gr}WFP1PbvO^n2wqdaTP~Pz{|M)i;Jq}#3|;Azds0#ZXs~=bm{I?VKRiD%)ZIO# zQ-8^3?>cZ{m?}x7;bkOn|3crtQ~CiT^;eV>S6B|kQ05n`qwz9z{m$PwFd`Vm=~%4K zZqr@@JXXyGCsNkp=u)t&h^&iWl?T3ahtmauKSUn8n@-z0wWK(9GodrT6XcZD3G0Zk z>>8*}p5+Hrs?6N}nbSr}{@O)C64!jIORrgt{e^8e!&Cfsj-m?pUa`VkM@zR9md_iI zlgfNoL!>a77*nZ6MPtERb@Y(oo9!0M5Xoe^_mq04b4;@sS!S+gDZZN;ew-1&9T}(F z374bL$M&*#mYS;_q&Fi%&3J2uv?7R+n--OU{xH}4~4^%xx^FF=Z)RC299#$CD9kY)5oe4GUa5k9Tk)1j5bXl2{XPk@2 zrR7g(ujB0%wP|yi%IWVpw^FAC`Kbs=b161eDpc%h16IdT;!1bnm8W4y9XMAS)K7Sg)G+aM zX&@i?EcsrQi!dnc=~}z?0#k!#t*v_IRrb}|__g~X0*Ux}SzpOp6qu?@0lR$S6(b?FqY6_?< zvuA>an}^DARG3Rhg zxLqIRNH~=z@J;ePk5eT$rF5t|BAs$;)81;2DuLdPq7A-#O`D1?6E2`w^~>lj6xvff zliH+xy&G>7p#K0y*SAMz+1(@@G**v8AIUq-1KVuv-b^%n)ZxLRS8Vy8??x){|^L=tY*5!hlDaO6cEriD0i zK=;yM_e?<`)E^)>Mt>Bjz~?SlIZ@*3;SThCgYe=(LwU-w7tfjUI7|L_#z{k|FmA#?H- zeSQYm$h(bX5z3O|WL{@OwY$g)xvooA43=c+ zu`~5Y%sw!0?X#u+W-c8?jq&vRlN-OXfpQjnl?r|NBk*K1Ui6}7rD9ZSwqd!xazmq+ zmL=gzj|~QXAe1WOkRZ7DX@E)x?Q1TpHkYiZ9K-Lz{p?)G{B>LBph7^N%(s$!YBdNb9 zfCmHh>SQ2HPV6ZfEo;&%sQk!$(Hi4Zb->yASw$ER;*Wqq+#y&-iS-7sox+G>16h2D zv;@a<(ew53agNy9P09a263w(J zDj)I|k%T!MzmOSHHp=0ju5u%z7*cv>(p4xo?C(Gn!$kLDYSLaAj)o{4`*X`5{jA4MJ#b}hj|DTBN1#P!G6(|zAjo%7tEF~#@=Cg3lC3=PF zb1{6*t$s=CVC3)e94;}-An*&;q9Hho&&nJzuHxFX-d22jBdH+w&fK8TL^JvPNFHX| z95Y@#*?Yj`3IT z4XB=P{36UUMt3kdt6IAuRE7fBwoRswIROb9AY-d{C}Mnht{m?kbjC4;6Xi< zOf8x>FY5ZlgjR^!WYQIOobt{>-@r{luaECjpx~l*eftv7&DBza9*^(_a1YZ`Y_f)LRM$B6>*L^tW)AE0x6o{Z!yQ|!oo7M#0%U@L zz0=)4YxDIk;Q3}VUU8|c5d{;Cn<4(6ghuuz27bW3UKVx!?W>(IAD4 zm2(Ban+~V6u3aHoDCc?+^ZM$_T-9R(0`0Y|mj~=&IbQRaDQ zOH!9J5eP#p*wRY@Y|iW5SZ|>^C@Gj21}#n8h7!+G2g3fnd*aC-E`fzeyTa=9O1WO>wd^f4P5YW6>j_{e=BLo5HnEH`S#f+CGMlHPo78RVGd-~Ord&0uhHNOT8w6WRarAMG?v|2$SNkj0 ziqKx0_7v@>`8pi_uEgq`i69{XgRN7&PZ}BEVew&4?P27RB)n&ZTJ^X)LrK_DGz_sw z60NgW)_4Xm_)Zq#?%HfCmZ8B?LV$stTDn>!@}ffsp-~hPASs}EQ$iLU#NFxtUI4Ef z{Q9}2X4gjuI9I_z23Y9RpJ)n#(cO7-Q+)smDdcD0-;~hOzBvIa9HO~-{vk;}`-NXd zNo2-m{mK8_{>{}~P``!>J$j0rznFIdfLU@LHccVlUVtNQHeJ4*Rnpcq;$U6gijc7R zNrplcI$QIj?MeniaY=kM$g!~sCO>@+euIrceYpHLtjGTUVm%H{uKxwrJJgf&+K^1$ zvb)}5=*T@HKo{-D-=>XCp@|`+R;7A1wE|UPiOGV%NNU;t*mxyw&eB^#L{ld@^y!<< zV_*pW6cupXsy_SRA7|$oKka|G_76x=!lD)`ln&&R`JP%yO{`N=?4Pq$HIGnpbp1S{ zm`O@iHm_E;1sWvV*io;Xqr1FLAC6{PEu0s!j3XaeXGiSv9~tZvL?;449H7YcEq~8? z5ZuZxsvS0yVaRUE2lae?Y`Y5H&sTG3>o2%+HtRz2WS=t3T9~|M#(iP?5A1%@-+sE6 z|C^>z_gcVp!VvGSYIk@Bou%?hisSA+Kf_F%bj1yyDkLIOuR|z22fiJgqyL1Yx+Wb| zQgQ||zya2$FjF66rX7+YdxM&JJAIXB&{P3ZyzF#V`Bud=(lhlV^!e|=(bFl@Iep(- zk1csCRfcjOO$=38#_fx8_CkELk{1&lPI!VHY;Ic;&CX#{VAJkEvgW;7*eq zdf#qw{C#_AYV*_Mf`5VG4|i|&Dk6)5_IR;}w4>h2Gt!@yzfn51irOhI;-T$;X+oLA zswbtVT-WH&!oJK6>^daBZu=+or2L#Yhbxg6gTrnmCJyb;7Ev+SSiI*Di;1&T(bCXL zJzQv?c{xv5{B!`{N?D?{H}Vh$5iEhe0D)-<8G&P54^f-bYl3uXTvP8yA;XbDhhc2` zlA1i*&@*5S3HE1Nt6I2HVYM>ce?JaV6l8oX54PFMOu`wjMmmmo zf_?}3roUowXLvl4L1d|@3l&P_t55RI1@y^R1jE+NSc;YO^;}n~X^VUL73LWYGSD1y zat&NLxGDB*x8>?7r<75<)XoX4jG_+)%VCd}rz&b8z^_yx79ME_N^IwN_m zS>rnF@r`tok;@ecu-DSfow5@MrD(2)k!7$E`zb6lMkxlHVe#?@T*ghLKj}E8d8y>` z?9x@BoavP0Mil8oV?8hbEakw6TXO2&Ldz*d7H;zCW6rky3ObxZz?Y=PvFq*FU@6gq zdr(0p8a}hukyhBT;K}MXta<8fm|%R@f~z&w9bTwV7j8f@_H|L#RVM{xw{u-+dq}OA z(+;c~qB!J7L2;N%3NhH43^+{gb3fKCyfGI=`2#C{r)LOFwjgZPmB7iOK` zHm6fChkF9ZZEC$XC7pj8BS%!X67$S#|5;|~3G-u?BnN4{Td$;b2#V0IBg(@>m;y5j zZVBwKY$;$^`Wp4TZ#rdG$a|gy*8=zB{nzivB2?G{qTOSON5LU%$=D!;wMSv_;B5iZ zy;DGc0aZM>eoY>@OI;e8s+pKssgJ&4z(Z(M%urPJ$h-cm`Q8JAcwsgn7_EqSF)|7% zaTE-59Jwc2*%8V(3hEdSE?If65eR(`+2TTPW-4|?(!7~yQZkabLi1o`o_inV`D@j2+t86nD3rEUlRWx#(3PP}i zFT@)t6pe)HPn@J?yO^3L`qLO>rn|gjI*L4>I zLY7uFuiqcKp5cQDfB8O;c+1Sbf<&O-uZ5w-$EiU=KX!g1aq^i9de<-KnG^=jpVFFOsRk8zyyZMO z=6-eCl1=qxdUXq6u13Bd++n1q(%f|XbDA>~x?|OJ&guHYqC^Q_p=BgiK_U&a{D~u! z5qBu3Dmj#o)mE9z6gCMY8}`R)=kH^2algnW&!p%Z-2xjWY>}{Mdj1pFS<1K1&_P_3 zi(Ki+J^zg?oxh;T@|zwl{bfpz(xr}P8UmMvWJ;14Bn1RsG!5hlD2$vj#w7fWziMgV z-lUm%-3BK~PkZWP1IO$Omy`6r@x1z8EcAY~C$ri?#e1n7-QdcCjNOxGY62a-nya}c z1$rZEClzvF+I@eth`t3&O?{t8_tMgN7lfK=!#{^8q%e}%(xnVu<2|3m;+-JY3h3UB zq88U65R)cm_lfR?AO2(sc=X0uiP-;WMc$W8>7D>zh+<@nf*D06?2rMZlijR}@LMek zjJT2WkVmuB!O!OKjNHHcapOPkPfq(9_eP3$gZ2_gSRrz8!KtSOp^%U4kX{@e>_gt2 zk+NC|(1;*{5v2^467R1tLTt|V@>mn<7!4ILCy(E0>REzgP|Oa zlmuMpq!agl-HgbNb&sDROhi`ExkNr!zaVJ-m1DMi;)}XPrLbgfNCYXsx&s z5C>W;Yb6(^Otws=5!ifLZL3}cC*PFxgsK5O!5t_5QBs~AI%F;8!i*VlG!8F(ME(<0 z1&40?DB+bjwIFX(#(%D*2&UE-xRo5Ca(n8j@(Z@lO7j6E_Qs`?5!p`V@>dnqpZj1L zaA!8V;UsjM@}Bu#ETkh~dJymjzRKU5{m`gN@fO$nL>JW3v8~~eU}Z?C83EMsxP^0O z!M=aP4d_z`sF-obe}iCJrD$U~U^*&fpJptpiB7}*tok1P)X;3=T-%ZeSf|;fVkRbY;?Js!= z;~JvNLTAKiJ;SzKp|y(0$r?i%X=+NsS?{dQSi99{sqRUD(BY*3xh8PAlD^w@9zKuB_2&<2ypDym@+V^)jt;m8!;mg#(tB16aD8=cDFuT@tG%8 zY$wl{bbrZs&>{ zd_gBrx1vd%h}6{L__IG^z)KGyNC=^1eC%0o*gd9vvEQ9>)UT9kD}@G!?4Gafwt zNU{Z}WoTgw%`zE?&)h&7lT{IVCjhERi}ewd8D#lEWEMrvrE8=IMdja0;rPg@tmt8s zE}JGjF)C1z@UwlHw{8?&FW2|+i6u9`-ltZ%zqwLOA#KGO^-K1}ZHcF1E$aTYl9??4 zvAjL+wcgdG>e-IMqU3#Riwg z1MZwXW%z$AM;Bzq-D$;YnK6o~ik|SjE0sQmT6cT|_D=L;<>2H8qH#;VW~s`Z>=C=s z%LE`)P_Mtq6R)HaeGAf0HqBJ#F9)64ye}NJI6BrC`HC86jS!!oIQn2)jGPHF zc02SMneci?Y?@&`i)b71V_%nZOQQXggsdB@c5$zL&hc&=@fxoUEQTr|tg4?&E>&<~kAAuW z`e6QP?p5j@#K^tHM7GG4T1m1bRKw%seRHPP=asvTvC@s}y)Q4^&bEalqF--Be5AYQ zA87*$SGiwb3&)u&4|KA08sDYS^K7FMHPw3zbd-q9W;WRDlZzVeTz{qQd6scR%=&u zA;7V7EaryG%!9{VYj?M~eM+}cp4KQOVKUS5uee(JZEg~Kp&5azbiRFE?$PQCtR;JP z=KYK*p@c5vQUhkhXD&zM&Vqa9X+qgjv(Tr{x>&9@IBt<8f-8-`i)K+rU{URRNglrV z_qEMOZ8epmO3X0ab!~^a$Td$$QTxjcao%s*HL6hI8i$E$??jfC+s-RtbI^&sQS78~ z4kCE_I+^u=rKQ)y^V5YuUDWfuQ<>2ZYKwfAg;M0+w}!qr9MQG4@^U`>Mqd(u#?q38 zkt4}JQ@wz87IZfrEzxE=e1mS|=?=^*#l!Y=JN}`N)Hudp0UGw=)q`snHd$+F*{A}m zKYqh!A#6w2Ew2M3KeAp-R$P3G;R>^znfJc!ep42G=*`$tEcJBQi2o5dAiO*iTc|T5 zjqa8nPS_bftq!lR0veX=$4$xb3<<|V2DIYjzUDq`<#+||24(RvC|-J z`izg0U)?Rjsg~{$@jQ{2t~5V%NAV**6=sr|W&+_)e$(+4h%-3PrJs8kCB(g&yCAo& zEM+*dh)xq&$agKCmLN<#P?xOxYdCVAHb!S+vE;>@Lep*k792rh{I(Zg;QZ79(j#ol zV;&e{K5`c9j7%UUx&Wu^WYSYqfFSI*iC$=FVW1b_c%kPdx>|?^CK`0x#CtWF$yq$) zlJ~-)4gS$TB%$iatI3La(k|k8GR$k+NXi8ot8q33g_9OaKTM5J05?!Tfd9IYSioea z9+(7lSJD~y^Oo_#hLK3NluzB1LHUrNlMoY=PaR@5C@X>a`S!g;uXbN~ND3gGUF0nP{vO(-Td=w2c-i zY2}OR6EPff-~q)5&R}kWlIVzzva!NCRCQ58e!=qAn(191r6Et;wclE6%GQW$4YU zW7!uyzvJI7{0B)1Wo(?rQ{R<1RtNOHhkK=3CdGOGlbz6CWtPpuXA52!s#hb_pW6rC zt)Tw|r6MLBbW&;nQosP#DJH4U(R0`NZ{Kbv&klf^JaW>Qxl>b>c^)}0{M0=Cs7~%V zll=--SYhS!G0l+h3`)J;Lan%1;T5Lsg){OQCdaxT`j3na4B{~?n^uC@l)guzE zbzF^!k6mW)TUnO}mWe=CE`A&7!5}=*aOmXES!J1|R9N5^0)%q9oct1YU zVqyzzY&3g%bc%0Xc}oE@769hO_u(`^+-2G=yA7*qqn>iBA zS?e(8&D%w#UQj*P&esJ<=rz-e@q;Thuk597b!SAyUlHt}u@@B?NdEM3wghz}pa%Nmd`Y0B z4#xh*g5MTP!&eY~qy72Px1EXDJD;JQ`ht3M(_-l4^i;OOdv;es#|2L`f{*8E9yMQt znY6S~q7o0b0qn|?oonfbsb2z59K8{vV|M}7U)9g-exfH>^s?~Ks4JS8NQG#4*!saG zmY%cZ72?n`Rucao5PmXNE76z%i=9>^3yH>)JZZhNl+@0A=6)5$o_e$>+Ls;a@rxS z-B%EYz5Jd=Z1YW(YVGP`YcP3{$5h!5nLjBfu-~_5UE3P(hj_N8`A4jlx2U^MSwx6& z(JjrZm^~fQ#uo1iOa)-<8(b$#Uq@#Pm2F3V=?FwO1=Lx!X!Y>+vO`{AT!@WUruzy^ zH(0>F@)s52nko2GH+XA86M+AcTU^L|Jq66?9&;meY2J52@NJJ6!jFQUVO&_uuyJ9= zz!Ip?P8hsBl5h*c-=twb>|^oUuxr-nO0P+^VhVOJb1k3fUi^fGUBo)FpsVJCTlBNR zUo}=oqztps4!d0eW+^&n-GjKpK2FfP@}r;>^kfvrNibqf3Nlp(`APO_qk@khlY6EA z!oQB;h7$Z4SjBqSnYtMRVOPnhyl~vLoz%v+z9WInBf09QL>bxoO(6lsG06cpP_DG3 zR2Rx{P}hGENB$hN@Y@u92@=8EmpxC@6o?HInuz-j+|mHpK)B5tqTekR zUoqKPSqIoVvyf)QYI>*(#Nf6c%#DAjg%TB{%7%czrQh2Ot*0v^yVIr8z$IZd3=Wqg zlQ`*sYakz#qMIijH>;kQPNB-SR2H@e*DFB3WGA7h%)3bYS2D1>+Fh7?@nL6rhLg+6 zgN)^xuTxFD4WQPaouOh;6lYZ^^>xat<@$^bRc+ULC4H%|Rrf^M-F(J`&c009@!UO0 zb;w96S~aXK`b8eN)E}agjDcPoUfSlt=Ud5wOmnE|4IenmXoHBvD(MMg=b*zrl7;G~g*bZs_LwIr}` zY@b5AL9ow!BMD_qvY5ELDD$Gh*7$(52#BSEM^y$RwCIBIdj{W{t=)Si{vG87UT{x= zsEu^h;t@SGZkW|e4TxHy^blneaIpFIT8B8O4>@RWr0|u;$78=6$SBJ}#VXTX|Mtg) zj*{{Rpp(fDfu=b}nJ{?&9NlnOCBAUc!j27FfDMF!Qs^)2Aa>i8)@+#}eD89rcka?8Ns@5x#r@6#0Q5R#>O0f7bBpp|e5ZqSXv{h(!`X!nef)OayVg*HOaQ zJm4-CogWaAiTHnZrpWfk(pnUe>>~B@IGM;gqpQBx|GeIbE-aje_2L*M4ygcBcN2NY zE@bbR@M!cC@-)wuSGSC`0f9BD{z+Uc^kmppg9ENG>$zt|Gt)PThv`H@?G{fwfTyd6 z{8Pv394=T4Tu>sK(zAvr=o-MR(2XHJF2e^y>vC|1(Ne4FVecZ?!r}qXcqz`p1(oiJ zI~yw|c_HSpMC{W}i(J$0V4o8YP{|stW=09#bk|u}Qxg{M-1mq2p-iwo>+0O68sk)L zFdQR4`7XG#n~Ug0Qf4jWuj{m(PP%^tn@jUo)O^}Z8fvaNQ)mP-7-opd0;f=A*DBmg z>LNMbsxeQp1oNVMYc($QwZ9h?o46S9&=FGP&mvvtcqS_a^-Q0(=R&v6_S4XYN#G2u zbF^YSKrMP^u&d2f#Ak#F>aYcfY1*m64FM{jcLANS&bMtl5dinZwo5~sUP?t%%s58;QUFhpd z8cj5jB@4dF$Q8m&Q~%m9`~wHz3wJg%roc#)EL;uy2S?mo39*w<*r%uTT6fwfhmfh>|DgeE7wBsfu+7QDb1FZ#9#a!4m;Z>WAU5s02@4TlCYF$lncsu^`b_=bEuHb&+dmWQQw@1AuDC)=3>nouOec{ z;G)(2HE=P9@!>cqI1CbW7?%Nga0PWUf-k&aiy#>kY%GWkiqv(7LF@#TpXSi*6?gry zeAXNW?W#yQ%m&FF%F{u_?4=8Kk6;cLr~Mpf{opyF(!o?;5a-s(3>(9EM-n+ErWi1W z+Tiky%;^>KbAm(33x3s^9V{Bj{zk0>=&O~wW)*qyOyomN1Pj8}5ri8j<;rcm)xS~X z9C!EX-d?OOEi7<-%X>Cdhs$w!QwwM(+^X@FDJgmPS$OwR>p_0T$LZ-aT;g&d<6^QR z0a(EjIyaQ^2t^_IjA-8Em1|=)2rj39C*5rKZ8sFkFGZr+!SA5O|0=N zii;(B5J3ZI+i<)!=D==~_k_7@NzP^Um60#b>{M?(-NF{xXJxrIz4X&*8c;jRZf|B~ zC*67&o+op75Yi4~?@u`=d#rV7BUzzT%^}$YExMj-?W^tjf@FDMqJ8H_{CKaHC2uad zcB|i|H3)_Mew@Kn{TUh?w18RSxDgOWrz5c5-4sm1Fa$!QC3^dQ-Mu$QH4?fBfFh*P zU%at|g2~ZUEBu?;7&+Blr8i9G?)RSG-W-t#zw`qP93t8ivR*KFHQO^F5i}hLQKSA! zVJ)7{ljFGXgi!qal$Ea)nwV4WK-Rb3n_JDhr-*8pe)&fV!u4z`=#3*cW;T!++F*yt z<~M%Z3Qnl^$OBhduAM@E@TxfBy<80wD!FJYE-&|#ZC-mk7SnNB=nK9KwRh(#;L2KR zhXSjOU4BYh=l9OJ185HZ<)$~pB-_OHU7_uQ&2&&2^hTS79CMm%mO-95*?SXz7U04O zG>!Ek^ev~Tw65x1(HI;!bYL)Wy3n`>#B1NEsr|Sl18ZqOYL~m-5&Y5)tei+BC&E)3 zPORcfkwv;%_Y;`PQ0s*;@Sm+PPGQw6KukHz`9L4-Ebv)z(I!w0Gm@61y_+8P#UcjU zV5h3g3nb~fh&d;+IDAG?4*VQP&pkqy<=Wu5$qW4A8?i{EoPd6|Fb#Zqvl)`RW%F3x ziL?EARC2F%^3w4*^E4kHp;%I z5lzPL82L<-qJka;*Ki=!y}!P6D8$+rMoVjwlA~!?fDA4q6eQ3f%)I^FAqg#qv@i>C z4D2sky-`eRNvMHE+Dzr!Z502!ax{a}Q3%8sIM?TcIq@O0k+&#o?R7(w0n`m~GiSnDA*Eu0yz(oaQru831U|K5SzfTo_H{26gAu z*=i5M4b%|+HGqE?TCsQh-Pv{KWB&~{DTXQi+&uCb(fV_Ds1?kZIrJBc)%PHD5x`6R zeQ*k{O|g|YUe1sRtzDmXJ)LfpIo(8bLH+&iOocyig8sbJei0eshrZWY*pBM$t{B3? zkCb**oys)GC9|a_(wl<8SqClST+*S6^@!Cak8|50Cy3+vniBkHp^A7LoYva9%=mHI z@P2BiKVbveAjlgA*OTX4zcp(xSBjoX9VTf2!LoKUrDi54vjF5w0BOMUsq>l&$7FSH z*FGpT8a)pMj3F78vLCMg1mRd_6$DI)W2e{Do%eg=iKEaBIr+=&CG?iVn6t~E0SQl5 z7{15NU-Mtk05?I3RiPRWPEPeoToRmK+8OB0l<$=ok4_>f<|fk}5uBxJaqKG837T|b z4e9pQSt@-nx7%h!P}%yIY`y8v>O)oeg-#7lC9hx#fEL`4)maGiC+xCfXTujGDC*Ed zTS`aL{2$fpT0c!4xz>#hvgar@M;RIRW2m&;>OIHOO(*dyevP|GhT!ticrGV}_0KC+ zl>Svoc&+dJ?D|gHj9@hf>J)k?(_114=3g2>pk=p#2UCxIghU7#J{4pvDUXIH{+D;|T7C3`bl4Vd{SIBp>3?_LgykaXiyumh z*j8Uwe(o75t4SA zO}{5zNTnLp)vCC4gw5r`CBP7**Ifi{jEkDoX?kvKh*ISmv(#65VDTK6%Gg zdB#tSp1S{J5=})e!9LD93b609#I2z2#Thy#S3)L*$Ab2ur*5b&*1Av9PTEk zYosPS7iDU?27!?o%~ea)NvTze&GNpJ0@wk_Of0rR%z^Udx^}lrwW+jD>d}^J4!KY4 zeF_=FOhErFO*x`W%O(^yzpiiPY&QS8xZG4?MGmYi|4_(&zQLFUYYe-EO(jig&SpWz zhrdnyonIySl(n&PM@NQc&S&f7Kn$b}sN8BiZ>6e_@ zIkmf}LK+)8nZ*Fg_!96qt!c|-8h6V!l0K1r*pAM9x=Vmg#AREx*j9bEhhl8=FYlGwi7sO1y6?FgN4S-5^-s5ZWr3eZXZ85ghj5Aw zEEQ!T-#o>?@Ux*g-F1q6@Aa=T8V)X_lC)z2S}>X&9$5#h@!ayTck0_?M*dG**BJ~4 z*M*4?(V~UuD@qV7tL>_*_f8O2S-VT3EEdtCeR}kktX`r-FA<&Soe)HHiRg>yUcyIy zeKX(tAT1ZH%C-_z=s8AaAvIQD;>gO(~C-+c;;cV_v*KtulDa-q@7Wa&ajnctKC}`EOSj zvJ%03Yx_BuxwUaV4i|hlc9&HroeUju;iyc-DC{JX<}pdWNkUtS;6B57KdLT z2B@o}wI6$9hRlRjWYX&YKf9su3xzZvE~pb< zPf4t_3-+tJg4RX>I&W``qmfcqrbQjO3MV zzrG54KWF>39M4Mb>3XirC;=Y3UARz7n}+mwG$E5!mj0r?msziipW0Q!%PyEb)e8G3 zcK`14gSHo?IQSGf0o}ZqQ}{(7Rp@8PzT-VzJZwf8S5G*!f!d77TG_!6&>S3CvZcT2 z#W&7!Km}YisLp*y?>GAFayL%0vy~&=u<0cl610R{8*yumAWv?0H{0$0=!TWW-8bG_ zIVEmUQ|j$cHsP{momOO2oZd(8w32BC3*3y@Rkj#*bJ@g!;*U9d-ZnU=vL$@U#DDX+ z4l5M%+VDlzn5k%sKnI2lz6zmN7-%fICQ^7v6Ky&bI% zd|6p2)^MRdsCvNgH5B|zL+8-LdyMHIGnpsa*IY2aM$II*_qls zB)`^b!K8?V(u@_x#TzcE&$Rk&?9^Jg#bgMORVo~aI=WmwHdtG4nRj#EF1V({5C)5EJI$kQd#c5%rjkWBu)947An7Ey{bx^!_cvV zoV1{J!5gqKBR2c}wPw>295D)SOk9WS>qG1#>!BBKe^Dp{*LX%3?TpQx>yhTaDSM0$ z=WMjO1iPFW1{bOVqv_r(fYrW_hE2KszN4_y28rrR3sk08jEpcd`j}J=)staeAFT3e zqA&%I`yn{!VE5)5Ogt%Se{t~R(Gb++7rD?*%Z9Gz$G9?X)$9{4?PaLAObKQwDs`F< z%_vMAd-yR|ds{}pE)uL5BG|ju^LII;M0-rCvSPPZDJzKGCc<-8ngri6!TpDR0eVtW z=EE_VaXzumTEk})Ivd1+cF>|WrX$N4=m^g^FVytaCorEB+8*EuPfjjla5ck`_=(yTcPQ#v2{Bo+$2T?LX1LU6CR+P_U=l{9uJ6 z0At2z8z)7zTRGUo_zys)kM-tjul+GDNMRj-01Y(gWSXzc9EHos+N0R)`>U9+<}w7ce{} zpAnXIl@B*a11H;l$wFF`YnDNskC_cht6*Qw#UeKBrJ-dEawWESrAH+PHL`{8AIjB} z=P65;aQZ5b>55bF@nU^CE?YkADbT0T4s%8u)6VjdJWA93ioN(D;0^NAj*UwV0)R&4 zCR#@pP7J@t9$ikN2$z}-^^Qh+9|!15KnL2Ojg@CIUk@D8Xex(buwJ5)&)rN$DN;&i z-jneC@@d-wMBNH>GqiaO5>f64C*6(Z%9=Hqz4@~h2f?}SZ2G+NJt