RLproject

The goal of this project was to investigate reinforcement learning(RL). This is another paragdim in machine learning along supervised learning and unsupervised learning. The notebooks contained in this file are a mixture of the notes written on RL along with the code used in a final report and a few derivations. Note the litrature here focuses explicitly on RL and does not go into deep reinforcement learning. The idea is that after completing the research one can return to this file at a later stage in order to recap on reinfircement learning litrature. I've placed it here in case anyone else is following the same path and the hope is that it could help them move faster than myself.

Most of these notebooks are written extremely informally as they were written primaly for myself to understand RL, however I hope that they can be some benefit to someone else.

Litrature on RL

Here I provide some of the litrature, classes/courses, talks/tutorials, and others (git source code and posts). There are two main text books on RL, Reinforcement Learning: An Introduction (2nd Edition) and Algorithms for Reinforcement Learning. The former is longer to read and is suitable for most but the second is more mathamatical and seemed to present the theory well. If Implementation is your intrest I would go with the former rather than the latter.

Textbooks:

• Reinforcement Learning: An Introduction (2nd Edition)
• Algorithms for Reinforcement Learning

Classes:

• David Silver's Reinforcement Learning Course (UCL, 2015)
• CS 8803 - Reinforcement Learning (Georgia Tech)
• CS885 - Reinforcement Learning (UWaterloo), Spring 2018

Online Courses:

• Reinforcement Learning in Python,
• Deep Reinforcement Learning in Python,

Talks/Tutorials:

• Introduction to Reinforcement Learning (Joelle Pineau @ Deep Learning Summer School 2016)
• Deep Reinforcement Learning (Pieter Abbeel @ Deep Learning Summer School 2016)
• Deep Reinforcement Learning ICML 2016 Tutorial (David Silver)
• Tutorial: Introduction to Reinforcement Learning with Function Approximation
• OpenAI Spinning Up
• Advanced Deep Learning & Reinforcement Learning (UCL 2018, DeepMind)

Other Projects:

• carpedm20/deep-rl-tensorflow
• matthiasplappert/keras-rl

Selected Papers:

• Deep Reinforcement Learning with Double Q-learning (2015-09)
• Prioritized Experience Replay (2015-11)
• Dueling Network Architectures for Deep Reinforcement Learning (2015-11)
• Asynchronous Methods for Deep Reinforcement Learning (2016-02)
• Deep Reinforcement Learning from Self-Play in Imperfect-Information Games (2016-03)
• Mastering the game of Go with deep neural networks and tree search