#Kohonen Self Organizing Maps (SOM) - python algorithms

This project contains a python implementation of several algorithms related to the self organizing maps of kohonen. For more information on Kohonen maps refer to (http://en.wikipedia.org/wiki/Self-organizing_map).

base/

contains base python code for manipulating datasets, defining the algorithms common interface and other machine learning tools and helpers (like cross validation, distance calculations, etc).

algorithms/

contains all algorithms implementation:

• kohonen: kohonen som maps algorithm implementation

variants:

• recsom: recursive self organizing maps variant implementation

other algorithms:

• knn: k-nearest neighbors algorithm implementation
• kmeans: k-means algorithm implementation
• stochastickmeans: stochastic k-means variant of k-means algorithm implementaion
• lvq: linear vector quantization algorithm implementation
• svm: support vector machines algorithm interface

datasets/

contains sample dataset to play with the algorithms and compare their results.

Note: In order to get all python scripts to work, the root directory (kohonen/) must be in the PYTHONPATH.