This repository is an playground to implement any type of machine learning algorihtm.

Currently Implemented Algorithms:

• Genetic Programming (Standard Version)

• Restricted Boltzmann Machine with dynamic numbers of hidden neurons
• Single Layer Perceptron (binary classification, multinominal classification has to be done with a one vs all strategy)
• Multi Layer Perceptron (binary and multinominal classification)

• Deep Belief Network (pretrained rbms stacked into a mlp)

• install python requirements in virtualenvironment based on requirements.txt.
• docs are created by running make html inside doc/ directory
• tests are run by executing nosetests in the root of the project (use nosetests --nocapture to also see print statements, to run a single test use e.g. nosetests path.to.module:Class.test_method)