A toy library of machine learning algorithms implemented in numpy for Andrew Ng's machine learning course. Not for production.

Currently has implementations for:

• Neural networks (multilayer perceptron)
• Collaborative filtering.

There are unit tests, but I haven't set up a unit test runner, so they have to be run individually:

$ python ml/algorithms/test_neural_networks.py
$ python ml/algorithms/collab_filter_test.py