scikit-neuralnetwork

Deep neural network implementation without the learning cliff! This library implements multi-layer perceptrons as a wrapper for the powerful pylearn2 library that's compatible with scikit-learn for a more user-friendly and Pythonic interface. Oh, and it runs on your GPU by default.

NOTE: This project is possible thanks to the nucl.ai Conference on July 20-22. Join us in Vienna!

---

Installation

You'll need to first install some dependencies manually. Unfortunately, pylearn2 isn't yet installable via PyPI and recommends an editable (pip -e) installation:

> git clone https://github.com/lisa-lab/pylearn2.git
> cd pylearn2; python setup.py develop

Once that's done, you can grab this repository and set your PYTHONPATH to point to the correct folder. A setup.py file is coming soon for the official version 0.1!

Demonstration

To run a visualization that uses the sknn.mlp.MultiLayerPerceptron just run the following command:

> PYTHONPATH=. python examples/plot_mlp.py --params activation

The datasets are randomized each time, but the output should be an image that looks like this...

Upcoming Features v0.1

• Full tests for sklearn Classifier and Regressor compatibility.
• Quick start in the README.rst file showing how to get an estimator.
• Allow using all layer types as hidden layers, not linear only for output.
• Better error checking for the layer specifications, useful messages otherwise.
• Installation using setup.py like all normal Python projects!