Python Neural Network - Handwritten digits classification

This project is a simple Python script which implements and trains a 2 layer neural network classifying handwritten digits using the MNIST database for both training and testing.

Usage

This script requires Python 3. By default, the script trains a NN with 300 hiddens units until convergence.

pip install scipy numpy
python neural_net.py

License

The code is liscensed under the MIT lisence. See LICENSE.md.

Yann LeCun and Corinna Cortes hold the copyright of MNIST database, which is a derivative work from original NIST datasets. MNIST database is made available under the terms of the Creative Commons Attribution-Share Alike 3.0 license.

The files data_testing and data_training contain the MNIST database saved using Python pickle. They are licensed under Creative Commons Attribution-Share Alike 3.0.