Simple Neural Network

This is a package for learning about neural networks through a simplified example. There are many other better packages for production use.

Chapters

The chapters folder contains a walk through for how to build up a neural network, starting with a simple linear classifier.

Src

The source for a simple neural network that can have a variable number of layers each with a variable number of neurons can be found in src.

Examples

There are examples using the code in src for classifying XOR and MNIST image data. There is also an html file for visualizing how a linear classifier adjusts to classify AND.

Further Learning

Andrew Ng's Coursera class

Andrey Kurenkov's Brief History of Neural Networks and Deep Learning Parts I, II, III, and IV

Kaggle