A Simple implementation of a Restricted Boltzmann Machine and Deep Belief Network

Peter Swire - swirepe.com

A Restricted Boltzmann machine is a type artificial neural network that is gaining a lot of popularity. They are used in everything from feature generation to image processing to simulating creativity.

A deep belief network is a series of Restricted Boltzmann Machines stacked on top of each other; the output of one is the input to the next.

This is a distilled version of a tool I use for generating features. It's written primarily for ease of understanding.

wiki

scholarpedia

dr. five minutes

deep learning

I'm releasing this into the public domain. If you find it helpful, pay it forward.