Innovation Award
 March 2010
Nominee
Vote |
Back propagation is a well known algorithm to implement neural networks.
It works by self-adjusting the weights of each neuron by propagating from the output to the input weights by evaluating the difference between the expected results and the current results during the training phase.
This class provides a PHP implementation of the back propagation algorithm.
Manuel Lemos |
This class can be used to implement neural networks using back propagation.
It can setup a neural network work with a given number of layers.
The class takes a data set and a test output data set and runs the neural network using back propagation to to adjust weights based on network errors.
No application links were specified for this class.

If you know an application of this package, send a message to the
author to add a link here.