Multilayer Backpropagation Neural Network

Implementation of the Multilayer Backpropagation Neural Network

The code implements the multilayer backpropagation neural network for tutorial purpose and allows the training and testing of any number of neurons in the input, output and hidden layers. Number of hidden layers can also be varied.
There are a total of three files with MLBPN_Train.m used for building and training the multilayer network on a desired input pattern and MLBPN_Test.m used for testing the trained neural network. You can provide your own patterns for training by modifying the DefinePattern.m file.

The learning rate, total iterations and activation function can all be changed if desired. The code provides you the ability to modify the forward and back propagation stages individually to allow for fast convergence on complex training data.