Home > lyngby > lyngby_nn_cforward.m

lyngby_nn_cforward

PURPOSE

lyngby_nn_cforward - Classifier neural network forward

SYNOPSIS

function [Y, O, H] = lyngby_nn_cforward(X, V, W);

DESCRIPTION

 lyngby_nn_cforward   - Classifier neural network forward 

       function [Y, O, H] = lyngby_nn_cforward(X, V, W);

       Input:  X   Input to neural network
               V   Input weights
               W   Output weights

       Output: Y   Output from the neural network after the softmax
                   is applied 
               O   Output from the neural network before the softmax
                   is applied
               H   Hidden layer output

       Implements:
           Y = softmax( tanh(XV)W )

       See also: lyngby_nn_cmain, lyngby_nn_ctrain

CROSS-REFERENCE INFORMATION

This function calls:

lyngby_nn_csoftmax lyngby_nn_csoftmax - Softmax for neural network classifier
lyngby_nn_tanh lyngby_nn_tanh - Faster hyperbolic tangent

This function is called by:

lyngby_nn_ccdc lyngby_nn_ccdc - NN-classifier cost function and derivative
lyngby_nn_cmain lyngby_nn_cmain - Main functions for classifier neural network
lyngby_nn_cpruneobd lyngby_nn_cpruneobd - Classifier neural network, OBD pruning
lyngby_nn_csoftline lyngby_nn_csoftline - Classifier neural network soft linesearch
lyngby_nn_ctrain lyngby_nn_ctrain - Classifier neural network training