Home > lyngby > lyngby_nn_edev.m

lyngby_nn_edev

PURPOSE

lyngby_nn_edev - Entropic neural network, input 1st der.

SYNOPSIS

function [dEV, H12] = lyngby_nn_edev(EY, H, X, V, W)

DESCRIPTION

 lyngby_nn_edev       - Entropic neural network, input 1st der.

        function [dEV, H12] = lyngby_nn_edev(EY, H, X, W)

       Input:  EY   Error on output
               H    Hidden layer output
               X    Neural network input
               W    Output weights

       Output: dEV  Derivative in a matrix
               H12  Intermidiate result: 1-H.^2

       First order derivative for input weights in the entropic
       neural network.

       See also: lyngby_nn_edew

CROSS-REFERENCE INFORMATION

This function calls:

This function is called by:

lyngby_nn_etrain lyngby_nn_etrain - Entropic neural network, training