Home > lyngby > lyngby_nn_epruneobd.m

lyngby_nn_epruneobd

PURPOSE

lyngby_nn_epruneobd - Pruning by Optimal Brain Damage (entropic)

SYNOPSIS

function [VMask,WMask] = lyngby_nn_epruneobd(X, T, V, W, Reg, arg1,arg2)

DESCRIPTION

 lyngby_nn_epruneobd  - Pruning by Optimal Brain Damage (entropic)

        function [VMask,WMask] = lyngby_nn_epruneobd(X, T, V, W, Reg,
           'PropertyName', 'PropertyValue')

        Input:        X      The datamatrix
               T      Target
               V      Input weights
               W      Output weights
               Reg    Regularization
               Property:
                  'DeleteFraction'   {0.05} Fraction to delete 
                                     of remaining weights

        Output:        VMask   Mask for input weights. Deleted weight is
                       zero, else one
               WMask   Mask for output weights.

CROSS-REFERENCE INFORMATION

This function calls:

lyngby_nn_eddevds lyngby_nn_eddevds - 2nd order, entropic, input, diag. sym.
lyngby_nn_eddewd lyngby_nn_eddewd - Entropic neural network, output, diag 2nd
lyngby_nn_eddru lyngby_nn_eddru - 2nd order der. of regularization for weights
lyngby_nn_eforward lyngby_nn_eforward - Entropic neural network, feed-forward
lyngby_nn_pruneobd lyngby_nn_pruneobd - Pruning by Optimal Brain Damage

This function is called by:

lyngby_nn_emain lyngby_nn_emain - Main function for Entropic Neural network