[Master Index] [Index for export/lyngby]

lyngby_nn_initvw

(export/lyngby/lyngby_nn_initvw.m)


Function Synopsis

[V, W] = lyngby_nn_initvw(N, arg1, arg2, arg3, arg4)

Help text

 lyngby_nn_initvw     - Initialize weights in neural network

       function [V, W] = lyngby_nn_initvw(N, ...
          'PropertyName', 'PropertyValue')

       Input:  N   Neural network structure

       Output: V   Input weights
               W   Output weights
               Property:
                  'Seed'  Seed for random generator
                  'Input' Input to the neural network. Used to scale
                          the weights 

       The neural network structure is containing the units in each
       layer. The first column contains the number of input units,
       the second column the number of hidden units, and the third
       column the number of output units. The first row contains the
       number of units counting a threshold unit.  

       N = [ [input+threshold hidden+threshold output] ; ...
             [input hidden output] ];

       For a two-layer feed-forward neural network with 2 inputs, 4
       hidden units, 1 output and with threshold units the structure
       should be defined as: 

       N = [ [ 3 5 1 ] ; [ 2 4 1] ]  

       See also: lyngby_nn_qmain, lyngby_nn_emain

Cross-Reference Information

This function is called by

Produced by mat2html on Wed Jul 29 15:43:40 2009
Cross-Directory links are: OFF