wolffd@0: wolffd@0: wolffd@0: wolffd@0: Netlab Reference Manual somtrain wolffd@0: wolffd@0: wolffd@0: wolffd@0:

somtrain wolffd@0:

wolffd@0:

wolffd@0: Purpose wolffd@0:

wolffd@0: Kohonen training algorithm for SOM. wolffd@0: wolffd@0:

wolffd@0: Synopsis wolffd@0:

wolffd@0:
wolffd@0: 
wolffd@0: net = somtrain{net, options, x)
wolffd@0: 
wolffd@0: wolffd@0: wolffd@0:

wolffd@0: Description wolffd@0:

wolffd@0: net = somtrain{net, options, x) uses Kohonen's algorithm to wolffd@0: train a SOM. Both on-line and batch algorithms are implemented. wolffd@0: The learning rate (for on-line) and neighbourhood size decay linearly. wolffd@0: There is no error function minimised during training (so there is wolffd@0: no termination criterion other than the number of epochs), but the wolffd@0: sum-of-squares is computed and returned in options(8). wolffd@0: wolffd@0:

The optional parameters have the following interpretations. wolffd@0: wolffd@0:

options(1) is set to 1 to display error values; also logs learning wolffd@0: rate alpha and neighbourhood size nsize. wolffd@0: Otherwise nothing is displayed. wolffd@0: wolffd@0:

options(5) determines whether the patterns are sampled randomly wolffd@0: with replacement. If it is 0 (the default), then patterns are sampled wolffd@0: in order. This is only relevant to the on-line algorithm. wolffd@0: wolffd@0:

options(6) determines if the on-line or batch algorithm is wolffd@0: used. If it is 1 wolffd@0: then the batch algorithm is used. If it is 0 wolffd@0: (the default) then the on-line algorithm is used. wolffd@0: wolffd@0:

options(14) is the maximum number of iterations (passes through wolffd@0: the complete pattern set); default 100. wolffd@0: wolffd@0:

options(15) is the final neighbourhood size; default value is the wolffd@0: same as the initial neighbourhood size. wolffd@0: wolffd@0:

options(16) is the final learning rate; default value is the same wolffd@0: as the initial learning rate. wolffd@0: wolffd@0:

options(17) is the initial neighbourhood size; default 0.5*maximum wolffd@0: map size. wolffd@0: wolffd@0:

options(18) is the initial learning rate; default 0.9. This parameter wolffd@0: must be positive. wolffd@0: wolffd@0:

wolffd@0: Examples wolffd@0:

wolffd@0: The following example performs on-line training on a SOM in two stages: wolffd@0: ordering and convergence. wolffd@0:
wolffd@0: 
wolffd@0: net = som(nin, [8, 7]);
wolffd@0: options = foptions;
wolffd@0: 
wolffd@0: 

% Ordering phase wolffd@0: options(1) = 1; wolffd@0: options(14) = 50; wolffd@0: options(18) = 0.9; % Initial learning rate wolffd@0: options(16) = 0.05; % Final learning rate wolffd@0: options(17) = 8; % Initial neighbourhood size wolffd@0: options(15) = 1; % Final neighbourhood size wolffd@0: net2 = somtrain(net, options, x); wolffd@0: wolffd@0:

% Convergence phase wolffd@0: options(14) = 400; wolffd@0: options(18) = 0.05; wolffd@0: options(16) = 0.01; wolffd@0: options(17) = 0; wolffd@0: options(15) = 0; wolffd@0: net3 = somtrain(net2, options, x); wolffd@0:

wolffd@0: wolffd@0: wolffd@0:

wolffd@0: See Also wolffd@0:

wolffd@0: kmeans, som, somfwd
wolffd@0: Pages: wolffd@0: Index wolffd@0:
wolffd@0:

Copyright (c) Ian T Nabney (1996-9) wolffd@0: wolffd@0: wolffd@0: wolffd@0: