Neural Network Toolbox™

Initialize SOM weights with principal components

Syntax

• weights = initsom(inputs,dimensions,positions)
  weights = initsom(inputs,dimensions,topologyFcn)
  

Description

initsompc initializes the weights of an N-dimensional self-organizing map so that the initial weights are distributed across the space spanned by the most significant N principal components of the inputs. Distributing the weight significantly speeds up SOM learning, as the map starts out with a reasonable ordering of the input space.

initsompc takes these arguments:

inputs	R x Q matrix of Q R-element input vectors
dimensions	D x 1 vector of positive integer SOM dimensions
positions	D x S matrix of S D-dimension neuron positions

and returns the following:

weights

S x R matrix of weights

Alternatively, initsompc can be called with topologyfcn (the name of a layer topology function) instead of positions. topologyfcn is called with dimensions to obtain positions.

Example

• inputs = rand(2,100)+[2;3]*ones(1,100);
  dimensions = [3 4];
  positions = gridtop(dimensions);
  weights = initsompc(inputs,dimensions,positions);
  

See Also

newsom, gridtop, hextop, randtop

Provide feedback about this page

initnw

initwb

© 1984-2008- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS