Winner-Relaxing Self-Organizing Maps
Disordered Systems and Neural Networks
2007-05-23 v2 Neural and Evolutionary Computing
Adaptation and Self-Organizing Systems
Neurons and Cognition
Abstract
A new family of self-organizing maps, the Winner-Relaxing Kohonen Algorithm, is introduced as a generalization of a variant given by Kohonen in 1991. The magnification behaviour is calculated analytically. For the original variant a magnification exponent of 4/7 is derived; the generalized version allows to steer the magnification in the wide range from exponent 1/2 to 1 in the one-dimensional case, thus provides optimal mapping in the sense of information theory. The Winner Relaxing Algorithm requires minimal extra computations per learning step and is conveniently easy to implement.
Cite
@article{arxiv.cond-mat/0208414,
title = {Winner-Relaxing Self-Organizing Maps},
author = {Jens Christian Claussen},
journal= {arXiv preprint arXiv:cond-mat/0208414},
year = {2007}
}
Comments
14 pages (6 figs included). To appear in Neural Computation