Magnification Laws of Winner-Relaxing and Winner-Enhancing Kohonen Feature Maps

Simple item page
dc.creatorClaussen, Jens Christian
dc.date2006-12-30
dc.date.accessioned2026-07-07T08:18:05Z
dc.date.available2026-07-07T08:18:05Z
dc.descriptionSelf-Organizing Maps are models for unsupervised representation formation of cortical receptor fields by stimuli-driven self-organization in laterally coupled winner-take-all feedforward structures. This paper discusses modifications of the original Kohonen model that were motivated by a potential function, in their ability to set up a neural mapping of maximal mutual information. Enhancing the winner update, instead of relaxing it, results in an algorithm that generates an infomax map corresponding to magnification exponent of one. Despite there may be more than one algorithm showing the same magnification exponent, the magnification law is an experimentally accessible quantity and therefore suitable for quantitative description of neural optimization principles.
dc.description6 pages, 3 figures. ESMTB 2002 Milano. For the extended journal version see cond-mat/0208414
dc.identifierhttps://arxiv.org/abs/cs/0701003
dc.identifierhttp://arxiv.org/abs/cs/0701003
dc.identifierpp. 17-22 in : V. Capasso (Ed.): Mathematical Modeling & Computing in Biology and Medicine, Miriam Series, Progetto Leonardo, Bologna (2003)
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/134312
dc.subjectNeural and Evolutionary Computing
dc.subjectInformation Theory
dc.titleMagnification Laws of Winner-Relaxing and Winner-Enhancing Kohonen Feature Maps
dc.typetext

Files

Collections

arXiv