Investigation of topographical stability of the concave and convex self-organizing map variant

Fabien Molle, Jens Christian Claussen*

*Corresponding author for this work

Research output: Chapter in Book/Published conference outputConference publication

Abstract

We investigate, by a systematic numerical study, the parameter dependence of the stability of the Kohonen Self-Organizing Map and the Zheng and Greenleaf concave and convex learning with respect to different input distributions, input and output dimensions.

Original languageEnglish
Title of host publicationArtificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings
PublisherSpringer
Pages208-215
Number of pages8
ISBN (Print)3540386254, 9783540386254
DOIs
Publication statusPublished - 2006
Event16th International Conference on Artificial Neural Networks, ICANN 2006 - Athens, Greece
Duration: 10 Sept 200614 Sept 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4131 LNCS - I
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference16th International Conference on Artificial Neural Networks, ICANN 2006
Country/TerritoryGreece
CityAthens
Period10/09/0614/09/06

Bibliographical note

© Springer-Verlag Berlin Heidelberg 2006

Fingerprint

Dive into the research topics of 'Investigation of topographical stability of the concave and convex self-organizing map variant'. Together they form a unique fingerprint.

Cite this