Summary
2007 International Symposium on Nonlinear Theory and its Applications
2007
Session Number:18PM2-B
Session:
Number:18PM2-B-3
Self-Organizing Mapping that considers neighborhood uniting
Mitsushi Yoshida, Daisuke Shima, Kaname Kurokawa, Hisashi Aomori, Mamoru Tanaka,
pp.284-287
Publication Date:2007/9/16
Online ISSN:2188-5079
DOI:10.34385/proc.41.18PM2-B-3
Summary:
Self-organizing indicates the system producing an own structure. Especially, the map system is called the self-organizing map (SOM). SOM can map to the low dimension by which the adjacency relation of the multidimensional data is maintained in nonlinearly. This method has been focused on because of the e?ectiveness for clustering, information compression, and visualization. On the other hand, since the SOM tends to compress the distance between data, the mapped data does not guarantee the actual distance relationship of the input space. Therefore, the problem is that an actual distance relationship in the input space is not expressed in the output space. In this paper, to solve the above problem, we propose the multidimensional lattice data addition learning model by which the concept of the neighborhood uniting is introduced to the study of the conventional self-organizing map.