Summary
International Symposium on Nonlinear Theory and its Applications
2008
Session Number:B2L-C
Session:
Number:B2L-C1
GHSOM with Ranking Mapping Scheme
Mitsushi Yoshida, Masatoshi Sato, Daisuke Shima, Hisashi Aomori, Mamoru Tanaka,
pp.-
Publication Date:2008/9/7
Online ISSN:2188-5079
Summary:
A self-organizing indicates the system producing an own structure. Especially, the map system is called the self-organizing map (SOM). The 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 effectiveness for clustering, information compression, and visualization and so on. And, the growing hierarchical self-organizing map (GHSOM) is efficient way to project input data onto output map using hierarchical structure in learning stage . However, most of the SOM and GHSOM projection methods are computationally expensive when the size of the data set becomes large. In this paper we present an intuitive and effective GHSOM projection method with comparatively low computational complexity for the purpose of cluster visualization. This method is called ranking mapping scheme (RMS). This method maps data vectors on the output space based on their responses to different prototype vectors. High-resolution maps can be obtained with a relatively small network size. The effectiveness of proposed method will be demonstrated using iris data set.