Summary
2007 International Symposium on Nonlinear Theory and its Applications
2007
Session Number:18PM2-B
Session:
Number:18PM2-B-4
Gray Scale Display of Input Data Using Shooting SOM
Masato Tomita, Haruna Matsushita, Yoshifumi Nishio,
pp.288-291
Publication Date:2007/9/16
Online ISSN:2188-5079
DOI:10.34385/proc.41.18PM2-B-4
PDF download (1.4MB)
Summary:
The Self-Organizing Map (SOM) is popular algorithm for unsupervised learning and is widely applied for many applications. In the previous study, we have proposed a new type of SOM algorithm, which is called Shooting SOM (SSOM) algorithm. The important feature of SSOM is that the neurons move like aiming at a target, namely, only some neurons near the cluster move toward the cluster to hit the area where input data are concentrated and 1-neighborhood neurons of the winner neuron get away a fraction of an inch from the cluster. Because of this feature, SSOM tends to self-organize each cluster along the figure of each cluster. We investigate the behavior of SSOM and apply SSOM to data visualization problems.