A3-3 Cluster Analysis with Smoothed Data Density Histogram on SOM

Abstract

We propose a cluster analysis method using self-organizing feature maps (SOM). SOM-based cluster analysis methods are attracted among a lot of studies on the automatic cluster analysis methods. In most situations, a histogram, i.e. a density distribution of data, obtained from the learnt SOM involves more excessive peaks and valleys than we envisioned by the true number of clusters. In this paper, a method to smooth the histogram is proposed to make cluster analysis easier and to eliminate the unwanted extreme values. The effectiveness and the validity of the proposed method for the automatic clustering are examined for several artificially generated data.