Kohonen self-organizing feature map and its use in clustering
Markus Törmä · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1994
Cluster analysis is an important part of pattern recognition. In this paper we present the applicability of one neural network model, namely Kohonen self-organizing feature map, to cluster analysis. The aim is to develop a method which could determine the correct number of clusters by itself. First, the general concept of neural networks and detailed introduction to Kohonen self-organizing feature map are discussed. Then, the suitability of Kohonen self- organizing feature map to cluster analysis is discussed and some simulations are presented.