MODIFICATION OF THE ALGORITHM FOR DETERMINING THE OPTIMAL NUMBER OF CLUSTERS IN CLUSTER ANALYSIS

摘要

Cluster analysis is a multivariate statistical method that aims to classification of objects into groups called clusters. In the current scientific literature, there are many cluster analysis methods that are used to classify objects into clusters. The aim of this article is to evaluate selected coefficients to determine the optimal number of clusters on real data sets and propose a new algorithm to determine the number of clusters, which aim is to increase the success rate in determining the optimal number of clusters. This algorithm is based on fuzzy clustering.