Determining the Number of Clusters by a Bayesian Approach
Zhu De-gang · International Journal of Applied Mathematics & Statistics/International journal of applied mathematics and statistics · 2013
Cluster analysis, which is the most well-known example of unsupervised learning, is a very popular tool for analyzing unstructured multivariate data. The methodology consists of various algorithms each of which seeks to organize a given data set into homogeneous clusters. It has always been a difficult problem to determine the number of clusters. This paper describes a Bayesian approach which can be used to find the best partition by maximizing the posterior likelihood. Experimental results on real-world data sets demonstrate useful properties of the proposed approach.