Gaussian Mixture Modeling (Gmm) for Cluster Analysis

Srinivas et al (2014) have suggested Automatic Merging of Clustering (AMOC) as the best optimal clustering algorithm in grouping six datasets viz., Iris, Glass, Breast Cancer, Halfmoon, Path based and Spiral. Each dataset could be made up made up of hidden patterns associated with small subgroups having weak commonality across the whole population forcing us to concentrate model building (a combined representation of unification of entire congregation of a given data set and segregating them into clusters) to have common representation for the newly formed clusters. This paper is aimed at applying finite mixture models (GMMs) for modelling the resultant clusters and to summarize the work done along with conclusions.