Study and analysis of particle swarm optimization for improving partition clustering

Abstract

Clustering is a widely used technique for finding the similar hidden patterns from a dataset. Many techniques are available for data clustering such as partition clustering, hierarchical clustering, density based clustering, and grid based clustering. This paper discusses various clustering techniques along with their benefits, drawbacks, characteristics, and applications. The paper also discusses various validity measures, which are useful in evaluating cluster quality. The paper discusses issues involved in Particle Swarm Optimization (PSO) and compares various variants of PSO that address the discussed issues. PSO can be applied to partition based clustering for improving performance and quality of resulting clusters. In that connection, the paper discusses about how PSO is useful to solve issues present in partition clustering. Moreover, the paper presents a survey of partition clustering using PSO. This paper would become useful to beginners and researchers in advancing the field of applying data clustering using PSO.