Data Clustering using Hybridization of Clustering Based on Grid and Density with PSO

Abstract

The purpose of this paper is to present a new clustering algorithm based on grid and density combined with particle swarm optimization (PSO). The algorithm is referred as hybridization of clustering based on grid and density with PSO (HCBGDPSO). Inspired by the influence function introduced in density-based clustering (DENCLUE) algorithm, a novel method for computing the density of grid cells is adopted in HCBGDPSO to achieve better precision instead of the method used in common grid-based clustering algorithm. Furthermore, PSO is combined in the algorithm to search the arbitrary-shape clusters. Finally, the results of the experiments indicate the effectiveness of the algorithm