Evolutionary dynamic particle swarm optimization for data clustering

Abstract

A clustering algorithm based on particle swarm optimization (PSO) and fuzzy theorem was introduced for data analysis. Clustering algorithms require users to set some parameters, such as the number of clusters k. However, it is unreasonable to expect users to specify a meaningful value of k if they lack prior knowledge of the data. This paper proposed an algorithm to determine the appropriate number of clusters and produced an associated set of cluster centers automatically. The proposed algorithm was compared with stand-alone PSO clustering and fuzzy c-means on three data sets. The results of the experiment showed that the proposed method was able to determine the number of clusters accurately, and to deliver favorable performance in the clustering of data.