Data clustering using hybrid particle swarm optimization

Abstract

Clustering is an important data mining task and has been explored extensively by a number of researchers for different application areas, such as text application and bioinformatics data. In this paper we propose the use of a novel algorithm for clustering data that we call hybrid particle swarm optimization with mutation (HPSOM), which is based on PSO. The HPSOM basically uses PSO and incorporates the mutation process often used in GA to allow the search to escape from local optima. It is shown how the PSO/HPSOM can be used to find the centroids of a user-specified number of clusters. The new algorithm is evaluated on five benchmark data sets. The proposed method is compared with the K-means (KM) clustering technique and the standard PSO algorithm. The results show that the algorithm is efficient and produces compact clusters. © 2012 Springer-Verlag.