The Improved K-Means with Particle Swarm Optimization

Abstract

In today’s world data mining has become a large field of research. As the time increases a large amount of data is accumulated. Clustering is an important data mining task and has been used extensively by a number of researchers for different application areas such as finding similarities in images, text data and bio-informatics data. Cluster analysis is one of the primary data analysis methods. Clustering defines as the process of organizing data objects into a set of disjoint classes called clusters. Clustering is an example of unsupervised classification. In clustering, K-Means (Macqueen) is one of the most well known popular clustering algorithm. K-Means is a partitioning algorithm follows some drawbacks: number of clusters k must be known in advanced, it is sensitive to random selection of initial cluster centre, and it is sensitive to outliers. In this paper, we tried to improve some drawbacks of K-Means algorithm and an efficient algorithm is proposed to enhance the K-Means clustering with Particle Swarm Optimization. In recent years, Particle Swarm Optimization (PSO) has been successfully applied to a number of real world clustering problems with the fast convergence and the effectively for high-dimensional data.

Keywords: Clustering, K-Means clustering, PSO (Particle Swarm Optimization), Hierarchical clustering.