Using k- Means algorithm to optimize clustering process based on Particle Swarm Optimization (PSO) algorithm

Clustering analysis is one of the data classification techniques, which try to group similar data within the possibly different classes and such grouping is done based on the similarity criterion among them. Similarity or non- similarity among data is usually expressed by a mathematical criterion. Many algorithms have been proposed for data clustering but due to several presuppositions, which are considered about features of data clustering so finding the comprehensive method for clustering has been changed into one of the unresolved problems. In this study, a new method has been created for clustering based on Particle Swarm Optimization and k-Means algorithms. It has been dealt with data clustering separately by means of PSO algorithm in the previous investigations so the present research tries to alleviate its weak points. The created algorithm has been implemented on Benchmark data and its results have been compared with the aforesaid clustering algorithm. The results of study indicate lesser sensitivity to the created primary particles and the higher value of silhouette measure for the suggested method

Keywords:

Clustering , Particle Swarm Optimization (PSO) Algorithm , K-means Algorithm , Silhouette Measure