A Novel Approach Towards K-Mean Clustering Algorithm With PSO

In this paper, the proposed approach is an unique combination of two most popular clustering algorithms Particle Swarm Optimization (PSO) and K-Means to achieve better clustering result. Clustering is a technique of grouping homogeneous objects of a dataset with aim to extract some meaningful pattern or information. K-Means algorithm is the most popular clustering algorithm because of its easy implementation and quick response. But it is inclined to produce local optimal solution due to its initial partition. The proposed method applied meta-optimization technique to overcome this limitation of K-Means with the help of PSO that offers a globalized search methodology but suffers from slow convergence near optimal solution. Here the proposed technique apply the result of PSO as the input seed of KMeans to obtain better result. Clustering performance of proposed algorithm is also evaluated of the basis of accuracy, execution time, quantization error, inter and intra cluster distance. Keywords-Clustering, K-Mean, PSO, Quantization Error, Inter and Intra Cluster Distance, Execution Time