A Novel Hybrid Clustering Techniques based on K-Means, PSO and Dynamic Optimization

Clustering is a process for partitioning datasets. This technique is a challenging field of research in which their potential applications pose their own special requirements. K-Means is the most extensively used algorithm to find a partition that minimizes Mean Square Error (MSE) is an exigent task. The Object Function of the K-Means is not convex and hence it may contain local minima. ACO methods are useful in problems that need to find paths to goals. Particle swarm optimization (PSO) is a global optimization algorithm for dealing with problems in which a best solution can be represented as a point or surface in an n-dimensional space. But PSO algorithm suffers from slow convergence near optimal solution. In this paper a new modified sequential clustering approach is proposed, which uses PSO in combination with K-Means & dynamic optimization algorithm for data clustering. This approach overcomes drawbacks of K-means, PSO technique, improves clustering and avoids being trapped in a local optimal solution. It was ascertained that the K-Means, PSO, KPSOK & dynamic optimization algorithms are proposed among these algorithms dynamic optimization results in accurate, robust and better clustering. General Terms The proposed techniques shows better results with combined approach with PSO compared to individual techniques implementation