A Comparative Study of Hard and Soft Clustering Using Swarm Optimization

Bijayalaxmi Panda, Soumya Sahoo, Sovan Kumar Patnaik Abstract— Cluster analysis is one of the major techniques in pattern recognition, which is basically considered as one of the unsupervised learning technique. We can apply clustering techniques in various areas like clustering medicine, business, engineering systems and image processing, etc.,The traditional hard clustering methods restrict that each point of the data set belongs to exactly one cluster. But fuzzy clustering proposed that the belongingness of each data points is based on a membership function.Now a days fuzzy clustering has been widely studied and applied in a variety of substantive areas.To find the global optimal solution we have also applied the concept of particle swarm optimization on K-means clusterings and modified particle swarm optimization on Fuzzy– c–means and performed a comparative study on four clustering algorithms on the basis of compactness,separability time complexity. .