A Comparative Study of Hard and Fuzzy Data Clustering Algorithms with Cluster Validity Indices

Data clustering is one of the important data mining methods. It is a process of finding classes of a data set with most similarity in the same class and most dissimilarity between different classes. The well known hard clustering algorithm (K -means) and Fuzzy clustering algorithm (FCM) are mostly based on Euclidean distance measure. In this paper, a comparative study of these algorithms with different distance measures such as Chebyshev and Chi-square is proposed. The new algorithms are tested on the four well known data sets such as Contraceptive Method Choice (CMC), Diabetes, Liver Disorders and Statlog (Heart) from the UCI repository. Experimental results show that FCM based on Chi-square distance measure gives better result than Chebyshev distance measure. We also propose the FCM algorithm based on σ -distance measure. The FCM algorithm is also tested with cluster validity indices such as partition coefficient and partition entropy. The results show that Chebyshev distance measure is reported maximum partition coefficient and minimum partition entropy than the other distance measures. This paper also provides a brief review of applications of K -means and Fuzzy c-means algorithms.