A new Differential Evolution based Fuzzy Clustering for Automatic Cluster Evolution

In this paper, the problem of finding the number of optimal cluster partitions in fuzzy domain has been countered. The fact motivated us to develop an algorithm on differential evolution for automatic cluster detection from the unknown data set. Here, assignments of points to different clusters are done based on a Xie-Beni index where the Euclidean distance takes into consideration. The cluster centers are encoded in the vectors, and the Xie-Beni index is used as a measure of the validity of the corresponding partition. The effectiveness of the proposed technique is demonstrated for two synthetic and two real life data sets. Superiority of the new method is demonstrated by comparing it with the variable length genetic algorithm based fuzzy clustering and well known fuzzy c-means algorithm.