Hierarchical Intuitionistic Fuzzy Possibilistic C Means Kernel Clustering Algorithm for Distributed Networks

Abstract

Advances in distributed networking have resulted in an explosion in size of modern datasets while storage and processing power continue to lag behind. This requires the need for algorithms that are efficient in terms of number of measurements and running time. To combat challenges associated with large datasets in distributed networks we propose hierarchical intuitionistic fuzzy possibilistic c-means kernel clustering algorithm. The algorithm executes hierarchically by performing clustering at each peer. The intuitionistic fuzzy degree and tipicality membership functions and weight-attributeentropy factor improves clustering performance. The experiments on artificial and real datasets establish the efficiency and effectiveness of the algorithm.