PERFORMANCE ANALYSIS OF CLUSTERING ALGORITHM ON ARBITRARY SHAPE AND DENSITY VARYING

Abstract

Clustering is the process that dividing the data into such group that contain similar data in one group and other data in another group. In simple way it separates the data with similar characteristic to make a cluster. In this several algorithms used for that can group the data by partition, hierarchy algorithm, k means algorithm, FCM algorithm, FCM sigma algorithm, standard FCM algorithm, Grid-based algorithm. Most popular algorithm k-means and FCM algorithm are used to partition the data into group, this two-algorithm having different approaches in k-means data will be included in one particular cluster whereas in FCM a data can be included in all existing cluster, here k means and FCM by default uses Euclidean distance measure. Here we using different distance measure to evaluate the performance analysis of k means and FCM algorithm using cosine distance measure, correlation distance measure, city block distance measure used in various dataset based on k-means clustering and FCM algorithm.