An efficient fuzzy C-least median clustering algorithm

In today’s reality ’World Wide Web’ is considered as the archive of extremely enormous measure of data. The substance and complexity of WWW are increasing day by day. Presently the circumstances are such that we are suffocating in data yet starving for knowledge. Because of these circumstances data mining is extremely important to get valuable data from WWW.Clustering data mining is the process of putting together meaning-full or use-full similar object into one group. It is a common technique for statistical data, machine learning and computer science analysis. Clustering is a kind of unsupervised data mining technique which describes general working behavior, pattern extraction and extracts useful information from time series data. In this paper we are discussing our new procedure for clustering called Fuzzy C-least median of squares algorithm which is an improvement to Fuzzy C-means (FCM) algorithm. As it is concerned with the least value among medians, it wipes out means squared error and eliminates the effect of outliers. We compared our clustering result got by applying FCM and FCLM by using Xie-Beni Index, Fukuyama-Sygeno Index and Partition Coefficient. The outcomes demonstrate a clear improvement of our algorithm than existing FCM algorithm