Fuzzy C-Means Clustering Algorithm Using Initial Centroid Selection Process In Data Mining Techniques

Abstract

Data mining technique is the discovering process of the large data sets for the pattern of making the group conjunction of machine learning, database design, and statistical reports to be analyzed. Data mining involves the steps are regression; summarization, Clustering, and association of large datasets through the various kinds of terminology should be used to form the bunching of datasets to group the datasets. The clustering method is a crucial way to calculate the distance between the centroids. The Clustering is the process of grouping the data, which is used to calculate the objects that are similar in characteristics and group together. This clustering method is used to choose the cluster centers and the centroids, which is calculate the distance among the objects. In this paper, we focus on the centroids to form the gap between the objects at a minimum requirement of clusters. The defined initial centroids are compared with the randomly selected initial centroids. By this way, the centroids distance calculated through the cluster formation of data. The initial clusters are augmenting the centroids, which is depends on the minimum distance. This initial centroid selection process enhanced the range among the objects. The overall selection of the centroids produces the better quality and the minimum range, Square Error, Precision and Recall calculation which is based on the threshold value. These clustering algorithm and the initial selection process used in the telecommunication and image segmentation based medical industry analysis. In fuzzy c means Clustering, considers each object with a member of initial clusters to the degree of nearness distance through the simulation of MAT LAB.