Case Study on Static k-Means Clustering Algorithm

Abstract

Data clustering is frequent research problem in many data mining applications. In this view, many clustering methods proposed in literature. One type of clustering is partitioning method which is centroid based technique. In this paper we are presenting the case study on conventional or static k-means partition clustering algorithm. Here we used static means the basic input parameter given to k-means is number of cluster (k), which constant for complete execution of data set. We need to decide the k values before algorithm starts and It does not changes, when there is a change in data set. We considered the some cases like distance measures, what is right number of clusters and relations between the algorithm parameters. We executed k-means algorithm on small data set and large data set and we presented the detailed steps for each case by showing the results