期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2015
卷号：4
期号：3
页码：754-758
出版社：Shri Pannalal Research Institute of Technolgy
摘要：On dealing with data means clustering plays a vital role, in which clustering algorithm and cluster validity are the most common linked approach used for cluster analysis. To improvise the performance, propose the validity index scheme in clustering algorithm and cluster validity. In general the concept of centroid ratio is for comparing the two obtained clu ster result. In the proposed system the centroid ratio in Pair wise Random Swap clustering algorithm, in order to avoid the local optimum problem faced by k-means algorithm. The methodology behind Centroid ratio is the swapping must be done between the convergence and perturbation in finding the nearest optimum by means of k-means. The centroid ratio is highly effective in MSE (Mean Square Error) which is fast and simple to process. The proposed system is applied with several database and compare the result with those traditional methods like Random Swap, Deterministic Random Swap, Repeated k-means or k-means++. As a result the proposed system is successfully attained the expected result and in order to improve its efficiency is productively applied in document clustering as well as color image quantization.
关键词：Data Clustering; K-means; random ; hdeterministic swap and clustering evaluation
