期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2017
卷号：6
期号：8
页码：17111
DOI：10.15680/IJIRSET.2017.0608204
出版社：S&S Publications
摘要：Clustering is the process of organizing objects into groups whose members are similar in some way andis very important technique in data mining as it has its applications spread extensively, e.g. marketing, biology, patternrecognition etc. So summarize the data stream in the real life with the online process is called as micro-cluster but itshows the density when we are combining the data in the one place. In the offline process we are using the modificationclustering algorithm to re-clustering into larger cluster. For that the middle of micro-cluster factor as the pseudo factorwith density randomly calculates their weight. That density data location of micro-cluster isn't always preserved theonline process. So used DBSTREAM, the primary micro-cluster based on online clustering aspect seize the densityamong micro-cluster thru shared density graph. We broaden and examine brand new technique to address this hasslesfor micro-cluster-based totally algorithms. The density facts on this graph are then exploited for re-clustering primarilybased on actual density among adjacent micro-clusters. For that shared density graph improves clustering first-classover different popular information flow clustering techniques which require the advent of a bigger variety of smallermicro-clusters to obtain similar results.
