期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：4
页码：453-456
语种：English
出版社：Ayushmaan Technologies
摘要：There are several techniques and algorithms are used for extractingthe hidden patterns from the large data sets and fnding therelationships between them. Clustering is one of the importanttechniques in data mining. Clustering algorithms are used forgrouping the data items based on their similarity the goal ofclustering is to group sets of objects into classes such that similarobjects are placed in the same cluster while dissimilar objects arein separate clusters. Outlier Detection is a very important researchproblem in data mining. Clustering algorithms are used for detectingthe outliers effciently The algorithms used in this research workare PAM (Partitioning around Medoid), CLARA (Clustering LargeApplications) AND CLARANS (Clustering Large ApplicationsBased on Randomized Search) and a new clustering algorithmENHANCED CLARANS for detecting outliers. In order to fnd thebest clustering algorithm for outlier detection several performancemeasures are used. The experimental results show that the outlierdetection accuracy is very good in the ECLARANS clusteringalgorithm compared to the existing algorithms. It has a very highaccuracy but still it takes time to be accurate. So by this researchwork this can also be done. The aim of this research is to reducethe time complexity of the ECLARANS.
