期刊名称:International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN:2320-9798
电子版ISSN:2320-9801
出版年度:2016
卷号:4
期号:2
页码:2427
DOI:10.15680/IJIRCCE.2016.0402204
出版社:S&S Publications
摘要:Clustering isan unsupervised learning technique which is used to group samples of data based on their features and properties of instances. In any clustering algorithm determining the number of clusters is a significant task which needs to be efficient to group the data with relatively similar characteristics. In this paper we use a method Gap statisticsalgorithmto determine the number of clusters for a Fuzzy C-means clustering algorithm[2]to group the samples of data. In gap statistics method we calculate the error measure for each sample of data and evaluate it with a reference value and depending on the evaluation we obtain the optimal number of clusters which can be applied to the Fuzzy C-means clustering
关键词:Fuzzy C-Means clustering; Gap Statistics; error measure; data characteristics; reference value