Measuring the congruence of fuzzy partitions in fuzzy c-means clustering

详细信息    查看全文

• 作者：Abdul Suleman ^{abdul.suleman@iscte.pt}
• 关键词：Fuzzy clustering ; Fuzzy c-means ; Reconstruction ; Cluster validity index ; Information criterion
• 刊名：Applied Soft Computing
• 出版年：2017
• 出版时间：March 2017
• 年：2017
• 卷：52
• 期：Complete
• 页码：1285-1295
• 全文大小：997 K
• 卷排序：52

文摘

We present a new cluster validity index for fuzzy c-means clustering that resorts to information theoretic principles. Overall, it measures the congruence between fuzzy c-partitions and the observed data. We tested its effectiveness using synthetic data as well as real life data. It potentially distinguishes noise data from well-structured data. Together with one available index, namely PCAES, it can help in solving a cluster validity problem.