Title :
A new cluster validity index for type-2 fuzzy c-means algorithm
Author :
Mema Devi, O. ; Begum, Shahin Ara
Author_Institution :
Dept. of Comput. Sci., Assam Univ., Silchar, India
Abstract :
Considering the growing application areas of type-2 fuzzy logic, this paper investigates the existing cluster validity indices that are suitable for FCM clustering algorithm. Based on the cluster fuzzy degree of FCM fuzzy set and the existing cluster validity indices, a new cluster validity index for the type-2 FCM called SM-index is proposed. The optimal partition or an optimal number of cluster, is obtained by maximizing the value of SM-index. The experimental results on the UCI and microarray data sets are reported to demonstrate the effectiveness of the proposed cluster validity index in appropriately determining the number of clusters.
Keywords :
fuzzy logic; fuzzy set theory; pattern clustering; FCM clustering algorithm; SM-index; UCI; cluster fuzzy degree; cluster validity index; microarray data set; optimal partition; type-2 fuzzy c-means algorithm; type-2 fuzzy logic; Clustering algorithms; Fuzzy logic; Indexes; Informatics; Linear programming; Partitioning algorithms; Uncertainty; Clustering; validity index and type-2 membership;
Conference_Titel :
Advances in Computing, Communications and Informatics (ICACCI), 2013 International Conference on
Conference_Location :
Mysore
Print_ISBN :
978-1-4799-2432-5
DOI :
10.1109/ICACCI.2013.6637497
