摘要:Fuzzy C-Means (FCM) is one of the most frequently used clustering method. However FCM has some disadvantages such as number of clusters to be prespecified and partition matrix to be randomly initiated which makes clustering result becomes inconsistent. Subtractive Clustering (SC) is an alternative method that can be used when number of clusters are unknown. Moreover, SC produces consistent clustering result. A hybrid method of FCM and SC called Subtractive Fuzzy CMeans (SFCM) is proposed to overcome FCM’s disadvantages using SC. Both SFCM and FCM are implemented to cluster generated data and the result of the two methods are compared. The experiment shows that generally SFCM produces better clustering result than FCM based on six validity indices. Keywords : Clustering, Fuzzy C-Means, Subtractive Clustering, Subractive Fuzzy C-Means
其他摘要:Fuzzy C-Means (FCM) is one of the most frequently used clustering method. However FCM has some disadvantages such as number of clusters to be prespecified and partition matrix to be randomly initiated which makes clustering result becomes inconsistent. Subtractive Clustering (SC) is an alternative method that can be used when number of clusters are unknown. Moreover, SC produces consistent clustering result. A hybrid method of FCM and SC called Subtractive Fuzzy CMeans (SFCM) is proposed to overcome FCM’s disadvantages using SC. Both SFCM and FCM are implemented to cluster generated data and the result of the two methods are compared. The experiment shows that generally SFCM produces better clustering result than FCM based on six validity indices. Keywords : Clustering, Fuzzy C-Means, Subtractive Clustering, Subractive Fuzzy C-Means