A Method to Ascertain Parameters of Samples and their Feature Weights in the Weighted Fuzzy Clustering

Abstract

The reasonable definitions of samples and their feature weights in weighted fuzzy clustering algorithm based on the thought of normalization and each computational formula are presented. Banding together with computational formulas of samples and their feature weights which derived in weighted FCM, we can get the regions of sample’s weight parameter() and sample feature’s weight parameter(). Then divide the regions into intervals, point out the clustering situations in different intervals and how changing of and affect the clustering result and the choice of feature.Try to explore the relationship between weightd parameter(、) and fuzzy constant(m). Finally, test result demonstrates the validity of the regions of parameter and its partition.