In this study, we propose a weighted mean subtractive clustering algorithm in which new cluster centers are derived by using weighted mean method on the data points around the center prototypes found by subtractive clustering. Comparisons between weighted mean subtractive clustering and other clustering alogrithms are performed on three datasets by using three indexes and visual methods. The experimental results show that weighted mean subtractive clustering finds more reasonable cluster centers and groups data better than other clustering alogrithms do.
展开▼
