作者: Kuo-Lung Wu , Miin-Shen Yang
DOI: 10.1016/S0031-3203(01)00197-2
关键词: Euclidean distance 、 Fuzzy clustering 、 k-medians clustering 、 Cluster analysis 、 Fuzzy logic 、 Robustness (computer science) 、 Correlation clustering 、 Mathematics 、 Machine learning 、 CURE data clustering algorithm 、 Artificial intelligence 、 Pattern recognition
摘要: Abstract In this paper we propose a new metric to replace the Euclidean norm in c-means clustering procedures. On basis of robust statistic and influence function, claim that proposed is more than norm. We then create two methods called alternative hard (AHCM) fuzzy (AFCM) algorithms. These types have robustness clustering. Numerical results show AHCM has better performance HCM AFCM FCM. recommend for use cluster analysis. Recently, algorithm successfully been used segmenting magnetic resonance image Ophthalmology differentiate abnormal tissues from normal tissues.