Novel clustering algorithm using K-means Chernoff faces

作者：Wang Jinjia^*; Hong Wenxue; Li Xin

来源：Chinese Journal of Scientific Instrument, 2007, 28(10): 1916-1920.

• chernoff faces
• multivariate data
• clustering
• K-means
• FCM

Chernoff faces are simplified, cartoon-like faces that can be used to graphically display complex multivariate data. Chernoff faces are clustering algorithms, which group together similar faces. Some disadvantage of Chernoff faces for clustering are the subjectivity, huge comparison workload and difficulty in assignment of facial features. So, a novel Chernoff face clustering algorithm is proposed, which combines with K-means or FCM clustering algorithms. The experiment results of IRIS data and vegetable oil data indicate that the proposed new algorithm is superior to traditional clustering algorithms.

• 出版日期2007
• 单位燕山大学