Semi-supervised Fuzzy Clustering and Its Application
Yang Xi-yan · Journal of Fujian Normal University · 2015
An extended form of semi-supervised fuzzy clustering algorithm is proposed,and its iterative solution is given.This new semi-supervised method uses class information of the labeled data effectively and reasonably to improve its classification ability.The iterative solutions of its membership degree and clustering centers have concise forms as those of FCM.Experiments on the cucumber data set show that the proposed algorithm is better than FCM,linear discriminant analysis and other two semi-supervised algorithms.