Fuzzy clustering with the generalized entropy of feature weights

Fuzzy c-means (FCM) is an important clustering algorithm. However, it does not consider the impact of different feature on clustering. In this paper, we present a fuzzy clustering algorithm with the generalized entropy of feature weights FCM (GEWFCM). By introducing feature weights and adding regularized term of their generalized entropy, a new objective function is proposed in terms of objective function of FCM. In GEWFCM, minimization of the dispersion within clusters and maximization of the generalized entropy of feature weights simultaneously obtain the optimal clustering results. Moreover, GEWFCM is viewed as a generalization of the maximum entropy-regularized weighted FCM (EWFCM). Experiments on data sets selected from University of California Irvine (UCI) machine learning repository demonstrate the effectiveness of presented method.