Research on Fuzzy C-Means Algorithm Based on the Information Entropy

Abstract

Aiming at the shortcomings of the Fuzzy C-Means (FCM) algorithm, a new improved FCM algorithm based on the information entropy has been proposed in this paper. Meanwhile, the entropy evaluation method is used to measure the varying degree of influence of the various attributes in cluster analysis. In the simulation experiments, two indexes called the fuzzy division coefficient FC(μ)and the average fuzzy entropy HC(μ) are used to evaluate the performance of the new FCM algorithm. The experimental results show that the FC(μ) and the HC(μ) are0.864 and -0.012 respectively which show that the improved FCM algorithm can obtain better classification effect than the FCM algorithm in practical application.