Visual Principle FCM Clustering Algorithm in Complex Network Application
FCM (Fuzzy C-Means) clustering algorithm is a prevailing used clustering analysis method, which was first proposed by Dunn, and promoted by Bezdek. At present it has been already widely used in image processing, the fuzzy modeling and the economical management and so on. But the fuzzy C average value clustering algorithm is based on distance, the clustering result was quite influenced by the isolated points in the data. Because FCM algorithm cannot distinguish the isolated points, once there exits one or more isolated points, the cluster center will move, moreover this algorithm needs to be given the parameter beforehand, which is the clustering number. The algorithm result is extremely sensitive to this parameter. Different value causes definitely different clustering result. But human eyes possess a singular aptitude to group objects and find the important structures in an efficient way in pattern recognition, image processing, and clustering analysis. In this paper we made some improvements to the FCM algorithm based on the density clustering method by simulating human vision. And we use the new algorithm on complex networks and analyze the characteristics of each cluster. It will be much easier for researching and exploration from decomposing a large network into several small networks by the method of clustering properly.
FCM cluster Algorithm vision principle Survival sector
Wang Zhongjun Mo Chunling
Wuhan University of Technology College of science, Wuhan 430063
国际会议
第六届管理学国际会议(Proceedings of ICM2007 the 6th International on Management)
武汉
英文
2000-2004
2007-08-03（万方平台首次上网日期，不代表论文的发表时间）