一种搜索模糊聚类全局最优解的Tabu算法  

A Tabu Search Algorithm for Fuzzy Clustering

作　　者：孙银慧[1] 白振兴[1] 董天明[2] 

机构地区：[1]空军工程大学工程学院 [2]94333部队

出　　处：《科技广场》2008年第10期6-9,共4页Science Mosaic

摘　　要：模糊聚类问题由于其非凸性而成为一个难以解决的数学问题。在解决模糊聚类问题时,会出现很多局部极小值和鞍点。因此,启发式的模糊C-均值算法是应用最为广泛的算法,其缺点是很容易陷入局部极小值。本文提出了一种搜索模糊聚类全局最优解的Tabu搜索算法,并比较这种新算法和模糊C-均值算法的性能。经过多次数据试验,证明Tabu搜索算法在搜索全局最优解时是很有效的。The Fuzzy Clustering Problem （FCP） is a mathematical program which is difficult to solve since it is nonconvex,which implies possession of many local minima.The fuzzy C-means heuristic is a widely known approach to this problem, but it is guaranteed only to yield local minima. In this paper,a new approach was proposed about this problem which is based on tabu search technique in order to find a global solution of FCE. The performance of the algorithm was compared with the fuzzy C-means algorithm＇s.it was proved effect.

关 键 词：模糊聚类 模糊C-均值算法 Tabu搜索技术 全局最优值 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]