文献类型：期刊文献

中文题名：基于混合遗传算法的K-Means最优聚类算法

英文题名：K-Means Optimal Clustering Algorithm Based on Hybrid Genetic Technique

作者：吕强[1];俞金寿[1]

机构：[1]华东理工大学自动化研究所,上海200237

年份：2005

卷号：31

期号：2

起止页码：219

中文期刊名：华东理工大学学报（自然科学版）

外文期刊名：Journal of East China University of Science and Technology

收录：CSTPCD;;EI(收录号：2005209110096);Scopus;北大核心:【北大核心2004】；CSCD:【CSCD2011_2012】；

语种：中文

中文关键词：数据挖掘;遗传算法;混沌优化;聚类

外文关键词：data mining; genetic algorithm; chaos optimization; clustering

摘要：针对遗传算法的K-Means聚类算法在遗传过程中容易受到适应度最大染色体的影响,存在过早收敛于局部最优值和遗传算法的局部搜索性能较差的问题,提出了结合混沌优化方法形成的混合遗传算法。仿真实验表明:该方法有效地克服了遗传算法的早熟问题,从而得到最优的聚类中心。
K-means algorithm based on genetic technique has a disadvantage that local optimal value is obtained earlier, because the largest fitness chromosome easily influences this algorithm in genetic process, and genetic algorithm possesses very poor local search performance. By combining the properties of both chaos optimization method and genetic algorithm, a new combinatorial optimization approach, the hybrid evolutional programming, is proposed in this paper. The experimental results show this algorithm avoids limitation of genetic algorithm.