计算机工程 ›› 2011, Vol. 37 ›› Issue (5): 13-15.doi: 10.3969/j.issn.1000-3428.2011.05.005

• 博士论文 • 上一篇    下一篇

基于人工免疫细胞模型的模糊聚类算法

王 磊,王 伟,李玉祥   

  1. (西安理工大学计算机科学与工程学院,西安 710048)
  • 出版日期:2011-03-05 发布日期:2012-10-31
  • 作者简介:王 磊(1972-),男,教授、博士,主研方向:人工免疫理论,智能计算,普适计算;王 伟、李玉祥,硕士研究生
  • 基金项目:
    国家自然科学基金资助项目(60603026)

Fuzzy Clustering Algorithm Based on Artificial Immune Cell Model

WANG Lei, WANG Wei, LI Yu-xiang   

  1. (School of Computer Science and Engineering, Xi’an University of Technology, Xi’an 710048, China)
  • Online:2011-03-05 Published:2012-10-31

摘要: 传统的模糊c均值算法需要提前输入聚类个数,但输入错误的聚类数会产生错误的聚类结果。为此,提出一种基于人工免疫细胞膜型的模糊聚类算法。引入种群规模迭代与模糊聚类迭代相结合的双迭代思路,利用种群规模迭代指导聚类数的自动生成,在每次种群规模迭代中加入模糊聚类迭代,同时将克隆选择、抗体免疫抑制等操作融入计算过程。理论分析与仿真结果表明,该算法能搜寻到正确的聚类个数,具有较好的聚类效果。

关键词: 模糊聚类, 人工免疫, 模糊c均值, 克隆选择, 抗体免疫抑制

Abstract: It is necessary to provide the number of clusters before it is used in the traditional Fuzzy c-Means clustering algorithm(FCM), but error number of clusters will make error result. Aiming at this proplem, this paper presents a novel clustering algorithm based on artificial immune cell model. With an idea of dual iteration of a combination of population size iteration and fuzzy clustering iteration, this algorithm inducts auto-formation of clusters number by use of the population size iteration and solves the problem of inputting the number of clusters in advance. The fuzzy clustering iteration which is introduced in each iteration of population size, adds clonal selection as well as antibody immune suppression operation. Theoretical analysis and simulation results show that the algorithm can get correct number of clusters, and get better clustering effect.

Key words: fuzzy clustering, artificial immune, fuzzy c-means, clonal selection, antibody immune suppression

中图分类号: