计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (24): 111-114.

• 数据库、信号与信息处理 • 上一篇    下一篇

一种新的基于进化计算的聚类算法

张俊溪1,吴晓军2   

  1. 1.西安航空技术高等专科学校 生物医电工程学院,西安 710077
    2.西北工业大学 自动化学院,西安 710072
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-08-21 发布日期:2011-08-21

New clustering algorithm based on evolutionary computation

ZHANG Junxi1,WU Xiaojun2   

  1. 1.Department of Biomedical Engineering,Xi’an Aerotechnical College,Xi’an 710077,China
    2.College of Automatic Control,Northwestern Polytechnical University,Xi’an 710072,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-08-21 Published:2011-08-21

摘要: 聚类是数据挖掘领域的重要研究内容之一。针对遗传聚类算法较好的稳定性与粒子群优化算法较强的局部搜索能力,在交叉、变异算子后叠加粒子群优化算子的方法实现了二者的结合,提出了GAPSO聚类算法,既保持了遗传算法的稳定性与泛化性的优势,又发挥了PSO算法收敛效率高的特点。通过对10组二维空间上的聚类样本进行实验研究显示,GAPSO聚类算法在收敛效率上显著优于GA聚类算法,在稳定性上优于PSO聚类算法。

关键词: 数据挖掘, 聚类, 遗传算法, 粒子群优化算法, 遗传粒子群优化算法(GAPSO)

Abstract: Cluster analysis which plays an important role in data mining,is widely used.It has important value both in theory and application.Considering the stability of the genetic algorithm and the local searching capability of particle swarm optimization in clustering,the two algorithms are combined.Particle swarm optimization operators are implemented after the crossover and mutation operators,and GAPSO clustering algorithm is put forwarded.Simulation results are given to illustrate the stability and convergence of the proposed method.GAPSO is proved to be easier to carry out,faster to converge and more stable than other methods.

Key words: data mining, cluster, genetic algorithm, particle swarm optimization, Genetic Algorithm Particle Swarm Optimization(GAPSO)