蚁群与遗传算法融合的聚类算法研究  

Research on clustering algorithm based on fusion of ant colony and genetic algorithm

作　　者：朱峰[1] 陈莉[1] 

机构地区：[1]西北大学信息科学与技术学院,陕西西安710127

出　　处：《西北大学学报（自然科学版）》2009年第5期745-749,共5页Journal of Northwest University（Natural Science Edition）

基　　金：陕西省自然科学基金资助项目(2003F23)

摘　　要：目的通过将蚁群与遗传算法融合,以解决蚁群聚类算法参数众多且与问题域相关,搜索容易出现停滞现象等问题。方法将主要影响蚁群聚类算法性能的5个参数作为遗传算法中的染色体进行编码。首先设计遗传算法的选择、交叉、变异算子,进而将用于聚类结果评价的F-measure函数作为适应度函数,通过多次迭代找出最优的参数组合。结果在仿真实验中,获得了较好的聚类效果。结论蚁群与遗传融合的聚类算法较蚁群聚类算法有更大的优势。Aim To present a clustering algorithm based on fusion of ant colony and genetic algorithm for overcoming the problems that ant clustering algorithm has many parameters and associated with the problem domain, prone to stagnation and other issues. Methods The proposed algorithm encode five parameters which influence the performance of ant clustering algorithm as chromosomes, design the selection, crossover and mutation operator, then use the F-measure function which is used for evaluate the results of clustering as fitness function. Eventually, the optimal combination of parameters is found through multiple iterations. Results Simulation experimental results show better clustering results of the proposed algorithm. Conclusion Clustering algorithm based on fusion of ant colony and genetic algorithms outperforms ant clustering algorithm.

关 键 词：蚁群聚类 遗传算法 算法融合 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]