首页 | 本学科首页   官方微博 | 高级检索  
     

基于扩散信息素的蚁群聚类算法及应用
引用本文:刘平,陈治平,林亚平,胡玉鹏.基于扩散信息素的蚁群聚类算法及应用[J].微计算机信息,2010(15).
作者姓名:刘平  陈治平  林亚平  胡玉鹏
作者单位:湖南大学软件学院;长沙学院信息与计算科学系;
摘    要:本文提出一种基于扩散信息素模型的全局收敛蚁群聚类算法,设计新的信息素更新机制与概率转移机制,适用于复杂的数据集分析。实验结果表明,新算法在聚类效果上比基本的蚁群聚类算法有较明显的改善。最后将新算法应用于电信运营商的客户数据分析中,用于建立客户细分聚类模型,对复杂客户数据集进行分类,取得了较理想的效果。

关 键 词:蚁群算法  聚类分析  信息素扩散模型  客户分类  

A Pheromone Diffusion based Ant Colony Clustering Algorithm and Applications
LIU Ping CHEN Zhi-ping LIN Ya-ping HU Yu-peng.A Pheromone Diffusion based Ant Colony Clustering Algorithm and Applications[J].Control & Automation,2010(15).
Authors:LIU Ping CHEN Zhi-ping LIN Ya-ping HU Yu-peng
Affiliation:LIU Ping CHEN Zhi-ping LIN Ya-ping HU Yu-peng(School Of Software,Hunan University,Hunan 410082,China)(Department Of Information , Computing Science,Changsha University,Hunan 410003,China)
Abstract:This paper proposed a pheromone diffusion model based global converging ant colony clustering algorithm(PD-CACCA),which not only designs the pheromone renewing mechanism but also the probability transferring scheme.PD-CACCA is suitable for analysis of complex data set.Experimental results show that the PD-CACCA algorithm can achieve distinct improvements compared to basic ant colony clustering algorithms.At last,the PD-CACCA is applied to analyze the customer data set for telecom operators,for building cust...
Keywords:ant colony algorithm  clustering  pheromone diffusion model  customer classifying  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号