一种基于信息素的蚁群聚类算法  

An Ant Colony Clustering Algorithm Based on Pheromone

作　　者：张建华[1] 赵东东[1] 江贺[1] 张宪超[1] 

机构地区：[1]大连理工大学软件学院

出　　处：《计算机工程与应用》2006年第20期157-159,163,共4页Computer Engineering and Applications

摘　　要：蚁群算法作为一种新型的优化方法,具有很强的适应性和鲁棒性。基于蚁群算法的聚类方法已经在当前数据挖掘研究中得到应用。文章提出了一个新颖策略来解决无人监督的数据聚类问题,利用信息素控制蚂蚁随机移动提高算法效率,采用运动速度各异的多个蚂蚁独立并行进行聚类来提高聚类质量。实验结果表明该方法是有效的。Ant colony algorithms are robust and adaptable as novel optimization methods.The ant-based clustering algorithm has currently applications in the data mining community.This paper presents a novel strategy to tackle unsupervised data clustering problems,in order to improve efficiency of algorithms which takes pheromone to control randomly moving.Each ant takes different types of moving speeds and independently clusters to improve quality of clustering.Results show that this method is impacfful.

关 键 词：蚁群算法 信息素 聚类 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]