基于蚁群算法的数据挖掘方法研究  

The Algorithm Based on ACA and Clustering Algorithm Combination

作　　者：张群[1] 熊英[1] 黄庆炬[1] 

机构地区：[1]湖北工业大学计算机学院,湖北武汉430068

出　　处：《湖北工业大学学报》2007年第2期5-9,共5页Journal of Hubei University of Technology

摘　　要：在研究了基本蚁群聚类模型、信息熵以及几个经典的聚类分析算法的基础上,针对传统K-means算法的不足,首先提出了一种基于信息素的k-means改进算法,该算法以基于信息素的转移概率为判断标准来进行聚类,减少了算法的参数个数,加快了聚类的进程.在深入研究了基于信息熵的LF改进算法的基础上,提出了一种蚁群聚类组合算法策略.As one of the most important domain of data mining, clustering analysis is mainly used to discover the valuable data distribution and data mode in the potential datum. On the basis of studies on basic clustering model, the theory of information entropy and two classical clustering analysis algorithms, an algorithm of K-means based on the pheromone are presented firstly. The algorithm works with the transformation probability to realize the clustering, which can reduce the number of the parameters and improve the speed of clustering.

关 键 词：数据挖掘 蚁群算法 K-MEANS算法 

分 类 号：TP311.12[自动化与计算机技术—计算机软件与理论]