一种新的混合聚类算法  

New Mixed Clustering Algorithm

作　　者：吴文丽[1] 刘玉树[1] 赵基海[1] 

机构地区：[1]北京理工大学,北京100081

出　　处：《系统仿真学报》2007年第1期16-18,共3页Journal of System Simulation

摘　　要：聚类是数据挖掘的主要技术之一,是一种无导师监督的模式识别方式。聚类分析就是按照数据间的相似程度,依据特定的准则将数据划分成不同子类。K-平均算法是经典的聚类算法。蚂蚁聚类算法是近来涌现的新的聚类算法,它通过模拟蚁群的智能行为进行聚类分析,已经在数据挖掘中得到应用。通过分析蚂蚁聚类算法和K-平均算法两种不同聚类算法的基本思想,将两种算法结合得到混合聚类算法,仿真实验证明混合聚类算法的算法性能优于蚂蚁算法和K-平均算法。Clustering is one of primary techniques in the filed of data mining. It is an unsupervised mode of pattern recognition. Clustering analysis is a division of data into similarity groups according to given rules. K-means algorithm is a classical clustering algorithm. Ant-clustering-algorithm which simulates intelligence actions of ant colony has been applied in the data mining. In this paper, a new mixed clustering algorithm was proposed which combined k-means algorithm with ant-clustering-algorithm on the base of the fundamental principle about two methods. The results of experiment demonstrate that this new mixed clustering algorithm has the advantage over the two others in performance.

关 键 词：群体智能 蚂蚁聚类算法 K-平均算法 混合聚类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]