基于蚁群智能的客户群偏好分析方法  

Analysis Method of Customers′ Preference Based on Ant Swarm Intelligence

作　　者：王卫平[1] 杨杰[1] 

机构地区：[1]中国科学技术大学商学院,安徽合肥230026

出　　处：《管理科学》2005年第4期54-57,共4页Journal of Management Science

摘　　要：采用蚁群智能的数据挖掘方法,根据客户对产品属性的偏好评分,对客户进行有效的分类。将客户偏好作为n维空间中的一个点,运用基于蚁群智能的聚类方法,根据预先设定的群体相似系数和拾起或放下概率进行聚类分析,在空间中采用递归算法,以获得不同偏好特征的客户群体模式。Confronted with the harsh competition in the market, whoever understands the preferences of customers, will stand out and win. This paper uses ant swarm intelligence to classify customers on the basis of their evaluation of product attributes. Flatly, suppose that customer preference is a point in N-Dimensional space. Then by applying the method of ant swarm intelligence, classify customers on the basis of the presupposed group resemblance coefficient and the drop-pickup probability. At last, use recursive arithmetic in the space to educe the customer groups of different preferences.

关 键 词：蚁群智能 客户偏好 聚类分析 群体相似度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]