客户关系管理中的混合数据聚类应用研究

聚类分析是客户关系管理（CRM）中非常重要的工具之一．提出一种新的算法，k-centers，通过重新定义相似度衡量和聚类中心更新方法，能够有效处理混合类型的CRM数据。以解决经典算法比如k-means等只适用于数值类型数据的不足，实验分析说明k-tenters算法能够准确把握不同类型客户的特征，为个性化的市场营销提供有力支持。

Study on the Application of Mixed-type Clustering in Customer Relationship Management

Cluster analysis is a very useful technique in customer relationship management(CRM).In this paper,a new clustering algorithm,k-centers,is proposed.It employs a new similarity measure and a new method on updating the centers.Thus,it removes the limitation of the k-means algorithm which is only efficient on numerical datasets,and it can efficiently handle mixed-type data in CRM.Experiments on a real-life dataset show that the k-centers algorithm can hold the characteristics of different kinds of customers,which provides great support for customized marketing.