客户分类与银行理财——关联分类算法在银行理财营销中的应用研究

Customer Segmentation and Banking Wealth Manangement——The Application of CBA in the Marketing of Banking Wealth Management Products

关联分类是数据挖掘中一种新的分类方法,集成了关联规则挖掘和分类的特点,具有较高的分类精度和较强的扩展性。本文针对银行理财产品营销,通过改进CBA算法,引入加权支持度以及加权置信度进行关联规则的挖掘,然后综合考虑这些规则与测试数据之间的最小差异度,并以此作为依据对客户的理财能力进行分类,为银行有效营销理财产品,提高收益,实现客户财富增长提供了参考。与原有算法相比,改进CBA算法虽然对于普通客户来说,分类精度略有下降,但是黄金客户和成长客户的分类准确性都有了较大的提高,对于开展针对性强的理财营销有一定现实意义。

Classification Based on Association （CBA） is a new classification method of data mining, which integrates the characteristics of mining and classification of the Association Rules and possesses high classification accuracy and strong ex- pansibility. In view of the marketing of personal wealth management products and by improving CBA, the paper introduces weighted support and weighted degree of confidence to the mining of the Association Rules and comprehensively considers the minimum diversity between the rules and the test data. On the basis of the minimum diversity, the paper assorts the cus- tomers＇ abilities of wealth management, which provides reference for effectively marketing wealth management products, im- proving earnings and increasing customers＇ wealth. In contrast to the original method, the accuracy of the improved CBA rel- atively decreases in identifying general customers but significantly increases in identifying the golden customers and the growing customers, whicb is consistent with the purpose of data mining, i.e. more exactly identifying the customers with high contribution, and is conductive to launching wealth management marketing with clear aim.