隐马尔可夫模型在客户关系动态研究中的应用

发布时间：2018-04-03 01:25

本文选题：隐马尔可夫模型　切入点：客户动态关系　出处：《商场现代化》2007年05期

【摘要】：客户关系管理(CRM)已经成为市场营销研究领域中的一个重要方面,其研究的方向大致有两个:一是以客户价值为核心的客户关系管理的应用研究;二是以客户行为为基础的客户关系动态研究。后者的研究更加抽象和一般化,其采用的主要工具是马尔可夫过程。但客户关系状态的不可观测性却不能够在单纯的马尔可夫过程得到解决,隐马尔可夫模型把马尔可夫过程和离散选择模型结合起来,用可以观测的客户购买行为来反向推断隐含的客户关系状态,从而推动了客户关系动态研究的进展。本文介绍了隐马尔可夫模型拟合客户动态关系的原理,并且用R语言对这一拟合过程进行了蒙特卡罗实验。
[Abstract]:CRM (customer relationship Management) has become an important aspect in the field of marketing research. There are two research directions: one is the application research of customer relationship management with customer value as the core;The second is the dynamic research of customer relationship based on customer behavior.The research of the latter is more abstract and general, and its main tool is Markov process.However, the unobservability of customer relationship can not be solved in the simple Markov process. The hidden Markov model combines the Markov process with the discrete selection model.Using observable customer purchase behavior to deduce the implicit customer relationship state in reverse, which promotes the research progress of customer relationship dynamics.This paper introduces the principle of hidden Markov model fitting customer dynamic relationship, and carries out Monte Carlo experiment on this fitting process with R language.
【作者单位】：西南财经大学工商管理学院西南财经大学工商管理学院
【分类号】：F274;F224

【相似文献】