基于信号博弈的GA在客户关系管理中的应用

针对客户关系管理环境下企业与客户之间双向信号传递的特点，构造了企业-客户信号博弈模型，以及求解该模型的GA（遗传算法）。基于企业-客户信号博弈的GA是一个求解企业和客户总体效用及其最优行动的工具。在该求解过程中，认为企业和客户是有限理性的，并通过GA中的选择、交叉、变异等操作逐步提高其对环境的适应度，从而最终选择可以为企业和客户带来总体效用最大的行动，即在交易过程中向对方发送可以带来最大效用的行为信号。这种工具为利用逆推归纳法得出的企业-客户信号博弈模型的贝叶斯均衡解提供了比较的依据。

Application of GA Based on Signaling Game in CRM

According to the characters of bi-direction information transferring between business and customer in CRM (Customer Relationship Management), a signaling game model and its GA (Genetic Algorithm) are established. The constructed GA based on the signaling game model is a tool to get the total optimal utilities of business and customer, as well as the signals that compose the total optimal utility. Business and customer are regarded as limit rational. The signals that can result in the optimal total utilities of business and customer are obtained through some operations of selection, over, and mutation. The application of the algorithm provides a comparative reference to the equilibrium of the game model of business-customer obtained by applying Backwards Induction.