企业-客户信号博弈及遗传算法的比较研究

针对客户关系管理中企业和客户之间合作博弈与非合作博弈相结合的特点，建立了由广告投入水平、产品价格、服务水平、咨询投入水平、当期购买数量和满意程度6个信号构成的企业一客户信号效用模型，并分别利用两种方法对其进行了求解。第一种方法根据博弈论中的逆推归纳法求出了该模型的均衡行动解，以解释企业和客户在交易中的理性选择行为；第二种方法构造了求解该模型的遗传算法，以求得企业和客户总体效用的最优解，以及构成最优效用的信号值，并利用Matlab对该优化过程给予了实现。这两种方法从不同的角度和出发点对客户关系管理环境下企业和客户的行为进行了解释和比较。

Comparative study of business-customer signaling game and its genetic algorithm

According to the characteristic of combination between cooperate game and non-cooperate game among business and customer in Customer Relationship Management (CRM), a business-customer signaling utility model was developed, which was composed of six signals such as advertisement input level, product price, service level sent by the business, consultation input level, purchase quantity in one time and satisfactory degree sent by the customer. The model was solved by two methods. One was to get the equilibrium by applying Backwards Induction, to explain the rational selections of business and customer. The other was to get the optimal resource configuration by applying genetic algorithm. The two methods explained and compared the activities of business and customer in CRM from the different standpoints.