决策支持下的不完全偏好信息多目标对策  

The Research for Multi-criteria Game with Imperfect Preference Information

作　　者：钟麟[1] 倪世宏[1] 钟卫[2] 张鹏[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038 [2]江西财经大学现代经济管理学院,江西南昌330013

出　　处：《数学的实践与认识》2013年第22期120-127,共8页Mathematics in Practice and Theory

摘　　要：针对多目标对策的不完全偏好信息,将理性策略概念和偏好规划理论引入到多目标对策研究中,把两人的多目标对策考虑为两个局中人同时面临的两个多目标决策问题,偏好信息用可行权重限定集合进行建模.多目标对策模型分为对策模型和决策模型两部分.在对策模型中,利用理性和偏好的共同知识,获得理性策略集.在决策模型中,局中人将策略选择作为使用私人偏好信息的多目标决策问题.最后,通过实例验算表明所提方法的有效性.In order to develop a multi-criteria game model that takes into account the imperfect information about preferences of players, the concept of rational strategies and the preference programming theory are introduced to this model. The imperfect preference information is modeled by the sets of possible weights. The multi-criteria game model is divided into two parts. First, a multi-criteria game model is used to derive conclusions based on common knowledge about rationality and preference, the solution concept for this game is sets of rational strategies. Second, a player analyzes the strategy selection as a multi- criteria decision problem based on private information about his preferences. Last, the model is verified by using an example.

关 键 词：多目标对策 不完全偏好信息 偏好规划 理性策略 

分 类 号：C934[经济管理—管理学]