摘要
在基于目标决策理论的基础上,分析决策者目标的不同类型及对应的目标效用函数,并针对决策者目标不同以及决策方案属性值为随机变量的风险型群决策问题,提出基于目标期望效用最大化的群体偏好集结方法。首先,该方法将属性的随机性、目标的随机性及决策者目标的满足程度转化为目标期望效用函数;其次,按期望效用类似的线性集结方式进行个体偏好的集结;最后,根据群体目标期望效用值形成方案排序。算例验证了该方法的有效性和实用性。
Based on target oriented decision theory, this paper analyzes the different targets of the decision makers and their corresponding utility function. For deferent targets and the random variable attributes of the risk group decision problem, a group preference aggregating approach is proposed based on target- oriented theory. By translating the attributes' and decision makers' uncertainty into target expected utility, the preference of each decision maker is achieved firstly. Then a linear approach is used to aggregate expert's preference and form a group target expected utility which can be used to determine the rank of alternatives. An example shows the feasibility and effectiveness of the proposed approach.
出处
《统计与信息论坛》
CSSCI
2009年第3期8-12,共5页
Journal of Statistics and Information
基金
国家自然科学基金重点资助项目<面向复杂大群体的群决策与支持平台>(70631004)
湖南省哲学社会科学基金项目<风险分析中专家主观概率判断的启发
集结与相关性研究>(08YBA022)
关键词
基于目标的决策
风险型多属性决策
群决策
效用函数
target based decision making
multi- attributes decision making under risk
group decision
utility function