摘要
针对方案属性值为随机变量、属性权重未知的风险型多属性群决策问题,根据决策者有无先验信息、专家估计方差是否可知等情形,提出了两类集结决策者和专家主观概率的多元Bayes模型。这些模型先将群体对属性值的估计集结成单一分布,然后用Monte Carlo模拟的方法,通过计算每个方案的排名期望值,得到各方案的排序。给出了应用该方法的具体步骤,实例分析验证了方法的有效性和实用性。
Aiming at the risky group decision making problem with a stochastic variable and unknown weight,two models of combining decision maker＇s and expert＇s subject probability are presented based on Bayesian theory,according to whether the decision maker has prior information and whether the expert＇s deviation is known.With those multivariate Bayesian aggregation models,a consensus probability distribution of each criterion＇s value is attained first.Then,by means of Monte Carlo simulation,a rank probability exception index is calculated which represents the probability of different ranks an alternative might get.Finally,an exception ranking score is defined as well,which can aid a decision maker to choose the best alternative.The solution procedure is given and the effectiveness of this approach is showed by an example.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2010年第5期971-975,共5页
Systems Engineering and Electronics
基金
国家自然科学基金(70971139
70921001)
湖南省自然科学基金(09JJ3130)资助课题
