基于模糊聚类的不同形式偏好信息的集结方法研究

针对具有5种不同形式偏好信息的群决策问题给出了一种分析方法.当专家给出的偏好信息是模糊互补判断矩阵、区间值、正互反矩阵、序关系值以及效用值时,首先把不同形式的偏好信息转化为模糊互补判断矩阵,然后,再根据模糊互补判断矩阵得出每个专家的方案排序值,据此对专家进行模糊聚类,根据聚类结果确定专家的权重,进而进行信息合成和方案选优,并用算例进行了验证.

Research on Aggregation Method with Different Forms of Preference Information Based on Fuzzy Clustering

This paper presents an analytical method of group decision making problem with 5 different forms of preference information,such as fuzzy complementary judgment matrix,interval value,positive reciprocal matrix,the order relation value and utility value.Firstly,the different forms of preference information are transformed into fuzzy complementary judgment matrix.Secondly,the expert ranking value is calculated according to the fuzzy complementary judgment matrix.Then,the experts are categorized,and the expert's weight is determined according to the clustering results.Lastly,the information is synthesized and the alternatives are ranked.A numerical example is illustrated to show the feasibility and effectivity of the developed model and method.