摘要
针对多属性多方案的大规模群体决策问题,提出了一种基于公理模糊集(AFS)理论确定决策成员权重的方法。该方法首先利用AFS理论的聚类算法对大群体决策成员进行聚类,得到每一类的语义描述以及每个决策成员隶属于各类的隶属程度。然后根据聚类结果以及各决策成员的个体意见之间的相对差异得出每位决策成员的权重。最后,利用得到的权重计算综合决策矩阵,获得决策方案的最终排序。该方法既避免聚类时阈值选取的主观性,又简化了决策流程。通过实例验证了该方法的可行性和有效性。
The method of determining weights of decision makers for multi-attribute and multi-ahernative large group decision making based on AFS is proposed. Firstly, the decision makers are classified by using AFS cluste- ring method, meanwhile, semantic descriptions of each class and the membership function of decision makers be- longing to each class are obtained. Then, the weight vectors of decision makers are derived by using the clustering result and the difference of opinion between decision makers. Finally, the ordering vectors of decision schemes are obtained by synthesizing the decision makers＇ weight vectors and decision matrices. This method avoids the empiri- cal selection of threshold value when clustering and simplifies the process of decision making. The example analysis verifies that the method is feasible and effective.
出处
《科学技术与工程》
北大核心
2015年第15期61-65,共5页
Science Technology and Engineering
基金
国家自然科学基金(61203283)
四川省人工智能重点实验室开放课题(2012RYJ02
2013RZJ03)
辽宁省自然科学基金(2014025004)
中央高校基本科研基金(3132014324)资助
关键词
多属性群决策
大群体
AFS理论
群体聚类
决策者权重
multi-attribute group decision-making large group AFS theory group clustering de-cision makers＇ weight
作者简介
田晓娟（1974-），女。研究方向：决策分析、模糊机器学习。E-mail：dongziyi@hotmail．com。
