The group consensus based evidential reasoning approach for multiple attributive group decision analysis
The group consensus based evidential reasoning approach for multiple attributive group decision analysis
复制标题
基于群体共识的多归因群体决策分析证据推理方法
DOI：
10.1016/j.ejor.2010.03.011
复制
发表时间：
2010-11-01
影响因子：
6.4
通讯作者：
Yang, Shan-Lin
中科院分区：
文献类型：
--
作者：
Fu, Chao;Yang, Shan-Lin
许多多属性决策分析问题既包含定量属性又包含定性属性，且具有各种不确定性，如无知、模糊、区间数据、区间置信度等。证据推理（ER）的方法，在20世纪90年代和近年来发展起来的，可以用来模拟这些问题。在本文中，ER方法扩展到群体共识（GC）的情况下，多属性群决策分析问题。为了构造和检查GC，首先开发了两个信念结构之间的相容性度量。考虑到两个专家的效用，他们的评估之间的兼容性自然是使用兼容性度量。基于两个专家评价的相容性，可以在特定的水平，可以是属性水平，替代水平，或全球水平的GC，可以被构建并在指定的时间内通过群体分析和讨论后到达。在GC条件下，研究了方案群体评价的形成、评价等级综合效用的实现以及扩展ER方法的性质和步骤。应用扩展的ER方法解决了一个工程项目管理软件选型问题，展示了其详细的实施过程，以及其有效性和适用性。(C)2010爱思唯尔有限公司版权所有。
Many multiple attribute decision analysis problems include both quantitative and qualitative attributes with various kinds of uncertainties such as ignorance, fuzziness, interval data, and interval belief degrees. An evidential reasoning (ER) approach developed in the 1990s and in recent years can be used to model these problems. In this paper, the ER approach is extended to group consensus (GC) situations for multiple attributive group decision analysis problems. In order to construct and check the GC, a compatibility measure between two belief structures is developed first. Considering two experts' utilities, the compatibility between their assessments is naturally constructed using the compatibility measure. Based on the compatibility between two experts' assessments, the GC at a specific level that may be the attribute level, the alternative level, or the global level, can be constructed and reached after the group analysis and discussion within specified times. Under the condition of GC, we conduct a study on the forming of group assessments for alternatives, the achievement of the aggregated utilities of assessment grades, and the properties and procedure of the extended ER approach. An engineering project management software selection problem is solved by the extended ER approach to demonstrate its detailed implementation process, and its validity and applicability. (C) 2010 Elsevier B.V. All rights reserved.