期刊
APPLIED SOFT COMPUTING
卷 98, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.asoc.2020.106757
关键词
Multi-criteria group decision making; Quantum probability theory; Linguistic distribution assessment; Extended generalized TODIM; Automobile recommendation
资金
- National Natural Science Foundation of China [72071045, 72071151, 71771051, 71701158, 71771001]
- MOE (Ministry of Education in China) Project of Humanities and Social Sciences [17YJC630114]
A new model for multi-criteria group decision making is developed, taking into account the influence of decision makers' opinions on each other and their subjective bounded rationality. This model integrates the extended generalized TODIM method and quantum decision theory to improve decision accuracy and stability through linguistic distribution assessments and a quantum possibilistic aggregation framework.
In most existing multi-criteria group decision making (MCGDM) problems, the decision makers (DMs) are usually regarded as independent and DMs' psychological behaviors are rarely considered. However, in many cases, the opinions of DMs are likely to influence each other and the DMs often hold subjective bounded rationality in the decision making process. To address this issue, a linguistic distribution behavioral MCGDM model integrating extended generalized TODIM (an acronym in Portuguese of interactive and multi-criteria decision making) method and quantum decision theory is developed. First, linguistic distribution assessments considering the sample size information are used to gather the group linguistic evaluations, which can integrally depict the quantitative distribution and qualitative vagueness. Second, an extended generalized linguistic distribution TODIM method is presented to reflect the DMs' bounded rational behaviors. It is used to calculate the dominance of each alternative. Third, a quantum possibilistic aggregation framework is constructed to explore the interference effects among DMs' opinions. In this process, DMs' opinions are viewed as synchronously occurred wave functions that interfere with each other and influence the aggregated result. Finally, a case study is examined to recommend the optimal automobile for customers based on distributed online reviews in newcar.xcar.com and autohome.com. Three separate scenarios with conjunctive, neural, and disjunctive modes are considered respectively. The sensitivity analysis is given to show the effect of risk factors and interference factors. The parameters show to be dominant in the final best automobile recommendation. The comparisons with some existing methods confirm the validity and stability of the proposed MCGDM method. (C) 2020 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据