Modeling heterogeneous multi-attribute emergency decision-making with Dempster-Shafer theory. (November 2021)

• Record Type: Journal Article
• Authors: Fei, Liguo; Feng, Yuqiang; Wang, Hongli
• Is Part Of: Computers & industrial engineering. Volume 161(2021)
• Publication Date: 2021-11
• Subjects: Heterogeneous multi-attribute decision-making -- Emergency decision-making -- Dempster-Shafer theory -- Heterogeneous information -- Medical supplier selection; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2021.107633
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: A heterogeneous multi-attribute decision-making framework is constructed based on Dempster-Shafer theory. The decision framework can reflect human-specific aspects in judgments, specifications or evaluations. Using DST enables decision makers to present information in a way that is more consistent with human cognition. Abstract: Conventional multi-attribute decision-making (MADM) methods, with regard to different attributes, typically require decision-makers (DMs) to use a unified information expression to evaluate the performance of various alternatives. This does not properly reflect the heterogeneity among attributes. This paper presents a novel heterogeneous multi-attribute decision-making approach based on Dempster-Shafer theory (DST). A human-centric heterogeneous multi-attribute emergency decision-making framework is constructed based on Dempster-Shafer structures. The term "human-centric" is emphasized here as the proposed framework encompasses aspects specific to human judgments, specifications, or evaluations. Considering the differences in DM backgrounds and preferences, our framework incorporates a heterogeneous decision environment. DST allows DMs to express decision information in a manner closely consistent with human cognition. A series of parameters capture the subjective attitudes of DMs in our framework, the embodiments of which are manifold: (1) they reveal the sensitivity of DMs to alternative ranks in the process of basic probability assignment generation from decision information; (2) they reflect the tolerance of DMs to uncertainty in the process of weight calculation from belief entropy; and (3) they facilitate analysis of DMs in the decision process based on belief intervals. The proposed approach is demonstrated on a emergency decision-making application for medical supplier selection in the context of COVID-19.