Multi-criteria decision analysis without consistency in pairwise comparisons. (June 2022)

• Record Type: Journal Article
• Authors: Carpitella, Silvia; Inuiguchi, Masahiro; Kratochvíl, Václav; Pištěk, Miroslav
• Is Part Of: Computers & industrial engineering. Volume 168(2022)
• Publication Date: 2022-06
• Subjects: Decision-Making Theory -- Decision Support Systems -- Analytic Hierarchy Process -- Intransitive Preferences -- Skew-symmetric Bi-linear Representation; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108089
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: A novel MCDM approach is proposed to treat complex inconsistent problems. The AHP is combined with the SSB representation of preferences into a new model. A real industrial case is solved to demonstrate the effectiveness of the approach. Abstract: Life itself is colorful and brings situations where making the right decision is a matter of compromise given the various criteria, often conflicting with each other. To handle such situations, a plethora of mathematical methods supporting decision-making has been developed. A little attention has been paid to cases where either criteria or expert preferences are not transitive by nature. Usually, standard decision-making methods handle such a case as an input error (input inconsistency). Being designed for consistent cases, standard methods may conclude in wrong results. We present a novel framework aimed at dealing with inconsistent preferences, without forcing experts to reconsider their initial judgments thus distorting their spontaneous assessments. A simulation analysis has been led to check the methodological validity of our proposal. Specifically, by setting different consistency ranges, thousands of experiments on simulated matrices confirm that our framework represents a valid alternative to the traditional practice. The applicability of the proposed approach has been eventually demonstrated through a real-world case study focused on supply chain management of a relevant industrial problem.