A stable multi-criteria decision model based on Markov chain. (September 2022)

• Record Type: Journal Article
• Authors: Fu, Chao; Ding, Xiaoyi; Chang, Wenjun
• Is Part Of: Computers & industrial engineering. Volume 171(2022)
• Publication Date: 2022-09
• Subjects: Multi-criteria analysis -- Markov chain -- Transition probability -- Stable solution -- Management platform supplier selection; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108436
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: A stable multi-criteria decision model is developed based on Markov chain. MCDM is demonstrated from the perspective of the Markov chain. Influence of unknown criterion weights (CWs) on alternative ranking is analyzed. Process of generating a unique stable solution from unknown CWs is designed. Difference between proposed model and other methods is highlighted by simulation. Abstract: This paper aims to generate a stable solution to a multi-criteria decision-making (MCDM) problem with unknown decision parameters. To this end, it proposes a multi-criteria decision model with interval numbers based on the Markov chain. For the construction of the proposed model, MCDM is demonstrated from the perspective of the Markov chain. Under these conditions, optimization problems are constructed for each alternative with a specified ranking based on unknown criterion weights, which are representative decision parameters. By using the optimized criterion weight matrix of each alternative derived from the constructed optimization problems, the transition probability of each alternative between two rankings is designed to form the transition probability matrix of each alternative. The relevant properties are theoretically analyzed. From the eigenvector of such a matrix, the stable ranking value distribution of each alternative is obtained, whose existence and uniqueness are theoretically proven. The stable solution is generated from the stable ranking distributions of alternatives. The proposed model is then used to analyze a management platform supplier selection problem, in which its applicability and validity are demonstrated. The meaningfulness of the proposed model is validated by simulation experiments and case comparison.