Abstract
The selection of the optimal supply is an open and crucial issue in supply chain management (SCM), which can be considered as a multi-criteria decision-making (MCDM) problem where the expression and processing of uncertain information could be involved. The purpose of this paper is to develop an elimination and choice translating reality (ELECTRE)-based MCDM method where the evaluation information is expressed and handled by a Dempster–Shafer theory (DST). DST is a primary methodology for uncertainty modeling. In this paper, the weight of the criteria and the performance of each alternative are expressed by linguistic terms and confidence levels, which are then converted to basic probability assignment (BPA) representations. To aggregate evaluations of different experts more rationally and efficiently, a discounting method in DST is presented based on the proposed concept of evidential reliability. In addition, as one family of MCDM models, the ELECTRE method is famous for its outranking relations to rank a set of alternatives. As an extension, synthetic weight, including subjective and objective weights, is applied to determine the concepts of concordance and discordance. The proposed DS-ELECTRE approach not only maintains the advantage of the DST that directly represents and handles uncertainty but also can play the role of the ELECTRE method in analyzing outranking relations among alternatives. An illustrative numerical example is conducted to demonstrate the effectiveness of the DS-ELECTRE method.
Highlights
Supplier selection has recently emerged as an active research field, which plays a crucial role in supply chain management (SCM)
In the supplier selection model based on Dempster–Shafer theory (DST) developed in this paper, the experts’ evaluation of criteria weights and alternatives are expressed by basic probability assignment (BPA), and the reliability-based discounting method we proposed is employed to obtain the decision matrix in the process of information fusion
To model multi-criteria decision-making (MCDM) problem for supplier selection, the study performed in this paper focuses on information characterized by uncertainty that is expressed by BPA in DST
Summary
Supplier selection has recently emerged as an active research field, which plays a crucial role in supply chain management (SCM). In the supplier selection model based on DST developed in this paper, the experts’ evaluation of criteria weights and alternatives are expressed by basic probability assignment (BPA), and the reliability-based discounting method we proposed is employed to obtain the decision matrix in the process of information fusion. Based on the decision matrix obtained by the improved combination rule in DST, in the novel ELECTRE method, the subjective and objective weights are calculated as the synthesis weight, which will be employed in the ranking process. To reduce the impact of uncertainty associated with evaluation information on fusion results, a reliability-based evidential discounting method is proposed in this paper. A new ELECTRE-based method is developed for solving MCDM problems in DST environments, in which decision makers can consider the evaluation itself without formality and can employ imperfect or insufficient knowledge of data. The second one is the Dubois-Prade’s definition of entropy in DST, which denotes the non-specificity measure
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
