The classification-based consensus in multi-attribute group decision-making

Abstract

In multi-attribute group decision-making problem (MAGDM), the existing consensus reaching process (CRP) is to obtain a consensus ranking of alternatives. However, these CRPs contradict some real-life MAGDM problems in which decision-makers do not need to rank alternatives and hope to classify the alternatives into several groups instead. Thus, in this paper we propose a new CRP in MAGDM, called the classification-based consensus reaching process (CCRP). First, we present a feedback method with minimum adjustments to generate the optimal adjusted individual matrices via a 0–1 mixed linear programming model for CCRP. Subsequently, we develop the interactive consensus reaching process based on the feedback method with minimum adjustments in CCRP. Finally, a practical example from China Undergraduate Mathematical Contest in Modeling and a simulation analysis are conducted to demonstrate the validity of the proposed CCRP.

Keywords:

• Multi-attribute group decision-making
• consensus
• feedback method
• minimum adjustments
• classification

Acknowledgement

Weijun Xu would like to acknowledge the financial support of grants (Nos. 71771091, 71471065) from NSF of China, Haiming Liang would like to acknowledge the financial support of grants (No.71601133) from NSF of China, and Yucheng Dong would like to acknowledge the financial support of grants (Nos. 71871149, 71571124) from NSF of China, and the grants (Nos. sksyl201705 and 2018hhs-58) from Sichuan University.

Disclosure statement

No potential conflict of interest was reported by the authors.

Figure 1. The consensus framework based on classifications in MADM.

Figure 2. The $\mathit{AM}$ values under different $q$ and $\alpha .$

Figure 3. The $\mathit{AM}$ values under different $\mathit{EL}$ and $\alpha .$