EFMCDM: Evidential fuzzy multicriteria decision making based on belief entropy

Abstract

Multicriteria decision making (MCDM) has become one of the most frequently applied decision making methodologies in various fields. However, uncertainty is inevitably involved in the process of MCDM due to the subjectivity of humans. To address this issue, a novel evidential fuzzy MCDM method, called EFMCDM, is proposed by integrating Dempster-Shafer theory with belief entropy. In particular, each criterion can be modeled as evidence, and all the alternatives compose the frame of discernment in the framework of Dempster-Shafer theory. To generate more appropriate basic probability assignments (BPAs) of the criteria, the EFMCDM method considers both the subjective and objective weighting of the criteria that are leveraged in MCDM problems. Thereafter, the classic Dempster's rule of combination is leveraged to fuse the multiple pieces of evidence into composite evidence. On this basis, the alternatives are ranked to determine the optimal alternative. In addition, the EFMCDM method can quantitatively model uncertainty and help to decrease the uncertainty caused by subjective human cognition to improve decision making. Finally, the rationality, effectiveness, and robustness of the EFMCDM method are demonstrated through experimental evaluations.