A fuzzy-cautious OWA approach with evidential reasoning

Multi-criteria decision making (MCDM) is to make decisions in the presence of multiple criteria. To make a decision in the framework of MCDM under uncertainty, a novel fuzzy - Cautious OWA with evidential reasoning (FCOWA-ER) approach is proposed in this paper. Payoff matrix and belief functions of states of nature are used to generate the expected payoffs, based on which, two Fuzzy Membership Functions (FMFs) representing optimistic and pessimistic attitude, respectively can be obtained. Two basic belief assignments (bba´s) are then generated from the two FMFs. By evidence combination, a combined bba is obtained, which can be used to make the decision. There is no problem of weights selection in FCOWA-ER as in traditional OWA. When compared with other evidential reasoning-based OWA approaches such as COWA-ER, FCOWA-ER has lower computational cost and clearer physical meaning. Some experiments and related analyses are provided to justify our proposed FCOWA-ER.