Fuzzy weak link approach to the two-stage DEA

Abstract

This paper refers to a recent approach to two-stage DEA called the weak link approach. It underlines the lack of solution uniqueness in this approach to DEA and the fact that in order for the solution to the weak link approach to be unique, the decision maker needs to express a preference on which Pareto solution would be most satisfactory. In this paper, we propose to use a fuzzy set approach called fuzzy bicriterial programming to help the decision maker to express this preference. Fuzzy bicriterial programming is explained and then applied to the weak link approach to the DEA. It is shown that for each candidate (Pareto) solution to the original weak link approach, there exists an expert opinion that can lead to the unequivocal selection of this solution due to the use of the fuzzy approach. The proposal is illustrated with examples.