Imprecise DEA Model Based on TOPSIS

Abstract

In view of the defect that TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) can not deal with imprecise data properly in multiple criteria decision making. And the weights are often hard to reflect the fact because of subjective preference of decision makers. We proposed a hybrid model combines TOPSIS and imprecise DEA model to improve the disadvantages. Ideal DMU and anti-ideal DMU were built, and the corresponding positive ideal point and negative ideal point were established. Imprecise DEA was set up and an upper-lower limit method is utilized to formulate the imprecise efficiency scores of the two hypothetical DMUs. Based on the distances between DMU0 and the two hypothetical DMUs respectively, imprecise relative closeness was formulated to rank the superiority of all DMUs. The imprecise DEA model based on TOPSIS can avoid too subjective weights and make the evaluation more rational. A numerical example indicated the efficiency of the hybrid measure.