Uncertain SBM data envelopment analysis model: A case study in Iranian banks

Data envelopment analysis (DEA) is a strong analytical tool and methodology for evaluating the relative efficiency of decision‐making units (DMUs). The DEA models require inputs and outputs, which are equipped with precise information. However, the real‐world inputs and outputs are probably so changeable and complicated that cannot be measured accurately. Consequently, this conflict leads to the analysis of uncertain DEA models. This paper aims to analyze the slacks‐based measure (SBM) model in an uncertain environment where the uncertain inputs and outputs are belief degree‐based uncertainty. The belief degree‐based uncertainty is useful for the cases in which no historical information of an uncertain event is available. As a solution methodology, the uncertain SBM model is converted to a crisp form by using three approaches separately: expected value model, expected value and chance‐constrained model, and dependent chance‐constrained model. Additionally, an alternative uncertain model is introduced to reveal differences of DMUs rankings in three crisp methods. Finally, an applied scenario regarding the Iranian banking system is documented to present the new models.