Data Envelopment Analysis with Sensitive Analysis and Super-efficiency in Indian Banking Sector

Data envelopment analysis (DEA) is non-parametric linear programming (LP) based technique for estimating the relative efficiency of different decision making units (DMUs) assessing the homogeneous type of multiple-inputs and multiple-outputs. The procedure does not require a priori knowledge of weights, while the main concern of this non-parametric technique is to estimate the optimal weights of inputs and outputs through which the proper classifications of DMUs are possible. DMUs classification with DEA has many challenges in the case of volatility in the values of inputs and outputs. Sensitivity classifications (either efficient or inefficient) as well as returns to scale (RTS) classification (CRS, IRS and DRS) of DMUs are the prominent and vital challenges in DEA studies. Flexible and feasible convex regions with changing values of the reference units from the reference set of inefficient DMUs. This paper has proposed the issues of sensitivities regarding the above mentioned classifications of DMUs and assessing the technical efficiencies by using SBM case of DEA models. Super-efficiency is estimated in case of input and output slacks approach measure and ranking was mad as per the super-efficiency score. Validity of the proposed model is carried with the suitable numerical illustration.