Executive Summary : | This project aims to develop novel models for stochastic data envelopment analysis (SDEA) involving random inputs and outputs data variables for the decision-making units (DMUs). Unlike the conventional data envelopment analysis (DEA), where the entire data is deterministic, the SDEA incorporates random noise or errors in reference technology, resulting in stochastic optimization models. One needs to model dependency among the stochastic variables. One such tool for doing so is through the copula theory. In this project, we aim to formulate new models for both the SDEA and InvSDEA set up in different underlying architecture (like network multi-stage DEA) and possess various features like undesirable outputs or inputs. We want to validate the proposed models by carrying their performance analysis on datasets from financial entities, especially banks, NBFC, or life insurance. We believe that our studies will unveil a holistic presentation in the performance evaluation of entities possessing data uncertainty represented by stochastic variables. |