Determine the Improvement Radius for an Extreme Efficient Unit by Efficiency Preserving of Other Extreme Units

Sensitivity and stability analysis is one of the most critical issues in data envelopment analysis (DEA) that states ‘what extent perturbations in the data can be tolerated before the DEA efficiency is changed’. The current paper provides a new approach for the sensitivity analysis of a specific extreme efficient decision-making unit (DMU) using new DEA models under variable returns to scale (VRS) technology. By these modified DEA models, we can determine the maximum decreasing changes in inputs and the maximum increasing changes in outputs for the extreme DMU so that the efficiency classification of all extreme DMUs remains unchanged. In our study, the data changes are only applied to the DMU under consideration while all the other DMUs remain fixed. The optimal solution of these models yields an improvement radius for the DMU under test. Finally, a numerical example illustrates the approach.

کلیدواژه ها:

Data envelopment analysis ، Sensitivity analysis ، Stability region ، Improvement radius.