Smoothed frontier to determine a single set of weights in CCR models

Abstract

Over time, the number and range of applications in the DEA have increased greatly. The resulting complexity of the relationships between the multiple inputs and outputs of decision making units (DMUs) in a study hinders the use of other methodologies. This makes DEA quite attractive for identifying the performance of these DMU's. The aim of this study is to define a method of solving the ambiguity in determining the weights of extremely efficient DMUs so that each DMU has a single set of weights, allowing for determination of the relative importance of each input or output. The proposal is to replace the original CCR DEA frontier, which is piece-wise linear, with a smoothed one, which is as close as possible to the first but continuously differentiable at all points. Smoothing allows for the calculation of trade-offs in the extreme-efficient DMUs and in Cross-Evaluation, among other applications.