A Novel Robust-DEA Approach for Efficiency Measurement of Heterogeneous Hybrid Networks under Uncertainty

Abstract

Over the past two decades, the use of network data envelopment analysis in real-world issues has attracted the attention of many researchers. This analysis is used when the output results of a decision unit are used as input to later units. Uncertainty in the inputs and outputs of each decision unit complicates the performance evaluation of such systems. In this paper, a new model of heterogeneous hybrid network data envelopment analysis is developed to measure the efficiency of decision units assuming the open structure of each decision unit, as well as the existence of interlayer relationships. In this model, constraints are defined in such a way that the number of units on the efficient boundary is limited. As a result, there is no need to use super-efficient models to determine overall performance. A robust approach has been used to deal with uncertainties in inputs, outputs and, interlayers. The application of the proposed model was studied to evaluate the performance of pistachio orchards in Yazd province, Iran, and the results were compared with traditional models. In traditional models, the number of units on the efficient boundary in the first decision unit was 5, and in the second stage was 2, which makes the performance evaluation conditional on the use of super-efficient models. With the implementation of the proposed model, no unit was placed on the efficient boundary. The results were approved by agricultural experts.