Estimation of efficiency scores with perturbation in data: an application for provincial gas companies in Iran

Abstract

One of the important problems for estimating the efficiency of decision making units (DMUs) is the existence of perturbations in the data. In this paper, for performance evaluation and ranking of DMUs with uncertain data, an integrated approach is presented based on robust data envelopment analysis (RDEA), stochastic frontier analysis (SFA) and principal component analysis (PCA). To calculate the efficiency of DMUs with perturbation in data, parametric (SFA) and non-parametric (RDEA-BCC) approaches are used. First, the efficiency scores are estimated from different RDEA-BCC and SFA models and then, the final scores and ranks are calculated by using PCA model. To reduce the inconsistency between the efficiencies generated by different SFA and RDEA models, the scores are considered as indicators for PCA model. The proposed model of this paper is implemented by using some actual data from provincial gas companies in Iran.