Abstract
This paper provides a new structure in data envelopment analysis (DEA) for assessing the performance of decision making units (DMUs). It proposes a technique to estimate the DEA efficient frontier based on the Arash Method in a way different from the statistical inferences. The technique allows decisions in the target regions instead of points to benchmark DMUs without requiring any more information in the case of interval/fuzzy DEA methods. It suggests three efficiency indexes, called the lowest, technical and highest efficiency scores, for each DMU where small errors occur in both input and output components of the Farrell frontier, even if the data are accurate. These efficiency indexes provide a sensitivity index for each DMU and arrange both inefficient and technically efficient DMUs together while simultaneously detecting and benchmarking outliers. Two numerical examples depicted the validity of the proposed method.
Correction
In the version originally published online figure 5 was mistakenly duplicated from figure 4 and there were small labelling errors in figures 1, 3 and 13. These have now been corrected in this final version
Correction
In the version originally published online figure 5 was mistakenly duplicated from figure 4 and there were small labelling errors in figures 1, 3 and 13. These have now been corrected in this final version
Acknowledgements
We are truly thankful to the two great referees for their important and useful suggestions to improve this paper significantly.