Data envelopment analysis with common weights: the weight restriction approach

The conventional data envelopment analysis suggests each decision-making unit selecting its most desirable weight. Applying these weights lets the units achieve their maximum performance. But, the performance of different units is achieved with different sets of weights. So, comparison and ranking of units on a common basis seems such an impossible challenge. However, the flexibility in choosing weights will make more than one efficient unit to be claimed as an efficient unit. In order to resolve these shortcomings, this paper proposes a method that only one common set of weights is obtained through this method. Toward this end, firstly, the efficiency of each unit is calculated, and then, the units are ranked by the efficiency scores earned from common weights. The weight restriction approach here not only generates positive weights but also prevents weights dissimilarity. The production of strictly positive weights through the proposed model makes it possible that no input and output variables are ignored.