Identifying a reference set using a restricted weights non-radial data envelopment analysis model

This research theoretically explores the set of efficient DMUs on the supporting hyperplane more formally, using a restricted weights non-radial data envelopment analysis (DEA) model. A range-adjusted measure (RAM) is used a representative of such non-radial models. To identify an appropriate projection (s), when multiple projections occur, strong complementary slackness conditions (SCSC) is introduced to the model. Generally, the addition of weight restrictions to models makes two difficulties as follows: 1)It is recognized that weight restrictions in DEA often leads to unacceptable results in computation of relative efficiency. 2)Restricted weights DEA models frequently encounter problems of inefficiency. However, the consideration to weight restrictions is affected the manager’s point of view and also decision making units (DMUs) are more accurate under the assessments. We then apply the combinational models to a real world data set and compare their solutions.