Development of open and non-homogenous network data envelopment analysis model

In this paper, data envelopment analysis (DEA) was discussed for non-homogenous mixed network structure where the outputs of each stage can enter the next stage in the same or another layer as inputs or come out from the system as final products. In this model, in addition to intermediate inputs, each stage can also have independent inputs. For this purpose, a mathematical model has been developed in which the independent inputs and final outputs for the components of the decision making units (DMUs) are examined. To show the efficiency of the model, real data for 20 DMUs were used and the results were compared with those of traditional models. The results obtained from the proposed method eliminate the shortcomings of traditional methods for proper detection of units that are on the efficient border (with an efficiency of one). The developed method can provide the managers and decision makers with a more accurate understanding of the performance of components of DMUs.