Increasing discrimination of network structure models by MCDEA at presence of undesirable output

In many real world DMUs have a two – stage network scenarios, which have not only inputs and outputs, but also intermediate measures and undesirable factors that exist in-between the two-stage operations. However, we would argue that existing two-stage DEA models use an unrealistic weighting system. The multiple criteria DEA can be used to improve discriminating power of classical DEA method and overcomes weak discrimination between DMUs. In this paper we use a modified multiple criteria two-stage DEA model which yields more realistic weights for the inputs and outputs and thus has better discrimination power than traditional two-stage DEA models at presence of undesirable outputs.