outputs estimation and efficiency improvement of two-stage network systems

Although data envelopment analysis (dea) is a mathematic based technique to estimate the efficiency score of decision making unit (dmu)s with multiple inputs and outputs, inverse dea (invdea) is a method to estimate inputs/outputs of dmus with respect to unchanged efficiency score. in the general point of view, dmus are considered as black-box, however there are units, which have internal process, that are called two-stage network systems. this paper proposes method for outputs estimation and efficiency improvement of dmus with two- stage network structure when inputs are revised. finally, a case study is presented to show the differences between proposed methods.