An Improvement on DEA Model for Many Efficient DMUs

Traditional data envelopment analysis (DEA) model can only evaluate the relative efficiency of decision making units (DMUs), some times it can not rank all the DMUs owing to existing many efficient DMUs possibly. In this article, a new method for ranking DMUs is proposed by improving DEA model. The method introduces the best and worst DMU respectively to change the production frontier, at the same time, the inputs and outputs of virtual decision making unit replace the DMUpsilas in the DEA model. So, two group of efficiency value can be obtained respectively. On the base, a cross efficiency model is given considering the manner and attitude of dealing with this problem. At last, an example testifies that the new method is feasible and effective.