The DEA Models And Evaluation Of Efficiency For Economic System

Data Envelopment Analysis,originated by A.Charnes and W.W.Cooper in 1978 is a new systems analysis method based on the conception of"evaluating the relative efficiency". It's a cross field of mathematics, operations research, mathematic economics and management science, which use the mathematical programming to evaluate the relative efficiency of decision making units (DMUs) with multiple outputs and multiple inputs. Thanks to its natural economic background and according to the methods, models and theory of DEA, we can use the input and output immediately to constitute a non-parametric DEA model and analysis it by the view of economics. Moreover, we can find a lot of management information. It makes the DEA attracts many scholars.Studying the DEA model is the base of other research in this area. This research summarizes the background of DEA firstly, expatiates its development of theory and application for the past 30 years. The content of this paper includes:1. Discuss the properties of units invariance and translation invariance. From the two sides of view in optimism and pessimism to evaluate the DMUs and proposes the two-side additive model method which satisfies the two desirable properties simultaneously. Finally, an example and some good results are given to illustrate this method.2. Study the super-efficiency of DEA models. This type of models mainly solves the issue that plural DMUs behave efficient if we use classical DEA models to evaluate the system. From the two sides of view in efficient frontier and anti-efficient frontier to evaluate the DMUs and achieve completely ranking. Applying this approach, the paper examines two numerical examples and proves the rationality of our method.3. Discuss the crossroad of DEA and AHP. Both of these two operations research methods are widely applied in the real world. This research integrates the objectivity of DEA and subjectivity of AHP and presents a new method which uses the idea of solving DEA model to solve the pair wise comparison matrix. Our method is brief, feasible and practicable for use.4. Establish an actual evaluating system and uses classical DEA models and new methods proposed by this research to evaluate the 8 industry clusters of electron and communication. Also, completely ranking is achieved and we present plentiful of management information and suggestion which will be propitious to industry regulation.