The Further Research On The DEA-Method

The theory of data envelopment analysis(DEA) is a new exploring field that Management Science combines with the Operation Research and Mathematical Economics. As a non-parameter method it is particularly suitable for dealing with the multi-input and multi-output evaluating problems of complex system. Because the DMUs'information full of economic meaning can be obtained by application of DEA theory, the DEA method is popular with all over the world. Although the diversity of its theoretical models and flexibility of application endow DEA with high vitality and practicality, it is necessary to develop and improve the DEA method for the sake of applications.Firstly, the efficiency value of the DEA method is useless in distinguishing the DEA—efficient DMUs. In accordance with the shortcoming, the DEA models with the parameters, namely PC2R and PBC2 model, are established. And the threshold conception being able to evaluate all the DMUs is provided. The threshold makes up for the shortcomings of the efficiency value of the DEA method.Secondly, the inputs and outputs of DMUs may be intervals because of insufficient information or measure error. For the reason, a new conception of maximum feasible-production set is proposed. And the interval-DEA model on the basis of the maximum feasible-production set, namely IC2R and IBC2 model, are established. And the interval efficient score being able to evaluate all the DMUs is provided. According to the idea of the extended DEA, the extended interval-DEA model on the basis of the maximum feasible-production set, namely EIC2R and EIBC2 model, are established. And the super-interval efficient score being able to evaluate all the DMUs is provided. The dissertation studies the DEA method by means of the method combining theory with practice. The work will develop the DEA method and theory and increase its practical value. The main contents and conclusions are as follow:1)The Chapter 1 systematicly introduces the main conception and the important conclusion of the DEA and the extended DEA for the dissertation research.2)At the Chapter 2, the the DEA models with the parameters, namely PC2R and PBC2 model, are brought forward. And the relationship between the efficiency value of the DMU and the parameters is discussed, and a new threshold index evaluating all the DMUs is provided.3)At the Chapter 3, because the inputs and outputs of DMUs may be intervals, a new conception of maximum feasible-production set is brought forward. And the interval-DEA model on the basis of the maximum feasible-production set, namely IC2R and IBC2 model, are established. And the interval efficient score being able to evaluate all the DMUs is provided.4)At the Chapter 4, the extended interval-DEA model on the basis of the maximum feasible-production set, namely EIC2R and EIBC2 model, are established. And the super-interval efficient score being able to evaluate all the DMUs is provided. It is parallel to the Chapter 3 in theory and framework.5)Furthermore, in Chapter 5 the empirical study is performed for validating the results from the work.6)Finally, the concluding remarks are presented. It is pointed out that problems are still explored about the DEA. The developing prospect of the DEA is also depicted according to the research effort.