The Study Of Classifying DMUs And Re-ranking Efficient DMUs And It's Application

Selecting DMUs and Re-ranking Efficient DMUs are two key problems of DEA. When evaluating the efficiency of DMUs by DEA,only if accurately selecting DMUs and re-ranking efficient DMUs, could get scientific result and valuable information.On Selceting DMUs, in order to retain the similarity of DMUs in the population of observations, should classify the DMUs before evaluating .The existing study either just present abstract ideas,or only solve partial problem.In order to get an all-around solution, we present the qualitative and quantitative approaches of classifying the DMUs, and propose the conditions when to ues qualitative and quantitative approach.On Re-ranking Efficient DMUs,we try to study from a new point.Because of the differences of input and output(reflact on the data of input indexes and output indexes ),there must be distinctions between the Efficient DMUs.The relation between the indexes that lead to this distinctions is grey relation,so we study it by grey relation analysis, calculate the Slope Relative Degree of the Efficient DMUs,and use the results to rank the Efficient DMUs.Finally, for validating the approches of Sellecting DMUs and Re-ranking Efficient DMUs we proposed , we construct an index system.for fund evaluating, and then investigate the perfermance of 46 open-ended funds in 2004,and get satisfactory result.