Determination of the Weights for the Ultimate Cross Efficiency Using Expert Scoring Method

摘要

— The theoretical research on the DEA (data envelopment analysis) cross-efficiency evaluation mainly focused on how to select one unique set of inputs and outputs weights among multiple sets of optimal weights for eachdecision making unit (DMU), but payed little attention to how to aggregate efficiencies in cross-efficiency matrix. The commonly used aggregation method is to aggregate them with equal weights without considering their difference. This paper deals with how to aggregate efficiencies in cross-efficiency matrix reasonably and proposes the use of expert scoringmethod to aggregate them. This method views efficiency values of each DMU in cross-efficiency matrix as its efficiency scores determined by different experts. Due to the distinction of different experts on the education background and work experience and other aspects, their efficiency scores to DMUs should be treated differently and allocated variant weights in the final overall assessment. The weight allocated to efficiency scores determined by one given expert is determined by its dissimilarity level to other experts. If the dissimilarity level of one given expert to other ones is huge, his authority will be questioned and it will be reasonable to allocate small weight to his efficiency scores. Finally, the numerical examples examinethe validity of the proposed method.