An Integrated DEA Ranking Method Based upon Benchmarking Sets

Zha Yong,Liang Liang,Xu Chuanyong,Yang Feng

2007-01-01

Abstract:Ranking method is a promising application in management science, which is under increasing interest in the context of DEA. Rank- ing approach has been actively sought by the researchers to improve the potential capabilities and to fully rank efficient, as well as ineffi- cient, decision making units. Distinguishing from the other researches, this article proposes integrated DEA ranking methods considering the preference of decision makers and management practice. Since the performance and process of some DMUs are widely recognized and viewed as the best practice by the decision makers, we include the preferable DMUs in a special reference set called benchmarking set (BS). Our problem is to find common weights, putting the best practice and preferable information of decision makers into consideration, as well as concerning about the efficiencies of other DMUs, which is more convincible and feasible and can easily be accepted by all DMUs than that derived from traditional DEA models. We expect the common weights resulted from the proposed integrated DEA models could guarantee those DMUs in BS DEA efficient. A numerical example concerning the Chinese universities is provided to illustrate the feasibility and convincement of the models.