摘要
通过一种改进的数据包络分析(DEA)模型———SE DEA模型,可以对决策单元进行分类,但在可变规模收益(VRS)条件下,这种模型可能不存在可行解.论文讨论了基于VRS的SE DEA模型不可行的充分必要条件,分析了有效决策单元的超有效和强超有效特性,实现对有效决策单元的进一步分类.最后给出了用基于VRS的SE DEA模型对决策单元进行分类的方法步骤,该分类方法改进和改善了传统的决策单元分类,可为决策者提供更加完备的决策信息.
Superefficiency data envelopment analysis (SEDEA)model,an improved DEA model, can be used to classify (DMU). But the SEDEA model under variable returnstoscale may be infeasible. This paper gives the sufficiency and necessity of the infeasibility of the SEDEA model. Meanwhile,the super efficient DMU and the strongly super efficient DMU are defined. Then the method of using SEDEA model to classify DMUs is put forward. This method is an improvement on the traditional method and can supply more information for decisionmakers.Therefore,it is practically useful.
出处
《天津大学学报（自然科学与工程技术版）》
EI
CAS
CSCD
北大核心
2003年第4期522-524,共3页
Journal of Tianjin University：Science and Technology
关键词
数据包络分析
SE—DEA模型
决策单元
分类
可变规模收益
不变规模收益
强超有效性
super-efficiency data envelopment analysis model
variable returns-to-scale
constant returns-to-scale
super efficient
strongly super efficient
classification
