考虑决策者风险偏好的机会约束DEA模型

Chance constrained DEA model considering decision maker's risk appetite

机会约束DEA模型打破了传统DEA模型对决策单元投入、产出的硬性约束条件,允许评价单元在既定的概率约束下超出前沿面,一般从统计角度将此概率设为某个足够小的置信水平.然而不同风险态度的决策者对评价单元具有不同的评价态度,因此,考虑决策者风险偏好的效率评价模型将更贴近现实.本文以机会约束DEA模型为基石,通过分析决策者感知风险与系统可测风险的相互关系,构建了考虑决策者风险偏好的机会约束DEA模型,定义了不同风险态度决策者的单元评价效率.数值分析在验证模型的有效性和可行性的同时,还揭示了考虑决策者风险偏好的单元效率与CCR效率之间的联系.

A chance constrained DEA model breaks the rigid restraints of inputs and outputs on DMUs, allo- wing some DMUs to exceed the production frontier on the assumption of certain probability of stochastic varia- bles （usually a sufficiently small confidence level）. Decision makers have different risk appetites towards dif- ferent DMUs or inputs （ outputs）, so the efficiency evaluation model considering decision maker＇ s risk appe- tite will be more applicable in reality. Based on a Chance constrained DEA model, this paper analyses the re- lationship between the perceived risk of decision makers and the measurable systematic risk, through which the chance constrained DEA model with decision maker＇ s risk appetite is constructed and the definitions of chance constrained DEA efficiency with decision maker＇ s appetite are defined. The numerical analysis not only proves the validity and feasibility of our model, but also reveals the relationship between the efficiencies calculated re- spectively by our model and the standard CCR model.