Traditional efficiency studies on network data envelopment analysis (DEA) consider decision-making units as black boxes that use a set of crisp inputs to produce a set of crisp outputs and that ignore intermediate measures and reverse flows. In real applications, however, we are faced with network systems with reverse flows in an uncertain environment. In this paper, therefore, a chance-constrained multistage DEA model is introduced to analyze the relative performances of supply chains and components in the presence of reverse flows and random factors. A real case in the sugar illustrates the proposed method. The results demonstrate the validity and applicability of the model.

中文翻译：

---

逆向随机供应链的绩效分析：基于 DEA 的方法

关于网络数据包络分析 (DEA) 的传统效率研究将决策单元视为黑匣子，它使用一组清晰的输入来产生一组清晰的输出，而忽略中间措施和反向流动。然而，在实际应用中，我们面临着在不确定环境中具有反向流动的网络系统。因此，在本文中，引入机会约束的多阶段 DEA 模型来分析供应链和组件在存在逆流和随机因素的情况下的相对性能。糖中的一个真实案例说明了所提出的方法。结果证明了模型的有效性和适用性。