The probability of occurrence, occurrence time and holding time make the estimation of the multiresource, multidimensional, transmissible and dynamic characteristics of supply-chain risk more valuable to achieving the goal of risk assessment. A dynamic Bayesian network model of supply-chain risk is constructed to describe the property of supply-chain risk; the Bayesian inference tool is then used to estimate the corresponding parameters by maximum likelihood and inference for supply-chain risks. It is concluded that supply-chain risk changes with time and would converge at a certain stable interval, occurrence time and holding time satisfy several Poisson processes.

中文翻译：

---

供应链系统的风险评估：一种概率推理方法

发生概率，发生时间和保持时间使得对供应链风险的多资源，多维，可传递和动态特征的估计对于实现风险评估的目标更加有价值。建立了供应链风险的动态贝叶斯网络模型来描述供应链风险的性质。然后使用贝叶斯推理工具通过最大可能性和供应链风险的推理来估计相应的参数。结论是，供应链风险随时间变化，并会在一定的稳定区间内收敛，发生时间和保持时间满足几个泊松过程。