作　　者： ; ; ;

机构地区： 天津大学管理与经济学部

出　　处： 《西安电子科技大学学报（社会科学版）》 2007年第4期32-39,共8页

摘　　要： 操作风险是指由不完善的或有问题的内部程序、人员及系统或外部事件所造成损失的风险。由于操纵风险具有构成更复杂、涉及诸多复杂因素、难以结构化、缺少历史数据等特点,因此难于建模与度量。贝叶斯网络是基于贝叶斯决策理论的因果建模技术。它能够用于建立操作风险度量系统并作为操作风险度量的基础。本文首先回顾了操作风险的相关内容。接下来介绍贝叶斯网络的概念、特点,以及贝叶斯网络模型的数学描述。通过实例演示了贝叶斯网络在银行操作风险方面的建模与应用。最后给出了一个操作风险管理的贝叶斯网络模型的框架。 Operational risk is the risk of loss resulting from inadequate or failed internal processes, people and systems or from external events. It is difficult to model and measure operational risk of bank. The Bayesian network has a great deal of advantages in modeling and measuring operational risk, which is induced by uncertain elements in a complicated system. After reviewing operational risk of bank and Bayesian networks, this paper illustrates how to use Bayesian networks to manage operational risk with examples, including measure risk, casual analysis, and scenario analysis etc. This paper also gives a Bayesian network model framework of operational risk.

关 键 词： 操作风险 贝叶斯网络 因果建模 情景分析 不确定性推理

领　　域： [哲学宗教]