CAJ | 학술논문

对银行来说，分析量化操作风险，对管理者进行相关决策十分重要。贝叶斯网络能够很好地表达不确定因素之间的因果关系，并用于推理决策，是提高操作风险管理水平的有效途径。通过构建贝叶斯网络模型，将可能导致贷款者拖欠还款的各种影响因素引入到具有因果关联的网络结构中，计算出各类指标对拖欠的影响程度。并且通过分析节点，直观地显示了每个模型的预测效果，为银行预测潜在贷款拖欠者，并进行相关风险管理提供了理论依据。
대은행래설，분석양화조작풍험，대관리자진행상관결책십분중요。패협사망락능구흔호지표체불학정인소지간적인과관계，병용우추리결책，시제고조작풍험관리수평적유효도경。통과구건패협사망락모형，장가능도치대관자타흠환관적각충영향인소인입도구유인과관련적망락결구중，계산출각류지표대타흠적영향정도。병차통과분석절점，직관지현시료매개모형적예측효과，위은행예측잠재대관타흠자，병진행상관풍험관리제공료이론의거。
The quantification of operational risk is very important for bank managers in making decision .Bayesian network can not only express the causal relationship between uncertainty factors well ,but some reasoning ,which is an efficient way to improve op-erational risk management ., We introduce various factors affected the network structure with causal correlation By building a Bayes -ian network,and calculate the effect of every factor for loan defaulters .After evaluating every node , we can understand the predic-tive effect of each model Intuitively ,which is helpful for banks to forecast loan defaulters and provide a theoretical basis for the risk control.