基于Bagging集成的个人信用风险评估方法研究

Research on Personal Credit Risk Assessment Method Based on Bagging Integration

在个人信用评估问题和风险日益剧增的背景下,为了能够高效地区分申请者的信用情况。从梯度提升树组合特征和集成算法的角度出发,提出一种基于Bagging集成算法的个人信用风险评估模型。为了验证梯度提升树组合生成特征的有效性,利用四个UCI数据集进行对比测试,结果表明通过增加新的组成特征,模型的鲁棒性更强。最后通过German和Credit两个数据集,与逻辑回归集成、支持向量机集成、随机森林集成、梯度提升树集成进行对比,验证了混合模型的有效性。

Under the background of increasing personal credit assessment problems and risks,it is necessary to distinguish applicants’credit conditions efficiently.From the perspective of the gradient-lifting tree combination features and integration algorithm,a personal credit risk assessment model based on Bagging integration algorithm is proposed.In order to verify the effectiveness of the gradient-boosting tree combination generating features,four UCI data sets were used for comparison testing.The results show that the model is more robust by adding new component features.Finally,through the German and Credit datasets,it is compared with logistic regression integration,support vector machine integration,random forest integration,and gradient-based tree integration to verify the effectiveness of the hybrid model.