基于C&RT-SVM的个人信用评估组合模型研究
Personal Credit Scoring based on C&RT and SVM Combined Model
本文针对个人信用评估单一模型中存在的不足，提出构建并行结构组合模型进行个人信用评估的方法。该模型将C&RT与SVM相结合，充分发挥单一模型各自的优势，利用加权投票的方法进行组合。在UCI德国个人信用数据集上的测试结果表明，组合模型在对样本的分类精度和稳定性上均优于单一模型，显著降低了消费信贷风险。
iming at the insufficiencies of single models in personal credit scoring, this paper presents a method for personal credit scoring by using combining forecast. The model combines C&RT with SVM, give full play to their respective advantages single model, using the method of weighted voting of confidence combined. Test results on the UCI Germany Personal credit data sets show that the combined model of the sample classification accuracy and stability are better than a single model, reducing the risk of consumer credit.
黄佳、龙海明
财政、金融计算技术、计算机技术
&RTSVM组合模型个人信用评估
&RTSVMcombined modelpersonal credit scoring
黄佳,龙海明.基于C&RT-SVM的个人信用评估组合模型研究[EB/OL].(2016-05-03)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/201605-12.点此复制
评论