摘要
针对个人信用等级的多分类问题进行了研究.通过建立个人信用风险评价指标体系,运用判别分析法构建关于样本的评分模型,得到判别得分;再用神经网络法对样本进行评分预测,得到对应得分,并对神经网络预测得分进行降序排列得到有序样本,最后进行有序样本最优分割,从而实现个人信用的等级划分.该模型在一定程度上有助于借贷者选择优质客户,从而降低信贷风险.
With the rapid growth of personal credit business,personal credit risk assessment is an urgent problem to be solved,and the personal credit rating classification problem is put under examination. Through the establishment of personal credit risk evaluation index system,using discriminant analysis method to construct a grading model for the samples,thus the discriminant score is obtained; then the neural network method is used to predict the scores for the samples,and the scores due are collected and such scores were to put into an descending order so an orderly sample is acquired. Finally,the optimal segmentation of the orderly sample is done so as to realize rating of personal credit. The model,to a certain extent,helps borrowers spot the high quality customers,thereby to reduce credit risk.
出处
《内江师范学院学报》
2018年第2期64-68,共5页
Journal of Neijiang Normal University
基金
四川省教育厅自然科学重点项目(13ZA0016)
西华师范大学博士科研启动基金(12B025)
西华师范大学英才科研基金项目(17YC381)
关键词
个人信用风险
判别分析
神经网络
评分等级
有序样本最优分割
personal credit risk
discriminant analysis
neural network
rating
optimal segmentation of ordered sample
