摘要
个人信用评估是现代商业银行个人信用管理的核心.本文将数据挖掘中的随机森林算法(Random Forests,RF)运用到现代个人信用评估模型中,实现了逐步优化和评估.实证分析的结果证明,随机森林模型具有较高的精确性和泛化能力,能够克服噪声数据的影响.通过对各特征变量的重要性评分,得到贷款期限和总额等对风险预测的准确率具有显著作用.
Personal credit assessment is the core of modem commercial bank personal credit management. In this paper, the Random Forest algorithm in data mining （the Random Forest, RF） apply to the modem personal credit evaluation model, realized step by step optimization and evaluation. Empirical analysis proves that the result of the random forest model has high accuracy and generalization ability, and can overcome the influence of the noise data. Through to the importance of each feature variables score, loan time limit and the total accuracy of risk prediction has a significant effect.
出处
《商丘师范学院学报》
CAS
2016年第12期12-15,共4页
Journal of Shangqiu Normal University
基金
国家社会科学基金"代际转移视角下缩小我国收入差距的路径与仿真模拟研究"(11CTJ006)资助项目
关键词
随机森林
特征变量
个人信用评估
R软件
random forests
characteristics of the variable
personal credit assessment
R software
