摘要
为了给消费者信贷决策提供合理依据,基于真实的客户信贷数据,运用逐步Logistic回归方法依据AIC准则进行变量选择建立经典决策树、条件推断树、随机森林、支持向量机等分类模型,并对4个分类模型的预测结果进行比较。结果表明:基于逐步Logistic回归建立的随机森林分类模型准确率达97%,预测效果最优;随机森林算法具有较高的分类精度,可以很好地应用在个人信用评估问题研究中。
In order to provide a reasonable basis for consumer credit decision-making,based on real customer credit data,the stepwise logistic regression method is used to select variables according to akaike information criterion to establish classification models such as classic decision trees,conditional inference trees,random forests,and support vector machines.The prediction results of four classification models are compared.The research shows that the accuracy rate of the random forest classification model based on stepwise logistic regression is 97%,and the prediction effect is the best;the random forest algorithm has high classification accuracy and can be well applied in the research of personal credit evaluation.
作者
李佳欣
Li Jiaxin(School of Mathematics and Statistics,Southwest University,Chongqing 400700,China)
出处
《湖南文理学院学报（自然科学版）》
CAS
2021年第1期5-8,57,共5页
Journal of Hunan University of Arts and Science(Science and Technology)
