Research on breach prediction for big data through hybrid ensemble learning and logistic regression
Xiaohan Zhang, Xu Chen · Journal of Physics Conference Series · 2021
Abstract The fast-growing online lending sector faces the risk of borrowers defaulting. Network lending data are often unbalanced. On the basis of EasyEnsemble algorithm, we proposed a hybrid ensemble algorithm, which combines the advantages of Bagging and Boosting. By comparing the performance of different learning models, the experimental results show that the hybrid ensemble algorithm has good classification accuracy and generalization ability. Further, we find that social capital has a negative effect on borrower default rate. This paper provides a relatively reliable method for the credit risk assessment of online borrowers, and also provides a new idea for risk monitoring in the field of online lending.