Research Article
Towards Personal Credit Default Prediction Method Based on Data Mining
@INPROCEEDINGS{10.4108/eai.2-6-2023.2334599, author={Yiran Xue}, title={Towards Personal Credit Default Prediction Method Based on Data Mining}, proceedings={Proceedings of the 2nd International Conference on Information Economy, Data Modeling and Cloud Computing, ICIDC 2023, June 2--4, 2023, Nanchang, China}, publisher={EAI}, proceedings_a={ICIDC}, year={2023}, month={8}, keywords={credit default financial loss prediction model evaluation metrics data mining}, doi={10.4108/eai.2-6-2023.2334599} }- Yiran Xue
Year: 2023
Towards Personal Credit Default Prediction Method Based on Data Mining
ICIDC
EAI
DOI: 10.4108/eai.2-6-2023.2334599
Abstract
This project focuses on assessing the user’s personal credit risk based on data mining techniques. This research is designed to help financial institutions predict whether borrowers will be able to repay all their loans within a given period, thereby reducing the financial losses caused by deviations between the risk assessment and the actual situation during the lending process. This study has collected a number of user data, and used five types of algorithms, such as DT (decision tree), NB (Naive Bayes) and LR (logistic regression), to build personal credit default prediction model respectively. Meanwhile, ACC (accuracy), AUC (area under the ROC curve) and KS values were selected as the model evaluation metrics. The experimental result shows that the DT model is the most suitable for personal credit default prediction.