Research on Personal Credit Evaluation Based on Machine Learning Algorithm

With the rapid development of Internet finance and consumer credit, it is necessary to establish a reasonable personal credit evaluation model. This paper discusses the accuracy of six general personal credit evaluation models based on the Give Me Some Credit data set from Kaggle. A multiple-average...

Full description

Saved in:
Bibliographic Details
Published in2021 6th International Symposium on Computer and Information Processing Technology (ISCIPT) pp. 48 - 52
Main Authors Liu, Shiqi, Wang, Ruiqi, Han, Yueji
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2021
Subjects
Algorithm contrast
Machine Learning
Machine learning algorithms
Neural networks
Personal Credit evaluation
Prediction algorithms
Predictive models
Stability analysis
Support vector machines
Training
Online AccessGet full text

Cover

More Information
Summary:With the rapid development of Internet finance and consumer credit, it is necessary to establish a reasonable personal credit evaluation model. This paper discusses the accuracy of six general personal credit evaluation models based on the Give Me Some Credit data set from Kaggle. A multiple-average machine learning model is established to improve the stability of the result. The research shows the NB classifier, Support Vector Machines, Random Forest and BP Neural Network have good prediction effect on individual credit default risk, while the accuracy of Logical Regression and Stepwise Regression is relatively low. The multiple averaging method can effectively reduce the volatility of the results, and the accuracy will not be significantly reduced under different proportion of training set and prediction set.
DOI:10.1109/ISCIPT53667.2021.00016