Exploring the Performance of XGBOOST and Artificial Neural Network in Personal Credit Default Prediction: An Empirical Study

As the global financial market continues to open up, credit rating work has become increasingly important in the financial industry. Not only enterprises and institutions, but now there is also a credit rating system for individuals. The emergence of personal credit rating technology helps banks and other institutions to predict credit defaults. However, most credit rating agencies use relatively backward technology, many subjective factors and computing restrictions lead to low reliability of rating results and low computing efficiency. To solve this problem, this paper proposes to use machine learning model XGBOOST and artificial neural network model as the solution. To explore the performance of the two models in credit default prediction, this study uses the credit data set provided by a certain platform in China to model and predict the two models respectively. Finally, the model is evaluated by experimental evaluation with AUC as the evaluation index.