The Comprehensive Analysis of Bank Loan Approval Prediction Based on Machine Learning Models

Deyi Li

2025

Abstract

In contemporary society, given the current economic depression, many households are experiencing significant financial pressure, making it increasingly challenging to meet large capital requirements. The price and cost of products is increasing, and the capital required for a lump-sum purchase has become prohibitively high. Therefore, obtaining a bank loan represents one of the most practical solutions to address this need. However, for banks, the volume of daily loan applications is substantial, making it impractical to approve every request. This necessitates the allocation of limited funds to reliable applicants who have a strong likelihood of repayment. Most banks consider a customer’s credit score as a critical factor in loan approval. However, when credit scores are unavailable in customer information, the alternative relevant data can be utilized to assess risk. Five models of machine learning, in this study, were applied to predict loan issuance status in both scenarios: including and excluding credit scores. The accuracy of prediction, precision of prediction, recall score of prediction, f1 and Area Under Curve (AUC) score of these models were compared and evaluated. When credit scores are available, the decision tree model attains the highest accuracy and AUC score compared to other models; among predictions made without credit scores, the random forest model provides the best comprehensive performance.

Paper Citation

in Harvard Style

Li D. (2025). The Comprehensive Analysis of Bank Loan Approval Prediction Based on Machine Learning Models. In Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE; ISBN 978-989-758-765-8, SciTePress, pages 527-531. DOI: 10.5220/0013701400004670

in Bibtex Style

@conference{icdse25,
author={Deyi Li},
title={The Comprehensive Analysis of Bank Loan Approval Prediction Based on Machine Learning Models},
booktitle={Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE},
year={2025},
pages={527-531},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013701400004670},
isbn={978-989-758-765-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE
TI - The Comprehensive Analysis of Bank Loan Approval Prediction Based on Machine Learning Models
SN - 978-989-758-765-8
AU - Li D.
PY - 2025
SP - 527
EP - 531
DO - 10.5220/0013701400004670
PB - SciTePress