USING MACHINE LEARNING IN BANKS TO FORECAST LOAN DEFAULTS: PART 2
Keywords:
banks, lending, loan defaults, machine learning, deep neural networks, Extreme Gradient Boosting algorithms.Abstract
For most traditional banks, lending activity remains the major source of revenue. However, there is always a possibility that some borrowers default on their loans and this situation creates credit risk for banks. To manage this risk, banks utilize different qualitative and quantitative methods to predict loan defaults. Currently, traditional algorithms such as logistic regression and decision trees are among the most popular models used to predict loan defaults. Their simple implementation and high accuracy make them preferred choice among credit analysts. However, rapid technological progress and increased computing capabilities of computers are creating opportunities to apply more advanced machine learning algorithms for managing credit risk. For instance, deep neural networks and recently developed Extreme Gradient Boosting algorithms (XGB) have been shown to exhibit high accuracy in wide range of classification tasks and have potential to exhibit high accuracy in loan default prediction.
References
Berdiyorov, B. (2024). Using Machine Learning in Banks to Forecast Loan Defaults. International Conference “Banklarda Raqamli Transformatsiya: Ta’lim Bilan Oʻzaro Hamkorlik”, TIFT University.
Odegua, R. (2020). Predicting Bank Loan Default with Extreme Gradient Boosting. Cornell University. Available from https://arxiv.org/abs/2002.02011 [Accessed 18 February 2020].
Lending Club Data (2025). Loan Default Dataset. Kaggle. Available from https://www.kaggle.com/wendykan/lending-club-loan-data/download
Kumar, U.A. (2018) Neural networks and statistical techniques: A review of applications. Expert Systems with Applications, 36(1), 2-17. Available from https://www.researchgate.net/publication/222546331_Kumar_UA_Neural_networks_and_statistical_techniques_A_review_of_applications_Expert_Systems_with_Applications_361_2-17
Zhou, L. and Wang, H. (2012). Loan Default Prediction on Large Imbalanced Data Using Random Forests, TELKOMNIKA Indonesian Journal of Electrical Engineering, 10(6), 1519-1525. Available from https://www.researchgate.net/publication/267864165_Loan_Default_Prediction_on_Large_Imbalanced_Data_Using_Random_Forests
Zurada, M. (2002). How Secure Are Good Loans: Validating Loan-Granting Decisions and Predicting Default Rates on Consumer Loans. The Review of Business Information Systems, 6(3), 65-84. Available from https://clutejournals.com/index.php/RBIS/article/view/4563/4654
