AI-Powered Loan Risk Assessment: Predicting Risk Levels Using Machine Learning and Customer Profiling

Authors- MMr. Y. Ravi Bhushan., C.Kavya Sri., B.Praharshitha., C.Sriyashaswi., S.V.Tejasri., A.Prudhvi Kedar.

Abstract-Banks are essential to the global financial system, relying on loan interest as a key source of income, but loan defaults can lead to significant losses. To mitigate this risk, machine learning algorithms can efficiently predict the likelihood of default before approving a loan. In this study, six models—Decision Tree, Random Forest, Support Vector Machine (SVM), Multi-layer Perceptron (MLP) Artificial Neural Network, Naive Bayes, and a stacking ensemble model—were trained using a dataset with 20 key factors commonly found in loan applications. The stacking ensemble model achieved the highest accuracy at 78.75%, while the Random Forest model performed similarly with 78.15% accuracy but greater efficiency. Key predictors of credit risk included credit amount, checking account status, customer age, loan duration, and loan purpose. These findings highlight the potential of machine learning models to enhance loan approval decisions and minimize financial risk for banks.

DOI: /10.61463/ijset.vol.13.issue2.266