LoanMatrix: An Ensemble Machine Learning Framework for Automated Loan Approval Prediction

Mahesh R. Bhurake, Amit S. Marchande and Tejas V. Joshi

International Journal of Scientific Research in Science and Technology, 2026, vol. 13, issue 3, 993-1002

Abstract: Loan approval in commercial banks still relies heavily on manual credit reviews, which are slow and often inconsistent. In this work, we built LoanMatrix, a credit scoring system that uses gradient boosting models to automate loan decisions. We tested three setups: LightGBM alone, CatBoost alone, and a combined ensemble using simple probability averaging. We also added SHAP explanations so banks can understand why a loan was approved or rejected. On a dataset of 32,581 loan records, our ensemble model reached 93.40% accuracy and an F1-score of 82.23%. One key benefit of the ensemble was that it cut prediction variance in half — from 2.8% to 1.4% — making decisions more stable across different applicant groups. CIBIL score and annual income were the top factors in predictions. The whole system runs under 1 millisecond per prediction, making it practical for real-time banking use.

Keywords: Credit risk assessment; gradient boosting; ensemble learning; LightGBM; CatBoost; Loan Default Prediction; Explainable AI; SHAP; Financial Machine Learning (search for similar items in EconPapers)
Date: 2026
References: Add references at CitEc
Citations:

Downloads: (external link)
https://ijsrst.com/home/article/view/IJSRST26133231 Abstract page (text/html)
https://ijsrst.com/home/article/download/IJSRST26133231/IJSRST26133231 Full text (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:etm:ijsrst:v13:y2026:i3:id:1692

DOI: 10.32628/IJSRST26133231

Access Statistics for this article

More articles in International Journal of Scientific Research in Science and Technology from Technoscience Academy
Bibliographic data for series maintained by Pankaj Sharma ().