The Prediction of Non-Performing Loans Using Artificial Intelligence - A literature Review | ||
| المجلة العلمية للبحوث والدراسات التجارية | ||
| Volume 39, Issue 3, September 2025, Pages 1467-1497 PDF (668.94 K) | ||
| Document Type: المقالة الأصلية | ||
| DOI: 10.21608/sjrbs.2025.348316.1855 | ||
| Authors | ||
| Doaa Rashad* ; Mohamed ِAbdel-Salam; Engy Yehia | ||
| Faculty of Commerce and Business Administration at Helwan University | ||
| Abstract | ||
| The profitability and stability of financial institutions are seriously threatened by Non-Performing Loans (NPLs), hence precise prediction of these loans is crucial for efficient risk control. This paper studies the development of prediction approaches, from statistical methods to Machine Learning (ML) and more recently, deep learning models in predicting Non-Performing Loans. Machine Learning models have shown more accurate results at spotting non-linear patterns than statistical methods, but they still struggle to analyze sequential or unstructured datasets efficiently. Results obtained by this paper showed that although machine learning methods such as Support Vector Machines and Decision Trees produced consistent results, deep learning techniques, especially Artificial Neural Networks (ANN), particularly show consistently better accuracy over a spectrum of applications. By evaluating recent prediction approaches, this study emphasizes the necessity of utilizing deep learning techniques to handle the changing complexities of credit risk assessment, particularly the prediction of Non-Performing Loans (NPLs) in financial institutions. | ||
| Keywords | ||
| Non-Performing Loans; Statistical Methods; Machine Learning; Deep Learning; Datasets | ||
|
Statistics Article View: 18 PDF Download: 12 |
||
