EUVIMEDEUVIMED entdecken
Crossref

Fraud Detection in Loan Applications Using Machine Learning: An Applied Study in the Saudi Financial Market

Elsedeeg Mohamed Elbasher Abdalla Ahmed, Mohamed Abdelsatar, Khaled Eskaf

Midocean Journal for Research and Studies · 12.06.2026 · 2 (3) · 330-346

DOI: 10.71335/krjgkf74 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Midocean Journal for Research and Studies, online erschienen am 12.06.2026. Verfasst von Elsedeeg Mohamed Elbasher Abdalla Ahmed und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

This study examines the role of machine learning techniques in improving fraud detection within consumer loan applications in the Saudi financial market, with a focus on evaluating the predictive performance and interpretability of different supervised learning models. Methodology: The study utilized a de-identified real-world dataset consisting of 4,400 consumer loan applications, including 2,110 fraudulent cases and 2,290 non-fraudulent cases. After data preprocessing, three supervised machine learning models were implemented and compared: Logistic Regression, Random Forest, and XGBoost. Model performance was evaluated using a holdout test split, while SHAP values were employed to interpret model decisions and provide audit explanations.

Results

The findings revealed that the tree-based models outperformed Logistic Regression in predictive capability, with XGBoost achieving the highest classification accuracy on the holdout test dataset. However, the study emphasized that achieving perfect performance on a single split is not sufficient for deployment, as additional validation is required to identify potential data leakage and ensure reliability using out-of-time samples. Furthermore, SHAP analysis enhanced the explainability of the model, although the interpretation of de-identified features requires further verification before practical audit implementation.

Zitieren

APA

Ahmed, E. M. E. A., Abdelsatar, M., & Eskaf, K. (2026). Fraud Detection in Loan Applications Using Machine Learning: An Applied Study in the Saudi Financial Market. Midocean Journal for Research and Studies, 2(3), 330-346. https://doi.org/10.71335/krjgkf74

Vancouver

Ahmed EMEA, Abdelsatar M, Eskaf K. Fraud Detection in Loan Applications Using Machine Learning: An Applied Study in the Saudi Financial Market. Midocean Journal for Research and Studies. 2026;2(3):330-346. doi:10.71335/krjgkf74

BibTeX
@article{Ahmed2026,
  author = {Ahmed, Elsedeeg Mohamed Elbasher Abdalla and Abdelsatar, Mohamed and Eskaf, Khaled},
  title = {Fraud Detection in Loan Applications Using Machine Learning: An Applied Study in the Saudi Financial Market},
  journal = {Midocean Journal for Research and Studies},
  year = {2026},
  volume = {2},
  number = {3},
  pages = {330--346},
  doi = {10.71335/krjgkf74}
}
Kostenlos starten →