Applying explainable artificial intelligence to interpret supervised ensemble learning models for robust credit card fraud detection
Shimaa S. Awad, Alyaa A. Hamza, Mohamed A. Sobh, Ayman M. Bahaa-Eldin · 2026
Read Applying explainable artificial intelligence to interpret supervised ensemble learning models for robust credit card fraud detection by Shimaa S. Awad, Alyaa A. Hamza, Mohamed A. Sobh, Ayman M. Bahaa-Eldin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13179347.txt.
Description
Credit card fraud detectors can be accurate and still useless to the humans who must act on their alerts, if nobody can explain the flags. The authors compare four supervised learning methods for fraud detection: logistic regression, random forest, and two boosting ensembles, XGBoost and LightGBM. To test generalizability, they run all four on three different public transaction datasets. Standard scores cover accuracy, precision, recall, and ranking quality.
Tree-based ensembles, led by XGBoost, give the best predictive results, while the linear method still improves clearly over baselines. The authors then apply the SHAP explanation framework to the top models, ranking which features drive the verdicts. The study pairs every performance figure with an interpretability readout, model by model and dataset by dataset. The authors offer the combined picture as a guide to building fraud systems that stay transparent and accountable.
Catalog record
| CREATOR | Shimaa S. Awad, Alyaa A. Hamza, Mohamed A. Sobh, Ayman M. Bahaa-Eldin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 175 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |