A Systematic Survey and Empirical Comparison of Hybrid Methods for Imbalanced Fraud Detection

This repository contains the code and datasets for our paper, "A Systematic Survey and Empirical Comparison of Hybrid Methods for Imbalanced Fraud Detection: Combining Resampling and Machine Learning," published in the AUT Journal of Mathematics and Computing.

Fraud detection faces a critical hurdle: severe class imbalance. Our work systematically explores hybrid frameworks that combine resampling techniques with machine learning to overcome this challenge.

🔗 Links

• Access the Paper: https://ajmc.aut.ac.ir/article_5913.html

🔑 Key Contributions

• Comprehensive survey of hybrid imbalance learning methods.
• Rigorous empirical study on auto insurance fraud data.
• Open-source code for reproducibility and further research.
• Evidence that resampling strategy profoundly impacts model performance.

📜 How to Cite

If you use this research, please cite our paper as follows:

Yousefimehr, B., Ghatee, M. (2026). 'A systematic survey and empirical comparison of hybrid methods for imbalanced fraud detection: Combining resampling and machine learning', AUT Journal of Mathematics and Computing, 7(1), pp. 85-116. doi: 10.22060/ajmc.2025.24642.1446