Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2502.02290 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914394162593792 |
|---|---|
| author | Lunghi, Daniele Molinghen, Yannick Simitsis, Alkis Lenaerts, Tom Bontempi, Gianluca |
| author_facet | Lunghi, Daniele Molinghen, Yannick Simitsis, Alkis Lenaerts, Tom Bontempi, Gianluca |
| contents | Adversarial attacks pose a significant threat to data-driven systems, and researchers have spent considerable resources studying them. Despite its economic relevance, this trend largely overlooked the issue of credit card fraud detection. To address this gap, we propose a new threat model that demonstrates the limitations of existing attacks and highlights the necessity to investigate new approaches. We then design a new adversarial attack for credit card fraud detection, employing reinforcement learning to bypass classifiers. This attack, called FRAUD-RLA, is designed to maximize the attacker's reward by optimizing the exploration-exploitation tradeoff and working with significantly less required knowledge than competitors. Our experiments, conducted on three different heterogeneous datasets and against two fraud detection systems, indicate that FRAUD-RLA is effective, even considering the severe limitations imposed by our threat model. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_02290 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FRAUD-RLA: A new reinforcement learning adversarial attack against credit card fraud detection Lunghi, Daniele Molinghen, Yannick Simitsis, Alkis Lenaerts, Tom Bontempi, Gianluca Machine Learning Artificial Intelligence Adversarial attacks pose a significant threat to data-driven systems, and researchers have spent considerable resources studying them. Despite its economic relevance, this trend largely overlooked the issue of credit card fraud detection. To address this gap, we propose a new threat model that demonstrates the limitations of existing attacks and highlights the necessity to investigate new approaches. We then design a new adversarial attack for credit card fraud detection, employing reinforcement learning to bypass classifiers. This attack, called FRAUD-RLA, is designed to maximize the attacker's reward by optimizing the exploration-exploitation tradeoff and working with significantly less required knowledge than competitors. Our experiments, conducted on three different heterogeneous datasets and against two fraud detection systems, indicate that FRAUD-RLA is effective, even considering the severe limitations imposed by our threat model. |
| title | FRAUD-RLA: A new reinforcement learning adversarial attack against credit card fraud detection |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2502.02290 |