Privacy-Preserving Federated Learning for Real-Time Fraud Detection in Payment Networks
Keywords:
Federated Learning; Real-Time Fraud Detection; Differential Privacy; Non-IID Data; Payment SystemsAbstract
The rapid adoption of real-time payment networks has intensified fraud risks, yet centralized detection systems conflictwith privacy regulations prohibiting cross-institutional data sharing. This article proposes AFLPFD, an AdaptiveFederated Learning Framework for Privacy-Preserving Real-Time Fraud Detection in Distributed Payment Networks, addressing the simultaneous challenges of privacy preservation
References
Victor Chang et al., "Digital payment fraud detection methods in digital ages and Industry 4.0," ScienceDirect, 2022. https://www.sciencedirect.com/science/article/abs/pii/S0045790622000465
Vankamamidi S Naresh and D Ayyappa, "Privacy-preserving federated credit risk models: evaluating differential privacy and homomorphic encryption techniques," NLM, 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC12864762/
