Federated Machine Learning For Fraud Detection In IoT-Based Credit Card Systems

Authors: Mrs N.S.Kulkarni, Santosh S Kalshetty

Abstract: With the proliferation of Internet of Things (IoT)-based payment systems, credit card fraud has become more sophisticated and widespread. Traditional centralized fraud detection methods often fall short due to latency, data privacy concerns, and scalability limitations. This paper proposes a novel Federated Machine Learning (FML) approach for real-time credit card fraud detection within IoT ecosystems. By leveraging federated learning, data remains on edge devices (e.g., PoS terminals, mobile devices), ensuring privacy while enabling collaborative model training. The proposed framework integrates anomaly detection, ensemble learning, and cloud-edge orchestration. Experimental results on real-world datasets demonstrate superior accuracy, recall, and privacy compared to traditional methods.

DOI: http://doi.org/