FEDERATED LEARNING IN ARTIFICIAL INTELLIGENCE: A PRIVACY-PRESERVING APPROACH FOR DISTRIBUTED MACHINE LEARNING SYSTEMS

Mehul J Vasava; Vrunda Gamit; Adesh V Panchal; Mihir M Patel; Hemangini Gohil

doi:10.29121/shodhkosh.v5.i5.2024.4817

Authors

Mehul J Vasava Assistant Professor, Department - CSE Deparment, College name - GEC, Patan
Vrunda Gamit Assistant Professor, Department - IT, College Name - Uka Tarsadia University
Adesh V Panchal Assistant Professor, Department - CSE Department, College - GEC, Patan
Mihir M Patel Assistant professor, Department - EC Department, College name - GEC, Patan
Hemangini Gohil Assistant professor, Department: CE, College name: Uka Tarsadia University

DOI:

https://doi.org/10.29121/shodhkosh.v5.i5.2024.4817

Keywords:

Federated Learning, Artificial Intelligence, Privacy Preservation, Distributed Machine Learning, Data Security, Edge Computing, Differential Privacy, Homomorphic Encryption, Iot, Gdpr

Abstract [English]

The advent of Artificial Intelligence (AI) and Machine Learning (ML) has revolutionized numerous industries by enabling systems to learn from data and make intelligent decisions. However, as the demand for data-driven models grows, so does the concern for data privacy and security. Traditional centralized learning paradigms collect data on a single server, creating substantial privacy risks, legal implications, and system inefficiencies. Federated Learning (FL) has emerged as a novel paradigm that enables model training across decentralized data sources without transferring raw data to a central server. This approach significantly enhances privacy preservation, reduces latency, and complies with data protection regulations such as GDPR and HIPAA. The paper delves into the core principles of FL, its system architectures, communication protocols, and privacy-preserving techniques such as differential privacy and homomorphic encryption. We also explore key applications across healthcare, finance, and IoT, highlighting both the opportunities and challenges in real-world implementation. Comparative analyses are presented between FL and traditional machine learning in terms of performance, privacy, and scalability. Finally, the paper addresses open research challenges and potential future directions to make federated learning more robust, scalable, and universally deployable.

References

Kairouz, P., McMahan, H. B., et al. (2019). "Advances and open problems in federated learning." arXiv preprint arXiv:1912.04977.

Bonawitz, K., Eichner, H., et al. (2019). "Towards federated learning at scale: System design." Proceedings of MLSys.

Li, T., Sahu, A. K., et al. (2020). "Federated optimization in heterogeneous networks." Proceedings of MLSys.

Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019). "Federated machine learning: Concept and applications." ACM Transactions on Intelligent Systems and Technology, 10(2), 1-19. DOI: https://doi.org/10.1145/3298981

Geyer, R. C., Klein, T., & Nabi, M. (2017). "Differentially private federated learning: A client level perspective." arXiv preprint arXiv:1712.07557.

Melis, L., Song, C., et al. (2019). "Exploiting unintended feature leakage in collaborative learning." IEEE Symposium on Security and Privacy. DOI: https://doi.org/10.1109/SP.2019.00029

Acar, D. A. E., Zhao, Y., et al. (2021). "Federated learning-based privacy-preserving and secure medical data sharing." Journal of Biomedical Informatics, 118, 103788.

Sav, S., & Seyfioglu, M. S. (2021). "A survey on communication-efficient federated learning." arXiv preprint arXiv:2106.11088.

Xu, J., Gursoy, M. E., & Velipasalar, S. (2021). "Hybrid differential privacy in federated learning: Achieving privacy with adaptive aggregation." Neural Computing and Applications.

Lu, Y., Huang, X., et al. (2020). "Blockchain and federated learning for privacy-preserved data sharing in industrial IoT." IEEE Transactions on Industrial Informatics, 16(6), 4177–4186. DOI: https://doi.org/10.1109/TII.2019.2942190