Federated Learning Overview: Frameworks, Challenges, and Future Directions
DOI:
https://doi.org/10.61173/rprnrk35Keywords:
Federated Learning, Data Privacy, Communication Efficiency, Edge ComputingAbstract
Federated Learning (FL) is a highly regarded distributed machine learning framework that aims to enable collaborative modeling among multiple parties without disclosing raw data. Compared to centralized learning, FL allows each participant to train models independently on-site and upload model update parameters rather than data to a central server, thereby improving model performance while effectively protecting data privacy. As artificial intelligence is increasingly applied in fields such as healthcare, finance, and industrial IoT, data privacy and compliance requirements are becoming increasingly stringent, highlighting the significant application potential and research value of federated learning. This paper systematically reviews the basic theories, core algorithms, and technical approaches of federated learning, focusing on research progress in areas such as communication efficiency, data heterogeneity, privacy protection, and trust mechanisms. The study also explored the development prospects and future directions of federated learning in the medical, financial, and precision computing fields. At the same time, we conducted a more in-depth analysis of the main challenges currently faced and analyzed potential future research directions in order to provide a reference for the continued development of this field.
References
[1] McMahan H B, Moore E, Ramage D, et al. Communicationefficient learning of deep networks from decentralized data. Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS). 2017.
[2] Yang Q, Liu Y, Chen T, et al. Federated machine learning: Concept and applications. ACM Transactions on Intelligent Systems and Technology, 2019, 10(2): 1–19.
[3] Li T, Sahu A K, Zaheer M, et al. Federated optimization in heterogeneous networks. Advances in Neural Information Processing Systems (NeurIPS). 2020.
[4] Acar D A, Zhao Y, Mattina M, et al. Federated learning based on dynamic regularization. International Conference on Learning Representations (ICLR). 2021.
[5] Sattler F, Wiedemann S, Müller K R, et al. Robust and communication-efficient federated learning from non-IID data. IEEE Transactions on Neural Networks and Learning Systems, 2020, 31(9): 3400–3413.
[6] Reisizadeh A, Mokhtari A, Hassani H, et al. FedPAQ: A communication-efficient federated learning method with periodic averaging and quantization. Proceedings of the 23rd International Conference on Artificial Intelligence and Statistics (AISTATS). 2020.
[7] Liu J, Kang J, Xu Y, et al. Recent advances on federated learning: A systematic survey. arXiv:2301.01299.
[8] Lu X, Li Y, Liu X, et al. Privacy-preserving asynchronous federated learning for edge network computing. IEEE Access, 2020, 8: 48970–48981.
[9] Scott S, Yu L, Zhao S, et al. PeFLL: Personalized federated learning by learning to learn. International Conference on Learning Representations (ICLR). 2024.
[10] Carlini N, Tramer F, Wallace E, et al. The secret sharer: Measuring unintended neural network memorization & extracting secrets. arXiv:1802.08232.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 by the authors.
This work is licensed under a Creative Commons Attribution 4.0 International License.
