Makine Öğreniminde Veri Gizliliği: Zorluklar ve Federe Öğrenme Çözümleri

Abstract

Federe Öğrenme (FL), veri yerelliğini koruyarak dağıtık cihazlar arasında iş birliğine dayalı model eğitimi gerçekleştirmekte ve bu yönüyle gizlilik açısından hassas uygulamalar için umut verici bir paradigma sunmaktadır. Bu makale, gizlilik ve mahremiyet koruma mekanizmalarına odaklanan FL çalışmalarına ilişkin yapılandırılmış ve kapsamlı bir inceleme sunmaktadır. Çalışma, öncelikle merkezi, merkezi olmayan, FedAvg, kümelenmiş, asenkron ve heterojen yaklaşımlar dâhil olmak üzere temel FL mimarilerini gözden geçirmekte ve bunların performans, ölçeklenebilirlik ve uygulama karmaşıklığı açısından karşılaştırmalı bir değerlendirmesini yapmaktadır. 2023-2025 dönemine ait güncel literatür analiz edilerek adalet, heterojenlik ve sistem düzeyinde güvenliğe ilişkin gelişen zorluklar vurgulanmaktadır. Ardından, Diferansiyel Gizlilik (DP), Homomorfik Şifreleme (HE), Güvenilir Yürütme Ortamları (TEE), Güvenli Toplama (SA) ve Güvenli Çok Taraflı Hesaplama (SMPC) gibi temel gizlilik yöntemleri karşılaştırmalı olarak incelenmektedir. Bu yöntemlerin hesaplama maliyeti, ölçeklenebilirlik ve koruma gücü açısından göreceli avantaj ve sınırlılıkları, sağlık, finans ve IoT gibi çeşitli uygulama alanları kapsamında tartışılmaktadır. Bulgular, tek bir mekanizmanın tam koruma sağlamadığını ve FL’de etkili gizlilik güvencesinin, verimlilik ile gizlilik arasında denge kuran hibrit yaklaşımlar gerektirdiğini ortaya koymaktadır. Son olarak, açık araştırma alanları ve gelecek yönelimler belirlenerek, gerçek dünya uygulamalarında heterojenlik ve dinamik katılım koşullarında gizlilik garantilerini sürdürebilen, uyarlanabilir, kaynak farkındalıklı ve güvene dayalı FL çerçevelerine duyulan ihtiyaç vurgulanmaktadır.

Keywords

• Federe Öğrenme
• Veri Gizliliği
• Makine Öğrenmesi
• Veri Gizliliği Koruyucu Teknikler

Data Privacy in Machine Learning: Challenges and Federated Learning Solutions

Abstract

Federated Learning (FL) enables collaborative model training across distributed devices while preserving data locality, making it a promising paradigm for privacy-sensitive applications. This paper presents a structured and comprehensive survey of FL studies with a focus on confidentiality and privacy-preserving mechanisms. This paper first reviews major FL architectures including centralized, decentralized, FedAvg, clustered, asynchronous, and heterogeneous approaches and provides a comparative discussion of their performance, scalability, and implementation complexity. Recent survey literature (2023-2025) is analyzed to highlight evolving challenges related to fairness, heterogeneity, and system-level security. Subsequently, key confidentiality methods are comparatively reviewed, including Differential Privacy (DP), Homomorphic Encryption (HE), Trusted Execution Environments (TEE), Secure Aggregation (SA), and Secure Multi-Party Computation (SMPC). Their relative trade-offs in computation cost, scalability, and protection strength are examined across diverse application domains, such as healthcare, finance, and IoT. The findings indicate that no single mechanism offers complete protection, and effective privacy assurance in FL requires hybrid approaches that balance efficiency with confidentiality. Finally, open research gaps and future directions are identified, emphasizing the need for adaptive, resource-aware, and trust-anchored FL frameworks capable of maintaining privacy guarantees under real-world heterogeneity and dynamic participation.

Keywords

• Federated Learning
• Data privacy
• Machine Learning
• Privacy-Preserving Techniques

References

1. Yin X., Zhu Y., and Hu J. (2021). A comprehensive survey of privacy-preserving federated learning: a taxonomy, review, and future directions, ACM Computing Surveys, 54, (6), 1-36.
2. Peng L. and Qiu M. (2024). AI in healthcare data privacy-preserving: enhanced trade-off between security and utility. Proceedings of Knowledge Science, Engineering and Management (KSEM), 349-360.
3. Paracha A, Arshad J, Farah M. B., and Ismail K. (2024). Machine learning security and privacy: a review of threats and countermeasures, EURASIP Journal on Information Security, 2024, (1).
4. Ma S., Cao Y., and Xiong L. (2021). Transparent contribution evaluation for secure federated learning on blockchain. 37th International Conference on Data Engineering Workshops (ICDEW), IEEE, 88-91.
5. Bonawitz K., Eichner H., Grieskamp W., Huba D., Ingerman A., Ivanov V., Kiddon C., Konečný J., Mazzocchi S., McMahan H. B., Van Overveldt T., Petrou D., Ramage D., and Roselander J. (2019). Towards federated learning at scale: system design, arXiv:1910.06664.
6. Zhang C., Xie Y., Bai H., Yu B., Li W., and Gao Y. (2021). A survey on federated learning, Knowledge-Based Systems, (216), 106775.
7. Truong N., Sun K., Wang S., Guitton F., and Guo Y. (2021). Privacy preservation in federated learning: an insightful survey from the GDPR perspective, Computers & Security, (110), 102402.
8. Kairouz P., McMahan H., B., Avent B., Bellet A., Bennis M., and et al. (2021). Advances and open problems in federated learning, arXiv:1912.04977.

9. Ha T., and Dang T. K. (2022). Inference attacks based on GAN in federated learning, International Journal of Web Information System, 18, (2-3), 117-136.
10. Liu Z., Guo J., Yang W., Fan J., and Zhao J. (2022). Privacy-preserving aggregation in federated learning: A survey, arXiv:2203.17005.
11. Zhang Y., Zeng D., Luo J., Xu Z., and King I. (2023). A Survey of trustworthy federated learning with perspectives on security, robustness, and privacy, arXiv:2302.10637.
12. Rafi T., H., Noor F., A.,Hussain T., and Chae D-K. (2023). Fairness and privacy-preserving in federated learning: A survey, arXiv:2306.08402.
13. Chhetri B., Gopali S., Olapojoye R., Dehbash S., and Namin A. S. (2023). A survey on blockchain-based federated learning and data privacy, arXiv:2306.17338.
14. Hasan J. (2023). Security and privacy issues of federated learning, arXiv:2307.12181.
15. Zhao J. C., Bagchi S., Avestimehr S., Chan K. S., Chaterji S., Dimitriadis D., Li J., Li N., Nourian A., and Roth H. R. (2024). Federated learning privacy: attacks, defenses, applications, and policy landscape – a survey, arXiv:2303.01347.
16. Li H., Ge L., and Tian L. (2024). Survey: federated learning data security and privacy-preserving in edge-internet of things, Artificial Intelligence Review, 57, 130.
17. Rahman R. (2025). Federated learning: A survey on privacy-preserving collaborative intelligence, arXiv: 2504.17703.
18. Jimenez-Gutierrez D. M., Falkouskaya Y., Hernandez-Ramos J. L., Anagnostopoulos A., Chatzigiannakis I., and Vitaletti A. (2025). On the security and privacy of federated learning: a survey with attacks, defenses, frameworks, applications, and future directions, arXiv:2508.13730.
19. Feng Y., Guo Y., Hou Y., Wu Y., Lao M., Yu T., and Liu G. (2025). A survey of security threats in federated learning, Complex & Intelligent Systems, 11, 165.
20. Soltani B., Zhou Y., Haghighi V., and Lui J. C. S. (2023). A survey of federated evaluation in federated learning. Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence (IJCAI), 6769-6777.
21. Konečný J., McMahan H. B., Ramage D., and Richtárik P. (2016). Federated optimization: distributed machine learning for on-device intelligence, arXiv:1610.02527.
22. AbdulRahman S., Tout H., Ould-Slimane H., Mourad A., Talhi C., and Guizani M. (2021). A survey on federated learning: the journey from centralized to distributed on-site learning and beyond, IEEE Internet of Things Journal, 8, (7), 5476-5497.
23. Xu Z., Zhang Y., Andrew G., Choquette C., Kairouz P., Mcmahan B., Rosenstock J., and Zhang Y. (2023). Federated learning of gboard language models with differential privacy. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, 5, 629-639.
24. Kasturi A., Sivaraju R., and Hota C. (2022). FedPeer: A peer-to-peer learning framework using federated learning, Edge Analytics, 517-525.
25. Yuan L., Wang Z, Sun L., Yu P. S., and Brinton C. G. (2023). Decentralized federated learning: a survey and perspective, arXiv preprint arXiv:2306.01603.
26. Liu Y., Shi Y., Li Q., Wu B., Wang X., and Shen L. (2024). Decentralized directed collaboration for personalized federated learning, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 23168-23178.
27. Zhao Y., Li M., Lai L., Suda N., Civin D., and Chandra V. (2022). Federated learning with non-IID data, arXiv preprint arXiv:1806.00582.
28. Pandey M., Pandey S., and Kumar A. (2022). Introduction to Federated Learning, Springer, Berlin, Germany, 1-17.
29. Lv Y., Ding H., Wu H., Zhao Y., and Zhang L. (2023). FedRDS: federated learning on non-IID data via regularization and data sharing, Applied Sciences, 13, (23), 12962.
30. Islam M. S., Javaherian S., Xu F., Yuan X., and Tizeng N-F. (2024). FedClust: tackling data heterogeneity in federated learning through weight-driven client clustering. In Proceedings of the IEEE International Conference on Parallel Processing (ICPP), 474-483.
31. Ghosh A., Chung J., and Yin D. (2021). An efficient framework for clustered federated learning. arXiv:2006.04088
32. Elhussein A., Gursoy G. (2023). Privacy-preserving patient clustering for personalized federated learning, arXiv:2307.08847.
33. Yoo J. H., Son H. M, Jeong H., Jang E. H., Kim A. Y., Yu H. Y., Jeon H. J., and Chung T-M. (2021). Personalized federated learning with clustering: non-iid heart rate variability data application, arXiv:2108.01903.
34. Nguyen T. and Thai M. T. (2023). Preserving privacy and security in federated learning, IEEE/ACM Transactions on Networking, 32,(1), 833-843.
35. Tang B., Xiao Y., Zhang L., Cao B., Tang M., and Yang Q. (2024). AFL-HCS: asynchronous federated learning based on heterogeneous edge client selection, Cluster Computing, 27, (5), 6247-6264.
36. Xie C. (2020). Asynchronous federated optimization, arXiv:1903.03934.
37. Nguyen J., Malik K., Zhan H., Yousefpour, A. Rabbat M., Malek M., and Huba D. (2022). Federated learning with buffered asynchronous aggregation, arXiv:2106.06639.
38. Odeh J. O., Yang X., Nawakanma C. I., and Dhelim S. (2024). Asynchronous privacy-preservation federated learning method for mobile edge network in industrial internet of things ecosystem, Electronics, 13, (9), 1610.
39. Yu L., Sun X., Albelaihi R., and Yi C. (2022). Latency aware semi-synchronous client selection and model aggregation for wireless federated learning, arXiv:2210.10311.
40. Xu C, Qu Y., Xiang Y, and Gao L. (2023). Asynchronous federated learning on heterogeneous devices: a survey, arXiv:2109.04269.
41. Rieke N., Hancox J., Fausto Milletari W L., Roth H., Albarqouni S., Bakas S., and et al. (2021). The future of digital health with federated learning, arXiv:2003.08119.
42. Islam M. S., Panta S., Xu F., Yuan X., Chen L., Tzeng N-F. (2025). SEAFL: enhancing efficiency in semi-asynchronous federated learning through adaptive aggregation and selective training, arXiv:2503.05755.
43. Fan B., Jiang J.,Su X.,Tarkoma S., and Hui P. (2024). A survey on model-heterogeneous federated learning: problems, methods, and prospects, arXiv:2312.12091.
44. Fallah A., Mokhtari A., and Ozdaglar A. (2020). Personalized federated learning: a meta-learning approach, arXiv:2002.07948
45. Tan A. Z., Yu H., Cui L., and Yang Q. (2022). Towards personalized federated learning, arXiv:2103.00710.
46. Gao D., Yao X., and Yang Q. (2022). A survey on heterogeneous federated learning, arXiv:2210.04505.
47. Zeng D., Xu Z., Liu S., Pan Y., Wang Q., and Tang X. (2024). On the power of adaptive weighted aggregation in heterogeneous federated learning and beyond, arXiv:2310.02702.
48. Pfeiffer K., Rapp M., Khalili R., and Henkel J. (2023). Federated learning for computationally-constrained heterogeneous devices: a survey, arXiv:2307.09182.
49. Lyu L. and Chen C. (2021). A Novel Attribute Reconstruction Attack in Federated Learning, arXiv:2108.06910.
50. Liang J., Wang R., Feng C., and Chang C. (2023). A survey on federated learning poisoning attacks and defenses, arXiv:2306.03397.
51. Banse A., Kreischer J., and Oliva i Jürgens X. (2024). Federated learning with differential privacy. arXiv:2402.02230.
52. Pan Y., Chao Z., He W., Jing Y., Hongjia L., and Liming W. (2024). FedSHE: privacy-preserving and efficient federated learning with adaptive segmented CKKS homomorphic encryption, Cybersecurity, 7, (1), 40.
53. Alam T. and Gupta R. (2023). Federated learning and its role in the privacy preservation of IOT devices, Future Internet, 14, (9), 246.
54. Zhang L., Duan B., Li J., Ma Z., and Cao X. (2024). A TEE-based federated privacy protection method: proposal and implementation, Applied Sciences-Basel, 14, (8).
55. Bonawitz K., Ivanov V., Kreuter B., Marcedone A., McMahan H. B., Patel S., Ramage D., Segal A., and Seth K. (2017). Practical secure aggregation for privacy-preserving machine learning. CCS '17: Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, 1175-1191
56. Hard A., Rao K., Mathews R., Ramaswamy S., Beaufays F., Augenstein S:, Eichner H:, Kiddon C., and Ramage R. (2019). Federated learning for mobile keyboard prediction. arXiv:1811.03604.
57. Bagdasaryan E., Veit A., Hua Y., Estrin D., and Shmatikov V. (2019). How to backdoor federated learning, arXiv:1807.00459.
58. Blanchard P., El Mhamdi E. M., Guerraoui R., and Stainer J. (2017). Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent, 31st Conference on Neural Information Processing Systems (NIPS 2017). 118-128
59. Yin D., Chen Y., Ramchandran K., and Bartlett P. (2021). Byzantine-robust distributed learning: towards optimal statistical rates. arXiv:1803.01498.
60. El Mhamdi E. H., Guerraoui R:,and Rouault S. (2018). The hidden vulnerability of distributed learning in byzantium, arXiv:1802.07927.
61. Fang M., Cao X., Jia J., and Gong N. (2020). Local model poisoning attacks to byzantine-robust federated learning, 29th USENIX Security Symposium. 1623-1640.
62. Yar G., Boboila S., Nita-Rotaru C., and Oprea A. (2023). Backdoor attacks in peer-to-peer federated learning. arXiv:2301.09732.
63. Cao X., Fang M., Liu J., and Gong N. Z. (2022). FLTrust: byzantine-robust federated learning via trust bootstrapping, arXiv:2012.13995.
64. Dwork C., and Roth A. (2014). The algorithmic foundations of differential privacy, Foundations and Trends in Theoretical Computer Science, 9, (3-4), 211-407.