Mitigating Label-Flipping Attacks in Federated Learning: Experiments and Client Selection Strategies

• João Pedro C. Batista UFES
• Eduardo M. M. Sarmento UFES
• Johann J. S. Bastos UFES
• Vinícius F. S. Mota UFES
• Rodolfo S. Villaça UFES

DOI: https://doi.org/10.5753/eries.2024.244627

Abstract

This article explores challenges affecting model efficacy in Federated Learning, particularly due to malicious clients engaging in attacks like ”label-flipping”. Through experiments in the MininetFed environment, it assesses the influence of these clients and the effectiveness of different client selection strategies and clustering algorithms in mitigating such specific attacks. The findings provide crucial insights for enhancing training process security and effectively safeguarding models in Federated Learning against internal threats.

References

Alves, V. R. M. et al. (2024). Seleção de clientes adaptativa baseada em privacidade diferencial para aprendizado federado. In SIMPÓSIO BRASILEIRO DE REDES DE COMPUTADORES E SISTEMAS DISTRIBUÍDOS (SBRC).

Bastos, J. J. S. et al. (2024). Mininetfed: Uma ferramenta para emulação e análise de aprendizado federado com dispositivos heterogêneos. In SIMPÓSIO BRASILEIRO DE REDES DE COMPUTADORES E SISTEMAS DISTRIBUÍDOS (SBRC).

de Souza, A. M. et al. (2023). Dispositivos, eu escolho vocês: Seleção de clientes adaptativa para comunicação eficiente em aprendizado federado. In SIMPÓSIO BRASILEIRO DE REDES DE COMPUTADORES E SISTEMAS DISTRIBUÍDOS (SBRC).

Jebreel, N. M., Domingo-Ferrer, J., Sánchez, D., and Blanco-Justicia, A. (2022a). Defending against the label-flipping attack in federated learning.

Jebreel, N. M. et al. (2022b). Lfighter: Defending against the label-flipping attack in federated learning. In Neural Networks. Elsevier.

Jiang, Y., Zhang, W., and Chen, Y. (2023). Data quality detection mechanism against label flipping attacks in federated learning. IEEE Transactions on Information Forensics and Security, 18:1625–1637.

Li, D., Wong, W. E., Wang, W., Yao, Y., and Chau, M. (2021). Detection and mitigation of label-flipping attacks in federated learning systems with kpca and k-means. In 2021 8th International Conference on Dependable Systems and Their Applications (DSA), pages 551–559.

Mammen, P. M. (2021). Federated learning: Opportunities and challenges. In Proceedings of ACM Conference (Conference’17). ACM.

Tolpegin, V., Truex, S., Gursoy, M. E., and Liu, L. (2020). Data poisoning attacks against federated learning systems.

Wang, T. et al. (2022). Federated learning framework based on trimmed mean aggregation rules. SSRN.