Abstract
Federated learning is a distributed setting where multiple participants jointly train a machine learning model without exchanging data. Recent work has found that federated learning is vulnerable to backdoor model poisoning attacks, where an attacker leverages the unique environment to submit malicious model updates. To address these malicious participants, several Byzantine-Tolerant aggregation methods have been applied to the federated learning setting, including Krum, Multi-Krum, RFA, and Norm-Difference Clipping. In this work, we analyze the effectiveness and limits of each aggregation method and provide a thorough analysis of their success in various fixed-frequency attack settings. Further, we analyze the fairness of such aggregation methods on the success of the model on its intended tasks. Our results indicate that only one defense can successfully mitigate attacks in all attack scenarios, but a significant fairness issue is observed, highlighting the issues with preventing malicious attacks in a federated setting.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bagdasaryan, E., Veit, A., Hua, Y., Estrin, D., Shmatikov, V.: How to backdoor federated learning. In: International Conference on Artificial Intelligence and Statistics, pp. 2938–2948. PMLR (2020)
Bhagoji, A.N., Chakraborty, S., Mittal, P., Calo, S.: Analyzing federated learning through an adversarial lens. In: International Conference on Machine Learning, pp. 634–643. PMLR (2019)
Blanchard, P., El Mhamdi, E.M., Guerraoui, R., Stainer, J.: Machine learning with adversaries: Byzantine tolerant gradient descent. In: Proceedings of the 31st International Conference on Neural Information Processing Systems (2017)
Chen, Y., Su, L., Xu, J.: Distributed statistical machine learning in adversarial settings: Byzantine gradient descent. Proc. ACM Measur. Anal. Comput. Syst. 1(2), 1–25 (2017)
Cohen, G., Afshar, S., Tapson, J., Van Schaik, A.: EMNIST: extending MNIST to handwritten letters. In: 2017 International Joint Conference on Neural Networks (IJCNN), pp. 2921–2926. IEEE (2017)
Fang, M., Cao, X., Jia, J., Gong, N.: Local model poisoning attacks to Byzantine-Robust federated learning. In: 29th USENIX Security Symposium (USENIX Security 2020), pp. 1605–1622 (2020)
Fung, C., Yoon, C.J., Beschastnikh, I.: Mitigating sybils in federated learning poisoning. arXiv preprint arXiv:1808.04866 (2018)
Kairouz, P., et al.: Advances and open problems in federated learning. arXiv preprint arXiv:1912.04977 (2019)
Krizhevsky, A., Hinton, G., et al.: Learning multiple layers of features from tiny images (2009)
Kusetogullari, H., Yavariabdi, A., Cheddad, A., Grahn, H., Johan, H.: ARDIS: a Swedish historical handwritten digit dataset. Neural Comput. Appl. 32(21), 16505–16518 (2020). https://doi.org/10.1007/s00521-019-04163-3
Li, L., Fan, Y., Tse, M., Lin, K.Y.: A review of applications in federated learning. Comput. Ind. Eng. 149, 106854 (2020)
Li, X., Huang, K., Yang, W., Wang, S., Zhang, Z.: On the convergence of FedAvg on Non-IID data. arXiv preprint arXiv:1907.02189 (2019)
Lim, W.Y.B., et al.: Federated learning in mobile edge networks: a comprehensive survey. IEEE Commun. Surv. Tutor. 22(3), 2031–2063 (2020)
Long, G., Tan, Y., Jiang, J., Zhang, C.: Federated learning for open banking. In: Yang, Q., Fan, L., Yu, H. (eds.) Federated Learning. LNCS (LNAI), vol. 12500, pp. 240–254. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-63076-8_17
Luo, J., et al.: Real-world image datasets for federated learning. arXiv preprint arXiv:1910.11089 (2019)
McMahan, B., Moore, E., Ramage, D., Hampson, S., Aguera y Arcas, B.: Communication-efficient learning of deep networks from decentralized data. In: Artificial Intelligence and Statistics, pp. 1273–1282. PMLR (2017)
Paszke, A., et al.: PyTorch: an imperative style, high-performance deep learning library. In: Advances in Neural Information Processing Systems, vol. 32, pp. 8026–8037 (2019)
Pillutla, K., Kakade, S.M., Harchaoui, Z.: Robust aggregation for federated learning. arXiv preprint arXiv:1912.13445 (2019)
Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014)
Sun, Z., Kairouz, P., Suresh, A.T., McMahan, H.B.: Can you really backdoor federated learning? arXiv preprint arXiv:1911.07963 (2019)
Wang, H., et al.: Attack of the tails: yes, you really can backdoor federated learning. arXiv preprint arXiv:2007.05084 (2020)
Yin, D., Chen, Y., Kannan, R., Bartlett, P.: Byzantine-robust distributed learning: towards optimal statistical rates. In: International Conference on Machine Learning, pp. 5650–5659. PMLR (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Roszel, M., Norvill, R., State, R. (2022). An Analysis of Byzantine-Tolerant Aggregation Mechanisms on Model Poisoning in Federated Learning. In: Torra, V., Narukawa, Y. (eds) Modeling Decisions for Artificial Intelligence. MDAI 2022. Lecture Notes in Computer Science(), vol 13408. Springer, Cham. https://doi.org/10.1007/978-3-031-13448-7_12
Download citation
DOI: https://doi.org/10.1007/978-3-031-13448-7_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-13447-0
Online ISBN: 978-3-031-13448-7
eBook Packages: Computer ScienceComputer Science (R0)