A Huber Loss Minimization Approach to Byzantine Robust Federated Learning
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (19) · 21806-21814
DOI: 10.1609/aaai.v38i19.30181 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Puning Zhao und 2 weiteren Autor:innen. Die Arbeit wurde bisher 28-mal zitiert (Crossref).
Abstract (Originalsprache)
Federated learning systems are susceptible to adversarial attacks. To combat this, we introduce a novel aggregator based on Huber loss minimization, and provide a comprehensive theoretical analysis. Under independent and identically distributed (i.i.d) assumption, our approach has several advantages compared to existing methods. Firstly, it has optimal dependence on epsilon, which stands for the ratio of attacked clients. Secondly, our approach does not need precise knowledge of epsilon. Thirdly, it allows different clients to have unequal data sizes. We then broaden our analysis to include non-i.i.d data, such that clients have slightly different distributions.
Zitieren
APA
Zhao, P., Yu, F., & Wan, Z. (2024). A Huber Loss Minimization Approach to Byzantine Robust Federated Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 38(19), 21806-21814. https://doi.org/10.1609/aaai.v38i19.30181
Vancouver
Zhao P, Yu F, Wan Z. A Huber Loss Minimization Approach to Byzantine Robust Federated Learning. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(19):21806-21814. doi:10.1609/aaai.v38i19.30181
BibTeX
@article{Zhao2024,
author = {Zhao, Puning and Yu, Fei and Wan, Zhiguo},
title = {A Huber Loss Minimization Approach to Byzantine Robust Federated Learning},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {19},
pages = {21806--21814},
doi = {10.1609/aaai.v38i19.30181}
}
EUVIMED