A First Order Meta Stackelberg Method for Robust Federated Learning (Technical Report)

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Li, Henger, Xu, Tianyi, Li, Tao, Pan, Yunian, Zhu, Quanyan, Zheng, Zizhan
Format: Preprint
Publié: 2023
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908718810005504
author Li, Henger
Xu, Tianyi
Li, Tao
Pan, Yunian
Zhu, Quanyan
Zheng, Zizhan
author_facet Li, Henger
Xu, Tianyi
Li, Tao
Pan, Yunian
Zhu, Quanyan
Zheng, Zizhan
contents Recent research efforts indicate that federated learning (FL) systems are vulnerable to a variety of security breaches. While numerous defense strategies have been suggested, they are mainly designed to counter specific attack patterns and lack adaptability, rendering them less effective when facing uncertain or adaptive threats. This work models adversarial FL as a Bayesian Stackelberg Markov game (BSMG) between the defender and the attacker to address the lack of adaptability to uncertain adaptive attacks. We further devise an effective meta-learning technique to solve for the Stackelberg equilibrium, leading to a resilient and adaptable defense. The experiment results suggest that our meta-Stackelberg learning approach excels in combating intense model poisoning and backdoor attacks of indeterminate types.
format Preprint
id arxiv_https___arxiv_org_abs_2306_13273
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A First Order Meta Stackelberg Method for Robust Federated Learning (Technical Report)
Li, Henger
Xu, Tianyi
Li, Tao
Pan, Yunian
Zhu, Quanyan
Zheng, Zizhan
Cryptography and Security
Computer Science and Game Theory
Recent research efforts indicate that federated learning (FL) systems are vulnerable to a variety of security breaches. While numerous defense strategies have been suggested, they are mainly designed to counter specific attack patterns and lack adaptability, rendering them less effective when facing uncertain or adaptive threats. This work models adversarial FL as a Bayesian Stackelberg Markov game (BSMG) between the defender and the attacker to address the lack of adaptability to uncertain adaptive attacks. We further devise an effective meta-learning technique to solve for the Stackelberg equilibrium, leading to a resilient and adaptable defense. The experiment results suggest that our meta-Stackelberg learning approach excels in combating intense model poisoning and backdoor attacks of indeterminate types.
title A First Order Meta Stackelberg Method for Robust Federated Learning (Technical Report)
topic Cryptography and Security
Computer Science and Game Theory
url https://arxiv.org/abs/2306.13273