arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.07026cs.LGcs.CRcs.DCstat.ML
收藏

联邦元学习的后门攻击

Backdoor Attacks on Federated Meta-Learning

University of Southern California(南加州大学)

Chien-Lun Chen, Leana Golubchik, Marco Paolieri

摘要:

联邦学习允许多个用户协作训练共享的分类模型,同时保护数据隐私。这种方法,模型更新由中央服务器聚合,被证明容易受到中毒后门攻击:恶意用户可以改变共享模型,任意分类来自给定类的特定输入。在本文中,我们分析了后门攻击对联邦元学习的影响,用户训练一个模型,该模型可以适应不同的输出类集,只使用几个例子。虽然原则上,适应能力可以使联邦学习框架对后门攻击更加鲁棒(当新的训练样本是良性的时),但我们发现即使是一次性攻击也可以非常成功,并且在额外的训练后仍然存在。为了解决这些漏洞,我们提出了一个防御机制的灵感来自匹配网络,其中输入的类是从其特征的相似性与支持集的标记的例子预测。通过从与联邦共享的模型中删除决策逻辑,后门攻击的成功率和持久性大大降低。

英文摘要:

Federated learning allows multiple users to collaboratively train a shared classification model while preserving data privacy. This approach, where model updates are aggregated by a central server, was shown to be vulnerable to poisoning backdoor attacks: a malicious user can alter the shared model to arbitrarily classify specific inputs from a given class. In this paper, we analyze the effects of backdoor attacks on federated meta-learning, where users train a model that can be adapted to different sets of output classes using only a few examples. While the ability to adapt could, in principle, make federated learning frameworks more robust to backdoor attacks (when new training examples are benign), we find that even 1-shot~attacks can be very successful and persist after additional training. To address these vulnerabilities, we propose a defense mechanism inspired by matching networks, where the class of an input is predicted from the similarity of its features with a support set of labeled examples. By removing the decision logic from the model shared with the federation, success and persistence of backdoor attacks are greatly reduced.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑