Unlearning Backdoor Attacks through Gradient-Based Model Pruning

arXiv:2405.03918v1 Announce Type: new
Abstract: In the era of increasing concerns over cybersecurity threats, defending against backdoor attacks is paramount in ensuring the integrity and reliability of machine learning models. However, many existing approaches require substantial amounts of data for effective mitigation, posing significant challenges in practical deployment. To address this, we propose a novel approach to counter backdoor attacks by treating their mitigation as an unlearning task. We tackle this challenge through a targeted model pruning strategy, leveraging unlearning …

abstract arxiv attacks backdoor challenges concerns cs.lg cybersecurity cybersecurity threats data deployment gradient however integrity machine machine learning machine learning models practical pruning reliability threats through type unlearning