Towards Certified Unlearning for Deep Neural Networks
Simple techniques extend certified unlearning to deep networks via inverse Hessian approximations, preserving guarantees for nonconvergence and sequential requests.
Published Aug 1, 2024arXiv ↗
Only vote on papers you've read. Sign in with GitHub to vote.
The paper advances certified unlearning for deep networks through efficient inverse Hessian approximations that preserve guarantees across nonconvex and sequential settings, but it buries epsilon bounds, omits dataset specifics, and leaves sequential error accumulation unexamined.
Abstract
In the field of machine unlearning, certified unlearning has been extensively studied in convex machine learning models due to its high efficiency and strong theoretical guarantees. However, its application to deep neural networks (DNNs), known for their highly nonconvex nature, still poses challenges. To bridge the gap between certified unlearning and DNNs, we propose several simple techniques to extend certified unlearning methods to nonconvex objectives. To reduce the time complexity, we develop an efficient computation method by inverse Hessian approximation without compromising certification guarantees. In addition, we extend our discussion of certification to nonconvergence training and sequential unlearning, considering that real-world users can send unlearning requests at different time points. Extensive experiments on three real-world datasets demonstrate the efficacy of our method and the advantages of certified unlearning in DNNs.