Towards Certified Unlearning for Deep Neural Networks

TL;DR

This paper extends certified unlearning techniques to deep neural networks, providing efficient computation methods and addressing nonconvergence and sequential unlearning challenges, validated through extensive experiments.

Contribution

It introduces simple techniques to adapt certified unlearning to nonconvex DNNs, including inverse Hessian approximation for efficiency and handling real-world unlearning scenarios.

Findings

Effective unlearning on three real-world datasets

Inverse Hessian approximation reduces computation time

Certified unlearning offers strong guarantees for DNNs

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.