Unlearning with Fisher Masking

Yufang Liu, Yuanbin Wu, Aimin Zhou

Keywords: machine unlearning, Fisher Information

TL;DR: We develop a new unlearning strategy based on Fisher Masking which shows strong unlearning performances across different datasets and deep neural network structures.

Abstract: Machine unlearning aims to revoke some training data after learning in response to requests from users, model developers, and administrators. Most previous methods are based on direct fine tuning, which may neither remove data completely nor retain full performances on the remain data. In this work, we find that, by first masking some important parameters before fine tuning, the performances of unlearning could be significantly improved. We propose a new masking strategy tailored to unlearning based on Fisher information. Experiments on various datasets and network structures show the effectiveness of the method: without any fine tuning, the proposed Fisher masking could unlearn almost completely while maintaining most of the performance on the remain data. It also exhibits stronger stability comparing with other unlearning baselines.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Social Aspects of Machine Learning (eg, AI safety, fairness, privacy, interpretability, human-AI interaction, ethics)