Towards Unbounded Machine Unlearning

• URL: http://arxiv.org/abs/2302.09880v3
• Date: Mon, 30 Oct 2023 10:14:02 GMT
• Title: Towards Unbounded Machine Unlearning
• Authors: Meghdad Kurmanji, Peter Triantafillou, Jamie Hayes, Eleni Triantafillou
• Abstract summary: We study unlearning for different applications (RB, RC, UP), with the view that each has its own desiderata, definitions for forgetting' and associated metrics for forget quality. For UP, we propose a novel adaptation of a strong Membership Inference Attack for unlearning. We also propose SCRUB, a novel unlearning algorithm, which is consistently a top performer for forget quality across the different application-dependent metrics for RB, RC, and UP.
• Score: 13.31957848633701
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deep machine unlearning is the problem of `removing' from a trained neural network a subset of its training set. This problem is very timely and has many applications, including the key tasks of removing biases (RB), resolving confusion (RC) (caused by mislabelled data in trained models), as well as allowing users to exercise their `right to be forgotten' to protect User Privacy (UP). This paper is the first, to our knowledge, to study unlearning for different applications (RB, RC, UP), with the view that each has its own desiderata, definitions for `forgetting' and associated metrics for forget quality. For UP, we propose a novel adaptation of a strong Membership Inference Attack for unlearning. We also propose SCRUB, a novel unlearning algorithm, which is the only method that is consistently a top performer for forget quality across the different application-dependent metrics for RB, RC, and UP. At the same time, SCRUB is also consistently a top performer on metrics that measure model utility (i.e. accuracy on retained data and generalization), and is more efficient than previous work. The above are substantiated through a comprehensive empirical evaluation against previous state-of-the-art.

Related papers

Err

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.