2023/02/20 by Meghdad Kurmanji, Peter Triantafillou, Kurmanji, Meghdad +3 · 65 citations
Computer Science · Mathematics · Medicine · Psychology · #Adaptation (eye) #Advanced Neural Network Applications #Artificial intelligence #Artificial neural network #COVID-19 diagnosis using AI #Cognitive psychology #Computer science #Computer security #Confusion #Cryptography and Security (cs.CR) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Forgetting #Generalization #Inference #Key (lock) #Machine Learning (cs.LG) #Machine learning #Mathematics #Psychology #Quality (philosophy) #Set (abstract data type)
paper · open access · doi:10.48550/arxiv.2302.09880
published in Warwick Research Archive Portal (University of Warwick) (University of Warwick)
openalex publication_date 2023/02/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
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.