MUSO: Achieving Exact Machine Unlearning in Over-Parameterized Regimes

Abstract

Machine unlearning (MU) is to make a well-trained model behave as if it had never been trained on specific data. In today's over-parameterized models, dominated by neural networks, a common approach is to manually relabel data and fine-tune the well-trained model. It can approximate the MU model in the output space, but the question remains whether it can achieve exact MU, i.e., in the parameter space. We answer this question by employing random feature techniques to construct an analytical framework. Under the premise of model optimization via stochastic gradient descent, we theoretically demonstrated that over-parameterized linear models can achieve exact MU through relabeling specific data. We also extend this work to real-world nonlinear networks and propose an alternating optimization algorithm that unifies the tasks of unlearning and relabeling. The algorithm's effectiveness, confirmed through numerical experiments, highlights its superior performance in unlearning across various scenarios compared to current state-of-the-art methods, particularly excelling over similar relabeling-based MU approaches.

Accepted by Machine Learning Journal

References in corpus (17)

• Neural Tangent Kernel: Convergence and Generalization in Neural Networks
• Understanding Black-box Predictions via Influence Functions
• The European Union general data protection regulation: what it is and what it means
• Privacy Risks of Securing Machine Learning Models against Adversarial Examples
• Characterizing Implicit Bias in Terms of Optimization Geometry
• Two models of double descent for weak features
• Systematic Evaluation of Privacy Risks of Machine Learning Models
• Making AI Forget You: Data Deletion in Machine Learning
• Approximate Data Deletion from Machine Learning Models
• On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning
• Theory on Forgetting and Generalization of Continual Learning
• Model Sparsity Can Simplify Machine Unlearning
• PUMA: Performance Unchanged Model Augmentation for Training Data Removal
• Can Bad Teaching Induce Forgetting? Unlearning in Deep Networks using an Incompetent Teacher
• Precise Asymptotic Analysis of Deep Random Feature Models
• Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening
• Learning to Unlearn: Instance-wise Unlearning for Pre-trained Classifiers