Stochastic Gradient Langevin Unlearning

arXiv:2403.17105v1 Announce Type: new
Abstract: ``The right to be forgotten'' ensured by laws for user data privacy becomes increasingly important. Machine unlearning aims to efficiently remove the effect of certain data points on the trained model parameters so that it can be approximately the same as if one retrains the model from scratch. This work proposes stochastic gradient Langevin unlearning, the first unlearning framework based on noisy stochastic gradient descent (SGD) with privacy guarantees for approximate unlearning problems under convexity …

abstract arxiv cs.cr cs.lg data data privacy gradient laws machine parameters privacy scratch stochastic type unlearning user data work