Online Structured Laplace Approximations for Overcoming Catastrophic Forgetting

Part of Advances in Neural Information Processing Systems 31 (NeurIPS 2018)

Bibtex »Metadata »Paper »Reviews »Supplemental »

Authors

Hippolyt Ritter, Aleksandar Botev, David Barber

Abstract

<p>We introduce the Kronecker factored online Laplace approximation for overcoming catastrophic forgetting in neural networks. The method is grounded in a Bayesian online learning framework, where we recursively approximate the posterior after every task with a Gaussian, leading to a quadratic penalty on changes to the weights. The Laplace approximation requires calculating the Hessian around a mode, which is typically intractable for modern architectures. In order to make our method scalable, we leverage recent block-diagonal Kronecker factored approximations to the curvature. Our algorithm achieves over 90% test accuracy across a sequence of 50 instantiations of the permuted MNIST dataset, substantially outperforming related methods for overcoming catastrophic forgetting.</p>