Interpolation Technique to Speed Up Gradients Propagation in Neural ODEs

Part of Advances in Neural Information Processing Systems 33 (NeurIPS 2020)

AuthorFeedback »Bibtex »MetaReview »Paper »Review »Supplemental »

Authors

Talgat Daulbaev, Alexandr Katrutsa, Larisa Markeeva, Julia Gusak, Andrzej Cichocki, Ivan Oseledets

Abstract

<p>We propose a simple interpolation-based method for the efficient approximation of gradients in neural ODE models. We compare it with reverse dynamic method (known in literature as “adjoint method”) to train neural ODEs on classification, density estimation and inference approximation tasks. We also propose a theoretical justification of our approach using logarithmic norm formalism. As a result, our method allows faster model training than the reverse dynamic method what was confirmed and validated by extensive numerical experiments for several standard benchmarks.</p>