Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations

Winnie Xu, Ricky T. Q. Chen, Xuechen Li, David Duvenaud
Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, PMLR 151:721-738, 2022.

Abstract

We perform scalable approximate inference in continuous-depth Bayesian neural networks. In this model class, uncertainty about separate weights in each layer gives hidden units that follow a stochastic differential equation. We demonstrate gradient-based stochastic variational inference in this infinite-parameter setting, producing arbitrarily-flexible approximate posteriors. We also derive a novel gradient estimator that approaches zero variance as the approximate posterior over weights approaches the true posterior. This approach brings continuous-depth Bayesian neural nets to a competitive comparison against discrete-depth alternatives, while inheriting the memory-efficient training and tunable precision of Neural ODEs.

Cite this Paper


BibTeX
@InProceedings{pmlr-v151-xu22a, title = { Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations }, author = {Xu, Winnie and Chen, Ricky T. Q. and Li, Xuechen and Duvenaud, David}, booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics}, pages = {721--738}, year = {2022}, editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel}, volume = {151}, series = {Proceedings of Machine Learning Research}, month = {28--30 Mar}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v151/xu22a/xu22a.pdf}, url = {https://proceedings.mlr.press/v151/xu22a.html}, abstract = { We perform scalable approximate inference in continuous-depth Bayesian neural networks. In this model class, uncertainty about separate weights in each layer gives hidden units that follow a stochastic differential equation. We demonstrate gradient-based stochastic variational inference in this infinite-parameter setting, producing arbitrarily-flexible approximate posteriors. We also derive a novel gradient estimator that approaches zero variance as the approximate posterior over weights approaches the true posterior. This approach brings continuous-depth Bayesian neural nets to a competitive comparison against discrete-depth alternatives, while inheriting the memory-efficient training and tunable precision of Neural ODEs. } }
Endnote
%0 Conference Paper %T Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations %A Winnie Xu %A Ricky T. Q. Chen %A Xuechen Li %A David Duvenaud %B Proceedings of The 25th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2022 %E Gustau Camps-Valls %E Francisco J. R. Ruiz %E Isabel Valera %F pmlr-v151-xu22a %I PMLR %P 721--738 %U https://proceedings.mlr.press/v151/xu22a.html %V 151 %X We perform scalable approximate inference in continuous-depth Bayesian neural networks. In this model class, uncertainty about separate weights in each layer gives hidden units that follow a stochastic differential equation. We demonstrate gradient-based stochastic variational inference in this infinite-parameter setting, producing arbitrarily-flexible approximate posteriors. We also derive a novel gradient estimator that approaches zero variance as the approximate posterior over weights approaches the true posterior. This approach brings continuous-depth Bayesian neural nets to a competitive comparison against discrete-depth alternatives, while inheriting the memory-efficient training and tunable precision of Neural ODEs.
APA
Xu, W., Chen, R.T.Q., Li, X. & Duvenaud, D.. (2022). Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations . Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 151:721-738 Available from https://proceedings.mlr.press/v151/xu22a.html.

Related Material