Using Autodiff to Estimate Posterior Moments, Marginals and Samples

Sam Bowyer, Thomas Heap, Laurence Aitchison
Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, PMLR 244:394-417, 2024.

Abstract

Importance sampling is a popular technique in Bayesian inference: by reweighting samples drawn from a proposal distribution we are able to obtain samples and moment estimates from a Bayesian posterior over latent variables. Recent work, however, indicates that importance sampling scales poorly — in order to accurately approximate the true posterior, the required number of importance samples grows is exponential in the number of latent variables [Chatterjee and Diaconis, 2018]. Massively parallel importance sampling works around this issue by drawing $K$ samples for each of the $n$ latent variables and reasoning about all $K^n$ combinations of latent samples. In principle, we can reason efficiently over $K^n$ combinations of samples by exploiting conditional independencies in the generative model. However, in practice this requires complex algorithms that traverse backwards through the graphical model, and we need separate backward traversals for each computation (posterior expectations, marginals and samples). Our contribution is to exploit the source term trick from physics to entirely avoid the need to hand-write backward traversals. Instead, we demonstrate how to simply and easily compute all the required quantities — posterior expectations, marginals and samples — by differentiating through a slightly modified marginal likelihood estimator.

Cite this Paper

BibTeX
@InProceedings{pmlr-v244-bowyer24a, title = {Using Autodiff to Estimate Posterior Moments, Marginals and Samples}, author = {Bowyer, Sam and Heap, Thomas and Aitchison, Laurence}, booktitle = {Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence}, pages = {394--417}, year = {2024}, editor = {Kiyavash, Negar and Mooij, Joris M.}, volume = {244}, series = {Proceedings of Machine Learning Research}, month = {15--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v244/main/assets/bowyer24a/bowyer24a.pdf}, url = {https://proceedings.mlr.press/v244/bowyer24a.html}, abstract = {Importance sampling is a popular technique in Bayesian inference: by reweighting samples drawn from a proposal distribution we are able to obtain samples and moment estimates from a Bayesian posterior over latent variables. Recent work, however, indicates that importance sampling scales poorly — in order to accurately approximate the true posterior, the required number of importance samples grows is exponential in the number of latent variables [Chatterjee and Diaconis, 2018]. Massively parallel importance sampling works around this issue by drawing $K$ samples for each of the $n$ latent variables and reasoning about all $K^n$ combinations of latent samples. In principle, we can reason efficiently over $K^n$ combinations of samples by exploiting conditional independencies in the generative model. However, in practice this requires complex algorithms that traverse backwards through the graphical model, and we need separate backward traversals for each computation (posterior expectations, marginals and samples). Our contribution is to exploit the source term trick from physics to entirely avoid the need to hand-write backward traversals. Instead, we demonstrate how to simply and easily compute all the required quantities — posterior expectations, marginals and samples — by differentiating through a slightly modified marginal likelihood estimator.} }
Endnote
%0 Conference Paper %T Using Autodiff to Estimate Posterior Moments, Marginals and Samples %A Sam Bowyer %A Thomas Heap %A Laurence Aitchison %B Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2024 %E Negar Kiyavash %E Joris M. Mooij %F pmlr-v244-bowyer24a %I PMLR %P 394--417 %U https://proceedings.mlr.press/v244/bowyer24a.html %V 244 %X Importance sampling is a popular technique in Bayesian inference: by reweighting samples drawn from a proposal distribution we are able to obtain samples and moment estimates from a Bayesian posterior over latent variables. Recent work, however, indicates that importance sampling scales poorly — in order to accurately approximate the true posterior, the required number of importance samples grows is exponential in the number of latent variables [Chatterjee and Diaconis, 2018]. Massively parallel importance sampling works around this issue by drawing $K$ samples for each of the $n$ latent variables and reasoning about all $K^n$ combinations of latent samples. In principle, we can reason efficiently over $K^n$ combinations of samples by exploiting conditional independencies in the generative model. However, in practice this requires complex algorithms that traverse backwards through the graphical model, and we need separate backward traversals for each computation (posterior expectations, marginals and samples). Our contribution is to exploit the source term trick from physics to entirely avoid the need to hand-write backward traversals. Instead, we demonstrate how to simply and easily compute all the required quantities — posterior expectations, marginals and samples — by differentiating through a slightly modified marginal likelihood estimator.
APA
Bowyer, S., Heap, T. & Aitchison, L.. (2024). Using Autodiff to Estimate Posterior Moments, Marginals and Samples. Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 244:394-417 Available from https://proceedings.mlr.press/v244/bowyer24a.html.

Related Material