Inference Networks for Sequential Monte Carlo in Graphical Models

Brooks Paige, Frank Wood
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:3040-3049, 2016.

Abstract

We introduce a new approach for amortizing inference in directed graphical models by learning heuristic approximations to stochastic inverses, designed specifically for use as proposal distributions in sequential Monte Carlo methods. We describe a procedure for constructing and learning a structured neural network which represents an inverse factorization of the graphical model, resulting in a conditional density estimator that takes as input particular values of the observed random variables, and returns an approximation to the distribution of the latent variables. This recognition model can be learned offline, independent from any particular dataset, prior to performing inference. The output of these networks can be used as automatically-learned high-quality proposal distributions to accelerate sequential Monte Carlo across a diverse range of problem settings.

Cite this Paper

BibTeX
@InProceedings{pmlr-v48-paige16, title = {Inference Networks for Sequential Monte Carlo in Graphical Models}, author = {Paige, Brooks and Wood, Frank}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {3040--3049}, year = {2016}, editor = {Balcan, Maria Florina and Weinberger, Kilian Q.}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/paige16.pdf}, url = {https://proceedings.mlr.press/v48/paige16.html}, abstract = {We introduce a new approach for amortizing inference in directed graphical models by learning heuristic approximations to stochastic inverses, designed specifically for use as proposal distributions in sequential Monte Carlo methods. We describe a procedure for constructing and learning a structured neural network which represents an inverse factorization of the graphical model, resulting in a conditional density estimator that takes as input particular values of the observed random variables, and returns an approximation to the distribution of the latent variables. This recognition model can be learned offline, independent from any particular dataset, prior to performing inference. The output of these networks can be used as automatically-learned high-quality proposal distributions to accelerate sequential Monte Carlo across a diverse range of problem settings.} }
Endnote
%0 Conference Paper %T Inference Networks for Sequential Monte Carlo in Graphical Models %A Brooks Paige %A Frank Wood %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-paige16 %I PMLR %P 3040--3049 %U https://proceedings.mlr.press/v48/paige16.html %V 48 %X We introduce a new approach for amortizing inference in directed graphical models by learning heuristic approximations to stochastic inverses, designed specifically for use as proposal distributions in sequential Monte Carlo methods. We describe a procedure for constructing and learning a structured neural network which represents an inverse factorization of the graphical model, resulting in a conditional density estimator that takes as input particular values of the observed random variables, and returns an approximation to the distribution of the latent variables. This recognition model can be learned offline, independent from any particular dataset, prior to performing inference. The output of these networks can be used as automatically-learned high-quality proposal distributions to accelerate sequential Monte Carlo across a diverse range of problem settings.
RIS
TY - CPAPER TI - Inference Networks for Sequential Monte Carlo in Graphical Models AU - Brooks Paige AU - Frank Wood BT - Proceedings of The 33rd International Conference on Machine Learning DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-paige16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 3040 EP - 3049 L1 - http://proceedings.mlr.press/v48/paige16.pdf UR - https://proceedings.mlr.press/v48/paige16.html AB - We introduce a new approach for amortizing inference in directed graphical models by learning heuristic approximations to stochastic inverses, designed specifically for use as proposal distributions in sequential Monte Carlo methods. We describe a procedure for constructing and learning a structured neural network which represents an inverse factorization of the graphical model, resulting in a conditional density estimator that takes as input particular values of the observed random variables, and returns an approximation to the distribution of the latent variables. This recognition model can be learned offline, independent from any particular dataset, prior to performing inference. The output of these networks can be used as automatically-learned high-quality proposal distributions to accelerate sequential Monte Carlo across a diverse range of problem settings. ER -
APA
Paige, B. & Wood, F.. (2016). Inference Networks for Sequential Monte Carlo in Graphical Models. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:3040-3049 Available from https://proceedings.mlr.press/v48/paige16.html.

Related Material