Probabilistic Flow Circuits: Towards Unified Deep Models for Tractable Probabilistic Inference

Sahil Sidheekh, Kristian Kersting, Sriraam Natarajan
Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence, PMLR 216:1964-1973, 2023.

Abstract

We consider the problem of increasing the expressivity of probabilistic circuits by augmenting them with the successful generative models of normalizing flows. To this effect, we theoretically establish the requirement of decomposability for such combinations to retain tractability of the learned models. Our model, called Probabilistic Flow Circuits, essentially extends circuits by allowing for normalizing flows at the leaves. Our empirical evaluation clearly establishes the expressivity and tractability of this new class of probabilistic circuits.

Cite this Paper

BibTeX
@InProceedings{pmlr-v216-sidheekh23a, title = {Probabilistic Flow Circuits: Towards Unified Deep Models for Tractable Probabilistic Inference}, author = {Sidheekh, Sahil and Kersting, Kristian and Natarajan, Sriraam}, booktitle = {Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence}, pages = {1964--1973}, year = {2023}, editor = {Evans, Robin J. and Shpitser, Ilya}, volume = {216}, series = {Proceedings of Machine Learning Research}, month = {31 Jul--04 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v216/sidheekh23a/sidheekh23a.pdf}, url = {https://proceedings.mlr.press/v216/sidheekh23a.html}, abstract = {We consider the problem of increasing the expressivity of probabilistic circuits by augmenting them with the successful generative models of normalizing flows. To this effect, we theoretically establish the requirement of decomposability for such combinations to retain tractability of the learned models. Our model, called Probabilistic Flow Circuits, essentially extends circuits by allowing for normalizing flows at the leaves. Our empirical evaluation clearly establishes the expressivity and tractability of this new class of probabilistic circuits.} }
Endnote
%0 Conference Paper %T Probabilistic Flow Circuits: Towards Unified Deep Models for Tractable Probabilistic Inference %A Sahil Sidheekh %A Kristian Kersting %A Sriraam Natarajan %B Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2023 %E Robin J. Evans %E Ilya Shpitser %F pmlr-v216-sidheekh23a %I PMLR %P 1964--1973 %U https://proceedings.mlr.press/v216/sidheekh23a.html %V 216 %X We consider the problem of increasing the expressivity of probabilistic circuits by augmenting them with the successful generative models of normalizing flows. To this effect, we theoretically establish the requirement of decomposability for such combinations to retain tractability of the learned models. Our model, called Probabilistic Flow Circuits, essentially extends circuits by allowing for normalizing flows at the leaves. Our empirical evaluation clearly establishes the expressivity and tractability of this new class of probabilistic circuits.
APA
Sidheekh, S., Kersting, K. & Natarajan, S.. (2023). Probabilistic Flow Circuits: Towards Unified Deep Models for Tractable Probabilistic Inference. Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 216:1964-1973 Available from https://proceedings.mlr.press/v216/sidheekh23a.html.

Related Material