On the definition and computation of causal treewidth

Yizuo Chen, Adnan Darwiche
Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence, PMLR 180:368-377, 2022.

Abstract

Causal treewidth is a recently introduced notion allowing one to speed up Bayesian network inference and to bound its complexity in the presence of functional dependencies (causal mechanisms) whose identities are unknown. Causal treewidth is no greater than treewidth and can be bounded even when treewidth is unbounded. The utility of causal treewidth has been illustrated recently in the context of causal inference and model-based supervised learning. However, the current definition of causal treewidth is descriptive rather than perspective, therefore limiting its full exploitation in a practical setting. We provide an extensive study of causal treewidth in this paper which moves us closer to realizing the full computational potential of this notion both theoretically and practically.

Cite this Paper

BibTeX
@InProceedings{pmlr-v180-chen22f, title = {On the definition and computation of causal treewidth}, author = {Chen, Yizuo and Darwiche, Adnan}, booktitle = {Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence}, pages = {368--377}, year = {2022}, editor = {Cussens, James and Zhang, Kun}, volume = {180}, series = {Proceedings of Machine Learning Research}, month = {01--05 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v180/chen22f/chen22f.pdf}, url = {https://proceedings.mlr.press/v180/chen22f.html}, abstract = {Causal treewidth is a recently introduced notion allowing one to speed up Bayesian network inference and to bound its complexity in the presence of functional dependencies (causal mechanisms) whose identities are unknown. Causal treewidth is no greater than treewidth and can be bounded even when treewidth is unbounded. The utility of causal treewidth has been illustrated recently in the context of causal inference and model-based supervised learning. However, the current definition of causal treewidth is descriptive rather than perspective, therefore limiting its full exploitation in a practical setting. We provide an extensive study of causal treewidth in this paper which moves us closer to realizing the full computational potential of this notion both theoretically and practically.} }
Endnote
%0 Conference Paper %T On the definition and computation of causal treewidth %A Yizuo Chen %A Adnan Darwiche %B Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2022 %E James Cussens %E Kun Zhang %F pmlr-v180-chen22f %I PMLR %P 368--377 %U https://proceedings.mlr.press/v180/chen22f.html %V 180 %X Causal treewidth is a recently introduced notion allowing one to speed up Bayesian network inference and to bound its complexity in the presence of functional dependencies (causal mechanisms) whose identities are unknown. Causal treewidth is no greater than treewidth and can be bounded even when treewidth is unbounded. The utility of causal treewidth has been illustrated recently in the context of causal inference and model-based supervised learning. However, the current definition of causal treewidth is descriptive rather than perspective, therefore limiting its full exploitation in a practical setting. We provide an extensive study of causal treewidth in this paper which moves us closer to realizing the full computational potential of this notion both theoretically and practically.
APA
Chen, Y. & Darwiche, A.. (2022). On the definition and computation of causal treewidth. Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 180:368-377 Available from https://proceedings.mlr.press/v180/chen22f.html.

Related Material