Learning from Pairwise Marginal Independencies

Johannes Textor Utrecht University, Alexander Idelberger, Maciej Liskiewicz
Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, PMLR R13:654-663, 2015.

Abstract

Dependency graphs (also called association graphs or bidirected graphs) represent marginal independencies amongst a set of variables. We give a characterization of the directed acyclic graphs (DAGs) that faithfully explain a given dependency graph in terms of their transitive closures, and use it to efficiently enumerate such structures. Our results map out the space of faithful causal models for given marginal independence relations, and show to which extent causal inference is possible without using conditional independence tests.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR13-university15o, title = {Learning from Pairwise Marginal Independencies}, author = {University, Johannes Textor Utrecht and Idelberger, Alexander and Liskiewicz, Maciej}, booktitle = {Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence}, pages = {654--663}, year = {2015}, editor = {Meila, Marina and Heskes, Tom}, volume = {R13}, series = {Proceedings of Machine Learning Research}, month = {12--16 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r13/main/assets/university15o/university15o.pdf}, url = {https://proceedings.mlr.press/r13/university15o.html}, abstract = {Dependency graphs (also called association graphs or bidirected graphs) represent marginal independencies amongst a set of variables. We give a characterization of the directed acyclic graphs (DAGs) that faithfully explain a given dependency graph in terms of their transitive closures, and use it to efficiently enumerate such structures. Our results map out the space of faithful causal models for given marginal independence relations, and show to which extent causal inference is possible without using conditional independence tests.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Learning from Pairwise Marginal Independencies %A Johannes Textor Utrecht University %A Alexander Idelberger %A Maciej Liskiewicz %B Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2015 %E Marina Meila %E Tom Heskes %F pmlr-vR13-university15o %I PMLR %P 654--663 %U https://proceedings.mlr.press/r13/university15o.html %V R13 %X Dependency graphs (also called association graphs or bidirected graphs) represent marginal independencies amongst a set of variables. We give a characterization of the directed acyclic graphs (DAGs) that faithfully explain a given dependency graph in terms of their transitive closures, and use it to efficiently enumerate such structures. Our results map out the space of faithful causal models for given marginal independence relations, and show to which extent causal inference is possible without using conditional independence tests. %Z Reissued by PMLR on 04 October 2026.
APA
University, J.T.U., Idelberger, A. & Liskiewicz, M.. (2015). Learning from Pairwise Marginal Independencies. Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R13:654-663 Available from https://proceedings.mlr.press/r13/university15o.html. Reissued by PMLR on 04 October 2026.

Related Material