Learning Acyclic Directed Mixed Graphs from Observations and Interventions

Jose M. Peña
Proceedings of the Eighth International Conference on Probabilistic Graphical Models, PMLR 52:392-402, 2016.

Abstract

We introduce a new family of mixed graphical models that consists of graphs with possibly directed, undirected and bidirected edges but without directed cycles. Moreover, there can be up to three edges between any pair of nodes. The new family includes Richardson’s acyclic directed mixed graphs, as well as Andersson-Madigan-Perlman chain graphs. These features imply that no family of mixed graphical models that we know of subsumes the new models. We also provide a causal interpretation of the new models as systems of structural equations with correlated errors. Finally, we describe an exact algorithm for learning the new models from observational and interventional data via answer set programming.

Cite this Paper

BibTeX
@InProceedings{pmlr-v52-pena16, title = {Learning Acyclic Directed Mixed Graphs from Observations and Interventions}, author = {Peña, Jose M.}, booktitle = {Proceedings of the Eighth International Conference on Probabilistic Graphical Models}, pages = {392--402}, year = {2016}, editor = {Antonucci, Alessandro and Corani, Giorgio and Campos}, Cassio Polpo}, volume = {52}, series = {Proceedings of Machine Learning Research}, address = {Lugano, Switzerland}, month = {06--09 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v52/pena16.pdf}, url = {https://proceedings.mlr.press/v52/pena16.html}, abstract = {We introduce a new family of mixed graphical models that consists of graphs with possibly directed, undirected and bidirected edges but without directed cycles. Moreover, there can be up to three edges between any pair of nodes. The new family includes Richardson’s acyclic directed mixed graphs, as well as Andersson-Madigan-Perlman chain graphs. These features imply that no family of mixed graphical models that we know of subsumes the new models. We also provide a causal interpretation of the new models as systems of structural equations with correlated errors. Finally, we describe an exact algorithm for learning the new models from observational and interventional data via answer set programming.} }
Endnote
%0 Conference Paper %T Learning Acyclic Directed Mixed Graphs from Observations and Interventions %A Jose M. Peña %B Proceedings of the Eighth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2016 %E Alessandro Antonucci %E Giorgio Corani %E Cassio Polpo Campos} %F pmlr-v52-pena16 %I PMLR %P 392--402 %U https://proceedings.mlr.press/v52/pena16.html %V 52 %X We introduce a new family of mixed graphical models that consists of graphs with possibly directed, undirected and bidirected edges but without directed cycles. Moreover, there can be up to three edges between any pair of nodes. The new family includes Richardson’s acyclic directed mixed graphs, as well as Andersson-Madigan-Perlman chain graphs. These features imply that no family of mixed graphical models that we know of subsumes the new models. We also provide a causal interpretation of the new models as systems of structural equations with correlated errors. Finally, we describe an exact algorithm for learning the new models from observational and interventional data via answer set programming.
RIS
TY - CPAPER TI - Learning Acyclic Directed Mixed Graphs from Observations and Interventions AU - Jose M. Peña BT - Proceedings of the Eighth International Conference on Probabilistic Graphical Models DA - 2016/08/15 ED - Alessandro Antonucci ED - Giorgio Corani ED - Cassio Polpo Campos} ID - pmlr-v52-pena16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 52 SP - 392 EP - 402 L1 - http://proceedings.mlr.press/v52/pena16.pdf UR - https://proceedings.mlr.press/v52/pena16.html AB - We introduce a new family of mixed graphical models that consists of graphs with possibly directed, undirected and bidirected edges but without directed cycles. Moreover, there can be up to three edges between any pair of nodes. The new family includes Richardson’s acyclic directed mixed graphs, as well as Andersson-Madigan-Perlman chain graphs. These features imply that no family of mixed graphical models that we know of subsumes the new models. We also provide a causal interpretation of the new models as systems of structural equations with correlated errors. Finally, we describe an exact algorithm for learning the new models from observational and interventional data via answer set programming. ER -
APA
Peña, J.M.. (2016). Learning Acyclic Directed Mixed Graphs from Observations and Interventions. Proceedings of the Eighth International Conference on Probabilistic Graphical Models, in Proceedings of Machine Learning Research 52:392-402 Available from https://proceedings.mlr.press/v52/pena16.html.

Related Material