BaCaDI: Bayesian Causal Discovery with Unknown Interventions

Alexander Hägele, Jonas Rothfuss, Lars Lorch, Vignesh Ram Somnath, Bernhard Schölkopf, Andreas Krause
Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:1411-1436, 2023.

Abstract

Inferring causal structures from experimentation is a central task in many domains. For example, in biology, recent advances allow us to obtain single-cell expression data under multiple interventions such as drugs or gene knockouts. However, the targets of the interventions are often uncertain or unknown and the number of observations limited. As a result, standard causal discovery methods can no longer be reliably used. To fill this gap, we propose a Bayesian framework (BaCaDI) for discovering and reasoning about the causal structure that underlies data generated under various unknown experimental or interventional conditions. BaCaDI is fully differentiable, which allows us to infer the complex joint posterior over the intervention targets and the causal structure via efficient gradient-based variational inference. In experiments on synthetic causal discovery tasks and simulated gene-expression data, BaCaDI outperforms related methods in identifying causal structures and intervention targets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v206-hagele23a, title = {BaCaDI: Bayesian Causal Discovery with Unknown Interventions}, author = {H\"agele, Alexander and Rothfuss, Jonas and Lorch, Lars and Somnath, Vignesh Ram and Sch\"olkopf, Bernhard and Krause, Andreas}, booktitle = {Proceedings of The 26th International Conference on Artificial Intelligence and Statistics}, pages = {1411--1436}, year = {2023}, editor = {Ruiz, Francisco and Dy, Jennifer and van de Meent, Jan-Willem}, volume = {206}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v206/hagele23a/hagele23a.pdf}, url = {https://proceedings.mlr.press/v206/hagele23a.html}, abstract = {Inferring causal structures from experimentation is a central task in many domains. For example, in biology, recent advances allow us to obtain single-cell expression data under multiple interventions such as drugs or gene knockouts. However, the targets of the interventions are often uncertain or unknown and the number of observations limited. As a result, standard causal discovery methods can no longer be reliably used. To fill this gap, we propose a Bayesian framework (BaCaDI) for discovering and reasoning about the causal structure that underlies data generated under various unknown experimental or interventional conditions. BaCaDI is fully differentiable, which allows us to infer the complex joint posterior over the intervention targets and the causal structure via efficient gradient-based variational inference. In experiments on synthetic causal discovery tasks and simulated gene-expression data, BaCaDI outperforms related methods in identifying causal structures and intervention targets.} }
Endnote
%0 Conference Paper %T BaCaDI: Bayesian Causal Discovery with Unknown Interventions %A Alexander Hägele %A Jonas Rothfuss %A Lars Lorch %A Vignesh Ram Somnath %A Bernhard Schölkopf %A Andreas Krause %B Proceedings of The 26th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2023 %E Francisco Ruiz %E Jennifer Dy %E Jan-Willem van de Meent %F pmlr-v206-hagele23a %I PMLR %P 1411--1436 %U https://proceedings.mlr.press/v206/hagele23a.html %V 206 %X Inferring causal structures from experimentation is a central task in many domains. For example, in biology, recent advances allow us to obtain single-cell expression data under multiple interventions such as drugs or gene knockouts. However, the targets of the interventions are often uncertain or unknown and the number of observations limited. As a result, standard causal discovery methods can no longer be reliably used. To fill this gap, we propose a Bayesian framework (BaCaDI) for discovering and reasoning about the causal structure that underlies data generated under various unknown experimental or interventional conditions. BaCaDI is fully differentiable, which allows us to infer the complex joint posterior over the intervention targets and the causal structure via efficient gradient-based variational inference. In experiments on synthetic causal discovery tasks and simulated gene-expression data, BaCaDI outperforms related methods in identifying causal structures and intervention targets.
APA
Hägele, A., Rothfuss, J., Lorch, L., Somnath, V.R., Schölkopf, B. & Krause, A.. (2023). BaCaDI: Bayesian Causal Discovery with Unknown Interventions. Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 206:1411-1436 Available from https://proceedings.mlr.press/v206/hagele23a.html.

Related Material