Identifying Confounding from Causal Mechanism Shifts

Sarah Mameche, Jilles Vreeken, David Kaltenpoth
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:4897-4905, 2024.

Abstract

Causal discovery methods commonly assume that all data is independently and identically distributed (i.i.d.) and that there are no unmeasured confounding variables. In practice, neither is likely to hold, and detecting confounding in non-i.i.d. settings poses a significant challenge. Motivated by this, we explore how to discover confounders from data in multiple environments with causal mechanism shifts. We show that the mechanism changes of observed variables can reveal which variable sets are confounded. Based on this idea, we propose an empirically testable criterion based on mutual information, show under which conditions it can identify confounding, and introduce CoCo to discover confounders from data in multiple contexts. In our experiments, we show that CoCo works well on synthetic and real-world data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-mameche24a, title = { Identifying Confounding from Causal Mechanism Shifts }, author = {Mameche, Sarah and Vreeken, Jilles and Kaltenpoth, David}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {4897--4905}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/mameche24a/mameche24a.pdf}, url = {https://proceedings.mlr.press/v238/mameche24a.html}, abstract = { Causal discovery methods commonly assume that all data is independently and identically distributed (i.i.d.) and that there are no unmeasured confounding variables. In practice, neither is likely to hold, and detecting confounding in non-i.i.d. settings poses a significant challenge. Motivated by this, we explore how to discover confounders from data in multiple environments with causal mechanism shifts. We show that the mechanism changes of observed variables can reveal which variable sets are confounded. Based on this idea, we propose an empirically testable criterion based on mutual information, show under which conditions it can identify confounding, and introduce CoCo to discover confounders from data in multiple contexts. In our experiments, we show that CoCo works well on synthetic and real-world data. } }
Endnote
%0 Conference Paper %T Identifying Confounding from Causal Mechanism Shifts %A Sarah Mameche %A Jilles Vreeken %A David Kaltenpoth %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-mameche24a %I PMLR %P 4897--4905 %U https://proceedings.mlr.press/v238/mameche24a.html %V 238 %X Causal discovery methods commonly assume that all data is independently and identically distributed (i.i.d.) and that there are no unmeasured confounding variables. In practice, neither is likely to hold, and detecting confounding in non-i.i.d. settings poses a significant challenge. Motivated by this, we explore how to discover confounders from data in multiple environments with causal mechanism shifts. We show that the mechanism changes of observed variables can reveal which variable sets are confounded. Based on this idea, we propose an empirically testable criterion based on mutual information, show under which conditions it can identify confounding, and introduce CoCo to discover confounders from data in multiple contexts. In our experiments, we show that CoCo works well on synthetic and real-world data.
APA
Mameche, S., Vreeken, J. & Kaltenpoth, D.. (2024). Identifying Confounding from Causal Mechanism Shifts . Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:4897-4905 Available from https://proceedings.mlr.press/v238/mameche24a.html.

Related Material