Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions

Amir Asiaee, Kaveh Aryan, James P. Long
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:161-184, 2026.

Abstract

Selective conformal prediction can yield substantially tighter uncertainty sets when we can identify calibration examples that are exchangeable with the test example. In interventional settings, such as perturbation experiments in genomics, exchangeability often holds only within subsets of interventions that leave a target variable "unaffected" (e.g., non-descendants of an intervened node in a causal graph). We study the practical regime where this invariance structure is unknown and must be estimated from data. Our main result quantifies how coverage degrades when the estimated safe calibration set accidentally includes interventions that affect the target, and gives a conservative correction when an upper bound on this error is available. Rather than learning a full causal graph, we learn only the intervention-target relationships needed to choose calibration interventions. We give algorithms for this partial learning task and evaluate them on synthetic structural equation models and {Replogle} {K562} {CRISPR}-interference data, where the experiments illustrate synthetic gains from selective calibration and finite-sample tradeoffs on real perturbation screens.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-asiaee26b, title = {Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions}, author = {Asiaee, Amir and Aryan, Kaveh and Long, James P.}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {161--184}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/asiaee26b/asiaee26b.pdf}, url = {https://proceedings.mlr.press/v337/asiaee26b.html}, abstract = {Selective conformal prediction can yield substantially tighter uncertainty sets when we can identify calibration examples that are exchangeable with the test example. In interventional settings, such as perturbation experiments in genomics, exchangeability often holds only within subsets of interventions that leave a target variable "unaffected" (e.g., non-descendants of an intervened node in a causal graph). We study the practical regime where this invariance structure is unknown and must be estimated from data. Our main result quantifies how coverage degrades when the estimated safe calibration set accidentally includes interventions that affect the target, and gives a conservative correction when an upper bound on this error is available. Rather than learning a full causal graph, we learn only the intervention-target relationships needed to choose calibration interventions. We give algorithms for this partial learning task and evaluate them on synthetic structural equation models and {Replogle} {K562} {CRISPR}-interference data, where the experiments illustrate synthetic gains from selective calibration and finite-sample tradeoffs on real perturbation screens.} }
Endnote
%0 Conference Paper %T Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions %A Amir Asiaee %A Kaveh Aryan %A James P. Long %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-asiaee26b %I PMLR %P 161--184 %U https://proceedings.mlr.press/v337/asiaee26b.html %V 337 %X Selective conformal prediction can yield substantially tighter uncertainty sets when we can identify calibration examples that are exchangeable with the test example. In interventional settings, such as perturbation experiments in genomics, exchangeability often holds only within subsets of interventions that leave a target variable "unaffected" (e.g., non-descendants of an intervened node in a causal graph). We study the practical regime where this invariance structure is unknown and must be estimated from data. Our main result quantifies how coverage degrades when the estimated safe calibration set accidentally includes interventions that affect the target, and gives a conservative correction when an upper bound on this error is available. Rather than learning a full causal graph, we learn only the intervention-target relationships needed to choose calibration interventions. We give algorithms for this partial learning task and evaluate them on synthetic structural equation models and {Replogle} {K562} {CRISPR}-interference data, where the experiments illustrate synthetic gains from selective calibration and finite-sample tradeoffs on real perturbation screens.
APA
Asiaee, A., Aryan, K. & Long, J.P.. (2026). Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:161-184 Available from https://proceedings.mlr.press/v337/asiaee26b.html.

Related Material