CausalGuard: Conformal Inference under Graph Uncertainty
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914585584336896 |
|---|---|
| author | Singh, Vikash Chen, Weicong Ganguly, Debargha Zhang, Yanyan Wang, Nengbo Sankar, Sreehari Hariri, Mohsen Nemecek, Alexander Song, Chaoda Wang, Shouren Zhang, Biyao Yang, Van Ayday, Erman Ma, Jing Chaudhary, Vipin |
| author_facet | Singh, Vikash Chen, Weicong Ganguly, Debargha Zhang, Yanyan Wang, Nengbo Sankar, Sreehari Hariri, Mohsen Nemecek, Alexander Song, Chaoda Wang, Shouren Zhang, Biyao Yang, Van Ayday, Erman Ma, Jing Chaudhary, Vipin |
| contents | Estimating treatment effects from observational data requires choosing an adjustment set, but valid adjustment depends on an unknown causal graph. Graph misspecification can cause under-coverage, while graph-agnostic conformal wrappers may regain nominal coverage only through large padding. We introduce CausalGuard, a structure-weighted conformal framework that calibrates after aggregating graph-conditional doubly robust pseudo-outcomes. Candidate DAGs are proposed from an LLM-derived edge prior, pruned by conditional-independence tests, and reweighted by Bayesian Information Criterion. A composite nonconformity score then calibrates the posterior-weighted pseudo-outcome. CausalGuard provides distribution-free finite-sample marginal coverage for this aggregated pseudo-outcome; under causal identification, overlap, conditional-mean nuisance stability, and concentration on target-aligned valid adjustment strategies, its conditional mean converges to the true Conditional Average Treatment Effect. Across five benchmarks, CausalGuard attains mean coverage above the nominal 90% level for the directly evaluable target and reduces width when graph-agnostic conformal baselines require large padding. Stress tests show that CausalGuard suppresses invalid collider adjustment and remains stable under misspecified priors when the retained candidate set is data-supported. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_21928 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CausalGuard: Conformal Inference under Graph Uncertainty Singh, Vikash Chen, Weicong Ganguly, Debargha Zhang, Yanyan Wang, Nengbo Sankar, Sreehari Hariri, Mohsen Nemecek, Alexander Song, Chaoda Wang, Shouren Zhang, Biyao Yang, Van Ayday, Erman Ma, Jing Chaudhary, Vipin Machine Learning Artificial Intelligence Methodology Estimating treatment effects from observational data requires choosing an adjustment set, but valid adjustment depends on an unknown causal graph. Graph misspecification can cause under-coverage, while graph-agnostic conformal wrappers may regain nominal coverage only through large padding. We introduce CausalGuard, a structure-weighted conformal framework that calibrates after aggregating graph-conditional doubly robust pseudo-outcomes. Candidate DAGs are proposed from an LLM-derived edge prior, pruned by conditional-independence tests, and reweighted by Bayesian Information Criterion. A composite nonconformity score then calibrates the posterior-weighted pseudo-outcome. CausalGuard provides distribution-free finite-sample marginal coverage for this aggregated pseudo-outcome; under causal identification, overlap, conditional-mean nuisance stability, and concentration on target-aligned valid adjustment strategies, its conditional mean converges to the true Conditional Average Treatment Effect. Across five benchmarks, CausalGuard attains mean coverage above the nominal 90% level for the directly evaluable target and reduces width when graph-agnostic conformal baselines require large padding. Stress tests show that CausalGuard suppresses invalid collider adjustment and remains stable under misspecified priors when the retained candidate set is data-supported. |
| title | CausalGuard: Conformal Inference under Graph Uncertainty |
| topic | Machine Learning Artificial Intelligence Methodology |
| url | https://arxiv.org/abs/2605.21928 |