Research map / Probabilistic and scientific modeling / Causal discovery
A Proxy Variable View of Shared Confounding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use some treatments as proxies for the shared unobserved confounder (instead of external proxies) to identify intervention distributions of other treatments | multiple treatments with shared unobserved confounding · theoretical identification (Theorem 1) | identifiability of P(y | do(aC)) | Helped |
| Estimate intervention distributions with the deconfounder (substitute confounder + outcome model) instead of solving the integral equation | deconfounder algorithm under shared confounding · theoretical correctness (Theorem 2) | correctness of estimated P(y | do(aC)) | Helped |
| Violate completeness condition (Assumption 1.2) by using a uniformly distributed confounder (ablation) | deconfounder ATE estimation in linear simulation · simulation with three treatments, one-dimensional U | RMSE of ATE estimate | Hurt |
| Satisfy completeness condition (Assumption 1.2) with normally distributed confounder (ablation) | deconfounder ATE estimation in linear simulation · simulation with three treatments, one-dimensional U | RMSE of ATE estimate | Helped |
| Satisfy identification conditions (Assumption 4.2) under selection bias with normal confounder (ablation) | deconfounder ATE estimation with selection bias · simulation with selection bias | ATE estimation quality | Helped |
| Violate Assumption 4.2 under selection bias (ablation) | deconfounder ATE estimation with selection bias · simulation with selection bias | ATE estimation quality | Hurt |
Closest papers by meaning
- Automating the Selection of Proxy Variables of Unmeasured Confounders (ICML 2024 · similarity 0.93)
- Proximal Causal Inference With Text Data (NeurIPS 2024 · similarity 0.91)
- Tight Partial Identification of Causal Effects with Marginal Distribution of Unmeasured Confounders (ICML 2024 · similarity 0.89)
- Identifying Causal Effects Under Functional Dependencies (NeurIPS 2024 · similarity 0.89)
- Fast Proxy Experiment Design for Causal Effect Identification (NeurIPS 2024 · similarity 0.89)
- Detecting and Measuring Confounding Using Causal Mechanism Shifts (NeurIPS 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.