Prediction-powered Generalization of Causal Inferences
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910472762032128 |
|---|---|
| author | Demirel, Ilker Alaa, Ahmed Philippakis, Anthony Sontag, David |
| author_facet | Demirel, Ilker Alaa, Ahmed Philippakis, Anthony Sontag, David |
| contents | Causal inferences from a randomized controlled trial (RCT) may not pertain to a target population where some effect modifiers have a different distribution. Prior work studies generalizing the results of a trial to a target population with no outcome but covariate data available. We show how the limited size of trials makes generalization a statistically infeasible task, as it requires estimating complex nuisance functions. We develop generalization algorithms that supplement the trial data with a prediction model learned from an additional observational study (OS), without making any assumptions on the OS. We theoretically and empirically show that our methods facilitate better generalization when the OS is high-quality, and remain robust when it is not, and e.g., have unmeasured confounding. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_02873 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Prediction-powered Generalization of Causal Inferences Demirel, Ilker Alaa, Ahmed Philippakis, Anthony Sontag, David Machine Learning Causal inferences from a randomized controlled trial (RCT) may not pertain to a target population where some effect modifiers have a different distribution. Prior work studies generalizing the results of a trial to a target population with no outcome but covariate data available. We show how the limited size of trials makes generalization a statistically infeasible task, as it requires estimating complex nuisance functions. We develop generalization algorithms that supplement the trial data with a prediction model learned from an additional observational study (OS), without making any assumptions on the OS. We theoretically and empirically show that our methods facilitate better generalization when the OS is high-quality, and remain robust when it is not, and e.g., have unmeasured confounding. |
| title | Prediction-powered Generalization of Causal Inferences |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2406.02873 |