A Neural Framework for Generalized Causal Sensitivity Analysis

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Frauen, Dennis, Imrie, Fergus, Curth, Alicia, Melnychuk, Valentyn, Feuerriegel, Stefan, van der Schaar, Mihaela
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917634585395200
author Frauen, Dennis
Imrie, Fergus
Curth, Alicia
Melnychuk, Valentyn
Feuerriegel, Stefan
van der Schaar, Mihaela
author_facet Frauen, Dennis
Imrie, Fergus
Curth, Alicia
Melnychuk, Valentyn
Feuerriegel, Stefan
van der Schaar, Mihaela
contents Unobserved confounding is common in many applications, making causal inference from observational data challenging. As a remedy, causal sensitivity analysis is an important tool to draw causal conclusions under unobserved confounding with mathematical guarantees. In this paper, we propose NeuralCSA, a neural framework for generalized causal sensitivity analysis. Unlike previous work, our framework is compatible with (i) a large class of sensitivity models, including the marginal sensitivity model, f-sensitivity models, and Rosenbaum's sensitivity model; (ii) different treatment types (i.e., binary and continuous); and (iii) different causal queries, including (conditional) average treatment effects and simultaneous effects on multiple outcomes. The generality of NeuralCSA is achieved by learning a latent distribution shift that corresponds to a treatment intervention using two conditional normalizing flows. We provide theoretical guarantees that NeuralCSA is able to infer valid bounds on the causal query of interest and also demonstrate this empirically using both simulated and real-world data.
format Preprint
id arxiv_https___arxiv_org_abs_2311_16026
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Neural Framework for Generalized Causal Sensitivity Analysis
Frauen, Dennis
Imrie, Fergus
Curth, Alicia
Melnychuk, Valentyn
Feuerriegel, Stefan
van der Schaar, Mihaela
Machine Learning
Unobserved confounding is common in many applications, making causal inference from observational data challenging. As a remedy, causal sensitivity analysis is an important tool to draw causal conclusions under unobserved confounding with mathematical guarantees. In this paper, we propose NeuralCSA, a neural framework for generalized causal sensitivity analysis. Unlike previous work, our framework is compatible with (i) a large class of sensitivity models, including the marginal sensitivity model, f-sensitivity models, and Rosenbaum's sensitivity model; (ii) different treatment types (i.e., binary and continuous); and (iii) different causal queries, including (conditional) average treatment effects and simultaneous effects on multiple outcomes. The generality of NeuralCSA is achieved by learning a latent distribution shift that corresponds to a treatment intervention using two conditional normalizing flows. We provide theoretical guarantees that NeuralCSA is able to infer valid bounds on the causal query of interest and also demonstrate this empirically using both simulated and real-world data.
title A Neural Framework for Generalized Causal Sensitivity Analysis
topic Machine Learning
url https://arxiv.org/abs/2311.16026