Interventional Processes for Causal Uncertainty Quantification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dance, Hugh, Orbanz, Peter, Gretton, Arthur
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913178317750272
author Dance, Hugh
Orbanz, Peter
Gretton, Arthur
author_facet Dance, Hugh
Orbanz, Peter
Gretton, Arthur
contents Reliable uncertainty quantification for causal effects is crucial in high-stakes applications, but remains challenging when the target is an entire function rather than a scalar estimand. In this work, we introduce a GP-based approach for uncertainty quantification of interventional functions. The central idea is to build on recent work representing interventional functions as an inner-product of observational functions in a reproducing kernel Hilbert space (RKHS), by constructing appropriate GP priors for such functions and inferring posteriors from observational data. Our approach yields closed-form posterior moments and tractable training and inference, while avoiding pathologies of previous GP prior constructions for RKHS functions. We further derive a practical procedure for posterior coverage calibration. Across synthetic benchmarks, causal Bayesian optimization tasks, and a large-scale real dataset, our method improves uncertainty quantification while remaining competitive in causal effect estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2410_14483
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Interventional Processes for Causal Uncertainty Quantification
Dance, Hugh
Orbanz, Peter
Gretton, Arthur
Machine Learning
Methodology
Reliable uncertainty quantification for causal effects is crucial in high-stakes applications, but remains challenging when the target is an entire function rather than a scalar estimand. In this work, we introduce a GP-based approach for uncertainty quantification of interventional functions. The central idea is to build on recent work representing interventional functions as an inner-product of observational functions in a reproducing kernel Hilbert space (RKHS), by constructing appropriate GP priors for such functions and inferring posteriors from observational data. Our approach yields closed-form posterior moments and tractable training and inference, while avoiding pathologies of previous GP prior constructions for RKHS functions. We further derive a practical procedure for posterior coverage calibration. Across synthetic benchmarks, causal Bayesian optimization tasks, and a large-scale real dataset, our method improves uncertainty quantification while remaining competitive in causal effect estimation.
title Interventional Processes for Causal Uncertainty Quantification
topic Machine Learning
Methodology
url https://arxiv.org/abs/2410.14483