Interventional Processes for Causal Uncertainty Quantification

Hugh Dance, Peter Orbanz, Arthur Gretton
10/18/2024
stat.MLcs.LGstat.ME

No official code found. Best implementation: HWDance/SSVGP (3 stars, Apache-2.0, last commit 2026-02-05).

Abstract

Reliable uncertainty quantification for causal effects is crucial in various applications, but remains difficult in nonparametric models, particularly for continuous treatments. We introduce IMPspec, a Gaussian process (GP) framework for modeling uncertainty over interventional causal functions under continuous treatments, which can be represented using reproducing Kernel Hilbert Spaces (RKHSs). By using principled function class expansions and a spectral representation of RKHS features, IMPspec yields tractable training and inference, a spectral algorithm to calibrate posterior credible intervals, and avoids the underfitting and variance collapse pathologies of earlier GP-on-RKHS methods. Across synthetic benchmarks and an application in healthcare, IMPspec delivers state-of-the-art performance in causal uncertainty quantification and downstream causal Bayesian optimization tasks.

View on arXivView PDF

Code Implementations(1)

HWDance/SSVGPCommunity100%
Apache-2.0

Spike and Slab Variational Gaussian Process (SSVGP) algorithm from Dance and Paige (2021) https://arxiv.org/abs/2111.04558

30Python, Jupyter NotebookFeb 20, 20228 months agoApache-2.0

Similar Papers

Instrumental and Proximal Causal Inference with Gaussian Processes

82% match2026

Kernel Model Validation: How To Do It, And Why You Should Care

80% match2025

Efficient Two-Stage Gaussian Process Regression Via Automatic Kernel Search and Subsampling

78% match2024

Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights

78% match2025

An AI-powered Bayesian generative modeling approach for causal inference in observational studies

78% match1 repo802025

Bayesian Causal Inference with Gaussian Process Networks

78% match2024

Cite this paper

@article{dance2024interventional,
  title  = {Interventional Processes for Causal Uncertainty Quantification},
  author = {Hugh Dance and Peter Orbanz and Arthur Gretton},
  year   = {2024},
  eprint = {2410.14483},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2410.14483v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2410.14483.svg)](https://codeofpaper.com/paper/2410.14483)

Discussion