Multivariate Causal Effects: a Bayesian Causal Regression Factor Model

2025-04-04

AI Summary

This study quantifies the multivariate causal effects of wildfire smoke on 27 chemical constituents of PM₂.₅ in the United States—moving beyond conventional univariate analyses focused solely on total PM₂.₅ concentration—to uncover inter-component dependence structures and their health implications. Method: We propose the first Bayesian causal inference framework for multivariate potential outcomes, featuring a treatment-specific latent factor model coupled with a probit stick-breaking prior to jointly address missing data imputation, adaptive latent structure learning, and causal effect identification. Contribution/Results: Simulation studies demonstrate accurate estimation of multivariate causal effects and treatment-heterogeneous latent structures. Empirical analysis provides the first systematic characterization of wildfire smoke’s differential causal impacts across PM₂.₅ chemical components and reveals their intrinsic correlation patterns. Our approach establishes a novel paradigm for fine-grained, compositionally informed air pollution health risk assessment.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Computer Vision: Multi-modal Vision

Application Category

• Web Mining and Content Analysis: Bridging structured and unstructured data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

The impact of wildfire smoke on air quality is a growing concern, contributing to air pollution through a complex mixture of chemical species with important implications for public health. While previous studies have primarily focused on its association with total particulate matter (PM2.5), the causal relationship between wildfire smoke and the chemical composition of PM2.5 remains largely unexplored. Exposure to these chemical mixtures plays a critical role in shaping public health, yet capturing their relationships requires advanced statistical methods capable of modeling the complex dependencies among chemical species. To fill this gap, we propose a Bayesian causal regression factor model that estimates the multivariate causal effects of wildfire smoke on the concentration of 27 chemical species in PM2.5 across the United States. Our approach introduces two key innovations: (i) a causal inference framework for multivariate potential outcomes, and (ii) a novel Bayesian factor model that employs a probit stick-breaking process as prior for treatment-specific factor scores. By focusing on factor scores, our method addresses the missing data challenge common in causal inference and enables a flexible, data-driven characterization of the latent factor structure, which is crucial to capture the complex correlation among multivariate outcomes. Through Monte Carlo simulations, we show the model's accuracy in estimating the causal effects in multivariate outcomes and characterizing the treatment-specific latent structure. Finally, we apply our method to US air quality data, estimating the causal effect of wildfire smoke on 27 chemical species in PM2.5, providing a deeper understanding of their interdependencies.

Problem

• Estimates wildfire smoke's causal effects on PM2.5 chemicals.
• Develops Bayesian model for multivariate causal inference.
• Addresses missing data in chemical species analysis.

Innovation

• Bayesian causal regression factor model
• Probit stick-breaking process prior
• Multivariate causal effects estimation