A General Bayesian Nonparametric Approach for Estimating Population-Level and Conditional Causal Effects

2025-11-28

AI Summary

This paper addresses causal inference under concurrent confounding bias and heterogeneous treatment effects (HTE) in observational data. We propose a Bayesian nonparametric method that constructs a nonparametric mixture model whose dependency structure is governed by confounders, without imposing parametric assumptions on the conditional distribution of potential outcomes. Leveraging data augmentation, the method enables efficient full-posterior inference, simultaneously estimating both the average treatment effect (ATE) and the conditional average treatment effect (CATE), while naturally quantifying uncertainty. Our key contribution lies in disentangling the dual role of confounders—governing both treatment assignment and outcome generation—thereby circumventing model misspecification risks inherent in conventional approaches. Extensive evaluations across multiple simulation settings and real-world datasets demonstrate that our method matches or outperforms state-of-the-art methods—including BART—in HTE estimation accuracy, confidence interval calibration, and robustness to model violations.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We propose a Bayesian nonparametric (BNP) approach to causal inference using observational data consisting of outcome, treatment, and a set of confounders. The conditional distribution of the outcome given treatment and confounders is modeled flexibly using a dependent nonparametric mixture model, in which both the atoms and the weights vary with the confounders. The proposed BNP model is well suited for causal inference problems, as it does not rely on parametric assumptions about how the conditional distribution depends on the confounders. In particular, the model effectively adjusts for confounding and improves the modeling of treatment effect heterogeneity, leading to more accurate estimation of both the average treatment effect (ATE) and heterogeneous treatment effects (HTE). Posterior inference under the proposed model is computationally efficient due to the use of data augmentation. Extensive evaluations demonstrate that the proposed model offers competitive or superior performance compared to a wide range of recent methods spanning various statistical approaches, including Bayesian additive regression tree (BART) models, which are well known for their strong empirical performance. More importantly, the model provides fully probabilistic inference on quantities of interest that other methods cannot easily provide, using their posterior distributions.

Problem

• Estimates causal effects from observational data flexibly
• Adjusts for confounding and models treatment effect heterogeneity
• Provides fully probabilistic inference on causal quantities

Innovation

• Bayesian nonparametric mixture model for causal inference
• Flexible conditional distribution modeling without parametric assumptions
• Efficient posterior inference via data augmentation technique