Conditional Counterfactual Mean Embeddings: Doubly Robust Estimation and Learning Rates

2026-02-04

AI Summary

This work addresses the limitation of existing methods that primarily focus on conditional average treatment effects and thus fail to capture the full distributional characteristics of potential outcomes. To overcome this, the paper proposes the Conditional Counterfactual Mean Embedding (CCME) framework, which, for the first time, enables full distributional modeling of conditional counterfactual distributions within a reproducing kernel Hilbert space (RKHS). The approach employs a two-stage meta-estimator compatible with any RKHS-valued regression technique—such as ridge regression, deep feature maps, or neural kernels—and enjoys both double robustness and finite-sample convergence guarantees. Empirical results demonstrate that CCME accurately recovers complex structures of conditional counterfactual distributions, including multimodality, offering significant theoretical and practical advantages.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Multiagent Systems: Mechanism Design

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

A complete understanding of heterogeneous treatment effects involves characterizing the full conditional distribution of potential outcomes. To this end, we propose the Conditional Counterfactual Mean Embeddings (CCME), a framework that embeds conditional distributions of counterfactual outcomes into a reproducing kernel Hilbert space (RKHS). Under this framework, we develop a two-stage meta-estimator for CCME that accommodates any RKHS-valued regression in each stage. Based on this meta-estimator, we develop three practical CCME estimators: (1) Ridge Regression estimator, (2) Deep Feature estimator that parameterizes the feature map by a neural network, and (3) Neural-Kernel estimator that performs RKHS-valued regression, with the coefficients parameterized by a neural network. We provide finite-sample convergence rates for all estimators, establishing that they possess the double robustness property. Our experiments demonstrate that our estimators accurately recover distributional features including multimodal structure of conditional counterfactual distributions.

Problem

• heterogeneous treatment effects
• counterfactual distribution
• conditional distribution
• distributional features
• potential outcomes

Innovation

• Conditional Counterfactual Mean Embeddings
• Reproducing Kernel Hilbert Space
• Double Robustness
• Neural-Kernel Estimator
• Heterogeneous Treatment Effects