dcFCI: Robust Causal Discovery Under Latent Confounding, Unfaithfulness, and Mixed Data

2025-05-10

AI Summary

Causal discovery from observational data faces robustness challenges due to latent confounding, unfaithfulness, and mixed variable types (continuous/discrete). Method: We propose dcFCI—the first nonparametric, data-compatible adaptation of the FCI algorithm—featuring: (i) a novel necessary and sufficient nonparametric PAG compatibility score; (ii) an anytime-terminating hybrid causal discovery framework unifying treatment of all three challenges; and (iii) integration of nonparametric dependence testing, PAG representation, and mixed-variable conditional independence assessment. Results: Experiments demonstrate that dcFCI significantly outperforms state-of-the-art methods on small-sample and heterogeneous real-world datasets, achieving high fidelity in recovering the true PAG. Moreover, its output Top-k PAG set quantifies structural uncertainty, enabling robust causal inference and decision-making.

Technology Category

• Reasoning under Uncertainty: Causality
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization
• Machine Learning: Causal Learning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Causal discovery is central to inferring causal relationships from observational data. In the presence of latent confounding, algorithms such as Fast Causal Inference (FCI) learn a Partial Ancestral Graph (PAG) representing the true model's Markov Equivalence Class. However, their correctness critically depends on empirical faithfulness, the assumption that observed (in)dependencies perfectly reflect those of the underlying causal model, which often fails in practice due to limited sample sizes. To address this, we introduce the first nonparametric score to assess a PAG's compatibility with observed data, even with mixed variable types. This score is both necessary and sufficient to characterize structural uncertainty and distinguish between distinct PAGs. We then propose data-compatible FCI (dcFCI), the first hybrid causal discovery algorithm to jointly address latent confounding, empirical unfaithfulness, and mixed data types. dcFCI integrates our score into an (Anytime)FCI-guided search that systematically explores, ranks, and validates candidate PAGs. Experiments on synthetic and real-world scenarios demonstrate that dcFCI significantly outperforms state-of-the-art methods, often recovering the true PAG even in small and heterogeneous datasets. Examining top-ranked PAGs further provides valuable insights into structural uncertainty, supporting more robust and informed causal reasoning and decision-making.

Problem

• Robust causal discovery under latent confounding and unfaithfulness
• Handling mixed data types in causal inference
• Assessing PAG compatibility with observed data nonparametrically

Innovation

• Nonparametric score for PAG compatibility with mixed data
• Hybrid algorithm dcFCI addressing confounding and unfaithfulness
• AnytimeFCI-guided search ranking and validating candidate PAGs