Structural Causal Models

Also known as SCM

Overview

Structural causal models represent how variables are generated by mechanisms, often organized in a graph, so they can answer questions about interventions and counterfactuals rather than correlations alone. Researchers learn these models from observational and experimental data using constraint- and score-based graph discovery, linear and non-Gaussian structural equation models, nonlinear neural or symbolic mechanisms, and invariance across environments; latent confounding, feedback, and time-varying systems extend the basic acyclic setting. The models support treatment-effect estimation, scientific mechanism analysis, fairness and recourse, anomaly diagnosis, and causal world models, with evaluation increasingly using synthetic benchmarks and intervention-based tests. A central challenge is that observational data often leave multiple structures or counterfactuals compatible with the evidence, so progress depends on explicit assumptions, informative interventions, and reliable uncertainty and benchmark evaluation.

Latest papers

151 papers10 papers in the last four weeks, down 33% on the four weeks before. 0.1% of all new papers.

May 29, 2024

Invariant Graph Representations for Continuous-Time Dynamic Graphs Under Distribution Shifts

Continuous-Time Dynamic Graphs (CTDGs) enable fine-grained modeling of evolving relational systems. However, most existing CTDG representation learning methods are tailored to in-distribution settings and exhibit limited robustness under out-of-distribution (OOD) shifts. Although recent causal approaches learn invariant representations via interventions, they are primarily designed for static or discrete-time graphs and become computationally prohibitive for CTDGs due to the combinatorial explosion of structural and temporal variations. To address these challenges, we propose CIR, a framework grounded in a novel structural causal model termed the ICCM. To avoid exhaustive interventions, we leverage the Normalized Weighted Geometric Mean (NWGM) to efficiently approximate interventional predictions. We further instantiate ICCM within a practical deep learning architecture that jointly captures invariant structural and temporal patterns through dedicated subgraph extractors, and maintains an environment memory bank to model distributional shifts across evolving contexts. Extensive experiments demonstrate that CIR consistently outperforms existing methods under diverse OOD scenarios.