Decomposing causality into its synergistic, unique, and redundant components

Abstract

Causality lies at the heart of scientific inquiry, serving as the fundamental basis for understanding interactions among variables in physical systems. Despite its central role, current methods for causal inference face significant challenges due to nonlinear dependencies, stochastic interactions, self-causation, collider effects, and influences from exogenous factors, among others. While existing methods can effectively address some of these challenges, no single approach has successfully integrated all these aspects. Here, we address these challenges with SURD: Synergistic-Unique-Redundant Decomposition of causality. SURD quantifies causality as the increments of redundant, unique, and synergistic information gained about future events from past observations. The formulation is non-intrusive and applicable to both computational and experimental investigations, even when samples are scarce. We benchmark SURD in scenarios that pose significant challenges for causal inference and demonstrate that it offers a more reliable quantification of causality compared to previous methods.

arXiv admin note: text overlap with arXiv:2310.20544

References in corpus (9)

• JIDT: An information-theoretic toolkit for studying the dynamics of complex systems
• Local information transfer as a spatiotemporal filter for complex systems
• The turbulent cascade in five dimensions
• Large scale flow effects, energy transfer, and self-similarity on turbulence
• Bits and Pieces: Understanding Information Decomposition from Part-whole Relationships and Formal Logic
• Convergent Cross-Mapping and Pairwise Asymmetric Inference
• On the representation and learning of monotone triangular transport maps
• Causality analysis of large-scale structures in the flow around a wall-mounted square cylinder
• Explicit Granger causality in kernel Hilbert spaces

Cited by in corpus (8)

• Chaos, coherence and turbulence
• Information-theoretic machine learning for time-varying mode decomposition of separated aerodynamic flows
• Convolutional causal learning for aerodynamic flows
• Interpreting deep learning-based stellar mass estimation via causal analysis and mutual information decomposition
• Causally coherent structures in turbulent dynamical systems
• Entropic Regression DMD (ERDMD) Discovers Informative Sparse and Nonuniformly Time Delayed Models
• Cause-and-effect approach to turbulence forecasting
• Quantifying Influence and Information Transfer in a Modified Vicsek Model with Non-reciprocal Interactions