Quantifying information transfer and mediation along causal pathways in complex systems

Abstract

Measures of information transfer have become a popular approach to analyze interactions in complex systems such as the Earth or the human brain from measured time series. Recent work has focused on causal definitions of information transfer excluding effects of common drivers and indirect influences. While the former clearly constitutes a spurious causality, the aim of the present article is to develop measures quantifying different notions of the strength of information transfer along indirect causal paths, based on first reconstructing the multivariate causal network (\emph{Tigramite} approach). Another class of novel measures quantifies to what extent different intermediate processes on causal paths contribute to an interaction mechanism to determine pathways of causal information transfer. A rigorous mathematical framework allows for a clear information-theoretic interpretation that can also be related to the underlying dynamics as proven for certain classes of processes. Generally, however, estimates of information transfer remain hard to interpret for nonlinearly intertwined complex systems. But, if experiments or mathematical models are not available, measuring pathways of information transfer within the causal dependency structure allows at least for an abstraction of the dynamics. The measures are illustrated on a climatological example to disentangle pathways of atmospheric flow over Europe.

20 pages, 6 figures

References in corpus (8)

• Complex networks in climate dynamics - Comparing linear and nonlinear network construction methods
• Kernel method for nonlinear Granger causality
• Exploration of synergistic and redundant information sharing in static and dynamical Gaussian systems
• Normalizing the causality between time series
• Synergy and redundancy in the Granger causal analysis of dynamical networks
• From brain to earth and climate systems: Small-world interaction networks or not?
• Optimal model-free prediction from multivariate time series
• Assessing directionality and strength of coupling through symbolic analysis: an application to epilepsy patients

Cited by in corpus (9)

• Detecting causal associations in large nonlinear time series datasets
• Causality Learning: A New Perspective for Interpretable Machine Learning
• High-recall causal discovery for autocorrelated time series with latent confounders
• A robust estimator of mutual information for deep learning interpretability
• Space-time nature of causality
• Inference of topology and the nature of synapses, and the flow of information in neuronal networks
• A Framework for Causal Discovery in non-intervenable systems
• Direct and Indirect Effects -- An Information Theoretic Perspective
• Bundled Causal History Interaction