Detecting causal associations in large nonlinear time series datasets

Abstract

Identifying causal relationships from observational time series data is a key problem in disciplines such as climate science or neuroscience, where experiments are often not possible. Data-driven causal inference is challenging since datasets are often high-dimensional and nonlinear with limited sample sizes. Here we introduce a novel method that flexibly combines linear or nonlinear conditional independence tests with a causal discovery algorithm that allows to reconstruct causal networks from large-scale time series datasets. We validate the method on a well-established climatic teleconnection connecting the tropical Pacific with extra-tropical temperatures and using large-scale synthetic datasets mimicking the typical properties of real data. The experiments demonstrate that our method outperforms alternative techniques in detection power from small to large-scale datasets and opens up entirely new possibilities to discover causal networks from time series across a range of research fields.

46 pages, 19 figures

References in corpus (14)

• Measuring Information Transfer
• Estimating Mutual Information
• Measuring and testing dependence by correlation of distances
• Consistency of the group Lasso and multiple kernel learning
• Estimating high-dimensional directed acyclic graphs with the PC-algorithm
• A Robust Method for Detecting Interdependences: Application to Intracranially Recorded EEG
• A class of Rényi information estimators for multidimensional densities
• Granger causality for state space models
• Causal Network Inference by Optimal Causation Entropy
• Quantifying causal influences
• Quantifying Causal Coupling Strength: A Lag-specific Measure For Multivariate Time Series Related To Transfer Entropy
• Graphical modelling of multivariate time series
• Quantifying information transfer and mediation along causal pathways in complex systems
• Optimal model-free prediction from multivariate time series

Cited by in corpus (43)

• Will Artificial Intelligence supersede Earth System and Climate Models?
• Autonomous inference of complex network dynamics from incomplete and noisy data
• Decomposing causality into its synergistic, unique, and redundant components
• Large-scale nonlinear Granger causality: A data-driven, multivariate approach to recovering directed networks from short time-series data
• Inferring species interactions using Granger causality and convergent cross mapping
• Inferring network properties from time series using transfer entropy and mutual information: validation of multivariate versus bivariate approaches
• Detecting climate teleconnections with Granger causality
• Visual Causality Analysis of Event Sequence Data
• Time Series Analysis via Network Science: Concepts and Algorithms
• Causally-informed deep learning to improve climate models and projections
• Machine Learning Applications in Cascading Failure Analysis in Power Systems: A Review
• netrd: A library for network reconstruction and graph distances
• Hypergraph reconstruction from dynamics
• Reconstructing regime-dependent causal relationships from observational time series
• Causality indices for bivariate time series data: a comparative review of performance
• Scalable Causal Structure Learning: Scoping Review of Traditional and Deep Learning Algorithms and New Opportunities in Biomedicine
• Measuring topological descriptors of complex networks under uncertainty
• Neural Additive Vector Autoregression Models for Causal Discovery in Time Series
• Predicting Fatigue Crack Growth via Path Slicing and Re-Weighting
• Community detection based on first passage probabilities
• Statistical Perspective on Functional and Causal Neural Connectomics: The Time-Aware PC Algorithm
• A data-driven framework for dimensionality reduction and causal inference in climate fields
• Consistent Causal Inference from Time Series with PC Algorithm and its Time-Aware Extension
• Causal network discovery by iterative conditioning: comparison of algorithms
• Insightful analysis of historical sources at scales beyond human capabilities using unsupervised Machine Learning and XAI
• Non-Asymptotic Guarantees for Reliable Identification of Granger Causality via the LASSO
• A Data-Driven Two-Phase Multi-Split Causal Ensemble Model for Time Series
• Demand Forecasting of Individual Probability Density Functions with Machine Learning
• Causality in extremes of time series
• Recurrence flow measure of nonlinear dependence
• Towards a robust approach to infer causality in molecular systems satisfying detailed balance
• Robust Distance Covariance
• Towards Physically Consistent Deep Learning For Climate Model Parameterizations
• Linear scaling causal discovery from high-dimensional time series by dynamical community detection
• Information Flow Rate for Cross-Correlated Stochastic Processes
• Temporal Latent Variable Structural Causal Model for Causal Discovery under External Interferences
• Detecting extreme event-driven causality
• Style-Aware Gloss Control for Generative Non-Photorealistic Rendering
• Causal pathway from AMOC to Southern Amazon rainforest indicates stabilising interaction between two climate tipping elements
• Identification of pressure points in modern power systems using transfer entropy
• Causal Intervention Sequence Analysis for Fault Tracking in Radio Access Networks
• Interpretable deep convolutional model for nonlinear multivariate time series in complex systems
• Assimilative Causal Inference