mlsanityData to 7 Oct 2026

Neural Additive Vector Autoregression Models for Causal Discovery in Time Series

arXiv 2010.09429 ↗First posted 19 Oct 2020cs.LGcited by 28

Not on the map. Why?

Abstract (arXiv)

Causal structure discovery in complex dynamical systems is an important challenge for many scientific domains. Although data from (interventional) experiments is usually limited, large amounts of observational time series data sets are usually available. Current methods that learn causal structure from time series often assume linear relationships. Hence, they may fail in realistic settings that contain nonlinear relations between the variables. We propose Neural Additive Vector Autoregression (NAVAR) models, a neural approach to causal structure learning that can discover nonlinear relationships. We train deep neural networks that extract the (additive) Granger causal influences from the time evolution in multi-variate time series. The method achieves state-of-the-art results on various benchmark data sets for causal discovery, while providing clear interpretations of the mapped causal relations.

Key facts. Cited by 28 papers in the map (data through 7 Oct 2026). First posted on arXiv 19 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 8; 2024: 5; 2025: 5; 2026: 7.

28 papers.

Sort:

  1. 2023A Survey on Causal Discovery Methods for I.I.D. and Time Series Datacited by 35
  2. 2022Rhino: Deep Causal Temporal Relationship Learning With History-dependent Noisecited by 30
  3. 2023Neural Structure Learning with Stochastic Differential Equationscited by 14
  4. 2022Deep Causal Learning: Representation, Discovery and Inferencecited by 11
  5. 2024Jacobian Regularizer-based Neural Granger Causalitycited by 9
  6. 2022Deep Recurrent Modelling of Granger Causality with Latent Confoundingcited by 8
  7. 2023CDANs: Temporal Causal Discovery from Autocorrelated and Non-Stationary Time Series Datacited by 7
  8. 2023Evaluating Temporal Observation-Based Causal Discovery Techniques Applied to Road Driver Behaviourcited by 5
  9. 2023Double Machine Learning Based Structure Identification from Temporal Datacited by 3
  10. 2024Sortability of Time Series Datacited by 3
  11. 2025DoFlow: Flow-based Generative Models for Interventional and Counterfactual Forecasting on Time Seriescited by 3
  12. 2023Causal Temporal Regime Structure Learningcited by 2
16 more
  1. 2025Flow based approach for Dynamic Temporal Causal models with non-Gaussian or Heteroscedastic Noisescited by 2
  2. 2023Hierarchical Topological Ordering with Conditional Independence Test for Limited Time Seriescited by 1
  3. 2024Bayesian Vector AutoRegression with Factorised Granger-Causal Graphscited by 1
  4. 2024Semiparametric inference for impulse response functions using double/debiased machine learningcited by 1
  5. 2025Generating Hypotheses of Dynamic Causal Graphs in Neuroscience: Leveraging Generative Factor Models of Observed Time Seriescited by 1
  6. 2026GRACE: Gated Refinement for Accurate Causal Edge Discovery in High-Dimensional Time Seriescited by 1
  7. 2023Efficient Interpretable Nonlinear Modeling for Multiple Time Seriescited by 0
  8. 2024LinBridge: A Learnable Framework for Interpreting Nonlinear Neural Encoding Modelscited by 0
  9. 2025SpinSVAR: Estimating Structural Vector Autoregression Assuming Sparse Inputcited by 0
  10. 2025A Gradient-based Causal Discovery Framework with Applications to Complex Industrial Processescited by 0
  11. 2026PRCD-MAP: Learning How Much to Trust Imperfect Priors in Causal Discoverycited by 0
  12. 2026Function-Valued Causal Influence in Nonlinear Time Seriescited by 0
  13. 2026Trajectory-Aware Node Contributions and the Limits of Static Controllabilitycited by 0
  14. 2026When Are Neural Interaction Discoveries Real? Identifiability, Recoverability, and a Pre-Fit Diagnosticcited by 0
  15. 2026LUCID: Learning Under Confounding for Inference and Discovery in Time Seriescited by 0
  16. 2026Forecast-Necessary Causal Discovery for Nonlinear Political Panel Data: Feedback, Functional Form, and the Dynamics of Democratizationcited by 0