Neural Additive Vector Autoregression Models for Causal Discovery in Time Series
This addresses the challenge of learning causal structures from observational time series data in scientific domains where linear assumptions fail, though it appears incremental as it builds on existing neural and additive approaches.
The paper tackles the problem of causal structure discovery in time series with nonlinear relationships, proposing Neural Additive Vector Autoregression (NAVAR) models that achieve state-of-the-art results on benchmark datasets.
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.