From Temporal to Contemporaneous Iterative Causal Discovery in the Presence of Latent Confounders
This work addresses causal discovery for time-series data with latent confounders, which is an incremental improvement over existing methods.
The authors tackled the problem of learning causal structures from observational time-series data with latent confounders by developing a constraint-based algorithm that learns long-term temporal relations before short-term ones, with contemporaneous relations last. This approach reduced the number of required statistical tests and led to higher accuracy on synthetic data and more plausible graphs on real-world data compared to state-of-the-art methods.
We present a constraint-based algorithm for learning causal structures from observational time-series data, in the presence of latent confounders. We assume a discrete-time, stationary structural vector autoregressive process, with both temporal and contemporaneous causal relations. One may ask if temporal and contemporaneous relations should be treated differently. The presented algorithm gradually refines a causal graph by learning long-term temporal relations before short-term ones, where contemporaneous relations are learned last. This ordering of causal relations to be learnt leads to a reduction in the required number of statistical tests. We validate this reduction empirically and demonstrate that it leads to higher accuracy for synthetic data and more plausible causal graphs for real-world data compared to state-of-the-art algorithms.