Understanding cause–effect relations is central in the natural and social sciences and a prerequisite for designing interventions and policies. The project develops novel methods for discovering cause–effect relations from non-experimental data by taking the so-called full Bayesian approach to statistical inference. We design algorithms that are able to carry out the required computations even in complex settings where previous algorithms would fail. We also study whether and how the direction of causation could be associated with the computational complexity of the corresponding causal mechanism. The results of the project advance efficient and reliable automated discovery of cause–effect relations.

2023

2026