Assumption violations in causal discovery and the robustness of score matching

Montagna F, Mastakouri AA, Eulig E, Noceti N, Rosasco L, Janzing D, Aragam B, Locatello F. Assumption violations in causal discovery and the robustness of score matching. arXiv, 2310.13387.

Download (ext.)

Preprint | Submitted | English
Author
Montagna, Francesco; Mastakouri, Atalanti A.; Eulig, Elias; Noceti, Nicoletta; Rosasco, Lorenzo; Janzing, Dominik; Aragam, Bryon; Locatello, FrancescoISTA
Department
Abstract
When domain knowledge is limited and experimentation is restricted by ethical, financial, or time constraints, practitioners turn to observational causal discovery methods to recover the causal structure, exploiting the statistical properties of their data. Because causal discovery without further assumptions is an ill-posed problem, each algorithm comes with its own set of usually untestable assumptions, some of which are hard to meet in real datasets. Motivated by these considerations, this paper extensively benchmarks the empirical performance of recent causal discovery methods on observational i.i.d. data generated under different background conditions, allowing for violations of the critical assumptions required by each selected approach. Our experimental findings show that score matching-based methods demonstrate surprising performance in the false positive and false negative rate of the inferred graph in these challenging scenarios, and we provide theoretical insights into their performance. This work is also the first effort to benchmark the stability of causal discovery algorithms with respect to the values of their hyperparameters. Finally, we hope this paper will set a new standard for the evaluation of causal discovery methods and can serve as an accessible entry point for practitioners interested in the field, highlighting the empirical implications of different algorithm choices.
Publishing Year
Date Published
2023-10-20
Journal Title
arXiv
Acknowledgement
We thank Kun Zhang and Carl-Johann Simon-Gabriel for the insightful discussions. This work has been supported by AFOSR, grant n. FA8655-20-1-7035. FM is supported by Programma Operativo Nazionale ricerca e innovazione 2014-2020. FM partially contributed to this work during an internship at Amazon Web Services with FL. FL partially contributed while at AWS.
Article Number
2310.13387
IST-REx-ID

Cite this

Montagna F, Mastakouri AA, Eulig E, et al. Assumption violations in causal discovery and the robustness of score matching. arXiv. doi:10.48550/arXiv.2310.13387
Montagna, F., Mastakouri, A. A., Eulig, E., Noceti, N., Rosasco, L., Janzing, D., … Locatello, F. (n.d.). Assumption violations in causal discovery and the robustness of score matching. arXiv. https://doi.org/10.48550/arXiv.2310.13387
Montagna, Francesco, Atalanti A. Mastakouri, Elias Eulig, Nicoletta Noceti, Lorenzo Rosasco, Dominik Janzing, Bryon Aragam, and Francesco Locatello. “Assumption Violations in Causal Discovery and the Robustness of Score Matching.” ArXiv, n.d. https://doi.org/10.48550/arXiv.2310.13387.
F. Montagna et al., “Assumption violations in causal discovery and the robustness of score matching,” arXiv. .
Montagna F, Mastakouri AA, Eulig E, Noceti N, Rosasco L, Janzing D, Aragam B, Locatello F. Assumption violations in causal discovery and the robustness of score matching. arXiv, 2310.13387.
Montagna, Francesco, et al. “Assumption Violations in Causal Discovery and the Robustness of Score Matching.” ArXiv, 2310.13387, doi:10.48550/arXiv.2310.13387.
All files available under the following license(s):
Copyright Statement:
This Item is protected by copyright and/or related rights. [...]

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data ISTA Research Explorer

Sources

arXiv 2310.13387

Search this title in

Google Scholar