Document Open Access Logo

Probabilistic and Causal Satisfiability: Constraining the Model

Authors Markus Bläser , Julian Dörfler , Maciej Liśkiewicz , Benito van der Zander

PDF HTML 5

Files

Thumbnail PDF
PDF
LIPIcs.ICALP.2025.144.pdf
  • Filesize: 0.98 MB
  • 20 pages
HTML5 Logo HTML (experimental)
LIPIcs.ICALP.2025.144.html

Document Identifiers

Related Versions

Subject Classification

ACM Subject Classification
  • Theory of computation → Complexity theory and logic
Keywords
  • Existential theory of the real numbers
  • Computational complexity
  • Probabilistic logic
  • Structural Causal Models

Metrics

  • Access Statistics
  • Total Accesses (updated on a weekly basis)
    0
    PDF Downloads
    0
    Metadata Views

Abstract

We study the complexity of satisfiability problems in probabilistic and causal reasoning. Given random variables X₁, X₂,… over finite domains, the basic terms are probabilities of propositional formulas over atomic events X_i = x_i, such as ℙ(X₁ = x₁) or ℙ(X₁ = x₁ ∨ X₂ = x₂). The basic terms can be combined using addition (yielding linear terms) or multiplication (polynomial terms). The probabilistic satisfiability problem asks whether a joint probability distribution satisfies a Boolean combination of (in)equalities over such terms. Fagin et al. [Fagin et al., 1990] showed that for basic and linear terms, this problem is NP-complete, making it no harder than Boolean satisfiability, while Mossé et al. [Mossé et al., 2022] proved that for polynomial terms, it is complete for the existential theory of the reals.
Pearl’s Causal Hierarchy (PCH) extends the probabilistic setting with interventional and counterfactual reasoning, enriching the expressiveness of the languages. However, Mossé et al. [Mossé et al., 2022] found that the complexity of satisfiability remains unchanged. Van der Zander et al. [van der Zander et al., 2023] showed that introducing a marginalization operator to languages induces a significant increase in complexity.
We extend this line of work by adding two new dimensions to the problem by constraining the models. First, we fix the graph structure of the underlying structural causal model, motivated by settings like Pearl’s do-calculus, and give a nearly complete landscape across different arithmetics and PCH levels. Second, we study small models. While earlier work showed that satisfiable instances admit polynomial-size models, this is no longer guaranteed with compact marginalization. We characterize the complexities of satisfiability under small-model constraints across different settings.

Cite As Get BibTex

Markus Bläser, Julian Dörfler, Maciej Liśkiewicz, and Benito van der Zander. Probabilistic and Causal Satisfiability: Constraining the Model. In 52nd International Colloquium on Automata, Languages, and Programming (ICALP 2025). Leibniz International Proceedings in Informatics (LIPIcs), Volume 334, pp. 144:1-144:20, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2025) https://doi.org/10.4230/LIPIcs.ICALP.2025.144

Author Details

Markus Bläser
  • Saarland University, Saarbrücken, Germany
Julian Dörfler
  • Saarland University, Saarbrücken, Germany
Maciej Liśkiewicz
  • University of Lübeck, Germany
Benito van der Zander
  • University of Lübeck, Germany

Funding

  • van der Zander, Benito: Work supported by the Deutsche Forschungsgemeinschaft (DFG) grant 471183316 (ZA 1244/1-1).

References

  1. Antonis Achilleos. NEXP-completeness and universal hardness results for justification logic. In International Computer Science Symposium in Russia, pages 27-52. Springer, 2015. URL: https://doi.org/10.1007/978-3-319-20297-6_3.
  2. Sanjeev Arora and Boaz Barak. Computational complexity: a modern approach. Cambridge University Press, 2009. Google Scholar
  3. Elias Bareinboim, Juan D. Correa, Duligur Ibeling, and Thomas Icard. On Pearl’s Hierarchy and the Foundations of Causal Inference, pages 507-556. Association for Computing Machinery, New York, NY, USA, 2022. URL: https://doi.org/10.1145/3501714.3501743.
  4. Markus Bläser, Julian Dörfler, Maciej Liśkiewicz, and Benito van der Zander. The existential theory of the reals with summation operators. In 35th Int. Symposium on Algorithms and Computation, ISAAC 2024, volume 322 of LIPIcs, pages 13:1-13:19. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2024. URL: https://doi.org/10.4230/LIPICS.ISAAC.2024.13.
  5. John Canny. Some algebraic and geometric computations in PSPACE. In Proceedings of the twentieth annual ACM symposium on Theory of computing, pages 460-467. ACM, 1988. URL: https://doi.org/10.1145/62212.62257.
  6. Gregory F. Cooper. The computational complexity of probabilistic inference using Bayesian belief networks. Artificial intelligence, 42(2-3):393-405, 1990. URL: https://doi.org/10.1016/0004-3702(90)90060-D.
  7. Paul Dagum and Michael Luby. Approximating probabilistic inference in Bayesian belief networks is NP-hard. Artificial Intelligence, 60(1):141-153, 1993. URL: https://doi.org/10.1016/0004-3702(93)90036-B.
  8. Julian Dörfler, Benito van der Zander, Markus Bläser, and Maciej Liśkiewicz. From probability to counterfactuals: the increasing complexity of satisfiability in Pearl’s Causal Hierarchy. In International Conference on Learning Representations (ICLR). PMLR, to appear, 2025. Available as ArXiv TR 2405.07373. Google Scholar
  9. Mathias Drton and Marloes H. Maathuis. Structure learning in graphical modeling. Annual Review of Statistics and Its Application, 4:365-393, 2017. Google Scholar
  10. Jeff Erickson, Ivor Van Der Hoog, and Tillmann Miltzow. Smoothing the gap between NP and ER. SIAM Journal on Computing, 53:FOCS20-102-FOCS20-138, 2024. Google Scholar
  11. Ronald Fagin, Joseph Y. Halpern, and Nimrod Megiddo. A logic for reasoning about probabilities. Information and computation, 87(1-2):78-128, 1990. URL: https://doi.org/10.1016/0890-5401(90)90060-U.
  12. Ronald Aylmer Fisher. Design of experiments. British Medical Journal, 1(3923):554, 1936. Google Scholar
  13. Clark Glymour, Kun Zhang, and Peter Spirtes. Review of causal discovery methods based on graphical models. Frontiers in genetics, 10:524, 2019. Google Scholar
  14. Dima Grigoriev and Nicolai Vorobjov. Solving systems of polynomial inequalities in subexponential time. J. Symb. Comput., 5(1/2):37-64, 1988. URL: https://doi.org/10.1016/S0747-7171(88)80005-1.
  15. Duligur Ibeling and Thomas Icard. Probabilistic reasoning across the causal hierarchy. In The 34th AAAI Conference on Artificial Intelligence, AAAI 2020, pages 10170-10177. AAAI Press, 2020. URL: https://doi.org/10.1609/AAAI.V34I06.6577.
  16. Duligur Ibeling, Thomas Icard, Krzysztof Mierzewski, and Milan Mossé. Probing the quantitative–qualitative divide in probabilistic reasoning. Annals of Pure and Applied Logic, 175(9):103339, 2024. URL: https://doi.org/10.1016/J.APAL.2023.103339.
  17. Duligur Ibeling, Thomas Icard, and Milan Mossé. On probabilistic and causal reasoning with summation operators. Journal of Logic and Computation, 2024. Google Scholar
  18. Daphne Koller and Nir Friedman. Probabilistic graphical models: principles and techniques. MIT press, 2009. Google Scholar
  19. Harry R. Lewis. Complexity results for classes of quantificational formulas. Journal of Computer and System Sciences, 21(3):317-353, 1980. URL: https://doi.org/10.1016/0022-0000(80)90027-6.
  20. Ni Y. Lu, Kun Zhang, and Changhe Yuan. Improving causal discovery by optimal Bayesian network learning. In Proc. of the AAAI Conference on Artificial Intelligence, volume 35(10), pages 8741-8748, 2021. URL: https://doi.org/10.1609/AAAI.V35I10.17059.
  21. Nicolai Meinshausen, Alain Hauser, Joris M. Mooij, Jonas Peters, Philip Versteeg, and Peter Bühlmann. Methods for causal inference from gene perturbation experiments and validation. Proceedings of the National Academy of Sciences, 113(27):7361-7368, 2016. Google Scholar
  22. Milan Mossé, Duligur Ibeling, and Thomas Icard. Is causal reasoning harder than probabilistic reasoning? The Review of Symbolic Logic, pages 1-26, 2022. Google Scholar
  23. Ignavier Ng, Yujia Zheng, Jiji Zhang, and Kun Zhang. Reliable causal discovery with improved exact search and weaker assumptions. Advances in Neural Information Processing Systems, NeurIPS, 34:20308-20320, 2021. URL: https://proceedings.neurips.cc/paper/2021/hash/a9b4ec2eb4ab7b1b9c3392bb5388119d-Abstract.html.
  24. Nils J. Nilsson. Probabilistic logic. Artificial Intelligence, 28(1):71-87, 1986. URL: https://doi.org/10.1016/0004-3702(86)90031-7.
  25. James D. Park and Adnan Darwiche. Complexity results and approximation strategies for MAP explanations. Journal of Artificial Intelligence Research, 21:101-133, 2004. URL: https://doi.org/10.1613/JAIR.1236.
  26. Judea Pearl. Probabilistic reasoning in intelligent systems: networks of plausible inference. Morgan Kaufmann, 1988. Google Scholar
  27. Judea Pearl. Causal diagrams for empirical research. Biometrika, 82(4):669-688, 1995. Google Scholar
  28. Judea Pearl. Causality. Cambridge University Press, 2009. Google Scholar
  29. Judea Pearl and Dana Mackenzie. The book of why: the new science of cause and effect. Basic books, 2018. Google Scholar
  30. Alexander Reisach, Christof Seiler, and Sebastian Weichwald. Beware of the simulated dag! causal discovery benchmarks may be easy to game. Advances in Neural Information Processing Systems, NeurIPS, 34:27772-27784, 2021. URL: https://proceedings.neurips.cc/paper/2021/hash/e987eff4a7c7b7e580d659feb6f60c1a-Abstract.html.
  31. Alexander Reisach, Myriam Tami, Christof Seiler, Antoine Chambaz, and Sebastian Weichwald. A scale-invariant sorting criterion to find a causal order in additive noise models. Advances in Neural Information Processing Systems, NeurIPS, 36:785-807, 2023. Google Scholar
  32. James Renegar. On the computational complexity and geometry of the first-order theory of the reals. Part I: Introduction. Preliminaries. The geometry of semi-algebraic sets. The decision problem for the existential theory of the reals. Journal of symbolic computation, 13(3):255-299, 1992. URL: https://doi.org/10.1016/S0747-7171(10)80003-3.
  33. Paul Rolland, Volkan Cevher, Matthäus Kleindessner, Chris Russell, Dominik Janzing, Bernhard Schölkopf, and Francesco Locatello. Score matching enables causal discovery of nonlinear additive noise models. In International Conference on Machine Learning, ICLM, pages 18741-18753. PMLR, 2022. URL: https://proceedings.mlr.press/v162/rolland22a.html.
  34. Dan Roth. On the hardness of approximate reasoning. Artificial Intelligence, 82(1-2):273-302, 1996. URL: https://doi.org/10.1016/0004-3702(94)00092-1.
  35. Marcus Schaefer. Complexity of some geometric and topological problems. In International Symposium on Graph Drawing, pages 334-344. Springer, 2009. URL: https://doi.org/10.1007/978-3-642-11805-0_32.
  36. Ilya Shpitser and Judea Pearl. Complete identification methods for the causal hierarchy. Journal of Machine Learning Research, 9(Sep):1941-1979, 2008. URL: https://doi.org/10.5555/1390681.1442797.
  37. Chandler Squires and Caroline Uhler. Causal structure learning: A combinatorial perspective. Foundations of Computational Mathematics, pages 1-35, 2022. Google Scholar
  38. Benito van der Zander, Markus Bläser, and Maciej Liśkiewicz. The hardness of reasoning about probabilities and causality. In Proc. Joint Conference on Artificial Intelligence (IJCAI 2023), 2023. Google Scholar
  39. Junzhe Zhang, Jin Tian, and Elias Bareinboim. Partial counterfactual identification from observational and experimental data. In International Conference on Machine Learning, pages 26548-26558. PMLR, 2022. URL: https://proceedings.mlr.press/v162/zhang22ab.html.
Any Issues?
X

Feedback on the Current Page

CAPTCHA

Thanks for your feedback!

Feedback submitted to Dagstuhl Publishing

Could not send message

Please try again later or send an E-mail
© 2023-2025 Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact