×

Decomposable context-specific models. (English) Zbl 07871665

Summary: We introduce a family of discrete context-specific models, which we call decomposable. We construct this family from the subclass of staged tree models known as CStree models. We give an algebraic and combinatorial characterization of all context-specific independence relations that hold in a decomposable context-specific model, which yields a Markov basis. We prove that a directed version of the moralization operation applied to the graphical representation of a context-specific model does not affect the implied independence relations, thus affirming that these models are algebraically described by a finite collection of decomposable graphical models. More generally, we establish that several algebraic, combinatorial, and geometric properties of decomposable context-specific models generalize those of decomposable graphical models to the context-specific setting.

MSC:

62R01 Algebraic statistics
62A09 Graphical methods in statistics
13P10 Gröbner bases; other bases for ideals and modules (e.g., Janet and border bases)
13P25 Applications of commutative algebra (e.g., to statistics, control theory, optimization, etc.)

References:

[1] Ananiadi, L. and Duarte, E., Gröbner bases for staged trees, Algebr. Stat., 12 (2021), pp. 1-20. · Zbl 1464.13023
[2] Boutilier, C., Friedman, N., Goldszmidt, M., and Koller, D., Context-specific independence in Bayesian networks, in Uncertainty in Artificial Intelligence (Portland, OR, 1996), Morgan Kaufmann, San Francisco, CA, 1996, pp. 115-123.
[3] Carli, F., Leonelli, M., Riccomagno, E., and Varando, G., The R package stagedtrees for structural learning of stratified staged trees, J. Stat. Softw., 102 (2022), pp. 1-30.
[4] Chickering, D. M., Heckerman, D., and Meek, C., A Bayesian approach to learning Bayesian networks with local structure, in Proceedings of the Thirteenth Conference on Uncertainty in Artificial Intelligence, UAI’97, , Morgan Kaufmann, San Francisco, CA, 1997, pp. 80-89.
[5] Collazo, R., Görgen, C., and Smith, J. Q., Chain Event Graphs, , CRC Press, Boca Raton, FL, 2018. · Zbl 1391.05001
[6] Corander, J., Labelled graphical models, Scand. J. Stat., 30 (2003), pp. 493-508. · Zbl 1034.62049
[7] Corander, J., Hyttinen, A., Kontinen, J., Pensar, J., and Väänänen, J., A logical approach to context-specific independence, in Logic, Language, Information, and Computation, Springer, Berlin Heidelberg, 2016, pp. 165-182. · Zbl 1478.03067
[8] Cox, D. A., Little, J., and O’Shea, D., An introduction to computational algebraic geometry and commutative algebra, in Ideals, Varieties, and Algorithms, 4th ed., , Springer, Cham, 2015. · Zbl 1335.13001
[9] Diaconis, P. and Sturmfels, B., Algebraic algorithms for sampling from conditional distributions, Ann. Statist., 26 (1998), pp. 363-397. · Zbl 0952.62088
[10] Duarte, E. and Görgen, C., Equations defining probability tree models, J. Symbolic Comput., 99 (2020), pp. 127-146. · Zbl 1451.13086
[11] Duarte, E., Marigliano, O., and Sturmfels, B., Discrete statistical models with rational maximum likelihood estimator, Bernoulli, 27 (2021), pp. 135-154. · Zbl 1465.62184
[12] Duarte, E. and Solus, L., Representation of Context-Specific Causal Models with Observational and Interventional Data, arXiv:2101.09271, 2022.
[13] Duarte, E. and Solus, L., A new characterization of discrete decomposable models, Proc. Amer. Math. Soc., 151 (2023), pp. 1325-1338. · Zbl 07640189
[14] Garcia, L. D., Stillman, M., and Sturmfels, B., Algebraic geometry of Bayesian networks, J. Symbolic Comput., 39 (2005), pp. 331-355. · Zbl 1126.68102
[15] Geiger, D., Heckerman, D., King, H., and Meek, C., Stratified exponential families: Graphical models and model selection, Ann. Statist., 29 (2001), pp. 505-529. · Zbl 1012.62012
[16] Geiger, D., Meek, C., and Sturmfels, B., On the toric algebra of graphical models, Ann. Statist., 34 (2006), pp. 1463-1492. · Zbl 1104.60007
[17] Görgen, C., Bigatti, A., Riccomagno, E., and Smith, J. Q., Discovery of statistical equivalence classes using computer algebra, Internat. J. Approx. Reason., 95 (2018), pp. 167-184. · Zbl 1451.62166
[18] Görgen, C. and Smith, J. Q., Equivalence classes of staged trees, Bernoulli, 24 (2018), pp. 2676-2692. · Zbl 1419.62124
[19] Janhunen, T., Gebser, M., Rintanen, J., Nyman, H., Pensar, J., and Corander, J., Learning discrete decomposable graphical models via constraint optimization, Stat. Comput., 27 (2017), pp. 115-130. · Zbl 1505.62200
[20] Koller, D. and Friedman, N., Probabilistic graphical models: Principles and techniques, in Adaptive Computation and Machine Learning, MIT Press, Cambridge, MA, 2009. · Zbl 1183.68483
[21] Lauritzen, S. L., Graphical Models, , The Clarendon Press, Oxford University Press, New York, 1996. · Zbl 0907.62001
[22] Leonelli, M. and Varando, G., Highly efficient structural learning of sparse staged trees, in International Conference on Probabilistic Graphical Models, , PMLR, 2022, pp. 193-204.
[23] Leonelli, M. and Varando, G., Context-specific causal discovery for categorical data using staged trees, in International Conference on Artificial Intelligence and Statistics, , PMLR, 2023, pp. 8871-8888.
[24] Maathuis, M., Drton, M., Lauritzen, S., and Wainwright, M., Handbook of Graphical Models, 1st ed., CRC Press, 2018.
[25] Nyman, H., Pensar, J., Koski, T., and Corander, J., Stratified graphical models: Context-specific independence in graphical models, Bayesian Anal., 9 (2014), pp. 883-908. · Zbl 1327.62030
[26] Pearl, J., Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference, Morgan Kaufmann, 1988.
[27] Pensar, J., Nyman, H., Koski, T., and Corander, J., Labeled directed acyclic graphs: A generalization of context-specific independence in directed graphical models, Data Min. Knowl. Disc., 29 (2015), pp. 503-533. · Zbl 1403.68206
[28] Poole, D. and Zhang, N. L., Exploiting contextual independence in probabilistic inference, J. Artificial Intelligence Res., 18 (2003), pp. 263-313. · Zbl 1056.68144
[29] Smith, J. Q. and Anderson, P. E., Conditional independence and chain event graphs, Artif. Intell., 172 (2008), pp. 42-68. · Zbl 1182.68303
[30] Sullivant, S., Toric fiber products, J. Algebra, 316 (2007), pp. 560-577. · Zbl 1129.13030
[31] Sullivant, S., Algebraic Statistics, , AMS, 2018. · Zbl 1408.62004
[32] Tikka, S., Hyttinen, A., and Karvanen, J., Identifying causal effects via context-specific independence relations, in Advances in Neural Information Processing Systems, Vol. 32, Curran Associates, 2019, pp. 2804-2814.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.