×

Toric ideals of characteristic imsets via quasi-independence gluing. (English) Zbl 07886937

Summary: Characteristic imsets are 0-1 vectors which correspond to Markov equivalence classes of directed acyclic graphs. The study of their convex hull, named the characteristic imset polytope, has led to new and interesting geometric perspectives on the important problem of causal discovery. In this paper, we begin the study of the associated toric ideal. We develop a new generalization of the toric fiber product, which we call a quasi-independence gluing, and show that under certain combinatorial homogeneity conditions, one can iteratively compute a Gröbner basis via lifting. For faces of the characteristic imset polytope associated to trees, we apply this technique to compute a Gröbner basis for the associated toric ideal. We end with a study of the characteristic ideal of the cycle and provide direction for future work.

MSC:

62Rxx Statistics on algebraic and topological structures
62E10 Characterization and structure theory of statistical distributions
62H22 Probabilistic graphical models
62D20 Causal inference from observational studies
62R01 Algebraic statistics
13P25 Applications of commutative algebra (e.g., to statistics, control theory, optimization, etc.)
13P10 Gröbner bases; other bases for ideals and modules (e.g., Janet and border bases)

References:

[1] 10.2140/astat.2020.11.5 · Zbl 1461.62221 · doi:10.2140/astat.2020.11.5
[2] 10.2140/astat.2021.12.1 · Zbl 1464.13023 · doi:10.2140/astat.2021.12.1
[3] 10.1214/aos/1031833662 · Zbl 0876.60095 · doi:10.1214/aos/1031833662
[4] 10.1007/978-1-4614-3719-2 · Zbl 1304.62015 · doi:10.1007/978-1-4614-3719-2
[5] 10.1162/153244303321897717 · Zbl 1084.68519 · doi:10.1162/153244303321897717
[6] 10.1016/j.jsc.2020.10.006 · Zbl 1461.62226 · doi:10.1016/j.jsc.2020.10.006
[7] 10.1016/j.aam.2020.102119 · Zbl 1457.92125 · doi:10.1016/j.aam.2020.102119
[8] 10.1016/j.aam.2023.102633 · Zbl 1531.92064 · doi:10.1016/j.aam.2023.102633
[9] 10.1007/s10801-013-0450-0 · Zbl 1314.13049 · doi:10.1007/s10801-013-0450-0
[10] 10.1007/978-3-319-95349-6_5 · doi:10.1007/978-3-319-95349-6_5
[11] 10.1080/10618600.2021.2020127 · Zbl 07633197 · doi:10.1080/10618600.2021.2020127
[12] 10.1137/21M1457205 · Zbl 07669656 · doi:10.1137/21M1457205
[13] 10.1007/BF01299745 · Zbl 0838.52015 · doi:10.1007/BF01299745
[14] 10.1006/jabr.1999.7918 · Zbl 0943.13014 · doi:10.1006/jabr.1999.7918
[15] ; Pearl, Judea, Causality: models, reasoning, and inference, 2000 · Zbl 0959.68116
[16] 10.1016/j.jalgebra.2020.09.032 · Zbl 1452.14003 · doi:10.1016/j.jalgebra.2020.09.032
[17] 10.1016/j.jsc.2015.07.003 · Zbl 1402.11121 · doi:10.1016/j.jsc.2015.07.003
[18] 10.1093/biomet/asaa104 · Zbl 07459733 · doi:10.1093/biomet/asaa104
[19] 10.1177/089443939100900106 · doi:10.1177/089443939100900106
[20] ; Studený, M.; Hemmecke, R.; Lindner, S., Characteristic imset: a simple algebraic representative of a Bayesian network structure, Proceedings of the fifth European Workshop on Probabilistic Graphical Models, 257, 2010
[21] 10.1016/j.ijar.2021.07.014 · Zbl 1520.68159 · doi:10.1016/j.ijar.2021.07.014
[22] 10.1090/ulect/008 · doi:10.1090/ulect/008
[23] 10.1016/j.jalgebra.2006.10.004 · Zbl 1129.13030 · doi:10.1016/j.jalgebra.2006.10.004
[24] 10.1090/gsm/194 · Zbl 1408.62004 · doi:10.1090/gsm/194
[25] 10.1007/s10994-006-6889-7 · Zbl 1470.68192 · doi:10.1007/s10994-006-6889-7
[26] ; Verma, T.; Pearl, J., Equivalence and synthesis of causal models, Proceedings of the sixth annual Conference on Uncertainty in Artificial Intelligence, 255, 1990
[27] 10.1016/B978-1-4832-8287-9.50049-9 · doi:10.1016/B978-1-4832-8287-9.50049-9
[28] 10.1080/00927879508825412 · Zbl 0836.13014 · doi:10.1080/00927879508825412
[29] 10.1137/130933848 · Zbl 1326.05057 · doi:10.1137/130933848
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.