Decomposing conditional independence ideals with hidden variables
Yulia Alexandr, Kristen Dawson, Hannah Friedman, Fatemeh Mohammadi, Pardis Semnani, and Teresa Yu

TL;DR
This paper analyzes determinantal ideals related to conditional independence models with hidden variables, providing explicit decompositions, conditions for primality, and combinatorial classifications.
Contribution
It introduces explicit ideal decompositions, identifies when components are prime, and derives formulas for their dimensions and classification.
Findings
Explicit decompositions of determinantal ideals for hidden variable models
Conditions under which components are prime ideals
Formulas for the dimensions and classification of components
Abstract
We study a family of determinantal ideals whose decompositions encode the structural zeros in conditional independence models with hidden variables. We provide explicit decompositions of these ideals and, for certain subclasses of models, we show that this is a decomposition into radical ideals by displaying Gr\"obner bases for the components. We identify conditions under which the components are prime, and establish formulas for the dimensions of these prime ideals. Moreover, we show that the components in the decomposition can be grouped into equivalence classes defined by their combinatorial structure, and we derive a closed formula for the number of such classes.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRough Sets and Fuzzy Logic · Bayesian Modeling and Causal Inference · Formal Methods in Verification
