Polyhedral aspects of score equivalence in Bayesian network structure learning
From MaRDI portal
Publication:2364495
Abstract: This paper deals with faces and facets of the family-variable polytope and the characteristic-imset polytope, which are special polytopes used in integer linear programming approaches to statistically learn Bayesian network structure. A common form of linear objectives to be maximized in this area leads to the concept of score equivalence (SE), both for linear objectives and for faces of the family-variable polytope. We characterize the linear space of SE objectives and establish a one-to-one correspondence between SE faces of the family-variable polytope, the faces of the characteristic-imset polytope, and standardized supermodular functions. The characterization of SE facets in terms of extremality of the corresponding supermodular function gives an elegant method to verify whether an inequality is SE-facet-defining for the family-variable polytope. We also show that when maximizing an SE objective one can eliminate linear constraints of the family-variable polytope that correspond to non-SE facets. However, we show that solely considering SE facets is not enough as a counter-example shows; one has to consider the linear inequality constraints that correspond to facets of the characteristic-imset polytope despite the fact that they may not define facets in the family-variable mode.
Recommendations
- Polyhedral approaches to learning Bayesian networks
- On polyhedral approximations of polytopes for learning Bayesian networks
- Learning Bayesian network structure: towards the essential graph by integer linear programming tools
- On open questions in the geometric approach to structural learning Bayesian nets
- Bayesian network structure learning with integer programming: polytopes, facets and complexity
Cites work
- scientific article; zbMATH DE number 420868 (Why is no real title available?)
- scientific article; zbMATH DE number 48812 (Why is no real title available?)
- scientific article; zbMATH DE number 1312984 (Why is no real title available?)
- scientific article; zbMATH DE number 1134987 (Why is no real title available?)
- scientific article; zbMATH DE number 2150792 (Why is no real title available?)
- scientific article; zbMATH DE number 3804333 (Why is no real title available?)
- scientific article; zbMATH DE number 1860211 (Why is no real title available?)
- A geometric view on learning Bayesian network structures
- Bayesian network structure learning with integer programming: polytopes, facets and complexity
- Characteristic imsets for learning Bayesian network structure
- Core-based criterion for extreme supermodular functions
- Efficient structure learning of Bayesian networks using constraints
- Learning Bayesian network structure: towards the essential graph by integer linear programming tools
- Lectures on Polytopes
- On open questions in the geometric approach to structural learning Bayesian nets
- On polyhedral approximations of polytopes for learning Bayesian networks
Cited in
(9)- Generalized Permutohedra from Probabilistic Graphical Models
- Core-based criterion for extreme supermodular functions
- Bayesian network structure learning with integer programming: polytopes, facets and complexity
- On open questions in the geometric approach to structural learning Bayesian nets
- Polyhedral approach to statistical learning graphical models
- The last dozen of years of or research in Czechia and Slovakia
- The dual polyhedron to the chordal graph polytope and the rebuttal of the chordal graph conjecture
- Towards using the chordal graph polytope in learning decomposable models
- Polyhedral approaches to learning Bayesian networks
This page was built for publication: Polyhedral aspects of score equivalence in Bayesian network structure learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2364495)
