Abstract: Relational models for contingency tables are generalizations of log-linear models, allowing effects associated with arbitrary subsets of cells in a possibly incomplete table, and not necessarily containing the overall effect. In this generality, the MLEs under Poisson and multinomial sampling are not always identical. This paper deals with the theory of maximum likelihood estimation in the case when there are observed zeros in the data. A unique MLE to such data is shown to always exist in the set of pointwise limits of sequences of distributions in the original model. This set is equal to the closure of the original model with respect to the Bregman information divergence. The same variant of iterative scaling may be used to compute the MLE in the original model and in its closure.
Recommendations
- Closure properties for relational systems with given endomorphism structure
- Closure, properties and closure properties of multirelations
- scientific article; zbMATH DE number 168430
- scientific article; zbMATH DE number 4076030
- scientific article; zbMATH DE number 175643
- scientific article; zbMATH DE number 4110691
- Relational, closure and partition powerset theories
- The poset of closures as a model of changing databases
- Mappings of closure relations
- A GALOIS CORRESPONDENCE BETWEEN CLOSURE SPACES AND RELATIONAL SYSTEMS
Cites work
- scientific article; zbMATH DE number 3157499 (Why is no real title available?)
- scientific article; zbMATH DE number 3481687 (Why is no real title available?)
- scientific article; zbMATH DE number 3509596 (Why is no real title available?)
- scientific article; zbMATH DE number 3519741 (Why is no real title available?)
- scientific article; zbMATH DE number 3602484 (Why is no real title available?)
- scientific article; zbMATH DE number 1134987 (Why is no real title available?)
- scientific article; zbMATH DE number 835749 (Why is no real title available?)
- scientific article; zbMATH DE number 3296905 (Why is no real title available?)
- Convex Polytopes
- I-divergence geometry of probability distributions and minimization problems
- Information projections revisited
- Iterative scaling in curved exponential families
- Maximum likelihood estimation in log-linear models
- On the toric algebra of graphical models
- Relational models for contingency tables
- Support sets in exponential families and oriented matroid theory
- Two algorithms for fitting constrained marginal models
Cited in
(9)- Rényi 100, quantitative and qualitative (in)dependence
- scientific article; zbMATH DE number 175643 (Why is no real title available?)
- Estimation and testing of multiplicative models for frequency data
- On the role of the overall effect in exponential families
- Hierarchical Aitchison-Silvey models for incomplete binary sample spaces
- On the closure properties of the class of full G-models of a deductive system
- Relational models for contingency tables
- Testing the fit of relational models
- scientific article; zbMATH DE number 3980540 (Why is no real title available?)
This page was built for publication: On the closure of relational models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q900826)
