On extended admissible procedures and their nonstandard Bayes risk

DOI10.1214/20-AOS2026zbMATH Open1486.62014arXiv1612.09305OpenAlexW3203906057MaRDI QIDQ2054475FDOQ2054475

Authors: Haosui Duanmu, Daniel M. Roy 

Publication date: 3 December 2021

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: For finite parameter spaces under finite loss, every Bayes procedure derived from a prior with full support is admissible, and every admissible procedure is Bayes. This relationship already breaks down once we move to finite-dimensional Euclidean parameter spaces. Compactness and strong regularity conditions suffice to repair the relationship, but without these conditions, admissible procedures need not be Bayes. Under strong regularity conditions, admissible procedures can be shown to be the limits of Bayes procedures. Under even stricter conditions, they are generalized Bayes, i.e., they minimize the Bayes risk with respect to an improper prior. In both these cases, one must venture beyond the strict confines of Bayesian analysis. Using methods from mathematical logic and nonstandard analysis, we introduce the class of nonstandard Bayes decision procedures---namely, those whose Bayes risk with respect to some prior is within an infinitesimal of the optimal Bayes risk. Among procedures with finite risk functions, we show that a decision procedure is extended admissible if and only if its nonstandard extension is nonstandard Bayes. For problems with continuous risk functions defined on metric parameter spaces, we derive a nonstandard analogue of Blyth's method that can be used to establish the admissibility of a procedure. We also apply the nonstandard theory to derive a purely standard theorem: when risk functions are continuous on a compact Hausdorff parameter space, a procedure is extended admissible if and only if it is Bayes.

Full work available at URL: https://arxiv.org/abs/1612.09305

Recommendations

• Note on a Fundamental Relationship Between Admissible and Bayesian Decision Rules
• scientific article; zbMATH DE number 4056771
• scientific article; zbMATH DE number 3923848
• scientific article; zbMATH DE number 53492
• Admissibility: Survey of a Concept in Progress

zbMATH Keywords

decision theorynonstandard analysiscomplete class theorems

Mathematics Subject Classification ID

Bayesian problems; characterization of Bayes procedures (62C10) Nonstandard measure theory (28E05) Complete class results in statistical decision theory (62C07) Admissibility in statistical decision theory (62C15)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Statistical decision theory and Bayesian analysis. 2nd ed
• Estimation with quadratic loss.
• Invariance, Minimax Sequential Estimation, and Continuous Time Processes
• Title not available (Why is that?)
• An Extension of Wald's Theory of Statistical Decision Functions
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Lexicographic probability, conditional probability, and nonstandard probability
• Admissible Estimators, Recurrent Diffusions, and Insoluble Boundary Value Problems
• Generalized Bayes estimators in multivariate problems
• Generalized Bayes Solutions in Estimation Problems
• Title not available (Why is that?)
• On finitely additive priors, coherence, and extended admissibility
• Star-Finite Representations of Measure Spaces
• The admissibility of the empirical distribution function
• Contributions to the Theory of Statistical Estimation and Testing Hypotheses
• Statistical Decision Functions
• Finitely-additive, countably-additive and internal probability measures.
• Foundations of a General Theory of Sequential Decision Functions
• A Necessary and Sufficient Condition for Admissibility
• A stepwise Bayesian procedure
• An Essentially Complete Class of Admissible Decision Functions
• Title not available (Why is that?)
• Title not available (Why is that?)
• Generalized Bayes Decision Functions, Admissibility and the Exponential Family

Cited In (7)

• Loeb extension and Loeb equivalence II
• Walrasian equilibrium theory with and without free-disposal: theorems and counterexamples in an infinite-agent context
• A Bayesian nonatomic game and its applicability to finite-player situations
• Integration with filters
• Mixing and hitting times for Gibbs samplers and other non-Feller processes
• On abstract economies with an arbitrary set of players and action sets in locally-convex topological vector spaces
• On existence of Berk-Nash equilibria in misspecified Markov decision processes with infinite spaces

This page was built for publication: On extended admissible procedures and their nonstandard Bayes risk

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2054475)