The e-posterior
- PMID: 36970821
- DOI: 10.1098/rsta.2022.0146
The e-posterior
Abstract
We develop a representation of a decision maker's uncertainty based on e-variables. Like the Bayesian posterior, this e-posterior allows for making predictions against arbitrary loss functions that may not be specified ex ante. Unlike the Bayesian posterior, it provides risk bounds that have frequentist validity irrespective of prior adequacy: if the e-collection (which plays a role analogous to the Bayesian prior) is chosen badly, the bounds get loose rather than wrong, making e-posterior minimax decision rules safer than Bayesian ones. The resulting quasi-conditional paradigm is illustrated by re-interpreting a previous influential partial Bayes-frequentist unification, Kiefer-Berger-Brown-Wolpert conditional frequentist tests, in terms of e-posteriors. This article is part of the theme issue 'Bayesian inference: challenges, perspectives, and prospects'.
Keywords: Bayes-frequentist debate; Savage–Dickey ratio; conditional inference; decision making; e-values; uncertainty quantification.
Similar articles
-
Prediction-based uncertainty quantification for exchangeable sequences.Philos Trans A Math Phys Eng Sci. 2023 May 15;381(2247):20220142. doi: 10.1098/rsta.2022.0142. Epub 2023 Mar 27. Philos Trans A Math Phys Eng Sci. 2023. PMID: 36970827 Review.
-
Can natural selection encode Bayesian priors?J Theor Biol. 2017 Aug 7;426:57-66. doi: 10.1016/j.jtbi.2017.05.017. Epub 2017 May 20. J Theor Biol. 2017. PMID: 28536034
-
Bayesian cluster analysis.Philos Trans A Math Phys Eng Sci. 2023 May 15;381(2247):20220149. doi: 10.1098/rsta.2022.0149. Epub 2023 Mar 27. Philos Trans A Math Phys Eng Sci. 2023. PMID: 36970819 Free PMC article. Review.
-
Empirical Bayes interval estimates that are conditionally equal to unadjusted confidence intervals or to default prior credibility intervals.Stat Appl Genet Mol Biol. 2012 Feb 21;11(3):Article 7. doi: 10.1515/1544-6115.1765. Stat Appl Genet Mol Biol. 2012. PMID: 22499708
-
Deep bootstrap for Bayesian inference.Philos Trans A Math Phys Eng Sci. 2023 May 15;381(2247):20220154. doi: 10.1098/rsta.2022.0154. Epub 2023 Mar 27. Philos Trans A Math Phys Eng Sci. 2023. PMID: 36970831
Cited by
-
A special issue on Bayesian inference: challenges, perspectives and prospects.Philos Trans A Math Phys Eng Sci. 2023 May 15;381(2247):20220155. doi: 10.1098/rsta.2022.0155. Epub 2023 Mar 27. Philos Trans A Math Phys Eng Sci. 2023. PMID: 36970829 Free PMC article. No abstract available.
LinkOut - more resources
Full Text Sources