A categorical foundation for Bayesian probability

DOI10.1007/S10485-013-9324-9zbMATH Open1302.60011arXiv1205.1488OpenAlexW3101547338MaRDI QIDQ742929FDOQ742929

Jared Culbertson, Kirk Sturtz

Publication date: 19 September 2014

Published in: Applied Categorical Structures (Search for Journal in Brave)

Abstract: Given two measurable spaces $H$ and $D$ with countably generated $sigma$-algebras, a perfect prior probability measure $P_H$ on $H$ and a sampling distribution $S:HightarrowD$, there is a corresponding inference map $I:DightarrowH$ which is unique up to a set of measure zero. Thus, given a data measurement $mu:1ightarrowD$, a posterior probability $widehat{P}_H=Icircmu$ can be computed. This procedure is iterative: with each updated probability $P_H$, we obtain a new joint distribution which in turn yields a new inference map $I$ and the process repeats with each additional measurement. The main result uses an existence theorem for regular conditional probabilities by Faden, which holds in more generality than the setting of Polish spaces. This less stringent setting then allows for non-trivial decision rules (Eilenberg--Moore algebras) on finite (as well as non finite) spaces, and also provides for a common framework for decision theory and Bayesian probability.

Full work available at URL: https://arxiv.org/abs/1205.1488

Recommendations

• A categorical approach to probability theory
• Categorical probabilistic theories
• Towards a categorical account of conditional probability
• scientific article; zbMATH DE number 7595320
• A Convenient Category for Higher-Order Probability Theory
• scientific article; zbMATH DE number 1746710
• Foundations of Bayesian theory
• Categorical Stochastic Processes and Likelihood
• Foundations of Bayesianism
• A note on probability metrics in a categorical setting

zbMATH Keywords

decision rulesBayesian probabilityGiry monadregular conditional probability

Mathematics Subject Classification ID

Bayesian problems; characterization of Bayes procedures (62C10) Axioms; other general questions in probability (60A05)

Cites Work

• Title not available (Why is that?)
• Real Analysis and Probability
• Statistical decision theory and Bayesian analysis. 2nd ed
• Probability Theory
• Title not available (Why is that?)
• Nuclear and trace ideals in tensored \(^*\)-categories
• The existence of regular conditional probabilities: Necessary and sufficient conditions
• Optimal transportation with capacity constraints
• Kleisli morphisms and randomized congruences for the Giry monad
• Perfect mixtures of perfect measures
• Commutative monads as a theory of distributions
• Disintegration and compact measures.
• Title not available (Why is that?)
• Perfect Probability Measures and Regular Conditional Probabilities

Cited In (16)

• A Convenient Category for Higher-Order Probability Theory
• Quantum GestART: identifying and applying correlations between mathematics, art, and perceptual organization
• The factorization of the Giry monad
• A non-commutative Bayes' theorem
• Title not available (Why is that?)
• Pointless Learning
• Weak and TV consistency in Bayesian uncertainty quantification using disintegration
• Categorical Stochastic Processes and Likelihood
• Giry and the Machine
• Belief propagation in monoidal categories
• Categorical approach to modelling and to coupling of models
• Conditional distributions for quantum systems
• From probability monads to commutative effectuses
• Multinomial and hypergeometric distributions in Markov categories
• A channel-based perspective on conjugate priors
• Disintegration and Bayesian inversion via string diagrams

This page was built for publication: A categorical foundation for Bayesian probability

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q742929)