A categorical foundation for Bayesian probability

From MaRDI portal




Abstract: Given two measurable spaces H and D with countably generated sigma-algebras, a perfect prior probability measure PH 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 widehatPH=Icircmu can be computed. This procedure is iterative: with each updated probability PH, 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.









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)