Disintegration of measures

Let \((X,B)\), \((Y,C)\) be measurable spaces and let \(\mu\) be a measure on the product measure space. A disintegration of \(\mu\) is a representation \[ \mu= \int\nu_y d\gamma(y),\tag{1} \] where for each \(y\) in \(Y\), \(\nu_y\) is a measure on \(X\). The meaning of this formula is that for each \(\phi\) in some specified class of functions on \(X\times Y\), \[ \int \phi d\mu= \int\biggl(\int \phi(x,y)d\nu_y(x)\biggr) d\gamma(y).\tag{2} \] In this paper, two theorems of this kind are proved: Theorem 1: Let \(T\) be the unit circle, \(Y\) a compact Hausdorff space, and \(\mu\) a probability measure on the Baire field of \(T\times Y\). There is a probability measure \(\gamma\) on \(Y\), and for almost every \(y\) in \(Y\) (with respect to \(\gamma\)) a probability measure \(\nu_y\) on \(T\) such that (2) holds for all continuous functions \(\phi\) on the product space. Theorem 2: Let \(X\) be a standard Borel space and \(Y\) a compact Hausdorff space. Give \(X\times Y\) the Borel structure, that is the product of the Borel field of \(X\) and the Baire field of \(Y\). Then every probability measure \(\mu\) on \(X\times Y\) has a disintegration, and (2) holds for all bounded measurable functions \(\phi\). For the entire collection see [Zbl 0879.00047].

Recommendations

Cited in

(17)

This page was built for publication: Disintegration of measures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4370038)