Probability measures and Milyutin maps between metric spaces (Q2518764): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
ReferenceBot (talk | contribs)
Changed an Item
 
(One intermediate revision by one other user not shown)
Property / arXiv ID
 
Property / arXiv ID: 0801.1721 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On exact atomless Milutin maps / rank
 
Normal rank
Property / cites work
 
Property / cites work: Topology of spaces of probability measures. I: The functors \(P_{\tau}\) and \(\hat P\) / rank
 
Normal rank
Property / cites work
 
Property / cites work: Topology of spaces of probability measures. II: Barycenters of Radon probability measures and the metrization of the functors \(P_{\tau}\) and \(\hat P\) / rank
 
Normal rank
Property / cites work
 
Property / cites work: The geometry of mappings of spaces of probability measures / rank
 
Normal rank
Property / cites work
 
Property / cites work: Open Mappings of Probability Measures and the Skorokhod Representation Theorem / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4040474 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Probability measures in topology / rank
 
Normal rank
Property / cites work
 
Property / cites work: Dense families of selections and finite-dimensional spaces / rank
 
Normal rank
Property / cites work
 
Property / cites work: Continuous selections. I / rank
 
Normal rank
Property / cites work
 
Property / cites work: Selected Selection Theorems / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5533878 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4761587 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On exact Milyutin mappings / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3288209 / rank
 
Normal rank

Latest revision as of 00:14, 29 June 2024

scientific article
Language Label Description Also known as
English
Probability measures and Milyutin maps between metric spaces
scientific article

    Statements

    Probability measures and Milyutin maps between metric spaces (English)
    0 references
    16 January 2009
    0 references
    The author proves that the functor \(\widehat P\) of Radon probability measures transforms open maps between completely metrizable spaces into soft maps. This result is applied to establish some properties of Milyutin maps between completely metrizable spaces. Let \(X\) be a metric space. Let \(\widehat P (X)\) denote the space of probability Radon measures on \(X\), equipped with the weak topology with respect to the space of bounded continuous functions on \(X\). The set of compactly supported probability Radon measures on \(X\) is denoted by \(P_{\beta}(X)\). \(\widehat P\) turns out to be a functor in the category of metrizable spaces and continuous maps; in particular, to every continuous function \(f : X \to Y\) one can associate a map \(\widehat P (f):\widehat P (X) \to\widehat P (Y)\), for more details, see e.g. [\textit{T. O. Banakh}, Mat. Stud. 5, 65--87 (1995; Zbl 1023.28502)]. Let \(X, Y\) be metric spaces. A continuous function \(f : X \to Y\) is a Milyutin map if there exists a continuous function \(g : Y \to\widehat P(X)\) such that \(\text{supp~} g(y ) \subseteq f^{-1}(y)\) for every \(y \in Y\). Such a \(g\) is called a choice map associated with \(f\). The set of all choice maps associated with \(f\) is denoted by \(Ch_{f} (Y, X)\) and it is equipped with the source limitation topology. A Milyutin map \(f\) is exact (atomless, resp.) if for every \(y \in Y\), \(\text{supp~}g(y )= f^{-1}(y)\) (\(g(y )\) is an atomless measure, resp.). A Milyutin map \(f : X \to Y\) is densely atomless if \[ \{g \in Ch_{f} (Y, X) : g(y ) \text{ is atomless for all }y \in Y\} \] is a dense \(G_{\delta}\) set in \(Ch_{f} (Y, X)\). A continuous function \(f : X \to Y\) is soft if for every metric space \(Z\), any closed set \(A \subseteq Z\) and continuous maps \(g : Z \to Y\), \(h : A \to X\) with \((f\circ h)= g|_{A}\) there exists a continuous map \(\widetilde g : Z \to X\) such that \(\widetilde g \) extends \(h\) and \(f \circ\widetilde g = g\). The author obtains the following results on \(\widehat P\) and Milyutin maps. Theorem. Let \(X, Y\) be complete metric spaces and let \(f : X \to Y\) be a continuous surjective open map. Then \(\widehat P (f ) :\widehat P (X) \to\widehat P (Y )\) is a soft map. The special cases of this result when \(X,Y\) are both compact or separable were established in \textit{V. V. Fedorchuk} [Russ. Math. Surv. 46, No. 1, 45--93 (1991; Zbl 0735.54033); translation from Usp. Mat. Nauk 46, No.~1(227), 41--80 (1991)] and \textit{T. Banakh} and \textit{T. Radul} [Mat. Stud. 11, No.~1, 17--30 (1999; Zbl 1023.28500)]. Theorem. Let \(X, Y\) be complete metric spaces. A continuous surjection \(f : X \to Y\) is an atomless Milyutin map if and only if there exists a complete subspace \(X_0 \subseteq X\) such that \(f_0 = f |_{X_{0}} : X_0 \to Y\) is an open surjection and all fibers of \(f_0\) are perfect sets. Moreover, for any such \(f\) there exists a continuous map \(f ^{\star} : P_{\beta} (Y) \to\widehat P(X)\) such that every \(\mu \in P_{\beta} (Y)\), \(f^{\star} (\mu)\) is atomless and \([\widehat P(f)](f^{\star}(\mu)) = \mu\). Theorem. Let \(X, Y\) be complete metric spaces. Then every continuous open surjection \(f : X \to Y\) with perfect fibers is a densely atomless Milyutin map. Theorem. Let \(X, Y\) be complete metric spaces and \(M\) be a separable metric space. Let \(f : X \to Y\) be a continuous open surjection and let \(\pi : X \to M\) be a continuous function. Then the choice maps \(h \in Ch_{f} (Y, X)\) such that \(\pi (\text{supp }h(y ))\) is dense in \(\pi (f^{-1}(y ))\) for every \(y\in Y\) form a dense \(G_{\delta}\) set in \(Ch_{f} (Y, X)\).
    0 references
    Radon probability measure functor
    0 references
    Milyutin map
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references