No label defined (Q4344739)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1036337
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 1036337 |
Statements
26 February 1998
0 references
Korovkin closure
0 references
normed vector lattice
0 references
\(T\)-Choquet boundary
0 references
0.8713740706443787
0 references
0.8526524901390076
0 references
The paper gives an extension of the Korovkin theorem, in which the identity operator is replaced by an arbitrary linear positive operator acting on a space of vector-valued functions. Consider the following notations. Let \(X\) be a compact Hausdorff space and \(E\) be a Dedekind complete normed vector lattice which contains an element \(e\) such that \(e>0\), \(|e|=1\) and \(|a|\leq|a|e\) for all \(a\in E\). Denote by \(B(X,E)\) the normed vector lattice of all \(E\)-valued bounded functions on \(X\) and by \(C(X,E)\) the linear sublattice of continuous functions. On \(B(X,E)\) it is considered the sup-norm denoted by \(|\cdot|\). Denote also, by \(\rho\), the function given by \(\rho(x)=e\), for all \(x\in X\). Let \(A(X,E)\) be a linear sublattice of \(C(X,E)\) which contains \(\rho\), let \(M\) be a linear subspace of \(A(X,E)\) which contains \(\rho\) and let \(T\) be a positive linear operator of \(A(X,E)\) into \(C(X,E)\). For a function \(f\in A(X,E)\) and \(x\in X\) denote \(f_T^*:= \inf\{T(h)(x)\); \(f\leq h\), \(h\in M\}\) and \(f_*^T:= \sup\{T(h)(x)\); \(h\leq f\), \(h\in M\}\). Set \(\widehat{M}_T(x):=\{f\in A(X,E)\); \(f_*^T(x)= f_T^*(x)\}\). Denote by \(R_x^T(M)\) the set of all \(M_T\) representing operators for \(x\), that is the set of all linear positive operators \(\mu: A(X,E)\to E\) such that \(\mu(g)= T(g)(x)\), for all \(g\in M\). Denote by \(\partial_M^T(X)\) the \(T\)-Choquet boundary of \(X\), defined by \(\partial_M^T(X):= \{x\in X\); \(R_x^T(M)= \{\delta_x\circ T\}\}\), where \(\delta_x\) is given by \(\delta_x(g):= g(x)\) for all \(g\in B(X,E)\). If \(A,B\) are normed spaces, \({\mathfrak L}\) is a class of mappings from \(A\) into \(B\), \(S\subset A\) and \(L\in{\mathfrak L}\), let denote by \(\text{Kor} ({\mathfrak L},S,L)\) the set of all \(f\in A\) with the property that if \(\{L_{\alpha,\lambda}\); \(\alpha\in D\), \(\lambda\in\Lambda\}\) is an arbitrary family of mappings in \({\mathfrak L}\), satisfying \(\lim_\alpha|L_{\alpha,\lambda}(g)- L(g)|=0\), uniformly in \(\lambda\in \Lambda\), for all \(g\in S\), then \(\lim_\alpha|L_{\alpha,\lambda}(f)- L(f)|=0\), uniformly in \(\lambda\in \Lambda\). Denote by \({\mathfrak I}[A,B]\), the class of all increasing mappings from \(A\) into \(B\) and by \({\mathfrak B}[A,B]\) the class of all positive linear operators from \(A\) into \(B\). The central result is (Theorem 3): The following assertions are equivalent: (a) \(\partial_M^T(X)=X\); (b) \(\text{Kor} ({\mathfrak I}[A(X,E), B(X,E)]; M,T)= A(X,E)\); (c) \(\text{Kor} ({\mathfrak B}[A(X,E), B(X,E)]; M,T)= A(X,E)\); (d) \(\widehat{M}_T= A(X,E)\).
0 references