Embedding theorems and strong approximation

From MaRDI portal





The authors consider the following subclasses of continuous functions: \(S_ p(\lambda):=f:\| \sum^{\infty}_{n=0}\lambda_ n| s_ n-f|^ p\| <\infty \},\) \(V_ p(\lambda):=\{f:\| \sum^{\infty}_{n=0}\lambda_ n| \tau_ n-f|^ p\| <\infty \},\) \(H^{\omega}:=\{f:\omega (f;\delta)=O(\omega (\delta))\}.\) They prove the assertions: Theorem 1. If \(p\geq 1\) and \(\{\lambda_ n\}\) is a monotonic (nondecreasing or nonincreasing) sequence of positive numbers satisfying the condition \(\lambda_ n/\lambda_{2n}\leq K\) \((n=1,2,...)\), then \(S_ p(\lambda)\subset V_ p(\lambda)\) holds. Theorem 2. Let \(\{\lambda_ n\}\) be a monotonic sequence of positive numbers, furthermore let \(\omega\) be a modulus of continuity and \(0<p<\infty\). Then condition (*) \(\sum^{n}_{k=1}(k\lambda_ k)^{- 1/p}=O(n\omega (1/n))\) implies (**) \(V_ p(\lambda)\subset H^{\omega}\). If \(p\geq 1\) and there exists a number Q such that \(0\leq Q<1\) and \(n^ Q\lambda_ n\uparrow\) holds, then conversely (**) implies (*). These theorems extended the results of \textit{V. G. Krotov} the first author [ibid. 40, 93-98 (1978; Zbl 0425.42005)]. In the case of the class \(F_ p(\lambda):=\{f:\| \sum^{\infty}_{n=0}\lambda_ n| \sigma_ n-f|^ p\| <\infty \}\) see the paper by \textit{J. Németh} [ibid. 48, 375-378 (1985)].











This page was built for publication: Embedding theorems and strong approximation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3697553)