Entropy numbers of finite dimensional mixed-norm balls and function space embeddings with small mixed smoothness (Q2663135): Difference between revisions

Given a compact set \(K\) in a quasi-Banach space \(Y,\) the \(k\)-entropy number \(e_{k}(K, Y)\) is defined as the smallest radius \(\varepsilon >0\) such that \(K\) can be covered with \(2^{k-1}\) copies of the ball \(\varepsilon B_{Y}.\) This concept can be also extended to operators in the following way. Given a compact operator \(T: X\rightarrow Y,\) where \(X, Y\) are quasi-Banach spaces, the \(k\)-entropy number of the operator \(T\) is defined as \(e_{k}(T)= e_{k} (T(B_{X}), Y).\) Entropy numbers constitute an interesting concept in approximation theory and appear in many approximation problems as the estimation of learning rates for machine learning problems, bounding \(s\)-numbers such as approximation Gelfand, or Kolmogorov numbers from below, among others. The aim of this contribution is to improve estimates for entropy numbers of embeddings between function spaces of dominating mixed smoothness. As particular cases one gets the scale of Besov spaces of dominating mixed smoothness and the scale of Triebel-Lizorkin spaces, which includes \(L_{p}\) and Sobolev spaces of mixed smoothness. In practice, the behavior of these numbers follows an asymptotic decay. This is also a well-known fact for \(s\)-numbers of these embeddings as approximation, Gelfand of Kolmogorov numbers (see [\textit{D. Dũng} et al., Hyperbolic cross approximation. Lecture notes given at the courses on constructive approximation and harmonic analysis, Barcelona, Spain, May 30 -- June 3, 2016. Cham: Birkhäuser (2018; Zbl 1414.41001)]). The dimension appears in the logarithmic term. The logarithmic term completely disappears in regimes of small smoothness. Indeed, the authors of the paper under review establish purely polynomial asymptotics bounds depending on the underlaying dimension only in the constant term. The key ingredient is a counterpart of the classical result [\textit{C. Schuett}, J. Approx. Theory 40, 121--128 (1984; Zbl 0497.41017)], as well as a general abstract version of such a theorem for operators between vector-valued sequence spaces given in [\textit{D. E. Edmunds} and \textit{Yu. Netrusov}, J. Approx. Theory 178, 13--21 (2014; Zbl 1286.47015)], that is carefully revisited. The cases where a Besov type sequence space compactly embeds into a Triebel-Lizorkin type sequence space as well as where a Triebel-Lizorkin type sequence space compactly embeds into a Besov type sequence space, respectively, are analyzed.