We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Numerical Analysis

Title: Infinite-Variate $L^2$-Approximation with Nested Subspace Sampling

Abstract: We consider $L^2$-approximation on weighted reproducing kernel Hilbert spaces of functions depending on infinitely many variables. We focus on unrestricted linear information, admitting evaluations of arbitrary continuous linear functionals. We distinguish between ANOVA and non-ANOVA spaces, where, by ANOVA spaces, we refer to function spaces whose norms are induced by an underlying ANOVA function decomposition. In ANOVA spaces, we prove that there is an optimal algorithm to solve the approximation problem using linear information. This way, we can determine the exact polynomial convergence rate of $n$-th minimal worst-case errors. For non-ANOVA spaces, we also establish upper and lower error bounds. Even though the bounds do not match in this case, they reveal that for weights with a moderate decay behavior, the convergence rate of $n$-th minimal errors is strictly higher in ANOVA than in non-ANOVA spaces.
Comments: Submitted to the MCQMC 2022 conference proceedings
Subjects: Numerical Analysis (math.NA)
Cite as: arXiv:2301.13177 [math.NA]
  (or arXiv:2301.13177v1 [math.NA] for this version)

Submission history

From: Kumar Harsha [view email]
[v1] Mon, 30 Jan 2023 18:44:20 GMT (56kb)

Link back to: arXiv, form interface, contact.