English

Mean dimension of radial basis functions

Numerical Analysis 2023-02-21 v1 Numerical Analysis Computation

Abstract

We show that generalized multiquadric radial basis functions (RBFs) on Rd\mathbb{R}^d have a mean dimension that is 1+O(1/d)1+O(1/d) as dd\to\infty with an explicit bound for the implied constant, under moment conditions on their inputs. Under weaker moment conditions the mean dimension still approaches 11. As a consequence, these RBFs become essentially additive as their dimension increases. Gaussian RBFs by contrast can attain any mean dimension between 1 and d. We also find that a test integrand due to Keister that has been influential in quasi-Monte Carlo theory has a mean dimension that oscillates between approximately 1 and approximately 2 as the nominal dimension dd increases.

Keywords

Cite

@article{arxiv.2302.09689,
  title  = {Mean dimension of radial basis functions},
  author = {Christopher Hoyt and Art B. Owen},
  journal= {arXiv preprint arXiv:2302.09689},
  year   = {2023}
}
R2 v1 2026-06-28T08:44:01.429Z