Superconvergence of kernel-based interpolation

DOI10.1016/J.JAT.2018.05.002MaRDI QIDQ1663841zbMATH OpenOpenAlexFDO

Authors Robert Schaback

Publication date 24 August 2018

Published in Journal of Approximation Theory (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1607.04219

zbMATH Keywords

convergenceerror boundsinterpolation

Mathematics Subject Classification ID

Interpolation in approximation theory (41A05) Eigenvalue problems for integral equations (45C05) Numerical interpolation (65D05)

Abstract: It is well-known that univariate cubic spline interpolation, if carried out on point sets with fill distance $h$, converges only like $calO(h^2)$ in $L_2[a,b]$ for functions in $W_2^2[a,b]$ if no additional assumptions are made. But superconvergence up to order $h^4$ occurs if more smoothness is assumed and if certain additional boundary conditions are satisfied. This phenomenon was generalized in 1999 to multivariate interpolation in Reproducing Kernel Hilbert Spaces on domains $Omegasubset{R}^d$ for continuous positive definite Fourier-transformable shift-invariant kernels on $R^d$. But the sufficient condition for superconvergence given in 1999 still needs further analysis, because the interplay between smoothness and boundary conditions is not clear at all. Furthermore, if only additional smoothness is assumed, superconvergence is numerically observed in the interior of the domain, but without explanation, so far. This paper first generalizes the "improved error bounds" of 1999 by an abstract theory that includes the Aubin-Nitsche trick and the known superconvergence results for univariate polynomial splines. Then the paper analyzes what is behind the sufficient conditions for superconvergence. They split into conditions on smoothness and localization, and these are investigated independently. If sufficient smoothness is present, but no additional localization conditions are assumed, it is proven that superconvergence always occurs in the interior of the domain. If smoothness and localization interact in the kernel-based case on $R^d$, weak and strong boundary conditions in terms of pseudodifferential operators occur. A special section on Mercer expansions is added, because Mercer eigenfunctions always satisfy the sufficient conditions for superconvergence. Numerical examples illustrate the theoretical findings.

Review text

In this paper the author gives a very interesting idea about the well known error bounds that he generalizes by an abstract theory and the known superconvergence results for univariate polynomial splines. He studies what is behind the sufficient conditions for superconvergence. If sufficient smoothness is present, but no additional localization conditions are assumed, it is numerically observed that superconvergence always occurs in the interior of the domain. Several examples are given.

Recommendations

• scientific article; zbMATH DE number 3942121
• Superconvergence points of polynomial spectral interpolation
• scientific article; zbMATH DE number 17127
• scientific article; zbMATH DE number 713465
• Superconvergence of Jacobi-Gauss-type spectral interpolation

Cites work

• Accuracy of radial basis function interpolation and derivative approximations on 1 D infinite grids
• An improved order of approximation for thin-plate spline intepolation in the unit disc
• An inverse theorem for compact Lipschitz regions in \(\mathbb {R}^d\) using localized kernel bases
• Approximation orders for interpolation by surface splines to rough functions
• Approximation orders for natural splines in arbitrary dimensions
• Closed form representations and properties of the generalised Wendland functions
• Comparisons of fundamental solutions and particular solutions for Trefftz methods
• Convolution roots of radial positive definite functions with compact support
• Extending the Range of Error Estimates for Radial Approximation in Euclidean Space and on Spheres
• Finite elements. Theory, fast solvers, and applications in solid mechanics. Transl. from the German by Larry L. Schumaker.
• High Order Local Approximations to Derivatives in the Finite Element Method
• Higher Order Local Accuracy by Averaging in the Finite Element Method
• scientific article; zbMATH DE number 5866652 (Why is no real title available?)
• scientific article; zbMATH DE number 5190601 (Why is no real title available?)
• scientific article; zbMATH DE number 44907 (Why is no real title available?)
• scientific article; zbMATH DE number 2204738 (Why is no real title available?)
• scientific article; zbMATH DE number 3254460 (Why is no real title available?)
• Improved error bounds for scattered data interpolation by radial basis functions
• L-splines
• Local accuracy for radial basis function interpolation on finite uniform grids
• Numerical analyses of the boundary effect of radial basis functions in 3D surface reconstruction
• Observations on the behavior of radial basis function approximations near boundaries
• On a method for interpolating functions on chaotic nets
• On a Polyharmonic Dirichlet Problem and Boundary Effects in Surface Spline Approximation
• Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree
• Radial Basis Functions
• Sampling and Stability
• Scattered Data Approximation
• Sobolev error estimates and a Bernstein inequality for scattered data interpolation via radial basis functions
• Stability of kernel-based interpolation
• Superconvergence in Galerkin finite element methods
• The L_p-approximation order of surface spline interpolation for 1 p 2
• The \(L_{2}\)-approximation order of surface spline interpolation
• The Mathematical Theory of Finite Element Methods

Cited in

(20)

• Discrete least-squares radial basis functions approximations
• Kernel-based interpolation at approximate Fekete points
• Sampling based approximation of linear functionals in reproducing kernel Hilbert spaces
• A novel class of stabilized greedy kernel approximation algorithms: convergence, stability and uniform point distribution
• Analysis of target data-dependent greedy kernel algorithms: convergence rates for \(f\)-, \(f \cdot P\)- and \(f/P\)-greedy
• scientific article; zbMATH DE number 3942121 (Why is no real title available?)
• scientific article; zbMATH DE number 1475033 (Why is no real title available?)
• A kernel-based meshless conservative Galerkin method for solving Hamiltonian wave equations
• Maximum likelihood estimation and uncertainty quantification for Gaussian process approximation of deterministic functions
• A type of super convergent formulas for numerical divided differences
• Asymptotic Bounds for Smoothness Parameter Estimates in Gaussian Process Interpolation
• Error bounds and the asymptotic setting in kernel-based approximation
• Optimal compactly supported functions in Sobolev spaces
• Sharp inverse statements for kernel interpolation
• Spectral equivalence of unsymmetric kernel matrices and applications
• Signal reconstruction using determinantal sampling
• Efficient numerical integration in reproducing kernel Hilbert spaces via leverage scores sampling
• Comparing scale parameter estimators for Gaussian process interpolation with the Brownian motion prior: leave-one-out cross validation and maximum likelihood
• Proving the stability estimates of variational least-squares kernel-based methods
• Kernel interpolation on generalized sparse grids

Describes a project that uses

Uses Software

• GaussQR

This page was built for publication: Superconvergence of kernel-based interpolation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1663841)