Showing 1–14 of 14 results for author: Narcowich, F J

Highly Localized RBF Lagrange Functions for Finite Difference Methods on Spheres

Authors: Wolfgang Erb, Thomas Hangelbroek, Francis J. Narcowich, Christian Rieger, Joseph D. Ward

Abstract: The aim of this paper is to show how rapidly decaying RBF Lagrange functions on the spheres can be used to create effective, stable finite difference methods based on radial basis functions (RBF-FD). For certain classes of PDEs this approach leads to precise convergence estimates for stencils which grow moderately with increasing discretization fineness. The aim of this paper is to show how rapidly decaying RBF Lagrange functions on the spheres can be used to create effective, stable finite difference methods based on radial basis functions (RBF-FD). For certain classes of PDEs this approach leads to precise convergence estimates for stencils which grow moderately with increasing discretization fineness. △ Less

Submitted 16 February, 2023; originally announced February 2023.

MSC Class: 65D12; 65N06

Interpolating splines on graphs for data science applications

Authors: John Paul Ward, Francis J. Narcowich, Joseph D. Ward

Abstract: We introduce intrinsic interpolatory bases for data structured on graphs and derive properties of those bases. Polyharmonic Lagrange functions are shown to satisfy exponential decay away from their centers. The decay depends on the density of the zeros of the Lagrange function, showing that they scale with the density of the data. These results indicate that Lagrange-type bases are ideal building… ▽ More We introduce intrinsic interpolatory bases for data structured on graphs and derive properties of those bases. Polyharmonic Lagrange functions are shown to satisfy exponential decay away from their centers. The decay depends on the density of the zeros of the Lagrange function, showing that they scale with the density of the data. These results indicate that Lagrange-type bases are ideal building blocks for analyzing data on graphs, and we illustrate their use in kernel-based machine learning applications. △ Less

Submitted 19 April, 2020; v1 submitted 27 June, 2018; originally announced June 2018.

Comments: 17 pages

MSC Class: 41A05; 41A15; 41A65

A Meshless Galerkin Method For Non-Local Diffusion Using Localized Kernel Bases

Authors: Richard B. Lehoucq, Francis J. Narcowich, Stephen T. Rowe, Joseph D. Ward

Abstract: We introduce a meshless method for solving both continuous and discrete variational formulations of a volume constrained, nonlocal diffusion problem. We use the discrete solution to approximate the continuous solution. Our method is nonconforming and uses a localized Lagrange basis that is constructed out of radial basis functions. By verifying that certain inf-sup conditions hold, we demonstrate… ▽ More We introduce a meshless method for solving both continuous and discrete variational formulations of a volume constrained, nonlocal diffusion problem. We use the discrete solution to approximate the continuous solution. Our method is nonconforming and uses a localized Lagrange basis that is constructed out of radial basis functions. By verifying that certain inf-sup conditions hold, we demonstrate that both the continuous and discrete problems are well-posed, and also present numerical and theoretical results for the convergence behavior of the method. The stiffness matrix is assembled by a special quadrature routine unique to the localized basis. Combining the quadrature method with the localized basis produces a well-conditioned, symmetric matrix. This then is used to find the discretized solution. △ Less

Submitted 11 January, 2016; originally announced January 2016.

Comments: 23 pages, 2, figures, submitted to Math Comp

An inverse theorem for compact Lipschitz regions in $R^d$ using localized kernel bases

Authors: Thomas Hangelbroek, Francis J. Narcowich, Christian Rieger, Joseph D. Ward

Abstract: While inverse estimates in the context of radial basis function approximation on boundary-free domains have been known for at least ten years, such theorems for the more important and difficult setting of bounded domains have been notably absent. This article develops inverse estimates for finite dimensional spaces arising in radial basis function approximation and meshless methods. The inverse es… ▽ More While inverse estimates in the context of radial basis function approximation on boundary-free domains have been known for at least ten years, such theorems for the more important and difficult setting of bounded domains have been notably absent. This article develops inverse estimates for finite dimensional spaces arising in radial basis function approximation and meshless methods. The inverse estimates we consider control Sobolev norms of linear combinations of a localized basis by the $L_p$ norm over a bounded domain. The localized basis is generated by forming local Lagrange functions for certain types of RBFs (namely Matérn and surface spline RBFs). In this way it extends the boundary-free construction of Fuselier, Hangelbroek, Narcowich, Ward and Wright. △ Less

Submitted 29 October, 2016; v1 submitted 12 August, 2015; originally announced August 2015.

MSC Class: 41A17; 41A27; 41A63

Direct and Inverse Results on Bounded Domains for Meshless Methods via Localized Bases on Manifolds

Authors: Thomas Hangelbroek, Francis J. Narcowich, Christian Rieger, Joseph D. Ward

Abstract: This article develops direct and inverse estimates for certain finite dimensional spaces arising in kernel approximation. Both the direct and inverse estimates are based on approximation spaces spanned by local Lagrange functions which are spatially highly localized. The construction of such functions is computationally efficient and generalizes the construction given by the authors for restricted… ▽ More This article develops direct and inverse estimates for certain finite dimensional spaces arising in kernel approximation. Both the direct and inverse estimates are based on approximation spaces spanned by local Lagrange functions which are spatially highly localized. The construction of such functions is computationally efficient and generalizes the construction given by the authors for restricted surface splines on $\mathbb{R}^d$. The kernels for which the theory applies includes the Sobolev-Matérn kernels for closed, compact, connected, $C^\infty$ Riemannian manifolds. △ Less

Submitted 6 September, 2017; v1 submitted 5 June, 2014; originally announced June 2014.

Comments: 29 pages. To appear in Festschrift for the 80th Birthday of Ian Sloan

MSC Class: 41A17; 41A27; 41A63

A Novel Galerkin Method for Solving PDEs on the Sphere Using Highly Localized Kernel Bases

Authors: F. J. Narcowich, Stephen T. Rowe, Joseph D. Ward

Abstract: We present a novel Galerkin method for solving partial differential equations on the sphere. The problem is discretized by a highly localized basis which is easily constructed. The stiffness matrix entries are computed by a recently developed quadrature formula unique to the localized bases we consider. We present error estimates and investigate the stability of the discrete stiffness matrix. Impl… ▽ More We present a novel Galerkin method for solving partial differential equations on the sphere. The problem is discretized by a highly localized basis which is easily constructed. The stiffness matrix entries are computed by a recently developed quadrature formula unique to the localized bases we consider. We present error estimates and investigate the stability of the discrete stiffness matrix. Implementation and numerical experiments are discussed. △ Less

Submitted 16 February, 2015; v1 submitted 21 April, 2014; originally announced April 2014.

Comments: 22 pages, 1 figure. Changes to previous version include new numerical experiments, updated figures, new styling, and re-written sections, and additional theory to address the use of local Lagrange functions for the discretization basis

Kernel Based Quadrature on Spheres and Other Homogeneous Spaces

Authors: E. Fuselier, T. Hangelbroek, F. J. Narcowich, J. D. Ward, G. B. Wright

Abstract: Quadrature formulas for spheres, the rotation group, and other compact, homogeneous manifolds are important in a number of applications and have been the subject of recent research. The main purpose of this paper is to study coordinate independent quadrature (or cubature) formulas associated with certain classes of positive definite and conditionally positive definite kernels that are invariant un… ▽ More Quadrature formulas for spheres, the rotation group, and other compact, homogeneous manifolds are important in a number of applications and have been the subject of recent research. The main purpose of this paper is to study coordinate independent quadrature (or cubature) formulas associated with certain classes of positive definite and conditionally positive definite kernels that are invariant under the group action of the homogeneous manifold. In particular, we show that these formulas are accurate -- optimally so in many cases -- and stable under an increasing number of nodes and in the presence of noise, provided the set X of quadrature nodes is quasi-uniform. The stability results are new in all cases. In addition, we may use these quadrature formulas to obtain similar formulas for manifolds diffeomorphic to $\mathbb{S}^n$, oblate spheroids for instance. The weights are obtained by solving a single linear system. For $\mathbb{S}^2$, and the restricted thin plate spline kernel $r^2 \log r$, these weights can be computed for two-thirds of a million nodes, using a preconditioned iterative technique introduced by us. △ Less

Submitted 21 November, 2012; originally announced November 2012.

Comments: 31 pages, 8 figures

MSC Class: 65D32; 65D30; 65D05; 41A05; 41A30; 41A63

Localized bases for kernel spaces on the unit sphere

Authors: E. Fuselier, T. Hangelbroek, F. J. Narcowich, J. D. Ward, G. B. Wright

Abstract: Approximation/interpolation from spaces of positive definite or conditionally positive definite kernels is an increasingly popular tool for the analysis and synthesis of scattered data, and is central to many meshless methods. For a set of $N$ scattered sites, the standard basis for such a space utilizes $N$ \emph{globally} supported kernels; computing with it is prohibitively expensive for large… ▽ More Approximation/interpolation from spaces of positive definite or conditionally positive definite kernels is an increasingly popular tool for the analysis and synthesis of scattered data, and is central to many meshless methods. For a set of $N$ scattered sites, the standard basis for such a space utilizes $N$ \emph{globally} supported kernels; computing with it is prohibitively expensive for large $N$. Easily computable, well-localized bases, with "small-footprint" basis elements - i.e., elements using only a small number of kernels -- have been unavailable. Working on $\sphere$, with focus on the restricted surface spline kernels (e.g. the thin-plate splines restricted to the sphere), we construct easily computable, spatially well-localized, small-footprint, robust bases for the associated kernel spaces. Our theory predicts that each element of the local basis is constructed by using a combination of only $\mathcal{O}((\log N)^2)$ kernels, which makes the construction computationally cheap. We prove that the new basis is $L_p$ stable and satisfies polynomial decay estimates that are stationary with respect to the density of the data sites, and we present a quasi-interpolation scheme that provides optimal $L_p$ approximation orders. Although our focus is on $\mathbb{S}^2$, much of the theory applies to other manifolds - $\mathbb{S}^d$, the rotation group, and so on. Finally, we construct algorithms to implement these schemes and use them to conduct numerical experiments, which validate our theory for interpolation problems on $\mathbb{S}^2$ involving over one hundred fifty thousand data sites. △ Less

Submitted 15 May, 2012; originally announced May 2012.

Comments: This article supersedes arXiv:1111.1013 "Better bases for kernel spaces," which proved existence of better bases for various kernel spaces. This article treats a smaller class of kernels, but presents an algorithm for constructing better bases and demonstrates its effectiveness with more elaborate examples. A quasi-interpolation scheme is introduced that provides optimal linear convergence rates

MSC Class: 41A05; 41A30; 41A63; 65D05

Journal ref: SIAM J. Numer. Anal. 51-5 (2013), pp. 2538-2562

Better bases for kernel spaces

Authors: E. J. Fuselier, T. C. Hangelbroek, F. J. Narcowich, J. D. Ward, G. B. Wright

Abstract: In this article we investigate the feasibility of constructing stable, local bases for computing with kernels. In particular, we are interested in constructing families $(b_ξ)_{ξ\inΞ}$ that function as bases for kernel spaces $S(k,Ξ)$ so that each basis function is constructed using very few kernels. In other words, each function $b_ζ(x) = \sum_{ξ\inΞ} A_{ζ,ξ} k(x,ξ)$ is a linear combination of sa… ▽ More In this article we investigate the feasibility of constructing stable, local bases for computing with kernels. In particular, we are interested in constructing families $(b_ξ)_{ξ\inΞ}$ that function as bases for kernel spaces $S(k,Ξ)$ so that each basis function is constructed using very few kernels. In other words, each function $b_ζ(x) = \sum_{ξ\inΞ} A_{ζ,ξ} k(x,ξ)$ is a linear combination of samples of the kernel with few nonzero coefficients $A_{ζ,ξ}$. This is reminiscent of the construction of the B-spline basis from the family of truncated power functions. We demonstrate that for a large class of kernels (the Sobolev kernels as well as many kernels of polyharmonic and related type) such bases exist. In fact, the basis elements can be constructed using a combination of roughly $O(\log N)^d$ kernels, where $d$ is the local dimension of the manifold and $N$ is the dimension of the kernel space (i.e. $N=#Ξ$). Viewing this as a preprocessing step -- the construction of the basis has computational cost $O(N(\log N)^d)$. Furthermore, we prove that the new basis is $L_p$ stable and satisfies polynomial decay estimates that are stationary with respect to the density of $Ξ$. △ Less

Submitted 3 November, 2011; originally announced November 2011.

Comments: 26 pages, 5 figures, 3 tables

MSC Class: 65D05; 65F08; 41A05

Polyharmonic and Related Kernels on Manifolds: Interpolation and Approximation

Authors: T. Hangelbroek, F. J. Narcowich, J. D. Ward

Abstract: This article is devoted to developing a theory for effective kernel interpolation and approximation in a general setting. For a wide class of compact, connected $C^\infty$ Riemannian manifolds, including the important cases of spheres and SO(3), we establish, using techniques involving differential geometry and Lie groups, that the kernels obtained as fundamental solutions of certain partial diffe… ▽ More This article is devoted to developing a theory for effective kernel interpolation and approximation in a general setting. For a wide class of compact, connected $C^\infty$ Riemannian manifolds, including the important cases of spheres and SO(3), we establish, using techniques involving differential geometry and Lie groups, that the kernels obtained as fundamental solutions of certain partial differential operators generate Lagrange functions that are uniformly bounded and decay away from their center at an algebraic rate, and in certain cases, an exponential rate. An immediate corollary is that the corresponding Lebesgue constants for interpolation as well as for $L_2$ minimization are uniformly bounded with a constant whose only dependence on the set of data sites is reflected in the mesh ratio, which measures the uniformity of the data. The kernels considered here include the restricted surface splines on spheres, as well as surface splines for SO(3), both of which have elementary closed-form representations that are computationally implementable. In addition to obtaining bounded Lebesgue constants in this setting, we also establish a "zeros lemma" for domains on compact Riemannian manifolds -- one that holds in as much generality as the corresponding Euclidean zeros lemma (on Lipschitz domains satisfying interior cone conditions) with constants that clearly demonstrate the influence of the geometry of the boundary (via cone parameters) as well as that of the Riemannian metric. △ Less

Submitted 19 April, 2011; v1 submitted 21 December, 2010; originally announced December 2010.

Comments: Final manuscript. Contains a new zeros lemma for regions on compact Riemannian manifolds with Lipschitz boundaries -- constants are in terms of parameters from a cone condition. 2 figures. 41 pages

MSC Class: 41A05; 41A63; 46E22; 46E35

Journal ref: Foundations of Computational Mathematics. October 2012, Volume 12, Issue 5, pp 625-670

Cardinal Interpolation with Gaussian Kernels

Authors: Thomas Hangelbroek, Wolodymyr Madych, F. J. Narcowich, J. D. Ward

Abstract: In this paper, interpolation by scaled multi-integer translates of Gaussian kernels is studied. The main result establishes $L_p$ Sobolev error estimates and shows that the error is controlled by the $L_p$ multiplier norm of a Fourier multiplier closely related to the cardinal interpolant, and comparable to the Hilbert transform. Consequently, its multiplier norm is bounded independent of the grid… ▽ More In this paper, interpolation by scaled multi-integer translates of Gaussian kernels is studied. The main result establishes $L_p$ Sobolev error estimates and shows that the error is controlled by the $L_p$ multiplier norm of a Fourier multiplier closely related to the cardinal interpolant, and comparable to the Hilbert transform. Consequently, its multiplier norm is bounded independent of the grid spacing when $1<p<\infty$, and involves a logarithmic term when $p=1$ or $\infty$. △ Less

Submitted 18 August, 2010; originally announced August 2010.

Comments: 18 pages

MSC Class: 41A15; 41A25; 41A63; 42B15

Journal ref: Journal of Fourier Analysis and Applications: Volume 18, Issue 1 (2012), Page 67-86

Kernel Approximation on Manifolds II: The $L_{\infty}$-norm of the $L_2$-projector

Authors: Thomas Hangelbroek, Fran J Narcowich, Xingping Sun, Joe D Ward

Abstract: This article addresses two topics of significant mathematical and practical interest in the theory of kernel approximation: the existence of local and stable bases and the L_p--boundedness of the least squares operator. The latter is an analogue of the classical problem in univariate spline theory, known there as the "de Boor conjecture". A corollary of this work is that for appropriate kernels th… ▽ More This article addresses two topics of significant mathematical and practical interest in the theory of kernel approximation: the existence of local and stable bases and the L_p--boundedness of the least squares operator. The latter is an analogue of the classical problem in univariate spline theory, known there as the "de Boor conjecture". A corollary of this work is that for appropriate kernels the least squares projector provides universal near-best approximations for functions f\in L_p, 1\le p\le \infty. △ Less

Submitted 16 December, 2010; v1 submitted 13 May, 2010; originally announced May 2010.

Comments: 25 pages; minor revision; new proof of Lemma 3.9; accepted for publication in SIAM J. on Math. Anal

MSC Class: 41A05; 41A63; 46E22; 46E35

Journal ref: SIAM J. Math. Anal. 43, pp. 662-684 (2011)

Kernel Approximation on Manifolds I: Bounding the Lebesgue Constant

Authors: Thomas Hangelbroek, Fran J. Narcowich, Joe D. Ward

Abstract: The purpose of this paper is to establish that for any compact, connected C^{\infty} Riemannian manifold there exists a robust family of kernels of increasing smoothness that are well suited for interpolation. They generate Lagrange functions that are uniformly bounded and decay away from their center at an exponential rate. An immediate corollary is that the corresponding Lebesgue constant will… ▽ More The purpose of this paper is to establish that for any compact, connected C^{\infty} Riemannian manifold there exists a robust family of kernels of increasing smoothness that are well suited for interpolation. They generate Lagrange functions that are uniformly bounded and decay away from their center at an exponential rate. An immediate corollary is that the corresponding Lebesgue constant will be uniformly bounded with a constant whose only dependence on the set of data sites is reflected in the mesh ratio, which measures the uniformity of the data. The analysis needed for these results was inspired by some fundamental work of Matveev where the Sobolev decay of Lagrange functions associated with certain kernels on Ω\subset R^d was obtained. With a bit more work, one establishes the following: Lebesgue constants associated with surface splines and Sobolev splines are uniformly bounded on R^d provided the data sites Ξare quasi-uniformly distributed. The non-Euclidean case is more involved as the geometry of the underlying surface comes into play. In addition to establishing bounded Lebesgue constants in this setting, a "zeros lemma" for compact Riemannian manifolds is established. △ Less

Submitted 13 May, 2010; v1 submitted 31 August, 2009; originally announced September 2009.

Comments: 33 pages, 2 figures, new title, accepted for publication in SIAM J. on Math. Anal

MSC Class: 41A05; 41A63; 46E22; 46E35

Journal ref: SIAM J. Math. Anal. Volume 42, Issue 4, pp. 1732-1760 (2010)

L^p Bernstein estimates and approximation by spherical basis functions

Authors: H. N. Mhaskar, F. J. Narcowich, J. Prestin, J. D. Ward

Abstract: The purpose of this paper is to establish L^p error estimates, a Bernstein inequality, and inverse theorems for approximation by a space comprising spherical basis functions located at scattered sites on the unit n-sphere. In particular, the Bernstein inequality estimates L^p Bessel-potential Sobolev norms of functions in this space in terms of the minimal separation and the L^p norm of the func… ▽ More The purpose of this paper is to establish L^p error estimates, a Bernstein inequality, and inverse theorems for approximation by a space comprising spherical basis functions located at scattered sites on the unit n-sphere. In particular, the Bernstein inequality estimates L^p Bessel-potential Sobolev norms of functions in this space in terms of the minimal separation and the L^p norm of the function itself. An important step in its proof involves measuring the L^p stability of functions in the approximating space in terms of the l^p norm of the coefficients involved. As an application of the Bernstein inequality, we derive inverse theorems for SBF approximation in the L^P norm. Finally, we give a new characterization of Besov spaces on the n-sphere in terms of spaces of SBFs. △ Less

Submitted 28 October, 2008; originally announced October 2008.

MSC Class: 41A17; 41A27; 41A63; 42C15