Approximation with rational and rescaled kernels: new steps forward
DOI:
https://doi.org/10.64700/mmm.119Keywords:
Radial basis function (RBF), variably scaled kernel, rational approximation, rescaled localized RBFAbstract
The aim of this paper is to present new insights into Rational Radial Basis Function (RRBF) approximations and their localized rescaled counterparts. After some necessary notation, we briefly recall the essence of stable computational techniques for overcoming the ill-conditioning of the kernel matrices, as well as the ones that explore stable bases. RRBFs are then introduced, and their benefits are described. We discuss Variably Scaled Kernels, which provide a more flexible tool for approximating with RBFs, substituting the shape parameter with a scaling function. This scaling function is then applied to the RRBFs to obtain a new family of RBFs that are more appropriate for discontinuous functions. Finally, the localized version of the RRBFs is recalled, for which we provide new insights into
an open conjecture regarding the sum of the cardinal function associated with the approximant in rescaled form.
References
[1] G. Audone, F. Della Santa, E. Perracchione and S. Pieraccini: A recipe for learning variably scaled kernels via discontinuous neural networks, J. Comput. Appl. Math., 469 (2025), 1–18.
[2] V. Bayona, N. Flyer, B. Fornberg and G. A. Barnett: On the role of polynomials in RBF-FD approximations: II. numerical solution of elliptic PDEs, J. Comput. Phys., 332 (2017), 257–273.
[3] F. Bernal, A. Safdari-Vaighani and E. Larsson: A radial basis function partition of unity method for steady flow simulations, J. Comput. Phys., 503 (2024), Article ID:112842.
[4] M. Bozzini, L. Lenarduzzi, M. Rossini and R. Schaback: Interpolation with variably scaled kernels, IMA J. Numer. Anal., 35 (2015), 199–219.
[5] C. Bracco, O. Davydov, C. Giannelli and A. Sestini: A positive meshless finite difference scheme for scalar conservation laws with adaptive artificial viscosity driven by fault detection, Comput. Math. Appl., 190 (2025), 103–121.
[6] M. D. Buhmann, S. De Marchi and E. Perracchione: Analysis of a new class of rational RBF expansions, IMA J. Numer. Anal., 40 (3) (2020), 1972–1993.
[7] G. Cappellazzo: Fast decaying polynomial reproduction, Master’s thesis, University of Padova, Padua (2023).
[8] R. Cavoretto, G. E. Fasshauer and M. McCourt: An introduction to the Hilbert-Schmidt SVD using iterated Brownian bridge kernels, Numer. Algorithms, 68 (2015), 393–422.
[9] C. S. Chen, A. Karageorghis and H. Zheng: Improved RBF collocation methods for fourth order boundary value problems, Commun. Comput. Phys., 27 (2020), 1530–1549.
[10] M. Chen, L. Ling and D. Yun: Proving the stability estimates of variational least-squares kernel-based methods, Comput. Math. Appl., 180 (2025), 46–60.
[11] O. Davydov, R. Schaback: Error bounds for kernel-based numerical differentiation, Numer. Math., 132 (2016), 243–269.
[12] O. Davydov, R. Schaback: Optimal stencils in Sobolev spaces, IMA J. Numer. Anal., 39 (2019), 398–422.
[13] S. De Marchi: On optimal center locations for radial basis function interpolation: computational aspects, Rend. Semin. Mat., Univ. Politec. Torino, 61 (2003), 343–358.
[14] S. De Marchi, G. Santin: A new stable basis for radial basis function interpolation, J. Comput. Appl. Math., 253 (2013), 1–13.
[15] S. De Marchi, A. Idda and G. Santin: A rescaled method for RBF approximation, Proceedings of Approximation Theory XV, Springer Proceedings on Mathematics and Statistics, 201, San Antonio, Texas (USA) (2017), 39–59.
[16] S. De Marchi, A. Martínez and E. Perracchione: Fast and stable rational RBF-based partition of unity interpolation, J. Comput. Appl. Math., 349 (2019), 331–343.
[17] S. De Marchi, A. Martínez, E. Perracchione and M. Rossini: RBF-based partition of unity methods for elliptic PDEs: Adaptivity and stability issues via variably scaled kernels, J. Sci. Comput., 79 (2019), 321–344.
[18] S. De Marchi, H. Wendland: On the Convergence of the rescaled localized radial basis function method, Appl. Math. Letters, 99 (2020), Article ID:105996.
[19] S. Deparis, D. Forti and A. Quarteroni: A rescaled localized radial basis function interpolation on non-cartesian and non-conforming grids, SIAM J. Sci. Comp., 36 (6) (2014), A2745–A2762.
[20] T. A. Driscoll, B. Fornberg: Interpolation in the limit of increasingly flat radial basis functions, Comput. Math. Appl., 43 (2002), 413–422.
[21] E. Farazandeh, D. Mirzaei: A rational RBF interpolation with conditionally positive definite kernels, Adv. Comput. Math., 47 (2021), Article ID:74.
[22] G. E. Fasshauer: Meshfree approximations methods with Matlab, World Scientific, Singapore (2007).
[23] G. E. Fasshauer, M. J. McCourt: Stable evaluation of Gaussian radial basis function interpolants, SIAM J. Sci. Comput., 34 (2012), A737–A762.
[24] G. E. Fasshauer, M. J. McCourt: Kernel-based approximation methods using Matlab,World Scientific, Singapore (2015).
[25] N. Flyer, E. Lehto, S. Blaise, G. B. Wright and A. St-Cyr: A guide to RBF-generated finite differences for nonlinear transport: shallow water simulations on a sphere, J. Comput. Phys., 231 (2012), 4078–4095.
[26] B. Fornberg, G. Wright: Stable computation of multiquadric interpolants for all values of the shape parameter, Comput. Math. Appl., 48 (2004), pp. 853-867.
[27] B. Fornberg, C. Piret: A stable algorithm for flat radial basis functions on a sphere, SIAM J. Sci. Comput., 30 (2007), 60–80.
[28] B. Fornberg, E. Larsson and N. Flyer: Stable computations with Gaussian radial basis functions, SIAM J. Sci. Comput., 33 (2011), 869–892.
[29] T. Hangelbroek, F. J. Narcowich, C. Rieger and J. D. Ward: An inverse theorem for compact Lipschitz regions in (R^d) using localized kernel bases, Math. Comp., 87 (2018), 1949–1989.
[30] H. Harbrecht, M. Multerer: Samplets: Construction and scattered data compression, J. Comput. Physics, 471 (2022), Article ID:111616.
[31] T. Hofmann, B. Schölkopf and A. J. Smola: Kernel methods in machine learning, Ann. Statist., 36 (3) (2008), 1171–1220.
[32] X. G. Hu, T. S. Ho and H. Rabitz: Rational approximation with multidimensional scattered data, Phys. Rev., 65 (2002), Article ID:035701.
[33] S. Jakobsson, B. Andersson and F. Edelvik: Rational radial basis function interpolation with applications to antenna design, J. Comput. Appl. Math., 233 (2009), 889–904.
[34] E. J. Kansa: Application of Hardy’s multiquadric interpolation to hydrodynamics, in: Proc. 1986 Simul. Conf., 4 (1986), 111–117.
[35] K. Kormann, C. Lasser and A. Yurova: Stable interpolation with isotropic and anisotropic Gaussians using Hermite generating function, SIAM J. Sci. Comput., 41 (2019), A3839–A3859.
[36] E. Larsson, B. Fornberg: Theoretical and computational aspects of multivariate interpolation with increasingly flat radial basis functions, Comput. Math. Appl., 49 (2005), 103–130.
[37] E. Larsson, E. Lehto, A. R. H. Heryudono and B. Fornberg: Stable computation of differentiation matrices and scattered node stencils based on Gaussian radial basis functions, SIAM J. Sci. Comput., 33 (2013), 869–892.
[38] E. Larsson, V. Shcherbakov and A. Heryudono: A least squares radial basis function partition of unity method for solving PDEs, SIAM J. Sci. Comput., 39 (2017), A2538–A2563.
[39] E. Larsson, R. Schaback: Scaling of radial basis functions, IMA J. Numer. Anal., 44 (2024), 1130–1152.
[40] E. Larsson: On the optimal shape parameter of Gaussian radial basis function interpolation, manuscript in preparation, (2026).
[41] X. Li, M. Chen, Z. Sun, L. Ling and S. Li: Energy-conserving Kansa methods for Hamiltonian wave equations, Appl. Math. Comput., 510 (2026), Article ID:129682.
[42] L. Ling, R. Opfer and R. Schaback: Results on meshless collocation techniques, Eng. Anal. Bound. Elem., 30 (2006), 247–253.
[43] M. J. L. Orr: Introduction to radial basis function networks, Tech. rep., Centre for Cognitive Sciences, University of Edinburgh, Edinburgh (1996).
[44] B. Raeisi, M. Fardi and M. A. Darani: RBF-based partition of unity methods for two-dimensional time-dependent PDEs: Numerical and theoretical aspects, Math. Comput. Simulation, 226 (2024), 152–171.
[45] C. Rieger, H. Wendland: Constructive approximation of high-dimensional functions with small efficient dimension with applications in uncertainty quantification, SIAM J. Numer. Anal., 64 (2026), 1–28.
[46] A. Safdari-Vaighani, A. Heryudono and E. Larsson: A radial basis function partition of unity collocation method for convection-diffusion equations arising in financial applications, J. Sci. Comput., 64 (2015), 341–367.
[47] S. Sarra: A rational radial basis function method for accurately resolving discontinuities and steep gradients, Appl. Math. Comput., 160 (2018), 131–142.
[48] S. Sarra: Local radial basis function methods: Comparison, improvements and implementation, J. Appl. Math. Phys., 11 (2023), 3867–3886.
[49] R. Schaback: Lower bound for norms of inverses of interpolation matrices for radial basis functions, J. Approx. Theory, 79 (1994), 287–306.
[50] R. Schaback: Error estimates and condition numbers for radial basis function interpolation, Adv. Comput. Math., 3 (1995), 251–264.
[51] R. Schaback: Native Hilbert spaces for radial basis functions. I., In: New developments in approximation theory, 132, Birkhäuser, Basel (1999), 255–282.
[52] R. Schaback: A unified theory of radial basis functions: Native Hilbert spaces for radial basis functions II, J. Comput. Appl. Math., 121 (2000), 165–177.
[53] R. Schaback: Multivariate interpolation by polynomials and radial basis functions, Constr. Approx., 21 (2005), 293–317.
[54] R. Schaback: Convergence of unsymmetric kernel-based meshless collocation methods, SIAM J. Numer. Anal., 45 (2007), 333–351.
[55] R. Schaback: All well-posed problems have uniformly stable and convergent discretizations, Numer. Math., 132 (2016), 597–630.
[56] B. Schölkopf, A. J. Smola: Learning with kernels: Support vector machines, regularization, optimization, and beyond, MIT Press, Cambridge (2002).
[57] V. Shcherbakov, E. Larsson: Radial basis function partition of unity methods for pricing vanilla basket options, Comput. Math. Appl., 71 (2016), 185–200.
[58] J. Slak, G. Kosec: On generation of node distributions for meshless PDE discretizations, SIAM J. Sci. Comput., 41 (2019), A3202–A3229.
[59] J. Slak, B. Stojanoviˇc and G. Kosec: High order RBF-FD approximations with application to a scattering problem, 4th International Conference on Smart and Sustainable Technologies (SpliTech 2019), Split & Bol. (Croatia) (2019).
[60] I. Tominec, E. Larsson, and A. Heryudono: A least squares radial basis function finite difference method with improved stability properties, SIAM J. Sci. Comput., 43 (2021), A1441–A1471.
[61] I. Tominec, M. Nazarov and E. Larsson: Stability estimates for radial basis function methods applied to linear scalar conservation laws, Appl. Math. Comput., 485 (2025), Article ID:129020.
[62] Z. Wang, M. Chen and J. Chen: Solving multiscale elliptic problems by sparse radial basis function neural networks, J. Comput. Phys., 492 (2023), Article ID:112452.
[63] B.-S. Wang, A. S. L. Chu, L. Ling and W. S. Don: Foliation structures and global flow dynamics of scalar hyperbolic conservation laws on manifolds: I. Geometry-compatible fluxes and numerical validation on sphere and torus, Commun. Nonlinear Sci. Numer. Simul., 152 (2026), Article ID:109415.
[64] H.Wendland: Scattered data approximation, Cambridge Monogr. Appl. Comput. Math., 17, Cambridge Univ. Press, Cambridge (2005).
[65] T. Wenzel, G. Santin and B. Haasdonk: Stability of convergence rates: kernel interpolation on non-Lipschitz domains, IMA J. Numer. Anal., 45 (2025), 777–798.
[66] T. Wenzel: Sharp inverse statements for kernel interpolation, Math. Comp., 95 (2026), 1389–1413.
[67] T. Wenzel, A. Iske: Spectral equivalence of unsymmetric kernel matrices and applications, BIT, 66 (2026), Article ID:14.
[68] T. Wenzel, G. Santin and B. Haasdonk: Analysis of structured deep kernel networks, J. Comput. Appl. Math., 476 (2026), Article ID:116975.
[69] A. Westermann, O. Davydov, A. Sokolov and S. Turek: Stability and accuracy of a meshless finite difference method for the Stokes equations, Internat. J. Numer. Methods Engrg., 126 (2025), Article ID:e70000.
[70] G. B. Wright, B. Fornberg: Stable computations with flat radial basis functions using vector-valued rational approximations, J. Comput. Phys., 331 (2017), 137–156.
[71] Z. Wu, R. Schaback: Local error estimates for radial basis function interpolation of scattered data, IMA J. Numer. Anal.,13 (1993), 13–27.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Stefano De Marchi, Elisabeth Larsson
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
