Samplet compression for conditionally positive definite kernels and universal Kriging

DOI10.48550/arXiv.2608.24283MaRDI QIDQ7284988zbMATH OpenarXivFDO

Authors Sara Avesani, Rüdiger Kempf, Michael Peters, Holger Wendland, Michael Multerer

Publication date 25 August 2026

Published in arXiv

Tldr We introduce a fast numerical method for conditionally positive definite (CPD) kernel approximation using samplets. By exploiting the vanishing moment property of samplets and their associated scaling distributions, we reduce an indefinite saddle-point system to a linear and triangular system that can be solved efficiently by sparse Cholesky factorization with O(N log N) cost.

Full work available at URL https://doi.org/10.48550/arXiv.2608.24283

arXiv classification

math.NA

Cites work

• A new variable shape parameter strategy for RBF approximation using neural networks
• A review of image-warping methods
• An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach
• Construction of generalized samplets in Banach spaces
• Construction of quasi-localized dual bases in reproducing kernel Hilbert spaces
• Convergence of the Smoothed Particle Hydrodynamics Method for a Specific Barotropic Fluid Flow: Constructive Kernel Theory
• Covariance Tapering for Interpolation of Large Spatial Datasets
• Extension of sampling inequalities to Sobolev semi-norms of fractional order and derivative data
• Fast evaluation of polyharmonic splines in three dimensions
• Interpolation of scattered data: distance matrices and conditionally positive definite functions
• Kernel techniques: From machine learning to meshless methods
• Model-Based Geostatistics
• Multiresolution local smoothness detection in non-uniformly sampled multivariate signals
• Multiscale scattered data analysis in samplet coordinates
• Numerical solution of saddle point problems by block {Gram--Schmidt} orthogonalization
• Samplet basis pursuit: multiresolution scattered data approximation with sparsity constraints
• Samplet limits and multiwavelets
• Samplets: construction and scattered data compression
• Samplets: wavelet concepts for scattered data
• Scattered Data Approximation
• Statistics for spatial data
• Sur l’erreur d’interpolation des fonctions de plusieurs variables par les D^m-splines
• The intrinsic random functions and their applications
• Theory and applications of the multiquadric-biharmonic method. 20 years of discovery 1968-1988
• Vector-Valued Gaussian Processes for Approximating Divergence- or Rotation-free Vector Fields

This page was built for publication: Samplet compression for conditionally positive definite kernels and universal Kriging

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284988)