Wavelet compression for sampling deep Gaussian processes

DOI10.48550/arXiv.2610.20266MaRDI QIDQ7395258zbMATH OpenarXivFDO

Authors Conor Osborne, M. Feischl, Oskar Broukal, Michael Feischl

Publication date 14 October 2026

Published in arXiv

Tldr We develop wavelet compression for deep Gaussian processes and show its convergence rate of exponential error decay when the Matérn kernel and wavelets are sufficiently smooth. We achieve log-linear time complexity in practice by choosing a suitable truncation level.

Full work available at URL https://doi.org/10.48550/arXiv.2610.20266

Mathematics Subject Classification ID

Gaussian processes (60G15) Probabilistic models, generic numerical methods in probability and statistics (65C20) Computational methods for sparse matrices (65F50) Numerical methods for wavelets (65T60)

arXiv classification

math.NA

Cites work

• A tail inequality for quadratic forms of subgaussian random vectors
• A Unifying Perspective on Non-Stationary Kernels for Deeper Gaussian Processes
• An Algorithm for Simulating Stationary Gaussian Random Fields
• Biorthogonal bases of compactly supported wavelets
• Composite Wavelet Bases with Extended Stability and Cancellation Properties
• Compression Techniques for Boundary Integral Equations---Asymptotically Optimal Complexity Estimates
• Computing a nearest symmetric positive semidefinite matrix
• Convergence rates of non-stationary and deep Gaussian process regression
• Deep Gaussian Process Priors for Bayesian Image Reconstruction
• Deep Gaussian Processes
• Diagnostics-driven nonstationary emulators using kernel mixtures
• Fast and Exact Simulation of Stationary Gaussian Processes through Circulant Embedding of the Covariance Matrix
• Fast sampling of parameterised Gaussian random fields
• Fast wavelet transforms and numerical algorithms I
• Gaussian Processes for Machine Learning
• How deep are deep Gaussian processes?
• Multilevel approximation of Gaussian random fields: Covariance compression, estimation and spatial prediction
• Multilevel approximation of Gaussian random fields: fast simulation
• Nonparametric Estimation of Nonstationary Spatial Covariance Structure
• On the Compressibility of Operators in Wavelet Coordinates
• Ten Lectures on Wavelets
• Using Nonorthogonal Lanczos Vectors in the Computation of Matrix Functions
• Wavelets on the interval and fast wavelet transforms

This page was built for publication: Wavelet compression for sampling deep Gaussian processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7395258)