Rate-Preserving Shrinking-Support Gaussian Process Prediction
Abstract
Gaussian process prediction is a central tool in spatial statistics, but standard implementations require dense matrix operations that become prohibitive for large datasets. We propose a scale-adjusted compactly supported working correlation for Gaussian process prediction, using the generalized Wendland family with a support radius ϕn that is allowed to decrease with the sample size n. Under fixed-domain asymptotics with quasi-uniform designs, a fixed support radius does not yield asymptotic sparsity, whereas a shrinking support radius can make the covariance matrix sparse. We show that ϕn and the regularization parameter can be jointly chosen so that the resulting predictor preserves the optimal integrated mean squared prediction error rate while reducing the number of nonzero covariance entries and the associated sparse matrix-vector multiplication cost. We also establish an analogous rate-preserving sparsification result for kernel ridge regression. Simulations and an ERA5 temperature application show that the proposed method achieves competitive prediction accuracy while retaining the computational efficiency provided by sparse linear algebra.
Related
- Gaussian Process Decorrelation for Spatiotemporal Deep Learning-Based Snow Water Equivalent Prediction
- GpFlash: Maximizing Speed and Minimizing Memory for Gaussian Process Training and Prediction
- A support preserving homotopy for the de Rham complex with boundary decay estimates
- Testing Gaussian process with applications to super-resolution
- A Gaussian Process Regression-Enhanced Two-Sweep Parameterized Power Method for Computing Extreme Eigenpairs
BibTeX
@misc{xiang2026ratepreserving,
title = {{Rate-Preserving Shrinking-Support Gaussian Process Prediction}},
author = {Xiaopeng Xiang and Wenlin Dai and Marc G. Genton and Wenjia Wang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.31284},
eprint = {2609.31284},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.31284},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Rate-Preserving Shrinking-Support Gaussian Process Prediction
AU - Xiaopeng Xiang
AU - Wenlin Dai
AU - Marc G. Genton
AU - Wenjia Wang
PY - 2026
DA - 2026/09/25
PB - arXiv
DO - 10.48550/arxiv.2609.31284
UR - https://arxiv.org/abs/2609.31284
AB - Gaussian process prediction is a central tool in spatial statistics, but standard implementations require dense matrix operations that become prohibitive for large datasets. We propose a scale-adjusted compactly supported working correlation for Gaussian process prediction, using the generalized Wendland family with a support radius ϕ_n that is allowed to decrease with the sample size n. Under fixed-domain asymptotics with quasi-uniform designs, a fixed support radius does not yield asymptotic sparsity, whereas a shrinking support radius can make the covariance matrix sparse. We show that ϕ_n and the regularization parameter can be jointly chosen so that the resulting predictor preserves the optimal integrated mean squared prediction error rate while reducing the number of nonzero covariance entries and the associated sparse matrix-vector multiplication cost. We also establish an analogous rate-preserving sparsification result for kernel ridge regression. Simulations and an ERA5 temperature application show that the proposed method achieves competitive prediction accuracy while retaining the computational efficiency provided by sparse linear algebra.
ER -