Minibatch Markov Chain Monte Carlo Algorithms for Fitting Gaussian Processes
Bayesian Analysis · 01.03.2026 · 21 (1) · 21-43
DOI: 10.1214/24-ba1459 ↗Kurz gefasst
Fachartikel in der Zeitschrift Bayesian Analysis, erschienen am 01.03.2026. Verfasst von Matthew J. Heaton und 1 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben.
Abstract (Originalsprache)
Gaussian processes (GPs) are a highly flexible, nonparametric statistical model that are commonly used to fit nonlinear relationships or account for correlation between observations. However, the computational load of fitting a Gaussian process is Opn3q making them infeasible for use on large datasets. To make GPs more feasible for large datasets, this research focuses on the use of minibatching to estimate GP parameters. Specifically, we outline both approximate and exact minibatch Markov chain Monte Carlo algorithms that substantially reduce the computation of fitting a GP by only considering small subsets of the data at a time. We demonstrate and compare this methodology using various simulations and real datasets.
Zitieren
APA
Heaton, M. J., & Johnson, J. A. (2026). Minibatch Markov Chain Monte Carlo Algorithms for Fitting Gaussian Processes. Bayesian Analysis, 21(1), 21-43. https://doi.org/10.1214/24-ba1459
Vancouver
Heaton MJ, Johnson JA. Minibatch Markov Chain Monte Carlo Algorithms for Fitting Gaussian Processes. Bayesian Analysis. 2026;21(1):21-43. doi:10.1214/24-ba1459
BibTeX
@article{Heaton2026,
author = {Heaton, Matthew J. and Johnson, Jacob A.},
title = {Minibatch Markov Chain Monte Carlo Algorithms for Fitting Gaussian Processes},
journal = {Bayesian Analysis},
year = {2026},
volume = {21},
number = {1},
pages = {21--43},
doi = {10.1214/24-ba1459}
}
EUVIMED