Gearing Gaussian process modeling and sequential design towards stochastic simulators
arXiv:2412.07306
arXiv preprint, December 2024
Abstract
This paper surveys and unifies Gaussian process modeling strategies for stochastic simulators under complex noise settings.
It covers heteroskedastic modeling, non-Gaussian noise, and quantile-oriented approaches, with emphasis on practical inference tradeoffs and data requirements.
The chapter also adapts sequential design procedures to these settings, including treatment of replication, and demonstrates the methodology on an epidemiological example.
Citation
BibTeX
@article{binois2024gearing,
title={Gearing Gaussian process modeling and sequential design towards stochastic simulators},
author={Binois, Mickael and Fadikar, Arindam and Stevens, Abby},
journal={arXiv preprint arXiv:2412.07306},
year={2024}
}