Goal-Oriented Lower-Tail Calibration of Gaussian Processes for Bayesian Optimization
Abstract
Gaussian process (GP) predictive distributions are commonly used in Bayesian optimization (BO) to guide the selection of evaluation points for expensive objective functions. The choice of kernel and hyperparameters has a strong influence on the exploration--exploitation trade-off. For minimization, sampling criteria such as expected improvement (EI) depend on both the probability mass below the current best value and the shape of the predictive distribution in this region. This article studies goal-oriented calibration of GP predictive distributions below a low threshold t in the noiseless setting, for standard GP models with hyperparameters selected by maximum likelihood. We consider two complementary forms of calibration below t for inputs distributed according to a reference measure μ: occurrence calibration over the design space and thresholded μ-calibration on sublevel sets of the form {x∈𝕏, f(x)≤ t}. We propose tcGP, a post-hoc method that combines these two forms of calibration for GP predictive distributions below t. With fixed GP hyperparameters, the exact EI sampling criterion based on tcGP generates a sequence of evaluation points that is dense in the design space. Experiments on standard benchmarks show improved lower-tail calibration and BO performance relative to standard GP models and globally calibrated GP models.
Related
- Isotropic Gaussian Processes Improve Vanilla Bayesian Optimization in High Dimensions
- Bayesian Estimation of Continuous Brain Connectivity with Log-Gaussian Cox Processes
- Bayesian Emulation of Multi-fidelity Earth System Modelling Using Hierarchical Gaussian Processes
- Fraud Detection via Bayesian Positive-Unlabeled Learning with Gaussian Processes and Multilayer Networks
- ISBO: Scalable Spatio-Temporal Bayesian Optimization with Log Gaussian Cox Process Models via the INLA-SPDE Approach
BibTeX
@misc{pion2026goaloriented,
title = {{Goal-Oriented Lower-Tail Calibration of Gaussian Processes for Bayesian Optimization}},
author = {Aurélien Pion and Emmanuel Vazquez},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2605.20145},
eprint = {2605.20145},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.20145},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Goal-Oriented Lower-Tail Calibration of Gaussian Processes for Bayesian Optimization
AU - Aurélien Pion
AU - Emmanuel Vazquez
PY - 2026
DA - 2026/09/09
PB - arXiv
DO - 10.48550/arxiv.2605.20145
UR - https://arxiv.org/abs/2605.20145
AB - Gaussian process (GP) predictive distributions are commonly used in Bayesian optimization (BO) to guide the selection of evaluation points for expensive objective functions. The choice of kernel and hyperparameters has a strong influence on the exploration--exploitation trade-off. For minimization, sampling criteria such as expected improvement (EI) depend on both the probability mass below the current best value and the shape of the predictive distribution in this region. This article studies goal-oriented calibration of GP predictive distributions below a low threshold t in the noiseless setting, for standard GP models with hyperparameters selected by maximum likelihood. We consider two complementary forms of calibration below t for inputs distributed according to a reference measure μ: occurrence calibration over the design space and thresholded μ-calibration on sublevel sets of the form {x∈𝕏, f(x)≤ t}. We propose tcGP, a post-hoc method that combines these two forms of calibration for GP predictive distributions below t. With fixed GP hyperparameters, the exact EI sampling criterion based on tcGP generates a sequence of evaluation points that is dense in the design space. Experiments on standard benchmarks show improved lower-tail calibration and BO performance relative to standard GP models and globally calibrated GP models.
ER -