Exploration strategies for Gaussian process bandit optimization with inference regret guarantees
Abstract
Existing cumulative regret guarantees for Bayesian optimization (BO) do not necessarily characterize the quality of the algorithm's final output. This gap is critical under noisy evaluations, where the goal is to return a single high-quality recommendation. To address it, we introduce efficient exploration strategies that can be incorporated into any BO procedure by occasionally replacing the acquisition query with an exploration query. To validate this approach, we bound the Gaussian process posterior standard deviation over the search space via the fill distance of the queries, which measures the largest region left unexplored. We then use this result to derive upper bounds on the inference regret, i.e., the gap between the global optimum and the function value at the algorithm's returned point. Empirically, these exploration strategies improve the final recommendations of existing BO algorithms on synthetic benchmarks, hyperparameter optimization, and a control task.
Related
- Risk-Averse Inference-Time Diffusion Model Alignment with Finite-Sample Guarantees
- Neural Estimation of Summary Statistics for Simulation-Based Inference with Non-Asymptotic Guarantees
- Certified Inference and Training for Deep Equilibrium Networks: A Continuation Framework with Polynomial Complexity Guarantees
- A note on the convergence guarantees of RLT-based algorithms for polynomial optimization
- Adaptive Conformal Inference Under Delayed Feedback: Coverage Guarantees and a Delay-to-Memory Diagnostic
BibTeX
@misc{kim2026exploration,
title = {{Exploration strategies for Gaussian process bandit optimization with inference regret guarantees}},
author = {Hwanwoo Kim and Jungwoo Hong and Kyungjae Lee},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.16187},
eprint = {2610.16187},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.16187},
}
RIS
TY - UNPB
TI - Exploration strategies for Gaussian process bandit optimization with inference regret guarantees
AU - Hwanwoo Kim
AU - Jungwoo Hong
AU - Kyungjae Lee
PY - 2026
DA - 2026/10/12
PB - arXiv
DO - 10.48550/arxiv.2610.16187
UR - https://arxiv.org/abs/2610.16187
AB - Existing cumulative regret guarantees for Bayesian optimization (BO) do not necessarily characterize the quality of the algorithm's final output. This gap is critical under noisy evaluations, where the goal is to return a single high-quality recommendation. To address it, we introduce efficient exploration strategies that can be incorporated into any BO procedure by occasionally replacing the acquisition query with an exploration query. To validate this approach, we bound the Gaussian process posterior standard deviation over the search space via the fill distance of the queries, which measures the largest region left unexplored. We then use this result to derive upper bounds on the inference regret, i.e., the gap between the global optimum and the function value at the algorithm's returned point. Empirically, these exploration strategies improve the final recommendations of existing BO algorithms on synthetic benchmarks, hyperparameter optimization, and a control task.
ER -