Bayesian Optimization for Intrinsically Noisy Response Surfaces

Anton van Beek · 2025

Read Bayesian Optimization for Intrinsically Noisy Response Surfaces by Anton van Beek (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00327.txt.

Description

Anton van Beek studied how well Bayesian optimization works when the thing being optimized is a real physical experiment with unavoidable noise. The method is popular for computer simulations, where it picks each new test smartly to find the best design quickly. Physical experiments, though, are noisy and budgets are tight, so users must make simplifying choices whose effects are poorly understood.

Van Beek ran controlled tests varying the sample count, the rule for choosing the next test, the correlation model, and the amount and shape of the noise. One correlation model, the Matérn, performed best across all test problems. The size of the experiment budget mattered most when picking other settings, and the results give experimenters practical guidance for noisy work.

Catalog record

CREATORAnton van Beek
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-00327

Browse the full library · About the Tower of Babel