Bayesian Optimization for Partially Known Systems using Hybrid Models

Eike Cramer · 2026

Read Bayesian Optimization for Partially Known Systems using Hybrid Models by Eike Cramer (2026) 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-2603-11199.txt.

Description

Eike Cramer and coauthors speed up Bayesian optimization for partially known systems by fusing its learned model with known physics equations. Standard Bayesian optimization learns a direct input-to-objective map and can need prohibitively many expensive trials, especially in high-dimensional nonlinear problems. Their hybrid writes out the known mechanistic equations and infers the missing pieces with a Gaussian process, entering the process as a constraint. On an in-silico distillation optimization it beats the standard loop clearly and sometimes converges in a single iteration.

Catalog record

CREATOREike Cramer
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE59166
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.11199

Browse the full library · About the Tower of Babel