Trust Region Constrained Bayesian Optimization with Penalized Constraint Handling

Raju Chowdhury, Tanmay Sen, Prajamitra Bhuyan, Biswabrata Pradhan · 2026

Read Trust Region Constrained Bayesian Optimization with Penalized Constraint Handling by Raju Chowdhury, Tanmay Sen, Prajamitra Bhuyan, Biswabrata Pradhan (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-2603-24567.txt.

Description

Chowdhury and co-authors propose a trust region method for constrained Bayesian optimisation in high-dimensional problems. Such problems are hard: costly evaluations, no gradients, and complex feasible regions. Their method converts constraints into a penalty term, giving one objective to model with a surrogate. A trust region limits the search near the current best point, which aids stability in high dimensions. Expected improvement picks each new evaluation point. On test problems, it finds high-quality feasible solutions with fewer evaluations than top rivals.

Catalog record

CREATORRaju Chowdhury, Tanmay Sen, Prajamitra Bhuyan, Biswabrata Pradhan
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27579
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.24567

Browse the full library · About the Tower of Babel