Bayesian Optimisation: Which Constraints Matter?

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lin, Xietao Wang, Ungredda, Juan, Butler, Max, Town, James, Rahat, Alma, Singh, Hemant, Branke, Juergen
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908722534547456
author Lin, Xietao Wang
Ungredda, Juan
Butler, Max
Town, James
Rahat, Alma
Singh, Hemant
Branke, Juergen
author_facet Lin, Xietao Wang
Ungredda, Juan
Butler, Max
Town, James
Rahat, Alma
Singh, Hemant
Branke, Juergen
contents Bayesian optimisation has proven to be a powerful tool for expensive global black-box optimisation problems. In this paper, we propose new Bayesian optimisation variants of the popular Knowledge Gradient acquisition functions for problems with \emph{decoupled} black-box constraints, in which subsets of the objective and constraint functions may be evaluated independently. In particular, our methods aim to take into account that often only a handful of the constraints may be binding at the optimum, and hence we should evaluate only relevant constraints when trying to optimise a function. We empirically benchmark these methods against existing methods and demonstrate their superiority over the state-of-the-art.
format Preprint
id arxiv_https___arxiv_org_abs_2512_17569
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Bayesian Optimisation: Which Constraints Matter?
Lin, Xietao Wang
Ungredda, Juan
Butler, Max
Town, James
Rahat, Alma
Singh, Hemant
Branke, Juergen
Machine Learning
Optimization and Control
Bayesian optimisation has proven to be a powerful tool for expensive global black-box optimisation problems. In this paper, we propose new Bayesian optimisation variants of the popular Knowledge Gradient acquisition functions for problems with \emph{decoupled} black-box constraints, in which subsets of the objective and constraint functions may be evaluated independently. In particular, our methods aim to take into account that often only a handful of the constraints may be binding at the optimum, and hence we should evaluate only relevant constraints when trying to optimise a function. We empirically benchmark these methods against existing methods and demonstrate their superiority over the state-of-the-art.
title Bayesian Optimisation: Which Constraints Matter?
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2512.17569