Bayesian Optimisation: Which Constraints Matter?
Xietao Wang Lin, Juan Ungredda, Max Butler, James Town, Alma Rahat, Hemant Singh, Juergen Branke
TL;DR
The paper tackles Bayesian optimization under decoupled, expensive constraints by introducing two acquisition strategies: dcKG, a fully decoupled constrained Knowledge Gradient, and cEI+, a faster variant that uses constrained EI to pick a location and dcKG to choose which function to evaluate there. Across synthetic benchmarks with varying constraint binding and evaluation costs, dcKG consistently outperforms state-of-the-art methods (including cKG and PESC), particularly when constraints are redundant or costs are heterogeneous. The work demonstrates that selectively evaluating only the informative objective/constraints can substantially reduce wasted evaluations and accelerate discovery of the optimum, with practical implications for engineering design and other expensive optimization tasks.
Abstract
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.