Bayesian Optimisation: Which Constraints Matter?
This work addresses expensive global optimisation problems with constraints, potentially improving efficiency in domains like engineering design, but it appears incremental as it builds on existing acquisition functions.
The paper tackles the problem of Bayesian optimisation with decoupled black-box constraints by proposing new variants of Knowledge Gradient acquisition functions that focus on evaluating only relevant constraints, and demonstrates their superiority over state-of-the-art methods in empirical benchmarks.
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.