Multi-Step Budgeted Bayesian Optimization with Unknown Evaluation Costs

TL;DR

This paper introduces a novel Bayesian optimization method that accounts for unknown and heterogeneous evaluation costs under budget constraints, improving decision-making in costly black-box function optimization.

Contribution

It proposes the budgeted multi-step expected improvement, a new acquisition function that effectively balances exploration and exploitation with unknown costs and budget limits.

Findings

Our method outperforms existing approaches on synthetic problems.

It effectively manages unknown evaluation costs in real-world scenarios.

The approach improves optimization efficiency under budget constraints.

Abstract

Bayesian optimization (BO) is a sample-efficient approach to optimizing costly-to-evaluate black-box functions. Most BO methods ignore how evaluation costs may vary over the optimization domain. However, these costs can be highly heterogeneous and are often unknown in advance. This occurs in many practical settings, such as hyperparameter tuning of machine learning algorithms or physics-based simulation optimization. Moreover, those few existing methods that acknowledge cost heterogeneity do not naturally accommodate a budget constraint on the total evaluation cost. This combination of unknown costs and a budget constraint introduces a new dimension to the exploration-exploitation trade-off, where learning about the cost incurs the cost itself. Existing methods do not reason about the various trade-offs of this problem in a principled way, leading often to poor performance. We formalize this claim by proving that the expected improvement and the expected improvement per unit of cost, arguably the two most widely used acquisition functions in practice, can be arbitrarily inferior with respect to the optimal non-myopic policy. To overcome the shortcomings of existing approaches, we propose the budgeted multi-step expected improvement, a non-myopic acquisition function that generalizes classical expected improvement to the setting of heterogeneous and unknown evaluation costs. Finally, we show that our acquisition function outperforms existing methods in a variety of synthetic and real problems.