Stratified Bayesian Optimization

TL;DR

Stratified Bayesian Optimization (SBO) is a new method for global optimization of noisy black-box functions that exploits the dependence on influential random inputs to improve efficiency and performance.

Contribution

This paper introduces SBO, a Bayesian optimization algorithm that leverages stratification based on influential random inputs to outperform existing methods.

Findings

SBO outperforms state-of-the-art Bayesian optimization methods.

Exploiting dependence on influential inputs reduces variance and improves optimization.

Numerical experiments validate the effectiveness of SBO.

Abstract

We consider derivative-free black-box global optimization of expensive noisy functions, when most of the randomness in the objective is produced by a few influential scalar random inputs. We present a new Bayesian global optimization algorithm, called Stratified Bayesian Optimization (SBO), which uses this strong dependence to improve performance. Our algorithm is similar in spirit to stratification, a technique from simulation, which uses strong dependence on a categorical representation of the random input to reduce variance. We demonstrate in numerical experiments that SBO outperforms state-of-the-art Bayesian optimization benchmarks that do not leverage this dependence.