The Parallel Knowledge Gradient Method for Batch Bayesian Optimization

TL;DR

This paper introduces the parallel knowledge gradient method, a novel batch Bayesian optimization algorithm that efficiently finds global optima faster than previous methods, especially in noisy settings.

Contribution

It develops a new batch Bayesian optimization algorithm that is one-step Bayes-optimal and provides an efficient computation strategy for practical use.

Findings

Faster convergence to global optima compared to previous methods

Effective in noisy evaluation environments

Demonstrated success on synthetic and real machine learning tasks

Abstract

In many applications of black-box optimization, one can evaluate multiple points simultaneously, e.g. when evaluating the performances of several different neural network architectures in a parallel computing environment. In this paper, we develop a novel batch Bayesian optimization algorithm --- the parallel knowledge gradient method. By construction, this method provides the one-step Bayes-optimal batch of points to sample. We provide an efficient strategy for computing this Bayes-optimal batch of points, and we demonstrate that the parallel knowledge gradient method finds global optima significantly faster than previous batch Bayesian optimization algorithms on both synthetic test functions and when tuning hyperparameters of practical machine learning algorithms, especially when function evaluations are noisy.