Skip to yearly menu bar Skip to main content


The Parallel Knowledge Gradient Method for Batch Bayesian Optimization

Jian Wu · Peter Frazier

Area 5+6+7+8 #146

Keywords: [ Gaussian Processes ] [ (Other) Machine Learning Topics ] [ Stochastic Methods ] [ (Other) Optimization ] [ Bayesian Nonparametrics ] [ (Other) Probabilistic Models and Methods ]


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.

Live content is unavailable. Log in and register to view live content