CSE17 - MS140-1: The Parallel Knowledge Gradient Method for Batch Bayesian Optimization(PDF) 17-SIAM-CSE17-MS140-1_031.pdf