Parallel Bayesian Optimization of Multiple Noisy Objectives with Expected Hypervolume Improvement

Neural Information Processing Systems 

Bayesian optimization (MOBO) is a sample-efficient approach for identifying the optimal trade-offs between the objectives.