A robust multi-objective Bayesian optimization framework considering input uncertainty

Bayesian optimization is a popular tool for optimizing time-consuming objective functions with limited number of function evaluations. In real-life applications like engineering design, the designer often wants to take multiple objectives as well input uncertainty into account find set robust solutions. While this an active topic in single-objective optimization, it less investigated multi-objective case. We introduce novel framework perform considering uncertainty. propose Gaussian Process model infer Bayes risk criterion quantify robustness, and we develop two-stage process search Pareto frontier, i.e., solutions that have good average performance under The complete supports various distributions takes full advantage parallel computing. demonstrate effectiveness through numerical benchmarks.