Final published version
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
}
TY - GEN
T1 - {PF}2 ES: Parallel Feasible Pareto Frontier Entropy Search for Multi-Objective Bayesian Optimization
AU - Qing, Jixiang
AU - Moss, Henry B
AU - Dhaene, Tom
AU - Couckuyt, Ivo
PY - 2023/4/25
Y1 - 2023/4/25
N2 - We present Parallel Feasible Pareto Frontier Entropy Search ({PF}2ES) — a novel information-theoretic acquisition function for multi-objective Bayesian optimization supporting unknown constraints and batch queries. Due to the complexity of characterizing the mutual information between candidate evaluations and (feasible) Pareto frontiers, existing approaches must either employ crude approximations that significantly hamper their performance or rely on expensive inference schemes that substantially increase the optimization’s computational overhead. By instead using a variational lower bound, {PF}2ES provides a low-cost and accurate estimate of the mutual information. We benchmark {PF}2ES against other information-theoretic acquisition functions, demonstrating its competitive performance for optimization across synthetic and real-world design problems.
AB - We present Parallel Feasible Pareto Frontier Entropy Search ({PF}2ES) — a novel information-theoretic acquisition function for multi-objective Bayesian optimization supporting unknown constraints and batch queries. Due to the complexity of characterizing the mutual information between candidate evaluations and (feasible) Pareto frontiers, existing approaches must either employ crude approximations that significantly hamper their performance or rely on expensive inference schemes that substantially increase the optimization’s computational overhead. By instead using a variational lower bound, {PF}2ES provides a low-cost and accurate estimate of the mutual information. We benchmark {PF}2ES against other information-theoretic acquisition functions, demonstrating its competitive performance for optimization across synthetic and real-world design problems.
M3 - Conference contribution/Paper
VL - 206
T3 - Proceedings of Machine Learning Research
SP - 2565
EP - 2588
BT - International Conference on Artificial Intelligence and Statistics
A2 - Ruiz, Francisco
A2 - Dy, Jennifer
A2 - van de Meent, Jan-Willem
PB - PMLR
ER -