Timezone: »
In many real-world scenarios, decision makers seek to efficiently optimize multiple competing objectives in a sample-efficient fashion. Multi-objective Bayesian optimization (BO) is a common approach, but many of the best-performing acquisition functions do not have known analytic gradients and suffer from high computational overhead. We leverage recent advances in programming models and hardware acceleration for multi-objective BO using Expected Hypervolume Improvement (EHVI)---an algorithm notorious for its high computational complexity. We derive a novel formulation of q-Expected Hypervolume Improvement (qEHVI), an acquisition function that extends EHVI to the parallel, constrained evaluation setting. qEHVI is an exact computation of the joint EHVI of q new candidate points (up to Monte-Carlo (MC) integration error). Whereas previous EHVI formulations rely on gradient-free acquisition optimization or approximated gradients, we compute exact gradients of the MC estimator via auto-differentiation, thereby enabling efficient and effective optimization using first-order and quasi-second-order methods. Our empirical evaluation demonstrates that qEHVI is computationally tractable in many practical scenarios and outperforms state-of-the-art multi-objective BO algorithms at a fraction of their wall time.
Author Information
Samuel Daulton (Facebook)
Research Scientist at Meta, PhD Candidate at Oxford. My research focuses on Bayesian optimization.
Maximilian Balandat (Facebook)
Eytan Bakshy (Facebook)
More from the Same Authors
-
2021 : Practical Policy Optimization with PersonalizedExperimentation »
Mia Garrard · Hanson Wang · Ben Letham · Zehui Wang · Yin Huang · Yichun Hu · Chad Zhou · Norm Zhou · Eytan Bakshy -
2021 : Optimizing High-Dimensional Physics Simulations via Composite Bayesian Optimization »
Wesley Maddox · Qing Feng · Maximilian Balandat -
2021 Poster: Multi-Step Budgeted Bayesian Optimization with Unknown Evaluation Costs »
Raul Astudillo · Daniel Jiang · Maximilian Balandat · Eytan Bakshy · Peter Frazier -
2021 Poster: Parallel Bayesian Optimization of Multiple Noisy Objectives with Expected Hypervolume Improvement »
Samuel Daulton · Maximilian Balandat · Eytan Bakshy -
2021 Poster: Bayesian Optimization with High-Dimensional Outputs »
Wesley Maddox · Maximilian Balandat · Andrew Wilson · Eytan Bakshy -
2020 : Contributed Talk 7: Distilled Thompson Sampling: Practical and Efficient Thompson Sampling via Imitation Learning »
Samuel Daulton · Hongseok Namkoong -
2020 Poster: BoTorch: A Framework for Efficient Monte-Carlo Bayesian Optimization »
Maximilian Balandat · Brian Karrer · Daniel Jiang · Samuel Daulton · Ben Letham · Andrew Wilson · Eytan Bakshy -
2020 Poster: Re-Examining Linear Embeddings for High-Dimensional Bayesian Optimization »
Ben Letham · Roberto Calandra · Akshara Rai · Eytan Bakshy -
2020 Poster: Efficient Nonmyopic Bayesian Optimization via One-Shot Multi-Step Trees »
Shali Jiang · Daniel Jiang · Maximilian Balandat · Brian Karrer · Jacob Gardner · Roman Garnett -
2020 Poster: High-Dimensional Contextual Policy Search with Unknown Context Rewards using Bayesian Optimization »
Qing Feng · Ben Letham · Hongzi Mao · Eytan Bakshy -
2020 Spotlight: High-Dimensional Contextual Policy Search with Unknown Context Rewards using Bayesian Optimization »
Qing Feng · Ben Letham · Hongzi Mao · Eytan Bakshy -
2019 : Invited Speaker: Eytan Bakshy »
Eytan Bakshy -
2017 Poster: Robust and Efficient Transfer Learning with Hidden Parameter Markov Decision Processes »
Taylor Killian · Samuel Daulton · Finale Doshi-Velez · George Konidaris -
2017 Oral: Robust and Efficient Transfer Learning with Hidden Parameter Markov Decision Processes »
Taylor Killian · Samuel Daulton · Finale Doshi-Velez · George Konidaris -
2016 Poster: Minimizing Regret on Reflexive Banach Spaces and Nash Equilibria in Continuous Zero-Sum Games »
Maximilian Balandat · Walid Krichene · Claire Tomlin · Alexandre Bayen