Affiliation:
1. University of Exeter, United Kingdom
2. Swansea University, United Kingdom
Abstract
The performance of acquisition functions for Bayesian optimisation to locate the global optimum of continuous functions is investigated in terms of the Pareto front between exploration and exploitation. We show that Expected Improvement (EI) and the Upper Confidence Bound (UCB) always select solutions to be expensively evaluated on the Pareto front, but Probability of Improvement is not guaranteed to do so and Weighted Expected Improvement does so only for a restricted range of weights.
We introduce two novel
-greedy acquisition functions. Extensive empirical evaluation of these together with random search, purely exploratory, and purely exploitative search on 10 benchmark problems in 1 to 10 dimensions shows that
-greedy algorithms are generally at least as effective as conventional acquisition functions (e.g., EI and UCB), particularly with a limited budget. In higher dimensions,
-greedy approaches are shown to have improved performance over conventional approaches. These results are borne out on a real-world computational fluid dynamics optimisation problem and a robotics active learning problem. Our analysis and experiments suggest that the most effective strategy, particularly in higher dimensions, is to be mostly greedy, occasionally selecting a random exploratory solution.
Publisher
Association for Computing Machinery (ACM)
Reference38 articles.
1. BoTorch: A framework for efficient Monte-Carlo Bayesian optimization;Balandat Maximilian;Advances in Neural Information Processing Systems,2020
2. How to Create Generalizable Results
3. A Limited Memory Algorithm for Bound Constrained Optimization
Cited by
47 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Imprecise Bayesian optimization;Knowledge-Based Systems;2024-09
2. Is greed still good in multi-objective Bayesian optimisation?;Proceedings of the Genetic and Evolutionary Computation Conference Companion;2024-07-14
3. Roadmap on data-centric materials science;Modelling and Simulation in Materials Science and Engineering;2024-07-03
4. A Surrogate-Assisted Coevolutionary Algorithm for Expensive Constrained Multiobjective Optimization;2024 IEEE Congress on Evolutionary Computation (CEC);2024-06-30
5. Bayesian Inverse Transfer in Evolutionary Multiobjective Optimization;ACM Transactions on Evolutionary Learning and Optimization;2024-06-28