Minimum Regret Search for Single- and Multi-Task Optimization

Minimum Regret Search for Single- and Multi-Task Optimization
复制标题

单任务和多任务优化的最小遗憾搜索

DOI:
--
复制
发表时间:
2016
期刊:
International Conference on Machine Learning
影响因子:
--
通讯作者:
J. H. Metzen
J. H. Metzen
中科院分区:
--
文献类型:
--
作者:
J. H. Metzen

文献摘要

被引文献

相似文献

我们提出了最小遗憾搜索(MRS),贝叶斯优化的一种新的采集功能。MRS与熵搜索(ES)等信息理论方法有相似之处。然而,虽然ES的目的是在每个查询在最大限度地提高信息增益方面的全球最大值,MRS的目的是在最小化的预期简单的遗憾,其最终建议的最佳。虽然经验ES和MRS在大多数情况下执行类似,但MRS比ES产生更少的具有高简单遗憾的离群值。我们提供了一个合成的单任务优化问题,以及一个模拟的多任务机器人控制问题的实证结果。
We propose minimum regret search (MRS), a novel acquisition function for Bayesian optimization. MRS bears similarities with information-theoretic approaches such as entropy search (ES). However, while ES aims in each query at maximizing the information gain with respect to the global maximum, MRS aims at minimizing the expected simple regret of its ultimate recommendation for the optimum. While empirically ES and MRS perform similar in most of the cases, MRS produces fewer outliers with high simple regret than ES. We provide empirical results both for a synthetic single-task optimization problem as well as for a simulated multi-task robotic control problem.