arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.11533cs.LG
收藏

MONGOOSE:基于元学习的路径平滑贝叶斯优化

MONGOOSE: Path-wise Smooth Bayesian Optimisation via Meta-learning

University of Bristol(布里斯托尔大学);University of Cambridge(剑桥大学)

Adam X. Yang, Laurence Aitchison, Henry B. Moss

AI总结:

MONGOOSE 针对评估点间移动成本较大的贝叶斯优化问题,利用元学习参数化策略生成平滑优化轨迹,以克服标准短视方法的跳跃路径缺陷并提升性能。

摘要:

在贝叶斯优化中,我们通常寻求最小化现实世界物理系统中出现的黑盒目标函数。评估此类黑盒目标函数的成本的一个主要来源,往往是为测量准备系统所需的努力。我们考虑一种常见场景:准备成本随着连续评估之间距离的增加而增长。在这种设定下,平滑的优化轨迹更受青睐,而标准短视(即一步最优)贝叶斯优化方法产生的跳跃路径是次优的。我们的算法 MONGOOSE 使用元学习的参数化策略来生成平滑的优化轨迹,在优化具有较大移动成本的函数时,相比现有方法取得了性能提升。

英文摘要:

In Bayesian optimisation, we often seek to minimise the black-box objective functions that arise in real-world physical systems. A primary contributor to the cost of evaluating such black-box objective functions is often the effort required to prepare the system for measurement. We consider a common scenario where preparation costs grow as the distance between successive evaluations increases. In this setting, smooth optimisation trajectories are preferred and the jumpy paths produced by the standard myopic (i.e.\ one-step-optimal) Bayesian optimisation methods are sub-optimal. Our algorithm, MONGOOSE, uses a meta-learnt parametric policy to generate smooth optimisation trajectories, achieving performance gains over existing methods when optimising functions with large movement costs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑