arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.01632cs.LGstat.ML
收藏

先验未知的时变高斯过程贝叶斯优化问题

Time-Varying Gaussian Process Bandits with Unknown Prior

University of Oxford(牛津大学);Toyota Motor Corporation(丰田汽车公司)

Juliusz Ziomek, Masaki Adachi, Michael A. Osborne

AI总结:

针对时变贝叶斯优化中函数先验未知、现有平稳问题优化方案无法适用的问题,提出PE-GP-UCB算法,推导了遗憾界,实验证明其效果优于多种现有方法。

摘要:

贝叶斯优化需要拟合高斯过程模型,而这又需要为未知的黑盒函数指定先验——绝大多数理论文献都假设该先验是已知的。然而,对于给定的黑盒函数,通常存在多种可能的先验,例如由持有不同观点的领域专家提出的先验。在某些情况下,用于选择先验的II型最大似然估计器具备一致性保证,但它并非普遍适用于所有类型的先验。如果问题是平稳的,我们可以采用Regret Balancing(遗憾平衡)方案来执行优化,但在时变问题的场景下,这类方案无法使用。为弥补现有研究的这一空白,我们提出了一种新算法PE-GP-UCB,它即便在不确切掌握函数先验的情况下,也能够求解时变贝叶斯优化问题。该算法基于如下事实:要么观测到的函数值与部分先验一致,此时很容易排除错误的先验;要么观测结果与所有候选先验都一致,此时我们的模型采用哪种先验都无关紧要。我们为所提出的算法推导了遗憾界。最后,我们在合成的和真实世界的时变问题上对算法进行了实证评估,结果表明其性能优于最大似然估计器、针对未知先验的完全贝叶斯处理方法以及Regret Balancing方案。

英文摘要:

Bayesian optimisation requires fitting a Gaussian process model, which in turn requires specifying prior on the unknown black-box function -- most of the theoretical literature assumes this prior is known. However, it is common to have more than one possible prior for a given black-box function, for example suggested by domain experts with differing opinions. In some cases, the type-II maximum likelihood estimator for selecting prior enjoys the consistency guarantee, but it does not universally apply to all types of priors. If the problem is stationary, one could rely on the Regret Balancing scheme to conduct the optimisation, but in the case of time-varying problems, such a scheme cannot be used. To address this gap in existing research, we propose a novel algorithm, PE-GP-UCB, which is capable of solving time-varying Bayesian optimisation problems even without the exact knowledge of the function's prior. The algorithm relies on the fact that either the observed function values are consistent with some of the priors, in which case it is easy to reject the wrong priors, or the observations are consistent with all candidate priors, in which case it does not matter which prior our model relies on. We provide a regret bound on the proposed algorithm. Finally, we empirically evaluate our algorithm on toy and real-world time-varying problems and show that it outperforms the maximum likelihood estimator, fully Bayesian treatment of unknown prior and Regret Balancing.

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。

↑