arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.07115stat.MLcs.LG
收藏

高斯过程后验的主成分分析

Principal component analysis for Gaussian process posteriors

Kyushu Institute of Technology(九州工业大学);The National Institute of Advanced Industrial Science and Technology(国立先进工业科学技术研究所);RIKEN AIP(理化学研究所AIP中心)

Hideaki Ishibashi, Shotaro Akaho

AI总结:

本文提出GP-PCA方法,通过在信息几何框架下将无限维GP后验降至有限维,实现低维空间估计,并基于变分推断提出近似方法,实验验证其作为元学习的有效性。

摘要:

本文提出了一种针对高斯过程(GP)后验的主成分分析扩展方法,记为GP-PCA。由于GP-PCA估计GP后验的低维空间,它可用于元学习(meta-learning),这是一种通过估计一组任务的结构来提升目标任务性能的框架。问题在于如何为具有无限维参数的GP集合定义结构,如坐标系和散度。在本研究中,我们在信息几何框架下,通过考虑具有相同先验的GP后验空间,将GP的无限性降至有限维情况。此外,我们提出了一种基于变分推断的GP-PCA近似方法,并通过实验证明了GP-PCA作为元学习的有效性。

英文摘要:

This paper proposes an extension of principal component analysis for Gaussian process (GP) posteriors, denoted by GP-PCA. Since GP-PCA estimates a low-dimensional space of GP posteriors, it can be used for meta-learning, which is a framework for improving the performance of target tasks by estimating a structure of a set of tasks. The issue is how to define a structure of a set of GPs with an infinite-dimensional parameter, such as coordinate system and a divergence. In this study, we reduce the infiniteness of GP to the finite-dimensional case under the information geometrical framework by considering a space of GP posteriors that have the same prior. In addition, we propose an approximation method of GP-PCA based on variational inference and demonstrate the effectiveness of GP-PCA as meta-learning through experiments.

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。

↑