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 1902.10664cs.LGcs.NAmath.NAstat.ML
收藏

基于高斯过程混合的主动学习局部函数复杂度

Local Function Complexity for Active Learning via Mixture of Gaussian Processes

Berlin Institute of Technology(柏林工业大学);Physikalisch-Technische Bundesanstalt(德国联邦物理技术研究院);BIFOLD-Berlin Institute for the Foundations of Learning and Data(BIFOLD-柏林学习与数据基础研究院);Korea University(高丽大学);Max Planck Institute for Informatics(马克斯·普朗克信息学研究所);RIKEN AIP(理化学研究所人工智能研究中心)

Danny Panknin, Stefan Chmiela, Klaus-Robert Müller, Shinichi Nakajima

AI总结:

针对真实数据非均匀性带来的统计推断挑战,本文提出基于高斯过程回归的局部函数复杂度估计方法,替代局部多项式平滑指标构建鲁棒可扩展的模型无关主动学习框架,在合成数据和小分子力场重构任务中实现最优性能并降低训练需求。

摘要:

真实世界数据中的非均匀性(例如由观测噪声水平变化或源函数结构复杂度差异导致的非均匀性)给统计推断带来了一系列独特挑战。在物理资源或计算时间有限的情况下,考虑这些非均匀性能大幅提升预测能力。本文借鉴源自局部多项式平滑(LPS)领域的局部函数复杂度(LFC)估计最新理论结果,建立了一种局部结构复杂度概念,并据此开发了一个模型无关的主动学习(AL)框架。由于依赖逐点估计,LPS模型类别在应对真实世界问题中常见的大输入空间维度时,既不具备鲁棒性也缺乏可扩展性。本文推导并估计了基于高斯过程回归(GPR)的LPS型LFC等效量,并将其用作上述框架的替代指标,以提升框架的鲁棒性和可扩展性。我们先在一个典型的低维合成数据集上评估了所提LFC估计在主动学习应用中的有效性,随后完成了重构小型有机分子量子化学力场这一具有挑战性的真实世界任务,证明该方法实现了当前最优性能,且训练需求显著降低。

英文摘要:

Inhomogeneities in real-world data, e.g., due to changes in the observation noise level or variations in the structural complexity of the source function, pose a unique set of challenges for statistical inference. Accounting for them can greatly improve predictive power when physical resources or computation time is limited. In this paper, we draw on recent theoretical results on the estimation of local function complexity (LFC), derived from the domain of local polynomial smoothing (LPS), to establish a notion of local structural complexity, which is used to develop a model-agnostic active learning (AL) framework. Due to its reliance on pointwise estimates, the LPS model class is not robust and scalable concerning large input space dimensions that typically come along with real-world problems. Here, we derive and estimate the Gaussian process regression (GPR)-based analog of the LPS-based LFC and use it as a substitute in the above framework to make it robust and scalable. We assess the effectiveness of our LFC estimate in an AL application on a prototypical low-dimensional synthetic dataset, before taking on the challenging real-world task of reconstructing a quantum chemical force field for a small organic molecule and demonstrating state-of-the-art performance with a significantly reduced training demand.

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。

↑