An additive approximate Gaussian process model for large spatio‐temporal data

An additive approximate Gaussian process model for large spatio‐temporal data
复制标题

大时空数据的加性近似高斯过程模型

DOI:
10.1002/env.2569
复制
发表时间:
2017
期刊:
影响因子:
1.7
通讯作者:
E. Kang
E. Kang
中科院分区:
环境科学与生态学3区
文献类型:
--
作者:
P. Ma;B. Konomi;E. Kang

文献摘要

参考文献

被引文献

相似文献

受大型地面臭氧数据集的启发,我们提出了一种新的计算高效的加性近似高斯过程。该方法结合了计算复杂度降低方法和可分离的协方差函数,可以灵活地捕获各种时空依赖结构。第一个组件能够捕获不可分离的时空变化,而第二个组件捕获可分离的变化。基于模型的分层公式,我们能够利用两个组件的计算优势并执行有效的贝叶斯推理。为了证明所提出方法的推理和计算优势,我们假设潜在时空协方差结构的各种场景进行了广泛的模拟研究。该方法还适用于分析美国东部地面臭氧的大型时空测量结果。
Motivated by a large ground‐level ozone data set, we propose a new computationally efficient additive approximate Gaussian process. The proposed method incorporates a computational‐complexity‐reduction method and a separable covariance function, which can flexibly capture various spatio‐temporal dependence structures. The first component is able to capture nonseparable spatio‐temporal variability, whereas the second component captures the separable variation. Based on a hierarchical formulation of the model, we are able to utilize the computational advantages of both components and perform efficient Bayesian inference. To demonstrate the inferential and computational benefits of the proposed method, we carry out extensive simulation studies assuming various scenarios of an underlying spatio‐temporal covariance structure. The proposed method is also applied to analyze large spatio‐temporal measurements of ground‐level ozone in the Eastern United States.
DOI: 10.1214/16-aoas931
发表时间: 2016-09
期刊: The annals of applied statistics
影响因子: --
作者:
Datta A;Banerjee S;Finley AO;Hamm NAS;Schaap M
通讯作者: Schaap M
DOI: 10.1007/s11222-010-9224-x
发表时间: 2012-05-01
影响因子: 2.2
作者:
Gramacy, Robert B.;Lee, Herbert K. H.
通讯作者: Lee, Herbert K. H.
DOI: 10.1080/10618600.2018.1537924
发表时间: 2019-03-21
影响因子: 2.4
作者:
Finley, Andrew O.;Datta, Abhirup;Banerjee, Sudipto
通讯作者: Banerjee, Sudipto
DOI: 10.1214/19-sts755
发表时间: 2021-02-01
影响因子: 5.7
作者:
Katzfuss, Matthias;Guinness, Joseph
通讯作者: Guinness, Joseph