Graphical Gaussian Process Models for Highly Multivariate Spatial Data.

Graphical Gaussian Process Models for Highly Multivariate Spatial Data.
复制标题

高度多元空间数据的图形高斯过程模型。

DOI:
10.1093/biomet/asab061
复制
发表时间:
2022
期刊:
影响因子:
2.7
通讯作者:
Banerjee,Sudipto
Banerjee,Sudipto
中科院分区:
数学2区
文献类型:
--
作者:
Dey,Debangan;Datta,Abhirup;Banerjee,Sudipto

文献摘要

参考文献

被引文献

相似文献

对于多变量空间高斯过程模型,互协方差函数的常规规范不利用相关变量间图来确保变量之间的过程级条件独立性。这是不可取的,特别是在高度多变量的设置中,其中流行的互协方差函数,如多变量Matérn函数,遭受维数灾难,因为参数和浮点运算的数量分别以二次和三次顺序随变量的数量而增加。我们提出了一类多变量的图形高斯过程,使用一个通用的构造称为拼接,工艺品从图形的互协方差函数,并确保变量之间的过程级条件独立。对于Matérn函数族,拼接产生多变量高斯过程,其单变量分量是Matérn高斯过程,并且符合由图形模型指定的过程级条件独立性。对于高度多变量设置和可分解的图形模型,拼接提供了大量的计算增益和参数降维。我们展示了图形Matérn高斯过程的实用性,联合建模高度多元的空间数据,使用模拟的例子和应用程序的空气污染建模。
For multivariate spatial Gaussian process models, customary specifications of cross-covariance functions do not exploit relational inter-variable graphs to ensure process-level conditional independence between the variables. This is undesirable, especially in highly multivariate settings, where popular cross-covariance functions, such as multivariate Matérn functions, suffer from a curse of dimensionality as the numbers of parameters and floating-point operations scale up in quadratic and cubic order, respectively, with the number of variables. We propose a class of multivariate graphical Gaussian processes using a general construction called stitching that crafts cross-covariance functions from graphs and ensures process-level conditional independence between variables. For the Matérn family of functions, stitching yields a multivariate Gaussian process whose univariate components are Matérn Gaussian processes, and which conforms to process-level conditional independence as specified by the graphical model. For highly multivariate settings and decomposable graphical models, stitching offers massive computational gains and parameter dimension reduction. We demonstrate the utility of the graphical Matérn Gaussian process to jointly model highly multivariate spatial data using simulation examples and an application to air-pollution modelling.
DOI: 10.1016/j.jmva.2011.05.010
发表时间: 2011-11
期刊: J. Multivar. Anal.
影响因子: --
作者:
Bo Li;Hao Zhang
通讯作者: Bo Li;Hao Zhang
DOI: 10.1093/biomet/asp078
发表时间: 2010-03-01
期刊: BIOMETRIKA
影响因子: 2.7
作者:
Apanasovich, Tatiyana V.;Genton, Marc G.
通讯作者: Genton, Marc G.
DOI: 10.5705/ss.202015.0309
发表时间: 2018
期刊: Statistica Sinica
影响因子: 1.4
作者:
W. Kleiber
通讯作者: W. Kleiber
DOI: 10.1007/s13253-020-00414-2
发表时间: 2020-09
期刊: Journal of Agricultural, Biological and Environmental Statistics
影响因子: --
作者:
G. A. Qadir;C. Euán;Ying Sun
通讯作者: G. A. Qadir;C. Euán;Ying Sun
DOI: 10.1007/s13253-018-00348-w
发表时间: 2019-09-01
影响因子: 1.4
作者:
Heaton, Matthew J.;Datta, Abhirup;Zammit-Mangion, Andrew
通讯作者: Zammit-Mangion, Andrew