Multi-Resolution Filters for Massive Spatio-Temporal Data

Multi-Resolution Filters for Massive Spatio-Temporal Data
复制标题

海量时空数据的多分辨率过滤器

DOI:
10.1080/10618600.2021.1886938
复制
发表时间:
2021
影响因子:
2.4
通讯作者:
Katzfuss, Matthias
Katzfuss, Matthias
中科院分区:
数学2区
文献类型:
--
作者:
Jurek, Marcin;Katzfuss, Matthias

文献摘要

参考文献

被引文献

相似文献

时空数据集的大小正在快速增长。例如,通过增加安装在卫星和飞机上的自动化传感器的数量,以提高分辨率来测量环境变量。利用这类通常有噪声和不完整的数据,目标是获得时空过程的完整地图,以及不确定性量化。这里我们主要研究线性高斯状态空间模型中的实时滤波推理。在每个时间点,状态都是在一个非常大的空间网格上评估的空间场,这使得使用卡尔曼滤波进行精确推理在计算上是不可行的。相反,我们提出了一种多分辨率过滤器(MRF),这是一种高度可伸缩的完全概率过滤方法,可以在所有尺度上解析空间要素。我们证明了MRF矩阵表现出一种特殊的块稀疏多分辨率结构,这种结构在时间上的滤波操作下保持不变。我们描述了与现有方法的联系,包括来自数值数学的层次矩阵。我们还讨论了用近似的Rao-Blackwell粒子滤波对时变参数的推断,其中积分似然是用马尔可夫随机场计算的。通过一个模拟研究和一个真实的卫星数据应用,我们证明了马尔可夫随机场的性能大大优于竞争的方法。补充材料包括用于再现模拟的Python代码、磁流变函数的一些详细性质和辅助理论结果。
Spatio-temporal datasets are rapidly growing in size. For example, environmental variables are measured with increasing resolution by increasing numbers of automated sensors mounted on satellites and aircraft. Using such data, which are typically noisy and incomplete, the goal is to obtain complete maps of the spatio-temporal process, together with uncertainty quantification. We focus here on real-time filtering inference in linear Gaussian state-space models. At each time point, the state is a spatial field evaluated on a very large spatial grid, making exact inference using the Kalman filter computationally infeasible. Instead, we propose a multi-resolution filter (MRF), a highly scalable and fully probabilistic filtering method that resolves spatial features at all scales. We prove that the MRF matrices exhibit a particular block-sparse multi-resolution structure that is preserved under filtering operations through time. We describe connections to existing methods, including hierarchical matrices from numerical mathematics. We also discuss inference on time-varying parameters using an approximate Rao-Blackwellized particle filter, in which the integrated likelihood is computed using the MRF. Using a simulation study and a real satellite-data application, we show that the MRF strongly outperforms competing approaches. Supplementary materials include Python code for reproducing the simulations, some detailed properties of the MRF and auxiliary theoretical results.
DOI: 10.1214/17-ba1069
发表时间: 2018-03-01
期刊: BAYESIAN ANALYSIS
影响因子: 4.4
作者:
Bradley, Jonathan R.;Holan, Scott H.;Wikle, Christopher K.
通讯作者: Wikle, Christopher K.
Rao-Blackwellized 粒子滤波器,用于随机捕食者-被捕食系统中的联合参数估计和生物量跟踪。
DOI: 10.3934/mbe.2014.11.573
发表时间: 2014
期刊: Mathematical biosciences and engineering : MBE
影响因子: --
作者:
L. Martin;G. Gilioli;Ettore Lanzarone;J. Míguez;S. Pasquali;F. Ruggeri;D. Ruiz
通讯作者: D. Ruiz
DOI: 10.1016/s0167-7152(96)00208-8
发表时间: 1997
影响因子: 0.8
作者:
M. Kanter
通讯作者: M. Kanter
DOI: 10.5705/ss.202018.0285
发表时间: 2017-10
期刊: Statistica Sinica
影响因子: 1.4
作者:
M. Katzfuss;Wenlong Gong
通讯作者: M. Katzfuss;Wenlong Gong
2010年里茨讲座:筛选效应何时成立?
DOI: --
发表时间: 2011
期刊:
影响因子: --
作者:
M. Stein
通讯作者: M. Stein