Fast Algorithms for Bayesian Inversion

Fast Algorithms for Bayesian Inversion
复制标题

贝叶斯反演的快速算法

DOI:
--
复制
发表时间:
2013
期刊:
影响因子:
--
通讯作者:
P. Kitanidis
P. Kitanidis
中科院分区:
--
文献类型:
--
作者:
Sivaram Ambikasaran;A. Saibaba;Eric F Darve;P. Kitanidis

文献摘要

参考文献

被引文献

相似文献

在本文中,我们回顾了一些使用贝叶斯方法解决大规模随机逆问题的快速算法。在简要介绍贝叶斯随机逆方法之后,我们回顾了以下计算技术来解决大规模问题:快速傅立叶变换、快速多极子方法(经典和黑盒版本),最后是分层矩阵方法。我们强调,这主要是一篇调查论文,提出了一些适用于大规模贝叶斯反演技术的快速算法,适用于地质统计学产生的应用。本文以研究生和计算工程师可以理解的水平呈现。因此,我们主要介绍算法思想和理论结果。
In this article, we review a few fast algorithms for solving large-scale stochastic inverse problems using Bayesian methods. After a brief introduction to the Bayesian stochastic inverse methodology, we review the following computational techniques, to solve large scale problems: the fast Fourier transform, the fast multipole method (classical and a black-box version), and finallym the hierarchical matrix approach. We emphasize that this is mainly a survey paper presenting a few fast algorithms applicable to large-scale Bayesian inversion techniques, applicable to applications arising from geostatistics. The article is presented at a level accessible to graduate students and computational engineers. Hence, we mainly present the algorithmic ideas and theoretical results.
DOI: 10.1029/2009wr008575
发表时间: 2010-07
影响因子: 5.4
作者:
Davina Pollock;O. Cirpka
通讯作者: Davina Pollock;O. Cirpka