GPU-accelerated particle methods for evaluation of sparse observations for inverse problems constrained by diffusion PDEs
GPU-accelerated particle methods for evaluation of sparse observations for inverse problems constrained by diffusion PDEs
复制标题
用于评估受扩散偏微分方程约束的逆问题的稀疏观测值的 GPU 加速粒子方法
DOI:
10.1016/j.jcp.2019.04.034
复制
发表时间:
2019
影响因子:
4.1
通讯作者:
Krometis, Justin
中科院分区:
文献类型:
--
作者:
Borggaard, Jeff;Glatt-Holtz, Nathan;Krometis, Justin
We consider the inverse problem of estimating parameters of a driven diffusion (e.g., the underlying fluid flow, diffusion coefficient, or source terms) from point measurements of a passive scalar (e.g., the concentration of a pollutant). We present two particle methods that leverage the structure of the inverse problem to enable efficient computation of the forward map, one for time evolution problems and one for a Dirichlet boundary-value problem. The methods scale in a natural fashion to modern computational architectures, enabling substantial speedup for applications involving sparse observations and high-dimensional unknowns. Numerical examples of applications to Bayesian inference and numerical optimization are provided.
DOI:
10.1137/19m1267544
发表时间:
2020
期刊:
SIAM/ASA Journal on Uncertainty Quantification
影响因子:
--
作者:
Borggaard, Jeff;Glatt-Holtz, Nathan;Krometis, Justin
通讯作者:
Krometis, Justin
DOI:
--
发表时间:
1993
期刊:
影响因子:
--
作者:
G. Whiffen;C. Shoemaker
通讯作者:
C. Shoemaker
DOI:
10.1098/rspa.2015.0142
发表时间:
2015-07-08
期刊:
Proceedings. Mathematical, physical, and engineering sciences
影响因子:
--
作者:
Hennig P;Osborne MA;Girolami M
通讯作者:
Girolami M