Stochastic Processes Under Linear Differential Constraints : Application to Gaussian Process Regression for the 3 Dimensional Free Space Wave Equation

by   Iain Henderson, et al.

Let P be a linear differential operator over π’ŸβŠ‚β„^d and U = (U_x)_x βˆˆπ’Ÿ a second order stochastic process. In the first part of this article, we prove a new simple necessary and sufficient condition for all the trajectories of U to verify the partial differential equation (PDE) T(U) = 0. This condition is formulated in terms of the covariance kernel of U. The novelty of this result is that the equality T(U) = 0 is understood in the sense of distributions, which is a functional analysis framework particularly adapted to the study of PDEs. This theorem provides precious insights during the second part of this article, which is dedicated to performing "physically informed" machine learning on data that is solution to the homogeneous 3 dimensional free space wave equation. We perform Gaussian Process Regression (GPR) on this data, which is a kernel based Bayesian approach to machine learning. To do so, we put Gaussian process (GP) priors over the wave equation's initial conditions and propagate them through the wave equation. We obtain explicit formulas for the covariance kernel of the corresponding stochastic process; this kernel can then be used for GPR. We explore two particular cases : the radial symmetry and the point source. For the former, we derive convolution-free GPR formulas; for the latter, we show a direct link between GPR and the classical triangulation method for point source localization used e.g. in GPS systems. Additionally, this Bayesian framework gives rise to a new answer for the ill-posed inverse problem of reconstructing initial conditions for the wave equation with finite dimensional data, and simultaneously provides a way of estimating physical parameters from this data as in [Raissi et al,2017]. We finish by showcasing this physically informed GPR on a number of practical examples.



There are no comments yet.


page 37

page 38

βˆ™ 05/29/2014

Functional Gaussian processes for regression with linear PDE models

In this paper, we present a new statistical approach to the problem of i...
βˆ™ 12/22/2020

Gaussian Process Regression constrained by Boundary Value Problems

We develop a framework for Gaussian processes regression constrained by ...
βˆ™ 06/16/2020

A Survey of Constrained Gaussian Process Regression: Approaches and Implementation Challenges

Gaussian process regression is a popular Bayesian framework for surrogat...
βˆ™ 06/21/2020

Statistical deconvolution of the free Fokker-Planck equation at fixed time

We are interested in reconstructing the initial condition of a non-linea...
βˆ™ 08/02/2018

Machine Learning of Space-Fractional Differential Equations

Data-driven discovery of "hidden physics" -- i.e., machine learning of d...
βˆ™ 10/17/2021

Gravitational wave surrogates through automated machine learning

We analyze a prospect for predicting gravitational waveforms from compac...
βˆ™ 09/28/2020

An inverse random source problem for the one-dimensional Helmholtz equation with attenuation

This paper is concerned with an inverse random source problem for the on...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.