MyJournals Home  

RSS FeedsEntropy, Vol. 22, Pages 152: Gaussian Process Regression for Data Fulfilling Linear Differential Equations with Localized Sources (Entropy)

 
 

28 january 2020 10:03:01

 
Entropy, Vol. 22, Pages 152: Gaussian Process Regression for Data Fulfilling Linear Differential Equations with Localized Sources (Entropy)
 


Specialized Gaussian process regression is presented for data that are known to fulfill a given linear differential equation with vanishing or localized sources. The method allows estimation of system parameters as well as strength and location of point sources. It is applicable to a wide range of data from measurement and simulation. The underlying principle is the well-known invariance of the Gaussian probability distribution under linear operators, in particular differentiation. In contrast to approaches with a generic covariance function/kernel, we restrict the Gaussian process to generate only solutions of the homogeneous part of the differential equation. This requires specialized kernels with a direct correspondence of certain kernel hyperparameters to parameters in the underlying equation and leads to more reliable regression results with less training data. Inhomogeneous contributions from linear superposition of point sources are treated via a linear model over fundamental solutions. Maximum likelihood estimates for hyperparameters and source positions are obtained by nonlinear optimization. For differential equations representing laws of physics the present approach generates only physically possible solutions, and estimated hyperparameters represent physical properties. After a general derivation, modeling of source-free data and parameter estimation is demonstrated for Laplace’s equation and the heat/diffusion equation. Finally, the Helmholtz equation with point sources is treated, representing scalar wave data such as acoustic pressure in the frequency domain.


 
184 viewsCategory: Informatics, Physics
 
Entropy, Vol. 22, Pages 153: Selection Consistency of Lasso-Based Procedures for Misspecified High-Dimensional Binary Model and Random Regressors (Entropy)
Entropy, Vol. 22, Pages 151: On the Information Bottleneck Problems: Models, Connections, Applications and Information Theoretic Views (Entropy)
 
 
blog comments powered by Disqus


MyJournals.org
The latest issues of all your favorite science journals on one page

Username:
Password:

Register | Retrieve

Search:

Physics


Copyright © 2008 - 2024 Indigonet Services B.V.. Contact: Tim Hulsen. Read here our privacy notice.
Other websites of Indigonet Services B.V.: Nieuws Vacatures News Tweets Nachrichten