Physics informed neural networks for elliptic equations with oscillatory differential operators

12/27/2022

∙

We consider standard physics informed neural network solution methods for elliptic partial differential equations with oscillatory coefficients. We show that if the coefficient in the elliptic operator contains frequencies on the order of 1/ϵ, then the Frobenius norm of the neural tangent kernel matrix associated to the loss function grows as 1/ϵ^2. Numerical examples illustrate the stiffness of the optimization problem.

READ FULL TEXT

Physics informed neural networks for elliptic equations with oscillatory differential operators

A certified wavelet-based physics-informed neural network for the solution of parameterized partial differential equations

Physics-informed Neural Networks for Elliptic Partial Differential Equations on 3D Manifolds

PINNs and GaLS: An Priori Error Estimates for Shallow Physically Informed Neural Network Applied to Elliptic Problems

Physics-informed neural networks with unknown measurement noise

Physics-informed neural networks for learning the homogenized coefficients of multiscale elliptic equations

Sobolev Acceleration and Statistical Optimality for Learning Elliptic Equations via Gradient Descent

Necessary Density Conditions for Sampling and Interpolation in Spectral Subspaces of Elliptic Differential Operators

Physics informed neural networks for elliptic equations with oscillatory differential operators

Related Research

A certified wavelet-based physics-informed neural network for the solution of parameterized partial differential equations

Physics-informed Neural Networks for Elliptic Partial Differential Equations on 3D Manifolds

PINNs and GaLS: An Priori Error Estimates for Shallow Physically Informed Neural Network Applied to Elliptic Problems

Physics-informed neural networks with unknown measurement noise

Physics-informed neural networks for learning the homogenized coefficients of multiscale elliptic equations

Sobolev Acceleration and Statistical Optimality for Learning Elliptic Equations via Gradient Descent

Necessary Density Conditions for Sampling and Interpolation in Spectral Subspaces of Elliptic Differential Operators