Deep learning-based partial differential equation(PDE) solvers have received much attention in the past few years. Methods of this category can solve a wide range of PDEs with high accuracy, typically by transforming the problems into highly nonlinear optimization problems of neural network parameters. This work reviews several deep learning solvers proposed a few years ago, including PINN, WAN, DRM, and VPINN. Numerical results are provided to make comparisons amongst them and address the importance of loss formulation and the optimization method. A rigorous error analysis for PINN is also presented. Finally, we discuss the current limitations and bottlenecks of these methods.
@article{arxiv.2407.00442,
title = {A Review of Neural Network Solvers for Second-order Boundary Value Problems},
author = {Ramesh Chandra Sau and Luowei Yin},
journal= {arXiv preprint arXiv:2407.00442},
year = {2024}
}