Critical Investigation of Failure Modes in Physics-informed Neural Networks
Preprint
Abstract
Several recent works in scientific machine learning have revived interest in the application of neural networks to partial differential equations (PDEs). A popular approach is to aggregate the residual form of the governing PDE and its boundary conditions as soft penalties into a composite objective/loss function for training neural networks, which is commonly referred to as physics-informed neural networks (PINNs). In the present study, we visualize the loss landscapes and distributions of learned parameters and explain the ways this particular formulation of the objective function may hinder or even prevent convergence when dealing with challenging target solutions. We construct a purely data-driven loss function composed of both the boundary loss and the domain loss. Using this data-driven loss function and, separately, a physics-informed loss function, we then train two neural network models with the same architecture. We show that incomparable scales between boundary and domain loss terms are the culprit behind the poor performance. Additionally, we assess the performance of both approaches on two elliptic problems with increasingly complex target solutions. Based on our analysis of their loss landscapes and learned parameter distributions, we observe that a physics-informed neural network with a composite objective function formulation produces highly non-convex loss surfaces that are difficult to optimize and are more prone to the problem of vanishing gradients.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equationsJournal of Computational Physics, 2019
- DGM: A deep learning algorithm for solving partial differential equationsJournal of Computational Physics, 2018
- The Deep Ritz Method: A Deep Learning-Based Numerical Algorithm for Solving Variational ProblemsCommunications in Mathematics and Statistics, 2018
- Deep Residual Learning for Image Recognition2016
- Solving differential equations with unsupervised neural networksChemical Engineering and Processing - Process Intensification, 2003
- Artificial neural networks for solving ordinary and partial differential equationsIEEE Transactions on Neural Networks, 1998
- Neural Network Differential Equation and Plasma Equilibrium SolverPhysical Review Letters, 1995
- Neural‐network‐based approximations for solving partial differential equationsCommunications in Numerical Methods in Engineering, 1994
- Multilayer feedforward networks with a nonpolynomial activation function can approximate any functionNeural Networks, 1993
- Multilayer feedforward networks are universal approximatorsNeural Networks, 1989