Achieving High Accuracy with PINNs via Energy Natural Gradient Descent
- Authors: Johannes Müller, Marius Zeinhofer
- Preprint year: 2023
- First public date: 2023-02-25
- arXiv: 2302.13163
- Status: Published
- Publication type: Conference paper
- Publication year: 2023
- Proceedings: Proceedings of the International Conference on Machine Learning (ICML 2023)
- Published version: Proceedings
Abstract
We propose energy natural gradient descent, a natural gradient method with respect to a Hessian-induced Riemannian metric as an optimization algorithm for physics-informed neural networks (PINNs) and the deep Ritz method. As a main motivation we show that the update direction in function space resulting from the energy natural gradient corresponds to the Newton direction modulo an orthogonal projection onto the model's tangent space. We demonstrate experimentally that energy natural gradient descent yields highly accurate solutions with errors several orders of magnitude smaller than what is obtained when training PINNs with standard optimizers like gradient descent or Adam, even when those are allowed significantly more computation time.