paper-with-me

홈 › Papers

Achieving High Accuracy with PINNs via Energy Natural Gradients

2023-02-25 · Johannes Müller, Marius Zeinhofer

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.

📄 PDF Abstract BibTeX arXiv:2302.13163

Code (1)

mariuszeinhofer/natural-gradient-pinns-icml23 공식 구현 jax

Tasks

Vocal Bursts Intensity Prediction

Methods 이 논문이 사용한 방법론

Adam 설명 없음
Natural Gradient Descent 설명 없음

Similar Papers 제목 키워드 기반

Improving Energy Natural Gradient Descent through Woodbury, Momentum, and Randomization

2025-05-17 · Andrés Guzmán-Cordero, Felix Dangel, Gil Goldshlager, Marius Zeinhofer

Natural gradient methods significantly accelerate the training of Physics-Informed Neural Networks (PINNs), but are often prohibitively costly. We introduce a suite of techniques to improve the accuracy and efficiency of…

Variational Monte Carlo

PINNs for Electromagnetic Wave Propagation

2025-12-29 · Nilufer K. Bulut arxiv

Physics-Informed Neural Networks (PINNs) solve physical systems by incorporating governing partial differential equations directly into neural network training. In electromagnetism, where well-established methodologies s…

PINN-DT: Optimizing Energy Consumption in Smart Building Using Hybrid Physics-Informed Neural Networks and Digital Twin Framework with Blockchain Security

2025-03-01 · Hajar Kazemi Naeini, Roya Shomali, Abolhassan Pishahang, Hamidreza Hasanzadeh 외

The advancement of smart grid technologies necessitates the integration of cutting-edge computational methods to enhance predictive energy optimization. This study proposes a multi-faceted approach by incorporating (1) D…

Deep Reinforcement Learningenergy management

Collocation-based Robust Variational Physics-Informed Neural Networks (CRVPINN)

2024-01-04 · Marcin Łoś, Tomasz Służalec, Paweł Maczuga, Askold Vilkha 외

Physics-Informed Neural Networks (PINNs) have been successfully applied to solve Partial Differential Equations (PDEs). Their loss function is founded on a strong residual minimization scheme. Variational Physics-Informe…

SetPINNs: Set-based Physics-informed Neural Networks

2024-09-30 · Mayank Nagda, Phil Ostheimer, Thomas Specht, Frank Rhein 외

Physics-Informed Neural Networks (PINNs) solve partial differential equations using deep learning. However, conventional PINNs perform pointwise predictions that neglect dependencies within a domain, which may result in …