Neural Galerkin schemes with active learning for high-dimensional evolution equations

DOI10.1016/j.jcp.2023.112588arXiv2203.01360OpenAlexW4387910988MaRDI QIDQ6117685

Benjamin Peherstorfer, Joan Bruna, Eric Vanden-Eijnden

Publication date: 21 February 2024

Published in: Journal of Computational Physics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2203.01360

zbMATH Keywords

Monte Carlopartial differential equationsmachine learningscientific computingactive learningdeep neural networks

Mathematics Subject Classification ID

Artificial intelligence (68Txx) Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65Mxx) Numerical methods for partial differential equations, boundary value problems (65Nxx)

Cites Work

• Unnamed Item
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• Recurrent neural network closure of parametric POD-Galerkin reduced-order models based on the Mori-Zwanzig formalism
• Analytical and numerical aspects of certain nonlinear evolution equations. III. Numerical, Korteweg-de Vries equation
• Dynamically orthogonal field equations for continuous stochastic dynamical systems
• From quantum to classical molecular dynamics: Reduced models and numerical analysis.
• A family of embedded Runge-Kutta formulae
• Theory and practice of finite elements.
• DGM: a deep learning algorithm for solving partial differential equations
• Model reduction and neural networks for parametric PDEs
• A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder
• Non-intrusive reduced order modeling of unsteady flows using artificial neural networks with application to a combustion problem
• Deep learning of dynamics and signal-noise decomposition with time-stepping constraints
• Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• A Survey of Projection-Based Model Reduction Methods for Parametric Dynamical Systems
• Online Adaptive Model Reduction for Nonlinear Systems via Low-Rank Updates
• Neural‐network‐based approximations for solving partial differential equations
• Geometric Subspace Updates with Applications to Online Adaptive Nonlinear Model Reduction
• The Shifted Proper Orthogonal Decomposition: A Mode Decomposition for Multiple Transport Phenomena
• Solving high-dimensional partial differential equations using deep learning
• Projection-based model reduction with dynamically transformed modes
• Nonlinear model reduction on metric spaces. Application to one-dimensional conservative PDEs in Wasserstein spaces
• Solving parametric PDE problems with artificial neural networks
• Evolution of Nonlinear Reduced-Order Solutions for PDEs with Conserved Quantities
• Rank-adaptive structure-preserving model order reduction of Hamiltonian systems
• Breaking the Kolmogorov Barrier with Nonlinear Model Reduction
• Model Reduction for Transport-Dominated Problems via Online Adaptive Bases and Adaptive Sampling
• Learning data-driven discretizations for partial differential equations
• Reduced basis techniques for nonlinear conservation laws
• Error Analysis of the Dynamically Orthogonal Approximation of Time Dependent Random PDEs
• Dynamical Model Reduction Method for Solving Parameter-Dependent Dynamical Systems
• Dynamical Low‐Rank Approximation
• Reduced basis methods for time-dependent problems