Learning Hamiltonian systems with mono-implicit Runge-Kutta methods

DOI10.1007/978-3-031-38271-0_55MaRDI QIDQ6178889zbMATH OpenOpenAlexFDO

Authors Håkon Noren

Publication date 16 January 2024

Published in Lecture Notes in Computer Science (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2303.03769

zbMATH Keywords

deep neural networksHamiltonian systemsinverse problemsmono-implicit Runge-Kutta

Mathematics Subject Classification ID

Numerical solution of inverse problems involving ordinary differential equations (65L09) Artificial neural networks and deep learning (68T07)

Abstract: Numerical integrators could be used to form interpolation conditions when training neural networks to approximate the vector field of an ordinary differential equation (ODE) from data. When numerical one-step schemes such as the Runge-Kutta methods are used to approximate the temporal discretization of an ODE with a known vector field, properties such as symmetry and stability are much studied. Here, we show that using mono-implicit Runge-Kutta methods of high order allows for accurate training of Hamiltonian neural networks on small datasets. This is demonstrated by numerical experiments where the Hamiltonian of the chaotic double pendulum in addition to the Fermi-Pasta-Ulam-Tsingou system is learned from data.

Recommendations

• Deep Hamiltonian neural networks based on symplectic integrators
• Symplectic learning for Hamiltonian neural networks
• Sympnets: intrinsic structure-preserving symplectic networks for identifying Hamiltonian systems
• Learning Hamiltonians of constrained mechanical systems
• A generalized framework of neural networks for Hamiltonian systems

Cites work

• A Class of Implicit Runge-Kutta Methods for the Numerical Integration of Stiff Ordinary Differential Equations
• A family of embedded Runge-Kutta formulae
• A proposal on machine learning via dynamical systems
• Deep Hamiltonian neural networks based on symplectic integrators
• Deep neural networks motivated by partial differential equations
• Geometric Numerical Integration
• Numerical Methods for Ordinary Differential Equations
• Numerical Optimization
• Optimal discrete and continuous mono-implicit Runge-Kutta schemes for BVODEs
• Order Results for Mono-Implicit Runge–Kutta Methods
• Stable architectures for deep neural networks
• Symplectic learning for Hamiltonian neural networks
• Variational learning of Euler-Lagrange dynamics from data

Cited in

(4)

• Symplectic learning for Hamiltonian neural networks
• Pseudo-Hamiltonian neural networks for learning partial differential equations
• A generalized framework of neural networks for Hamiltonian systems
• Learning dynamical systems from noisy data with inverse-explicit integrators

This page was built for publication: Learning Hamiltonian systems with mono-implicit Runge-Kutta methods

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6178889)