Machine learning of continuous and discrete variational ODEs with convergence guarantee and uncertainty quantification

Abstract

The article introduces a method to learn dynamical systems that are governed by Euler--Lagrange equations from data. The method is based on Gaussian process regression and identifies continuous or discrete Lagrangians and is, therefore, structure preserving by design. A rigorous proof of convergence as the distance between observation data points converges to zero and lower bounds for convergence rates are provided. Next to convergence guarantees, the method allows for quantification of model uncertainty, which can provide a basis of adaptive sampling techniques. We provide efficient uncertainty quantification of any observable that is linear in the Lagrangian, including of Hamiltonian functions (energy) and symplectic structures, which is of interest in the context of system identification. The article overcomes major practical and theoretical difficulties related to the ill-posedness of the identification task of (discrete) Lagrangians through a careful design of geometric regularisation strategies and through an exploit of a relation to convex minimisation problems in reproducing kernel Hilbert spaces.

References in corpus (13)

• On Learning Hamiltonian Systems from Data
• Nonparametric inference of interaction laws in systems of agents from trajectory data
• Symplectic Learning for Hamiltonian Neural Networks
• Symplectic Gaussian Process Regression of Hamiltonian Flow Maps
• Solving and Learning Nonlinear PDEs with Gaussian Processes
• Modified Equations for Variational Integrators
• Variational Learning of Euler-Lagrange Dynamics from Data
• Physically Consistent Learning of Conservative Lagrangian Systems with Gaussian Processes
• Hamiltonian Neural Networks with Automatic Symmetry Detection
• Discrete Lagrangian Neural Networks with Automatic Symmetry Discovery
• Learning particle swarming models from data with Gaussian processes
• Learning of discrete models of variational PDEs from data
• Error Analysis of Kernel/GP Methods for Nonlinear and Parametric PDEs