Deep Hamiltonian networks based on symplectic integrators

Abstract

HNets is a class of neural networks on grounds of physical prior for learning Hamiltonian systems. This paper explains the influences of different integrators as hyper-parameters on the HNets through error analysis. If we define the network target as the map with zero empirical loss on arbitrary training data, then the non-symplectic integrators cannot guarantee the existence of the network targets of HNets. We introduce the inverse modified equations for HNets and prove that the HNets based on symplectic integrators possess network targets and the differences between the network targets and the original Hamiltonians depend on the accuracy orders of the integrators. Our numerical experiments show that the phase flows of the Hamiltonian systems obtained by symplectic HNets do not exactly preserve the original Hamiltonians, but preserve the network targets calculated; the loss of the network target for the training data and the test data is much less than the loss of the original Hamiltonian; the symplectic HNets have more powerful generalization ability and higher accuracy than the non-symplectic HNets in addressing predicting issues. Thus, the symplectic integrators are of critical importance for HNets.

References in corpus (14)

• On the difficulty of training Recurrent Neural Networks
• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations
• Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
• Inferring solutions of differential equations using noisy multi-fidelity data
• Multistep Neural Networks for Data-driven Discovery of Nonlinear Dynamical Systems
• Gradient Descent Converges to Minimizers
• Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning
• Hamiltonian Graph Networks with ODE Integrators
• Symplectic Recurrent Neural Networks
• Theory II: Landscape of the Empirical Risk in Deep Learning
• Equivariant Hamiltonian Flows
• Hamiltonian Generative Networks
• Neural Canonical Transformation with Symplectic Flows
• Quantifying the generalization error in deep learning in terms of data distribution and neural network smoothness

Cited by in corpus (7)

• Symplectic Neural Networks in Taylor Series Form for Hamiltonian Systems
• Port-Hamiltonian Neural Networks for Learning Explicit Time-Dependent Dynamical Systems
• Symplectic Learning for Hamiltonian Neural Networks
• Locally-symplectic neural networks for learning volume-preserving dynamics
• Inverse modified differential equations for discovery of dynamics
• A hybrid approach for solving the gravitational N-body problem with Artificial Neural Networks
• Benchmarking Energy-Conserving Neural Networks for Learning Dynamics from Data