Symplectic Learning for Hamiltonian Neural Networks

Abstract

Machine learning methods are widely used in the natural sciences to model and predict physical systems from observation data. Yet, they are often used as poorly understood "black boxes," disregarding existing mathematical structure and invariants of the problem. Recently, the proposal of Hamiltonian Neural Networks (HNNs) took a first step towards a unified "gray box" approach, using physical insight to improve performance for Hamiltonian systems. In this paper, we explore a significantly improved training method for HNNs, exploiting the symplectic structure of Hamiltonian systems with a different loss function. This frees the loss from an artificial lower bound. We mathematically guarantee the existence of an exact Hamiltonian function which the HNN can learn. This allows us to prove and numerically analyze the errors made by HNNs which, in turn, renders them fully explainable. Finally, we present a novel post-training correction to obtain the true Hamiltonian only from discretized observation data, up to an arbitrary order.

arXiv comment10 pages, 4 figures; Source code, datasets and pre-trained models available at https://github.com/SpaceAbleOrg/symplectic-hnn

References in corpus (20)

• Adam: A Method for Stochastic Optimization
• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Decoupled Weight Decay Regularization
• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• Machine Learning for Fluid Mechanics
• Enforcing Analytic Constraints in Neural-Networks Emulating Physical Systems
• Hamiltonian Neural Networks
• The Challenge of Machine Learning in Space Weather Nowcasting and Forecasting
• Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge
• Deep Network Approximation for Smooth Functions
• Symplectic ODE-Net: Learning Hamiltonian Dynamics with Control
• Explicit symplectic approximation of nonseparable Hamiltonians: algorithm and long time performance
• Symplectic Recurrent Neural Networks
• Symplectic Neural Networks in Taylor Series Form for Hamiltonian Systems
• Hamiltonian Generative Networks
• EM-like Learning Chaotic Dynamics from Noisy and Partial Observations
• Data-driven Prediction of General Hamiltonian Dynamics via Learning Exactly-Symplectic Maps
• Deep Hamiltonian networks based on symplectic integrators
• Sparse Symplectically Integrated Neural Networks

Cited by in corpus (3)

• Symplectic integration of learned Hamiltonian systems
• Constrained Hamiltonian Systems and Physics-Informed Neural Networks: Hamilton-Dirac Neural Networks
• Machine learning of continuous and discrete variational ODEs with convergence guarantee and uncertainty quantification