Survey on the Techniques of Physics-Informed Machine Learning

This survey, in the format of an interactive zettelkasten-type vault, will constitute a survey of the techniques in the literature on physics-informed machine learning with a focus on dynamical systems and the Hamiltonian formalism. The main pages are:

• Solving Differential Equations using Neural Networks
• Traditional Methods for solving Hamiltonian Systems
• Hamiltonian Neural Networks and related architectures
• SympNet and other symplectic, structure-preserving networks
• Other architectures

Below are a few, additional ideas that are pertinent to my research:

• Dimension-Preserving Networks and BuckiNet
• Symbolic Regression Methods