Symplecticity-preserving prediction of parameter-dependent Hamiltonian dynamics by Generalized Kernel Interpolation

DOI10.48550/arXiv.2606.18937MaRDI QIDQ7266588zbMATH OpenarXivFDO

Authors Robin Herkert, Tobias Ehring, Bernard Haasdonk

Publication date 17 June 2026

Published in arXiv

Tldr This paper extends kernel-based symplectic predictors to parameter-augmented settings where learned flow-map surrogates depend on physical parameters and macro time-step sizes via a product kernel ansatz. The resulting large-step predictors are symplectic by construction for every fixed parameter instance, with convergence analysis and error bounds carried over from the non-augmented framework.

Full work available at URL https://doi.org/10.48550/arXiv.2606.18937

arXiv classification

math.NA

This page was built for publication: Symplecticity-preserving prediction of parameter-dependent Hamiltonian dynamics by Generalized Kernel Interpolation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7266588)