Learning piecewise-smooth dynamical systems

DOI10.48550/ARXIV.2608.19785MaRDI QIDQ7285066arXivFDO

Authors Davide Murari, Erik Jansson, Carola-Bibiane Schönlieb, Chris Budd OBE

Publication date 20 August 2026

Published in arXiv

Tldr This paper presents a machine learning framework to identify piecewise-smooth dynamical systems from trajectory data, which includes detecting discontinuities and characterizing their behavior using geometry-constrained neural networks.

Full work available at URL https://doi.org/10.48550/ARXIV.2608.19785

Mathematics Subject Classification ID

Discontinuous ordinary differential equations (34A36) Approximation by arbitrary nonlinear expressions; widths and entropy (41A46) Artificial neural networks and deep learning (68T07) System identification (93B30)

arXiv classification

cs.LGmath.DSmath.NA

Cites work

• An analysis of the periodically forced PP04 climate model, using the theory of non-smooth dynamical systems
• Data-driven discovery of coordinates and governing equations
• Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks
• Deep learning for universal linear embeddings of nonlinear dynamics
• Discontinuous neural networks and discontinuity learning
• Discovering governing equations from data by sparse identification of nonlinear dynamical systems
• Error bounds for approximations with deep ReLU networks
• Grazing bifurcations and transitions between periodic states of the PP04 model for the glacial cycle
• Learning Hamiltonians of constrained mechanical systems
• Long Short-Term Memory
• Modeling with nonsmooth dynamics
• Neural Ordinary Differential Equations
• Numerical differentiation of noisy data: A unifying multi-objective optimization framework
• Numerical differentiation of noisy, nonsmooth data
• Sympnets: intrinsic structure-preserving symplectic networks for identifying Hamiltonian systems
• Universal Differential Equations for Scientific Machine Learning

This page was built for publication: Learning piecewise-smooth dynamical systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7285066)