Adaptive Symmetry Discovery for Dynamical System Identification
Not on the map. Why?
Abstract (arXiv)
Dynamical systems model trajectory data generated by fixed underlying dynamics, with applications ranging from biology to physics. Especially in scientific settings, dynamical systems are not generic but often exhibit symmetries imposed by physical laws, formalized through equivariance with respect to group actions. The identification problem concerns recovering the parameters of a system from observed trajectories. In this work, we study adaptive symmetry discovery for dynamical system identification and address how a system can be identified from a single trajectory when it is equivariant with respect to an unknown symmetry group. To this end, we first show that for known symmetries, the system can be identified from a significantly shorter single trajectory than in the generic setting, and we precisely characterize this improvement. We then consider the automatic symmetry discovery setting, proposing a method to learn the symmetry group directly from a single trajectory and incorporate it into the identification procedure, achieving the same optimal trajectory length as in the known-symmetry case. Our analysis relies on tools from group representation theory and the expander properties of Cayley graphs, and may be of independent interest for the study of symmetries in dynamical systems.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 41 of its 59 references that are in the map. First posted on arXiv 8 Aug 2026.
Builds on
41 of 59 references.
References · 41
- 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,131
- 2021Frame Averaging for Invariant and Equivariant Network Designcited by 166
- 2018Learning Without Mixing: Towards A Sharp Analysis of Linear System Identificationcited by 158
- 2022Equivariance with Learned Canonicalization Functionscited by 106
- 2018Near optimal finite time identification of arbitrary linear dynamical systemscited by 89
- 2021Automatic Symmetry Discovery with Lie Algebra Convolutional Networkcited by 82
- 2020Learning nonlinear dynamical systems from a single trajectorycited by 46
- 2021Learning Partial Equivariances from Datacited by 44
- 2021Physics-Guided Deep Learning for Dynamical Systems: A Surveycited by 43
- 2023Generative Adversarial Symmetry Discoverycited by 40
- 2023Latent Space Symmetry Discoverycited by 29
- 2023The Exact Sample Complexity Gain from Invariances for Kernel Regressioncited by 27
29 more
- 2022Single Trajectory Nonparametric Learning of Nonlinear Dynamicscited by 26
- 2022Learning from many trajectoriescited by 24
- 2023Learning Layer-wise Equivariances Automatically using Gradientscited by 23
- 2024Lie Algebra Canonicalization: Equivariant Neural Operators under arbitrary Lie Groupscited by 23
- 2023A Unified Framework to Enforce, Discover, and Promote Symmetry in Machine Learningcited by 22
- 2021Frame Averaging for Equivariant Shape Space Learningcited by 18
- 2024Equivariance via Minimal Frame Averaging for More Symmetries and Efficiencycited by 18
- 2024Learning Infinitesimal Generators of Continuous Symmetries from Datacited by 17
- 2024Symmetry-Informed Governing Equation Discoverycited by 14
- 2024Symmetry Discovery Beyond Affine Transformationscited by 14
- 2023Sample Complexity Bounds for Estimating Probability Divergences under Invariancescited by 11
- 2024On the Ability of Deep Networks to Learn Symmetries from Data: A Neural Kernel Theorycited by 10
- 2024Symmetry Discovery for Different Data Typescited by 7
- 2025Discovering Symbolic Differential Equations with Symmetry Invariantscited by 6
- 2025Achieving Approximate Symmetry Is Exponentially Easier than Exact Symmetrycited by 6
- 2025Learning with Exact Invariances in Polynomial Timecited by 5
- 2025Explicit Discovery of Nonlinear Symmetries from Dynamic Datacited by 5
- 2025Latent Mixture of Symmetries for Sample-Efficient Dynamic Learningcited by 5
- 2025AtlasD: Automatic Local Symmetry Discoverycited by 4
- 2025Governing Equation Discovery from Data Based on Differential Invariantscited by 4
- 2025Research Program: Theory of Learning in Dynamical Systemscited by 4
- 2025Continuous Symmetry Discovery and Enforcement Using Infinitesimal Generators of Multi-parameter Group Actionscited by 3
- 2025Control-Oriented System Identification: Classical, Learning, and Physics-Informed Approachescited by 3
- 2026Data Augmentation: A Fourier Analysis Perspectivecited by 3
- 2025Learning finite symmetry groups of dynamical systems via equivariance detectioncited by 2
- 2025Learning Equivariant Functions via Quadratic Formscited by 2
- 2020Optimal control with learning on the fly: a toy problemcited by 1
- 2021Controlling Unknown Linear Dynamics with Bounded Multiplicative Regretcited by 1
- 2024Almost Optimal Agnostic Control of Unknown Linear Dynamicscited by 1
Built on by
Citing papers per year: 2026: 2.
2 papers.