Scalable extended dynamic mode decomposition using random kernel approximation
From MaRDI portal
Abstract: The Koopman operator is a linear, infinite-dimensional operator that governs the dynamics of system observables; Extended Dynamic Mode Decomposition (EDMD) is a data-driven method for approximating the Koopman operator using functions (features) of the system state snapshots. This paper investigates an approach to EDMD in which the features used provide random approximations to a particular kernel function. The objective of this is computational economy for large data sets: EDMD is generally ill-suited for problems with large state dimension, and its dual kernel formulation (KDMD) is well-suited for such problems only if the number of data snapshots is relatively small. We discuss two specific methods for generating features: random Fourier features, and the Nystrom method. The first method is a data-independent method for translation-invariant kernels only and involves random sampling in feature space; the second method is a data-dependent empirical method that may be used for any kernel and involves random sampling of data. We first discuss how these ideas may be applied in an EDMD context, as well as a means for adaptively adding random Fourier features. We demonstrate these methods on two example problems and conclude with an analysis of the relative benefits and drawbacks of each method.
Recommendations
- A data-driven approximation of the koopman operator: extending dynamic mode decomposition
- On convergence of extended dynamic mode decomposition to the Koopman operator
- A kernel-based method for data-driven Koopman spectral analysis
- Data-driven approximation of the Koopman generator: model reduction, system identification, and control
- The mpEDMD Algorithm for Data-Driven Computations of Measure-Preserving Dynamical Systems
Cites work
- A computational method to extract macroscopic variables and their dynamics in multiscale systems
- A data-driven approximation of the koopman operator: extending dynamic mode decomposition
- A kernel-based method for data-driven Koopman spectral analysis
- Advanced Lectures on Machine Learning
- Analysis of Fluid Flows via Spectral Properties of the Koopman Operator
- Applied and computational measurable dynamics
- Applied Koopmanism
- Compressed sensing and dynamic mode decomposition
- Diffusion maps
- Diffusion maps, spectral clustering and reaction coordinates of dynamical systems
- Dynamic mode decomposition of numerical and experimental data
- Generalized Inverses of Partitioned Matrices
- Hamiltonian Systems and Transformation in Hilbert Space
- Isostables, isochrons, and Koopman spectrum for the action-angle representation of stable fixed point dynamics
- Linearly recurrent autoencoder networks for learning dynamics
- Nonlinear Dimensionality Reduction
- On dynamic mode decomposition: theory and applications
- On the use of Fourier averages to compute the global isochrons of (quasi)periodic dynamics
- Optimal mode decomposition for unsteady flows
- Spectral analysis of nonlinear flows
- Spectral properties of dynamical systems, model reduction and decompositions
- Variants of dynamic mode decomposition: boundary condition, Koopman, and Fourier analyses
Cited in
(27)- On convergence of extended dynamic mode decomposition to the Koopman operator
- On robust computation of Koopman operator and prediction in random dynamical systems
- Geometric considerations of a good dictionary for Koopman analysis of dynamical systems: cardinality, ``primary eigenfunction, and efficient representation
- Tensor-based computation of metastable and coherent sets
- Data-driven approximation of the Koopman generator: model reduction, system identification, and control
- tgEDMD: approximation of the Kolmogorov operator in tensor train format
- Eigendecompositions of transfer operators in reproducing kernel Hilbert spaces
- Extension of dynamic mode decomposition for dynamic systems with incomplete information based on t-model of optimal prediction
- A kernel-based method for data-driven Koopman spectral analysis
- Sparsity-promoting algorithms for the discovery of informative Koopman-invariant subspaces
- Numerical methods to evaluate Koopman matrix from system equations
- Robust approximation of the stochastic Koopman operator
- Randomized Dynamic Mode Decomposition
- Kernel methods for detecting coherent structures in dynamical data
- Koopman operator and its approximations for systems with symmetries
- Two methods to approximate the Koopman operator with a reservoir computer
- The mpEDMD Algorithm for Data-Driven Computations of Measure-Preserving Dynamical Systems
- Generalizing dynamic mode decomposition: balancing accuracy and expressiveness in Koopman approximations
- Ensemble forecasts in reproducing kernel Hilbert space family
- Reachability of Koopman Linearized Systems Using Random Fourier Feature Observables and Polynomial Zonotope Refinement
- Learning nonparametric ordinary differential equations from noisy data
- Deep learning enhanced dynamic mode decomposition
- Spatiotemporal GEIM for ultra-real-time prediction of coupled multi-physics in reactor transients using sparse observations
- Reachability of Koopman linearized systems using explicit kernel approximation and polynomial zonotope refinement
- Linear quadratic control of nonlinear systems with Koopman operator learning and the Nyström method
- Falsification using reachability of surrogate Koopman models
- Data-driven optimal prediction with control
This page was built for publication: Scalable extended dynamic mode decomposition using random kernel approximation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5230606)
