Kernel Analog Forecasting: Multiscale Test Problems
From MaRDI portal
Abstract: Data-driven prediction is becoming increasingly widespread as the volume of data available grows and as algorithmic development matches this growth. The nature of the predictions made, and the manner in which they should be interpreted, depends crucially on the extent to which the variables chosen for prediction are Markovian, or approximately Markovian. Multiscale systems provide a framework in which this issue can be analyzed. In this work kernel analog forecasting methods are studied from the perspective of data generated by multiscale dynamical systems. The problems chosen exhibit a variety of different Markovian closures, using both averaging and homogenization; furthermore, settings where scale-separation is not present and the predicted variables are non-Markovian, are also considered. The studies provide guidance for the interpretation of data-driven prediction methods when used in practice.
Recommendations
- Analog forecasting with dynamics-adapted kernels
- Operator-theoretic framework for forecasting nonlinear time series with kernel analog techniques
- Nonparametric forecasting: a comparison of three kernel-based methods
- Combining Forecasts with Nonparametric Kernel Regressions
- scientific article; zbMATH DE number 1796933
- Kernel-based prediction of non-Markovian time series
- Prediction of autoregressive processes via the reproducing kernel spaces
- scientific article; zbMATH DE number 2185586
Cites work
- A computational strategy for multiscale systems with applications to Lorenz 96 model
- A note on diffusion limits of chaotic skew-product flows
- Almost sure invariance principle for nonuniformly hyperbolic systems
- Analog forecasting with dynamics-adapted kernels
- Analysis of Fluid Flows via Spectral Properties of the Koopman Operator
- Asymptotic analysis for periodic structures
- Asymptotic theory of mixing stochastic ordinary differential equations
- Bi-stochastic kernels via asymmetric affinity functions
- Consistency of spectral clustering
- Data-driven model reduction and transfer operator approximation
- Data-driven spectral analysis of the Koopman operator
- Data-driven spectral decomposition and forecasting of ergodic dynamical systems
- Delay-coordinate maps and the spectra of Koopman operators
- Deterministic homogenization for fast-slow systems with chaotic noise
- Deterministic Nonperiodic Flow
- Diffusion maps
- Diffusion maps, spectral clustering and reaction coordinates of dynamical systems
- Dynamic mode decomposition of numerical and experimental data
- Ergodic theory, dynamic mode decomposition, and computation of spectral properties of the Koopman operator
- Error estimates for spectral convergence of the graph Laplacian on random geometric graphs toward the Laplace-Beltrami operator
- Extracting macroscopic dynamics: model problems and algorithms
- From graph to manifold Laplacian: the convergence rate
- Geometric harmonics: a novel tool for multiscale out-of-sample extension of empirical functions
- scientific article; zbMATH DE number 45848 (Why is no real title available?)
- Koopman analysis of the long-term evolution in a turbulent convection cell
- Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
- Machine learning for prediction with missing dynamics
- Modeling of missing dynamical systems: deriving parametric models using a nonparametric framework
- Multiscale Methods
- On the Approximation of Complicated Dynamical Behavior
- Operator-theoretic framework for forecasting nonlinear time series with kernel analog techniques
- Principles of multiscale modeling
- Reproducing kernel Hilbert space compactification of unitary evolution groups
- Scikit-learn: machine learning in Python
- Smooth approximation of stochastic differential equations
- Spectral analysis of nonlinear flows
- Spectral properties of dynamical systems, model reduction and decompositions
- The Lorenz attractor exists
- Theory of Reproducing Kernels
- Variable bandwidth diffusion kernels
Cited in
(17)- Ensemble Kalman inversion for sparse learning of dynamical systems from time-averaged data
- Operator-theoretic framework for forecasting nonlinear time series with kernel analog techniques
- Analog forecasting with dynamics-adapted kernels
- Stochastic methods for the prediction of complex multiscale phenomena
- Data-driven prediction of multistable systems from sparse measurements
- Modern Koopman theory for dynamical systems
- A framework for machine learning of model error in dynamical systems
- Multiscale extensions for enhancing coarse grid computations
- The mpEDMD Algorithm for Data-Driven Computations of Measure-Preserving Dynamical Systems
- Active Operator Inference for Learning Low-Dimensional Dynamical-System Models from Noisy Data
- Operator inference with roll outs for learning reduced models from scarce and low-quality data
- Quantum Mechanics for Closure of Dynamical Systems
- Learning to Forecast Dynamical Systems from Streaming Data
- Rigorous data‐driven computation of spectral properties of Koopman operators for dynamical systems
- Ensemble forecasts in reproducing kernel Hilbert space family
- Learning nonparametric ordinary differential equations from noisy data
- Ensemble Kalman methods: a mean-field perspective
This page was built for publication: Kernel Analog Forecasting: Multiscale Test Problems
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5006465)
