Active Operator Inference for Learning Low-Dimensional Dynamical-System Models from Noisy Data
From MaRDI portal
Abstract: Noise poses a challenge for learning dynamical-system models because already small variations can distort the dynamics described by trajectory data. This work builds on operator inference from scientific machine learning to infer low-dimensional models from high-dimensional state trajectories polluted with noise. The presented analysis shows that, under certain conditions, the inferred operators are unbiased estimators of the well-studied projection-based reduced operators from traditional model reduction. Furthermore, the connection between operator inference and projection-based model reduction enables bounding the mean-squared errors of predictions made with the learned models with respect to traditional reduced models. The analysis also motivates an active operator inference approach that judiciously samples high-dimensional trajectories with the aim of achieving a low mean-squared error by reducing the effect of noise. Numerical experiments with high-dimensional linear and nonlinear state dynamics demonstrate that predictions obtained with active operator inference have orders of magnitude lower mean-squared errors than operator inference with traditional, equidistantly sampled trajectory data.
Recommendations
- Sampling Low-Dimensional Markovian Dynamics for Preasymptotically Recovering Reduced Models from Data with Operator Inference
- Operator inference for non-intrusive model reduction of systems with non-polynomial nonlinear terms
- Operator inference and physics-informed learning of low-dimensional models for incompressible flows
- Bayesian operator inference for data-driven reduced-order modeling
- Operator inference of non-Markovian terms for learning reduced models from partially observed state trajectories
Cites work
- A data-driven approximation of the koopman operator: extending dynamic mode decomposition
- A framework for the solution of the generalized realization problem
- A new selection operator for the discrete empirical interpolation method -- improved a priori error bound and extensions
- A survey of projection-based model reduction methods for parametric dynamical systems
- An `empirical interpolation' method: Application to efficient reduced-basis discretization of partial differential equations
- An eigensystem realization algorithm for modal parameter identification and model reduction
- Approximation of Large-Scale Dynamical Systems
- Data-driven model order reduction of quadratic-bilinear systems.
- Data-driven operator inference for nonintrusive projection-based model reduction
- Data-driven parametrized model reduction in the Loewner framework
- Data-driven reduced order modeling for time-dependent problems
- Data-Driven Sparse Sensor Placement for Reconstruction: Demonstrating the Benefits of Exploiting Known Patterns
- Deep learning of dynamics and signal-noise decomposition with time-stepping constraints
- Discovering governing equations from data by sparse identification of nonlinear dynamical systems
- Dynamic mode decomposition of numerical and experimental data
- Effectively subsampled quadratures for least squares polynomial approximations
- Extracting Sparse High-Dimensional Dynamics from Limited Data
- Finite sample properties of system identification methods
- From Fixed-X to Random-X Regression: Bias-Variance Decompositions, Covariance Penalties, and Prediction Error Estimation
- scientific article; zbMATH DE number 6678650 (Why is no real title available?)
- scientific article; zbMATH DE number 3997615 (Why is no real title available?)
- scientific article; zbMATH DE number 1445419 (Why is no real title available?)
- Inexact solves in interpolatory model reduction
- Kernel Analog Forecasting: Multiscale Test Problems
- Learning low-dimensional dynamical-system models from noisy frequency-response data with Loewner rational interpolation
- Lift \& learn: physics-informed machine learning for large-scale nonlinear dynamical systems
- Missing Point Estimation in Models Described by Proper Orthogonal Decomposition
- Model reduction of bilinear systems in the Loewner framework
- Modeling Systems Based on Noisy Frequency and Time Domain Measurements
- Non-intrusive reduced order modeling of nonlinear problems using neural networks
- Nonlinear model reduction via discrete empirical interpolation
- Nonlinear System Identification: A User-Oriented Road Map
- On dynamic mode decomposition: theory and applications
- On the sample complexity of the linear quadratic regulator
- On the Scalar Rational Interpolation Problem
- Operator inference of non-Markovian terms for learning reduced models from partially observed state trajectories
- Physics-informed regularization and structure preservation for learning stable reduced models from data with operator inference
- Probabilistic error estimation for non-intrusive reduced models learned from data of systems governed by linear parabolic partial differential equations
- Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations. Application to transport and continuum mechanics.
- Reduced order modeling for nonlinear structural analysis using Gaussian process regression
- Refined perturbation bounds for eigenvalues of Hermitian and non-Hermitian matrices
- Robust data-driven discovery of governing physical laws with error bars
- Sampling Low-Dimensional Markovian Dynamics for Preasymptotically Recovering Reduced Models from Data with Operator Inference
- Sparse dynamics for partial differential equations
- Spectral analysis of nonlinear flows
- Stability of discrete empirical interpolation and gappy proper orthogonal decomposition with randomized and deterministic sampling points
- SubTSBR to tackle high noise and outliers for data-driven discovery of differential equations
- Tangential interpolation-based eigensystem realization algorithm for MIMO systems
- Vector Fitting for Matrix-valued Rational Approximation
Cited in
(11)- Operator inference of non-Markovian terms for learning reduced models from partially observed state trajectories
- Learning quantities of interest from dynamical systems for observation-consistent inversion
- Learning low-dimensional dynamical-system models from noisy frequency-response data with Loewner rational interpolation
- Sampling Low-Dimensional Markovian Dynamics for Preasymptotically Recovering Reduced Models from Data with Operator Inference
- Operator inference with roll outs for learning reduced models from scarce and low-quality data
- Operator inference driven data assimilation for high fidelity neutron transport
- Learning reduced-order quadratic-linear models in process engineering using operator inference
- Learning stochastic reduced models from data: a nonintrusive approach
- Bayesian learning with Gaussian processes for low-dimensional representations of time-dependent nonlinear systems
- Learning latent space dynamics with model-form uncertainties: a stochastic reduced-order modeling approach
- Exact operator inference with minimal data
This page was built for publication: Active Operator Inference for Learning Low-Dimensional Dynamical-System Models from Noisy Data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6113944)
