Data-driven model reduction, Wiener projections, and the Koopman-Mori-Zwanzig formalism

Abstract

Model reduction methods aim to describe complex dynamic phenomena using only relevant dynamical variables, decreasing computational cost, and potentially highlighting key dynamical mechanisms. In the absence of special dynamical features such as scale separation or symmetries, the time evolution of these variables typically exhibits memory effects. Recent work has found a variety of data-driven model reduction methods to be effective for representing such non-Markovian dynamics, but their scope and dynamical underpinning remain incompletely understood. Here, we study data-driven model reduction from a dynamical systems perspective. For both chaotic and randomly-forced systems, we show the problem can be naturally formulated within the framework of Koopman operators and the Mori-Zwanzig projection operator formalism. We give a heuristic derivation of a NARMAX (Nonlinear Auto-Regressive Moving Average with eXogenous input) model from an underlying dynamical model. The derivation is based on a simple construction we call Wiener projection, which links Mori-Zwanzig theory to both NARMAX and to classical Wiener filtering. We apply these ideas to the Kuramoto-Sivashinsky model of spatiotemporal chaos and a viscous Burgers equation with stochastic forcing.

Substantial revisions, including additional references. To appear in Journal of Computational Physics

References in corpus (9)

• Data-driven non-Markovian closure models
• Nonparametric forecasting of low-dimensional dynamical systems
• Non-Markovian Closure Models for Large Eddy Simulations using the Mori-Zwanzig Formalism
• Data-adaptive harmonic spectra and multilayer Stuart-Landau models
• Machine-learning error models for approximate solutions to parameterized systems of nonlinear equations
• Model Reduction with Memory and the Machine Learning of Dynamical Systems
• Data-driven model reduction for stochastic Burgers equations
• Parametric Reduced Models for the Nonlinear Schrödinger Equation
• Modeling of Missing Dynamical Systems: Deriving Parametric Models using a Nonparametric Framework

Cited by in corpus (33)

• On closures for reduced order models $-$ A spectrum of first-principle to machine-learned avenues
• Data-Driven Variational Multiscale Reduced Order Models
• Machine Learning for Prediction with Missing Dynamics
• Reduced-Order Models for Coupled Dynamical Systems: Data-driven Methods and the Koopman Operator
• Data-driven molecular modeling with the generalized Langevin equation
• Theoretical tools for understanding the climate crisis from Hasselmann's program and beyond
• Time Series Forecasting Using Manifold Learning
• Identification of MIMO Wiener-type Koopman Models for Data-Driven Model Reduction using Deep Learning
• Kernel-based Prediction of Non-Markovian Time Series
• Conditional Gaussian Nonlinear System: a Fast Preconditioner and a Cheap Surrogate Model For Complex Nonlinear Systems
• Nonequilibrium Statistical Mechanics and Optimal Prediction of Partially-Observed Complex Systems
• Data-driven construction of stochastic reduced dynamics encoded with non-Markovian features
• Resampling with neural networks for stochastic parameterization in multiscale systems
• A Causality-Based Learning Approach for Discovering the Underlying Dynamics of Complex Systems from Partial Observations with Stochastic Parameterization
• Data Driven Learning of Mori-Zwanzig Operators for Isotropic Turbulence
• Data-driven model reduction for stochastic Burgers equations
• CEBoosting: Online Sparse Identification of Dynamical Systems with Regime Switching by Causation Entropy Boosting
• CGNSDE: Conditional Gaussian Neural Stochastic Differential Equation for Modeling Complex Systems and Data Assimilation
• Koopman Operator, Geometry, and Learning
• Modeling Partially Observed Nonlinear Dynamical Systems and Efficient Data Assimilation via Discrete-Time Conditional Gaussian Koopman Network
• Shock trace prediction by reduced models for a viscous stochastic Burgers equation
• Hypoellipticity and the Mori-Zwanzig formulation of stochastic differential equations
• Effective Mori-Zwanzig equation for the reduced-order modeling of stochastic systems
• Stability Preserving Data-driven Models With Latent Dynamics
• Data-driven Mori-Zwanzig modeling of Lagrangian particle dynamics in turbulent flows
• Learning effective good variables from physical data
• Data-driven stochastic modeling of coarse-grained dynamics with finite-size effects using Langevin regression
• ISALT: Inference-based schemes adaptive to large time-stepping for locally Lipschitz ergodic systems
• A Martingale-Free Introduction to Conditional Gaussian Nonlinear Systems
• NySALT: Nyström-type inference-based schemes adaptive to large time-stepping
• A simple hypocoercivity analysis for the effective Mori-Zwanzig equation
• Constrained Policy Optimization for Stochastic Optimal Control under Nonstationary Uncertainties
• Koopman operator-based discussion on partial observation in stochastic systems