Learning stochastic closures using ensemble Kalman inversion
From MaRDI portal
Abstract: Although the governing equations of many systems, when derived from first principles, may be viewed as known, it is often too expensive to numerically simulate all the interactions they describe. Therefore researchers often seek simpler descriptions that describe complex phenomena without numerically resolving all the interacting components. Stochastic differential equations (SDEs) arise naturally as models in this context. The growth in data acquisition, both through experiment and through simulations, provides an opportunity for the systematic derivation of SDE models in many disciplines. However, inconsistencies between SDEs and real data at short time scales often cause problems, when standard statistical methodology is applied to parameter estimation. The incompatibility between SDEs and real data can be addressed by deriving sufficient statistics from the time-series data and learning parameters of SDEs based on these. Following this approach, we formulate the fitting of SDEs to sufficient statistics from real data as an inverse problem and demonstrate that this inverse problem can be solved by using ensemble Kalman inversion (EKI). Furthermore, we create a framework for non-parametric learning of drift and diffusion terms by introducing hierarchical, refinable parameterizations of unknown functions, using Gaussian process regression. We demonstrate the proposed methodology for the fitting of SDE models, first in a simulation study with a noisy Lorenz '63 model, and then in other applications, including dimension reduction in deterministic chaotic systems arising in the atmospheric sciences, large-scale pattern modeling in climate dynamics, and simplified models for key observables arising in molecular dynamics. The results confirm that the proposed methodology provides a robust and systematic approach to fitting SDE models to real data.
Recommendations
- Data-driven closures for stochastic dynamical systems
- Statistical Learning of Nonlinear Stochastic Differential Equations from Nonstationary Time Series using Variational Clustering
- Detecting stochastic governing laws with observation on stationary distributions
- ATLAS: a geometric approach to learning high-dimensional stochastic systems near manifolds
- Systematic physics constrained parameter estimation of stochastic differential equations
Cited in
(28)- Systematic physics constrained parameter estimation of stochastic differential equations
- Empirical evaluated SDE modelling for dimensionality-reduced systems and its predictability estimates
- Data-driven closures for stochastic dynamical systems
- Bayesian learning of stochastic dynamical models
- Ensemble Kalman inversion for sparse learning of dynamical systems from time-averaged data
- Generative ensemble regression: Learning particle dynamics from observations of ensembles with physics-informed deep generative models
- Generative stochastic modeling of strongly nonlinear flows with non-Gaussian statistics
- ATLAS: a geometric approach to learning high-dimensional stochastic systems near manifolds
- Statistical Learning of Nonlinear Stochastic Differential Equations from Nonstationary Time Series using Variational Clustering
- Mitigating Model Error via a Multimodel Method and Application to Tropical Intraseasonal Oscillations
- A framework for machine learning of model error in dynamical systems
- Reservoir computing with error correction: long-term behaviors of stochastic dynamical systems
- A causality-based learning approach for discovering the underlying dynamics of complex systems from partial observations with stochastic parameterization
- Detecting stochastic governing laws with observation on stationary distributions
- Recovering critical parameter for nonlinear Allen–Cahn equation by fully discrete continuous data assimilation algorithms *
- Quantum Mechanics for Closure of Dynamical Systems
- Filtering dynamical systems using observations of statistics
- Learning about structural errors in models of complex dynamical systems
- CGNSDE: conditional Gaussian neural stochastic differential equation for modeling complex systems and data assimilation
- Neural dynamical operator: continuous spatial-temporal model with gradient-based and derivative-free optimization methods
- Bayesian experimental design for model discrepancy calibration: an auto-differentiable ensemble Kalman inversion approach
- The fundamental subspaces of ensemble Kalman inversion
- Modeling partially observed nonlinear dynamical systems and efficient data assimilation via discrete-time conditional Gaussian Koopman network
- Forecast error growth: a dynamic-stochastic model
- Model discovery on the fly using continuous data assimilation
- Data-driven stochastic closure modeling via conditional diffusion model and neural operator
- CGKN: a deep learning framework for modeling complex dynamical systems and efficient data assimilation
- Relaxation-based schemes for on-the-fly parameter estimation in dissipative dynamical systems
This page was built for publication: Learning stochastic closures using ensemble Kalman inversion
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5023844)
