Active operator inference for learning low-dimensional dynamical-system models from noisy data

by   Wayne Isaac Tan Uy, et al.

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.



There are no comments yet.


page 20


Sampling low-dimensional Markovian dynamics for pre-asymptotically recovering reduced models from data with operator inference

This work introduces a method for learning low-dimensional models from d...

Operator inference of non-Markovian terms for learning reduced models from partially observed state trajectories

This work introduces a non-intrusive model reduction approach for learni...

Non-intrusive Nonlinear Model Reduction via Machine Learning Approximations to Low-dimensional Operators

Although projection-based reduced-order models (ROMs) for parameterized ...

Reduced operator inference for nonlinear partial differential equations

We present a new scientific machine learning method that learns from dat...

Learning low-dimensional dynamical-system models from noisy frequency-response data with Loewner rational interpolation

Loewner rational interpolation provides a versatile tool to learn low-di...

Markov Neural Operators for Learning Chaotic Systems

Chaotic systems are notoriously challenging to predict because of their ...

Error-in-variables modelling for operator learning

Deep operator learning has emerged as a promising tool for reduced-order...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.