Sharp Spectral Rates for Koopman Operator Learning

Vladimir R Kostic, Karim Lounici, Pietro Novelli, massimiliano pontil

Keywords: Statistical Learning Theory, Dynamical Systems

TL;DR: We derive minimax optimal statistical rates for learning Koopman operator and combine them with spectral perturbation theory to provide strong guarantees for learning Koopman eigenvalues and eigenfunctions.

Abstract:

Non-linear dynamical systems can be handily described by the associated Koopman operator, whose action evolves every observable of the system forward in time. Learning the Koopman operator and its spectral decomposition from data is enabled by a number of algorithms. In this work we present for the first time non-asymptotic learning bounds for the Koopman eigenvalues and eigenfunctions.
We focus on time-reversal-invariant stochastic dynamical systems, including the important example of Langevin dynamics.
We analyze two popular estimators: Extended Dynamic Mode Decomposition (EDMD) and Reduced Rank Regression (RRR). Our results critically hinge on novel {minimax} estimation bounds for the operator norm error, that may be of independent interest. Our spectral learning bounds are driven by the simultaneous control of the operator norm error and a novel metric distortion functional of the estimated eigenfunctions. The bounds indicates that both EDMD and RRR have similar variance, but EDMD suffers from a larger bias which might be detrimental to its learning rate. Our results shed new light on the emergence of spurious eigenvalues, an issue which is well known empirically. Numerical experiments illustrate the implications of the bounds in practice.

Submission Number: 6738