LGOCMLFeb 16, 2022

Single Trajectory Nonparametric Learning of Nonlinear Dynamics

arXiv:2202.08311v229 citations
AI Analysis

This work provides theoretical guarantees for nonparametric learning in dynamical systems, which is incremental as it builds on existing information-theoretic methods to derive rate-optimal bounds.

The paper tackles the problem of learning nonlinear dynamics from a single trajectory using nonparametric least squares estimation, establishing optimal error bounds that scale as T^{-1/(2+q)} for stable processes and specific hypothesis classes.

Given a single trajectory of a dynamical system, we analyze the performance of the nonparametric least squares estimator (LSE). More precisely, we give nonasymptotic expected $l^2$-distance bounds between the LSE and the true regression function, where expectation is evaluated on a fresh, counterfactual, trajectory. We leverage recently developed information-theoretic methods to establish the optimality of the LSE for nonparametric hypotheses classes in terms of supremum norm metric entropy and a subgaussian parameter. Next, we relate this subgaussian parameter to the stability of the underlying process using notions from dynamical systems theory. When combined, these developments lead to rate-optimal error bounds that scale as $T^{-1/(2+q)}$ for suitably stable processes and hypothesis classes with metric entropy growth of order $δ^{-q}$. Here, $T$ is the length of the observed trajectory, $δ\in \mathbb{R}_+$ is the packing granularity and $q\in (0,2)$ is a complexity term. Finally, we specialize our results to a number of scenarios of practical interest, such as Lipschitz dynamics, generalized linear models, and dynamics described by functions in certain classes of Reproducing Kernel Hilbert Spaces (RKHS).

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes