Saved in:
Bibliographic Details
Main Authors: Racz, Daniel, Petreczky, Mihaly, Daroczy, Balint
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.13766
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913840643440640
author Racz, Daniel
Petreczky, Mihaly
Daroczy, Balint
author_facet Racz, Daniel
Petreczky, Mihaly
Daroczy, Balint
contents We derive a finite-sample probabilistic bound on the parameter estimation error of a system identification algorithm for Linear Switched Systems. The algorithm estimates Markov parameters from a single trajectory and applies a variant of the Ho-Kalman algorithm to recover the system matrices. Our bound guarantees statistical consistency under the assumption that the true system exhibits quadratic stability. The proof leverages the theory of weakly dependent processes. To the best of our knowledge, this is the first finite-sample bound for this algorithm in the single-trajectory setting.
format Preprint
id arxiv_https___arxiv_org_abs_2503_13766
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A finite-sample bound for identifying partially observed linear switched systems from a single trajectory
Racz, Daniel
Petreczky, Mihaly
Daroczy, Balint
Machine Learning
Systems and Control
We derive a finite-sample probabilistic bound on the parameter estimation error of a system identification algorithm for Linear Switched Systems. The algorithm estimates Markov parameters from a single trajectory and applies a variant of the Ho-Kalman algorithm to recover the system matrices. Our bound guarantees statistical consistency under the assumption that the true system exhibits quadratic stability. The proof leverages the theory of weakly dependent processes. To the best of our knowledge, this is the first finite-sample bound for this algorithm in the single-trajectory setting.
title A finite-sample bound for identifying partially observed linear switched systems from a single trajectory
topic Machine Learning
Systems and Control
url https://arxiv.org/abs/2503.13766