Saved in:
Bibliographic Details
Main Authors: Bryutkin, Andrey, Levine, Matthew E., Urteaga, Iñigo, Marzouk, Youssef
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.11535
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918131727859712
author Bryutkin, Andrey
Levine, Matthew E.
Urteaga, Iñigo
Marzouk, Youssef
author_facet Bryutkin, Andrey
Levine, Matthew E.
Urteaga, Iñigo
Marzouk, Youssef
contents Standard Bayesian approaches for linear time-invariant (LTI) system identification are hindered by parameter non-identifiability; the resulting complex, multi-modal posteriors make inference inefficient and impractical. We solve this problem by embedding canonical forms of LTI systems within the Bayesian framework. We rigorously establish that inference in these minimal parameterizations fully captures all invariant system dynamics (e.g., transfer functions, eigenvalues, predictive distributions of system outputs) while resolving identifiability. This approach unlocks the use of meaningful, structure-aware priors (e.g., enforcing stability via eigenvalues) and ensures conditions for a Bernstein--von Mises theorem -- a link between Bayesian and frequentist large-sample asymptotics that is broken in standard forms. Extensive simulations with modern MCMC methods highlight advantages over standard parameterizations: canonical forms achieve higher computational efficiency, generate interpretable and well-behaved posteriors, and provide robust uncertainty estimates, particularly from limited data.
format Preprint
id arxiv_https___arxiv_org_abs_2507_11535
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Canonical Bayesian Linear System Identification
Bryutkin, Andrey
Levine, Matthew E.
Urteaga, Iñigo
Marzouk, Youssef
Machine Learning
Systems and Control
Computation
Standard Bayesian approaches for linear time-invariant (LTI) system identification are hindered by parameter non-identifiability; the resulting complex, multi-modal posteriors make inference inefficient and impractical. We solve this problem by embedding canonical forms of LTI systems within the Bayesian framework. We rigorously establish that inference in these minimal parameterizations fully captures all invariant system dynamics (e.g., transfer functions, eigenvalues, predictive distributions of system outputs) while resolving identifiability. This approach unlocks the use of meaningful, structure-aware priors (e.g., enforcing stability via eigenvalues) and ensures conditions for a Bernstein--von Mises theorem -- a link between Bayesian and frequentist large-sample asymptotics that is broken in standard forms. Extensive simulations with modern MCMC methods highlight advantages over standard parameterizations: canonical forms achieve higher computational efficiency, generate interpretable and well-behaved posteriors, and provide robust uncertainty estimates, particularly from limited data.
title Canonical Bayesian Linear System Identification
topic Machine Learning
Systems and Control
Computation
url https://arxiv.org/abs/2507.11535