Stable spline identification of linear systems under missing data

DOI10.1016/J.AUTOMATICA.2019.108493MaRDI QIDQ6119719zbMATH OpenOpenAlexFDO

Authors Gianluigi Pillonetto, Alessandro Chiuso, Giuseppe De Nicolao

Publication date 20 February 2024

Published in Automatica (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1908.03913

zbMATH Keywords

Gaussian processeskernel-based regularizationlinear system identificationmissing dataradial basis functions kernelsstable spline imputationstable spline kernels

Mathematics Subject Classification ID

Numerical computation using splines (65D07) Numerical radial basis function approximation (65D12) System identification (93B30) Linear systems in control theory (93C05)

Abstract: A different route to identification of time-invariant linear systems has been recently proposed which does not require committing to a specific parametric model structure. Impulse responses are described in a nonparametric Bayesian framework as zero-mean Gaussian processes. Their covariances are given by the so-called stable spline kernels encoding information on regularity and BIBO stability. In this paper, we demonstrate that these kernels also lead to a new family of radial basis functions kernels suitable to model system components subject to disturbances given by filtered white noise. This novel class, in cooperation with the stable spline kernels, paves the way to a new approach to solve missing data problems in both discrete and continuous-time settings. Numerical experiments show that the new technique may return models more predictive than those obtained by standard parametric Prediction Error Methods, also when these latter exploit the full data set.

Recommendations

• A new kernel-based approach for linear system identification
• Generalized system identification with stable spline kernels
• Kernels for linear time invariant system identification
• Regularized linear system identification using atomic, nuclear and kernel-based norms: the role of the stability constraint
• Linear system identification using the sequential stabilizing spline algorithm

Cites work

• A new kernel-based approach for linear system identification
• A new look at the statistical model identification
• Autoregressive spectral analysis when observations are missing
• Estimation of Time Series Models in the Presence of Missing Data
• Exact likelihood of vector autoregressive-moving average process with missing or aggregated data
• Frequency domain system identification with missing data
• Gaussian processes for machine learning.
• scientific article; zbMATH DE number 44386 (Why is no real title available?)
• scientific article; zbMATH DE number 44406 (Why is no real title available?)
• scientific article; zbMATH DE number 45848 (Why is no real title available?)
• scientific article; zbMATH DE number 3567782 (Why is no real title available?)
• scientific article; zbMATH DE number 840151 (Why is no real title available?)
• Identification of ARX-models subject to missing data
• Kernel methods in system identification, machine learning and function estimation: a survey
• Maximum Likelihood Fitting of ARMA Models to Time Series with Missing Observations
• On the estimation of transfer functions, regularizations and Gaussian processes-revisited
• Prediction error identification of linear systems: a nonparametric Gaussian regression approach
• Regression and time series model selection in small samples
• The exact likelihood of an autoregressive-moving average model with incomplete data

Cited in

(1)

• Kernel-based identification of asymptotically stable continuous-time linear dynamical systems

This page was built for publication: Stable spline identification of linear systems under missing data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6119719)