4.7 Article

Learning Causal Estimates of Linear Operators From Noisy Data

Related references

Note: Only part of the references are listed.
Article Automation & Control Systems

Kernel-Based Models for System Analysis

Henk J. van Waarde et al.

Summary: This article introduces a computational framework for identifying nonlinear input-output operators and regularizing them based on suitable input-output properties. By modeling in a reproducing kernel Hilbert space, the biased identification problem is solved, which is a departure from the prevailing state-space framework. This kernel-based framework can better integrate the requirements of data modeling and system analysis.

IEEE TRANSACTIONS ON AUTOMATIC CONTROL (2023)

Article Automation & Control Systems

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

Matteo Scandella et al.

Summary: In this paper, a non-parametric approach to infer a continuous-time linear model from data is proposed, which automatically selects a proper structure of the transfer function and guarantees system stability. Through benchmark simulation examples, the proposed approach is shown to outperform state-of-the-art continuous-time methods, even in critical cases where short sequences of canonical input signals are used for model learning.

INTERNATIONAL JOURNAL OF CONTROL (2022)

Article Automation & Control Systems

Deterministic error bounds for kernel-based learning techniques under bounded noise

Emilio Tanowe Maddalena et al.

Summary: This paper discusses the problem of reconstructing a function from noise-corrupted samples, analyzing two kernel algorithms- kernel ridge regression and epsilon-support vector regression. By establishing finite-sample error bounds and providing numerical examples, the connection between these algorithms and Gaussian processes is explored, aiming to bridge the gap between non-parametric kernel learning and system identification for robust control.

AUTOMATICA (2021)

Article Automation & Control Systems

Learning linear modules in a dynamic network using regularized kernel-based methods

Karthik Raghavan Ramaswamy et al.

Summary: The method presented in the paper uses regularized kernel-based methods for identification, reducing computational complexity and showing potential in local module identification.

AUTOMATICA (2021)

Article Geosciences, Multidisciplinary

Identification of linear response functions from arbitrary perturbation experiments in the presence of noise - Part 1: Method development and toy model demonstration

Guilherme L. Torres Mendonca et al.

Summary: The method developed in this study can identify linear response functions with only one perturbation experiment and a control experiment, using regularization theory to address the ill-posed problem. By comparing frequency spectra, the noise level needed for proper estimation of the regularization parameter can be determined, especially for monotonic functions. The method's robustness and advantages are investigated using a toy model, and it can also be applied to other types of linear ill-posed problems.

NONLINEAR PROCESSES IN GEOPHYSICS (2021)

Article Automation & Control Systems

A shift in paradigm for system identification

Lennart Ljung et al.

INTERNATIONAL JOURNAL OF CONTROL (2020)

Article Automation & Control Systems

Kernel-based identification of non-causal systems with application to inverse model control

Lennart Blanken et al.

AUTOMATICA (2020)

Article Automation & Control Systems

On the mathematical foundations of stable RKHSs

Mauro Bisiacco et al.

AUTOMATICA (2020)

Article Automation & Control Systems

Recursive estimation for sparse Gaussian process regression

Manuel Schuerch et al.

AUTOMATICA (2020)

Article Automation & Control Systems

Regularized LTI system identification in the presence of outliers: A variational EM approach

Martin Lindfors et al.

AUTOMATICA (2020)

Article Automation & Control Systems

Continuous-Time DC Kernel-A Stable Generalized First-Order Spline Kernel

Tianshi Chen

IEEE TRANSACTIONS ON AUTOMATIC CONTROL (2018)

Article Automation & Control Systems

Maximum Entropy Kernels for System Identification

Francesca Paola Carli et al.

IEEE TRANSACTIONS ON AUTOMATIC CONTROL (2017)

Article Automation & Control Systems

Regularized nonparametric Volterra kernel estimation

Georgios Birpoutsoukis et al.

AUTOMATICA (2017)

Article Automation & Control Systems

KERNELS FOR LINEAR TIME INVARIANT SYSTEM IDENTIFICATION

Francesco Dinuzzo

SIAM JOURNAL ON CONTROL AND OPTIMIZATION (2015)

Article Automation & Control Systems

Kernel methods in system identification, machine learning and function estimation: A survey

Gianluigi Pillonetto et al.

AUTOMATICA (2014)

Article Automation & Control Systems

On the estimation of transfer functions, regularizations and Gaussian processes-Revisited

Tianshi Chen et al.

AUTOMATICA (2012)

Article Computer Science, Theory & Methods

Optimal learning rates for least squares regularized regression with unbounded sampling

Cheng Wang et al.

JOURNAL OF COMPLEXITY (2011)

Article Mathematics, Applied

VECTOR VALUED REPRODUCING KERNEL HILBERT SPACES AND UNIVERSALITY

C. Carmeli et al.

ANALYSIS AND APPLICATIONS (2010)

Article Statistics & Probability

A REPRODUCING KERNEL HILBERT SPACE APPROACH TO FUNCTIONAL LINEAR REGRESSION

Ming Yuan et al.

ANNALS OF STATISTICS (2010)

Article Automation & Control Systems

A new kernel-based approach for linear system identification

Gianluigi Pillonetto et al.

AUTOMATICA (2010)

Review Statistics & Probability

Kernel methods in machine learning

Thomas Hofmann et al.

ANNALS OF STATISTICS (2008)

Article Computer Science, Artificial Intelligence

A unifying view of Wiener and Volterra theory and polynomial kernel regression

Matthias O. Franz et al.

NEURAL COMPUTATION (2006)

Article Mathematics, Applied

VECTOR VALUED REPRODUCING KERNEL HILBERT SPACES OF INTEGRABLE FUNCTIONS AND MERCER THEOREM

Claudio Carmeli et al.

ANALYSIS AND APPLICATIONS (2006)

Review Mathematics, Applied

Regularization networks and support vector machines

T Evgeniou et al.

ADVANCES IN COMPUTATIONAL MATHEMATICS (2000)