System identification: theory for the user
System identification: theory for the user
Optimal estimation theory for dynamic systems with set membership uncertainty: an overview
Automatica (Journal of IFAC)
Worst-case control-relevant identification
Automatica (Journal of IFAC) - Special issue on trends in system identification
On Tikhonov regularization, bias and variance in nonlinear system identification
Automatica (Journal of IFAC)
Finite-dimensional approximation of Gaussian processes
Proceedings of the 1998 conference on Advances in neural information processing systems II
Advanced lectures on machine learning
Mercer theorem for RKHS on noncompact sets
Journal of Complexity
Nonparametric identification of population models via Gaussian processes
Automatica (Journal of IFAC)
Bayes and empirical Bayes semi-blind deconvolution using eigenfunctions of a prior covariance
Automatica (Journal of IFAC)
Input estimation in nonlinear dynamical systems using differential algebra techniques
Automatica (Journal of IFAC)
Comparing different approaches to model error modeling in robust identification
Automatica (Journal of IFAC)
Non-stationary stochastic embedding for transfer function estimation
Automatica (Journal of IFAC)
From experiment design to closed-loop control
Automatica (Journal of IFAC)
Prediction error identification of linear systems: A nonparametric Gaussian regression approach
Automatica (Journal of IFAC)
Automatica (Journal of IFAC)
On the estimation of transfer functions, regularizations and Gaussian processes-Revisited
Automatica (Journal of IFAC)
A Bayesian approach to sparse dynamic network identification
Automatica (Journal of IFAC)
Consistent identification of Wiener systems: A machine learning viewpoint
Automatica (Journal of IFAC)
Estimation of building occupancy levels through environmental signals deconvolution
Proceedings of the 5th ACM Workshop on Embedded Systems For Energy-Efficient Buildings
Regularized spectrum estimation using stable spline kernels
Automatica (Journal of IFAC)
Hi-index | 22.16 |
This paper describes a new kernel-based approach for linear system identification of stable systems. We model the impulse response as the realization of a Gaussian process whose statistics, differently from previously adopted priors, include information not only on smoothness but also on BIBO-stability. The associated autocovariance defines what we call a stable spline kernel. The corresponding minimum variance estimate belongs to a reproducing kernel Hilbert space which is spectrally characterized. Compared to parametric identification techniques, the impulse response of the system is searched for within an infinite-dimensional space, dense in the space of continuous functions. Overparametrization is avoided by tuning few hyperparameters via marginal likelihood maximization. The proposed approach may prove particularly useful in the context of robust identification in order to obtain reduced order models by exploiting a two-step procedure that projects the nonparametric estimate onto the space of nominal models. The continuous-time derivation immediately extends to the discrete-time case. On several continuous- and discrete-time benchmarks taken from the literature the proposed approach compares very favorably with the existing parametric and nonparametric techniques.