Kernel Selection in Linear System Identification: Part II: A Classical Perspective
2011 (English)In: Proceedings of the 50th IEEE Conference on Decision and Control, 2011, 4326-4331 p.Conference paper (Refereed)
In this companion paper, the choice of kernels for estimating the impulse response of linear stable systems is considered from a classical, “frequentist”, point of view. The kernel determines the regularization matrix in a regularized least squares estimate of an FIR model. The quality is assessed from a mean square error (MSE) perspective, and measures and algorithms for optimizing the MSE are discussed. The ideas are tested on the same data bank as used in Part I of the companion papers. The resulting findings and conclusions in the two papers are very similar despite the different perspectives.
Place, publisher, year, edition, pages
2011. 4326-4331 p.
Kernel, Least squares approximation, Mean square error methods, Identification
IdentifiersURN: urn:nbn:se:liu:diva-95596DOI: 10.1109/CDC.2011.6160722ISBN: 978-1-61284-799-3ISBN: 978-1-61284-800-6OAI: oai:DiVA.org:liu-95596DiVA: diva2:636442
50th IEEE Conference on Decision and Control, Orlando, FL, USA, 12-15 December, 2011
FunderSwedish Research CouncilEU, European Research Council