On the worst-case divergence of the least-squares algorithm | NOVA. The University of Newcastle's Digital Repository

Home

Description	Size	Format
Author final version	189 KB	Adobe Acrobat PDF	View/Open

Please use this identifier to cite or link to this item: http://hdl.handle.net/1959.13/30303

Title: On the worst-case divergence of the least-squares algorithm
Author/Creator: Ackay, Huseyin; Ninness, Brett
Description: In this paper, we provide a H ∞ norm lower bound on the worst-case identification error of least-squares estimation when using FIR model structures. This bound increases as a logarithmic function of model complexity and is valid for a wide class of inputs characterized as being quasi-stationary with covariance function falling off sufficiently quickly.
Relation: Systems & Control Letters Vol. 33, Issue 1, p. 19-24
Publisher Link: http://dx.doi.org/10.1016/S0167-6911(97)00093-5
Date: 1998
Publisher: Elsevier Science
Keyword(s): least-squares; identification in H ∞; time-domain data; divergance
Resource Type: journal article
Identifier: http://hdl.handle.net/1959.13/30303
Identifier: ISSN:0167-6911
Language: eng

53 Visitors

57 Hits

NOVA. The University of Newcastle's Digital Repository