Title :
On the worst-case divergence of the least-squares algorithm
Author :
Akcay, Huseyin ; Ninness, Brett
Author_Institution :
Arcelik AS, Istanbul, Turkey
Abstract :
We provide an 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
Keywords :
H∞ optimisation; covariance matrices; discrete time systems; identification; least squares approximations; linear systems; time-domain analysis; FIR model structures; H∞-norm; H∞ optimisation; SISO systems; covariance function; discrete time systems; divergence; identification; least-squares algorithm; linear time invariant systems; lower bound; time domain analysis; worst-case divergence; Amplitude estimation; Estimation error; Finite impulse response filter; Gaussian processes; H infinity control; Noise level; Stochastic resonance; Stochastic systems; System identification; Time domain analysis;
Conference_Titel :
American Control Conference, 1998. Proceedings of the 1998
Conference_Location :
Philadelphia, PA
Print_ISBN :
0-7803-4530-4
DOI :
10.1109/ACC.1998.703276
