On the worst-case divergence of the least-squares algorithm

@inproceedings{Akay1998OnTW,
  title={On the worst-case divergence of the least-squares algorithm},
  author={H{\"u}seyin Akçay and Brett Ninness},
  year={1998}
}
In this paper, we provide a H 1 {norm lower bound on the worst{case identiication 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 oo suuciently quickly. 

From This Paper

Topics from this paper.

Citations

Publications citing this paper.

References

Publications referenced by this paper.
SHOWING 1-10 OF 14 REFERENCES

Asymptoti varian e expressions for identi ed bla kbox transferfun tion models

  • L. Ljung.
  • IEEE Transa tions on Automati Control
  • 1995

Estimation of model quality

  • J. R. Partington
  • Automati a
  • 1995

Membership set estimation : size , optimalinput , omplexity and relations with least - squares

  • R. Tempo, H. Cho
  • 1995

Properties of least - squares estimates in set membershipidenti ation

  • G. C. Goodwin
  • 1995

The sample omplexityof worstase identi ation of F . I . R . systems

  • T. Theodosopoulos M. A. Dahleh, J. N. Tsitsiklis
  • Syst . Contr . Lett .
  • 1995

Worst - ase analysis of the least - squares method and relatedidenti ation methods

  • A. Tikku
  • Syst . Contr . Lett .
  • 1995

Least - squares type algorithms for identi ation inthe presen e of modelling un ertainty

  • K. M. Nagpal
  • IEEE . Trans . Automat . Contr .
  • 1994

Some new properties of Fourier onstantsMath

  • J. E. Littlewood
  • Ann .
  • 1993

The least - squares identi ation of FIR systemssubje t to worstase noise

  • H. Hjalmarsson
  • Syst . Contr . Lett .
  • 1993

Similar Papers

Loading similar papers…