A Gauss-Newton-like optimization algorithm for "weighted" nonlinear least-squares problems

@article{Guillaume1996AGO,
  title={A Gauss-Newton-like optimization algorithm for "weighted" nonlinear least-squares problems},
  author={Patrick Guillaume and Rik Pintelon},
  journal={IEEE Trans. Signal Processing},
  year={1996},
  volume={44},
  pages={2222-2228}
}
The Gauss-Newton algorithm is often used to minimize a nonlinear least-squares loss function instead of the original Newton-Raphson algorithm. The main reason is the fact that only first-order derivatives are needed to construct the Jacobian matrix. Some applications as, for instance multivariable system identification, give rise to “weighted” nonlinear least-squares problems for which it can become quite hard to obtain an analytical expression of the Jacobian matrix. To overcome that struggle… CONTINUE READING
Highly Cited
This paper has 37 citations. REVIEW CITATIONS

Citations

Publications citing this paper.
Showing 1-10 of 20 extracted citations

References

Publications referenced by this paper.
Showing 1-7 of 7 references

Identification of multiinput multioutput systems using frequency - domain models

  • P. Guillaume, J. Schoukens
  • 1992

Nonlinear least squares and nonlinearily constrained optimization,

  • P. E. Gill, W. Murray
  • Numevical Analysis. Dundee,
  • 1991

Description of a maximum likelihood estimator in the frequency domain for multiinput , multioutput systems and its application to flight flutter analysis

  • J. Schoukens
  • Mech . Syst . Signal Processing
  • 1990

Matrix Compututions

  • G. H. Golub, C. F. Van Loan
  • SIAM J. Sci. Stat. Comput.,
  • 1990

The State of the Art

  • D. Jacobs, Ed
  • Numerical Analysis. London: Academic,
  • 1977

An adaptive n o n linear least - squares problem

  • D. M. Gay, R. E. Welsch
  • ACM Truns . Murh . Sofmare
  • 1976

Similar Papers

Loading similar papers…