The Solution of Generalized Least Squares Problems

  • G. Alistair Watson
Part of the International Series of Numerical Mathematics book series (ISNM, volume 75)

Abstract

A problem frequently encountered in empirical sciences is that of establishing a causal relationship between experimental variables. This involves firstly the selection of a suitable model for the process under consideration containing a number of free parameters, and secondly the choice of values of these parameters to give a best fit, in an appropriate sense, to the available data. The usual procedure is to treat one of the problem variables as being the ’dependent’ variable, and to attribute errors to the observed values of that variable. The parameters are then chosen so as to make these errors small in some sense: for example a commonly used method is to minimize the sum of squares.

Keywords

Line Search Hessian Matrix Quadratic Programming Problem Descent Direction Linear Manifold 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Al-Baali, M. and Fletcher, R.. Variational methods for non-linear least squares, Journal of the Oper.. Res. Soc. (to appear)Google Scholar
  2. 2.
    Britt, H.I. and Luecke, R.H. The estimation of parameters in nonlinear, implicit models,. Technometrics.15(1973), 233 – 247.CrossRefGoogle Scholar
  3. 3.
    Dennis, J.E. Jr., Gay, D.M. and Welsch, R.E. An adaptive nonlinear least squares algorithm, A.C.M. Trans. Math. Soft-ware7(1981), 348 – 368CrossRefGoogle Scholar
  4. 4.
    Draper, N.R. and Smith, H. Applied Regression Analysis ( Second Edition ), Wiley, New York (1981).Google Scholar
  5. 5.
    Fletcher, R. Practical Methods of Optimization, Vol. II, Constrained Optimization, Wiley, Chichester (1981).Google Scholar
  6. 6.
    Golub, G.H. Least squares, singular values and matrix approximations, Aplikace Mathematiky13(1968), 44 – 51Google Scholar
  7. 7.
    Nazareth, L. An adaptive method for minimizing a sum of squares of nonlinear functions, IIASA Report WP-83-99 (1983)Google Scholar
  8. 8.
    O’Neill, M., Sinclair, I.G. and Smith, F.J. Polynomial curve fitting when abscissas and ordinates are both subject to error, The Computer Journal12(1969), 52 – 56.Google Scholar
  9. 9.
    Powell, D.R. and Macdonald, J.R. A rapidly convergent -iterative method for the solution of the generalized non-linear least squares problem, The Computer Journal 15 (1972), 148 – 155.Google Scholar
  10. 10.
    Macdonald, J.R. Review of some experimental and analytical equations of state, Rev. Mod. Phys. 41 (1969), 306 – 349.Google Scholar
  11. 11.
    Macdonald, J.R. and Powell, D.R. Discrimination between equations of state, J. Res. Nat. Bur. Stand. Series A75A(1971), 441 – 453.Google Scholar
  12. 12.
    Rao, C.R. Matrix approximations and reduction of dimensionality in multivariate statistical analysis, in Multivariate Analysis V (ed. P.R. Krishnaiah) North Holland (1980)Google Scholar
  13. 13.
    Ruhe, A. and Wedin, P.-A.. Algorithms for separable non-linear least squares problems, SIAM Rev.22(1980), 318 – 339.CrossRefGoogle Scholar
  14. 14.
    Spath, H. Orthogonal least squares fitting with linear manifolds, preprint (1984).Google Scholar
  15. 15.
    Tarantola, A. and Valette, B. Generalized nonlinear inverse problems solved using the least squares criterion, Rev. Geo. and Space Phys.20(1982), 219 – 232.CrossRefGoogle Scholar
  16. 16.
    Watson, G.A. The solution of generalized least squares problems, University of Dundee, Department of Mathematical Sciences Report N.A./86 (1985).Google Scholar

Copyright information

© Birkhäuser Verlag Basel 1985

Authors and Affiliations

  • G. Alistair Watson
    • 1
  1. 1.Department of Mathematical SciencesUniversity of DundeeDundeeScotland

Personalised recommendations