Nonconvex minimization calculations and the conjugate gradient method

  • M. J. D. Powell
Conference paper
Part of the Lecture Notes in Mathematics book series (LNM, volume 1066)

Abstract

We consider the global convergence of conjugate gradient methods without restarts, assuming exact arithmetic and exact line searches, when the objective function is twice continuously differentiable and has bounded level sets. Most of our attention is given to the Polak-Ribière algorithm, and unfortunately we find examples that show that the calculated gradients can remain bounded away from zero. The examples that have only two variables show also that some variable metric algorithms for unconstrained optimization need not converge. However, a global convergence theorem is proved for the Fletcher-Reeves version of the conjugate gradient method.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. A. Buckley (1982), "Conjugate gradient methods", in Nonlinear Optimization 1981, ed. M.J.D. Powell, Academic Press (London).Google Scholar
  2. R. Fletcher (1980), Practical Methods of Optimization, Vol. I: Unconstrained Optimization, John Wiley & Sons (Chichester).MATHGoogle Scholar
  3. R. Fletcher and C.M. Reeves (1964), "Function minimization by conjugate gradients", The Computer Journal, Vol. 7, pp. 149–154.MathSciNetCrossRefMATHGoogle Scholar
  4. M.J.D. Powell (1977), "Restart procedures for the conjugate gradient method", Math. Programming, Vol. 12, pp. 241–254.MathSciNetCrossRefMATHGoogle Scholar
  5. J.R. Thompson (1977), "Examples of non-convergence of conjugate descent algorithms with exact line searches", Math. Programming, Vol. 12, pp. 356–360.MathSciNetCrossRefMATHGoogle Scholar
  6. G. Zoutendijk (1970), "Nonlinear programming, computational methods", in Integer and Nonlinear Programming, ed. J. Abadie, North-Holland Publishing Co. (Amsterdam).Google Scholar

Copyright information

© Springer-Verlag 1984

Authors and Affiliations

  • M. J. D. Powell

There are no affiliations available

Personalised recommendations