Alternating multi-step quasi-Newton methods for unconstrained optimization.

*(English)*Zbl 0886.65064The authors consider multistep quasi-Newton methods for unconstrained optimization. These methods were introduced in three earlier papers of the authors, where they showed how an interpolating curve in the variable-space could be used to derive an appropriate generalization of the secant equation normally employed in the construction of quasi-Newton methods. One of the most successful of these multistep methods employs the current approximation to the Hessian to determine the parametrization of the interpolating curve and, hence, the derivatives required in the generalized updating formula. However, certain approximations were found to be necessary in the process in order to reduce the level of computation required (which must be repeated at each iteration) to acceptable levels.

In this paper, the authors show how a variant of this algorithm, which avoids the need for such approximations, may be obtained. This is accomplished by alternating, on successive iterations, a single-step and a two-step method. The results of a series of experiments, which show that the new algorithm exhibits a clear improvement in numerical performance, are reported.

In this paper, the authors show how a variant of this algorithm, which avoids the need for such approximations, may be obtained. This is accomplished by alternating, on successive iterations, a single-step and a two-step method. The results of a series of experiments, which show that the new algorithm exhibits a clear improvement in numerical performance, are reported.

Reviewer: J.Guddat (Berlin)

##### Software:

minpack
PDF
BibTeX
XML
Cite

\textit{J. A. Ford} and \textit{L. A. Moghrabi}, J. Comput. Appl. Math. 82, No. 1--2, 105--116 (1997; Zbl 0886.65064)

Full Text:
DOI

##### References:

[1] | Broyden, C.G., Quasi-Newton methods and their application to function minimization, Maths. comp., 21, 368-381, (1967) · Zbl 0155.46704 |

[2] | Broyden, C.G.; Broyden, C.G., The convergence of a class of double-rank minimization algorithms, parts I and II, J. inst. math. appl., J. inst. math. appl., 6, 222-231, (1970) · Zbl 0207.17401 |

[3] | Fletcher, R., A new approach to variable metric algorithms, Comput. J., 13, 317-322, (1970) · Zbl 0207.17402 |

[4] | Fletcher, R., Practical methods of optimization, (1987), Wiley New York · Zbl 0905.65002 |

[5] | Ford, J.A.; Moghrabi, I.A., Alternative parameter choices for multi-step quasi-Newton methods, Optim. meth. software, 2, 357-370, (1993) |

[6] | Ford, J.A.; Moghrabi, I.A., Multi-step quasi-Newton methods for optimization, J. comput. appl. math., 50, 305-323, (1994) · Zbl 0807.65062 |

[7] | Ford, J.A.; Moghrabi, I.A., Further investigation of multi-step quasi-Newton methods, Scientia iranica, 1, 327-334, (1995) · Zbl 0972.65039 |

[8] | Ford, J.A.; Moghrabi, I.A., Minimum curvature multi-step quasi-Newton methods, Comput. math. appl., 31, 179-186, (1996) · Zbl 0874.65046 |

[9] | Ford, J.A.; Moghrabi, I.A., An alternating multi-step quasi-Newton method for unconstrained optimization, () · Zbl 0807.65062 |

[10] | Ford, J.A.; Saadallah, A.F., A rational function model for unconstrained optimization, (), 539-563 · Zbl 0666.90062 |

[11] | Goldfarb, D., A family of variable metric methods derived by variational means, Math. comp., 24, 23-26, (1970) · Zbl 0196.18002 |

[12] | MorĂ©, J.J.; Garbow, B.S.; Hillstrom, K.E., Testing unconstrained optimization software, ACM trans. math. software, 7, 17-41, (1981) · Zbl 0454.65049 |

[13] | Shanno, D.F., Conditioning of quasi-Newton methods for function minimization, Math. comp., 24, 647-656, (1970) · Zbl 0225.65073 |

[14] | Shanno, D.F.; Phua, K.H., Matrix conditioning and nonlinear optimization, Math. programming, 14, 149-160, (1978) · Zbl 0371.90109 |

This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.