Methods of Conjugate Directions Versus Quasi-Newton Methods.

Abstract

It is shown that algorithms for minimizing an unconstrained function F(x), x belongs to E sup n, which are solely methods of conjugate directions, can be expected to exhibit only an n or (n - 1) step superlinear rate of convergence to an isolated local minimizer. This is contrasted with quasi-Newton methods which can be expected to exhibit every step superlinear convergence. Similar statements about a quadratic rate of convergence hold when a Lipschitz condition is placed on the second derivative of F(x). (Author)

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1971
Accession Number
AD0731718

Entities

People

  • Garth P. Mccormick
  • Klaus Ritter

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Behavior And Behavior Mechanisms
  • Behavioral Disciplines And Activities
  • Behavioral Sciences
  • Canada
  • Continents
  • Convergence
  • Cooperation
  • Geographic Regions
  • Group Dynamics
  • Mathematics
  • New Brunswick

Fields of Study

  • Mathematics

Readers

  • Approximation Theory.
  • Linear Algebra