A Study of Conjugate Gradient Methods.

Abstract

A number of new properties of algorithms of the Conjugate Gradient type are shown, paying particular attention to methods which utilize variable metric information in determining the conjugate gradient search directions. A comprehensive discussion of conjugate gradient methods is attempted, and each algorithm is presented within the context of other existing algorithms, and approach which provides fresh insights and some new algorithms.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1978
Accession Number
ADA066391

Entities

People

  • Jorge Nocedal
  • L. Nazareth

Organizations

  • Stanford University

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Computational Science
  • Convergence
  • Differential Equations
  • Eigenvalues
  • Equations
  • Information Systems
  • Iterations
  • Linear Algebra
  • Linear Systems
  • Management Information Systems
  • Nonlinear Programming
  • Operations Research
  • Optimization
  • Partial Differential Equations
  • Polynomials
  • Sequences

Readers

  • Computational Linguistics
  • Computational Modeling and Simulation
  • Linear Algebra