Some Iterative Procedures to Obtain Nonlinear Least Squares Estimates.

Abstract

This review paper discusses the basic theory and underlying motivation, the algorithm, and the problems of convergence for some of the common techniques now in use to obtain nonlinear, least squares, estimates. The methods considered are the method of steepest descent, the Newton-Raphson method, the Gauss-Newton method, Hartley's modified Gauss-Newton method, and Marquardt's method. Special attention is paid to Marquardt's method.

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1976
Accession Number
ADA024437

Entities

People

  • R. P. Kelley

Organizations

  • University of Missouri

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Behavior And Behavior Mechanisms
  • Convergence
  • Human Behavior
  • Mathematical Analysis
  • Mathematics
  • Motivation

Fields of Study

  • Mathematics

Readers

  • Approximation Theory.
  • Operations Research