Accession Number:

ADA101356

Title:

Partitionability of Implicit Least Squares Model Fitting Problems

Descriptive Note:

Technical rept.

Corporate Author:

ARMY BALLISTIC RESEARCH LAB ABERDEEN PROVING GROUND MD

Personal Author(s):

Report Date:

1981-05-01

Pagination or Media Count:

35.0

Abstract:

Non-linear least squares model fitting problems differ from general optimization problems due to the special structure of the least squares normal equations. In simple least squares problems, that structure algorithmic simplification can be achieved. More general least squares problems can often be made partitionable by proper parameter manipulations. In this report, sufficient partitionability conditions are derived and parameter manipulations are discussed for the achievement of partitionability.

Subject Categories:

  • Statistics and Probability

Distribution Statement:

APPROVED FOR PUBLIC RELEASE