A comparison of least-squares solutions to overdetermined linear systems

Date

3-1991

Degree

Bachelor of Science in Applied Mathematics

College

College of Arts and Sciences (CAS)

Adviser/Committee Chair

Lynie B. Dimasuay

Abstract

The concern of the study was to apply three least-squares solution methods to overdetermined systems of linear equations. these three are: QR Decomposition, Singular Value Decomposition and LU Decomposition. A program for each method was developed using the Turbo Pascal 5.0. the Computational cost and the accuracy of each method were deserved by entering eleven overdetermined systems in matrix form into the developed programs. After which, the results were analyzed and served as measures for comparison. Out of the eleven matrices entered, three were found out to be singular. However, SVD and LU were able to obtain approximate linear least-squares solutions. Armed with bases for comparison and the results, it was deduced that SVD was the most accurate as compared to the QR and LU. The fastest in giving results is the LU Decomposition.

Language

English

Location

UPLB Main Library Special Collections Section (USCS)

Call Number

Thesis

Document Type

Thesis

This document is currently not available here.

Share

COinS