Accession Number : AD0749105

Title :   Stability of the Solution of Definite Quadratic Programs,

Corporate Author : TEXAS UNIV AUSTIN CENTER FOR NUMERICAL ANALYSIS

Personal Author(s) : Daniel,James W.

Report Date : SEP 1972

Pagination or Media Count : 18

Abstract : The paper studies how the solution of the problem of minimizing Q(x) = 1/2(x sup T)Kx - (k sup T)x subject to Gx < or = g and Dx = d behaves when K, k, G, g, D, and d are perturbed, say by terms of size epsilon, assuming that K is positive definite. It is shown that in general the solution moves by roughly epsilon if G, g, D, and d are not perturbed; when G, g D, and d are in fact perturbed, much stronger hypotheses allow one to show that the solution moves by roughly epsilon. Many of these results can be extended to more general, nonquadratic, functionals. (Author)

Descriptors :   (*QUADRATIC PROGRAMMING, ALGORITHMS), CONVEX SETS, INEQUALITIES, PERTURBATION THEORY, THEOREMS

Subject Categories : Operations Research

Distribution Statement : APPROVED FOR PUBLIC RELEASE