- Published 2010 in Math. Meth. of OR

The problem of minimizing f̃ = f +p over some convex subset of a Euclidean space is investigated, where f(x) = x Ax + b x is a strictly convex quadratic function and |p| is only assumed to be bounded by some positive number s. It is shown that the function f̃ is strictly outer γ-convex for any γ > γ∗, where γ∗ is determined by s and the smallest eigenvalue of A. As consequence, a γ∗-local minimal solution of f̃ is its global minimal solution and the diameter of the set of global minimal solutions of f̃ is less than or equal to γ∗. Especially, the distance between the global minimal solution of f and any global minimal solution of f̃ is less than or equal to γ∗/2. This property is used to prove the rough support property of f̃ and some generalized optimality conditions.

Showing 1-10 of 24 references

Highly Influential

11 Excerpts

Highly Influential

5 Excerpts

Highly Influential

7 Excerpts

Highly Influential

7 Excerpts

Highly Influential

5 Excerpts

Highly Influential

2 Excerpts

Highly Influential

3 Excerpts

Highly Influential

1 Excerpt

Highly Influential

1 Excerpt

Highly Influential

1 Excerpt

@article{Phu2010GlobalIO,
title={Global infimum of strictly convex quadratic functions with bounded perturbations},
author={Hoang Xuan Phu and Vo Minh Pho},
journal={Math. Meth. of OR},
year={2010},
volume={72},
pages={327-345}
}