Year: 2012
Author: Fusheng Wang, Chuanlong Wang, Li Wang
Journal of Computational Mathematics, Vol. 30 (2012), Iss. 3 : pp. 262–278
Abstract
In this paper, a new trust region algorithm for minimax optimization problems is proposed, which solves only one quadratic subproblem based on a new approximation model at each iteration. The approach is different from the traditional algorithms that usually require to solve two quadratic subproblems. Moreover, to avoid Maratos effect, the nonmonotone strategy is employed. The analysis shows that, under standard conditions, the algorithm has global and superlinear convergence. Preliminary numerical experiments are conducted to show the efficiency of the new method.
You do not have full access to this article.
Already a Subscriber? Sign in as an individual or via your institution
Journal Article Details
Publisher Name: Global Science Press
Language: English
DOI: https://doi.org/10.4208/jcm.1109-m3567
Journal of Computational Mathematics, Vol. 30 (2012), Iss. 3 : pp. 262–278
Published online: 2012-01
AMS Subject Headings:
Copyright: COPYRIGHT: © Global Science Press
Pages: 17
Keywords: Trust-region methods Minimax optimization Nonmonotone strategy Global convergence Superlinear convergence.