A NEW TRUST-REGION ALGORITHM FOR FINITE MINIMAX PROBLEM |
| |
Authors: | Fusheng Wang Chuanlong Wang Li Wang |
| |
Institution: | [1]Department of Mathematics, Taiyuan Normal University, Taiyuan 030012, China [2]Department of Mathematics, University of California, San Diego, USA |
| |
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. |
| |
Keywords: | Trust-region methods Minimax optimization Nonmonotone strategy Globalconvergence Superlinear convergence |
本文献已被 维普 等数据库收录! |
| 点击此处可从《计算数学(英文版)》浏览原始摘要信息 |
|