Extending Mehrotra and Gondzio higher order methods to mixed semidefinite-quadratic-linear programming

被引：5

作者：

Haeberly, JP

Nayakkankuppam, MV

Overton, ML

机构：

[1] Fordham Univ, Dept Math, Bronx, NY 10458 USA

[2] NYU, Courant Inst, Dept Comp Sci, New York, NY USA

[3] NYU, Courant Inst, Dept Comp Sci, New York, NY USA

来源：

OPTIMIZATION METHODS & SOFTWARE | 1999年 / 11-2卷 / 1-4期

基金：

美国国家科学基金会;

关键词：

D O I：

10.1080/10556789908805748

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

We discuss extensions of Mehrotra's higher order corrections scheme and Gondzio's multiple centrality corrections scheme to mixed semidefinite-quadratic-linear programming (SQLP). These extensions have been included in a solver for SQLP written in C and based on LAPACK. The code implements a primal-dual path-following algorithm for solving SQLP problems based on the XZ + ZX search direction and Mehrotra's predictor-corrector method. We present benchmarks showing that the use of the higher order schemes yields substantial reductions in both the number of iterations and the running time of the algorithm, and also improves its robustness.

引用

页码：67 / 90

页数：24