A GENERAL-PURPOSE PARALLEL ALGORITHM FOR UNCONSTRAINED OPTIMIZATION

被引:13
|
作者
Nash, Stephen G. [1 ]
Sofer, Ariela [1 ]
机构
[1] George Mason Univ, Dept Operat Res & Appl Stat, Fairfax, VA 22030 USA
基金
美国国家科学基金会;
关键词
nonlinear optimization; truncated Newton method; parallel computing; conjugate-gradient method;
D O I
10.1137/0801032
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
This paper describes a general-purpose algorithm for unconstrained optimization that is suitable for a parallel computer. It is designed to be as easy to use as traditional algorithms for this problem, requiring only that a (scalar) subroutine be provided to evaluate the objective function and its gradient vector of first derivatives. The algorithm used is a block truncated-Newton method. Truncated-Newton methods are a class of methods that compromise on Newton's method so that large problems can be solved. Enhancements to the basic method suitable for a parallel computer are described. These include a revised data storage scheme, new preconditioning and initialization strategies to accelerate the method, a parallel line search, revised stopping rules for the inner algorithm, and a new "nonlinearity" test to determine the adequacy of the quadratic model. Numerical results are presented to illustrate the performance of the method, and comparisons are made with other scalar and parallel algorithms.
引用
收藏
页码:530 / 547
页数:18
相关论文
共 50 条