Parallel computing in bound constrained quadratic programming

被引:0
|
作者
D'Apuzzo, M [1 ]
De Simone, V [1 ]
Marino, M [1 ]
Toraldo, G [1 ]
机构
[1] Univ Naples 2, Dpt Math, Caserta, Italy
关键词
D O I
暂无
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We describe our research activity concerning the development of parallel software for numerical optimization. We present some design and computational experiences on the solution of convex bound constrained quadratic programming problems on high performance machines. We compare the projected gradient method with the potential reduction algorithm in order to develop an efficient parallel software.
引用
收藏
页码:479 / 491
页数:13
相关论文
共 50 条