摘要
假设可供使用的处理机p+q台,将其分成两组,两组处理机之间进行异步并行计算,由此提出了一种求解非凸函数极小的并行算法。若目标函数连续可微,且其一阶导数Lipschitz连续,证明了并行拟牛顿算法的全局收敛性。
p+q sets of processors are assumed, which are available for use .They are divided into 2 groups, for which asynchronous and parallel calculations are conducted. A method is proposed for solving non-convex function minimals. If the target function is continuously differential and Lipschitz continuous in first order derivative global convergence is demonstrated for the parallel pseudo-Newtonian algorithm.
出处
《长江大学学报(自然科学版)》
CAS
2005年第1期1-3,i001,共4页
Journal of Yangtze University(Natural Science Edition)
关键词
拟牛顿方法
并行算法
全局收敛
非凸极小
pseudo-Newtonian algorithm
parallel algorithm
global convergence
non-convex minimal