A RESTARTING DIRECTION FOR THE CONJUGATE GRADIENT METHOD  

作　　者：孙麟平 

机构地区：[1]Department of Mathematics,Nanjing University,Nanjing 210093,PRC

出　　处：《Numerical Mathematics A Journal of Chinese Universities(English Series)》1995年第2期154-163,共10页

基　　金：The Project Supported by National Natural Foundation of China

摘　　要：The main purpose of this paper is to provide a restarting direction for improving on the standard conjugate gradient method.If a drastic non-quadratic behaviour of the objective function is observed in the neighbour of xk,then a restart should be done.The scaling symmetric rank-one update with Davidon’s optimal criterion is applied to generate the restarting direction.It is proved that the conjugate gradient method with this strategy retains the quadratic termination.Numerical experiments show that it is successful.The main purpose of this paper is to provide a restarting direction for improving on the standard conjugate gradient method.If a drastic non-quadratic behaviour of the objective function is observed in the neighbour of xk,then a restart should be done.The scaling symmetric rank-one update with Davidon's optimal criterion is applied to generate the restarting direction.It is proved that the conjugate gradient method with this strategy retains the quadratic termination.Numerical experiments show that it is successful.

关 键 词：OPTIMIZATION QUASI-NEWTON CONJUGATE GRADIENT RESTART 

分 类 号：O241[理学—计算数学]