欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
利用前一步得到的曲率信息代替xk到xk＋1段二次模型的曲率给出一个具有和BFGS类似的收敛性质的类BFGS算法，并揭示新算法与自调比拟牛顿法的关系．从试验函数库CUTE中选择标准试验函数，对比标准BFGS算法及其它改进BFGS算法进行数值试验．试验结果表明这个新算法的表现有点象自调比拟牛顿算法．
英文摘要：
A modified BFGS-type algorithm is proposed, where the current average curvature between line segment xk to xk＋1 is approximated hy the nearest past iteration. The algorithm preserves the convergence properties of the famous BFGS algorithm. The relation between the new algorithm and a self-scaling quasi-Newton algorithm is revealed. Compared with the classic BFGS and other modified BFGS algorithms, numerical results for standard test problems from CUTE are presented, which indicate that the numerical performance of the new algorithm is somewhat like the self-scaling quasi-Newton algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：