| 一种基于二阶导数的 BP 算法 |
| |
| 引用本文： | 迟彦惠,齐欢,张希承,唐建国.一种基于二阶导数的 BP 算法[J].华中科技大学学报(自然科学版),1998(3). |
| |
| 作者姓名： | 迟彦惠 齐欢 张希承 唐建国 |
| |
| 作者单位： | 华中理工大学数学系 |
| |
| 摘 要： | 根据神经网络模型的结构特点，将能量函数的二阶导数与最速下降方向相结合，构造出一种新型的ＢＰ算法，该算法比梯度法收敛快，较牛顿法计算量小．它适合于计算结构复杂的ＢＰ神经网络模型，理论分析表明该算法行之有效，计算机仿真达到了理想的效果．
|
| 关 键 词： | 最速下降方向 BP算法 神经网络 能量函数 梯度法 牛顿法 |
A New BP Algorithm Based on Second Derivative |
| |
| Chi Yanhui M.E., Dept. of Math.,HUST,Wuhan ,China. Qi huan Chang Xicheng Tang Jianguo.A New BP Algorithm Based on Second Derivative[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,1998(3). |
| |
| Authors: | Chi Yanhui ME Dept of Math HUST Wuhan China Qi huan Chang Xicheng Tang Jianguo |
| |
| Institution: | Chi Yanhui M.E., Dept. of Math.,HUST,Wuhan 430074,China. Qi huan Chang Xicheng Tang Jianguo |
| |
| Abstract: | Based on the architectural features of the neural network, a new BP algorithm is developed by combining the gradient direction with the second derivative of the energy function. It is shown that the rate of convergence of the algorithm proposed is faster the that of gradient method, and its amount of work is less than that of Newton method. It is fairly suitable to the calculation of a BP neural network model of complicated architecture. |
| |
| Keywords: | descent direction BP algorithm neural network energy function gradient method Newton method |
| 本文献已被 CNKI 等数据库收录！ |
