| 适合于BP学习的共轭梯度法 |
| |
| 引用本文： | 田军,虞厥邦,杨灿. 适合于BP学习的共轭梯度法[J]. 电子科技大学学报(社会科学版), 1994, 0(5) |
| |
| 作者姓名： | 田军 虞厥邦 杨灿 |
| |
| 作者单位： | 电子科技大学光电子技术系 |
| |
| 摘 要： | 基于ＬＭＳ的标准ＢＰ算法收敛速度极慢，而共轭梯度法要求精确的线性搜索，这在神经网络的高维权空间中是难以实现的。本文提出了一种新的ＢＰ学习算法，它采用一种对线性搜索要求不高的改进的共轭梯度法与一种简单的不精确线性搜索相结合，极大地提高了ＢＰ学习速度。经多次测试表明，与标准ＢＰ算法相比，该算法的效率提高了二个数量极。
|
| 关 键 词： | 神经网络，BP学习，共轭梯度法，无约束最优化 |
CONJUGATE GRADIENT ALGORITHM SUITABLE FOR BP LEARNING |
| |
| Tian Jun , Yu Juebang, Yang Can. CONJUGATE GRADIENT ALGORITHM SUITABLE FOR BP LEARNING[J]. Journal of University of Electronic Science and Technology of China(Social Sciences Edition), 1994, 0(5) |
| |
| Authors: | Tian Jun Yu Juebang Yang Can |
| |
| Abstract: | The LMS- based conventional backpropagation(BP)algoritim is very slow inconvergence speed in the high-dimensional weight space of neural network, the exact linesearches required by conjugate gradient algorithm is very difficult in implementation. This paperpresents a novel BP algorithm, which greatly improves leaming speed by using a modifiedconjugate gradient algorithm incorporating a simple inexact line sedch algorithm. Given manytesting results show that the efficiency of the algorithm relative to conventional BP is of morethan two order magnitude. |
| |
| Keywords: | neural network BP learning conjugate gradient algorithm unconstrained optimization |
| 本文献已被 CNKI 等数据库收录！ |
