【摘要】 提高BP网络的训练速率是改善BP网络性能的一项重要任务。该文在误差反向传播算法 (BP算法 )的基础上提出了一种新的训练算法 ,该算法对BP网络的传统动量法进行了修改 ,采用动态权值调整以减少训练时间。文章提供了改进算法的仿真实例 ,仿真结果表明用该方法解决某些问题时 ,其相对于BP网络传统算法的优越性。更多还原

【Abstract】 It is the most important task to improve the convergence rate of BP neural networks. This paper presents a new training algorithm based on error back propagation (BP Algorithm).This algorithm modulates the traditional momentum BP by adopting dynamic adjusting weights for reducing the training time. This article also provides simulation examples for modified BP algorithm and these results prove the new method is much better than the traditional methods for solving some problems.更多还原