BP神经网络算法的改进  

The Improvement of Learning Algorithm of BP Neural Networks

作　　者：孙建业 王辉[1] 

机构地区：[1]哈尔滨工业大学

出　　处：《系统工程与电子技术》1994年第6期41-46,共6页Systems Engineering and Electronics

摘　　要：本文分析了ＢＰ算法所面临的问题，给出了一种改进算法，说明了它们的原理和应用环境，将它们与传统的ＢＰ算法作了比较，并通过一实例验证：应用改进算法可以大幅度地提高ＢＰ神经网络的学习速度，这对ＢＰ神经网络的应用有较大实际意义。After analyzing the problems of BP algorithm,the authors present an improved algorithm,illustrate its principle and application circumstances,and compare it with the traditional BP algorithm,Simulation results show that the improved algorithm can enhance the learning speed of BP neural networks by a wide margin. This is of great significance to the application of BP neural networks.

关 键 词：神经网络 BP算法 算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]