摘要
BP算法现在已成为目前应用最广泛的神经网络学习算法,它在函数逼近、模式识别、分类、数据压缩等领域有着更加广泛的应用,但存在收敛较慢问题.笔者在文中简述了BP算法原理,针对BP算法的收敛性问题,提出了几点改进措施.
The BP(back propagation) algorithm is a neural network learning algorithm, it is applied extensively in function approximation, mode distinguishing, classification, data compression et, but it has a question of convergence. In this paper, based on describing the principle of the BP algorithm, the convergence is discussed deeply, and several improvements to BP neural network are proposed.
出处
《重庆交通学院学报》
2005年第1期143-145,共3页
Journal of Chongqing Jiaotong University
