广义同余神经网络的算法改进与性能分析

Analysis on the Characteristics of Generalized Congruence Neural Networks with an Improved Algorithm

摘要: 对广义同余神经网络(GCNN)的性能进行了深入的分析研究,提出了一种改进的广义同余学习算法,并将该算法与当前广泛使用的标准BP网络算法进行了比较。计算机数字实例模拟表明,该算法具有学习速度快、拟合效果好等特点。
- 神经网络 /
- 同余式 /
- 算法 /
- BP网络
Abstract: An improved learning algorithm for GCNN is proposed and presented in this paper based on a study on the characteristics of GCNN. In addition, a comparison between GCNN and the currentwide used BP net algorithms is performed. Computer simulation using concrete examples shows that the suggested algorithm provides satisfied fitting features and faster learning speed.
- neural networks /
- congruence /
- algorithm /
- BP networks

点击查看大图