摘要
对传统的ＢＰ算法进行了改进，提出了基于二阶导数的ＢＰ算法，大大减少了样本训练的迭代次数，从而提高了网络运算速度．对异或问题和蜢虫分类问题的计算，取得了满意的效果．
A new BP algorithm with introduction of second derivative is proposed to improve the conventional BP algorithm. The iteration of training samples is reduced. The computation rate of the algorithm in the networks is raised. The algorithm is applied to discussing XOR problem and the classification of midges.
出处
《华中理工大学学报》
CSCD
北大核心
1999年第6期90-92,共3页
Journal of Huazhong University of Science and Technology
基金
国家自然科学基金
关键词
神经网络
快速分类
BP算法
梯度法
学习算法
neural network
fasting classification
BP algorithm
gradient method
differentiate analysis
