摘要
将相关性剪枝算法(CPA)和变学习率、附加动量方法结合提出了一种基于CPA的改进的BP神经网络剪枝算法.实验结果表明,改进的算法可以降低训练步数,加快神经网络的收敛速度,在测试数据集上的均方误差也得到了进一步的优化.
BP（back propagation）neural network is one of the most widely used artificial neural networks.It is known that the performance of the BP neural network depends mainly on its structure.What＇s more,the BP neural network takes a long time to achieve convergence and the results may fill in local optimum.By combining CPA,variable learning rate and additional momentum,we propose an improved BP neural network pruning algorithm named as LMCPA neural network in this paper.Experimental results have shown that the performance of the neural network has been improved by LMCPA algorithm.
出处
《西南大学学报（自然科学版）》
CAS
CSCD
北大核心
2016年第3期165-170,共6页
Journal of Southwest University(Natural Science Edition)
基金
国家自然科学基金项目(71201129)
关键词
BP神经网络
相关性剪枝算法
变学习率
附加动量
bp neural network
correlation pruning algorithm（CPA）
variable learning rate
additional momentum
