摘要
为克服和改进传统的BP算法的不足,发挥神经网络和遗传算法各自的优势,提出了一种基于遗传算法的神经网络二次训练方法。将遗传算法应用于神经网络的权值训练中,并用神经网络二次训练得到最终结果,降低了计算时间,是一种比较有效的方法。
To overcome and improve on the lack of the traditional BP algorithm, and bring into play the advantage of artificial neural network and genetic algorithm, the paper bring forward a kind of second training method of the artificial neural network based on genetic algorithm. The method bring genetic algorithm into the learn weight training artificial neural network, and get the final result by the second training of method of the artificial neural network. This method reduces the time of computing and is more efficacious.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第2期316-318,共3页
Computer Engineering and Design
关键词
BP算法
人工神经网络
遗传算法
二次训练
学习
权值
BP algorithm
artificial neural network （ANN）
genetic algorithm （GA）
second training
learn
weight
