摘要
引荐了一种自动优化神经网络的新方法。这种启迪方法综合采用了相关有效算法,通过快速自底向上构造神经网络算法,可以获得优化结构的神经网络,即时选定参数算法动态优化神经网络的学习参数,并且快速交叉校验算法为解决过度适应问题提供了捷径。实验证明,这种启迪方法能自动有效地优化神经网络,与其它算法相较而言,具有更好的归纳性能、优化的网络结构和更快的学习速度。
A new approach for the automated design of optimal neural network is presented.The heuristic approach employs several efficient algorithms,such as construction of optimal network architecture via fast cascade-correlation,dynamic optimization of learning parameters via simultaneous determination,avoiding overfitting problem via fast cross-validation.The experimental results show that the heuristic approach can automatically design optimal neural network with good generalization capability and optimal network and short training time in comparison with other algorithms.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第23期34-37,77,共5页
Computer Engineering and Applications
关键词
神经网络
瀑流关联
自底向上
学习参数优化
交叉校验
Neural Network,cascade-correlation,bottom-up,learning parameters optimization,cross-validation