摘要
为了克服传统BP网络的不足,发挥共轭梯度法和梯度下降法各自的优势,本文介绍了在城市地下水水质评价模型中采用集成BP网络、用共轭梯度法和梯度下降法相结合的混合算法来训练网络的方法。通过实例表明,将此网络模型应用于城市地下水水质评价是有效的、可行的。
To overcome limitation of the traditional BP network,and bring into play the advantage of conjugate gradient algorithm and gradient descent algorithm. The article introduces the method to train the network adopting the integrated BP network, the conjugate gradient algorithm combined with the gradient descent algorithm in the evaluation model of groundwater quality in the city. The result shows that the network model is effective and feasible applied in the evaluation of groundwater quality in the city.
出处
《地下水》
2006年第5期36-37,共2页
Ground water
关键词
传统BP网络
共轭梯度法
梯度下降法
集成BP网络
城市地下水水质评价
Traditional BP network, the conjugate gradient algorithm, the gradient descent algorithm and the integrated BP network and the evaluation of groundwater quality
