摘要
本文结合工程实例研究了人工神经网络输入层参数对网络训练和测试的影响 ,表明输入层参数变化对于网络测试并无显著影响 。
In this work, the influence of parameters in input layer on network training and testing is studied through an engineering example. It is shown that parameter increase or decrease in number in input layer has not significant effect on testing network, and some parameters are enough to train and test network.
出处
《土工基础》
2004年第6期49-51,共3页
Soil Engineering and Foundation
关键词
基坑
围护结构
变形预测
人工神经网络
excavation, retaining structure, deformation prediction, artificial neural network
