摘要
基于正则化理论,通过添加一个正则化因子,构造了一个正则化时间序列模型.算例分析表明,新的正则化时间序列分析模型可改善时间序列分析模型中信息矩阵的病态程度,避免时间序列分析模型产生的不适定性,在一定程度上起到了稳定所求参数的作用.
Based on regularization theory, a regularization factor is added to time series model which constructs a regularization time series model. Example analysis shows that the ill-condi- tioned matrix is improved in regularization time series＇ analysis model, which avoids the ill- posed time series＇ analysis model. Analysis of engineering example shows that this new regularized time series＇ analysis model has an effect which makes the parameters stable to some extent.
出处
《交通科学与工程》
2009年第4期11-15,共5页
Journal of Transport Science and Engineering
基金
湖南省科技厅科技计划资助项目(2008SK3054)
关键词
时间序列
自回归模型
正则化
路基沉降
time series
auto regression model
regularization
subgrade settlement
