摘要
分析了大坝安全监测数据处理中自变量因子之间的多重相关性对传统的多元回归模型的影响.运用偏最小二乘回归分析的基本原理和交叉有效性检验方法,建立了基于偏最小二乘法的回归模型.工程实例表明,偏最小二乘回归模型能够有效消除因子相关性对模型回归系数估计和回归分析效果的影响,取得的结果较为合理.
In dam safety monitoring data analysis,the multiple correlation among independent variables is inevitable.The influence of multiple correlation on traditional multivariate regression model is analyzed.The basic principle and cross validation test of the partial least squares regression（PLSR） are introduced,and the model based on PLSR is established.The given example shows that the PLSR model can effectively overcome the influence of multiple correlation on regression coefficients estimate and effect of regression analysis,so the result is more reasonable.
出处
《三峡大学学报（自然科学版）》
CAS
2011年第4期19-21,共3页
Journal of China Three Gorges University：Natural Sciences
基金
河海大学水文水资源与水利工程科学国家重点实验室专项基金(2009586012)
关键词
大坝安全监测
多重相关性
多元回归
偏最小二乘回归
dam safety monitoring
multiple correlation
multivariate regression
partial least squares regression
