一种改进的线性回归模型

An improved linear regression model

本文阐述了一种改进的线性回归模型——偏最小二乘回归模型的基本原理,并结合实例指出当变量间存在着线性关系时,普通的多元线性回归建模方法会失效;通过对偏最小二乘回归模型和主成分回归模型的系数估计精度及预测精度两方面的比较,说明要消除变量间的线性关系,前者优于后者。

The paper firstly presents the basic principles of an improved linear regression model which is partial least-squares regression model. Through analyzing some cases, we find that the common method of multiple linear regression model will he failure when the linear relations exist in variables. The coefficient and forecasting accuracy of partial least-squares regression model and principal component regression are estimated and compared. The results show that, to eliminate the linear relationship between variables, the former model is better than the latter one.