岭回归和主成分回归下的农业总产值因素分析

Agriculture output factor analysis with ridge regression and principal component regression

选取7个影响吉林省农业总产值的因素,运用SAS软件建立了农业总产值的多元回归模型。为解决经典线性回归模型的多重共线性问题,运用主成分回归模型和岭回归模型对其进行了修正,最后对这两个修正模型进行了比较分析,得出岭回归模型相对较优的结论。

Seven factors influencing agricultural output of Jilin Province are selected and then the multiple regression model for the output is established with SAS software.To solve the multicollinearity problem in the classical linear regression model,we adjust the model with both the principal component regression and ridge regression model.The improved the models are compared and it comes to a conclusion that the ridge regression has better performance.