一类改进广义岭估计的实例研究

Improved Generalized Ridge Regression and Its Application

摘要: 本文对广义岭回归方法进行了改进，并在真实数据中进行了实例验证。改进广义岭估计主要对广义岭估计在线性回归模型中当存在若干较大异常值影响模型精度的情况进行了修正，加入适当的修正参数，使模型达到对数据更精确的拟合以及预测作用，并对中俄贸易数据进行了实例验证。

Abstract: In this paper, the generalized ridge regression method is improved and verified by real data. The improved generalized ridge regression mainly corrects the generalized ridge estimation in the linear regression model when there are some large outliers that affect the accuracy of the model. Appropriate correction parameters are added to make the model achieve more accurate data fitting and prediction effects. Finally, we verified it in trade data.