剔除相关性的最小二乘理论研究

The research on the Least Squares Removed Correlation

提出了一种新的回归模型,剔除相关性的最小二乘,并给出建模步骤。该模型能有效的克服变量间的相关性,兼顾到变量的筛选并克服回归系数反常的现象。

This paper presents a new linear regress model, the least squares removed correlation, then giveing the model of founding method, it will effectively overcome the correlation between variables and take account of the variable selection ,as well as Overcome the abnormal phenomenon of regression coefficients.