多元共线性引发的问题及其实际的处理方法

PROBLEM CAUSED BYMULTVARIATE COLLINEARITY AND METHOD OF DEALING WITH THE PROBLEM PRACTICALLY

多元线性回归分析中，要求自变量之间是相互独立的，但在解决实际问题时，往往不能满足此要求，而出现有一些自变量“共线”的情况．文中主要阐述多元共线的含义、多元共线时给统计分析所带来的困难。

Multivariate linear regression analysis requires independent variables to be independent each other. But in some cases of solving practical problems, independent variables are collinear and not independent each other. This paper mainly expounds the sense of multivariate collinearity and the difficult problems in statistical analysis caused by multivariate collinearity, and contributes some method for reference to overcoming the difficulty.