线性模型的新的两参数估计
A new two-parameter estimation in linear model

基于最小二乘估计在复共线性存在时的不足，提出线性模型参数估计的一种新方法——新的两参数估计. 首先，给出该估计的若干条基本性质——最小二乘估计的线性变换、具有有偏性、压缩性和可容许性. 其次，在均方误差矩阵的意义下，证明该估计优于最小二乘估计和旧的两参数估计.
Basing on the shortcoming of least squares estimator when multicollinearity occurs，we introduce a new method for parameter estimation in linear model-new two-parameter estimation. First of all，some properties of new two-parameter estimation are discussed-it’s a linear transformation of least squares estimator，it’s biased，comprehensed and admissible. Furthermore，we show that our new two-parameter estimation is superior，in the mean square error matrix sense，to the least squares estimator and the old two-parameter estimation