广义压缩最小二乘估计参数的选择准则

文［１］中提出了广义压缩最小二乘估计，并讨论了它的可容许性及在均方误差准则下的最优性．本文进一步讨论它的其它性质，提出广义压缩最小二乘估计参数选择的Ｑ（ｃ）准则，并给出按此准则选择参数的方法和引入模糊主成分系数的概念，从而在对各种有偏估计的优、缺点进行比较的基础上提出了当参数选择效果不够理想时的一些修正方法

The Q ( c ) Criterion of Choosing Parameter about the Generalized Shrunken Least Squares Estimate

The generalized shrunken least squares estimate is proposed and its admissibility,under the MSE criterion is studied in paper . In this article,we investigate some of its other properties,propose the Q(c) criterion of choosing parameter about the generalized shrunken least squares estimate,give the method of choosing parameter under Q(c) criterion and some other methods,propose vague principal component coefficient,Based on a comparison of the variously biased estimate,we propose some methods of correction when the result of estimating is not satisfying.