论Panel模型谱分解特征根估计量之比

    On the Ratio of Estimators of Panel Model's Spectral Decomposition Eigenvalues

    • 摘要: 为了改进Panel模型参数估计的常用方法,分析了Panel模型谱分解后2个特征根之比,给出了此比例的几种调整方法.使用这些方法导出了模型方差分量的若干不变估计类中MSE准则下的最优估计;还导出了模型回归系数的一个两步无偏估计类,并且给出调整比例的表达式,从而在类中降低了估计量方差的上界.

       

      Abstract: In order to improve the usual methods of parameter estimate in the Panel model, the ratio of two spectral eigenvalues of Panel model is analysed and several methods to adjust the ratio are provided. Applying these methods, it derives the best estimator under MSE criterion in some subclasses of invariant estimator of variance component in the model; and derives a class of unbiased two-stage estimator for the regression coefficient. Also, expression of adjusting the ratio so as to decrease the covariance matrix in this class is given.

       

    /

    返回文章
    返回