摘要
讨论了变系数回归模型中变系数的加权最小二乘估计和变系数的加权可估函数的线性估计的最优性;推广了Gauss-Markov定理;并且构造出参数σ2的估计量.
In this paper, optimal property for weighted least squares estimate of the varying-coefficient and linear estimation of weighting estimable function of the varying-coefficient in the varying-coefficient regression model are discussed. And the extension of theorem of Gauss-Markov is obtained, then the estimator of paremeterσ2 is also obtained.
出处
《湘南学院学报》
2005年第2期15-19,22,共6页
Journal of Xiangnan University
基金
湖南省教育厅科学研究项目(03G007)
关键词
变系数回归模型
加权可估
加权线性无偏估计
相合估计
未知参数
varying-coefficient regression model
weighting estimation
weighting estimation and linear estimation
