摘要
本文讨论线性回归模型中约束条件下最小二乘估计的影响问题,利用约束下的Welsch-Kuh统计量研讨了剔除第i组数据(x＇_4,y_4)后,对x_i和x_y处拟合值的影响,另又用Cook统计量讨论了第i组数据对约束最小二乘估计的影响。
This paper discusses the influence problem of least squares estimator for the linear regression model under the restrioted condition Using the W-K statistic under the restricted condition, we study the inflaences on the fitted values at x_i and x_i respectively while the i-th case (x＇_i, y_i) is deleted. The Cook's statistic is also used to discuss the i-th case's influence on least squares estimator under the restricted condition.
出处
《应用概率统计》
CSCD
北大核心
1991年第2期136-142,共7页
Chinese Journal of Applied Probability and Statistics
基金
山西省自然科学基金
