摘要
文章针对线性模型中设计阵存在复共线性,提出了参数估计的一种新的几乎无偏两参数估计,在均方误差矩阵准则下,给出了新估计优于最小二乘估计、几乎无偏岭估计、几乎无偏Liu估计等估计的充分条件。通过数值模拟验证了相关理论结果。
Aiming'at the existence of multicollinearity in the design matrix of the linear model, this paper proposes a new al- most unbiased two-parameter estimation. Under the criterion of mean square error matrix, the paper gives sufficient conditions for the new estimation over the least-square estimation, almost unbiased ridge estimation, almost unbiased Liu estimation and so on. Finally the paper makes a numerical simulation to verify the relevant theoretical results.
出处
《统计与决策》
CSSCI
北大核心
2017年第24期18-21,共4页
Statistics & Decision
基金
河南省基础与前沿技术研究项目(142300410401)
河南省科技厅软科学项目(162400410106)
关键词
线性模型
复共线性
均方误差矩阵
两参数估计
几乎无偏两参数估计
linear model
multicollinearity
mean square error matrix
two-parameter estimation
almost unbiased two-pa-rameter estimation