多元线性模型回归系数的压缩LS估计*
The Compression LS Estimate of Regression Coefficient in Multivariate Linear Model
-
摘要: 本文采用压缩LS估计 来估计设计矩阵呈病态的多元线性模型的回归系数B.通过k值的选取,可使 =Vec 的均方误差MSB小于β=Vec(B)的LS估计β*的MSE.证明了 具有可容许性、抗干扰性和有效性,并给出了实际应用中选取k值的方法.Abstract: In this paper,compression LS estimate of the regression coefficient B is considered when the design matrix present ill-condition in multivariate linear model.The MSE(mean square error)of the estimate =Vec is less than theMSE of LS estimate β* of the regression coefficient β= Vec(B) by choosing the pa-rameter k.Admissibility,numerical stability and relative efficiency of )are proved.The method of determining k value for practical use is also suggested.
-
[1] 陈希孺、王松桂.《近代回归分析-原理、方法及其应用》,安徽教育出版社,合肥(1987). [2] 曾志斌,多元广义岭估计及K值选取的Q(C)准则,福建师范大学学报(自然胖学版),6(4)(1990).25-33. [3] 曾志斌,多元线性摸型回归斌数的根方估计,福建师范失学学报(自然科学版),7(3)(1991),35-40. [4] 王松桂,《线性模型的理论及其应用》,安徽教育出版社,合肥(1987). [5] 孙维广,《矩阵扰动分析》,科学出版社,北京(1987). [6] Bloomfield,P.and G.S.Waston,The efficiency of least squares,Biometrika,62(1975),121-128. [7] 刘爱义、王松桂,线性模型中LS估计的一种新的相对效率,应用概率统计.5(2)(1987),97-104. [8] Aroold,S.F.,Theory of Linear Models and Multivariate Analisis,John Wiley,New York(1981). [9] 杨虎,单参数主成分回归估计,高校应用数学学报,4(1)(1989),74-80.
计量
- 文章访问数: 2116
- HTML全文浏览量: 107
- PDF下载量: 506
- 被引次数: 0