期刊文献+

数据的剔除对缺落值模型的影响

THE INFLUENCE OF DELETING DATA ON MISSING OBSERVATIONS MODEL
下载PDF
导出
摘要 从距离角度和效率角度研究了一组数据和多组数据缺落值估计的影响,建立了剔除一组数据时Cook距离与复相关系数的等式关系,剔除多组数据时Cook距离与广义相关系数的不等式及广义方差比度量与Hoetelling广义相关系数的等式关系. The influences of one case or several cases on missing observations are studied from the Cook distance and efficiency points of view. When one case is deleted,the equality relation between Cook distance and coefficient of total correlation is built.Also when several cases are deleted, we have found the inequality relation between Cook distance and generalized correlation coefficient. The equality relation betweengeneralized variance ratio and Hoetelling correlation coefficient is obtained too.
出处 《南京大学学报(数学半年刊)》 CAS 2005年第2期349-354,共6页 Journal of Nanjing University(Mathematical Biquarterly)
关键词 缺落值线性模型 缺落值估计 COOK距离 广义方差比 复相关系数 广义相关系数 missing observations linear model, estimate of missing observations, Cook distance, generalized variaace ratio, coefficient of total correlation, generalized correlation coefiicieat
  • 相关文献

参考文献4

  • 1Wang S G. Estimates of MissingObservations and Invariant Variance Subspace[J]. Acta. Math.Appl. Sin., 1985, 8(3):329-333.
  • 2Henderson H V and Searle S R. On Deriving the Inverse of Sum of Matries[J]. SIAMReview, 1981,23(1): 53-59.
  • 3Ni G X. Theory and Method of Common Matrix[M]. Shanghai Science and Technology Press,1984.
  • 4Zhang Y T. Generalized Correlation Coefficient and Its Application[J]. Acta Math. Appl.Sin.,1979, 1(3): 312-320.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部