南京林业大学学报(自然科学版) ›› 1997, Vol. 21 ›› Issue (04): 77-80.doi: 10.3969/j.jssn.1000-2006.1997.04.016

• 研究论文 • 上一篇    下一篇

线性回归模型中自变量变换与强影响点

童春发   

  1. 南京林业大学基础课部
  • 出版日期:1997-12-18 发布日期:1997-12-18

TRANSFORMATIONS OF EXPLANATORY VARIABLES AND INFLUENTIAL CASES IN LINEAR REGRESSION MODEL

Tong Chunfa   

  1. Basic Courses Department Nanjing Forestry University Nanjing 210037
  • Online:1997-12-18 Published:1997-12-18

摘要: <正>提出了单个数据或一组数据对线性回归模型中自变量变换参数极大似然估计影响的诊断方法,并把该方法同Atkinson的方法作了比较。

Abstract: A diagnostic method is proposed for assessing the influence of an individual case and a group of cases on the likelihood estimate of the transformation parameters for the explanatory variables in linear regression models. The method is compared with one proposed by Atkinson (1982).